- Split View
-
Views
-
Cite
Cite
Abstracts, European Heart Journal - Cardiovascular Imaging, Volume 17, Issue suppl_1, May 2016, Pages i1–i80, https://doi.org/10.1093/ehjci/jew093
- Share Icon Share
1575
Extracellular volume associates with outcomes more strongly than native or post-contrast myocardial T1
Abstract
Objectives: Myocardial fibrosis in noninfarcted myocardium is emerging as a principal phenotype of vulnerability to adverse events such as mortality and hospitalization for heart failure (HHF), but its optimal noninvasive measurement remains uncertain despite consistently robust histologic validation data for extracellular volume fraction (ECV). We therefore compared ECV, native T1, post contrast T1, the gadolinium contrast partition coefficient (lambda), and the presence of nonischemic scar in their associations with mortality and HHF outcomes.
Method: To quantify of myocardial fibrosis, we performed T1 mapping (MOLLI) in basal and mid short axis slices with cardiovascular magnetic resonance (CMR) before contrast and 12-30 minutes post contrast bolus in 1185 consecutive patients without amyloidosis, hypertrophic or stress cardiomyopathy. We assessed associations with outcomes using Kaplan-Meier plots and chi square values from univariable Cox regression models. All standard T1 mapping parameters were obtained: native and post contrast myocardial T1, the partition coefficient lambda, and ECV. ECV = (1-hematocrit) · [ΔR1myocardium]/[ΔR1bloodpool], where R1 = 1/T1 Late gadolinium enhancement imaging with phase sensitive reconstruction identified nonischemic scar.
Results: Over a median of 1.7 years, 111 individuals experienced events after CMR: 55 HHF events and 74 deaths. ECV yielded better separation of Kaplan-Meier curves in a dose dependent fashion (Figure) and also stronger associations with the combined endpoint of death or HHF. The ECV chi square (77.3, p < 0.001) was at least twice as large as the Native T1 chi square (37.5, p < 0.001), the lambda chi square (34.8, p < 0.001) and nonischemic scar (chi square = 20.5, p<0.001). Post-contrast T1 was not associated with outcomes, even when adjusting further for time after contrast bolus, renal function, and patient weight (chi square <3, p >0.10).
Conclusion: Analogous to histologic previously published validation data, quantitative ECV myocardial fibrosis measures associated with outcomes far stronger than other surrogate measures outcome measures such as native T1, post contrast T1 and nonischemic scar on LGE images. These data suggest that ECV is the noninvasive metric of choice to measure myocardial fibrosis.
Figure. Kaplan-Meier Plots for T1 mapping parameters.
1507
Cardiac Magnetic Resonance measured Extracellular Volume Independently Predicts Adverse Outcome in Heart Failure with Preserved Ejection Fraction
Abstract
Background: Cardiac magnetic resonance (CMR) T1 mapping allows accurate non-invasive quantification of extracellular matrix volume (ECV). Accumulation of extracellular matrix (ECM) is a key pathophysiological mechanism in heart failure with preserved ejection fraction (HFpEF). HFpEF accounts for approximately 50% of all HF cases and is associated with a poor prognosis. However, the prognostic relevance and hemodynamic consequences of CMR-ECV in HFpEF are unknown.
Materials and Methods: Between January 2011 and August 2015 we prospectively enrolled 108 HFpEF patients. All patients underwent CMR T1 mapping for ECV measurement, using the Modified Look-Locker Inversion recovery (MOLLI) sequence. Right heart catheterization was performed in 92 (85%) patients.
Results: The study cohort was prospectively followed for up to three years, during which 33 (30.6%) had a cardiac event (hospitalization and/or death for cardiac reason). Patients with ECV ≥27.6% had more events (p = 0.011), higher NT-proBNP levels (p = 0.018), larger right atria (p = 0.025) compared to patients with ECV <27.6%. By univariable Cox analysis ECV ≥ 27.6% (p = 0.040), presence of pleural effusion (p = 0.001), NYHA class III or IV (p = 0.005), NT-proBNP (p = 0.001), systolic pulmonary artery pressure (sPAP) (p = <0.001), diastolic PAP (p = 0.001), mean PAP (p = <0.001), stroke volume (p = 0.041), venous oxygen saturation (SvO2) (p = 0.028), transpulmonary gradient (p = <0.001), diastolic pressure gradient (p = 0.002), pulmonary pulse pressure (p = <0.001) and pulmonary vascular resistance (PVR) (p = <0.001) were significantly associated with outcome. By multivariable Cox analysis only ECV ≥27.6% (p = 0.049), presence of pleural effusion (p = 0.001), NYHA class III or IV (p = 0.012), NT-proBNP (p = 0.036) and PVR (p = <0.001) remained significant predictors of cardiac events. ECV significantly correlated with clinical, CMR and hemodynamic parameters: 6MWD (R = -0.272, p = 0.006), NYHA class (R = 0.220, p = 0.023), NT-proBNP (R = 0.362, p = <0.001), pulmonary artery diameter (R = 0.262, p = 0.006), right atrial volume (P = 0.299, p = 0.002), mean right atrial pressure (P = 0.246, p = 0.020), SvO2 (P = -0.241, p = 0.023) and stroke index (P = -0.234, p = 0.030).
Conclusions: CMR-ECV independently predicts outcome in HFpEF and is correlated with severity of symptoms as well as exercise capacity. Reduction of ECV may therefore be a promising therapeutic target.
1457
Cardiac MRI Under Percussive Ventilation: A New Promising Technique
Abstract
Background: High-resolution 3D cardiac MR is in general based on respiratory compensation techniques to allow for long acquisitions extending beyond a breath-hold. Such compensation mechanisms are, however, not always accurate. High frequency percussive ventilation (PV) administers small volumes of air, so-called “percussions”, with adjustable pressures and at high frequencies, thereby potentially replacing spontaneous breathing and allowing for a long period of apnea.
Aim: To implement the PV technique in a MR environment and to test its potential for 3D high-resolution coronary MRA during long periods of apnea without the need to apply respiratory compensation techniques.
Material and Methods: PV was delivered by Transrespirator® (Percussionaire® Corporation,Idaho,USA), which was adapted to be MRI compatible by integrating tubes allowing its installation outside the MRI room. One metallic component of the Phasitron was replaced by a MRI-compatible part. The phasitron represents an open circuit in order to protect from the risk of baro-trauma. The entire system is pneumatic, works without electronics, is not influenced by electromagnetic forces, and it ventilates non-invasively without sedation. One volunteer (38 y) and one patient (55 y) with a thymic lesion underwent a MRI (1.5T Aera,Siemens,Germany) under percussive ventilation (PV-MRI). Participants were trained to maintain their breath during 3 preliminary sessions before undergoing the PV-MRI. During PV-MRI, the apnea periods ended at the discretion of the participant. Heart rate and O2 saturation were monitored during the whole MRI. The breathing cycle amplitude reduction was monitored by using a conventional MR navigator. Turbo-SE, VIBE, and coronary MRA (3D cartesian/radial SSFP, isotropic resolution: 1.15mm3/1.3mm3) were performed without contrast medium during 3 sets of PV apnea on inspiration.
Results: PV-MRI was well tolerated in both cases. Mean apnea duration was 10.5 min (range 8.5-12.0 min) in the healthy volunteer and 5.8 min (range 5.0-6.2 min) in the patient. The coronary arteries were well visualized by the 3D-SSFP sequence as shown in Fig. 1A/B. With this technique, lung volumes were “frozen” in full inspiration and pulmonary vessels were also well visualized with PV-MRI (Fig. 1C).
Conclusion: PV is feasible and well tolerated in the MRI environment and allows for apnea durations of up to 10 minutes. Coronary MRA yielded high-resolution images of the coronary arteries without the need to apply any respiratory compensation. Larger series are needed to evaluate the potential of PV-MRI.
Figure 1. A/B/C: 30/8.5/20-mm thick MIP of a 3D-SSFP acquisition without respiratory compensation acquired during a 6.2 minutes apnea with PV.
1644
Histological Validation of Cardiac Magnetic Resonance for the Evaluation of Myocardial Fibrosis after Heart Transplantation in Children
Abstract
Objectives: Diffuse fibrosis may impact myocardial function in children after heart transplantation (HTx). Quantification of diffuse myocardial fibrosis with cardiac magnetic resonance (CMR) is increasingly applied to patients with heart disease and has been shown to correspond well to fibrosis degrees by histology. However, comparisons with histological fibrosis burdens in children are lacking. The aim of this study was to correlate CMR markers of fibrosis to the gold standard of myocardial biopsy.
Methods: 17 patients (9.7 ± 6.3 years; 8 girls) after HTx were prospectively enrolled. Twenty endomyocardial biopsy (EMB) procedures (with 5 myocardial samples each) and CMRs were performed on the same day, at a median of 2.2 years post HTx, 0.02-12.6 years). EMB was done as per institutional post HTx protocol. All 100 specimens were stained with picrosirius red. The degree of fibrosis (%fibrosis) was calculated as ratio of fibrosis area to total myocardial area. Endocardium and scars at previous biopsy sites were excluded from analysis.
Native and post-contrast T1 values were measured within the interventricular septum (IVS) on a mid-ventricular short axis slice, using a modified look-locker inversion recovery approach (MOLLI). Extracellular volume (ECV) was calculated from pre- and post-contrast T1 times and the patients' hematocrit. Native T1 and ECV were correlated with %fibrosis.
Results: Mean %fibrosis was 10.0 ± 3.4%. Mean septal native T1 values were 1008 ± 32ms. Mean ECV within the IVS was 0.30 ± 0.03. %fibrosis and showed a moderate correlation with T1 (r = 0.53, p < 0.05) and with ECV (r = 0.46, p < 0.05), respectively.
Conclusions: Fibrosis markers by CMR correlate with histological degrees of fibrosis on EMBs in children after HTx. The agreement between CMR and histology may have been impacted by of inflammation and edema during rejection on T1 and ECV measurements. Further, poor the EMBs may be an imperfect representation of the septal myocardium, especially when taken from previous biopsy sites. Overall, this study suggests that native T1 and CMR can be utilized as markers of diffuse myocardial fibrosis in children.
1493
First Pass Perfusion Reserve Index in Paediatric Patients with Arterial Switch for Transposition of Great Arteries
Abstract
Background: Arterial switch operation is the treatment of choice for correction of transposition of the great arteries (TGA). Coronary anatomy is the most impacting early prognostic factor after surgery. We previously described potential mechanisms of late coronary complications according to coronary position after surgical transfer. As systematic screening protocol is not worldwide defined, data about late coronary outcome are lacking. Noninvasive first pass perfusion with cardiac magnetic resonance (cMRI) is sensitive technique for evaluation of myocardial perfusion in ischemic heart disease.
Objectives: Here we aimed to apply a comprehensive cMRI protocol in children and adolescents to assess coronary artery complications in late follow up after arterial switch operation for TGA.
Methods: Between 2010 and 2014, 66 patients were prospectively enrolled (median age 13.9± 5.5 years) into two groups according to the presence (9/66)or absence (57/66) of coronary arteries abnormalities at previous computer tomography or angiography screening at 5 ± 2 years. Coronary position and anatomy were analysed as previously described. Semi-quantitative evaluation of myocardial perfusion was performed by the analysis of myocardial First-Pass perfusion images at rest and during adenosine infusion and the segmental perfusion reserve index (PRI) was calculated. Late gadolinium enhancement (LGE) was also studied.
Results: Eleven patients (11/66= 16%) had perfusion defects (PRI < 1.5). In 9/11 perfusion defects were in anterior-anteroseptal or anterolateral segments vs 2/11 in inferior-inferoseptal-inferolateral segments. In 16 (16/66 = 24%) patients left coronary artery was reimplanted in clock position 11 or 12. Interestingly, these patients with left coronary in position 11-12 had more frequent perfusion defects than other patients (7 vs 2, p = 0.004). Moreover, considering only the myocardial segments irrigated by left coronary artery, semiquantitative perfusion defects were found only in patients with left coronary artery in position 11 and 12 (7 vs 0, p = 0.0002). These patients had significantly lower PRI in myocardial segments irrigated by left coronary artery myocardial than other patients (2.0 ± 1 vs 2.7 ± 1, p < 0.05). LGE was positive in 2 cases. No patients with decreased PRI had LGE. All patients with decreased PRI were asymptomatic and did not show echocardiographic or functional signs of myocardial ischemia.
Conclusion: cMRI can provide a useful tool for detecting asymptomatic myocardial perfusion delay. PRI analysis results confirm the presence of a high-risk group of patients. Clinical implication of early detection of perfusion delay should be investigated.
1652
Myocardial Fibrosis is Prevalent in Obstructive Sleep Apnea and Associated with Hospitalization for Heart Failure or Death
Abstract
Background: Myocardial fibrosis may be prevalent and mediate increased rates of hospitalization for heart failure (HHF) and mortality in patients with obstructive sleep apnea (OSA). We quantified myocardial fibrosis by extracellular volume fraction (ECV) and examined associations between these outcomes. Since myocardial fibrosis is reversible, this data may inform disease mechanisms and potential targets for therapy in OSA that remain uncertain.
Methods: We enrolled 1,094 consecutive referred patients for cardiovascular magnetic resonance (CMR) without amyloidosis, congenital heart disease, stress cardiomyopathy, or hypertrophic cardiomyopathy and 16 healthy controls. We quantified ECV in noninfarcted myocardium and tracked outcomes prospectively, including a subgroup of 176 patients with available polysomnography. For survival analyses, the logrank test compared Kaplan-Meier curves, and multivariable Cox regression models measured associations with outcomes.
Results: Myocardial fibrosis was prevalent in OSA, where 97 of 324 (30%) patients exhibited elevated ECV > =30%. While patients with OSA exhibited higher ECV compared to healthy controls (27.6% (IQR 25.2-30.5) versus 24.1% (IQR 23.1%-25.7%, p = <0.001)), ECV in OSA was not significantly higher compared to patients with known or suspected heart disease referred for CMR (p = 0.45). OSA disease severity measured by apnea-hypopnea index (AHI) was weakly correlated with left ventricular mass index (R2 = 0.03, p = 0.01), systolic blood pressure (R2 = 0.03, p = 0.049), and diastolic blood pressure (R2 = 0.034 p = 0.04). AHI was not significantly associated with ECV, end-diastolic volume index, ejection fraction, B-type Natriuretic Peptide, Glomerular Filtration Rate (GFR), or HHF/mortality events (p < =0.09 for all).
Over a median follow-up of 2.3 years (IQR 1.2-3.2 years) there were 62 HHF events and 108 deaths after CMR in 147 individuals. There was no difference in rates of HHF or death (p = 0.15) in patients with or without OSA (Figure 1). Yet, in OSA patients, myocardial fibrosis was significantly associated with increased event rates (Figure 2) adjusting for age, GFR, myocardial infarction size, ejection fraction (HR 1.6 95%CI 1.2-2.2 for every 5% increase in ECV, p = 0.003).
Conclusions: Myocardial fibrosis is prevalent in OSA and associated with HHF or death. Quantifying myocardial fibrosis in OSA patients may improve cardiovascular risk stratification and provide a phenotype for targeted therapeutic intervention.
Figure 1.
Figure 2.
1496
Myocardial substrates underlyng early ventricular arrhythmias in st-elevation acute myocardial infarction: the role of cardiac magnetic resonance
Abstract
Background: Is well note that scar detected by Cardiac Magnetic Resonance (CMR) is a predictor of late and no early ventricular arrhythmias due to re-entry mechanism. Recent studies suggested that myocardium salvaged may represent a substrate for early one. CMR is able to individualize myocardial area at risk and salvaged myocardium.
Aims: Analyze if substrates detected by CMR are able to predict arrhythmias in patients with ST-segment elevation myocardial infarction.
Methods: We enrolled consecutive patients presenting acute coronary syndrome with persistent ST-segment elevation and treated by primary PCI within 12 hours after symptoms onset and who underwent a complete CMR protocol including edema and LGE sequences within 10 days from admission. We excluded patients with late myocardial infarction (>12 h), pre-infarction angina and poor image quality and not eligible for CMR. All patients were monitored to detect early (<48 hours) major ventricular arrhytmias as ventricular fibrillation (VF) and susteined ventricular tachycardia (SVT).
Results: 30 patients (75% male, median age 63 y.o.) were enrolled. During first 48 hours 8 patients had at least one major arrhythmias (3 SVT, 3 VF pre PCI and 3 VF post PCI). The overall population were divided into two groups in according with the presence (8) or absence (22) of arrhythmias. Time pain-to-balloon (233 vs 170 minutes), Troponin I peak (72 vs 97 ug/L) and culprit lesion were not different between the two groups. The arrhythmic one had a quite normal left ventricle ejection fraction compared to the non arrhythmic one (55% vs 49%, p = 0.045). The presence and the global amount of edema, myocardial necrosis (LGE), microvascular obstruction and haemmoragic area were not difference into the two groups. The only parameter that demonstrate a significant association with the presence of early ventricular major arrhythmias was salvaged myocardium (21% vs 5%, p = 0.035). An amount of 10% salvaged myocardium was the critical substrate mass to determine the presence early ventricular arrhythmias (specificity 77%, sensitivity 75%).
Conclusion: From our data not area at risk but a critical mass of salvaged myocardium was associated with the presence of major early ventricular arrhythmias. Probabily due to an heterogenic substrate related to peri-infarct region.
Figure 1.
1416
Cardiac magnetic resonance predicts atrial fibrillation occurrence in patients with hypertrophic cardiomyopathy
Abstract
Objectives: Patients with hypertrophic cardiomyopathy (HCM) are known to be more prone to atrial fibrillation (AF). We sought to identify predictors of AF onset in HCM patients with the use of cardiac magnetic resonance (CMR).
Methods: We prospectively enrolled 259 patients with HCM, no prior history of atrial arrhythmia, and no AF episode on a 24-hour Holter performed at inclusion. 165 patients (63%) underwent CMR using Cine SSFP and Late gadolinium enhanced (LGE) imaging (3D IR FLASH). On the left ventricle (LV), maximal wall thickness, LV volumes and ejection function (EF) were measured. LGE images were analyzed to detect LV fibrosis, and LGE patterns were categorized as either ischemic-like or non-ischemic, the latter being further characterized as either midwall patchy LGE in hypertrophied segments or focal LGE on LV-RV junctions. In addition left atrial (LA) dimensions and volumes (maximum, pre-A and minimum volumes) were measured and used to compute the following functional parameters: LA total ejection fraction (EF), conduit fraction, active emptying fraction, and expansion index. Patients were followed clinically and with 24-hour Holter recordings every 6 months. Univariable and multivariable regression analyses were performed to identify clinical and imaging predictors of AF occurrence.
Results: During a mean follow up of 27 ± 11 months 31 patients (12%) developed AF. In the CMR group AF was detected in 19 patients (12%). Characteristics in patients with vs. without AF occurrence are shown in the Table. Sequential Cox multivariable analysis, after elimination of potential multicollinearity, identified CMR independent predictors of AF occurrence: LA diameter (HR = 1.09, 95%CI: 1.01-1.16, P = 0.02), LA EF (HR = 0.94, 95%CI: 0.90-0.99, P = 0.008) and LV LGE (HR = 4.53, 95%CI: 1.02-20.06, P = 0.046). Among clinical characteristics, age was the only independent predictor (HR = 1.03, 95%CI: 1.00-1.06, P = 0.03). Univariable analysis documented LGE at RV-LV junctions as the pattern with the closest relationship with AF (71% in pts with AF occurrence vs. 37% without, P = 0.006).
Conclusion: LA size, LA emptying function and LV fibrosis on CMR are independent predictors of AF occurrence in patients with HCM. The mechanisms linking the ventricular HCM phenotype to AF should be analyzed in future studies.
1469
T1 and T2 mapping cardiovascular magnetic resonance to monitor inflammatory activity in patients with myocarditis
Abstract
Background: Myocarditis is a heterogeneous disease with a large spectrum of possible clinical courses ranging from complete healing to end stage heart failure. Thus, there is a need for reliable tools to monitor disease activity and to identify patients with persistent inflammation who are at risk for an adverse outcome. T1 and T2 mapping are emerging cardiovascular magnetic resonance (CMR) tools for quantitative tissue characterization in myocarditis. This study evaluated the use of T1 and T2 mapping CMR to monitor disease activity during a 12-month follow-up period after first manifestation of myocarditis.
Methods: This study included 37 patients with the first clinical manifestation of rigorously defined acute myocarditis, who underwent CMR at 1.5 Tesla at baseline (BL) in the acute phase, after 3 months (FU-I, n = 27) and after 12 months (FU-II, n = 16). T1 and T2 mapping CMR were performed in addition to a conventional CMR myocarditis protocol. T1 quantification was performed on three short axes using a modified Look-Locker inversion-recovery (MOLLI) variant before and 15 minutes after administration of 0.075 mmol/kg Gadolinium-BOPTA. T2 quantification was performed using a free-breathing, navigator-gated gradient- and spin-echo hybrid sequence (GRASE). Maps were generated using a dedicated plug-in for the OsiriX software to calculate myocardial T2, native/post-contrast T1 and extracellular-volume-fraction (ECV) values.
Results: Median left ventricular ejection fraction (LVEF) improved from 56 (32-67) % at BL to 63 (49-68) % at FU-I (p < 0.01). Median global myocardial T2 significantly decreased from 61 (57-64) ms at BL to 58 (55-63) ms at FU-I (p < 0.05). Similarly, median global myocardial native T1 decreased from 1100 (1070-1135) ms at BL to 1059 (1039-1113) ms at FU-I (p < 0.05). Furthermore, global myocardial ECV was significantly lower at FU-I with 26 (24-30) % compared to 29 (26-32) % at BL (p < 0.05). However, there were no further significant changes in LVEF, global myocardial T2, native T1 and ECV during the further course from FU-I to FU-II.
Conclusion: T1 and T2 mapping CMR techniques offer novel insights into the course of myocardial inflammation in myocarditis. Our findings suggest that the major decrease of inflammatory activity in myocarditis occurs within the first 3 months after first clinical manifestation. T1 and T2 mapping techniques have great potential for monitoring disease activity in myocarditis.
1480
Impact of electronic coaching on cardiovascular risk reduction in a high-risk primary prevention population – A cardiovascular magnetic resonance sub-study
Abstract
Background: Cardiovascular disease (CVD) remains one of the main causes of mortality globally. Innovative techniques are required to tackle the anticipated rise in CVD due to rising obesity, diabetes and an ageing population. Personalized electronic (e-) coaching utilizes platforms such as the internet and email to help motivate healthy living. Cardiovascular magnetic resonance (CMR) imaging offers a range of powerful imaging parameters that can be used as outcome measures in clinical trials at reduced cost and follow-up duration and can inform performance of larger scale randomized controlled trails (RCT).
Aims: The aim of the study was to assess the clinical effectiveness of e-coaching in a high risk, asymptomatic primary prevention cohort using surrogate markers of left ventricular (LV) mass and pulse wave velocity.
Methods: Between July 2013 and November 2015 we conducted a single center RCT of 402 participants, comparing e-coaching in addition to standard of care (SOC) vs. SOC alone. E-coaching included a personalized webpage with heart health advice for each subject based on their individual lifestyle and risk factors, progress charts with goals and achievements and email reminders with tips. The SOC consisted of face-to-face advice on risk factor and medication based on ESC and NICE guidelines.
Estimated 10-year CVD risk of 10% or more was required with no prior history of CVD. In the sub-study CMR (Philips 1.5T) along with Vicorder device based carotid-femoral pulse wave velocity (PWV) and CVD risk assessment was performed at baseline and 6 month follow-up in the first 100 participants (1:1 allocation based on treatment arm) consenting to CMR and with no scanning contraindications.
Results: Average age was 65.8 years in the e-coaching and 64.1 in the SOC group. See table 1 for other baseline data. Compared to baseline both groups showed reduction in LV mass (e-coach -2.01g vs. -1.27g, p= 0.74, figure 1) and PWV (e-coach -0.30 m/s vs. -0.77m/s, p= 0.08). Findings were consistent when LV mass was indexed to body surface area. Global risk scores, body mass index, glucose and blood pressure also showed improvement but with no statistical difference between the two groups.
Conclusion: Personalized e-coaching does not show clinical effectiveness in CVD risk reduction when combined with current standard of care. Evidence of effectiveness of a potentially costly health technology is vital before recommending its widespread use in resource limited health care systems.
Figure. Change in LV mass in the e-coaching and standard of care groups over 6 months.
1598
Anatomical and functional evaluation of postinterventional pulmonary vein stenosis by magnetic resonance imaging
Abstract
Background: Pulmonary vein stenosis (PVS) after radiofrequency ablation for atrial fibrillation remains a challenge despite considerable technological advances. We aimed to evaluate anatomical and functional PVS assessment by MRA and MR perfusion imaging.
Material and Methods: 57 patients (61 studies) with a known PVS were evaluated. Patients were scanned in a 1.5T MRI scanner. All patients received a contrast-enhanced three-dimensional angiography and high-resolution k-t accelerated CMR lung perfusion was performed in 28 patients. In 50 patients previous studies (MRA or CTA) were available for comparison. Three dimensional reconstructions of the left atria were performed and pulmonary vein branches were assessed. Pulmonary perfusion deficits were correlated to PV subbranches.
Results: MRA was safe in all patients. Underlying heart rhythm did not affect image quality. In all patients with a perfusion deficit the corresponding PV subbranch could be ientified. In 3 patients without previous imaging MRA alone did not allow detect occlusion of a subbranch but a corresponding perfusion deficit existed. In one patient a perfusion deficit could be diagnosed despite PV collateralization.
Conclusions: Anatomical and functional PVS assessment by MRA and MR perfusion imaging has the potential to improve diagnosis and evaluation of long term outcome. MRA has the potential to miss occlusion of PV subbranches and can lead to inaccurate diagnoses.
Figure. (A) First dynamic of contrast-enhanced three-dimensional cardiovascular magnetic resonance angiography of the pulmonary veins (p.a.view) showing total occlusion of the left superior PV (B) High-resolution k-t accelerated cardiovascular magnetic resonance lung perfusion imaging identified an extensive perfusion deficit of the left upper lobe (coronal orientation).
1364
Reduced infarct-adjacent wall thickening and impaired restperfusion in the area at risk of successfully reperfused acute myocardial infarction
Abstract
Background: Myocardial infarction (MI) leads to heart failure in a substantial number of patients. Investigation of the infarct-adjacent segments can offer better insights into left ventricular remodeling. Our study aims to quantify radial wall strain (WS) and myocardial blood flow (MBF) with cardiac magnetic resonance (CMR) in the infarct zone (IZ), borderzones (BZ) and remote myocardium (RM) in a large patient population in the acute phase of MI.
Methods: In this substudy of the NOMI-trial (ClinicalTrials.gov identifier: NCT01398384) using the placebo arm, 68 patients underwent CMR with T2-weighted, cine, dual bolus first pass perfusion, and gadolinium enhancement (LGE) imaging within 48 to 72 hours of acutely revascularized ST-elevation MI in a 1.5 tesla MR scanner. TIMI 2-3 flow was achieved in all patients after PCI. Corresponding basal and midventricular SAX slices of the CMR acquisitions were aligned and divided into 12 corresponding equiangular (30°) segments. WS was measured as the percentage of radial wall thickening, absolute MBF was quantified using Fermi deconvolution. Values for the IZ were calculated as the mean of segments with LGE. The infarct-adjacent 30 degree segments were defined as BZ-inAAR or BZ-outAAR, depending on their location resp. in- or outside the area at risk on T2-weighted imaging. Values for the RM were calculated as the mean of the remaining segments. Apical slices seldomly included more than one myocardial zone and were excluded from analysis.
Results: One hundred thirty six slices were analyzed. WS was impaired to a similar extent in BZ-inAAR (28.5%, 12.0-49.5) and BZ-outAAR (29.0%, 10.7-46.2, p = 0.141) compared to RM (49.5%, 34.9-59.7, p < 0.003), and even more in the IZ (15.1%, 1.6-29.5, p < 0.003). MBF was impaired in the BZ-inAAR and IZ only (0.39 ml/g/min, 0.33-0.53 and 0.30 ml/g/min, 0.22-0.41 resp.), compared to BZ-outAAR and RM (0.49 ml/g/min, 0.35-0.62 and 0.44 ml/g/min, 0.34-0.59 resp., p < 0.003).
Conclusions: WS is impaired in the infarct-adjacent segments regardless of perfusion, so probably as a result of geometric tethering. Myocardial restperfusion in the area at risk is still reduced in successfully revascularized acute MI.
Figure. WS is presented on the left, MBF on the right. WS is decreased to a similar extent in both borderzones(left), although perfusion is normal in the BZ-outAAR, whereas MBF in the BZ-inAAR is significantly impaired (right). * = p < 0.003, ns = not significant.
1580
Correlation between circulating microRNA 29 and diffuse myocardial fibrosis, assessed by T1 mapping, in patients affected by non ischemic dilative cardiomyopathy
Abstract
Background: Myocardial fibrosis is a fundamental process in the development of myocardial dysfunction in cardiomyopathies, leading to myocardial remodeling and poor outcome. Circulating microRNAs could prove to be novel biomarkers for cardiovascular diseases. In particular, miR-29 is the best-characterized direct regulator of extracellular matrix protein synthesis. Recently, it has been associated with diffuse myocardial fibrosis in hypertrophic cardiomyopathy (HCM).
Aim of the Study: To assess the correlation between microRNA 29a and diffuse interstitial myocardial fibrosis in patients affected by non-ischemic dilative cardiomyopathy (NIDCM).
Methods: Eleven consecutive patients who were referred for screening to Policlinico San Donato, Cardiac Imaging Unit, were prospectively enrolled. All patients underwent CMR, performed with LGE quantification to detect regional fibrosis and T1 mapping for diffuse fibrosis. Peripheral blood samples were taken before CMR exam to obtain hematocrit, useful for extracellular volume
(ECV) calculation. Plasma samples (10 min centrifugation at 400 g followed by a further 10 min at 600 g) were stored at −80 °C for RNA isolation.
Results: Patients affected by NIDCM were 42% males (57,8 ± 13,5 years old), Left Ventricular End Diastolic Volume (LVEDV index ml/m2) was 112,7 ± 15, Left ventricular mass index was 104 ± 9,2, mean LV Ejection Fraction was 36,9 ±13,3, LGE % was 3,7± 7,7. No significant difference was found in microRNA 29a levels between DCM and controls (n = 42) (1,10± 0,7 vs 1,25 ± 0,94, p:ns). In DCM patients, an inverse correlation between microRNA29a and left ventricular stroke volume (SV) was found (Rho = -0,711, p:0,021). Mean septum ECV was 0,33 ± 0,03. No correlation was found between microRNA29a levels and macroscopic fibrosis, as assessed by LGE percentage. However, we found a strong inverse correlation, in DCM patients, between microRNA 29 and ECV (Rho = -0,672, p:0,047), and a tendency of negative correlation with blood T1 mapping post (Rho = -0,594, p:0,092).
Conclusions: For the first time it has been reported a strong negative correlation between microRNA-29 a and ECV in NIDCM patients, probably reflecting its role as a negative regulator of genes encoding collagens and extracellular matrix proteins. This finding suggests that microRNA-29a could be a circulating biomarker for early detection of diffuse myocardial fibrosis, as assessed by T1 mapping. A larger sample size is needed to confirm our data and to determinate the role of this biomarker in DCM patients.
1435
Association of Smoking with Myocardial Injury and Clinical Outcome in Patients Undergoing Mechanical Reperfusion for ST-Elevation Myocardial Infarction
Abstract
Aims: There is evidence suggesting a positive effect of cigarette smoking on myocardial tissue reperfusion and clinical outcomes in patients with myocardial infarction (“smoker's paradox”). We aimed to evaluate the relationship of smoking with cardiac magnetic resonance (CMR)-determined myocardial salvage and damage as well as clinical outcome in patients undergoing primary percutaneous coronary intervention (PPCI) for ST-elevation myocardial infarction (STEMI).
Methods and Results: This multicenter study included 727 consecutive STEMI patients reperfused within 12 hours after symptom onset. CMR imaging parameters (area-at-risk [AAR], infarct size [IS], myocardial salvage index [MSI], microvascular obstruction [MVO]) and intramyocardial haemorrhage [IMH] were compared according to admission smoking status. Major adverse cardiac events (MACE) rates at 12 months after infarction were compared between groups. Three hundred and thirty-nine (46.6%) patients were current smokers. There was no difference in the extent of AAR (35 [24-47] vs. 37 [27-49] % of left ventricular volume [LV], p = 0.10), IS (16 [8-25] vs. 17 [10-26] %LV, p = 0.21), MSI (53 [29-70] vs. 52 [34-71], p = 0.47), MVO (0 [0-1.7] vs. 0 [0-1.6] %LV, p = 0.91), or in the frequency of IMH (42% vs. 39%. p = 0.58) between smokers and non-smokers. Smokers had lower MACE (3.8% vs. 8.2%, p = 0.01) rate. However, adjustment for differences in baseline risk factors, attenuated the association of smoking with MACE markedly (HR = 0.71, 95% CI 0.36 to 1.38, p = 0.31).
Conclusion: Smoking is not associated with PPCI efficacy (myocardial salvage) or irreversible myocardial damage in patients with STEMI. The lower MACE rate of smokers was entirely explained by differences in baseline risk characteristics, thus challenging the existence of a “smoker's paradox”.
1640
Assessing the risk of late cardiotoxicity in low risk breast cancer survivors receiving contemporary anthracycline treatment: a 6 year 100 patient study
Abstract
Background: Chemotherapy cures patients with cancer but anthracyclines (AC) are cardiotoxic. In clinical practice left ventricle ejection fraction (EF) changes are used to detect cardiotoxicity. However, routine surveillance should not be limited to assess EF alone due to EF limitation.
Objectives: To explore the late cardiac effects of AC using advanced cardiac imaging and biomarkers in low-risk breast cancer survivors.
Methods: 132 women from a former study of AC (intermediate dose – mean dose 399 ± 85.4 mg/m2) breast cancer therapy were traced. After exclusions (died, recurrent disease, pregnancy, did not wish to attend), 98 attended. These had been selectively recruited to have no cardiovascular risk factors and no previous radio/chemo-therap. All subjects underwent baseline pre-chemotherapy cardiac magnetic resonance (CMR) imaging. Subjects were compared with an age-matched control group of healthy volunteers (HV, n = 34) at follow-up. All underwent resting ECG, blood biomarkers, advanced echocardiography and CMR structure, function (with the original 2005 protocol) and T1 mapping (MOLLI) for ECV quantification.
Results: 98 women attended for follow up (age 55 ± 8.5 years, mean 6 years post AC treatment), and no participants reported cardiac symptoms or fulfilled criteria for cardiotoxicity. Across the AC group, EF fell by -3.7 ± 4.2% (range from -13 to +8%) over 6 years follow-up, although all measurements remained within the normal range. Patients with higher dose of AC (≥450mg/m2) had greater EF reduction compare to lower doses (4.6 ± 3.7% versus 2.7 ± 4.4%, p = 0.030). There were no differences in standard CMR parameters between the AC and HV groups, except the LA was larger in the AC group. Native myocardial T1 was greater in AC (1052 ± 30 vs 1037 ± 31 msec, p = 0.012). Mean ECV calculations were similar (0.28 ± 0.029 vs 0.28 ± 0.029, p = 0.803), however, values were slightly higher when left (vs right) radiotherapy had been performed (0.29 ± 0.030 vs 0.28 ± 0.023, p = 0.017). Echocardiography found lower systolic and diastolic myocardial velocities and global longitudinal strain parameters - but within the normal reference range - with greatest reductions in the AC group who had experienced an EF drop ≥5% by CMR.
Conclusion: Chemotherapy with AC causes a small non-clinical reduction in the cardiac function (however measured) with a dose-response relationship, as does left sided radiotherapy. These changes are however small and may be missed by EF measurements alone.
Figure. Left: AC subjects exposed to higher anthracycline doses shower greater EF reduction compared with low dose. Centre: myocardial native T1 was higher in the AC patients. Right: No statistically differences in the myocardial ECV mean values were observed.
1511
Risk stratification in sarcoidosis: Incidence of cardiac sarcoidosis in individuals diagnosed with extra-cardiac disease by cardiovascular magnetic resonance
Abstract
Introduction: Cardiac death is the leading cause of mortality in patients with sarcoidosis. In those with extra-cardiac sarcoidosis diagnosis, little is known of the incidence of cardiac sarcoidosis (CS). Myocardial infiltration of cardiac sarcoidosis is a marker in risk stratification of these patients. Late gadolinium enhancement (LGE) by cardiovascular magnetic resonance (CMR) imaging has the ability to identify the myocardial infiltration of cardiac sarcoidosis (CS). We therefore investigated the incidence of CS in patients with extra-cardiac sarcoidosis without cardiac manifestations using LGE-CMR.
Methods: 72 patients with clinically and radiologically diagnosed pulmonary sarcoidosis, without signs or symptoms of cardiovascular involvement, were enrolled over a three-year period. LGE-CMR was performed on all patients. The presence of LGE in the left ventricular myocardium was considered diagnostic for CS. Patients were classified as CS or non-CS based on the LGE-CMR findings.
Results: Mean age was 51 ± 19 years, with a female predominance (61%). CS was detected in 19 patients (23%). When compared to those patients without LGE, patients with CS had higher rates of previously documented non-sustained ventricular tachycardia (22% vs. 5%), a greater prevalence of an abnormal ECG (52% vs. 18%) and a higher heart rate during CMR (82bpm vs. 72bpm).
Conclusion: CMR should be considered as an adjunct to conventional diagnostic workup in all patients with pulmonary sarcoidosis. Long-term follow-up of consecutive patients with isolated CS is needed to determine the natural history and rates of ventricular arrhythmias.
1334
Patterns of late gadolinium enhancement in Brugada syndrome
Abstract
Background: There is increasing evidence to suggest that the Brugada ECG pattern is a marker of subtle structural heart disease. We characterised a cohort of patients with Brugada syndrome (BrS) using cardiac magnetic resonance (CMR) late gadolinium enhancement techniques.
Methods: BrS was diagnosed according to international guidelines. CMR data from 78 BrS patients (44 + /-15 years; 64% male) was compared with data from 78 normal controls (42 + /-14 years; 64% male).
Results: Right ventricular (RV) ejection fraction was lower (61 + /-8% vs 64 + /-5%; p = 0.004) and RV end-systolic volume was greater (31 + /-10ml/m2 vs 28 + /-6ml/m2; p = 0.015) in BrS patients compared with normal controls, but these values remained within the normal range. Overall LGE was demonstrated in 8% BrS patients (midwall LGE in 5%) and no normal controls (p = 0.028; figure 1). In BrS patients with midwall LGE there were no other features of cardiomyopathy at the time of CMR but further genetic testing has revealed a desmoplakin mutation in one patient and evolution of T wave inversion throughout all the precordial ECG leads during two year follow-up in another. Both of these are major criteria for arrhythmogenic right ventricular cardiomyopathy (ARVC), but as yet neither patient fulfills diagnostic criteria for this condition.
Conclusion: Some BrS patients demonstrate midwall LGE consistent with an underlying cardiomyopathic process. This lends further support to the presence of subtle structural abnormalities in BrS. It suggests a degree of phenotypic overlap with ARVC in some cases and may serve as an early marker for evolution of a cardiomyopathic phenotype over time.
Figure 1. Cardiac magnetic resonance imaging in Brugada syndrome (BrS). Panels A and B demonstrate late gadolinium enhancement (LGE) in a 45 year old male patient with BrS. Patchy midwall LGE is demonstrated in the mid and apical lateral wall, basal and apical inferolateral wall of the left ventricle and mid anteroseptum. Panels C and D demonstrate midwall LGE in the mid inferoseptum of his 53 year old sister who also has BrS. Panels E and F demonstrate midwall LGE in the basal and mid septum and anteroseptum of a 40 year old man with BrS who was subsequently found to have a desmoplakin gene mutation.
1591
Detailed Left Atrial Assessment in Anderson Fabry Disease
Abstract
Introduction: Left atrial (LA) function is increasingly recognised as an important marker in multiple cardiac pathologies. Left ventricular hypertrophy and diastolic dysfunction are common in Fabry Disease (FD), and histological LA abnormalities are identifiable post-mortem. The aim of this study was to assess LA function in a cohort of patients with FD, with and without left ventricular hypertrophy.
Methods: 25 genetically-confirmed patients with FD (15 without LVH, 10 with LVH, all in sinus rhythm) and 15 age and gender matched healthy volunteers (HV); FD (13 male) 46.2 ± 13.8 years, HV (7 male) 39.7 ± 14.3 years, underwent CMR at 1.5T (Avanto, Siemens AG, Erlangen, Germany). A standardised clinical CMR protocol was used in all. Biplane LA volumes were obtained from 4- and 2-chamber SSFP cine images (cmr42, Circle Cardiovascular Imaging Inc., Calgary, Canada). Mean atrial strain and strain-rate parameters were also derived from these sequences using dedicated CMR feature tracking software (Diogenes® TomTec, Germany).
Results: Results are shown in table 1. Those with LVH were older, reflecting the natural history of FD. LA volumes and volume-derived left atrial functional parameters (total, passive and active ejection fractions and LA expansion index) did not differ between LVH-negative FD patients and HVs. However those with LVH had significantly increased volumes, with reduction in total and passive ejection fractions and reduced LA expansion index, reflecting reduced conduit and reservoir functions.
Those without LVH had significantly reduced total atrial strain and early negative strain rate, reflecting reduced LA and LV compliance, respectively. Early passive strain and negative strain rate, and LA expansion index reduced further in LVH positive patients, reflecting worsening conduit and reservoir function. Strain markers of booster pump function, known to increase in hypertrophic cardiomyopathy, showed no difference between the groups.
Discussion: LA function, particularly conduit function which reflects LV compliance, is impaired in Fabry disease, even in the absence of left ventricular hypertrophy and increased LA volume. Disturbances in volumetric indices of LA function were seen only in those with LVH, and likely reflect a later stage of disease. LA strain parameters of conduit and reservoir function, however, may have a role in the detection of early cardiac involvement in Fabry Disease
. | Group A: Healthy Volunteers (n = 15) . | Group B: Fabry Disease, No LVH (n = 15) . | Group C: Fabry Disease, LVH (n = 10) . | p- value (Group A vs. B) . | p-value (Group A vs. C) . | p-value (Group B vs. C) . |
---|---|---|---|---|---|---|
Gender | Male = 7 | Male = 8 | Male = 5 | |||
Age (years) | 39.7 ± 14.4 | 40.1 ± 12.4 | 57.4 ± 7.9 | |||
Maximum LV Wall Thickness (mm) | 9.6 ± 0.9 | 10.3 ± 3.1 | 17.3 ± 6.1 | 0.43 | <0.0001 | 0.002 |
LV Ejection Fraction (%) | 63.8 ± 2.7 | 66.5 ± 6.2 | 68.0 ± 10.5 | 0.14 | 0.16 | 0.68 |
LA VOLUMES | ||||||
LA Max Volume/BSA(ml) | 39.3 ± 12.1 | 37.7 ± 12.9 | 49.9 ± 10.3 | 0.73 | 0.06 | 0.04 |
LA Min Volume/BSA(ml) | 16.3 ± 5.8 | 15.7 ± 5.1 | 27.6 ± 7.0 | 0.77 | 0.0007 | 0.0002 |
LA Pre-atrial Contraction Volume/BSA (ml) | 26.4 ± 8.2 | 27.0 ± 8.7 | 40.7 ± 9.6 | 0.84 | 0.0002 | 0.003 |
LA FUNCTION | ||||||
Total Ejection Fraction (%) | 58.8 ± 6.5 | 58.1 ± 4.0 | 44.6 ± 9.1 | 0.72 | 0.0004 | <0.0001 |
Reservoir | ||||||
LA expansion index (%) | 1.48 ± 0.36 | 1.40 ± 0.23 | 0.85 ± 0.31 | 0.52 | 0.0007 | 0.0001 |
Total strain (%) | 34.86.6 | 28.25.4 | 24.45.6 | 0.005 | 0.002 | 0.15 |
Peak systolic strain rate (-1) | 1.31 ± 0.47 | 1.08 ± 0.27 | 1.03 ± 0.22 | 0.12 | 0.16 | 0.65 |
Conduit | ||||||
Passive Ejection Fraction (%) | 32.67.4 | 27.111.2 | 18.68.4 | 0.12 | 0.0007 | 0.09 |
Passive Strain (%) | 20.1 ± 6.1 | 17.4 ± 6.0 | 11.0 ± 4.4 | 0.24 | 0.002 | 0.02 |
Peak early negative strain rate (-1) | -1.07 ± 0.31 | -0.84 ± 0.31 | -0.5 ± 0.16 | 0.04 | 0.0002 | 0.01 |
Booster Pump | ||||||
Active Ejection Fraction | 38.6 ± 9.1 | 41.4 ± 9.6 | 32.0 ± 8.7 | 0.42 | 0.12 | 0.04 |
Active Strain (%) | 13.8 ± 5.6 | 11.1 ± 4.2 | 13.5 ± 4.0 | 0.13 | 0.86 | 0.22 |
Peak late negative strain rate (-1) | -1.0 ± 0.37 | -0.85 ± 0.45 | -0.95 ± 0.29 | 0.34 | 0.78 | 0.59 |
. | Group A: Healthy Volunteers (n = 15) . | Group B: Fabry Disease, No LVH (n = 15) . | Group C: Fabry Disease, LVH (n = 10) . | p- value (Group A vs. B) . | p-value (Group A vs. C) . | p-value (Group B vs. C) . |
---|---|---|---|---|---|---|
Gender | Male = 7 | Male = 8 | Male = 5 | |||
Age (years) | 39.7 ± 14.4 | 40.1 ± 12.4 | 57.4 ± 7.9 | |||
Maximum LV Wall Thickness (mm) | 9.6 ± 0.9 | 10.3 ± 3.1 | 17.3 ± 6.1 | 0.43 | <0.0001 | 0.002 |
LV Ejection Fraction (%) | 63.8 ± 2.7 | 66.5 ± 6.2 | 68.0 ± 10.5 | 0.14 | 0.16 | 0.68 |
LA VOLUMES | ||||||
LA Max Volume/BSA(ml) | 39.3 ± 12.1 | 37.7 ± 12.9 | 49.9 ± 10.3 | 0.73 | 0.06 | 0.04 |
LA Min Volume/BSA(ml) | 16.3 ± 5.8 | 15.7 ± 5.1 | 27.6 ± 7.0 | 0.77 | 0.0007 | 0.0002 |
LA Pre-atrial Contraction Volume/BSA (ml) | 26.4 ± 8.2 | 27.0 ± 8.7 | 40.7 ± 9.6 | 0.84 | 0.0002 | 0.003 |
LA FUNCTION | ||||||
Total Ejection Fraction (%) | 58.8 ± 6.5 | 58.1 ± 4.0 | 44.6 ± 9.1 | 0.72 | 0.0004 | <0.0001 |
Reservoir | ||||||
LA expansion index (%) | 1.48 ± 0.36 | 1.40 ± 0.23 | 0.85 ± 0.31 | 0.52 | 0.0007 | 0.0001 |
Total strain (%) | 34.86.6 | 28.25.4 | 24.45.6 | 0.005 | 0.002 | 0.15 |
Peak systolic strain rate (-1) | 1.31 ± 0.47 | 1.08 ± 0.27 | 1.03 ± 0.22 | 0.12 | 0.16 | 0.65 |
Conduit | ||||||
Passive Ejection Fraction (%) | 32.67.4 | 27.111.2 | 18.68.4 | 0.12 | 0.0007 | 0.09 |
Passive Strain (%) | 20.1 ± 6.1 | 17.4 ± 6.0 | 11.0 ± 4.4 | 0.24 | 0.002 | 0.02 |
Peak early negative strain rate (-1) | -1.07 ± 0.31 | -0.84 ± 0.31 | -0.5 ± 0.16 | 0.04 | 0.0002 | 0.01 |
Booster Pump | ||||||
Active Ejection Fraction | 38.6 ± 9.1 | 41.4 ± 9.6 | 32.0 ± 8.7 | 0.42 | 0.12 | 0.04 |
Active Strain (%) | 13.8 ± 5.6 | 11.1 ± 4.2 | 13.5 ± 4.0 | 0.13 | 0.86 | 0.22 |
Peak late negative strain rate (-1) | -1.0 ± 0.37 | -0.85 ± 0.45 | -0.95 ± 0.29 | 0.34 | 0.78 | 0.59 |
1634
Role of cardiac magnetic resonance in the diagnosis of ARVC/D mimics
Abstract
Introduction: ARVC/D is a rare, genetic heart disease, with autosomal dominant trasmission. Diagnostic 2010 Task Force Criteria include CMR. The purpose of our study is to evaluate the role of CMR in patients with suspected ARVC/D in confirming/ruling out the diagnosis or identify mimics of ARVC/D.
Materials and Methods: Consecutive subjects referred for a clinical CMR during a 12-month period we included (N = 2481). We identified patients referred for CMR with suspected ARVC/D on the basis of symptoms and clinical presentation, family history of ARVC/D or sudden cardiac death, abnormal ECG or TTE.
Results: N = 124 patients (5% of the 12-month CMR referral cohort) were referred with suspected ARVC/D were identified. In 14% of the patients (n = 17) CMR was able to identify the final diagnosis. In addition CMR was able to establish in large part of patients a normal heart structure. N = 4 patients (3%) met criteria for ARVC/D: n = 3 major and n = 1 minor, n = 11 patients (8.8%) had disease could mimic ARVC/D “ARVC/D mimics”, n = 2 patients had evidence of previous MI and n= 70 (56%) had a normal heart structure.
Additional diagnosis identified were congenital absence of pericardium (n = 1) (Fig.1), dilated cardiomyopathy (n = 1), left ventricular non compaction cardiomyopathy (n = 1), left ventricular non compaction cardiomyopathy plus myocardial infarct (n = 1), arrhythmogenic left ventricular cardiomyopathy (ALVC, variant of ARVC/D) (n = 1), anomalous venous return (n = 1) (Fig.1), atrial septal defect with left to right shunting (n = 1) (Fig.2), asymmetric pectus excavatum (n = 1) (Fig.2), athletic heart (n = 2) and sarcoidosis (n = 1), all mimicking ARVC/D.
Conclusions: Our study demonstrates the role of CMR not only in detecting abnormalities compatible with ARVC/D but also in identifying a variety of diseases that can mimic ARVC/D.
Figure 1. (A)(B) Congenital absence of pericardium. (A) 4-chamber view, heart is
displaced in the left side of the thorax, with apex pointing posteriorly; axial HASTE, lung interposed between the ascending aorta and the pulmonary artery (arrow) (B). (C)(D) Anomalous venous return. (C) 4-chamber view, RV dilated; (D) left upper pulmonary vein (arrow-head) on the left side of aortic arch draining upwards toward the brachiocephalic trunk.
Figure 2. (A)(B) ASD. Large superior interatrial defect is seen with blood flowing from the left to the right (arrow-head) (C)(D) Asymmetric pectus excavatum. Right ventricle distorted by the abnormal chest morphology (arrow)
1321
Comparison of transtlioracic ecliocardiography versus cardiac magnetic for implantable cardioverter defibrillator therapy in primary prevention strategy dilated cardiomyopathy patients
Abstract
Background: Implantable cardioverter-defibriIIators(ICDs) has been proved as a valid primary prevention strategy to reduce mortaIity in patients with dilated cardiomyopathy (DCM) with reduced Ieft ventricula rejection function (LVEF) <35%. Cardiac magnetic resonance (CMR) is now considered the gold standard technique for LVEF assessment and it provides important information on tissue characterization such as late gadoIinium enhancement (LGE). Several studies have shown differences between CMR and TIE evaluation. The aim of this study is to determine whether LV evaIuation and LGE detection by CMR are superior to conventional TIE measurements for risk stratification of DCM patients evaluated for ICD impIantation in primary prevention strategy.
Methods: 270 consecutive DCM patients (Meanage 63 ± 13 yo, maIe 220 patients) referred to our Institution to be evaluated for ICD impIantation in primary prevention were enroIIed. All patients under went both the and CMR left ventricle end-diastoIic (LVEDV) and end-systoIic (LVESV) voIumes and LVE festimation. Additionally, LGE. presence was a isodetected by CMR. All patients were foIIowed-up for the major adverse cardiacevents (MACE) defined as a combined end point of ventricular tachycardia, ventricular fibrillation and sudden cardiacdeath.
Results: All patients performed both tests successfully. The mean follow-up was 850 ± 330 days. The showed a lower LVEDV (86 ± 28 vs. 131 ± 41 ml/m2) and LVEDV (57 ± 21 vs. 93 ± 40 ml/m2) and a higher LVEF (35 ± 10 vs. 31 ± 9 %) as compared to CMR (p < 0.0001). MACE occurred in 68 patients (25 %). Patients experienced MACE showed a higher LVEDV-IIE (94 ± 28 vs. 84 ± 28 ml/m2, p:0.01), LVESV-IIE (64 ± 27 vs. 55 ±23 ml/m2, p:0.003), LVEDV-CMR (141 ± 43 vs. 128 ± 41 ml/m2, p:0.01), LVESV-CMR (105 ± 42 vs. 90 ± 39 ml/m2, p:0.003), Iower LVEF-CMR (29 ± 10 vs. 32 ± 9 %, p:0.0027) and a higher LGE prevalence (67 vs. 44 %, p:0.0009) as compared to patients without MACE. At multivariate anaIysis, LVEF-CMR [HR:2.3 (1.6-3.01)] and presence of LGE [HR: 4.08 (2.15-6.02)] were independently associated with MACE (p < 0.001). In the subset of patients with LVEF-TTE > 35 %, the addition of LVEF-CMR and LGE provides anetre classification improvement (NRI) of 42 % and 26%, respectiveIy, interms of outcomes.
Conclusions: LVEF and LG Eestimation by CMR might provide additional prognostic stratification as compared to The that could identify a subset of subjects in whom ICD implantation is still indicated despite LVEF-TTE > 35%.
1393
Validation of aortic in-vitro strain measurement by Magnetic Resonance Imaging with realistic abdominal aortic aneurism phantom
Abstract
Introduction: Preoperative diagnostic protocols of abdominal aortic aneurysm (AAA) are today mainly based on the measurement of the aortic maximum diameter. This measurement is insufficient because the diameter is not a discriminant variable for predicting the rupture of the aorta. Recent works show the importance of determining the wall stress both due to the aortic shape, the pressure and the blood flow. The problem is very complex and requires the implementation of sophisticated models taking into account the heterogeneity of tissues and the complexity of flow. Then it is essential to validate the capacity of existing medical imaging systems to provide reliable measurements that will be introduced in these models.
Method: The aim of this study is to verify the MRI's ability to provide reliable measurements, firstly for the deformation of the aortic wall, and secondly for the blood flow. Cine MRI acquisitions (SSFP sequence) enabled to accurately determine the geometry and deformation of the aneurysm and 4D flow MRI measurements (thanks to a 3D PC-MRI sequence) were used to quantify the velocities of the fluid. Measurements were carried out in vitro with an experimental device that simulated hemodynamic circulation on a realistic AAA phantoms in silicone that have properties closest as possible as the actual physiological conditions. The sequences were prospectively gated with the pressure signal given by a pressure sensor. The deformations of the phantom wall were then determined through solid modeling with Abaqus software in which the geometry and internal stresses determined by CFD modeling (pressure, Wall Shear Stress) were introduced (see Fig. 1). The deformations were compared with measurement made by stereovision. Moreover, blood flow estimation with MRI was compared with the fluid modelling performed with ANSYS software.
Results: The obtained deformations with MRI are close to those obtained by stereovision with relative deviations less than 15% (see Fig.2). The comparison pixel to pixel between 4D Flow IRM and CFD is difficult because the spatial discretizations are different for the both technics. However the results of the mean instantaneous velocities are very similar between the two approaches throughout the cycle. Conclusion Our approach makes it possible to validate the ability of MRI to perform deformation measurements on phantom carried out with homogeneous materials and having complex geometries in which complex flow circulate. The first results showed the performance of the MRI to provide reliable raw data that can be included in the theoretical models.
Figure 1. The different steps of the strain calculation
Figure 2. Maximal main strain comparison between stereovision and modelling obtained from IRM measurement
1474
A novel method of Segment Length Tracking providing regional strain measures from standard CMR cine images in CRT candidates
Abstract
Background: Myocardial deformation analysis, providing strain measures on a regional level throughout the left ventricle (LV), is increasingly recognized for its additional value in the selection of heart failure (HF) patients for Cardiac Resynchronization Therapy (CRT). Although Cardiovascular Magnetic Resonance (CMR) myocardial tagging (MT) is considered to be the gold standard in quantifying LV wall deformation, this specialized technique is predominantly used for scientific purposes only. CMR cine imaging, on the other hand, has become routine clinical practice in most CRT candidates. Accordingly, the purpose of this study is to evaluate the accuracy and robustness of regional strain measurements using a novel method of segment length tracking (SLT) on standard cine images in CRT candidates compared to MT.
Methods: Twenty-seven HF patients (age 65 ± 10 y, 16 men) with left bundle branch block (LBBB) underwent CMR examination with both steady-state free-precession (SSFP) cine imaging and MT imaging to determine regional strains in the septum and the lateral wall. SLT was performed by measuring segment length between two endocardial anatomic landmarks (trabeculae) throughout all phases on the mid-LV short axis cines, relative to the end-diastolic segment length (figure 1A). This measure of frame-to-frame segment length change was compared to circumferential strain measurements in the tagged mid-layer of the same short axis slice position, analyzed using inTag software (figure 1B).
Results: Regional end-systolic strain values showed excellent agreement between both methods (figure 2). End-systolic strain measures showed large variation from -18.0% in the lateral wall up to +11.7% in the septum. Over this range of 29.7%, the mean difference between both methods was -2.0 ± 6.8% with the cine analysis structurally producing slightly lower strain values compared to the tagging analysis. Reproducibility of end-systolic strain measures by SLT were excellent for both the intra-observer analysis (ICC 0.96) and the inter-observer analysis (ICC 0.92).
Conclusions: Introducing the SLT post-processing technique on standard CMR cine images offers both accurate and robust circumferential strain measures compared to the gold standard MT technique. Future studies will have to focus on the prognostic value of SLT based strain parameters in CRT candidates.
Figure A illustrates the SLT method on cine images. Figure B illustrates the MT analysis on tagged images (same patient.) ED, end-diastolic; avo, aortic valve opening; ES, end-systolic; mvo, mitral valve opening.
1623
T1 mapping can quantify the area-at-risk and infarct size – no need for T2 mapping or conventional LGE imaging in acute STEMI at 1.5T
Abstract
Background: Native T1 mapping has been shown to accurately delineate the area-at-risk (AAR) at 1.5T in dogs and at 3T in STEMI patients when compared to T2 mapping. T1 mapping-based inversion recovery (IR) synthetic late gadolinium enhancement (LGE) could also accurately quantify chronic myocardial infarct (MI) size. We evaluated whether T1 mapping could accurately quantify the AAR from the native T1 maps and the acute MI size from the post-contrast T1 mapping-based IR synthetic LGE.
Methods: CMR was performed in 22 reperfused STEMI patients on a Siemens 1.5T scanner at 2 ± 1 days. Left ventricular (LV) short axis T2 maps, native T1 (MOLLI) maps, motion-corrected, free-breathing single shot SSFP averaged PSIR LGE (FB MOCO PSIR 10 mins post Dotarem) and post contrast T1 maps (15 mins post Dotarem) were acquired. The MOLLI prototype generated 2 sets of IR images (magnitude-reconstructed IR: MagIR and phase sensitive IR: PSIR) inline at different inversion times (TI). Using CVI42 software, the images from the MagIR and PSIR sets with the optimal TI were chosen to obtain 2 sets of synthetic LGE images. Figure 1 shows an example of a patient with an anterior STEMI. MI size and AAR were quantified using 5SD and 2SD thresholds from the remote myocardium respectively and expressed as percentage of the LV.
Results: The AAR by T2 was 42.7 ± 13.0% and by T1 was 42.6 ± 13.2%, P = 0.86 with an excellent correlation (R = 0.99) and agreement (bias = 0 ± 4.6%). The MI size by FB MOCO PSIR LGE was similar to MagPSIR synthetic LGE (24.8 ± 14.7% versus 24.6 ± 14.3%, P = 0.51) but marginally higher than PSIR synthetic LGE (24.8 ± 14.7% versus 24.1 ± 14.1%, P = 0.01). There was an excellent correlation and agreement between both the FB MOCO PSIR LGE and MagPSIR synthetic LGE (R= 0.99, bias = 0 ± 2.5%) and PSIR synthetic LGE (R = 0.99, bias = 0 ± 2.4%).
Conclusions: T1 mapping allows the quantification of the AAR from the native T1 maps and MI size from the synthetic LGE images derived from the post-contrast T1 maps. Scanning time could be shortened by not doing T2 mapping or LGE.
Figure 1.
1373
Reliability and reproducibility of trans-valvular flow measurement by 4D flow magnetic resonance imaging in acute myocardial infarct patients: two centre study
Abstract
Background: 4-dimensional (4D) flow MRI has advantage over 2-dimensional PC-MRI as it allows retrospective valve tracking, multi-planar dynamic phase contrast reconstruction to estimate stroke volumes (SV) reliably with good reproducibility in healthy subjects. However, this has not been demonstrated in patients with regional wall motion abnormality. We hypothesise that the SV assessed by 4D flow MRI through the mitral valve (MV) and the aortic valve (AV) will be consistent. We also investigated mitral valve (MV) and aortic valve (AV) SV and also the inter-observer reliability for intracardiac flow in acute myocardial infarction (AMI) patients.
Methods: Fifteen patients underwent CMR at 1.5T (Ingenia CV, Philips Healthcare, Best, The Netherlands). CMR Protocol included: 2-chamber, 3-chamber, 4-chamber cines and 4D flow MRI with isotropic voxel size (3 × 3 × 3mm3), parallel imaging (SENSE 2), velocity sensitivity Venc 150cm/s in all three directions and using echo-planar imaging (EPI) to factor of 5 for read-out acceleration. Free breathing was allowed and no respiratory motion correction was used. Retrospective gating was used and 30 cardiac phases were reconstructed. Images were analysed by two assessors (MH and PG) from two sites blinded to each other. Retrospective valve tracking with measurement planes positioned perpendicular to the inflow direction on 2-, 3- and 4-chamber cines was used to calculate SV. Background correction was used from velocity sampled in the myocardium.
Results: Mean MV SV was 68 ± 15 ml (PG) and 67 ± 17 ml (MH) (p = 0.90). Mean AV SV was 70 ± 20 ml (PG) and 73 ± 14 ml (MH) (p = 0.96). Eight (53%) patients had mitral regurgitation, four (27%) had aortic regurgitation and four (27%) had tricuspid regurgitation. All regurgitation was graded as trivial/mild (MR fraction - 6.7 ± 2 %; AR fraction - 1.9 ± 4 %). SV through MV and AV did not differ (PG: intraclass correlation coefficient (ICC) was 0.94, CI 0.81–0.97 p = 0.65; MH: ICC = 0.86, CI 0.46–0.96 p = 0.35). Coefficient of variation (CV) of inter-observer variability for MV SV was 6.3% and the concordance correlation coefficient was 0.93 CI (0.81-0.97) with accuracy of 0.99. CV of inter-observer variability for AV SV was 6.4% and the concordance correlation coefficient was 0.90 CI (0.72-0.96) with accuracy of 0.98.
Conclusion: 4D flow MRI with retrospective valve tracking provides reliable assessment of transvalvular flow in AMI patients with high inter-observer agreement.
Figure 1. Scatter diagram for MV and AV stroke volume demonstrating inter-observer reproducibility.
1588
Insights into hypertensive heart disease phenotypes: spectrum of myocyte, interstitial and vascular changes by cardiovascular MRI
Abstract
Introduction: The European Society of Cardiology classified hypertensive heart disease (HHD) into 4 left ventricular (LV) phenotypes by indexed LV mass (LVM), mass : volume ratio (M:V) and indexed end diastolic volume (EDV) (Table 1). All forms of HHD carry varying degrees of adverse cardiovascular prognosis, but their underlying mechanisms are poorly understood. We sought to investigate the presence and extend of ultra–structural myocardial changes and variation in central aortic function using CMR relaxometry and voxel–tracking myocardial strain techniques.
Methods: 88 hypertensive patients (49 ± 14 years, 57% male, SBP: 167 ± 30mmHg, DBP: 96 ± 14 mmHg) underwent CMR (1.5T) and were compared with 29 age–and sex–matched normotensive controls (47 ± 14years, 59% male, SBP: 128 ± 12mmHg, DBP: 79 ± 10mmHg). Native and post–contrast T1–mapping was performed. Circumferential myocardial strain was calculated with voxel–tracking software. Aortic compliance and distensibility were also estimated.
Results: Please see Figure 1. At a structural level, increased LV mass in eccentric LVH and concentric LVH resulted from: i) significantly increased myocardial cell volume (eccentric LVH: 78 ± 19 vs concentric LVH: 73 ± 15 vs remodeling: 55 ± 9 respectively, P < 0.05) and ii) significantly increased interstitial volume (eccentric LVH: 33 ± 10 vs concentric LVH: 30 ± 10 vs remodeling: 19 ± 2 respectively, P < 0.05). Functionally, eccentric LVH and concentric LVH were associated with significantly impaired peak circumferential strain (eccentric LVH: –12.8 ± 4.6 vs concentric LVH: – 15.5 ± 3.1 vs remodeling: –17.1 ± 3.2 vs controls: –17.4 ± 2.6% respectively, P < 0.05), with evidence of systolic and diastolic strain rate impairment. Despite similar BP severity as LVH phenotypes, LV remodeling was not associated with significant intracellular/interstitial expansion (native T1: 1029 ± 45 vs 1024 ± 41ms, P = 0.67) or myocardial dysfunction compared to normotensive controls but was associated with reduced aortic compliance and distensibility.
Conclusions: The extent of interstitial fibrosis is different across hypertensive heart disease phenotypes. LVH, in particular eccentric LVH, is associated with significant myocardial interstitial fibrosis and systolic/diastolic strain impairment. LV remodeling is associated with normal myocardial, but abnormal aortic, function. Our results may explain the poor cardiovascular prognosis with hypertensive LVH.
Figure 1.
1412
Myocardial partition coefficient of gadolinium: A comparison between patients with acute myocarditis, chronic infarction and healthy volunteers
Abstract
Introduction: Cardiovascular magnetic resonance (CMR) imaging allows distinction of myocarditis (MC) from myocardial infarction (MI). The tissue-blood partition coefficient (PC) of gadolinium, derived from T1 measurements, may reflect tissue injury, increasing in proportion to the expansion of the extracellular volume by edema or fibrosis. In this study, we compared the PC of gadolinium in the overall myocardium, in scarred areas and in unscarred (remote) areas in healthy volunteers, in patients with acute MC and in patients with chronic MI. We also correlated the PC with the myocardial scar burden.
Methods: Healthy volunteers (10, males (53.5 %), 49.2 ± 12.3 years), chronic MI patients (19, males (94.7 %), 62.7 ± 8.0 years) and acute MC patients (15, males (80%), 46.9 ± 15.6 years) underwent late gadolinium enhancement (LGE) CMR imaging and MOLLI T1 measurements with administration of gadolinium (Gd-BOPTA) at 3T (Magnetom Skyra, Siemens Healthcare, Erlangen). T1 measurements were performed before, 5 and 15 minutes after Gd-BOPTA injection, in 3 short-axis slices, analyzing both the entire myocardium and the remote myocardium after exclusion of the scar, as defined by LGE. The PC for Gd-BOPTA was computed. The percent of LGE volume (signal intensity > 5 SD above the remote myocardium signal intensity) was used to measure the scar burden.
Results: The PC was higher in the overall left ventricular myocardium of MC patients compared to chronic MI and volunteers (0.43 ± 0.07 vs. 0.40 ± 0.05, p = 0.08 vs. 0.40 ± 0.06, p = 0.09, respectively). PC in MC scars was significantly higher than in the remote myocardium (0.60 ± 0.12 vs. 0.40 ± 0.08, p = 0.002), and significantly lower than in chronic MI scars (0.60 ± 0.12 vs. 0.85 ± 0.21, p = 0.002). There was no difference between PC in the remote myocardium of MC and chronic MI patients compared to healthy volunteers. The correlation between % volume of LGE and PC in LV was strong and significant in MC and chronic MI patients with R = 0.89, p = 0.04 and R = 0.74, p = 0.03 for MC and chronic MI patients respectively (figure 1).
Conclusion: The PC of Gd-BOPTA is significantly elevated in areas of acute and chronic myocardial injury. Based on PC results, MC may show diffuse myocardial injury while chronic MI scars correspond to focal areas with a higher degree of gadolinium uptake, suggesting denser replacement fibrosis. The longitudinal change in PC and its relation to the evolution of MC may be studied to use it as an index of disease activity
Figure 1. Correlation between partition coefficient and LGE volume in percent of LV volume
1386
A comparison of circumferential strain results from multiple software packages in healthy subjects
Abstract
Background: Cardiovascular Magnetic Resonance (CMR) is an important modality to assess cardiac function. CMR is now the reference standard modality to assess myocardial global parameters such as cardiac mass, volume and ejection fraction but it can also assess myocardial regional deformation parameters such as strain and strain rate throughout the cardiac cycle. An accurate assessment of regional parameters could allow earlier detection of cardiac disease onset thus playing a critical role in patient management and improving quality of life. An increasing number of studies make use of feature tracking techniques as a quantitative post processing method. They derive motion deformation parameters from standard steady-state free precession sequence, which is part of routine clinical imaging protocols. The main objective of this study is to compare circumferential strain measurements obtained from different feature tracking and tagging software packages in healthy subjects.
Methods: 41 healthy subjects were prospectively enrolled to undergo CMR; one study was excluded for insufficient image quality. All images were obtained using a 1.5T Achieva Philips scanner (Best, the Netherlands) and a dedicated 32-channel cardiac coil. Steady state free precession breath hold cine images were obtained in the short axis plane (basal, mid and apical levels). Matching short axis tagged images were obtained using complementary spatial modulation of magnetization (CSPAMM), with a tag separation of 7.5 mm and a tag grid angle of 90o. Tagging is considered as the reference standard method to measure strain from CMR images.
Endocardial and epicardial boundaries of the left ventricle were drawn manually at end diastole and end systole. Strain and strain rate values were calculated semi-automatically using the following software packages: 2D Cardiac Performance Analysis, MR (TomTec Imaging Systems, Munich, Germany), CVI42 (Circle Cardiovascular Imaging Inc. Calgary, Canada) and CIM-FT (Auckland MRI Research Group, New Zealand). Tagged images were analyzed using CIMTag2D software (CIMTag2D v.8.1.2, Auckland MRI Research Group, New Zealand). All statistical analysis was done using SPSS (IBM Corporation, Armonk, New York, USA).
Results: Results of global circumferential strain and strain rate means are given in Table 1 and Table 2.
There were significant differences (P < 0.05) in most circumferential (basal, mid, apical) strains measured by Tomtec, CVI42 and CIM-FT compared to CIMTag measurements. There were also statistically significant differences for most circumferential (basal, mid, apical) strain rates measured by FT-software packages compared to CIMTag results. Strain rate results measured by Tomtec showed significant differences for 3 parameters, whilst 3 parameters showed no significant difference compared to CIMTag results. Mean strain results measured by CIMTag showed higher values in most parameters than FT-software packages results.
Figure 1 shows a box plot of the peak absolute values of the global circumferential strain for the mid short axis slice. The peak strain measured by CVI42 is more spread out compared to other software packages, which indicates more variability. Apical and basal slices showed similar trends. Moreover, apical and mid circumferential strain results measured by CIMTag were higher than those measured by FT-software packages.
Conclusions: FT- software packages showed significant differences in circumferential strain measurements when compared to CIMTag, with only a few parameters in agreement. Higher variability was observed for CVI42 compared to other software packages. There is a clear need for a reference standard method of validation, ideally based on a numerical phantom to assess the accuracy of these software packages in order to facilitate their use in a clinical setting. Currently outputs from different FT software cannot be directly compared and a need for standardization exists.
Funding: This project is funded by a Saudi scholarship to the first author.
Parameters . | CIMTag . | Feature Tracking . | ||
---|---|---|---|---|
CVI42 . | Tomtec . | CIM-FT . | ||
SAX-basal | 17.78 ± 2.12 | 20.27 ± 5.89 | 15.62 ± 2.98 | 16.92 ± 2.50 |
SAX-mid | 19.49 ± 3.98 | 16.68 ± 4.06 | 14.22 ± 2.84 | 15.25 ± 2.37 |
SAX-apical | 19.59 ± 5.43 | 17.33 ± 4.15 | 17.48 ± 4.02 | 18.48 ± 4.48 |
Parameters . | CIMTag . | Feature Tracking . | ||
---|---|---|---|---|
CVI42 . | Tomtec . | CIM-FT . | ||
SAX-basal | 17.78 ± 2.12 | 20.27 ± 5.89 | 15.62 ± 2.98 | 16.92 ± 2.50 |
SAX-mid | 19.49 ± 3.98 | 16.68 ± 4.06 | 14.22 ± 2.84 | 15.25 ± 2.37 |
SAX-apical | 19.59 ± 5.43 | 17.33 ± 4.15 | 17.48 ± 4.02 | 18.48 ± 4.48 |
Parameters . | CIMTag . | Feature Tracking . | |||
---|---|---|---|---|---|
. | CVI42 . | Tomtec . | CIM-FT . | ||
SAX-basal | RP | −0.76 ± 0.12 | −1.21 ± 0.78 | −0.75 ± 0.18 | −0.64 ± 0.12 |
P | 0.71 ± 0.15 | 1.33 ± 1.04 | 0.86 ± 0.21 | 0.55 ± 0.17 | |
SAX-mid | RP | −0.88 ± 0.12 | −0.85 ± 0.27 | −0.76 ± 0.18 | −0.65 ± 0.13 |
P | 0.76 ± 0.17 | 0.99 ± 0.34 | 0.69 ± 0.19 | 0.46 ± 0.12 | |
SAX-apical | RP | −0.93 ± 0.11 | −0.79 ± 0.21 | −1.07 ± 0.63 | −0.87 ± 0.32 |
P | 0.86 ± 0.21 | 0.98 ± 0.25 | 0.99 ± 0.66 | 0.61 ± 0.37 |
Parameters . | CIMTag . | Feature Tracking . | |||
---|---|---|---|---|---|
. | CVI42 . | Tomtec . | CIM-FT . | ||
SAX-basal | RP | −0.76 ± 0.12 | −1.21 ± 0.78 | −0.75 ± 0.18 | −0.64 ± 0.12 |
P | 0.71 ± 0.15 | 1.33 ± 1.04 | 0.86 ± 0.21 | 0.55 ± 0.17 | |
SAX-mid | RP | −0.88 ± 0.12 | −0.85 ± 0.27 | −0.76 ± 0.18 | −0.65 ± 0.13 |
P | 0.76 ± 0.17 | 0.99 ± 0.34 | 0.69 ± 0.19 | 0.46 ± 0.12 | |
SAX-apical | RP | −0.93 ± 0.11 | −0.79 ± 0.21 | −1.07 ± 0.63 | −0.87 ± 0.32 |
P | 0.86 ± 0.21 | 0.98 ± 0.25 | 0.99 ± 0.66 | 0.61 ± 0.37 |
Figure 1. Absolute peak global circumferential strain for mid-short axis slice measured by different software packages.
1362
Cardiac Involvement in Patients With Different Rheumatic Disorders
Abstract
Purpose: There is much controversy about the prevalence of cardiac involvement in patients with rheumatic disorders. This is of importance, since consequences of cardiac involvement range from different treatment regimes to adverse outcomes. Consequently, our primary aim was to evaluate the prevalence of cardiac involvement in different forms of rheumatic disorders by CMR.
Methods: 297 patients underwent CMR. Cardiac involvement was defined by presence of late gadolinium enhancement (LGE). Patients with a history of myocardial infarction and/or prior revascularization procedure were excluded. Patients were further divided into 5 subgroups: (1) ANCA associated vasculitis, (2) non-ANCA associated vasculitis, (3) autoimmune connective tissue disorders, (4) arthritis, and (5) sarcoidosis.
Results: We enrolled n = 63 patients in the ANCA associated vasculitis group, 34 of them (54%) were LGE positive. LGE showed a non-ischemic pattern in 33 patients (97%), isolated ischemic LGE pattern was present in only 1 patient. The Non-ANCA associated vasculitis group consisted of 32 patients, in 7 patients (22%) LGE could be detected. LGE was non-ischemic in 6 patients (86%), isolated ischemic LGE was present in only 1 patient. In the group with connective tissue disorders 15 out of 110 patients (14%) showed LGE: Of these, 13 patients (87%) showed non-ischemic LGE, 2 patients isolated ischemic LGE, 2 patients had both non-ischemic and ischemic pattern. The Arthritis group consisted of 47 patients, in 10 of them (21%) LGE was detected. Non-ischemic LGE pattern was found in 8 patients (80%), ischemic LGE was present in 3 patients (30%), in one patient LGE-CMR revealed both LGE patterns. 45 patients were in the sarcoid group, 11 of them (24%) showed LGE: 8 patients had non-ischemic LGE (73%), 3 patients ischemic LGE (27%). Comparing prevalence of cardiac involvement in the five different rheumatic subgroups, the highest prevalence of cardiac involvement (54%) was detected in patients with ANCA-associated vasculitis. Conversely, in our cohort of autoimmune connective tissue disorders, LGE CMR revealed the lowest prevalence of cardiac involvement.
Conclusion: There is a wide variation in the prevalence of cardiac involvement in different groups with rheumatic disorders. In patients with ANCA associated vasculitis cardiac involvement seems to be common (54%), whereas patients with autoimmune connective tissue disorders seem to have a low prevalence of cardiac involvement (14%).
1366
Gender differences in the development of cardiac complications: a multicentric prospective study in a large cohort of thalassemia major patients
Abstract
Introduction: We aimed to prospectively assess if the male gender was associated with an higher risk of progressive cardiac iron accumulation, development of biventricular dysfunction and myocardial fibrosis assessed by CMR, and development of cardiac complications including heart failure (HF), arrhythmias and pulmonary hypertension (PH).
Methods: We considered 1711 TM patients (899 females, 31.09 ± 9.08 years), consecutively enrolled in the Myocardial Iron Overload in Thalassemia (MIOT) Network. Myocardial iron overload was assessed by the multislice multiecho T2* technique. Biventricular function was quantified by cine images. Late gadolinium enhancement (LGE) images were acquired to detect myocardial fibrosis.
Results: Although having a similar risk of accumulating iron, males showed a significant higher risk of developing cardiac dysfunction, heart failure, arrhythmias and cardiac complications globally considered (Table 1). Figure 1 shows the Kaplan-Meier curves for the outcomes for which the male sex was a significant prognosticator. Until 20-30 years of follow-up (FU) the two lines (male and female sex) were almost overlapping while after they clearly diverged.
Patients were divided in two groups based on the FU duration. A significant gender-specific difference in the frequency of ventricular dysfunction and cardiac complications appeared for patients followed for at least 20 years. So, two subgroups of patients were identified: patients followed for less than 20 years and patients followed for more than 20 years. In the first subgroup males and females had a comparable risk of developing cardiac iron overload, ventricular dysfunction and cardiac complications. Conversely, if a FU longer than 20 years was considered, males exhibited a significant higher risk of having ventricular dysfunction, heart failure, arrhythmias, and cardiac complications.
Conclusion: Females seem to tolerate iron toxicity better, possibly as an effect of reduced sensitivity to chronic oxidative stress. According to the International Guidelines, TM patients should perform a complete cardiac evaluation every year. Our study suggested that in females older than 20 years the FU may be performed every 24 months, thus reducing health care costs.
1646
Comparison of T1-mapping, T2-weighted and contrast-enhanced cine imaging at 3.0T CMR for diagnostic oedema assessment in ST-segment elevation myocardial infarction
Abstract
Background: Myocardial salvage index (MSI) derived using cardiac magnetic resonance (CMR) can be used to gauge success of reperfusion strategies; it remains a strong predictor of adverse remodeling and prognosis post STEMI. However, various limitations of sequences used to typically image myocardial oedema have made it challenging to achieve reliable AAR detection in the majority in clinical trials. Optimal detection of reversible myocardial injury could improve the reliability of MSI determination and its prognostic relevance.
Objective: To compare the robustness of myocardial oedema detection using T2-weighted short-tau inversion recovery (T2w-STIR), free-breathing motion-corrected (moco) T1-mapping and contrast-enhanced steady state free precession (CE-SSFP) sequences on 3.0T CMR in patients presenting with STEMI.
Methods: Forty-five patients underwent CMR 1-5 days following presentation with STEMI. AAR was quantified using semi-automatic thresholding on T2w-STIR and short-axis CE-SSFP cine images and resulting parametric colour maps from moco-T1 Modified Look Locker Inversion Recovery (MOLLI) sequences and compared using analysis of variance (ANOVA). Pearson's correlation coefficient was used to assess correlation. Inter-sequence agreement was assessed using the Bland-Altman method, coefficient of variation (CoV) and two-way mixed-effect intra-class correlation coefficient (ICC) for absolute agreement.
Results: Patient characteristics are presented in Table 1. AAR assessed using the three methods was not significantly different (p = 0.988) with excellent correlation and good agreement, although with wide limits of agreement (see Table 2). Correlation and agreement between AAR derived using moco-MOLLI and CE-SSFP were excellent. However, the diagnostic imaging rate obtained with T2w-STIR (76%) was poor compared with CE-SSFP (87%) and highest with moco-T1-mapping (98%), p = 0.008.
Conclusions: moco-T1-mapping using MOLLI may be the more robust than T2w-STIR and CE-SSFP for detecting reversible myocardial injury and determination of MSI for the prediction of functional recovery following STEMI. The poor diagnostic imaging rate of T2w-STIR may limit its usefulness in large clinical studies. CE-SSFP maybe a useful alternative to T2w-STIR for AAR detection where T1-mapping is not available. Our findings require validation in larger multi-centre studies.
1375
Evaluation of Tissue Changes in Remote Noninfarcted Myocardium after Acute Myocardial Infarction using T1-mapping
Abstract
Background: Following acute myocardial infarction (MI) there is an extensive inflammatory reaction in infarcted myocardium. However, less is known about the inflammatory response in noninfarcted remote regions, and its role in cardiac remodelling. The present study assesses the temporal changes in T1-relaxation parameters (as markers of edema and cellularity) of the remote myocardium after acute MI, and determines its relationship with cardiac remodelling.
Methods: In this prospective study, 42 patients with acute MI treated with primary PCI were included and underwent CMR after 4-6 days and 3 months. Cine imaging, late gadolinium enhancement, and T1-mapping (MOLLI) was performed at 1.5T. T1-maps were acquired at the level of the infarcted area, before and 8 and 25-minutes after bolus (0.2 mmol/kg) injection of a gadolinium (Gd)-based contrast. T1 relaxation times were measured in the myocardial tissue opposite of the infarcted area. Gd-clearance rate was calculated by dividing the change in T1 between the 2 post-contrast time points by the time interval. Native T1 relaxation times were corrected for heart rate variability during data acquisition and were normalized to blood T1.
Results: Native T1 values of remote myocardium had significantly decreased from baseline to follow-up (997 ± 35 msec to 986 ± 27 msec, p < 0.01)(Figure 1). There were, however, no changes from baseline to follow-up in 8-min post-contrast T1 (p = 0.48), 25-min post-contrast T1 (p = 0.16), and Gd-clearance rate (p = 0.95). At baseline, remote native T1 values significantly differed between subgroups of patients, and were higher in patients with microvascular obstruction (MVO) (p = 0.02), an anterior MI (p = 0.04), absence of ST segment resolution (STR) (0.04), and large MI size (p = 0.01)(Figure 2). In a multivariable linear regression model, remote zone native T1 values at baseline were inversely associated with left ventricular ejection fraction (LVEF) at baseline (B: -0.08, p < 0.01), but were not associated with changes in LVEF or LV volumes.
Conclusion: In remote noninfarcted myocardium, a small change in native T1, but not in post-contrast T1 or Gd-clearance rate, was observed between baseline and 3 months after acute MI. Remote native T1 values at baseline were associated with markers of reperfusion injury and were independently associated with worse LVEF post-MI.
Figure 1. Native normalized T1 relaxation times in remote myocardium 4–6 days (base) and 3 months (FU) after MI.
Figure 2. Differences in remote native T1 values between patients with (open squares) and without (black squares) MVO, anterior MI, no STR, and large MI.
1377
Right ventricular long axis strain – The prognostic value of a novel parameter in non-ischemic dilated cardiomyopathy using standard cardiac magnetic resonance imaging
Abstract
Background: Long axis strain (LAS) has been shown to be a fast assessable parameter for the analysis of left ventricular longitudinal function in cardiac magnetic resonance (CMR). However, the prognostic value of right ventricular long axis strain (RV-LAS) using CMR is unknown. This study investigates the association of RV-LAS with outcome in patients with non-ischemic dilated cardiomyopathy (DCM).
Methods: 350 consecutive patients with DCM were prospectively included and analyzed retrospectively. RV-LAS was assessed by measuring the length between the epicardial border of the left ventricular apex and the middle of a line connecting the origins of the tricuspidal valve leaflets in enddiastole and endsystole (figure 1). Values for RV-LAS were calculated according to the strain formula.
Results: The endpoint, a combination of cardiac death, heart transplantation, cardiac decompensation or sustained ventricular arrhythmias, occurred in 85 patients during a mean follow-up period of 3.8 ± 0.1 years. The mean values of RV-LAS were significant reduced in patients with endpoint (-10.6 ± 4.4 vs. -8.3 ± 3.9, p = <0.0001). On multivariate analysis, RV-LAS remained an independent predictor of the endpoint (HR 1.1, p = 0.004). ROC analysis yielded a RV-LAS criterion of -8.9% for the endpoint. Kaplan-Meier survival curves showed a reduced outcome in patients with RV-LAS ≤-8.9% (p < 0.0001).
Conclusion: Assessment of RV-LAS in CMR is an independent predictor of outcome in patients with DCM and offers incremental information beyond standard CMR parameters including LGE.
Figure 1. RV-LAS: The length between the epicardial border of the LV apex and the middle of a line connecting the origins of the tricuspidal valve leaflets was measured in both enddiastole and endsystole.
1389
The role of the right ventricular insertion point in heart failure patients with preserved ejection fraction: Insights from a cardiovascular magnetic resonance study
Abstract
Background: In pulmonary hypertension (PH), increased afterload for the right ventricle (RV) is reported to induce fibrosis at the RV insertion point (RVIP), detectable by cardiac magnetic resonance (CMR) using late gadolinium enhancement (LGE). In contrast to LGE imaging, T1-mapping, a new CMR technique, allows quantitative assessment of myocardial native T1 times and extracellular volume (ECV). However, the role of T1-mapping of the RVIP in humans is unknown and the prognostic value has never been investigated.
Methods: We investigated 116 patients with heart failure and preserved ejection fraction (HFpEF), a patient population frequently suffering from PH, who underwent CMR including T1-mapping. Of these, 100 (86%) underwent right heart catheterization (RHC) for hemodynamic assessment. Native T1-times were measured at the anterior and inferior RVIP and ECV was calculated.
Patients were followed for 21 ± 18 months and the prognostic value of T1-mapping of the RVIPs was investigated by Cox-regression analysis.
Results: A total of 82% suffered from PH (mean pulmonary artery pressure, mPAP, ≥25mmHg). In 30%, LGE was detectable at the anterior RVIP, which was associated with higher mPAP as compared to patients without LGE (39 ± 13mmHg vs. 32 ± 9mmHg, p = 0.045), however, this was not seen for the inferior RVIP (34 ± 11mmHg vs. 31 ± 10mmHg, p = 0.118), where in 77% of patients LGE was present.
Native T1-times were 987 ± 78ms at the anterior and 1017 ± 89ms at the inferior RVIP and ECV was 32 ± 6% and 35 ± 7%, respectively. There was a significant correlation between mPAP and native T1 times as well as ECV of the anterior RVIP (r = 0.354, p < 0.001 and r = 0.248, p = 0.038) and the inferior RVIP (r = 0.248, p = 0.026 and r = 0.289, p = 0.015).
In total, 23% experienced a cardiovascular event. By Kaplan-Meier analysis, LGE was significantly associated with reduced survival when present at the anterior RVIP (log-rank, p = 0.011) but not at the inferior RVIP (log-rank, p = 0.207).
By multivariable Cox-regression, native T1-times at the anterior, but not the inferior, RVIP above the median were significantly associated with outcome (p = 0.037), even after adjusting for age, atrial fibrillation, diabetes, NT-proBNP levels, RV size and function, and presence of LGE at the anterior RVIP.
Conclusion: T1-mapping at the anterior RVIP provides new insights in the pathological processes in HFpEF patients and might detect alterations at early stages in these patients. Additionally, it is a predictor of outcome.
1398
Myocardial fibrosis associates with B-type natriuretic peptide levels and outcomes more than wall stress
Abstract
Background: B-type natriuretic peptide (BNP) is a nonspecific cardiac risk marker attributed to hemodynamic wall stress that strongly associates with adverse outcomes, but whether myocardial fibrosis (MF) is more strongly associated with both BNP and risk than wall stress is uncertain. Such data might demonstrate MF as a novel etiology of BNP elevation and further promote the emerging paradigm of MF as a therapeutic target in cardiomyopathy and heart failure.
Methods: We measured MF with histologically validated extracellular volume fraction (ECV) measures in 1176 consecutive adult individuals referred for clinical cardiovascular magnetic resonance (CMR). Linear regression quantified associations with log transformed BNP. Among those at risk with elevated BNP (≥100 pg/mL), Cox regression models quantified associations with the combined endpoint of all-cause mortality or hospitalization for heart failure (HHF) using a standardized definition, blinded to ECV and wall stress data. All multivariable models (linear or Cox) adjusted for demographics, EF and volumes, body mass, renal function, atrial fibrillation, mitral regurgitation, ischemic cardiomyopathy, and myocardial infarction size.
Results: Median age was 56 years (IQR 44-66). Median ECV was 27.6 (IQR 25.5-30.7%). Both ECV and wall stress associated with log BNP (p < 0.001) in univariable linear regression models, but ECV was more strongly associated (t value 17.2 vs. 15.5). In multivariable linear regression models, ECV remained strongly associated with log BNP (t value 10.6, p < 0.001), whereas wall stress was no longer associated with log BNP (t value 0.7, p = 0.51). In 497 patients with BNP ≥ 100 pg/mL, 103 events occurred (43 HHF and 78 deaths including 16 deaths after HHF) over a median of 2.3 years. Wall stress did not associate with outcomes. In contrast, ECV was the variable most strongly associated with adverse events (HR 1.36 95%CI 1.15-1.61) for every 5% increase in multivariable Cox models.
Conclusion: Myocardial fibrosis (MF) is more strongly associated with BNP than wall stress. In those with elevated BNP, ECV was robustly associated with outcomes but wall stress was not. These data suggest that MF is a key mediator of cardiomyopathy and vulnerability. Since MF is known to be reversible, MF may be a promising therapeutic target in heart failure and cardiomyopathy.
1478
Prognostic Value of Pulmonary Blood Volume by Contrast-Enhanced Magnetic Resonance Imaging in Heart Failure Outpatients – The PROVE-HF Study
Abstract
Background: In patients with heart failure (HF) early diagnosis of subclinical congestion is key to ensure prompt and effective treatment, and to prevent recurrent hospitalizations. Pulmonary blood volume (PBV) is a novel magnetic resonance (MRI) tool for the quantitative evaluation of pulmonary congestion.
Purpose: To prospectively assess the prognostic value of PBV in a cohort of HF outpatients.
Methods: Forty-four consecutive patients (34 men, 60 ± 12 years) and 31 age- and sex-matched healthy controls underwent contrast-enhanced cardiac MR. PBV was calculated as the product of stroke volume and number of cardiac cycles for an intravenous bolus of gadolinium contrast to pass through the pulmonary circulation, as determined by first-pass perfusion imaging.
Results: As compared to healthy controls, chronic HF outpatients showed significantly higher PBV index (PBVI, 317 ± 112 vs. 379 ± 146, ml/m2, p = 0.03) and pulmonary transit time (PTT, 6.6 ± 1.8 vs. 8.4 ± 2.9 sec, p = 0.004). PBVI was significantly associated with echocardiographic indices of diastolic dysfunctioN. Namely, PBVI showed a moderate positive correlation with tissue-Doppler E/E' ratio (R2 = 0.391, P < 0.001) and systolic pulmonary artery pressure, as assessed by CW-Doppler (R2 = 0.255, P < 0.001).During a median follow-up period of 39 ± 20 months, 13 patients (29%) reached the composite end-point of cardiovascular death, HF hospitalization or sustained ventricular arrhythmias/appropriate ICD intervention. Using a cut-off point of PBVI >541 ml/m2, corresponding to 2SD above the mean of healthy controls, Kaplan-Meier event-free survival rates were significantly higher in patients below (81%) as compared with patients above (14%) this cut-off (P < 0,001). On multivariate analysis PBVI was the only independent predictor of the composite end-point (HR 8.3- 95% CI: 2.01-35.66; p = 0.004).
Conclusions: PBVI is a novel approach to quantitatively determine pulmonary intravascular blood pool and may be used to evaluate the severity of pulmonary congestion in HF patients undergoing cardiac MRI examination.
1370
Magnetic Resonance Adenosine Perfusion Imaging as Gatekeeper of Invasive Coronary
Abstract
Background: Current guidelines for the diagnosis and management of patients with stable coronary artery disease (CAD) strongly support the performance of non-invasive imaging techniques for the detection of myocardial ischemia prior to revascularization procedures. This recommendation originates from the strong evidence base showing the lack of prognostic benefit from percutaneous coronary interventions (PCI) over optimal medical therapy in patients without verification of myocardial ischemia. On the other hand, it could be demonstrated that patients with functionally significant coronary artery stenoses do benefit from revascularization. Cardiac magnetic resonance imaging (CMR) has emerged to be a diagnostic modality of choice for the detection of myocardial ischemia with high sensitivity and specificity. We therefore designed this prospective and randomized trial to compare a CMR-driven vs. angiography-driven management of patients with stable CAD and pathologic stress ECG concerning major cardiac endpoints, futile angiographies and quality of life.
Methods: Consecutive patients suspicious of initial manifestation or progress of CAD were prospectively enrolled in this study. After obtainment of written consent, subjects were randomized 1:1 into two groups. Group I directly underwent coronary angiography. PCI was performed in case of ≥70% stenosis in a coronary vessel with ≥2 mm diameter. Group II underwent contrast-enhanced (Dotarem, Guerbet, France) adenosine-perfusion CMR at 3 Tesla. Patients that exhibited reversible ischemia were then sent to PCI. All patients received optimal medical therapy.
After sample size estimation and power analysis, follow-up was planned to be 3 years for every patient.
Primary endpoint was defined as cardiac death and non-fatal myocardial infarction. Secondary endpoints were number of coronary angiographies with and without PCI and change in symptoms/quality of life. Symptoms were objectified by the Seattle Angina Questionnaire (SAQ) at baseline and after each year of follow-up.
Results: After randomization, group I consisted of 101 patients, 93 of these patients were treated according to protocol. In group II, 95 out of 99 patients received the initially planned treatment. In group I, there 50 (53.8%) coronary angiographies revealed no obstructive CAD. The other 43 (46.2%) subjects were treated with PCI. In the CMR group, 68 (71.6%) patients did not exhibit reversible ischemia and were thus treated conservatively. The other 27 (28.4%) patients with reversible ischemia were sent to PCI. SAQs showed no significant baseline difference between both groups. One year results are expected in January 2016.
Conclusion: This is a randomized and prospective clinical trial with three-years follow-up period, which will prove whether a CMR-driven management of patients with stable CAD is non-inferior compared to primary coronary angiography regarding hard endpoint as myocardial death or non-fatal myocardial infarction and quality of life assessed by the SAQ. One year follow-up data will be available for presentation during the EuroCMR meeting 2016.
1509
Influence of non-invasive hemodynamic CMR parameters on maximal exercise capacity in surgically untreated patients with Ebstein's anomaly
Abstract
Introduction: Ebstein's anomaly is often associated with RV dysfunction. Data on RV function in surgically untreated patients are however rare. Since a good quality of life correlates with a good exercise capacity in daily life, we investigated non-invasive quantitative data derived from cardiovascular magnetic resonance (CMR) and its impact on maximal exercise capacity in patients with surgically untreated Ebstein's anomaly.
Methods: We investigated 54 unoperated patients with Ebstein's anomaly, age 5 to 69 years (median 30 years) and examined these patients with CMR and cardiopulmonary exercise testing (CPET). We compared seventeen CMR parameters with CPET parameters. We performed univariate and multivariate analysis with the focus on the maximal exercise capacity in these patients. For the maximal exercise capacity peak oxygen uptake as the percentage of normal (peakVO2%) was selected. The following CMR volume and flow parameters were correlated to peakVO2%. Both right and left ventricular ejection fraction (RVEF and LVEF), the indexed enddiastolic and endsystolic volumes (RVEDVi, RVESVi, LVEDVi and LVESVi) as well as the indexed stroke volumes (RVSVi and LVSVi), the total normalized right and left heart volumes (volume of the atrium and the ventricle together) as well as the total right to left heart volume ratio (R/L-ratio). Indexed flow data as the antegrade (PA ante, Aorta ante) and the indexed net flow (PA netto, Aorta netto) in the pulmonary artery and the aorta as well as its normalized values on heart rate (CI-PA, CI-Aorta) were used.
Results: RVEF (r2 0.2788), PA netto (r2 0.2330), and PA ante (r2 0.1912) showed the best correlation with peakVO2% (all p < 0.001). Further significant linear correlation could also be demonstrated with CI-PA, LVEF, LVSVi, Aorta netto, RVESVi and Aorta ante. All other parameters did not show a significant correlation with peakVO2%. Multivariate analysis for RVEF and PA netto revealed a r2 of 0.4350.
Conclusions: CMR parameters reflecting cardiac function as RVEF and LVEF and flow data of cardiac forward flow best correlate to peakVO2%. The evaluation of the indexed net flow in the pulmonary artery and the overall function of the right ventricle (RVEF) best predicts the maximal exercise capacity in patients with Ebstein's anomaly.
1356
Proximal aortic stiffening in Turner patients is more pronounced in the presence of a bicuspid valve. A segmental functional MRI study
Abstract
Objective: To study structural and functional segmental aortic properties in Turner syndrome (TS) patients.
Methods: Aortic abnormalities contribute to increased morbidity and mortality of women with Turner syndrome. MRI allows segmental study of aortic elastic properties.
Using MRI, we performed pulse wave velocity (PWV) and distensibility measurements of the thoracic and abdominal aorta in 55 TS-patients, aged 13 to 59 years, and in a control population (n = 31; aged 21 to 58 years). We investigated the contribution of TS on aortic stiffness in our entire cohort, in aortic valve-morphology subgroups, and in the younger and older subgroups.
Results: Compared to controls, TS patients had higher PWV of the entire aorta (F = 7.3; p = 0.009) and thoracic aorta (TA-PWV; F = 10.5; p = 0.002), but not of the abdominal aorta (F = 0.3; p = 0.619). Ascending aortic (AA) diameters (but not more distal aortic diameters) were also significantly larger (F = 4.2; p = 0.043). No differences in distensibility were found. The increase in TA-PWV and AA diameters was more pronounced in bicuspid compared to tricuspid TS patients. Similarly, bicuspid TS patients also showed a decreased proximal aortic distensibility. In exploratory analyses, the observed proximal aortic stiffening and dilatation were largely similar when comparing younger and older TS patients.
Conclusion: Turner patients exhibit a predominant stiffening and dilatation of the proximal aorta, more pronounced when their aortic valve is bicuspid. These abnormalities are present at an early age, suggesting an aortic wall disease inherent to the TS, and only limited effect of accelerated ageing in the young adult Turner patient.
1503
Flow pattern and vascular distensibility of the pulmonary arteries in patients after repair of tetralogy of Fallot. Insights from 4D flow CMR
Abstract
Objectives: Pulmonary regurgitation is a frequent sequela after repair of tetralogy of Fallot (TOF). The regurgitant flow may lead to changes in the flow profile and in size and distensibility of the pulmonary arteries (Pas).
We sought to assess Pa flow and distensibility in TOF patients (pts) by cardiac magnetic resonance (CMR) and to correlate them with the flow patterns provided by 4D flow CMR.
Methods: 18 TOF pts (mean age 28 ± 11yrs, weight 63 ± 12 kg) and 9 control subjects (age 17 ± 7yrs, weight 63 ± 24kg) underwent CMR. 2D Phase-contrast (PC) images were acquired through-plane in the main (MPA), right (RPA) and left pulmonary artery (LPA). A 4D PC dataset was acquired covering all the great arteries. Vessel areas and quantitative flow were measured on the 2D PC images. The flow patterns in MPA, RPA and LPA were qualitatively assessed for presence of helix or vortex on the reconstructed 4D images. Flow parameters, size and distensibility of the Pas were compared between TOF pts and controls and in the TOF group between RPA and LPA with regard to helix/vortex.
Results: In TOF pts, MPA mean regurgitant fraction (RF) was 25 ± 17%. Compared to controls, both Pas were larger and distensibility was higher in LPA (p 0.048) but not in RPA. RF was greater in LPA than in RPA (29 ± 19% vs 16 ± 14%, p 0.001) and LPA area was larger than RPA area (316 ± 134 vs 257 ± 116 mm2/m2, p 0.0342). LPA net flow was lower than RPA net flow (p 0.0005). Distensibility was similar in LPA and RPA and significantly correlated with RF, regurgitant flow and minimum area in both Pas branches.
By 4D flow, vortex was observed in the LPA in 72% (13/18) of TOF pts, but not in normals. Helical flow was present in 44% (8/18) of pts and in 11% (1/9) of normals. Presence of vortex in the LPA was independent from any other parameter; LPA helix was more frequent in pts with higher distensibility (p 0.04).
RPA presented helical flow in 77% (14/18) of TOF pts and in 55% (5/9) of controls. Vortex was only detected in 11% (2/18) of TOF pts. RPA helical flow was not correlated to other parameters.
Conclusion: In pts after TOF repair, Pas size and distensibility are mainly determined by the amount of regurgitant flow and less by specific flow patterns, such as vortex or helix. Characteristic flow patterns are found in LPA and RPA, which seem to be more related to the geometry of the pulmonary bifurcation and its branches than to quantitative flow parameters.
1516
Myocardial deformation characteristics of the systemic right ventricle after atrial switch operation for transposition of the great arteries
Abstract
Introduction: The atrial switch operation (Senning) has been the main surgical repair technique for d-transposition of the great arteries for many years. The Senning procedure results in a subsystemic morphologic right ventricle (RV) and a subpulmonary morphologic left ventricle (LV). This can be regarded as a model for the effects of long-term pressure overload on the RV, and of ultimately decreased afterload on the LV. We sought to determine the impact of these chronically altered loading conditions on the myocardial deformation of the RV and LV.
Methods: 26 patients after Senning (age 28.4± 7.5y) and 18 normal controls (age 22.2± 11.4y; p = 0.034) underwent cardiac magnetic resonance (CMR) imaging. 2D SSFP cine images were acquired in an horizontal long axis and in a short axis covering both ventricles and post-processed with a feature tracking software (TomTec 2D CPA). Global circumferential strain was measured on a short axis mid-ventricular slice. Global longitudinal strain was measured in a long axis, separately for each ventricle.
Results: When comparing RV in either position, subsystemic circumferential strain was higher than subpulmonary circumferential strain (-16.1± 2.9% vs. -13.1± 4.3%; p < 0.01), and subsystemic longitudinal strain was lower than subpulmonary longitudinal strain (-12.8± 3.3% vs. -18.3± 3.8%; p < 0.001). In contrast, LV global strain in subsystemic vs. subpulmonary position was similar: LV circumferential strain (-23± 13.1% vs. -20.2± 3.9%; n.s.); LV longitudinal strain (-17.5± 4.6% vs. -16.1± 5.3%; n.s.).
The subsystemic RV showed lower circumferential (-16.1± 2.9% vs. -23± 13.1%; p < 0.05) and lower longitudinal strains (-12.8± 3.3% vs. -17.5± 4.6%; p < 0.001) than the subsystemic LV. The subpulmonary LV exerted greater circumferential strains (-20.2± 3.9% vs. -13.1± 4.3%; p < 0.001) but similar longitudinal strains compared to the subpulmonary RV (-16.1± 5.3% vs. -18.3± 3.8%; n.s.).
Conclusions: In discordant ventriculo-arterial connection, the subsystemic RV adapts to the increased afterload with an increase in circumferential strain and an impaired longitudinal deformation. This may represent the effect of a positive interventricular interaction due to the shared circumferential fibers, since the LV shows higher circumferential strain than the RV even in subpulmonary position.
1633
Three-dimensional vortex formation in patients with a Fontan circulation: evaluation with 4D flow CMR
Abstract
Introduction: In the normal heart, formation of ring-shaped vortex flow has been suggested to help efficient blood flow during early (E) and late (A) diastolic filling of the left ventricle (LV), while altered vortex formation can be associated with LV abnormalities and inefficient blood flow [1,2,3]. Patients with a Fontan circulation have various underlying cardiac anatomy and it is unknown how this relates to vortex formation during diastolic filling. The purpose of this study was to assess three-dimensional (3D) vortex formation in patients with a Fontan circulation with different underlying pathologies, using four-dimensional (4D) flow CMR.
Methods: 11 patients with a Fontan circulation (age 15 ± 5 years) were evaluated by 4D flow CMR. Whole-heart 4D Flow CMR scans were performed on a 3 Tesla MRI (Ingenia, Philips Healthcare) with three-directional velocity-encoding of 150 cm/s in all directions, acquired voxel size of 3 × 2.6 × 3 mm3 and 30 phases reconstructed over one cardiac cycle. 3D vortex formation was identified in the LV at E- and A-diastolic filling, using the Lambda2 (λ2)-method [2]. Vortex formation during E- and A-diastolic filling was visually assessed.
Results: Vortex flow was present in all 11 patients at peak E-filling and in 10 patients at peak A-filling. One patient (with no A-wave) had no vortex during A-filling. Table 1 shows the versatile vortex formation in these patients. Normal “ring-shaped”' vortices were present in four patients. A double vortex ring (“eight-shaped”) was present in four patients with two functional atrioventricular (AV)-valves (Figure 1a). One patient showed a large vortex ring along the ventricular septal defect (VSD) with protrusion in both ventricles (Figure 1b). Triangular vortices were present in two hypoplastic left heart syndrome (HLHS) patients (Figure 1c). Two patients showed a vortex ring with reversed septal-lateral orientation compared to other patients, which could be related to the asymmetric length of the AV-valve leaflets [3].
Conclusion: Altered 3D vortex formation is present in patients with a Fontan circulation during E- and A-filling and can be related to the different underlying pathologies. Future studies have to reveal the influence of this abnormal vortex formation on cardiac function.
References:
1. Kilner et al. Nature 2000
2. Elbaz et al. JCMR 2014
3. Calkoen et al. JTCVS 2015
1483
Mitral valve prolapse: arrhythmogenic substrates by cardiac magnetic imaging
Abstract
Introduction: Mitral valve prolapse (MVP) is a common valvular disorder with a benign prognosis in the vast majority of patients, but with an incidence of sudden cardiac death (SCD) twice than in the general population. Recent data support the theory that left ventricular (LV) fibrosis constitutes the substrate of SCD in MVP patients. Cardiac Magnetic Resonance (CMR) represents a noninvasive imaging modality that provides a comprehensive characterization of both the valve and the myocardium.
Aim: To assess the presence of myocardial substrates as scar in patients with MVP and arrhythmias.
Methods: We enrolled consecutive patients with MVP diagnosed by an 2D echocardiography. For each one we detect the presence of ventricular arrhythmias by a 24 hours 12-leads ECG continuous monitoring. All patients underwent complete CMR scan including post contrast sequences. Complex ventricular arrhythmias (LV-CVA) were defined as non sustained or sustained ventricular tachycardia exclusively with right bundle branch block (RBBB) morphology suggesting an origin from LV. Patients with moderate or severe mitral valve regurgitation or other cardiomyopathies were exclude.
Results: a total of 52 patients were enrolled (33 female, median age 44 years); a bileaflet MVP was found in 31 (60%). On the basis of presence of significant ventricular arrhythmias, a group of “arrhythmic MVP” (N:33) compared to “non-arrhythmic MVP” (N:19) patients were identified. No differences on ventricular volumes and function were found. On post-contrast sequences, LV LGE was identified in 36 (69%). The LGE location was at the level of papillary muscles in 28 patients (54%) and LV infero-basal segment, near posterior valve leaflet annulus, in 26 (50%). The presence of LV-LGE was associated with LV-CVA [32 (97%) vs 4 (21), p < 0.001].
Conclusions: arrhythmic MVP patients are characterized by a myocardial substrate of electrical instability detectable by CMR, such as LGE both of papillary muscle and infero-basal LV wall. This points may represent area stretched by abnormal valve excursion leading to a local fibrosis, as confirmed also by post mortem analysis. The management of MVP patients with LGE by CMR, whether drugs, catheter ablation or ICD, remains a clinical challenge requiring further studies.
1596
Increased local wall shear stress after coarctation repair is associated with descending aorta pulse wave velocity: evaluation with CMR and 4D flow
Abstract
Background: Coarctation of the aorta (CoA) is a local narrowing of the descending aorta (AO), typically distal from the left subclavian artery. Curative operational treatment is available and survival in childhood is good. However, in adulthood complications including hypertension, restenosis and aneurysm formation are not uncommon. In this explorative CMR study, we sought to investigate whether an association exists between AO wall compliance (expressed in pulse wave velocity (PWV) and distensibility) and 3D wall shear stress (WSS) in the AO in children after CoA repair.
Methods: 19 patients aged 12.8 ± 3 years underwent 4D flow CMR on 3T MRI (Ingenia, Philips Healthcare). 14 patients had a bicuspid aortic valve and none of the patients had clinical indication for re-intervention. 19 healthy volunteers aged 13.2 ± 3 years were included to assess normal values of PWV. PWV was determined from high-temporal 1-directional velocity encoding, for both proximal AO (ascending AO plus aortic arch) as well as the descending AO. Distensibility was measured at the ascending AO from the lumen area distension and the brachial pulse pressure. 3D WSS was determined from 4D flow CMR using CAAS MR 4Dflow v1.0 software (Pie Medical Imaging).
Results: There was no difference in age between volunteers and patients (p = 0.901). Proximal AO PWV (4.3 ± 0.9 m/s; in controls vs 5.0 ± 1.4 m/s in patients) and descending AO PWV (4.1 ± 0.9 m/s vs 3.9 ± 0.9 m/s) were not significantly different between both groups (p = 0.499 and p = 0.538 respectively). Ascending AO distensibility and proximal AO PWV were correlated (r = -.62; p = 0.008) in patients as expected, but not in volunteers due to the narrow range in both parameters. No relationship was found between overall WSS and PWV (r = -.053 p = 0.846) in entire AO but peak WSS in the aortic arch and proximal AO PWV were correlated (r = -.550 p = 0.027). The average ratio of peak WSS near the repaired CoA to average WSS in the entire AO was 2.6 ± 0.5 and this ratio was correlated to descending AO PWV (r = .518; p = 0.04).
Conclusion: Regional AO wall compliance measured by PWV was not significantly different between young controls and patients after CoA repair. However, at the repaired site, local WSS is more than doubled in patients, and this is associated with PWV distal to the CoA site. The AO seems to be locally affected in the distal rather than proximal AO. Altered wall compliance and WSS may in part explain the long term sequela in patients after CoA repair.
1636
Three-dimensional wall shear stress assessed by 4Dflow CMR in bicuspid aortic valve disease
Abstract
Background: Ascending aorta (AAo) aneurysms may be partially caused by altered flow patterns associated with bicuspid aortic valve (BAV). Wall shear stress (WSS) has been hypothesized as an important biomarker for aortic dilatation. Thus, specific WSS patterns could potentially explain the development of a certain phenotype in BAV patients. The aim of our study was to analyze differences in WSS patterns using 3D maps according to the BAV morphotype.
Methods: Eighty-six BAV patients with no severe valvular disease and aortic diameters under 50 mm were enrolled. Fusion phenotype was right-left (RL-BAV) in 65 patients, and right-non coronary (RN-BAV) in 21. All subjects underwent 4D Flow MRI with retrospective cardiac gating at 1.5 T. Eight double-oblique analysis planes were equally distributed in the AAo between the sinotubular junction and the origin of the brachiocephalic trunk. Peak-systolic WSS was calculated in each plane, as described by Stalder et al. 3D WSS maps were estimated by longitudinally interpolating these WSS values. Peak velocity, flow eccentricity and rotational flow were evaluated in three slices located at proximal (S2), mid (S4) and distal AAo (S8). Calculations were performed using custom Matlab software.
Results: Table 1 summarizes differences between RL-BAV and RN-BAV in peak velocity, eccentricity and rotational flow at different aortic levels. RN-BAV had higher peak velocities and rotational flow. Different BAV phenotypes presented different outflow jet direction that correlated with the distribution of WSS within the aortic wall. Thus, RN-BAV patients presented a maximum WSS from proximal left posterior-to-distal right aortic wall (figure 1b) while the maximum WSS was from left posterior-to-right anterior at proximal-medial levels in RL-BAV patients (figure 1a).
Conclusion: Maximum AAo-WSS region varies according to the BAV morphotype. 3D WSS maps can precisely illustrate these surface variations in WSS and can be a useful tool to better understand the pathophysiology of aortic dilatation in these patients.
1464
Cardiac Amyloidosis and Aortic Stenosis – The Convergence of Two Aging Processes
Abstract
Background: Senile cardiac amyloidosis (SCA), once thought to be a rare disease, has been commonly diagnosed with Cardiac MRI (CMR). Recent autopsy series in octogenarians with severe aortic stenosis who received transcatheter aortic valve replacement reported prevalence up to 30%.
Objectives: To evaluate the prevalence of SCA detected by CMR in patients with confirmed moderately-severe symptomatic aortic stenosis (AS) and its association with clinical outcomes.
Methods: 95 consecutive AS patients (85% with severe AS) who underwent CMR study (1.5T Siemens Magnetom Espree) and echocardiogram within 15 days (median 6 days, IQR 0-15) were included in the analysis. Society of Thoracic Surgery predicted risk of mortality (STS-PROM) was calculated for each patient using over 40 clinical parameters. SCA was identified when characteristic pattern involving either subendocardial or diffuse myocardial LGE was observed. Cox-proportional hazards model performed after adjusting for potential confounders to evaluate the independent prognostic role of AS + SCA.
Results: A total of 9 patients (8 males, 89%) had AS + SCA (10% prevalence). Clinical, imaging and outcomes are listed in Table 1. Patients with AS + SCA were older than those with only AS (87 ± 5 vs. 68 ± 5 years, p < 0.0001) and had higher STS-PROM (6.8 ± 4.2 vs. 3.3 ± 3.2 %, p = 0.003). Over median follow-up of 10 months (IQR: 3- 19 months), there were 28 deaths (30%) and 39 AVRs (31 surgical and 8 transcatheter). There was no difference in the percentage of AVR performed in AS + SCA vs. isolated AS (33% vs. 42%, p = 0.73). Despite adjustment for AVR (HR = 0.51, 95% CI 0.22-1.18, p = 0.11) and STS-PROM (HR = 1.16, 95% CI 1.07-1.25, p < 0.0001), presence of AS + SCA was an independent predictor of all-cause mortality (HR = 4.02, 95% CI = 1.3-12.7, p = 0.02).
Conclusion: SCA is not uncommon in octogenarians presenting with symptomatic severe AS. One year all-cause mortality in AS + SCA patients was 3 times higher than in patients with isolated severe AS (Figure 1). Whether this risk can be modulated by additional therapies, beyond AVR, requires future studies.
Figure 1. Kaplan-Meier Curves Comparing All-Cause Mortality in AS vs. AS + SCA Patients
1630
Blood T1 variability explained in healthy volunteers: an analysis on MOLLI, ShMOLLI and SASHA
Abstract
Background: Native myocardial T1 is known to be affected by variables such as age, gender, heart rate and partial voluming from blood pool. Blood T1 itself has a wider variability. We aimed to investigate causes of blood T1 variability.
Methods: 77 healthy volunteers with no known cardiovascular condition underwent CMR at 1.5T (Siemens, Avanto). Mid ventricular short axis native T1 maps by MOLLI (with T1* reconstruction in addition), ShMOLLI and SASHA were acquired. Hematocrit (Hct), iron profile and lipid profile were acquired immediately prior to the scan. CVI42 (Calgary, Canada) was used for analysis of the maps. A ROI was drawn in the blood pool on the MOLLI T1 map, avoiding papillary muscles, and was copied onto the MOLLI T1*, ShMOLLI and SASHA T1 maps.
Results: Complete datasets of blood and maps were available for all 77 volunteers (mean age 49 ± 14, range 20-76, 49% males). Mean ± SD of blood T1 by MOLLI T1 was 1638 ± 78ms, MOLLI T1* 1686 ± 111ms, ShMOLLI T1 1543 ± 77ms and SASHA 1584 ± 100ms. There was a negative correlation between blood T1 and Hct (R2 0.530, coeff. -0.728, p < 0.0001). Hct, iron, HDL-cholesterol, ferritin, triglicerides (TG), LDL-cholesterol and total iron binding capacity (TIBC) resulted to be significant at univariate analysis while this was not the case for albumin and total cholesterol. The multivariate analysis performed including only the significant variables showed that Hct, iron and HDL-cholesterol are significantly correlated with blood T1 by MOLLI T1 and T1*, ShMOLLI and SASHA (Table 1).
Conclusions: In health, Hct, iron and HDL-cholesterol explain almost all (90%) of blood T1 variability with anaemia and low iron increasing T1 but with HDL reducing it.
1408
Myocardial deformation on CMR predicts adverse outcomes in carcinoid heart disease - a new marker of risk
Abstract
Introduction: Carcinoid heart disease (CHD) is a frequent and adverse complication of carcinoid syndrome due to right ventricular (RV) failure. Medical therapy has a 2-year survival of 20% and while surgical valve replacement is effective in improving symptoms and may increase survival, peri-operative risk remains 15-20%. Echocardiography is considered gold standard for assessing CHD and data on the role of cardiacmagnetic resonance imaging (CMR) are limited despite recognised advantages in assessment of theright heart. This study aims to assess the role of CMR in the assessment of CHD.
Methods: 50 consecutive patients with proven neuroendocrine tumours were referred with elevated NT pro-BNP to the European Centre of Excellence for Neuroendocrine Tumours in Birmingham between 2005-2015. At referral, all subjects underwent comprehensive left ventricular (LV) and RV assessment with CMR (1.5T Siemens Avanto), including deformation (Tissue Tracking, Circle cvi42), and late gadolinium enhancement (LGE).
Results: 36 patients were diagnosed with CHD and 14 without (CHD-neg). RV valve disease was universal in CHD: severe tricupid regurgitation (97%), severe pulmonary regurgitation (86%). On CMR, RV end-diastolic volume (EDV) and end-systolic volume (ESV) were increased (120 ± 30 vs. 67 ± 14 ml/m2, p < 0.01; 49 ± 20 vs. 11 ± 3ml/m2, p < 0.01) but with no difference in RVEF (60 ± 14% vs. 60 ± 9% p = 0.92). There was early evidence of ventricular-ventricular interaction, with reduction in LVEDV (53 ± 16 vs. 72 ± 16 ml/m2, p < 0.01) and LVESV (19 ± 10 vs. 28 ± 16ml/m2, p < 0.05) in CHD but no difference in LVEF (67 ± 8% vs. 63 ± 14%, p = 0.3). There was no difference in LV global longitudinal strain (GLS) or circumferential strain (GCS) between groups. RV LGE indicative of endocardial plaques was present in 6/36 (17%) but not observed in CHD-neg. Diffuse LV LGE was present 5 CHD patients. Over follow up (median 1.3 years [0.6-3.1]), 20 patients with CHD died. These patients had a lower GCS (14.8 ± 4.6% vs. 18.2 ± 4.4%, p < 0.05) and lower GLS (14.7 ± 4.7% vs. 18.3 ± 4.4%, p < 0.05) on CMR but no difference in ventricular size, EF or NT-proBNP. In a logistic regression model, LV GLS remained an independent predictor of death.
Conclusion: Significant increase in RV size and the presence of RV plaques are measurable on CMR early following referral with CHD. This is sufficient to adversely affect LV filling and global deformation, which may contribute to effort intolerance and adverse outcomes.
1492
Myocardial Perfusion Reserve and Global Longitudinal Strain in Early Rheumatoid Arthritis
Abstract
Objectives: Rheumatoid arthritis (RA) is a systemic inflammatory condition associated with increased cardiovascular mortality compared with the general population. Proposed mechanisms for this increased mortality include coronary microvascular dysfunction due to immune dysregulation and systemic inflammation.
First pass myocardial perfusion CMR allows quantification of myocardial blood flow (MBF) from which myocardial perfusion reserve (MPR) can be derived. In the absence of epicardial coronary artery disease, reduced MPR represents coronary microvascular dysfunction. We therefore hypothesised MPR would be reduced in RA patients.
Additionally, we hypothesised that abnormalities in left ventricular (LV) deformation would be evident in RA patients, as LV mass has been reported to be reduced in established disease.
Methods: Twelve patients with newly diagnosed, treatment naïve RA and 12 healthy volunteers (HV) underwent CMR at 3.0T (Phillips Achieva TX). Both groups had no history of coronary artery disease or additional risk factors for this. Dual bolus resting and stress perfusion imaging was performed (0.1mmol/kg Gadolinium DTPA) and MBF values for the mid ventricular slice were estimated using Fermi constrained convolution (PMI v 0.4 [Sourbron, 2009]). MPR was calculated by dividing stress MBF by rest MBF. Left ventricular ejection fraction (LVEF) and global longitudinal strain (GLS) by feature tracking were calculated using bSSFP images (CVI 42, Circle Cardiovascular Imaging, Calgary, Canada).
Results: Mean age of RA patients was 48 ± 16 and mean age of HV 47 ± 14. Of the RA patients, were 4 were male and 8 were female. Of the HV, 2 were male and 10 were female. Mean values for MPR were 1.98 ± 0.79 and 1.99 ± 0.72 (P = 0.88) for RA and HV respectively. Mean values for LVEF were 63 ± 4 and 62 ± 4% (P = 0.48) respectively for RA patients and HV. Mean values for GLS were -20.2 and -21 ± 0.4 (P= 0.396) for RA and HV respectively.
Conclusion: These preliminary findings from the CADERA trial suggest that, newly diagnosed, treatment naïve RA patients have no detectable abnormalities on perfusion CMR compared with HV. Therefore, whilst present in established RA, coronary microvascular dysfunction may not yet have developed in early RA. Additionally, no abnormalities of LV systolic function were evident in newly diagnosed RA and may be a later manifestation of the disease.
1500
Exercise CMR to differentiate athlete's heart from patients with early dilated cardiomyopathy
Abstract
Introduction: It can be difficult to distinguish between early stage dilated cardiomyopathy (DCM) and left ventricular (LV) dilation with mildly reduced LV ejection fraction (EF) which is a relatively common observation amongst endurance athletes (EAs). This has important consequences for clearance for competitive sport participation since DCM may precipitate fatal ventricular arrhythmias.
Methods: We prospectively included 9 EAs and 9 asymptomatic patients with mild DCM (7 familial DCM and 2 partially recovered severe DCM), all LVEF 40-55% at inclusion. In all subjects beta-blockers were withheld 24 hours before evaluation. First, cardiopulmonary exercise testing was performed to determine maximal power (Pmax). Then, cardiac magnetic resonance (CMR) imaging was performed at rest and during bicycle exercise at 25%, 50% and 66% of Pmax to determine left and right ventricular (RV) end-diastolic and end-systolic volumes, from which ejection fraction (EF) was calculated.
Results: At rest, RVEF was lower in EAs than in DCM patients, whereas LVEF was similar (Figure 1A). At peak exercise, RVEF was similar in both groups. However, LVEF augmentation was observed in EAs but was attenuated in DCM patients. Receiver-operator characteristic curves demonstrated that a cut-off value of 11.2% for the increase in LVEF from rest to peak exercise (ΔLVEF) had a sensitivity of 89% and specificity of 89% to differentiate EAs from DCM patients [AUC = 0.89(0.72-1.06); P = 0.005], whereas resting LVEF (cut-off: 50.8%) was not predictive (Figure 1B).
Conclusion: Evaluation during exercise facilitates the differentiation between athlete's heart and DCM. An athlete's heart demonstrates good myocardial contractile reserve as compared to subjects with early DCM in whom LV functional reserve is reduced.
Figure 1.
1559
Real-Time, x-mri guidance to optimise left ventricular lead placement for delivery of cardiac resynchronisation therapy
Abstract
Suboptimal LV lead placement contributes to cardiac resynchronization therapy (CRT) non-response. Cardiac MR (CMR) identifying myocardial scar & dyssynchrony allows targeted LV lead placement. Currently CMR is performed prior & separate to implantation. We describe a novel guidance platform, integrated within an XMR facility enabling CMR data acquisition, processing & analysis with immediate overlay onto live fluoroscopy.
Methods and Results: Short & long axis cines & late gadolinium sequences are segmented and projected onto a 16-segment AHA model demonstrating location, transmurality & burden of myocardial fibrosis. Regional motion analysis demonstrates the latest mechanical activating segments. This process is performed seamlessly while the patient is within the XMR suite allowing segmented LV geometry/scar to be immediately co-registered with fluoroscopy. CS venography is overlaid onto the 3D-shell to identify the preferential targets for LV stimulation. This platform has been successfully used in 6 patients with no significant increase in procedure time or radiation dose compared with our historical controls (89 ± 18 vs.113 ± 28 minutes & 826 ± 191 vs. 950 ± 450 cGycm2, p = ns). Late, target segments avoiding scar were paced in all cases. Visualization of the target informed positioning of the LV lead and enabled identification of the optimal poles to use for biventricular stimulation (Figure).
Conclusion: Performing and processing CMR data in real time is feasible and may aid optimal LV lead placement for CRT delivery. This can be accomplished in a single sitting with no delay between CMR and CRT implant.
1560
The role of Cardiac magnetic resonance imaging in patients undergoing ablation for ventricular tachycardia- Defining the substrate and visualizing the outcome
Abstract
Background: Pre- interventional Cardiac Magnetic Resonance (CMR) imaging may help to identify the underlying arrhythmogenic substrate in patients undergoing radiofrequency (RF) ablation for ventricular tachycardia (VT). However, the extent and characteristics of structural damage induced by the ablation itself and its association with acute procedural outcome remain poorly defined so far.
Methods: 17 patients (82% male, median age 57.5± 11 years, 6 with known coronary artery disease) who underwent VT ablation received CMR scans pre- and post- ablation. CMR was performed at 1.5 Tesla including cine, T2-weighted, 3D- early and late- gadolinium- enhancement (LGE)- imaging in standard cardiac geometries. LGE and microvascular obstruction (MO) were expressed as percentage of total left ventricular mass. The electrophysiological (EP) procedures were performed according to local protocols using the Carto©- Mapping- system in all cases. Non- inducibility of VTs at the end of the procedure was defined as the primary endpoint. RF ablation was performed using a 3.5 mm Thermocool catheter, flow rate 30 ml/min.
Results: In 13 patients, substrate defined as positive LGE, could be described on pre-interventional CMR imaging (11 endocardial, 1 endo- and epicardial, 1 strictly epicardial; mean LGE 22.6% ± 16). Mean left ventricular ejection fraction (LV-EF) before ablation was 45% ±11. Successful ablation was performed in all cases using a mean maximum energy of 55 ±11 Watt. Epicardial access was used in 3 patients.
On postinterventional scans, myocardial edema was visible in 9 patients. MO was present in all patients (mean MO 6 ± 2%, 7 cases with a MO >5%) Mean postinterventional LV-EF showed to significant difference to baseline (42%15, p = 0.4) and decreased in only 1 patient (MO> 5%). The extent of MO correlated significantly with the applied maximum energy (R = 0.625, p = 0.007) and the mean ablation time (R = 0.512, p = 0.036). A 100% transmural ablation lesion was observable in 4 patients.
Conclusion: Apart from reliably defining the arrhythmogenic substrate preinterventionally in patients with VTs undergoing ablation, postprocedural CMR imaging reveals acute myocardial injury in the form of edema and MO induced by the ablation itself. The extent of the lesions directly correlates with the time and the maximum power of the applied RF energy. Whether the lesion characteristics on are associated with short- and long- term outcome needs to be further assessed.
1590
Impact of cardiovascular magnetic resonance on clinical management and decision-making of out of hospital cardiac arrest survivors with inconclusive coronary angiogram
Abstract
Background: Non-traumatic out of hospital cardiac arrest (OHCA) is the leading cause of death worldwide. Urgent coronary angiography is a class IB recommendation, as 2/3 of cases are secondary to acute coronary syndrome. However diagnosis and management of patients with unobstructed coronaries or unidentified culprit lesion on angiogram is challenging. We sought to assess the additive role of cardiac magnetic resonance (CMR) in patients with an inconclusive coronary angiogram and to determine the best CMR parameter in predicting clinical impact.
Methods: We analysed the CMR registry data of consecutive patients surviving non-traumatic OHCA, undergoing urgent angiogram and CMR from two tertiary cardiology centres. The study focused on the analysis of patients with an inconclusive angiogram, defined as unobstructed coronaries or coronary artery disease (CAD) without a clear culprit lesion. Clinical impact of CMR was defined as a change in diagnosis, as compared to a multi- parametric pre-CMR diagnosis, or a change in management, which could be a change in medication or the performance/avoidance of invasive procedures.
Results: Out of 174 OHCA survivors referred for CMR following urgent angiogram, we identified 110 patients (63%,84 male,mean age55 ± 17) with inconclusive angiogram: 73 patients(66%) had unobstructed coronaries and 37(34%) had CAD with no clear culprit. Diagnosis based on CMR findings was ischemic heart disease in 43 patients (39%), non-ischemic heart disease in 33 (30%), a structurally normal heart was found in 24 (22%) and non-specific findings in 10 (9%). Overall, CMR had a clinical impact in 72% of patients, determining a change in diagnosis in 25% of patients, a change in management in 30%, and a change in both in 17%. CMR led to revascularization in 20% of patients and to ICD implantation in 10%; an invasive procedure was avoided in 12% of patients. In a multivariate model including clinical and imaging parameters, the strongest predictors of the clinical impact of CMR were regional wall motion abnormality and the analysis of LGE sequences (p 0.042,95% CI 1.05-9.99; p < 0.0001,95% CI 5.46-73.7,respectively).
Conclusions: CMR had a clinical impact in more then two thirds of OHCA survivors with an inconclusive coronary angiogram. The analysis of LGE sequences was the strongest independent predictor of clinical impact following CMR. Given its additional role CMR should be incorporated in the clinical-diagnostic work-up of this group of patients.
1561
Detection of coronary stenosis at rest using Oxygenation-Sensitive Magnetic Resonance Imaging
Abstract
Objective: To determine whether Oxygenation-Sensitive Magnetic Resonance Imaging (MRI) can identify functionally significant coronary artery disease (CAD) without the need for stress.
Background: In the setting of CAD, microvascular dilation has been shown to occur in post-stenotic myocardium at rest. Microvascular volume is a determinant of signal intensity with Oxygenation-Sensitive (O-S) MRI. Therefore, we postulated that O-S imaging at rest could detect microvascular heterogeneity, and thereby identify coronary stenosis without recourse to physiological or pharmacological stress.
Methods: Rest O-S images were acquired at 3 Tesla using a T2-prepared ECG-gated steady-state free precession sequence and assessed quantitatively using a resting O-S index (interquartile range of segmental O-S signal intensities/median segmental signal intensity). First-pass perfusion images were also acquired at rest and stress (intravenous adenosine, 140 µg/kg/min) and assessed quantitatively using model-independent deconvolution. A threshold resting O-S index to identify myocardial ischaemia at the patient level (as defined quantitatively as ≥1 myocardial segment with hyperaemic myocardial blood flow ≤1.6ml/min/g) was first determined in a derivation arm comprising 25 patients with known CAD and 20 healthy volunteers. To determine diagnostic performance, this threshold was then applied in a separate validation group comprising 57 patients with suspected CAD referred for diagnostic angiography.
Results: Receiver-operating characteristic curve analysis defined a threshold resting O-S index of 16.1% to identify patients with myocardial ischaemia (72% sensitivity and 78% specificity; area under the curve 0.72 ± 0.08, p = 0.0078). Application of this threshold in the validation arm of the study yielded sensitivity 88%, specificity 58% and diagnostic accuracy 75% for the identification of subjects with ischaemia, and 82%, 61%, and 75%, respectively, for identifying those with significant coronary stenosis. At the patient level, there was reasonable agreement between resting O-S imaging and both first-pass perfusion (kappa = 0.51) and quantitative coronary angiography (kappa = 0.48).
Conclusions: O-S imaging indicates the presence of microvascular heterogeneity at rest in CAD. Assessment of microvascular heterogeneity offers a potential means for identifying the anatomical and functional significance of CAD without the need for stress.
1385
Longitudinal strain assessment in dilated cardiomyopathy patients using a novel accelerated DENSE sequence
Abstract
Objectives: Strain assessments are prognostically valuable in patients with dilated cardiomyopathy (DCM), particularly longitudinal strain which can be abnormal prior to the development of marked systolic functional impairment. Displacement encoding with stimulated echoes (DENSE) is an accurate and reproducible technique to measure myocardial strain, but its clinical utility is limited by long breath hold times or navigator gating which can be impractical and time consuming. We developed an accelerated DENSE sequence and sought to evaluate its utility in longitudinal strain assessment in DCM.
Methods: We modified a cine spiral DENSE sequence to selectively excite a reduced field of view thereby reducing breath hold time by 30%. 12 DCM patients (8 males, mean age 55 ±15 years) underwent a 2D cine DENSE imaging using this novel sequence in a horizontal long-axis slice and vertical long-axis slice (Siemens Skyra 3T). DCM was diagnosed based on CMR criteria of LV dilation and impaired ejection fraction according to age and gender adjusted nomograms.
Images were acquired at 3.5x3.5x8.0mm3 spatial resolution, 30ms temporal resolution and 2 direction encoding at 0.06cycles/mm. The FOV was 224mm2 with a breath hold duration of 14 RR intervals. Global longitudinal strain was extracted from the DENSE data using contour strains in a Matlab post-processing software from the University of Virginia.
Results: All patients were able to sustain the breath hold duration for the accelerated DENSE acquisition and had complete CMR assessment of myocardial strain. Mean left ventricular ejection fraction (LVEF) was 47 ± 7.8%.
Table 1 shows global peak strain and time to peak strain from horizontal and vertical long axis images for each subject. Peak longitudinal strain measurements show strong negative correlations with LVEF for both vertical (r = -0.84, p = <0.001) and horizontal (r = -0.73, p = 0.007) long axis measurements.
Figure 1 shows an example of longitudinal strain curves for a DCM patient (subject 10) with severely impaired systolic function (LVEF 36%) compared to a healthy volunteer (LVEF 68%).
Conclusion: We demonstrate the feasibility of use of a novel accelerated DENSE sequence to obtain longitudinal strain measurements in patients with DCM in a clinically acceptable breath hold duration.
Figure 1. Longitudinal strain curves from a healthy volunteer compared to a patient with DCM (HLA = horizontal long axis, VLA = vertical long axis).
1407
Simultaneous T1 and T2 cardiac quantification with CABIRIA: initial clinical experience
Abstract
Objectives: The purpose of this study was to evaluate the performance of the cardiac balanced-SSFP inversion recovery with interleaved sampling acquisition (CABIRIA) [1] sequence on a cohort of patients with different cardiac pathologies. This sequence provides simultaneous T1 and T2 quantification. The homogeneity of its quantitative results across the different regions of the heart was assessed and the global T1 and T2 values were compared across pathologies.
Methods: A cohort of 14 subjects was included in the study. Among these, 2 had no cardiac condition, 5 suffered from myocarditis, 4 from Hypertrophic Cardiomyopathy (HCM), and 3 from Pulmonary Hypertension (PAHT). A pre-contrast short-axis CABIRIA acquisition (Resolution: 1.7x1.7x7mm^3, Flip angle: 35°, 15 acquisitions per inversion pulse, repeated 2 times) was performed in the context of a routine cardiac MR examination on a 1.5 Avanto scanner (Siemens Healthcare, Erlangen, Germany). The T1 and T2 maps were analyzed offline by means of a custom segmentation interface that divided the view into 6 segments, following the American Heart Association guidelines (see fig. 1). Each segment of each map was treated as a data point and differences across the segments were analyzed as a measure of robustness of the CABIRIA method. Differences across pathologies were also quantitative analyzed. Analysis of Variance (ANOVA) testing was used in both cases.
Results: No significant differences were found for T1 maps across the different segments (p = 0.35) and moderately significant differences were found for T2 maps (p = 0.03). Significant differences were found across diagnoses for T1 values (p < 0.001), with values of 1225 ± 121, 1108 ± 91, 1207 ± 107, and 1220 ± 88ms for normal, HCM, Myocarditis and PAHT. Nonsignificant differences were found across diagnoses for T2 values (p = 0.06), with respective values of 93 ± 28, 75 ± 21, 78 ± 23, and 81 ± 30 ms (Fig. 2).
Conclusion: The method showed a higher robustness in T1 mapping with respect to T2 mapping, which can be hypothesized as a result of magnetic field inhomogeneities across the different segments. Our findings resulted in a higher T1 values for healthy subjects with respect to other pathologies. This is somewhat surprising, but this can be due to the small number of subjects included in this control group, which also lead to a higher spread in values as depicted in Fig. 2.
Reference:
1. Santini et al. Magn Reson Med, 74: 365–371, 2015
1423
Head-to-head comparison of acceleration algorithms in 4-dimensional flow CMR
Abstract
Background: Standardisation and validation of four-dimensional (4D) flow CMR acquisition methods is lacking. Our aim was to compare three acceleration methods in 4D flow CMR: 4D segmented fast-gradient-echo with segmentation factor 2 (4D-TFE), 4D non-segmented gradient-echo with echo-planar imaging (EPI, factor 5) (4D-FFEPI) and 4D kt-BLAST accelerated TFE (factor 5; 11 training lines).
Methods: CMR was performed in two institutions (Leiden and Leeds) on identical 1.5T systems (Ingenia CV, Philips Healthcare). Methods were compared in a phantom (straight tube with 7mm diameter with stationary flow ranging between 1.07-2.48 l/min) and 15 volunteers. In the phantom setup, the 3 4D flow sequences were compared to 2-dimensional phase contrast (2D PC). In volunteers, the CMR protocol included: cines, 2D PC at the aortic valve (AV) and mitral valve (MV) and whole-heart free-breathing (no respiratory motion correction) 4D flows. Field of view, slices, phases (30), voxel size and VENC were the same for each subject. 4D-FFEPI and 4D-TFE used retrospective ECG gating while 4D kt-BLAST used prospective ECG-triggering. In volunteers, stroke volumes (SV) at MV and AV were obtained by retrospective valve tracking. In volunteers, net acquisition time for each 4D flow sequence was recorded, as well as a visual grading of image quality on a four-point scale: 0, no artefacts to 3, non-evaluable.
Results: The phantom setup showed that all 4D flows underestimated flow volumes when compared to 2D PC (4D-FFEPI -3 ± 1%; 4D-TFE -9 ± 1%; 4D kt-BLAST -10 ± 3%). In vivo validation was performed in 15 volunteers (44 ± 11yrs old and 60% male). Acquisition times are presented in Figure 1. 4D-FFEPI and 4D kt-BLAST had minimal artifacts. For 4D-TFE however, 8 AV and 5 MV assessments were non-evaluable because of phase dispersion artefacts. All methods showed significant difference between AV SV and MV SV (Table 1). Non-significant underestimation in 4D flow peak velocity was found at static positions: 4D-FFEPI: 2 ± 26 cm/s with CoV 26%, 4D-TFE: 11 ± 41 cm/s with CoV 40%, 4D kt-BLAST: 7 ± 30 cm/s with CoV 31%.
Conclusion: 4D EPI accelerated 4D flow had the shortest acquisition time, best agreement with the 2D-PC and the lowest variability. Consistency between MV and AV net forward flow was marginally but not significantly better with 4D-FFEPI (10.5%) than with 4D-TFE (11%). 4D kt-BLAST demonstrated significant underestimation of NFF and CoV was poor. 4D-TFE was most susceptible to artefacts.
Figure 1. Time taken (in minutes) for acquisition using the three 4D Flow accelerated techniques.
1502
Left ventricular function and size evaluated by hybrid cardiac positron emission tomography-magnetic resonance: Intraindividual comparison of left ventricular ejection fraction and ventricular volumes derived by two modalities
Abstract
Objectives: Hybrid cardiac imaging using combined positron emission tomography (PET) and magnetic resonance (MR) systems is an upcoming imaging technology, currently under evaluation for clinical use. The aim of this work is to intraindividually compare results from left ventricular function and left ventricular volumes evaluated by an integrated PET-MR system using a dual tracer protocol.
Methods: Dual tracer [13N]ammonia ([13N]NH3) (750 ± 230MBq) and 2-[18F]fluoro-2-deoxy-D-glucose ([18F]FDG) (310 ± 92MBq) cardiac PET-MR was performed in 13 patients for evaluation of cardiac viability in ischemic heart disease. Left ventricular ejection fraction, endiastolic volumes (EDV), and endsystolic volumes (ESV) were derived from 3 Tesla cardiac MR as well as from the simultaneously acquired [13N]NH3 and [18F]FDG PET images. MR data were calculated from short axis cine images in the endiastolic and endsystolic phase during breath hold. PET data were derived from the myocardial phase during free breathing using 16 gates in a time frame of 15 minutes for [13N]NH3 and 10 or 20 minutes for [18F]FDG.
Results: Mean values for LVEF derived from MR, [13N]NH3 PET and [18F]FDG PET were 45 ± 16%, 48 ± 17% and 46 ± 17%. Mean values for EDV were 188 ± 66ml for MR, 171 ± 85ml for [13N]NH3 PET and 159 ± 86ml for [18F]FDG PET. For ESV MR derived values were 122 ± 68ml; [13N]NH3 PET values were 101 ± 76ml and [18F]FDG PET values were 96 ± 78ml respectively. LVEF showed an excellent correlation between MR and [13N]NH3 PET (R = 0.96; mean difference 4.6%; range 0-14%) as well as between MR and [18F]FDG PET (R = 0.89; mean difference 6%; range 1-17%). In addition, significant correlations have been observed between EDV obtained from MR and by [13N]NH3 PET (R = 0.86; mean difference 39ml; range 7-106ml), and between EDV measured by MR and [18F]FDG PET (R = 0.83; mean difference 43ml; range 0-109ml). Similarly, the comparison of ESV values acquired from MR with those acquired from [13N]NH3 PET and [18F]FDG PET revealed an excellent correlation (R = 0.94; mean difference 29ml; range 7-60ml) and (R = 0.88; mean difference 32ml; range 13-82ml) respectively; all P < 0.001.
Conclusion: Left ventricular volumes and left ventricular function simultaneous derived from an integrated PET-MR do show an excellent correlation between the data derived from the MR and the PET system with slightly better results for [13N]NH3 PET compared to [18F]FDG PET when considering MR as the gold standard.
1510
Left Atrium assessed by Cardiovascular Magnetic Resonance at 1.5 and 3 Tesla – age and gender effects
Abstract
Background: The impact of the left atrium (LA) for the prediction of therapeutic success as well as its impact on clinical outcome e.g. in cardiomyopathies is well known. Currently, the use of volumes and/or function is different in most trials but a standardization would be helpful to offer a uniform view on disease understanding.
Cardiovascular magnetic resonance (CMR) is a well established technique of cardiac morphology and accepted as the gold standard, but often disparaged as a very time‐consuming method.
We aim to provide LA parameters including LA ejection fraction (EF) in male and female healthy volunteers at 1.5 and 3 Tesla using a time effective approach.
Methods: We analyzed 203 healthy volunteers at 1.5 and 3 Tesla using cine‐SSFP‐imaging based on the biplanar approach. Simpsons rule was used to quantify the atria in the 2‐and 4‐chamber view excluding pulmonary veins and the left atrium appendage in LA Systole and LA Diastole (see also figure 1).
Using cvi42 we calculated mean values and standard deviation for LAEDV, LAESV, LASV, LAEF and all volumes were normalized to Body‐surface‐area (BSA) and Height (H) and subdivided by gender.
We analyzed the volunteers according to three age groups with group 1 from 20 to 39 years, group 2 from 40 to 59 years and group 3 60 years orolder to compare LA values at different ages.
Results: We included 105 male and 77 female healthy volunteers (age group 1 50 male and 26 female, age group 2 36 and 31 and group 3 19 and 20).
There was no significant difference in the absolute volumes inour volunteers between field strengths (see table 1).
There were no gender‐related differences in normalized parameter with males having a LAEDV/BSA of 33 ml/m2 and a LAESV/BSA of 12 ml/m2 and women of 34 ml/m2 and 11 ml/m2 (p > 0.6), respectively. But the absolute values were different (LAEDV and LAESV (p < 0.01) with men having larger volumes.
We found a significant decrease in LAEDV/BSA and LAEDV/H in older age with age group1 having a LAEDV/BSA of 36 ml/m2, decreasing to a LAEDV/BSA of 33 ml/m2 in age group 2 respectively of 30 ml/m2 in age group 3 (see also figure 1).
Summary: Our main findings are that LA–size decreases during the life–cycle in healthy volunteers and that normalization to body–size helps to overcome sex–related differences. There is no dependency on field strengths.
Figure 1. Decreasing LAEDV/BSA with older age.
1514
Comparison of Free Breathing Cardiac MRI Radial technique to the Standard Multi breath-hold cine SSFP CMR technique for the assessment of LV Volumes and Function
Abstract
Background: Cardiac MRI (CMRI) is generally accepted as the gold standard for left ventricular (LV) volumes and function assessment. Standard cine SSFP sequences require prolonged, repeated breath holds and therefore may be challenging, particularly for debilitated, pediatric and elderly patients. The compressed sensing, Radial cine (RC) k‐space acquisition technique can provide a good image quality without the need for breath holding. The purpose of this study was to compare volume and function results between free breathing RC technique and the standard multi–breath-hold cine SSFP technique.
Methods: 24 patients who underwent clinically indicated cardiac MRI (CMRI) were included in the study. In all patients both standard cine and radial cine sequences were performed. The images were analyzed using CMR42 semiautomatic tool of the CVI42 software, version 5.0.0 for both LV and RV end diastolic volume (EDV), end systolic volume (ESV) and ejection fraction (EF) and myocardial mass for LV. The RC and the standard breath-hold SSFP techniques were compared using Pearson correlation and Bland-Altman analyses. Values of p < 0.05 were considered statistically significant.
Results: The average age was 45.7 ± 18, there were 15 males and 9 females. Results of Pearson correlations and Bland-Altman analyses (P values for all correlations were <0.05) are summarized in the table below.
Conclusions: The results demonstrate the feasibility of applying the RC compressed sensing strategy to evaluate LV and RV volume and function with high accuracy in a variety of patients, without the need for breath holding, in considerably shorter scan times.
1536
Self-navigated free-breathing isotropic 3D whole heart phase sensitive inversion recovery magnetic resonance without navigator for detection of myocardial infarction
Abstract
Background: Cardiac magnetic resonance (CMR) is considered as gold standard for detection of scar after myocardial infarction. Navigator techniques allow acquisition of 3D scar information however have limited scan efficiency. This study evaluates the accuracy of a self-navigated isotropic free-breathing 3D whole-heart CMR acquisition technique with integrated, contrast independent iterative approach to 1D self navigation with phase sensitive inversion recovery (3DSN-PSIR) to detect myocardial scar tissue.
Methods: Patients presenting with chronic myocardial infarction detected by late gadolinium enhancement on standard 2D phase sensitive inversion recovery sequences (2D-PSIR) underwent a CMR exam with 3DSN-PSIR on a 1.5T clinical MRI scanner (MAGNETOM Aera, Siemens Healthcare, Germany). Data acquisition was performed 15 minutes after administration of 0.2mmol/kg of Gadobutrol in free-breathing during the most quiescent systolic phase with a prototype ECG-triggered and segmented respiratory self-navigated 3D radial trajectory implementing an inversion-recovery (IR) prepared acquisition, a reference scan, and a phase-sensitive reconstruction. Imaging parameters were: TR/TE 2.9/1.45ms, FOV 220mm3, resolution 1.4mm3, RF excitation angles 115°(IR) and 8°(reference), bandwidth 900 Hzx/px, fat saturation, inversion time (TI) 250-320ms. A total of about 14000 radial readouts were acquired for each 3DSN-PSIR with 100% respiratory efficiency. Short axis slices were reconstructed from 3DSN-PSIR datasets to yield the same slice thickness (10mm) as the 2D-PSIR images. Scar tissue was segmented both on 2D-PSIR and the reconstructed 3DSN-PSIR slices.
Results: Six patients were included (patient characteristics see table). 3DSN-PSIR successfully corrected for respiratory motion in all acquisitions. All scars visualized by 2D-PSIR could be identified by 3DSN-PSIR (example see figure). 3D and 2D scar volume did not differ: 3D: 59 ± 28ml vs. 2D: 47 ± 18ml, p = 0.07. Bland Altman-analyses and correlations showed a good agreement of quantification of scar volume obtained by 3DSN-PSIR compared to standard 2D-PSIR: 8.9 ± 6.6ml, r = 0.99, p < 0.001, intraobserver -3.2 ± 9.6 ml, r = 0.94, p= 0.005; interobserver -11.2 ± 14.7 ml, r = 0.88, p = 0.032.
Conclusions: These preliminary data show that detection of myocardial scar by 3DSN-PSIR is feasible and shows a good agreement with standard 2D-PSIR. This technique should be investigated in future larger studies.
1547
Assessment of Right Ventricular Strain Using Myocardial Deformation Recovery Semi Automated Technique: Initial Experience and Normal Values
Abstract
Objective: Assessment of Right Ventricluar (RV) strain by cardiac magnetic resonance (CMR) imaging may add prognostic information, beyond global RV function in patients with pulmonary hypertension, right sided valvular pulmonary tricuspid regurgitation, and complex congenital heart disease. Currently echocardiography is the main modality to assess RV strain and is limited to RV free wall.
Methods: Dedicated RV Myocardial Strain Software: Myocardial deformation recovery from the cine MRI first involves constructing a deformable model based on the endocardial and the epicardial tracing in the reference frame, assuming the myocardium is nearly incompressible. In each of the subsequent frames the displacements of the control points of the model are determined using feature tracking method with the incompressible model constraint. Using this method, not only the myocardial boundaries, but the entire myocardial tissue is tracked. CMR of 37 patients without known cardiac pathology and preserved biventricular function were analyzed retrospectively.
Strain analysis: Horizontal long axis was used for calculation of longitudinal strain (fig 1). Short axis at the base, mid and apex of the RV was used to calculate circumferential strain (fig 2). Intraventricular septum was not included in the strain calculation. Endocardial and epicardial contours were drawn in the phase with most distinct myocardium boundaries; RV trabeculations were carefully excluded. The software automatically propagated contours throughout all phases. Longitudinal (Ell) and circumferential strain (Ecc) was computed.
Results: The age range was 4-72 years with an average of 31 (± 20.69). Analysis time per patient was < 5 minutes. Average Ell was -21.92 (± 3.42). Average Ecc for basal, mid and apical segments were as follows: -10.60 (±3.42), -10.94 (± 3.68), -11.84 (±3.28). There was no significant difference between the basal, mid and apical levels. The intraclass correlation coefficient (ICC) was 0.785 for Ell, and 0.877, 0.943, and 0.902 respectively for the Ecc at the base, mid, and apical levels. There was no correlation between age and strain.
Conclusion: Rapid analysis of RV Ell and Ecc can be accomplished with excellent intraobserver variability. Further studies are needed to determine performance and utility of this dedicated RV Deformation Recovery Strain software in the clinical settings.
1586
Tissue tracking myocardial deformation analysis and prediction of left ventricular remodeling in acute myocardial infarction
Abstract
Background: The novel tissue tracking analysis software (TT) for cardiac magnetic resonance (CMR) allows the study of myocardial deformation parameters based on routinely acquired CMR sequences. It was our purpose to evaluate the feasibility of this technique in ST-elevation acute myocardial infarction (STEMI) patients and its ability to predict left ventricular remodeling.
Methods: A total of 100 STEMI patients were prospectively studied with a 1.5 T CMR scan within 3-5 days of successful primary percutaneous intervention. Routine SSFP sequences from 3 long-axis and a stack of short-axis slices were analyzed with TT (CVI42, Circle Cardiovascular Imaging Inc.) and peak systolic global longitudinal, circumferential and radial strain values were studied and correlated with basal ejection fraction and infarct size as determined with late gadolinium enhancement imaging. Follow-up CMR was performed at 6-months to assess left ventricular volumes. Left ventricular remodeling was defined as a 20% or greater increase in left ventricular end-diastolic volume (LVEDV).
Results: During follow-up, 27 patients presented with left ventricular remodeling (mean increase in LVEDV 36 ± 12%, minimum 22%, maximum 78%). Patients with left ventricular remodeling presented lower basal left ventricular ejection fraction and higher relative infarct size and microvascular obstruction mass, as well as worse global longitudinal, circumferential and radial strain (table 1). Global peak systolic circumferential strain was the best predictor of left ventricular remodeling with a sensitivity of 70% and a specificity of 74% for a cut-off value of -15.9% (AUC 0.7).
Conclusions: CMR myocardial deformation analysis with the TT software provides a new tool for evaluation of STEMI patients and prediction of left ventricular remodeling. Global peak systolic circumferential strain was the strongest predictor with a cut-off value of -15.9%.
1589
Investigating strategies for optimal 31P MRS clinical cardiac at 3T: Initial Results
Abstract
Background: The metabolic information obtained from 31P MR cardiac spectra has been demonstrated to be useful in the failing heart [1], in patients with Dilated Cardiomyopathy (DCM) [2]. However technological and physiological challenges prevent wider clinical use of the technique. Prohibitively expensive high field MR scanners, requiring specialist hardware and post-processing knowledge are required to obtain adequate signal to noise (SNR) to quantify the Phosphocreatine (PCr) and Adenosine Tri-phosphate (ATP) signals. Respiratory and cardiac motion is known to degrade spectral quality. Two approaches often employed to mitigate motion effects involve positioning patients prone and ECG gating; which prolongs scan times (> 30 mins.) Lying prone for long periods is not well tolerated by cardiac patients.
Objective: We investigated a number of strategies to optimise 31P spectral acquisition both ECG gated and un- gated to obtain myocardial spectra from healthy subjects using a 1H-31P cardiac surface coil on a 3T Siemens Trio MRI scanner.
Methods: Following coil spatial coil sensitivity testing using a 31P 2-D Chemical Shift Imaging (CSI) free induction decay (FID) sequence gated and un-gated ECG, free breathing, nuclear Overhauser enhanced (nOe), cardiac was used to acquire 31P spectra in un-combined mode from five healthy subjects with parameters [TR/TE: 440/2.3ms FA: 90]. Additionally an in-house PCr-optimised sequence, with saturation bands [TR/TE: 500/2.3ms FA: 45] was used to acquire spectra from a single volunteer. Finally, the impact of nOe on spectral quality was also investigated. Spectra were acquired in the prone position with fitting and quantification carried out using Siemens software. Approval was obtained from the local research ethics committee.
Results: Figure 1 shows the voxel site in the anterior myocardium from which spectra were obtained in under 14 minutes from the simple loop coil. Figures 2 and 3 show the impact of PCr flip angle and TR optimisation in the same volunteer. An increase in PCr integral of more than 10% was obtained, although ATP resonances were diminished. The application of nOe the increased PCr integral by 61%.
Discussion and Conclusion: With careful patient positioning, utilising nOe, quantifiable free-breathing, non ECG-gated spectra were obtained from a simple loop coil in under 14 minutes. The default settings (combined spectra) rendered PCr and ATP peaks unobservable. Similarly, ATP spectral peaks acquired without nOe were also unquantifiable.
Figure 1. IV septal voxel site
Figure 2. 31P spectrum from loop coil
Figure 3. Flip Angle optimised spectrum
1620
Quantitative Criteria for the Diagnosis of the Congenital Absence of Pericardium by Cardiac Magnetic Resonance
Abstract
Background: Congenital absence of the left ventricular pericardium (LCAP) is a rare and poorly known cardiac malformation. Diagnosis of LCAP is usually performed by Cardiac Magnetic Resonance (CMR) using only qualitative criteria. The aim of the present study was to establish quantitative criteria for the accurate diagnosis of LCAP on CMR.
Methods: We enrolled 9 consecutive patients affected by LCAP (mean age 26 ± 8 years, 7 males), 13 healthy controls, 13 patients with dilated cardiomyopathy (DCM), 12 patients with hypertrophic cardiomyopathy (HCM) and 13 patients with right ventricular overload (RVO). All patients underwent CMR. For each patients we measured left and right ventricular volumes, mass, atrium, end-diastolic and end-systolic longitudinal diameters (DLD, SLD) and transverse diameters (DTD, DTS) the whole heart volume change (WHVC), and the cardio-vertebral angle (CVA).
The whole-heart volume was measured in end-systole and end-diastole and difference. Whole-heart volume change (WHVC), was expressed in percentage of the end-diastolic volume. The cardio-vertebral angle (CVA) was also measured in the end-diastolic phase of the long axis cine stack as the angle identified between a line directed from the anterior margin of vertebral body to cardiac apex and the major axis of vertebral body
Results: LCAP had higher left ventricular and right ventricular DLD and SLD than all the other groups. Yet, much greater significant difference was found for the WHVC and CVA.
The WHVC was significantly higher in LCAP (21.9 ± 5.4), compared to healthy subject (8.6 ± 2.4, p < 0.001), DCM (7.1 ± 1.8, p < 0.001), HCM (9.3 ± 2.4, p < 0.001) and RVO (8 ± 2.4, p < 0.001). CVA was significantly lower in LCAP (104 ± 14°) than healthy controls (140 ± 11°, p < 0.001), DCM (139 ± 5°,p < 0.001), HCM (150 ± 6°, p < 0.001) and RVO (131 ± 8°, p<0.001).
The sensitivity and specificity of DLD and SLD of both ventricles, WHVC and CVA have been evaluated at ROC curves analysis.
WHVC had the highest sensitivity and specificity to diagnose LCAP using a threshold of >13% Statistical analysis demonstrated the statistical equivalence of interobserver and intraobserver WHVC measurements.
Conclusions: In LCAP the systo-diastolic WHVC is an optimal quantitative criteria to identify the patient with LCAP, beside the standard qualitative features and the indirect morphologic signs on CMR imagines or as a replacement when these are doubts.
1632
Widespread tissue injury during acute myocardial infarction: evidence from advanced CMR relaxometry
Abstract
Objective: In patients with ST elevation myocardial infarction (STEMI), myocardial tissue injury is not restricted to the territory supplied by the culprit artery but it affects also the remote myocardium supplied by unaffected arteries.(1) The aim was to investigate the T1 characteristics in infarcted and remote myocardium comparing it with normal healthy volunteers.
Methods: 30 patients (mean age 61 ± 10years and 70% males) with STEMI and successful revascularisation by percutaneous coronary intervention (PPCI) and 20 healthy volunteers (with no previous medical history), were included. Each patient underwent clinical CMR at 1.5 T, T1 mapping (MOLLI) pre and post contrast within 48hours after PPCI. The healthy volunteers underwent native T1 mapping only, as no contrast was given. The pre and post contrast T1 values and ECV were evaluated in each of the 16 AHA myocardial segments.
Results: Out of 480 myocardial segments in STEMI patients, 143 were affected and the rest were unaffected or remote (no regional wall motion abnormality, no oedema and no late gadolinium enhancement). The mean native T1 and the mean ECV of the remote myocardium was significantly lower compared to infarcted myocardium (1054 ± 65msec vs 1153 ± 85ms, p < 0.0001 and 0.32 ± 0.06 vs 0.46 ± 0.11ms, p < 0.0001 respectively). However when compared to 320 normal myocardial segments from healthy volunteers, the mean native T1 in the remote myocardium was significantly higher (1054 ± 65msec vs 1028 ± 48msec, p < 0.0001). The mean ECV of the remote myocardium was also higher than normal reference standard (0.32 ± 0.06 vs 0.25 ± 0.04). (2)
Conclusions: This is the first study looking at the impact of STEMI on remote myocardium via non-invasive tissue characterisation by advanced CMR relaxometry technique. Our study highlights that in STEMI the remote myocardium is also affected when compared to normal healthy myocardium. Albeit the myocardial tissue injury in the infarcted territory is significantly greater than remote myocardium. These findings may have significant future implications in the treatment of STEMI, including targeting the remote myocardial inflammation.
References:
1. Abbate A, Bonanno E, Mauriello A, et al. Widespread myocardial inflammation and infarct-related artery patency. Circulation. 2004;110(1):46-50.
2. Dabir D, Child N, Kalra Aet al. Reference values for healthy human myocardium using a T1 mapping methodology: results from the International T1 Multicenter cardiovascular magnetic resonance study. J Cardiovasc Magn Reson. 2014;16:69.
1322
Computed tomography coronary angiography verSus sTRess cArdiac magneTic rEsonance for the manaGement of sYmptomatic revascularized patients: a cost effectiveness study (STRATEGY study)
Abstract
Background: Computed tomography coronary angiography (CTCA) and stress cardiac magnetic resonance (stress-CMR) are both suitable for diagnosing obstructive coronary artery disease (CAD) in symptomatic patients for chest pain with previous history of revascularization. However, the evaluation of performance of non-invasive tests has to take in account the subsequent diagnostic, testing and medical procedures derived, clinical outcomes and cumulative costs and radiation exposure rather than their diagnostic accuracy alone. The aim of this study is to compare the clinical and economic outcomes of using anatomical (CTCA) versus a functional strategy (stress-CMR) in revascularized symptomatic patients for chest pain.
Methods and Materials: four hundred revascularized symptomatic patients for chest pain were addressed to CTCA (n: 200, mean age 68 +- yo, male 168) or stress-CMR (n:200, mean age 66 +- 9 yo, male 177) and followed-up in terms of downstream non invasive tests, invasive coronary angiography (ICA) and revascularization procedure, medical costs for CAD management, cumulative effective radiation dose and major adverse cardiac events (MACEs) defined as composite endpoints of non fatal myocardial infarction and cardiac death.
Results: all patients performed both tests successfully. The mean follow-up for CTCA and stress-CMR groups were similar (772 + -398 vs 794 + -345 days, p:ns). Compared with stress CMR strategy, CCTA was associated with an increased likelihood of subsequent mean number of further non-invasive tests (1.04 vs. 0.81, p < 0.01), cardiac catheterization (40% vs. 30%, p< 0.05). No differences were found in terms of subsequent percutaneous coronary interventions (PCI) (30% vs. 26%, p:0,37) but stress-CMR group was associated with a favorable trend of PCI/ICA rate (86% vs.75%, p:0.08) and MACEs (4% vs. 8.5%, p:0.06). Moreover, CTCA strategy showed a higher mean cost per patient [2329.46 vs. 2617.47 Euro (-11%), p < 0.05] and a lower mean effective radiation dose [22 vs. 4.4 mSv (-80% 10 yop), p < 0.01].
Conclusions: despite similar on risk profiles, revascularized patients initially evaluated with CTCA after PCI had more downstream non invasive and invasive testng, higher CAD-related spending and effective radiation exposure as compared to patients evaluated with stress-CMR with comparable MACEs.
1339
Comparison of low- versus high-dose of gadobutrol for late gadolinium enhancement imaging at 1.5 Tesla: a clinical feasibility study
Abstract
Background: Gadolinium-based contrast agent (GBCA) enhanced imaging forms the basis of contemporary cardiac magnetic resonance (CMR) imaging. There is a continuous drive to reduce doses due to possible accumulation of GBCA in human tissue and rare but potentially severe side-effects. Gadobutrol represents a second-generation macrocyclic, non-ionic GBCA, with favorable physicochemical properties resulting in a high relaxivity allowing for possible dose reduction.
Aim: To assess intra- and inter-reader comparability of late gadolinium enhancement (LGE) imaging and image quality of 0.1 mmol/kg versus 0.2 mmol/kg gadobutrol dose in a cohort of consecutive patients.
Materials and Methods: Thirty-two consecutive patients (22 males and 10 females; median age, 57.5 years, range 19-87 years) with an indication for LGE CMR imaging at 1.5 Tesla were prospectively enrolled. LGE left ventricular short axis images were acquired 5 minutes after 0.1 mmol/kg of gadobutrol administration (“low-dose”). Then, additional 0.1 mmol/kg gadobutrol were administered, LGE short axis imaging was repeated 10 minutes after administration (“high-dose”). A four-point scale was used to grade visual signal intensity, image quality, scar-to-myocardium and scar-to-blood contrast.
Results: From the initial 32 patients enrolled in the study, 3 were excluded for artifacts. Global scar detection showed no statistical difference between readers (p = 0.06), doses (p = 1) and their interaction (p = 1). When LGE positive segments were divided in ischaemic and non-ischaemic, segmental extension showed no statistically significant difference between readers (p = 0.35), doses (p = 0.13) or their combined effect (p = 0.39). Transmurality of ischaemic segments showed no significant difference as well between readers (p = 0.62), doses (p = 0.17) or their combined effect (p = 0.4). Signal intensity, image quality and scar-to-myocardium contrast were rated similar for dose (p = 0.20), reader (p = 0.30) or their interaction (p = 0.20). Scar-to-blood contrast showed no differences between readers (p = 0.06) but was significantly higher with the low-dose (p < 0.05).
Conclusions: Low-dose gadobutrol LGE imaging may represent a cost effective, safe and tolerable alternative to high-dose. Our results showed nearly identical results between readers for the detection of enhancement, image quality and signal intensity with a better delineation between scar and blood when compared to 0.2 mmol/kg body weight gadobutrol.
1347
Multi-parametric Cardiac Magnetic Resonance for Prediction of Cardiac Complications in Thalassemia Intermedia: a Prospective Multicenter Study
Abstract
Introduction: Cardiovascular Magnetic Resonance (CMR) has an established role in managing and predicting prognosis of patients with Thalassemia Major (TM). Thalassemia Intermedia (TI) is a milder variant of thalassemia showing a different clinical and prognostic profile; pulmonary hypertension (PH) is a more common complication.
We prospectively determined the predictive value of CMR parameters, including measurement of right ventricular mass, for cardiac complications in TI.
Methods: We considered 342 TI patients enrolled in the Myocardial Iron Overload in Thalassemia (MIOT) Network; about half of them (178/302, 58.9%) were transfusion-dependent. Myocardial and liver iron overload were measured by T2* multiecho technique. Atrial dimensions, left and right ventricular mass and systolic function were quantified by cine images. Late gadolinium enhancement (LGE) images were acquired to detect myocardial fibrosis [6].
Results: Twenty-three patients were excluded because a cardiac complication was present at the time of first CMR, so we prospectively followed 319 patients (mean age at first scan 38.02 ± 11.69 years, 165 females).
Mean follow-up time was 52.24 ± 24.87 months. Cardiac events were recorded in 22 patients (6.9%): heart failure (HF) in 1 patient, arrhythmias in 12 patients, pulmonary hypertension (PH) in 7 patients and myocardial infarction (MI) in 2 patients. Due to the low number of events, only arrhythmias, PH and cardiac complications globally considered were taken as cardiac outcomes for univariate and multivariate analysis.
In the multivariate analysis RV hypertrophy was the only independent predictive factor for arrhythmias (HR = 33.83, 95% CI = 6.07-188.74, P < 0.0001) and PH (HR = 73.33, 95% CI = 10.00-537.57, P < 0.0001). RV hypertrophy (HR = 24.12, 95% CI = 5.09-114.12, P < 0.0001) and myocardial fibrosis by LGE (HR = 6.59, 95% CI = 1.33-32.67, P = 0.021) were independent prognostic factors for cardiac complications in the multivariate analysis. The Figures display the Kaplan–Meier curves showing the impact of the independent predictive factors on each outcome.
Conclusions: For the first time we studied the prognostic value of RV mass as part of multiparametric CMR imaging in a population of TI patients. RV hypertrophy identified patients at high risk for arrhythmias and PH. Both RV hypertrophy and fibrosis were independent predictive factor for cardiac complications. Measurement of RV mass should be part of the multi-parametric CMR study of TI patients.
1461
Prognostic value of Cardiovascular Magnetic Resonance derived indexes of myocardial fibrosis in heart transplant recipients
Abstract
Background: The myocardium of transplanted hearts is characterized by interstitial inflammation and damage followed by death of myocytes. These lesions result in replacement and interstitial fibrosis which in the long term can impair ventricular function.The aim of the present study was to assess the association of tissue characterization using cardiovascular magnetic resonance (CMR) parameters with cardiovascular adverse events in patients who survived the first year after orthotopic cardiac transplantation (HTx).
Methods and Results: We studied 70 subjects age 48 ± 13 year, 19 F. Average time from HTx 6.5 ± 6 yrs. Myocardial T1 was determined using MOLLI protocol at 1.5 tesla consisting of 3 inversion blocks generating 3-3-5 single-shot True-FISP images. Pre- and post-Gadolinium (GD) injection short-axis T1 maps were acquired within single-breathholds. The myocardial T1 mapping values (ms) were drawn from AHA 16 regions of the left ventricle. Myocardial T1 mapping values pre- and post-GD were normalized to the T1 of blood and haematocrit to calculate the Extra Cellular Volume fraction (ECV). Late gadolinium enhancement was quantified as relative percentage of the total left ventricular mass (LGE%). A composite of major adverse cardiovascular events: cardiovascular death, hospital admission for congestive heart failure, arrhythmias, coronary revascularization (MACE) were recorded during the follow-up period after CMR median 1.73 years. Data are reported as mean ± SD. Twelve MACE were recorded (2 deaths). Patients with MACE had a higher pre-contrast T1: 1047 ±75 vs 989 ± 57 msec, p = 0.004, higher LGE%: 9 ± 10 vs 2.7 ± 3.6, p = <0.001 and higher ECV: 34 ± 0.03% vs 29 ± 0.03%, p = 0.0001. ROC curves selected thresholds of 1016 msec (Sensitivity 67; Specificity 72) for pre-contrast T1 mapping, 30% (Sensitivity 83; Specificity 59) for ECV and 2.77% (Sensitivity 83; Specificity 68) for LGE% to discriminate patients with MACE. Kaplan-Meier survival curves demonstrated that pre-contrast T1 > 1016 msec (logrank test: p = 0.04), ECV >30% (logrank test: p = 0.02) and LGE% > 2.77% (logrank test: p = 0.0007) were associated with an increased risk of sustaining an index event during follow up.
Conclusions: CMR derived indexes of fibrosis can help to stratify HTx patients with increased risk of MACE and CV death.
1523
The role of CMR in the acute phase of hospitalization: changing paradigms
Abstract
Background: Cardiac Magnetic Resonance (CMR) is an invaluable tool in the assessment of ischaemic and non-ischaemic heart disease. The incremental value of CMR in chronic cardiac conditions has already been demonstrated. However, there is limited evidence evaluating the impact of CMR in the acute hospitalised cardiac patients. We sought to evaluate the impact of CMR on the diagnosis and clinical decision-making in acute hospitalised patients.
Methods: We evaluated the 1 year registry data of 2481 consecutive scans (Jan 2014-Dec2014) at a large tertiary cardiothoracic center and identified 283 patients refered for inpatient CMR scan. CMR protocol included short axis and long axis cines, T2 weighted oedema sequences, early and late gadolinium enhancement(LGE) images. Definitions for “significant clinical impact” of CMR included change in pre-CMR diagnosis, influence on hospitalization period, change in medication, as well as influence on decision making for invasive medical procedures such as CABG, angiography and ICD implantation.
Results: Of the 283 patients, 8 (2.8%) were excluded due to poor image quality and/or incomplete scans, leaving a sample of 275 patients (66% male, mean age 59yrs, with mean ejection fraction of 46% ± 19. Patients were referred to CMR for further assessessment of ischemic heart disease, cardiomyopathy or congenital heart disease.
Overall, CMR had a significant clinical impact on 68% of the patients. This included a completely new diagnosis in 27% of the patients, change in management in 31% and a total of 10% of patients had both a new diagnosis and a change in management. CMR results led to invasive procedures on 27%, avoided invasive procedures on 16%, and had an influence on hospital discharge on 15% of the patients. 84% of the patients had echocardiography prior to CMR. CMR confirmed the echo diagnosis in 11%, complemented the echo findings by additing significant new information in 41% and changed the echo diagnosis in 30% of the cases.
In a multivariable model that included clinical and imaging parameters, age and presence of LGE were the only independent predictor of “significant clinical impact” (LGE p-value .007, OR 2.782, CI 1.328-5.828)
Conclusions: CMR had a significant impact in patient's diagnosis and management in 68% of acutely hospitalised patients. The presence of LGE was the only independent predictor of significant clinical impact following CMR.
1542
Preoperative CMR-based score predict ventricular response after surgical left ventricular reconstruction in ischemic heart failure patients
Abstract
Objectives: Left ventricular (LV) remodeling is a complex process following acute myocardial infarction (MI) and leading to a dilated and distorted LV cavity. Surgical LV reconstruction (SVR) aims to reduce LV end-diastolic volume by excluding scar tissue and to restore a more physiological LV geometry, thereby improving LV function and clinical status. However, factors determining outcome are still not completely understood. Cardiac magnetic resonance (CMR) is the gold standard imaging technique to show necrotic myocardium (scar tissue) in vivo and to quantify LV function. Based on the observation that the total amount and the regional distribution of scar tissue could affect postoperative reverse remodeling, we suggest a new score including preoperative CMR parameters assumed to affect LV function (CMR score) in order to predict postoperative outcome.
Methods: Patients with previous MI eligible for SVR were studied by contrast-enhanced CMR before and six months after surgery. The CMR protocol included standard sequences for anatomical and functional evaluation of the ventricles and post-contrast sequences for late gadolinium enhancement (LGE) evaluation. The CMR score includes the following preoperative CMR parameters: LV end-diastolic volume, LV end-systolic volume, RV end-systolic volume, LV diastolic sphericity index, LV mass and wall thickness, LGE mass and distribution.
Results: We studied 18 patients (mean age 65 ± 10 years, 16 males). CMR score showed an inverse, significant statistical correlation with LV ejection fraction (LVEF) (rho = -0.60, p value = 0.009) and with right ventricular ejection fraction (RVEF) at 6-month follow-up (rho = -0.67, p value = 0.002) (see Figure).
Conclusion: Our preliminary data suggest that anatomical and functional variables from preoperative CMR could predict the effect of SVR on LV and RV function. CMR analysis of LGE mass and distribution could play a role in determining postoperative cardiac function. This study suggests the clinical usefulness of CMR score that, however, needs to be validated in a larger population.
1555
Excellent response rate to cardiac resynchronization therapy guided with magnetic resonance imaging
Abstract
Objectives: Cardiac resynchronization therapy (CRT) is a well-established method for patients with drug refractory symptoms of heart failure. However, there is still substantial number of non-responders to CRT. We aimed to prove that cardiac magnetic resonance (CMR) can accurately identify the target segment for CRT implantation and therefore increase the response rate.
Methods: The study population consisted of 45 patients (age 63 ± 12y, 71% males) undergoing CRT implantation. All patients underwent 2D echocardiography (ECHO) including strain analysis on Vivid E9 (GE Healthcare) and CMR on Siemens Trio 3T scanner including tagging sequence with strain analysis (inTag Sofware). The target segment for lead implantation was defined as the latest site of mechanical contraction (both ECHO and CMR) with < 25% myocardial scar determined by the late gadolinium enhancement (CMR). CRT implantation was guided by CMR results only. We aimed to achieve the lead/ target segment exact concordance. All patients were followed up including repeat ECHO and clinical assessment. Responders were defined as LV end-systolic diameter reduction > 10% together with NYHA class improvement ≥ 1 at 12 months follow-up.
Results: The average length of the follow up was 20 ± 7 months with 11.1% death rate and 8.9% rate of hospitalizations for heart failure. Symptoms at baseline were NYHA II in 40% and NYHA III in 60%. Left bundle branch block (LBBB) (Strauss criteria) was present in 88.9%, mean QRS duration was 165 ± 14 ms. CMR derived mean LV EF was 24 ± 9%, end-diastolic volume 219 ± 58 ml, end-systolic volume 169 ± 59 ml. Image quality for target segment detection was suboptimal in 2 patients (4.4%) in CMR study. The response rate was 100% in the subgroup of 18 patients with LBBB and the exact concordance of the lead position/target segment assessed by CMR. The response rate was 59% in 22 patients with LBBB and discordant lead position/target segment. There were 5 patients without LBBB with the response rate 0%, independent of the lead position/target segment concordance. The response rate was 81% in patients with LBBB and concordant lead/target segment assessed retrospectively by ECHO and 69% in patients with LBBB and discordant lead/target segment (ECHO). Fig 1,2
Conclusions: CMR enables detection of the site of latest LV mechanical activation in 95.6%. Successful trans-venous targeted lead placement results in excellent response rate, but remains challenging and surgical placement might be an option.
1626
The ECG as a predictor of arrhythmogenic substrate on Cardiac Magnetic Resonance Imaging in patients undergoing ablation for premature ventricular contractions
Abstract
Background: The most likely origin of premature ventricular contractions (PVC) may be deduced from surface ECG analysis during pre-procedural planning of an electophysiological (EP) procedure. Apart from purely benign forms of increased ventricular ectopy, myocardial substrate (e.g. regional fibrosis) may be present in certain cases which will significantly impact the invasive treatment approach. Cardiac magnetic resonance imaging (CMR) can reliably identify fibrotic target lesions and, hence, may assist in adequate patient selection and procedural planning.
Methods: 48 patients (60% male, median age 56 ± 14.8, median PVC count 18748 ± 13175) referred for radiofrequency (RF) ablation of PVCs at our institution who underwent CMR before EP study were included in the analysis. A 12 lead electrocardiogramm (ECG) documentation of PVCs was available in all patients. CMR was performed at 1.5Tesla using cine, T1/T2-weighted and 3D-delayed enhancement imaging in standard cardiac geometries. RF ablation was performed according to local protocols using a 3-D-mapping tool (Carto®) and a 3.5 mm Thermocool catheter, flow rate 30 ml/min.
Results: On surface ECG right bundle branch block (RBBB) morphology was present in 50% of patients. Fibrotic substrate on CMR was present in 11 patients. In univariate analysis, RBBB morphology on surface ECG significantly predicted the presence of substrate (p = 0.01). Multiple PVC morphologies (≥2) were present in 6 patients and also predicted substrate on CMR (p = 0.006) Patients with substrate showed no significant difference with regard to baseline characteristics including known structural heart disease or left ventricular ejection fraction (43% ± 13 vs. 50% ± 11, p = 0.09). CMR-identified ventricular substrate was verified to represent the site of origin of the clinical PVCs during EP study in all patients. Succesful RF ablation of the clinical PVC (reduction of PVC burden >95%) was achieved in 93.8% of all cases.
Mean ablation time was signifiantly longer and the mean maximum applied energy was higher in patients with ventricular substrate (1479 ± 971s vs. 722 ± 744s, p = 0.012 and 50 ± 7Watt vs. 43 ± 10 Watt, p = 0.04 respectively).
Conclusion: ECG morphology of PVCs is a significant predictor of the presence of myocardial subtrate on CMR imaging. In patients with RBBB morphology and/or multiple PVC patterns CMR should be performed routinely before ablation in order to guide the electrophysiological procedure and improve success rate.
1649
Comparison of T1-mapping at 3.0T CMR and angiographic APPROACH score for area at risk assessment in ST-segment elevation myocardial infarction
Abstract
Background: Myocardial salvage index (MSI) derived using cardiac magnetic resonance (CMR) is a strong independent predictor of adverse remodeling and prognosis post STEMI. It is typically imaged using a T2-weight short tau inversion recovery sequence, which is susceptible to an array of artifacts limiting its usefulness to evaluate MSI as a reliable surrogate marker of myocardial injury. T1-mapping has recently come to the fore as a viable alternative but it requires acquisition of, often time consuming, additional images and post-processing. Optimal detection of reversible myocardial injury could improve the reliability of MSI determination and its prognostic relevance.
Objective: To compare the robustness of myocardial oedema detection using free-breathing motion-corrected (moco) T1-mapping on 3.0T CMR and the validated ‘APPROACH’ angiographic area at risk (AAR) score in patients presenting with STEMI.
Methods: Forty-five patients underwent CMR 1-5 days following presentation with STEMI. AAR was quantified using resulting parametric colour maps from moco-T1 Modified Look Locker Inversion Recovery (MOLLI) sequences and compared to the AAR derived in a core-lab using the APPROACH score in a using analysis of variance (ANOVA). Pearson's correlation coefficient was used to assess correlation. Inter-sequence agreement was assessed using the Bland-Altman method, coefficient of variation (CoV) and two-way mixed-effect intra-class correlation coefficient (ICC) for absolute agreement.
Results: Patient characteristics are presented in Table 1. AAR assessed using the two methods was significantly different (p = 0.02) with moderate to poor correlation and agreement (see Table 2). However, the diagnostic imaging rate obtained with both techniques was excellent
Conclusions: moco-T1-mapping using MOLLI at 3.0T CMR and angiographic APPROACH score are not inter-changeable for detecting reversible myocardial injury and determination of MSI for the prediction of functional recovery following STEMI. Further work is required to assess the test-retest repeatability and overall reliability of each technique.
1340
Pathological correlates of left bundle branch disease in patients with non-ischemic cardiomyopathy: a cardiovascular magnetic resonance study
Abstract
Background: The pathological correlates of intraventricular conduction delays in patients with non-ischemic cardiomyopathy (NICM) have been scarcely investigated.
Aim: To assess left ventricular (LV) structural, functional and tissue abnormalities associated with left-system intraventricular conduction delay, either left anterior hemiblock (LAHB) or complete left bundle branch block (LBBB), in a cohort of NICM patients submitted to cardiovascular magnetic resonance (CMR).
Materials and Methods: Twelve lead electrocardiogram and CMR were performed in 196 consecutive patients with NICM. The presence and extent of myocardial fibrosis was evaluated with late gadolinium enhancement (LGE) technique. The presence of left-bundle disease (LBD) including LBBB or LAHB, right-bundle branch block, aspecific conduction delay and normal intraventricular conduction (nIVC) was recorded.
Results: Compared to nIVC patients, those with LBD were older (66 vs 59.5 years, p = 0.001), had greater LV volumes (p = 0.035 for end-diastolic and p = 0.009 for end-systolic volume) and mass (92 vs 85 g/m2, p = 0.034), and showed lower LV ejection-fraction (33 vs 40%, p = 0.008). LGE was observed more commonly in LBD than nIVC patients and was more often located in the septum (all P < 0.001). At multivariate analysis, septal LGE (but not fibrosis in other LV regions) was independently associated with a higher likelihood of LBD (OR 6.1, 95% CI 2.9 to 12.7, P < 0.001), even after correction for LV volumes, mass and ejection fraction.
Conclusions: In NICM, the presence of LBD is associated with worse LV remodeling and dysfunction than nIVC. Septal fibrosis yielded a 6-fold greater likelihood of LBD independently of the degree of LV dilatation and systolic dysfunction.
1342
Myocardial remodelling and fibrosis in nonischaemic dilated cardiomyopathy: insights from cardiovascular magnetic resonance
Abstract
Background: In nonischaemic dilated cardiomyopathy (DCM), myocardial fibrosis can be detected by cardiovascular magnetic resonance (CMR) as late gadolinium enhancement (LGE) and has been associated with worse prognosis. We investigated whether LGE is associated with left ventricular (LV) reverse remodelling in DCM.
Methods: Ninety-seven DCM patients (age 56 ± 14 years, 62 male) with LV ejection fraction ≤50% were enrolled and underwent baseline CMR; patients with ischaemic, valvular, congenital heart disease, other cardiomyopathies or contraindications to CMR were excluded. After a median 29-month follow-up (interquartile range 16-46) on optimal medical therapy, all patients underwent a second CMR; patients who died, underwent device implantation or declined a second CMR, were also excluded from the study. LGE was quantified on post-contrast CMR images as percentage of LV mass. LV reverse remodelling was defined as a decrease >10% of LV end-systolic volume at follow-up.
Results: Mean LV ejection fraction was 37 ± 10% at baseline, 44 ± 12% at follow-up. LGE was present in 56 (58%) patients at baseline (median 6% of LV mass, interquartile range 3-11%), without significant differences at follow-up (median 7%, interquartile range 4-11%, p = NS). Patients experiencing LV reverse remodelling during follow-up (n = 53, 55%) presented a baseline worse LV ejection fraction (34 ± 11%) than patients not experiencing LV reverse remodelling (41 ± 8%, p < 0.01), greater LV end-diastolic volume (125 ± 38 vs. 112 ± 26 ml/m2, p = 0.05), worse right ventricular ejection fraction (54 ± 13% vs. 59 ± 10%, p = 0.03), but significantly less fibrosis (median 5% of LV mass, interquartile range 2-9% vs. median 9%, interquartile range 3-15%, p < 0.01); no age or gender differences were observed (p = NS). Multivariate regression analysis showed that LGE extent at baseline CMR was a negative predictor of LV reverse remodelling (p < 0.01), even after correction for age, New York Heart Association class, LV volumes and systolic function.
Conclusions: In DCM patients, LGE extent was a negative independent predictor of LV reverse during follow-up, irrespective of the initial clinical status and the severity of ventricular dysfunction.
1411
The association between fibrosis and contractile dysfunction in hypertrophic cardiomyopathy assessed by cardiovascular magnetic resonance
Abstract
Introduction: At present the diagnosis of hypertrophic cardiomyopathy (HCM) is made when a segment measures ≥15mm using cardiac imaging. Using cardiovascular magnetic resonance (CMR) extracellular volume (ECV) mapping techniques it is possible to detect early phenotypic manifestation of HCM, such as extracellular matrix expansion, which can occur before overt hypertrophy. It is unknown whether strain impairment occurs early in the disease process when there is ECV expansion with overt hypertrophy.
Methods: 50 patients with HCM underwent CMR studies at 3.0T including cine imaging in multiple planes, T1 mapping for calculation of ECV and late gadolinium enhancement (LGE) imaging. For each segment of the American Heart Association (AHA) model of each subject segment thickness, the presence of LGE and strain by feature tracking were measured.
Results: Circumferential strain by feature tracking (Ecc-FT) was lower in segments with ECV expansion (-17.0 ± 10.3 vs -19.9 ± 8.4%, N = 783 P < 0.001). However in segments <15mm Ecc-FT was not significantly different between those with and without ECV expansion (-19.7 ± 9.2 vs -20.6 ± 7.9%, N = 684, P = 0.43). In segments <15mm Ecc-FT was significantly lower in those with LGE (-12.9 ± 8.2 vs -20.9 ± 8.1%, N = 684, P < 0.001).
There were significant correlations between Ecc-FT and both segment thickness and ECV (Rs = 0.616 and 0.176, P < 0.001 for both). However on multivariable linear regression only segment thickness had a significant association with Ecc-FT (beta = -0.54, p < 0.001).
Conclusion: Regional strain impairment is predominantly associated with the degree of hypertrophy and replacement fibrosis assessed by the presence of LGE. In non-hypertrophied segments strain is not significantly impaired by the presence of interstitial fibrosis detected by ECV expansion on T1 mapping. Therefore the presence of interstitial fibrosis may be a more useful method than impairment of strain of identifying HCM in subjects with borderline LV hypertrophy.
Figure. (A) Gross hypertrophy (>15mm) in the septum. (B) FT strain map of the same slice, shown in systole. Colour scale ranges from akinetic (red) to >20% strain (blue). (C) LGE imaging showing replacement fibrosis in inferoseptum (white arrow). (D) ECV map ranging from 0 (blue) to 100% (red). Note the replacement fibrosis (black arrow) and the non-hypertrophied anterior and anterolateral walls have marked ECV expansion (35-45%) yet have normal radial strain on FT imaging.
1622
Persistent myocardial inflammation due to intramyocardial haemorrhage in reperfused STEMI as a precursor to adverse LV remodelling - insights from multi-parametric mapping
Abstract
Background: Microvascular obstruction (MVO) and intramyocardial haemorrhage (IMH) still occur in reperfused ST-segment elevation myocardial infarction (STEMI) patients and are associated with adverse left ventricular (LV) remodelling and worse clinical outcomes. We aimed to use T1, T2 and T2* mapping to understand the changes occurring in the myocardial infarct (MI) zone and the remote myocardium in reperfused STEMI patients with and without adverse (LV) remodelling.
Methods: 40 reperfused STEMI patients underwent a CMR at 4 ± 2 days and at 5 ± 2 months post PPCI on a Siemens 1.5T scanner. Native T1 (MOLLI), T2, T2* and post contrast T1 maps were acquired. Co-registered and motion-corrected automated ECV maps were generated offline. 20 matched healthy volunteers served as controls. IMH and chronic iron deposition were defined as a hypointense core with T2* < 20ms. Adverse LV remodelling was defined as ≥20% increase in LV end-diastolic volume.
Results: 26 out of 40 patients (65%) had MVO. Figure 1 shows an example of the multi-parametric maps of an inferior STEMI with MVO and IMH and chronic iron deposit on the follow-up scan (arrows). The T2* in the MI zone in patients with IMH was 13 ± 3ms compared to 30 ± 6ms in those without (P < 0.001). All patients with adverse LV remodelling (8/40, 20%) had MVO with IMH. 87% of patients with IMH on the acute scans developed chronic iron deposition and displayed higher T2 values in the surrounding infarct zone compared with those without at follow-up (T2: 61 ± 4ms versus 53 ± 4ms, P = 0.001). ECV in remote myocardium (ECVRemote) was higher on the acute scan when compared to control (27.9 ± 2.1% versus 26.4 ± 2.1%, P = 0.01), but this difference was non-significant at follow-up (27.0 ± 2.1% versus 26.4 ± 2.1%, P = 0.30). In the 8 patients with adverse LV remodelling, the ECVRemote was higher on both the acute and follow-up scan when compared to those without adverse LV remodelling (acute scan: 29.5 ± 1.4% vs. 27.4 ± 2.0%; P = 0.01; follow-up scan 28.6 ± 1.5% vs. 26.6 ± 2.1%; P = 0.02). There was no difference in T2 values in the remote myocardium in these patients.
Conclusions: Most patients with IMH developed chronic iron deposition. This is associated with persistent inflammation on T2 mapping and adverse LV remodelling. Adverse LV remodelling is associated with remote ECV expansion. IMH - either acutely or chronically–is a potential therapeutic target to prevent inflammation and subsequent adverse LV remodelling.
Figure 1
1566
Semiquantitative analysis of low and high b value DWI for detecting myocardial edema in acute myocarditis
Abstract
Objectives: Evaluate the clinical feasibility and accuracy of semiquantitative analysis of DWI with low and high b value in patients with acute myocarditis compared to T2 STIR.
Methods: 20 patients with confirmed acute myocarditis and 17 control subjects were submitted to cardiac MRI. The cardiac MRI protocol included short-axis black blood T2 STIR, delayed enhancement and DWI. MRI studies were performed using a 3 Tesla and 1,5T Magnet.
We performed a whole cardiac SSh EPI DWI sequence including three b values (0, 50 and 300 mm2/s) with cardiac trigger, acquired in diastole and apnea.
Semiquantitative analysis of the signal intensity (SI) and ADC map by plotting ROIs on areas of inflamed myocardium showing high signal intensity on both b50 and b300 mm2/s (SIb300; ADCb300) and in those regions with high signal intensity on b50 mm2/s that loss signal at b300 mm2/s (SIb50; ADCb50). The ratio of signal intensity of T2 STIR myocardium vs pectoral muscle (T2 Ratio), SI and ADCb50 Ratio (SIb50/SIpectoral; ADCb50/ADCpectoral), SI and ADCb300 Ratio (SIb300/SIpectoral; ADCb300/ADCpectoral) were obtained.
Results: Significant differences between patients with acute myocarditis in comparison to normal controls were found in T2 Ratio (2.40 ± 0.674 vs. 1.74 ± 0.29: p < 0.001), SIb300 (803.55 ± 640.29 vs. 891.17 ± 220.9; p: 0.018) and SIb300 Ratio (2.59 ± 1.61 vs. 1.43 ± 0.69; p < 0.001). SIb50 and ADC based measurements did not show significant differences among both groups.
ROC curves revealed a high area under the curve of T2 Ratio (0.874; p < 0.001); SIb300 (0.725; p: 0.020) and SIb300 Ratio (0.828; p < 0.001).
Conclusion: DWI (SIb300 and SIb300 Ratio) detect myocardial edema in a similar manner to T2 STIR and with excellent interobserver agreement.
1567
Value of Cardiac MRI In Detecting Coronary Artery Disease In Newly Diagnosed Systolic Dysfunction
Abstract
Introduction: Coronary angiography (CA) is systematically realized in case of newly diagnosed systolic dysfunction (SyD) but 75-80% are normal. Subendocardial enhancement in Cardiac-MRI (CMR) is known as highly specific for ischemic disease, and its value to diagnose Coronary Artery Disease (CAD) has been suggested in small series.
Objective: Determine the value of CMR in detecting CAD for patients with SyD (left ventricular ejection fraction LVEF ≤45%) without obvious etiology.
Methods- From 2009 to 2015, 305 patients were retrospectively included with newly diagnosed LVEF ≤45% unexplained by clinical, electrical and echographic evaluation. A Coronary Angiography (CA) and CMR had to be performed within 3 months. 40 CMR and 40 CA were re-interpreted to evaluate the fiability of the interpretations. 4 groups were defined, depending on CAD diagnosed or not by CA (« CA+ or –») or CMR (« CMR+ or - »).
Results: CA was normal for 247 (81%) patients and identified 58 (19%) ischemic heart diseases; CMR showed a late subendocardial enhancement in 85 (28%) patients. 20 (6,5%) patients were revascularized. Most of CMR and CA were concordant: 215 (70%) patients « CA- and CMR- » and 56 (18%) patients « CA+ and CMR+ ». 2 (0,7%) patients were « CA+ but CMR- »; none were revascularized. 32 (10%) patients were « CA- but CMR+ »; for most of them, diagnosis of MINOCA (Myocardial Infarction and Non-Obstructive Coronary Arteries) was withheld: reperfused acute myocardial infarction, coronary spasm … Sensibility of CMR compared to CA was 97%, with a good concordance in terms of coronary territory too; specificity was 87%. Sensibility of CA compared to CMR was 64% and specificity 99%. For 84 (28%) patients CMR allowed another diagnosis: myocarditis, LV non-compaction, Tako-Tsubo …
Conclusions: CMR detected 97% of significant CAD at CA in patients with newly diagnosed LVEF ≤ 45%, and recognized a lot of MINOCA not diagnosed by CA. Every patients revascularized had late gadolinium enhancement. Thus, in this indication, CMR seems more and more interchangeable with CA.
1570
Usefulness of cardiac magnetic resonance in tuberous sclerosis complex
Abstract
Objective: Tuberous sclerosis complex (TSC) is a rare autosomal dominant inherited syndrome, which may affect many organs, most commonly brain, kidneys, lungs and heart. The most common cardiac manifestation is the rhabdomyoma, a benign tumour that usually undergoes spontaneous regression in the first years of life. The aim of this study was to review the cardiac magnetic resonance (CMR) findings in TSC adult patients.
Methods: TSC adult patients referred to CMR between April 2012 and January 2015 were analysed. Five patients were referred, but in one patient CMR could not be done due to claustrophobia. A transthoracic echocardiography (TTE) was performed before CMR (1.5T Avanto Dot, Siemens, Erlangen and Signa LX 9.1, General Electric, Milwakee). The CMR protocol included the following sequences: b-steady-state free precession (b-SSFP) cine sequence, T1 black-blood turbo spin-echo (BB-TSE), T1 BB-TSE with fat saturation (fat-sat), myocardial perfusion with a saturation-recovery turbo field-echo sequence and late gadolinium-enhancement sequence (LGE).
Results: Mean age was 36,6 ± 11,9 years and 4/5 were female. All patients had brain involvement, with medical history of epilepsy in 4/5. Renal involvement with angiomyolipomas was present in 3/5. No patient had cardiovascular symptoms. Intermittent preexcitation in the electrocardiogram was present in one patient but without history of tachycardia. Echocardiography and CMR findings are summarized in the table. In 3/4 patients, CMR showed at least one intramyocardial area of hypointensity in b-SSFP and hyperintensity in T1 BB-TSE with a marked signal decrease in T1 BB-TSE fat-sat (figure). In the echocardiography these areas presented as rather hyperechogenic (figure). Finally, global and segmental functions were normal and no pathological enhancements were found.
Conclusion: Adult patients with TSC may have myocardial fatty infiltrates probably due to a rhabdomyoma regression. CMR is very useful in TSC because it provides a better characterization of cardiac disease.
1578
Papillary muscles offer further insight into hypertrophied hearts: a cardiovascular magnetic resonance study
Abstract
Background: The aetiology of left ventricular hypertrophy (LVH) is often difficult to determine. The left ventricular papillary muscles (LVPM) sometimes appear differentially hypertrophied, particularly in Fabry disease (FD) where storage appears detectable by cardiovascular magnetic resonance (CMR) T1 mapping.
Aim: To characterise LVPM mass across a range of heart diseases with LVH, and to gain further insight into the mechanisms of LVPM hypertrophy in FD using T1 mapping.
Methods: 538 cases were retrospectively recruited from research CMR cohorts: 125 FD, 85 hypertrophic cardiomyopathy (HCM), 67 amyloid, 82 aortic stenosis (AS), 40 hypertension, and 79 controls. LVPM contribution to total LVM was measured using manual contouring on CMR short axis cine images. T1 values (septal and LVPM) were measured using ShMOLLI sequences in FD and controls.
Results: LVPM contribution to LVM was the highest in the LVH + ve FD group and significantly increased compared to all other groups (FD: 13 ± 3%, HCM 10 ± 3%, amyloid 8 ± 2%, AS 7 ± 3%, hypertension 7 ± 2%, controls 7 ± 1%; p < 0.001). LVH + ve HCM also had significantly increased LVPM. In the LVH-ve cohorts, only FD had significantly increased LVPM (11 ± 3%; p < 0.001).
In FD there was relative concordance in septal and LVPM T1 values. LVH + ve FD: when the septal T1 was low, LVPM T1 was low in 90%. LVH-ve FD: when the septal T1 was normal, the LVPM T1 was normal in 70% (indicating no detectable storage), and when the septal T1 was low, 75% had a low LVPM T1 (indicating storage). Despite these findings, LVPM contribution to LVM was similar between the low septal T1 and normal septal T1 groups in LVH-ve FD (11 ± 3% vs 10 ± 3%, p = 0.08).
Conclusion: Disproportionate hypertrophy of LVPMs in LVH + ve hearts is seen in Fabry and HCM. This phenomenon also occurs in LVH-ve FD. Low T1 is not always present in FD LVPM hypertrophy, implicating other pathological mechanisms.
1627
Diagnostic and clinical implications of CMR timing (early versus late) in patients with troponin positive acute coronary syndromes and unobstructed coronary arteries
Abstract
Background: 7-15% of patients with acute coronary syndrome (ACS) have unobstructed coronary arteries. In these patients cardiac magnetic resonance (CMR) can identify different underlying aetiologies.
Aim: To evaluate the diagnostic and decision making implications of CMR timing (early versus late) in patients with ACS and unobstructed coronary arteries.
Methods: Registry data on 204 consecutive patients (mean age 55yrs, 51% males) with troponin positive ACS and unobstructed coronary arteries, referred for a CMR (Sept 2011-Jul 2014) were evaluated. An “early” CMR (<2weeks from presentation) was performed in 98 patients and 106 patients underwent a “late” (>2weeks from presentation) comprehensive CMR scan. Table 1. “Significant clinical impact” was predefined as change in diagnosis/management. Propensity matching was performed between early and late CMR groups to minimise any selection bias due to the differences in clinical characteristics.
Results: Overall, a cause for the troponin rise was found in 70% of patients and CMR had a significant clinical impact in 66%, including leading to change in the final diagnosis in 54%. In a multivariable model that included clinical and imaging parameters, presence of late gadolinium enhancement (LGE) and age were the only independent predictors of “significant clinical impact” (LGE OR 2.3, p = 0.02). Table 2. In a propensity score analysis, 58 pair of patients was matched for early and late CMR. The diagnostic pick up rate in the “early” was significantly higher than “late” group (88% vs 50% p < 0.0001). Myocarditis (33%) was the most common diagnosis in the “early” group, whereas myocardial infarction (MI) (22%) was most prevalent in the matched “late” group. The prevalence of MI was not different in the “early” and “late group (22% vs 26% respectively, p = 0.83). The clinical impact also improved significantly in the early group compared to the propensity score matched late group. (76% vs 51%, p = 0.01).
Conclusion: In a large cohort of patients with troponin positive ACS and unobstructed coronary arteries CMR was able to establish a final diagnosis in overall 70%. CMR made a significant additive clinical impact on management and diagnosis in 66% of patients, with LGE being the best independent predictor of impact. Moreover, the diagnostic value as well as the clinical impact of CMR improved significantly when carried out within 2 weeks from presentation.
1323
Evaluation of right ventricular transverse strain and strain rate in patients with acute ST-segment elevation myocardial infarction: a cardiac magnetic resonance feature tracking study
Abstract
Background: Impaired right ventricular (RV) systolic function after acute ST-segment elevation myocardial infarction (STEMI) is associated with poor clinical outcomes. CMR feature tracking technique enables quantification of myocardial deformation and becomes promising method. The aim of this study was to evaluate longitudinal and transverse strain and strain rate of RV free wall in patients with acute isolated left ventricular STEMI using CMR feature tracking.
Methods: A total of 38 participants (mean age 59 ± 10.8 years; 60% male) were recruited into the study and underwent CMR at 1.5 Tesla (Siemens Magnetom Aera). TomTec 2D Cardiac Performance Analysis software was used to analyze mean and segmental longitudinal and transverse strain and strain rate of the RV free wall. Patients with RV myocardial infarction (LGE of RV myocardium) were excluded.
Results: RV free wall longitudinal strain and strain rate parameters did not reach statistical significance. Patients with acute STEMI demonstrated lower mean RV free wall transverse strain (12.40 ± 8.10% vs 21.70 ± 7.32%, p = 0.0001) and strain rate (1.08 ± 0.33 s-1 vs 1.51 ± 0.38 s-1, p = 0.001) than controls. We also performed segmental analysis. Transverse strain was significantly impaired in the basal (15.52 ± 8.81% vs 24.52 ± 11.29%, p = 0.03), mid (11.16 ± 8.34% vs 18,21 ± 7.57%, p = 0.02) and apical (10.52 ± 11.53% vs 22.37 ± 10.33%, p = 0.001) segments of RV free wall in patients with STEMI comparing with controls. Transverse strain rate was lower in the basal (0.97 ± 0.44 s-1 vs 1.51 ± 0.43 s-1, p = 0.001) and mid (0.91 ± 0.31 s-1 vs 1.36 ± 0.41 s-1, p = 0.01) segments in STEMI group. There was no difference between groups for apical transverse strain rate of RV free wall.
Conclusions: This study demonstrates that transverse strain and strain rate of RV free wall are impaired in patients with acute isolated left ventricular STEMI while longitudinal RV systolic function remains normal.
1333
Cardiac resynchronization in ischemic heart failure patients: a comparison between therapy guided by cardiac magnetic resonance imaging and 2D-speckle tracking echocardiography
Abstract
Aims and Objectives: In heart failure (HF) patients, cardiac resynchronization therapy (CRT) has beneficial effects on symptoms, left ventricular (LV) systolic function and survival. In ischemic HF patients non-response rate to CRT is higher as compared to non-ischemic HF patients. Aim of study was to explore role of late gadolinium enhancement cardiac MRI (LGE-CMR) and 2D-speckle tracking echocardiography to guide CRT implantation in ischemic HF patients.
Methods and Materials: 90 ischemic HF patients were divided in 2 groups. Group 1: 30 patients undergone to LGE-CMR and 2D-speckle tracking echocardiography at baseline. Scar transmurality was assessed on LGE-CMR images using 17-segment model. Transmural scar was defined as having >50% of thickness in any of 17 segments. Viable segment was defined when scar tissue was < 51% of its thickness. From 2D-speckle tracking echocardiography was derived global longitudinal strain bulls eye of time to peak longitudinal strain. Target zone for LV lead placement was defined most delayed viable segment based on information derived from LGE-CMR and 2D-speckle tracking echocardiography. Group 2: 60 patients were implanted according current clinical practice with LV lead preferably placed in (postero-)lateral vein without any imaging information. Response to CRT was defined as reduction of LV end-systolic volume ≥15% at 6 months follow-up.
Results: Baseline characteristic were not different between the two study groups. Response rate to CRT was significantly higher in group 1 (73% vs 52%, p = 0.045).
Conclusions: LGE-CMR together with 2D-speckle tracking echocardiography are important tools to guide CRT implantation in ischemic HF patients in order to improve response rate to CRT.
1338
Cardiac magnetic resonance versus bisphosphonate scintigraphy for diagnosis of cardiac amyloidosis
Abstract
Background: Cardiac amyloidosis (CA) is a disease still under-diagnosed, requiring a multidisciplinary approach. Cardiac magnetic resonance (CMR) imaging, with late gadolinium enhancement (LGE) and bisphosphonates scintigraphy (BS) represent two valuable tools for CA diagnosis. Currently, no comparison has been performed between LGE CMR and single photon emission computed tomography (SPECT) findings.
Aim: To compare SPECT global and regional radiotracer myocardial uptake and LGE distribution in patients with transthyretin-related CA (TTR-CA) or light-chain cardiac amyloidosis (AL-CA). To assess CMR and BS feasibility in patients with CA and to test diagnostic accuracy of hydro-methylene-bisphosphonates in CA.
Results: Eighty-two patients with clinical suspicion of CA underwent BS between 2011-2015. CMR was not performed in 35 of them (43%). Twenty-five patients (30%) were excluded because of CMR-related contraindications (implanted with CMR incompatible devices, claustrophobia, advanced chronic kidney disease). Forty-six patients (57%) underwent both CMR and BS representing study population. TTR-CA and AL-CA groups did not show statistical differences in terms of biventricular volumes, masses and myocardial thickness. Moreover, no differences were noted in terms of LGE presence with the exception of LGE apex-to-base gradient that was observed in 26% of TTR-CA and in no AL-CA patient (p < 0.05).
With regard to BS specificity and sensitivity CA generic diagnosis was 100% and 66% respectively, whereas CMR presented 100% specificity and sensitivity. When a specific diagnosis of TTR-CA was considered, BS specificity and sensitivity was 100% and 95% respectively, while CMR specificity and sensitivity was 67% and 100% respectively. SPECT semi-quantitative evaluation, assessed as number of beats per voxel per minute (bvm) > 9851 allowed to identify segments likely comprising amyloid deposits with 87% specificity and sensitivity. Comparison between SPECT and CMR showed a 93% concordance for the left ventricle (p < 0.0001) and 84% concordance (p < 0.005) for the right ventricle.
Conclusions: Combined CMR and SPECT use allows exploiting CMR accuracy for CA diagnosis and SPECT diagnostic accuracy for TTR-CA detection. SPECT and CMR segmental comparison of cardiac amyloid is highly concordant for both ventricles in TTR-CA.
1341
Strain relaxation index, a novel tagged MRI-derived diastolic function parameter, is impaired in metabolic syndrome
Abstract
Background: Strain relaxation index (SRI) is a recently described diastolic function parameter derived from tagged magnetic resonance imaging (MRI) that independently predicts heart failure. Subclinical structural and functional cardiac changes in patients with metabolic syndrome (MetS) are being increasingly recognized and are associated with increased insulin resistance.
Objective: To evaluate the association between tagged-MRI derived left ventricular diastolic functional parameters and insulin resistance, across the continuum from MetS to type 2 diabetes mellitus (T2DM).
Methods: This cross-sectional study included 1,558 subjects from the Multi-Ethnic Study of Atherosclerosis (MESA) baseline examination who underwent diastolic function analysis by tagged-MRI. Diastolic parameters were early diastolic strain rate (EDSR), peak torsion recoil rate (pTRR) and SRI (defined as the difference between post-systolic and systolic times of the strain peaks, divided by the EDSR peak). Insulin resistance was determined by homeostatic model assessment of insulin resistance (HOMA-IR). Subgroups were defined assuming categories of increasing insulin resistance (IR): 1) no MetS and no T2DM; 2) presence of MetS without T2DM; 3) presence of T2DM. Multivariate linear regression was used to determine the relationship of MetS and IR categories with diastolic function, including age, gender and ethnicity as co-variates.
Results: The final population included 53% males, mean age of 65.5 ± 9.7 years. Subgroups 1, 2 and 3 were comprised by 937 (60%), 407 (26%) and 214 (14%) subjects respectively. HOMA-IR increased from subgroup 1 (1.8 ± 0.9) toward subgroup 2 (3.0 ± 1.8) and subgroup 3 (6.5 ± 27.5), p for trend <0.001. Significant decrease in EDSR and pTRR and an increase in SRI, meaning worse diastolic function, were found across the groups (Table 1). In the multivariate regression analysis, MetS was associated with worse SRI (β-coefficient = 0.263; 95% CI 0.086-0.441), but not with EDSR and pTRR. Considering subgroup 1 as reference, MetS without T2DM was associated with worse SRI (β-coefficient = 0.274; 95% CI 0.074-0.474), and a tendency to impaired SRI in diabetic patients was found (β-coefficient = 0.215; 95% CI: -0.040;0.471).
Conclusion: MetS was associated with worse left ventricular diastolic function, assessed by SRI. Our findings highlight the significant changes observed in patients with impaired glucose homeostasis, even without the presence of T2DM.
1349
Global Longitudinal Strain Predicts Chronic Myocardial Infarction in Patients with Normal Ejection Fraction
Abstract
Objective: Chronic myocardial infarction (MI) detected by late gadolinium enhancement (LGE) is associated with significant mortality and morbidity [1]. Limited subendocardial infarction may not lead to reduction in ejection fraction (EF) and a regional wall motion abnormality (RWMA) may not be evident.
Global longitudinal strain (GLS) is impaired independently of EF in a number of conditions, enabling early detection of disease. Strain imaging predicts final infarct size in MI and is superior to LVEF in predicting morbidity and mortality [2]. We hypothesised that subjects with chronic MI but normal EF would have impaired GLS compared to healthy volunteers.
Methods: Twenty patients with chronic MI (defined as subendocardial hyperenhancement on LGE) and normal LVEF and 20 healthy volunteers underwent CMR at either 1.5T or 3.0T (Philips Achieva TX). Standard bSSFP cine images were used to calculate LV dimensions and GLS by feature tracking (CVI 42, Circle Cardiovascular Imaging Calgary, Canada). LGE imaging was performed in all patients (0.2mmol/kg Gadolinium DTPA).
Results: Patients were matched for age (59.78 ± 12 vs 59.59 ± 5.4 p = 0.95) and for ejection fraction (see figure A) (60.4 ± 3.8 vs 62.2 ± 3.5 p= 0.11). Visual evidence of RWMA was present in 13/20 (65%) of chronic MI patients and 0/20 healthy volunteers. GLS (see figure A) was significantly lower in patients with chronic MI than in those without (-16.07 ± 3.9 vs -19.79 ± 2.3 p= 0.001).
Conclusion: GLS is impaired in patients with chronic MI but normal LVEF. GLS identifies abnormalities in LV systolic contraction not apparent with LVEF alone. It may reveal chronic MI in patients with contraindications to gadolinium-based contrast or prognostication of this subset of chronic MI patients. GLS could be used to detect chronic MI by alternative imaging modalities.
References:
1. EB S, JJ C, Sigurdsson S, et al. Prevalence and prognosis of unrecognized myocardial infarction determined by cardiac magnetic resonance in older adults. JAMA 2012;308:890–6.http://dx.doi.org/10.1001/2012.jama.11089
2. Antoni ML, Mollema S a., Delgado V, et al. Prognostic importance of strain and strain rate after acute myocardial infarction. Eur Heart J 2010;31:1640–7. doi:10.1093/eurheartj/ehq105
1352
Optimal Dose Of Dobutamine During Low-Dose Dobutamine Stress Echocardiography In Correctly Identify Viable Segments On Cardiovascular Magnetic Resonance
Abstract
Background: Partial thickness late gadolinium enhancement (LGE) on cardiovascular magnetic resonance (CMR) and biphasic response on low-dose dobutamine stress echocardiography (DSE) can predict viable myocardium. However little is known about the optimal dose of dobutamine during low dose DSE at which biphasic response first appear to correctly identify viable segments on CMR.
Objective: This study aims to look at the ability of different doses of dobutamine during low-dose DSE to correctly identify viable segments on CMR by late gadolinium enhancement.
Method: Patients with chronic left ventricular dysfunction scheduled for elective revascularization were prospectively selected. They underwent DSE using incremental doses of 2.5mcg/kg, 5mcg/kg, 7.5mcg/kg, 10mcg/kg, 15mcg/kg and 20mcg/kg. Blood pressure (BP), ECG and heart rate were continuously monitored. This was followed by late gadolinium study using standard viability protocol on 1.5T CMR. Number of thickening and non-thickening segments visualized at each stage of DSE were compared to the number of segments with LGE on CMR based on 16-segment model. Data was analyzed using SPSS.
Result: 80 myocardial segments were analyzed. Number of thickening segments were highest at 7.5mcg/kg of dobutamine. After 7.5mcg/kg of dobutamine, there was a negative association between systolic BP, diastolic BP and heart rate with increasing dobutamine doses. In CMR, 44 segments had less than 50% thickness subendocardial LGE and 41 segments had more than 50% thickness of subendocardial LGE. At baseline, the mean sensitivity of DSE to correctly identify number of viable segments on CMR was 25% with a specificity of 79%. With incremental dose of dobutamine, the sensitivity and specificity increased, reaching its highest at 7.5mcg/kg of dobutamine (sensitivity and specificity 80%). At 10mcg/kg of dobutamine onwards, the sensitivity and specificity dropped to lowest level of 52% and 45% respectively at 20mcg/kg of dobutamine.
Conclusion: At 7.5mcg/kg of dobutamine, biphasic response is best demonstrated and has the highest sensitivity and specificity to correctly identify viable segments on CMR. Higher dobutamine doses may precipitate haemodynamic compromise in patients with critical CAD.
1368
Absolute wall thickening and left ventricular ejection fraction–a unifying theory of myocardial contraction and heart failure?
Abstract
Introduction: No single well–established hypothesis for the mechanisms of heart failure exists. Left ventricular ejection fraction (LVEF) is widely used to dichotomise heart failure into reduced and preserved LVEF cohorts, with the latter often believed to be due to diastolic dysfunction. Mathematical modeling has demonstrated that LVEF is determined by myocardial shortening and end–diastolic wall thickness (EDWT). We hypothesized that a common myocardial mechanism occurs in heart failure regardless of cause and we aimed to quantify the relationship of absolute wall thickening (AWT) on LVEF across a range of global cardiac pathologies with varying degrees of EDWT using cardiac magnetic resonance (CMR).
Methods: 183 subjects underwent CMR at 1.5T, including 53 with idiopathic dilated cardiomyopathy (DCM), 36 with amyloid cardiomyopathy, 55 with hypertension (HTN) and 39 controls. Ventricular volumes and LVEF were calculated. EDWT and end–systole wall thickness (ESWT) were measured in basal and mid myocardial segments from long axis cines. AWT was defined as (ESWT–EDWT). Relative AWT (rAWT) was (AWT/ED LV internal diameter). Longitudinal strain was estimated from modified 6–point mean mitral annular plane systolic excursion. Circumferential strain was estimated using an established equation.
Results: Please see Table 1. The study included subjects with normal (controls: 64 ± 7% and HTN: 66 ± 8%), moderately reduced (Amyloid: 49 ± 16%) and severely reduced (DCM: 30 ± 11%) LVEF. There were significant differences in indexed EDV (controls: 77 ± 18ml/m2, DCM: 133 ± 27ml/m2, HTN: 85 ± 16ml/m2 and amyloid: 86 ± 34ml/m2, P < 0.0001) and EDWT (controls: 8 ± 1mm, DCM: 8 ± 1mm, HTN: 11 ± 3mm and amyloid: 14 ± 3mm, P < 0.0001) between subgroups. There was weak positive correlation between EDWT and LVEF (R = 0.185, P < 0.05)(Figure 1A). Reduced longitudinal (R = –0.457, P < 0.0001) and circumferential strain (R = –0.710, P < 0.0001) were associated with significantly reduced AWT. AWT was strongly related to LVEF (R = 0.812, P < 0.0001). Changes in rAWT accounted for 77% of LVEF variability across the cohort (R2 = 0.766)(Figure 1B).
Conclusion: We demonstrate, for the first time, that LVEF is determined by changes in rAWT across a range of global LV diseases of varying EDWT and EDV. AWT depends on EDWT and the degree of longitudinal and circumferential strain. Our findings offer a potentially unifying mechanism for changes in LVEF and refine our understating of heart failure with preserved and reduced LVEF.
1376
Transient St Elevation in Acs Like Myocarditis
Abstract
Background: Chest pain in acute myocarditis is common and generally associated with some ST elevation and rise of myocardial damage markers. In some patients further chest pain recurs hours or days after initial symptom, sometimes but not invariably associated with pericarditis; some previous reports suggest that coronary vasospasm may be a clinically relevant occurence in acute myocarditis cases.
Methods: We prospectively collected pertinent data in all ACS like myocarditis patients that had a CMR scan performed in between 2 weeks after symptom onset. EKG was continuously monitored for at least the first 72 hours post admission.
Results: 25 patients, mean age 46 years, were collected. Among these, 2 had recurrence of chest pain during the first 48 hours, with concomitant transient ST elevation > 2mm, both persisting for a few minutes, resembling variant angina. Coronary angiography (in the absence of chest pain) did not show any luminal defect nor irregularities of vessel walls; myocardial markers were rised and CMR showed non coronary late Gadolinium enhancement (LGE) pattern with mid wall-epicardial localisation; overall left ventricular systolic function was normal. T2 sequences and LGE failed to show any pericardial inflammation.
Our study put together with previous data suggests coronary vasospasm as a mechanism of recurrent chest pain in patients with myocarditis.
Conclusions: coronary vasospasm could be a not uncommon, clinically relevant event in acute myocarditis patients with ACS like pattern.
1379
Patients after Fontan with a “total cavopulmonary connection” Fontan modification develop more collateral flow compared to “old-fashioned” Fontan modifications
Abstract
Objectives: Fontan palliation has evolved over time. Among many modifications creation of “total cavopulmonary connection (TCPC) ” Fontan such as intracardiac tunnel or extracardiac conduit has been proposed as a superior alternative to “old-fashioned” Fontan modifications such as RA-PA, or RA-RV tunneling. Collateral flow is a well-known residuum in patients with Fontan. However, no studies have examined collateral flow between different types of Fontan modifications. So the aim of this study was to compare collateral flow of patients with a TCPC to collateral flow of patients with “old-fashioned” Fontan modifications.
Methods: We conducted a retrospective study on patients with any type of Fontan, who underwent a routine clinical CMR at our institution over a 6-year period with collateral flow quantification. The entire cohort was divided into two groups: 32 patients with TCPC, (median age 15(2-60)yrs.); 7 patients with “old-fashioned” Fontan modifications, (median age 31(21- 35)yrs.). Patients with a fenestration were excluded. We determined collateral flow by using the commonly used equation: flow volume aorta ascendens–(flow volume superior vena cava + flow volume inferior vena cava)
Results: Collateral flow was larger in patients with a TCPC compared to patients with “old-fashioned” Fontan modifications [median 0.6((-0.1)-1.7) vs. 0.2((-0.04)-0.5) L/min/m2, p < 0.006]. Additionally, cardiac index measured as the flow volume in the aorta ascendens was also larger in patients with a TCPC compared to patients with “old-fashioned” Fontan modifications [median 3.2(2.2-4.9) vs. (2.6(1.9-3.3) L/min/m2, p < 0.01].
Conclusions: Patients years after any type of Fontan modifications have substantial collateral flow. However, patients with a TCPC have significantly more collateral flow compared to patients with “old-fashioned” Fontan modifications. This leads to a significantly higher work load of the heart as demonstrated by the larger cardiac index of patients with a TCPC compared to patients with “old-fashioned” Fontan modifications. We suggest that the larger collateral flow in TCPC may be due to different flow characteristics in the superior vena cava compared to “old-fashioned” Fontan modifications.
1387
A MRI–derived 3D patient specific model for fibrosis quantification in atrial fibrillation
Abstract
Introduction: Atrial fibrillation (AF) is the most common type of arrhythmia and the mechanisms that sustain it are not yet clearly identified. Catheter ablation is a promising therapy for AF. However, to achieve durable restoration of sinus rhythm, multiple procedures may be required. Early studies have suggested both extensive atrial tissue fibrosis and association between scar gaps and pulmonary vein (PV) reconnection sites as possible causes of the poor outcomes of the AF catheter ablation. In this study, in order to assist the electrophysiologist in patient selection and ablation procedure planning, we developed a 3D patient–specific left atrium (LA) model integrating anatomical and structural information derived from magnetic resonance angiography (MRA) and delayed–enhanced MR imaging (DE–MRI).
Materials and Methods: Thirty–five patients with paroxysmal AF were enrolled in the study and MRA and DE–MRI images were acquired. A patient–specific anatomical model was derived by MRA data, applying an edge–based level set approach guided by a phase–based edge detector (figure A, upper panels). A multimodality affine registration based on mutual information was then applied to register MRA into the spatial domain of DE–MRI (figure A, bottom panels). Once affine registration parameters were obtained, the corresponding intensity gray level information derived from the DE–MRI was overlapped on the registered 3D surface LA model, allowing the 3D visualization of LA fibrosis location and extent (figure B). In order to obtain a qualitative validation, the 3D LA models were compared with voltage maps reconstructed during the ablation procedures.
Results: The 3D patient specific model obtained through MRA segmentation and registration of DE–MRI data was feasible in all patients. Time required for the analysis was about 30 min for each patient. An example of the qualitative comparison between high enhanced regions in the 3D LA model with fibrosis information and the low voltage areas in the electroanatomical map is shown in figure C.
Conclusion: Preliminary qualitative validation of the 3D LA model including structural information seems a promising tool for a correct fibrosis localization and quantification. Next steps include assessment of the proposed tool to quantify scar location and extent for patient selection and catheter ablation planning.
1391
Scar burden and survival in patients with ischemic cardiomyopathy and poor LV ejection fraction
Abstract
Background: It is unclear whether in patients with ischemic cardiomyopathy and low EF (EF < 40%) late revascularization might be beneficial. We investigated the prognosis in a cohort of patients with ischemic cardiomyopathy and EF < 40% following late revascularization (LR) compared to a cohort of patients receiving optimal medical therapy alone (MED). We also investigated whether scar burden associates with prognosis in these 2 groups.
Methods: Patients with ischemic cardiomyopathy and low EF, defined as <40%, who underwent CMR in a single institution were prospectively followed up since 2003. 25 consecutive patients who underwent LR and 25 consecutive patients who had MED are reported here. Patient demographics, CMR characteristics and scar burden calculated using the 5SD and FWHM methods (CMR 42, Circle CI) were analyzed by blinded observers.
Results: Patient demographics and CMR results including scar burden are shown in table 1.
There was no difference in survival between LR and MED (p = 0.62) as shown in the Kaplan Meier survival curve in figure 2. Scar burden did not associate with prognosis in either the LR (FWHM p = 0.983; 5SD p = 0.572) or the MED group (FWHM p = 0.952, 5SD p = 0.896). In the LR no other variables were significant. However in the MED NYHA Class was significantly associated with prognosis with higher NYHA conferring worse survival (HR = 3.2, p = 0.011, 95% CI 1.3-7.9, per 1 category increase in NYHA)
Conclusion: LR did not offer a prognostic benefit in patients with ischemic cardiomyopathy and low EF. Scar burden did not associate with prognosis in either LR or MED group. In the MED group, patients with worse NYHA had worse prognosis, but this was not seen in the LR group.
Figure 1. Kaplan Meier survival curve of patients who underwent Late revascularization (Late Revasc) or medical therapy alone (No Revasc), indicating that there is no significant survival benefit in survival between the two groups.
1392
Relation of inflammatory markers with myocardial and microvascular injury in patients with reperfused ST- elevation myocardial infarction
Abstract
Background: In patients with acute ST-elevation myocardial infarction (STEMI), elevated concentrations of inflammatory markers are correlated with worse clinical outcome. The aim of this study was to comprehensively investigate the relationship of circulating markers of inflammation with myocardial and microvascular damage after STEMI.
Methods: In 111 consecutive STEMI patients, blood samples were obtained on admission and from day 1 to day 4 after primary percutaneous coronary intervention (PPCI) and analyzed for high-sensitivity C-reactive protein (hs-CRP), white blood cell count (WBCc) and fibrinogen. Cardiac magnetic resonance imaging was performed within the first week and 4 months after PPCI for assessment of myocardial function and damage.
Results: Peak concentrations of hs-CRP (20.5[9.6-44.4]mg/L), WBCc (12.4[10.5-15.3]G/L) and fibrinogen (3640[3150-4550]mg/L) showed significant correlations with both infarct size (r = 0.31 to 0.41; p < 0.01) and left ventricular (LV) ejection fraction (r = -0.29 to -0.39; p< 0.01) assessed in the acute as well as chronic stage following STEMI. Furthermore, peak concentrations of these inflammatory markers were significantly higher in patients with microvascular obstruction (MVO) compared to patients without MVO (p ≤ 0.01). C-statistics revealed that the prognostic values of all three biomarkers for the prediction of large chronic infarct size (>8% of LV myocardial mass) were moderate without significant differences (area under the curve (AUC): hs-CRP = 0.73 (95%CI 0.63-0.82), WBCc = 0.67 (95%CI 0.56-0.78) and fibrinogen = 0.70 (95%CI 0.59-0.80); all p > 0.12). Combination of inflammatory markers did not significantly increase the AUC (p > 0.05).
Conclusion: In STEMI patients treated with PPCI, increased levels of hs-CRP, WBCc and fibrinogen are associated with decreased LV function and more pronounced myocardial damage at baseline and 4 months after infarction.
1406
Equivalence of segmented conventional and fast single-shot late gadolinium enhancement (LGE) techniques for
Abstract
Background: Segmented single-slice/single-breath-hold 2D phase-sensitive inversion recovery (2D-PSIR) sequences are the gold standard for evaluation of myocardial fibrosis. Aim of this study was to assess the accuracy of single-shot LGE sequences to detect and quantify myocardial fibrosis in patients with ischemic and non-ischemic etiology of LGE.
Methods: Patients with chronic myocardial infarction/coronary artery disease (CAD; n = 121), myocarditis (n = 35) and hypertrophic cardiomyopathy (HCM) (n = 20) were prospectively enrolled. Image studies were performed on 1.5T Siemens Magnetom Avanto fit. After administration of contrast agent (gadoteridol 0.2mmol/kg for CAD and HCM, gadopentetate dimeglumin 0.2mmol/kg for myocarditis), LGE images were acquired ECG-gated in short axis slices (slice thickness 7mm, gap 0mm) using 4 different LGE sequences (Figure 1): (1) conventional segmented 2D phasesensitive inversion recovery in single-slice/single-breath-hold technique (TR 744ms, TE 5.17ms, Flip Angle 30°), (2) single-breath-hold 3D-IR sequence (TR 924ms, TE 1.06ms, Flip Angle 50°), (3) 3D-SSFP sequence in breath-hold and (4) non-breath-hold technique (TR 700ms, TE 1.05ms, Flip Angle 65°). For all techniques, inversion time was individually adjusted to null the remote myocardium. LGE was quantitatively assessed using a semi-automated threshold method: positive LGE was defined as signal intensity 6 standard deviations (SD) above signal intensity of remote myocardium for CAD and 3 SD for myocarditis/HCM. Differences in size of LGE-positive areas (in gram) were analyzed using Bland-Altman-Analysis.
Results: There was no significant difference in size of LGE-positive areas between the segmented 2D-PSIR and 3D-IR, 3DSSFP LGE sequence (Table 1A), independent of the underlying etiology. Comparing breath-holding to free-breathing technique in the 3D-SSFP LGE sequence, scar size in patients with CAD is smaller in free-breathing sequence (mean difference 0.44g, p = 0.01), whereas no differences were found in HCM (Table 1B).
Acquisition times were significantly shorter for 3D-IR (23.3 s ± 7.4 s) and 3D-SSFP (22.1 s ± 7.0 s) as compared to 2D-PSIR (365.2 s ± 94.9 s).
Conclusions: Fast 3D-SSFP, 3D-IR and conventional segmented 2D-PSIR sequences are equivalent techniques for the assessment of myocardial fibrosis, independent of an ischemic or non-ischemic etiology. In CAD, free-breathing 3DSSFP technique results in slightly smaller infarction size compared to acquisition in single breath-hold.
1410
Cardiac Mri Appearances of Tuberculosis - A Review of Varied Presentations in India
Abstract
Objective: Pictorial representation of different types of presentation of cardiac tuberculosis on cardiac MRI seen in our institute.
Introduction: Cardiovascular tuberculosis is a relatively infrequent condition in European society but is commonly seen in the Indian subcontinent. Classical presentation is with pericardial involvement with some cases showing myocardial involvement. Cardiac MRI is extremely useful in assessing patients with suspected cardiac tuberculosis as echocardiography is often non diagnostic. It can also be used in following up of patients to assess response to therapy.
Methods and Materials: Retrospective review of all the cardiac MR examinations performed in our institute during the period of January 2014 to December 2015. All patients with proven tuberculosis were included and analysed.
Results: In our institute, Cardiac Tuberculosis on CMR was seen in the following major patterns:
1. Mass like presentation in both atria.
2. Pericarditis- Calcific and non calcific
3. Myocarditis
Conclusion: Although cardiac tuberculosis is a rare disease, the recognition of this entity and its varied presentation is important for diagnosis, management and follow up. As this being one of the treatable condition.
1415
Atheroma burden, cardiac remodelling and epicardial fat: A comparison between healthy South Asian and European adults using Whole Body Cardiovascular MR
Abstract
Objectives: South Asians (SAs) have a higher risk of coronary artery disease, congestive cardiac failure and stroke, but paradoxically lower prevalence of peripheral arterial disease (PAD) than Western Europeans (WEs). The aim of this study was thus to evaluate early changes in systemic atherosclerotic burden and cardiac remodelling in healthy South Asians compared with Western Europeans using whole body cardiovascular MRI (WB-CVMR).
Methods: 19 SA and 38 age, gender and BMI matched WE were recruited. All were ≥40 years, free from cardiovascular disease (CVD) and with a 10-year risk of CVD <20% as assessed by the ATPIII risk score. WB-CVMR was performed which comprised a whole body angiogram (WBA) and cardiac magnetic resonance (CMR). These were performed on a 3T MRI (Tim Trio, Seimens, Germany) scanner using whole body coils following dual phase injection of gadoteric acid (Guerbet, France). A standardized atherosclerotic score (SAS) was calculated from the WBA, while indexed left ventricular mass and volumes, and indexed epicardial and thoracic fat volumes were calculated from the CMR.
Results: SAs exhibited a significantly lower iliofemoral atheroma burden (regional SAS 0.0 ± 0.0 vs 1.9 ± 6.9, p = 0.048) and a trend towards lower overall atheroma burden (WB SAS 0.7 ± 0.8 vs 1.8 ± 2.3, p = 0.1). They had significantly lower indexed left ventricular mass (46.9 ± 11.8 vs 56.9 ± 13.4ml/m2, p = 0.008), end diastolic volume (63.9 ± 10.4 vs 75.2 ± 11.4ml/m2, p = 0.001), end systolic volume (20.5 ± 6.1 vs 24.6 ± 6.8ml/m2, p = 0.03) and stroke volume (43.4 ± 6.6 vs 50.6 ± 7.9ml/m2, p = 0.001), but with no significant difference in ejection fraction, mass-volume ratio or global functioning index. These differences persisted after accounting for CVD risk factors. Despite smaller heart volumes the volume of epicardial fat deposition was similar between the two groups (2.6 ± 1.4 vs. 2.7 ± 1.7 cm2/m2, p = 0.8).
Conclusions: South Asians have a lower peripheral atherosclerotic burden and smaller hearts than Western Europeans even in a healthy low-intermadiate risk population. Thus the paradoxical high risk of CVD compared with PVD risk may be due to an adverse cardiac haemodynamic status incurred by the smaller heart rather than atherosclerosis.
1418
Symptomatic Ventricular Arrhythmias: Diagnostic Yield of Cardiac Magnetic Resonance
Abstract
Objectives: To assess the diagnostic yield of cardiac magnetic resonance (CMR) in patients with ventricular arrhythmias.
Methods: From 2014 to 2015, consecutive patients referred to our institution for the management of symptomatic non-sustained ventricular tachycardia (NSVT), sustained ventricular tachycardia (SVT), ventricular fibrillation (VF) or aborted sudden cardiac death (SCD) were prospectively included. A complete diagnostic work-up was performed including clinical symptoms, personal and family history, ECG and/or Holter, echocardiography, and at the discretion of the referring cardiologist exercise testing and coronary angiography. CMR was performed in all patients, including cine SSFP and late gadolinium enhancement (LGE) imaging (breath-holded 3D IR FLASH). The diagnostic yield of CMR was assessed by comparing diagnoses obtained with vs. without CMR information, as defined by 2 observers in consensus, non-CMR diagnoses being blinded from CMR results.
Results: 217 consecutive patients (age 54 ± 17 years, 55 women) were included (69[32%] NSVT, 123[57%] SVT, 25[12%] VF or SCD). Dominant clinical symptoms were palpitation in 75(35%), lightheadedness in 39(18%), syncope in 35(16%), chest pain in 25(12%), SCD in 25(12%), dyspnea in 15(7%), and cardiogenic shock in 3(2%). The diagnostic work-up comprised exercise testing in 180(83%), and coronary angiography in 137(63%). Diagnoses without CMR information were no structural heart disease (SHD) in 123(57%), ischemic in 57(27%), non-ischemic dilated cardiomyopathy (DCM) in 15(7%), post-inflammatory in 3(1%), arrhythmogenic right ventricular cardiomyopathy (ARVC) in 9(4%), hypertrophic cardiomyopathy in 4(2%), and other in 6(3%). In the whole population CMR led to a modification of diagnosis in 40/217(18%). The diagnostic yield of CMR was higher in patients with no SHD as per pre-CMR assessment (diagnostic change in 32/123[26%]) than in patients with SHD (diagnostic change in 8/94[9%]). Modifications in diagnosis were influenced by LGE imaging in 36/40(90%), and by cine imaging in 4/40(10%). When including CMR information, novel diagnoses included post-inflammatory scar in 20(50% of all diagnostic changes), ischemic in 6(15%), no SHD in 3(8%), ARVC in 3(8%), DCM in 2(5%), and other in 6(15%).
Conclusions: The incremental value of CMR in the diagnostic management of patients with ventricular arrhythmias (NSVT, SVT and SCD) is high, particularly in those with no known SHD as per pre-CMR assessment.
1421
CMR assessment of aortic stiffness in asymptomatic low risk patients with type 2 diabetes mellitus
Abstract
Introduction: Arterial Stiffness (AS) is a non-invasive marker of cardiovascular disease which results in a reduction in aortic distensibility (AD) and an increase in pulse wave velocity (PWV). Previous studies have demonstrated increased AS is associated with adverse outcomes. Advanced age and type 2 diabetes mellitus (T2DM) are associated with increased AS however the influence of other risk factors in T2DM and how they interact is not known. We hypothesised that poor control of diabetes and higher ambulatory blood pressure (ABP) would be associated with increased arterial stiffness.
Methods: We recruited 80 asymptomatic patients with type 2 diabetes from primary care who were not taking any antihypertensive medication and 10 healthy controls. All subjects underwent assessment of risk factors, 24 hour ABP and CMR at 3.0T with Philips Achieva system. CMR protocol included cine aortic imaging planned at the level of the main pulmonary artery and retrospectively gated, velocity encoded phase-contrast cines.
AD was calculated from an average of contours drawn on the ascending and descending aorta. PWV was derived from velocity-time curves and the distance between the ascending and descending aorta using the transit-time method.
Results: AD was lower (2.04 ± 1.02 10-3mmHg-1 vs 3.18 ± 1.32 10-3mmHg-1, P = 0.005) however the difference in PWV was not significant (7.81 ± 2.77 m/s vs 6.97 ± 1.93 m/s, P = 0.38) in T2DM than in controls.
In subjects with T2DM there were significant negative correlations between AD and age (R = -0.64, P < 0.001), duration of diabetes (R = -0.23, P = 0.045) and mean systolic ABP (R = -0.26, P = 0.02). PWV had a significant correlation with increasing age (R = 0.50, P < 0.001) and duration of diabetes (R = 0.31, P = 0.01). There was no significant association between AS and total cholesterol, smoking, hsCRP or the presence of persistent microalbuminuria.
Conclusion: In this cohort of low risk, asymptomatic patients with T2DM there was evidence of increased aortic stiffness compared to controls. In these patients aortic stiffness was influenced primarily by age, duration of T2DM and systolic ABP.
Table showing AD and PWV of 10 healthy controls and 80 patients with T2DM.
. | Control . | T2DM . | P value . |
---|---|---|---|
Ascending aorta AD, 10-3mmHg-1 | 2.95±1.81 | 1.90±1.19 | 0.06 |
Descending aorta AD, 10-3mmHg-1 | 3.42±1.13 | 2.19±1.04 | 0.002 |
Average AD, 10-3mmHg-1 | 3.19±1.32 | 2.04±1.02 | 0.005 |
Aortic PWV, m/s | 6.98±1.93 | 7.81±2.77 | 0.38 |
. | Control . | T2DM . | P value . |
---|---|---|---|
Ascending aorta AD, 10-3mmHg-1 | 2.95±1.81 | 1.90±1.19 | 0.06 |
Descending aorta AD, 10-3mmHg-1 | 3.42±1.13 | 2.19±1.04 | 0.002 |
Average AD, 10-3mmHg-1 | 3.19±1.32 | 2.04±1.02 | 0.005 |
Aortic PWV, m/s | 6.98±1.93 | 7.81±2.77 | 0.38 |
1436
Shock index as a predictor of myocardial damage and clinical outcome in ST-elevation myocardial infarction
Abstract
Background: Data on the prognostic value of the shock index in patients with ST-elevation myocardial infarction (STEMI) are scarce. Furthermore, the relationship of shock index with myocardial damage is unknown. Aim of this study was therefore to evaluate the association of the shock index with markers of myocardial damage and clinical outcome in patients with reperfused STEMI.
Methods and Results: This multicenter study analyzed 791 patients. Patients were categorized in 2 groups according to admission shock index (optimized cut-off = 0.62). Infarct severity was determined by cardiac magnetic resonance imaging. Major adverse cardiac events (MACE) were defined as a composite of death, reinfarction and new congestive heart failure within 12 months. Patients with elevated admission shock index (n = 321 [40.6%]) had a significantly larger area-at-risk (37.6 [27.8-50.4]% of left ventricular volume [LV] vs. 34.3 [24.5-46.0]%LV, p = 0.02), larger infarct size (19.5 [10.7-28.0]%LV vs. 14.9 [7.7-22.3]%LV, p < 0.001), lower myocardial salvage index (46.2 [27.9-64.5] vs. 53.5 [36.5-75.2], p < 0.001), and a larger extent of microvascular obstruction (0.3 [0.0-2.2]%LV vs. 0.0 [0.0-1.4]%LV, p = 0.01). An elevated shock index was associated with reduced MACE-free survival (p < 0.001). Furthermore, admission shock index was identified as independent predictor of MACE (hazard ratio = 2.92 [1.24-4.22], p < 0.01).
Conclusion: STEMI patients with an elevated admission shock index had more pronounced myocardial and microvascular damage. Moreover, shock index was independently associated with MACE at 12 months after STEMI.
1451
Combined biomarker testing for the prediction of microvascular obstruction after primary percutaneous coronary intervention for acute ST-segment elevation myocardial infarction
Abstract
Background: Microvascular obstruction (MVO) as detected by cardiac magnetic resonance (CMR) imaging indicates microvascular destruction with subsequent adverse clinical outcome after reperfused ST-segment elevation myocardial infarction (STEMI). The predictive value of different biomarkers for the occurrence of MVO is insufficiently studied. This study compared the prognostic value of admission and peak concentrations of routinely available biomarkers for the detection of MVO after reperfused STEMI.
Methods: One hundred and twenty-eight STEMI patients undergoing primary percutaneous coronary intervention (PPCI) were enrolled in this single-center, prospective, observational study. CMR was performed within the first week after infarction to assess infarct characteristics, including MVO. Admission and peak concentrations of high-sensitivity cardiac troponin T (hs-cTnT), creatine kinase (CK), N-terminal pro-B-type natriuretic peptide (NT-proBNP), high-sensitivity C-reactive protein (hs-CRP), lactate dehydrogenase (LDH), aspartate transaminase (AST) and alanine transaminase (ALT) were measured.
Results: MVO was detected in 69 patients (54%). Peak concentrations of hs-cTnT, CK, hs-CRP, LDH, AST and ALT showed similar prognostic value for the prediction of MVO (area under the curve (AUC) = 0.77, 0.77, 0.68, 0.79, 0.78 and 0.73, all p > 0.050), whereas the prognostic utility of peak NT-proBNP was lower (AUC = 0.64). Combination of these biomarkers did not result in higher predictive value as compared to hs-cTnT alone (p = 0.349).
Conclusion: hs-cTnT, CK, hs-CRP, LDH, AST and ALT peak concentrations provided similar prognostic value for the prediction of MVO, whereas the prognostic utility of NT-proBNP was lower. Combining these biomarkers could not further improve prognostic utility compared to hs-cTnT alone.
1452
A novel oscillometric technique compared with cardiac magnetic resonance for the assessment of aortic pulse wave velocity in ST-segment elevation myocardial infarction
Abstract
Background: Pulse wave velocity (PWV) is the proposed gold-standard for the assessment of aortic elastic properties. A recently developed oscillometric device, which allows for non-invasive and cost-effective measurement of aortic PWV, showed a moderate to good agreement with cardiac magnetic resonance imaging (CMR) in healthy volunteers. Now we compared the two methods in patients presenting with ST-segment elevation myocardial infarction (STEMI).
Methods: We assessed aortic PWV in 60 mechanically reperfused STEMI patients using two different methods. The oscillometric method (PWVOSC) is based on mathematical transformation of brachial pressure waveforms, oscillometrically determined using a common cuff (Mobil-O-Graph, I.E.M. Stolberg, Germany). Phase-contrast CMR imaging (1.5 Tesla scanner, Siemens, Erlangen, Germany) at the level of the ascending and abdominal aorta was performed to determine PWVCMR with the use of the transit time method.
Results: The mean age of the study population was 57 ± 11 years; 11 (19%) were female. Median PWVOSC was 7.4 m/s (IQR 6.8 - 8.9 m/s) and median PWVCMR was 6.3 m/s (IQR 5.7 - 8.2 m/s) (p < 0.001). A strong correlation was detected between both methods (r = 0.724, p < 0.001). Bland-Altman analysis revealed a bias of 0.62 m/s (upper and lower limit of agreement: 3.84 m/s and -2.61 m/s). The coefficient of variation between both methods was 21%.
Conclusion: In mechanically reperfused STEMI patients, aortic PWV assessed non-invasively by transformation of brachial pressure waveforms showed an acceptable agreement with the CMR-derived transit time method.
1456
Aorto-pulmonary collaterals evaluated by CMR is associated to reduced “effective” cardiac index late after Fontan palliation
Abstract
Objective: Aorto-pulmonary collaterals (APCs) are frequent in patients with uni-ventricular heart (UVH). Their clinical significance remains controversial and the mechanism that leads to their formation are poor known. Cardiac output (CO) in a Fontan circulation is decreased compare to normal population. Quantitative assessment of APCs blood flow and evaluation of cardiac output (CO) and cardiac (CI) using cardiac magnetic resonance (CMR) has been already validated.
Aim: to evaluate the factors associated to APCs flow (QAPCs) assessed by CMR in patients late after Fontan palliation and their impact on “effective” CI and on clinical outcome.
Methods: Form our CMR database we identified all patients with Fontan intervention who underwent CMR. Patients with a complete set of flow measurements allowing calculation of APC flow were included. QAPCs was calculated as (left pulmonary veins flow + right pulmonary veins flow) - (right pulmonary artery flow + left pulmonary artery flow), assessed by means of through-plane phase-contrast.Values were normalized to body surface area. Effective CI was calculated as (QAo flow- QAPCs)/BSA. Medical and surgical history and clinical status were recorded.
Results: Sixty-five patients post Fontan palliation were included in the study (36 M, age: 19 ± 10). Follow-up from Fontan palliation at CMR study was 12 ± 7 years, range: 1-31 years. Median QAPCs was 734 ml/min/m2 (range 106-3000), corresponding to a median 22% (4-73) of systemic blood flow. QAPCs didn't correlate neither with age at CMR study and at Fontan palliation neither with rest O2 sat, ventricular volumes, ejection fraction and mass. QAPCs was not associated to systemic ventricular type (right, left or 2 complex ventricles), atrial arrhythmias, atrio-ventricular valve regurgitation. Meanwhile QAPCs correlates with indexed aortic flow, r= 0.6, P< 0.001 but inversely correlates with effective CI, r= -0.39, P = 0.001. QAPCs correlate inversely also with left pulmonary artery area (r= -0.37 P= 0. 07).
Conclusions: New CMR techniques allow reliable quantification of QAPCs. After Fontan palliation QAPCs has a wide range and is associated to reduced left pulmonary artery area. From our data increase of APC flow is associated to further reduction of effective systemic blood flow. Further studies are needed to better define the clinical impact of APC and the indication of its embolization.
1458
Evaluation of pulmonary transit time and Pulmonary Blood Volume with first-pass perfusion CMR imaging in adult with repaired Congenital Heart disease
Abstract
Background: Management of Adults With Repaired Congenital Heart Disease is still challenging. Heart failure secondary to residual anatomical sequels as well arrhythmic events are not rare in this population. Pulmonary transit time (PTT) is significantly prolonged in patients with acquired heart disease and heart failure. Cardiac magnetic resonance (CMR) has emerged as an accurate tool to quantify PTT of intravenous contrast agents and Pulmonary blood volume (PBV). The aim of this study was to assess PTT and PBV in adults with Repaired cono-truncal heart defect, to determine if they reflect conventional indexes of ventricular dysfunction and to assess their association to adverse cardiac event.
Methods: twenty-seven patients with repaired cono-truncal anomalies, mainly Tetralogy of Fallot (n = 19), 16 men (31 ± 11 years) and 13 age-and sex-matched healthy controls underwent CMR examination comprehensive of first-pass perfusion. PBV was calculated as the product of stroke volume and number of cardiac cycles for an intravenous bolus of gadolinium contrast to pass through the pulmonary circulation, as determined by cardiac-gated first-pass perfusion imaging. Twenty-one patients underwent also a cardio-pulmonary test (CPT). PTT and indexed PBV (PBVi) correlated to conventional indexes of ventricular function. Cardiac event before the study or during follow up as well as subsequent re-operation were reported.
Results: PTT correlates inversely with LVEF (r= -0.46,P= 0.01), with LVESV (R= 0.48; P = 0.01) and with VECO2 slope at CPT (r= 0.5, P = 0.02); PBVI correlates with LVESV (R= 0.38; P = 0.04) and with VECO2 slope at CPT (r= 0.49, P = 0.02). Using a cut-off of PTT >10.3 sec, corresponding to 2SD above the mean of healthy controls, six patients above it have lower LVEF in comparison with those with PTT≤ 10.3 (median 48%, range 37-57% vs median 57%, range 37-73%) and experienced more cardiac events before CMR (p = 0.04) and during follow-up (p = 0.03). The occurrence of cardiac events and/or reinterventions during the follow-up was associated to higher PBVi (342 ± 104 sec vs 148 ± 144, P = 0.01).
Conclusions: pulmonary intravascular blood pool (PBVi) and the time it flows through (PTT), can be assessed non invasively by means of contrast-enhanced MR. Both PTT and PBVi were associated to cardiac event at study and at follow-up in our group of adult with repaired cono-truncal anomalies. Further studies in a larger population are warranted to assess its prognostic value.
1459
Prognostic value of the cardiac magnetic resonance as a predictor of improvement in ventricular function after TakoTsubo syndrome
Abstract
Introduction: Tako-Tsubo syndrome (TTS) is a reversible cardiomyopathy defined by criteria established by the Mayo Clinic and by the Japanese Society, which are echocardiographic, electrocardiographic and analytical criteria, having excluded coronary disease, pheochromocytoma or myocarditis. The resolution of the wall motion abnormalities could take days or even months. Cardiac magnetic resonance imaging (CMR) can help to predict the degree of recovery of ventricular function on the basis of the findings of the baseline study on the income.
Methods: The objective was to analyze the value of the CMR as a predictor of the degree of improvement in ventricular function (LVEF) in the follow-up. All patients (P) were included from January 2009 to April 2015 with the diagnostic code 429.83 (TS, ICD-9 - MC) who held a CMR in income.
Results: We included 25 P that met the diagnostic criteria of STK of the Mayo Clinic and who had performed a CMR during hospitalization. The average age was 69.6 years, 92% were women and 56% presented stress as a trigger. The majority of the cases were presented with killip class I (48%). 12% had a history of ischemic heart disease. Of these 25 P, 52% of them presented a normal CMR (Group 1). The remaining 48% (Group 2) presented some alteration in tissue characterization (8% edema, 28% enhancement transmural/ subendocardial, 12% subepicardial). The average initial LVEF in Group 1 was 34% and in Group 2, 45%, with statistically significant difference (p = 0.005). The average recovery of ventricular function (assessed with echocardiogram) in Group 1 was 25 points, while in Group 2 it was 13 points, this difference being statistically significant (p = 0.008). With respect to enzyme elevation, Tpn T US or CK, there were no significant differences (p = 0.621, and p = 0.269 respectively).
Conclusions: In this serie of 25 P with Tako-Tsubo Cardiomyopathy according to criteria of Mayo Clinic, the CMR allows to classify them into 2 groups, those with alterations in tissue characterization and those without, being the last group the one that present a more decreased income LVEF, although with a greater degree of recovery. These findings would reclassify up to 40% of patients (10 patients) in other diagnoses such as myocarditis or ischemic disease in the presence of coronary arteries without significant lesions, and must be considered as a new diagnostic criterion in future classifications.
1462
Diagnostic performance of ECG detection of left atrial enlargement in patients with arterial hypertension relative to the cardiac magnetic resonance gold-standard: impact of obesity
Abstract
Introduction: Guidelines on the management of arterial hypertension suggest that an electrocardiogram (ECG) be routinely performed. ECG may demonstrate evidence of left atrial enlargement (LAE), which has adverse prognostic implications. We sought to determine the accuracy of 5 ECG criteria of LAE in a hypertensive cohort relative to cardiac magnetic resonance (CMR) and to investigate the confounding effect of obesity.
Methods: Consecutive referrals for CMR (1.5T) from a tertiary hypertension clinic were reviewed. Patients with any concomitant cardiac pathology were excluded. ECGs were assessed, blinded to the CMR data, for: 1) P wave >110ms, 2) P mitrale (notched P wave with inter-peak duration >40ms), 3) P wave axis <30°, 4) Area of negative P terminal force in lead V1 (NPTF-V1) > 40ms·mm and 5) Positive P terminal force in aVL (PPTF-aVL) >0.5mm. Maximal LA volume index (LAVI), excluding LA appendage and pulmonary vein confluence, was measured by the biplane area-length method (previously validated against LA short axis cine stack). LAE was defined as ≥55ml/m2 (>2 standard deviations above published references values). Sensitivity, specificity, positive predictive value, negative predictive value and accuracy were calculated. Area under the receiver operator curve analysis was performed.
Results: 130 patients were included (age: 51.4 ± 15.1 years, 47% male, 51% obese, systolic blood pressure: 171 ± 29mmHg, diastolic blood pressure: 97 ± 15mmHg). The prevalence of LAE by CMR was 26% and by ECG varied from 1% (P mitrale) to 27% (P axis < 30o), and was 46% when ≥1 ECG LAE criteria were present (Table 1). There was no significant difference in mean LAVI when ≥1 ECG LAE criterion was present compared to when no ECG LAE criteria were present (47 ± 15 vs 50 ± 15 ml/m2, P = 0.235). All the individual ECG LAE criteria were more specific than sensitive (Table 1), with specificities ranging from 70% (P axis <30o) to 99% (P mitrale). Obesity attenuated the specificity of most of the individual ECG LAE criteria (Table 2). Obesity correlated with significant lower specificity (48% vs 65%, P < 0.05) and a trend towards lower sensitivity (59% vs 43%, P = 0.119) when ≥1 ECG criteria of LAE were present.
Conclusion: Individual ECG criteria of LAE in hypertension are specific, but not sensitive, for identifying anatomical LAE, relative to CMR. LAE in hypertension should not be excluded on the basis of the ECG, particularly in obese subjects.
1463
Utility of cardiac magnetic resonance imaging for diagnosis of cardiac sarcoidosis and prediction of therapeutic effects in patients with complete heart block and implanted magnetic resonance-conditional pacemaker: A multicenter study
Abstract
Aims: Complete heart block (CHB) is a serious consequence of cardiac sarcoidosis (CS), requiring early diagnosis and effective anti-inflammatory treatment. This study aimed to survey the prevalence of CS in CHB patients implanted with magnetic resonance-conditional pacemaker (MRCP) and to evaluate whether cardiac MR (CMR) findings could predict responders to corticosteroid therapy.
Methods and Results: Fifty CHB patients implanted with MRCP were enrolled in this study. CMR was performed to detect delayed enhancement (DE) as a sign of myocardial fibrosis and increased T2-weighted signal (T2-WS) as a sign of myocardial edema. We defined CS based on both histological diagnosis of extra-cardiac sarcoidosis and positive findings on CMR. Eight (16%) of the 50 patients diagnosed with CS.All of the CS patients were female, and the incidence of CS in females aged 40-69 years was 63% (5 of 8 patients). In the CS group, angiotensin-converting enzyme levels were significantly higher (18.1 ± 4.1 U/L vs. 13.5 ± 4.7 U/L, p = 0.014) and the left ventricular ejection fraction was significantly lower (57 ± 7 % vs. 62 ± 6 %, p = 0.048) in the CS group than in the idiopathic CHB group. Overall, 5 of 8 CS patients recovered from CHB after corticosteroid therapy; before corticosteroid therapy, all these patients exhibited smaller area of DE and increased T2-WS, and showed normal wall motion and thickness in the interventricular septum.
Conclusions: CS was diagnosed in 63% of female, middle-aged patients with CHB requiring pacemaker implantation. The extent of cardiac involvement detected by CMR might predict the response to corticosteroid therapy.
1467
Cardiac magnetic resonance late gadolinium enhancement in patients with genetic dilated cardiomyopathy
Abstract
Introduction: Dilated cardiomyopathy (DCM) has an estimated prevalence of 1:2500 in adult population, with genetic etiology explaining 30-50% of “idiopathic” cases. Genetic causality is difficult to identify because of the scarcity of distinctive red flags. In recent years, cardiac magnetic resonance (CMR) has emerged as a valuable imaging modality in this field. However, it's utility in diagnosing genetic DCM remains largely unknown. In this work we aimed to describe CMR findings in genetically characterized DCM patients.
Methods: We included patients with idiopathic and familial DCM, that underwent a comprehensive CMR with a 3-T scanner (Siemens, Erlangen, Germany), as part of the diagnostic work-up. Left ventricular (LV) volumes, ejection fraction (LVEF) and mass were measured using dedicated software (ARGUS Software™, Siemens Healthcare Global). LV late gadolinium enhancement (LGE) presence, pattern and location were assessed; extensive fibrosis was defined as LGE in ≥ 3 LV segments.
Molecular analysis included the search of mutations in LMNA/C, MYH7, MYBPC3, TNNT2, ACTC1, TPM1, CSRP3, TCAP, SGCD, PLN, MYL2, MYL3, TNNI3, TAZ and LBD3 genes. Pathogenicity was assessed by comparisons with mutations previously described, functional tests and segregations studies.
Results: We analyzed 73 patients, 46.6% with familial DCM, 52.8% men, with mean LVEF of 34 ± 11% and LV end-diastolic volume of 128 ± 34mL. We identified 18 genetic variants in 17 distinct patients. Eleven patients presented variants with pathogenicity criteria.
Comparing patients with or without genetic variants, we observed no difference in CMR parameters. Focusing on patients with mutations in MYBPC3, TNNT2 and MYH7 genes, we found only a trend toward an association of MYH7 mutations with LGE (p = 0.057)–with a significant predilection for septum involvement (p = 0.042), and with the presence of non-compaction (p = 0.057). There was no relationship between the remaining CMR variables.
Conclusion: LGE might have some utility in the clinical recognition of patients with genetic DCM, namely those with MYH7 mutations, although additional studies are warranted to confirm these findings. Nevertheless, the exclusion of other causes of LV dysfunction and the use of more recent CMR tools, as interstitial fibrosis assessment, support the continued exploration of this technique in the evaluation of genetic/familial DCM patients.
1471
2.Left ventricular hypertrophy in hypertensive patients–comparison of Cardiac Magnetic Resonance and Echocardiographic analysis of morphological and functional LV-parameters
Abstract
Summary:
Aim: The aim of this research is to present usefulness of Cardiac Magnetic Resonance (CMR) to assess left ventricular hypertrophy in hypertensive patients and comparison of echocardiographic and Cardio MR results.
Material and Methods: We analysed a group consisting of 48 hypertensive patients, with suspected left ventricular hypertrophy. Cardio MR was performed by 1,5 Tesla Scanner.
We assessed left ventricular parameters such as: Ejection Fraction, End-Diastolic Volume and End-Systolic Volume. We also assessed left ventricular diameters as well as left ventricular posterior wall and interventricular septum thickness. We also calculated left ventricular mass (LVM) and left ventricular mass index (LVMI).
Results: In most cases (66,7%) patients had significant hypertrophy of the left ventricle. They also had higher LVM than the normal. We affirmed statistically significant changes of left ventricular's morphological parameters and LVM.
The significant correlation was found between posterior wall diameter and LVM, as well as between interventricular septum and LVM. The significant correlation was found between posterior wall diameter and left ventricular Ejection Fraction and End-Systolic Volume.
To compare Echocardiography analysis with CMR we have to conclude that LV mass and Ejection Fraction measured by CMR is usually less than echocardiographic LV mass and EF, but EDV and ESV is usually more than in Echocardiography.
Conclusion: In conclusion we can say that CMR is an effective method to detect LVH in hypertensive patients. There are significant differences between CMR and Echocardiographic results of LVH parameters.
1472
Is Angiographic Perfusion Score assessed in patients with acute myocardial infarction correlated with Cardiac Magnetic Resonance infarct size and N-terminal pro-brain natriuretic peptide in 6-month follow-up
Abstract
Summary:
Aim: Aim of the study was to analyze the correlation between Angiographic Perfusion Score (APS) and Cardio MR infarct size and left ventricular function parameters in 6 month follow-up.
Material and Methods: APS proposed as a simple, angiographic score linking epicardial and myocardial perfusion parameters before and after percutaneous coronary intervention (PPCI) is a predictor of shortterm outcome in patients with ST-segment elevation myocardial infarction (STEMI) treated with PCI.
In a cohort of 68 patients with STEMI treated with PCI APS was calculated for infarct-related artery based on angiographic parameters and was defined as the sum of the TIMI flow grade and the TMPG before and after PCI (range of points from 0 to 12). Full perfusion was defined as APS ≥ 10.
All patients received trombolytics.
Cardiac magnetic resonance (CMR) parameters and N-terminal pro-brain natriuretic peptide (NT pro-BNP) were assessed at 6 months.
Results: Median APS was 7.5 points. APS ≥ 10 was present in 42% of patients. The significant correlation was found between APS and: CMR infarct size, CMR LV ejection fraction, LV end-diastolic volume index, LV end-systolic volume index, NT pro-BNP. Patients with APS ≥ 10 had significantly lower infarct size, LV volumes, higher EF and lower NT pro-BNP.
Conclusion: APS assessed in patients with STEMI treated with PCI is a good predictor of infarct size and left ventricular function in 6-month follow-up. There was a significant correlation of APS and CMR parameters and NT pro-BNP after 6 months.
1476
Cardiac Magnetic Resonance Patterns of Left Ventricular Diastolic Function In Hypertrophic Cardiomyopathy
Abstract
Purpose: Assessment of diastolic function with cardiac magnetic resonance (CMR) is challenging and underused, probably based on the belief that it requires complex calculations and special software. Although some diastolic function parameters are easy to obtain with routine cine CMR, information on their significance is currently lacking. The purpose of this study was to compare the diastolic function parameters obtained with CMR in a cohort of patients with hypertrophic cardiomyopathy (HCM) with those seen in a control healthy group.
Methods: A total of 31 patients with HCM and 23 healthy control group subjects with good quality CMR imaging studies, were analyzed. Diastolic function by CMR was obtained with left ventricle (LV) automatic segmentation of routine cine CMR images by tracing endocardial and epicardial borders of LV in short axis views. Time-volume curves were obtained and the following diastolic function parameters were calculated: Peak Filling Rate (PFR), time to PFR and diastolic volume recovery (proportion of diastole required to recover 80% stroke volume) (DVR80). Groups were compared by parametric and non-parametric tests depending on the normal distribution of the samples. A ROC curve was obtained to identify the best cutoff values for diastolic dysfunction.
Results: LV filling curves were successfully obtained in all patients analyzed. Mean age in HCM group was 65 ± 11.6 years old and median age in healthy group was 33 (IQR: 14) years. LV mass was higher in patients with HCM (186.9 ±70.45 gr vs 108.22 ±39.37 gr; p < 0.001). PFR was lower in HCM group compared to control group (400 IQR: 300 vs 600 IQR: 300 ml/sec; p = 0.001). Diastolic filling intervals were significantly prolonged in patients with HCM: tPFR (223 IQR: 169 vs 158 IQR: 63 msec; p = 0.008); DVR80 (83 IQR: 12 vs 57 IQR: 15%; p < 0.001). DVR80 was the parameter that obtained a higher area under the ROC curve (AUC = 0.89 ± 0.04; p < 0.001). A cutoff value of 69.5% DVR80 identified the presence of diastolic dysfunction with 84% sensitivity and 87% specificity.
Conclusions: All CMR simple markers of diastolic dysfunction are significantly affected in patients with HCM. DVR80 was the most sensitive and specific parameter of diastolic dysfunction in these patients.
1477
Impact of platelet volume on thrombus burden and tissue reperfusion in patients with STEMI treated with primary angioplasty: MRI study
Abstract
Objective: Mean platelet volume (MPV), one of the indices of platelet reactivity has been shown to be related to impaired angiographic reperfusion in ST-segment elevation myocardial infarction (STEMI) patients treated with primary angioplasty. However data regarding MPV and its association with tissue level reperfusion in the setting of STEMI are limited. We aimed to investigate whether MPV on admission is associated with microvascular obstruction (MVO) assessed by cardiac magnetic resonance imaging (MRI) in STEMI patients with high thrombus burden of infarct related artery (IRA).
Methods: Based on SMART MRI study, out of 70 eligible patients, cardiac MRI was not performed in 5 patients because of patient's refusal, and in 5 patients due to chronic renal insufficiency. Thus, cardiac MRI was performed in 60 patients 10 ± 6 days after index STEMI, and MVO and IS were available in only 37 patients. All patients underwent thrombectomy and primary angioplasty for STEMI. MRI MVO and infarct size (IS) were measured 8 min after gadolinium injection with late enhancement sequences and were analyzed quantitatively at a blinded core laboratory. Thrombus burden of IRA were evaluated by angiography and by optical computed tomography after thrombectomy. Patients were grouped into two as with (n = 14, 38%) and without MRI MVO. Admission MPV and residual platelet reactivity (RPR, assessed by light transmittance aggregometry) after 60 mg prasugrel loading dose, were measured and compared between two groups.
Results: Admission MPV was similar in patients with and without MVO (11.4 ± 1.0 fL vs. 11.0 ± 0.9 fL respectively, p = 0.214) as well as RPR (ADP 10 µmol, 36.0% ± 18.3% vs. 32.6% ± 14.3%, p = .529). A significant Spearman's correlation was found between MPV and thrombus burden of IRA (p = 0.036), but not between MPV and MVO (p = 0.141). After multivariate analysis, including age, MPV, RPR, IS and door-to-balloon time, the only significant predictor of MVO was the severity of IS as peak creatine kinase (p = 0.003), but not the MPV.
Conclusions: in setting of contemporary primary angioplasty for STEMI using thrombectomy and high-potency antiplatelet drug prasugrel, the MPV on admission is associated with high thrombus burden of IRA but it doesn't seem able to influence tissue reperfusion evaluated by MRI.
1479
Right ventricle systolic function assessment and its prognostic implications in cardiac amyloidosis
Abstract
Background: Cardiac involvement in amyloidosis is due to extracellular deposition of misfolded proteins, leading to a restrictive cardiomyopathy. From well-built evidence we know that left ventricle systolic dysfunction (LVSD) is associated with a dismal prognosis. Nonetheless, the prognosis impact of right ventricle systolic dysfunction in this subset is less clear.
Methods: From 2012 to 2014, consecutive patients with a diagnosis of cardiac amyloidosis, who had a complete transthoracic echocardiographic (TTE) and cardiac magnetic resonance study (CMR), within 2 months, were retrospectively analyzed. Left and right ventricle systolic function was determined by both imaging modalities. TTE measurements included left ventricle ejection fraction (LVEF), left ventricle global longitudinal strain (LVGLS), right ventricle ejection fraction (RVEF) and right ventricle global longitudinal strain (RVGLS). By CMR left and right ventricle ejection fraction (CMR LVEF and RVEF, respectively) were determined. Prognostic outcome was defined by a composite endpoint of cardiac mortality and hospital readmission.
Results: 18 patients presented a complete echocardiographic study. Of those, 8 patients also had a contemporary CMR study (50% females, mean age 70.6 ± 10.9 yo). Primary amyloidosis was diagnosed in 50% and 62.5% were in NYHA class ≥III. By TTE, mean LVEF was 57.1 ± 14.3%, LVGLS was -9.1 ± 4.4%, RVEF was 57.1 ± 16.7% and RVGLS -13.7 ± 9.4%. By CMR, mean LVEF and RVEF was 55.1 ± 16.2% and 46.6 ± 17.9%, respectively. Mean follow-up time was 15.4 ± 15.1 months and the composite endpoint occurred in 62.5% of patients. In this cohort, patient with adverse events presented a significantly lower RVEF (46.4 ± 6.6 vs 69 ± 4.2%, p = 0.007) and worst RVGLS (-6.5 ± 3.2 vs -20.8 ± 0.7%, p = 0.007) by TTE and lower RVEF (37.3 ± 10.3 vs 70.5 ± 2.1%, p = 0.013) by CMR. Patients with adverse outcomes presented lower LVEF and worst LVGLS, by TTE and lower LVEF by CMR, although it did not achieved statistical significance (55.0 ± 12.2 vs 56.0 ± 25.4%, -7.3 ± 2.4 vs -14.4 ± 6.2%, 53.4 ± 11.9 vs 71 ± 12.7%, all p > 0.05).
Conclusions: In our cohort, a right ventricle systolic dysfunction measured by TTE and CMR conferred a dismal prognosis in cardiac amyloidosis patients. This finding heightens the importance of consider this sometimes overlooked tool in the routine evaluation of these patients.
1484
Cardiac MRI - an important tool in the evaluation of multsystemic inflamatory diseases. An Erdheim-Chester Disease case report
Abstract
Introduction: It has been found by MRI and CT studies that cardiac involvement in multisystemic disease occurs more often than expected, in different forms, and that cardiovascular manifestations are underdiagnosed. Cardiovascular involvement in multisystemic disease as in the Erdheim-Chester disease (ECD) is seen in the majority of patients and is associated with a reduced response to treatment and an overall poor prognosis. The prognosis of ECD remains poor, with mortality rates of 60% at 32 months after diagnosis, mainly due to cardiopulmonary causes.
Case Report: A 49-year-old man, affected by ECD, was referred to our department for a correct assessment of the degree of cardiac involvement through cardiac magnetic resonance imaging (MRI). The patient presented with a medical history of bone, retroperitoneal involvement and with heart failure symptoms. The final diagnosis of ECD was confirmed by proximal right tibia biopsy which showed the presence of infiltrated foamy histiocytes, CD 68 + CD1a -.
The MRI study at the admission showed diffusely thickened pericardium with strong enhancement of the pericardial layers after gadolinium administration with nodular delayed enhancement involving the epicardium. Real-time CMR in cardiac short-axis during free breathing showed increased ventricular coupling with inspiratory septal inversion and increased right-sided septal motion at onset of expiration. Analysis with Argus software from short axis SSFP cine images demonstrated reduced left ventricular ejection fraction (LVEF 47%) with reduced stroke volume (36 ml) and reduced ejection fraction of the right ventricle (RVEF 35%).
Cardiac MRI 6 months after treatment with INTERFERON demonstrated the normalization of the ejection fraction of the left ventricle and of the right ventricle with persistence of late-enhanced pericardium with gadolinium. The patient is now 12 months post-diagnosis and clinically stable with improved peripheral edema and dyspnea.
Conclusions: This is a case of heart failure as a manifestation of cardiac involvement in a multisystemic disease. Cardiac MRI is an important tool for a correct assessment of ECD. This technique allows a precise definition of localization and extension of the fibrotic involvement. An accurate evaluation of the cardiovascular system is mandatory because if affected by ECD, the prognosis will worsen. Cardiac MRI represents also the best imaging exam for the evaluation of the response to the treatment.
1485
Predictive value of cardiac magnetic resonance for future adverse cardiac events in patients with ST-segment elevation myocardial infarction
Abstract
Purpose: The aim of the study was to determine prognostic value of infarct size (IS), microvascular obstruction (MVO) and left ventricular ejection fraction (EF) evaluated by cardiac magnetic resonance imaging (CMR), markers of necrosis, electrocardiographic and echocardiographic parameters in patients with ST-elevation myocardial infarction (STEMI) treated by primary percutaneous coronary intervention (pPCI).
Material and Methods: 85 STEMI patients treated by pPCI within 12 hours after symptom onset had an evaluation of CMR within three days. Markers of necrosis were measured every 8 hours from the admission to the hospital.
Electrocardiography (Selvester and Aldrich score) were assessed at the admission, after pPCI and before CMR and echocardiography (WMSI) within the first day after pPCI. CMR was performed on the 1.5T system within 96 hours after pPCI. Morphology and function of LV was estimated by steady-state free precession (SSFP). To evaluate the infarct size and MVO, late gadolinium enhancement (LGE) sequence was (10-15 min after administration of gadobutrol). Infarct size was defined as area above 50% of the maximal signal intensity within LGE (FWMH–full-width half maximum). MVO was described as the area of absence or hyperenhancement of myocardium surrounded by LGE. IS and MVO were determined by planimetry and summation of discs method. All patients were observed 24 months to evaluate major adverse cardiovascular events (MACE). The primary endpoint was defined as a composite of death, myocardial reinfarction, coronary reintervention and congestive heart failure.
Results: Multiple regression analysis showed that that left ventricular ejection fraction, IS and MVO were the strongest independent predictors of MACE (p = 0.0001, p = 0.00001 and p = 0.028 respectively). IS larger than 20% of left ventricular mass predicts MACE with a sensitivity of 85% and specificity of 86%. The presence of MVO also predicts MACE within 2-year follow-up with a sensitivity and specificity of 80%. WMSI, troponin I serum concentration, ECG derived parameters do not provide accurate information of cardiovascular events.
Conclusions: CMR has predictive value for future cardiovascular events. Infarct size, MVO and left ventricular ejection fraction evaluated by CMR are independent predictors of outcome after STEMI.
1486
Time-to-treatment but not thrombectomy influence infarct size and microvascular obstruction in patients with acute ST-segment elevation myocardial infarction treated with primary coronary intervention
Abstract
Background: Infarct size and microvascular obstruction evaluated by cardiac magnetic resonance (CMR) have a prognostic value in patients with ST-segment elevation myocardial infarction (STEMI) treated with primary percutaneous coronary intervention (pPCI). Time to treatment and successful reperfusion are essential in those patients. Some trials have shown that thrombectomy improves surrogate and clinical outcome. Lastly thrombectomy is questioned to be effective in patients with STEMI.
The aim of the study: The aim of the study was to evaluate the influence of time-to-treatment and manual thrombectomy on infarct size (IS) and microvascular obstruction (MVO) in patients with STEMI.
Material and methods: We examined 85 patients (mean age 59 ± 11 years; 59 males and 26 females) with first STEMI treated with pPCI within 12 hours from symptoms onset. Infarct related artery TIMI flow and MBG were evaluated after pPCI. CMR was performed on the 1.5T system within 96 hours after pPCI. Morphology and function of myocardium was estimated by steady-state free precession (SSFP) sequence. To evaluate the infarct size and MVO, a late gadolinium enhancement (LGE) sequence was performed (10-15 min after administration of gadobutrol). Infarct size was defined as area above 50% of the maximal signal intensity within LGE (FWMH–full-width half maximum), MVO was described as the area of absence or hyperenhancement of myocardium surrounded by LGE. IS and MVO were determined by planimetry and summation of discs method.
Results: The mean size of infarct size in study group was 16.52 ± 10.54% of LV myocardial mass. The presence of MVO was observed in 28 patients (32.94%) with a mean size of 8 ± 6% of infarct size. Patients treated in 90 minutes from the symptoms onset had lower infarct size than patients treated after 360 minutes from the beginning of chest pain (p = 0.048). Manual thrombectomy during pPCI was performed in 37 patients (43.53%) and these patients had mean ejection fraction 54.36 ± 8.8% with IS estimated on CMR 16.96 ± 9.8% of myocardial mass and MVO 3.85% of infarct size. The patients treated by pPCI alone, without thrombectomy, had similar EF (54.8 ± 10.7%; p = ns) and CMR findings: IS = 16.25 ± 10.05% of LV myocardial mass (p = ns) and MVO = 4.84% of infarct size (p = ns).
Conclusion: Rapid treatment of patients with STEMI results in smaller infarct size evaluated by CMR. Manual thrombectomy during pPCI does not improve CMR outcomes in STEMI patients.
1489
Primary PCI versus Early Routine Post Fibrinolysis PCI for ST Elevation Myocardial Infarction
Abstract
Background: The rationale for Pharmaco-invasive strategy is that many patients have a persistent reduction in flow in the infarct-related artery. The aim of the present study is to assess safety and efficacy of pharmaco-invasive strategy compared to primary PCI and ischemia driven PCI on degree of myocardial salvage and outcomes.
Methods and Results: Sixty patients with 1st attack of acute STEMI within 12h were randomized to 4 groups (15 patients each): primary PCI for patients presented to PPCI-capable centers (group I), transfer to PCI if presented to non-PCI capable center (group II), pharmaco-invasive strategy (group III) and fibrinolytic (streptokinase) and ischemia driven PCI (group VI). The primary endpoint is the infarct size measured by cardiac MRI 3-5 days post MI. Pharmaco-invasive strategy led to significant reduction in infarction size and major adverse cardiac and cerebrovascular event (MACCE) compared to group IV and decrease -but not significant- in infarction size and MACCE compared to group I and II. Minor bleeding was significantly higher in pharmaco-invasive group compared to other groups.
Conclusions: Pharmaco-invasive strategy resulted in effective reperfusion and smaller infarction size in patients with early STEMI who could not undergo primary PCI within 2 hours after the first medical contact. This can provides a wide time window for PCI when the application of primary PCI within the optimal time limit is not possible. However, pharmaco-invasive strategy was associated with a slightly increased risk of minor bleeding.
1490
Evaluation of ventricular function in Fontan patients undergoing feature tracking magnetic resonance strain
Abstract
Introduction: Feature tracking strain (FTS) is a new technique to assess cardiac function from cardiac magnetic resonance (CMR). We compared FTS with conventional function parameters in single ventricle subjects with Fontan physiology undergoing CMR.
Methods: 18 Fontan subjects (mean age 17.6 + /-9.2 years, post Fontan period 14.2 + /-8.2 years, 13/18 morphologic right ventricle, 5/18 morphologic left ventricle) underwent a CMR study. Ventricular volumes and cardiac output were calculated offline (Medis QMass advanced edition, the Netherlands). Offline strain and strain rate (SR) analyses were performed (TomTec Image Arena, Germany) on the 4-chamber and short-axis views at the basal, mid, and apical levels of the ventricle.
Results: Basal endocardial circumferential strain/SR (-11 + /-9 % and -0.8 + /-0.5 1/s) were lower than it at the mid (-17 + /-6 % and -1.1 + /-0.5 1/s; p = 0.02 and 0.02) and apical (-26 + /-9 % and -1.9 + /-1.0 1 /s; p = 0.001 and 0.003) levels. There was correlation between average endocardial longitudinal strain/SR and ventricular end-diastolic volume (r = 0.73 and 0.68). At the mid and apical levels, there was correlation between average endocardial circumferential strain/SR and ventricular end-systolic volume, and ejection fraction (r = 0.53 to 0.85). There was also correlation between average endocardial radial strain/SR and ventricular stroke volume at the mid level (r = 0.69 and 0.77).
Conclusions: There is correlation between strain/SR and measures of ventricular volume, ejection fraction and cardiac output in single ventricle subjects with Fontan physiology. FTS is independent of inadequate acoustic windows unlike echocardiography and could have clinical relevance. Analysis of regional strain may helpful in understanding myocardial mechanics in the single ventricle in further studies.
1491
Impacts of atrialized right ventricle and left ventricular displacement in Ebstein's anomaly on left ventricular function assessed by cardiovascular MRI
Abstract
Background: Left ventricular (LV) failure is common in Ebstein's anomaly, though remains poorly understood. The objective of this study was to investigate whether atrialized right ventricle (ARV) and LV displacement impact LV function in Ebstein's anomaly by cardiovascular MRI (CMR).
Methods: We analyzed CMR studies in patients with Ebstein's anomaly for measures of severity, including ARV, functional (FRV) and total right ventricular (TRV) volumes, LV volumes, LV and RV ejection fraction (EF). And, we analyzed LV longitudinal, radial and circumferential strain/strain rate in accordance with the 16-segment model proposed by the American Society of Echocardiography. We also measured ARV global longitudinal strain. We related those CMR values to LV failure in Ebstein's anomaly. The MRI imaging was obtained using the 1.5-Tesla system (Siemens MAGNETOM Sonata). The analyses were performed using a workstation (Medis QMass advanced edition and Tomtec Image Arena).
Results: Twelve unrepaired patients (mean age 17 + /-7 years, 75 % female) were included. LV end-diastolic volume was normal (87 + /-24 ml/m2) and LVEF (59 + /-6 %) was preserved. TRV end-diastolic volume (188 + /-127 ml/m2) was dilated and FRVEF (37 + /-16 %) was reduced. At basal and middle cavity of LV, anterior radial strain were higher than it of the septal segments (p = 0.01 and 0.02). At middle and apical cavity of LV, septal circumferential strain were higher than it of the anterior segments (p = 0.001 and <0.05) on the contrary. ARV, FRV and TRV were negatively correlated with LVEF (r = -0.58 to -0.88). However, ARV, FRV and TRV were positively correlated with LV global circumferential strain at middle cavity (r = 0.62 to 0.87), and ARV global longitudinal strain was with LVEF (r = 0.77).
Conclusions: The apparent basal and middle septal radial hypokinesis with middle and apical septal circumferential hyperkinesis observed in most patients is likely attributable to anterior cardiac displacement rather than true paradoxical motion. LVEF was negatively affected by RV volumes; however, right and left ventricular interaction was indeed proved in Ebstein's anomaly.
1494
Final diagnosis for patients presenting with chest pain, electrocardiographic changes or troponin rise and normal coronary arteries: insights from Cardiovascular MRI in our population
Abstract
Literature states that about 75% of patients with chest pain, electrocardiographic (EKG) alteration at admission or troponin (Tn) rise and normal coronary arteries (NCA: normal or stenosis <50%) are diagnosed of myocarditis in CMR.
The purpose of this review was to know the real final diagnosis of patients suffering from this condition in our centre.
Methods: All patients referred to CMR from January 2012 to October 2015 with these criteria were included. Studies were performed within the first week of admission on a 1.5 T or 3T scanners. SSFP cines, T2-weighted black-blood fat-sat images and DE-segmented GE IR images 10-20 minutes after iv gadolinium-DTPA administration (0.15mmol/kg) were performed in 3 long-axis and 11-15 parallel short-axis views. All images were reviewed and analyzed off-line with specialized post-processing software (Philips IntelliSpace Portal®, Philips).
Results: 84 patients met the inclusion criteria. Medium age was 43yo. Males 71%. Main EKG alterations were ST segment elevation (STE) (57%) and T wave inversion (25%). Mean peak TnI levels (12h of admission) was 9.4ng/ml. Mean LVEF was 57% (range 20-84%). Regional wall motion abnormalities were present in 56 cases (63%). High signal in T2w images was detected in 65% of cases and LGE in 60 cases (71%). Final CMR diagnosis resulted in myocarditis (48%), myocardial infarct (MI) (18%), Takotsubo (13%), dilated cardiomyopathy (DCM) (3.6%) and cardiac sarcoid and pericarditis in 1.2%. 8 cases were normal (9.6%). CMR was not conclusive in 1.
Myocarditis patients had the highest peak TnI levels (15.5 vs 4.1; p = 0.02). STE was more frequent in Takotsubo and pericarditis (100%) than myocarditis (54.3%) or MI (35.7%) [p = 0.032]. Interestingly, only 11.4% of patient with STE at admission had final diagnosis of MI. Edema was more frequent in myocarditis (80%) and Takotsubo (81%) as compared to the rest [p < 0.01], being its extention more important in Takotsubo, myocarditis and MI (mean segments affected: 5.7, 3.2 and 1.53 respectively; p = 0.02). Patients with myocarditis and pericarditis were significantly younger (33yo vs 48yo; p < 0.01). DCM patients had larger ventricles (iLVEDV 154ml vs 80ml) and poorer LVEF (26% vs 61%) [p < 0.01].
Conclusions: Less than half of patients of our cohort had final diagnosis for myocarditis. A not insignificant proportion of patients had myocardial infarcts despite NCA (18%). More interestingly, only 11.4% of of patients that presented with STE at admission had MI.
1495
Early Predictive Factors of LV Remodeling after STEMI; Assessment by Coronary Angiogram and Cadiovascular Magnetic Resonance
Abstract
Despite advances in acute myocardial infarction (MI) care, left ventricular remodeling (LVR) after revascularization is one of the major determining factor for heart failure development and cardiovascular events. The aims of our National multicentre Regsitry PREGICA-MRI was to explore angiographic and cardiovascular magnetic resonance (CMR) parameters that were predictive factors of LVR. LVR was defined as a >20% increase in 6-month end-diastolic volume (EDV) from baseline values.
CMR has been recently recognized as one of the best imaging technique for left ventricular (LV) function assessment, myocardial viability and infarct size (IS) or microvascular obstruction (MVO) measurement. It therefore be an optimal imaging modality to predict LVR, assessing IS and MVO.
Material and Methods: Forty six patients (mean age 56,7 ± 10 years) presenting with first episode of acute MI and treated by primary PCI <12h of chest pain onset were studied
CMR was performed within the first week (day 4 ± 2) of ST elevation MI and repeated at a mean time of 6.0 ± 1.4 months. Steady state free precession cines, T2 STIR, first pass perfusion (FP), delayed hyper-enhancement (DHE) images were obtained on a 1.5-T MR system.
Ventricular volumes, ejection fraction (EF), LV mass, and FP perfusion defects were recorded. IS and MVO was measured on DHE images. MVO was defined as areas of hypo-enhancement within hypersignal of necrosed myocardium in DHE images.
Revascularization of culprit lesion was performed by primary PCI within 12 hours of chest pain onset. Myocardial blush index (MBI) and TIMI score was recorded and Index of Microcirculatory Resistance (IMR) was measured using a pressure guide wire.
Results: LVR 6 months after myocardial infarction was found in 16 patients (35%) of the cohort PREGICA-MRI.
IMR level was significantly higher in patients associated with LVR (73,95 versus 27,23 p < 0,03).
A significant risk of LVR was found in patients with IMR > 40 (OR 15 (1,030-218,4), p = 0,03).
LVR was associated with patients having a large transmural infarct on DHE images (OR 25,46 (1,397 - 463,8), p = 0,002), higher number of akinetc segments (6,26 versus 4, p = 0,001), and higher number of segments with MVO (4,43 versus 2,26, p = 0,03).
Conclusions: High level of IMR, number of segments with MVO and transmural DHE can be considered as predictive factors of LVR 6 months after acute MI and early revascularization.
1497
The Pathobiologic Mechanisms and the Prognostic Meaning of t wave Inversion in Acute Myocarditis. a Study Performed by Cardiac Magnetic Resonance
Abstract
Background: The pathobiologic mechanisms as the prognostic meaning of electrocardiographic (ECG) T-wave inversion (TWI) occurring in a subgroup of patients with acute myocarditis remain to be elucidated. Contrast-enhanced cardiac magnetic resonance (CMR) offers the potential to identify myocardial tissue changes such as edema and/or fibrosis. A pathophysiologic link between dynamic TWI and transient myocardial edema in other cardiac disease such asTako-tsubo cardiomyopathy was already described.
Aims: The aims are understand the myocardial substrates that underlie TWI in myocarditis and investigate the prognostic value during follow up.
Methods: We enrolled consecutive patients with acute myocarditis according European Society of Cardiology position statement. All patients to be enrolled had to using a comprehensive CMR protocol which included T2 weighted sequences for myocardial edema and post constrast sequences for necrosis. A 12-lead ECG was collected at the same day. After six months patients underwent clinical follow up consisting of echocardiogram, ECG and -in case of reduced ejection fraction- a second CMR.
Results: We studied 76 consecutive patients (median age 34 years). At the time of CMR, TWI was observed in 21 (27%) patients. There was a statistically significant association of TWI with the median number of left ventricular (LV) segments showing either any pattern of myocardial edema (transmural and non-trasmural) [3(2-4) vs 5(3-7);p = 0.015] or myocardial late-gadolinium-enhancement [3(2-4) vs 4(3-7); p = 0.002]. Transmural myocardial edema involving ≥2 LV segments was found in 17/21 (81%) patients with TWI versus 13/55 (24%) patients without TWI (p < 0.001) and remained the only independent predictor of TWI at multivariable analysis (OR = 9.96; 95%CI = 2.71-36.6; p = 0.001). Overall, topographic concordance between the location of TWI across the ECG leads and the regional distribution of transmural myocardial edema was 88%. There was no association between acute TWI and reduced LV ejection fraction (<55%) at 6-months of follow-up.
Conclusion: LV transmural myocardial edema as evidenced by CMR was demonstrated to be the anatomical substrate associated with the occurence of TWI in patiens with acute myocarditis, As an expression of reversible myocardial edema, the development of TWI during the acute disease phase doesn't represent an adverse marker since doesn't predict systolic LV dysfunction at 6 months follow-up.
1501
The Influence of Left Atrial Function on Exercise Tolerance in Patients with Heart Failure and Preserved Ejection Fraction: A Cardiac Magnetic Resonance Feature Tracking Study
Abstract
Background: Heart failure with preserved ejection fraction (HFpEF) is a syndrome which accounts for impaired exercise capacity in a growing number of patients. Intrinsic ventricular stiffness constant Beta is a known factor resulting in impaired exercise capacity. Cardiovascular magnetic resonance myocardial feature tracking (CMR-FT) is a novel tool measuring atrial strain and phasic performance from routine cine CMR images using standard steady-state free precession sequences (SSFP). The interplay between left atrial (LA) function and exercise capacity in HFpEF patients remains to be elucidated.
Methods: We performed CMR-FT in 20 patients with HFpEF and 10 patients without heart failure symptoms. LA function can be analyzed using 2- and 4-chamber views for assessment of LA reservoir function (total strain [εs], peak positive strain rate [SRs]), LA conduit function (passive strain [εe], peak early negative strain rate [SRe]) and LA booster pump function (active strain [εa], late peak negative strain rate [SRa]). Invasive pressure-volume-loops were obtained with a conductance catheter during basal conditions, transient preload reduction and handgrip exercise. Diastolic stiffness constant Beta was extrapolated during transient preload reduction. During exercise testing on an upright cycle ergometer functional capacity (max Watts) and maximum oxygen uptake (VO2max ml/kg*min) was measured.
Results: Patients with HFpEF showed greater left atrial volumes (LAV max 48 ± 11 vs. 36 ± 13 ml/m2, p < 0.05) and lower exercise capacity (92 ± 29 vs. 168 ± 41 Watts, p < 0.001) and oxygen uptake (17 ± 6 vs. 33 ± 15 ml/kg*min, p < 0.001). LA reservoir and conduit function were significantly lower in HFpEF patients (εs 23 ± 6 vs. 28 ± 6%, p < 0.05; εe 9 ± 5 vs. 15 ± 5%, p < 0.01). HFpEF patients had a higher intrinsic stiffness constant Beta (0.036 ± 0.006 vs. 0.021 ± 0.008, p < 0.001). Beta showed a weak correlation with VO2max (r = -0.37, p < 0.05). Strain measurement for LA conduit function showed the best correlation with VO2max (r = 0.824, p < 0.001). Multivariate regression analysis including Beta, isovolumetric relaxation constant Tau and LA conduit function (εe) was performed. In this model, εe remained the only predictor of VO2max (R= 0.81, R2 = 0.67, p < 0.001).
Conclusion: LA dysfunction, especially impaired conduit function, significantly contributes to impaired functional capacity in HFpEF patients. These results propose LA dysfunction as another important contributor to HFpEF, independent of LV stiffness.
1504
Microvascular Obstruction in Patients with Anterior ST-Elevation Myocardial Infarction who Underwent Primary Percutaneous Coronary Intervention: Predictors and Impact on the Left Ventricular Function
Abstract
Introduction: Microvascular obstruction (MVO) is an independent predictor of left ventricular remodelling and mortality following ST-segment Elevation Myocardial Infarction (STEMI). Cardiac Magnetic Resonance (CMR) is a method for quantitative evaluation of Myocardial Blood Flow (MBF) that might provide superior predictive data in the assessment of MVO.
Hypothesis: We sought to compare angiographic and CMR-derived assessment of MVO to determine the optimal methodology that best correlates with left ventricular function post-STEMI, as well as predictors of MVO.
Methods: Thirty patients presented with anterior STEMI were successfully reperfused with final TIMI 3 flow in the infarct-related artery (IRA). Following primary Percutaneous Coronary Intervention (PCI), angiographic assessment of microvascular perfusion was performed with Myocardial Blush Grade (MBG). Assessment of CMR-derived MVO was performed from day 2 - 4 after primary PCI. MVO volume was evaluated with late gadolinium enhancement CMR sequences. Infarction size was also measured. Echocardiography was done within 48 hours after Primary PCI and repeated after 3 months for assessment of the Left Ventricular ejection fraction (LVEF). Patients were followed up clinically for Major Adverse Cardiac Events (MACE) during hospital stay and up to 3 months. 46 %. MBG showed no correlation to CMR-derived MVO. There was a strong positive correlation between infarction size and MVO volume (p = 0.001) in CMR. There was more delayed peaking of Creatine Kinase (CK) total and (CK)-MB in patients with CMR-derived MVO compared to patients with no CMR-derived MVO (p = 0.016). On multivariate analysis, of all angiographic and CMR variables, CMR-derived MVO was the strongest predictor of LVEF at 90 days (p = 0.004). CMR-derived MVO had a positive association with occurrence of MACE (p = 0.02). The only procedural predictor of CMR-derived MVO was pre-stenting balloon angioplasty (p < 0.05).
Conclusions: CMR-derived MVO in late gadolinium enhancement sequences strongly predicts left ventricular function following Anterior STEMI at 90 days. CMR-derived MVO is associated with a lower LVEF, increased infarct size, and a greater risk of MACE. Pre-stenting balloon angioplasty strongly affects CMR-derived MVO after Primary PCI.
1508
Histological Validation of ECV Quantification by Cardiac Magnetic Resonance T1 Mapping in Cardiac Amyloidosis
Abstract
Background: Cardiac amyloidosis (CA) is caused by accumulation of amyloid fibrils in the myocardium, which leads to an increase in extracellular volume (ECV). Cardiac magnetic resonance (CMR) T1 mapping allows accurate non-invasive ECV measurement. However, it has not been investigated weather CMR-ECV accurately quantifies ECV in CA.
Materials and Methods: Between July 2011 and November 2015 21 CA patients were enrolled. The study population consisted of 7 (33.3%) wild-type transthyretin (TW-TTR) and 14 (66.6%) light chain (AL) CA patients. All patients underwent myocardial biopsy (EMB) and CMR. EMB specimens were stained with Modified Trichrome and ECV was quantified via ImageJ software using a color-threshold macro. CMR-ECV was quantified with T1 mapping using the Modified Look-Locker Inversion recovery (MOLLI) sequence. Spearman's correlation and Bland-Altman plots were used for correlation analysis and assessment of agreement between histological ECV (H-ECV) and CMR-ECV.
Results: All CA patients (N = 21) had a CMR-ECV median of 48.5% (23.6%-71.0%), an H-ECV median of 50.2% (10.7%-80.6%), a combined (CMR-ECV + H-ECV) median of 48.3% (18.8%-75.5%), a median difference between CMR-ECV and H-ECV of -0.376 (-40.7-18.1) and a correlation coefficient of R = 0.752; p= <0.001. For WT-TTR-CA patients (N = 7) results were as follows. Median CMR-ECV of 48.6% (40.8%-70.5%), median H-ECV of 54.2% (32.1%-80.6%), combined ECV median of 54.4% (36.5%-75.5%), a median difference between CMR-ECV and H-ECV of 10.132 (-9.8-18.1) and a correlation coefficient of R = 0.714; p = 0.071. AL-CA patients (N = 14) showed a median CMR-ECV of 47.8% (23.6%-71.0%), a median of H-ECV of 46.8% (10.7%-69.7%), a combined median of 45.9% (18.8%-69.0%), a median difference between CMR-ECV and H-ECV of -0.746 (-40.7-8.4) and a correlation coefficient of R= 0.807; p= <0.001.
Conclusions: We are the first to histologically validate CMR-ECV in cardiac amyloidosis patients. Our results show excellent correlation and good agreement of CMR-ECV with H-ECV in the whole study population as well as in subgroup analysis for AL-CA and WT-TTR-CA. CMR-ECV quantification by T1 mapping accurately reflects ECV in cardiac amyloidosis patients. As many current phase II and phase III trials for treatment of CA are under way, ECV measurement via CMR may provide important information on weather CA therapies can indeed reduce ECV in CA. Therefore, CMR-ECV quantification might be an appealing method in evaluating success.
1513
Comparative Evaluation of Flow Quantification Across the Atrioventricular Valve in Patients with Functional Univentricular Heart After Fontan's Surgery and Healthy Controls: Measurement by 4D Flow Magnetic Resonance Imaging and Streamline Visualization
Abstract
Purpose: To evaluate the inflow pattern and flow quantification in patients with functional univentricular heart after Fontan's operation using 4D flow magnetic resonance imaging (MRI) with streamline visualization when compared with the conventional 2D flow approach.
Method: Seven patients with functional univentricular heart after Fontan's operation and twenty-three healthy controls underwent 4D flow MRI. In two orthogonal two-chamber planes, streamline visualization was applied, and inflow angles with peak inflow velocity (PIV) were measured. Trans-atrioventricular flow quantification was assessed using conventional 2D multiplanar reformation (MPR) and 4D MPR tracking the annulus and perpendicular to the streamline inflow at PIV, and they were validated with net forward aortic flow.
Results: Inflow angles at PIV in the patient group demonstrated wide variation of angles and directions when compared with the control group (p < 0.01). The use of 4D flow MRI with streamlines visualization in quantification of the trans-atrioventricular flow had smaller limits of agreement (2.2 ± 4.1mL; 95% limit of agreement -5.9-10.3mL) when compared with the static plane assessment from 2DFlow MRI (-2.2 ± 18.5mL; 95% limit of agreement agreement -38.5-34.1mL). Stronger correlation was present in the 4D flow between the aortic and trans-atrioventricular flow (R2 correlation in 4D flow: 0.893; in 2D flow: 0.786).
Conclusions: Streamline visualization in 4D flow MRI confirmed variable atrioventricular inflow directions in patients with functional univentricular heart with previous Fontan's procedure. 4D flow aided generation of measurement planes according to the blood flood dynamics and has proven to be more accurate than the fixed plane 2D flow measurements when calculating flow quantifications.
Figure 1. MRI 4D flow with streamline visualization (a,b) Healthy volunteer four-chamber view with streamline visualization by velocity encoding (c,d) A patient with functional univentricular heart, two-chamber view with streamline visualization at diastole and systole demonstrating regurgitant jet
Figure 2. Agreement between atrioventricular valve forward flow and aorta. Bland–Altman graphs between aortic flow and effective forward flow in healthy patients with (a) 2D flow (b) 4D flow and patients with functional univentricular heart with (c) 2D flow and (d) 4D flow.
1515
Does arterial switch for d-transposition of the great arteries alter myocardial deformation of the ventricles?
Abstract
Introduction: The arterial switch operation (ASO) is currently the surgical technique of choice for repair of d-transposition of the great arteries. The main pulmonary artery is moved forward (Lecompte maneuver) and its branches are stretched on either side of the ascending aorta. The coronary arteries are removed from and reinserted into the aorta. We sought to assess myocardial deformation changes in the right (RV) and left ventricles (LV) as signs of subclinical myocardial functional impairment after ASO and re-positioning of the coronary arteries.
Methods: Patients after ASO and normal controls underwent cardiac magnetic resonance (CMR) imaging including 2D SSFP for ventricular function. 2D SSFP cine images were post-processed with a feature tracking software (TomTec 2D CPA). Global circumferential strain was measured on short axis mid-ventricular slices and global longitudinal strain on horizontal long-axis images, separately for each ventricle. Patients with pulmonary arteries stenoses or history, symptoms, or CMR findings suspicious for coronary compromise were excluded.
Results: Eighteen patients after ASO (age 16.8± 6.7y) were compared to 18 normal controls (age 22.2± 11.4y; p = 0.098). RVs of ASO patients showed lower longitudinal strains (-14.1± 6.4% vs. -18.3± 3.8%; p < 0.05) but higher circumferential strains (-16.6± 3.2% vs. -13.1± 4.3%; p < 0.01) compared to normal RVs. LV longitudinal strain (-15.4± 5.1% vs. -17.5± 4.6%; n.s.) and LV circumferential strain (26± 5.6% vs. -23± 13.1%; n.s.) were not significantly different in patients vs. controls. There were no differences between ASO patients and controls regarding ejection fractions of the RV (54± 6% vs. 52%± 5%; n.s.) and LV (58± 8% vs. 60%± 5%; n.s.) or regarding end-diastolic volumes of the RV (91± 21ml/m2 vs. 94± 12ml/m2; n.s.) and LV (87± 26ml/m2 vs. 80± 11ml/m2; n.s.) indexed to body surface area, respectively.
Conclusions: LV deformation is preserved after the ASO operation, despite coronary artery surgery. In contrast, even in the absence of significant pulmonary artery stenosis, RV deformation is altered with decreased global longitudinal strain and increased circumferential strain, while preserving RV volume and ejection fraction. This may be the result of abnormal ventriculo-arterial coupling after the Lecompte maneuver and its changes in the outflow tract geometry.
1527
Accuracy of T1 Mapping by multi-professional CMR operators to predict myocardial infarct
Abstract
Background: Native T1 mapping is increasingly used in clinical and research applications to distinguish between normal myocardium and fibrosis or scar. Online reconstruction of T1 maps on the scanner console is now widely available and can delineate areas of myocardial infarction prior to late gadolinium enhancement (LGE) imaging. Such advance knowledge may help in optimising timings and obtaining targeted views during LGE. The aim of the study was to establish the feasibility and accuracy of detecting myocardial infarction from T1 maps obtained during live scanning and to compare the performance of clinicians and non-clinicians.
Methods: Native T1-maps were generated on a Philips scanner (1.5T Ingenia CV, Philips Healthcare) during live scanning. The colour scales for the maps were set between 200msec and 1600msec. 8 CMR professionals of varying scanning ability (4 Cardiologists, 3 Radiographers, 1 Medical Physicist) received basic T1 mapping reading instructions to identify abnormal myocardium with infarct. Each individual then assessed 40 segments (in total 320 segments by 8 assessors) of short-axis native T1 maps and were asked blindly to identify normal, area(s) of infarct and affected coronary territories. Sensitivity and specificity to detect infarction on subsequent LGE images was determined.
Results: In total 320 segments involving 10 cases were assessed by eight readers. There were no significant differences in detecting myocardial infarction between the three groups readers (cardiologists: sensitivity of 87%, 95% CI 66-98 and specificity of 67%, 95% CI 46-76; radiographers: sensitivity of 83%, 95% CI 69-92 and specificity of 75%, 95% CI 54-89; physicist: sensitivity of 83% and specificity of 75%). The accuracy of all observers to identify infarction on LGE and to identify the correct coronary artery territory are listed in Table 1. The highest sensitivity and specificity were demonstrated for inferior infarction while sensitivity to detect septal infarction was low.
Conclusions: T1 Mapping is an additional tool that may be used by the CMR operator to predict infarct size and location prior to LGE. Information obtained pre-LGE is beneficial where patients have a limited tolerance to breath-hold commands and the targeting of infarct areas over normal myocardium is a priority during a limited window of opportunity for LGE. In this pilot study all clinicians and non-clinicians perform similarly.
1531
Detecting hypertensive heart disease: the additive value of cardiovascular magnetic resonance imaging
Abstract
Introduction: Arterial hypertension is common. Hypertensive heart disease (HHD) is associated with significant morbidity and mortality. European Society of Hypertension/Cardiology guidelines advise screening for HHD in hypertensive subjects, to aid risk stratification which may impact clinical management. Cardiac magnetic resonance (CMR) is current non–invasive gold standard for assessing ventricular structure and function. We aimed to determine the additive value of CMR in hypertensives who had already undergone echocardiography.
Methods: 85 subjects from the tertiary hypertension clinic with preceding echo underwent 1.5T CMR. Left ventricular mass and volumes were estimated from short–axis steady–state free precession cines. LVH was defined on the basis of echo and CMR normal reference ranges. LV wall thickness in the mid–cavity interventricular septum (IVS) and the mid–cavity posterolateral wall (PLW) were compared between modalities. Gadolinium was administered to all subjects undergoing CMR. The presence and pattern of myocardial late gadolinium enhancement (LGE) was documented.
Results: 85 subjects were included (60% men, office systolic BP: 165 ± 28mmHg, office diastolic BP: 94 ± 14mmHg). Overall, there was no difference in prevalence of LVH by echo compared to CMR (68% vs 66%, P = 0.746). However, there was a discrepancy between echo and CMR in 28%. Relative to CMR gold– standard, echo over–diagnosed LVH in 15% and missed LVH in 13%. The diagnostic performance of echo at detecting LVH was as follows: specificity 55%, sensitivity 80%, positive predictive value 78%, negative predictive value 59% and overall accuracy 72%. The echo measurement of PLW was >2mm than the corresponding CMR measurement in 45%. Ischaemic LGE was present in 7% of subjects (Figure 1A) and non–ischaemic LGE was present in 9% (Figure 1B).
Conclusion: Although there was no significant difference in overall prevalence of hypertensive LVH between echocardiography and CMR, echocardiography over–diagnosed and under–diagnosed LVH in an important minority of patients. Echocardiographic measurement of PLW was particularly prone to error relative to CMR. LGE tissue characterization is unique to CMR and identified ischaemic and non–ischaemic myocardial fibrosis is an important minority of hypertensive patients. Our findings support an extended role of CMR in hypertension where documenting in the presence/absence of HHD may have clinical management implications.
1534
Diagnostic Performance of Cardiac Magnetic Resonance Strain Parameters in Assesment of Myocardial Ischemia
Abstract
Recently new tools for analysis of myocardial strain by cardiac magnetic resonance (CMR) have been developed and must be analyzed to establish their clinical utility. We aim to study usefulness and diagnostic performance of CMR strain parameters in patients with early ischemia.
Methods: patients with chest pain and CMR dipyridamole stress test conducted prior catheterization were included. All of them had positive stress test and coronary artery stenosis ≥70% confirmed by coronary angiogram. CMR stress protocol included functional study with steady state free precession sequences in cardiac long and short axis at rest and after administration of 0,84mg/kg of intravenous dipyridamole and also 4-6 slices of spoiled fast gradient echo sequences acquired post- stress every two beats during al least two minutes at the same time of administration of 0,15mmol/kg of God-BOPTA to study myocardial perfusion.
Inducible wall motion abnormalities (IWMAs) and also global and segmental circumferential (C) and radial strain and their time to peak (TTP) were measured with dedicated software. Changes of these parameters after stress were also analyzed. Segmental ischemia analysis was performed using as standard reference the presence of inducible perfusion deficit along with stenosis ≥70% in the responsible coronary artery. Assessment of strain, inducible wall motion abnormalities (IWMAs) and perfusion were performed blindly by three independent observers using AHA standard myocardial segmentation.
Results: 203 basal and medium myocardial segments were studied. CMR strain parameters with significant association, AUROC, cut- off points and diagnostic value are shown in the table. Assessment of IWMAs in this setting showed low sensitivity, 30%, and positive predictive value 56% but high specificity and negative predictive value 90% and 76% respectively (P > 0.05).
Conclusions: 1.CTTP after stress and its change were the parameters with best diagnostic performance for diagnosis of early ischemia by CMR strain. 2. Despite their modest specificity computational assessment of strain had better diagnostic performance than visual assessment of IWMAs.
Parameters . | AUROC . | p . | Limits of confidence . | Cut-off . | Se/Sp . |
---|---|---|---|---|---|
Stress CTTP | 0.63 | 0.016 | 0.52-0.65 | 351 | 90/30 |
CTTP change | 0,33 | 0.002 | 0.23-0.43 | -163 | 93/15 |
Parameters . | AUROC . | p . | Limits of confidence . | Cut-off . | Se/Sp . |
---|---|---|---|---|---|
Stress CTTP | 0.63 | 0.016 | 0.52-0.65 | 351 | 90/30 |
CTTP change | 0,33 | 0.002 | 0.23-0.43 | -163 | 93/15 |
CTTP: circumferential time to peak.
CTTP change: changes in CTTP stress vs rest.
1535
Relationships between left ventricular filling pressures and longitudinal dysfunction with myocardial fibrosis in uncomplicated hypertensive patients
Abstract
Objective: We explored relationships between left ventricular filling pressures (LVFP) and longitudinal dysfunction with myocardial fibrosis in hypertensive patients without left ventricular (LV) hypertrophy.
Methods: Thirty-five hypertensive patients without echocardiographically determined LV hypertrophy underwent a complete echo-Doppler exam including Speckle Tracking and cardiac magnetic resonance with late gadolinium enhancement (LGE).
Results: LGE was identified in 13 patients (37%): mainly non-subendocardial and involving always the ventricular septum. Regardless the type of diastolic pattern, E/e' was higher (p < 0.0001) in patients with LGE than in those without LGE. All patients with LGE had E/e' > 8. Elevated LVFP (E/e' ≥ 13 or E/e'> 8 with left atrial volume index ≥34 ml/m2) were found in 39% of LGE patients (all showing LGE at the right ventricular attachment to ventricular septum) and only in 5% of patients without LGE (p = 0.01). Tissue Doppler s' velocity (p = 0.0003) and global longitudinal strain (GLS) (p = 0.0005) were lower in presence of LGE. LGE volume varied substantially with the type of quantification used (Figure), but no difference of manual with 6 standard deviations (SD), 5SD and full width at half maximum (FWHM) was found. Manual quantification showed high intra-observer and inter-observer reproducibility (mean difference and 95% CI = - 0.42± 0.82%, ICC = 0.94, p = 0.004 and -0.37± 0.81%, ICC = 0.94, p = 0.005; respectively). 6SD and FWHM were the semi-automated techniques with the highest intra- and inter-observer reproducibility (both p < 0.0001). E/e' correlated with mean LGE volume quantified manually (r = 0.75, p = 0.003), FWHM (r = 0.81, p = 0.0007) and 6SD (r = 0.74, p = 0.004). GLS correlated with LGE quantified manually (r = 0.66, p = 0.01), FWHM (r = 0.66, p = 0.01) and 6SD (r = 0.66, p = 0.01). s' correlated with LGE quantified manually (r = 0.64, p = 0.01), FWHM (r = 0.68, p = 0.01) and 6SD (r = 0.66, p = 0.01). By adjusting for age and relative wall thickness in separate multilinear regression analyses, LGE presence was independently associated with E/e' (β = 0.74; p <0.0001) and GLS (β = 0.61; p = 0.003) or s' (β = -0.59; p < 0.005).
Conclusion: Even in absence of LV hypertrophy, in hypertensive patients LVFP and GLS are independently associated with evidence and magnitude of LGE. LVFP increase and longitudinal dysfunction should be considered as indicators of the extent of myocardial fibrosis induced by pressure overload.
1539
Predictive Clinical Factors of Tissue Damage Severity in Reperfused Acute Myocardial Infarction as Visualized by Cardiac Magnetic Resonance
Abstract
Primary percutaneus coronary intervention (PCI) has been shown to decreased mortality in patients with ST -segment elevation myocardial infarction (STEMI) reperfused by primary PCI, nevertheless its usefulness to preserve myocardial tissue is influenced by some clinical and technical factors. We aim to study which factors are associated with size and severity of myocardial damage as visualized by cardiac magnetic resonance (CMR) in patients with STEMI reperfused by primary PCI.
Methods: Consecutive patients with STEMI were studied by CMR in a first week post -STEMI and clinical and imaging data were collected. CMR protocol included short and long axis b-SSFP cine sequences, STIR sequences in identical slices to cine sequences to quantify myocardial mass at risk (RA) and also early (i.e. 1-2 minutes) and late (i.e. 5-8 minutes) post contrast LGE- FGE sequences to asses mass of microvascular obstruction and necrosis as shown in fig1. Quantification of necrosis and OMV mass was performed with dedicated software.
Necrosis ratio was derived as percentage of risk area (NR) and microvascular damage was analyzed as percentage of necrosis mass (MVOR). Demographic data, cardiovascular risk factors, TIMI flow prior PCI, successful trombectomy, severity of associated coronary artery disease and delay of revascularization (i.e. total ischemic time (IT), door to device time, pain to drug time and transfer delay) were also analyzed. Linear regression multivariate analysis was employed to find the best predictive model for cellular and vascular myocardial damage as visualized by CMR.
Results: Seventy one patients, 55 (76%) males, mean age 65± 13 years were included. All but one of patients had TIMI 3 flow in culprit artery after PCI. Necrosis ratio was significantly associated with pain to drug time (PTD) as shown by the algorithm: NR = 0.3PTD-173. (R = 0.3, P = 0.04). Nevertheless total ischemic time (IT) was the only independent predictor of microvascular damage as shown by the algorithm: OMVR= IT 0.5 (R = 0.52, p < 0.001)
Conclusions: 1. Necrosis ratio is significantly associated to time to drug in multivariate analysis. 2. OMV ratio is significantly predicted by total ischemic time.
OMV ratio: (OMV mass in early LGE sequences/Necrosis mass in late LGE) sequences)x100.
Figure 1. left panel: LGE-FGE late enhancement sequences: right panel LGE-FGE early enhancement sequences.
1541
Which CMR derived parameter predicts better the need of invasive treatment in aortic coarctation?
Abstract
Severity of aortic coarctation (AC) is actually evaluated using clinical and imaging parameters in a comprehensive manner. We aim to analyze which cardiac magnetic resonance (CMR) derived parameters predicts better the need of surgical or percutaneous treatment of aortic coarctation.
Methods: Consecutive patients with AC referred for CMR were selected. Patients without focal stenosis or loss of follow up were excluded. Clinical data were obtained reviewing medical records with a range of follow up (1-70 months). CMR protocol includes: 1.Angiographic evaluation with multiplanar reconstructions to estimate the minimum diameter and area at proximal aorta indexed by body surface area and relative stenosis (RE) referred to the maximum diameter of distal descending aorta; 2. Physiological evaluation with phase-contrast sequences quantifying maximum velocity post-coarctation, derived gradient and flow pattern in thoracic aorta expressed as flow increase rate (FIR= distal thoracic aorta flow–proximal thoracic aorta flow)/proximal thoracic aorta flow %.)
Results: 105 patients 21 (78%) male, mean age (25+ 15 years), 12 excluded, 5 due to suceptibility artefacts, 2 due to absence of focal stenosis and five for losing follow up. Ninety seven patients, 50 of them with native AC were finally studied. Twenty seven (28%) required treatment during follow up and in 9 of them (9.3%) was percutaneous. Area under receiver operator characteristics curves (AUROC), statistical significance, cut-off points and confidence limits are shown in the table. Associated variables were entered into a multivariate model in order to find the best model to predict the need for intervention within a year after CMR. The best diagnostic model included both RE and FIR and entering FIR in the evaluation accounts for 11.5 % more patients correctly predicted over only anatomic evaluation. In addition if Cox model is used for evaluation FIR is the only independent predictor with HR de 1.1 for every 1% increase in FIR.
Conclusions: 1. Flow pattern in thoracic descending aorta expressed as flow increase rate is the CMR parameter that predicts better the need of invasive treatment of aortic coarctation with HR of 1.1 every 1% increase in FIR. 2.Entering FIR in the evaluation accounts for 11.5 % more patients correctly predicted over anatomic evaluation using ER within the first year after CMR.
Variables . | AUROC . | P . | Confidence . | Cut-off . | S/E % . |
---|---|---|---|---|---|
Minimum diameter(i) | 0.72 | <0,05 | 0.14 -0.42 | 5mm/m2 | 92/53 |
Peak gradient | 0.60 | 0,3 | 0.43 -0.77 | 20mmHg | 77/50 |
Flow increase rate% | 0.90 | <0,001 | 0.82-0.98 | 15% | 88/99 |
Maximum velocity | 0.60 | 0,3 | 0.43 -0.77 | 2.2m/seg. | 77/50 |
Relative stenosis | 0.85 | <0,001 | 0.75 -0.95 | 48% | 82/79 |
Minimum area (i) | 0.84 | <0.001 | 0.72-0.95 | 33mm2 | 89/73 |
Variables . | AUROC . | P . | Confidence . | Cut-off . | S/E % . |
---|---|---|---|---|---|
Minimum diameter(i) | 0.72 | <0,05 | 0.14 -0.42 | 5mm/m2 | 92/53 |
Peak gradient | 0.60 | 0,3 | 0.43 -0.77 | 20mmHg | 77/50 |
Flow increase rate% | 0.90 | <0,001 | 0.82-0.98 | 15% | 88/99 |
Maximum velocity | 0.60 | 0,3 | 0.43 -0.77 | 2.2m/seg. | 77/50 |
Relative stenosis | 0.85 | <0,001 | 0.75 -0.95 | 48% | 82/79 |
Minimum area (i) | 0.84 | <0.001 | 0.72-0.95 | 33mm2 | 89/73 |
1543
Contrast-enhanced magnetic resonance tomography in patients with supraventricular tachyarrhythmias
Abstract
The aim of this study was quantitative assessment of the relationship of signs of damage to the atrial myocardium revealed with using contrast-enhanced MRI (CE MRI), with the risk of development of supraventricular tachyarrhythmias.
Methods: The study included 34 patients (29 men, 5 women, aged 54 ± 9). Patients were assembled into two clinical groups: 1 - with single supraventricular and ventricular extrasystoles or with episodes of atrial fibrillation (14); 2 - with frequent episodes of paroxysmal supraventricular tachyarrhythmias (20) and the control group, without signs of clinically significant atherosclerosis (9 persons).
All patients underwent CE MRI of the heart (contrast agent (CA) was injected at a dose of 0.1 mmol/1 kg body weight). According to the MRI the volume of the left atrium (LA), cm3; index of enhancement (IE) T1-WI of the largest area in the LA; the volume of accumulation of CA in the area of its greatest accumulation cm3; total volume of the accumulation of CA in the LA, cm3 were determined.
Results: In Group 2 the accumulation of CA was of multifocal character in 17 of 20 patients (p <0,05, criterion χ2), whereas in Group 1 multifocal character was seen only in 1 patient. In Group 2 the volume of LA was significantly higher compared to Group 1 (94.3 ± 10.2 cm3 и 55.3 ± 9.5 cm3, respectively). Patients in whom the volume of the LA was more than 80 cm3 tended to develop persons atrial tachyarrhythmia (probability 90%). A statistically significant difference was determined by the IE between the groups (1.43 ± 0.07 vs. 1.15 ± 0.04): the IE > 1.27, meant belonging to Group 2. The volume of CA accumulation in area of the greatest accumulation (0.516 ± 0.111 and 0.279 ± 0.085 cm3, p < 0.05) and total amount of accumulation of the LA (0.769 ± 0,212 and 0.325 ± 0.103 cm3, p<0.05) were significantly higher in Group 2. The combination of signs: atrium expansion >80 cm3, the IE > 1.27, the multiple character of the accumulation of CA - points to the inevitability of the development of supraventricular tachyarrhythmias.
Conclusion: CE MRI of atrial myocardium is an important additional method of investigation and prediction of development of supraventricular arrhythmias.
1546
Prognostic Value of CMR Imaging Biomarkers on Outcome in Peripheral Arterial Disease: a 6-year Follow-up Pilot Study
Abstract
Objectives: To explore whether cardiovascular magnetic resonance (CMR) imaging biomarkers show prognostic value of outcome after six year follow-up in patients with symptomatic peripheral arterial disease (PAD).
Methods: Forty-two consecutive patients (mean age 64 ± 10 years), referred for contrast-enhanced MR angiography (CE-MRA) evaluation in the work-up for PAD, were included. In all patients, a comprehensive CMR examination was performed consisting of multi-station CE-MRA of the infra-renal aorta and run-off vessels, carotid vessel wall imaging from dual-inversion black blood imaging, cardiac cine imaging using steady-state free-precession, and assessment of the proximal and distal aortic pulse wave velocity (PWV) from two high-temporal through-plane velocity-encoded MRI acquisitions placed perpendicular to the aorta. After a follow-up period of 72 ± 5 months patients were categorized for outcome: all-cause mortality, cardiac event (coronary artery disease, heart failure, myocardial infarction) or cerebral event (stroke, transient ischemic attack).
Results: At baseline, mean MRA stenosis class (i.e., the average stenosis severity visually scored over 27 standardized segments on a 5-point scale) over all patients was 1.6 ± 0.5. Mean proximal aorta PWV was 9.8 ± 6.3m/s and descending aorta PWV 9.1 ± 3.8m/s. Mean left ventricular ejection fraction (LVEF) from short-axis planimetry was 58 ± 10% and mean carotid normalized wall index 0.46 ± 0.07. During follow-up, six patients had died, six patients experienced a cardiac event and three patients a cerebral event. Binomial logistic regression analysis showed mean stenosis class as a significant independent predictor for all-cause mortality (beta 3.0 ± 1.3, p = 0.02). Descending aorta PWV and age were interrelated with PAD stenosis severity and associated with all-cause mortality, but were not significant independent predictors. For occurrence of a cardiac event, LVEF was a significant independent predictor (beta -0.20 ± 0.08, p = 0.008). None of the CMR imaging biomarkers was a significant independent predictor for cerebral event.
Conclusion: Mean MRA stenosis class was a significant independent predictor for all-cause mortality in patients with symptomatic PAD after a follow-up period of six years. LVEF was a significant independent predictor for cardiac morbidity in this patient population. Further studies with larger patient groups and a longer follow up period have to confirm these results of the present pilot study.
1549
Dobutamine-Stress-CMR in Young Adults after Arterial Switch Operation as Neonates
Abstract
Introduction: Neonatal arterial switch operation (ASO) is the standard of care in patients with simple transposition of the great arteries (TGA). Follow up can be complicated by stenosis or kinking of the reinserted coronary arteries causing myocardial ischemia, and, after Lecompte maneuver, stenosis or kinking of the central pulmonary arteries (PA) resulting in abnormal left/right pulmonary blood flow distribution (PBFD).
The aim of this study is to evaluate the rate and extend of stress induced myocardial ischemia and of abnormal differential PBFD at rest and under dobutamine stress.
Methods: 23 patients (age 18-28; 16 male) underwent CMR (1,5 Tesla) at rest (coronary scan, cine, phase contrast flow) and under dobutamine/atropine stress (cine, phase contrast flow pulmonary perfusion), followed by angiography and late gadolinium enhancement (LGE).
Results: Coronary arteries: 1x known prox. LAD occlusion (collateralized via RCA; coronary type AB1), 1x proximal LCA-occlusion, LIMA-bypass). No other stenosis or kinking of the proximal coronary arteries. Coronary classification (Sauer/Gittenberger-de Groot): 14x A1 (normal type), 5x AB1 (RCX originates from RCA), 4x B1 (right single ostium, LCA retroaortal). 3 patients were reclassified compared to the operation report.
Dobutamine-stress: 21/23 reached target heart rate; 1x abort at 120/min (20 µ dobu) due to complex ventricular extrasystoles, 1x abort at 160/min due to problems with ECG-triggering. 1 patient with known LAD occlusion showed strong chest pain and hypokinesia in 1 segment under maximum stress.
Pulmonary perfusion at rest: 2 patients showed relevant stenosis of the mean PA. 2 patients had abnormal PBFD (>2:1). 3 patients had prior intervention for RVOT/PA-stenosis. (PBFD > 2:1). -> cumulative rate of relevant RVOT/PA stenosis 7/23 (30%).
Pulmonary perfusion under dobutamine stress: there was no relevant difference of PBFD under dobutamine stress compared to rest-PBFD (Friedman test: p= 0,75; see table). On individual patient level, no worsening of abnormal PBFD was found.
Conclusion: CMR found no new proximal coronary stenosis or kinking. Coronary type was reclassified in 3 patients. 2/23 DSMR were pathologic. 4 patients had relevant RVOT/PA stenosis, 2 patients had abnormal pulmonary blood flow distribution at rest (PBFD >2:1). PBFD did not worsen under dobutamine stress.
LGE: no myocardial scarring. 1 patient had intramyocardial fibrosis.
1553
Impact of posteromedial papillary muscle infarction on mitral regurgitation after ST-segment elevation myocardial infarction
Abstract
Introduction: Recently there is a growing interest in the role of papillary muscle infarction (PMI) as a cause of mitral regurgitation and unfavorable outcome after ST-segment elevation myocardial infarction (STEMI). We hypothesized that PMI of the posteromedial papillary muscle might be associated with a higher incidence of mitral regurgitation (MR) after STEMI.
Methods: We included 242 patients with first STEMI who underwent a late-enhancement cardiac magnetic resonance scan within a median of 2 (IQR: 2-4) days and echocardiography within 3 (IQR: 2-5) days after primary angioplasty for the index event. PMI was scored based on the extent of LE in at least 2 consecutive short axis slices (AL-PMI: anterolateral PMI, PM-PMI: posteromedial PMI). MR was reviewed and graded by at least one experienced observer based on the recommendations of the European Association of Echocardiography.
Results: PMI occurred in (40%) of the study population. 23% had isolated PM-PMI, 7% AL-PMI and 11% had infarction of both PM. We observed a higher incidence of PM-PMI (28% versus 13%) and a lower incidence in AL-PMI (4% versus 13%) in patients with MR. Patients with isolated PM-PMI had higher odds for the occurrence of MR than patients with no-PMI, AL-PMI or both-PMI (OR: 2.62, CI: 1.27–5.40, p < 0.01), whereas isolated AL-PMI resulted in a decreased odds-ratios for MR (OR: 0.267, CI: 0.1–0.751, p < 0.01). Furthermore age above the median of 56 years resulted in an increase in the odds for MR (OR: 2.400, CI: 1.391–4.140, p < 0.01). In patients with non-anterior STEMI patients with PM-PMI had increased odds for the occurrence of MR (OR: 3.19, CI: 1.21–8.40, p < 0.02).
Conclusion: Isolated PM-PMI increases, and isolated AL-PMI decreases, the odds for MR after STEMI independent of infarct location.
1556
Role of cardiac magnetic resonance imaging in assessment of left ventricular hypertrophy
Abstract
Background: Determination of the etiology of left ventricular hypertrophy (LVH) is a common clinical problem that can be challenging, because pathological forms of LVH present with overlapping phenotypes in various diseases of the myocardium.
Left ventricular hypertrophy is frequently encountered in the setting of uncontrolled systemic hypertension and aortic stenosis. While aortic stenosis is a straightforward differential diagnosis, other conditions can be difficult.
In some cases, the differentiation between hypertensive heart disease and other cardiomyopathies causing ventricular hypertrophy may be difficult.
A number of other conditions can mimic LVH such as infiltrative diseases, hypertrophic cardiomyopathy (HCM), Fabry's disease, cardiac sarcoidosis, aortic stenosis, and exercise-induced ventricular hypertrophy or athlete's heart.
The unique capacity of CMR to non-invasively characterize myocardial tissue represents an important advantage over other non-invasive imaging modalities.
Echocardiography is still the first-line imaging modality for assessment of myocardial disease, but it has its limitations, including restricted echo windows and poor tissue characterization ability. With technical advances in the field of cardiovascular magnetic resonance imaging, it has been shown to be helpful in reaching a more precise diagnosis in most of the cases.
Patients and Methods: This study was conducted on 54 patients who showed various patterns of left ventricular hypertrophy by echocardiography. All patients were examined by cardiac magnetic resonance imaging (Philips Achieva 1.5T). Sequences included ECG-gated SSFP, black-blood imaging, phase contrast flow imaging, and late gadolinium enhancement.
Results: MRI confirmed the diagnosis of HCM in 18 patients, assessing the various phenotypes and patterns of left ventricular hypertrophy. It showed left ventricular outflow tract obstruction in 9 of those 18 patients, and midcavitary obstruction in 6 patients. Myocardial fibrosis was detected in 9 of the 18 patients by late gadolinium enhancement.
Eleven patients were diagnoses by MRI as having cardiac amyloidosis, another 9 patients as having other types of restrictive cardiomyopathies further assessed by laboratory and pathologic testing.
In 3 patients, left ventricular hypertrophy was attributed to aortic stenosis, and the remaining 9 patients the cause of LVH was attributed to systemic hypertension.
Conclusion: Cardiac MRI is a valuable tool in assessment of left ventricular hypertrophy, for assessment of ventricular size, function, global and regional mass, presence of outflow tract and cavitary obstruction, as well as myocardial fibrosis. The overall acquired data can help guide the diagnosis and management of the underlying cause of left ventricular hypertrophy.
42-year-old male patient with hypertension and palpitations. 4C and SA images captured from SSFP sequences, showing LVH, mild pericardial effusion, and a mitral regurgitation jet with bi-atrial dilatation. Ventricular volumes were normal with low-normal systolic function, with diastolic dysfunction.
LGE image in SA plane acquired after challenging TI scout, showing dark blood-pool with global subendocardial LV and RV wall enhancement. Features were suggestive of amyloidosis.
50-year-old female patient suffering from hypertension and rheumatoid arthritis. LGE images in SA and 4C plane, showing diffuse ring of subendocardial LV wall enhancement. Suggested diagnosis was amyloidosis versus Loeffler endocarditis, biopsy and laboratory investigations were requested.
19 year old female patient suffering from palpitation and effort intolerance. 4C and SA images captured from SSFP sequences. 4C image in systole, showing LVH mainly midcavitary and apical, with obliteration of LV cavity in systole. SA image at the apical region in diastole also showing LVH. LGE image in SA plane, showing no evidence of myocardial fibrosis. Features were suggestive of apical HCM.
1569
Using intrinsic Cardiac Shear Waves to measure Myocardial Stiffness: Preliminary results from Patients with Heart failure with preserved Ejection Fraction
Abstract
Background: Heart Failure with preserved Ejection Fraction (HFpEF) is common and challenging to diagnose. Transient Magnetic Resonance Elastography (tMRE) has been shown to detect stiffness in many organs and so we have applied this concept in vivo to measure myocardial stiffness.
Methods: Aortic valve closure results in a shear wave that propagates through the myocardium. The speed of this wave can be measured if the sequence is timed to the exact valve closure time. Individual sequences are ECG and navigator gated and take approximately 90 seconds, the complete scan including planning takes approximately 40 minutes. Increased myocardial stiffness results in increased speed of shear wave propagation.
Our study has three parts: sequence development and implementation on a clinical 3T MR scanner (Philips Medical Systems), a volunteer study (n = 5) to assess cardiac and respiratory motion compensation strategies and the sensitivity of the motion encoding gradients; and a patient study (n = 8) comparing with traditional echocardiographic measures of diastolic dysfunction and serum NT proBNP.
Results: The tMRE sequence was successfully developed and implemented. Motion-encoded images (motion sensitized gradient at 165Hz) demonstrated myocardial wall shear waves associated with the aortic valve closure at 329ms (range 280-375ms) after the R wave. tMRE successfully showed a difference in speed of propagation between volunteers and patients (speed volunteers 12.6m/s ± 2.0m/s, speed patients 21.4m/s ±4.9m/s, p < 0.05, Figure 1). One patient who had invasive pressure volume loop studies was found to have increased left ventricular end diastolic pressure (LVEDP) and prolonged tau (represents the exponential decay of the ventricular pressure during isovolumic relaxation)–this patient had a faster speed of sheer wave propagation in tMRE. There was no clear correlation shown between the speed of propagation in patients and the degree of diastolic dysfunction (measured by degree of LVH, LA dilatation and E/E') but the Pearson coefficient with plasma NT-proBNP was 0.63 in patients presenting with acute heart failure.
Conclusions: We have successfully developed and applied a new technique to quantify myocardial stiffness in vivo from shear waves generated by aortic valve closure.
tMRE is patient friendly and does not require a transducer. We have shown that tMRE has potential to be an important diagnostic tool for the early detection of myocardial stiffness.
1571
Relationship of cerebrovascular reactivity and MRI pattern of carotid atherosclerotic plaque
Abstract
Objectives: The ultrastructure of carotid and coronary atherosclerotic plaques is now discussed among factors determining risk of future ishaemic events. The aim of our study was to compare in patients with carotid atherosclerosis the structure of atherosclerotic plaques imaged using MRI with cerebral vascular reactivity quantified by echo.
Methods: The patients population comprised 21 persons (as old as 64 ± 13) with hemodynamically significant carotid artery stenosis > 70 % of lumen, quantified with high-resolution MR imaging of arterial wall. To assess the cerebrovascular reactivity transcranial ultrasound of middle cerebral artery was carried out with quantification of blood flow parameters at rest and during the stress tests: first with breath holding and then with hyperventilation, with calculation of reactivity index.
From visual evaluation of MRI of carotid stenoses four types of plaques were assembled: mixed plaque with severe fibrosis and lipid core, plaque with lipid component, with intraplaque hemorrhage and calcified plaque.
To assess the patterns of autoregulative reaction to the stress test the following conventional patterns were used: positive reaction (reactivity index 1.1-1.4), negative reaction (reactivity index 0.9-1.1) and paradoxical reaction (reactivity index of less than 0.9).
Results: It was found that at breath holding hypercapnic test there is a statistically significant association between MRI plaque structure and type of the reaction (p = 0.081), whereas at the test with hyperventilation no significant differences were found. At breath hold test in 56 % of patients with MR images of plaque hypointense on T1-weighted scans and hyperintense on T2-weighted ones unidirectional positive response was observed; pathologic unidirectional negative response or multidirectional response were observed in patients with MR evidence of hemorrhagic plaques hyperintense on T1-weighted scans and hypointense on T2-weighted ones.
Conclusion: Atherosclerotic lesions of the internal carotid artery with intra-plaque hemorrhage are accompanied with critical deterioration of cerebrovascular reactivity. Probably any hemorrhage to carotid plaque could be assumed as critical risk factor and requires removal of the plaque by carotid endarterectomy. High-resolution MRI of carotid arterii may be suggested as screening test in all patients of risk groups of atherosclerosis.
1577
Feasibility study of an MR conditional pedal ergometer for cardiac stress MRI–preliminary results in healthy volunteers and patients with suspected coronary artery disease
Abstract
Background: Stress cardiac magnetic resonance imaging (CMR) is increasingly being used to assess chest pain in patients with known or suspected coronary artery disease and is most often performed by use of pharmacological agents such as adenosine and dobutamine. However, exercise stress testing provides a physiological challenge to the myocard and is generally regarded as superior to pharmacological stress testing. The purpose of this study was to evaluate the use of an MR conditional pedal ergometer for cardiac stress CMR in healthy volunteers as well as patients with suspected coronary artery disease.
Methods: All participants (10 healthy volunteers (45 ± 15, 10% women) and 11 patients (60 ± 8, 27% women) with known or suspected coronary heart disease) underwent exercise stress MRI at a whole-body 3 Tesla MR system. CMR protocol included first-pass perfusion (FPP) and real-time CINE imaging before and after exercise stress with a MR conditional pedal ergometer and late enhancement (LE) PSIR sequences after stress as well. Patients were examined in short term prior to elective coronary angiography (CAG). Image-quality regarding artifacts as well as visual analysis of left ventricular wall motion abnormalities (LVWMA) and FPP stress defects were rated by two experienced radiologists, functional parameters were determined from short-axis cine MR sequences before and after exercise stress.
Results: 88% of real-time Cine studies (90% of Rest-Cine images, 85% of Stress Cine images) were regarded with good to slightly impaired image-quality, whereas there was no significant difference of Cine image quality between patients and healthy volunteers (p > 0.05). Left ventricular ejection fraction (LV EF) improved from resting to stress conditions in both groups, healthy volunteers (LV EF 62 ± 6 to 65 ± 7 p =0.04) as well as patients (LV EF 55 ± 6 to 60 ± 7, p = 0.003), whereas EDV and ESV did not show significant alterations. 5/11 patients showed regional LVWMA on stress images. Out of these patients two patients get coronary stenting of corresponding vessels whereas one patient showed an instent restenosis at CAG with subsequent balloon-PTCA and another patient was conducted to bypass grafting. However, stress FFP sequences revealed only in two of the patients with LVWMA a subendocardial stress defect. LE sequences showed a subendocardial scar from silent myocardial infarction in one patient.
Conclusion: Cardiac stress CMR by the use of a MR conditional pedal ergometer is feasible with the ability to acquire images with good to slightly impaired quality. Exercise stress CMR provides physiological dynamic conditions to evaluate contractile reserve as well as detection of LVWMA. However, stress defects at FPP seem to fail a reliable presentation in this setting. Further improvements are necessary and confirmations in larger trials are needed.
1581
Pulmonary valve replacement for severe pulmonary stenosis has a positive effect on left ventricular remodeling
Abstract
Background: Right ventricular (RV) volumes and function play a key role in the decision-making on pulmonary valve replacement (PVR) in patients with pulmonary valve stenosis (PS) or regurgitation (PR). Although left ventricular (LV) function has been described to be equally important as RV function in these patient's symptomatology and prognosis, little is known about LV response to PVR. Therefore, we aim to evaluate LV remodeling after PVR in PS or PR patients.
Methods: We performed cardiovascular magnetic resonance imaging (CMR) before and after PVR in 18 consecutive patients; nine with severe PR (mean age at intervention 27 ± 16), and ten with severe PS (mean age at intervention 27 ± 8). Outcome parameters were change in LV end-diastolic (ΔLVEDV), LV end-systolic (ΔLVESV) volumes, and ejection fraction (ΔLVEF), before and after PVR.
Results: Most patients had congenital aortic stenosis, with previous Ross procedures (6 patients) or Tetralogy of Fallot (5 patients). Fifteen patients had undergone previous PVR (6 in the PR group; 9 in the PS group). After PVR, patients with severe PR showed only minor change in LVEDV, in LVESV and in LVEF (ΔLVEDV 1.5 ml/m2, p = ; ΔLVESV 0.1 ml/m2; ΔLVEF 1.2%; p = 0.7). However, there was an important trend in positive remodeling of the LV in patients who underwent PVR with severe PS (ΔLVEDV 8.5 ml/m2; ΔLVESV -1.1 ml/m2; ΔLVEF 5.1%; p = 0.08).
Conclusion: Patients who undergo PVR for PS seem to benefit more from biventricular remodeling compared to those who undergo PVR for PR. This superior LV remodeling could be due to superior remodeling of the pressure overloaded RV, with subseqent disburden of the LV, or to intrinsic characteristics of the LV, which was pre-operatively protected from overload by the PS.
1582
The RV after cardiac surgery, more resilient than thought: multiparametric quantification shows altered rather than reduced function
Abstract
Background: Right ventricular (RV) function declines after cardiac surgery, particularly long axis function. Three principal speculated causes are suggested: (a) loss of constraint when opening pericardium; (b) intra-operative ischemia (c) post-operative tethering. The impact of relative contributors of RV function (AV plane descent, apical movement) is obscure. The complexity of cardiac operation is also thought to contribute to RV impairment. We sought to untangle these factors by comparing pre- and post-surgery RV function changes in a cohort of patients with aortic valve replacement (AVR) and 1/3 had in addition grafting (CBAG).
Method: 81 patients were scanned prior to, and 1 year following AVR (n = 59) or combined AVR and CABG (n = 22). RV volume and function were calculated with standard techniques. RV longitudinal function was assessed as follows, Figure 1a.
• A - Tricuspid annular displacement: TAD, distance measured between tricuspid annulus in diastole and systole
• B - TAPSEout, distance between tricuspid annulus and a fixed point on chest wall in diastole and systole
• C - RV apical motion, distance between the RV apex and a fixed point on the anterior chest wall in systole and diastole.
• D- TAPSEin, distance between tricuspid annulus and RV apex in diastole and systole (includes apical motion)
Results: RV volumes and ejection fraction were in normal limits, pre and post, Table 1.
RV longitudinal function however assessed using tricuspid annular displacement and TAPSEout declined after surgery by 41% (p < 0.0001) and 43% (p < 0.0001) respectively. Apical motion increased post operatively by 52% (p < 0.0001), meaning that overall TAPSEin only fell by 21% (p < 0.0001), Figure 1b.
RVEF had weak correlation with each of the markers of RV longitudinal function, (r = 0.30 for TAD, p < 0.01) and considering a value of TAD <16mm alone, 84% of patients were classified as RV dysfunction post-operatively, despite preserved RVEF. There were no significant differences between the AVR + CABG and AVR alone groups in any of the measures of RV function (RVEF: 65% ± 9% versus 64% ± 9%, p = 0.88) respectively.
Conclusion: The effect of cardiac surgery on RV function is overstated, tricuspid annular displacement falls, but apical basal contraction increased in compensation. Doing additional grafts has no incremental detrimental effect on RV function. Ischaemia is not the cause of changes - tethering and/or loss of pericardial constraint appear to be the principal mechanisms of RV change.
Figure 1. Right Ventricle after Cardiac Surgery: Methods and change in function.
1584
Usefulness of cardiovascular magnetic resonance to differentate coronary artery disease from non ischemic cardiomyoptathy in patients with heart failure
Abstract
Introduction: We evaluated the capacity of cardiovascular magnetic resonance (CMR) to distinguish between underlying coronary artery disease (CAD) from non ischemic cardiomyoptathy in patients (pts) with left ventricular (LV) dysfunction.
Methods: We prospectively studied 140 consecutive patients with heart failure and an ejection fraction (EF) ≤40% assessed by echocardiogrpahy referred for stress CMR (1.5 Tesla). Standard protocol consisted of: 1) assessment of myocardial function at rest; 2) pharmacological stress induced either by dobutamine or by adenosine; 3) assessment of late gadolinium enhancement (LGE) sequences.
Results: All tests were performed in 140 patients, 114 males (81%), mean age 64.42 ± 12.6 years. Stress was induced using adenosine in 98 patients (70 %) and with dobutamine in 42 patients (30%). Mean duration with adenosine was 46 ± 8 minutes and with dobutamine 58 ± 8 minutes. LVEF was 30.8 ± 8.6%. Myocardial ischemia was present in 20 pts (14%). Subendocardial or transmural LGE was found in 90 (64%) pts, corresponding to a previous myocardial infarction (MI). Intramoycardial LGE was only seen in 5 pts (4%), in 3 with dilated cardiomyopathy (DCM) and in 2 with myopericarditis. No LGE was observed in 35 pts (25%) with DCM.
Conclusion: With the use of CMR we could clearly distinguish pts with CAD from those with non ischemic cardiomyopathy. Of note, a high percentage of pts with DCM had no LGE, suggeting a potential recuperation of the LV function.
1593
What does CMR add to the ESC Risk Prediction Model to Assess the Occurrence of Sudden Cardiac Death in Patients with HCM?
Abstract
Objective: Hypertrophic cardiomyopathy (HCM) is a complex genetic heart disease. Although the overall risk for sudden cardiac death is relatively low some patients with HCM die suddenly from fatal arrhythmic events. The recently published ESC clinical risk prediction model uses clinical parameters to estimate the probability of sudden cardiac death (SCD) at 5 years. Since this risk prediction model does not take into account CMR parameters, the aim of our study was to compare patients with low, intermediate and high ESC risk scores of SCD according to CMR characteristics.
Methods: In 149 HCM patients the ESC risk score was retrospectively calculated according to the following formulas: Prognostic index = [0.15939858 x maximal wall thickness(mm)] - [0.00294271 x maximal wall thickness2 (mm2)] +[0.0259082 x left atrial diameter (mm)] + [0.00446131 x maximal(rest/Valsalva) left ventricular outflow tract gradient (mm Hg)] +[0.4583082 x family history SCD] + [0.82639195 x NSVT] +[0.71650361 x unexplained syncope] - [0.01799934 x age at clinical evaluation (years)] and probability of SCD at 5 years =1–0.998exp (Prognostic index). CMR characteristics were compared among patients with low (risk of SCD < 4%), intermediate (risk of SCD 4- < 6%) and high risk (risk of SCD ≥ 6%).
Results: Of 149 patients with HCM 121 (81%) belonged to the low, 18 (12%) to the intermediate and 10 (7%) to the high risk group. The probability of SCD at 5 years was 2 ± 1% in the low risk, 5 ± 1% in the intermediate and 12 ± 8% in the high risk group. Comparing CMR characteristics, left and right ventricular EF as well as left and right ventricular volumes were comparable among the different risk groups. However, patients with a high ESC risk score revealed a significantly higher extent of fibrosis defined as extent of LGE 29 ± 22 compared to patients with an intermediate risk score and an exent LGE % of 12 ± 12 (p = 0.02) and patients with a low risk score and an extent LGE % of 10 ± 11% (p < 0.001).
Conclusion: Patients with a high clinical risk of SCD according to the ESC clinical risk prediction model reveal a higher extent of LGE compared to patients with HCM and lower risk scores. Therefore, the extent of LGE seems to be an additional risk marker in these patients.
1597
Detecting Progression of Diffuse Interstitial Fibrosis in Alstrom Syndrome
Abstract
Introduction: Alstrom Syndrome (ALMS) is a rare inherited disorder caused by mutation in ALMS1 gene. The syndrome is a multi-system disorder with exaggerated features of metabolic syndrome and although rare, provides a monogenic model for end-organ fibrosis and as a paradigm for the effects of severe metabolic syndrome. Adults with ALMS have a high risk of death from heart failure in their twenties due to a cardiomyopathy which (in small post-mortem series) is characterised by coarse fibrosis on histology. Our previous work has identified expansion of the extracellular space (ECV) consistent with diffuse interstitial fibrosis in over half of asymptomatic ALMS patients compared to controls. The aim of this study was to investigate the longitudinal change in ECV and assess the impact on ventricular structure and function.
Methods: A prospective longitudinal cohort study of patients attending the national service for ALMS at the Centre for Rare Disease in Birmingham from 2012. At referral and on annual follow up, all subjects underwent comprehensive LV and RV assessment with cardiac MRI (CMR 1.5T Siemens Avanto). The presence of diffuse interstitial myocardial fibrosis was assessed using native myocardial T1 relaxation mapping and extracellular volume (ECV) in the LV septum (MOLLI) using cvi42® (Circle Cardiovascular Imaging). Coarse replacement fibrosis was assessed using standard late gadolinium enhancement imaging.
Results: In total 14 patients (male gender 71%, age 28 ± 8years) had baseline and follow up data (median 1.7 [1.1-2.8] years). CMR data is presented in Table 1. The native LV myocardial T1 values and ECV were increased in the septum at basal and mid levels at follow up. Left ventricular mass increased (54 ± 9g/m2 vs. 62 ± 12g/m2) but with a reduction in septal myocardial intracellular volume (ICV 0.74 ± 0.06 vs. 0.68 ± 0.04, p < 0.05) suggesting ECV expansion rather than myocyte hypertrophy was the driver. There were no differences in LV or RV volumes or RVEF. Four patients had LGE; two patients had focal at RV insertion points LGE and two patients had mid-wall LGE in the basal infero-lateral segments.
Conclusion: ALMS is associated with increases in ECV and progressive change in T1 values over time that reflects progression of diffuse interstitial fibrosis in asymptomatic adults. Cross-sectional studies have identified ECV as a biomarker of cardiovascular “vulnerability” but longitudinal tracking has the potential to highlight those at greatest risk.
1612
Diffuse fibrosis in the ventricles of patients with transposition of great arteries late after atrial switch
Abstract
Objectives: In adult patients with transposition of great arteries late after atrial switch (Mustard or Senning), the subaortic morphologic right ventricle (RV) is hypertrophic, while the subpulmonary morphologic left ventricle (LV), is usually hypotrophic.
The extent of diffuse fibrosis in the RV and LV in these patients remains unclear. Therefore, the aim of this study was to determine myocardial extracellullar volume (ECV) in both ventricles in these patients.
Methods: We determined ECV by cardiac magnetic resonance (CMR) in 10 patients (36.8 ± 5.3 years old), without relevant pulmonary stenosis, late after atrial switch by acquiring T1-maps of the myocardium before and 10 minutes after injection of Gadolinium-based contrast agent.
Results: ECV of the LV was significantly increased compared to the RV (LV 36 ± 4% vs. RV 27 ± 3%, p = 0.05).
Conclusions: In patients late after atrial switch, ECV of the hypotrophic, subpulmonary morphologic LV is significantly increased compared to ECV of the hypertrophic, subaortic morphologic RV. ECV of the RV is in the upper normal range. Increased ECV in the LV may be due to diffuse fibrosis, induced by long time reduced activity of LV, or simply due to relatively higher ECV surrounding hypotrophic myocardial cells of the LV.
1631
Utility of Cardiac Magnetic Resonance in the diagnosis and stratification of arrhythmic risk in patients with confirmed or suspected ventricular arrhythmias
Abstract
Purpose: The etiological diagnosis of cardiac arrhythmias is often difficult. Cardiac Magnetic Resonance (CMR) as gold standard in anatomical and functional cardiac evaluation can be used as a fundamental technique in the diagnostic evaluation of patients (P) with cardiac arrhythmias. The purpose of the study is to determine the impact of CMR in the diagnosis and stratification of arrhythmic risk in P with confirmed or suspected ventricular arrhythmias.
Methods: Prospective study during 3 years where we evaluate all P who underwent CMR for etiological clarification of aborted sudden cardiac death, ventricular arrhythmias, suspected structural heart disease with high arrhythmic potential or unexplained recurrent syncope. We excluded all P who already had etiological diagnosis by other diagnostic methods. Acquisitions were performed in all P to evaluate anatomically and functionally the right and left ventricles and to look for the presence of late enhancement.
Results: A total of 54 P were evaluated, of which 32 were male. The mean age was 45,8 ± 16,58 years. The indications for the CMR were: aborted sudden cardiac death due to ventricular fibrillation - 4 P; sustained Ventricular Tachycardia (VT) - 8 P; non-sustained VT - 3 P; very frequent ventricular extrasystoles - 25 P; suspected structural heart disease with high potential arrhythmic–5 P and unexplained syncope - 9 P.
In only one patient it was not possible complete the CMR due to gating artifacts conditioned by the presence of cardiac arrhythmias and respiratory synchrony. In 7 P, we verified the presence of gating artifacts conditioned by the presence of cardiac arrhythmias which precluded the accurate measurement of ventricular volumes, but did not affect the diagnostic accuracy. Overall, CMR was normal in 39 P, in 5 P we found nonspecific changes deserving a future control and in 10 P it was possible to establish a definitive diagnosis which was unknown before the realization of the CMR. The diagnoses established were: arrhythmogenic right ventricular dysplasia - 2 P, left ventricular non compaction - 3 P, silent myocardial infarction sequel–2 P; acute myocarditis–1 patient; hypertrophic cardiomyopathy–1 patient; and amyloidosis–1 patient.
Conclusions: CMR is a technique with high spatial resolution, feasible and safe which allowed the diagnosis and risk stratification in 17% of our study population of patients with high arrhythmic potential when the first-line tests are normal.
1635
Size matters: pulmonary veins geometry by cardiac magnetic resonance imaging in atrial fibrillation patients
Abstract
Background: The study of pulmonary vein (PV) anatomy is a crucial step prior to successful ablation therapy for atrial fibrillation (AF).
PV anatomy is variable in the number of independent veins draining to the LA, in their ostial shape and course, and this can be visualised with Cardiovascular Magnetic Resonance (CMR). PV anatomy was investigated in a small series of consecutive patients referred for a clinical CMR exam before ablation therapy for AF.
Methods: 10 consecutive patients aged 56 ± 11.6 years old were referred for a pre-procedural clinical CMR from Oct 2015 to Jan 2016.
Imaging was performed using a 3 Tesla scanner (Siemens Avanto, Siemens, Erlangen, Germany). Magnetic resonance angiograms (MRAs) were obtained with a breath-hold 3D fast spoiled gradient-echo imaging sequence in the coronal plane, after the administration of gadolinium contrast agent. All images were analyzed by a single investigator. The ostial cross-sectional diameters, area and perimeter were measured for each PV: left upper (LUPV), left lower (LLPV), right upper (RUPV) and right lower (RLPV). Eccentricity index (a measure of the cross-sectional shape of the vein) was calculated as a ratio maximum over minimum ostial diameter, and defined as follows: round (≤ 1.2);; oval (>1.2 and ≤1.4);; flat (>1.4).
Results: PVs measurements are presented in Table 1. Left PVs were more often flat (40% LUPV, 70% LLPV) or oval (40% LUPV, 30% LLPV), with a minimum diameter in the antero-posterior direction. Inversely, right PVs were rounder (80% RLPV;; 50% RUPV) or oval shaped (20% RLPV;; 30% RUPV). LLPV was significantly different from RLPV in all the measurements (p < 0.05), and from RUPV in all the measurements except for the maximum diameter. No significant differences were found between LLPV and LUPV.
Conclusions: The cross-sectional appearance of the left lower pulmonary vein differs significantly from that of the other veins. It often appears elliptically deformed which appears to be related to an interposition between the descending aorta and the left atrium. As both anatomical structures are under pressure in arterial hypertension, the relationship between atrial fibrillation and arterial hypertension may in part be related to this deformation. The prognosis of atrial fibrillation may also be related to this anomaly. Thus, further studies comparing AF population with a control group of healthy subjects are needed to better investigate a possible pathogenetic implication of our finding.
Figure 1. Example of LLPV course.
1642
How do the differences in Left Ventricular wall measurements from Echocardiography and CMR in patients with Hypertrophic Cardiomyopathy affect current Sudden Cardiac Death Risk Scores?
Abstract
Background: Hypertrophic Cardiomyopathy (HCM) is defined by the presence of increased Left ventricular (LV) wall thickness not solely explained by abnormal loading conditions. All studies investigating the prognostic value of LV hypertrophy (LVH) in HCM and the recent ESC Sudden Cardiac Death (SCD) risk stratification have used echocardiography. Studies have suggested improved accuracy measuring LV wall thickness using cardiac magnetic resonance (CMR) due to improved spatial resolution. In our institution patients diagnosed with HCM are referred for both echocardiography and CMR. Their annual risk stratification usually involves echocardiography.
Objectives: Objectives of the study were to assess the presence of any systematic difference with echocardiography and CMR in patients with HCM and to assess these results on patients' SCD risk scores.
Methods: Maximal LV wall thickness measurements were recorded by four operators on 50 consecutive patients referred for echocardiography and CMR. Maximal LV wall measurements were taken at 12 points in each patient.
Results: There were 600 CMR LV wall thicknesses, compared to only 421 that could be measured on echocardiography and for direct comparison only matched data was used. There was excellent agreement on the location of maximal LV wall thickness and the 50 measures of myocardial maximal LV wall thickness were higher on echo that CMR (echo mean 1.9cm, SD 0.5; CMR 1.6, SD 0.3 (p < 0.01)). The Bland Altman plots confirmed improved inter-observer reliability with CMR than echo (Figure 1A and B).
The SCD risk score was calculated for all 50 patients and using the different maximal wall thicknesses from echo and CMR, there was a mean difference in 5-year risk SCD of 0.49%, SD 0.45 (p = 0.37) (Figure 2).
Conclusion: Echocardiography systematically over reports maximal LV wall thickness in patients with HCM, when compared to CMR. Despite this, when LV maximal wall thickness results are entered into established SCD risk models, there is no statistically significant difference. Clinicians need to be aware of this disparity between imaging modalities.
Figure 1. A: Bland Altman Plot for LV wall thickness measurements measured by echocardiography. B: Bland Altman Plot for LV wall thickness measurements measured by CMR.
Figure 2. Bland Altman Plot for the difference between ESC SCD risk score for patient using echocardiography and CMR
1651
Noninvasive assessment of intracardiac viscous energy loss in Fontan patients from 4D Flow CMR
Abstract
Introduction: In the normal heart, ventricular viscous energy loss (EL) is the kinetic energy that is lost due to frictional forces between blood and surrounding structures in the ventricle and induced by blood viscosity.[1] Patients with a Fontan circulation have abnormal morphology of the atrioventricular (AV) valves and ventricles, which may promote inefficient flow-structure interaction that can increase viscous energy loss. The purpose of this study was to noninvasively assess intracardiac EL in Fontan patients with various underlying pathologies 4D flow CMR compared to the left ventricle in healthy volunteers.
Methods: 11 patients with a Fontan circulation (age 15.0 ± 5.1 years) and 11 volunteers (15.4 ± 5.7 years) underwent whole-heart 4D Flow CMR at 3T (VENC =150 cm/s, spatial resolution 3 × 2.6 × 3 mm3, 30 retrospectively reconstructed phased over one cardiac cycle). The systematic ventricle in patients and left ventricle (LV) in volunteers were segmented and EL was computed over diastole (EL_diastole), systole (EL_systole) and total cardiac cycle (EL_total), all in Joule, using the Navier-Stokes energy equations [1]. Measured EL was then normalized by stroke volume (m3). Normal range of LV viscous energy loss was derived from volunteers as the 95%CI (confidence interval). Patients' ventricular EL parameters were classified as: below, within or above 95%CI. An independent samples T-test was used to compare EL_total between both patients and controls.
Results: Patients showed significantly higher EL_total than controls (mean ± SD, 8.8 ± 3.4 vs. 4.5 vs. 0.84, P-value 0.001). Detailed results are shown in Figure 1. In three patients, EL_total was in the normal range (95% CI 3.34-6.04): two patients with hypoplastic left heart syndrome (HLHS) and one patient with tricuspid atresia (TA) with concordant VA connection. EL_diastole was in the normal range in four patients (95% CI 2.05-3.68): three HLHS patients and one patient with transposition of the great arteries (TGA). One patient with TA with concordant VA connection was below the normal range. In two patients, EL_systole was in the normal range (95% CI 0.98-2.57): one HLHS patient and one TA patient with concordant VA connection. Highest values of different EL parameters were consistently found in a TA patient with discordant VA connection.
Conclusion: Patients with a Fontan circulation show elevated ventricular viscous energy loss compared to healthy LV.
Reference:
1. Elbaz et al. MRM 2016 (in press)
Figure 1. Viscous energy loss in patients with a Fontan circulation.
1653
Behcet and Myocardial Infarction: A Rare Combination
Abstract
Introduction: Myocardial infarction during Behçet disease is rare. And may be the first manifestation.
Case Report: A 45-years-old man with a history of angio-behcet since 2009 and ischemic heart disease treated by interventricular artery stenting in 2008, was admitted with signs of congestive heart failure. The electrocardiogram showed an inverted T waves in lateral leads. Echocardiography demonstrated left ventricular dilatation with akinetic movement of the septum and the anterior wall. It also revealed a severe left ventricular dysfunction. A cardiac MRI was realized to differentiate beteween specific inflammatory myocardial damage and myocardial damage resulting from a coronary artery involvement, this technique showed a transmural late enhancement in the territory of the interventricular artery and part of a territory of the circumflex artery. These data was confirm by the coronary angiography that demonstrated a intra stent occlusion with a good down stream bed. Considering the patient history of Behcet a medical treatment was required including colchicine and heart failure traitement.
Discussion: Coronary involvement is exceptional and predominantly affects males with an average age of 40.The lesion is frequently proximal, located on one coronary artery and it primarily affects the interventricular artery.
The diagnosis of coronary vasculitis should be considered in a yong patient without traditional risk factor of atherosclerosis.
About 20 cases of myocardial infarction were reported in the litterature but the etiopathogeny, the causal relationsheep and the treatment are yet unknown.
Conclusion: Our patient highlight the importance of MRI in the distinguishing between specific inflammatory myocardial damage and myocardial infarction thereby providing a better medical decisions.
1328
Impact of New Cerebral Ischemic Lesions On the Occurrence of Delirium after Transcatheter Aortic Valve Implantation
Abstract
Aims: To investigate the impact of new cerebral ischemic injury evaluated by diffusion weighted magnetic resonance imaging (DWI) on the occurrence of postoperative delirium (POD) after transcatheter aortic valve implantation (TAVI).
Methods: All TAVI patients participating in prospective studies that involved postprocedural cerebral DWI were identified. Periprocedural, clinical and imaging data had been prospectively collected. Cerebral DWI was performed within 5 days after TAVI. Quantification of cerebral injury was performed by measurement of the number and volume of lesions. The occurrence of POD was assessed using the Diagnostic and Statistical Manual of Mental Disorder 4rd Edition criteria.
Results: Post-procedural cerebral DWI was performed in 103 patients. A total of 673 new DWI-lesions (IQR: 5[2-8]) were detected in 94 (91%) patients after TAVI. POD occurred in 15 (14.5%) patients, and 47% (n = 7) of these cases were associated with at least one peri-procedural complications, including permanent pacemaker implantation (n = 2), atrial fibrillation (n = 5), acute kidney injury stage II/III (n = 1), major vascular or major/ life-threatening bleeding (n = 3). POD was associated with higher number of DWI-lesions regardless of complications (in uncomplicated TAVI: 9[7-24] vs.4[2-7], p = 0.008; and in complicated TAVI: 7[2-9] vs.5[1-7], p = 0.611) and longer hospital-stay (6[5-8] vs. 5[4-6], p = 0.038). Factors independently associated with POD were higher number of DWI-lesions (Odds Radio [OR] 1.09; 95% confidence interval [CI] 1.02-1.17), pre-existing cerebral global atrophy (OR 4.72; 95% CI 1.06-21.11) and logistic EuroSCORE (OR 1.09; 95% CI 1.01-1.18).
Conclusions: Higher number of cerebral ischemic lesion detected with DWI is associated with the occurrence of postoperative delirium after transcatheter aortic implantation. Patients with postoperative delirium showed longer hospital stay.
1329
Heart T2* assessment to measure iron overload using different software tools in patients with Thalassemia Major
Abstract
Aims and Objectives: To evaluate the correlation and the interchangeability between a generic mapping T2* and specific protocols T2* in the myocardium with iron overload (IO) in patients with Thalassemia Major (TM).
Methods and Materials: Within Myocardial Iron Overload in Thalassemia (MIOT) project database we extract 60 TM patients with global heart T2* values covering entire clinical range: 20 without IO, 20 with moderate IO, 20 with severe IO. Midpapillary short axis GRE multi-echo images of the left ventricle acquired by a GE 1.5T scanner were processed by a general purpose T2* mapping tool (SEGMENTS T2* module 1.9, Medviso, Lund, SE) and by a dedicated tool validated in IO measurements in TM (HIPPO MIOT, FTGM, Pisa, IT). T2* value in the midventricular septum was evaluated by manually tracing a ROI in the T2* map in SEGMENT and by averaging the T2* values in segments 8 and 9 in HIPPO MIOT. Manual truncation of later echoes was used in HIPPO MIOT in patients with severe IO if needed.
Results: Table shows the paired-t-test analysis results for measurement in the midventricular septum. A significant difference between the two programs was detected in patients with severe IO, leading to a significant difference in the whole population. The Bland-Altman plot confirmed statistical analysis findings.
Conclusions: The two software tools are not interchangeable in measurement of low T2* values: different image analysis algorithms are exploited in the two tools, ROI-based mono-exponential fitting with manual truncation in HIPPO MIOT and pixel-wise hybrid fitting method in SEGMENT.
1332
Hypertrabeculated Left Ventricle at Cardiac Magnetic Resonance Imaging: β-Thalassemia Major vs. Left Ventricular Non -Compaction Disease
Abstract
Background: Cardiac magnetic resonance (CMR) is used to differentiate left ventricle noncompaction (LVNC) from other pathological and physiological conditions characterized by prominent LV trabeculae. The hypertrabeculated LV reported in β-thalassemia major (β-TM) patients has not been considered so far as a potential differential diagnosis. We tested the diagnostic accuracy of CMR imaging in differentiating the hypertrabeculated left ventricle (LV) in patients with β-TM from LVNC disease.
Methods: CMR cine images were analyzed in 10 patients previously diagnosed with LVNC and in 38 patients with β-TM. Two CMR diagnostic criteria were applied at end-diastole: a ratio of noncompacted to compacted myocardium (NC/C ratio) >2.5 at a segmental level and a percentage of trabeculated LV mass >20% of global LV mass.
Results: In β-TM patients at least 1 positive NC/C segment was found in 51% of cases. Compared with LVNC patients, in β-TM patients the non-compaction areas were less frequent (7% vs. 37% of overall myocardial segments, p < 0.0001), with a similar distribution within the LV (predominant at the apex and postero-lateral wall, uncommon at the septum) precluding differential diagnosis. To distinguish LVNC from β-TM patients, a NC/C ratio of >2.5 showed low specificity (58%) whereas a trabeculated LV mass >20% was more accurate (sensitivity 100%, specificity 87%). Best specificity (92%) was obtained with a trabeculated LV mass percentage of >26%.
Conclusions: Differentiation of LVNC from hypertrabeculated LV in β-TM patients may depend on the selected CMR criterion. In our patient population, trabeculated LV mass percentage showed to be better than NC/C ratio.
1335
Aortic Regurgitation following Transcatheter Aortic Valve Implantation (TAVI): a CMR Study of two prosthesis designs
Abstract
Background: Aortic regurgitation following TAVI is an independent predictor of mortality in the postoperative period and has prompted development of novel bioprostheses designed to facilitate precise deployment and minimise paravalvular regurgitation. CMR offers full volumetric quantitation independent of the number or eccentricity of regurgitant jets.
Aim: To compare the degree of early post-procedure aortic regurgitation following implantation of the Medtronic CoreValve and Boston Scientific Lotus Valve systems in patients treated for severe symptomatic aortic stenosis.
Methods: All patients underwent an identical 1.5T CMR scan (Intera, Phillips Healthcare, Best, Netherlands) post-procedure prior to discharge. Through-plane phase contrast velocity encoded imaging was performed perpendicular to the aortic valve jet at a level above the aortic prosthesis. Aortic flow was subsequently quantified offline (QFlow version 7.2, Medis Medical Imaging Systems, Leiden, The Netherlands) and regurgitant fraction (%) derived. AR was defined as regurgitant fraction of none/trivial ≤5%, mild 6-15%, moderate 16-25%, moderate-severe 26-48%, and severe >48% The invasive AR index derived at the time of TAVI was also recorded ([(Diastolic BP-LVEDP)/Systolic BP] x100).
Results: 18 Medtronic CoreValve patients (age 79.6 ± 6.3years, 67% male, EuroSCORE II 4.59 ± 3.59%) and 24 Boston Lotus TAVI patients (age 78.6 ± 8.7years, 54% male, EuroSCORE II 4.07 ± 3.37%) were studied at a median interval of 6 days following implantation. The residual peak aortic pressure gradient observed following Boston Lotus implantation was higher than that following Medtronic CoreValve (25.8 ± 12.1 vs.15.0 ± 5.5mmHg, p = 0.001). The absolute regurgitant fraction was significantly lower following Boston Lotus implantation than following Medtronic CoreValve (Lotus: 4.0 ± 3.5% vs. CoreValve: 11.7 ± 7.2%, p = 0.001). There was a negative correlation between invasive AR index and regurgitant fraction from CMR imaging (Spearman's coefficient -0.4, p = 0.016).
Conclusions: CMR directly quantifies AR following TAVI and there is agreement with invasive AR measurements. Significantly less aortic regurgitation but a higher residual peak pressure gradient was observed following the Boston Lotus valve compared with the Medtronic CoreValve. Further work is required to assess whether this translates into a mortality benefit.
1336
Incremental value of semi-quantitative evaluation of myocardium perfusion with 3T stress cardiac MRI
Abstract
Background: 3T MRI has been adopted by some centers as the primary choice for assessment of myocardial perfusion over conventional 1.5T MRI. However, there is no data published on the potential additional value of incorporating semi-quantitative data from 3T MRI. This study sought to determine the performance of qualitative 3T stress magnetic resonance myocardial perfusion imaging (3T-MRMPI) and the potential incremental benefit of using a semi-quantitative perfusion technique in patients with suspected coronary artery disease (CAD).
Methods: Fifty eight patients (41 men; mean age: 59 years) referred for elective diagnostic angiography underwent stress 3T MRMPI with a 32-channel cardiac receiver coil. The MR protocol included gadolinium-enhanced stress first-pass perfusion (0.56mg/kg, dipyridamole), rest perfusion, and delayed enhancement (DE). Visual analysis was performed in two steps. Ischemia was defined as a territory with perfusion defect at stress study but no DE or a territory with DE but additional peri-infarcted perfusion defect at stress study. Semi-quantitative analysis was calculated by using the upslope of the signal intensity-time curve during the first pass of contrast medium during dipyridamole stress and at rest. ROC analysis was used to determine the MPRI threshold that maximized sensitivity. Quantitative coronary angiography served as the reference standard with significant stenosis defined as >70% diameter stenosis. Diagnostic performance was determined on a per-patient and per-vessel basis.
Results: Qualitative assessment had an overall sensitivity and specificity for detecting significant stenoses of 77% and 80%, respectively. By adding MPRI analysis, in cases with negative qualitative assessment, the overall sensitivity increased to 83%. The impact of MPRI differed depending on the territory; with the sensitivity for detection of left circumflex (LCx) stenoses improving the most after semi-quantification analysis, (66% versus 83%).
Conclusions: Pure qualitative assessment of 3T MRI had acceptable performance in detecting severe CAD. There is no overall benefit of incorporating semi-quantitative data; however a higher sensitivity can be obtained by adding MPRI, especially in the detection of LCx lesions.
1343
Left ventricular morphological quantification with single shot and free-breathing SSFP cine imaging compared with standard breath-hold SSFP cine imaging
Abstract
Introduction: Multiple breath hold cine balanced steady state free procession (SSFP) imaging is the reference standard technique for the cardiovascular magnetic resonance (CMR) assessment of left ventricular mass (LVM), volumes and ejection fraction (LVEF). Single shot (SS) cine imaging is an alternative for left ventricular (LV) volumetric assessment in a shorter acquisition time, but still requires several breath-holds. Modern SSFP imaging allows excellent image quality, even when free breathing (fbSSFP). We evaluated the feasibility and accuracy of fbSSFP with signal averaging and SS imaging for LV quantitation in comparison with the reference technique of multiple breath-hold SSFP.
Methods: We prospectively recruited 26 consecutive patients (male n = 18(69%),age 49 ± 21yrs) attending for evaluation of various cardiac conditions using a 1.5T Philips Ingenia scanner. Multi-slice, multi-phase cine imaging using standard SSFP with retrospective gating was used to cover the LV in short axis (typically: 12 slices, 8mm thick, 2mm gap, 30 phases, TR/TE 4.4/2.2ms, flip angle 60°, temporal resolution 27ms, 1.1x1.1mm in-plane resolution). The identical pulse sequence was repeated with 4 signal averages during free-breathing (fbSSFP) followed by a SS-acquisition (15 phases, temporal resolution 118ms, 1.5x1.5mm in-plane resolution,1-2 breath-holds; all phases acquired in 1 heartbeat). LV end diastolic volume (LVEDV), LVM & LVEF were quantified by manually contouring endo and epicardial borders using dedicated computer software(cmr42 ). Bland-Altman analysis was used to assess the agreement of SS and fbSSFP with SSFP. Percent agreement of SS and fbSSFP with SSFP was calculated by dividing the SD of the difference between measurements by the mean. Mean differences were compared using Student's t test.
Results: All 78 LV stack acquisitions were of suitable image quality (Fig 1). Results can be seen in Table 1 and Fig 1. Agreement for fbSSFP vs SSFP was 94.5%, 93.6% and 93.8% for LVEDV, LVM and LVEF respectively. Agreement for SS vs. SSFP was 92.3%, 90.5% and 93.2% for LVEDV, LVM and LVEF respectively. Differences in quantitation led to a change in LVEF category (by one grade) in 2 patients for the fbSSFP acquisition and 3 patients in the SS acquisition group.
Conclusion: Both SS and fbSSFP imaging allow diagnostic quality imaging with good reproducibility and agreement with the reference standard SSFP. These two techniques may be helpful in patients unable to breath-hold.
1344
Changes of cardiac iron and function during pregnancy in transfusion-dependent thalassemia patients
Abstract
Objectives: Our aim was to assess the changes in cardiac and hepatic iron overload and in morpho-functional cardiac parameters by Magnetic Resonance Imaging (MRI) in transfusion-dependent (TD) thalassemia patients who got pregnant and interrupted their chelation treatment.
Methods: We considered 17 women (14 with thalassemia major and 3 with thalassemia intermedia) enrolled in the Myocardial Iron Overload in Thalassemia (MIOT) project who had a pregnancy with successful delivery and who performed a MRI scan before and after the pregnancy. The T2* technique was used for iron overload assessment. For the heart, a multislice multiecho approach was used. Liver T2* was converted into liver iron concentration (LIC). Atrial areas and biventricular function were quantified in a standard way by cine images.
Results: The pre-pregnancy MRI was performed 15.02 ± 5.31 months before the delivery and the post-partum MRI 5.73 ± 4.45 months later. For 16 new-mothers the post-partum MRI was performed after the restart of the chelation therapy (3.95 ± 4.10 months later). One new-mother performed the post-partum MRI 3 months before restarting the chelation therapy. The table shows the MRI parameters at the two MRIs. The pre-pregnancy and the post-partum global heart T2* values and number of pathological segments were comparable. Two patients with a normal global heart T2* value (>20 ms) before pregnancy showed a pathological post-partum value. After pregnancy there was a significant increase of MRI liver iron concentration (LIC) values. At the pre-partum MRI six (35.3%) patients had a MRI LIC < 3 mg/g/dw while at the post-partum MRI all patients had a pathological MRI LIC.Among the biventricular function parameters, there was a significant increase of right ventricular (RV) end-systolic volume index and a significant reduction of RV ejection fraction.
Conclusion: In TD patients cessation of chelation therapy can cause rapid iron overload. Pregnant women with thalassemia should be monitored carefully for iron loading and cardiac status before they embark upon a pregnancy and afterwards and consideration should be given to offering desferrioxamine immediately after delivery. In women showing severe iron overload before pregnancy desferrioxamine should be started after the middle of the second trimester. The negative impact on the RV parameters could reflect the effect of the high cardiac output state independent of the physiological changes during pregnancy.
1346
Significant improvement of survival by T2* MRI in thalassemia major
Abstract
Introduction: In 2004 seven Italian centers reported survival data for patients with thalassemia major (TM) and showed that heart disease due to iron overload was the most common cause of death. In the same years the accurate and noninvasive assessment of cardiac siderosis was made possible in Italy by the introduction of the T2* cardiovascular magnetic resonance (CMR). We aimed to evaluate if the deployment of T2* CMR had an impact on the mortality rate.
Methods: Four centers contributed to the present study, updating the data of the enrolled patients until August 31, 2010. For the patients who died, the date of the death represented the end of the study. 577 patients (264 females and 313 males) were included. All patients were born on or after January 1, 1960 and mean age at the follow-up was 28.04 ± 10.88 years.
Results: One-hundred and fifty-nine (27.6%) patients died, 124 of whom (77.9%) died before the year 2000.
Dead patients were significantly younger and they were more frequently males. Dead patients started chelation therapy significantly later. Dead patients showed an higher frequency of HIV, arrhythmias and heart failure. According to the Cox model, the following variables were identified as significant univariate prognosticators for the death: male sex (HR = 1.87, 95%CI = 1.34-2.60, P < 0.0001), HIV (HR = 2.55, 95%CI = 1.25-5.20, P = 0.010) and heart failure (HR = 8.86, 95%CI = 6.37-12.31, P < 0.0001).
MRI was not performed in 406 patients (70.4%) and no patient had been scanned before his/her death. Among the survivors, MRI was not performed in the 59% of the cases (P < 0.0001). The absence of an MRI scan was a significant univariate prognosticator for death (HR = 43.25, 95%CI = 11.32-165.33, P < 0.0001).
The study was restricted to the patients dead after 2004 (19/159 = 12%) or followed until August 2010 (N = 357). In this subgroup of 376 patients, MRI was not performed in the 52.4% of the survivors and in all dead patients (P < 0.0001). The absence of a MRI exam was reconfirmed as a strong predictive factor for death (HR = 49.37, 95%CI = 1.08-2263.24, P = 0.046). The Kaplan–Meier curve is showed in the Figure. The log-rank test revealed a significant difference in the curves (P < 0.0001).
Conclusions: Our data suggests that the use of T2* CMR, that enables individually tailored chelation regimes reducing the likelihood of developing decompensated cardiac failure, allowed the reduction of cardiac mortality in chronically transfused TM patients.
1350
The impact of trans-catheter aortic valve implantation induced left-bundle branch block on cardiac reverse remodelling
Abstract
Background: Left-bundle branch block (LBBB) is common following trans-catheter aortic valve implantation (TAVI) and has been linked to increased mortality, although whether this is due to less favourable cardiac reverse remodelling is unclear.
Methods: 48 patients undergoing TAVI for severe aortic stenosis were evaluated. 24 patients with new LBBB (LBBB-T) following TAVI were matched with 24 patients with a narrow post-procedure QRS (nQRS). Patients underwent cardiovascular magnetic resonance (CMR) imaging prior to and 6m post-TAVI. Measured cardiac reverse remodelling parameters included left ventricular (LV) size, ejection fraction (LVEF) and global longitudinal strain (GLS). Inter- and intra-ventricular dyssynchrony was determined using time to peak radial strain derived from CMR Feature Tracking.
Results: Change in LV indexed end-systolic volume (LVESVi), LVEF and GLS was significantly different between the two groups (LVESVi: nQRS -7.9 ± 14.0 vs. LBBB-T -0.6 ± 10.2ml/m2, p = 0.020, LVEF: nQRS +4.6 ± 7.8 vs LBBB-T -2.1 ± 6.9%, p = 0.002; GLS: nQRS -2.1 ± 3.6 vs. LBBB-T +0.2 ± 3.2%, p = 0.024) (Figure 1). The nQRS group had a significant improvement in LVEF (54.1 ± 11.5 to 58.7 ± 9.0%, p = 0.010) and GLS (15.6 ± 3.9 to 17.7 ± 2.7, p = 0.010) at follow-up. There was significant post-procedure inter- and intra-ventricular dyssynchrony in the LBBB-T group (inter: LBBB-T 130 ± 73ms vs nQRS 23 ± 86ms, p = <0.001; intra: LBBB-T 118 ± 103ms vs. nQRS 13 ± 106ms, p= 0.001) (Figure 2). Baseline LVEF (beta -0.414, p = 0.015) and post-procedure QRS (beta -0.422, p = <0.001) remained significant independent predictors of change in LVEF on multiple regression analysis. Baseline LVEF (beta = -0.502, p = 0.001), baseline GLS (beta -1.02, p = <0.001) and post-procedure QRS (beta = 0.322, p = 0.001) were independent predictors of a change in GLS at 6 months.
Conclusion: LBBB-T is associated with less favourable cardiac reverse remodelling. In view of this, every effort should be made to prevent TAVI-induced LBBB, especially as TAVI is now being extended to a younger, lower risk population.
Figure 1. (left) Mean cardiac remodeling parameters pre and post-procedure. Relationship between post-procedure QRS duration and change in LVESVi, LVEF % GLS.
Figure 2. (above) Radial strain at peak systole in nQRS (A) and LBBB-T (B)
1351
Value of magnetic resonance myocardial perfusion imaging in patients with indeterminate coronary computed tomography angiography results
Abstract
Introduction: Coronary computed tomography angiography (coronary CTA) and cardiac magnetic resonance myocardial perfusion imaging (CMR-MPI) are established diagnostic tools for non-invasive assessment of coronary artery disease (CAD). While coronary CTA offers high-resolution anatomic images of coronary anatomy, CMR-MPI provides information on ischemia. The aim o our study was to evaluate the contribution of CMR-MPI in patients with indeterminate coronary CTA due to severe coronary calcification or the presence of coronary artery lesions of intermediate severity.
Methods: 70 patients consecutive with known (n = 16) or suspected CAD (n = 56) were included. All patients had undergone coronary CTA with a standard protocol for clinical reasons (Siemens Somatom Force CT or Siemens Cardiac Sensation 64). CMR-MPI with adenosine stress (1.5 T Siemens Magnetom) was performed due to an indeterminate coronary CTA result, either due to the presence of intermediate stenoses (50% - 70% luminal narrowing in coronary CTA) or impaired evaluability doe to coronary calcium. CMR-MPI data were analysed visually and with semi- quantitative evaluation of the myocardial perfusion index (MPRI).
Results: In 43 patients (61%), intermediate stenoses were present in coronary CTA, while severe coronary calcium impaired avaluability in 27 patients (39%, mean calcium score: 1501 ± 896). Visual analysis of CMR-MPI revealed myocardial ischemia in 16 patients (23%), while semiquantitative analysis demonstrated decreased MPRI (≥ 2 segments with MPRI < 1.5) in 33 patients (47%). In 17 patients invasive coronary angiography was performed, in 11 significant CAD was confirmed. Two patients denied invasive coronary angiography (figure 1).
Conclusion: Adenosine stress Magnetic Resonance Myocardial Perfusion imaging permits to rule out ischemia in a considerable number of patients with intermediate stenoses or severe calcification in coronary CT angiography. It should therefore be considered a useful gatekeeper to invasive angiography following indeterminate coronary CT angiography examinations.
1353
Gender differences in response to Transcatheter Aortic Valve implantation in patients with severe aortic stenosis assessed by feature tracking
Abstract
Introduction: Aortic stenosis (AS) results in left ventricular hypertrophy, myocyte dysfunction and fibrosis as a physiological response to excessive afterload. Transcatheter Aortic Valve implantation (TAVI) is an effective treatment for severe AS, however controversy exists whether gender predisposes to improved myocardial function post procedure. Feature-tracking cardiovascular magnetic resonance (FT-CMR) is a novel method for quantification of myocardial strain and strain rate that can identify subtle changes in left ventricular (LV) mechanics prior to overt changes in LV ejection fraction (EF). Our aim was to compare the acute impact of TAVI on LV function between males and females defined by strain measurements.
Methods: 59 patients, 34 males and 25 females matched for age (78.4 ± 7.19years vs 78.8 ± 9.7years p = 0.876), EuroSCORE II (5.19 ± 4.33% vs 4.26± 2.99% p = 0.788) and presence of baseline myocardial fibrosis (p = 0.829) with severe, symptomatic AS were prospectively enrolled prior to TAVI. All patients underwent an identical 1.5T CMR protocol (Intera or Ingenia, Phillips Healthcare, Best, The Netherlands) at baseline and at a median of 4 days following TAVI. Endocardial and epicardial borders were traced manually on the end-diastolic slice, then strain and strain rate measurements were obtained using commercially available post-processing software (CVI42, Circle Cardiovascular Imaging, Calgary, Alberta, Canada).
Results: TAVI was associated with a significant and comparable reduction in peak aortic pressure gradient and LV mass index in both men and women (Table 1). There was no significant difference in extent of pressure drop (p = 0.374) or extent of regression of LV mass index (p = 0.095) between men and women. There was no significant change in left ventricular end diastolic volume indexed (LVEDVI) or LVEF observed acutely following TAVI, in men or women (Table 1). Longitudinal strain (Ell), circumferential strain (Ecc) and circumferential strain rate (SRcc) did not change significantly in men or women (Table 2). However, TAVI was associated with a significant increase in longitudinal strain rate (SRll) in males (p = 0.012) that was not observed in females (p = 0.184) (Table 2).
Conclusion: TAVI was associated with structural reverse remodelling that was comparable between males and females. Functional recovery in LV mechanics following TAVI may occur earlier in males than in females, although this may reflect the higher LVMI that men have at baseline prior to TAVI.
1354
A qualitative assessment of first-pass perfusion bolus timings in the assessment of myocardial ischemia: A magnetic resonance study
Abstract
Purpose: To determine if a qualitative visual assessment of CMRI first pass perfusion (FPP) bolus arrival and transit times identified FPP differences between normal cardiac function and impaired function in patients with ischemic heart disease. Further, to determine if this technique is consistent with that of more time consuming ROI-based assessments.
Methods: Twenty-six patient volunteers were retrospectively included in this study. Each participant was assigned into one of two groups–Normal Viability Report (NVR) or Ischemic Heart Disease (IHD). The NVR group contained 11 patients (5 male: 6 female; mean age 40 (range 17- 63) years), the IHD group, 15 patients (13 male: 2 female; mean age 71 (range 56-89) years). IHD was defined as late gadolinium enhancement (LGE) in one or more of the mid-ventricular, short-axis coronary territories (AHA model), consistent with ischemic aetiology. Volunteers were imaged on a 1.5T MRI Scanner (Aera, Siemens, Erlangen, Germany). All images were acquired at end expiration breath-hold using an optimised CMR protocol. Patients were injected with a standard 20ml (0.5mmol/ml) bolus of Gadoteric Acid (Guerbet, Villepinte, France) administered at a consistent rate of 4.0ml/s for all patients, followed by a saline flush of 20ml.
Results: Bolus arrival and transit times, into and through the cardiac chambers, were visually and semi-quantitatively assessed and compared. The qualitative and semi-quantitative assessments resulted in almost identical timing values for both patient cohorts. However, the mean FPP bolus arrival time was slower for the IHD cohort relative to the NVR cohort, regardless of the analysis technique (p < 0.01 for RA arrival, and p < 0.005 for LV arrival). Similarly the mean FPP bolus transit time was slower for the IHD cohort relative to the NVR cohort, regardless of the analysis technique (p < 0.05). A clear difference was also noted for mean FPP bolus arrival and transit times between the NVR and IHD cohorts. Additionally, there were marked differences in LV structure and function between the two cohorts, providing supportive evidence that qualitative delayed FPP bolus time measurements may be a useful predictor of LV dysfunction.
Conclusion: A qualitative visual assessment of bolus arrival and transit times can identify FPP differences between normal cardiac function and impaired function in patient volunteers and this technique is consistent with that of ROI- based assessments.
1355
MRI prospective survey on cardiac iron and function and on hepatic iron in non transfusion-dependent thalassemia intermedia patients treated with desferrioxamine or non chelated
Abstract
Background: Few studies have evaluated the efficacy of iron chelation therapy in thalassemia intermedia (TI) patients. Our study aimed to prospectively assess by quantitative Magnetic Resonance imaging (MRI) the efficacy of Desferrioxamine (DFO) and its advantages with respect to the absence of chelation therapy in non transfusion-dependent (NTD) TI patients.
Methods: Among the 185 TI patients enrolled in the MIOT (Myocardial Iron Overload in Thalassemia) network and with a MRI follow-up (FU) study at 183 months, we selected 65 NTD patients. Cardiac iron overload was assessed by the multislice multiecho T2* technique. LV function parameters were quantified by cine SSFP sequences. Liver T2* values were converted into liver iron concentration (LIC) values.
Results: We considered 18 patients who have not received any chelation therapy (50% males; mean age: 37.83 ± 14.29 years) and 33 patients who had received DFO alone between the two MRI scans (51.5% males; mean age: 38.85 ± 7.83 years).The two groups were comparable for age, sex and baseline MRI data.
No patient treated with DFO had cardiac iron. At baseline only one non-chelated patient showed a pathological global heart T2* value (<20 ms) and he recovered at the FU. The percentage of patients who maintained a normal global heart T2* value was 100% in both groups.
Among the biventricular functional parameters, a significant decrease in the right ventricular ejection fraction was detected in DFO patients (-3.48 ± 7.22%; P = 0.024).
In patients with hepatic iron at baseline (MRI LIC ≥3 mg/g/dw), the reduction in the MRI LIC values was significant only in the DFO group (-2.20 ± 4.84 mg/g/dw; P = 0.050). The decrease in MRI LIC values was comparable between the groups (P = 0.155).
Conclusions: In this small population of sporadically or non transfused TI patients, DFO therapy showed no advantage in terms of cardiac iron and but its administration allowed and improvement in right ventricular function and hepatic iron overload.
The changes in cardiac T2* values and in the global systolic biventricular function were not significantly different between the two groups (Table below).
1358
Coronary Calcification Compromises Myocardial Perfusion Irrespective of Luminal Stenosis
Abstract
Aim: The aim of this study was to evaluate the relationship between coronary artery calcification (CAC) assessed by multi-detector computed tomography (MDCT) and myocardial perfusion assessed by cardiac magnetic resonance imaging (CMR) in a group of symptomatic patients.
Method: Retrospective analysis of 120 patients (age 65.1 ± 8.9 years, 88 males) who presented with atypical chest pain to Bethanien Hospital, Frankfurt, Germany, between 2007 and 2010 and who underwent CAC scoring using MDCT, CMR and conventional coronary angiography. Patients were divided into those with high grade (HG) stenosis (n = 67, age 65.1 ± 9.4 years) and those with no HG stenosis (n = 53, age 65.1 ± 8.6 years).
Results: There were more males with HG stenosis (82.1% vs. 62.3%, p = 0.015), in whom the percentage and number of perfusion segments were higher at rest (37.3% vs. 17%, p = 0.014) but not different with stress (p = 0.83) from those with no-HG stenosis. Thirty four patients had myocardial perfusion abnormalities at rest and twenty six patients developed perfusion defects with stress. Stress-induced myocardial perfusion defects were 22.4% sensitive and 79.2% specific for detecting HG stenosis. The CAC score was lower in patients with no-HG stenosis compared to those with HG stenosis (p < 0.0001). On the ROC curve, a CAC score of 293 had a sensitivity of 71.6% and specificity of 83% in predicting HG stenosis [(AUC 0.80 (p < 0.0001)]. A CAC score of 293 or the presence of at least 1 segment myocardial perfusion abnormality was 74.6% sensitive and 71.7% specific in detecting HG stenosis, the respective values for the two abnormalities combined being 19.4% and 90.6%. The severity of CAC correlated with the extent of myocardial perfusion in the patient group as a whole with stress (r = 0.22, p = 0.015), particularly in those with no-HG stenosis (r = 0.31, p = 0.022). A CAC score of 293 was 31.6% sensitive and 87.3% specific in detecting myocardial perfusion abnormalities.
Conclusion: In a group of patients with exertional angina, coronary calcification is more accurate in detecting high grade luminal stenosis than myocardial perfusion defects. In addition, in patients with no stenosis the incremental relationship between coronary calcium score and the extent of myocardial perfusion suggests coronary wall hardening as an additional mechanism for stress-induced angina other than luminal narrowing. These findings might have a clinical impact on management strategies of these patients other than conventional anti-anginal therapy.
Risk factor distribution in the total study population divided into those with HG stenosis and no HG stenosis
Risk factors . | Total n = 120 . | HG stenosis n = 67 . | no-HGstenosis n = 53 . | p-value . |
---|---|---|---|---|
Males, n(%) | 88 (73.3) | 55 (82.1) | 33 (62.3) | 0.015 |
Hypercholesterolaemia, n(%) | 27 (22.5) | 18 (26.9) | 9 (17.0) | 0.198 |
Hypertension, n(%) | 45 (37.5) | 24 (35.8) | 21 (39.6) | 0.669 |
Smoking, n(%) | 18 (15.0) | 11 (16.4) | 7 (13.2) | 0.625 |
Diabetes, n(%) | 14 (11.7) | 8 (11.9) | 6 (11.3) | 0.916 |
Obesity, n(%) | 3 (2.5) | 1 (1.5) | 2 (3.8) | 0.427 |
Family history of CVD, n(%) | 20 (16.7) | 11 (16.4) | 9 (17.0) | 0.934 |
Prior MI, n(%) | 34 (28.3) | 25 (37.3) | 9 (17) | 0.014 |
Risk factors . | Total n = 120 . | HG stenosis n = 67 . | no-HGstenosis n = 53 . | p-value . |
---|---|---|---|---|
Males, n(%) | 88 (73.3) | 55 (82.1) | 33 (62.3) | 0.015 |
Hypercholesterolaemia, n(%) | 27 (22.5) | 18 (26.9) | 9 (17.0) | 0.198 |
Hypertension, n(%) | 45 (37.5) | 24 (35.8) | 21 (39.6) | 0.669 |
Smoking, n(%) | 18 (15.0) | 11 (16.4) | 7 (13.2) | 0.625 |
Diabetes, n(%) | 14 (11.7) | 8 (11.9) | 6 (11.3) | 0.916 |
Obesity, n(%) | 3 (2.5) | 1 (1.5) | 2 (3.8) | 0.427 |
Family history of CVD, n(%) | 20 (16.7) | 11 (16.4) | 9 (17.0) | 0.934 |
Prior MI, n(%) | 34 (28.3) | 25 (37.3) | 9 (17) | 0.014 |
The difference in myocardial perfusion between the two groups at rest and with stress
. | CMR perfusion defect (n, %) . | ||
---|---|---|---|
. | 0 segment . | > 1 segment . | p-value . |
At rest | |||
HG stenosis | 42 (62.7) | 25 (37.3) | 0.014 |
noHG stenosis | 44 (83.0) | 9 (17) | |
At stress | |||
HG stenosis | 52 (77.6) | 15 (22.4) | 0.83 |
noHG stenosis | 42 (79.2) | 11 (20.8) |
. | CMR perfusion defect (n, %) . | ||
---|---|---|---|
. | 0 segment . | > 1 segment . | p-value . |
At rest | |||
HG stenosis | 42 (62.7) | 25 (37.3) | 0.014 |
noHG stenosis | 44 (83.0) | 9 (17) | |
At stress | |||
HG stenosis | 52 (77.6) | 15 (22.4) | 0.83 |
noHG stenosis | 42 (79.2) | 11 (20.8) |
The difference in CAC between the two groups
. | CAC (n, %) . | ||||
---|---|---|---|---|---|
. | 0-99 . | 100-399 . | 400-999 . | ≥ 1000 . | p . |
HG stenosis | 12 (17.9) | 12 (17.9) | 20 (29.9) | 23 (34.3) | <0.0001 |
No HG-stenosis | 29 (54.7) | 17 (32.1) | 4 (7.5) | 3 (5.7) |
. | CAC (n, %) . | ||||
---|---|---|---|---|---|
. | 0-99 . | 100-399 . | 400-999 . | ≥ 1000 . | p . |
HG stenosis | 12 (17.9) | 12 (17.9) | 20 (29.9) | 23 (34.3) | <0.0001 |
No HG-stenosis | 29 (54.7) | 17 (32.1) | 4 (7.5) | 3 (5.7) |
The relationship between CAC level and the number of segments showing myocardial perfusion, particularly in patients with no HG stenosis
. | Rest . | Stress . | ||
---|---|---|---|---|
All | r = 0.066 | p = 0.476 | r = 0.221 | p = 0.015 |
HG-stenosis | r = 0.049 | p = 0.696 | r = 0.189 | p = 0.125 |
No HG-stenosis | r = 0.149 | p = 0.288 | r = 0.314 | p = 0.022 |
. | Rest . | Stress . | ||
---|---|---|---|---|
All | r = 0.066 | p = 0.476 | r = 0.221 | p = 0.015 |
HG-stenosis | r = 0.049 | p = 0.696 | r = 0.189 | p = 0.125 |
No HG-stenosis | r = 0.149 | p = 0.288 | r = 0.314 | p = 0.022 |
The ROC curve for calcium score to predict the coronary artery stenosis, AUC = 0.80 (95% CI 0.82-0.88, p < 0.0001), cut-off value is CAC = 293 with sensitivity 71.6% and specificity 83%.
1359
Non–contrast three–dimensional magnetic resonance imaging for pre–procedural assessment of aortic annulus dimensions in patients undergoing transcatheter aortic valve implantation
Abstract
Objectives: Multidetector computed tomography (CT) has generally been demonstrated as the most accurate modality for pre–procedural aortic annulus assessment in patients undergoing transcatheter aortic valve implantation (TAVI). A major drawback of CT consists in the use of contrast dye, which may be unsafe in TAVI patients who frequently are affected by renal failure. In this setting, we sought to appraise the accuracy of non– contrast three–dimensional (3D) magnetic resonance imaging (MRI) in comparison with CT.
Methods: This is a single–center prospective study enrolling 48 consecutive patients with severe aortic stenosis screened for TAVI, who underwent non–contrast 3D MRI and standard cardiac CT for pre–TAVI aortic annulus assessment. Annular area measurements were obtained at mid–diastole for MRI, whereas maximum systolic area was assessed by CT only. Systolic MRI dimensions were modelled, by adding a corrective factor (+7.5% of the area), which represents the relative difference of systolic and diastolic CT area dimensions according to our local data and previous studies. Tests of correlation and agreement were performed.
Results: The mean subjective MRI image quality was 3.2 ± 0.8. According to CT, 20 patients (43.5%) had severe or massive annular calcifications. Figure 1 depicts Bland–Altman plots (Figure 1A) and Passing and Bablock regression analysis (Figure1B) showing no relevant differences of cross–sectional area measurements for corrected systolic MRI vs. systolic CT measurements (mean bias 0.09 cm2; 95% limits of agreement: –0.48 to 0.67). Higher bias was seen in patients with severe or massive annular calcifications (Figure1 C and D). Disagreement between CT and MRI in terms of transcatheter valve size was reported in 3 patients (6.5%). All of them had severe or massive annular calcifications.
Conclusions: In patients referred for TAVI, aortic annulus measurements using corrected systolic MRI compare favorably with those made at CT. Caution should be applied when choosing transcatheter valve size in the presence of severe or massive annular calcification as they reduce the accuracy of MRI measurements.
1360
“Systolic ventricularization” of the left atrium with bileaflet mitral valve prolapse: impact on quantification of mitral regurgitation
Abstract
Objective: To assess the impact of bileaflet mitral valve prolapse on the quantification of mitral regurgitation (MR).
Methods: MR severity was quantified in 23 patients with bileaflet mitral valve prolapse by calculating the regurgitant volume (RVol= left ventricular stroke volume (LVSV) minus the reference SV) using cardiac magnetic resonance. LVSV was calculated as the difference between the diastolic and systolic LV volumes using the Simpson's method (from the mitral annulus to the apex: standard SV). A corrected LV systolic volume was also calculated as the sum of the standard systolic LV volume plus the prolapse volume, i.e the volume included between the mitral annulus and the prolapsing mitral valve (corrected SV). The prolapse volume was calculated as the area of the basal LV slice times the prolapse height measured on a 2-chamber cine slice (Fig 1). We compared the differences in MR grading with these two methods, using the right ventricular (RV) SV as a reference.
Results: 23 patients (mean age 40 ± 23 years, 8 males) were included, and 13 had no detectable MR jet. The mean prolapse height was 10 ± 4 mm. RVSV was measured in all patients and showed an excellent agreement with the aortic SV measured by phase contrast (bias 0.2 ml, 95%-CI -7 to +8 ml; n = 14). In the 13 patients with no detectable MR jet, the corrected SV showed an excellent agreement with the reference SV (bias 0.6 ml; 95%-CI -4.0 to +5.2 ml) and was therefore consistent with the diagnosis of absent MR. Conversely, the standard SV showed a significant overestimation (Fig 2A; bias 16.3 ml; 95%-CI -5.1 to 37.9 ml). Likewise, over the whole group, the standard and corrected methods provided significantly different results for LV ejection fraction (67 ± 7% vs 53 ± 9%, p < 0.0001), RVol (27 ± 17 vs 7 ± 12 ml, p < 0.0001) and regurgitant fraction (RF, 27 ± 16 vs 9 ± 15%, p < 0.0001). The distribution of the MR grades - absent or trace (RF < 5%), mild (RF 5-29%), moderate (RF 30-50%) and severe MR (RF > 50%) - among the 23 patients was 0/13/8/2 with the standard and 12/8/3/0 with the corrected method (Fig 2B). Overall, the 2 methods were concordant in only 3 (13%) patients, as the standard method overestimated MR severity by 1 grade in 19 (83%) and by 2 grades in 1 (4%) patient.
Conclusion: In patients with bileaflet mitral prolapse, SV corrected for the prolapse volume better reflects the true LVSV. This may improve MR severity grading by preventing a systematic overestimation of RVol and RF.
Figure 1.
Figure 2.
1361
CMR assessment of left ventricular remodeling 6 months after percutaneous edge-to-edge repair using Mitraclip
Abstract
Objectives: Percutaneous edge-to-edge repair with MitraClip has been recommended as an alternative to conventional mitral valve repair in symptomatic severe mitral regurgitation at very high surgical risk. We aimed to assess left ventricular remodeling following MitraClip.
Methods: Consecutive patients scheduled for MitraClip were prospectively studied. Transthoracic echocardiography and CMR at 1.5 Tesla were performed before and 6 months after MitraClip procedure. Mitral insufficiency was graded using color Doppler flow mapping (grade 1-4). Left ventricular volumes were measured using short-axis SSFP CMR images covering the left ventricle.
Results: Forty consecutive patients (median 72 years (range 41-87 years); M/F: 27/13) at high surgical risk (logistic EuroSCORE 17,4% (5-62%)) with symptomatic severe (grade 3 or 4) mitral regurgitation (ischemic cardiomyopathy n = 29, non-ischemic dilated cardiomyopathy n = 8, mitral valve prolapse n = 3) underwent MitraClip procedure. MitraClip reduced mitral regurgitation (residual grade 1 in 10 patients, grade 2 in 20 patients and grade 3 in 7 patients), but failed in 3 patients (residual grade 4). CMR could not be performed in 16 patients due to a contraindication (pacemaker (n = 3), CRT-P (n = 1), CRT-D (n = 4), ICD (n = 6), claustofobia (n = 2)). One patient died during 6 months follow-up due to intractable heart failure. MitraClip caused intracavitary metallic artefacts at the base of the heart, but did not impede endocardial border delineation. No change in left ventricular volumes was measured following successful MitraClip implantation (enddiastolic volume: 250 ml (102-448 ml) versus 245,5 ml (79-489 ml) NS; endsystolic volume: 147 ml (31-376 ml) versus 161.5 ml (40-395 ml) NS and ejection fraction: 31% (10-73%) versus 31% (16-61%) NS). After mitraClip failure negative remodeling occurred within 6 months follow-up in 2 patients (enddiastolic volume: 330 and 250 ml versus 370 and 260 ml respectively, endsystolic volume: 224 and 184 ml versus 258 and 213 ml respectively) and a left ventricular assist device was implanted due to intractable heart failure in the third patient (no CMR because of ICD).
Conclusions: Despite reduction of mitral regurgitation, no change of left ventricular volumes was measured at 6 months follow-up. In MitraClip failure short term negative remodeling and/or clinical deterioration occurred at 6 months follow up. Device implantation accounted for CMR contraindications mainly.
1363
Accuracy of Transthoracic Echocardiography (TTE) in comparison with Cardiac Magnetic Resonance (CMR)
Abstract
Objective: to evaluate the accuracy of TTE in the evaluation of proximal thoracic aorta diameters in comparison withCMR in cohort of hypertensive patients.
Background: Proximal thoracic aorta dilatation is a pathologic process potentially associated to arterial hypertension (AH). Accurate and reproducible measurements of proximal thoracic aorta diameters are essential to diagnosis and follow-up. Transthoracic echocardiography (TTE) represents the most feasible test to suspect aortic dilatation, whereas the cardiac magnetic resonance (CMR) is usually considered the gold standard for aortic examination. Methods: In 61 hypertensive patients of our Hypertension Ambulatory, TTE and CMR diameter measurements were performed. We specifically compared the two measurements at the level of Sinuses of Valsalva (SoV) and maximal Ascending Aorta (Asc) diameter.
Results: Our patients showed a wide range of aortic diameters (SoV: 27.5-54 mm; Asc 26-54 mm). TTE and CMR diameters were significantly related (SoV: r 0.924 p < 0.001; Asc: r 0.944 p < 0.001). Mean difference between CMR and TTE diameters was 2.71 ± 2.31 mm at the level of SOV and 0.89 ± 1.9 mm at the level of maximal Asc diameter, with a TTE systematic overestimation by Leading-edge to Leading-edge (LE)convention. Also diameters estimated by Inner Edge-to-inner Edge (IE) convention showed a good correlation with CMR measurements (SoV: r 0.917 p < 0.001; Asc: r 0.942 p < 0.001). Comparing the two approaches, IE showed a better correlation with CMR measurements than LE convention at the level of maximal Aortic diameter (AscMAX_LE r 0.86 vs AscMAX_IE r 0.877, p<0.001), with a good agreement inter and intra observer (mean Interclass Correlation > 98 %, mean difference 0.53 ± 1.78 mm).
Conclusion: TTE is an accurate and costeffective technique for the assessment and follow-up of proximal thoracic aorta pathology, in particular at Asc level, allowing use of TTE as an alternative to MRI in patients with a good acoustic window.
1374
CMR for myocardial iron overload assessment: a new calibration curve from the MIOT project
Abstract
Introduction: The measurement of myocardial iron by T2* CMR has been established as fundamental to the best practice management of thalassemia. However, iron calibration data in humans is limited and CMR calibration varies according to instrumentation and technique. The aim of this study was to calibrate the T2*-CMR technique for noninvasive cardiac iron assessment, by considering a segmental approach.
Methods: Four human hearts were studied from transfusion-dependent patients after their death. A multislice multiecho T2* approach was adopted. After CMR, used as guidance, the heart was cut in three short-axis slice and each slice was cut into different equiangular segments, the same ones in which the T2* was assessed. Tissue iron concentration in the segments was measured with inductively coupled plasma atomic emission spectroscopy.
Results: T2* and iron concentration were overall assessed in 36 myocardial segments: 6 in the first heart (year 2004), 6 in the second one (year 2004), 8 in the third one (year 2005), and 16 in the fourth one (year 2010), Figure 1A shows the segmental iron concentration (in milligrams per gram dry weight) plotted versus the correspondent segmental T2* value (in milliseconds). As expected, the relationship was not linear. In Figure 1B the R2* values (R2* = 1000/T2*, in s-1) were considered. Regression analysis yielded a linear calibration of the following form: [Fe]R2* =0 .0079 × R2* - 0.1262 (R-square = 0.999).
Conclusions: We found an excellent linear agreement between R2* and cardiac iron and we provided a new calibration equation in humans.
1381
Can Speckle Tracking Imaging Reveal Myocardial Iron Overload in Thalassemia Major? A Combined Echocardiography and Cardiac Magnetic Resonance Study
Abstract
Background: Cardiac complications related to myocardial iron overload (MIO) remain the main cause of morbidity and mortality in thalassemia major (TM). Cardiac magnetic resonance (CMR) is a unique non-invasive technique to quantify MIO and the multislice T2* technique allows the identification of different patterns of MIO. Unfortunately, the availability for CMR scans in validated centers is still limited in many countries. On the other hand, echocardiography is a widely diffused, non-expensive and feasible technique, with elevated accuracy in the evaluation of cardiac function and morphology. Nevertheless, standard echocardiographic examination fails in detecting MIO until a substantial reduction of the left ventricular ejection fraction (LVEF) occurs. The evaluation of myocardial deformation by two-dimensional speckle tracking imaging (2DSTI) demonstrated a great accuracy in detecting subtle myocardial dysfunction in many different pathologic conditions. We aimed to investigate the role of 2DSTI in the detection of MIO in TM patients.
Methods: We recruited 31 TM patients (15 males, 37.87 ± 9.64 years) enrolled in the Myocardial Iron Overload in Thalassemia (MIOT) Network. All patients underwent to CMR and to echocardiography in the same day. CMR was performed with a 1.5T scanner and segmental and global T2* values were measured with a previously validated software (HIPPO MIOT®). Values of GLS were derived from the three apical views, while radial and circumferential strain were obtained from the three parasternal short axis views.
Results: Mean global heart T2* was 35.91 ± 12.69 ms (range:5.63-47.13 ms). Six patients (19.4%) showed a pathologic global heart T2* value (<20 ms). GLS showed a significant correlation with T2* values (R = -0.401; P = 0.025) and the percentage of patients with altered GLS (lower or equal -19) was significantly higher in the group with a significant MIO than in the group with no significant MIO (83% vs 28%, P = 0.05). Logistic regression demonstrated that patients with impaired GLS had a significant higher risk of showing pathological T2* values (Odds-ratio-OR = 12.86, 95%CI = 1.27-130.54; P = 0.031).
No relation was observed between GLS, age and sex, and between T2* values, LVEF, radial strain and circumferential strain.
Conclusions: Left ventricular GLS can be useful in detecting subtle myocardial dysfunction due to MIO in TM patients, demonstrating a significant correlation with MIO detected by CMR.
1382
Native myocardial T1 mapping in patients with pulmonary hypertension and age matched volunteers
Abstract
Background: Studies on porcine models [1] and patients with pulmonary hypertension (PH) [2] show T1 elevation at the RV insertion and correlations with markers of cardiac remodelling and function, indicating cardiac remodelling is accompanied by myocardial tissue changes. However the clinical significance of tissue changes remains uncertain.
Objectives: To establish whether myocardial T1 is elevated in patients with PH and whether myocardial T1 changes are linked to cardiac remodelling and function.
Methods: 202 consecutive patients with suspected PH underwent cardiac MRI on a 1.5 GE scanner between 11.2014-09.2015, including a Modified Look Locker Inversion Recovery sequence (MOLLI). 36 healthy volunteers underwent MOLLI sequences. MOLLI images were analysed using small, central regions of interest.
Results: 119 of the patients with no residual motion had a diagnosis PH with pulmonary wedge pressure (PCWP) ≤ 15mmHg, 23 patients had left heart disease (PH-LHD) and 16 had no PH. All volunteers were well registered post-registration.
Myocardial T1 was elevated in the interventricular insertion points (superior p < 0.001, inferior p < 0.001) in patients with PH, compared to volunteers (see Table 1 and Figure 1).
RV insertion point T1 was found to correlate with RV mass (superior, r = 0.394, p = 0.002), RV end diastolic volume index (inferior, r = 0.204, p = 0.041) and RV end systolic volume index (inferior, r = 0.197, p = 0.048). No significant correlations were found with LV volume, stroke volume, mass or ejection fraction.
Myocardial T1 in patients with PH-LHD correlated with LV end systolic volume (septal, r = -0.520, p = 0.016; inferior RV insertion point r = -0.442, p = 0.045; RV free wall r = -0.648, p = 0.001) and LV ejection fraction (septal r = -0.539, p = 0.013; superior RV insertion point r = -0.452, p = 0.040; inferior RV insertion point r = 0.446, p = 0.043 and RV free wall r = -0.654, p = 0.001).
Conclusion: Native T1 was found to be elevated in patients with PH. and this study confirms that myocardial changes accompany right ventricular remodelling. T1 values were not found to correlate with RV function. Further work to assess the sensitivity of myocardial T1 to change in patients on therapy and the prognostic significance of myocardial T1 is warranted.
References:
1. Spruijt, O., et al. 2015.Journal of Cardiovascular Magnetic Resonance, 17(Suppl 1): Q44.
2. Garcia-Alvarez, A., et al. Jacc-Cardiovascular Imaging, 8(1): p. 76-82
1384
A Insidious Line Between Thalassemia Intermedia And Left Ventricular Non-Compaction Disease: The Role Of Cardiac Magnetic Resonance
Abstract
Objectives: We tested the diagnostic accuracy of cardiac magnetic resonance (CMR) imaging for the differential diagnosis between the left ventricle non-compaction (LVNC) and the negative heart remodeling in β-Thalassemia intermedia (β-TI) patients. In particular, we verified whether the diastolic non-compacted to compacted myocardium (NC/C) ratio criterion could actually discriminate the abnormal trabeculations observed in β-TI from LVNC patients and we compared this diagnostic approach with the CMR criterion suggested by Grotoff M et al., which has been reported to be highly sensitive and specific for the LVNC diagnosis.
Methods: CMR images were analyzed in 180 patients with β-TI enrolled in the Myocardial Iron Overload in Thalassemia (MIOT) Network and 20 patients with LVNC studied at FTGM MRI Lab in Pisa from 2012 to 2014.
The CMR diagnostic criteria applied in β-TI patients were: a modified CMR Petersen's criterion proposed by Piga based on a more restrictive ratio of diastolic NC/C >2.5 at a segmental level and the Grothoff's criteria (percentage of trabeculated left ventricular myocardial mass (LV-MM) ≥25% of global LV mass and a total LV-MMI NC ≥15 g/m2).
In the 20 patients with LVNC the final diagnosis was performed based on Grotoff's criteria and on the clinical/functional criteria for LVNC to further increase the pre-test probability of the disease.
Results: In β-TI patients at least 1 positive NC/C segment was found in 18 patients (10%).
Compared with LVNC patients, in 18 β-TI patients the non-compaction areas were less frequent (3.70 ± 2.22 vs 1.62 ± 1.16; P = 0.007). The LV-MM NC percentage and LV-MMI NC g/m2 were significantly higher in LVNC than in β-TI patients 27.21 ± 2.45% vs. 10.88 ± 3.96%, P < 0.001; 20.35 ± 5.60 % vs.7.30 ± 4.77 %, P < 0.001 ).
None of the β-TI patients fulfilled the Grothoff's criteria.
The table shows the comparison of CMR parameters. LVNC patients showed significant lower LV stroke volume index and LV ejection fraction and they had an higher frequency of myocardial fibrosis detected by the LGE technique, although the statistical significance was not reached.
Conclusions: Differentiation of LVNC from hypertrabeculated LV in β-TI patients due to a negative heart remodeling depends on the selected CMR criterion. Based on our data in all β-TI patients with a NC/C ratio >2.5 we suggest to use Grothoff's criteria to improve the specificity of the diagnosis of LVNC.
1388
Pulmonary Artery : Ascending Aorta Diameter - An Important and Easily Measureable Prognostic Parameter
Abstract
Background: The pulmonary artery (PA) diameter and its relation to the ascending aorta (PA:Ao ratio) by cardiovascular magnetic resonance (CMR) or computed tomography (CT) have been identified as a non-invasive marker for pulmonary hypertension in heart and lung disease. However, its prognostic value is largely unknown.
Methods and Results: 650 consecutive patients (47.2% female, mean age 56.1 ± 17.1 years) referred to CMR were prospectively enrolled. Diameters of the great arteries were measured in axial black blood images, a PA:Ao ratio ≥1.0 was defined as cut-off for further analysis. The primary endpoint was defined as a composite of cardiovascular hospitalization and death.
In total, 131 (20.2%) patients presented with a PA:Ao ratio ≥1.0, which was associated with female gender (p = 0.010), atrial fibrillation (p < 0.001), diabetes (p < 0.001), impaired renal function (p < 0.001), higher NT-proBNP levels (p < 0.001), left ventricular volume (p = 0.023), right ventricular size (p = 0.002), and function (p < 0.001). Patients were followed for 17.8 ± 12.9 months, during which 110 (16.9%) experienced an event. Patients with a PA:Ao ratio ≥1.0 showed significantly higher event rates by Kaplan-Meier analysis (log-rank, p < 0.001). By multivariable Cox-regression analysis, a PA:Ao ratio ≥1.0 was independently associated with outcome, in addition to age, NT-proBNP serum levels, and right ventricular size.
Conclusion: The PA:Ao ratio is an easily measurable parameter by CMR and CT. A ratio ≥1.0 identifies patients at risk, presumably due to elevated pulmonary artery pressures. Based on these results, the PA:Ao ratio should routinely be assessed in CMR and CT scans.
1394
Novel carotid artery ultrasound index–Extra-media thickness and a well-established cardiac magnetic resonance fat quantification method
Abstract
Background: Epicardial is a well-established surrogate marker of cardiovascular diseases and complications. Extra-media thickness (EMT) is a novel ultrasound index related to arterial adventitia and adipose tissue.
Objectives: We aimed to evaluate the association between carotid EMT and epicardial fat (EF) and their relation to cardiovascular risk and metabolic syndrome (MS).
Methods: Eighty consecutive patients (age: 52.3 ± 15 years; males 66%) scheduled for cardiac magnetic resonance (CMR) were prospectively included in the study. Anthropometric parameters, CMR indices of EF and pericardial fat (PF), both common carotid arteries EMT, and ultrasound indices of visceral fat were measured in patients.
Results: In our study group, 57% of patients represented a very high cardiovascular risk, overweight or obesity was found in 70%, high body fat in 47%, and MS in 60% of individuals. Mean EMT (690 ± 120 µm) was significantly associated with EF area (r = 0.45; p < 0.001) and PF area (r = 0.35; p < 0.001). Among all fat indices, only EMT (MS+ 744 ± 145 µm vs. MS- 648 ± 90 µm; p = 0.002) and EF area (MS+ 888 ± 461 mm2 vs. MS 658 ± 343 mm2; p = 0.02) were significantly increased in patients with MS compared with individuals without MS. Multivariable regression analysis also showed that mean EMT is independently associated with number of cardiovascular risk factors (b = 0.005; p < 0.001). Moreover, very high cardiovascular risk subjects showed significantly increased EMT/BMI (p < 0.001) and EF area/BMI (p < 0.05) ratios. However, there was no significant association between EMT/BMI and EF area/BMI values (p = ns).
Conclusions: Our study showed the first findings on the associations between a novel ultrasound index EMT and EF assessed in a reference method of CMR. Carotid EMT may be a new surrogate marker, which may provide additional data on cardiometabolic risk beyond that derived form a well-established EF alone.
1403
Validation of CMR-derived LVOT diameters against direct in-vivo measurements
Abstract
Background: Left outflow tract (LVOT) size is necessary information for planning many interventions, including the Ross operation and percutaneous aortic valve replacement. Noninvasive measurement of LVOT diameter is routinely obtained from computed tomography or cardiovascular magnetic resonance (CMR) studies, but such measurements have not been validated in-vivo. CMR is radiation-free and provides a comprehensive assessment of the heart and aorta prior to aortic valve replacement. We sought to determine the agreement between different CMR measurements of LVOT and direct in-vivo measurement during cardiac surgery.
Methods: We included all patients undergoing a Ross operation who had both a comprehensive preoperative CMR study and recording of intraoperative LVOT diameters. All CMR studies were performed using a 1.5T scanner (Philips Intera) using a standard protocol including orthogonal long-axis LVOT cine-CMR images, double-oblique short-axis LVOT cine images planned from these views and a 3D steady-state free precession (3DSSFP) electrocardiogram and respiration-gated non-contrast magnetic resonance angiography covering the LVOT and thoracic aorta. Since the LVOT is ellipse-shaped, minimal and maximal LVOT diameters were measured and averaged in long-axis cine-CMR, in short-axis cine-CMR and from multiplanar reconstructed 3DSSFP images. The LVOT was measured in-vivo using Hegar dilators. Diameters were compared with in-vivo measurements using Passing and Bablock linear regressions.
Results: We identified 15 patients (mean age 44.3, 23-58 years, 6 females) with complete records undergoing a Ross operation (9 aortic stenosis, 1 aortic regurgitation, 5 mixed disease). All patients had a bicuspid or unicuspid aortic valve. The best agreement (table) was found between the 3DSSFP minimal LVOT diameter (0.89, 95%CI 0.49-1.29) and the mean 3DSSFP diameter (0.87, 0.59-1.33). All CMR measurements overestimated the LVOT diameter, however the bias was smallest for the mean 3DSSFP diameter (+1.63, -9.47 to 9.91 mm).
Conclusion: CMR LVOT measurements should be obtained from reconstruction of gated 3D datasets. Mean and minimal LVOT diameters best agree with the in-vivo measurements and could therefore be more relevant to report than the maximal LVOT diameter.
1409
Early myocardial perfusion measured by CMR in acute myocardial infarction treated by primary PCI–a postconditioning study
Abstract
Objectives: The aim of the study was to investigate the first pass perfusion in STEMI patients randomized to PCI with and without postconditioning. The CMR perfusion findings were intended to relate to the functional parameters of enddiastolic volume (EDV), endsystolic volume (ESV) and ejection fraction (EF), and infarct size (IS), area at risk (AaR) and myocardial salvage all measured with cardiac MRI.
Materials and Methods: 225 patients recruited from the POSTEMI study [1] were included. All patient had first time STEMI with symptoms less that 6 hrs, TIMI flow 0-1 in the occluded artery and minimal collaterals to the periphery of the vessel. CMR was performed at a 1.5 T scanner (Philips Intera) 1-5 days (mean 2.1) after the PCI. Magnevist 0.05 mmol/ kg body weight was injected at 3 ml/s. Signal intensity (SI) was measured in the infarcted area and in the normal myocardium and maximum contrast enhancement (MCE) was calculated (SI at baseline-maximal SI after contrast)/SI at baseline x100).
Results: There were no significant differences in first pass perfusion between patients treated with standard PCI and patients treated with additional postconditioning. Combining all patients the infarcted myocardium showed a significantly lower MCE compared to non-infarcted control myocardium (93 ± 55 vs. 114 ± 50; p < 0,001). Patients were divided into four quartiles of MCE in the infarcted myocardium (MCEi). MCE in the non-infarcted myocardium (MCEn) was significantly decreased in patients with low MCEi compared to patients with high MCEi (table). Patients with high MCEi had significantly lower EDV and ESV and higher EF than patients with low MCEi (table). Also the IS and AaR were smaller, myocardial salvage was higher, and occurrence of microvascular obstruction (MVO) was lower.
Conclusion: There were no differences in MCE of the infarcted area between patients treated with standard PCI and patients treated with additional postconditioning. The level of MCE correlated with EDV, ESV and EF as well as with the IS, AaR, myocardial salvage and occurrence of MVO. Patients with reduced MCE in infarcted myocardium also had impaired MCE in non-infarcted myocardium.
Reference:
1. Limalanathan S. et al. Rationale and design of the POSTEMI (postconditioning in ST-elevation myocardial infarction) study. Cardiology, 2010. 116(2): p. 103-9.
1420
Assessment of paravalvular aortic regurgitation after transcatheter aortic valve implantation using cardiac magnetic resonance imaging: a comparative study with echocardiography and angiography
Abstract
Aims: Assessment of paravalvular aortic regurgitation (AR) after transcatheter aortic valve implantation (TAVI) using Edwards SAPIEN XT valve remains challenging using transthoracic echocardiography (TTE) or angiography. Cardiac magnetic resonance imaging (c-MRI) has a low intraobserver and interobserver variability in the assessment of regurgitant volumes and might be more reliable to assess AR post-TAVI. We therefore aimed to evaluate the value of c-MRI to assess paravalvular AR after TAVI.
Methods and Results: Between February 2012 and March 2013, 132 consecutive patients underwent successful TAVI using exclusively Edwards SAPIEN XT prosthesis. AR was evaluated by c-MRI, TTE and angiography in 45 patients (27 women, mean age 84.1 ± 7.1 years). Angiography was performed immediately after TAVI whereas TTE and c-MRI were performed one month after implantation. At baseline c-MRI, the mean aortic regurgitant fraction (ARF) was 21.3 ± 12.5%. A significant AR (> grade II) was present in 24 (56%) patients using c-MRI (30%< ARF > 50%) whereas it was only observed in 18 (40%) and 12 (27%) patients using TTE and angiography, respectively. Interestingly, there was a poor correlation between c-MRI and TTE (r = 0.16, p = 0.28) and between c-MRI and angiography (r = 0.30, p = 0.06). In contrast, there was a good correlation between TTE and angiography (r = 0.6, p < 0.001). TTE underestimated AR by one degree in 9 patients, and by two degrees in 6 patients as compared to c-MRI.
Conclusions: The results of our study suggest that TTE and angiography may underestimate the severity of AR after TAVI as compared to c-MRI. Furthermore, there is a poor correlation between c-MRI and TTE or angiography. Further studies are mandatory to confirm our results in a larger population.
1422
Left atrial strain measured by feature tracking predicts left ventricular end diastolic filling pressure
Abstract
Introduction: Left ventricular end-diastolic filling pressure (LVEDP) is an invasive measure of LV function obtained at cardiac catheterisation (CC) that predicts prognosis and guides therapeutic strategy. Echocardiographic E/E' ratio has been shown to be inaccurate for estimation of LVEDP. Feature-tracking cardiovascular magnetic resonance (FT-CMR) is a novel method for quantification of myocardial deformation and can be used to quantitatively assess left atrial (LA) function. Currently there is no validated MRI parameter that estimates LVEDP. We hypothesised that LA strain correlates to LVEDP.
Methods: 14 patients in sinus rhythm, with severe AS underwent a 1.5T CMR protocol (Ingenia, Phillips Healthcare, Best, The Netherlands). LVEDP was recorded at the time of CC by standard techniques. 4 chamber and mid ventricular short axis steady state free procession cine images were obtained: LA endocardial and epicardial borders were traced manually on the end-diastolic slice and strain measurements were calculated using commercially available post-processing software (CVI42, Circle Cardiovascular Imaging, Calgary, Alberta, Canada).
Results: Patients were divided into 2 groups: low EDP (13 ± 2.4mmHg) and high EDP (36.1 ± 3.4mmHg) (p < 0.01). Both groups were evenly matched for baseline demographics (Table 1). Peak atrial longitudinal strain (PALS) was significantly different between low EDP and high EDP group (-21.7± 8.5 versus -11.1± 2.1% p = 0.01) (Figure 1). In multivariable analysis PALS was a determinant of LVEDP independent of other factors (Table 2). There was a moderate negative correlation between increasing invasive LVEDP and PALS (Pearson's correlation coefficient -0.647, p = 0.009).
Conclusion: LA function (PALS) as measured by FT-CMR is independently associated with LVEDP and may have a role in predicting LV filling pressures via a routine CMR protocol.
1426
Validation of extracellular volume equation by serial cardiac magnetic resonance imaging measurements in patients with varying hematocrit
Abstract
Background: Extracellular volume (ECV) measurement by cardiac magnetic resonance imaging (CMR) is an emerging marker of myocardial extracellular matrix. ECV is calculated from native and post contrast T1 measurements in myocardium and blood, and corrects for hematocrit (hct) values in the following equation: ECV = (1-hct)*λ, where λ =ΔR1 myocardium/ΔR1 blood, and R1= 1/T1. A larger hct leads to a smaller ECV if the other parameters are constant. Cross sectional case control studies demonstrate that ECV is elevated in various diseases such as myocarditis, dilated and hypertrophic cardiomyopathy and amyloidosis, and it has been shown that ECV correlates with histopathologic extent of myocardial fibrosis. However, whether variation in ECV measurements in longitudinal studies is explained by co-variation of hct, remains unknown.
Methods and Results: Sixty-two women (mean age 50.7 ± 9.3) with early breast cancer and with no serious comorbidity scheduled for adjuvant chemotherapy with the cardiotoxic anthracycline epirubicin were examined with serial CMR at baseline (visit 1) and after first cycle of 5-fluorouracil, epirubicin and cyclophosphamide (visit 2). Epirubicin dose ranged from 60-100 mg/m2. T1-relaxation times were assessed in one midventricular short axis slice of the left ventricle before and 15 minutes after 0.2 mmol/kg gadolinium contrast media using the Modified Look-Locker inversion recovery sequence (MOLLI) at 1.5T Achieva; Philips Medical Systems. Hct decreased significantly from visit 1 to visit 2 (mean change -0.02 ± 0.02, p < 0.001).
In the same period, λ also decreased significantly (mean change - 0.025 ± 0.04, p < 0.001). However, ECV only had a minor decrease, from 0.274 to 0.268 (mean change - 0.006 ± 0.023, p = 0.033) (Figure). Change in hct from visit 1 to visit 2 did not correlate significantly with the change in ECV (r= -0,180, p = 0,162) or change in λ, (r= 0,173, p = 0,178).
Conclusion: Longitudinal measurements of ECV by cardiac CMR in breast cancer patients undergoing chemotherapy show that early changes in hct do not correlate with a change in ECV, suggesting that change in ECV is independent of variations in hematocrit.
1427
Assessing diastolic function applying Cardiovascular Magnetic Resonance - comparison with the gold standard
Abstract
Objective: To identify a CMR parameter that allows the diagnosis of left ventricular (LV) diastolic dysfunction (DD) as defined by the gold standard [1].
Background: Heart failure (HF) with preserved ejection fraction (HFPEF) is prevalent in up to 50% of HF. In contrast to HF with reduced ejection fraction survival of HFPEF could not be improved in the last decades, although mortality is comparable. This may be attributed to the challenges of accurately making the diagnosis of DD.
Methods: We identified patients with (DD+) and without (DD-) DD based on the current guidelines by quantification of E/E' by echocardiography, invasive LV enddiastolic pressure (LVEDP) and NTproBNP [1]. CMR was performed at 1.5 T (Siemens, Avanto). Using standard cine-SSFP LV- and left atrial (LA) morphology was assessed in long (LAX) and short axis (SAX). Diastolic strain rate (SR) parameters were analysed in high temporal resolution cine SSFP (64 phases) by tissue tracking (TT) cvi42 prototype 5.3.0., circle, Canada (figure 1).
SR was also assessed based on SSFP tagging using additional prototype sequences ((SPAMM (TR 21.14ms) and CSPAMM (TR 42.28ms)): slice thickness 6mm, matrix 256x256mm2, FOV 300x300mm2, Flip angle 20°, tag spacing 7mm) in 3 SAX and 4CV.
Tissue phase mapping (TPM) was analysed in 3 SAX to evaluate diastolic myocardial peak velocities using Matlab.
Transmitral inflow was evaluated using phase contrast (PC) techniques (TR 17.44ms, 64 phases, in plane resolution 1.3x1.3mm2, VENC 120cm/s).
Results: From 725 screened patients 22 patients could be clearly defined as DD- and 15 as DD+ (table 1).
There were no differences between DD+ and DD- regarding LV-morphology, SR by tagging and transmitral flow.
DD+ showed increased LA-volume (DD-: 74.6ml ± 13.8 vs. DD: + 93.3ml ± 26.2 p = 0.013).
Further differences were identified by a segmental strain analysis: DD+ had significantly reduced radial (E-Peak: DD-: -14.5% ± 6.7 vs. DD + : -10.9% ± 5.9, p = 0.03; A-Peak: DD-: -4.8% ± 1.3 vs. DD + : -3.7% ± 1.4 p = 0.03) and circumferential (A-Peak: DD-: 3.8% ± 1.2 vs. DD + : 2.8% ± 0.8; p = 0.006) SR in the basal inferolateral wall assessed by TT. In the same segment D+ showed lower peak radial myocardial velocity by TPM (DD-: -3.7ms ± 0.7 vs. DD + : -2.8ms ± 1.0; p = 0.02).
Conclusions: LA-size and quantification of myocardial relaxation help to identify patients with diastolic dysfunction.
Reference:
1. Paulus, W.J., et al. Eur Heart J, 2007. 28(20):p. 2539-50.
1475
Role of Adenosine Stress Cardiac Mri in the Setting of Chronic Total Occlusion of Coronary Arteries
Abstract
Background: Conflicting data exists on the usefulness of a perfusion stress test prior to reopening a chronically occluded coronary artery. We sought to test whether a stress MR may improve patient selection for Chronic Total Occlusion (CTO) reopening over a conventional Late Gadolinium Enhancement (LGE) study.
Methods: 66 CTO patients without any other relevant coronary artery stenosis (or already treated if present at baseline angiography), underwent an adenosine stress MR before a reopening attempt. ECG and echocardiographic data were available for all. Patients were considered suitable for the reopening procedure in case of myocardial ischemia and/or viable myocardium subtended to the CTO.
Presence of a previous myocardial infarction (MI) was defined at ECG: Q waves > 40 msec; at echo: akinetic areas with reduced wall thickness; at stress MR: presence of ischemic LGE. A semi-quantification of LGE was assessed considering the percentage of myocardial mass with a score based on a 16 segment model: for each segment the score ranged from 0 to 4, according to the cut-offs of <25%, 26-50%, 51-75%, >76%. Each point accounted for 1.56% of LV mass.
Presence of myocardial ischemia (a subendocardial perfusion defect lasting for at least 5 heart beats during adenosine infusion) was diagnosed only in subsegments without LGE. With a score range from 0 to 2 (perfusion defect > 51% of wall thickness), each point account for 3.12% of LV mass.
Ischemia was graded as mild (<7%), moderate (7-15%) or severe (>15%) according to the extension of hypo-perfused myocardial wall.
Results: PREVIOUS MI: using MR as the gold standard for the diagnosis, the prevalence was globally high: 70% in our series. Both ECG and echo significantly underestimate this data (40% and 45% respectively).
LGE was absent in 20 pt (30%), <25% in 18%, 26-50% in 31% and >76% in 18% of patients; we observed a trend between the extension of LGE and the increase in Ventricular volumes and the decrease in ejection fraction. (Table 1)
ISCHEMIA: adenosine stress MR showed a perfusion abnormality in 100% of CTO territory. When we combined the extension of LGE with ischemia, we found that the grade of the latter was almost independent from LGE transmurality. (Figure 1)
Conclusions: Cardiac MRI can explain the partial benefit of CTO reopening in some patients (insufficient viable or ischemic myocardium), and should be performed before a CTO reopening.
LGE . | no LGE (20/66 - 30%) . | LGE ≤ 25% (13/66 - 18%) . | 25% < LGE < 75% (20/66 - 31%) . | LGE ≥ 75% (13/66 - 18%) . | 70% . |
---|---|---|---|---|---|
LV Volume(ml/m2) | 81.2 | 93 | 98 | 97.3 | |
LVEF | 63.5 | 57.2 | 52.3 | 49 | |
RV Volume (ml/m2) | 49.2 | 60.7 | 73.5 | 78 | |
RVEF | 65 | 61 | 65.8 | 62 | |
LA Area (cm2) | 23 | 22.3 | 25 | 25 | |
ECG- Q waves (prevalence %) | 25 (5pz) | 23 (3pz) | 62 (12pz) | 54 (7pz) | 40% |
ECHO- akynesia (prevalence %) | 5 (1pz) | 38 (5pz) | 67 (14pz) | 81 (10pz) | 45% |
LGE . | no LGE (20/66 - 30%) . | LGE ≤ 25% (13/66 - 18%) . | 25% < LGE < 75% (20/66 - 31%) . | LGE ≥ 75% (13/66 - 18%) . | 70% . |
---|---|---|---|---|---|
LV Volume(ml/m2) | 81.2 | 93 | 98 | 97.3 | |
LVEF | 63.5 | 57.2 | 52.3 | 49 | |
RV Volume (ml/m2) | 49.2 | 60.7 | 73.5 | 78 | |
RVEF | 65 | 61 | 65.8 | 62 | |
LA Area (cm2) | 23 | 22.3 | 25 | 25 | |
ECG- Q waves (prevalence %) | 25 (5pz) | 23 (3pz) | 62 (12pz) | 54 (7pz) | 40% |
ECHO- akynesia (prevalence %) | 5 (1pz) | 38 (5pz) | 67 (14pz) | 81 (10pz) | 45% |
1520
Aortic Elasticity Indexes by Magnetic Resonance Predict Progression of Ascending Aorta Dilation
Abstract
Background: Aortic distensibility and pulse wave velocity (PWV) assessed by magnetic resonance (MRI) are under investigation as parameters to evaluate the indication forascending aorta replacement. Recently, the maximum rate of systolic distension (MRSD) was proposed as a new index of aortic elasticity. The aim of this longitudinal study was to assess the determinants of ascending aorta (AA) growth rates over time focusing on the possible impact of parameters of aortic elasticity in patients with AA dilation.
Methods: We performed multiple MRI in 65 patients with AA dilation. The median time between the first and the last MRI was of 17 months (25°-75° 12-30 months). A significant increase in AA diameter was defined as a ≥2 mm increase in the diameter was detected from the baseline and the last MRI.
Results: An increase in AA diameter was found in 42 (68%) patients (AAD+ group) and absent in 20. The median increase was 0.16 (25°-75° 0.32-0.7) mm/month. The AAD+ group had a lower MRSD (4.6 ± 2.2 VS 7.4 ± 2.0, p <0.001) but non-different PWV and distensibility. The MRSD showed 93.7% specificity and 75.6% sensitivity for the prediction of increase. Kaplan-Meyer analysis showed that patients with MRSD ≤ 6 had lower progression-free survival times than the entire population (p < 0.002) and in the subgroup of patients with bicuspid aortic valve (p <0.05). MRSD was the best predictor of AA dilation progression.
Conclusions: MRSD is an index of aortic strain rate and is a valuable predictor for progression in AA dilation.
1522
Combined atrioventricular assessment of diastolic function by cardiac magnetic resonance
Abstract
Background: The study of diastolic function represents a crucial element in the evaluation of patients affected by cardiovascular diseases, particularly in patients with heart failure with preserved ejection fraction. Echocardiography is the keystone in the detection of diastolic dysfunction (DD). Cardiac magnetic resonance (CMR) offers unique diagnostic features including elevated definition of the endocardial border in respect to blood pool and the opportunity to obtain cine sequences with good temporal resolution.
Aims: To evaluate the role of CMR in the evaluation of diastolic function in a cohort of subjects with various degrees of diastolic dysfunction as assessed by standard echocardiography examination.
Methods: We recruited 22 patients with heterogeneous heart diseases associated with diastolic dysfunction and 18 healthy controls. Diastolic function was evaluated with a commercial echocardiographic system following the current guidelines (EACVI/ASE). Subjects were subdivided in 4 groups based on the result of echocardiography: group A= normal pattern; group B= I grade DD; group C= II grade DD and group D= III grade DD. All the patients underwent to CMR with a 1.5 T scaner. Short axis cine-SSFP images covering the entire left atrium and the left ventricle were acquired and to obtain left atrium (LA) and left ventricle (LV) volumes in all cardiac phases, obtaining V/time curves and the dV/time with a validated software.
Results: LA emptying fraction (LAeF) resulted the best parameter to distinguish patients with different degrees of DD (P < 0.001). Analysis of the receiver operating curve (ROC) showed that LAeF ≥43% identified patients with normal diastole or with I grade DD with 100% sensitivity and 77% specificity (95% CI 58.9-90.4; AUC 0.95; P < 0.0001), while a value ≤38% identified patients with III grade DD with 100% sensitivity and 80% specificity (95% CI 39,8 - 100,0; AUC 0.94; P<0.0001). dV/time of atrial late peak index and dV/time of ventricular late peak index were the best parameter to distinguish I grade from normal and II grade with best cut-offs >2.96 (95% CI 31,6 - 86,1; AUC 0.76; sens 61.5; spec 89.3; P = 0.003) and >1.98 (95% CI 42,4 - 80,6; AUC 0.81; sens 91.6; spec 62.9; P < 0.0001) respectively.
Conclusions: Analysis of cardiac volumes by CMR can provide useful information about the status of diastolic function. Further investigation is necessary for a better validation of the method.
1537
Safety, image quality and clinical utility of cardiac magnetic resonance in patients with antiarrhythmic devices
Abstract
Recent technological advances are changing the setting of cardiac magnetic resonance imaging (CMR) in patients with cardiac implantable electronic devices (CIEDs), but several issues regarding safety and usefulness of CMR remain in the agend of cardiologists PURPOSE: to report safety issues and imaging facts in a cohort of CIEDs patients.
Methods: Eighty four patients, 19 (23%) of them with implantable cardioverter defibrillator (ICD) and 65 (77%) with pacemaker (PM) were included. The need of CMR and CIEDs characteristics were carefully studied prior scan in order to assess risk-benefit profile. All approved scans were performed with a developed safety protocol. Changes of CIEDs parameters, size and type of artefacts and clinical value of CMR were analysed.
Results: No patients had symptoms or significant changes in device parameters. Signal loss area (fig 1) around device was significantly correlated with type and size of CIEDs. Sixty patients (71%) had also out of phase artefacts due to severe magnetic field inhomogeneities. Usual CMR data could be reported in 81% of patients. Clinical value of CMR al utility was significantly related with type and size of CIEDs but only their volume is an independent predictor. [OR:1.1;p < 0.02;IC:1.01-1.2].
Conclusions: In this setting CMR can be performed with acceptable risk-benefit profile if routine safety protocol was applied. Clinical value of CMR is associated with device characteristics but only device volume is an independent predictor of unsuccessful scan.
Figure 1. Signal loss area quantified in basal cardiac short axis slice of SSFP sequence using a region of interest whose center is placed in the center of device and includes all of the Signal loss area in a 73-year-old man with a PM (left panel) and a 63-year-old man with an ICD (right panel).
Figure 2. Regression plots showing relationship between size of device and susceptibility artefacts.
1538
Usefulness of cardiac magnetic resonance to predict the need for surgical procedures in patients with mitral regurgitation
Abstract
Nowadays, evaluation of patients with mitral regurgitation (MR) requiring surgical procedures is made using clinical and echocardiographic data but cardiac magnetic resonance (CMR) has been gradually performed in more clinical settings allowing us to study p with MR of any etiologies and severity degrees.
Purpose: To asses utility of CMR to predict the need of surgical treatment in MR patients
Methods: consecutive patients with any degree of MR referred to CMR and with echocardiographic study within three months were prospectively included. Patients with other causes of volume overload were excluded. Demographic data, clinical status (NYHA) and also imaging data were analyzed. Follow up was conducted using computerized medical records.Echocardiographic severity of MR was estimated in a comprehensive manner as recommended by American society of echocardiography. CMR protocol included: quantification of left and right ventricular ejection fraction and stroke volume (LVSV, RVSV) by short axis b-SSFP sequences using Simpson method to allow estimation of regurgitate volume (RV)i.e. RV = LVSV-RVSV and regurgitant fraction (RF) as the ratio of the RV divided by the LVSV i.e. RF (%) = (RV÷ LVSV) × 100 and also quantification of anatomic regurgitate orifice (ARO) by planimetry using four slices of phase images acquired perpendicular to ventricular septum including atrio-ventricular plane. RV and ARO were indexed by body surface area (i)
Results: 72 patients with median follow p of 30 months were included, 74%male, 39 (54%) needed surgery during follow up and in 28 of them (71%) reparative procedures were conducted. Primary MR was found in 32 (44%) of patients. Associated CMR parameters, cut-off points and diagnostic performance were shown in the table. These variables along with clinical and echocardiographic data were included in a multivariate Cox model which shows that NYHA clinical status [HR: 1.6 p = 0,007], iRV [HR:1.3 p = 0.03] and echocardiographic assessment of MR severity[HR:3.5 p = 0.04] were independent predictors of the need for surgical procedures.
Conclusion: 1. CMR evaluation of MR severity provides additional information over clinical and echocardiographic data in MR patients under clinical evaluation. 2. iRV with a cut –off point of 25ml/m2 is the most powerful predictor of the need for surgical procedures.
Variable . | AUROC . | P . | Confidence interval . | Cut-off . | Sen/Spe . | |
---|---|---|---|---|---|---|
LVEF | 0.673 | 0.018 | 0.530 | 0.816 | 60% | 57/ 68 |
iARO | 0.655 | 0.033 | 0.518 | 0.793 | 0.25cm2/m2 | 67/68 |
RF | 0.679 | 0.014 | 0.547 | 0.811 | 40% | 70/55 |
iRV | 0.740 | 0.001 | 0.619 | 0.861 | 25ml/m2 | 73/61 |
Variable . | AUROC . | P . | Confidence interval . | Cut-off . | Sen/Spe . | |
---|---|---|---|---|---|---|
LVEF | 0.673 | 0.018 | 0.530 | 0.816 | 60% | 57/ 68 |
iARO | 0.655 | 0.033 | 0.518 | 0.793 | 0.25cm2/m2 | 67/68 |
RF | 0.679 | 0.014 | 0.547 | 0.811 | 40% | 70/55 |
iRV | 0.740 | 0.001 | 0.619 | 0.861 | 25ml/m2 | 73/61 |
LVEF: Left ventricle ejection fraction; ARO: anatomic regurgitant orifice in phase-sequences; RV: regurgitant volume; RF: regurgitant fraction; i: indexed by body surface area
1550
Normal T1, T2, T2* and extracellular volume reference values in healthy volunteers at 3 Tesla cardiac magnetic resonance
Abstract
Background and Aim: Myocardial T1, T2 and T2* imaging techniques become more and more used in clinical practice. While normal values for T1, T2 and T2* times are well established for 1.5 Tesla (T) cardiac magnetic resonance (CMR), data for 3T remain scarce. Therefore we sought to determine normal reference values and inter and intraobserver reproducibility for native T1, T2, T2* mapping and extracellular volume (ECV) at 3T in healthy volunteers.
Methods: 55 healthy volunteers aged 20 to 80 years old (mean age = 50 ± 17 years, 62% men) underwent left-ventricular T1 (11 images, 18 heart-beats 3-(3)-3-(3)-5 steady state free precession Modified Look Locker Imaging (MOLLI)), T2 (spin echo-imaging with 8 echos ranging from 8 to 64ms) and T2* (gradient echo imaging with 8 echos ranging from 2.5 to 15 ms) mapping at 3T CMR (Philips Ingenia 3T). T1 mapping by MOLLI was repeated 15 min after administration of 0.2 mmol/kg Gadovist. ECV was computed as = (1-Ht)* ((1/T1) myo-post–(1/T1) myo-pre)/(1/T1) blood-post–(1/T1) blood-pre). T1, T2 and T2* times and ECV were measured by 2 observers in a region of interest (ROI) encompassing the whole mid-ventricular short axis slice as well as in a septal ROI. Inter- and intraobserver reproducibility were estimated by intraclass correlation coefficient (ICC).
Results: Mean normal values and inter/ intraobserver reproducibility are shown in the table below.
Conclusions: We provide normal myocardial T2, T2*,T1 and ECV reference values for 3T CMR. Myocardial T2, T2* and T1 had high inter and intraobserver reproducibility and were significantly different from those reported at 1.5 Tesla CMR. By contrast, ECV values in the whole myocardium were not significantly different from those reported at 1.5 T.
1551
Comprehensive intra-ventricular myocardial deformation strain analysis in healthy volunteers: implications for regional myocardial disease processes
Abstract
Ejection fraction (EF) is a traditional marker of systolic function. However it may not detect early, subtle cardiac disease with regional predilection. Aim of the study was to define regional intra-ventricular variation in myocardial strain in a cohort of healthy volunteers grouped per age and gender using Tissue-tracking cardiac magnetic resonance(CMR).
Methods: 94 healthy volunteers were recruited (equal gender distribution from 20-79 years). CMR at 1.5T was performed. Tissue-tracking software (CVI42, Circle Cardiovascular Imaging Inc.) estimated myocardial strain from the long-axis and the short-axis steady-state free precession (SSFP) cine images (Figure.1).The entire cohort was analysed by two independent readers. Inter-observer variability was determined by intra-class correlation coefficient (ICC). Myocardial segments were defined in accordance with the American Heart Association 16-segment model. Regional variations between basal, mid-cavity and apical segments as well as septal and lateral segments were assessed. Statistical analysis was performed using unpaired two-tailed T tests and one-way ANOVA with post-hoc correction as appropriate (p < 0.05).
Results: Inter-observer reproducibility analyses were excellent (ICC [95th confidence interval]: 0.918 [0.875-0.946]). Women had significantly greater longitudinal strain compared to men (-21.4 ± 2.1% vs -19.8 ± 2.7%, P < 0.001) and age is correlated significantly with increasing radial (R = 0.481, P < 0.001)and circumferential strain (R = -0.514, P < 0.001). Regional variation in strain was observed (Table.1) with a significant increase in deformation of the apical segments compared to the basal and mid-cavity segments for both radial and circumferential strain. The lateral wall had significantly larger circumferential and radial strain values compared to the septum in all subjects, as well as across all age and gender subgroups.
Conclusion: This is the first study to demonstrate that there is a gradient, increasing toward the apex, in circumferential and radial strain with CMR-derived myocardial strain, as previously showed with echocardiography. We also showed that the lateral wall consistently has significantly greater radial and circumferential strain values. Our findings are important as a comprehensive understanding of normal intra-ventricular regional variation is needed before this new tool can be implemented in daily clinical practice for the assessment of regional ventricular wall motion abnormalities.
1557
Elastic properties changes of aorta in patients with dilatation of the ascending aorta evaluated by Magnetic Resonance
Abstract
Objectives: Aortic stiffness is an independent cardiovascular risk factor. Magnetic resonance (MR) allows evaluation of aortic elastic properties by different indexes such as distensibility, and the pulse wave velocity (PWV) in three sections: aortic arch (aPWV), descending aorta (dPWV) and total PWV (tPWV).
We sought to changes of indexes of elastic properties of the aorta in patients with widening of ascending aorta.
Methods: We have examined 110 patients. AD (aortic dilatation) group: 90 patients with thoracic aorta dilatation > 4cm (27 (29%) women and 63 (71%) men) and control group: 20 patients without history of cardiac diseases (4 (20%) women and 16 (80) man). Distensibility, elastic modulus, stiffness index, compliance and aPWV, dPWV and tPWV were measured by the analysis of CMR images acquired using a 1.5 T (Magneto Avanto, Simens).
Results: pPWV and tPWV was higher in AD patients (10,2 ± 5,5 vs 7,2 ± 2,6, p = 0,03; 8,6 ± 3,2 vs 7,0 ± 1,9, p = 0,04), there was no difference in dPWV (8,0 ± 4,1 vs 7,5 ± 2,2) (Tab.2.). We found significant correlation between tPWV and patients age both in AD group (r = 0,499, p = 0,0003) and controls (r = 0,604, p = 0,006). In AD group there were 2 patients with thrombus, 24 with atherosclerotic plaques, 6 with bicuspid aortic valve and 38 with aortic regurgitation. tPVW in patients with thrombus and those with atherosclerotic plaques was higher compared to patients without these findings (16,9 ± 1,7 vs 7,8 ± 3,9, p = 0,001 for thrombus and 14,1 ± 3,8 vs 9,1 ± 3,9, p = 0,004 for plaques). The presence of aortic regurgitation or bicuspid valve did not affect tPWV value.
Conclusions: This study confirmed that physiological ageing is associated with a progressive impairment of the elastic properties of the aortic wall. Results of this study may be useful for the early identifcation of subjects with impaired aortic wall properties providing referral values of elasticity indexes assessed by CMR in different classes of age.
1558
The prevalence of active myocarditis assessed by cardiovascular magnetic resonance in patients with clinically suspected myocarditis
Abstract
Background: Myocarditis is still a significant problem in our daily clinical practice. The diagnosis of myocarditis remains challenging due to heterogeneity of clinical presentation. Therefor the real prevalence of the disease remains unknown. Cardiovascular Magnetic Resonance (CMR) has become a main non-invasive tool for assessing myocardial inflammation.
Methods: 143 consecutive patients with a clinical suspicion of myocarditis were evaluated by CMR according to Lake Louis criteria.
Results: CMR criteria for active myocarditis were fulfilled by 49 patients (34%). Additionally CMR examination was repeated in 5 cases with primarily negative result due to high clinical probability of myocarditis. One of them revealed active myocarditis. Totally 50 out of 143 patients (35%) were diagnosed with active myocarditis. Considering the CMR fulfilled criteria 71 patients (50%) had late gadolinium enhancement patterns typical for myocarditis. That may suggest that some of them had healed myocarditis at the time of examination.
Conclusions: The prevalence of active myocarditis is relatively high in patients with clinically suspected myocarditis. Myocardial inflammation may play a major role in every other patient with clinically suspected myocarditis.
1563
Quantitative assessment of myocardial scar heterogeneity using texture analysis to risk stratify post–MI patients for ICD insertion
Abstract
Following myocardial infarction (MI), one of the most serious complications is the development of ventricular arrhythmias, which can lead to sudden cardiac death (SCD). ICD therapy can be used to alleviate this risk. Currently ICD placement is determined by left ventricular ejection fraction (LVEF), which is the best available risk stratifier for ventricular arrhythmia and SCD in these patients. However, the majority of patients who develop ventricular arrhythmias have a preserved LVEF, whilst many with poor LVEF are implanted with ICDs inappropriately. In the present study, we use Texture Analysis (TA) to evaluate LGE–CMR (late gadolinium enhancement–cardiac MRI) images of myocardial scar in order to determine whether tissue heterogeneity indices could be used to accurately stratify arrhythmic risk. In this retrospective blinded analysis, 80 patients (age, 61 ± 12 years) with past medical history of MI were followed–up for a mean duration of 475 days (range, 4 to 1367 days) following a LGE–CMR scan that displayed myocardial scar. Ventricular tachycardia, ventricular fibrillation and unexplained syncope events were recorded. TA was performed on regions of interest delineating scar on short axis LGE images in all patients (Figure 1). The analysis comprised an initial filtration step, which highlighted objects of 2–6 pixels of radius using a Laplacian filter [spatial scaled factor (SSF) 2–6] followed by histogram analysis of pixel intensity, from which a set of 6 statistical parameters (kurtosis, skewness, entropy, mean of positive pixels, standard deviation (SD) and mean intensity) could be derived. Significantly different statistical parameters in group–wise comparison revealed that when a coarse filter was applied (SSF5 or SSF6), kurtosis was significantly (p <0.05) higher than those patients without arrhythmia/syncope following MI (no events group). Furthermore, there was a significantly reduced skewness in the fine SSF2 filter level and reduced SD in the SSF0, SSF2, SSF5 and SSF6 filter levels in the group of patients with events (Figure 2). This study concluded that TA of myocardial scar identified by LGE in CMR images determines significant differences in specific statistical parameters between event–free patients and patients who underwent ventricular arrhythmic events following MI. Further study is required to explore whether this can be a more accurate method of risk stratifying post–MI patients for ICD implantation than current methods.
1564
Gender differences in exercise capacity and LV remodeling in response to pressure overload in aortic stenosis
Abstract
Objectives: The aim of this study was to compare the remodeling response to aortic stenosis (AS) between men and women with initially asymptomatic AS, using cardiac magnetic resonance (CMR) imaging.
Methods: Subjects with asymptomatic moderate to severe AS (≥2 of: AVA <1.5 cm2, peak pressure gradient >36 mm Hg or mean PG >25 mm Hg) were recruited in this prospective, multi-centre, observational study conducted within the United Kingdom between April 2012 and November 2014. All patients underwent an echocardiogram, stress CMR and bicycle cardiopulmonary exercise testing (CPET).
Results: 174 patients (133 male, mean age 66.2 ± 13.3 years) were recruited. There was no difference in resting haemodynamics, incidence of co-morbidities and common cardiovascular medication use between the genders. Female patients were slightly younger but with similar severity of AS. There was no difference in the total exercise duration, haemodynamic response to exercise or the incidence of a positive test between the genders. However, female patients achieved a higher %predicted workload and %peak VO2, despite lower absolute values of workload and peak VO2.
Male patients were found to have significantly higher LV volumes and mass, with greater degree of concentric remodeling than females. They demonstrated both lower systolic (LVEF, PSSL + C, PSSRL) and diastolic function (PEDSRL + C). Both stress and rest myocardial blood flow (MBF) were significantly lower in male patients, with no difference in myocardial perfusion reserve (MPR). However, when MBF was multiplied by the LVMI to give the total MBF, there was no difference between the genders. Markers of focal fibrosis were higher in males, with a greater incidence and % of LGE. For diffuse fibrosis, there was no difference in native T1, but the ECV was found to be slightly higher in female patients.
Conclusion: There are clear differences between the genders in their exercise capacity and LV remodeling response to AS, with male patients demonstrating a greater degree of LV remodeling and fibrosis, with lower %predicted exercise capacity than females, for a similar degree of AS.
Abbreviations: BSA:body surface area, AV Vmax: aortic valve maximal velocity, MPG: mean pressure gradient, AVAI: aortic valve area index, LVEDVI = left ventricular end-diastolic volume indexed to BSA, LVESVI = left ventricular end systolic volume index, LVSI = left ventricular stroke volume index, LVEF = left ventricular ejection fraction, LVMI = left ventricular mass index, LAVI = left atrial volume index, MPR = myocardial perfusion reserve, MBF = myocardial blood flow, LGE = late gadolinium enhancement, ECV = extracellular volume fraction, PSS:peak systolic strain, PSSR:peak systolic strain rate, PEDSR: peak early diastolic strain rate, L:longitudinal, C:circumferential (* p < 0.05)
1572
Myocardial wall stress as a novel CMR measure to assess cardiac function
Abstract
Background: There are inherent and technical limitations in assessing cardiac function with current cardiovascular magnetic resonance (CMR) techniques (left ventricular ejection fraction (LVEF) and myocardial tagging). We developed a novel measure of myocardial wall stress based on flow rate and examined its clinical utility.
Method: Healthy controls (n = 10; 42 [34, 62] years), patients with genotype + /phenotype+ (G + /P+) hypertrophic cardiomyopathy (HCM; n = 5; 63 [50, 63] years]; LVEF 58 [50, 72] %) and G + /P+ dilated cardiomyopathy (DCM; n = 5; 44 [31, 50] years; LVEF 27 [21, 43] %) underwent CMR scan (3T, Philips Ingenia). Flow rate was estimated using a 4D cardiac model reconstructed from short-axis cines (Image Arena, TomTec and an in-house program; Panel A). Rate of change in pressure-normalized wall stress at any instant during cardiac cycle (s-1) was derived: , where Vmyo is the myocardial volume and is the flow rate from the 4D model (Panel A).
Results: Systolic and diastolic wall stress did not correlate with age or LVEF (P > 0.05 for all). Patients with G + /P+ HCM had the lowest systolic wall stress compared to G + /P+ DCM and controls (3.9 [2.3, 4.1] versus 4.2 [3.8, 5.6 versus 7.4 [6.0, 8.4] s-1, respectively; P < 0.0001), and these differences were not related to differences in LVEF (Panel C). Similar findings were observed in early and late diastolic wall stress; and without overlapping of interquartile range between healthy controls and patients (P < 0.0001 for all).
Conclusion: A novel measure of wall stress based on conventional CMR sequences holds potential of assessing systolic and diastolic cardiac function, and may provide incremental value over LVEF.
1573
Feature tracking cardiac magnetic resonance to assess LV mechanics in pressure and volume overload
Abstract
Background: In aortic valve disease, left ventricular (LV) dimensions and ejection fraction are important hallmarks for decision making. However, the effects of pressure overload (aortic stenosis) or/and volume overload (aortic regurgitation) lead to different LV remodelling (concentric and eccentric hypertrophy, respectively) which may LV mechanics differ. We aimed at characterizing LV mechanics (longitudinal strain) using feature tracking cardiac magnetic resonance (FT-CMR) in patients with various degrees of aortic stenosis and aortic regurgitation, and preserved LV ejection fraction (LVEF).
Methods: Seventy-one patients (14 with normal valve function, 29 with aortic stenosis and 28 with aortic regurgitation), mean age 45 ± 19 years, 70% men, who underwent clinically indicated CMR and showed preserved LV ejection fraction (>50%) were included. LV end-diastolic (EDV) and end-systolic volumes (ESV), LVEF and mass were measured on steady-state free precession (SSFP) cine images. FT-CMR analysis was performed offline using tissue-tracking software (CVI42, Circle Cardiovascular Imaging Inc.) to estimate LV global longitudinal strain (GLS) from two long-axis SSFP cine images (Figure 1). To correct for the LV remodelling process, LV GLS was corrected for LV EDV.
Results: There were significant differences in LV volumes, mass and ejection fraction across the 3 groups of patients (Table 1): patients with aortic regurgitation showed significantly larger LV volumes, and lower LVEF compared to patients with normal aortic valve function and patients with aortic stenosis. There were no differences in LV GLS across the groups. However, after correcting for LV EDV, patients with aortic regurgitation showed more impaired LV GLS as compared to the other groups.
Conclusions: LV mechanics differ significantly across the different types of aortic valve dysfunction (normal, stenosis and regurgitation), with aortic regurgitation showing the most impaired LV GLS corrected for LV end-diastolic volume, despite preserved LVEF.
*p < 0.001 vs. Aortic stenosis and normal; †p < 0.001 vs. Aortic stenosis; #p < 0.001 vs. Aortic regurgitation and Aortic stenosis.
Figure 1. Assessment of LV GLS with FT-CMR. From two long-axis SSFP cine images, the time-GLS curve is obtained and peak LV GLS is determined.
1574
Safety, feasibility and clinical impact of Cardiovascular Magnetic Resonance in patients with MR-conditional devices
Abstract
Background: Cardiovascular magnetic resonance (CMR) is increasingly used for diagnostic work-up. Implanted cardiac devices were considered unsuitable for CMR. However, with the development of MR-conditional devices access to CMR has increased, despite concerns regarding image quality and diagnostic accuracy. We aimed to assess the safety, feasibility and clinical impact of CMR in patients wearing MR-conditional devices.
Materials and Methods: We retrospectively enrolled patients wearing MR-conditional devices who underwent CMR in a 1.5T scanner (Jun 2012-Nov 2015). Every sequence was analysed by two independent observers and the presence of artefact (0 = no, 1 = minor, and 2 = major artefacts impacting the image quality and interpretation). Inter-observer agreement was assessed per sequence and as overall judgement on scan quality and interpretation. Clinical impact of CMR was defined as a change in diagnosis and a change in management. All devices were interrogated before and after CMR.
Results: We enrolled 46 consecutive patients (28 male, mean age 56 ± 16 years); 22 patients (48%) had MR-conditional pacemaker and 24 (52%) implantable loop recorders. All CMR scans were successfully completed and diagnostic; 22 scans (48%) were not affected by artefacts, 17 (37%) had minor artefacts and 7 (15%) major. Additional FLASH sequences were performed in 9 patients (20%) to overcome artefacts. No change in device parameters was reported after the scan. There was moderate agreement between the observers regarding the judgement on quality and interpretation, both overall (kappa0.454, p < 0.0001) and per sequence, with the exception of long-axis post-contrast sequences, for which it was only fair (kappa0.284, p0.005) (Table1). SSFP cine was the sequence most affected by artefacts. A 16-segment analysis showed that artefacts mainly affect the mid-apical left ventricular anterior wall and anteroseptum (Figure1). CMR had a clinical impact in 26 patients (57%), determining a change in diagnosis in 16 (35%), a change in management in 5 (11%) and a change in both in 5 patients (11%).
Conclusion: In patients with MR-conditional devices, CMR is safe and feasible and the images obtained are of diagnostic quality despite a degree of artefacts present. The diagnostic quality of the scan was such that the CMR findings resulted in major clinical impact from which this patient population could directly benefit from.
. | Cohen's kappa . | p . |
---|---|---|
HASTE | 0.378 | 0.001 |
Long Axis Cine | 0.356 | 0.001 |
Short Axis Cine | 0.532 | 0.000 |
EGE | 0.398 | 0.003 |
Long Axis LGE | 0.284 | 0.005 |
Short Axis LGE | 0.516 | 0.000 |
Overall Judgement | 0.454 | 0.000 |
. | Cohen's kappa . | p . |
---|---|---|
HASTE | 0.378 | 0.001 |
Long Axis Cine | 0.356 | 0.001 |
Short Axis Cine | 0.532 | 0.000 |
EGE | 0.398 | 0.003 |
Long Axis LGE | 0.284 | 0.005 |
Short Axis LGE | 0.516 | 0.000 |
Overall Judgement | 0.454 | 0.000 |
Figure 1. Left ventricular segmental analysis to assess artefacts interference. Sixteen-segment model showing that artefacts mostly affect the mid-apical left ventricular anterior and anteroseptal walls, on cine and post-contrast sequences, respectively.
1576
T1 Mapping at 1-Year Following Aortic Valve Replacement: Baseline Geometry Defines Magnitude of Fibrosis Regression
Abstract
Background: Ventricular morphology in severe aortic stenosis (AS) can be (1) normal to (2) concentric remodeling, (3) concentric hypertrophy and (4) eccentric hypertrophy (Figure 1A). We designed the RELIEF-AS Study and used T1 mapping to measure the cellular and fibrosis volume associated with these pattern of remodelling at baseline and the changes that occur one year after valve replacement.
Methods: 123 patients (age 69 ± 11 years; 56% male) with symptomatic, severe AS (AVAi 0.4 ± 0.1cm2/m2) underwent CMR prior to AVR. Patients were categorized by remodelling patterns. 95 patients attended repeat CMR at one year. T1 mapping (ShMOLLI) was performed pre and 15 minutes post-contrast in 3 short axis views for global ECV (excluding segments with infarct-pattern scar). Matrix volume (LVM * ECV) and cell volume (LVM * [1-ECV]) were calculated.
Results: LVM regressed by 20% (171 ± 63g to 136 ± 42g, p < 0.001) at one year. This regression was greatest in patients with eccentric hypertrophy, figure 1B, but even patients with normal geometry displayed LVM regression. Overall, there was a 17% reduction in matrix (48 ± 19ml to 40 ± 13ml, p < 0.001) and a 23% reduction in cell volume (117 ± 42ml to 90 ± 28ml, p < 0.001)–meaning the ECV increased over the year (28.7 ± 2.6% to 30.3 ± 3.2%, p < 0.001). The greatest cell and matrix regression was seen in patients with concentric and eccentric hypertrophy (p < 0.001; Figure 1C and 1D).
Discussion: Within the 20% reduction in LVM at one year post AVR, myocardial fibrosis and cellular regression occurs. But the extent depends on the pre-op morphology with the greatest cell and matrix regression seen in concentric and eccentric hypertrophy. This data supports the position that human diffuse fibrosis is dynamic and measurable by CMR - a key biological result and proof-of-concept for drug development targeting myocardial fibrosis.
Figure 1. Baseline remodelling pattern (A) predicts overall mass (B), cellular (C) and matrix regression (D) after AVR (p < 0.001; figures B-D show mean ± 2SE).
1583
Normal values of LV global myocardial mechanics using two and three-dimensional cardiovascular magnetic resonance
Abstract
Background: Impairment of left ventricular (LV) deformation is an independent predictor of cardiovascular outcomes in various cardiomyopathies despite normal LV ejection fraction. Cardiovascular magnetic resonance myocardial feature tracking (CMR-FT) has recently proposed as a robust method that can provide quantitative measurements of myocardial mechanics. Our aim was to provide reference values for myocardial deformation from 2 and 3 Dimensional (2D, 3D) CMR-FT in a population of control subjects.
Methods: 75 control subjects underwent a CMR study (1.5 T). CMR-FT was applied to standard short and long axis views of vector ECG gated CMR cine SSFP sequences. 2D and 3D global peak longitudinal, circumferential and radial systolic strain values (GLS, GCS, GRS respectively) were measured. All CMR analysis was performed on commercially software (Circle CVI 42®, Calgary, Canada).
Results: 57% of the subjects were men with a mean age of 48 ± 16 years. 2D global peak systolic values were -19.0 ± 2.6% for longitudinal, − 18.6 ± 2.9% for circumferential and 37,1 ± 8.5% for radial strain. 3D global peak systolic values were -15.8 ± 2.1% for longitudinal, −17.9 ± 2.9% for circumferential and 41.8 ± 11% for radial strain. No significant differences were found between men and women, however age was moderately correlated with all LV strains parameters (figure). Compared to 3D, 2D strains values were higher for GLS and GCS (mean differences 3.1 ± 1.9, p < 0.001 and 0.7 ± 3.5, p = 0.08 respectively) while 2D GRS were lower compared to 3D values (mean difference 4.7 ± 8.5,p < 0.001).
Conclusions: Myocardial mechanics using CMR-FT is a novel technique applied to standard cine CMR images providing a reliable quantification of global LV deformation. Normal ranges of global 2 and 3D LV strains are needed to implement its use in clinical practice.
Figure 1
1585
Prediction of infarct transmurality in acute myocardial infarction based on cardiac magnetic resonance deformation analysis
Abstract
Aims: Tissue tracking deformation analysis (TT) provides a new tool for the quantitative study of myocardial function with cardiac magnetic resonance (CMR). It was our purpose to evaluate whether CMR deformation parameters allowed assessment of infarct transmurality and how they correlated with established echocardiographic deformation parameters.
Methods: CMR was performed in 120 patients with ST-segment elevation acute myocardial infarction within 3-5 days of admission and successful primary percutaneous revascularization. Using a 1.5T CMR scanner, routine SSFP images from 3 long-axis and a stack of short-axis slices were analyzed with TT (CVI42, Circle Cardiovascular Imaging Inc.) for peak systolic longitudinal, circumferential and radial segmental strain and post-systolic peak strain, according to the 16-segment AHA model. Infarct transmurality determined with late gadolinium enhancement was used to define no infarction (0%), non-transmural infarction (1-49%) and transmural infarction (50-100%). Additionally, 112 patients also underwent echocardiography and speckle tracking analysis for longitudinal systolic strain.
Results: Deformation analysis was possible in 97,9% of the segments (n= 1880) with CMR and in 96,2% (n= 1724) with echocardiography. With increasing infarct transmurality, peak systolic strain parameters significantly worsened and occurence of post-systolic peak strain increased (table 1). Circumferential strain was the strongest discriminator of transmural versus non-transmural infarction with a sensitivity of 85% and a specificity of 80% (cut-off value of -14,0%, AUC 0,887), while echocardiography longitudinal strain presented a sensitivity of 83% and a specificity of 77% (cut-off value of -12,3% and AUC 0,863). Longitudinal strain assessed with CMR and echocardiography showed poor correlation (r = 0,4, p < 0,001), low bias (-0,33%) and moderate limits of agreement (±7%) and interclass correlation (ICC 0,66, p < 0,001).
Conclusion: CMR-TT analysis of myocardial deformation allows distinction between non-transmural and transmural infarction. Circumferential strain with a cut-off value of -14% was the best predictor of transmural infarction.
1595
Measuring invasive blood pressure by catheters guided solely by Cardiovascular Magnetic Resonance by using a new guidewire without the need of a hybrid CMR-fluoroscopy suite
Abstract
Background: Blood pressure or blood pressure gradients cannot be evaluated accurately by routine cardiovascular magnetic resonance. However, blood pressure can be measured using invasive fluid-filled catheters guided by flouroscopy in conventional catheter-laboratories. First clinical approaches have also been made using so-called hybrid cardiovascular magnetic resonance (CMR)-fluoroscopy suites. Therefore, the aim of this study was to test the feasibility of measuring blood pressure using fluid-filled catheters solely by CMR guidance without the need of a hybrid CMR-fluoroscopy suite.
Methods: In this ongoing study, patients scheduled for routine clinical CMR and combined diagnostic and interventional catheterization by flouroscopy were included. Up-to-date 6 patients with untreated or recurring coarctation of the aorta and 4 further patients with Fontan Circulation, RVOT-Conduit and pulmonary artery hypertension (2 female, median age: 23 years, range: 13 to 55 years) were included into the study.
Interim Results: Blood pressure or a blood pressure gardient was measured succesfully by fluid-filled catheters guided solely by CMR using a guidewire (MRWire, nano4imaging, Aachen, Germany) designed for use in the MR environment in all patients. No guidewire-related adverse event occurred.
Conclusions: This study shows that invasive blood pressure can be measured relatively easily using fluid-filled catheters solely by CMR guidance without the need of a hybrid MR-fluoroscopy suite. In this ongoing study, up-to-date no guidewire-related adverse event occurred.
1599
Influence of active and passive cardiac implants on CMR image quality: results from a consecutive patient series
Abstract
Background: Cardiac magnetic resonance imaging of patient with active and passive cardiac implants has emerged as being feasible and safe despite previous contraindications. Artifacts caused by the systems containing ferromagnetic material can impair the diagnostic quality. We aimed to provided an overview over imaging quality depending on type of device and imaging sequence.
Material and Methods: 50 patients having been implanted with either a pacemaker, implantable cardiodefibrillator (ICD) or implantable loop recorder (ILR) were evaluated. Patients were scanned in a 1.5T MRI scanner and image quality was graded on a 4-point scale. Steady-state freeprecession and turbo field echo sequences were used to assess cardiac function. For myocardial late gadolinium-enhanced (LGE) imaging, 3D inversion prepared spoiled gradient-echo sequences were used.
Results: CMR was safe in all patients. There were no changes in pacing thresholds, sensing or lead impedance. Imaging artifacts were more prominent in SSFP then in TFE sequences. Devices implanted on the right side of the sternum did not impair image quality. ICD devices created larger artifacts then pacemakers and ILR did not impact imaging at all. Number of evaluable cardiac segments correlated with distance between device and heart.
Conclusions: When adhering to strict safety precautions CMR can be safe and feasible inpatients carrying active and passive cardiac devices. While heavily depending on location and type of device imaging quality is generally good enough for diagnostic purposes.
Figure. LGE Scar imaging of the right ventricle. A pacemaker lead in the RV, a device artifact and a scar of the inferior wall are visible.
1600
Reproducibility of aortic 4D flow measurements in healthy volunteers
Abstract
Objective: 4D flow MR can provide quantitative information about hemodynamic parameters. One such parameter, wall shear stress (WSS), has been associated with vascular wall remodelling and has been frequently applied in clinical studies over the recent years. However, data on the measurement variability of WSS is limited. We evaluated the reproducibility of 4D flow derived parameters with an automatic time-resolved 3D vessel segmentation model by a scan-rescan analysis.
Methods: 8 healthy volunteers (age 26.9 ± 2.7 yrs) underwent a 4D flow MRI scan and subsequently a rescan, after a 15-minute break (3.0T, Philips Healthcare; retrospective ECG and respiratory navigator gating). From 3D PC-MR images, a 3D aortic volume was automatically segmented after defining the vessel of interest by three markers using CAAS MR 4Dflow v1.0 software (Pie Medical Imaging). An aortic centerline was automatically generated in the 3D aortic volume and 3 analysis planes were placed along the centerline on anatomic landmarks (mid-ascending aorta, aortic arch, proximal descending aorta) (Figure 1). Hemodynamic parameters were determined and a reproducibility analysis (Bland-Altman) was performed for: maximum velocity (Vmax), mean velocity (Vmean), peak flow rate, cardiac output (CO) and 3D maximum WSS (WSSmax) and 3D mean WSS (WSSmean).
Results: Heart rate (HR) was similar for the scan (63.1 ± 6.6 bpm) and rescan (62.6 ± 6.4 bpm) in the volunteers (p = 0.79). Scan-rescan data showed good reproducibility for Vmean, flow rate and CO (<10% variation compared with average values and good intraclass correlation (ICC)) and moderate reproducibility for Vmax and WSSmean (<12.5% variation, moderate ICC). Although WSS estimations during the cardiac cycle showed similar trend, the variability in the magnitude of WSSmax was high, with low ICC (Table 1).
Conclusion: Reproducibility of standard 4D flow parameters in this scan-rescan study in healthy volunteers was good. However, the variability of planar WSS measures was high and out of clinically acceptable margins. Factors that potentially contribute to this lower WSS reproducibility: 1. Operator dependent factors: plane placement; manual defined markers for automatic vessel segmentation 2. Patient factors: although averaged HR was similar, individual variations could have contributed. Assessment of the observer variability needs to be done to evaluate an observer-dependent contribution. A potential better alternative is a segmental WSS approach.
1601
An automatic approach to extract 4D flow hemodynamic markers: application in BAV-affected patients
Abstract
Objectives: 4D flow CMR is able to assess aortic hemodynamics, but available commercial software still lack of standardized automatic protocols for post-processing of hemodynamic markers [Markl, 2014]. For this purpose, we used an in-house platform with ad-hoc automatic algorithms to assess BAV hemodynamics with minimal user-dependent tasks.
Methods: Ten subjects (5 healthy volunteers, 5 BAV patients, age- and sex- matched) were analyzed to test the effectiveness of the proposed approach. CMR acquisitions were performed at John Radcliffe Hospital (Oxford, United Kingdom) on a 3.0 T MR system (Trio, Siemens, Erlangen, Germany). 4D flow sequences were performed along oblique-sagittal planes entirely encompassing the whole thoracic aorta (Fig. 1a), using prospective ECG-gating, respiratory navigator and the lowest non-aliasing VENC. An innovative semi-automated segmentation algorithm was adopted to extract the volume of interest of the aorta (Fig. 1b). Then, cross-sectional planes were automatically placed along the centerline of the vessel (Fig. 1c). Hemodynamic markers (e.g. velocity peaks, flow displacement, flow jet angle, wall shear stress) were computed considering three time-frames over the peak systole. “Hot-spots” corresponding to the maximum values of the computed markers were automatically identified. Mann-Whitney non-parametric tests were performed to assess differences between the two populations.
Results: The multiplane analysis identified expected significant differences (p-value < 0.05) between the two groups: BAV velocity fields were characterized by asymmetrical distributions with a higher variability in the peak values (217 ÷ 410) with respect to the controls (120 ÷ 151 cm/s). BAVs presented a marked eccentricity (0.19 ÷ 0.28 vs 0.03 ÷ 0.11) and deflection (31.4 ÷ 50.2 vs 12.1 ÷ 17.2 °) of the systolic jet: these data corroborated the occurrence of localized wall shear stress “hot-spots” (31.4 ÷ 50.2 vs 13.6 ÷ 19.6 Pa). The lumen area was also markedly above the physiological values (821 ÷ 1559 vs 465 ÷ 658 cm2). Furthermore, supporting the clinical relevance of the approach, the segmentation process approximately lasted 10 minutes, whereas the centerline extraction and variables computation required 2.5 ± 0.5 minutes.
Conclusion: In the proposed work, we investigated the relevance of an automatic protocol for 4D flow data post-processing within the aorta, able to extract BAVs hemodynamic alterations and statistically significant “hot-spots” against healthy subjects.
Figure 1. Schematic representation of developed the semi-automatic pipeline: (a) raw data loading; (b) semi-automatic segmentation of the aorta; (c) automatic definition of the vessel centerline and cross-sectional planes; (d) automatic computation of hemodynamic variables and “hot-spots” identification.
1602
Global myocardial mechanics with 2 and 3-Dimensional cardiovascular magnetic resonance feature tracking in patients with myocarditis
Abstract
Background: Myocarditis is an important cause of cardiac morbidity and mortality. Left ventricular ejection fraction is not impaired in the majority of patients. Impairment of left ventricular (LV) deformation is an independent predictor of cardiovascular outcomes in various cardiovascular conditions despite normal LV ejection fraction Cardiovascular magnetic resonance myocardial feature tracking (CMR-FT) has recently proposed as a robust method that can provide quantitative measurements of myocardial mechanics with good reproducibility. The aim of our study was to evaluate 2D and 3D left ventricular (LV) CMR-FT parameters in patients with myocarditis.
Methods: 18 patients with clinical diagnosis of myocarditis and 32 controls subjects underwent a CMR study in a 1.5 Tesla scanner. CMR-FT was applied to standard short axis and 2-, 3- and 4-chamber views of vector-ECG gated CMR cine SSFP sequences. 2 and 3 Dimensional (2D and 3D) global peak longitudinal, circumferential and radial systolic strain values (GLS, GCS, GRS respectively) were measured. All CMR analysis was performed on commercially available software (Circle CVI 42®, Calgary, Canada).
Results: Patients and controls were well matched for age and gender. Patients with myocarditis showed impaired LVEF (myocarditis, vs. controls, % 57 ± 8 vs. 62 ± 5, p < 0.05), but not significant differences in LV end-diastolic volumes or LV mass. Myocarditis group exhibited attenuation of all 2D and 3D LV CMR-FT parameters (table). The subgroup of patients (n = 13) with normal LVEF (FEVI > 53%) showed significantly impaired 2D-GLS, 2D-GRS and 3D GRS (myocarditis, vs. controls, 2D-GLS,%: 16 ± 1.8 vs. 18.7 ± 2.1, p < 0.001; 2D-GRS,%: 25.8 ± 4 vs. 33.6 ± 6.1, p < 0.001; and 3D GRS,%: 31.9 ± 7.5 vs. 40.8 ± 10.8, p < 0.01).
Conclusion: LV deformation parameters are impaired in patients with myocarditis, even in patients with normal LVEF, providing novel insight into the pathophysiology of the disease. Its potential to predict clinical outcomes needs further evaluation.
1603
A CMR-based clinician-friendly assessment of in vivo left ventricle hemodynamics
Abstract
Objectives: 4D flow can provide a deeper insight into intra-cardiac hemodynamics, although laborious post-processing and high operator expertise are still required [Markl, 2014] holding back its usability in the daily clinical-practice. We herein proposed a clinically oriented platform to exploit 4D flow analysis starting from an easy-to-use segmentation of left ventricle (LV).
Methods: CMR acquisitions were performed at the Multimodality IRCSS San Donato (Milan, Italy) on a 1.5 T MR scanner (Aera, Siemens, Erlangen, Germany). 2D CINE short-axis slices were defined along the LV axis from the apex over the mitral valve (MV) annulus (Fig. 1a). 4D flow acquisitions were performed using oblique-sagittal planes encompassing the LV and the ascending aorta (Fig. 1b) with prospective ECG-gating, respiratory navigator and the lowest non-aliasing VENC. CINE images were used for fast semi-automated segmentation of the LV endocardium (ROILv) based on Otsu clusterization [Otsu, 1979] and 2D region growing (Fig. 1a). Two subjects (Males, age 65-58) were analyzed to test the effectiveness of the proposed approach. ROILv was extracted at the end-diastole frame to account for the maximum LV volumetric dilation. This mask was then automatically mapped into the 4D flow acquisition volume (Fig. 1b) to allow for data post-processing. User-defined cross-sections were used to automatically compute hemodynamics markers (e.g. velocity vectors and contours, streamlines, vorticity and wall shear stress maps) within the ROILv.
Results: MV position and diastolic filling direction well matched the CINE-based ROILv (Fig. 1c), confirming the reliability of the automatic registration. The diastolic net flow was quantified near the MV position: E/A ratios of 2.5 and 0.7 were found, suggesting the occurrence of high atrial-pressure filling and impaired relaxation, respectively. Velocity peaks (<100 cm/s) were identified on 2D velocity contours. Streamlines were computed to help visualizing the formation of sub-annular vortexes.
Furthermore, ROILv segmentation lasted less than 30 seconds and hemodynamic parameters were computed nearly real-time for a single cross-section, pointing out the clinical relevance of the platform.
Conclusion: We implemented a reliable user-friendly pipeline to ease 4D flow data post-processing exploiting clinical-conventional 2D sequences for fast LV segmentation. The preliminary hemodynamic analysis proved able to quantify intra-cardiac in vivo markers nearly real-time.
Figure 1. Definition of the steps implemented to ease left-ventricular 4D flow data post-processing: (a) 2D CINE images used for semi-automatic volumetric segmentation; (b) 4D flow volume for left ventricle automatic registration; (c) extraction of hemodynamic markers during the diastolic filling phase of the left ventricle.
1604
Reproducibility of left atrial strain using cardiovascular magnetic resonance myocardial feature tracking
Abstract
Background: Left atrium (LA) size is increasingly recognized to have an important role in determining prognosis and risk stratification in different cardiomyopathies. Reduced LA function is associated with diastolic dysfunction, myocardial fibrosis and with atrial arrhythmias. Cardiovascular magnetic resonance myocardial feature tracking (CMRFT) has recently been introduced for the evaluation of LA function. Standardization of post-processing, including assessment of reproducibility and discrimination between health and disease, is necessary for implementation of this technique in clinical practice. The aim of this study is to test inter and intra-observer reproducibility of LA deformation function in a group of healthy controls and patients with hypertrophic cardiomyopathy (HCM).
Methods: 10 patients with HCM and 10 healthy subjects underwent assessment of LA function by CMRFT. Two independent observers measured LA CMRFT in 2, 3 and 4-chamber views of vector-ECG gated CMR cine SSFP sequences, using cvi42® (Circle Cardiovascular Imaging, Calgary, Canada). Longitudinal strain (LS) and radial strain (RS) of the LA was analysed in all subjects. Intra- and inter-observer reproducibility of was assessed using Bland-Altman methods.
Results: Intra- and inter-observer agreements for LS (r = 0.96, mean difference (MD) ± SD = 0.71 ± 3.6; r = 0.92, MD ± SD = -6.9 ± 6.3 respectively, p < 0.001 for both) and RS (r = 0.96, mean difference (MD) ± SD = 0.16 ± 1.5; r = 0.91, MD ± SD = -2.61 ± 2.3 respectively, p < 0.001 for both) values across the whole cohort were high. Similarly, the intra- and inter-observer coefficients of variation (CoV) for LS (0.71%, 0.71%) and for RS: 0.71 and 0.72%) values were low. Compared to controls, LS and LD were significantly impaired in HCM (control vs. HCM, LS, %: 29.3 ± 11.8 vs. 15.7 ± 8.5; RS,%: 15.8 ± 4.2 VS. 10.24 ± 4.7 p < 0.01 for both).
Conclusions: We demonstrate that LS and RS provide accurate, reproducible and easy to apply measurements of left atrial deformation. Both parameters can detect LA dysfunction in a model of cardiomyopathy such us HCM.
1605
The severity of myocardial infarction in STEMI, determined by transmurality of infarct and infarct characteristics, impacts on myocardial T2 values
Abstract
Background: Infarcts in STEMI are heterogeneous. Infarcts can be subendocardial or transmural. Infarct characteristics vary from oedematous and salvaged myocardium, to infarcted tissue, to regions if microvascular obstruction (MVO) and intramyocardial hemorrhage (IMH). T2 mapping is a sequence that can identify infarcted and oedematous tissue. The aims of this study were to identify if there were differences in the T2 values of infarct by transmurality and the different infarct characteristics.
Methods: 50 patients with STEMI were prospectively recruited to the study. CMR was performed at day 2 following PPCI. Myocardial oedema was identified by T2w STIR. Infarct was identified 10 minutes following contrast administration on LGE imaging. MVO was identified as an hypointense region located within the hyperintense infarct. IMH was identified as a hypointense core on T2w STIR. T2 values were measured in these different infarct characteristics. In addition, infarct transmurality and T2 values were assessed per myocardial segment. All patients provided informed written consent and the study was approved by the regional ethics committee.
Results: There is a difference between T2 values for each characteristic (Figure 1). The T2 values in all infarct characteristics are significantly different from remote unaffected myocardium. Infarcted myocardium has a significantly higher T2 value than salvaged myocardium (p< 0.0001). MVO has a significantly decreased T2 value from infarcted myocardium (p = 0.03), with IMH having an even lower T2 value (p = 0.004). T2 values were different according to transmurality of LGE as compared to unaffected segments (Figure 2). LGE 0-25% median 56.6ms, IQR 53.9-60.1ms; LGE 25-50% median 60.1ms, IQR 55.0-67.0ms; LGE 50-75% median 63.7ms, IQR 60.3ms-69.4ms; LGE 75-100% median 64.7ms, IQR 60.0-69.8ms, (p < 0.0001).
Conclusions: T2 values are affected by severity of infarction as determined by transmurality of infarction and infarct characteristics.
Figure 1. T2 value for each characteristic of myocardial infarction.
Figure 2: T2 value according to transmurality of LGE
1606
MicroRNA as potential biomarkers of acute myocardial damage following STEMI
Abstract
Background: Cardiovascular magnetic resonance (CMR) can identify myocardial characteristics following ST segment elevation myocardial infraction (STEMI) that are predictors of poor prognosis. These parameters include early and persistent microvascular obstruction (MVO), and infarct size (IS). Certain circulating plasma microRNAs (mRNA) have been suggested as potential biomarkers in STEMI. The aim of the study was to determine the association between CMR characteristics of poor prognosis following STEMI and mRNA 133a, 208b, 194 and 214.
Methods: 50 patients with STEMI were prospectively recruited to the study. Patients had plasma mRNA 133a, 208b, 194 and 214 measured at 24 hours following onset of pain. CMR was performed at day 2 following STEMI. 46 patients had analyzable plasma mRNA. Early MVO was measured on dynamic imaging following contrast, persistent MVO was measured at 10 minutes following contrast administration The correlation between MVO and IS and mRNAs were assessed using Spearman Correlation, All patients provided informed written consent and the study was approved by the regional ethics committee.
Results: None of the 4 mRNA tested had a significant association with early MVO. mRNA 214 significantly correlated with persistent MVO. mRNA 133a and 208b significantly correlated with infarct size (Table 1).
Conclusions: Circulating plasma mRNA 133a and 208b may potential biomarkers to predict infarct size following STEMI.
Funding: This study was funded by the David Telling Charitable Trust and the NIHR Bristol Cardiovascular Biomedical Research Unit.
. | Early MVO . | Persistent MVO . | Infarct Size . |
---|---|---|---|
mRNA 133a | r = -0.17 (p = 0.26) | r = 0.22 (p = 0.13) | r = 0.32 (p = 0.03) |
mRNA 208b | r = -0.37 (p = 0.81) | r = 0.18 (p = 0.23) | r = 0.38 (p = 0.01) |
mRNA 194 | r = 0.10 (p = 0.51) | r = -0.01 (p = 0.96) | r = 0.02 (p = 0.87) |
mRNA 214 | r = -0.01 (p = 0.96) | r = 0.35 (p = 0.02) | r = 0.26 (p = 0.08) |
. | Early MVO . | Persistent MVO . | Infarct Size . |
---|---|---|---|
mRNA 133a | r = -0.17 (p = 0.26) | r = 0.22 (p = 0.13) | r = 0.32 (p = 0.03) |
mRNA 208b | r = -0.37 (p = 0.81) | r = 0.18 (p = 0.23) | r = 0.38 (p = 0.01) |
mRNA 194 | r = 0.10 (p = 0.51) | r = -0.01 (p = 0.96) | r = 0.02 (p = 0.87) |
mRNA 214 | r = -0.01 (p = 0.96) | r = 0.35 (p = 0.02) | r = 0.26 (p = 0.08) |
1607
Myocardial blush grade is associated with microvascular obstruction on CMR following STEMI
Abstract
Background: Traditionally angiographic measures at the time of primary percutaneous intervention (PPCI) have been used to assess the microcirculatory flow following restoration of flow to the epicardial infarct related artery (IRA); myocardial blush grade (MBG) is an established method. Microvascular obstruction (MVO) by cardiovascular magnetic resonance (CMR) is a predictor of poor prognosis following ST segment elevation myocardial infarction (STEMI).
Methods: 50 patients with STEMI were prospectively recruited to the study. Patients were included if they presented within 12 hours of pain, had TIMI 0/1 flow, and were proceeding with PPCI in a large epicardial vessel. MBG was assessed at the end of the PPCI procedure. CMR was performed at day 2 following PPCI; persistent MVO was measured at 10 minutes following contrast administration. The association between MBG with MVO was assessed using the Mann-Whitney test. All patients provided informed written consent and the study was approved by the regional ethics committee.
Results: MBG was split into 2 groups: MBG 0 or 1 and MBG 2 or 3. The MVO was indexed to infarct size (MVO/ infarct size). MBG was significantly associated with MVO (p = 0.04). (Figure 1)
Conclusions: MBG following successful PPCI is associated with MVO by CMR.
Funding: This study was funded by the David Telling Charitable Trust and the NIHR Bristol Cardiovascular Biomedical Research Unit.
Figure 1. MBG vs MVO (p=0.04)
1608
4D Flow CMR imaging: Comparison of conventional parallel imaging and variable density k-t acceleration
Abstract
Introduction: 4D flow CMR imaging is a phase contrast MRI technique with 3-directional velocity encoding that enables measurement and visualization of 3D blood flow (1) and quantitative analysis of hemodynamic parameters such as flow velocities and wall shear stress (WSS). We sought to investigate the impact of variable density k-t acceleration (2,3) (kat-ARC VD) on acquisition time, image quality and quantification.
Methods: We performed 4D flow CMR on 6 healthy volunteers using a Discovery 750w system (GE Healthcare, USA). Two scans without contrast were performed, one with conventional parallel imaging by Autocalibrating Reconstruction (ARC) and one with kat-ARC VD. The imaging parameters were identical for both measurements except the acceleration factors as below: 84 slices of 2.4mm, field of view of 380x228 mm, matrix of 180x 180 pixels, TR 2.1ms, flip angle 8 degrees. For the ARC sequence, the acceleration factor was 2, leading to an acquisition time of 14-16 minutes. For the kat-ARC VD sequence, the central part of k-space was accelerated by a factor 2, the outer part of 8, reducing the scan time to 5 minutes. We measured signal-to-noise ratio (SNR) for all time points on the magnitude images. Within the same ROI, flow speeds were calculated in the three flow-encoded directions. The total flow velocity was calculated as a vector sum from the three individual components.
Blood flow visualization,quantification of blood flow, velocities and WSS have been performed using CAAS MR 4D Flow (Pie Medical Imaging, The Netherlands).
Results: SNR was higher on the ARC measurement (p < 0.05). Despite SNR differences the reconstructed flow curves were comparable for both techniques as shown in Fig. 1. Fig. 2 shows an example of the streamline, vector fields and WSS. Visual inspections showed that the source images were better on the ARC acquisitions. However, flow quantification and wall shear stress were comparable by both methods.
Conclusions: With kat-ARC VD, less k-space data can be acquired than in ARC, leading to faster scanning time. Despite an observed decrease in SNR with kat-ARC VD, we show that reconstructed flow graphs are hardly influenced by the two different techniques. However, visual inspections appear still better with the conventional ARC. Further quality control will be performed for clinical routine implementation.
References:
1. Dyverfeldt et al. JCMR 2015.
2. Okuda et al. ISMRM. 2015.
3. Lai P et al. JCMR 2015.
1609
In-vitro comparison of segmented-gradient-echo versus non-segmented echo planar imaging 4D Flow CMR: validation of flow volume and 3D vortex ring assessment
Abstract
Background: 4D Flow CMR enables in-vivo acquisition of all flow velocity components allowing analysis of flow patterns. Intra-cardiac vortex ring flow pattern has gained interest as a potential marker of cardiac health [1]. Different acceleration approaches for 4D Flow CMR acquisition are currently in use, but comparison of these techniques and their impact on vortex ring properties is lacking. Using a previously presented vortex ring pulsatile phantom [2], this work aims to compare two 4D Flow CMR acceleration protocols, a non-segmented echo planar imaging (EPI) versus segmented fast gradient-echo (TFE) approach with respect to: 1) stroke volumes (SV) relative to 2D phase contrast (PC) and 2) 3D vortex ring flow properties.
Methods: A pulsatile flow phantom (57 bpm) (Figure 1) generating a symmetric vortex ring propagating in a water tank was used [2]. Two pump settings with two stroke ratios (VFR = stroke_length/nozzle_diameter) are used [2]: a low (VFR1) and a high stroke (VFR3). Conventional 2D PC was performed inside the nozzle and perpendicular to the flow direction. 4D Flow CMR using either EPI or TFE acquisitions were performed. CMR was performed at 1.5T and geometric properties were kept equal between both scans (spatial resolution 3 × 3 × 3mm3, 40 phases reconstructed, VENC 100cm/s). EPI factor 5 was applied versus segmentation factor 2 for TFE. True temporal resolution was 32ms for EPI and 50ms for TFE. Total SV was first measured from 2D PC and used as reference. SV of 4D Flow EPI and TFE were measured after reformatting the same 2D plane. Percentage errors (%) in SV between 4D Flow and 2D PC were computed. 3D vortex rings were mathematically extracted and their propagation speed (VPS) in cm/s and circularity index (VCI) were compared between TFE and EPI [3].
Results: Acquisition time for EPI was 5.4 minutes, for TFE 11 minutes. Detailed results are listed in Table 1 and Figure 1. The relative error in SV was higher for TFE than EPI. Vortex properties (VPS and VCI) were not significantly different between TFE and EPI for both pump settings.
Conclusions: Global vortex ring properties are in good agreement between both EPI and TFE while a potential discrepancy might exist in their stroke volume measurements. 4D Flow EPI, with only half the acquisition time of TFE, shows promise as an accelerated sequence towards a clinically feasible acquisition duration.
References:
1. Gharib et al. PNAS 2006.
2. Toger et al. MRM 2015.
3. Elbaz et al. JCMR 2014.
1614
Not just 2D but also 4D flow measurements in pulsatile phantom are accurate and reproducible
Abstract
Introduction: 4DFlow is an emerging technique with high potential in the evaluation of congenital and acquired heart diseases. The validation of accuracy and reproducibility of flow measurements in 4DFlow is of key importance for the clinical use of this technique. Here, we compare 2D and 4DFlow using a robust pulsatile flow phantom setup with different tube configurations and heart rates.
Methods:Acquisition: A closed-circuit pulsatile flow phantom was built; composed by an industrial membrane flow pump, two counter-flow parallel silicone tubes (inflow tube at isocenter and center of FoV and outflow tube out of isocenter) and a Coriolis flow meter (Endress + Häuser), to measure a ground-truth net flow measure with a precision of about ±1ml/s. Two configurations were tested: 1) with tubes parallel to the z axis of the scanner (axial), and 2) with tubes placed in a double oblique.
Flow phantom scans were performed on a GE 3T MR750w MR scanner (Waukesha, WI). Retrospectively gated PC; “breathhold” and “free-breathing” (with 3 averages) 2D through-plane FastCINE and kat ARC 4D Flow were scanned [1]. Table 1 summarizes the flow phantom configuration setups and the main MRI protocol parameters.
Analysis: PC net flow measurements were evaluated using CVI42 software (Calgary, Canada) for 2D and Arterys (San Francisco, CA) for 4DFlow. For both, background phase correction (BPC) using static phantom correction [2] and image-based correction [3] was applied. Statistical analyses were performed in Microsoft Excel. Intraclass Correlation Coefficients (ICC) was used to evaluate reproducibility. The Bland-Altmann plot was used to compare the obtained net flow results using MR to the ground-truth value measured by the flow meter in percentage difference as a measure of accuracy.
Results: At the position of the outflow tube (>8cm from isocenter) before BPC the highest background phase error of 13% for the lower VENC was found. However, Figure 1 shows a mean percentage error of less than 8% (less than 6% for 4DFlow) with respect to the ground-truth flow for both tubes and for all configurations after BPC. Reproducibility was excellent obtaining ICC = 0.97 for 20 datasets.
Discussion: 2D PC and 4DFlow across multiple imaging conditions and setup configurations for a pulsatile flow phantom setup are reliable and accurate.
References:
1. Lai P et al. ISMRM 2015.
2. Chernobelsky et al. 2007.
3. Tan et al. ISMRM 2014.
1615
Diffusion Tensor Imaging: Comparison of Hypertrophic Cardiomyopathy, Hypertension and Healthy Cohorts
Abstract
Introduction: It is important to differentiate hypertrophic cardiomyopathy (HCM) from other aetiologies of left ventricular hypertrophy (LVH). In vivo diffusion tensor imaging (DTI) may improve diagnostic accuracy through novel myocardial tissue characterisation. Progression of myocyte orientation is determined via the helix angle gradient (HAG); myocardial organisation is assessed by fractional anisotropy (FA); mean diffusivity (MD) is the diffusion coefficient; and myolaminar function is assessed through secondary eigenvector angulation (E2A). We sought to determine whether DTI could detect clinical differences between HCM, HTN and healthy controls.
Method: We recruited 25 HCM patients, 13 patients with hypertension and 15 age-matched controls. Three short-axis mid ventricular slices were acquired with multiple breath holds, during the systolic and diastolic pauses with a diffusion weighted, stimulated echo EPI sequence. Data was processed to calculate global systolic and diastolic HAG, FA, MD and E2A values.
Results: Both systolic and diastolic HAG were significantly less in HCM compared to HTN and controls (Systole: 6.2 ± 1.0 v 7.4 ± 0.9 v 8.6 ± 0.8°/mm, p < 0.001; Diastole: 6.5 ± 1.0 v 7.9 ± 0.9 v 8.9 ± 1.09°/mm, p < 0.001). Combining all 3 cohorts, there was an inverse correlation between HAG and wall thickness (Diastole: R2 -0.61, p < 0.001, Systole: R2 -0.70). Diastolic FA was less in HCM compared to controls (0.51 ± 0.05 v 0.55 ± 0.04, p = 0.009), but not significantly different to HTN (0.53 ± 0.04). Systolic MD was greater in HCM compared to controls (1.07 ± 0.13 v 0.91 ± 0.13x103mm2s-1, p = 0.002) but not significantly different to HTN (0.97 ± 0.14 x103mm2s-1). Systolic and diastolic E2A were greater in HCM compared to HTN and controls (Systole: 62 ± 5° v 56 ± 6° v 53 ± 4°, p < 0.001; Diastolic: 47 ± 7° v 35 ± 6° v 26 ± 5°, p < 0.001).
Summary: DTI may aid differentiation of HCM from other LVH aetiologies. Phasic E2A values, especially, were significantly different between cohorts. This protocol was, however, unable to detect myocardial disarray via FA. Reasons include both technical factors such as resolution, SNR and strain effects, and the small study sample. Future work aims to fully address the impact of strain on DTI parameters.
1624
Impact of myocardial fibrosis measured by cardiac magnetic resonance imaging on reverse left ventricular remodelling after transcatheter aortic valve implantation
Abstract
Background: Interstitial myocardial fibrosis (MF) is common in patients with aortic stenosis and can be visualised by delayed contrast-enhanced (DCE) cardiac magnetic resonance imaging (MRI). Data from patients undergoing surgical aortic valve replacement (AVR) suggest an inverse correlation between MF and an improvement of left ventricular (LV) after AVR. We previously observed reverse LV remodeling after transcatheter aortic valve implantation (TAVI). The aim of the current study was to evaluate the impact of MF on post-TAVI LV remodeling.
Methods: 36 patients treated with TAVI using a Sapien XT, Sapien 3 or Lotus prosthesis between 7/14 and 3/15 who underwent post-TAVI MRI were assessed. We performed baseline MRI at a median of 9 days and follow-up at 6 months after TAVI. MF was assessed using DCE imaging, analysing the presence and pattern–infarct-type/non-ischemic - of DCE, the number of affected segments (NoS) following the 17-segments-model, and a severity score multiplying the NoS with DCE with a factor of 1, 2 or 3 in accordance with the visual amount of DCE in each segment.
Results: The mean age of the patients was 79.6 ± 6.2 years; 58% were women. We observed an increase of LVEF from baseline to follow-up (59.9% vs. 63.9%, p = 0.006) and a reduction of LV mass (138.7g vs. 123.4g, p = 0.0007). LV enddiastolic volume (EDV) remained unchanged (127.6ml vs. 120.5ml, p = 0.19).
Only 2 patients of the study cohort (6%) had no evidence of DCE at baseline. In the remaining 34 patients (94%) non-ischemic DCE was observed with a mean NoS of 3.6 ± 2.0 and a mean score of 5.7 ± 3.5. Non-ischemic DCE was predominantly distributed in the basal and midventricular septal and basal inferolateral segments. Infarct-type DCE was detected in 7 patients (19%), with a median of two segments involved. No correlation was observed between the NoS with DCE and changes of LVEF (r = 0.02, p = 0.89), EDV (r = -0.08, p = 0.63) or LV mass (r = -0.28, p = 0.10). Equally, using the severity score, no correlation was found with changes of LVEF (r = 0.02, p = 0.92), EDV (r = 0.01, p = 0.95) or LV mass (r = -0.18, p = 0.30). No differences in functional or structural LV remodeling were observed when the cohort was divided along the median into patients with a DCE score ≤5 and >5 and patients with NoS ≤3 and patients with NoS >3, respectively.
Conclusions: In this elderly TAVI cohort, no significant correlation was seen between the presence and severity of MF visualized by DCE and LV functional or structural remo
1625
Prosthetic valve regurgitation after transcatheter aortic valve implantation with new-generation devices compared to surgical aortic valve replacement–a cardiac magnetic resonance imaging flow measurement analysis
Abstract
Background: Transcatheter aortic valve implantation (TAVI) has become a widely used alternative for patients with severe symptomatic aortic stenosis who are considered to be at high risk for cardiac surgery. Post-TAVI paravalvular aortic regurgitation (AR) has been frequently observed, whereas AR after surgical aortic valve replacement (SAVR) is thought to be a rare finding. However, contemporary transcatheter valves have been modified to prevent relevant paravalvular leaks. Cardiac magnetic resonance imaging (MRI) is a non-invasive, observer-independent method for the quantification of post-TAVI AR, and has been shown to be more reliable than echocardiography. Thus, we sought to compare the frequency and severity of AR after TAVI with new-generation devices compared to SAVR.
Methods: Patients aged over 65 years who underwent transfemoral TAVI using the Edwards Sapien 3 valve or the Lotus valve and patients who underwent SAVR using bioprosthetic valves between July 2014 and September 2015 and who had no contraindications for MRI were included into the analysis. TAVI patients underwent cardiac MRI at a median of 10 days after valve implantation and the SAVR patients at a mean of 24 days after surgery. Phase-contrast MRI was conducted to quantify the degree of AR. A calculated regurgitant fraction (RF) ≤15% was graded as I (mild), 16–30% as II (moderate), 31–50% as III (moderate to severe) and > 50% as grade IV (severe) AR, a RF <1% was classified as no AR (grade 0).
Results: 56 patients were included into the TAVI group and 30 patients into the SAVR group. TAVI patients were significantly older than SAVR patients (80.6 ± 7.1 years vs. 74.1 ± 4.4 years, p < 0.001) and more frequently women (52% vs. 20%). Among TAVI patients, the majority (n = 42, 75%) had mild post-procedural paravalvular AR. Moderate AR was observed in 4 patients (7%) and 10 patients (18%) had no AR using MRI flow measurements. Similarly, 24 patients of the surgical AVR group had mild AR (80%), moderate AR was present in 1 patient (3%), and no AR was observed in the remaining 5 patients (17%). The median RF in the TAVI group was 3.12% (IQR 1.45 to 5.85), which was almost identical to the median RF of 3.45% (IQR 1.52 to 5.85) observed in the SAVR group (p = 0.84).
Conclusions: Using MRI flow measurements, patients treated with new-generation transcatheter aortic valves with an adaptive seal have similar rates of paravalvular AR compared to patients treated with surgical aortic valve replacement.
1637
Assessment of Aortic and Pulmonary Artery stiffness in Patients with COPD using Cardiac Magnetic Resonance
Abstract
Background: COPD has been associated with increased cardiovascular risk, although mechanisms are still investigated. One of the proposed theories is that the increased systemic induced inflammation and accelerated aging caused by the chronic respiratory process may increase arterial stiffness.
Aim: We aimed to evaluate aortic and main pulmonary artery distensibility in patients with COPD in comparison to a normal age-matched control group.
Methods: We recruited 51 subjects, of which 29 with diagnosis of COPD and FEV/FVC <70%, and 22 normal age-matched controls (mean age 64 ± 10 years). MRI images were acquired using a 1.5T scanner, and analysed using CVI42 software. Left ventricle and right ventricle function and volumes were evaluated using short axis SSFP cine. Aortic and pulmonary distensibility were measured using a validated method that take in consideration the arteries maximal and minimal areas from axial SSFP cines acquired perpendicular to both vessels.
Results: Aortic distensibility was reduced in the COPD patients compared to control (0.0022610 X10 -3mm Hg -1 vs 0.004337 X10 -3mm Hg-1, p = 0.003). Pulmonary distensibility was also reduced in COPD patients (0.195344mmHg) vs control (0.277923mmHg) (p = 0.005). The distensibility of descending aorta was similar in both groups (p= 0.06). Ejection fraction and biventricular volumes were also similar in the two groups.
Conclusion: Patients with COPD have significantly increased aortic and main pulmonary stiffness measured by cardiac magnetic resonance. This was observed in the presence of normal LV/RV systolic function in both groups. Reduced aortic and pulmonary artery distensibility could represent the early phase changes in cardiovascular function in these patients but further research is needed.
1638
Myocardial Mechanics implications using 2D Cardiovascular Magnetic Resonance in Aortic Regurgitation
Abstract
Background: The implications of left ventricular (LV) strain in patients with aortic regurgitation (AR) are not well known. Cardiovascular magnetic resonance myocardial feature tracking (CMR-FT) has recently been proposed as a robust method that can provide quantitative measurements of myocardial mechanics. The purpose of this study was to measure LV strain changes in AR patients and to compare its results with a healthy population.
Methods: 46 patients underwent a CMR study (1.5T). Of these, 14 (30.4%) had severe AR and 11 (23.9%) had moderate AR. 21 (45.7%) healthy individuals were included as controls. CMR-FT was applied to standard short and long axis views of CMR cine SSFP sequences. 2D global peak longitudinal, circumferential and radial systolic strain values (GLS, GCS, GRS respectively) were measured. CMR analysis was performed on Circle CVI 42® (Calgary, Canada) software.
Results: AR and healthy control patients were demographically similar. Absolute GLS, GCS, GRS values decreased as AR severity increased (Figure1). Thus, patients with severe AR showed significantly lower GLS, GCS, and GRS values compared to healthy controls (Table1). Moderate AR patients showed a significantly lower GLS absolute value (16.83 ± 4.5 Vs 19.51 ± 2.8, p = 0.046), and a trend to lower GCS (17.67 ± 4.27 Vs 19.66 ± 2.67, p = 0.115) and GRS (GRS–Short Axis 33.48 ± 11.4 Vs 39.16 ± 8.76, p = 0.127) (GRD–Long Axis 31.67 ± 12.14 Vs 37.11 ± 9.14, p = 0.164) values.
Conclusions: LV strain measurement by CMR-FT has shown a relationship with AR degree. Its values decreased as AR severity increased. Further studies are needed to confirm this finding and to clarify its clinical implication.
1639
Delineation of myocardial infarction & viability by 12 lead ECG vs cardiac magnetic resonance
Abstract
Background: Q-waves on 12 lead ECG is considered a marker of a large and/or transmural myocardial infarction (MI). Late gadolinium enhancement (LGE) cardiovascular magnetic resonance (CMR) accurately identifies myocardial infarction and has become the gold standard for the assessment of myocardial viability. However, CMR is not universally available and clinicians often have to make assumptions on viability based solely on the presence and location of ECG Q-waves.
Aim: To determine the diagnostic accuracy of Q-waves on 12 lead ECG to identify myocardial scarring and predict regionality. To ascertain the CMR predictors of Q-wave.
Methods: Data was collected on consecutive patients referred for a stress CMR with suspected ischaemic heart disease (from April 2013 to Mar 2014). Patients with non-ischemic heart disease were excluded from the analysis. Pathological Q-waves was defined as deflection > 25% of the subsequent R wave, or being > 40ms in width and > 2 mm in amplitude in >1 corresponding lead, Q waves in any 2 or more precordial leads from V1-V4 reflected LAD territory. A comprehensive CMR protocol was used. Transmural infarction was defined as >50% LGE. Univariate and multivariate logistic regression analyses were performed to determine the relation between the presence/absence of Q waves and CMR variables.
Results: 498 patients were included (mean age of 64 ± 12 years, 71% males). 290 patients demonstrated MI, 157 were transmural and 133 sub-endocardial based on CMR LGE. The overall diagnostic accuracy of Q-wave as a marker of transmural MI was 66% and the diagnostic accuracy of Q waves as a predictor of previous MI (composite of sub-endocardial and transmural) was only 55%. Table 1. 126 had pathological Q-waves on 12 lead ECG, 40% had LAD territory Q waves, 55% non-LAD and 5% a combination. Of those with anterior (LAD) Q waves, 68% demonstrated LAD territory LGE and in non-LAD Q waves, 67% demonstrated a non-LAD territory infarct by LGE. Univariate predictors of the presence of a Q wave included any LGE, spatial extent of scar tissue, transmurality(>50 or >75% LGE), and total scar score. On multivariate analysis, total scar score and >75% thickness LGE remained significant for prediction of the presence of a Q wave on the ECG. Table 2.
Conclusion: Our study demonstrates that the presence of pathological Q-waves on 12 lead ECG is not only a poor marker of myocardial scarring, but also a poor predictor of viability and regionality of infarction when compared to CMR. The study also highlights that presence of Q wave on the ECG correlates only with total scar score and >75% wall thickness LGE. In their clinical decision making process, clinicians needs to be aware of the limitations of ECG Q-waves.
. | Sensitivity (%) . | Specificity (%) . | Positive Predictive value (%) . | Negative Predicitive value (%) . | Accuracy (%) . |
---|---|---|---|---|---|
Q waves vs transmural MI | 36.3 | 79.8 | 45.3 | 73.1 | 66.1 |
Q waves vs any MI | 32.8 | 85.1 | 75.4 | 47.6 | 54.6 |
. | Sensitivity (%) . | Specificity (%) . | Positive Predictive value (%) . | Negative Predicitive value (%) . | Accuracy (%) . |
---|---|---|---|---|---|
Q waves vs transmural MI | 36.3 | 79.8 | 45.3 | 73.1 | 66.1 |
Q waves vs any MI | 32.8 | 85.1 | 75.4 | 47.6 | 54.6 |
. | Univariate . | Multivariate . | ||||
---|---|---|---|---|---|---|
Parameter . | Odds . | 95% CI . | P- value . | Odds . | 95%CI . | P-value . |
Age | 1.0 | .98-1.02 | .706 | 1.0 | .98-1.01 | .976 |
Sex | 1.23 | .76-1.99 | .393 | 0.86 | .50-1.47 | .592 |
LGE | 2.99 | 1.84-4.84 | <0.001 | 1.76 | 0.83-3.33 | .147 |
>50% LGE | .343 | 0.22-0.58 | <0.001 | 0.96 | 0.47-1.93 | .899 |
>75% LGE | .533 | 0.30-0.92 | .024 | 2.29 | 1.05-4.99 | .037 |
No. of segments | 1.19 | 1.11-1.26 | <0.001 | 0.82 | 0.65-1.04 | .104 |
Scar score | 1.07 | 1.05-1.10 | <0.001 | 1.15 | 1.04-1.27 | .004 |
. | Univariate . | Multivariate . | ||||
---|---|---|---|---|---|---|
Parameter . | Odds . | 95% CI . | P- value . | Odds . | 95%CI . | P-value . |
Age | 1.0 | .98-1.02 | .706 | 1.0 | .98-1.01 | .976 |
Sex | 1.23 | .76-1.99 | .393 | 0.86 | .50-1.47 | .592 |
LGE | 2.99 | 1.84-4.84 | <0.001 | 1.76 | 0.83-3.33 | .147 |
>50% LGE | .343 | 0.22-0.58 | <0.001 | 0.96 | 0.47-1.93 | .899 |
>75% LGE | .533 | 0.30-0.92 | .024 | 2.29 | 1.05-4.99 | .037 |
No. of segments | 1.19 | 1.11-1.26 | <0.001 | 0.82 | 0.65-1.04 | .104 |
Scar score | 1.07 | 1.05-1.10 | <0.001 | 1.15 | 1.04-1.27 | .004 |
1641
Regional variation in native T1 values in normal healthy volunteers?
Abstract
Background: Recent advances in CMR allow quick tissue characterisation based on the absolute quantifiable differences in recovery rates of longitudinal magnetisation by T1 mapping. T1 mapping is a robust and highly reproducible sequence, which is increasingly used in research as well as clinical examination. The reference normal range may be appropriate for comparing it with disease with global involvement however its usage in diseases with regional affection may be erroneous.
Aim: We assessed the regional variation in native T1 in healthy human myocardium.
Methods Healthy subjects (n = 20; mean age 41 years), with no previous medical history, were examined. Further inclusion criteria for all were absence of regular medication and subsequently normal findings of routine CMR. All subjects underwent T1 mapping using a uniform imaging set-up (modified Look- Locker inversion recovery, MOLLI, using scheme 3(3)3(3)5)) on 1.5 Tesla (T) Avanto (Siemens) scanner. Native T1-maps were acquired in 4, 3 and 2chamber long axis as well as base, midcavity and apical short axis slices. The images were motion corrected. T1 values were calculated by drawing region of interest in the 16 AHA segments obtained from the short axis slices using the Argus software (Siemens), with caution in not including the blood pool.
Results: The mean global native T1 values were 1028 ± 48 msec using the 16segment approach (average of all the 16segments) and 1021± 45msec from the midcavity short axis. There were significant differences between the native T1 values in the basal and midcavity segments (240 segments) compared to the apical segments (80 segments)(1023 ± 42 msec vs 1043 ± 62msec, respectively, p = 0.0009). There was also significant difference in the T1vaules in the basal and midcavity septum (80segments) compared to basal and midcavity lateral wall (80segments), 1033 ± 37 vs 1009 ± 43msec, p= 0.0002.
Conclusion: Our study has shown significant regional variation in the native T1 values in different segments of healthy normal myocardium. It is unclear whether this is a genuine difference or whether this is due to susceptibility artifacts, and further larger studies are warranted.
1645
Feasibility of myocardial strain assessment using tissue tracking at 3.0T CMR following ST-segment elevation myocardial infarction
Abstract
Background: Early detection of subclinical contractile dysfunction after ST-segment elevation myocardial infarction (STEMI) may allow targeted therapy to improve prognosis. Global myocardial strain can predict recovery, adverse LV remodeling and prognosis after STEMI. Strain can be measured by cardiac magnetic resonance (CMR) using various techniques but its widespread clinical utility is limited by the need to acquire additional sequences, time-consuming post-processing or diastolic tag fading in the case of myocardial tagging.
Objective: To assess the feasibility of myocardial strain assessment using a single (mid-ventricular) or three-slice (base/mid/apical) approach and comparing with strain derived from complete left ventricular (LV) coverage using 3.0T CMR in patients presenting with STEMI.
Methods: 183 patients underwent CMR 1-5 days following presentation with STEMI. Circumferential strain (Ecc) was measured using Tissue Tracking (TT), a novel plug-in for cmr42 (Circle Cardiovascular Imaging, Calgary, Canada), that computes strain on routinely acquired steady-state free precession (SSFP) cine sequences, using (1) all slices, (2) three (base, mid and apical) slices and (3) a single mid-LV slice from a complete short-axis (SAX) SSFP cine stack. Endocardial and epicardial contours and insertion points were manually defined at end-diastole. LV extent was defined on the long-axis 4-chamber slice. Ecc derived using a mid-LV slice, three (base/mid/apical) or all slices were compared using analysis of variance (ANOVA). Pearson's correlation coefficient was used to assess correlation. Inter-technique agreement was assessed using the Bland-Altman method, coefficient of variation (CoV) and two-way mixed-effect intra-class correlation coefficient (ICC) for absolute agreement.
Results: Patient characteristics are presented in Table 1. Ecc assessed using the three methods was not significantly different (p = 0.936) with excellent correlation and good agreement overall (see Table 2). Correlation and agreement between Ecc derived using all versus three slices were excellent.
Conclusions: Ecc derived using TT on three routinely acquired SAX LV slices may be used in lieu of ‘whole stack’ analysis for detecting early contractile dysfunction following STEMI. This may reduce analysis time considerably and facilitate migration of strain assessment in to the CMR routine for widespread clinical application. Our findings require validation in large multi-centre studies.
1648
Diagnostic Impact of Cardiac Magnetic Resonance in patients with acute chest pain, troponin elevation and no significant angiographic coronary artery disease
Abstract
Purpose: The absence of significant coronary lesions is described in approximately 10%-15% of hospitalized patients (P) with suspected acute coronary syndromes (ACS). The diagnostic and therapeutic approach of this P is often difficult. The main objective of this study is to characterize the diagnostic impact of Cardiac Magnetic Resonance (CMR) in P admitted with chest pain, troponin elevation and no significant angiographic coronary artery disease.
Methods: Prospective study conducted over a period of three years. During this period we performed CMR on all P with acute chest pain and troponin elevation and no significant angiographic coronary artery disease. All P underwent morphological and functional CMR and we looked for the presence of late enhancement. (LE) Cardiovascular risk factors, ST segment alterations on admission and maximum troponin value were evaluated. The D established after coronariography (C) and before CMR were compared with the definitive D established after CMR. The definitive D was established through the integration of clinical presentation and the findings of the CMR.
Results: We evaluated 49 P (27 men, mean age 53 ± 18,6 years).The presence of hypertension was 63,3%, diabetes mellitus 8,2%, dyslipidaemia 42,8%, smoking 36,7% and obesity 20,4%. On admission 25 P (51%) had ECG with ST segment elevation. Twelve of these P were referred to primary angioplasty. The average of maximum troponin I level was 6,36 ± 7,7 ng/mL. After C and before CMR the D established were: myocarditis in 22 P, Acute myocardial infarction (AMI) with spontaneous reperfusion in 22 P and takotsubo myocardiopathy (TM) in 5 P. In CMR the presence of late enhancement was found in 36 patients (73,5%). According to the distribution pattern of LE, it was possible to establish the following definitive diagnosis: myocarditis in 27 P, AMI with spontaneous reperfusion in 8 P, hypertrophic cardiomyopathy in 4 P and TM in 4 P. CMR allowed the maintenance of presumptive diagnostic in 19 P (38,8%) and it was possible to established a different diagnostic in 22 P (44,9%). In only 8 P (16,3%) it was not possible to establish a definitive diagnosis, although CMR excluded a type 1 AMI.
Conclusions: In our study population CMR allowed the establishment of a different diagnosis or exclude the presumptive diagnosis in 61,2%% of the P, which has very different therapeutic and prognostic implications. CMR should be considered in the differential diagnosis in these cases.