Evaluation of extracorporeal shock wave therapy for refractory angina pectoris with quantitative analysis using cardiac magnetic resonance imaging: a short communication

Patients of 18 years or older suffering from refractory angina pectoris where included in the study. Refractory angina pectoris is defined as having angina pectoris NYHA class 3–4 based on documented ischaemia without any possible revascularisation option left, either PCI or CABG, under a maximum tolerable dose of medication. Patients were excluded from the study when they suffered from severe heart valve disease, intraventricular thrombus, a malignancy in the area of treatment, chronic lung disease (including emphysema and pulmonary fibrosis), a myocardial infarction in the 3 months prior to the start of the study, had a contraindication for undergoing CMR, or when the patient was pregnant.

The protocol was approved by the medical ethics committee of the VU University Medical Center and all patients provided written informed consent.

After refraining from intake of competitive antagonists of adenosine in the 24 hours prior to examination, CMR was performed on a clinical 1.5 Tesla scanner (Avanto, Siemens, Erlangen, Germany) using a phased-array cardiac receiver coil. All images were ECG gated and acquired during mid-expiratory breath-holds of 10 to 15 seconds, depending on the heart rate.

Using the four-chamber long-axis cine in the end-systolic phase, 3 short-axis slice positions were determined for myocardial perfusion imaging, at 25, 50 and 75 % of the distance between the mitral valve annulus and the apex. Prior to imaging, patients received intravenous adenosine at a rate of 140 µg/kg/min for 3 minutes. Image acquisition was initiated simultaneously with the administration of a 0.1 mmol/kg bolus of a gadolinium-based contrast agent (Magnevist, Schering AG, Berlin, Germany) at a rate of 3 ml/sec. Images were acquired at the 3 aforementioned levels during each cardiac cycle, for a total duration of 50 cardiac cycles. Adenosine infusion was discontinued immediately after the images were obtained.

At least 10 minutes after stopping adenosine infusion and myocardial perfusion acquisition, resting myocardial perfusion imaging was performed, using the same sequence parameters and contrast dose (total cumulative dose 0.2 mmol/kg). The sequence parameters used for cine and myocardial perfusion imaging are mentioned in Table 1.

Shockwave therapy was applied using the Cardiospec (Medispec, Germantown, Maryland). The location of the ischaemic zone was detected by consensus of 2 experienced observers based on visual inspection of CMR, using regional function. This area was translated into the 16-segment model of echo and divided into three parts. In order to make sure we covered the entire ischaemic zone we enlarged the treatment area. Therefore, more regions were treated than located on CMR. The treatment protocol covered a total period of 9 weeks in which patients were treated in weeks 1, 5 and 9. During each treatment week the patient visited the outpatient clinic 3 times, resulting in a total of 9 treatments.

Patients were positioned in a supine position and connected to continuous ECG monitoring. An S3 ultrasound probe, which was linked to a Sonos 4500 (Philips, Best, the Netherlands) and connected to the Cardiospec, was positioned on the chest of the patient so that the ischaemic treatment zone was located in the centre of the ultrasound image. The depth of the ischaemic zone was measured using a caliper on the ultrasound machine. Based on these measurements a balloon with a shockwave electrode inside and attached to the Cardiospec was placed on a patient’s chest at the appropriate position. Using an ellipsoid reflector inside the balloon enabled us to focus the shockwave with the right intensity to any possible depth. The intensity used was based on pre-clinical studies where an optimal dosage-effect relation was shown at approximately 10 % of the intensity used with lithotripsy, as was mentioned earlier [9]. Since the depth and location of the machine could be adjusted we were able to apply the same amount of energy to two remote distinct treatment zones. The balloon was inflated with saline water (5 % NaCl) until a good connection with the skin was achieved. Prior to inflation of the balloon a large amount of gel was applied to the skin in order to prevent any air becoming trapped between the skin and the balloon. Once installed, a total of 100 shocks per spot were applied. The amount of spots treated during a session depended on the size of the ischaemic zone, with every spot covering an area of approximately 1 by 1 cm.

Analysis was performed off-line using dedicated software (MASS v.5.1 2010-EXP beta, Medis, Leiden, the Netherlands) by consensus of 2 experienced observers. All analyses were performed on short-axis images with a 16-segmental distribution of the myocardium. Using the cine images, myocardial volumes during the end-diastolic and end-systolic phase were calculated. From the volumes, stroke volume and ejection fraction were calculated. The location of the treatment zone was based on visual inspection by consensus of 2 experienced observers.

Myocardial perfusion was evaluated quantitatively by calculation of the maximal relative upslope [%], using the maximum upslope [au·s⁻¹] of the myocardial signal-intensity-versus-time curves and by correcting for baseline signal intensity [au] and the arterial input by the upslope of the blood pool curve [au·s⁻¹] (Fig. 1). As mentioned above, in order to make sure we treated the entire ischaemic zone we enlarged the treatment area, resulting in treatment of healthy segments. To reduce the effect of these treated healthy segments and to reduce the effect of artifacts on analysis we divided the 3 short-axis views (base, mid and apical) into 12 segments per axis. Previous studies have already proven the relative upslope to be an accurate and reproducible parameter for quantitative perfusion analysis [16‐18].

Prior to the start of shockwave treatment all patients were asked to answer the Seattle Angina questionnaire to estimate NYHA class, the use of nitroglycerin and the frequency of angina pectoris. The same questionnaire was used 3 months after final treatment during follow-up CMR.

Categorical data are presented as frequencies (percentage), and continuous data as mean ± SD. Comparisons between means of groups were performed using Student’s T‑test (2 groups) and the Wilcoxon signed-rank test in case of ordinal variables. Obtained data were analysed using SPSS 21.0 (Chicago, USA).

Baseline and follow-up results are mentioned in Table 2. There was no significant difference in left ventricular volumes and ejection fraction between baseline and follow-up results. We did not find a significant difference in the myocardial perfusion ratio index, which is the relative ratio of the maximum upslope of stress to rest normalised to left ventricle input: 0.79 ± 0.11 vs 0.72 ± 0.24 (p = 0.40). For analysis of the myocardial perfusion ratio index the left ventricle was divided into a base, mid and apical part during analysis using CMR and each part contained 12 segments, in total 36 segments per patient. In one patient the left ventricular base images during follow-up were not suitable for analysis. Therefore, a total of 276 segments out of the initially available 288 segments (= 96 %) were analysed. Based on visual inspection, a total of 77 segments were marked as ischaemic segments and received treatment. Also, when group comparison was performed between the 77 treated and 199 untreated segments based on the myocardial perfusion ratio index there was no significant difference: 0.80 ± 0.22 vs 0.76 ± 0.31 (p = 0.42).

Neither NYHA class or frequency of nitroglycerin showed a significant improvement in patients analysed with CMR (p = 0.102 and p = 0.480, respectively). Although, a trend to NYHA class improvement was observed after treatment.

When the total group of 15 patients was analysed, as shown in Fig. 2, we found a significant improvement in NYHA class and reduction in nitroglycerin use (p = 0.034 and p = 0.012 respectively). Frequency of chest pain did not show a significant improvement (p = 0.405).