Short versus conventional hydration for prevention of kidney injury during pre-TAVI computed tomography angiography

The study population consisted of consecutive patients with symptomatic aortic valve stenosis and impaired renal function who underwent pre-TAVI CTA. All patients were screened for CI-AKI prevention [12]. Inclusion criteria were chronic kidney disease (CKD) classification 3a or above (estimated glomular filtration rate (eGFR) <60 ml/min/1.73 m² based on the modification of diet in renal disease formula [14]), age >18 years and planned for CTA prior to TAVI. Exclusion criteria were known iodine allergy, contrast administration <7 days, already on haemodialysis, multiple myeloma, Waldenstrom’s disease and left ventricular ejection fraction <20% (not eligible to be randomised to conventional hydration).

The study was designed as an open-label single-centre, unblinded, randomised non-inferiority trial and followed CONSORT (Consolidated Standards for Reporting Trials) guidelines for non-inferiority trials [15]. The protocol was registered in the Netherlands Trial Register (NTR5014), complies with the Declaration of Helsinki and was approved by the local ethics committee. All participants provided written informed consent before randomisation.

Randomisation was performed electronically at admission on a 1:1 ratio to either a short hydration protocol (SHORT; 1 h sodium bicarbonate prior to CTA, 1.4% 3 ml/kg/h) or conventional hydration (CONV; 24 h sodium chloride, 0.9% 1 ml/kg/h, 8 h prior to CTA, 16 h post-CTA). Block randomisation was used (block size 4, 6 and 8), stratified into four groups corresponding to CKD classification 3a, 3b, 4 and 5. The eGFR ranges per group were (3a) 45–59, (3b) 30–44, (4) 15–29 or (5) <15 ml/min/1.73 m².

Prior to CTA, patients were instructed to temporarily discontinue all nephrotoxic medication on the day of CTA. According to hospital protocol all patients with impaired renal function in TAVI work-up were hospitalised on the day of CTA. Hydration volumes were based on body weight. If patients developed acute heart failure, hydration was discontinued according to hospital protocol. Contrast administration at CTA was performed with 90 ml iopromide 300 mg I/ml (Ultravist 300; Bayer, Leverkusen, Germany; low osmolar, non-ionic) at a rate of 5 ml/s. No other CI-AKI preventive measures were used. Directly prior to the start of hydration blood serum and urine samples were collected. The level of dyspnoea was scored using the Borg CR10 scale at baseline and within 1 h after completing the hydration. The follow-up blood sample was collected 2–5 days after CTA at the outpatient clinic or general practitioner and sent to the regional laboratory service for analysis.

The primary outcome was the percentage change in serum creatinine between the two hydration protocols at 2–5 days after contrast administration as compared to baseline creatinine. Based on previous research the non-inferiority margin for the percentage change was set at a 10% difference between the two hydration groups [16].

Predefined secondary outcomes were the absolute change in creatinine, the occurrence of CI-AKI (defined as an increase in creatinine >25% or 44.2 µmol/l within 2–5 days [17]), an increase on the self-reported level of dyspnoea according to the Borg scale and, as a marker for acute heart failure, the absolute change in N-terminal prohormone of brain natriuretic peptide (NT-proBNP) within 2 and 5 days. In addition we analysed the mean change in eGFR and hydration volumes.

The trial was designed as a non-inferiority study and was powered on the relative change in creatinine based on a prior hydration trial [16]. A predefined non-inferiority margin of 10% was used. The one-sided two-sample t-test of equivalence of means was used to compare means. With a‑one sided α = 0.025 and β = 0.15, it was calculated that 35 patients per arm would be sufficient (70 patients in total). To correct for a dropout of approximately 15% the target sample size was determined as 80 patients.

Study results were analysed on a modified intention to treat (mITT) basis; all patients that did not receive the allocated intervention (hydration) or with failed follow-up measurements were excluded from the main analysis. Baseline categorical variables were presented as numbers with percentages, continuous variables as means with standard deviations (SD) or medians with inter-quartile ranges (IQR). Non-inferiority for the percentage change in creatinine was calculated using a two-sided, two-sample t-test. The mean of the difference with its 95% confidence interval was compared with the predefined non-inferiority margin; a mean difference between the two hydration protocols of less than 10% was considered non-inferior. A p-value for the non-inferiority test was calculated using a one-sided, two-sample t-test. All secondary outcomes were calculated using superiority testing applying two-sided t-tests or Fisher exact test as appropriate.

Hydration with the short protocol has the advantage of lower fluid load administration, decreasing the risk of acute heart failure. In our trial, only one patient hydrated with the conventional protocol developed acute heart failure. We used NT-proBNP as a marker for decompensation and found no difference between the two hydration arms at 2–5 days after contrast administration. Shortness of breath was a more sensitive symptom of developing acute heart failure in this population, increasing the importance of dyspnoea assessment. Although the study sample was not powered to detect any statistical differences on the Borg CR10 dyspnoea scale, we did see a substantial difference in patients with increased dyspnoea between the two groups.

This single-centre study had its limitations. We had a small but appropriate sample size for the primary endpoint. Although there were no patients in our study who developed CI-AKI, we are aware that the definition of CI-AKI is based on the cut-off values for serum creatinine change and differs across guidelines [19‐21]. Our primary outcome was the percentage change in serum creatinine. We used a common definition for the development of CI-AKI, which is not the most strict definition but is predictive for long-term mortality and corresponds closely to a high degree of kidney injury according to the RIFLE criteria (risk, injury, failure, loss of kidney function, and end-stage kidney disease ) and the Acute Kidney Injury Network (AKIN) [19‐21]. Even when analysing our data using the most strict definitions to detect kidney injury, none of the patients were considered to have developed CI-AKI. This is remarkable, since most studies with follow-up after CTA report at least a small percentage of patients developing CI-AKI. A possible reason could be that despite the broad inclusion criteria, most patients in our study had a relatively ‘high’ eGFR >30 ml/min/1.73 m², with a lower risk of developing CI-AKI. Nevertheless, this distribution of eGFR represents our current TAVI cohort and the results can be interpreted as those of a real-world TAVI population. Furthermore, this is a population at high risk of developing renal injury based on their high age and a high prevalence of diabetes mellitus, hypertension and albuminuria. This would make it easier to extend the results to other patient groups at lower risk for CI-AKI.

Lastly, although we examined dyspnoea, this study was underpowered to detect a significant difference in patient-perceived exertion. The trend towards more dyspnoea in patients hydrated with the conventional protocol was clear but not significant.

We included a homogeneous but complex population of high age, with severe aortic valve stenosis and multiple comorbidities, prone to develop CI-AKI. Since the short hydration protocol was feasible and safe for these patients, a more general population is likely to benefit equally from the same strategy.

New indications for CTA with similar amounts of contrast, such as coronary CTA, are emerging. Most likely implementation of a short hydration protocol will induce more flexible CTA planning in an outpatient setting, a reduction of admission days as well as of the risk of hospital-acquired complications, and a reduction of costs while improving patient comfort [16].With the current technological advances in CTA, the volume of contrast agent necessary per study for quality imaging is decreasing [22, 23]. It is conceivable that with current techniques, the risks for CI-AKI after CTA will diminish over time.

Our findings are in line with recent guideline changes by the Dutch Society for Radiology and embedded in the guideline database of the Dutch Society for Clinical Specialists [24].