Baseline Predictors of Renal Failure in Transcatheter Aortic Valve Implantation.

BACKGROUND

Acute kidney injury (AKI) post transcatheter aortic valve implantation (TAVI) is associated with worsened short- and long-term outcomes. We sought to identify significant baseline predictors of AKI and establish a high-risk group within patients enrolled in the multicenter SWISS-TAVI cohort.

METHODS AND RESULTS

A total of 526 patients who underwent TAVI were included in our analysis. Patients on hemodialysis were excluded. Within the first week after valve implantation, fifty patients (9.5%) developed AKI. There was a significantly higher prevalence of diabetes mellitus in the AKI group (45% vs 28%; P=.02). The odds ratio (OR) for patients suffering from diabetes mellitus who developed AKI was 1.9 after multivariable binary regression analysis (95% confidence interval, 1.018-3.553; P=.04). Chronic kidney disease (CKD) stage ≥4 was more prevalent in the AKI group (26% vs 14%; P=.04). Every 1 mg/dL creatinine above normal level at baseline increased AKI risk by a factor of 1.6 (OR, 1.605; 95% CI, 1.111-2.319; P=.01). Age, gender, body mass index, history of dyslipidemia, and history of hypertension were similar between the groups. In the diabetic population of 155 patients (29.5%), AKI developed in 22 patients (14.2%), compared with the non-diabetic population of 370 patients (70.5%), where AKI developed in 27 patients (7.3%). In the diabetic population, an elevation by 1 mg/dL in baseline creatinine was an independent predictor of developing kidney injury (OR, 2.061; 95% CI, 1.154-3.683; P=.02, while in non-diabetic patients, neither baseline glomerular filtration rate, CKD grade, STS score, EuroScore II, ACEF score, nor procedural contrast usage were predictors of AKI.

CONCLUSION

Diabetics with CKD stage ≥4 (as defined by the Kidney Disease: Improving Global Outcomes criteria) constitute a high-risk group for developing AKI after TAVI. In this high-risk subgroup, baseline creatinine in combination with amount of contrast agent used were strong risk factors for developing AKI. AKI in non-diabetics was less predictable by baseline characteristics.