ABSTRACT
BACKGROUND: Blood transfusion is associated with acute kidney injury (AKI) after transcatheter aortic valve implantation (TAVI). We sought to elucidate in more detail the relation between blood transfusion and AKI and its effects on short- and long-term mortality. METHODS AND RESULTS: Nine hundred ninety-five patients with aortic stenosis underwent TAVI with the Medtronic CoreValve or the Edwards Valve in 7 centers. AKI was defined by the Valve Academic Research Consortium (absolute increase in serum creatinine ≥0.3 mg/dL [≥26.4 µmol/L] or ≥50% increase ≤72 hours). Logistic and Cox regression was used for predictor and survival analysis. AKI occurred in 20.7% (n=206). The number of units of blood transfusion ≤24 hours was the strongest predictor of AKI (≥5 units, OR, 4.81 [1.45-15.95], 3-4 units, OR, 3.05 [1.24-7.53], 1-2 units, OR, 1.47 [0.98-2.22]) followed by peripheral vascular disease (OR, 1.48 [1.05-2.10]), history of heart failure (OR, 1.43 [1.01-2.03]), leucocyte count ≤72 hours after TAVI (OR, 1.05 [1.02-1.09]) and European System for Cardiac Operative Risk Evaluation (EuroSCORE; OR, 1.02 [1.00-1.03]). Potential triggers of blood transfusion such as baseline anemia, bleeding-vascular complications, and perioperative blood loss were not identified as predictors. AKI and life-threatening bleeding were independent predictors of 30-day mortality (OR, 3.15 [1.56-6.38], OR, 6.65 [2.28-19.44], respectively), whereas transfusion (≥3 units), baseline anemia, and AKI predicted mortality beyond 30 days. CONCLUSIONS: AKI occurred in 21% of the patients after TAVI. The number of blood transfusions but not the indication of transfusion predicted AKI. AKI was a predictor of both short- and long-term mortality, whereas blood transfusion predicted long-term mortality. These findings indicate that outcome of TAVI may be improved by more restrictive use of blood transfusions.