Letter by Karkouti and Beattie Regarding Article “Aprotinin Does Not Increase the Risk of Renal Failure in Cardiac Surgery Patients”
To the Editor:
We read with interest the observational study by Furnary and associates1 on the risks of aprotinin and cardiac surgery. Although we agree with the principal conclusion of the study, which is that observational studies assessing the risks of aprotinin in cardiac surgery may be unduly influenced by selection bias, it is our opinion that the study contained several errors relating to our previously published study on the same topic.2
First, we did not find aprotinin to be independently associated with renal failure requiring dialysis; rather, we only found an association between aprotinin and renal dysfunction (as measured by the highest creatinine within seven days after surgery).2
Second, we did not adjust for in-hospital red blood cell (RBC) transfusion because it is more an outcome than a confounder. That is, many patients require RBC transfusions after cardiac surgery as a direct consequence of perioperative complications such as progressive renal dysfunction. Patients with progressive renal dysfunction, for example, may receive more RBC transfusions because: (1) their hospital stay is prolonged; (2) they undergo more blood tests; and (3) clinicians may increase their RBC transfusion trigger to enhance oxygen delivery to the failing organs. By adjusting for in-hospital RBC transfusions in their analysis, Furnary and associates1 may have inadvertently masked any harmful renal effects of aprotinin. Adjusting for perioperative transfusions, on the other hand, may arguably be valid.
Third, our finding that aprotinin is associated with renal dysfunction is now confirmed by meta-analyses of randomized controlled trials.3 It is simply not correct that the Food and Drug Administration issued the warning on the renal risks of aprotinin based solely on the results of the observational studies. Rather, the FDA concluded that “aprotinin increases the risk for renal dysfunction, but the data do not establish an increased risk for renal failure requiring dialysis” based on the totality of clinical data, including the available randomized controlled trials.4
Finally, although it is true that randomized controlled trials did not find an association between aprotinin and renal failure requiring dialysis, it is important to note that these studies were underpowered (singly or in combination) to detect the effects of aprotinin on this and other infrequent complications. Furnary and associates,1 in their analysis of the subgroup of patients who did not receive RBC transfusions (n=5729), clearly demonstrate the issue of inadequate power. Using this large subgroup, which is comparable in size to the total number of subjects in placebo-controlled randomized controlled trials of aprotinin,3 they were unable to obtain a precise estimate of the association between aprotinin and renal failure, as is evident by the wide confidence interval (0.40 to 4.33) around their point estimate of the odds ratio (1.31). Thus, on the basis of the totality of clinical data, one cannot exclude the possibility that aprotinin causes renal failure requiring dialysis. An adequately powered randomized clinical trial including subjects who are at high risk for renal failure is urgently needed to answer this question.
Both authors have received speakers’ fees from Bayer, and Dr Beattie has acted as a consultant to Bayer.
Brown JR, Birkmeyer NJ, O'Connor GT. Meta-analysis comparing the effectiveness and adverse outcomes of antifibrinolytic agents in cardiac surgery. Circulation. 2007; 115: 2801–2813.
US Food and Drug Administration. FDA Advisory Committee briefing document: joint meeting of the Cardiovascular and Renal Drugs Advisory Committee and the Drug Safety and Risk Management Committee. Available at: http://www.fda.gov/ohrms/dockets/ac/07/briefing/2007-4316b1-00-index.htm. Accessed September 19, 2007.