Aprotinin in 2008

�There are three kinds of lies: lies, damned lies, and statistics.’ Benjamin Disraeli 1804 - 1881

�Statistics: The only science that enables different experts using the same figures to draw different conclusions.’ Evan Esar 1899 - 1995

Has aprotinin still a place in cardiac surgery? The short answer is “yes.”

Perhaps the most controversial issue relating to cardiac surgery today relates the use of aprotinin. With all the recent information, one might well ask “Is there a place for aprotinin in current practice?”

A plethora of negative publicity has followed the publication in January 26, 2006 of an article in the New England Journal of Medicine. In a somewhat striking publication, it was accompanied by a wealth of venomous publicity in the lay press [1]. Additionally, legal websites touted for business. Mangano, using data collected from 1996 – 2000 (The McSPI EPI II database) reported an association of aprotinin use with an increase in mortality, renal dysfunction and MI after cardiac surgery. This was purportedly in patients undergoing revascularisation surgery for ST elevation myocardial infarction. This indication does not appear to be supported in other publications from the same database.

Subsequent events have included numerous publications that refute the findings of the NEJM paper, and some that support them. The FDA has convened 2 review panels with a wealth of experienced clinicians, including the authors of many of these reports and statisticians appearing to give evidence. After the first of these meetings in September, 2006, changes to the labelling for aprotinin in the USA, which usually flows on to other countries were agreed. These related to the possibility of anaphylaxis mostly related to re-exposure, and to the risk of renal dysfunction . The second meeting in September 2007, had access to FDA conducted analyses of data from Mangano’s database as well as details of the i3 study, somewhat controversially withheld from the first meeting in September 2006.

In this age of evidence-based medicine, we should look at the available evidence and how it is presented. This is where the difficulty in assessing the evidence lies. Much of the evidence may not be directly comparable. Much of the evidence lies in the minds of the users and non-users.

Firstly, Henry and the Cochrane team updated their review of antifibrinolytic therapy in cardiac surgery. The effectiveness of aprotinin for reduction of blood transfusion and return to the operating room for bleeding was confirmed. In their evaluation of 2 lysine analogues and aprotinin, Henry and colleagues noted that the incidence of MI, stroke, renal dysfunction and mortality was not increased in the 211 RCTs of non-urgent patients analysed. They comment “these results conflict with the results of recently published non-randomised studies.” [2]. Other meta-analyses from Sedrakyan looking at CABG patients only, concurred with the Cochrane review [3].

Brown reporting on 138 trials showed efficacy and an increased incidence of raised creatinine [4]. In the patient cohorts studied, there is little doubt there is a high degree of efficacy and a low rate of apparent drug related side effects.

Secondly, we are increasingly aware of the effects of blood transfusions. From the early 80s when HIV infection was found to be transmitted via blood products to the current day when TRALI is increasingly recognised. Koch has examined the impact of red cell transfusion on survival after CABG and found increased mortality [5]. Surgenor looked at the responses to anaemia, felt by his group in an earlier study to be the cause of morbidity and mortality[6]. The response to anaemia was transfusion. Both transfusion and anaemia contributed to low output failure with transfusion having the greater risk [6]. Kulier, a co-author with Mangano, reported an almost identical finding with anaemia being a contributor to poor outcome but exacerbated by the subsequent need for transfusion [7]. Of note, this report was based on data collected in the McSPI EPI II database.

One might expect then that this fact would be acknowledged in the other reports from the same database. In the 2006 paper, FFP use had an odds ratio for the renal outcome of 2.4, similar to that for aprotinin for the renal composite outcome [1]. Aprotinin did not appear as a risk factor. Of note, Euroscore was calculated for each patient and was a predictor of increased mortality [7]

In higher risk groups, such as aortic surgery, there is little or no comparative data on the impact of aprotinin. All studies that refer to the use of aprotinin vary in quality from small observational