'Recombinant factor VIIa as adjunctive therapy for bleeding control in severely injured trauma patients: Two parallel randomized, placebo-controlled, double blind clinical trials.' Boffard KD, Riou B, Warren B, Choong PIT, Rizoli S, Rossaint R, Axelsen M, Kluger Y. J Trauma 2005; 59: 8-18.
In the opinion of Dr. Morris Blajchman, this article represents a classic example of what Dr. Richard Horton (Editor of The Lancet) has called "information laundering". The results of this important study, based on the data relevant to the clearly stated primary outcome, are clearly negative. Thus the data in the article show clearly that rFVIIa does not alter the transfusion requirements of severely bleeding trauma patients. Yet the authors of this article suggest that it is a positive study.
It is important to note that the clearly stated primary endpoint was the number of RBC units transfused during the 48-hour period after the first dose of trial product. This primary endpoint is clearly stated however, nowhere in the conclusions or in the subsequent discussion do the authors adequately relate to the data relevant to this primary endpoint. This article reports the data from a randomized, placebo-controlled, double-blind trial in two types (blunt versus penetrating) severely bleeding trauma patients (n=301). Each type of trauma patient was randomized to receive either rFVIIa, or placebo, in addition to their standard treatments. The first dose of rFVIIa (200 µg/kg) followed the transfusion of eight RBC units, with additional rFVIIa doses (100 µg/kg on each occasion) given 1 and 3 hours later.
The primary endpoint data are given in Table 2 (on lines 3 and 6) and these data clearly indicate no difference in the RBC transfusion rate between control and rFVIIa treated patients, for either trauma cohort. In the blunt trauma patients the median (range) of RBC units transfused for all patients (n=72) was 7.8 (0-48) RBC units for the patients receiving rFVIIa versus 7.2 (0-35) RBC units for the patients (n=64) receiving placebo (p=0.07). In fact, there was a higher RBC transfusion rate trend for patients receiving rFVIIa! In the patients with penetrating trauma the comparable results for all patients were: rFVIIa patients (n=69) 4.0 (0-37) RBC units versus 4.8 (0-41) RBC units for placebo (n=61) patients (p=0.24). There are no differences in any secondary endpoints which include mortality, thromboembolic deaths, ventilator-free days, or extent of multi-organ failure. Thus, there is clearly no benefit for trauma patients receiving rFVIIa compared to control patients, yet the authors conclude "Among blunt trauma patients, rFVIIa significantly reduced the need for RBC transfusion and massive blood transfusions". This conclusion is based on an analysis of the blunt trauma patients, including only those who were alive 48 hours after being entered into the trial, an outcome clearly not the designated primary endpoint of the study.
It is clear that this manuscript had been inadequately vetted prior to its acceptance for publication. Moreover, it could have benefited greatly from an external statistical review. The article is clearly one whose conclusions have been grossly distorted. Could this be the consequence of the influence of the sponsors who wanted to ensure that this article had a positive outcome in favour of rFVIIa?