I appreciate the fact that Novo has responded to my paper describing the thrombotic adverse events reported by MedWatch. These authors correctly recognized the caution I repeatedly stated in interpreting these low-prevalence events. The limitations of MedWatch are clearly and extensively discussed in my paper, using data to potentially understand the reported higher prevalence of thrombosis with recombinant factor VIIa (rFVIIIa) vs. factor VIII inhibitor bypass activity (FEIBA).
The inclusion of adverse events from clinical trials does not in any way invalidate the MedWatch numbers of thrombotic events. What is stated in Sallah et al.'s letter  demonstrates that patients with either no inherent coagulation defect or acquired defects (Table 1 in ) develop thrombotic complications. These cases may very well differ from those of FEIBA. Thus, if one then calculates the relative risk of thrombotic adverse events with more than likely comparable diagnosis with FEIBA, the ratio is close to 1 : 1. Thus, the conclusion that FEIBA has more, or that rVIIA has fewer, thrombotic complications is still incorrect.
The statement that non-inhibitor patients receiving rVIIa are at more of a thrombotic risk than inhibitor patients with life-threatening bleeds or undergoing surgery is conjecture. In fact, not until all the patients who develop thrombotic complications with either agent are carefully, and independently, analyzed, with knowledge of the setting and cumulative dose, will one be able to be more confident regarding the relative risks in recipients. There may well be quite a difference in the risk of thrombosis in coagulation-deficient vs. non-coagulation-deficient patients. The number of doses used in the calculations of my paper were those published by both Baxter and Novo.
Sallah et al.  suggest that the differing mechanism of action of the two biological agents might explain the reported differences as well as differences among the recipients. There are no data to suggest this. The ages of the two reported groups are comparable. The list of investigated patients and their diagnosis, as well as of patients receiving off-label rVIIa, simply reveals that thrombosis occurs in these patients as well as in hemophilia patients.
I concur that the editorial  accompanying my article was inaccurate. My paper  never stated that all the rVIIa and FEIBA patients were hemophiliacs with inhibitors. I also clearly stated that both biological agents had a very low prevalence of thrombotic events, and did not suggest that either gave an unnecessary risk of clotting.
I also agree that only continued surveillance, as well as appropriate trials, will assure us of the relative safety of these agents, critical for hemostasis.