I think richard pazdur has an opening at the FDA for you? ;)
all true what you say. I'd just add that sometimes the details on the biologic plausibility of a drug drug interaction add weight to handicapping chances of prospectively hitting an endpoint that was indeed retrospectively found. in this case, in addition to the empiric data showing those pts without heparin did better, we know the following:
1. heparin can antagonize the antiinflammatory effects of ATIII (this was not really appreciated until after the kybersept study was designed)
2. at least some of the bneficial effects of ATIII were muted by bleeding, and eliminating heparin will surely decrease bleeds
so I would place more weight on the kybersept retrospective analysis, but again everything you say is true and this is still a relatively high risk study moving forward
as for reading into how elective heparin use may have biased the data - its almost impossible to say. in large part, heparin use is physician and center dependent. one could also speculate that those pts on heparin were either deemed to be at higher risk of DVT (obese, prolonged immobilization, etc.) and/or at lower risk of bleeding (after all who woudl give heparin if coag factors are severely depleted), and hence perhaps slightly better prognostic category. between the inherent variation in practice patterns, and whatever one can read into pt prognosees based on heaprin vs no heparin allocation, its almost impossible to say how randomizing heaprin use would have altered the outcome - hence the need for a prospective trial
jmofwiw
AI
Market Data 
Markets 
