Biotech Values

[urche]

xarelto in ACS data

Genisi,
With due respect, I want to revisit some of the data from the ATLAS trial and shed a little more of my perspective, not that it's any more correct than yours.

First, I still see the new dose as problematic in that the 5 mg BID dose seemed to be significantly more harmful than the 2.5 mg BID dose, yet the prior A fib data suggested even higher doses might be needed there. Such discordancy leads me to suspect that additional study will be required and the risk with the previously suggested a fib doses may be higher than we thought.

My second point is on numbers needed to treat and harm. Based on the data in TheHeart.org article you cited, as you pointed out,

number-needed-to-treat-to-benefit (NNT) - 56 (one death prevented for every 56 patients treated for two years), which is quite good.

I believe that applies for the combined doses. For the lower dose (2.5 mg BID) dose I calculate NNT=62.5 to prevent one CV death, including MI and stroke. I agree that is a reasonable number. But, it is important to put in perspective also the number needed to harm. For the same lower dose subgroup, looking at the "major non-CABG bleeding" risk (1.8% with rivaroxaban vs .6% with placebo), if I am calculating correctly, NNH=83.

So, we are dealing with numbers needed to save a life and cause major bleeding that are really pretty close. Generally survival trumps bleeding on the human value meter. But, I submit to you that a lot of people, if given the choice between a low chance of a sudden cardiac death (in fact most of the deaths in the treatment group were SCD, not recurrent MI or strokes) or major bleeding (knowing that 25% of those bleeds were intracerebral hemorrhages) many reasonable people would opt to avoid the hemorrhagic stroke. And doctors and lawyers will be scrutinizing the bleeding complications with pathologic interest.

So, my point is that IMO the decision to use riva in ACS is far from clear cut.

Regards,
Urche