ONXX / FDA Concerns:
All known on study deaths were 38 in total, with 21 due to disease progression. That leaves 17 deaths in play.
Of those, 10 were due to cardiac (n=7; 5 identified by applicant, 2 additional by FDA) or "cardiac possibly contributory to death" (n=3, only identified by FDA). So 10/526 = 1.9%. Of the 7 deaths, 5 died on cycle 1 of drug administration, 1 on cycle 2 and 1 on cycle 4.
The other one the FDA appears to have pointed out is Hepatic Failure, which occurred in 2 patients out of 526. I'm not sure why they've actually singled this one out, since it is on par with the other found causes of death (multi-organ failure, sepsis, intracranial hemorrhage, pneumonia, renal failure). The company did a small study to show that carfilzomib did not influence the PK of CYP3A4 substrates (the only one that carfilzomib was found to "moderately" inhibit, in vitro).
They do make a bigger deal of "cardiac disorders" in the briefing (30/526 patients, 6%), although I'm unsure how certain findings would relate given that the patient is on drug for a relatively brief period of time versus the typical trajectory of the disease (example: aortic valve stenosis). If I squinted hard to see a pattern, I would say exclude people with coronary artery disease from the drug. In the real world, the safety signals here should ideally be compared to Velcade, which also has cardiac effects. The disadvantage, in the case of these single arm trials, is that you don't have a comparator arm for safety. In one velcade trial, treatment-emergent cardiac disorders were 15% and 13% in the velcade and dexamethasone groups, respectively. Doesn't look as bad in that light, especially considering that patients in the carfilzomib trial did have the option of receiving low dose dexamethasone occasionally.
All in all, you can tell by the way that the FDA is wording their document that they are not intending to approve. They say things such as the patients who were "unresponsive or intolerant" to all known drugs "was a very small fraction off the ITT population (69/266; 26%)" or that the ORR in the primary efficacy study was "only 22%" while making a very big deal about adverse events that were present at or less than 1%. In a sense this is their mandate to keep patients safe, but the tone seems to be one where they're stacking the deck against the drug a little bit.
It's too bad, because this appears to be a reasonable case of the sponsor following the accelerated approval guidelines. The phase 3 studies are under way, so the sponsor is acting in good faith.
News 
Market Data 
Discover 
