Pharma’s Cutting Edge

I was a bit surprised to learn yesterday of FDA’s desire for more cardiovascular safety data before approving Pargluva. We knew from the advisory committee hearing that FDA had concerns about the safety profile of the drug. Two of the eight questions to the committee concerned cardiovascular safety. The preclinical studies suggested no major cardiac toxicity concerns beyond those seen with PPAR gamma agonists such as rosi- and pioglitazone. Cardiac myopathy, a theoretical adverse effect of PPAR alpha agonists at high doses, was not seen preclinically, except for an increased incidence of cardiac hypertrophy in the 6-month rat toxicology study. See the FDA’s toxicology briefing document for details.

Muraglitazar Deaths

The clinical experience with muraglitazar indicated the potential for congestive heart failure and suggested the possibility of excess cardiac mortality. The table above shows mortality reported during muraglitazar clinical trials.

Note that there were 9 cardiovascular-related deaths among subjects assigned to muraglitazar (0.28% incidence) compared with 1 among placebo (0.17% incidence). There were none among the 823 subjects who were assigned to pioglitazone 15 mg or 30 mg daily. Five of the 9 on muraglitazar were at the highest dose tested. Furthermore, 4 of the nine were temporally related to the presence of CHF. On the other hand, three CV-related deaths on muraglitazar were sudden without autopsy and three others were non-cardiac (stroke and PE). Right away you see FDA’s dilemma. Here they have a drug with a cardiac mortality signal suggesting roughly 2x risk over placebo with no mortality observed during active comparator treatment. However, it would not be unreasonable to exclude from consideration the three CV-related deaths that were non-cardiac, since mechanistically the concern is in the heart. So, we’d be left with five muraglitazar events, yielding an incidence of 0.185%, which is very similar to the placebo incidence. Thus, the excess cardiac mortality disappears.

FDA was also likely concerned about an apparent dose-relationship. Doses above those recommended for use are often used off-label, abd FDA will consider toxcity at higher than recommended doses if they believe off-label use to be significant. Below are serious adverse events (SAEs) froma dose-ranging trial of muraglitazar. Sorry if it’s hard to read.

Muaglitazar Dose Response

Note the higher incidence of CV-related SAEs with doses of muraglitazar over 5 mg.

Cardiac SAEs at 5 mg were also higher than placebo or active comparator when muraglitazar was combined with metformin or sulfonylurea. Furthermore, broken out by preferred AE term, myocardial infarction occurred three times more often with muraglitazar (all doses) compared with pioglitazone (all doses). And the incidence of CHF was double that of pioglitazone (30 mg) versus muraglitazar (5 mg).

Details of the safety profile are available in the FDA’s medical officer review.

Overall, I can see why FDA would be concerned. But I think two factors probably played most against immediate approval. The first is the efficacy profile of the drug. Frankly, it was disappointing and wasn’t good enough to justify any meaningful increase in risk beyond the PPAR gammas already approved. The second is FDA’s current approach to risk. In the WSJ article cited here, a Deutsche Bank analyst said he doesn’t think this approvable decision reflects on FDA’s current state of mind as regards risk mitigation. I disagree. I think that two years ago this drug would have met with approval accompanied by a rigorous postmarketing committment and surveillance program. Today it received an approvable with a likely approval hinging on further clinical trials before launch. I noted with skepticism commentary that approval could be delayed up to a year. In fact, approval could be delayed for much longer than a year depending on what is asked of BMS. If an outcomes study is required, forget it. That’s several years in the making at a minimum. However, if a surrogate study will suffice (and I can’t imagine which surrogate would be acceptable) a year is possible but unlikely.

October 19th, 2005 | Category: Drug Safety, FDA, For Investors, R&D, Technology | Subscribe to comments | Leave a comment | Trackback URL