Expert Financial Analysis and Reporting

Yesterday, the FDA advisory committee voted 9 to 5 against approval of Lorcaserin. I listened to the day long proceedings and also tuned in to the conference call with investors that the company held this morning. I have a number of takeaways.

The panel and FDA are charged with weighing the benefits of the drug against the risks. In its briefing documents, the FDA acknowledged that lorcaserin did meet the criteria for efficacy in its trials and that its therapeutic benefits were modest, but good enough. The panel also seemed to feel that its efficacy was modest although some panel members stated that the drug did have clinical benefit. If one thinks of a teeter totter with benefits on one side and risks on the other, the benefit site is a bit light so that risks are magnified.

The panel seemed to have the most problems in dealing with a carcinogenicity signal seen in a pre-clinical rat study. In one strain of rats, mammary tumors were seen although interestingly these were disproportionate in males. Arena’s expert on the interpretation of animal carcinogenicity studies, Dr. Gary Williams, argued that these results were not attributable to lorcaserin but were due to physiological mechanisms in the rats. It was also pointed out that there were no cancer signals in concurrent mice carcinogenicity studies and that in the clinical trials the incidence of cancer was 2.5% in the lorcaserin group and 2.3% in the placebo group. The panel did not have anyone skilled in animal carcinogenicity study interpretation and they seemed to struggles with the issue and as to whether the rat findings signal a risk of lorcaserin causing breast cancer.

Arena stated that the results of the rat carcinogenicity studies became available while the phase III human trials were ongoing. The company immediately went to the FDA and showed them the data. The FDA allowed the phase III trials to continue. The data was not released to investors at the same time and only became known when the briefing documents were released. The company stated that it felt that this information was not material, but it was obviously wrong. This information caused a great deal of pressure on the stock when it became known and was a major concern for the FDA (highlighted in briefing documents) and the panel members. I saw one lawsuit filed against Arena yesterday alleging the withholding of material information.

Arena was also criticized for having a patient population in its trial that was not representative of the population in which it would be used if it became commercially available. Obesity is frequently a co-morbid condition with diabetes and cardiovascular disease. Some panel members felt that the patients studied in this trial were relatively more healthy than the ones that will use the drug in the real world. Arena does have a large trial ongoing in diabetes patients called BLOOM DM. If successful, this may assuage the committee’s concerns on safety and efficacy in sicker patients. Topline results will be available before year end.

I was a little surprised that there was not more emphasis on valvulopathy. The panel seemed to feel that Arena had done as much as possible to evaluate this risk in its phase III trials. The incidence of valvulopathy in the clinical trials was 2.6% in the lorcaserin patients and 2.7% in the control group. There was some discussion on the statistical analysis. In a clinical study, the risk of an event relative to the control group is stated and in the case of lorcaserin this was 1.07 or a 7% greater relative risk for valvulopathy. Statisticians also look at the 95% interval in which they are confident that there is only a 5% risk that the actual incidence will fall outside of the range. In the case of lorcaserin in its phase III trials the upper bound of the confidence interval was 1.55 and the lower bound was 0.74. This means that with 95% confidence the risk of valvulopathy for lorcaserin relative to placebo falls within a range of 26% reduction in relative risk to a 55% increase in relative risk. The company and FDA had tentatively defined the upper bound to be 1.50 so that lorcaserin just missed. However, this was so arbitrary that I don’t put much stock in it.

There was also not much talk about the lack of data on use of lorcaserin with phenteramine. I expected much more concern than the panel actually showed.

Where do we go from here? I think I can say with 99.9999% certainty that lorcaserin will not be approved on its PDUFA date of October 22^nd. I think that the FDA will definitely want to look at the BLOOM DM data available in late 2010. This should provide a lot of information on sicker patients that the panel was concerned about. Assuming this trial is positive, I think the FDA will conclude that the product is effective (it already has) and that the side effect issue of valvulopathy can be handled with some kind of REMS, a risk evaluation strategy. I don’t think the FDA will ask for another large human trial.

The carcinogenicity issue is difficult to predict. The panel was relatively naïve on animal carcinogenicity and reacted emotionally rather than scientifically. The FDA is much better equipped to make a rational decision. While I think that Arena’s arguments were persuasive on carcinogenicity, I am troubled that the FDA focused the panel’s attention on the issue. The FDA might ask for more pre-clinical carcinogenicity studies. There is a possibility that the FDA might also ask for a drug interaction study with phenteramine although there was nothing in the meeting to suggest that this might be the case.

Net, net I think that investors will await the results of the BLOOM DM study and guidance from the FDA on whether any new animal carcinogenicity or human studies are needed. I think that the stock will be in limbo until then and that probably means late this year before investors feel they can make an informed judgment. I think that based on all that I have heard that lorcaserin is approvable. The question is when and this hinges on whether the FDA requires additional animal or human trials.