FDA approves Gemzar for ovarian cancer despite its lack of efficacy

Townsend Letter for Doctors and Patients, Nov, 2006 by Ralph W. Moss

This summer, the Food and Drug Administration (FDA) took the rare step of overruling one of its own advisory panels and approved the drug Gemzar (known generically as gemcitabine) for the treatment of recurrent ovarian cancer. Earlier this year, the prestigious Oncologic Drugs Advisory Committee (ODAC) strongly recommended against approval of the drug for this indication. (1) But the FDA--under the acting directorship of Andrew C. von Eschenbach, MD--approved Gemzar in combination with carboplatin for women whose disease had relapsed six months or more after their initial therapy.

[ILLUSTRATION OMITTED]

In the past, FDA decisi on-makers have generally followed the advice of their advisory panels, although they are not legally bound to do so. In March, a spokesperson for Eli Lilly, the manufacturer of Gemzar, said that the company was "really disappointed" with the panel's decision. So what happened behind the scenes between March's disappointment and July's triumphant announcement of approval? Lilly spokesman Dr. Gregory Clarke stated that the company had provided the regulatory agency with additional information. "This is a situation where we were able to provide the FDA with additional analysis of the data that may have given a little more insight on the benefits," he said. Meanwhile, FDA officials have refused to comment on the controversial decision. It seems decidedly odd that this "additional analysis" was provided directly to FDA leaders and not to the expert panel or the general public.

This is also the first time that the FDA has approved an ovarian cancer drug based upon progression-free survival (PFS), rather than on the basis of an improvement in overall survival (OS). The difference between these two measures is not trivial. Although it sounds like a highly desirable outcome, "progression-free survival" merely represents an interval of variable length between the administration of treatment and the point at which the cancer inexorably starts to advance again. Such delayed tumor growth does not necessarily equate to an extension of the patient's lifespan. Indeed, tumors that are arrested temporarily can then advance more rapidly once they begin growing again, and patients may actually live no longer than they would if they had not received the treatment.

The FDA panel found that the combination of Gemzar and carboplatin increased median progression-free survival in patients by just 2.8 months when compared with carboplatin alone. However, the key fact was that there was no difference in the overall length of patients' survival. The Gemzar-added group also experienced more adverse effects.

For those reasons, in March, this panel voted 9-2 against recommending approval of the drug for ovarian cancer. The panel also commented on the weakness of Lilly's trial data and criticized the way the company conducted the 356-person clinical study (Pierson 2006). It was a landmark finding and seemed to indicate a new determination on the part of advisors to draw the line--to tell both the drug industry and regulators that they would have to provide patients with real survival benefits if they expected to reap the huge financial rewards that generally follow approval. "The main issue is whether adding 2.8 months to median progression-free survival at a cost of additional toxicity with no apparent effect on survival is a sufficient basis for Gemzar approval for this use," an FDA spokesperson said at the time. (2)

Eli Lilly countered, however, that the purpose of the drug was to prevent the cancer from worsening, not necessarily to help patients live longer. But, logically, if the Gemzar-treated patients die at the same time as those denied this additional drug, how can that constitute an improvement in their condition? Didn't their condition worsen significantly after the drug stopped working, so that they then died at the same time as those who weren't given the drug?

The company's other arguments in favor of Gemzar are convoluted in the extreme. Dr. Clark also asserted, in Gemzar's defense, that the drug allowed patients to go longer without being treated again and therefore could help spare them some side effects of chemotherapy. This is twisted logic, since they would now have to face the side effects of both Gemzar and carboplatin in combination--a regimen that carries a greater array of side effects than either drug administered alone.

When he was asked why Lilly did not mention in its press release the drug's failure to prolong life, Dr. Richard Gaynor, head of cancer research at Lilly, said: "I'm not sure that needed to be the focus." Of course not, as long as you're not the unlucky woman with ovarian cancer. Apparently, for Lilly executives, the focus should be kept not on prolongation of life but on the illusory benefit of "progression-free survival," which will bring concomitant benefits to their bottom line.

Dr. Gaynor also claimed that it is often difficult to determine if, and by how much, a particular cancer drug helps prolong life because seriously ill cancer patients often switch back and forth between different treatments. But isn't that precisely why rigorous clinical trials were invented in the first place, to tease apart and analyze the problems posed by the various treatments that patients are offered?