Substantial Evidence for Provenge Efficacy: FDA Panel

May 1, 2007

An FDA advisory panel has supported the requested approval of the cancer vaccine Provenge (sipuleucel-T, Dendreon) for the treatment of asymptomatic, metastatic, hormone-refractory prostate cancer.

GAITHERSBURG, Maryland—An FDA advisory panel has supported the requested approval of the cancer vaccine Provenge (sipuleucel-T, Dendreon) for the treatment of asymptomatic, metastatic, hormone-refractory prostate cancer. [As ONI went to press, Dendreon received an "approvable" letter from FDA requesting additional clinical data in support of efficacy.] Although members of the Cellular, Tissue, and Gene Therapies Advisory Committee did not formally vote to recommend approval, its members did tell FDA that the data submitted to them showed that Provenge is reasonably safe and has substantial evidence of efficacy.

The decision came after the panel heard presentations by Dendreon and an FDA review team on two similarly designed phase III trials submitted by Dendreon. Neither study met its primary endpoint of improved time to progression. However, post hoc analyses found that the larger of the studies showed a significant advantage for Provenge in overall survival after 3 years. Overall survival in the second trial did not reach significance.

Studies 1 and 2 were randomized, double-blind, placebo-controlled, multicenter trials. Eligibility criteria included metastatic prostate cancer without vis-ceral metastases, tumor progression despite androgen deprivation, no cancer-related pain, no systemic steroids or prior immunotherapy, and an ECOG status of 0 or 1. The primary endpoint for both studies was time to progression.

The two trials used 2:1 randomization, with patients to receive Provenge or placebo three times, with 2 weeks between each treatment. Study 1 randomized 82 men (median age 73; 89% white) to Provenge and 45 (median age 71; 93.3% white) to placebo. The treatment group had a nonsignificant median time to progression of 11.0 weeks, compared with 9.1 weeks in the placebo arm (P = .085).

After study 1 failed to meet its primary endpoint, researchers halted enrollment in study 2 following accrual of 98 of 120 planned patients with similar age and racial characteristics as those in study 1. The difference in time to progression in the second study also was nonsignificant, 10.9 weeks for Provenge vs 9.9 weeks for placebo (P = .719). The two trials found no significant regression in tumor size among the treated patients.

Although neither study specified overall survival as an endpoint, a post hoc analysis of the study 1 data showed an overall median survival benefit for the Provenge arm, compared with placebo, of 25.9 months vs 21.4 months (P = .01). "Overall survival is the least biased, least variable, and most clinically meaningful assessment of an oncology product," said Mark Frohlich, MD, Dendreon's vice president of clinical affairs.

An analysis of study 2 found a median overall survival of 19 months for the treatment arm vs 15.7 months for the placebo group, a difference that failed to reach statistical significance (P = .331). "It should be noted that the survival time in this study was shorter than the counterpart in study 1, which suggests that significant populations in these two studies may not be exactly the same," remarked FDA clinical reviewer Ke Liu, MD, PhD. Dendreon also presented combined overall survival data from the two trials, showing a significant advantage for Provenge, 23.2 weeks vs 18.9 weeks for placebo (P = .011)

Safety Analysis

In its safety analysis, FDA relied on data from the 146 Provenge patients and 76 placebo patients from the two studies. Common adverse events occurring in more than 10% of the treated patients were chills (58% vs 8% of placebo patients); pyrexia (33% vs 7%); headache (19% vs 7%); asthenia (15% vs 7%); nausea (15% vs 8%); paraesthesia (13% vs 9%); vomiting (12% vs 3%), and dyspnea (11% vs 3%). Most adverse events were grade 1-2 and resolved within 48 hours.

Panel members focused primarily on the cardiovascular accidents (CVAs) seen in the two trials, although the differences between the treatment and placebo arms did not reach statistical significance. There were 3 CVAs in the Provenge patients (2%) vs 0 in the placebo group.

The sponsor also submitted CVA data on the three completed studies for the proposed indication: androgen-independent metastatic prostate cancer (studies 1-3), showing 17 CVAs in 345 treated patients (4.9%) vs 3 in 172 placebo patients (1.7%) (P = .092). Deaths attributed to CVAs occurred in 7 and 2 patients, respectively (P = .724).

The panel voted unanimously that Provenge is a reasonably safe drug. However, several members expressed concern about the CVA data and the drug's potential for serious harm. They urged FDA to prod the sponsor to further explore the risk of CVAs associated with Provenge. One panelist declared the agency should not approve the drug until Dendreon guaranteed it would complete an ongoing 500-patient phase III randomized trial, known as IMPACT, that has overall survival as its primary endpoint; some panelists said this study would provide definitive answers to their concerns.

Committee members were less certain about the efficacy of Provenge, with several expressing concern about the size of the trial, the reliability of the survival findings, and whether the lack of meaningful minority participation might hide a lack of efficacy in nonwhites. Howard I. Scher, MD, chief of genitourinary oncology, Memorial Sloan-Kettering Cancer Center, succinctly summed up the survival issue: "The question boils down to: Is this a drug effect or is it related to the patient population?" The committee majority favored the drug-effect alternative and voted 13-to-4 that substantial evidence supports the drug's use in its intended patient population.