ODAC advises FDA to wait for survival data on Orplatna

BETHESDA, Maryland—Members of the Oncologic Drugs Advisory Committee (ODAC) unanimously recommended that the Food and Drug Administration delay considering the approval of Orplatna (satraplatin, GPC Biotech) until the agency receives overall survival data from the ongoing SPARC clinical trial.

GPC Biotech had sought accelerated approval for its drug for the treatment of hormone-refractory (androgen-independent) prostate cancer (HRPC) that had failed prior chemotherapy on the basis of findings from a protocol-specified preliminary analysis that showed a 33% reduction in the risk of disease progression (reported in Oncology NEWS International March 2007, page 1).

[Editors' Note: A week after the meeting, GPC Biotech announced that it had withdrawn its New Drug Application (NDA) for Orplatna due to the ODAC decision. The company anticipates overall survival results to be available within 6 months. "However, this timing is based on an extrapolation of death rates in the trial and may change," GPC Biotech said in a press release. Bernd R. Seizinger, MD, PhD, CEO of GPC Biotech, commented: "We are focusing our efforts on the overall survival results from SPARC and integrating these data into the strongest possible NDA submission. If these data are positive, we plan to submit an NDA as quickly as possible."]

The ODAC panel examined five signifi-cant issues raised by FDA about the Orplatna NDA (Table) and sought answers from company executives before the 13-to-0 vote. In their questions and discussions, members made clear their concerns with several aspects of the study.

A key issue was whether the Orplatna endpoint data presented served as an adequate surrogate likely to predict that the drug will increase overall survival. "We need to know that," John T. Farrar, MD, of the University of Pennsylvania and a temporary voting member of the panel, told the sponsor. "We don't know now that long-term treatment won't lead to something else, kidney failure or other things, that would lessen survival."

The final survival analysis will take place after 700 deaths occur. In May, GPC Biotech officials notified FDA that they expected to have the survival data by the end of 2007. However, Marcel Rozencweig, MD, the company's chief medical officer, told ODAC that deaths among SPARC participants had slowed to five or fewer per month. "If that trend continues, it is going to take much longer to reach the 700 events," he said.

Richard Pazdur, MD, director of FDA's Office of Oncology Drug Products, cautioned the committee that "this time difference should not be the tie breaker" in the panel's decision whether to advise FDA to delay approval consideration until it has the survival findings.

Another confounding factor for FDA and the committee was the "composite" endpoint that researchers had developed to assess progression-free survival (PFS), which consisted of a trio of events that did not include prostate-specific antigen (PSA) progression. "The FDA has no prior experience with this endpoint," Dr. Pazdur said. "This was clearly communicated to the applicant during the development phase."

The Satraplatin and Prednisone Against Cancer (SPARC) trial protocol defined PFS as the time to disease progression or death. Disease progression, in turn, was defined as the first occurrence of one of three events:

• Radiologic progression, based on new bone or soft tissue lesions.

• Skeletal events, such as bone fracture, spinal cord or nerve root compression, or a change of anticancer therapy for bone pain.

• Symptomatic progression, such as worsening pain, increased opioid use, or weight loss attributable to cancer.

During planning for the trial, "FDA recommended that survival be the [sole] primary endpoint, and the study was powered for survival," said FDA reviewer Martin Cohen, MD. SPARC's final protocol included both the composite PFS and overall survival as primary endpoints.

Daniel Petrylak, MD, associate professor of medicine at Columbia University Medical Center and one of the SPARC principal investigators, explained the reasoning behind the composite PFS endpoint. "When we designed the SPARC trial, we wanted to come up with measures of progression that had been evaluated in other studies, and we recognized that PSA level was not an endpoint that FDA was going to accept," he said. "We basically wanted to be as rigorous as we could to find every way we could to look at clinically relevant endpoints."

Orplatna is a novel oral organoplatinum complex active against cell lines resistant to taxane and other platinum compounds. At present, only docetaxel (Taxotere) has demonstrated an increased survival among hormone-refractory prostate cancer patients. About 90% of the 27,050 US men projected to die of prostate cancer this year will have bone metastases. In a survey of 409 HRPC patients reported at the 2006 American Society of Clinical Oncology meeting, 60% said they experienced bone pain all the time or every day.

Given these facts, Orplatna meets an unmet medical need, said Nicholas J. Vogelzang, MD, director of the Nevada Cancer Institute, speaking on behalf of GPC Biotech. "These patients clearly need new and well-tolerated therapies that will slow the course of their disease, reduce pain, and improve the quality of their life," Dr. Vogelzang said.

To support its New Drug Application, GPC Biotech presented data from a protocol-specified preliminary analysis of SPARC, a multicenter, randomized, double-blind, placebo-controlled study conducted at 170 sites in 16 countries.

Researchers randomized 950 patients with advanced hormone-refractory prostate cancer who had progressed following one (and only one) prior systemic chemotherapy regimen to treatment with either Orplatna plus prednisone (n = 635) or placebo plus prednisone (n = 315). Placebo patients were not crossed over to Orplatna after progression.

SPARC Results

The early analysis, which focused primarily on the PFS composite endpoint, showed that Orplatna provides hormone-refractory prostate cancer patients statistically significant and clinically relevant benefits, Dr. Rozencweig said.

The Orplatna-treated patients had significantly fewer PFS events—83.1% (528) vs 87.0% (274) in the placebo—and a significant 36% delay in time to pain progression. Fewer Orplatna patients experienced pain progression as well (35% vs 43% for placebo).

Mean PFS in the Orplatna arm was 24.9 weeks vs 16.2 weeks for placebo. Median PFS was 11.1 weeks vs 9.7 weeks, respectively (HR 0.67, P < .0001. The researchers found a marked difference in the percentage of patients who were progression free at 6 and 12 months: 30% and 16% , respectively, for Orplatna vs 17% and 7% for placebo.

The interim analysis found a trend in overall survival that favored Orplatna, based on 463 deaths, with a median overall survival of 61.3 weeks for Orplatna vs 57.3 weeks for placebo. The most significant toxicities in the Orplatna patients were grade 3 and 4 neutropenia, 21.1% and 4.1% respectively, and grade 3 and 4 thrombocytopenia (21.8% combined).

Both the composite endpoint concept and the specific components of the SPARC PSF endpoint came in for criticism from ODAC members. "When you're measuring multiple different things, it adds up to further confusion rather than enlightenment," said Michael C. Perry, MD, director of hematology/medical oncology, University of Missouri.

Dr. Farrar expressed concern that patients might have guessed their drug assignment based on their side effects or blood counts, thus unblinding the pain results. "We know that patients who suffer side effects have a greater placebo effect with regard to pain," he said. "Given the fact that there was a high rate of serious side effects in the treatment group, I would argue that the entire pain benefit could be explained potentially by patients having a placebo effect."