Washington Editor

SILVER SPRING, MD – Hope overrode the FDA's concerns about clinical trial conduct when an advisory committee gave a thumbs up Tuesday to Novartis Pharmaceuticals Corp.'s Afinitor and Pfizer Inc.'s Sutent as treatments for pancreatic neuroendocrine tumors (pNET).

The FDA made it clear that the Oncologic Drugs Advisory Committee was not voting to recommend approval, but rather to indicate whether Afinitor (everolimus) and Sutent (sunitinib) demonstrated a sufficient benefit-risk profile given the problems with the trials. Nevertheless, the vote opens the door for the first new approvals for the indication in nearly 30 years.

The committee voted unanimously in support of Afinitor's risk-benefit and 8-2 for Sutent. But most of the yes votes for Sutent were accompanied by a call for more data, as the Phase III placebo-controlled trial was halted early, calling into question the robustness of the results. Several of the panel members suggested a head-to-head trial between the two drugs.

Because pNET is such a rare disease, more than 75 percent of the patients go five to seven years after experiencing symptoms before they are diagnosed, Grace Goldstein, the chief operating officer for the Carcinoid Cancer Foundation Inc., told the panel. By that time, their cancer has grown and spread.

The FDA referred to pNET as an "indolent" disease, but when the cancer spreads to other parts of the body, it becomes aggressive and difficult to treat. The five-year survival rate for patients with advanced disease is 27 percent.

"Having a new medication they can take is life-affirming," Goldstein said, adding that pNET patients need hope so they can fight their disease rather than succumb to it.

Jim Epperlein, a patient representative on the panel, echoed Goldstein's message. Should his cancer progress, "this gives me a chance," he said of Sutent. "This gives me hope."

But the FDA questioned whether either treatment was as effective as the companies claimed, pointing to irregularities in how the trials were conducted.

In the case of Sutent, New York-based Pfizer prematurely terminated the trial on the advice of an independent data monitoring committee (DMC), which claimed early data demonstrated progression-free survival (PFS), the primary endpoint.

According to the trial protocol, the first interim review of the data was to occur after 130 events had occurred. The DMC recommended termination after 73 events.

"The earlier a trial is stopped for efficacy the more likely the observed magnitude of treatment effect is an overestimate," FDA statistician Jenny Zhang told the panel.

Gideon Blumenthal, a medical officer at the FDA, said the early termination also called into question the data for the secondary endpoints, overall survival and patient-reported outcomes. Missing data and potential unblinding further confounded patient outcomes.

Wyndham Wilson, chairman of the committee and chief of the Lymphoma Therapeutics Section at the National Cancer Institute's Center for Cancer Research, agreed, saying, "Early looks can confound the ability of this committee to evaluate how robust a result is."

After reading from the FDA's guidance on DMC conduct, Wilson chided the Sutent DMC for its multiple early looks at the trial data. It is a "little bit of a poster child for everything that should not have been done," he added.

In the DMC's defense, its chairman, Robert Maki, an oncologist at the Memorial Sloan-Kettering Cancer Center, reminded the advisory panel that this was a placebo-controlled trial and the DMC had to act as the guardian of the patients on placebo. He also noted the known toxicities with Sutent, a multikinase inhibitor approved for gastrointestinal stromal tumors and advanced kidney cancer.

When the DMC noted the number of deaths, 15, in the placebo arm was three times the number in the Sutent arm, "I was very upset," Maki said. That difference is what prompted the DMC to recommend termination of the trial.

That termination, coupled with the number of censored patients and crossover from the placebo arm into the treatment arm, had committee members giving half-hearted support for the drug, primarily because there are so few options for pNET patients.

Streptozocin was approved as a treatment in 1982 under much looser standards, but it has limited benefit. Octreotide was approved several years later to treat some of the symptoms of pNET.

In voting yes, Epperlein said he hoped this would not set a precedent for other sponsors to end trials early as that can end up wasting more time for cancer patients who are running out of time.

Pfizer expects to hear from the FDA sometime this year on whether it will approve the new indication for Sutent, Mace Rothenberg, senior vice president of clinical development and medical affairs for the company's oncology unit, told BioWorld Today after the meeting.

Pleased with the vote, he said he could understand the concerns raised by the FDA and the panel members. However, it's important to consider the totality of the data and the context in which it was emerging, he added, referring to the increased number of deaths and adverse events in the placebo arm.

The company plans to work with the FDA to ensure it has all the information it needs to finalize its review.

The agency also had concerns about the conduct of two Phase III trials for Novartis' Afinitor, which is already approved for other indications, including renal-cell carcinoma. The East Hanover, N.J.-based firm initially had filed a supplemental new drug application to include two types of neuroendocrine tumors – pNET and carcinoid – but updated the application Friday following feedback from the FDA to include only pNET. (See BioWorld Today, April 11, 2011.)

That's not surprising, given that the carcinoid tumor study, RADIANT-2, failed to hit statistical significance in the primary endpoint of PFS when tested in combination with Sandostatin LAR (octreotide) vs. Sandostatin plus placebo. The Phase III study in pNET patients did better. Data from that trial, RADIENT-3, showed that Afinitor plus best supportive care more than doubled PFS vs. placebo, extending the median time without tumor growth from 4.6 months to 11 months.

FDA reviewers raised concerns that the two studies in similar tumor types should yield such different data and questioned changes made in the trial designs. In its presentation, Novartis pointed out that the FDA had advised it in 2006 that the natural histories and chemosensitivities are different for pNET and carcinoid tumors.

Everolimus targets mTOR, a protein that acts as a regulator of tumor cell division, blood vessel growth and cell metabolism. Everolimus, which was granted priority review in the U.S., is being reviewed by regulatory authorities in several other countries as a treatment for advanced NET of gastrointestinal, lung or pancreatic origin, Novartis said.

Pfizer won European approval last year for Sutent in pNET. (See BioWorld Today, Jan. 30, 2006.)