Is AZ, Merck's Lynparza approvable in pancreatic cancer sans survival benefit? FDA's about to decide

When AstraZeneca and Merck & Co. unveiled in February that a study of Lynparza in pancreatic cancer had read out positively, there was much excitement in the oncology world. However, detailed data presented later revealed one caveat: The PARP inhibitor may have delayed disease progression, but it didn’t top placebo at prolonging patients’ lives.

Now, an independent FDA advisory panel is set to vote Tuesday on whether the data are strong enough to warrant an approval. After that, the FDA will soon make its final decision.

Despite the mixed showing, along with concerns about trial design that FDA reviewers outlined in a briefing document (PDF), one group of analysts figures Lynparza still has a good chance of snatching a go-ahead in germline BRCA-mutated pancreatic cancer that hasn’t progressed after an initial round of chemo.

RELATED: ASCO: AZ, Merck's Lynparza fends off pancreatic cancer, cutting progression risk in half

According to phase 3 data from the Polo trial presented in June at the American Society of Clinical Oncology (ASCO) annual meeting, Lynparza cut the risk of disease worsening or death by 47% in the indication. Patients on Lynparza went a median 7.4 months without their cancer progressing compared with 3.8 months for the placebo arm, meeting the study’s primary endpoint.

However, no significant difference was observed in survival benefits between the two groups. An interim analysis on about half of the trial population who had died showed Lynparza patients lived a median 18.9 months, versus 18.1 months for those on placebo, according to a paper concurrently published in The New England Journal of Medicine. At the time, investigators noted that overall survival may have been affected by placebo recipients who crossed over to receive active treatment after their cancer worsened.

Is the observed effect on progression-free survival (PFS) clinically meaningful, considering a lack of evidence of an effect on overall survival (OS)? That's what the FDA will ask of the oncology experts Tuesday—and the answer is likely yes, an SVB Leerink team led by analyst Andrew Berens argued in a Friday note to clients.

First, the lack of alternative treatment options may tilt the balance in Lynparza’s favor, Berens noted. There’s no approved therapy indicated for gBRCA-mutated pancreatic cancer, and there are also no therapies specifically greenlighted for the maintenance of metastatic disease at all, as chemotherapy is currently still the standard of care for this tough-to-treat cancer.

The Polo data received “strong positive sentiment” among physicians at ASCO, Berens said. To that point, the National Comprehensive Cancer Network (NCCN) just in November added Lynparza to its updated treatment guidelines for pancreatic cancer, elevating the drug from a now-removed footnote that suggested physicians “consider” using it to a full recommendation in gBRCA-mutated patients. It’s a “strong endorsement that clinicians view the data as compelling, practice-changing and worthy of approval,” Berens said.

However, all those arguments rest on Polo being an accurate reflection of Lynparza’s effect. Turns out, the FDA has questions about that, too.

In its prepared briefing document for the expert panel, FDA staffers suggest the trial is too small, with 154 patients altogether. Of the 62 patients enrolled in the control arm, six had non-measurable disease and four had missing data on disease burden at baseline, they noted.

RELATED: AACR: Clovis' Rubraca follows Lynparza's big pancreatic cancer win with phase 2 data

The reviewers also referred to an “imbalance between arms for some known baseline factors that may affect prognosis” as evidence that the results may have unfairly favored Lynparza. Specifically, the Lynparza arm had a higher proportion of patients with ECOG performance status of zero (71% versus 61%), which means they were in better physical condition.

What’s more, the limitation of current imaging technology to accurately measure tumor burden in the pancreas is also cited as a concern for using PFS—as compared to OS—in this indication, an even more important factor as the majority of patients with disease progression in Polo were identified based on local disease advancements.

Considering all these factors, Berens' team still believes the panel will endorse Lynparza in the indication, particularly following the NCCN inclusion. The FDA may decide to defer a final decision, which is currently scheduled to arrive Dec. 28 based on a six-month priority review, but Berens called it “an unlikely scenario.”