In a somber end to an FDA advisory committee’s two-day scouring of multiple drugmakers’ oncology data, Pfizer’s pitch to expand Talzenna into a broader prostate cancer population has failed to impress.

Late Wednesday, eight outside experts on the FDA’s Oncologic Drugs Advisory Committee delivered a unanimous 8-to-0 ‘No’ vote on the question of whether Pfizer’s PARP inhibitor Talzenna has a favorable benefit-risk profile in metastatic castration-resistant prostate cancer (mCRPC) patients without homologous recombination repair (HRR) gene mutations.

Patients without HRR mutations make up the majority of the mCRPC patient population at about 70%, according to briefing docs released ahead of the meeting.

Pfizer is aiming to expand Talzenna’s existing prostate cancer nod—in combination with its Astellas-partnered androgen receptor inhibitor Xtandi—beyond the HRR mutant population where it’s already approved.

Pfizer’s phase 3 Talapro-2 trial, which compared a Talzenna-Xtandi cocktail with Xtandi alone, recently met both progression-free survival and overall survival endpoints in a first-line mCRPC patient population regardless of HRR status.

That said, the FDA has called out Pfizer for the “suboptimal design” of its study, laying into the New York drugmaker for failing to have a statistically powered test in the trial’s biomarker-negative population.

In turn, there is a chance that the study could have yielded false conclusions in the HRR-negative population, the FDA cautioned earlier this week.
 

'Shoot an arrow and paint a target around it'

During a clarifying question portion of Wednesday’s meeting, the FDA’s longtime oncology czar, Richard Pazdur, M.D., didn’t mince words about his misgivings around Pfizer’s data, arguing that the drugmaker’s study did not adequately evaluate Talzenna in non-HRR-mutant patients, which make up the bulk of the mCRPC population.

“To put the elephant in the room, this would be like somebody coming to us for a huge indication—70% of this population—coming to us with a randomized study, having it completed, and then saying, ‘oops, I forgot to do a statistical analysis plan,’” he said.

“In this situation where you’re talking about a huge number of patients, you don’t need proof—you need robust proof,” Pazdur added. Attempts to draw those conclusions after the fact are “tantamount to somebody shooting an arrow on the wall and then painting a target around it,” he claimed.

In justifying the company’s trial design and the results, Pfizer’s genitourinary therapeutic area head Dana Kennedy cited “the best science and clinical data in 2018 suggesting similar efficacy in biomarker positive and negative populations.” She noted that the risk of false positives on the study’s primary progression-free survival endpoint was split evenly between the unselected and HRR-selected cohorts.

Kennedy pointed to results from the final analysis of Talapro-2, which found that the Talzenna-Xtandi combo charted a roughly 20% reduction in the risk of death versus Xtandi alone in first-line mCRPC. The trial also showed that Pfizer’s cocktail helped extend patients’ lives by a median time of 45.8 months versus 27 months in the solo Xtandi arm—again, in all patients regardless of HRR status.

“Multiple definitions for non-HRRm will be presented today,” Kennedy added. “Regardless of which definition is used—either protocol defined, the sponsors’ or FDA exploratory definitions—we observe a consistent improvement across primary and secondary endpoints. This supports the overall treatment effect in this subgroup and is unlikely to be due to chance.”

Despite Pfizer’s best efforts to make a case for Talzenna in a broader mCRPC population, the outside experts on the advisory committee panel unanimously adopted the FDA’s position when the meeting finally came to a vote.

“There was concern on my part, as well as many others, about full characterization of the non-mutant population, the statistical assumptions that went into making the case that there was a benefit,” William Gradishar, M.D., from Northwest University’s Feinberg School of Medicine, said in justifying his vote.

“Because this study was not powered to test efficacy in the patients without HRR mutations, I thought this weakened the strength of the conclusions we could draw,” Columbia University Medical Center’s Neil Vasan, M.D., Ph.D., added.

“As a field, we need to commit to formally evaluating whether targeted therapies benefit patients with and without the biomarker,” Vasan said. “Simply put, precision oncology demands precision trials.”

The experts also raised concerns about Talzenna’s toxicity in the study’s population.

“In the overall population of non-HHRm, you have 8 out of 10 patients that had grade three or higher toxicity, 4 out of 10 patients needing a transfusion,” Daniel Spratt, M.D., of Case Western Reserve University, said of Pfizer’s data package.

“Again, the toxicity can’t be ignored here,” added the National Cancer Institute’s Ravi Madan, M.D. “Just because we can manage a toxicity doesn’t mean it doesn’t impact patient lives, even if we warn them about it ahead of time.”

Given those concerns, the panelists delivered a firm vote against Talzenna in non-HRR-mutant prostate cancer patients.

The FDA is not required to follow the advice of its expert panels when it comes time to make an approval decision, though the regulator often does.
 

2-day adcomm questioning

Pfizer’s vote of no confidence capped off a grueling two-day meeting during which the FDA’s outside experts also pored over data on cancer meds from Roche, Johnson & Johnson and UroGen.

In UroGen’s case, the experts voted 4 to 5 that an intravesical solution of the company’s chemotherapy Jelmyto, also known as UGN-102, had a favorable benefit-risk profile in recurrent low-grade intermediate-risk non-muscle invasive bladder cancer (LG-IR-NMIBC).

“While we are disappointed by today’s outcome, we continue to believe our clinical data support UGN-102 for the treatment of recurrent LG-IR-NMIBC, a disease with no FDA-approved therapies,” Liz Barrett, UroGen’s CEO, said in a statement. UroGen’s drug will go before the FDA for approval by June 13.

Roche was also hit with a losing vote in its bid to advance a combination therapy using Columvi in patients with previously treated diffuse large B-cell lymphoma (DLBCL) who aren’t good fits for stem cell transplant.

On Tuesday, the FDA’s oncology advisory panel voted 8 to 1 that the results from Roche’s late-stage Starglo trial weren’t applicable to U.S. patients, citing an imbalance between Asian and non-Asian regions included in the study. The study looked at the combination of Columvi and the chemotherapy regimen GemOx in second- or later-line transplant-ineligible DLBCL patients.

Only J&J’s Darzalex Faspro passed muster with the outside experts, who voted 6 to 2 on Tuesday in favor of the med’s benefit-risk profile in patients with high-risk smoldering multiple myeloma (SMM). SMM is an asymptomatic precursor to active myeloma.

If approved, Darzalex would become the first medication specifically approved to treat SMM.