乌鸦传媒

Adding a PD-1 inhibitor to PARP inhibitor maintenance therapy failed to slow progression or improve survival in newly diagnosed advanced ovarian cancer, a large randomized trial showed.

In fact, median progression-free survival (PFS) decreased from 20.2 months with rucaparib (Rubraca) monotherapy to 15.0 months with the addition of nivolumab (Opdivo), translating into a hazard ratio of 1.29 (95% CI 1.08-1.53). An interim analysis of overall survival (OS) also showed a trend toward a worse outcome with the combination.

Treatment exposure to rucaparib was longer with monotherapy, and the combination was associated with substantially more toxicity, reported Bradley Monk, MD, of Florida Cancer Specialists & Research Institute in West Palm Beach, at the European Society for Medical Oncology congress in Barcelona.

"So, [the combination] was associated with increased toxicity and did not extend the PFS benefit of rucaparib monotherapy as first-line maintenance treatment," said Monk. "The efficacy of monotherapy, though, was consistent with what we've shown, with a hazard ratio of 0.52 with 2 years of additional follow-up."

The present negative results add to a checkered history for immuno-oncology strategies in advanced ovarian cancer, said invited discussant Philipp Harter, MD, of the University of Essen in Germany.

"We started several trials with a high spirit and a high motivation, and we wanted to see comparable results, like in melanoma and like in cervical cancer and endometrial cancer," said Harter.

Instead, the efforts to convert "cold" ovarian cancers into hot ones with immunotherapy have been met mostly with failure, beginning with the JAVELIN randomized trials of avelumab (Bavencio), which were "completely negative." It was thought that perhaps a different drug would make a difference, said Harter, so other studies were conducted with different immune checkpoint inhibitors.

So far, the most "positive" negative trial has been , which evaluated atezolizumab (Tecentriq) plus bevacizumab (Avastin) and showed a nonsignificant 20% reduction in the survival hazard versus chemotherapy in recurrent disease.

A precedent for combining a PD-1/PD-L1 pathway blocker and a PARP inhibitor exists in the trial, which evaluated chemotherapy with or without atezolizumab (Tecentriq) followed by niraparib (Zejula) maintenance with or without atezolizumab. The results showed no benefit with the addition of the PD-L1 inhibitor, but "at least it wasn't detrimental," said Harter.

The MEDIOLA trial evaluated olaparib (Lynparza) and durvalumab (Imfinzi) with or without bevacizumab in recurrent ovarian cancer and showed a 34% response rate with the PARP inhibitor and the PD-L1 inhibitor, increasing to about 90% with the addition of bevacizumab. The same triplet therapy also produced positive results in the DUO-O trial.

Additional ongoing trials may provide more insight into the best immuno-oncology strategy for ovarian cancer, said Harter.

The ATHENA-COMBO trial reported by Monk had the goal of building on the positive results of the ATHENA-MONO trial, which showed a doubling of median PFS (20 vs 9 months) with rucaparib maintenance versus placebo in newly diagnosed patients with platinum-sensitive ovarian cancer. Rucaparib maintenance performed "absolutely as expected" in ATHENA-COMBO, said Harter.

The discontinuation rate with nivolumab was 10 or more percentage points higher than most other trials of immuno-oncology in ovarian cancer. Perhaps a different immune checkpoint inhibitor might have been more tolerable, Harter suggested.

"There's more to come," he said. "We have two additional trials investigating the role of checkpoint inhibitors in primary ovarian cancer. I'm really looking forward to seeing all of those data."

ATHENA-COMBO investigators enrolled patients with newly diagnosed stage III/IV ovarian cancer at 294 sites in 24 countries.

All patients underwent primary surgery or interval debulking and received frontline platinum-containing chemotherapy that resulted in partial or complete response. They were randomized 4:4:1:1 to rucaparib plus nivolumab (n=436), rucaparib plus placebo (n=427), placebo plus nivolumab, or double placebo. The primary endpoint was investigator-assessed PFS.

Landmark PFS analyses showed a consistent advantage for rucaparib-placebo beginning at 12 months (63% vs 56% with rucaparib-nivolumab) and continuing through 48 months (33% vs 26%). In contrast, rucaparib-placebo reduced the PFS hazard by 46% versus the small group of patients who received two placebos (20.2 vs 9.2 months).

The interim OS analysis yielded median values of 49.4 months for the combination and 58 months for rucaparib-placebo (HR 1.13, 95% CI 0.93-1.38).

Treatment exposure was 75% longer with rucaparib (8.4 months, including a 1-month lead in, vs 4.6 months for nivolumab starting at month 2). Patients in the placebo arm had a higher dose intensity as compared with the nivolumab arm (0.89 vs 0.84).

Grade ≥3 treatment-emergent adverse events (AEs) and AE-related treatment interruptions occurred more often in the nivolumab arm. AE-related treatment discontinuation rates were 25.4% with the combination and 14.7% with rucaparib-placebo. More than twice as many fatal treatment-emergent AEs occurred with the combination (nine vs four). The most common treatment-emergent AEs leading to discontinuation were increased liver enzymes, anemia, asthenia, neutropenia, thrombocytopenia, febrile neutropenia, rash, and nausea.

Comment