“Despite notable success with the incorporation of bevacizumab and atezolizumab in the treatment of other solid tumors ... the first such study in ovarian cancer did not meet its first primary endpoint” of extending PFS in the intention-to-treat population or in patients positive for programmed death-ligand 1 (PD-L1), said investigator Kathleen Moore, MD, of the University of Oklahoma in Oklahoma City.
Dr. Moore presented this study, IMagyn050/GOG 3015/ENGOT-OV39, at the European Society for Medical Oncology Virtual Congress 2020.
Explaining the negative results
The rationale for this study was good, said discussant Isabelle Ray-Coquard, MD, PhD, of the University Claude Bernard Lyon I in Villeurbanne, France, who was an investigator on the study but not an author on the meeting report.
Unfortunately, the study’s results were ultimately negative. A prior phase 3 trial of an immune checkpoint inhibitor in ovarian cancer also had negative results. In the JAVELIN OVARIAN 100 trial, adding avelumab to chemotherapy did not improve PFS, and the trial was stopped early for futility.
Dr. Ray-Coquard posed the question of whether checkpoint inhibitors are “dead” for epithelial ovarian cancer but said the answer isn’t clear.
There might be something unique about the tumor microenvironment that shields ovarian cancer from the immune system, or perhaps the PD-L1 pathway is the wrong target for immunotherapy.
It might also be that the first-line setting is the wrong place for checkpoint inhibitors, and they might work better in the second line, Dr. Ray-Coquard said.
Another possibility is that the delayed effect of immunotherapy means that overall survival – which isn’t yet mature for IMagyn050 – might be a better primary outcome than PFS.
Patient and treatment details
IMagyn050 enrolled 1,301 patients with newly diagnosed, stage III/IV epithelial ovarian, primary peritoneal, or fallopian tube cancer.
Patients were randomized to atezolizumab (n = 651) or placebo (n = 650) in combination with bevacizumab, paclitaxel, and carboplatin every 3 weeks for six cycles. For cycles 7-22, patients received bevacizumab with placebo or atezolizumab.
Most patients had ovarian cancer (73% in the placebo arm and 75% in the atezolizumab arm), followed by fallopian tube cancer (17% and 15%, respectively) and primary peritoneal cancer (10% and 9%, respectively).
In both arms, 31% of patients had stage IV disease, and 60% were PD-L1 positive (at least 1% of tumor cells staining positive).
Treatment was given in the neoadjuvant setting for 25% of patients in both arms and after primary cytoreductive surgery for the remaining patients.
Efficacy and safety
In the intention-to-treat population, the median PFS was 19.5 months with atezolizumab and 18.4 months in the placebo arm. In PD-L1–positive patients, the median PFS was 20.8 months and 18.5 months, respectively.
The differences in PFS were not statistically significant or clinically meaningful, Dr. Ray-Coquard said.
She noted that results were not stratified by BRCA status, which has been linked to PFS in ovarian cancer, and a potential imbalance between the groups might have contributed to the negative results.
Overall survival data won’t be mature until 2023, but, in the first interim analysis, “there was no apparent difference in the curves,” Dr. Moore said.
In the intention-to-treat population, the median overall survival was not reached in either treatment arm. In the PD-L1–positive population, the median overall survival was 31.2 months in the placebo arm and was not reached in the atezolizumab arm.
Adverse events were consistent with the known safety profiles of atezolizumab and the other drugs, Dr. Moore said. There were more serious events and more events leading to discontinuation with atezolizumab.
Adverse events more common with atezolizumab included febrile neutropenia, pyrexia, thyroid abnormalities, and rash, including severe cutaneous reactions. Colitis, pancreatitis, and infusion-related reactions were relatively infrequent but more common with atezolizumab, Dr. Moore said.
What the future holds
Dr. Ray-Coquard said there are signals in IMagyn050 that warrant follow-up, including a trend toward improved PFS among women who had high-grade nonserous clear cell histology. In this group, the median PFS was 12.3 months in the placebo arm and 13.6 months in the atezolizumab arm (hazard ratio, 0.64).
Additionally, there was a significant PFS improvement in women with at least 5% of their tumor cells staining positive for PD-L1. The median PFS was 20.2 months in the placebo arm but was not reached in the atezolizumab arm (HR, 0.64; P = .0278).
Dr. Ray-Coquard said we need to know who these PD-L1–high patients are in terms of BRCA status, histology, and residual disease.
She went on to say that companies haven’t given up on checkpoint inhibitors for ovarian cancer. Phase 3 trials are testing the atezolizumab/bevacizumab/chemotherapy combination for late relapsed disease, atezolizumab with niraparib and chemotherapy for recurrent ovarian cancer, and nivolumab with rucaparib following response to chemotherapy.
The IMagyn050 study was funded by F. Hoffmann-La Roche, the company developing atezolizumab. Dr. Moore and Dr. Ray-Coquard disclosed relationships with Roche and many other companies.
SOURCE: Moore K et al. ESMO 2020, Abstract LBA31.