Introduction/Background Immune checkpoint inhibitors have generally shown disappointing results in phase 3 trials in ovarian cancer and their role in combination with standard therapy is unclear; however, subgroup analyses have yielded unexpected results. We explored clinical outcomes according to investigator-selected chemotherapy in the ANITA (NCT03598270) phase 3 trial.

Methodology Eligible patients had measurable high-grade serous, endometrioid or undifferentiated ovarian cancer with first or second relapse >6 months after platinum-based chemotherapy. Standard therapy comprised an investigator-selected carboplatin doublet (paclitaxel, gemcitabine or pegylated liposomal doxorubicin [PLD]) for 6 cycles, followed by maintenance niraparib in patients with a complete or partial response or stable disease. Patients were randomised 1:1, stratified by carboplatin doublet, platinum treatment-free interval (TFIp), BRCA status and PD-L1 status (SP142 IHC assay), to receive either atezolizumab (1200 mg every 3 weeks or equivalent) or placebo throughout standard therapy. The primary endpoint was progression-free survival (PFS). Subgroup analyses according to carboplatin partner were prespecified.

Results In the gemcitabine subgroup, PFS favoured atezolizumab (hazard ratio 0.49, 95% CI 0.28–0.87). Similar effects were not observed in the PLD or paclitaxel subgroups (table 1). Patients in the gemcitabine subgroup were more heavily pretreated (with the highest frequency of prior PARP inhibition and 2 prior chemotherapy lines in the placebo group), received a lower median carboplatin dose (inherent in the protocol) and had shorter median PFS and lower objective response rate (ORR) than patients in the PLD or paclitaxel subgroups.

Conclusion The apparent difference in atezolizumab treatment effect according to investigator-selected chemotherapy doublet, including more favourable outcomes with atezolizumab in the gemcitabine subgroup, may be attributable to selection bias towards gemcitabine in poorer-prognosis patients and the small sample sizes of the paclitaxel and gemcitabine subgroups. Ongoing translational work may reveal further differences and hypotheses.

Disclosures Study drug and funding for this investigator-initiated study are provided by F. Hoffmann-La Roche and GSK.

FH reports honoraria/consultation fees from Novocure, PharmaMar, AstraZeneca, Roche, Tesaro, GSK, Clovis, amedes, NEO NewOncology and Zai Lab.

BP reports honoraria/consultation fees from GSK, MSD, AstraZeneca, Clovis and PharmaMar, participation in company-sponsored speaker’s bureau for Clovis, PharmaMar, GSK and MSD and travel grants/attendance at scientific meetings from GSK, MSD, AstraZeneca and Clovis.

RS reports grants/research support from AstraZeneca, honoraria/consultation fees from GSK, Eisai and Seagen, participation in company-sponsored speaker’s bureau for Novartis, GSK, Eisai and Clovis Oncology and non-financial support from MSD, GSK and Novartis.

NC reports grants/research support from AstraZeneca, GSK and Roche, honoraria/consultation fees from AstraZeneca, Clovis, Eisai, GSK, Immunogen, Mersana, MSD, Nuvation BIO, OncXerna Therapeutics, Inc., Roche and Novocure and participation in company-sponsored speaker’s bureau for AstraZeneca, Clovis Oncology, GSK, MSD/Merck and Eisai.

SH reports honoraria/consultation fees from and participation in company-sponsored speaker’s bureau for AstraZeneca, Merck, Novartis, MSD Oncology, BMS, Sanofi, Gilead and GSK and travel/accommodation/expenses from Gilead, Roche, GSK and MSD Oncology.

GM reports honoraria/consultation fees and travel grants from Roche and GSK.

FJ reports honoraria/consultation fees from GSK, AstraZeneca, MSD, Eisai, Seagen and Novocure.

MQ has no potential conflicts of interest to report.

LMansi reports honoraria/consultation fees from Lilly, MSD, Novartis, Eisai, Pfizer and Roche, participation in company-sponsored speaker’s bureau for MSD and Roche and travel grant and scientific meeting attendance for Pfizer.

PE-G reports grants/research support from GSK and honoraria/consultation fees/participation in company-sponsored speaker’s bureau for GSK, Clovis, PharmaMar, AstraZeneca and MSD.

J-SF reports honoraria/consultation fees from AstraZeneca, Daiichi Sankyo, Pfizer, Lilly, Novartis, GSK, MSD, Eisai and Seagen.

EG reports honoraria/consultation fees and participation in company-sponsored speaker’s bureau for AstraZeneca-MSD, Clovis Oncology, GSK-Tesaro, PharmaMar and Roche.

LManso has no potential conflict of interest to report.

VC reports advisory board honoraria from AstraZeneca, GlaxoSmithKline, MSD and PharmaMar and speaker honoraria from AstraZeneca and MSD.

LS-L has no potential conflicts of interest to report.

MEO reports honoraria/consultation fees from Merck, MSD, GSK, AstraZeneca, Clovis, PharmaMar and Eisai.

DD has no potential conflicts of interest to report.

FM reports grants/research support from Roche, AstraZeneca, Novartis, Eisai, Gilead, MSD, Vaccibody, GSK, Lilly, Seagen and Daiichi Sankyo and honoraria/consultation fees from Roche, AstraZeneca, Gilead, MSD, GSK, Clovis, Novartis, Pfizer, Lilly, PharmaMar, Genomic Health, Myriad, Seagen, Daiichi Sankyo, Stemline and Menarini.

EGM has no potential conflicts of interest to report.

AGM reports honoraria/consultation fees from Alkermes, Amgen, AstraZeneca, Clovis, Genmab, GSK, Hedera Dx, Immunogen, Kartos, Karyopharm, Illumina, Mersana, MSD, Novartis, Novocure, Oncoinvent, PharmaMar, Roche, SOTIO, SUTRO, Seagen, Takeda and Tubulis, participation in company-sponsored speaker’s bureau for AstraZeneca, Clovis, GSK, Immunogen, Mersana, MSD, Novocure, PharmaMar, Roche and Takeda and grants/research supports from GSK, Roche, ISCIII and AECC.

Baseline characteristics, carboplatin exposure and outcomes according to chemotherapy regimen and treatment arm