848MO - Phase II open-label randomized study of pembrolizumab with or without bevacizumab in platinum-resistant recurrent/metastatic nasopharyngeal carcinoma (R/M NPC)

Background

Patients with platinum-resistant (Pt-R) R/M NPC have poor prognosis. In these patients, pembrolizumab is not superior to chemotherapy. VEGF is overexpressed in NPC and suppresses anti-tumour immune response; prior single arm studies show impressive tumour response with anti-VEGF and anti-PD1 combination. We conducted a trial to compare the efficacy of pembrolizumab with bevacizumab priming vs pembrolizumab in NPC.

Methods

Patients with Pt-R R/M NPC were randomized 1:1 to 3-weekly pembrolizumab 200mg (Arm A) or with bevacizumab 7.5mg/kg administered 1 week prior to each dose (Arm B), till progression (PD), intolerance, or up to 2 years. Crossover was allowed from Arm A to Arm B at PD. The primary endpoint was tumor response (ORR) assessed by RECIST 1.1 based on intention to treat. Secondary endpoints included safety, progression-free survival (PFS) and overall survival (OS). Translational studies were done on serial tumour and blood sampling to determine the immunophenotypic effects of treatments.

Results

48 patients were randomised equally to both study arms. The median follow-up was 28.3m (IQR: 15.9-55.9). Arm B had higher ORR (58.3% vs 12.5% [95% CI for difference: 18.8-72.9, p<0.001]), and longer PFS (median of 13.8m vs 1.6m [HR 0.25, 95% CI 0.13-0.50, p<0.001]) than Arm A. Median OS was 18.5m vs 11.7m (HR 0.62, 95% CI 0.29-1.30, p=0.101) and grade 3 adverse events (AEs) occurred in 25% (6/24) vs 8.3% (2/24) in Arms B and A respectively. Incidence of immune related AEs was similar with no grade 4-5 AEs. More patients in Arm B had >80% reduction in baseline plasma EBV DNA during treatment (70.8% vs 21.1%, p=0.001). 13 patients crossed from Arm A to Arm B: 5 (38.5%) had PR, 5 (38.5%) had SD, with median PFS of 5.0m (95% CI 2.8m-NR). Multiplex IHC of tumour showed dramatic increase in density of CD4+ and PAX5+ B cells 8 days after bevacizumab in the good responders with more modest CD8+ T cells. Arm B had increased immune cell infiltrate after 2 cycles of treatment compared with Arm A.

Conclusions

Bevacizumab and pembrolizumab is tolerable and more efficacious than pembrolizumab alone in Pt-R NPC. Mechanistically, bevacizumab enhances immune cell infiltration, specifically humoral immune responses.

Clinical trial identification

NCT03813394.

Editorial acknowledgement

Legal entity responsible for the study

National University Cancer Institute, Singapore.

Funding

National Medical Research Council.

Disclosure

W.Q. Chong: Financial Interests, Personal, Advisory Board: Merck; Financial Interests, Personal, Invited Speaker: MSD, Ipsen; Financial Interests, Institutional, Research Funding: MSD. R.A. Soo: Financial Interests, Personal, Advisory Board: Abbvie, Amgen; Financial Interests, Advisory Board: AnHeart, AstraZeneca, Bayer, BMS, Boehringer Ingelheim, Daiichi Sankyo, GSK, J INTS BIO, Janssen, Lily, Merck, Merck Serono, Novartis, Pfizer, Puma, Roche, Sanofi, Taiho, Takeda, Thermo Fisher, Yuhan Corporation; Financial Interests, Invited Speaker: Chugai; Financial Interests, Research Grant: AstraZeneca, Boehringer Ingelheim, Pfizer. B. Goh: Financial Interests, Personal, Advisory Board: Bayer, Otsuka, Adagene, Novartis, Pfizer; Financial Interests, Personal, Research Funding: Bayer, Kyowa Hakko Kirin, Genentech/ Roche; Financial Interests, Institutional, Research Funding: BMS, MSD oncology, Adagene, MSD oncology, ALX Oncology, Taiho pharmaceutical; Financial Interests, Personal, Invited Speaker: MSD Oncology, Novartis, Amgen; Financial Interests, Personal, Stocks/Shares: Gilead science, Regeneron, Seattle Genetics/ Estella's. All other authors have declared no conflicts of interest.