Pembrolizumab Plus Chemo Shows Encouraging Activity in NEPC

At the 2026 European Society for Medical Oncology (ESMO) Congress, Gunhild von Amsberg, MD, professor of uro-oncology in the Department of Hematology and Oncology and the Martini-Klinik at University Medical Center Hamburg-Eppendorf in Germany, presented updated results from cohort I of the phase 1b/2 KEYNOTE-365 trial (NCT02861573).¹ In patients with metastatic neuroendocrine prostate cancer (NEPC), adding pembrolizumab (Keytruda) to carboplatin and etoposide produced a higher objective response rate (ORR) and longer overall survival (OS) than chemotherapy alone.

NEPC is an aggressive form of prostate cancer, for which no standard of care for metastatic disease has been validated in a randomized study, and outcomes with the commonly used platinum-based chemotherapy remain poor. Pembrolizumab has received FDA orphan drug designation for NEPC but is not approved for this indication. Initial data from 48 patients in this cohort, presented in 2025, showed an ORR of 25% with pembrolizumab plus chemotherapy vs 10% with chemotherapy alone.²

Among patients with disease measurable by RECIST v1.1, the ORR by blinded independent central review (BICR) was 28% (95% CI, 14%-45%) with pembrolizumab plus chemotherapy (n = 36), compared with 15% (95% CI, 4%-34%) with chemotherapy alone (n = 27). All responses were partial: 10 patients in the pembrolizumab arm and 4 in the chemotherapy arm, with no complete responses. Disease control rates were 50% (95% CI, 33%-67%) and 48% (95% CI, 29%-68%), respectively. The median duration of response was 6.2 months (range, 2.0+ to 10.2+) with pembrolizumab and 4.1 months (range, 2.8+ to 6.2) with chemotherapy alone.

In a post hoc analysis by PD-L1 status, the ORR was 40% (95% CI, 16%-68%) vs 18% (95% CI, 2%-52%) for patients with a combined positive score (CPS) of at least 1. For patients with a CPS below 1, it was 15% (95% CI, 3%-38%) vs 13% (95% CI, 2%-40%).

The median OS was 10.2 months (95% CI, 8.0-11.4) with pembrolizumab plus chemotherapy, compared with 7.1 months (95% CI, 4.0-8.2) with chemotherapy alone (HR, 0.33; 95% CI, 0.17-0.63). The 6-month OS rates were 85% and 59%, respectively. The OS benefit appeared greatest in patients with a CPS of at least 1 (median, 11.0 vs 4.8 months; HR, 0.16; 95% CI, 0.05-0.49). In those with a CPS below 1, the median OS was 8.1 vs 7.1 months (HR, 0.49; 95% CI, 0.21-1.11).

Additionally, the median radiographic progression-free survival (rPFS) by BICR was 5.1 months (95% CI, 4.1-6.3) vs 4.0 months (95% CI, 2.1-5.0; HR, 0.66; 95% CI, 0.36-1.21), and 6-month rPFS rates were 49% and 26%. For patients with a CPS of at least 1, the median rPFS was 6.0 vs 3.8 months (HR, 0.40; 95% CI, 0.16-1.01). For those with a CPS below 1, it was 5.1 vs 4.1 months (HR, 0.87; 95% CI, 0.38-1.97).

Cohort I enrolled 70 patients, and 40 were assigned to pembrolizumab plus chemotherapy vs 30 to chemotherapy alone. Random assignment was initially 1:1 and changed to 3:1 after a protocol amendment took effect on February 7, 2025. Patients received carboplatin at area under the curve 5 every 3 weeks and etoposide 100 mg/m² on days 1 to 3 every 3 weeks for 4 to 6 cycles. Patients in the experimental arm also received pembrolizumab 200 mg every 3 weeks for up to 35 cycles. Randomization was stratified by ECOG performance status.

Eligible adults had de novo or treatment-emergent metastatic NEPC confirmed by central histology review, no prior platinum-based chemotherapy for NEPC, and an ECOG performance status of 0 or 1. Most patients had small cell carcinoma: 80% in the pembrolizumab arm and 67% in the chemotherapy arm. The median age was 69 (range, 48-82) and 67 years (range, 47-86), respectively. The coprimary end points were ORR per RECIST v1.1 by BICR and safety. Secondary end points included duration of response, rPFS by BICR per PCWG3-modified RECIST v1.1, and OS. The median follow-up was 15.4 months (range, 0.8-43.5) at the November 19, 2025, data cutoff.

Treatment-related adverse events (TRAEs) occurred in 95% of patients receiving pembrolizumab plus chemotherapy and 90% of those receiving chemotherapy alone. Grade 3 to 5 TRAEs occurred in 63% and 80%, respectively, and led to discontinuation in 8% and 10%. No treatment-related deaths occurred in the pembrolizumab arm. One death in the chemotherapy arm (3%) was caused by tumor lysis syndrome. The most common TRAEs with pembrolizumab plus chemotherapy were anemia (65%), neutropenia/neutrophil count decreased (53%), thrombocytopenia/platelet count decreased (48%), nausea (38%), fatigue (35%), and alopecia (33%). The median duration of therapy was 4.2 months (range, 0-11.8) with pembrolizumab plus chemotherapy and 2.3 months (range, 0.1-4.0) with chemotherapy alone.

The investigators concluded that efficacy favored pembrolizumab plus chemotherapy across PD-L1 subgroups and that its safety profile was consistent with the known profiles of each agent, with no new safety signals. They cautioned that results should be interpreted carefully because of the small sample sizes, and said further analysis is planned.

References

1. von Amsberg G, Emmenegger U, Arslan C, et al. Updated results from the phase 1b/2 KEYNOTE-365 cohort I: pembrolizumab plus carboplatin and etoposide chemotherapy or chemotherapy alone for neuroendocrine prostate cancer. Presented at: 2026 European Society for Medical Oncology Congress; October 23-27, 2026; Madrid, Spain. Abstract 2217RO.
2. Von Amsberg G, Emmenegger U, Robbrecht DGJ. Phase 1b/2 KEYNOTE-365 cohort I: pembrolizumab (pembro) plus carboplatin and etoposide chemotherapy (chemo) or chemo alone for metastatic neuroendocrine prostate cancer (NEPC). J Clin Oncol. 2025;43(suppl 16):5059. doi:10.1200/JCO.2025.43.16_suppl.5059