In this oral presentation, Johann S. De Bono, MD, Ph.D shared data from the first three cohorts of KEYNOTE-199, a study of pembrolizumab 200 mg every 3 weeks for 35 cycles or until progression. The primary endpoint was overall response rate by an independent central review for cohorts 1 and 2. Key secondary endpoints included disease control rates by PCWG3 modified RECIST, duration of response, overall survival, and safety.
The first cohort was a PD-L1 positive population, as defined by the Agilent pharmDX 22C3 assay. Positivity was defined as a combined positive score greater than or equal to 1. 131 patients have been enrolled in cohort 1 with a median follow up time of 8.1 months. Cohort 2 is the PD-L1 negative cohort – 67 patients have been enrolled and treated with a median follow up of 7.9 months. Cohort three is a bone predominant cohort with no measurable disease per RECIST, regardless of PD-L1 status.
At baseline, all three cohorts had median ages between 68-70. Cohort 1 and 2 had a higher PSA mean (308 and 346 respectively), compared with the bone predominant cohort which was expected. About 25-32% of patients had 2 or more lines of chemotherapy, 22-26% had 2 or more lines of anti-endocrine therapies, and about a quarter of patients have had enzalutamide and abiraterone.
The waterfall plot showed that both PD-L1 positive and PD-L2 negative patients had PSA responses. 10% of patients had a PSA decline between 30 to 100% from baseline. When all three cohorts were combined, 11% of patients had a PSA decline greater than 50%. In terms of the radiographic responses by central review, in the PD-L1 positive population, 2 patients (2) had a CR, 5 (4%) patients had a PR, and 22 (17%) patient had stable disease. In the PD-L1 negative, population, similar results were shown (See table below), with no CRs.
During the abstract discussion session by Dr. Douglas G. McNeel noted this lack of difference for overall response between cohort 1 and 2 especially given that this trial format was designed to enrich for a response in the PD-L1 positive cohort. Another interesting point by Dr. McNeel was that after genomic testing, there were more patients who responded who didn’t have BRCA1/2 or ATM, or other DDR genes (4), compared to those did have DNA repair defects (2). This is also similar to what was reported by Dr. Graff in the earlier mentioned abstract3.
In conclusion, single pembrolizumab has anti-tumor activity in a small proportion of patients with prostate cancer. However, PD-L1 status alone does not appear to separate out the responders from non-responders. Additionally, DNA repair defects also did not accurately predict which patients would respond. I think that further biomarker work is necessary to help select for patients who will derive benefit from single-agent pembrolizumab and that further research in combination therapy with immunotherapy as a backbone is necessary.
- Topalian SL, Hodi FS, Brahmer JR, et al. Safety, activity, and immune correlates of anti–PD-1 antibody in cancer. New England Journal of Medicine 2012;366:2443-54.
- Hansen A, Massard C, Ott PA, et al. Pembrolizumab for patients with advanced prostate adenocarcinoma: Preliminary results from the KEYNOTE-028 study. Annals of Oncology 2016;27:725PD-PD.
- Graff JN, Alumkal JJ, Thompson RF, et al. Pembrolizumab (Pembro) plus enzalutamide (Enz) in metastatic castration resistant prostate cancer (mCRPC): Extended follow up. American Society of Clinical Oncology; 2018.
Presented by: Johann S. De Bono, MD, Ph.D., FRCP, Royal Marsden Hospital
Written by: Jason Zhu, MD. Fellow, Division of Hematology and Oncology, Duke University, at the 2018 ASCO Annual Meeting - June 1-5, 2018 – Chicago, IL USA