Pembrolizumab for Treatment-Refractory Metastatic Castration-Resistant Prostate Cancer: Multicohort, Open-Label Phase II KEYNOTE-199 Study

Abstract

Purpose: Pembrolizumab has previously shown antitumor activity against programmed death ligand 1 (PD-L1)-positive metastatic castration-resistant prostate cancer (mCRPC). Here, we assessed the antitumor activity and safety of pembrolizumab in three parallel cohorts of a larger mCRPC population.

Methods: The phase II KEYNOTE-199 study included three cohorts of patients with mCRPC treated with docetaxel and one or more targeted endocrine therapies. Cohorts 1 and 2 enrolled patients with RECIST-measurable PD-L1-positive and PD-L1-negative disease, respectively. Cohort 3 enrolled patients with bone-predominant disease, regardless of PD-L1 expression. All patients received pembrolizumab 200 mg every 3 weeks for up to 35 cycles. The primary end point was objective response rate per RECIST v1.1 assessed by central review in cohorts 1 and 2. Secondary end points included disease control rate, duration of response, overall survival (OS), and safety.

Results: Two hundred fifty-eight patients were enrolled: 133 in cohort 1, 66 in cohort 2, and 59 in cohort 3. Objective response rate was 5% (95% CI, 2% to 11%) in cohort 1 and 3% (95% CI, < 1% to 11%) in cohort 2. Median duration of response was not reached (range, 1.9 to ≥ 21.8 months) and 10.6 months (range, 4.4 to 16.8 months), respectively. Disease control rate was 10% in cohort 1, 9% in cohort 2, and 22% in cohort 3. Median OS was 9.5 months in cohort 1, 7.9 months in cohort 2, and 14.1 months in cohort 3. Treatment-related adverse events occurred in 60% of patients, were of grade 3 to 5 severity in 15%, and led to discontinuation of treatment in 5%.

Conclusion: Pembrolizumab monotherapy shows antitumor activity with an acceptable safety profile in a subset of patients with RECIST-measurable and bone-predominant mCRPC previously treated with docetaxel and targeted endocrine therapy. Observed responses seem to be durable, and OS estimates are encouraging.

Trial registration: ClinicalTrials.gov NCT02787005.

FIG 1.

Patient disposition.

FIG 2.

Waterfall plots of change from baseline in tumor size and prostate-specific antigen (PSA) level. (A) Best percentage change from baseline in the sum of longest diameters of target lesions as assessed by RECIST v1.1 by central review in the 199 patients enrolled in cohorts 1 and 2. Asterisks represent the 23 patients in cohort 1 and 10 patients in cohort 2 who were not evaluable for change from baseline in tumor size because they did not have one or more evaluable postbaseline imaging assessment. (B) Best percentage change from baseline in PSA level in the 243 patients enrolled in cohorts 1, 2, and 3 who had a baseline PSA measurement. Asterisks represent the 27 patients in cohort 1, 12 patients in cohort 2, and 11 patients in cohort 3 who had a baseline PSA measurement but did not have one or more postbaseline PSA measurement. Changes > +100% were truncated at +100%.

FIG 3.

Antitumor activity. (A) Swimmer plot of patients with confirmed response as assessed by Prostate Cancer Working Group 3 (PCWG3)–modified RECIST v1.1 per central review in cohorts 1 (n = 7) and 2 (n = 2) or with stable disease or noncomplete response/nonprogressive disease of 6 or more months duration in cohorts 1 (n = 10), 2 (n = 10), and 3 (n = 23). Patients who died after month 6 without evidence of disease progression before death were considered to have stable disease for 6 months or longer. The end of the bar indicates the time to the last imaging assessment. (B) Kaplan-Meier estimates of radiographic progression-free survival (rPFS) as assessed by PCWG3-modified RECIST v1.1 per central review in cohorts 1 (n = 133), 2 (n = 66), and 3 (n = 59). (C) Kaplan-Meier estimates of overall survival (OS) in cohorts 1, 2, and 3.