Pembrolizumab plus enzalutamide in castration- resistant prostate cancer that exhibited enzalutamide resistance led to positive antitumor signals and a favorable safety profile, according to results of the phase II KEYNOTE-199 trial presented at the 2020 Genitourinary Cancers Symposium.
Julie N. Graff, MD
Pembrolizumab (Keytruda) plus enzalutamide (Xtandi) in castration- resistant prostate cancer (mCRPC) that exhibited enzalutamide resistance led to positive antitumor signals and a favorable safety profile, according to results of the phase II KEYNOTE-199 trial presented at the 2020 Genitourinary Cancers Symposium.1
“The addition of pembrolizumab to enzalutamide after enzalutamide resistance showed modest antitumor activity and durable response in patients with RECIST-measurable and bone-predominant mCRPC,” Julie N. Graff, MD, associate professor of medicine at Oregon Health & Science University in Portland, said during a presentation of the data.
Cohorts 4 (n = 81) and 5 (n = 45), consisting of patients with RECIST-measurable disease and bone metastasesonly/bone-predominant disease, respectively, were included in the current analysis. The primary end point was objective response rate (ORR) by RECIST 1.1 per blinded-independent central review in cohort 4. Secondary end points included disease control rate (DCR), radiographic progression-free survival (rPFS), prostate-specific antigen (PSA) response rate, overall survival (OS), and safety. Median time from enrollment to data cutoff was 15.3 months (range, 6.7-20.7) in cohort 4 and 19.1 months (range, 6.520.9) in cohort 5.
Among patients in cohort 4, 53% had a reduced target lesion, with 24% of patients having ≥30% reduction from baseline. The ORR for this cohort was 12%, composed of 2 complete responses (CRs; 2%) and 8 partial responses (PRs; 10%). Stable disease of any duration occurred in 38%, and the DCR for this group was 51%.
The duration of response in cohort 4 was 6.3 months (95% CI, 2.5+ to 13.4), with 60% of responses lasting longer than 6 months.
Patients in cohort 5 had a DCR of 51%, comprised exclusively of non-CR and non-PR changes to the target lesion from baseline. In both cohorts, the PSA response rate was 14%, with individual rates of 16% and 9% in cohorts 4 and 5, respectively. Forty percent of patients had any PSA decrease from baseline, with 14% having ≥50% decrease (TABLE).
“For some of the responders, the PSA was nearly a 100% response,” Graff noted, looking at a waterfall plot of PSA response.
Median rPFS in cohort 4 was 4.2 months (95% CI, 2.5-6.0) and 4.4 months (95% CI, 3.4-6.2) in cohort 5. Corresponding rates at 12 months were 17% and 23%.
Median OS was not reached (NR; 95% CI, 15.9NR) in cohort 4 and 18.8 months (95% CI, 14.0NR) in cohort 5. Rates of 12-month OS were 70% and 75%, respectively.
“Of 81 patients enrolled in cohort 4… 13 patients are still on treatment,” Graff said. In cohort 5, she said, “there are still 6 patien ts of those 45 patients [who are undergoing] treatment.”
Of treatment-related adverse events, fatigue occurred most often (24% of patients), followed by hyperthyroidism (16%); rash (15%); pruritis, decreased appetite, and maculopapular rash (10% each); diarrhea (9%); and nausea (7%). Another notable but unexpected event was rash, with any-grade events occurring at a rate of 15%, of which 2% were grade ≥3.
“We didn’t think about rash when we used these agents together. The great majority of the time, it was controlled with oral or topical steroids. However, 1 patient did have to receive intravenous steroids,” Graff said.
Immune-mediated events were observed in 29% of patients, including hypothyroidism in 16%, severe skin reaction in 6%, hyperthyroidism in 3%, and 2% each of adrenal insufficiency, myocarditis, pneumonitis, type 1 diabetes mellitus, colitis, and hypophysitis.
Patients in cohorts 4 and 5 had median ages of 74 years and 69 years, respectively. Corresponding PSA values at baseline were 31 ng/mL (range, 0.4-1667) and 19 ng/mL (range, 1.4-1750), and two-thirds of patients in both cohorts had a Gleason score ≥8. Most patients had prior enzalutamide exposure lasting ≥6 months. The most frequently reported ECOG performance status was 0 in cohort 4 at 53%, followed by 1 in 43% and 2 in 4%. In cohort 5, 58% of patients had an ECOG performance status of 1, followed by 38% at 0 and 4% at 2. More patients in cohort 4 had PD-L1positive disease at 40% versus 20% in cohort 5.
“I’d like to point out that patients in cohort 4 had a better performance status. They also were more likely to have PD-L1 positivity,” Graff said. “Interestingly, out of 81 patients, 19 had liver metastases, which is a poor prognostic indicator.”
Prior to this analysis, single-agent pembrolizumab demonstrated durable activity in patients with prostate cancer whose tumors were PD-L1 positive in cohorts 1 and 2 of KEYNOTE-199 and in the phase Ib KEYNOTE-028 trial.2,3
Results of a phase II trial of 28 patients that Graff presented at the 2018 American Society of Clinical Oncology Annual Meeting showed that pembrolizumab added to enzalutamide demonstrated a 25% radiographic response rate for patients with measurable disease and PSA response reduction ≥50% in 18% of patients.4
“The combination is now being studied in a phase III trial, KEYNOTE-641 [NCT03834493], which is for patients who are enzalutamide naïve,” Graff concluded. This trial will evaluate enzalutamide plus either pembrolizumab or placebo in a randomized fashion to determine OS and rPFS in patients with mCRPC.