Patient-reported outcomes (PROs) in KEYLYNK-010: Pembrolizumab (pembro) plus olaparib (ola) vs abiraterone acetate (abi) or enzalutamide (enza) for patients (pts) with previously treated metastatic castration-resistant prostate cancer (mCRPC).

131 Background: The phase 3, randomized KEYLYNK-010 trial (NCT03834519) of pembro + ola vs next-generation hormonal agent (NHA) abi or enza did not significantly improve rPFS or OS in molecularly unselected pts with mCRPC treated with prior NHA and docetaxel. The study was stopped for futility after the second prespecified interim analysis. PROs for pembro + ola vs NHA in KEYLYNK-010 are presented. Methods: Pts were randomly assigned 2:1 to receive pembro 200 mg IV Q3W for ≤35 cycles (~2 y) + ola 300 mg orally BID or NHA (either abi 1000 mg orally QD + prednisone 5 mg orally BID, if pt previously received enza, or enza 160 mg orally QD if pt previously received abi). PROs were evaluated in pts who received ≥1 dose of study treatment and had ≥1 PRO assessment. FACT-P and BPI-SF were administered at baseline, Q3W until wk 24, Q6W until wk 72, and Q12W thereafter for ≤2 y. Time to pain progression (TTPP) based on BPI-SF was a prespecified secondary end point. Prespecified exploratory end points included least squares mean (LSM) change from baseline to wk 15 for FACT-P total and subscales scores (FACT-G total, TOI, FAPSI-6, FWB, PWB, and PCS) and BPI-SF scores (pain interference, pain severity, and worst pain), and time to deterioration (TTD) and overall improvement rate in FACT-P total and subscale scores. Differences were evaluated using 2-sided nominal P values not controlled for multiplicity. Results: A total of 793 pts were randomly assigned to pembro + ola (n = 529) or NHA (n = 264). As of January 18, 2022, median follow-up was 18.7 mo (range, 6.1-31.7). In all randomized pts, completion rate for FACT-P and BPI-SF at baseline and wk 15 was >84% and >57%, respectively. No differences were observed in the median TTPP for pembro + ola (13.5 mo [95% CI, 9.7-NR]) vs NHA (12.0 mo [95% CI, 10.1-NR]; HR, 0.95 [0.72-1.26]). No LSM differences were observed in FACT-P total scores (pembro + ola, –4.62 [95% CI, –6.47 to –2.77] vs NHA, –5.86 [95% CI, –8.58 to –3.13]) or BPI-SF scores (Table). There were no differences in TTD in FACT-P total, FACT-G total, TOI, FAPSI-6, FWB, PWB, and PCS scores between groups. A numerically higher proportion of pts had improved + stable FACT-P total scores for pembro + ola (44.0%) vs NHA (39.0%). FACT-P and BPI-SF scores were generally maintained across all evaluated time points up to wk 81. Conclusions: No clinically meaningful changes from baseline were observed in HRQoL or disease-related symptom scores with either pembro + ola or NHA. PRO scores were generally similar between pembro + ola and NHA at all analyzed time points, suggesting HRQoL was maintained in heavily pretreated pts receiving pembro + ola. Clinical trial information: NCT03834519 . [Table: see text]