Clinical and genomic predictors of sipuleucel-T (sip-T) outcomes in men with metastatic castration-resistant prostate cancer (mCRPC).

Abstract

228 Background: Survival outcomes with sip-T in mCRPC are variable, and mechanisms driving these differences remain unknown. Tumor genomics influencing immune response may explain outcome heterogeneity, with immunogenic features (e.g. CDK12 mutation [mut]) potentially conferring benefit, while immune-evasive alterations (e.g. MYC, PTEN, AR amplification) may predict poorer outcomes. Identifying clinical and genomic correlates of differential outcomes may enhance patient selection and optimize sip-T efficacy. Methods: This single-center retrospective study included men who received ≥1 sip-T infusion at Duke Cancer Institute (2014-2024) with tumor genomic data in the Duke Molecular Registry of Tumors. The primary endpoint was overall survival (OS), defined as time from first sip-T infusion to death from any cause. Secondary endpoints included real-world progression-free survival (PFS), time to next therapy, and PSA response. Kaplan-Meier estimates were used to calculate median time-to-event and 36-mo survival. Elastic net was used to identify and rank factors predictive of OS, which were then evaluated in a multivariable cox model. Results: Among 429 men treated with sip-T, 185 (43%) had molecular data (tumor tissue 43%, cfDNA 57%). Median age was 70 yrs (48-95); most were ECOG 0 (60%) and White (85%) or Black (14%). Frequent alterations included TP53 (48%), AR amplification (27%), PTEN loss (20%), and MYC gain (10%). Overall, median OS and PFS were 44.0 and 3.6 mos, respectively. MYC gain emerged as the strongest predictor of worsened OS. Among men with MYC gain (n=18) vs those without (n=167), median OS was 16.0 vs 44.0 mos and PFS 2.5 vs 3.9 mos. Median time to next therapy was 4.1 vs 7.4 mos, and median change in PSA post-sip-T was +111% vs +25%. At 36 mos, survival probabilities were: no detectable mutations 82%, MSS 64%, PTEN mut/loss 56%, TP53 mut/loss 51%, AR amplification 46%, MYC gain 28%, and RB1 mut/loss 25%. In multivariable analysis, longer time from diagnosis to therapy (HR 0.99, p=0.001) and higher hemoglobin (HR 0.83, p=0.013) were independently associated with improved OS, whereas ECOG PS >0 (HR 1.52, p=0.027), older age (HR 1.03, p=0.024), and MYC gain (HR 2.53, p=0.002) or AR amplification (HR 1.58, p=0.044) predicted worse OS. Conclusions: Among men with mCRPC treated with sip-T, MYC and AR gain had strong and significant independent associations with rapid post-treatment progression and poor OS. Lack of anemia, good PS, longer time since diagnosis, and no detectable genetic alterations associated with improved OS, consistent with prior studies of other mCRPC therapies. Rapid tumor proliferation and progression likely explain inferior outcomes and limited benefits, and concurrent or alternative therapies should be considered. Multicenter external validation of these findings is planned to better define the subset of men most likely to benefit from sip-T.