Assessment of PSA responses and changes in the rate of tumor growth (g-rate) with immune checkpoint inhibitors in US Veterans with prostate cancer

Abstract

Background: The value of immune checkpoint inhibitors (PD1/PDL1 inhibitors; ICI) in treating prostate cancer (PC) is limited. We examined data from US Veterans with PC to assess disease response to ICIs as monotherapy or combined with abiraterone or enzalutamide. We compared results with reference datasets to assess ICI efficacy in the real-world. Methods: We queried the VA corporate data warehouse (CDW) to identify Veterans with a diagnosis of PC who received ICI for any malignancy and had ≥1 PSA measurement while receiving ICI. To evaluate ICI monotherapy, we restricted analysis to Veterans who had not received LHRH agonists/antagonists, PC-directed medical therapy, or radiation/extirpative surgery of the bladder/prostate within and preceding the duration of ICI administration. For ICI combination analysis, we identified Veterans who received abiraterone or enzalutamide for PC while on ICI. We calculated rates of tumor (PSA) growth (g-rates), comparing them to a 1:2 matched reference cohort. Results: We identified 787 Veterans with PC and ≥1 PSA measurement while receiving an ICI. The median duration of ICI therapy was 155 days. 223 Veterans received ICI monotherapy, with only 17(8%) having a reduction in PSA (median decline=43%). 12 (5%) had PSA declines >30% (PSA30) which included 6 (3%) who had PSA reductions greater than 50% (PSA50). Median g-rates for ICI plus abiraterone (n=20) or enzalutamide (n=31) were 0.000689/d-1 and 0.002819/d-1, respectively, and were statistically insignificant compared to g-rates of matched cohorts receiving abiraterone (g=0.000925/d-1, p=0.73) or enzalutamide (g=0.001929/d-1, p=0.58) alone. Conclusion: Our data align with clinical trial data in PC, demonstrating limited benefit from ICI monotherapy and predicting no survival benefit from simultaneous administration of abiraterone or enzalutamide with an ICI using g-rate. We demonstrate the value of estimating g-rates and of our reference database in approaching challenging clinical questions and as aids in drug development.