AbstractBackgroundAndrogen receptor (AR) gene alterations, as detected by circulating tumor cell‐free DNA (cfDNA) genomic profiling, have been shown to emerge after a variable duration of androgen signaling inhibition. AR alterations were associated with inferior outcomes on treatment with androgen receptor pathway inhibitors (ARPI) in the first line metastatic castration‐resistant prostate cancer (mCRPC) setting in a phase 2 trial. Here in, we assessed the impact of these AR alterations on survival outcomes in a real‐world patient population of mCRPC experiencing disease progression on an ARPI.MethodsIn this IRB‐approved retrospective study, consecutively seen patients with a confirmed diagnosis of mCRPC, with disease progression on a treatment with ARPIs in the first line mCRPC setting, with no prior exposure to an ARPI in the castration sensitive setting, and with available cfDNA profiling from a CLIA certified laboratory were included. Patients were categorized based on AR status: wild‐type (AR_wt) or alteration‐positive (AR₊). The objective was to correlate overall survival (OS) after disease progression on the first‐line ARPI with the presence or absence of AR alterations. Kaplan–Meier and Cox Regression Tests were used as implemented in R‐Studio (v.4.2).ResultsA total of 137 mCRPC patients were eligible: 69 with AR_wt versus 68 with AR₊. The median OS posttreatment with the first ARPI was significantly higher for AR_wt than AR₊ patients (30.1 vs. 15.2 mos; p < 0.001). Of 108 patients who received a subsequent line of therapy, 63 received an alternate ARPI (AR₊ 39 vs. 24 AR_wt), while 20 received a taxane‐based therapy (11 AR₊ vs. 9 AR_wt). Among patients receiving an alternate ARPI, AR₊ had numerically shorter OS (16.8 vs. 30.4 mos, p = 0.1). Among patients receiving taxane‐based regimens, the OS was not significantly different between AR₊ and AR_wt (14.5 vs. 10.1 mos, p = 0.18).ConclusionIn this real‐world study, mCRPC patients with AR alterations on cfDNA had inferior OS after disease progression on the first ARPI, compared to those who did not, and may impact outcomes on a subsequent ARPI but not on subsequent taxane‐based therapy received. By providing survival estimates for patients with or without AR alterations, our data may aid in patient counseling, prognostication, treatment decision, and for designing future clinical trials in this setting.