Real-world (RW) characteristics, management, and outcomes of genomically defined mismatch repair deficient (MMR-d) and aggressive variant prostate cancer (AVPC).

e17042 Background: AVPC (≥ 2 alterations in PTEN / TP53 / RB1 ) and MMR-d are genomically distinct subsets of prostate cancer with poor response to androgen-pathway therapies (ARPi) and docetaxel (DTx). Methods: This prospective RW multicenter study (Alberta, Canada) investigates the incidence, characteristics, systemic treatment (tx), and outcomes of pts with metastatic (m) and/or castration resistant prostate cancer (CRPC) receiving genomic testing (2018-2025). Pts with AVPC or MMR-d were compared with pts with no biomarker detected (NBD: lacking AVPC, MMR-d and HRD). Pt characteristics, time to CRPC (tCRPC) from androgen deprivation therapy (ADT) initiation, survival post-mCRPC (OS), time to PSA progression (TTPP), and radiographic PFS (rPFS) were assessed using univariate tests of association and Kaplan-Meier analyses. Results: Of 178 cases, 19 (12%) were AVPC (18 PTEN, 4 RB1, 17 TP53), 19 (12%) MMR-d (4 EPCAM, 3 MLH1, 6 MSH2, 6 MSH6, 2 PMS2), and 140 (76%) NBD. HRD was observed concurrently in 42% (8/19) each of AVPC [ATM/PTEN/TP53 (n = 2); BRCA2/PTEN/TP53 (n = 3); BRCA2/TP53/RB1 (n = 2); PALB2/PTEN/TP53 (n = 1); RAD51C/TP53/RB1 (n = 1)] and MMR-d [ATM/MLH (n = 1); ATM/EPCAM/MSH6 (n = 1); BRCA2/MSH6 (n = 1); BRCA2/MSH2 (n = 2); CDK12/MSH6 (n = 1); CDK12/MSH2 (n = 1); 1 PALB2/MSH6 (n = 1)]. Excluding concurrent HRD, 11 AVPC (11 PTEN, 2 RB1, 9 TP53), 11 MMR-d (3 EPCAM, 1 MLH1, 2 MSH2, 4 MSH6, 2 PMS2), and 140 NBD were analyzed. Concurrent HRD did not impact tCRPC or OS for AVPC or MMR-d. Clinicopathological features (age, Gleason ≥8, disease volume, visceral or de novo metastases, pathology, PSA) were similar across groups. tCRPC was 16.6, 18.4 and 29.9 mo for AVPC, MMR-d and NBD, respectively (p = 0.6). AVPC had a significantly shorter OS (11.9 mo), compared to MMR-d (13.6 mo) and NBD (37 mo); p < 0.01). 97/178 pts with mCSPC received treatment intensification (66 ARPi [26 APA; 31 ABI; 9 ENZA), 19 ARPi + DTx [8 ABI; 11 DARO], and 12 DTx). 1L mCRPC tx consisted of 82% (41/50) ARPi [2 APA; 20 ABI; 19 ENZA), 16% (8/51) DTx, and 2% (1/50) platinum chemotherapy (P). 2L for mCRPC pts was 77% (39/55) DTx, 24% (13/55) ARPi [6 ABI; 7 ENZA], 3% (2/55) cabazitaxel, and 2% (1/55) P. Conclusions: This RW study identified a high proportion of pts harbouring AVPC or MMR-d were also concurrently HRD. Yet, AVPC is a distinct aggressive subtype with poor outcomes, demonstrating a short OS and tCRPC, and rapid TTPP and rPFS on both ARPi and Dtx. No pts received AKT inhibitors and a single pt with MMR-d (PMS2) received a self-funded immune checkpoint inhibitor, underlining the need for earlier comprehensive biomarker testing in M1 or CRPC and further trial development in these unique subtypes. AVPC n (%) MMR-dn (%) NBDn (%) p 1L Single Agent ARPi N 9/11(81) 8/11 (72) 90/126 (71) 0.3 TTPP 8.3 14.8 26.6 <0.01* R-PFS 8.2 10.3 21.4 <0.01* 2L DTx N 4/7 (57) 5/6 (83) 34/47 (72) 0.8 TTPP 1.4 11.1 10.2 <0.01* R-PFS 4.2 3.0 7.1 0.5