Genomic alterations and their pathologic responses in high-risk localized prostate cancer (HRLPC) in subprotocol 1 of the Genomic Umbrella Neoadjuvant study (GUNS).

Abstract

403 Background: GUNS (NCT04812366) is a multicenter adaptive phase II trial evaluating 24 weeks of biomarker-selected, neoadjuvant androgen receptor pathway inhibitor (ARPI) combination therapies on depth of pathologic response (complete response [pCR] or <5 mm minimal residual disease [MRD]) in HRLPC. After 8 weeks of LHRHa + apalutamide (APA), participants are assigned to 1 of 4 sub-protocols (SP) combining 16 weeks of an ARPI doublet with drugs defined by specific genomic alterations (e.g. SP-2, docetaxel for RB1 , PTEN , TP53 loss; SP-3, niraparib for DNA-repair def ; SP-4, atezolizumab for mismatch-repair def ). SP-1 randomises men without these aggressive genomic alt and includes those that enhance AR activity (ETS fusions, FOXA1 , SPOP ) to either SP-1a (LHRHa + APA) or SP-1b (LHRHa + APA + abiraterone acetate/prednisone). SP-1 tests the hypothesis that ARPI triplet intensification, in cancers with AR-associated genomic alt , will increase depth of pathologic response. Methods: From 9/2021 to 8/2024, GUNS enrolled 95 and 30 men at University of BC and Toronto, respectively. Diagnostic biopsies underwent Tempus’ CLIA-certified 648-gene panel DNA sequencing (seq) and whole-transcriptome RNA-seq. This analysis focuses on 46 men enrolled to the 1 st stage of SP-1 who completed neoadjuvant therapy and surgery. Results: DNA-seq from 105/125 patients reveals a genomic landscape dominated by AR-associated alterations ( 36% ETS fusion, 23% FOXA1, 13% SPOP ). Other frequently altered genes were TP53 (14%), PTEN (12%), and BRCA2 (9%). Transcriptomes largely cluster in alignment with ETS fusions and SPOP status. ETS fusions were associated with PCS2-luminal-subtype and decreased proliferative signatures, whereas most other genomic alt were associated with PCS1-luminal-subtype. AR signatures associated with SPOP and FOXA1 mutations but not ETS fusions. 46 men, equally balanced for high-risk features and genomic alt , were randomized to SP-1a or SP-1b. Undetectable pre-surgery PSA levels trended higher in SP-1b (16/23) vs. SP-1a (11/23), but was not statistically significant (p = 0.12). While there were no pCR, MRD rates were significantly higher in SP-1b compared to SP-1a (43% vs 13%, p=0.012, odds ratio = 5.9). Degenerative scores (morphologic indicators of treatment stress) also averaged higher in SP-1b vs. SP-1a. Positive margin (17%) and lymph node (35% vs 26%) status were similar in both arms. Although genomic PTEN alt were assigned to SP-2, 7 patients in SP-1 were found to be PTEN neg by IHC and 6 were non-MRD; PTEN-IHC neg trended more common in non-MRD (21%) than MRD (8%) cases. Conclusions: SP-1 associated genomic alt (ETS fusion, FOXA1, SPOP) are the most frequent alterations in GUNS. Significantly higher rates of MRD in SP-1 patients treated with an ARPI triplet vs. doublet are of interest and support further evaluation with 2 nd stage expansion. Clinical trial information: NCT04812366 .