Decipher risk stratification of radiorecurrent prostate cancer: Correlative analysis of the F-SHARP trial of salvage reirradiation.

Abstract

419 Background: A third of patients with biochemical recurrence after radiation (RT) have intraprostatic radiorecurrence (IPR) on PSMA PET/CT. Patients with IPR have worse metastasis-free survival - a surrogate for progression to lethal prostate cancer (PCa). We previously reported the results from our F-SHARP clinical trial that demonstrated salvage reirradiation using focal dose-escalated high dose rate (HDR) brachytherapy is safe and effective. Here, we examine the Decipher score to determine if it could be a tool to risk stratify patients with IPR. Methods: F-SHARP (NCT03312972) is a multi-institutional phase I/II trial of focal dose-escalated salvage HDR for IPR. Patients were recruited from 2017-2023 at 3 centers. Eligibility criteria included a history of localized PCa treated with any form of definitive RT and biopsy-proven IPR with no regional or distant metastasis. Of the 62 participants, 37 consented for the biomarker correlative study and 31 (50%) had sample data passing quality control for Decipher analysis (Veracyte, San Diego, CA). De-identified data from 146,940 patients tested (2016-2024) with the Decipher prostate genomic classifier were retrieved from the GRID registry (NCT02609269) and used to create a matched cohort based on NCCN risk at diagnosis. Univariable Cox proportional hazards models were used to compare oncologic, CTCAE v4.03 toxicity, and EPIC-26 hrQoL events by Decipher risk group. Results: The biomarker cohort had similar baseline characteristics to the overall trial cohort. 30% received ADT with initial RT (73% external beam, 27% LDR brachytherapy). At recurrence, 71% had high Decipher risk (median score 0.67) as compared to only 35% (median score 0.48) in the matched GRID cases (n=130,760). Median time from initial RT to enrollment in Decipher low (<0.45) was 16.9 years, compared to 8.0 and 7.4 years in the intermediate (0.45-0.6) and high (>0.6) score patients. Median follow up was 32.3 months. Decipher score was not associated with toxicity or quality of life post-HDR (all p>0.05). As depicted in the table, higher Decipher score was associated with an increased risk of biochemical progression-free survival (bFS; p=0.03), local recurrence-free survival (LRFS; p=0.04), and radiographic progression-free survival (rPFS; p=0.01). At 3 years, bPFS was 43% vs. 75%, LRFS was 58% vs. 100%, and rPFS was 42% vs. 100% for Decipher high vs. lower risk (<0.6). Conclusions: Salvage reirradiation is a growing indication for RT in PCa. This is the first use of genomic risk stratification in this setting. Nearly a third of patients with IPR have a lower Decipher risk score, and our data suggest especially favorable outcomes with salvage HDR. Future studies to determine how Decipher risk stratification can be used to tailor further treatment intensification with systemic therapy for those most at risk of reirradiation failure are warranted. Clinical trial information: NCT03312972 . Endpoint Hazard Ratio per 0.1 unit (95% CI) bPFS 1.70 (1.05-2.75) LRFS 2.30 (1.02-5.18) rPFS 2.47 (1.23-4.97)