Genomic predictors of response and resistance to 177-Lu-PSMA-617 using circulating tumor DNA.

Abstract

220 Background: There is a critical need to identify biomarkers of response and resistance to 177Lu-PSMA-617. Post-hoc analyses of VISION have explored the prognostic and predictive value of select clinical parameters. However, we are still in the early stages of deciphering which molecular changes are associated with favorable or adverse treatment outcomes. Methods: We conducted a retrospective analysis of patients with metastatic castration-resistant prostate cancer (mCRPC), who received 177Lu-PSMA-617 at Mayo Clinic in Rochester, Minnesota, in the interval of March 2022 to March 2023. We included patients who provided research authorization, had a baseline PSA of at least 2 ng/mL, received at least 2 cycles of treatment, and had circulating tumor DNA (ctDNA) genomic profiling performed by Guardant (83 gene panel) within 60 days of treatment initiation. Clinical, laboratory, genomic, and radiomic parameters were abstracted by trained research personnel. Two independent analyses of “response” and “resistance” were performed. Response to treatment was defined as 50% or greater decrease in PSA from baseline (PSA50 response: dichotomized yes versus no), and resistance was defined as non-decreasing or rising PSA during treatment (PSA non-response: dichotomized yes versus no). Associations between gene alterations and response and resistance were evaluated via Fisher’s exact tests. Results: Between March 2022 and March 2023, 273 patients received cycle 1 of 177Lu-PSMA-617. A total of 108 patients met the inclusion criteria. The median number of cycles at data cutoff was 6 (IQR: 4,6) with 57 (53%) patients receiving all 6 planned cycles. The median PSA at baseline was 16.65 ng/ml (IQR: 5.20, 66.12). A total of 65 patients (60%) experienced a PSA50 response. No molecular alterations were associated with a higher rate of PSA50 response; however, alterations in ARID1A, AR, and CHEK2 were associated with lack of PSA50 response (Fisher’s exact p < 0.05). Of the 108 evaluable patients, 20 (19%) met criteria for PSA non-response. Alterations in AR, ARID1A, MYC, TP53, CHEK2, ATM, EGFR, and NOTCH1 were significantly associated with PSA non-response (Fisher’s exact p < 0.05); in each case, the presence of the alteration was positively associated with resistance (i.e., more instance of PSA non-response). No biomarkers were found to be inversely associated with PSA non-response. Conclusions: ctDNA assessed by commercially available, targeted sequencing panels prior to treatment may be useful in identifying the patients least likely to benefit from 177Lu-PSMA-617. Further work is ongoing to integrate ctDNA findings with clinical parameters in the hopes of understanding the factors mediating response and resistance to therapy.