Prospective Characterization of Circulating Tumor Cells in Hormone Sensitive Oligometastatic Prostate Cancer Patients on a Metastasis-Directed Therapy Trial

Abstract

Prospective data have shown that metastasis-directed therapy (MDT) can alter the natural history of oligometastatic disease. In hormone-sensitive prostate cancer (HSPC), the clinical effect of MDT has been validated by STOMP, ORIOLE and SABR-COMET phase II trials. Circulating tumor cells (CTCs) are likely the source for the formation of macroscopic metastases. CTCs may provide an approach for identifying subgroups of patients with oligometastatic HSPC (oligoHSPC) that would benefit most from MDT. Our main goal was to evaluate the feasibility of CTC detection and subtypes in oligoHSPC patients that may benefit from MDT. ORIOLE randomized men with recurrent HSPC with 1-3 metastases to observation (Obs) vs. stereotactic ablative radiotherapy (SABR) MDT. Blood samples were prospectively collected at baseline (D0) and 6-mos (D180) and shipped for analysis on Epic Sciences liquid biopsy platform (Epic Sciences, San Diego, CA). Machine learning algorithms identified CTCs and characterized androgen receptor (AR) and PSMA expression. Association with clinical factors and outcomes were examined. Biochemical failure-free survival (BFFS) event was a PSA rise of at least 2 ng/mL and 25% above nadir. Progression-free survival (PFS) was a composite endpoint including BFFS event, radiologic progression (RECIST v1.1); symptomatic progression; initiation of ADT; or death. Comparisons of patient and tumor characteristics performed by two-sample t-tests. Survival curves were generated by the Kaplan-Meier method and evaluated by the log-rank test. Effect of SABR on post-SABR on CTC levels were calculated by McNemar test. A total of 82 samples were collected in ORIOLE: 70 SABR (35 D0 and 35 D180) and 12 Obs (7 D0 and 5 D180). 30/42 men had CTCs detected on D0 (71%; AR+ = 7, PSMA+ = 13) and in 26/40 on D180 (65%; AR+ = 9, PSMA+ = 8). Median follow-up was 41.7-mos. There was no association between CTC presence or subtypes (AR+ or PSMA+) with Gleason score or PSA. PFS was significantly lower in the patients with AR+ vs. AR- CTCs on D0 in the SABR arm (p = 0.011, median PFS: AR+ = 9.3- vs. AR+ = 27.1-mos). The median BFFS trended towards a difference for AR+ = 12.9- vs. AR- = 29.2-mos (D180, p = 0.058). SABR had no effect on the presence or subtypes of CTC at D180. Baseline and dynamic CTC levels and their subtypes in oligoHSPC from the ORIOLE randomized trial of MDT was examined. AR+ CTCs at baseline and 6-mos were correlated with clinical outcomes following SABR. Longer follow-up, further analysis and a greater number of patients are needed for a more comprehensive conclusion.