Circulating tumour cells (CTCs) and PSMA PET correlates in the phase I PRINCE trial of ¹⁷⁷Lu-PSMA-617 plus pembrolizumab for metastatic castration resistant prostate cancer (mCRPC).

5027 Background: The Phase I PRINCE trial (NCT03658447) is evaluating the efficacy of 177 Lu-PSMA-617 plus pembrolizumab for mCRPC. The utility of serial monitoring of CTCs and PET as biomarkers of prognosis and clinical benefit of 177 Lu-PSMA-617-based therapy remains unknown. Methods: 36 of 37 pts with high PSMA expression on PSMA PET underwent serial CTC collections in conjunction with PSMA PET at baseline, every 12 weeks for 48 weeks and every 24 weeks thereafter. CK+, CD45- CTCs were enumerated from 3ml of blood and stained for PSMA (Epic Sciences platform). Associations between PSMA+ CTC counts, PET molecular tumor volume (MTV), total lesional activity (TLA; MTVxSUVmean) were assessed by Spearman correlation. Cox models assessed the association of CTC and PSMA PET parameters with radiographic progression-free survival (rPFS) and PSA PFS. A subset of pre-treatment CTCs underwent single cell low-pass whole genome sequencing to characterize copy number aberrations. Results: 32/36 pt (89%) had detectable CTCs (median 7, range 0-514) with 23 (64%) being PSMA+ (median 1, range 0-224) at baseline. At w12, 23/33 (70%) had CTCs detected with 10 (30%) being PSMA+. Baseline PSMA+ CTC count and MTV were moderately correlated (r s = 0.57, p < 0.001). Of 22 evaluable pts with baseline PSMA+ CTC, 18 (82%) showed decrease by w12 with clearance in 13. This paralleled reductions in MTV (-18% med relative change, IQR: -57 to -1) and TLA (-48% med relative change, IQR: -77 to -28). Total CTC and PSMA+ CTC counts at baseline, and PET parameters were not associated with PSA PFS or rPFS. Clearance of PSMA+ CTC by w12 (13/22 pts) was associated with improved rPFS (med NR vs 3.0 mos, HR 0.23, 95%CI 0.07-0.74, p = 0.007) and PSA PFS (med 11.2 vs 3.5 mos, HR 0.28, 95%CI 0.11-0.73, p = 0.006). Persisting PSMA-neg CTCs at w12 trended to worse rPFS (med 4.1 vs 12.3 mos, p = 0.11) and PSA PFS (med 5.1 vs 12.3 mos, p = 0.07). Of pts not progressing by w12, decrease in PSA (HR 0.83 per 10% decrease, 95%CI 0.74-0.93, p < 0.001), MTV (HR 0.85 per 10% decrease, 95%CI 0.75-0.96, p = 0.008), MTV > 30% decrease (HR 0.30, 95% CI 0.08-1.08, p = 0.05) and TLA (HR 0.88 per 10% decrease, 95%CI 0.78-1.00, p = 0.04) associated with improved rPFS beyond w12. Pre-treatment CTCs (18 pts) exhibited genomic heterogeneity and frequent loss of PTEN, TP53 and RB1. Pts with compound TP53 and RB1 loss at baseline nonetheless had PSMA+ CTCs in high proportion (med 91.3% of total CTCs). Conclusions: PSMA PET-positive mCRPC is associated with high rates of PSMA+ CTCs which decline with 177 Lu-PSMA-617 plus pembrolizumab in parallel with PSMA MTV/TLA. Despite imaging suitability for therapy, CTCs had heterogenous PSMA expression and genomic alterations associated with aggressive disease. Early changes in PSMA+ CTCs and MTV/TLA were associated with outcomes and may aid in determining clinical activity of LuPSMA-based therapy.