Initial report of a randomized trial of post-prostatectomy prostate cancer radiotherapy using either fluciclovine ( ¹⁸ F) or PSMA ( ⁶⁸ Ga) PET/CT for target dose-escalation.

Abstract

326 Background: Fluciclovine PET-guided prostate cancer (PCa) radiotherapy (XRT) improves failure-free survival (FFS) over conventional imaging (CI) alone in post-prostatectomy (RRP) recurrence (EMPIRE-1, PMID: 33971152). In this randomized trial we explored dose-escalation (DE) to sites of PET uptake (not done in EMPIRE-1) using either fluciclovine or 68 GaPSMA PET/CT-guided XRT and compared cancer control to the fluciclovine PET arm of EMPIRE-1 2Y failure free survival (FFS) of 79.6%. Methods: From 2019-2023, 140 pts w/ PCa with detectable PSA post-RRP & negative CI were stratified by: (a) PSA (<1.0 v ≥ 1.0 ng/mL), (b) adverse path [+ECE, +SV, +margin, +node] (0 v any) & (c) ADT (Y v N) & randomized to XRT directed by fluciclovine (Arm 1) v 68 GaPSMA (Arm 2). In both Arms, XRT decisions were PET determined: (A) extrapelvic (EP) uptake (no XRT); (B) pelvic uptake (XRT to pelvis + prostate bed [PB]); (C) PB only uptake (XRT to PB); & (D) no uptake (XRT to PB). Pelvic dose: 45.0-50.4/1.8 Gy (with optional DE {to PET uptake} up to 56 Gy); PB dose: 64.8-70.2 Gy (with optional DE up to 76 Gy). Failure was defined as in EMPIRE-1. Primary endpoint, declared a priori, was 2Y FFS comparing entire cohort [Arms (1+2)] to the 79.6% 2Y FFS of the fluciclovine PET arm of EMPIRE-1 using Z-test. KM curves for Arms 1 & 2 were compared using log-rank test. Univariate (UV) & multivariable (MV) analyses were performed by Cox proportional hazards model on demographic, disease, & treatment factors. Provider-reported [acute & late, GI & GU] toxicities were compared using χ 2 test. Results: 140 pts were enrolled (Arm 1: 70; Arm 2: 70). Arms were balanced on age, race, PSA, GG, ECE, SV, margin, node, & ADT use. 5 pts in Arm 1 and 1 pt in Arm 2 withdrew before PET. For pts completing XRT (Arm 1: 59; Arm 2: 60), median FU was 2.00 years. For primary endpoint, 2Y FFS for Arm (1+2) 87.4 v 79.6% target from EMPIRE-1 (p=0.018). There were no significant differences in 2Y FFS between Arm 1 v Arm 2 (88.2% v 86.9%, p=0.604). PET uptake in Arms 1 v 2 were: EP: 5 v 8; pelvic+/-PB: 10 v 10; PB only: 47 v 24; none: 3 v 27. Typical PB DE (n=74) was 74 Gy & pelvis DE (n= 16) was 55 Gy. For Arm 1 v Arm 2, covariates reaching p <0.10 on UV analysis were: +SV (p=0.010), +margin (p=0.039), PB uptake (p=0.068), & PB boost (p=0.085), and on MV analyses were: +SV (p=0.004), ADT (p=0.067), +margin (p=0.069), & PB boost (p=0.096). Toxicity was similar & low in both Arms 1 & 2. Conclusions: Although fluciclovine had higher diagnostic yield in the PB & 68 GaPSMA had higher yield for EP disease, both radiotracers had similar yield in the pelvic LN, & both radiotracers had similar impact on FFS in this setting. Integrating either fluciclovine or 68 GaPSMA into post-RRP XRT planning for DE to sites of PET uptake in PB or pelvis resulted in improved FFS over a prior trial in which no DE was performed; other factors may also have contributed to this observed difference. Clinical trial information: NCT03762759 .