199 Background: 177Lu-PSMA-617 is approved for post taxane, post ARPI mCRPC based on the VISION trial. With the emergence of PSMAfore and earlier line data, there is growing interest in if radioligand therapy should be moved upstream, particularly in taxane naive, ARPI resistant disease. To inform sequencing decisions, we compared treatment effect estimates in post taxane versus taxane naive settings using pooled randomized evidence. Methods: Randomized controlled trials evaluating 177Lu-PSMA-617 in mCRPC with available hazard ratios for radiographic progression free survival (rPFS) or overall survival (OS) were included. Trials were grouped by taxane exposure status (post taxane vs taxane naive). Fixed effect models were used for figure presentation, and Hartung–Knapp random effects models were reported in text to account for small study numbers. PSMAfore was included using IPCW adjusted OS estimates to address crossover. Results: Four trials (VISION, PSMAfore, TheraP, Satapathy; n=1,559) contributed to rPFS pooling. The fixed effect pooled HR for rPFS was 0.52. In taxane naive trials (PSMAfore and Satapathy), HR was 0.54 (95% CI, 0.43 - 0.64), while post-taxane trials (VISION and TheraP) showed HR 0.50 (95% CI, 0.40 - 0.64). Using HK adjustment, the pooled HR for rPFS was 0.54 (95% CI 0.10–2.93) in taxane naive disease and 0.50 (95% CI 0.03–9.03) post taxane. OS synthesis using IPCW adjusted PSMAfore and VISION showed a pooled HR of 0.62 (95% CI 0.52–0.73). All estimates favored radioligand therapy with consistent direction of effect across subgroups. Conclusions: The relative benefit of 177Lu-PSMA-617 appears comparable in taxane naive and post taxane mCRPC, suggesting that treatment activity is preserved when used earlier in the disease course. These findings support a biology driven sequencing model in which PSMA PET based selection, rather than taxane exposure history alone, may guide placement of radioligand therapy. Prospective sequencing focused trials will be essential to validate optimal timing and ensure equitable implementation.
Dirican et al. (Sun,) studied this question.