Koichiro Kimura, Vishnu Murthy, Andrew F. Voter, Lilja B. Sólnes, Abuzar Moradi Tuchayi, Surekha Yadav, Lela Theus, Andrew T. Nguyen, Vinícius Ludwig, Adrien Holzgreve, Lena M. Unterrainer, Tristan R. Grogan, Johannes Czernin, MD Thomas A. Hope, Andrei Gafita, Jérémie Calais
Pretherapeutic visual and quantitative indicators derived from prostate-specific membrane antigen (PSMA) PET/CT have been proposed as predictors of response to [177Lu]Lu-PSMA-617 (177Lu-PSMA) therapy in patients with metastatic castration-resistant prostate cancer. This study aimed to evaluate and compare the prognostic performance of these indicators in a cohort treated under the U.S. Expanded-Access Program. Methods: This retrospective analysis included 88 patients with metastatic castration-resistant prostate cancer from 3 U.S. institutions (University of California Los Angeles, University of California San Francisco, and Johns Hopkins) enrolled in the Expanded-Access Program who underwent baseline PSMA PET/CT before receiving 177Lu-PSMA. We assessed visual indicators—such as the visual PSMA PET tumor–to–salivary gland ratio, tumor heterogeneity, and intensity scores—and quantitative metrics including total tumor volume (TTV), total tumor SUVmean, total tumor SUVmax, total lesion uptake (TTV × total tumor SUVmean), total lesion quotient (TTV ÷ total tumor SUVmean), and the quantitative PSMA PET tumor–to–salivary gland score. Associations with clinical outcomes—a 50% or greater prostate-specific antigen decline (PSA50), prostate-specific antigen (PSA) progression-free survival (PFS), and overall survival (OS)—were analyzed using univariate and multivariate models. Predictive performance was evaluated via the concordance index from Cox proportional hazards regression. Results: After a median follow-up of 36.1 mo (95% CI, 33.8–37.8 mo), the PSA50 rate was 43%, the median PSA PFS was 4.5 mo (95% CI, 3.7–7.2 mo), and the 2-y PSA PFS rate was 7% (95% CI, 3–16%). The median OS was 12.5 mo (95% CI, 10.4–17.1 mo), and the 2-y OS rate was 29% (95% CI, 20–40%). Among the evaluated metrics, total tumor SUVmean showed the highest predictive accuracy for PSA50 (area under the curve, 0.81; 95% CI, 0.73–0.90). In multivariate analyses adjusted for clinical factors, a higher total tumor SUVmean was independently associated with improved PSA PFS (hazard ratio, 0.58; 95% CI, 0.39–0.86; P = 0.007) and OS (hazard ratio, 0.54; 95% CI, 0.34–0.86; P = 0.009). Total tumor SUVmean also yielded higher concordance index values compared with models based on clinical variables alone (PSA PFS, 0.667 vs. 0.594; OS, 0.687 vs. 0.661, respectively). Conclusion: Baseline total tumor SUVmean on PSMA PET/CT provides independent prognostic information beyond clinical parameters and may serve as a useful biomarker for patient selection and treatment personalization with 177Lu-PSMA.