Tumor Segmentation on PSMA PET/CT Predicts Survival in Biochemical Recurrence of Prostate Cancer: A Retrospective...

Background: PSMA PET/CT imaging has become a cornerstone in the management of prostate cancer, particularly in the setting of biochemical recurrence (BCR). While semi-quantitative parameters such as SUVmean have been evaluated as prognostic biomarkers in metastatic castration-resistant prostate cancer (mCRPC), particularly after the VISION trial, the prognostic role of volumetric measures such as Total Molecular Volume (TMV) remain largely unexplored, especially in earlier stages of the disease such as in biochemical recurrence following primary treatment. Methods: This retrospective monocentric study included 84 patients with BCR who underwent PSMA PET/CT imaging between 2020 and 2021 using either [⁶⁸Ga]Ga-PSMA-11(Ga-PSMA) or [¹⁸F]-PSMA-1007 (F-PSMA) as tracers. Total tumor burden was assessed through manual 3D segmentation to derive whole-body Total Molecular Volume (wb TMV) and Total Lesion PSMA (wb TL-PSMA). Clinical and imaging variables were correlated with overall survival (OS) and progression-free survival (PFS) using Cox regression models. Kaplan-Meier analyses were performed based on wb TMV and thresholds determined by the Youden index. Results: A PSMA PET/CT correlation for BCR was identified in 69% of patients, with comparable detection rates between tracers (Ga-PSMA 67% vs. F-PSMA 63%, p = 0.7). A higher wb TMV was significantly associated with worse OS (HR 2.20, p < 0.001) and PFS (HR 2.01, p < 0.001) in univariable analyses. In multivariable models, log₂(wb TMV) remained an independent prognostic factor for PFS (HR 1.78, p = 0.005). Patients with log₂(wb TMV) > 2.87 exhibited significantly poorer survival outcomes. A PSA at diagnosis > 17 ng/mL also predicted shorter PFS. Conclusions: Tumor segmentation from PSMA PET/CT imaging provides powerful prognostic information in patients with biochemical recurrence of prostate cancer, independently of the tracer used. The wb TMV represents a promising volumetric biomarker for future risk stratification and therapeutic decision-making, particularly in earlier stages of prostate cancer progression, where predictive imaging biomarkers remain largely undefined.