Semiautomatically Quantified Tumor Volume Using ⁶⁸Ga-PSMA-11 PET as a Biomarker for Survival in Patients with...

Area of Science:

• Nuclear Medicine
• Oncology
• Medical Imaging

Background:

• Prostate-specific membrane antigen (PSMA)-targeting PET imaging is a standard for prostate cancer staging.
• Quantifying PSMA PET-derived whole-body tumor volume for survival prediction remains challenging.
• A validated software solution for PSMA PET biomarker quantification is needed.

Purpose of the Study:

• To develop and evaluate a software for semiautomated quantification of PSMA PET biomarkers.
• To assess the clinical utility of whole-body tumor volume (PSMA_TV50) as a prognostic biomarker.
• To investigate the correlation between semiautomated and manual quantification methods.

Main Methods:

• A research prototype software was developed for semiautomated PSMA PET quantification.
• Neural networks were trained for organ segmentation to exclude physiologic PSMA uptake.
• PSMA PET/CTs from 40 patients treated with 177Lu-PSMA-617 were analyzed for PSMA_TV50 and other biomarkers.

Main Results:

• Semiautomated quantification highly correlated with manual readings for PSMA_TV50 (R²=1.000) and SUVmax (R²=0.988).
• High interobserver agreement was achieved for the semiautomated workflow (PSMA_TV50 ICC=1.000, SUVmax ICC=0.997).
• PSMA_TV50 was a significant independent predictor of overall survival in patients receiving 177Lu-PSMA-617 therapy (HR=1.004, P=0.002).

Conclusions:

• The proposed software enables reproducible and easy quantification of PSMA PET biomarkers like PSMA_TV50.
• PSMA_TV50 is a promising and significant predictor of overall survival in advanced prostate cancer.
• This tool can enhance the clinical utility of PSMA PET imaging in treatment response assessment.