Predictive Model Estimates Dosimetry for PSMA Therapy

A proof-of-concept machine-learning model used pre-therapy 18F-PSMA PET/CT to predict absorbed radiation dose to tumors and organs, News-Medical reports. The study enrolled nine patients with metastatic castration-resistant prostate cancer, contributing 57 tumors, 36 salivary glands, and 18 kidneys for analysis, News-Medical reports. The work was presented at the Society of Nuclear Medicine and Molecular Imaging 2026 Annual Meeting, News-Medical reports.

According to News-Medical, researchers developed a mixed-effects machine learning model that combined uptake-based PET metrics, radiomic features, and clinical biomarkers to predict post-therapy dosimetry. Predictions were compared with dosimetry calculated after one cycle of 177Lu-PSMA radiopharmaceutical therapy to assess accuracy, per News-Medical. The article quotes Amit Nautiyal, PhD, saying, "18 F-PSMA PET/CT is already routinely performed and widely available in prostate cancer patients, but its potential to predict treatment radiation dose has not previously been explored," News-Medical reports.

Early-stage models that link pre-therapy imaging to dosimetry can reduce reliance on post-therapy imaging workflows and speed decision cycles for radiopharmaceuticals. Observed patterns in similar translational studies show that combining classical uptake metrics with radiomics and patient-level random effects often improves sample-efficient prediction while raising the need for external validation.

For practitioners, a validated pre-therapy dosimetry predictor would simplify patient selection and toxicity risk assessment for PSMA-targeted radioligand therapy across centers that already perform 18F-PSMA PET/CT. Observers should note the study's small sample size and single-cycle comparison when weighing generalizability.

Look for larger, multi-center validation cohorts, prospective comparisons across imaging protocols, and tests of model calibration against multi-cycle dosimetry. News outlets and conference follow-ups that report validation results will be key indicators of clinical readiness.