Nomogram-based risk classification for predicting response to metastasis-directed stereotactic body radiotherapy in PSMA PET-staged oligorecurrent prostate cancer (PORTAL): an international, retrospective cohort study

Abstract

Background Metastasis-directed therapy (MDT) can defer androgen deprivation therapy (ADT) in oligorecurrent prostate cancer, but outcomes vary and there is a paucity of risk stratification tools. We aimed to develop and validate a nomogram to predict ADT-free survival. Methods In this international, retrospective cohort study across ten hospitals and academic institutions in Europe (Austria, Italy, Latvia, the Netherlands, Poland, and Portugal), we included men aged 18 years and older with histologically confirmed prostate cancer (all subtypes) previously treated with curative-intent local therapy, who developed metachronous oligorecurrent, hormone-sensitive disease (≤5 prostate-specific membrane antigen [PSMA] PET-detected pelvic nodal or distant metastases or both) and received stereotactic body radiotherapy to all lesions without adjuvant systemic therapy. The primary endpoint was the proportion of patients remaining ADT-free at 1 year of follow-up. Ten clinical predictors were evaluated in a multivariable Cox model with backward elimination. Internal validation with bootstrapping assessed model performance (C-index and calibration), followed by external validation. Findings A total of 1461 men treated between July 1, 2014, and Dec 31, 2023, were screened, of whom 586 were included in the development cohort (307 [52%] pelvic nodal and 279 [48%] distant metastases); median follow-up was 37 months (IQR 24–58). The external validation cohort (henceforth validation cohort) comprised 131 patients (57 [44%] pelvic nodal, 74 [56%] distant), with a median follow-up 43 months (24–57). ADT-free survival at 1 year was 84·3% (95% CI 81·4–87·3) in the development cohort and 92·8% (88·4–97·4) in the validation cohort. Independent predictors of earlier ADT initiation were higher pre-MDT PSA (hazard ratio 1·05 [95% CI 1·03–1·08]), shorter PSA doubling time (0·97 [0·95–0·98]), three to five lesions (HR 1·74 [1·33–2·28]), and distant metastases (HR 1·45 [1·18–1·78]). Model discrimination was 0·66 (95% CI 0·65–0·68) in the development cohort and 0·65 (95% CI 0·55–0·75) in the external validation cohort. Risk stratification separated patients into three prognostic groups—low risk, intermediate risk, and high risk (p<0·0001). Interpretation This nomogram predicts ADT-free survival after MDT in oligorecurrent prostate cancer. While individual-level discrimination was modest, risk-group stratification showed meaningful separation, supporting its potential use in treatment selection. Funding None.