Nodal maximum standardised uptake value improves PSMA ‐ PET / CT ‐based nodal staging in intermediate‐ and high‐risk prostate cancer

Objectives To evaluate whether nodal maximum standardised uptake value (SUV max ) improves the positive predictive value (PPV) of prostate‐specific membrane antigen (PSMA) positron emission tomography (PET)/computerised tomography (CT) for detecting lymph node invasion (LNI) in prostate cancer (PCa), and to develop clinical decision tools to guide decision‐making for patients with low‐ and high‐volume nodal disease. Patients and Methods This international multicentre study included patients with histopathologically confirmed PCa who underwent preoperative PSMA‐PET and robot‐assisted radical prostatectomy with extended pelvic lymph node dissection (2016–2023). Sensitivity, specificity, PPV, and negative predictive value (NPV) for LNI detection were calculated, defining a positive test as non‐physiological PSMA uptake classified according to the Prostate Cancer Molecular Imaging Standardised Evaluation (PROMISE) molecular imaging Tumour‐Node‐Metastasis (miTNM) system. Receiver operating curves analysis identified optimal nodal SUV max cut‐offs for pelvic LNI. Logistic regression assessed predictors for pathological N1 stage (pN1) and high‐volume nodal disease (four or more positive nodes). Clinical decision tools were developed to stratify patients into three risk groups for pN1 and high‐volume nodal disease. Results A total of 521 patients were included, with a median (interquartile range) age of 66 (61–71) years and prostate‐specific antigen (PSA) level of 9.9 (6.6–17.0) ng/mL. The PSMA‐PET showed 45.0% sensitivity, 94.3% specificity, 64.7% PPV, and 88.3% NPV for LNI detection. Adding nodal SUV max ≥4.9 improved the PPV to 81.1% and sensitivity to 71.4% and, combined with a PSA level ≥10 ng/mL, magnetic resonance imaging (MRI) T‐stage ≥T3a, and miT‐stage ≥T3a, identified 95% of pN1 cases in the high‐risk group. High‐volume nodal disease was found in 3.8%, including 1.1% in miN0 patients. Nodal SUV max ≥7.2, MRI T‐stage ≥T3a, and miN2 (multiple suspicious LNs) were predictors for high‐volume nodal disease and, combined in a clinical decision tool, excluded all low‐risk patients for high‐volume nodal disease. Conclusion Nodal SUV max improves PPV of PSMA‐PET for nodal staging and aids in excluding high‐volume nodal disease. Clinical decision tools integrating nodal SUV max with relevant clinical parameters demonstrate potential to guide individualised nodal management. These findings warrant external validation in larger cohorts.