Prostate-specific membrane antigen (PSMA) PET imaging has become a key tool in the detection and staging of prostate cancer, offering improved diagnostic capabilities over traditional imaging methods. However, variability in how these scans are interpreted has posed challenges for consistent clinical decision-making. The introduction of the PSMA Reporting and Data System (PSMA-RADS) aimed to bring standardisation to image analysis, enhancing communication and improving diagnostic reliability. A multicentre prospective study has now compared the performance of PSMA-RADS version 1.0 and version 2.0 across different clinical contexts, demonstrating the benefits of the updated framework.
Diagnostic Accuracy Gains in Complex Cases
The study evaluated 443 men undergoing ⁶⁸Ga-PSMA-11 PET/CT imaging, divided into three groups: new diagnosis, biochemical recurrence (BCR) and follow-up. PSMA-RADS version 2.0 showed marked improvements in diagnostic performance, especially in the more complex BCR and follow-up cases. In the follow-up group, diagnostic accuracy rose from 88.7% with version 1.0 to 94.7% with version 2.0. Sensitivity in this group also improved, increasing from 87.7% to 95.7%. For BCR cases, accuracy advanced from 92.6% to 95.7%, while sensitivity reached 97.3% with version 2.0. In newly diagnosed patients, improvements were smaller but still notable. Specificity increased from 76.9% to 89.7%, and overall accuracy rose from 95.9% to 97.4%.
A key development in version 2.0 was the introduction of the PSMA-RADS 5T category, designed to capture treated metastases more accurately. This refinement helped to reduce ambiguity in cases where previously treated lesions remained metabolically active. As a result, 25.9% of BCR cases and 47.4% of follow-up cases were reclassified under 5T. These reclassifications led to improved alignment with prostate-specific antigen (PSA) dynamics and more accurate discrimination of treatment response, particularly in complex or previously treated cases.
Improved Consistency and Lesion Classification
Both versions of PSMA-RADS were associated with high interrater reliability, although version 2.0 offered improvements in specific areas. Agreement was strongest for prostate and bone assessments, where intraclass correlation coefficients (ICC) reached values between 0.85 and 0.92. These figures indicate consistent interpretations across radiologists, contributing to the overall robustness of the system. Soft tissue evaluation, however, remained a challenge. ICC values for soft tissue ranged from 0.36 to 0.50 in both versions, highlighting a need for further refinement in this area.
Must Read: 18F-PSMA and 18F-NaF PET/CT for Metastases Detection in Prostate Cancer
The introduction of the 5T category in version 2.0 proved particularly useful in improving lesion characterisation and reducing misclassification. In version 1.0, many treated lesions were grouped with active malignancies under the PSMA-RADS 5 category, leading to potential diagnostic uncertainty. By separating treated metastases, version 2.0 provided a clearer clinical picture and helped clinicians distinguish between treatment effects and active disease. These changes contributed to more nuanced assessments, particularly valuable in patient monitoring and treatment planning.
The study also explored the frequency distribution of PSMA-RADS categories. In version 1.0, category 5 dominated across all groups. With version 2.0, the introduction of 5T redistributed a significant number of cases, particularly in follow-up patients, where the proportion of category 5 findings dropped from 70.0% to 31.6%. This shift indicates greater precision in assigning findings to appropriate categories, enhancing clinical utility without compromising sensitivity.
PSA Correlation and Clinical Relevance
PSMA-RADS version 2.0 showed stronger correlations with PSA values and changes across all clinical groups. In newly diagnosed patients, baseline PSA correlation increased modestly, while in BCR and follow-up groups, changes in PSA (ΔPSA) showed much stronger associations with PSMA-RADS scores. The correlation coefficient for ΔPSA in the BCR group improved from 0.58 to 0.72, and in the follow-up group from 0.61 to 0.74. These correlations were statistically significant and remained robust after adjusting for age, tumour stage, Gleason score and treatment history.
The updated version also improved the alignment between PSA-based treatment response and imaging findings. Among follow-up patients, the proportion of cases showing concordance between a PSA reduction of 50% or more and PSMA-RADS response increased from 82.3% in version 1.0 to 88.7% in version 2.0. These improvements reinforce the clinical value of PSMA-RADS version 2.0 in monitoring therapy and evaluating disease progression.
The study’s high interrater and intrarater reliability across most anatomical regions further supports the use of PSMA-RADS in routine clinical practice. The consistent performance of version 2.0 across various subgroups and imaging challenges demonstrates its robustness. However, areas such as soft tissue assessment and lower specificity in BCR cases remain limitations that require further enhancement in future iterations.
PSMA-RADS version 2.0 represents a meaningful advancement in standardised PSMA PET/CT reporting for prostate cancer. It delivers improved diagnostic accuracy, especially in challenging BCR and follow-up settings, and offers clearer lesion characterisation through the introduction of the 5T category. Enhanced alignment with PSA dynamics and robust interrater reliability further establish its clinical relevance. Nonetheless, limitations remain in the interpretation of soft tissue findings and in specificity for recurrent disease. Future refinements should focus on these areas, potentially incorporating quantitative metrics and advanced technologies to further strengthen diagnostic confidence and clinical decision-making.
Source: Radiology: Imaging Cancer
Image Credit: iStock
