A new meta-analysis revealed significantly higher accuracy for prostate-specific membrane antigen (PSMA) positron emission tomography (PET) in detecting biochemical recurrence (BCR), lymph node metastasis and bone metastasis in comparison to multiparametric MRI (mpMRI) and computed tomography (CT).
For the meta-analysis, recently published in Academic Radiology, researchers reviewed data from 67 studies and a total cohort of 7,637 patients. The meta-analysis authors also assessed the use of artificial intelligence (AI) for prostate cancer (PCa) BCR detection based on 21 studies.
The researchers found that PSMA PET provided the highest accuracy in PCa BCR detection (89 percent) in contrast to AI (76 percent), mpMRI (71 percent) and CT (45 percent).
For lymph node metastasis detection, PSMA PET provided 89 percent accuracy in comparison to 79 percent for mpMRI and 72 percent for CT. The meta-analysis authors also pointed out enhanced accuracy in detection of bone metastasis for PSMA PET (95 percent) versus mpMRI (85 percent), CT (81 percent) and bone scan (80 percent).
“This meta-analysis underscores the superior diagnostic accuracy of PSMA PET compared to (conventional imaging modalities), particularly in detecting BCR of PCa. PSMA PET consistently demonstrates higher diagnostic accuracy in identifying metastasis in lymph nodes, bones, and visceral organs,” wrote study co-author Qun Tang, M.D., who is affiliated with the Medical School at the Hunan University of Chinese Medicine in Hunan, China, and colleagues.
While the researchers emphasized the strength of PSMA PET for metastatic PCa detection, they noted that mpMRI provided enhanced detection of local PCa recurrence at 93 percent in comparison to PSMA PET (84 percent) and CT (77 percent).
“mpMRI’s high soft-tissue resolution allows it to distinguish between residual prostate tissue, scar tissue, and recurrent tumors, making it particularly valuable for detecting local recurrence after (radical prostatectomy),” added Tang and colleagues.
Three Key Takeaways
- Superior diagnostic accuracy with PSMA PET. The meta-analysis authors determined that PSMA PET offers enhanced accuracy compared to mpMRI and CT for detecting biochemical recurrence (BCR), lymph node metastasis, and bone metastasis in prostate cancer.
- mpMRI remains the best tool for local recurrence detection. After prostatectomy, mpMRI provides the best detection of local PCa recurrence due to its superior soft-tissue resolution, outperforming PSMA PET and CT in this setting.
- More research needed for clinical application of AI in prostate cancer imaging. AI demonstrates moderate accuracy for BCR detection but requires further refinement and the development of explainable models to improve clinical adoption and reliability.
For the assessment of AI for PCa BCR detection, the meta-analysis authors found pooled sensitivity, specificity and AUC of 77 percent, 76 percent and 79 percent respectively. Noting wide variation in cohort size with the reviewed studies assessing AI, the researchers emphasized continued research to enhance the clinical utility of AI.
“ … The development of explainable AI models is crucial for their clinical deployment, as transparent decision-making processes will enhance trust and adoption in healthcare settings,” maintained Tang and colleagues.
(Editor’s note: For related content, see “PSMA PET/CT Facilitates Better Long-Term Survival Rates After Salvage Radiotherapy for Recurrent PCa,” “The Reading Room Podcast: Emerging Trends with Theranostics in Prostate Cancer, Part 2” and “Study Emphasizes PSA and PSMA PET Tumor Volume Assessment for Predicting mHSPC Progression After Apalutamide and ADT.”)
In addition to sample size variation with the AI studies, the authors conceded a predominance of retrospective studies overall in the reviewed research.