Nomograms that incorporate PSMA PET imaging may offer improved long-term risk stratification over established clinical risk scores for patients with prostate cancer (PCa).
In an international retrospective study, recently reported in Lancet Oncology, researchers reviewed PSMA PET data from 11,154 patients to develop and subsequently compare the aforementioned nomograms to clinical risk scoring models from the European Association of Urology (EAU), the National Comprehensive Cancer Network (NCCN) and others. Both nomograms incorporated disease group, distant metastasis detection, total tumor load and PSMA expression score, according to the study.
The study authors found that the visual PSMA PET nomogram achieved an 83 percent C-index and the quantitative PSMA PET nomogram had an 84 percent C-index for long-term risk stratification in patients with PCa.
In comparison to the EAU risk scoring model, the visual PSMA PET nomogram and quantitative nomogram provided 12 percent (51 percent vs. 63 percent) and 14 percent higher AUCs (51 percent vs. 65 percent), respectively, for risk stratification in patients with biochemical recurrence of PCa.
For patients with metastatic hormone-sensitive PCa, the authors found that the PSMA PET (PPP3) nomograms offered 15 and 14 percent AUCs than the GAFITA risk score (developed by Gafita and colleagues in 2021). The researchers also determined that both PSMA PET nomograms had a seven percent higher AUC (84 percent vs. 77 percent) than a clinical risk model from the National Comprehensive Cancer Network (NCCN) for risk stratification across all PCa groups.
“Current risk models rely on conventional imaging. Risk assessment needs to be urgently updated as entry of PSMA-PET findings into STARCAP or NCCN is no longer applicable. Head-to-head comparison of clinical risk scores showed superior prognostic accuracy of PPP3 nomograms compared with EAU risk scores for patients with biochemically recurrent prostate cancer, the GAFITA nomogram for patients with metastatic hormone-sensitive prostate cancer, and the NCCN risk score for all disease groups pooled,” maintained lead study author Madeleine J. Karpinski, MSc, who is affiliated with the Department of Nuclear Medicine at the German Cancer Consortium and the National Center for Tumor Diseases at University Hospital Essen in Essen, Germany, and colleagues.
Three Key Takeaways
• PSMA PET–based (PPP3) nomograms outperform traditional risk scores. Both visual and quantitative PSMA PET nomograms demonstrated higher prognostic accuracy (C-index ~83–84 percent) and significantly improved AUCs versus EAU and other clinical models, particularly in biochemical recurrence.
• Consistent performance across disease states. The PPP3 nomograms provided superior or comparable risk stratification across multiple prostate cancer settings (biochemical recurrence, metastatic hormone-sensitive, and pooled populations), and performed similarly to leading models (e.g., STARCAP, GAFITA) in initial staging and mCRPC.
• Potential to unify and modernize risk stratification. Incorporating PSMA PET metrics (tumor burden, metastases, PSMA expression) enables a single framework applicable across all PCa stages, supporting longitudinal risk tracking and highlighting the need to update conventional imaging–based scoring systems.
The researchers also found that the PPP3 nomograms achieved nearly equivalent AUCs to the STARCAP risk score for initial staging of PCa (65 percent and 63 percent vs. 62 percent) and the aforementioned GAFITA risk score for those with metastatic castration-resistant prostate cancer (mCRPC) (70 percent and 71 percent. vs. 72 percent).
“Superior or equal performance of PPP3 nomograms compared with established clinical risk scores is important for clinical implementation. In contrast to established clinical risk scores, PPP3 provides prognostication for all disease groups, which will simplify clinical reporting and allow for risk tracking in individual patients over time,” added Karpinski and colleagues.
(Editor’s note: For related content, see “The Reading Room Podcast: Current Insights and Emerging Trends in Molecular Imaging and Theranostics for Prostate Cancer,” “What New Research Reveals About 18F-Flotufolastat PET/CT for Detecting Biochemical Recurrence and Bone Metastasis in PCa” and “Comparative PET Imaging Trial Shows Increased Detection of PCa Recurrence with 64Cu-SAR-bisPSMA.”)
In regard to study limitations, the authors acknowledged some missing clinical data for determination of Gleason grade group and TNM staging as well as a lack of standardized timing for PET imaging and pharmaceutical type. They also conceded that their study findings need to be externally validated in a prospective trial.
