How a Post-RP MMAI Model Performed in a Long-Term Survivor Population

An MMAI prognostic model demonstrated meaningful risk stratification in a Mayo Clinic long-term survivor cohort, with findings that extend the model's validated performance horizon well beyond what most existing prognostic tools have been formally tested against. In the third segment of 5-part series, R. Jeffrey Karnes, MD, the Dr. Anson Clark Professor of Urology at Mayo Clinic in Rochester, Minnesota, presents the headline results: Among men classified as MMAI High, the probability of distant metastasis at 10 years approached 15%, compared with less than 5% in the MMAI Low group—representing an approximately fivefold difference in risk. In multivariable analysis, MMAI carried strong hazard ratios for distant metastasis and, importantly, the CAPRA-S score was no longer an independent predictor once MMAI was included in the model.

Karnes explains that the ability of MMAI to add prognostic value above and beyond CAPRA-S is particularly significant given that CAPRA-S is freely calculable from routinely available clinical variables. That the digital pathology component of the MMAI model—its analysis of hematoxylin and eosin–stained slide morphology—contributed incremental information beyond what clinical variables alone could capture suggests the model is detecting biologic features not fully reflected in standard pathologic grading. In the subgroup of patients who developed biochemical recurrence, MMAI retained its prognostic significance: Men with a high score who experienced recurrence had a metastasis risk of approximately 20% on follow-up, compared with less than 10% for those with a low score. Karnes further notes that prior studies incorporating MMAI into banked tissue from the NRG 9601 and 0534 salvage radiation trials showed a benefit from salvage radiotherapy in patients with high scores, with a possible additional signal for concurrent hormonal therapy.

The model also maintained its prognostic performance at 15 years of follow-up, a time horizon at which events such as metastasis and prostate cancer–specific death continue to occur, as demonstrated by the long-term data from the ProtecT trial. Karnes emphasizes that this durability is clinically meaningful, because patients undergoing RP today are generally healthy men with normal life expectancies who warrant sustained oncologic surveillance. Even in the approximately 75 patients with detectable PSA immediately after surgery—a population that often creates acute management uncertainty—the MMAI model maintained its ability to discriminate between those at higher and lower risk for future metastasis, supporting its utility across the full spectrum of post-RP clinical presentations.