Evaluating the MMAI Biomarker in Patients With mHSPC

A study from Mark Markowski, MD, PhD, and colleagues presented at the 2024 American Society of Clinical Oncology Annual Meeting conducted a prognostic validation of the digital, pathology-based, multi-modal artificial intelligence (MMAI) biomarker in patients with metastatic hormone-sensitive prostate cancer (mHSPC).

Researchers noted that the ArteraAI MMAI prognostic biomarker performs risk stratification of patients with localized prostate cancer using histopathology images and clinical data. They evaluated data from a subset of patients in the randomized, phase 3 CHAARTED trial.

The MMAI model created scores and analyzed prognostic ability for overall survival (OS) in the Cox proportional hazard model. The model included 4 defined prognostic cohorts, including M0LV, M0HV, M1LV, and M1HV. Dr. Markowski and colleagues used multivariable analysis (MVA), including MMAI risk group and treatment to analyze the prognostic cohorts.

From the CHAARTED study, 456 (57.7%) patients were found to have evaluable digital histopathological images. Of those patients, 370 (81.1%) were MMAI-high and 86 (18.9%) were MMAI-intermediate/low risk. Among the participants, 57 were classified as M0LV, 29 as M0HV, 66 as M1LV, and 242 as M1HV, with an MMAI-high proportion of 56.1%, 69.0%, 86.4%, and 92.6%, respectively.

The 5-year OS was 39%, 58%, 83%, respectively (P<.001), in the MMAI high, intermediate, and low cohorts. Through MVA, investigators found that the MMAI-high risk cohort related to OS (95% CI, 1.10-2.84; P=.02) following adjustments for treatment arm, volume status, and stage at diagnosis. HV groups were also connected with OS after MVA.

“The ArteraAI MMAI model was found to be prognostic for OS among men with mHSPC in CHAARTED,” the researchers wrote. “This effect persisted when controlling for treatment, metastatic burden, and metastatic status at diagnosis. On MVA, M0HV and M1HV also maintained prognostic value. This analysis is exploratory in nature, to be followed by development of a model optimized for advanced prostate cancer.”