TOPLINE

An AI-powered biomarker based on histology alone predicted survival among men with metastatic hormone-sensitive prostate cancer and identified those who benefitted from escalation to triplet therapy.

METHODOLOGY

• Clinical trials have shown that treatment intensification — adding androgen receptor pathway inhibitors and chemotherapeutics such as docetaxel to standard androgen deprivation therapy (ADT) — can improve survival outcomes for patients with metastatic prostate cancer. But it’s unclear which patients stand to benefit most from escalation to triplet therapy.
• Researchers developed a prognostic biomarker using a computational histology AI platform, which extracted quantitative histomorphologic features from whole-slide images of diagnostic specimens from patients with metastatic hormone-sensitive prostate cancer.
• The tool was developed using data from 507 patients who participated in the phase 3 CHAARTED trial and validated in 684 participants from the phase 3 ENZAMET trial. Development involved constructing a signature of features associated with overall survival in CHAARTED, with a continuous risk score dichotomized into favorable-risk (90%) and unfavorable-risk (10%) groups.
• Performance was assessed in ENZAMET using multivariable Cox proportional hazards models adjusting for treatment, age, performance status, Gleason score, PSA levels, volume of disease, and timing of metastases.

TAKEAWAY

• In the validation cohort, 5-year overall survival rates were 39% vs 69% among patients with unfavorable-risk scores vs those with favorable-risk scores.
• After adjustment for clinical variables, patients with unfavorable-risk scores had worse overall survival (hazard ratio [HR], 2.34; P < .001) and worse progression-free survival (HR, 2.17; P < .001) compared with those with favorable-risk scores.
• Among patients with unfavorable risk, those treated with triplet therapy (ADT, androgen receptor pathway inhibitor, and docetaxel) had improved overall survival (HR, 0.45; P = .047) and progression-free survival (HR, 0.42; P = .012) compared with those receiving doublet therapy (no docetaxel). Patients with favorable risk had similar outcomes with either therapy.
• The interaction between the biomarker and treatment intensification was significant for overall survival (P = .003) and progression-free survival (P < .001), suggesting the biomarker predicted benefit with triplet over doublet therapy.

IN PRACTICE

“This biomarker was developed specifically in the metastatic setting, incorporated histology alone, remained independently prognostic beyond conventional clinicopathologic risk factors, and was associated with benefit from treatment intensification for high-risk disease,” the study authors wrote. “This tool has the potential to advance prognostication in [metastatic prostate cancer].”

SOURCE

The study, led by Christopher J. Sweeney, MBBS, Adelaide University, Adelaide, Australia, was published online in JCO Precision Oncology.

LIMITATIONS 

Use of androgen receptor pathway inhibitors in the validation cohort but not in the development cohort induced a calibration discrepancy with shifted absolute event rates. Black patients, who are more likely to die from prostate cancer, were underrepresented in ENZAMET. Because docetaxel treatment was not randomized, the biomarker-treatment interaction findings represent a post hoc secondary analysis of heterogeneity of treatment effect, warranting future study in randomized cohorts of doublet and triplet therapy. 

DISCLOSURES

The study received support from Valar Labs, Inc., and several co-authors were employed by the company. Sweeney disclosed receiving research funding from Janssen Biotech, Inc.; Astellas Pharma Inc.; Bayer; and other companies. Additional disclosures are noted in the original article.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.