Development and validation of a digital pathology artificial intelligence (DPAI)-based biomarker predicting risk of Gleason grade group reclassification for patients who are candidates for active surveillance

Aims

Active surveillance (AS) allows selected men with localized prostate cancer to defer curative therapy and reduce treatment morbidity. Conversion from AS to treatment is commonly triggered by Gleason grade group (GGG) upgrading on confirmatory biopsy. We developed and validated a digital pathology artificial intelligence (DPAI) biomarker to predict GGG upgrading in AS-eligible patients.

Materials & Methods

The DPAI model was trained using histopathology image features from diagnostic biopsies of 998 patients and validated in an independent cohort of 296 patients meeting criteria for AS. Logistic regression estimated the probability of confirmatory-biopsy GGG increase, and feature selection identified the most predictive variables.

Results

AI-GUR (Artificial Intelligence-Gleason Upgrade Risk) predicted GGG reclassification at confirmatory biopsy (OR 1.60; p=0.0003), and provided information beyond conventional stratification (risk group, CAPRA) and cribriform morphology (all p<0.01). Predicted risks were similar across time from diagnosis (∼10-15% to ∼85% at 1, 1.5, or 2 years; p for time=0.50), consistent with initial biopsy mischaracterization rather than time-dependent progression.

Conclusions

AI-GUR provides individualized estimates of confirmatory-biopsy GGG upgrading for AS candidates. Using DPAI may improve shared decision-making by complementing standard clinicopathologic tools and molecular testing using the same biopsy specimen, while informing the likelihood of grade upgrade at confirmation.

Summary Points

Active surveillance (AS) can reduce treatment-related morbidity in localized prostate cancer (PCa), and confirmatory biopsy is used to address uncertainty related to grade reclassification and potential mischaracterization at diagnostic biopsy.

We developed and validated AI-GUR (Artificial Intelligence–Gleason Upgrade Risk), a digital pathology AI (DPAI) biomarker to predict Gleason grade group (GGG) reclassification at confirmatory biopsy in AS-eligible patients.

The model was trained in 998 patients and validated in an independent cohort of 296 patients, all candidates for AS by guideline criteria.

In validation, AI-GUR significantly predicted GGG reclassification (OR 1.60, p=0.0003) and added information beyond conventional risk stratification tools (risk group, CAPRA) and cribriform morphology (p-values <0.01).

Predicted risk was invariant with time since diagnosis (∼10–85% at 1, 1.5, or 2 years; p-value for time=0.50), consistent with capturing mischaracterization at initial biopsy rather than time-dependent progression; AI-GUR provides individualized risk estimates that may improve shared decision-making for AS candidates using the same biopsy material used for conventional stratification and advanced molecular prognostic testing, with a limitation of archived material.