Artificial Intelligence (AI)-Driven Screening of Equivocal Prostate Immunohistochemistry (IHC) Cases: Development and Validation of a Screening and Cancer Detection Framework.

Abstract

Artificial intelligence (AI) has been proposed as a solution to meet increasing demand for the diagnostic services of pathologists (B.I., C.F., V.M., B.G.N., X.J.Y.). Prostate biopsies are a significant source of this demand, and a substantial fraction of these biopsies require immunohistochemistry (IHC) staining, which adds work, time, and cost to the diagnostic process. Equivocal cases, which often prompt the use of IHC, not only present the greatest challenge to AI cancer detection tools but also represent the area where properly trained systems could offer the most clinical value by reducing the need for ancillary testing for cancer confirmation. From August 2021 to April 2023, we scanned 25,570 slides from 1641 patients. We investigated the performance of institutionally developed models for prostate cancer detection using digital pathology images, with an aim for reducing work, turnaround time, and costs related to the ordering of IHC equivocal cases. We advanced complementary sensitive and specific models to screen these challenging cases, aiming to identify slides that could be confidently diagnosed without any ancillary IHC tests. Additionally, we compared the performance of a prostate-specific model to a general-purpose foundation model for screening and cancer detection. Our screening models correctly classified 55% of challenging equivocal blocks where IHC was ordered with a 1.4% error rate. We found that the foundation model achieved higher screening rates (ie, percentage of cases where IHC could be avoided), but this came at the cost of uniformly higher error rates and significantly greater computational demands. When trained as a standalone prostate cancer detection system, our model demonstrated high concordance with pathologist ground truth, achieving an area under the curve of 98.5%, sensitivity of 95.0%, and specificity of 97.8%. Computational models can aid in the diagnosis of prostate cancer and can effectively screen challenging prostate biopsy cases, reducing unnecessary IHC utilization and helping to optimize pathology workflows. Plain language summary Traditional pathologic diagnosis of prostate cancer can be labor intensive. Equivocal cases in particular often require special testing (immunohistochemistry [IHC]) to establish a diagnosis, which introduces further delay and cost. This study develops an artificial intelligence system specifically designed to screen equivocal cases and reduce the need for IHC. Our model serves as a second-read tool to help optimize pathology workflow and reduce turnaround time and costs by flagging cases where IHC can be safely avoided.