Detection of Prostate Cancer in 3-Dimensional Pathology Datasets via Generative Immunolabeling

Recent advancements in nondestructive 3-dimensional (3D) pathology offer a complement to standard histology by enabling comprehensive volumetric analyses of intact clinical specimens (eg, biopsies). Previous studies have demonstrated the added prognostic value of 3D pathology for prostate cancer (PCa) risk stratification by correlating 3D microarchitectural features with long-term patient outcomes. However, these analyses relied on coarse manual annotations of cancer-enriched regions for downstream analysis without fine-grained delineation between often-intermixed cancerous and benign glands. To automate and improve the process of delineating between benign and PCa-enriched regions in 3D pathology data sets, we have developed a 3D computational pipeline: Synthetic Immunolabeling for Generative Heatmaps of Tumor (SIGHT). SIGHT relies on deep learning-based 3D image translation models, trained in a fully supervised manner, to convert hematoxylin and eosin analog 3D pathology data sets into multiplexed 3D immunofluorescence data sets that facilitate tumor detection. Our implementation of SIGHT synthetically labels 2 cytokeratin markers that are differentially expressed in cancerous and benign prostate glands, which are used to generate explainable 3D heatmaps of cancer-enriched regions in prostate tissues. Validation of SIGHT against ground truth annotations from a panel of genitourinary pathologists (X.F., M.D., L.D.T.) yields an average F1 score of 0.88, which is comparable with the average interpathologist agreement F1 score of 0.90. To demonstrate how SIGHT can automate and improve a pipeline for risk stratification based on 3D pathology data sets of PCas, we developed preliminary machine classifiers of recurrence risk based on 3D glandular histomorphometric features from 75 patients. Volumetric glandular analysis in SIGHT-identified cancer-enriched regions versus all tissue regions yields a Kaplan-Meier hazard ratio of 3.57 (CI, 1.6-7.9) versus 0.92 (CI, 0.45-1.89).