Abstract 5406: A deep learning enabled digital pathology platform for characterization of cancer-associated thrombosis of ovarian cancer

Abstract

Abstract Purpose: Cancer-associated thrombosis (CAT), characterized by activation, extravasation, and aggregation of platelets in tumor microenvironment, is a common complication observed in ovarian cancer patients. A third of ovarian cancer patients have elevated platelet counts that are associated with poor prognosis. However, the exact mechanism by which platelets preferentially extravasate into tumor microenvironment is unknown, partly owing to lack of tools that can accurately segment and measure platelet counts found in tumor tissues. In this study, we hypothesize that tumor secreted SDF-1 interacts with CXCR4 receptors on platelets and initiates platelet chemotaxis towards tumor microenvironment. We developed a digital pathology (DP) platform with automated whole tissue confocal imaging and deep learning (DL) model that can be used to automatically segment out platelets and calculate platelet density with high precision and consistency in the mouse ovarian cancer tissue sections for supporting hypothesis evaluation. Methods: A DL algorithm to accurately segment out platelets was developed based on the U-Net structure using our original datasets from the murine ovarian cancer model. Tumor-bearing mice were injected with 5 mg/kg of Plerixafor, a CXCR4 inhibitor, daily for 4 weeks. Then, tumor nodules were resected, stained with CD41a antibody, and imaged using the whole slide confocal microscopy. Using these high-resolution images, we performed manual annotation of platelets to generate the mask dataset and performed further data augmentation to increase the diversity of datasets. We trained our model using 85% of these images and performed validation using the remaining 15%. Results: Using our DLDP platform, the validation dataset achieved pixel-wise accuracy of 98.7%. A receiver operating characteristic (ROC) curve demonstrated an area under the ROC curve (AUC) of 0.99 with high sensitivity and specificity. Intersection over union (IoU) score of 0.724 was obtained. A statistically significant (p=0.0467), 7.03% reduction in the platelet density was shown in the Plerixafor-treated mouse ovarian tumor tissues (n=8) compared to the control tumor tissues (n=9). Note that the traditional threshold method was not able to find the statistical significance (p=0.0966) in the same dataset. Conclusions: We developed a DLDP platform to segment and quantitate tens of thousands of platelets from whole slide mouse ovarian cancer tissues automatically with high accuracy and consistency. Applying the DLDP platform, we demonstrated that treatment with Plerixafor leads to overall reduction in platelet infiltration within ovarian murine tumors. With high-throughput, accurate, and automated detection of platelets, our DLDP platform can serve as a powerful tool for characterization of CAT by evaluation of platelet density as a potential prognostic biomarker for ovarian cancer. Citation Format: Ju Young Ahn, Wendolyn Carlos-Alcalde, Matthew Vasquez, Min Soon Cho, Vahid Afshar-Kharghan, Stephen T. Wong. A deep learning enabled digital pathology platform for characterization of cancer-associated thrombosis of ovarian cancer. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5406.