Abstract 5438: The combination of artificial intelligence and BLEACH&STAIN multiplex fluorescence immunohistochemistry facilitates automated prostate and breast cancer detection

Abstract Background: Automated prognosis marker assessment in prostate and breast cancer using immunohistochemistry is currently hampered by the lack of a reliable differentiation between benign and malignant glands. To evaluate the patient’s risk in routine clinical practice in prostate cancer prognosis parameters such as the Gleason grading, that are accompanied by a high interobserver variability are used. In breast cancer multi-gene panels are used that are influenced by fluctuating tumor purity. A reproducible prognostic evaluation is lacking in both tumor entities. Design: To enable automated prognosis marker quantification, we have developed and validated a framework for automated prostate and breast cancer detection that comprises three different artificial intelligence analysis steps and an algorithm for cell-distance analysis of multiplex fluorescence immunohistochemistry. Pan-cytokeratin (panCK) antibodies were used to detect epithelial cells and antibodies directed against Myosin and p63 were used to identify basal cells. Results: The optimal distance between Myosin+ and p63+ basal cells and benign panCK+ cells was identified as 25 µm in breast and 23 µm in prostate cancer and used to exclude benign glands from the analysis combined with several deep learning-based algorithms. Our framework discriminated normal glands from malignant glands with an AUC of 0.96 in breast and 0.98 % in prostate cancer. The approach for automated prostate and breast cancer detection, by excluding benign gland from the analysis, improved the predictive performance of prognosis markers significantly (p<0.001). To compare the multiplex fluorescence immunohistochemistry-based (mfIHC) automated prognosis marker assessment with “classical” bright field-based automated prostate cancer detection for prognosis marker assessment in a cohort of 30 biopsies from routine clinical practice prognosis markers were manually assessed and set as gold standard. An excellent agreement between the mfIHC-based automated cancer detection and the reference manual cancer identification was found (intraclass correlation [ICC]: 0.94 [95% CI 0.87 - 0.97]). Conclusion: Automated prostate and breast cancer identification enables fully automated prognosis marker assessment in routine clinical practice using deep learning and multiplex fluorescence immunohistochemistry. BLEACH&STAIN as well as other multiplex fluorescence immunohistochemistry approaches that enable the simultaneous analysis of 20+ biomarkers can be used to established prognosis panels that can be now assessed in an automated manner. Citation Format: Tim Mandelkow, Gisa Mehring, Elena Bady, Maximilian Lennartz, Christoph Fraune, Frank Jacobsen, Natalia Gorbokon, Till Krech, Eike Burandt, Anne Menz, Ria Uhlig, Till S. Clauditz, Guido Sauter, Markus Graefen, Sarah Minner, Niclas C. Blessin. The combination of artificial intelligence and BLEACH&STAIN multiplex fluorescence immunohistochemistry facilitates automated prostate and breast cancer detection. [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 5438.