Abstract 7382: Artificial intelligence derived features of collagen architecture identify high risk HPV associated oropharyngeal squamous cell carcinoma patients

Abstract Introduction: Human papilloma virus (HPV) induced oropharyngeal squamous cell carcinoma (OPSCC) is a common subtype of head and neck carcinoma. Dysregulation of the extracellular matrix (ECM), an important part of the tumor microenvironment (TME), is shown in HPV related cancers. We studied collagen fibers (CF), the amplest component of the ECM, and evaluated whether their characteristics within the TME can offer a prognostic value in OPSCC. Methods: Whole slide images (WSIs) of OPSCC patients were obtained from Washington University (WU, n=107), Houston Veterans Administration Hospital (HV, n=94), Cleveland Clinic (CC, n=336), John Hopkins University (JH, n=121) and Vanderbilt Medical Centre (VM, n=158). For the identified CF in tumor-stroma regions of WSIs, the extracted CF features included: CF fragmentation measure, CF bundling percentage, CF rigidity measure, CF anisotropy index, and CF density index. For survival analysis using overall survival (OS) and Disease-Free Survival (DFS) as endpoints, the median risk score in the training set (WU) was applied for risk stratification in the validation sets (HV, CC, JH, VM) by means of a Least Absolute Shrinkage and Selection Operator-Cox regression model. Multivariable Cox Proportional Hazards Regression was done for both OS and DFS. Results: The patients defined as “high risk” had worse OS than those identified as “low risk” (HV (p=0. 03), CC (p=0. 03), JH (p=0. 02), VM (p=0. 04) ). Multivariable analysis showed that CF features were prognostic independent of T/N stages, age, sex, and smoking status with HR=1. 99 (95% CI=1. 46-2. 72, p=0. 00001) for OS and HR=1. 73 (95% CI=1. 21-247, p=0. 002) for DFS (Table 1). Table 1: Multivariable Survival Analysis for OS and DFS, HR=Hazard Ratio, CI=Confidence Interval Variable Overall Survival Disease Free Survival HR (95% CI) P value HR (95% CI) P value Age (65 years vs. =65 years) 0. 99 (0. 93-1. 06) 0. 78444 0. 96 (0. 82-1. 12) 0. 62864 Sex (Female vs. Male) 0. 56 (0. 29-1. 10) 0. 09042 0. 50 (0. 22-1. 13) 0. 09385 Smoking status (Yes vs. No) 1. 85 (1. 31-2. 62) 0. 00051 1. 03 (0. 72-1. 47) 0. 86871 T-stage (T3/4 vs T1/2) 1. 86 (1. 38-2. 50) 0. 00004 1. 91 (1. 36-2. 69) 0. 00018 N-stage (N+ vs N0) 1. 76 (1. 31-2. 37) 0. 00016 2. 04 (1. 46-2. 86) 0. 00003 Collagen fiber features (High risk vs low risk) 1. 99 (1. 46-2. 72) 0. 00001 1. 73 (1. 21-2. 47) 0. 00286 Conclusion: Artificial intelligence derived features of CF architecture can help identify high vs low risk patients in HPV related OPSCC. These findings deserve further confirmation in larger prospectively collected OPSCC data sets and could impact future treatment paradigms. Citation Format: Reetoja Nag, Haojia Li, Germán Corredor, Pingfu Fu, Nabil F. Saba, Mehmet A. Bilen, Tilak Pathak, Paula Toro, Mojgan Mokhtari, Krunal Pandav, Michael Gilkey, Shlomo Koyfman, Deborah Chute, Wade L. Thorstad, Mitra Mehrad, Patricia D. Castro, Andrew G. Sikora, Rebecca D. Chernock, Jingqin Luo, Mitchell Frederick, Vlad Sandulache, David J. Adelstein, James Lewis Jr. , Anant Madabhushi. Artificial intelligence derived features of collagen architecture identify high risk HPV associated oropharyngeal squamous cell carcinoma patients abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts) ; 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84 (6Suppl): Abstract nr 7382.