Digital pathology and image analysis of p53 biomarker in lymphomas using two algorithms: correlation with genotype and visual inspection

p53 immunohistochemistry (IHC) is widely used as a rapid surrogate for detecting <i>TP53</i> mutations, with <i>TP53</i> mutations being a key biomarker for poor outcomes in lymphomas. We developed two algorithms using digital quantification tools to assess p53 expression from whole slide images of 77 lymphoma samples. An experienced pathologist visually evaluated the p53 slides, classifying cases as likely wild-type or mutated <i>TP53</i> genotype. We correlated the results of the algorithms and visual inspection with the actual <i>TP53</i> genotype. For cases with p53 overexpression (likely missense mutations), the algorithms achieved 86.7% sensitivity and 98.2% specificity (visual inspection: 80% and 95.2%). For cases with reduced p53 expression (likely 'other' mutations), the algorithms showed 92.7% sensitivity and 100% specificity (visual inspection: 40% and 95.8%). This study demonstrates that combining digital pathology with digital quantification tools-based algorithms can reliably predict <i>TP53</i> genotype from p53 IHC patterns, with comparable or slightly superior performance to an experienced pathologist.