Validation of p53 Immunohistochemistry (PAb240 Clone) in Canine Tumors with Next-Generation Sequencing (NGS) Analysis

SIMPLE SUMMARY: TP53 is a tumor-suppressor gene that is involved in cell cycle arrest, and its mutation is an event that is recognized to cause and contribute to carcinogenesis. Increased immunohistochemical expression of p53 in tumors has been associated with mutation of the gene; therefore, this can be an important tool to detect p53 anomalies. In this study, the expression of p53 was compared to NGS analysis in order to verify if p53 could predict TP53 gene mutation. The results indicate that with a 10% threshold, the sensitivity of immunohistochemistry was 60%, the specificity was 86.7%, and the accuracy was 76%. This suggests that we can expect up to 25% inaccurate predictions. Moreover, a substantial decrease in sensitivity was observed, moving the threshold to 50%. Overall, these results suggest that IHC can predict mutation and that the 10% threshold can be considered appropriate. ABSTRACT: In human medicine, p53 immunohistochemistry (IHC) is a common method that is used for the identification of tumors with TP53 mutations. In veterinary medicine, several studies have performed IHC for p53 in canine tumors, but it is not known how well it actually predicts the mutation. The aim of this study was to estimate the accuracy of the IHC method for p53 (clone PAb240) using a lab-developed NGS panel to analyze TP53 mutations in a subset of malignant tumors in dogs. A total of 176 tumors were analyzed with IHC and then 41 were subjected to NGS analysis; among them, 15 were IHC positive and 26 were negative, and 16 out of 41 (39%) were found to be inadequate for NGS analysis. Excluding the non-evaluable cases at NGS, of the remaining eight IHC-positive cases, six were mutants and two were wild-type. Among the 17 IHC-negative cases, 13 were wild type, and 4 were mutants. The sensitivity was 60%, specificity was 86.7%, and the accuracy was 76%. These results suggest that when using IHC for p53 with this specific antibody to predict mutation, up to 25% wrong predictions can be expected.