A Strategy for defining biologically relevant levels of p53 protein expression in clinical samples with reference to...

Numerous studies have investigated p53 immunohistochemistry as a diagnostic, prognostic or predictive marker in various neoplasms. However, the literature is confused and contradictory paying little attention to pivotal aspects of p53 biology nor basic immunohistochemical principles. Here we highlight the effect of varying antibody concentration on p53 immunohistochemical expression. Six uterine endometrioid carcinomas, one uterine serous carcinoma (USC) and a proliferative endometrium were stained with antip53 antibody D07 at varying dilutions. Cases were scored on a scale of 0-6 depending on the percentage of positive nuclei. Greater than 95% of epithelial cells in the USC were positive at all dilutions. The proliferative endometrium exhibited positive staining of >95% of epithelial cells at low dilutions but at high dilutions most epithelial cells were negative. The proportion of positive tumor nuclei in the endometrioid carcinomas varied markedly with antibody concentration. We illustrate that the signal obtained from p53 immunohistochemistry is dependent on antibody concentration. This has diverse implications. First, the p53 labeling index is unreliable without attention to such issues. Second, comparisons between studies are not valid since different antibody concentrations (and other variables) are used. Thirdly, standardization between laboratories is necessary if clinical utility is to be attached to markers including p53. We propose a strategy for determining the optimal p53 antibody concentration.