Patterns of p53 mutations separate ovarian serous borderline tumors and low- and high-grade carcinomas and provide support for a new model of ovarian carcinogenesis -: A mutational analysis with immunohistochemical correlation

The infrequent association of serous borderline tumors (SBTs) with invasive serous carcinoma has led to the view that SBTs are unrelated to invasive serous carcinoma. Nonetheless, mortality associated with SBTs is generally attributed to malignant transformation, and traditionally these tumors have been designated as "carcinomas of low malignant potential." Previous immunohistochemical studies evaluating p53 expression and molecular genetic studies evaluating mutational status have reported that p53 overexpression and mutations are infrequent in SBTs and occur in as many as 50% to 80% of invasive serous carcinomas. The different methodologies for determining p53 status and the failure to correlate the findings with tumor grade make these studies difficult to interpret. The current study was undertaken to overcome these deficiencies and to reconcile the relationship of SBTs to invasive serous carcinoma by performing a morphologic, immunohistochemical, and molecular genetic analysis comparing SBTs with low- and high-grade serous carcinoma. The molecular genetic analysis used a highly stringent, carefully designed nucleotide-sequencing method. A total of 96 sporadic serous tumors including 25 SBTs (11 atypical proliferative serous tumors and 14 intraepithelial low-grade serous carcinomas [noninvasive micropapillary serous carcinomas, MPSCs]), 12 low-grade serous carcinomas (invasive MPSCs), and 59 high-grade serous carcinomas were analyzed for their p53 mutational status of exons 5 to 9. Functional mutations, defined as mutations resulting in the alteration of the structure of the encoded protein, were detected in 30 of 59 (50.8%) high-grade serous carcinomas and 1 (8.3%) of 12 low-grade invasive serous carcinomas compared with 2 (8%) of 25 SBTs, both of these in intraepithelial low-grade serous carcinomas (non-invasive MPSCs). The similar frequency of p53 mutations in SBTs and low-grade invasive serous carcinomas in contrast to the significantly higher frequency of p53 mutations in high-grade serous carcinomas (P < 0.0005) suggests a common lineage for SBTs and low-grade invasive serous carcinomas and supports the view that SBTs are unrelated to the usual type of invasive serous carcinoma, which is a high-grade neoplasm. Mutational status was also correlated with p53 immunoreactivity. Although p53 immunoreactivity is generally higher in those specimens containing mutant p53, immuno-staining is neither sufficiently specific nor sensitive enough to predict p53 mutations. The molecular genetic findings confirm our hypothesis of dual pathways of serous carcinogenesis based on previous analyses of KRAS and BRAF mutations on the same set of cases in which KRAS and BRAF Mutations were found in 60% of SBTs and low-grade serous carcinoma but not in high-grade serous carcinomas. Based on these studies, we have proposed a model of serous carcinogenesis in which SBTs are the precursors of low-grade serous carcinomas whereas the usual type of invasive serous carcinoma is a high-grade neoplasm that develops "de novo" from in situ alterations in epithelial inclusion cysts.