作者
M. Herman Chui,Deyin Xing,Felix Zeppernick,Zoe Q. Wang,Charlotte Gerd Hannibal,Kirsten Frederiksen,Susanne K. Kjær,Leslie Cope,Robert J. Kurman,Ie Ming Shih,Tian Li Wang,Russell Vang
摘要
Although risk factors have been established for the development of serous carcinoma after a diagnosis of serous borderline tumor (SBT), comprising atypical proliferative serous tumor (APST) (ie, conventional SBT) and noninvasive low-grade serous carcinoma (niLGSC) (ie, micropapillary SBT), subsequent invasive carcinoma still occurs in a subset of women who are not at increased risk. Whether subsequent serous carcinoma in women with a prior SBT represents malignant progression/recurrence or an independent primary tumor is unclear, and the combined clinicopathologic and molecular features of SBTs and their subsequent carcinomas have not been fully characterized. In this study, we analyzed a cohort of 42 women initially diagnosed with SBT who subsequently developed serous carcinoma of a total of 1025 cases of ovarian SBT from a nationwide population-based cohort. Review of the diagnostic slides was performed from this subset of SBTs and matched metachronous invasive serous carcinomas (39 low grade, 3 high grade). DNA was extracted from tissue blocks available for 41 cases (both SBT and carcinoma, n=36; SBT only, n=3; carcinoma only, n=2). Samples were subjected to digital droplet PCR to analyze mutation hotspots in KRAS (codon 12) and BRAF (V600E), which are frequently found in low-grade serous tumors. Eighty-one percent of SBTs (34/42) were APST, and 19% (8/42) were niLGSC. Forty percent of cases (17/42) were FIGO stage I, the majority of which were APST (14/17; 82%). The median time to development of carcinoma was 9 years (range, 0.6 to 25 y). Mutations in SBTs were distributed as follows: 5/39 (13%) BRAF mutant, 22/39 (56%) KRAS mutant, and 12/39 (31%) wild-type for both genes. There was a significant relationship between SBT gene mutation and histologic type, with BRAF mutations occurring exclusively in APST and a higher frequency of niLGSC among SBTs wild-type for BRAF and KRAS ( P =0.01). The diffuse presence of tumor cells with abundant eosinophilic cytoplasm was significantly associated with the BRAF mutation ( P =0.001). Mutational analyses of matched SBT/carcinoma pairs revealed concordant profiles in 33/36 (92%) cases, of which 19 (53%) were KRAS mutant, 4 (11%) were BRAF mutant, and 10 (28%) were wild type for both genes. The 3 discordant cases consisted of a wild-type niLGSC with a subsequent BRAF -mutant invasive LGSC, a KRAS G12V -mutant APST with a KRAS G12C -mutant LGSC, and a BRAF -mutant APST with subsequent development of a KRAS G12D -mutant high-grade serous carcinoma. In conclusion, some women with SBTs can subsequently develop serous carcinoma, occasionally over 10 years later. Most subsequent carcinomas are low grade, but a small subset can be high grade. The type of gene mutation in SBT correlates with various histologic features. While most cases of serous carcinoma developing after a diagnosis of SBT probably represent tumor progression, a minority are independent primary tumors, presumably arising from endosalpingiosis.