[Origin of ovarian epithelial inclusions and its relationship with the development of low-grade serous carcinoma].

Objective: To investigate the possible origin of ovarian epithelial inclusions and its relationship with the low-grade ovarian serous carcinoma.

Methods: By comparatively evaluating the morphologic (secretory and ciliated cell distribution) and immunophenotypic [using paired box gene 8 (PAX8), tubulin, calretinin, and Ki-67 as first antibodies] attributes of ovarian epithelial inclusions, the normal tubal epithelium, and the ovarian tumors, all adnexal tissues from a total of 198 patients were studied, including 116 adnexae removed for non-neoplastic indications, 53 serous cystadenomas, 44 serous borderline tumors, and 41 low-grade serous carcinomas, which were collected from Qilu Hospital of Shandong University and University of Arizona in USA. Immunohistochemical single staining was used to detect the expressions of PAX8, tubulin, calretinin, and Ki-67 in the two groups, while immunohistochemical double staining of PAX8/calretinin was used to figure out the immunophenotype of various ovarian epithelial inclusions in a more intuitive way.

Results: With immunohistochemical single staining of PAX8 and calretinin, the vast majority (90%, 54/60) of ovarian surface epithelia displayed a mesothelial phenotype [calretinin(+), PAX8(-)], whereas 10% (6/60) of the cases displayed foci with tubal phenotype [calretinin(-), PAX8(+)]. In contrast, most (79%, 728/921) of the ovarian epithelial inclusions displayed a tubal phenotype, though 21% (193/921) of the ovarian epithelial inclusions showed a mesothelial phenotype. It was further proved by immunohistochemical double staining of PAX8/calretinin. Secretory and ciliated cells were found in the ovarian epithelial inclusions with tubal phenotype. There was a progressive increase in the secretory/ciliated cells ratio and proliferative index, from ovarian epithelial inclusions/cystadenomas to borderline tumors to low-grade serous carcinoma, according to the expression of tubulin and Ki-67.

Conclusions: The findings make a strong argument that the ovarian epithelial inclusions displaying a tubal phenotype with PAX8(+), calretinin(-) is likely derived from fallopian tube rather than through Mullerian metaplasia from ovarian surface epithelium. The increasing trend of secretory/ciliated cells ratio and proliferative index from ovarian epithelial inclusions/cystadenomas to borderline tumors to low-grade serous carcinoma indicates that the latter is a clonal expansion of secretory cells. Genetic and molecular studies are needed to further confirm these findings.