Search

Expression of neuroendocrine (NE) markers in primary ovarian non-NE epithelial tumors has rarely been evaluated. The aim of our study was to evaluate the expression of the most widely used NE markers in these neoplasms and to determine any prognostic significance of NE marker expression. The cohort consisted of 551 primary ovarian tumors, including serous borderline tumors, low-grade serous carcinomas, high-grade serous carcinomas (HGSC), clear cell carcinomas, endometroid carcinomas, mucinous borderline tumors, and mucinous carcinomas. Immunohistochemical analysis was performed using antibodies against INSM1, synaptophysin, chromogranin, and CD56 on tissue microarray. Positivity for INSM1, synaptophysin, chromogranin, and CD56 was most frequently observed in mucinous tumors (48.7%, 26.0%, 41.5%, and 100%, respectively). The positivity for these NE markers was mostly restricted to nonmucinous elements distributed throughout the tumor. The mucinous borderline tumor and mucinous carcinomas groups had similar proportions of positivity (mucinous borderline tumor: 53%, mucinous carcinomas: 39%). In the other tumor types, except for HGSC, there was only focal expression (5%–10%) or negativity for NE markers. HGSC showed high CD56 expression (in 26% of cases). Survival analysis was only performed for CD56 in HGSC as this was the only group with sufficient positive cases, and it showed no prognostic significance. Except for mucinous tumors, expression of NE markers in non-NE ovarian epithelial tumors is low. CD56 expression in HGSC occurs frequently but is without diagnostic or prognostic value.

In gynecological neoplasms, immunohistochemical (IHC) expression of p53 is generally an accurate predictor of TP53mutation status if correctly interpreted by the pathologist. However, the literature concerning cut-offs, frequency and prognostic significance of p53 staining in ovarian mucinous tumours is limited and heterogeneous. We performed an analysis of 123 primary ovarian mucinous tumours including mucinous borderline tumours (MBT), mucinous carcinomas (MC), and tumours with equivocal features between MBT and MC. We assessed p53 expression for the three recognised patterns of aberrant staining in ovarian carcinoma [overexpression (‘all’), null and cytoplasmic] but using a recently suggested cut-off for aberrant overexpression in ovarian mucinous tumours (strong nuclear p53 staining in ≥12 consecutive tumour cells) and correlated the results with next generation sequencing (NGS) in all qualitatively sufficient cases (92/123). Aberrant p53 expression was present in 25/75 (33.3%) MBT, 23/33 (69.7%) MC (75% of MC with expansile invasion and 61.5% with infiltrative invasion), and 10/15 (66.7%) tumours equivocal between MBT and MC. Regarding the 92 tumours with paired IHC and mutation results, 86 showed concordant results and six cases were discordant. Three discordant MBT cases showed aberrant expression but were TP53wild-type on sequencing. Three cases had normal p53 expression but contained a TP53mutation. Overall, IHC predicted the TP53mutation status with high sensitivity (94.1%) and specificity (92.7%). The accuracy of IHC was 93.5% with a positive predictive value of 94.1% and a negative predictive value of 92.7%. When comparing MC cases with wild-type TP53versus those with TP53mutation, there were no significant differences concerning disease-free survival, local recurrence-free survival, or metastases-free survival (p>0.05). In the MBT subgroup, there were no events for survival analyses. In conclusion, using an independent large sample set of ovarian mucinous tumours, the results of our study confirm that the suggested refined cut-off of strong nuclear p53 staining in ≥12 consecutive tumour cells reflect high accuracy, sensitivity and specificity for an underlying TP53mutation but the TP53mutation status has no prognostic significance in either MC or MBT.

Uterine tumor resembling ovarian sex cord tumor (UTROSCT) is a rare tumor of uncertain lineage and low malignant potential. Most tumors behave in a benign manner, but a subset of UTROSCT exhibit an aggressive clinical course with recurrences and metastases. The recurrent molecular alterations in UTROSCT mostly represent gene fusions involving NCOA1-3. We performed a comprehensive clinicopathological, morphologic, immunohistochemical, and molecular analysis on a cohort of 35 UTROSCT. The tumors exhibited various architectural patterns (diffuse, corded/trabecular, tubular, sertoliform, fascicular, whorled, nested, microfollicular, and pseudoglandular), often in combination. The immunohistochemical analysis confirmed the polyphenotypic immunoprofile, often with coexpression of sex cord–stromal, smooth muscle, and epithelial markers, as well as hormone receptors. Next-generation sequencing RNA analysis revealed recurrent NCOA1-3gene fusions in 22/32 analyzed cases (69%), including ESR1::NCOA3(11/22), GREB1::NCOA2(7/22), ESR1::NCOA2(3/22), and GREB1::NCOA1(1/22). Tumor mutation burden was low in all cases. The fusion-positive cases exhibited statistically significant association with whorled architecture, conversely necrosis was associated with fusion-negative status. We did not find a significant relationship between any architectural pattern and GREB1alterations, but the NCOA2-altered tumors were associated with pseudoglandular architecture. The GREB1-altered cases occurred in older patients and tended to be more often intramural masses compared with ESR1-altered cases. On the contrary, the ESR1-altered cases presented more often like submucosal or polypoid tumors. Two tumors exhibited aggressive behavior with recurrent disease. Both of these cases harbored a GREB1::NCOA2fusion. Unsupervised hierarchical cluster analysis of our cohort revealed 2 main clusters. The tumors with GREB1or NCOA2fusion cluster together, suggesting that there are underlying molecular differences between these cases and cases with ESR1::NCOA3fusion or without fusion. Our findings contribute to the growing knowledge about a rare neoplasm with currently uncertain biological behavior.