Uterine Tumor Resembling Ovarian Sex Cord Tumor: A Distinct Entity Characterized by Recurrent NCOA2/3 Gene Fusions.

Uterine tumor resembling ovarian sex-cord tumor (UTROSCT) is a rare and distinctive neoplasm of unclear histogenesis, and uncertain malignant potential. These neoplasms morphologically resemble sex-cord stromal tumors of the ovary, and possess a polyphenotypic immunophenotype. Their molecular pathogenesis has yet to be elucidated; notably, however, tumors lack alterations found in other uterine tumors bearing sex-cord-like differentiation, such as endometrial stromal sarcoma. Following identification of an index patient with an ESR1-NCOA3 fusion gene by RNA-sequencing, we undertook a retrospective review for additional cases of UTROSCT. We identified a total of 4 patients, with an average age of 53 years (range, 38 to 68 y). RNA-sequencing was performed in all cases, revealing an ESR1-NCOA3 fusion in 2 cases and one case each with related ESR1-NCOA2 and GREB1-NCOA2 fusions. Each of the tumors showed histologic and an immunophenotype features within the previously reported spectrum of UTROSCT; interestingly, one case contained prominent spindle cell fascicles and another was largely comprised of sheets of small round cells. Our results demonstrate UTROSCT are defined by recurrent fusions involving NCOA2 or NCOA3, a finding that is directly amenable to diagnostic evaluation. This study confirms UTROSCT is molecularly distinct from endometrial stromal sarcoma, and raises intriguing new questions into the pathogenesis of these neoplasms and possible relationship with other NCOA fusion-positive uterine tumors.

Differential Immunoreactivity of p16 in leiomyosarcomas and leiomyoma variants.

Several studies have now examined the cDNA expression profiles of healthy myometrium, leiomyomas (LM), and leiomyosarcomas (LMS). This has produced a list of candidate genes that might be useful tools for distinguishing these entities from each other. The potential candidates identified from this body of research include insulinlike growth factor 1, h-caldesmon, cytokeratin 18, and the cyclin-dependent kinase 4 inhibitor, p16. To determine whether the immunohistochemical expression of these proteins could aid in the diagnosis of LMS and LM variants, we constructed a tissue microarray consisting of cases of healthy myometrium (n = 10), LM (not otherwise specified and variants; n = 47), and LMS (n = 8), and then measured the immunoreactivity of each of these proteins. The cases were scored on the basis of staining intensity (weak, moderate, or strong) and extent (focal or diffuse) and were assigned a final score from 0 to +3. Immunostaining for p16 was statistically stronger in LMS than in LM and its subtypes (P < 0.001). Specifically, the p16 immunostaining score in LMS cases (n = 8) was at least +2, whereas the p16 immunostaining scores in all LM cases (n = 47) were either 0 (n = 35) or +1 (n = 12). The expression of the remaining antibodies did not show a statistically significant difference between the 2 groups. Furthermore, none of the markers studied showed any differences among the LM variants. The results of this study confirm the overexpression of p16 in LMS and suggest that p16 can serve as a reliable immunohistochemical marker in distinguishing uterine LMS from LM and its benign variants.