Immunophenotype of uterine tumor resembling ovarian sex cord tumor (UTROSCT): Case series and meta-analysis of the literature

Objectives: This study aimed to define the frequency of positivity of several immunohistochemical markers in uterine tumor resembling ovarian sex cord tumor (UTROSCT). Methods: All consecutive UTROSCT cases were retrieved from consultation files of one of the authors. Histological and immunohistochemical slides were reviewed. In addition, three electronic databases were searched from their inception to January 2024 for all studies assessing the immunophenotype of UTROSCT. Exclusion criteria were: sample size < 10 patients, overlapping patient data, reviews. Endometrial stromal tumors with sex cord-like areas (formerly called "type I UTROSCT") were excluded. Immunohistochemical markers assessed in >= 10 cases in at least 3 different series were included. For each marker, pooled positivity rate was calculated by using a random effect model; mean labeling index was calculated for Ki67. Results: Thirty UTROSCT cases were retrieved Six studies were included, and 30 new cases were obtained, with a total of 181 UTROSCTs. Fourteen immunohistochemical markers were assessed. Pooled positivity rates were (in descending order): CD56 = 97 %, progesterone receptor = 91 %, estrogen receptor = 85.5 %, WT1 = 84 %, widespectrum cytokeratins = 78.7 %, CD99 = 77 %, desmin = 74.5 %, calretinin = 70.6 %, smooth muscle actin = 56.4 %, inhibin = 44.5 %, CD10 = 41 %, caldesmon = 21.9 %, Melan-A/MART-1 = 10.4 %. Mean Ki67 labeling index was 8.7 %. Conclusions: Immunophenotypically, UTROSCT is less consistent than ovarian sex cord tumors and overlaps with other mesenchymal and epithelial tumors; an integrated clinico-pathological and immunohistochemical evaluation appears necessary for a correct diagnosis.