Identification of the most sensitive and robust immunohistochemical markers in different categories of ovarian sex cord-stromal tumors.

Different immunohistochemical sex cord-stromal markers have been previously studied in various types of ovarian sex cord-stromal tumors; however, the sensitivity for sex cord-stromal lineage may vary between markers, and some markers may not be as sensitive in some types of sex cord-stromal tumors compared with other tumors in this spectrum of neoplasms. The goals of this study were to determine which immunohistochemical markers are the most sensitive and immunohistochemically robust for sex cord-stromal lineage within a given type of ovarian sex cord-stromal tumor, and to establish whether there are substantial differences of expression of these markers between different types of sex cord-stromal tumors. Immunohistochemical stains for markers which have known variable specificity for sex cord-stromal lineage [inhibin, calretinin, MART-1/melan-A, CD99, steroidogenic factor 1 (SF-1, adrenal 4-binding protein), and WT1], were performed in 127 cases of 5 different types of ovarian sex cord-stromal tumors: adult granulosa cell tumor (n=32), Sertoli cell tumor (n=27), Sertoli-Leydig cell tumor (n=18), steroid cell tumor (n=25), and fibroma/fibrothecoma (n=25). All cases in each type of sex cord-stromal tumor expressed SF-1. Inhibin and calretinin were expressed in all groups of tumors but with a lesser frequency (56% to 100% and 36% to 100% of cases, respectively). All types of tumors except steroid cell tumor expressed WT1. Fibroma/fibrothecoma was the only type of tumor that did not express CD99. The only tumor groups that showed expression of MART-1 were Sertoli-Leydig cell tumor (restricted to the Leydig cell component) and steroid cell tumor (94% and 96% of cases, respectively). The type of sex cord-stromal tumor that was least frequently positive for several of the different markers studied was fibroma/fibrothecoma. Among all tumor groups combined, inhibin and WT1 were the 2 markers showing the most diffuse expression. Likewise, the single marker showing the most optimal combination of diffuse and strong staining (immunohistochemical composite score: possible range, 1 to 12) varied between tumors: adult granulosa cell tumor-inhibin (score 10.0); Sertoli cell tumor-WT1 (score 10.8); Sertoli-Leydig cell tumor (Sertoli cell component)-WT1 (score 10.4); steroid cell tumor-inhibin (score 11.2); and fibroma/fibrothecoma-WT1 (score 8.9). We conclude that most immunohistochemical sex cord-stromal markers have sufficient sensitivity for sex cord-stromal lineage. Although each of the different types of sex cord-stromal tumors has a slightly unique immunoprofile in terms of frequency and extent of expression, these differences are relatively minor for most types of tumors with certain exceptions (eg, WT1 is not diagnostically useful in steroid cell tumor; CD99 is not diagnostically useful in fibroma/fibrothecoma; the only sex cord-stromal tumor for which MART-1 is diagnostically useful is steroid cell tumor; inhibin and calretinin are less diagnostically useful in fibroma/fibrothecoma than in the other types of tumors, but expression in fibrothecoma was higher than in fibroma). SF-1 is the most sensitive sex cord-stromal marker among the most common types of sex cord-stromal tumors. Given the findings relating to sensitivity and extent of expression in this study, and known specificity in the literature, the most informative sex cord-stromal markers to be used for the distinction from nonsex cord-stromal tumors are inhibin, calretinin, SF-1, and WT1 (the exact number of markers to be used should be based on the degree of difficulty of the case and level of experience of the pathologist); however, the utility of immunohistochemistry for the diagnosis of fibroma/fibrothecoma is somewhat limited.