Sclerosing Stromal Tumor of the Ovary: MR-Pathologic Correlation in Three Cases

Ovarian sex cord stromal tumors are classified, inter alia, as granulosa stromal cell tumors, Sertoli's stromal tumors, and steroid cell tumors. SSTs, along with granulosa cell tumors, thecomas and fibromas, are granulosa stromal cell tumors (6), and were first described as a distinct entity among ovarian sex cord stromal tumors by Chalvardjian and Scully in 1973 (7). They are known to have the following characteristic clinical features: first, they usually occur in a younger age group, among those aged 27 or 28 years (1, 4, 5); whereas other types of stromal tumors are most common in the fifth and sixth decades, about 80% of SSTs are encountered during the second and third decades (1-3, 5, 8). In our study, two of three patients were aged less than 30, and the other was 39. Second, the most common presenting symptom is menstrual irregularity (3-5), and this occurred in all our three patients. Third, ascites occurs, but is rare (2, 4, 7, 9). In all our three cases, however, there was some ascites around the mass or in the cul-de-sac, and in several previous case reports the presence of ascites was also noted (1-2). To date, malignant SST has not been reported (1, 4). In our study, all three SSTs originated from the left ovary, and in several previously published reports, occurrence on this same side was also described (1-3, 5). In another study, however, this was not the case (4).

Hormonal activity has been reported in a few documented cases (1, 9-10), and though none of our patients underwent hormonal assay, they demonstrated no symptoms suggesting hormonal activity.

In a previous report by Lee et al. (4), SSTs were shown at ultrasonography to be solid and cystic adnexal masses with centrally located, multiple, round or cleft-like cysts, and two patients in our study, in one of whom the cyst was entirely solid and homogeneously echogenic, showed similar features. Among the cases described by Lee et al., ascites was uncommon, occurring in only two of the seven. Ascites was, however, demonstrated at ultrasonography in all our patients. In their report, color Doppler ultrasonography of SSTs revealed prominent vascularity in the peripheral portion and central intercystic spaces; power Doppler ultrasonography was not performed. In one of our patients, however, this modality showed a peculiar finding: the so-called "spoke-wheel appearance" , i.e., peripherally located arc-shaped large vessels and multiple vertically oriented centripetal vascular flow. We believe that because of its higher sensitivity to flow signals, power Doppler ultrasonography permits continuous delineation of vascular flow.

The T2-weighted MRI findings of SSTs are a round or oval-shaped mass with hyperintense cystic components, and a heterogeneous solid component of intermediate to high signal intensity (1, 2, 5, 8). At dynamic contrast-enhanced imaging, early peripheral enhancement with centripetal progression has been reported (2, 5, 8).

Microscopically, SSTs are characterized by a number of distinctive features: 1) a pseudolobular growth pattern, in which cellular areas are separated by edematous and collagenous hypocellular areas; 2) collagenous sclerosis within the cellular areas; 3) marked vascularity, with a "hemangiopericytomatous" pattern; and 4) heterogeneity of the cell population (1, 2). Except for this last-mentioned, our MR findings clearly reflected these pathologic features. At T2-weighted imaging, pseudolobulation was represented by an admixture of hypointense lesions set against a background of hyperintense stroma; pronounced sclerosis within the cellular nodules correlated with the hypointense areas. Prominent vascularity within cellular areas explained the marked and early tumoral enhancement we observed. Pathologic specimens obtained from our patients showed that the cystic areas were small; in fact the areas of high signal intensity seen at T2-weighted imaging and the unenhanced areas observed at dynamic imaging correlated with areas of hypocellularity and edema, and the thick or thin peripheral rim corresponded to the compressed ovarian cortex and capsule. In an earlier report, Ihara et al. (1) concluded that the presence of a thick peripheral rim consisting of compressed ovarian stroma can help differentiate SSTs from other stromal tumors. SSTs, they claimed, have a thick peripheral rim for two reasons: first, they occur in younger women with large ovaries; second, the tumors are slow growing. Fibromas and thecomas, on the other hand, are common in older women with atrophied ovarian stroma that are hardly visible, even if present at a tumor's periphery. However, our report showed that the peripheral rim of SSTs varies in thickness, and may be very thin. In one of our patients (case 3), the peripheral ovarian cortex contained three functional cysts.

The differential diagnosis of SSTs should include thecoma/fibroma, metastases, and malignant epithelial ovarian tumors. Fibromas/thecomas usually show low signal intensities at T2-weighted imaging, and slow and prolonged enhancement at dynamic MRI. Ovarian metastases and malignant epithelial tumors usually occur in older patients, and dynamic MR imaging does not usually reveal progressive centripetal enhancement.

In summary, SSTs are mixed cystic and solid tumors with central round or cleft-like cysts and show quite specific imaging findings, especially at power Doppler ultrasonography and dynamic MRI. Findings of large peripherally located vessels and centripetal vascular flow ("spoke-wheel appearance") at power Doppler ultrasonography, and early peripheral enhancement with centripetal progression at dynamic MRI, are usual. Acknowledgment of these specific findings of SSTs permits the accurate preoperative diagnosis of these benign tumors and promotes less invasive surgery (such as laparoscopic tumor excision) rather than oophorectomy.