Myoid Gonadal Stromal Tumor

A Clinicopathologic Study of Three Cases of a Distinctive Testicular Tumor

Chia-Sui Kao, MD; Thomas M. Ulbright, MD


Am J Clin Pathol. 2014;142(5):675-682. 

In This Article


Sex cord–stromal tumors of the testis are rare, representing approximately 4% to 6% of all adult testicular tumors.[24–26] Even rarer are pure spindle cell tumors, lacking a sex cord component and showing myofilaments, pinocytic vesicles, and lipid droplets on electron microscopy as well as SMA and S100 protein immunoreactivity.[1–5] Weidner,[3] after ultrastructural study, proposed the termmyoid gonadal stromal tumor for this neoplasm based on its myogenic differentiation (thin filaments with focal densities and α-SMA immunoreactivity), similar to peritubular myoid cells present in the normal testis, which also contain subplasmalemmal micropinocytotic vesicles, thin filaments with focal densities, and reactivity for α-SMA and desmin. Given the rarity of this tumor, previous reports were restricted to one or two cases each, and all were clinically benign.[1–5] In addition, the nomenclature for this tumor varied from "unusual gonadal stromal tumor" and "testicular stromal tumor with myofilaments" to "unclassified sex cord–stromal tumor with predominance of spindle cells." To our knowledge, our study is the largest contribution regarding this rare entity and provides additional clinicopathologic information concerning this neoplasm.

Our three patients all had single testicular masses without serum tumor marker elevations or hormonal dysfunction. The gross appearance was a well-circumscribed mass with a yellow or tan appearance. The tumor cells were spindled or fusiform and arranged in tight, short, and occasionally broad, intersecting fascicles. The nuclei were uniform and elongated with inconspicuous to small nucleoli and had occasional grooves. The cytoplasm was ill-defined and pale to lightly eosinophilic. Follow-up of our patients showed all were alive with no evidence of disease.

Immunohistochemistry demonstrated consistent strong and diffuse positive reactivity for SMA, S100 protein, and FOXL2; positivity for the first two was considered one of the defining criteria for this neoplasm.[4] SF-1 was also diffusely positive in all three cases but with less intensity, and there was variable positivity for desmin and WT1. Inhibin and calponin were both focally and weakly positive. SOX9, h-caldesmon, and calretinin, on the other hand, were consistently negative. The only similarities between the tumors and the peritubular myoid cells of the adjacent nonneoplastic testis were reactivities for SMA, calponin, and desmin (variable). The peritubular myoid cells were consistently negative for S100 protein, FOXL2, SF-1, inhibin, and WT1 but positive for h-caldesmon. The difference in immunostaining pattern argues against myoid gonadal stromal tumors showing peritubular myoid cell differentiation.

Aside from peritubular myoid cells, the origin from primitive gonadal stromal cells—more specifically, intertubular mesenchymal cells that underwent myogenic differentiation—has also been hypothesized.[2,4,27–29] The tumor's consistent strong and diffuse positivity for FOXL2, together with its negative reactivity for SOX9, provides support that myoid gonadal stromal tumor is a pure stromal neoplasm rather than of mixed sex cord–stromal lineage since SOX9 is a relatively sensitive and specific marker for sex cord tumor elements compared with FOXL2, which is a better marker for "stromal" tumors in addition to granulosa cell tumors (C.S. Kao, T.M. Ulbright, and M.T. Idrees, unpublished observations).[11,12,14] SF-1, on the other hand, is a very sensitive but less specific "general" marker for both sex cord and stromal elements.[15] Our findings support the proposition made initially by Greco et al2 that myoid gonadal stromal tumors are derived from intertubular primitive mesenchymal cells that undergo myogenic differentiation.

Even though myoid gonadal stromal tumors have been shown to contain myofilaments and desmosomes on electron microscopy, similar to smooth muscle tumors or those with myoepithelial differentiation, h-caldesmon was consistently negative while calponin was only focally and weakly positive. Its S100 protein reactivity also contrasts with the negativity of leiomyoma. These findings provide further support that myoid gonadal stromal tumor is a distinct entity separate from leiomyoma and most likely not of myoepithelial origin.

The differential diagnosis of myoid gonadal stromal tumor includes leiomyoma, testicular fibrothecoma, and unclassified sex cord–stromal tumor; Table 4 provides a summary of their individual immunohistochemical properties. The main differential between myoid gonadal stromal tumor and leiomyoma has been presented in detail by Du et al[4] and is only briefly discussed here. Myoid gonadal stromal tumor lacks the diffuse broad fascicles, abundant eosinophilic cytoplasm, and cigar-shaped nuclei often accompanied by perinuclear vacuoles seen in leiomyoma (Table 2). Furthermore, myoid gonadal stromal tumor shows S100 protein positivity and lacks h-caldesmon reactivity, while the opposite profile is observed in leiomyoma.

The variably present collagen bundles may be somewhat akin to those seen in testicular fibrothecoma, but in the latter, these bundles are typically thicker and more diffuse, sometimes forming hyaline plaques, similar to those seen in the ovarian counterpart.[30] The pattern of immunostaining is also helpful in differentiating myoid gonadal stromal tumor from testicular fibrothecoma, as has been described previously.[30] Myoid gonadal stromal tumor shows consistent strong and diffusely positive immunostaining for S100 protein and SMA with focal or negative reactivity for inhibin. Testicular fibrothecoma, on the other hand, shows frequent patchy to diffuse inhibin and calretinin staining with variable, patchy S100 protein reactivity.[30] In addition, testicular fibrothecoma shows positive reactivity with SOX9, FOXL2, and SF-1 (C.S. Kao, T.M. Ulbright, and M.T. Idrees, unpublished observations), whereas myoid gonadal stromal tumor is positive for FOXL2 and SF-1 but negative for SOX9, supporting the hypothesis that these two are separate entities. SMA is commonly positive in testicular fibrothecoma and does not aid in the distinction from myoid gonadal stromal tumor.[30]

In our opinion, spindle cell tumors of the testis with sex cord differentiation appreciable on H&E- or reticulin-stained slides without distinct tubule or follicle formation should be diagnosed as "unclassified sex cord–stromal tumor," even with S100 protein and/or SMA positivity since these reactivities are nonspecific and may also be observed in other types of sex cord–stromal tumors.[5,7,9,30,31] We have excluded tumors with sex cord differentiation from our series. Careful light microscopic examination of standard H&E-stained sections permits the identification of subtle sex cord elements in unclassified, spindle cell–predominant sex cord–stromal tumors. These usually stand out as darker, tight groupings of cells with rounder nuclei and indistinct, typically pale, often scant cytoplasm; they may be arranged in nests or short cords Image 1F. Such foci can be highlighted by a relative lack of reticulin fibers, contrasting with the surrounding pericellular stromal pattern (Image 1F, inset), and also by stronger reactivity for inhibin, although overall, the immunoprofile of these two entities may show significant overlap. Sometimes Sertoli cell, Leydig cell, and granulosa cell tumors may have a prominent component of spindle cells, but adequate sampling permits appreciation of more characteristic tumor features. A diagnosis of myoid gonadal stromal tumor should therefore be reserved for pure spindle cell tumors lacking sex cord differentiation and show, at the minimum, strong and diffuse positivity for both S100 protein and SMA.

A literature search for testicular tumors fulfilling our diagnostic criteria revealed six cases (Table 3).[1–5] Combining our cases with those previously reported (a total of nine) shows that testicular myoid gonadal stromal tumors are unilateral and small (100% <4 cm) and occur in a wide age range (4–59 years; mean, 37 years; median, 43 years) but mostly in young to middle-aged men. Occasional mitotic figures occur, but they lack features associated with malignant behavior established for tumors in the sex cord–stromal category.[32–34] Follow-up information for all nine patients (3–60 months; mean, 25 months; median, 12 months) shows that these tumors have behaved in a benign fashion. Hence, the biologic potential for small tumors (<4 cm) with low mitotic activity is extremely favorable. We cannot reliably assess prognostic features indicative of aggressive behavior at this stage given that all tumors fulfilling our diagnostic criteria have been benign.

In summary, we report the first series of myoid gonadal stromal tumor with detailed clinicopathologic findings, including new information on immunohistochemical properties, together with a review of previously reported cases and differential diagnosis. Furthermore, the new information on its immunophenotype provides support that this distinctive neoplasm originates from intertubular primitive mesenchymal cells that undergo myogenic differentiation rather than peritubular myoid cells. The experience, although limited, has been entirely benign and justifies a conservative approach in the management of patients with these tumors so long as the morphologic findings are within the spectrum of the cases we have herein summarized. Although one may argue that the distinction from other sex cord–stromal tumors of the testis is subtle, the combined morphologic, immunohistochemical, and ultrastructural findings are unique and clearly defined by Du et al,[4] warranting separation from other tumors. We believe our series of myoid gonadal stromal tumors will aid in the recognition of this rare entity and encourage further reports to expand our understanding of its behavior.