Granular cell tumor (GCT) is a rare tumor of neural/Schwann cell origin that was first described in 1945 by Ravich et al.GCT can occur at any age and any part of the body. The most common site is the tongue, followed by the the breast, upper respiratory tract (throat and bronchus), and gastrointestinal tract (esophagus, large intestine and perianal, stomach, small intestine, and bile duct), rarely occurring in the orbit. The prognosis is poor. The local recurrence rate of malignant GCT is 32%-70%, the metastasis rate is 50%-62%, and the mortality rate is 39%-65%(1,2). Due to the clinical course of GCT and pathologic examination result is not completely consistent, diagnostic criteria for differentiating benign from malignancy in GCT has been controversial.
Studies have found that many GCTs that are initially thought to be benign have subsequently recurred or metastasized, and the clinical course is malignant.
Benchekroun has done relevant retrospective studies and found that 2% to 3% of GCT cases are histologically benign but clinically malignant.Therefore, many researchers believe that "metastasis" is the criterion to define malignant GCT (3).After summarizing the histological morphology of 46 cases of malignant GCT, Fanburg-Smith et al. (1) proposed six histological features of malignant GCT :(1) increased nucleo-plasma ratio, (2) pleomorphic nuclei, (3) neoplastic necrosis, (4) spindle-cell morphological changes, (5) vacuolar nuclei and large nucleoli.(6) More than two mitotic figures per 10 high-power fields (HPFs).They considered GCT to be "atypical GCT" if it exhibited 1-2 of these features, and "malignant GCT" if it exhibited 3-6 features.Several other diagnostic criteria have been proposed by subsequent researchers.For example, Nasser et al. (4) argued that the six morphological indicators in the Fanburg-Smith criteria did not have the same weight on the malignant biological behavior of GCT tumors.
They suggested that mitotic image count and neoplastic necrosis are the key indicators that really affect the biological behavior of GCT, and therefore proposed that GCT be classified as benign or with uncertain malignant potential based on the presence or absence of "necrosis" and/or "more than two mitotic images per 10 high-magnification fields."Machado et al. (3) found that not all GCT with malignant histological features would have metastasis, and the malignant histological morphology was not completely equivalent to the clinical course of malignancy.Therefore, they propose to classify tumors with malignant histological features as "GCT with increased risk of metastasis" and tumors without malignant features as "GCT with little metastatic potential".Both groups considered metastasis to be the only diagnostic criterion for malignant GCT.
In our view, these differences can be addressed by Gamboa's classification. As early as 1955, Gamboa raises two types of malignant GCT (5).One is classified as Gamboa type I, which is clinically malignant but pathologically benign, and the other is classified as Gamboa type II, which is clinically and histologically malignant.In practice, the Gamboa classification and the Fanburg-Smith standard can be used in combination.If it is confirmed that the tumor has metastasized, a malignant GCT should be diagnosed, even if its morphology is "benign."Otherwise, the Fanburg-Smith criteria should be adopted to classify the tumor into benign/atypical/malignant GCT.We report a case of male orbital GCT with increased nucleus to plasma ratio, pleomorphic cells, vesicular nuclei, some with distinct nucleoli, local spindle cell changes, neoplastic necrosis, and mitotic image of 2/10HPF. Although no tumor metastasis has been found in this case so far, according to Fanburg-Smith's diagnostic criteria, this case should be diagnosed as malignant granulosa cell tumor.
Differential diagnosis:
This case should be distinguished from many other tumors in the orbit in terms of morphology and location.(1) Malignant peripheral nerve sheath tumors (MPNSTs) and epithelioid malignant peripheral nerve sheath tumors.MPNST also originates from neural/Schwann cells and can occur in the orbit.It is characterized by a high density of spindle-shaped cells, tapering of the nucleus at both ends, and cell pleomorphism (6).Epithelioid MPNST is characterized by epithelioid cell morphology with eosinophilic cytoplasm.MPNST also expresse Schwann cell markers (S100 and SOX10), but most of them showed focal or patchy staining, while the stainings in this case are diffuse and consistent.Secondly, MPNST is often accompanied by the deletion of H3K27Me3 expression, and epithelioid MPNST is usually accompanied by the deletion of INI-1 expression. In this case, there was no deletion of H3K27Me3 and INI-1 expression.Moreover, in addition to expressing Schwann cell markers (S100 and SOX10), MPNST usually does not express calretinin, TFE3, CD68 and other markers, which can help distinguish MPNST from GCT. (2) Malignant melanoma.Melanoma can present with spindle or epithelioid morphology, often with distinct nucleoli and expression (S100 and SOX10), and needs to be differentiated from GCT. In terms of S-100, Melan-A and HMB-45 markers commonly used in the diagnosis of malignant melanoma, S-100 is the most sensitive, and S-100 is usually positive in malignant melanoma, and in most cases, Melan-A and HMB-45 expression are also common in melanomas .In some cases (such as metastatic melanoma or spindle cell melanoma), the latter two markers (Melan-A and HMB-45) may not be present, which can cause diagnostic confusion if pigmentation is not evident at this time (7). This case does not express melanin markers, no pigmentation, and has cytoplasmic specific eosinophilic particles and strong expression of calretinin, TFE3, CD68 and other markers, which can be distinguished from melanoma. In addition, it has been reported in literature that GCT can have positive expression of TFE3 in tumor nucleus, but no relevant report has been found in malignant melanoma, which can be one of the indicators to help differential diagnosis (8).(3) Morphologically, it should also be differentiated from sarcomatoid carcinoma and rhabdomyosarcoma. In this case, the epithelial marker PCK is negative, and the myogenic marker Desmin is not expressed, which can help to differentiate from these two.
Due to the lack of large series, long-term follow-up studies on GCT treatment, and no clinical trials related to oncology, no definitive conclusions can be drawn about the best treatment and follow-up for GCT.At present, the main treatment plan for GCT is local surgical resection.Regional lymph node dissection can be performed for malignant GCT.Studies have found that the therapeutic effect of chemotherapy and radiotherapy on malignant GCT is uncertain, and there may be some adverse reactions.Treatment options for unresectable malignant GCT or metastatic GCT have been limited.A recent study reported that pazopanib had some effect on metastatic malignant GCT.PIK3CA gene mutations have been identified in a subset of malignant GCT, thus providing a new strategy for the treatment of malignant GCT with targeted inhibitors (such as PI3K/AKT/mTOR inhibitors) alone or in combination with other regimens (9).