EMCs were first defined as a class of solid tumours by Enzinger et al. [1], and they constitute < 1% of all soft-tissue sarcomas [2]. EMCs are commonly seen in male adults, with a mean age of 50 years. Lesions are most frequently found in the deep soft tissues of the proximal extremities, especially the lower extremities; however, in some cases, EMCs can also affect the trunk, head and neck, abdominal wall, paravertebral soft tissues and bones [7–10].
Histologically, EMCs can be categorised into classical EMCs and two variants, cellular EMC and solid non-myxoid EMC [11]. Cellular EMCs account for ~ 29% of all EMCs [12] and share the same nodular structure with classic EMC; however, the cellular variant is characterised by an abundance of compact tumour cells and a limited myxoid matrix, with the tumour cells in round or polygonal shapes, enriched in hyperacidophilic or hypochromic cytoplasm and having well-defined, centred, or deviated nuclei. Immunohistochemically, vimentin expression exhibits a diffuse staining pattern, with the expression of S-100, CD117, syn and NSE differing from each other. Moreover, no antibody of relative specificity is detected, and in most cases, the expression level of ki-67 is < 10%. Ultrastructural studies reveal that EMC cells are rich in mitochondria and contain a well-developed Golgi apparatus, numerous smooth vesicles and dense core granules; synaptophysin expression is detected by immunoelectron microscopy, demonstrating the neuroendocrine differentiation of EMC cells [8, 13]. A recent study [14] reported positivity for INSM1 in up to 90% of EMC cases, which is considered evidence for the neuroendocrine differentiation of EMC cells; however, INSM1 expression was not detected in this case. In terms of molecular genetics, NR4A3 rearrangement is involved in ~ 90% of EMC cases but is absent in other types of sarcomas, which may facilitate the development of an EMC-specific technique for differential diagnosis. In patients with NR4A3 rearrangement, NR4A3 is mostly fused with EWSR1 (about 62–75%) [4–6] and less frequently with TAF15 (27%), TCF12 (4%), TFG and FUS. Other studies have reported HSPA8-NR4A3 translocation [15], a novel t(2;22)(q34;q12) EWSR1 translocation [16] and SMARCA2-NR4A3 fusion [17].
As described above, the tumour cells, in this case, had a nodular structure and underwent infiltrative growth, with the invasion of the peripheral fat and muscle tissues. There were abundant fibrous tissues between the nodules, most of which were identified as solid nodules (> 90%) and comprised of round or fusiform cells. The tumour cells had uniform size, round nuclei with well-defined nucleoli and eosinophilic cytoplasm, and their nuclear fission was 2/10 HPF. The nodular structure partly resembled that of classic EMC, but the tumour cells were loosely arranged, forming reticular or crosswise layers in the myxoid matrix. Local necrosis was observed. It's a huge challenge to diagnose this tumor because of the lack of specificity in morphological features. There are many soft tissue tumours with similar morphology to this tumour The main differential diagnosis includes cellular EMC, proximal-type epithelioid sarcoma, extrarenal malignant rhabdoid tumour, epithelioid angiosarcoma, malignant solitary fibrous tumour, extraosseous Ewing's sarcoma, desmoplastic small round cell tumour, metastatic dedifferentiated chordoma, poorly differentiated synovial sarcoma and epithelioid malignant peripheral nerve sheath tumour. IHC staining plays an essential role in this process. In this case, the IHC staining results suggested that the patient was diffusely positive for CD117, vimentin, CD56 and NSE, focally positive for desmin, with a ki-67 level of ~ 40% and negative for other markers. Despite the effectiveness of IHC staining, a definitive diagnosis of EMC cannot solely depend on positive IHC results for CD117, vimentin, CD56 and NSE as specific markers, but it gave us a diagnosis clue at least, so NGS was performed in the tumor, and a site specific to EWSR1 exon 7-NR4A3 exon 2 fusion was subsequently identified. Based on this, a diagnosis of cellular EMC was confirmed by considering the combined results of the morphological, immunophenotype and molecular analyses.
Interestingly, along with EWSR1 exon7-NR4A3 exon2 fusion, the NGS detected KIT exon 13 mutations; moreover, it was noted that the IHC staining results demonstrated a strong diffuse expression of CD117. Only a few studies have discussed the role of CD117 in IHC staining or the KIT gene in the diagnosis of EMC. Hornick et al. [18] reported that 2 of 20 patients tested positive for CD117 (one focally positive case and one diffusely positive case), without clarifying the expression intensity or detection of the KIT gene. In the study by Subramanian et al. [19], based on IHC and ISH ( In Situ Hybridization) tests, 8 of 19 EMC patients were focally positive for CD117, and 6 of 11 EMC patients were diffusely positive for CD117, but the expression intensity or scope was not discussed; the diffuse positive cases (6/11) underwent screening of KIT exons 9, 11, 13 and 17 for mutations, and the results suggested that there were no mutations in these exons. The IHC staining results in the study by Stacchiotti et al. [20] suggested the presence of CD117 expression in 6 of 9 EMC cases, including four cases of weak focal expression of CD117, while the western blot test results in the same study indicated the KIT expression without phosphorylation; yet, no KIT gene analysis was conducted. Subsequently, Urbini et al. [21] reported the case of an EMC patient (1/20) with KIT exon 11 mutations, which, apart from this paper, is the only reported EMC case with KIT gene mutations; however, no IHC staining results were provided in the study. In this case, the presence of KIT exon 13 mutations was established based on the NGS and morphological findings, i.e. abundant tumour cells, scarce myxoid stroma and strong diffuse expression of anti-CD117 antibody in tumour cells. We consider that CD117-positive (CD117⁺) cellular EMC might have a higher frequency of KIT gene mutations. EMC subtypes were not elaborated in the above-mentioned studies, which involved patients without KIT gene mutations but who tested positive for anti-CD117 antibody or those with KIT gene mutations before anti-CD117 antibody testing. Moreover, to the best of our knowledge, no study has explored the association of anti-CD117 antibody and the KIT gene with EMC subtypes. Therefore, it is necessary to study more EMC cases positive for anti-CD117 antibody and cellular EMC cases positive for CD117.
In a gastrointestinal stromal tumour, the location of KIT mutations is associated with its biological behaviour, with exons 11 and 13 mutations providing evidence regarding its malignant biological behaviour [22]. KIT exon 13 mutations are relatively rare, accounting for 0.8–4.1% of all KIT mutations, and patients with KIT exon 13 mutations benefit from sunitinib therapy [23]. A patient with EMC who had KIT exon 11 mutations and had never received sunitinib was reported to benefit from sunitinib therapy [21]. In this case, the patient experienced KIT exon 13 mutations but was not administered sunitinib; therefore, the therapeutic activity of sunitinib as part of the treatment for this patient remains unclear. However, the findings of the preceding studies indicate that KIT mutations are potential targets for EMC treatment. Further studies with larger sample size and follow-up data are needed to verify this viewpoint.
Presently, wide surgical resection is the mainstay of treatment for patients with EMC, and opinions are mixed concerning the treatment of cellular EMC. The prognosis of EMC remains controversial. In the 1970s, Enzinger et al. [1] viewed EMCs as low-grade malignant tumours when they first introduced this class of solid tumours. Later, the long-term follow-up study of 10 patients with EMC by Saleh et al. [24] revealed that 70% of the patients with EMC had an average survival of 10 years. Meis-Kindblom et al. [12] reviewed 177 EMC cases treated by the Armed Forces Institute of Pathology and Sahlgren University Hospital over 20 years for analysing the biological behaviour and prognosis of EMC. They discovered that EMC had a rather long disease course and was prone to metastasis and local recurrence; moreover, the case-death ratio was relatively high, and the biological behaviour differed from that of other low-grade malignant soft-tissue sarcomas. The 10-year survival rate was 70% among the patients with EMC, and the prognosis of EMC was not associated with histological grade or proliferative markers but with the location and size of the tumour in elderly patients. In this case, it was only about 2 months after the surgery, so the significance of follow-up results were limited. We will continue to follow up.
In conclusion, molecular detection is an indispensable technique for the diagnosis of EMC, especially rare variants like cellular EMC. The KIT mutations that were reported in this case report may offer fresh insights regarding EMC treatment.