Ovarian mucinous tumors with mural nodules are rare. Such mucinous epithelium can be benign, borderline or malignant. Although there have been sporadic reports of malignant mural nodules, such as clear cell carcinoma, carcinosarcoma, and sarcoma, the most common type of mural nodular malignancy is anaplastic carcinoma . The mural nodules of anaplastic carcinoma can be divided into three types: rhabdomyoid, spindle cell-like and pleomorphic [2, 4]. The three cases of anaplastic carcinoma in our report were spindle cell-like.
Case 1was a young woman with SLMNs. In some areas, spindle cells showed infiltrative growth, and immunohistochemical staining showed diffuse expression of AE1/AE3, p53 was negative for mutation type (Figs. 2A and 2B). Matias-Guiu et al.  proposed that cytokeratin immunohistochemical staining should be used to distinguish SLMNs and anaplastic carcinoma mural nodules. CK staining was negative or focal positive in SLMNs, while strong positive diffuse staining was detected in anaplastic carcinoma nodules. The heterogeneous expression of AE1/AE3 and p53 in Case 1 (Figs. 2C and 2D) suggested that there were two components in the mural nodule. Therefore, the patient was diagnosed as mucinous adenocarcinoma with sarcomalike and anaplastic carcinoma mural nodules. In a literature review, Shao et al.  reported that most benign SLMNs are associated with anaplastic carcinoma mural nodules. Case 1 was young and the anaplastic carcinoma was spindle-shaped, which is not easy to distinguish from spindle cells in SLMNs, making this very easy to overlook in the diagnosis process. Therefore, it is necessary to take sufficient samples for careful observation.
The differential diagnosis between SLMNs and sarcomatous or anaplastic carcinoma mural nodules is generally considered to be particularly important. SLMNs have a good prognosis because of their inert clinicopathological characteristics. The formation of SLMNs may be the result of differentiation of undifferentiated mesenchymal cells beneath the mucinous epithelium into epithelial cells, elicited by some stimulation, for example, intramural hemorrhage or cyst contents. Zheng et al.  reported that the proliferation of multinucleated giant cells was mostly located between the basement membrane of the mucous epithelium and the stromal cells in the bleeding area of the cystic wall in sarcomalike cases with anaplastic carcinoma. It was suggested that the formation of SLMNs might be the result of differentiation of undifferentiated stromal cells beneath the mucous epithelium under external stimulation. However, the latest research on mural nodules showed that there were tumor cell groups with abnormal expression of p53 and MTAP in SLMNs, and molecular detection suggested that SLMNs are not a benign reactive process and all mural nodules in mucinous ovarian tumors should be regarded as potentially malignant neoplasms. Therefore all mural nodules in mucinous ovarian need to be thorough sampling and rigorous clinical correlation to exclude extraovarian spread [3, 7]. Based on the results of HE morphological evaluation and molecular detection of the former, the molecular detection results may be closer to the nature of mural nodules. Therefore, SLMNs may not be the result of reactive process of mucinous tumors, but the clonal derivatives of ovarian mucinous tumors.
Case 1refused to expand the scope of the operation and postoperative chemotherapy. Pleural thickening and pleural effusion were found 11 months after the operation. Pleural biopsy showed adenocarcinoma and immunohistochemistry showed metastasis. Adenocarcinoma cells were found in pleural effusion and ascites, and the patient died 12 months after surgery.
In Case 1, both mucinous adenocarcinoma and anaplastic carcinoma nodules were negative p53 mutations, while in Cases 2 and 3, p53 was wild-type (Figs. 2E and 2F). The synchronous expression of p53 in epithelia and mural nodules suggested that mural nodules may be homologous with mucinous tumors, which is the result of dedifferentiation of mucinous tumors and tumor lesions. This conjecture is consistent with reports by Desouki et al.  and Zheng et al. . Shao et al.  and others also speculated that the mural nodules of anaplastic carcinoma may be the result of dedifferentiation of mucinous tumors, in which p53 mutation plays a key role. Chapel et al.  reported that only 1 of 13 patients had p53 mutations in mural nodules, while no mutations were detected in mucinous tumors. There were eight cases with the same p53 mutation, indicating that p53 mutation does not play a key role in the occurrence of mural nodules. Although the immunohistochemical expression of p53 was not completely consistent with the results of molecular detection, no p53 mutation was found in the three cases of mural nodules of anaplastic carcinoma reported here. NGS analysis of Case 3 also confirmed the absence of mutations p53.
Case 3 was diagnosed as ovarian mucinous tumor by frozen section examination. The appendectomy was prophylactic, and the appendiceal disease was diagnosed as broad-based serrated adenoma, which is rare and belongs to precancerous lesions. Two rare pathological types occurred in the same patient at the same time, indicating the possibility of genetic changes. Therefore, NGS detection was carried out. No germline mutation was found. Only missense mutations of PIK3CA and PTEN were detected.
Among seven cases of mucinous tumors with mural nodules of anaplastic carcinoma, Mesbah Ardakani et al.  identified six cases with KRAS mutation. Therefore, it is inferred that the mural nodules of anaplastic carcinoma may originate from tumors with KRAS mutation. However, in 13 cases of mural nodules analyzed by Chapel et al. , only nine cases had KRAS missense mutations, and two had PIK3CA missense mutations. In Case 3, no KRAS mutation was found, while a PIK3CA missense mutation was identified. Therefore, KRAS mutation may not be associated with the formation of mural nodules. The PTEN mutation suggests that there signaling pathways other than the KRAS pathway are involved in tumorigenesis.
Recent studies suggest that loss of the SWI/SNF protein may play an important role in the de-differentiation of gynecological tumors. Mutations of SMARCA4, SMARCB1, ARIDLA and ARIDLB are common in undifferentiated carcinoma, including undifferentiated components of de-differentiated carcinoma. It is reported that detecting the corresponding loss of protein expression can be regarded as a surrogate method for mutation of these genes. Chaudet et al.  found that nine of 25 mural nodules lacked expression of one or more SWI/SNF proteins, which was retained in all related mucinous tumors. Furthermore, the clinical outcomes of patients with SWI/SNF expression deficiency were often poor. In Case 3, immunohistochemical staining showed no loss of INI-1 expression in mucinous tumors and mural nodules, indicating that there was no SMARCB1 mutation in this case (Fig. 2G).
Among the three cases reported here, blood NGS analysis was performed for only one case, and the other two cases were not tested for molecular markers. Moreover, NGS analysis of blood cannot be used to detect mucoepidermic and mural nodules.
In conclusion, young patients with SLMNs morphology, full pathological analysis should be undertaken to avoid the omission of malignant mural nodules. The expression of p53 in the mural nodules of anaplastic carcinoma was consistent with that in mucinous tumors. The serrated lesions of the appendix and ovarian mucinous tumor occurred simultaneously. No germline mutations were detected by NGS analysis, although there were missense mutations of PIK3CA and PTEN, indicating that this was a sporadic case.