Primary Colonic Medullary Carcinoma With Exclusive Squamous Differentiation

Discussion

In this study, we report the first description of a colonic MC with exclusive squamous differentiation. Examination of the clinicopathologic findings in this case support the diagnosis. While colonic MC is rare, it is still much more common than a primary colonic SCC or metastatic SCC to the colon. Female sex, right colon location, and no prior history of malignancy additionally support the diagnosis. Lack of lymph node metastasis in such a locally advanced tumor would be much more commonly seen in MC than a primary or metastatic SCC. Of course, the histomorphology is key to diagnosis. The word “medullary” was coined to reflect the histological features of this tumor being solid, with nested, organoid, or trabecular growth pattern mimicking endocrine carcinomas. Cytologically, the tumor cells are classically small, round and uniform with scant eosinophilic cytoplasm and rounded nuclei with small central nucleoli (1). Grossly, MC is usually a locally advanced polypoid mass with central ulceration, often 50% larger in size than non-medullary malignancies, and predominantly involving the right colon (2, 3). Instead of having an infiltrative pattern typically seen in SCC or other conventional CRC, MC has pushing borders with increased intratumoral and peritumoral lymphocytic infiltration. Further, ancillary tests demonstrating dMMR/MSI-H and BRAF mutation are confirmatory (11). Despite intercellular bridging, desmoplastic stromal reaction, and diffuse squamous marker expression (CK5/6 and p40) that masqueraded itself as a SCC, the diagnosis of MC is reasonable and should be managed accordingly in our patient.

Complete and exclusive squamous cell differentiation in an otherwise classic MC not only posed diagnostic challenges, but also unveiled a previously unrecognized plasticity of differentiation in this tumor (12). Indeed, MC has been reported to frequently lose its colonic markers (12, 13). Aberrant nuclear and cytoplasmic positivity for a neural marker Calretinin in MC has been reported (13, 14). It is worth noting that the World Health Organization and the American Joint Committee on Cancer (AJCC) specifically separate MC from other undifferentiated CRC for its uniform morphology and molecular defects (15, 16). Thus, when Nguyen et al. described a case of “MC” with aberrant lymphoid marker expression, it was noted that the reported tumor had pleomorphic morphology and microsatellite stable (MSS) phenotype that would best be classified as an undifferentiated carcinoma of colon rather than MC (17). Nevertheless, their case also exemplified that CRC, no matter poorly differentiated or undifferentiated, can present with aberrant differentiation completely deviating from its colonic lineage.

Distinction between a primary MC and SCC is critical for clinical follow-up and management. Compared to other histologically similar entities or undifferentiated carcinomas, the prognosis of MC is more favorable despite frequent large tumor size and local aggressiveness (2). Surgical resection is the main treatment modality (1, 3). Adjuvant chemotherapy is not usually required unless lymph node or distant metastases are present. In addition, the dMMR/MSI-H status may render it responsive to an immune checkpoint inhibitor (18). In contrast, the diagnosis of a large poorly differentiated SCC in visceral organs often portends a poor prognosis (9). The preferred treatment of colonic SCC is surgical resection, with the recommendation of additional of chemotherapy (8). Pure SCC in the right colon is extremely rare and has been mostly associated with adenosquamous carcinoma of the colon (19). It was hypothesized that the squamous differentiation in colonic tumors may be associated with pluripotent stem cell differentiation, malignant transformation of persistent ectopic embryonal nests of ectodermal cells, or squamous differentiation of a preexisting colonic adenoma, which may be also the underlying mechanism in our case (20).

Besides SCC, other differential diagnosis to be considered in our case includes colorectal neuroendocrine carcinoma (NEC), which often exhibits organoid, nested, or trabecular growth pattern similar to those of MC (13). Negative neuroendocrine markers can help differentiating MC from NEC. With the characteristic increased lymphocytic infiltration, a high-grade lymphoma should also be ruled out, especially when the MC demonstrates aberrant lymphoid marker expression (11). Given the lack of mucosal surface precursor lesions, other differential diagnoses to be considered include epithelioid mesenchymal tumors such as epithelioid gastrointestinal stroma tumor (GIST), or malignant gastrointestinal neuroectodermal tumor (GNET), which may morphologically mimic MC, but can be distinguished by ancillary studies (21).

In summary, we report the first case of a colonic MC with exclusive squamous differentiation, masquerading as a primary colonic SCC or metastatic SCC from the lung. The correct diagnosis of MC was dependent on an integrated analysis of the clinical picture, histomorphology, immunophenotype, and molecular studies. The take-home message is to always keep MC in the differential diagnosis when dealing with a poorly differentiated neoplasm of the colon, especially when located in the right colon. In addition, this unusual case appears to open a Pandora’s box for unlimited differentiation potential in this type of tumor, which was not previously recognized. Future studies may focus on the pathogenesis of aberrant squamous differentiation and its implication in patient survival, which may or may not be different than an otherwise orthodox colonic MC.