Merkel cell carcinoma (MCC) is a rare neuroendocrine cutaneous malignancy with mortality rates of over one-third, which is double that of melanoma.1,2 MCC shares cytologic, histopathologic, and immunohistochemical features with a group of small round blue cell (SRBC) tumors, making differentiation difficult. The trabecular pattern, named for its resemblance to the trabeculae of cancellous bone (Fig. 1A), is anecdotally cited as a characteristic feature of MCC (Fig. 1B). Although many of the entities in the SRBC differential do not typically involve the skin, even some of the least common non-MCC SRBC tumors can metastasize to the skin,3 and metastatic foci may share histopathological features with the primary lesion. Broadening the inclusion parameters to typically non-cutaneous SRBC tumors strengthens the sensitivity of findings. We therefore sought to analyze growth patterns of MCCs and other SRBC tumors to assess the sensitivity and specificity of the trabecular pattern.
MATERIALS AND METHODS
This is an institutional review board–approved retrospective observational study conducted on 153 cases, including 74 non-MCC SRBC tumors and 79 MCCs. The cases were consecutively retrieved from the archives of the Ackerman Academy of Dermatopathology, New York, NY, and the Department of Pathology of the Medical University of South Carolina (MUSC), representing a period between 2002 and 2016. No patient identifiers or medical records were collected or maintained. An expert dermatopathologist assessed the growth pattern(s) of 79 immunohistochemically confirmed cases of cutaneous MCC and 74 non-MCC SRBC tumors. An example of each pattern is demonstrated in Figure 2. Non-MCC SRBC cases encompassed both primary and metastatic disease, including core biopsy (10/74) and excision (64/74) specimens. All MCC cases were biopsy specimens (eg, shave, punch). Because specimen sampling represents only a portion of the tumor, we allowed for varying proportions of trabecular features. For example, we counted trabecular features as present in a nephroblastoma with only one small focus of trabecular growth in 1 of 21 sections. Areas of crush artifact were avoided, and cases were excluded if there was insufficient tissue. The non-MCC cases had been extensively evaluated during initial workup, to include expanded immunohistochemistry and even molecular/genetic analysis when necessary. Any ambiguous or questionable cases were not included.
To simplify our study, we only assessed whether a trabecular pattern was present or absent, and intentionally did not assess depth, cytology, other patterns of distribution beyond those described in Table 1, concurrent epidermal atypia, etc. Our sole purpose was to challenge the anecdotal association of the trabecular pattern with MCC.
Contingency tables were generated, and sensitivities, specificities, likelihood ratios, and diagnostic odds ratios were calculated. All tests were statistically significant with a 2-tailed P-value of <0.05. Limitations of this study include its retrospective nature, and reliance of the original diagnostic classification of each neoplasm.
The frequency of histopathologic growth patterns observed in MCC and non-MCC SRBC tumors are summarized in Table 1. Multiple patterns were often noted in a single specimen. In non-MCCs, the most common histopathologic pattern was diffuse, followed by large anastomosing nests, infiltrative, and small islands. The trabecular pattern was rarely observed in non-MCC SRBC tumors. By contrast, the trabecular pattern was most common (72.2%) in MCCs. The presence of “any” trabecular features had a sensitivity of 72.2% and a specificity of 87.9% for MCC, with positive and negative likelihood ratios of 5.9 and 0.32, respectively. The odds ratio for a diagnosis of MCC over non-MCC was 18.7 with 95% confidence interval (7.9–43.9). If cases with only a small isolated focus of trabecular features were discounted, the sensitivity decreased to 68.4%, but the specificity rose to 93.2%, and the positive likelihood ratio, negative likelihood ratio, and odds ratio became 10.1, 0.34, and 29.8, respectively. No other histopathologic pattern was found to be more prevalent than trabecular in MCCs.
Merkel cells were first described in 1875 by Friedrich Sigmund Merkel and were coined “touch corpuscles.”4 In 1972, Cyril Toker reported “primary neuroendocrine carcinoma,”5 and Toker coined the term “trabecular carcinoma.”2,5
The incidence of MCC in the United States has been increasing. It tripled between 1986 and 2001 and now, approximately 1500 new cases are diagnosed each year.6,7 Clinically, MCC appears as a reddish-blue papule or nodule of new onset.8 Clinical features of MCC can be summarized in the mnemonic AEIOU (asymptomatic, expanding rapidly, immunosuppression, older than 50 years, ultraviolet-exposed/fair skin).8
Typically, MCC is histologically composed of SRBCs with scant cytoplasm and tightly packed/molded nuclei in sheets or a trabecular array. Nuclear molding, apoptotic cells, and mitoses are often present. Characteristic immunohistochemical staining patterns include cytokeratin 20 with a paranuclear dot-like staining, and positive staining with low molecular weight keratin, neurofilament, neuron-specific enolase, CD56, and epithelial membrane antigen.9 Neuroendocrine stains such as synaptophysin and chromogranin are typically positive as well.10 Negative staining for thyroid transcription factor-1 (TTF-1) and CD45/lymphocyte common antigen helps to exclude small cell lung carcinoma and lymphoma.9 The presence of Merkel cell polyomavirus can also be used as a differentiating feature.11
Our findings suggest that the presence of a trabecular pattern, particularly if present in more than small foci, correlates with a diagnosis of MCC and helps to exclude other tumors. This could potentially reduce the reliance on broad immunohistochemical panels, and may be especially useful in cases with aberrant patterns of immunostaining. Because small cell lung carcinoma can demonstrate a trabecular pattern, TTF-1 staining may still be prudent.
1. Calonje E, Brenn T, Lazar A, et al. McKee's Pathology of the Skin. 4th ed. Edinburgh, United Kingdom: Saunders; 2012.
2. Becker JC, Schrama D, Houben R. Merkel cell carcinoma
. Cell Mol Life Sci. 2009;66:1–8.
3. Larbcharoensub N, Pongtippan A, Pangpunyakulchai D, et al. Sister Mary Joseph nodule caused by metastatic desmoplastic small round cell tumor: a clinicopathological report. Mol Clin Oncol. 2016;5:557–561.
4. Merkel F. Tastzellen and Tastkoerperchen bei den Hausthieren und beim Menschen [in German]. Arch Mikrosc Anat. 1875;11:636–652.
5. Toker C. Trabecular carcinoma of the skin. Arch Dermatol. 1972;105:107–110.
6. Hodgson NC. Merkel cell carcinoma
: changing incidence trends. J Surg Oncol. 2005;89:1–4.
7. Lemos B, Nghiem P. Merkel cell carcinoma
: more deaths but still no pathway to blame. J Invest Dermatol. 2007;127:2100–2103.
8. Wang TS, Byrne PJ, Jacobs LK, et al. Merkel cell carcinoma
: update and review. Semin Cutan Med Surg. 2011;30: 48–56.
9. McNiff JM, Cowper SE, Lazova R, et al. CD56 staining in Merkel cell carcinoma
and natural killer-cell lymphoma: magic bullet, diagnostic pitfall, or both? J Cutan Pathol. 2005;32:541–545.
10. Saini AT, Miles BA. Merkel cell carcinoma
of the head and neck: pathogenesis, current and emerging treatment options. Onco Targets Ther. 2015;8:2157–2167.
11. Feng H, Shuda M, Chang Y, et al. Clonal integration of a polyomavirus in human Merkel cell carcinoma
. Science. 2008;319:1096–1100.