Elsevier

Human Pathology

Volume 42, Issue 10, October 2011, Pages 1385-1390
Human Pathology

Original contribution
Immunohistochemistry for Merkel cell polyomavirus is highly specific but not sensitive for the diagnosis of Merkel cell carcinoma in the Australian population

https://doi.org/10.1016/j.humpath.2010.12.013Get rights and content

Summary

Recent studies have demonstrated a high frequency of detection of Merkel cell polyomavirus in Merkel cell carcinoma. However, most of these studies are from European or North American centers that have relatively low sun exposure and may have a higher incidence of virus-driven oncogenesis compared with the highly sun-exposed but predominantly fair-skinned Australian population. We performed immunohistochemistry for Merkel cell polyomavirus on 104 cases of Merkel cell carcinoma and 74 cases of noncutaneous small cell–undifferentiated carcinoma from 3 major Australian centers. Nineteen (18.3%) cases of Merkel cell carcinoma showed positive staining for Merkel cell polyomavirus versus 1 (1.3%) of small cell–undifferentiated carcinoma. All 15 cases (14.3%) of Merkel cell carcinoma with areas of mixed squamous differentiation showed negative staining. We found positive staining in only 3 (7.7%) of 39 Merkel cell carcinoma from the head and neck (the most sun-exposed area) versus 16 (24.6%) of 65 of tumors from other sites (P < .05). Our findings support the concept of a Merkel cell polyomavirus–driven and a non-Merkel cell polyomavirus–driven (primarily sun-dependent) pathway in Merkel cell carcinoma carcinogenesis, with the latter being significantly more frequent in Australia and in mixed squamous–Merkel cell carcinoma (which is also more frequent in Australia). Although immunohistochemistry for Merkel cell polyomavirus seems to be highly specific in all populations, the low incidence of Merkel cell polyomavirus–positive Merkel cell carcinoma in a highly sun-exposed population limits its diagnostic utility in this setting.

Introduction

Merkel cell carcinoma (MCC) is a rare primary neuroendocrine carcinoma of the skin. MCC is a biologically aggressive tumor that has shown an increasing incidence in recent years, with a 3-fold rise observed over a 15-year period in North America [1]. MCC occurs most commonly in the sun-exposed skin of older white patients [1], [2], [3], [4], [5]. There is also a strong epidemiologic association between MCC and immunosuppression [2], [3], [5], [6]. Histologically, MCC is characterized by trabeculae, sheets, and nest of small cells with scant cytoplasm, a salt and pepper chromatin pattern with nuclear moulding, and frequent mitoses and apoptotic bodies (Figs. 1). It may occur in intimate association with an in situ or invasive squamous cell carcinoma. There are no firm data on the rate of mixed squamous-MCC differentiation, but it has been estimated as occurring in 5% to 10% of all MCC [7], [8]. MCC can be morphologically identical to small cell–undifferentiated carcinoma (SCUC) of any site, making the role of ancillary tests such as immunohistochemistry valuable in the differentiation between the 2 entities [2].

Recently, Feng et al [9] described a new human polyomavirus that they named Merkel cell polyomavirus (MCPyV). They strongly implicated MCPyV in the pathogenesis of MCC by using digital transcriptome subtraction to detect MCPyV sequences in 8 (80%) of 10 MCC. In 6 of these cases, viral DNA was integrated in the tumor in a clonal pattern, suggesting that MCV infection and integration preceded neoplasia. Subsequently, various studies primarily based in European and North American centers have shown the frequency of detection of MCPyV DNA using polymerase chain reaction (PCR) to be in the range of 40% to 100% [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28] (Table 1). A small study of Australian patients suggested that the prevalence may be only 24% in this country [13]. This study examined a total of 37 MCC tumor samples from North America and Australia. MCPyV DNA was detected in 16 (43%) of 37 cases. Of 16 tumors originating from North America, 11 (69%) were positive for MCPyV, compared with only 5 (24%) of 21 specimens from Australia. The authors propose that this may be because greater sun exposure in Australia results in a non-MCPyV–driven pathway of carcinogenesis being relatively more frequent.

To date, most studies investigating the prevalence of MCPyV in MCC have used PCR to detect viral sequences. However, routine PCR is beyond the capabilities of most surgical pathology laboratories. Furthermore, PCR may be so sensitive that it detects low-level carrier states that may not be associated with MCC [28], [29]. Recently, a mouse monoclonal anti-MCPyV antibody CM2B4 that recognizes the MCPyV T antigen was developed [29]. This antibody has been shown to be highly sensitive and specific for the presence of a high-viral copy number in cells. The antibody is commonly negative in tissue with a low-viral copy number [7], [29]. Therefore, it has been suggested that this antibody would be useful in the differential diagnosis of MCC versus SCUC of other primary sites and of distinguishing MCPyV-driven tumors from those that arise through a different pathway [7], [29].

Using CM2B4, we sought to investigate whether immunohistochemistry has diagnostic value in the Australian population given the initial suggestion that MCPyV virus may be a less important cause of neoplasia in this highly sun-exposed and predominantly fair-skinned population. We analyzed the efficacy of immunohistochemistry for MCPyV in the differential diagnosis of MCC from SCUC of other sites, primarily of lung origin. We went on to investigate whether MCPyV-positive tumors demonstrated different morphologic or clinical features to MCPyV-negative MCC, particularly whether they were less likely to occur in more heavily sun-exposed areas of the body and to be associated with mixed squamous differentiation.

Section snippets

Materials and methods

All cases of MCC from 3 major teaching hospitals in Sydney, Australia, for which tissue was available in paraffin blocks, were identified. These cases were identified by searching the departments' computerized databases for a minimum of 10 years (Royal North Shore Hospital, 12 years; Westmead Hospital, 16 years; St Vincent's Hospital, 10 years). The cases were retrieved and reviewed by an experienced pathologist (A.G.) to confirm the diagnosis of MCC based on combined clinical, morphologic, and

Results

There were 104 patients with MCC for whom tissue was available from the 3 centers. Forty-six (44.2%) patients were female and 58 (55.8%) were male, with a mean age of 73.8 years. Thirty-nine tumors (37.5%) were located on the head and neck, 13 (12.5%) on the upper limbs, 15 (14.4%) on the lower limbs, 9 (8.7%) on the trunk, and 25 (24.0%) were from other locations. The site was unknown for 3 (2.9%) cases (Table 2). Fifteen (14.3%) of the MCC tumors showed at least focal in situ or invasive

Discussion

The recent description of MCPyV by Feng et al [9] has been followed by multiple studies that have supported the notion of a strong epidemiologic link between MCPyV and MCC. Most of these studies have shown rates of MCPyV between 60% and 90%. However, these studies have been based in North American and Europe–areas with significantly lower rates of sun exposure–driven carcinogenesis than that in Australia. In our study, only 19 (18.3%) of 104 cases of Australian cases of MCC showed positive

References (29)

  • J. Kaae et al.

    Merkel cell carcinoma: incidence, mortality, and risk of other cancers

    J Natl Cancer Inst

    (2010)
  • K.J. Busam et al.

    Merkel cell polyomavirus expression in Merkel cell carcinomas and its absence in combined tumors and pulmonary neuroendocrine carcinomas

    Am J Surg Pathol

    (2009)
  • H. Feng et al.

    Clonal integration of a polyomavirus in human Merkel cell carcinoma

    Science

    (2008)
  • A. Kassem et al.

    Frequent detection of Merkel cell polyomavirus in human Merkel cell carcinomas and identification of a unique deletion in the VP1 gene

    Cancer Res

    (2008)
  • Cited by (99)

    View all citing articles on Scopus

    Disclosure/conflict of interest: The authors declare no conflict of interest.

    View full text