Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Epidemiology of Merkel Cell Polyomavirus and Involvement in Human Cancer

Merkel cell polyomavirus (MCPyV) is the so far only human polyomavirus that is linked to the etiology of a human cancer, namely to Merkel cell carcinoma (MCC). MCPyV was first described in 2008, when a previously unidentified polyomavirus was found in 8 of 10 of MCC tumours [1]. The discovery of MCPyV was made using a novel methodology called digital transcriptome subtraction, developed by the same research team [2]. MCPyV was demonstrated to be monoclonally integrated in the human genome in MCC [1].

Presence of the genome of MCPyV in MCC tumors from around the world has been confirmed by several independent research groups (Table 15.1). Most studies have found an MCPyV detection rate in MCC of around 70–80 %, similar to the original report by Feng et al. [1]. MCPyV infection is also common on healthy human skin [3–5] and can be found all over the body. However, on skin sites other than the MCC tumors, the viral loads of MCPyV are typically more than 60 times lower [6]. The detection rate of MCPyV DNA in tissue samples depends not only on the study population’s geographical location but also on the tissue sample storage conditions and PCR type and primer-set involved. The detection rate is higher, when fresh-frozen samples are used rather than formalin-fixed paraffin-embedded samples [7–11].

Table 15.1 Merkel cell polyomavirus DNA in Merkel cell carcinomas and in control tissues
Full size table

The seroprevalence of MCPyV, as measured using ELISA or Luminex assays detecting MCPyV-specific IgG antibodies, is uniformly high (ranging from 46 to 88 %) in the populations studied [12–17]. The serological assays are not sufficiently standardized to allow conclusions on whether variability in seroprevalences are attributable to methodological aspects of the serology or reflect true epidemiological differences between the populations studied. However, it is clear that MCPyV infection occurs already in early childhood [14, 16, 18]. Kean et al. reported a 20.5 % MCPyV seroprevalence among 1–5-year-old children [14], Tolstov et al. a 43 % MCPyV seroprevalence among children 2–5 years [16], whereas we have found a 32 % MCPyV seroprevalence among Swedish children 1–12 years [19]. Early school age seems to be an important age for McPyV antibodies (seroconversion). Chen et al. found that while children 1–4 years of age had a seroprevalence of 9 %, the seroprevalence was 35 % among children age 4–13 years of age. Also, 33 % of MCPyV-seronegative children at the age up to 3 years were found to have seroconverted when tested again 5–8 years later [18]. Rates of acquisition appear to be substantial also among adults. In a cohort of 117 MCPyV-seronegative males, 31 seroconverted over a 4-year follow-up period, corresponding to a 6.6 % annual seroconversion rate. Once seroconversion has taken place, the MCPyV IgG levels remain detectable up to 25 years after the exposure. No signs, symptoms, or routine diagnostic test results were associated with MCV infection, indicating that the primary infection is mostly asymptomatic [20].

The exact mode of MCPyV transmission is unclear. Intrauterine transmission does not appear to occur, as no MCPyV DNA was detectable in miscarried or aborted fetuses [21]. MCPyV DNA is common on the human skin [3, 7] and frequent presence of MCPyV DNA on environmental surfaces that has been in contact with human skin suggests that virus may be shed from the surface of infected skin. In a study of 60 environmental surface samples, 45 (75.0 %) were positive for MCPyV DNA and in a few of these samples the viral DNA was even protected from DNase degradation, suggesting that it represented viral DNA encapsidated inside infectious virus particles [22]. Apart from the skin, MCPyV DNA has also been found at high levels in the upper aerodigestive tract, in the digestive system, and in the saliva, but was less frequently found in lung and genitourinary system samples [6]. However, MCPyV has been found in the lower respiratory tract [23] as well as on the tonsils [18]. Presence of MCPyV DNA appears to accumulate with age, being more frequent in adults than in children [24]. MCPyV can also persist in inflammatory monocytes and spread along the monocyte migration routes [25]. MCPyV has also been found in the lymphatic system [26]

The mechanism whereby MCPyV can cause MCC has been suggested to be a 2-step process involving a random viral integration into the host genome and a mutation in the Large T antigen (LT) region that eliminates viral replication capacity but retains the Rb-binding capacity [27, 28]. As a majority of the population has been infected with MCPyV, but MCC is very rare disease, events in addition to infection itself (e.g., UV-exposure or immunosuppression) must be essential for the development of MCC. Patients with MCPyV-positive MCC generally have higher levels of serum IgG antibodies specific for the MCPyV virion than do healthy individuals [15, 16, 29] and elevated MCPyV virion antibodies are detectable many years before the MCC tumor is diagnosed [19]. Presence of high MCPyV-specific antibody levels is strongly associated with high MCPyV viral loads [13, 30], suggesting that a persistent presence of a high-viral load MCPyV infection is an important step in the carcinogenic pathway [9, 16]. Serum IgG levels against the MCPyV T antigens are elevated in 40.5 % of patients with MCC, but the levels of antibodies to the T antigen appear to follow the extent of the growing tumor [31]. Although this implies that antibodies to the T antigen are more an effect of the tumor than a marker of tumor etiology, MCPyV T antigen antibodies could represent a clinically useful tumor marker to indicate disease status. In contrast, the MCPyV-neutralizing antibodies to the virion are, especially in females, elevated already decades before the tumor arises, indicating that the neutralizing antibodies are markers of an etiologic factor on the causal pathway to the MCC tumor [19].

Epidemiology of MCC and Other MCPyV-Related Diseases

MCC is a rare but highly aggressive neuroendocrine skin malignancy that affects elderly and immunosuppressed individuals. The incidence of primary MCC in the USA is 3.4 per million person-years [32]. The number of MCC cases has increased about threefold during the past 20 years, attributable mostly to the aging population [33, 34]. In Denmark, the MCC incidence between 1995 and 2006 was 2.2 cases per million person-years [35] and in Japan it was1.45 per million [36]. The incidence rates of MCC in Denmark had increased 5.4-fold over the 18 year period from 1986 until 2003 [37]. There is a strong association between MCC and white/fair-skinned individuals, advanced age, and sun exposure. At least in Scandinavia, MCC occurs more often among females than males [35, 38]. MCC incidence is increased in both AIDS and posttransplantation populations and tends to develop at a younger age in these populations [32]. The prognosis is poor and dependent on stage at diagnosis, with 5-year survival rates of patients with localized, regional, and distant disease reported to be 64 %, 39 %, and 18 %, respectively [39]. Clinically, MCC typically presents as a fast growing, nontender, red to violet papule or nodule on sun-exposed areas of the skin, especially in the head and neck region. As also other skin malignancies present with such features, histological examination of biopsy specimens is required for the diagnosis [40]. MCC is not only a skin malignancy but can appear also in numerous other sites such as lymph nodes, oral cavity, breast, vaginal walls, and salivary glands. It has been demonstrated that also extracutaneous MCC harbor polyomavirus DNA [41].

Patients diagnosed with MCC are at increased risk of a second cancer, particularly for other skin cancers [35, 38, 42, 43]. Some studies have also found increased risks for and for chronic lymphocytic leukemia (CLL) and multiple myeloma [42, 43], but this has not been confirmed in all populations [35, 38, 42, 43]. MCC has also been found to be increased as secondary cancer after CLL, Hodgkin, and non-Hodgkin lymphomas [35].

Squamous cell carcinoma (SCC) of the skin can occur simultaneously with MCC [44–46]. MCC arising in association with a concomitant SCC of the skin seem to be negative for MCPyV DNA [47]. One study found MCPyV DNA in pure MCC, but none of four tumors with a combined MCC + SCC morphology were positive for MCPyV (P = 0.001) [48]. However, another study of two mixed MCC–SCC lesions found MCPyV DNA in both of them [49]. Approximately 15 % of SCCs of the skin from immunocompetent individuals have been reported to contain MCPyV DNA [13, 50, 51]. After immunosuppression, MCPyV is more reactivated among patients with non-melanoma skin cancers compared to healthy individuals [52]. MCPyV DNA has also been detected in SCC tumor tissues. Rollison et al. found that 55 (38 %) of 145 SCC tumors were MCPyV DNA positive and that MCPyV DNA-positive SCC cases were more frequently MCPyV seropositive [53].

MCPyV DNA has also been reported to be associated with skin cancers developing among subjects with the rare hereditary immunodeficiency disease epidermodysplasia-verruciformis (EV). These tumors also contain human papillomaviruses (HPV) and it was suggested that MCPyV and HPV may act as synergistic factors in the development of these tumors [54].

There are inconclusive data on MCPyV in other diseases. MCPyV was detected in 5/19 CLL cases, but only in 13.4 % of healthy controls (P < 0.04) [55]. In situ hybridization studies have found MCPyV DNA in CLL cells [56]. A very low level of MCPyV DNA in was found in 33 % of CLL/ small lymphocytic lymphomas [57].

It is clear that not all MCC cases contain MCPyV DNA, suggesting that MCC may exist as two etiologically distinct subtypes: virus-related and virus-unrelated [27]. MCPyV-negative MCC have more irregular nuclei (P < 0.001) and more abundant cytoplasm (P = 0.001) than MCPyV-positive MCC, which have uniform, round nuclei and scant cytoplasm [48]. MCPyV-positive and -negative MCCs also have different prognosis. MCC patients with high-viral load MCPyV DNA have a better prognosis than MCC tumor with little or no viral DNA [9, 58, 59]. However, two reports did not find any association between MCPyV status and prognosis of MCC [60, 61].

Therapeutics Against MCPyV-Associated Cancer and Prevention of Infection

MCPyV infection activates the BIRC5 gene, which encodes the survivin oncoprotein, and is highly upregulated by LT sequestration of Rb. A small molecule survivin inhibitor (YM155) has been tried as therapy for MCPyV-MCC and found to prolong survival of mice bearing MCC tumors [62].

A candidate prophylactic vaccine based on MCPyV virus-like particles (VLPs) has been found to, in mice, elicit antibody responses that robustly neutralize MCV reporter vectors in vitro, suggests that a VLP-based vaccine could be effective for preventing MCPyV infection [15]. There is also a candidate DNA vaccine containing the MCPyV Large T antigen gene (aa1-258) (pcDNA3-LT) that has been found to generate antitumor effects against a transplantable MCPyV LT carrying murine melanoma cell line in vaccinated C57BL/6 mice [63].

Conclusion

In just 5 years since the discovery of MCPyV, intensive research has established a basic knowledge regarding the epidemiology, immunology, and pathogenesis of the virus and clarified that this virus has an etiological role in a majority of MCC cases. The list of infections known to cause cancer in man has been continuously expanding and the MVPyV demonstrates that modern research technology now can enable discovery, validation, and basic characterization in just a few years.