1. Merkel Cell Carcinoma: Introduction
Merkel cell carcinoma (MCC) is a rare and highly aggressive neuroendocrine cancer of the skin. MCC is thought to originate from the nerve-associated Merkel cell touch receptors, which are in the layer of basal cells at the deepest portion of the epidermis [
1]. With a 33% mortality rate, MCC is deadlier than other more common forms of skin cancer [
2]. MCC is associated with local recurrence, regional metastasis, and distant metastasis to the brain, bone, liver, lung, and heart [
3]; it is the second most common cause of skin cancer death after melanoma, with an estimated cause-specific death rate of 0.6 per 100,000 persons (in 2006) [
4,
5]. In addition, the death rate of MCC has been rising rapidly—in the past two decades, deaths due to MCC have increased more than 3-fold [
6]. In 1992, cytokeratin 20 (CK-20) was identified as a key diagnostic immunohistochemical marker for MCC [
7,
8]. Before diagnosis using CK-20, arduous and intensive electron microscopy was required to accurately distinguish MCC from other carcinomas [
9]. Thus, many earlier cases of MCC plausibly were misdiagnosed or unreported. In conjunction with the increasingly aging population, the discovery of CK-20 may partially explain the surge in MCC cases in recent decades [
10].
The primary risk factors for MCC include older age, white race, exposure to UV radiation, and immunosuppression [
11,
12,
13]. With an average age at onset of approximately 74 to 76 years, MCC in the United States is more common among elderly white men with decreased immune function [
14]. The higher incidence in non-Hispanic whites may be attributable to their relatively lower amounts of protective skin pigment [
5]. Notably, a Danish study reported a higher MCC incidence rate among females, reflecting the higher proportion of women in these older populations [
15]. The high mortality rate of MCC may be partially explained by the advanced age at onset in combination with decreased immune function of the elderly as a function of aging [
16,
17].
In this paper, we review the role of immunosuppression in the development of MCC. The data highlighted in this report include research on MCC in the immunosuppressed setting from the United States, Australia, and European countries. Given that most of the MCC literature is limited to white patients, future research on this tumor in other populations is warranted. In a retrospective study of MCC from mainland China, none of the 22 patients identified with MCC had human immunodeficiency virus (HIV)/AIDS, lymphoma, or any other common forms of immunosuppression [
18]. Authors from areas with low inherent incidence rates of nonmelanoma skin cancers (e.g., Korea, Japan) also report few cases of MCC, making it difficult to ascertain relationships in these areas of the world [
19,
20,
21].
Most reports of MCC in the setting of immunosuppression are relatively brief compared with well-described skin cancer counterparts such as malignant melanoma. Both MCC and malignant melanoma share risk factors (e.g., UV exposure, immunosuppression), but melanoma has a higher incidence rate and MCC generally has a worse overall prognosis [
14,
15,
22]. A large population-based study found a significant 3-fold increased risk (95% CI, 1.74–4.95) of MCC development in patients with malignant melanoma [
22]. The immunosuppressed microenvironment of malignant melanoma may be conducive for development of secondary malignancies, including MCC [
23].
1.1. MCC and Immunosuppression
Immunosuppressed individuals constitute approximately 10% of the MCC patient population [
24]. Immunosuppression increases the risk of MCC and appears to be associated with a worse prognosis [
25]. Systemic immune suppression is associated with poorer MCC survival, independent of stage at presentation (
i.e., local, regional, distant), with a 3-year MCC-specific survival nearly half that of nonimmunosuppressed individuals [
25]. Likewise, other research has indicated that MCC survival is linked with strong intratumoral immune responses [
26]. Reports have described MCC onset at an earlier age in immunosuppressed patients, particularly organ transplant recipients (OTRs) and patients with HIV/AIDS [
13,
27]. Another study did not report a statistically significant difference in age of MCC onset in immunosuppressed patients, likely because the majority of these individuals had chronic lymphocytic leukemia, a cancer typically presented in older adults. [
24]. Immunosuppression-related risk factors for MCC development include UV-induced immunosuppression, organ transplantation, HIV/AIDS, autoimmune diseases, and lymphoproliferative disorders (LPDs) [
13,
22,
28,
29].
1.2. Merkel Cell Polyomavirus
The Merkel cell polyomavirus (MCPyV) was discovered in 2008, suggesting a link between MCC and immune suppression. Like all polyomaviridae, MCPyV is a small, circular, nonenveloped, double-stranded DNA virus that integrates into the tumor genome in a clonal manner. Although most individuals are naturally exposed to MCPyV, very few have MCC; therefore, other factors such as an immunosuppressed state likely contribute to viral integration, mutagenesis, and carcinogenesis [
30]. Other tumors that have viral origins include Kaposi sarcoma and Burkitt lymphoma; these also have a higher incidence in immunodeficient patients [
31].
The large T-antigen expressed in MCC tumors in the truncated form inhibits retinoblastoma tumor suppressor genes and promotes cell division. Large T-antigen mutations are characteristic of MCPyV-positive MCC tumor cells [
32]. MCPyV is present in 75% to 80% of MCCs, whereas it was identified in 16% of control tissue samples [
33,
34]. MCPyV has been detected in other cutaneous tumors, but in some reports, the evidence indicating the presence of MCPyV was mixed [
33,
34,
35,
36,
37].
MCC development appears to have two distinct etiologic pathways: MCPyV mediated and non-MCPyV mediated [
38,
39]. Analyses of primary and metastatic MCC tumors show that an infiltration by T cells (CD8
+, CD4
+, and CD3
+) and increased immune response transcripts are associated with tumor regression and improved prognosis [
25,
40,
41]. The high intratumoral T-cell counts are even associated with better prognosis in MCPyV-negative MCC [
42]. The effect of MCPyV on the clinical course of patients with MCC is uncertain. Two studies have shown that individuals with MCPyV-positive tumors have a better prognosis than those with MCPyV-negative tumors [
39,
43]. Accordingly, MCPyV-negative MCCs specifically harbor mutations in the tumor suppressor gene TP53 that are linked to worse outcomes in cancers because of resistance to cancer therapies [
44,
45]. In contrast, other studies have observed no statistically significant survival differences between patients with MCPyV-positive and MCPyV-negative tumors [
46,
47]. These conflicting findings may be explained by differences in geographical origin and timing of patient cohort collection, chance, or technical factors such as different MCPyV assays.
MCC develops despite the presence of both humoral and cellular responses against MCPyV infection. Antibodies against MCPyV viral capsid proteins, particularly immunoglobulin G, are detected in up to 80% of healthy adults (>50 years old) [
48]. MCC tumors notably do not express viral capsid proteins. Some have proposed that immunocompromised patients with MCC have higher antibody titers and viral loads [
49]. Humoral immune responses may promote antitumor activity, although its contribution is unproven [
26]. Meanwhile, cellular responses target MCPyV through virus-reactive CD8
+ and CD4
+ T cells [
50]. Comparative studies that have investigated MCPyV and other polyomaviruses in immunosuppressed individuals such as OTRs have found relatively low levels of MCPyV and were inconclusive [
31,
51]. Better understanding of the pathogenesis of MCPyV and its interaction with the immune response could enable more effective treatments for MCC.
2. MCC and UV-Induced Immunosuppression
UV radiation is a known risk factor for many skin cancers [
52]. It can cause mutagenic, carcinogenic, and immunosuppressive effects [
53,
54]. UV radiation promotes DNA damage, induces the immunosuppressive cytokines interleukin-1 and tumor necrosis factor α, and generates reactive oxygen species; the decrease in DNA repair and subsequent immunosuppression contribute to carcinogenesis [
55]. Solar radiation is a major risk factor for MCC. Accordingly, MCC typically develops in sun-exposed skin surfaces, notably the head and neck, followed by the extremities [
4,
5].
UV-B rays are less prevalent than UV-A rays, but they are much more intense and destructive [
56]. UV-B induces mutations in the tumor suppressor p53 and Ha-ras genes, which increase the risk of cancer [
57]. The UV-B index, the quantitative geographic measurement of radiation exposure, is positively associated with MCC incidence across US cities [
58]. This observation is supported by a study showing greater prevalence of non-MCPyV-mediated MCC in Australia than in North America [
59], possible because of the increased sun exposure in Australia.
In 2010, Mogha
et al. [
60] demonstrated a molecular link between MCC and sun exposure. The mRNA transcript of MCPyV small t antigen had a dose-dependent increase after UV radiation (in the form of solar-simulated radiation). MCC patients with MCPyV VP1 antibody titers >10,000 had a significantly better prognosis than controls [
61]. In addition, progression-free survival was observed in groups with higher antibody levels. Further, UV-induced mutations (
i.e., pyrimidine dimers) impair helicase and prevent replication, which in turn could promote the survival of MCPyV. Specifically, there is a high frequency of pyrimidine dimer substitutions in large T-antigen mutations [
32].
UV-A has also been reported to induce MCC. This long-wavelength UV corresponds to deeper penetrance beyond the epidermis into the dermis and is a significant contributor to UV-induced immunosuppression [
62]. In a nationwide US study (1975–1998), two of 1,380 patients with psoriasis treated with methoxsalen (psoralen) and UV-A photochemotherapy had MCC develop more than 20 years later [
63]. Calzavara-Pinton
et al. [
64] also identified two immunosuppressed patients with MCC development after high-dose UV-A1 (320–400 nm) phototherapy. On a molecular level, analyses of MCC cell lines showed oxidative damage induced by environmental factors, especially UV-A, which resulted in chromosomal imbalances [
57].
6. MCC and Autoimmune Diseases
Autoimmune disorders impair the natural immune response, creating an environment that makes an affected individual particularly vulnerable to secondary malignancies [
129]. Several autoimmune disorders have been linked to an increased incidence of MCC (
Table 4). In a Swedish retrospective registry analysis, Hemminki
et al. [
11] reported that the risk of MCC was significantly increased after ankylosing spondylitis (SIR, 15.62; 95% CI, 2.95–46.25), inflammatory bowel disease (SIR, 4.02; 95% CI, 1.82–7.66), and Crohn disease (SIR, 4.38; 95% CI, 1.38–10.31). Meanwhile, the risk of MCC was not significantly different after rheumatoid arthritis (SIR, 2.42; 95% CI, 0.96–5.01). Other studies observed increased risk of MCC in patients with rheumatoid arthritis, which was perhaps at least partially attributable to immunosuppression after chronic use of systemic corticosteroids [
12,
73]. The use of tumor necrosis factor inhibitors in the treatment of systemic autoimmune diseases may further exacerbate immunosuppression [
97]. MCC has been described after the administration of the anti-tumor necrosis factor therapy rituximab [
130].
The majority of information on autoimmune disorders and MCC is scattered among case reports, which makes it difficult to establish associations [
131,
132]. However, MCC development has been documented in patients with autoimmune hepatitis, systemic lupus erythematosus, Behçet syndrome, and chronic sarcoidosis and recieving immunosuppressive medications [
131,
133,
134,
135,
136]. Paraneoplastic autoimmune syndromes, including Lambert-Eaton myasthenic syndrome, have also been linked to MCC, although a relationship with these syndromes and the clinical outcome of MCC is not well established [
137].
Table 4.
Summary of relative risks of MCC occurring after various forms of immunosuppression.
Table 4.
Summary of relative risks of MCC occurring after various forms of immunosuppression.
Reference | Study Type (Country) | Mode of Immunosuppression | Findings a |
---|
Lanoy et al. 2010 [12] | Registry analysis. (USA) | Autoimmune disease (RA) | 79/1977 MCC patients had RA develop
Odds ratio, 1.39 (95% CI, 1.10–1.75) |
Cirillo et al. 2012 [71] | Single-center analysis. (Italy) | Autoimmune disease (RA) | 3/48 patients with RA had MCC develop during immunosuppressant corticosteroid treatment Statistical significance not determined |
Hemminki et al. 2012 [11] | Registry analysis (Sweden) | Autoimmune diseases (AS, IBS, CD, RA) | 3/112,541 patients with AS had MCC (SIR, 15.62; 95% CI, 2.95–46.25) 9/68,915 patients with IBS had MCC (SIR, 4.02; 95% CI, 1.82–7.66) 5/35,422 patients with CD had MCC (SIR, 4.38; 95% CI, 1.38–10.31) 7/71,963 patients with RA had MCC (SIR, 2.42; 95% CI, 0.96–5.01). |
Engels et al. 2002 [13] | Registry analysis (USA) | AIDS | 6/30,9365 patients with AIDS had MCC (relative risk, 13.4; 95% CI, 4.9–29.1) |
Lanoy et al. 2009 [125] | Registry analysis (USA) | AIDS | 17/497,142 male patients with AIDS had MCC (SIR, 11; 95% CI, 6.3–17) |
Lunder and Stern. 1998 [63] | Multicenter analysis (USA) | UV-A phototherapy + psoralen | 3/1380 (0.2%) patients with psoriasis receiving UV-A + psoralen had MCC develop ~100-fold increase |
Calzavara-Pinton et al. 2010 [64] | Retrospective analysis (Italy) | UV-A phototherapy | 2 immunosuppressed patients had MCC develop after UV-A1 phototherapy treatment
Statistical significance not determined |
Sahi et al. 2012 [29] | Registry analysis, 1994–2009 (Finland) | Statins (HMG-CoA-reductase inhibitors) | 50/454,935 statin users had MCC develop
Age ≤60 y: SIR, 3.16; 95% CI, 0.65–9.23
Age 60–74 y: SIR, 1.94; 95% CI, 1.23–2.90
Age ≥75 y: SIR, 0.89; 95% CI, 0.67–0.92 |
Howard et al. 2006 [22] | Registry analysis, 1986–2002 (USA) | Malignant melanoma | 16/70,604 patients with melanoma had MCC (SIR, 3.05; 95% CI, 1.74–4.95) |
7. Recommendations
Current treatment of MCC is assessed on the basis of the clinical stage of the tumor at presentation [
138]. Biopsies, including the low-risk sentinel lymph node biopsy and CT/PET scans are used as diagnostic tests to assess the stage of MCC. While sentinel lymph node biopsy is a reliable staging procedure in many cancers including melanoma, its value in MCC patients remains unclear [
139]. Conventional routes of MCC treatment include surgery, radiotherapy, and chemotherapy. Surgical excision is the first-line treatment for localized MCC, often in combination with radiotherapy or chemotherapy (or both) for regional or distant metastases [
140,
141]. The effect of chemotherapy on the outcome of MCC is controversial; some evidence even suggests that certain forms of chemotherapy may be detrimental [
142]. Autologous peripheral blood stem cell transplantation has also been proposed as a mode of treatment, though its effects on MCC are contradictory. One case report achieved a 6-month complete remission of metastasized MCC using a combination of autologous peripheral blood stem cell transplantation and high-dose chemotherapy [
143]. Meanwhile, another case report found an onset of fatal, metastatic MCC following autologous peripheral blood stem cell transplantation for non-Hodgkin’s lymphoma [
144]. Further research needs to be conducted in order to determine the effect of stem cell transplantation on MCC.
Multidisciplinary management is key in treating all patients with MCC, especially those with immunosuppression. Annual skin surveillance examinations should be implemented, particularly for patients with solid organ transplantation, HIV/AIDS, a history of LPDs, or other iatrogenic forms of immunosuppression. A lower threshold for biopsy of new or changing skin lesions, especially those in sun-exposed regions, is warranted [
141]. Furthermore, education and awareness of MCC is crucial for both clinicians and at-risk individuals. Clinicians and immunodeficient individuals may also benefit from remembering the acronym AEIOU: asymptomatic/lack of tenderness, expanding rapidly (<3 months), immunosuppression, older than 50 years, and UV-exposed site location. A study of 195 patients showed that 89% met at least three of the AEIOU criteria, 52% met four or more, and 7% met all five [
24]. Moreover, it is important for immunosuppressed patients to perform monthly self-examinations, avoid tanning beds, and see a dermatologist annually. UV (sun) protection is essential for immunosuppressed patients and should be part of their daily routine. In addition, vitamin D supplementation may be beneficial, especially for OTRs [
72,
145].
Decreasing immunosuppression by modifying immunosuppressive treatments may be beneficial when deemed safe. The administration of interleukin-2 in combination with HAART to bolster T-cell immunity and prevent metastatic spread of MCC in HIV-infected individuals has been suggested [
127]. The use of statins, which are immunosuppressive agents, has also been linked to a 3-fold increased risk of MCC development in individuals younger than 60 years [
29,
146].
Immunotherapy has potential use in MCPyV-positive cases. Proposed novel treatments include the use of immune-regulatory cytokines, adoptive cellular infusions, topical immune modulators, and DNA vaccines [
26]. Several studies on the application of antiviral type I interferons and tumor necrosis factor on patients with MCC are promising, but more clinical studies are needed [
147,
148,
149]. More research on MCPyV and its role in MCC development in patients with a history of autoimmune diseases, HIV/AIDS, and organ transplantation would lend insight on MCC pathogenesis and origin.
8. Conclusions
Immunosuppressed patients have greater risk of malignancies, including MCC, and also have increased risk of worse cause-specific survival in the setting of MCC. Long-term immunosuppression appears to increase the risk of MCC development and also results in earlier age at onset, more aggressive course, and a worse prognosis. Most reports of MCC in immunosuppressed patients are brief. The limited numbers of MCC cases in the setting of immunosuppression make associations difficult to ascertain. Dermatologic surveillance is crucial in severely immunosuppressed individuals who have undergone organ transplantation or have a history of LPDs, HIV/AIDS, or autoimmune diseases. Immunosuppressive agents appear to increase susceptibility to MCC development. Better understanding of how MCPyV modulates the immune system will be beneficial for developing effective immunotherapies for MCC. Additionally, better understanding of the pathogenesis, prognosis, and management of MCC is necessary for the future direction and recommendations for treatment, surveillance, and outcomes in immunocompromised patients.