Merkel cell polyomavirus sequences are frequently detected in nonmelanoma skin cancer of immunosuppressed patients



Recently, a new human polyoma virus has been identified in Merkel cell carcinomas (MCC). MCC is a highly aggressive neuroendocrine nonmelanoma skin cancer (NMSC) associated with immunosuppression. Clonal integration of this virus which was termed Merkel cell polyoma virus (MCPyV) was reported in a number of MCC. Squamous cell carcinoma (SCC) and basal cell carcinoma (BCC) are also NMSC and are the most frequent cancers in the setting of immunosuppression. A unique group of 56 NMSC from 11 immunosuppressed patients and 147 NMSC of 125 immunocompetent patients was tested for MCPyV by DNA PCR, targeting the Large T Antigen and the structural Viral Protein 1. NMSC included SCC, BCC and Bowen's disease (BD). In addition, normal skin and 89 colorectal cancers were tested. MCPyV specific sequences were significantly more frequently found in NMSC of immunosuppressed patients compared to immunocompetent patients (p < 0.001). In particular BD and BCC revealed a significant increased association of MCPyV of immunosuppressed patients (p = 0.002 and p = 0.006). Forty-seven of 147 (32%) sporadic NMSC were MCPyV positive. Interestingly, 37.5% (36/96) of sporadic BCC of immunocompetent patients were MCPyV positive. No MCPyV was detected within normal skin and only 3 out of 89 of additionally tested colorectal cancers were MCPyV positive. Our data show that MCPyV is a frequently reactivated virus in immunocompromized patients. How MCPyV contributes to the pathogenesis of NMSC, i.e., BD, SCC and BCC, in immunosuppressed patients and in addition, potentially to the pathogenesis of a subset of sporadic BCC needs further investigations. © 2009 UICC

Squamous cell carcinoma (SCC) and basal cell carcinoma (BCC) are nonmelanoma skin cancers (NMSC) and in this order constitute the most frequent cancers associated with immunosuppression in transplant recipients.1–6 According to the steadily increasing number of transplant operations performed each year in the European Union and the United States, post-transplant skin cancer is a leading medical issue in current transplantation medicine. To date a number of risk factors for the increasing number of NMSC under immunosuppression have been identified.2 In addition to SCC and BCC, other NMSC, i.e., sebaceous cancers, cutaneous lymphomas and Merkel cell carcinomas (MCC) occur more frequently in post-transplant patients.7, 8 MCC has been described relatively recently and is a rare but very aggressive malignant neuroendocrine skin cancer of the elderly and immunosuppressed.8–10 Very recently, Feng et al. reported the identification of a new human polyoma virus which was designated Merkel cell polyomavirus (MCPyV) based on its detection in MCC by digital transcriptome subtraction technique.11 They reported the presence of MCPyV in 8 of 10 human MCC and also clonal integration of the viral DNA in 6 of 8 MCPyV-positive MCC. Analyzing the first large number patient cohort of MCC by PCR using diverse oligonucleotides targeting different parts of the MCPyV genome we were able to confirm the presence of MCPyV in 77% of 39 MCC recently.12 In addition, we identified a unique 90 bp deletion of the VP1 gene of MCPyV in one MCC leading to a predicted loss of 28 amino acids.12 The prevalence of MCPyV associated MCC has been confirmed by others to various extent recently.13, 14 Because of the high incidence of MCPyV in MCC and the association of NMSC and immunosuppression we were interested to investigate the presence of MCPyV, comparing immunosuppressed and immunocompetent patients. In the present study, we tested 56 NMSC of 11 immunosuppressed patients, including Bowen's disease (BD), SCC and BCC for the presence of MCPyV and compared the results to sporadic BCC (n = 96), sporadic BD (n = 30) and sporadic SCC (n = 30). In addition, we tested colorectal cancer resection specimens (n = 89) and normal skin obtained from plastic surgery (n = 6).

Material and methods

Study subjects

The study included representative formalin-fixed and paraffin-embedded resection and biopsy specimens of 203 NMSC of 136 patients. Details of clinicopathologic parameters are included in Table I and within the result section. The study was conducted according to the national ethic guidelines. The investigation protocol was approved by the ethical review board of the Institution (nr. 267/08) and patients signed informed consent prior to the surgical procedure.

Table I. Detailed Clinicopathological Data of 11 Patients with 56 NMSC
Pat. IDAgeSexTXDTXImmunsuppressive ThDXLOCALß-GVP1LT3
  1. ISP, immunsuppressed patient; m, male; f, female; TX, transplantation; HTX, heart transplantation; NTX, kidney transplantation; AIG, autoimmune glomerulonephritis; DTX, date of transplantation; n.a., not applicable; DX, diagnosis; BD, Bowen's disease; SCC, squamous cell carcinoma; BCC, basal cell carcinoma; Local, localisation;, ß-G, beta-Globin PCR; LT3, Large T Antigen 3-PCR; VP1, Viral Protein1-PCR; Cyclosp. A, Cyclosporin A; Tacrol., Tacrolimus.

ISP152mHTX2000Cyclosp. ABCCnose+
ISP259mHTX2002Cyclosp. ABCClower leg+++
ISP368fNTX1986Cyclosp. ABCCnose+
ISP452mNTX1995Tacrol., PrednisolonBDthigh+++
ISP571mNTX2004Tacrol., PrednisolonSCCface+
ISP674mAIGn.a.Cyclosp. A, DecortinBCChead++
ISP758mNTX1979Cyclosp. A, SirolimusSCCshoulder++
      BCCupper arm+
ISP868fNTX1991Cyclosp. A, PrednisonBDlower leg+++
ISP965mNTX1991Cyclosp. A, Decortin,SCChead+
      SCClower arm+
ISP1044mNTX1992Tacrol., Cyclosp. ASCChand+
ISP1171fNTX1990 a.1997Cyclosp. ASCCface+++

DNA extraction

First, H&E stained sections of all specimens were reviewed by at least two of the six experienced dermatopathologists (AzH, CD, DN, KT, MW, WW) to select paraffin material containing more than 95% tumour tissue. Two consecutive 5 μm paraffin sections from each specimen were subjected to DNA extraction. In brief, after deparaffination the tissues were lysed by proteinase K overnight (56°C) until complete tissue lysis, and DNA was extracted using the DNeasy Tissue kit® (Qiagen, Hilden, Germany). Purified DNA was measured in a spectrophotometer (Nanodrop, ND1000; PeqLab, Erlangen, Germany) and directly used for PCR.

MCPyV detection by PCR

DNA quality was confirmed by β-globin PCR using the GH20 (5′- GAA GAG CCA AGG ACA GGT AC -3′) and PCO4 (5′- CAA CTT CAT CCA CGT TCA CC -3′) primer set.

PCR was performed with 150 ng of genomic DNA using the AmpliTaq Gold™ (Roche) DNA polymerase in a final volume of 50 μl. For MCPyV detection we used the LT3 and VP1 primer sets as published.11 DNA and PCR mixtures were prepared and kept in separate rooms. Water instead of DNA template was used for PCR negative controls containing all other PCR components.

SYBR green Real-Time PCR

RT-PCR and data analyses were performed in a total volume of 25 μl using 96-well microwell plates and an ABI PRISM 7000 Sequence detector (Applied BioSystems, Foster City, CA). Each reaction contained 6 μl of purified DNA, 12.5 μl QuantiFast SYBR Green PCR Kit (QIAGEN GmbH, Germany), 200 nM of the respective primers LT3 and ß-globin. To reach a total volume of 25 μl per well, DNase-RNase-free distilled water (QIAGEN) was added. To check for amplicon contamination, every run contained at least three controls in which nuclease free H2O was substituted for template:

The reaction was run online at 50°C for 2 min and 95°C for 10 min, followed by 40 cycles at 95°C for 15 s and 60°C for 1 min.

All PCRs were performed in triplicate. To evaluate the efficiency of the amplification, a standard curve was constructed using the threshold cycle (CT) versus 5-fold dilution of one of the positive controls. The specificity of the reaction is given by the detection of the Tms of the amplification products immediately after the last reaction cycle. Results were analyzed with the melting curve analysis software (Applied BioSystems) provided with the ABI PRISM 7000 sequence detector.

Sequence analyses

Selected PCR products were submitted to automated nucleotide sequencing in an ABI 3130XL genetic analyser (ABI, Darmstadt, Germany) to confirm the viral origin of the PCR products. DNA sequences were compared to the reference sequences of the NCBI Entrez Nucleotide database gb|EU375803.1 Merkel cell polyomavirus isolate MCC350 or gb|EU375804.1 Merkel cell polyomavirus isolate MCC339, using the NCBI Blast program.

Statistical analysis

SPSS15.0 software package (SPSS, Inc., Chicago, IL) was used for statistical analyses. MCPyV positivity distributions between tumours of immunosuppressed and immunocompetent patient groups were compared for the result of LT3 or VP1 and LT3 and VP1-PCR by means of the chi square test. p-values <0.005 were considered as statistically significant.


Clinicopathological findings

In the present study, we have analysed 203 NMSC of 136 patients for the presence of MCPyV. Fifty-six NMSC of 11 immunosuppressed patients (mean age: 62 years; range 44–74 years; 8 male and 3 female). Eight patients were immunosuppressed due to renal transplantation, 2 due to heart transplantation and 1 due to autoimmune renal disease (ISP6; Table I). In addition, we analysed 147 NMSC of 125 immunocompetent patients of which 88 were male and 37 female. The mean age of this group was 76.3 years (range: 31–96 years). In detail, the study included 96 patients with BCC (66 male and 30 female) with (according to the current WHO-classification) 78 nodular BCC, 8 superficial BCC, 7 infiltrative BCC, 2 basosquamous and 1 keratotic BCC.15 As precursor lesions of invasive SCC we included 23 BD of 17 patients (11 male and 6 female). The mean age of the BD patients was 76 years (range: 51–96 years). The details of the BD patients and additional 28 SCC of 12 patients (11 male, 1 female) are given in Table II. The mean age of the SCC patients was 80 years (range: 72–92 years).

Table II. Summary of Clinicopathological Data of Immuncompetent Patients with Bowen's Disease (BD) and Squamous Cell Carcinoma (SCC)
DiagnosisPat. IDAgeSexLocalizationB-GLT3VP1
  1. ß-G, beta-globin; LT3, Large T antigen 3; VP1, Viral protein 1; + = positive; − = negative; m = male; f = female.

BDBD172mleft shoulder+
 BD251mright groin+
 BD373fleft thigh+
    right thigh+
 BD495mright ear+
 BD582fright temple+
 BD695fright buccal lat.+
    right buccal++
 BD777mright ear+
 BD887mleft buccal med+
 BD976mright thigh++
 BD1077mleft back of hand+
    left thorax+
 BD1162fleft back of foot+
 BD1282fleft back of hand++
    left back of hand+
 BD1368mleft calf+
 BD1466fleft forehead+
 BD1571mleft lower leg+
 BD1662mright lower leg+
 BD1796mright shoulder+
SCCSCC192mhead left++
 SCC276mright lower leg++
    right thigh+
    left hand+
 SCC392mhead right++
    left ear+
    abdomen left+
    back left+
 SCC475mright back of hand+
 SCC575fleft lower arm+
 SCC680mleft ear+
    left back of hand+
    right ear+
 SCC882mforehead right+
    right ear+
    right buccal++
 SCC1172mleft temple+
    right shoulder+
    left hand+
    left hand++

Overall detection of MCPyV in NMSC of immunosuppressed and immunocompetent patients

We tested the DNA of 203 NMSC of 136 patients for the presence of MCPyV sequences by using DNA PCR directed against the Large T antigen (LT3) and the Viral Protein 1 (VP1; Fig. 1a and 1b). Specific MCPyV PCR products were found in 32% of the sporadic NMSC of the immunocompetent patients. In contrast, 62% of NMSC of the immunosuppressed patients were positive for MCPyV by either LT3 or VP1 PCR, i.e., the prevalence of MCPyV compared to NMSC in immunocompetent patients was significantly (p < 0.001; 1.9-fold) increased (Table III). It is of special interest that also the number of tumours positive for both, LT3 and VP1 was significantly increased (p < 0.001; 4.2-fold) in the NMSC of the immunosuppressed patients.

Figure 1.

(a) Representative results of the PCR products (308 bp) using NMSC DNA with LT3 primers. M indicates molecular weight marker, − = water control, SCC = squamous cell carcinoma, BCC = basal cell carcinoma, BD = Bowen's disease. (b) Representative results of the PCR products (351bp) using NMSC DNA with VP1 primers. M indicates molecular weight marker, − = water control, SCC = squamous cell carcinoma, BCC = basal cell carcinoma, BD = Bowen's disease.

Table III. Prevalence of Specific MCPYV-PCR Products in Nonmelanoma Skin Cancer (NMSC) Associated with and Without Immuno-Suppression
NMSC immunosuppressionMCPyV-PCRNMSC sporadicp-value
  1. LT3, Large T antigen; VP1, Viral Protein 1; p-value Chi-square test according to Pearson.

All (n = 56) All (n = 147) 
35 (62.5%)LT3 or VP147 (32%)<0.001
21 (37.5%)LT3 and VP113 (8.8%)<0.001
According to histology:
 Bowen's disease (n = 13)  Bowen's disease (n = 23) 
  9 (69%)LT3 or VP1  4 (17.4%)0.002
  5 (38.5%)LT3 and VP1  0 (0%)0.001
 Squamous cell carcinoma (n = 25)  Squamous cell carcinoma (n = 28) 
  13 (52%)LT3 or VP1  7 (25%)0.043
  7 (28%)LT3 and VP1  0 (0%)0.003
 Basal cell carcinoma (n = 18)  Basal cell carcinoma (n = 96) 
  13 (72.2%)LT3 or VP1  36 (37.5%)0.006
  9 (50%)LT3 and VP1  13 (13.5%)<0.001

MCPyV detection in NMSC of immunosuppressed patients

Out of 18 BCC of the immunosuppressed patients 13 (72.2%) were tested positive for MCPyV sequences either by LT3 or VP1 PCR. Compared to BCC of immunocompetent patients MCPyV sequences were found 2 fold more frequently in BCC of the immunosuppressed patients (p = 0.006). Fifty percent of the BCC of the immunosuppressed patients revealed PCR products for LT3 and VP1 when compared to BCC of the immunocompetent patients of which only 13.5% were positive for LT3 and VP1 (p < 0.001).

It is of interest that the number of MCPyV-positive SCC was 2 fold higher in the group of the immunosuppressed patients (Table III; p = 0.043). In addition, none of the sporadic SCC was tested positive by both, LT3 and VP1 PCR, whereas 26.9% of the SCC of the immunosuppressed patients were significantly more frequently positive for both, i.e., LT3 and VP1 PCR products (p = 0.003).

The difference between sporadic NMSC and NMSC of immunosuppressed patients was most striking in the BD group. Here 69% of BD were MCPyV positive in the immuno- suppressed group compared to only 17.4% in the immunocompetent group revealing a significant difference between these two groups (p = 0.002). In addition, 38.5% of BD of the immunosuppressed patients tested positive for both, LT3 and VP1 PCR products in contrast to the immunocompetent group of M. Bowen in which none of the positive cases was positive for both, LT3 and VP1 (p = 0.001).

MCPyV detection in sporadic NMSC

We analyzed the tumour tissue of 96 BCC of patients who according to the available clinical files had no history of immunosuppression. Testing for the presence of MCPyV we found that 37.5% (n = 36) of the sporadic BCC were positive in either the LT3 or VP1 DNA PCR (Table III). Only 13 BCC (13.5%) were tested positive for MCPyV in both PCRs, i.e., LT3 and VP1. In 11 BCC (11.4%) only PCR products for LT3 and in an additional 12 BCC (12.5%) only VP1 PCR products were detected.

In contrast, MCPyV sequences in SCC were detected at a much lower frequency (Table II). Out of the 28 SCC of nonimmunosuppressed patients only 7 (25%) tested positive for the presence of MCPyV by either LT3 or VP1 PCR. Five of these were positive both by LT3 PCR and 2 by VP1 PCR. In contrast to sporadic BCC none of the SCC of the immunocompetent patients was positive for both, LT3 and VP1. The results of the amplification of LT3 and VP1 from the DNA extracted from the lesions of BD did not reveal a significant difference compared to SCC. We have analyzed 23 BD of 17 patients of which only 4 were positive for MCPyV in either LT3 (n = 2) or VP1 (n = 2) PCR. Again, none of the 4 MCPyV positive cases was tested positive for LT3 and VP1 together (Table II).

MCPyV viral load analyses

In accordance to Garneski et al.14 we found a marked variability in the amount of viral DNA present in the MCC previously tested positive for MCPyV12 and in 30 NMSC of the immunosuppressed patients. The mean relative copy number changes of 30 NMSC tested was 5-fold lower compared to the mean relative copy number changes of 20 MCC.

MCPyV sequence analyses

Sequence analyses of selected PCR-products identified the PCR products amplified with the MCPyV primers as MCPyV sequences with only minor changes in the nucleotide order.


SCC and BCC are the most frequent tumours associated with immunosuppression. MCPyV is a recently discovered human polyomavirus which is associated with MCC. MCC is a rare and aggressive malignant neuroendocrine skin cancer of the elderly but its risk is increased by a factor of 40 in immunosuppressed recipients of organ transplants.16 Because of this association we tested a large number of NMSC of immunosuppressed and immunocompetent patients for the presence of MCPyV. Viral DNA was detected by PCR using the previously described LT3 and VP1 primers, which had already successfully been applied in recent studies for MCPyV associated MCC.11, 12 In the immunosuppressed patients MCPyV was 2-fold more frequent than in immunocompetent patients. MCPyV was 2-fold more frequent in NMSC of immunosuppressed patients, compared to that observed in NMSC of immunocompetent patients by either LT3- or VP1-PCR (p < 0.001). The positivity of NMSC either positive for LT3 or VP1 is an interesting an important finding to discuss. On the one hand it may point to an incongruity of the primers, which could be expected as well for LT3 as for VP1. On the other hand it might reflect sequence heterogeneity of the MCPyV as previously described.11 The number of newly identified human polyoma viruses has been increased during the last 2 years.11, 17, 18 Analyzing the number of NMSC tested positive for both, LT3- and VP1-PCR, MCPyV was even 4.3-fold more frequent in the NMSC of immunosuppressed patients (p < 0.0001). This finding possibly reflects an increase in viral nucleic acids of MCPyV due to enhanced viral replication because of the reduced capability in the immunosuppressed patients to keep the virus in latency. This might be similar to Epstein-Barr virus (EBV) associated post-transplantation associated diseases (PTLD) in which it has been shown that, based on viral gene expression patterns, EBV switches from latent to lytic infection and thus the viral load increases.19, 20

The significant increase of MCPyV positivity in BD of immunosuppressed patients and the negative finding of both, LT3 and VP1 PCR products, in the same lesion from immunocompetent patients might also reflect enhanced viral replication in a precursor lesion of invasive SCC in those patients of the former group.

Although it is well known that different immunosuppressive treatment modalities are associated with different frequencies of NMSC, no significant association could be established with the immunosuppressive treatment of the patients of our study and the presence of MCPyV. This is mainly due to the relatively small number of the unique group of immunosuppressed patients investigated in the present study. To study the possible impact of different immunosuppressive treatment modalities on the association of MCPyV-related NMSC either retrospective analyses on recently published clinical trials or prospective studies are needed.

Surprisingly, 37.5% of sporadic BCC were MCPyV positive in either LT3 or VP1 PCR. This certainly points to a role of MCPyV in a significant subset of sporadic BCC. No association could be established with distinct histopathological subtype of BCC and MCPyV. In a small number of BCC (n = 24) Becker et al. reported the finding of MCPyV in 3 cases (12.5%) using real time PCR directed against the small T antigen of MCPyV.13 The lower frequency in this study compared to ours is most likely reflected in the number of BCC tested and the additional primer used in our study targeting the VP1 gene of MCPyV. The same applies to the recent finding of Garneski et al. reporting SCC of which 2 (13.3%) were MCPyV positive.14 These results are close to our findings of MCPyV in 7 out of 28 sporadic SCC. Our results are further substantiated by the negative finding of MCPyV in normal skin, by the infrequent finding of MCPyV in colorectal carcinomas (3/89; 3.4%) and by our recently reported negative finding of MCPyV DNA in 45 healthy blood donors.12 MCPyV shares highest sequence homologies with the lymphotropic polyoma virus (LPyV) and the hamster polyoma virus (HPyV).11, 21, 22 The latter has been shown to induce so called “epitheliomas” in experimental animal models, which are similar to human BCC. In humans the pathogenesis of NMSC, namely SCC and BCC is closely linked to actinic skin damage which leads to local immunosuppression either enabling MCPyV infection or reactivation of latent MCPyV thus possibly contributing to already characterized steps in the carcinogenesis of NMSC.23, 24

Our data show that MCPyV is a frequently reactivated virus in immunocompromized patients. How MCPyV contributes to the pathogenesis of NMSC, i.e., BD, SCC and BCC, in immunosuppressed patients and potentially to the pathogenesis of a subset of sporadic BCC needs further investigations, e.g., integration or signature deletions of the T-antigen.11, 25 In addition, future studies are needed to investigate matched pairs of NMSC and normal skin of individual patients in order to assess further evidence on the causal relation of MCPyV and immunosupression related NMSC. According to our results MCPyV is a highly potential candidate tool to monitor the development of NMSC under immunosuppression for clinicians and pathologists.


We like to express our thanks to Anja Schoepflin for her technical support. We also thank Prof. Harald zur Hausen, DKFZ, Heidelberg, Germany, for helpful comments and reading the manuscript.