Article Text

other Versions

Human polyomavirus DNA detection in keratoacanthoma and Spitz naevus: no evidence for a causal role
  1. Linnea Haeggblom,
  2. Joar Franzén,
  3. Anders Näsman
  1. Department of oncology and pathology, Karolinska Institutet, Stockholm, Sweden
  1. Correspondence to Dr Anders Näsman, Department of Oncology and Pathology, Karolinska Institutet, Stockholm 177 76, Sweden; anders.nasman{at}


Keratoacanthomas (KA) and Spitz naevus (SN) are both lesions with unknown aetiology; therefore, the possibility of a viral involvement, more specifically the involvement of human polyomaviruses (HPyV), was investigated. In total, 22 cases of KA and 25 cases of SN were tested for the presence of HPyVs. DNA was extracted and amplified by multiplex PCR and thereafter tested with a multiplex bead-based assay for HPyVs (BKPyV, JCPyV, KIPyV, WUPyV, MCPyV, TSPyV, HPyV6, 7 and 9) and two primate viruses (SV40 and LPyV). HPyV DNA was found in 20 of the 47 lesions. There was no significant difference in HPyV DNA detection frequency between patients diagnosed with KA and patients diagnosed with SN, nor any over-representation of a specific HPyV type in any of the two patient categories. In conclusion, evidence for a specific aetiological role of any of the above tested HPyVs in either KA or SN was not disclosed.

Statistics from


Elderly and immunosuppressed people are more prone to develop keratoacanthomas (KA). These squamous cell tumours have a distinct clinical course, with a rapid growth phase followed by regression and a distinct histopathology, although some authors claim that the KA lesion is a variant of a well-differentiated squamous cell carcinoma.1 ,2 Due to the histopathology and the risk group of patients, it has been proposed that some of these tumours may be virus-associated. Thus, attempts have been made to link KA to, for example, mucosal and cutaneous human papillomaviruses, but the results have unfortunately not been convincing.3 Recently, some researchers have instead suggested an association to human polyomaviruses (HPyV), suggesting that especially HPyV6 may play a role in the emergence of KA.4 However, these results are still controversial, and others have reported a lower HPyV6 prevalence in KA.5 Hence, here, we wanted to study the role of HPyV in a broader perspective in KA and compare the results with a control group of skin tumours of another origin.

Spitz naevus (SN) is a distinct melanocytic benign lesion with unknown aetiology. The histomorphology resembles to some extent a malignant melanoma, but the clinical course is benign. In many cases, a halo phenomenon has also been observed surrounding the lesion.1 ,6 SN is described as particularly common in young adults and children, but a recent published study showed that approximately two-thirds of patients with SN were over 20 years of age.7 Nevertheless, due to the unknown aetiology, the predominance in younger patients and the fact that a halo phenomenon often is observed, a viral cause could theoretically be possible. Since the role of HPyV in SN, to our knowledge, has never been studied, we wanted here to scrutinise these tumours for the presence of HPyV DNA in comparison with a group of tumours derived from an epithelial origin.

The field of HPyV is emerging, and today there are 13 different HPyV types described, where the majority have been disclosed in the last 10 years. HPyVs are all small non-enveloped DNA viruses made up of approximately 4K base pairs and have an oncogenic and pathogenic potential in humans.8 Merkel cell carcinoma has been associated to Merkel cell polyoma virus (MCPyV),9 trichodysplasia spinulosa to HPyV8 (or trichodysplasia spinulosa polyomavirus, TSPyV)10 and HPyV7, to a specific rash occurring in transplant patients,11 while the pathogenic role of the other known viruses is more uncertain.

HPyV 6 and 7 were isolated from skin swab samples obtained from healthy individuals,12 while the other remaining viruses have been detected elsewhere, and described to occur in samples from various locations in the skin.13

Subsequently, in this study, the presence of DNA from BKPyV, JCPyV, KIPyV, WUPyV, MCPyV, TSPyV, HPyV6, 7 and 9 was evaluated in two distinct different histopathological diseases, KA and SN, in a case-to-case comparison setting to assess the possibility of a viral association in these diseases.

Material and methods

Patient samples

Paraffin-embedded biopsies from patients diagnosed with KA and SN at the Karolinska University Hospital were identified and obtained from the pathology unit. In total, 22 KA diagnosed during 2014 and 25 SN diagnosed between 2005 and 2014 were obtained (table 1). Age data were extracted from the hospital registry. The study was performed according to ethical permission 2012/83-31/2 obtained from the Stockholm Regional Ethical Committee at Karolinska Institutet.

Table 1

Study population (number of patients and their mean age) presented per diagnosis and as a total

DNA extraction and HPyV DNA detection

Paraffin-embedded slides, 10 μm thick, were deparaffinised in xylene and rehydrated in absolute ethanol. Blank samples, only containing paraffin, were added to every fifth sample. DNA was then extracted with the Roche high-pure RNA paraffin kit according to the manufacturer’s instruction, without RNAse treatment, as previously described.14 Subsequently, presence of HPyV DNA was assessed by a multiplex Luminex method, described elsewhere.11 Briefly, a multiplex PCR was performed, containing primers covering both the large T and the VP1 region for simultaneous detection of BKPyV, JCPyV, KIPyV, WUPyV, MCPyV, TSPyV, HPyV6, 7 and 9. and two primate polyomaviruses (SV40 and LPV) and the housekeeping gene β-globin. PCR products were then detected with the Luminex method. As previously described, a median fluorescent index value over 2× background +20 was considered a positive value.11

Statistical analysis

χ2 test was used to compare proportions of HPyV-positive samples between the diagnostic categories. Two-sided p values were reported. An independent two-sided t-test was used to compare the mean age between HPyV-positive and HPyV-negative samples.


In total, 22 and 25 tissue sections from patients with KA and SN, respectively, were included in this study. All samples from the lesions had amplifiable DNA (amplifiable β-globin), and 20 of the 47 lesions showed presence of HPyV DNA (table 1).

There was no statistical difference in HPyV DNA detection frequency between patients diagnosed with KA and patients diagnosed with SN. Furthermore, in both patient categories, there were no difference in age between patients with HPyV DNA positive samples and patients with HPyV DNA negative samples (table 1). However, as already known, patients with KA were significantly older than patients with SN (data not shown).

Moreover, there was no significant over-representation of any HPyV type in any of the two patient categories (table 2).

Table 2

Distribution of HPyV types between keratoacanthomas (KA) and Spitz naevus (SN)


In the present case-to-case comparison study, a fairly low and a similar prevalence of HPyV, without any disease-specific domination, in both KA and SN was observed, suggesting that HPyVs are not associated to either KA or SN.

Available paraffin-embedded biopsies from two distinct skin lesions with different origins were obtained here from the pathology department at the Karolinska university hospital and tested for the presence of HPyV DNA in a case-to-case comparison. Since KA arises from the epithelial cells in the epidermis and SN emerges from melanocytes,1 it is unlikely that a single viral agent would be a causative factor for both diseases. In this study, we were not able to show any difference in the HPyV DNA distribution between the different diseases, suggesting that the HPyV detection prevalence in this study only represents a viral background flora and not a disease causative agent. Furthermore, patients with virus-associated lesions are often older/immune-suppressed, which is also observed in Merkel cell carcinoma.15 In other cases, patients with, for example, human papillomavirus-induced oropharyngeal cancer tend to be younger instead.16 Here, there were no differences in age between the patients with KA or SN, with or without the presence of HPyVs, thus possibly supporting the statement above that the tested HPyVs are not associated to the development of KA or SN.

Similar to other studies,5 ,17 a fairly higher number of MCPyV-positive KA was observed, as compared with other known HPyV types. A recent report by Scola et al found a similar MCPyV DNA prevalence in KA (29%), and similar prevalence numbers in other skin-examined lesions (eg, Mb Bowen, basal cell carcinoma, actinic keratosis, squamous cell carcinomas, microcystic adnexal carcinomas and atypical fibroxanthoma), suggesting that MCPyV is commonly found on the human skin, but not necessarily linked to disease—except in Merkel cell carcinoma.5 Besides MCPyV, HPyV6 was also detected, but with much lower prevalence numbers and lower viral loads, and no additional HPyV types were reported.5

Nevertheless, recently Beckervordersandforth et al4 suggested a link between HPyV6 and KA. In their study, 42.3% of all 59 examined KA were positive for HPyV6 DNA tested by PCR, which also was confirmed with FISH hybridisation. Likewise, higher prevalence numbers of HPyV6 were also reported in basal cell carcinomas, squamous cell carcinomas and trichoblastomas.4 In contrast, in the present study, HPyV6 DNA was not detected in KA or SN. Although our study was slightly smaller in size and the PCR strategy differed between the two studies, the dissimilarities observed with regard to HPyV6 are puzzling, and larger materials are needed before the suspected role of HPyV6 in KA fully can be understood.

Notably, a fairly low BKPyV and KIPyV prevalence was observed, especially in SN. To our knowledge, there are no reports investigating the prevalence of KIPyV in human skin lesions, but the virus has been examined for, and not been detected in, mucosal melanomas.18 ,19 Presence of BKPyV has previously been detected in skin samples from healthy individuals, and in a former study, as many as 76% of biopsies from normal skin disclosed the presence of BKPyV DNA when using a very sensitive technique (PCR followed by Southern blot).20 Here, lower prevalence numbers were detected.

In conclusion, in this study, a specific role of HPyV (BKPyV, JCPyV, KIPyV, WUPyV, MCPyV, TSPyV, HPyV6, 7 and 9) was not observed in either KA or SN. This does not exclude the possibility that these lesions still may be associated to yet other identified or unidentified viruses, and further studies are warranted.


The authors want to thank Professor Tina Dalianis for clinical material, and Professor Tina Daliais and Associate Professor Torbjörn Ramqvist for technical and laboratory support.


View Abstract


  • Handling editor Cheok Soon Lee

  • Contributors LH and AN planned the study. LH and JF ran the experiments. LH, JF and AN evaluated the data. LH, JF and AN. wrote and approved the manuscript.

  • Funding This work was financially supported by the Stockholm Council (SLL; clinical postdoctoral appointment), KI, Stiftelsen Sigurd och Elsa Goljes Minne and Magnus Bergvalls stiftelse.

  • Competing interests None.

  • Ethics approval 2012/83-31/2.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.