Human papillomavirus infections in nonmelanoma skin cancers from renal transplant recipients and nonimmunosuppressed patients

BACKGROUND

Nonmelanoma carcinomas of the skin represent the most frequent cancers among the Caucasian population worldwide. They occur with high frequency in renal allograft recipient patients after prolonged immunosuppression.

PURPOSE

We analyzed tumors obtained from both immunosuppressed and nonimmunosuppressed patients for human papillomavirus (HPV) DNA.

METHODS

Twenty-nine specimens of nonmelanoma carcinomas of the skin were obtained from 19 renal allograft recipient patients; these included 20 specimens of squamous cell carcinoma (SCC) from 11 patients, five specimens of basal cell carcinoma (BCC) from four patients, and four specimens of carcinoma in situ (CIS) from four patients. Forty-one specimens of nonmelanoma carcinomas of the skin were obtained from 32 nonimmunosuppressed patients; these included 26 SCC specimens from 19 patients, 11 BCC specimens from nine patients, and four keratoacanthoma (benign epithelial tumor) specimens from four patients. A polymerase chain reaction method involving use of degenerate oligonucleotide primers, in which the conserved region of the open reading frame of the HPV L1 (major capsid protein) gene is amplified, was used to amplify total cellular DNA purified from individual tumors. The DNA of each specimen was subjected to 16 different amplification reactions; different primer combinations were used in order to increase the sensitivity and specificity of HPV detection. Resulting products were probed with a radioactively labeled, degenerate oligonucleotide. HPV-specific DNA was either sequenced directly after elution from the gel or amplified with semi-nested, degenerate primers, after which the products were cloned and sequenced. Sequences were compared with all known papillomavirus sequences.

RESULTS

Thirteen (65%) of the 20 SCC specimens and three of the five BCC specimens from immunosuppressed (renal allograft recipient) patients contained identifiable HPV-related sequences, among them 13 putative novel HPV genomes. In addition, all other malignant tumor specimens from this patient group revealed faint signals upon amplification and hybridization; the origin of these signals has not been identified in the present study. In nonimmunosuppressed patients, eight (31%) of 26 SCC specimens and four (36%) of 11 BCC specimens contained sequences of HPV types. Two putative novel HPV sequences could be identified in this group. Faint signals of yet undetermined origin were observed in eight of the SCC specimens and in two of the BCC specimens. Two of four keratoacanthoma specimens contained sequences of known HPV type. (Keratoacanthoma is a nonmalignant lesion for which the natural history has not been defined.) The spectrum of HPV types in both groups of patients differed substantially.

CONCLUSIONS

These data point to the frequent presence of HPV sequences in SCCs and BCCs of the skin. The etiologic relationship of these infections to the respective malignant tumors remains to be evaluated.

IMPLICATIONS

The presence of HPV DNA in a large percentage of specimens of nonmelanoma carcinomas of the skin from immunosuppressed patients, as well as from nonimmmunosuppressed patients, renders a papillomavirus infection as a possible factor in the etiology of this disease.

Prevailing papillomavirus types in non-melanoma carcinomas of the skin in renal allograft recipients

The role of human papillomavirus (HPV) in the aetiology of in situ and invasive carcinoma of the genital tract is well established. In the rare disorder epidermodysplasia verruciformis (EV), in which patients develop extensive warts of unusual types and multiple cutaneous squamous cancers on light-exposed skin, current evidence suggests a probable role for a specific group of EV HPVs in the carcinogenic process. Determination of the possible role of HPV in the aetiology of non-melanoma skin cancers (NMSCs), which occur frequently in immunosuppressed organ allograft recipients, has been limited, until recently, by the lack of availability of a sensitive detection system for a wide range of cutaneous HPV types. We have used a combination of 2 sets of PCR primers to examine 68 benign and malignant tumours collected over a 12-year period from 25 renal allograft recipients. Cloning and sequencing of the PCR products were carried out to distinguish HPV DNA from cellular sequences. A combination of these techniques revealed HPV DNA in all viral warts, 65% of keratoses, 91% of intra-epidermal cancers and 91% of invasive squamous cancers. Both cutaneous and EV HPV types were detected, including 18 novel types. In 4 patients with multiple cancers, the most prevalent types were in the EV group: HPV 20, 23, 38 and 2 novel types, DL40 and DL267 (related to HPV 10 and 38, respectively). These 5 HPV types were present in a total of 73% of all malignant lesions tested. The technique described represents a reliable method of HPV DNA detection in NMSC. The EV group of HPVs predominate in the cancers, but the multiplicity of HPV types detected with double infection in some lesions suggests virus/virus in addition to virus/host interaction in the carcinogenic process.

Detection of epidermodysplasia verruciformis-like human papillomavirus types in malignant and premalignant skin lesions of renal transplant recipients

To evaluate the putative role of human papillomaviruses (HPV) in the development of skin cancer in renal transplant recipients, a series of skin biopsies from premalignant and malignant skin lesions was analysed using the polymerase chain reaction. Four different consensus primer pairs were used. HPV DNA was detected in five of 24 cases of squamous cell carcinoma, in one of three cases of Bowen's disease, in none of four basal cell carcinomas, in two of seven cases of actinic keratosis and in one of five cases of keratoacanthoma. Typing by direct sequencing of the amplified HPV DNA was possible in seven of nine cases, and revealed epidermodysplasia verruciformis (EV)-associated HPV types, or HPV types related to EV-associated types. Hence, HPV DNA could be detected in a significant proportion of (pre)malignant skin tumours in renal transplant recipients. The finding that some of the detected HPV types were as yet uncharacterized EV-related types, suggests that HPV DNA could be present in a higher percentage of lesions, and might be detected with refinement of the techniques.

Cutaneous verruciform xanthoma: a report of five cases investigating the etiology and nature of xanthomatous cells

Verruciform xanthoma is a rare clinicopathologic entity of uncertain etiology that occurs primarily in the oral mucosa. Aggregates of foam cells in the submucosal stroma or papillary dermis in association with verrucous epithelial hyperplasia are the hallmark of this lesion. Extraoral (cutaneous) occurrence of verruciform xanthoma is much rarer and has been reported mostly in the genital skin. Five cases of extraoral cutaneous verruciform xanthoma (three from the scrotum, one from the penis, and one from the nose) and one histologic "simulant" (from skin of the nose) were studied. The lesions were solitary, raised, or polypoid with cup-shaped craters filled with parakeratotic cells that blended into keratinocytes of an acanthotic and papillomatous epidermis. There was a neutrophilic infiltrate of varying intensity between plump parakeratotic cells and keratinocytes, near the surface of the epidermis. Aggregates of foam cells were present in the papillary dermis, which was highly vascular. A plasma cell predominant infiltrate was seen at the base in a bandlike fashion. Despite the architectural resemblance of verruciform xanthoma to verrucous mucocutaneous lesions related to human papillomavirus infection, it was not detected by either immunohistochemistry, in situ hybridization, polymerase chain reaction, or Southern blot analysis in any case. The foam cells were weakly positive for cytokeratin and for Factor XIIIa but negative for S-100 protein. The KP1 and Mac 387 immunostain showed focal weak staining in foam cells. We postulate that a cascade of events pursue after initial keratinocytic damage attracting neutrophils, with subsequent phagocytosis of necrotic keratinocytic debris by dermal dendrocytes, eventually leading to the ultimate manifestation of the lesion as verruciform xanthoma. The etiologic agent remains elusive, but based on our findings, we conclude that verruciform xanthoma is most likely not a human papillomavirus-associated squamoproliferative lesion and that the foam cells, a histologic hallmark of the lesion, are most likely derived from dermal dendritic cells.

Identification of human papillomavirus in keratoacanthomas

BACKGROUND

Keratoacanthomas are benign, clinically distinct skin tumors that may infiltrate and show cellular atypia. A viral etiology has been suggested, and the aim was to search for human papillomavirus (HPV) in keratoacanthomas.

METHODS

From 21 immunosuppressed organ transplant recipients and 11 non-immunosuppressed patients, 72 fresh biopsies with diagnosis of keratoacanthomas were analyzed. For detection of cutaneous and genital HPV DNA, single-tube nested "hanging droplet" polymerase chain reaction (PCR) and another PCR (GP5+ and 6+) were used, respectively.

RESULTS

Among 21 immunosuppressed patients, 71% (15/21) harbored HPV DNA at least in one sample. Of the keratoacanthoma lesions, 55% (33/60) were HPV DNA positive. Fourteen samples from eight immunosuppressed patients contained HPV types 5, 9, 10, 14, 19, 20, 21, 38, 49, 80, putative HPV types as HPVvs20-4, HPVvs75, and HPVvs92 and FA16.1, FA23.2, FA37, FA75, and FA81. Among 11 non-immunosuppressed patients, 36% (4/11) harbored HPV DNA at least in one sample, and 33% (4/12) of their keratoacanthomas were HPV DNA positive. In total, HPV DNA was detected in 51% (37/72) of the keratoacanthomas.

CONCLUSIONS

By the use of PCR, cutaneous HPV DNA was detected in 51% (37/72) of the keratoacanthomas. No predominating HPV type or genital HPV type was identified. The role of HPV in keratoacanthomas remains thus elusive.

Unbiased approach for virus detection in skin lesions

To assess presence of virus DNA in skin lesions, swab samples from 82 squamous cell carcinomas of the skin (SCCs), 60 actinic keratoses (AKs), paraffin-embedded biopsies from 28 SCCs and 72 kerathoacanthomas (KAs) and fresh-frozen biopsies from 92 KAs, 85 SCCs and 92 AKs were analyzed by high throughput sequencing (HTS) using 454 or Ion Torrent technology. We found total of 4,284 viral reads, out of which 4,168 were Human Papillomavirus (HPV)-related, belonging to 15 known (HPV8, HPV12, HPV20, HPV36, HPV38, HPV45, HPV57, HPV59, HPV104, HPV105, HPV107, HPV109, HPV124, HPV138, HPV147), four previously described putative (HPV 915 F 06 007 FD1, FA73, FA101, SE42) and two putatively new HPV types (SE46, SE47). SE42 was cloned, sequenced, designated as HPV155 and found to have 76% similarity to the most closely related known HPV type. In conclusion, an unbiased approach for viral DNA detection in skin tumors has found that, although some new putative HPVs were found, known HPV types constituted most of the viral DNA.

The Mastomys natalensis papillomavirus: nucleotide sequence, genome organization, and phylogenetic relationship of a rodent papillomavirus involved in tumorigenesis of cutaneous epithelia

Mastomys natalensis is a rodent of African origin afflicted with a very high incidence of skin tumors (keratoacanthomas and squamous carcinomas), which are associated with a papillomavirus, M. natalensis papillomavirus (MnPV). We have determined the genomic sequence of MnPV, which has a size of 7687 bp. The genomic organization is similar to that of other papillomaviruses, with open reading frames E6, E7, E1, E2, and E4 in the early and L2 and L1 in the late region. Due to an unusually large hinge region, the transcriptional activator E2 has a size of 542 amino acids rather than 400 to 460 amino acids, as in other papillomaviruses. An open reading frame E5 coding for a small hydrophobic membrane protein is missing, as is the case for some cutaneous human papillomaviruses (HPV). This fact, together with the composition of cis-responsive elements in its long control region and phylogenetic evaluation of segments of its E6, E1, and L1 genes, indicates a relationship of MnPV to the cottontail rabbit papillomavirus and several HPV types found in lesions of cutaneous epithelia, in particular to those that are associated with epidermodysplasia verruciformis. MnPV may be a useful model system for tumorigenesis of cutaneous epithelia in humans.

Expression of cell-cycle proteins p53, p21 (WAF-1), PCNA and Ki-67 in benign, premalignant and malignant skin lesions with implicated HPV involvement

A series of 120 biopsies from benign (verruca vulgaris and keratoacanthoma), premalignant (actinic keratosis and extragenital Bowen's disease) and malignant (squamous cell carcinoma) skin lesions were studied immunohistochemically for the expression of cell-cycle proteins p53, p21 (WAF-1), PCNA and Ki-67. The presence of human papillomavirus (HPV) DNA in these samples had been analysed previously using in situ hybridization (ISH) and PCR. Moderate to intense expression of both PCNA and Ki-67 was present in most of the lesions studied. PCNA staining was extensive in the epidermis underneath the layers where abundant HPV DNA staining was shown in HPV DNA-positive verrucas. In keratoacanthomas, p21 and PCNA expression remained low, despite intense p53 expression. In actinic keratosis, only half of the specimens showed overexpression of p53 associated with moderate or intense expression of PCNA. In extragenital Bowen's lesions, all these cell-cycle markers were overexpressed, but in squamous cell carcinomas, they were heterogeneously expressed and showed no correlation with tumour differentiation. Our results suggest a mechanism by which HPV can reactivate the host genes (leading to cell proliferation) to support its own DNA replication. Also p21 might start keratinocyte differentiation in areas where HPV DNA replication starts. Cell proliferation remained active in actinic keratosis and Bowen's lesions, emphasizing the precancer character of these lesions in contrast with the benign nature of keratoacanthoma and verruca vulgaris.

Posttransplant skin cancer: a possible role for p53 gene mutation but not for oncogenic human papillomaviruses

BACKGROUND

Loss of p53 tumor suppressor function is a critical step in the development of diverse malignancies, including skin cancers in nonimmunosuppressed patients where UV-specific p53 gene mutations have been identified. In tumors associated with human papillomavirus (HPV), such as cervical carcinoma, p53 may be inactivated instead by binding to a viral oncoprotein.

OBJECTIVE

Our purpose was to examine the hypothesis that HPV may play an analogous role in the development of posttransplant skin cancer.

METHODS

p53 Immunoreactivity, suggestive of p53 gene mutation, was examined by immunocytochemistry. Oncogenic HPV DNA was detected by polymerase chain reaction.

RESULTS

Comparable p53 immunoreactivity was seen in skin tumors from both transplant and nontransplant patients. HPV DNA was not demonstrated in any tumor specimen.

CONCLUSION

Our data do not implicate oncogenic HPV in posttransplant skin cancer. p53 Gene mutation, rather than HPV-induced p53 degradation, may be more significant in the development of these tumors.

Identification of DNA sequences of both genital and cutaneous HPV types in a small number of keratoacanthomas of nonimmunosuppressed patients

BACKGROUND

Viral infection was suggested to be etiologically involved in skin tumor development. Data on the association of human papillomavirus (HPV) with keratoacanthomas are still conflicting.

OBJECTIVE

Analysis of HPV infection in keratoacanthomas of the general population.

METHODS

HPV DNA was detected by nested PCR. To include a broad range of both cutaneous and mucosal HPV types, HPV PCR was performed with two sets of degenerate primers.

RESULTS

Considering only beta-globin-positive specimens, HPV DNA was detected in 20% of the specimens obtained from 18% of the patients. The spectrum of HPV types detected contains HPV types 6, 14, 16, 35, 58 and 61. In 1 case, the underlying HPV type was not identified. In 1 specimen with transition towards squamous-cell carcinoma, HPV 16 was detected.

CONCLUSIONS

HPV is probably not generally associated with the etiology of keratoacanthoma but may be relevant in individual cases. Oncogenic HPV types may be cofactors for malignant transformation of initially benign skin lesions like keratoacanthomas.

Frequent detection of human polyomavirus 6 in keratoacanthomas

BACKGROUND

The recent discovery of the Merkel cell polyomavirus and its consistent association with Merkel cell carcinoma has drawn attention to the numerous recently discovered polyomaviruses and their possible involvement in the etiopathogenesis of non-melanoma skin cancer (NMSC). Data on the recently discovered human polyomavirus 6 (HPyV6) and its role in NMSC are sparse and in part controversial.

METHODS

In the present study we tested a large number (n = 299) of NMSC specimens for the presence of human polyomavirus 6 (HPyV6) by DNA PCR and HPyV6 fluorescence in situ hybridization (FISH). In detail, 59 keratoacanthomas (KA), 109 basal cell carcinomas (BCC), 86 squamous cell carcinomas (SCC) and 45 trichoblastomas (TB) were tested for the presence of HPyV6.

RESULTS

HPyV6 DNA PCR and subsequent sequence analysis revealed that 25 KAs (42.3 %), 23 BCCs (21.1 %), 8 SCCs (9.3 %) and 10 TBs (22.2 %) were HPyV6 positive. The presence of HPyV6 DNA was visualized and validated on the single cell level within the histomorphological context by HPyV6 fluorescence in situ hybridization.

CONCLUSIONS

The high frequency of HPyV6 DNA in 42.3 % of KA possibly points to a role for HPyV6 in the etiopathogenesis of KAs. Although the detection rate of HPyV6 DNA in BCCs and TBs is within the previously reported detection range in normal skin, it does not exclude a possible role for HPyV6 in the carcinogenesis in a significant subset of these skin tumors.

Detection of human papillomavirus DNA in keratoacanthomas by polymerase chain reaction

The etiology of keratoacanthomas is unknown, but human papillomavirus (HPV) has been suspected to be involved in the pathogenesis of this lesion because koilocytic changes may be observed and because HPV has been found in cutaneous squamous-cell carcinomas and premalignant keratoses in immunosuppressed patients. We analyzed DNA extracted from 39 keratoacanthomas from 22 "at-risk" patients (nine patients undergoing UV light and/or anthralin therapy for psoriasis, 10 solid organ transplant recipients, one patient with xeroderma pigmentosa, one patient with acquired immunodeficiency syndrome, and one patient undergoing therapy for non-Hodgkin's lymphoma) for the presence of HPV. The results were compared with analyses of DNA extracted from 30 keratoacanthomas from 28 patients at no known increased risk for these lesions. Using polymerase chain reaction (PCR) primers designed to detect multiple HPV types (including 6, 11, 16, 18, 31, and 33), HPV was detected in seven keratoacanthomas from six of the at-risk patients and in eight sporadic keratoacanthomas from eight patients without risk factors. HPV was also present in one of 26 nonlesional skin controls. Statistical analysis showed a significant difference in the prevalence of HPV DNA sequences found in keratoacanthomas compared to normal control skin (p = 0.038). The presence of virus by PCR could not be predicted by histologic evaluation. Sequence analysis showed the presence of HPV types 11, 13, 24, 33, and 57. Although these results confirm the frequent presence of HPV in keratoacanthomas, the role of this virus in the etiology and pathogenesis of these lesions remains to be elucidated.

Nine-year follow-up of a case of Grzybowski type multiple keratoacanthomas and failure to demonstrate human papillomavirus

We describe a patient with a 9-year history of generalized eruptive keratoacanthoma (KA) of the Grzybowski type whose multiple skin lesions showed steady progression, resulting in a sclerotic, mask-like facial expression and ectropion. Eleven tumour biopsies representing lesions of different stages and localizations (erupting and regressing KAs, biopsies from non-involved light-protected and light-exposed skin, dermatosclerosis and squamous cell carcinomas) were analysed for human papillomavirus (HPV) sequences using a polymerase chain reaction approach capable of detecting the majority of all presently known HPV genotypes. None of the biopsy specimens proved to be HPV-positive, although HPV was detected in weakly and heavily affected control specimens by the method applied. These findings suggest an HPV-independent aetiology of this rare type of multiple KA.

Known HPV types have no association with keratoacanthomas

Certain HPV types have been linked to the genesis and development of premalignant and malignant skin diseases. There have been several contradictory reports on the role of HPV infections in the development of keratoacanthomas (KAs). To further study the involvement of HPVs in the aetiology of KAs, we investigated paraffin-embedded specimens of 80 biopsies of KAs for the presence of HPV 1, 2, 3, 4, 5, 7, 26, 37, 38, 47 and 59 DNA by in situ hybridization (ISH) with biotinylated probes under high stringency conditions (Tm-10 degrees C). Every fourth biopsy specimens was also examined by polymerase chain reaction (PCR) with consensus primers targeting the HPV E1 and L1 regions. The positive cases were further studied by direct DNA sequencing. All specimens proved to be negative for all HPV DNAs studied by ISH. Three out of 20 cases produced in positive PCR amplifications when consensus primers targeting the L1 region were used. However, the same samples remained negative with general primers targeting the E1 region. The DNA sequence analysis of the PCR-positive products showed a 76% homology with HPV type 17. Our results suggest that the known HPV types are unlikely to have any role in the aetiology of KAs.

Role of human papillomavirus in cutaneous oncogenesis

There is a strong association between the human papillomavirus and cutaneous squamous cell carcinoma. If this association was merely random, one would expect an equal distribution of human papillomavirus types among affected individuals. However, only specific types of human papillomavirus are consistently found in cutaneous and genital squamous cell carcinomas. Immunosuppressed individuals clearly have a much higher incidence of cutaneous carcinomas. Immunosuppression, either local or systemic, not only decreases immune surveillance but may also dictate the amount and type of virus each individual may carry. Epidermodysplasia verruciformis and other rare hereditary disorders that combine specific immune defects and an increased incidence of malignancy are very useful models that clearly fulfill a multistep theory of oncogenesis. The precise mechanism of oncogenesis in these select human papillomavirus types is not yet fully understood. Intracellular interactions with the recently described tumor suppressor proteins may prove to be the primary site of action of these oncogenic viruses. Environmental cocarcinogens and activation of oncogenes are clearly important if not essential factors in human papillomavirus-associated tumors. As our knowledge and understanding of malignant transformation grows, it becomes apparent that this is a complex multistep process.

HPV 6-Positive Giant Keratoacanthoma in an Immunocompetent Patient

Keratoacanthoma (KA) is a clinically distinct, rapidly growing lesion that generally presents as a solitary crateriform nodule in sun-exposed areas in elderly, fair-skinned individuals. A KA larger than 20-30 mm is referred to as giant keratoacanthoma, a relatively rare lesion especially in young patients. Such lesions grow rapidly with possible destruction of underlying tissues. In addition to ultraviolet exposure, KAs have also been associated with chemical carcinogens, chemical peels, genetic factors, chronic skin conditions that produce scarring, trauma and thermal burns. Immunosuppressed patients, especially after transplantation, also develop KAs. A viral etiology has been suggested but not confirmed. We encountered a case of giant keratoacanthoma (greater than 50 mm in diameter) with induration of underlying structures on the upper lip of a 39-year-old male sailor. The patient reported sudden appearance and rapid enlargement of the lesion in only three weeks. Biopsy of the cutaneous lesion and the characteristic clinical history suggested the diagnosis of keratoacanthoma. Total excision with primary closure of the defect by a nasolabial advancement flap was performed. Histological examination of the tumor mass confirmed the diagnosis of KA with infiltrative growth and perineural invasion. Immunosuppression was excluded by blood analyses, as were HIV, syphilis and hepatitis infections. Only low-risk genital HPV type 6 was detected in the lesion, suggesting a possible cocarcinogenic effect of HPV and UV light in a chronically sun-exposed patient.

Reinitiated expression of EJras transgene in targeted epidermal cells of transgenic rabbits by cottontail rabbit papillomavirus infection

Transgenic rabbits carrying the EJras oncogene have been established in our laboratory (Am. J. Pathol. 155 (1999) 315). The expression of the ras gene is targeted to the epidermal keratinocytes using the upstream regulatory region (URR) of the cottontail rabbit papillomavirus (CRPV). All of the transgenic rabbits develop keratoacanthomas at multiple sites in the skin at 2-3 days after birth, and the tumors spontaneously regress in 1.5-2 months. With regression of the keratoacanthomas, the rabbits appear normal and EJras expression is undetectable in their skin. To determine if CRPV infection would reinitiate the expression of the EJras transgene and make the rabbits more sensitive to tumorigenesis, the rabbits were infected with CRPV at 2 months of age when the keratoacanthomas had regressed. This study shows that CRPV infection of the transgenic rabbit skin could shorten the latency required for CRPV papilloma initiation, and significantly increase the tumor growth and persistence rate compared with non-transgenic rabbits. Furthermore, EJras expression became detectable in the CRPV induced papillomas in transgenic rabbits, but not in the papillomas of non-transgenic rabbits. These results indicate that CRPV infection is able to reinitiate the expression of the CRPV URR controlled EJras oncogene carried by the transgenic rabbits and that the expression of EJras can enhance the tumorigenesis of CRPV infection.

Human papillomavirus and widespread cutaneous carcinoma after PUVA photochemotherapy

BACKGROUND

Oral psoralen with UV-A (PUVA) photochemotherapy is known to cause cutaneous malignancies and has been associated with cutaneous immunosuppression. Human papillomavirus infection has also been associated with cutaneous malignancies and with immunosuppressed individuals. We therefore sought evidence of human papillomavirus infection in a patient with a long history of PUVA therapy and multiple cutaneous malignancies.

OBSERVATIONS

During a 15-year period, an otherwise healthy patient with psoriasis who had undergone a 10-year course of PUVA photochemotherapy developed 13 squamous cell carcinomas, eight lesions diagnosed as "squamous cell carcinoma vs keratoacanthoma," 14 other keratoacanthomas, six basal cell carcinomas, one melanoma in situ, and 18 other keratinocytic dysplasias. Twenty-two of the 30 lesions tested for human papillomavirus DNA by polymerase chain reaction were positive for type 16/18, including six of the seven basal or squamous cell carcinomas tested.

CONCLUSION

We hypothesize that PUVA therapy-induced immunosuppression may play an important role in PUVA-related carcinogenesis by affecting the extent and pathogenicity of human papillomavirus infection.

Solitary keratoacanthomas in immunocompetent patients: no detection of papillomavirus DNA by polymerase chain reaction

The aetiology of keratoacanthoma (Ka) is unknown, but human papillomavirus (HPV) has been implicated in the pathogenesis of this lesion. To evaluate the role of HPV in the development of KA in the general population, 20 tissue specimens were analysed by polymerase chain reaction (PCR). To include a broad range of both cutaneous and mucosal HPV types, PCR was performed with two sets of degenerate primers. No HPV-DNA sequences were detected in any lesions analysed. These results do not support the hypothesis that HPV is involved in the etiology and pathogenesis of the KA in immunocompetent patients.