1. Institute of Virology, University of Cologne, Cologne, Germany
2. ^*Department of Dermatology, University of Cologne, Cologne, Germany
3. ^†Department of Dermatology, Medical Academy, Warsaw, Poland

Correspondence: Dr Pawel G. Fuchs, Institute of Virology, University of Koeln, Fuerst-Pueckler-Str. 56, 50935 Koeln, Germany. Email: pawel.fuchs@medizin.uni-koeln.de

Received 14 December 1999; Revised 3 March 2000; Accepted 5 April 2000.

Abstract

DNA of human papillomaviruses has frequently been detected in nonmelanoma skin cancers, raising the question of a possible causal contribution of these tumor viruses to skin carcinogenesis. Basal cell carcinomas are the most common nonmelanoma skin cancers; however, so far they are only poorly analyzed with regard to human papillomavirus infection. We searched for human papillomavirus-DNA in 69 biopsies from 61 immunocompetent basal cell carcinoma patients from two geographic locations in Europe using six different polymerase chain reaction primer systems. We could demonstrate human papillomavirus-DNA in 43.5% of the tested tumors. Human papillomavirus positivity did not seem to correlate with the duration of disease or patients' age. The vast majority of virus types in the biopsies belonged to the group of epidermodysplasia verruciformis-associated human papillomavirus. Of 31 sample pairs tested for human papillomavirus-DNA in tumors as well as in perilesional healthy skin, seven carried viral sequences in lesional and healthy skin and three only in the basal cell carcinoma. Six of the seven human papillomavirus-positive basal cell carcinoma/healthy skin pairs contained identical human papillomavirus types in tumors and histologically normal tissue. Forty basal cell carcinoma patients were additionally analyzed for IgG antibodies against virus-like particles of three representative epidermodysplasia verruciformis–human papillomavirus types: 8, 15, and 36. No statistically significant differences could be detected between human papillomavirus antibody prevalences of basal cell carcinoma patients and of dermatologically healthy individuals. Moreover, serologic findings did not correlate with the detection of specific human papillomavirus types in tumors. Our results seem to suggest that the occurrence of human papillomavirus-DNA in basal cell carcinoma does not reflect a major etiologic role of human papillomavirus in this cancer.