OBJECTIVES: To evaluate the role of HPV in oral carcinogenesis, we examined the prevalence of HPV in malignant, potentially malignant and normal oral epithelium and studied the relation of HPV prevalence with other factors obtained from the patient's records.MATERIALS AND METHODS: Our material consisted of 291 tissue specimens from 258 individuals. From every individual formalin fixed and paraffin embedded tissues were examined by nested Polymerase Chain Reaction (NPCR) for the detection of HPV DNA and by immunohistochemistry (IHC) for the in situ detection of HPV L1 protein. Positive PCR products were sequenced in order to type HPVs. Also 33 fresh tissues were obtained, fixed and used to detect HPV particles by transitional electron microscopy (TEM).RESULTS: HPV was detected in 32.9% of the tissue specimens by NPCR, in 4.7% by immunohistochemistry and in 28.1% by TEM. In detail, by nested PCR HPV L1 DNA was detected in 40% of normal tissues, 40% of fibromas, 35.8% of non-dysplastic leukoplakias, 31.6% of dysplastic leukoplakias and 22.2% of oral squamous cell carcinomas. The HPV viral load of 96.5% of the samples was very low (1 viral copy per 10(2)-10(4) cells). HPV16 prevails in all histological groups in 89-100%.CONCLUSION: We conclude that HPV does not seem, from the specific sample examined, to play a substantial role in oral carcinogenesis. However, it cannot be excluded that HPV could be involved in oral carcinogenesis only in cases with high viral load or at early stages of carcinogenesis possibly through the hit-and-run mechanism.