Introduction: Adequate brushing of oral mucosa is important for accurate human papillomavirus (HPV) detection in potentially malignant (oral leukoplakia [OL], oral lichen planus [OLP]) and malignant (oral squamous cell carcinoma [OSCC]) lesions. Since various factors may limit the adequacy of oral brushing and, consequently, the accuracy of HPV detection, modified sampling procedures should be evaluated for their effect on HPV frequency and/or types detected.
Aim: To compare the HPV frequency in samples obtained by brushing the lesion site with the frequency in samples obtained by brushing an apparently normal adjacent site. The correlation between HPV frequency and keratinization of the site affected by the lesion, as well as sociodemographic variables (age, sex, smoking and drinking habits), was also examined.
Methods: HPV DNA was detected in brushing samples from 50 patients with OL, 49 with OLP, and 17 with OSCC. Polymerase chain reaction (PCR) amplification was performed by MY09/MY11 and GP05+/GP06+ primers; the HPV type was identified by DNA sequencing and a reverse hybridization (line probe) assay. Data were analyzed by the Z test, the Fisher's exact test, the chi-square test, odds ratio (OR), and a logistic regression model.
Results: HPV DNA was detected in 22% of samples from lesion sites and in 16% of samples from adjacent sites (p = 0.22) in patients with OL, in 24.5% and 22.4% of samples from lesion and adjacent sites, respectively, in patients with OLP (p = 0.40), and in 35.3% and 41.2% of samples from lesion and adjacent sites, respectively, in patients with OSCC (p = 0.36). Lesions adjacent to HPV-positive normal sites had an increased rate of HPV detection (OR = 30; 95% CI 9.57, 94.1). HPV-18 was the most frequent genotype, followed by HPV-6, -16, -33, and -53. HPV prevalence was reduced in lesions at keratinized sites (14.5%) compared with non-keratinized sites (34.4%; p = 0.007; OR = 0.32; 95% CI 0.13, 0.81).
Discussion: In patients with OL, OLP, or OSCC, a high prevalence of HPV infection was shown in apparently normal sites adjacent to lesion sites infected by HPV. The lower HPV frequency in lesions at keratinized sites suggests that HPV detection by lesion brushing is affected by keratinization. The keratinized epithelium may be less susceptible to HPV infection or, alternatively, the highly proliferative activity in non-keratinized sites may predispose to HPV infection.
Conclusion: Results from this study indicate that taking samples from normal sites adjacent to oral lesions may be of value in HPV detection, particularly when the lesions are located at keratinized sites. This sampling procedure may allow more accurate diagnosis of HPV infection compared with sampling only the lesion site, and may also represent a reliable method to investigate the biological characteristics of HPV infection and related oral carcinogenesis.