Is p16^INK4A overexpression a reliable biomarker for oncogenic human papillomavirus infection in normal tonsils?

4467

Recent analyses have provided evidence that human papillomavirus (HPV) is an etiological risk factor for head and neck squamous cell carcinomas (HNSCC). Oncogenic HPV, mainly type 16, has been strongly associated with tonsillar carcinomas and appears to be of prognostic significance. Because HPV-positive tonsillar carcinomas also accumulate p16^INK4A this inhibitor of cell proliferation might be a candidate biomarker for clinical diagnosis. Data on the value of p16^INK4A overexpression to predict HPV infection in normal and dysplastic tonsil tissue, however, are scarce. Here we studied the prevalence of p16^INK4A overexpression by using immunohistochemistry on 4 μm thick formalin-fixed, paraffin-embedded sections from normal tonsil samples of 263 patients (mean age: 29.2; range 12–70 years). These patients were treated in 1997-1998 for non-oncologic reasons, which mostly included tonsillectomy. In addition, genomic DNA was extracted from these samples for HPV-specific PCR analysis. Strong p16^INK4A-positive and a number of weakly positive and negative samples (n=28) were subsequently screened by PCR for the presence of adenovirus and cytomegalovirus DNA. All 263 samples contained squamous cell epithelium and 187/263 harbored lymphoid tissue, 177 of which contained tonsillar crypt epithelium. Diffuse cytoplasmic and nuclear p16^INK4A staining was only identified in crypt epithelium (49/177; 27.7%) and germinal centers (52/187; 27.8%), which was strongly associated with each other (p=0.001). PCR analysis of all 263 tonsil samples revealed 8 HPV-positive cases (3%), including one with HPV 16 and 18, one with HPV 16, 18 and 68, and one with HPV 27. Only the former case was associated with p16^INK4A overexpression, whereas the remaining 7 HPV-positive cases were p16^INK4A negative. PCR analysis revealed no detection of adenovirus and cytomegalovirus DNA in 28 tonsil samples (17 strongly p16^INK4A positive, 5 weakly positive and 6 negative). Our results show a rather high percentage of normal tonsillar crypt epithelia with increased p16^INK4A expression, which only in 1 case appears to be associated with HPV. In contrast, 7 p16^INK4A-negative cases were positive by PCR analysis. Thus, p16^INK4A should be used with caution as biomarker for HPV on normal tonsil tissue. Other mechanisms of p16^INK4A activation are therefore implicated, which do not point to adenovirus or CMV infection. This study was supported in part by the Medical Research Foundation (“Profileringsfonds”) of the University Hospital Maastricht, grant no. PF126.