p16(INK4a) as a complementary marker of high-grade intraepithelial lesions of the uterine cervix. I: Experience with squamous lesions in 189 consecutive cervical biopsies.

AIM: To test the usefulness of p16(INK4a) immunostaining for improving the diagnostic accuracy of cervical punch biopsies referred to a routine laboratory setting during the investigation of women with abnormal Papanicolaou smears.
METHODS: A total of 188 consecutive and unselected colposcopically directed cervical biopsies and a single contemporaneous cervical polyp were accessioned prospectively over a 3-month period, step-serially sectioned and examined by H&E and immunostained for p16(INK4a). The clinical context, results of concurrent Papanicolaou smears/ThinPrep slides and Digene hybrid capture tests for high-risk human papillomavirus (HPV) subtypes, as well as follow-up cervical smears/ThinPrep, biopsies and loop excisions of transformation zones or cone biopsies were all correlated with the morphological and immunohistochemical findings.
RESULTS: Seventy-seven biopsies (40.7%) displayed a high-grade squamous intraepithelial lesion (HGSIL; cervical intraepithelial neoplasia [CIN] 2-3), 27 (14.3%) showed a low grade squamous intraepithelial lesion (HPV +/- CIN1) and 85 (45%) showed a range of non-dysplastic (inflammatory or reactive) changes. Diffuse strong parabasal immunostaining for p16(INK4a), suggestive of integrated high-risk HPV DNA into the host genome, was observed in 81 biopsies (42.9%, including the cervical polyp) and correlated (>90%) with HGSIL in the H&E sections. Only one case revealed irreconcilable discordance between the histological features and this strong parabasal immunostaining pattern. Focal and weaker midzonal or superficial p16(INK4a) immunostaining, suggestive of episomal HPV infection, was noted in 19 biopsies (10%) and these biopsies exhibited a range of histological changes but predominantly low grade squamous intraepithelial lesion (LGSIL). No staining of the squamous epithelium was seen in 89 biopsies (47.1%). Again, only one case revealed irreconcilable discordance between the histological features and this negative immunostaining pattern. On review of all cases where discordant results were noted between the H&E appearances and expected p16(INK4a) immunostaining, we found 26 cases (13.7%) in which this discordance prompted justifiable modification of the original diagnosis.
CONCLUSIONS: Thus, within a routine diagnostic laboratory, p16(INK4a) immunostaining appears to be a very useful adjunctive test in the examination of colposcopically directed cervical biopsies, in the diagnostic cascade of women investigated for abnormal Papanicolaou smears. It is possible, as further data accumulate concerning the importance of integration of high-risk HPV DNA into the host cell genome and the reliability with which this can be identified by p16(INK4a) immunostaining, that this will become the diagnostic 'lesion of interest', replacing the subjective histological grading of cervical dysplasia, in the management of such patients; i.e., the discriminatory watershed between continued surveillance and active intervention.