Expression of p16 INK4A in Papanicolaou smears containing atypical squamous cells of undetermined significance from the uterine cervix

Abstract

Objective This study aimed to verify one of the major diagnostic dilemmas in routine Papanicolaou (Pap) smears. Atypical squamous cells of undetermined significance (ASCUS) were tested immunocytochemically for p16 INK4A (p16), and results were correlated with follow-up biopsies for more accurate diagnoses. Methods The study included 66 Pap smears of ASCUS diagnostic categories, all of which were correlated histologically. The cytological diagnoses of ASCUS were further classified cytologically according to the 2001 Bethesda System, as “atypical squamous cells of undetermined significance (ASC-US)” or “atypical squamous cells which cannot exclude HSIL (ASC-H).” All Pap smears were decolorized and immunostained with the primary anti-p16 antibody, clone E6H4. Immunoreactivity for p16 was correlated with histological sections (which were also immunostained for comparison) in a semiblind fashion. Results Of the 66 smears containing ASCUS, 47 (71%) were reclassified as ASC-US and 19 (29%) as ASC-H. Follow-up biopsies revealed that 21 (32%) cervices had no obvious abnormalities but only reactive changes. A significant proportion of histological diagnoses were CIN1/LSIL (24 cases, 36%), CIN2 or 3/HSIL (17 cases, 26%), squamous cell carcinoma (two cases, 3%), or AIS/adenocarcinoma (two cases, 3%). The p16 immunocytochemical stain was reactive in 40 (60.6%) of 66 smears: either weakly/sporadically (18 cases, 45%) or strongly positive (22 cases, 55%). Conversely, 26 (39.4%) of the smears were negative for p16 and displayed predominantly reactive changes. However, five cases of LSIL and one of HSIL were negative for p16. From the results of p16 immunoreactivity of atypical cells for detection of biopsy-proved significant lesions (HSIL or higher), this analysis was highly sensitive (sensitivity, 95%) and specific (specificity, 96%) and had favorable positive (91%) and negative (98%) predictive values. Conclusion On the basis of both morphological and immunostaining patterns, there is a clear association between strong p16 immunostaining of atypical cells in the smears and the presence of significant lesions in the cervix, except in two patients. Similarly, there is a clear association between the lack of p16 expression and the absence of cervical lesions. The p16 immunocytochemical stain can be applied successfully to conventional Pap smears and may serve as a useful biomarker in ASCUS-containing smear diagnoses. This may offer a more objective parameter to help clarify this ambiguous area in gynecological cytopathology.