Immunocytochemistry in liquid-based cervical cytology: analysis of clinical use following a cross-sectional study.

Cytological screening for cervical cancer is hampered by imperfect sensitivity and low inter-observer reproducibility. Human papillomavirus (HPV) testing lacks specificity as a primary screening method. Studies indicate that immunocytochemical detection of alterations caused by HPV in the host cells can optimise screening. Here, the potential of p16(INK4a) (cyclin-dependent kinase inhibitor p16) and MIB-1 (Ki-67 proliferation marker) as adjunct molecular markers for cervical lesions was investigated in a prospective, cross-sectional study of 500 samples in the framework of opportunistic screening in Flanders, Belgium. A consecutive series of 200 samples and 100 samples from the cytological categories ASC, LSIL and HSIL were investigated. Surepath samples were interpreted according to the Bethesda 2001 reporting system. HPV testing was done with MY09/MY11 consensus PCR. Immunocytochemistry for p16(INK4a) and MIB-1 was performed with an automated staining protocol. The number of immunoreactive cells/1,000 cervical cells was assessed. There was a higher mean number of p16(INK4A) and MIB-1 immunoreactive cells/1,000 cells in HSIL (4.06 +/- 1.93 and 11.13 +/- 2.83, respectively) compared to other cytological categories. Both markers showed a large spread in counts, for all categories. In cases of HSIL without immunoreactive cells for either marker, low cellularity and long-term storage in water were often the cause of false negativity. This study confirms that positive staining for p16(INK4a) and MIB-1 is highly correlated with presence of high-grade lesions. These markers could be used as adjuncts to increase the sensitivity of cytological screening as well as the specificity of the HPV test. However, clear methodological standards are needed for optimal performance of immunocytochemistry in a clinical setting.