Expression of MN/CA9 protein in Papanicolaou smears containing atypical glandular cells of undetermined significance is a diagnostic biomarker of cervical dysplasia and neoplasia.

Abstract

Despite the enormous impact that Papanicolaou (Pap) smear screening has had on the incidence of cervical carcinoma in developed countries, there is still an unacceptably high frequency of occurrence of this cancer. In part, this is due to human error associated with cytologic diagnoses of Pap smears. Also, the use of new sampling devices, such as the cytobrush, has increased the complexity of diagnosing benign and neoplastic cervical cytology. This is particularly apparent in the diagnosis of atypical glandular cells of undetermined significance (AGUS). Approximately 40% of AGUS diagnoses have a corresponding significant lesion at biopsy follow-up, and 60% do not. There is clearly a need for an adjunct to cytologic diagnosis that can readily identify AGUS smears that are diagnostic of significant lesions. The authors have identified the MN/CA9 antigen as a strong candidate for an adjunct biomarker. A total of 245 Pap smears of all AGUS diagnostic categories with histologic confirmation were studied. The median age of the patients was 39 years. The Bethesda system classification (AGUS-favor reactive, AGUS-not otherwise specified, and AGUS-favor neoplastic) was used. All of the Pap smears were decolorized and immunostained with monoclonal antibody to MN/CA9 antigen by the immunoperoxidase technique. The results of MN/CA9 immunoreactivity were correlated with the histologic data in a semiblinded fashion. The follow-up biopsies showed that a high percentage (70%) of patients had low and high grade cervical intraepithelial neoplasia lesions, respectively (CIN I and CIN II or III). Clinically significant lesions-adenocarcinoma in situ/carcinoma (AIS/CA) and CIN II or III-were found in 50% of the cases. Among these, 11% were AIS/CA. In the three subcategories of AGUS diagnosis, the AGUS-not otherwise specified showed the broadest range of lesions in the follow-up biopsies. Three patterns of MN/CA9 immunoreactivity were observed in the Pap smears: 1) atypical cells, 2) normal endocervical cells only, and 3) all cells negative. All Pap smears that were MN/CA9 positive were histologically confirmed to be clinically significant lesions or CIN I; in addition, there were a very small number (n = 12) of cases of atypia. None of the benign lesions showed MN/CA9 expression in the corresponding Pap smears. Furthermore, the pattern of atypical cell immunostaining identified all cases with significant lesions (AIS/CA and CIN II or III) in the cervices. Conversely, the majority of CIN I cases (82%) and all cases of atypia showed positive immunostaining restricted to normal endocervical cells only. There is a clear association between MN/CA9 immunostaining of atypical cells and the presence of significant lesions in the cervix. Similarly, there is a clear association between lack of expression of MN/CA9 and the absence of cervical lesions. However, the screen does not allow discrimination between CIN I and atypia. The authors also found that, based on the combined patterns of morphology and immunostaining, they are able to discriminate between AIS and CIN II or III in AGUS Pap smear diagnoses. Thus, expression of the MN/CA9 antigen is indeed a discriminator of significant lesions in AGUS Pap smear diagnoses.