Feasibility of a concanavalin A-peroxidase labeling method to detect cancerous and precancerous lesions of the uterine cervix.

For the detection of cancerous and precancerous lesions in cervical cytopathology, the feasibility of a concanavalin A-peroxidase labeling procedure was tested and compared with the Papanicolaou method. To this end, the percentage of labeled flattened epithelial cells with a morphologically normal appearance present in cervical cell suspensions was determined. It was found that the mean labeling percentage of the control group was 71% (SD, 11%). The means for mild, moderate, and severe dysplasia groups were, respectively, 54% (SD, 19%), 48% (SD, 13%), and 44% (SD, 16%). The mean for the carcinoma in situ group was 32% (SD,11%), and for the squamous cell carcinoma group 16% (SD, 5%). It appeared that the labeling percentage gradually decreases with increasing atypia of the epithelium as confirmed by histological observation. A complete dissimilarity was found between healthy individuals and cancer patients. In a follow-up study it was found that the mean labeling percentage did not alter in cases of an unchanged stage of disease. A reestablishment of the normal concanavalin A-peroxidase labeling percentage often appeared once the cancerous or precancerous lesion was treated. In conclusion, the concanavalin A-peroxidase labeling method can be considered as a supplementary technique to the Papanicolaou method for the early detection of cervical cancer. It reduces the effect of sampling and screening errors of the Papanicolaou method, and it allows a more objective cytological diagnosis. In addition, it may possess prognostic significance.