The Ki-67 Labeling Index Is Not a Useful Predictor for the Follow-up of Cervical Intraepithelial Neoplasia 1

OBJECTIVE: Our aim was to determine whether the Ki-67 immunostaining pattern, present on diagnosis of cervical intraepithelial neoplasia (CIN), predicts the change from low-grade to high-grade CIN over a 2-year period after diagnosis. MATERIALS AND METHODS: Of 59 cervical biopsy samples from 59 patients diagnosed as having cervical CIN, 35 were diagnosed as CIN 1 and 24 were diagnosed as CIN 2 or CIN 3. The Ki-67 immunostain showed immunopositive cells in the upper two thirds of the epithelium in all specimens. Two hundred nuclei were counted in 25 high-power fields in each specimen, including all of the epithelial layers, to determine the mean number of Ki-67-positive cells. In situ hybridization was used to demonstrate and type human papillomavirus. The chi test, Fisher exact test, Student t test, one-way analysis of variance, and Tukey test were used for statistical analysis, with significance set at p < .05. RESULTS: The mean Ki-67 labeling index for CIN 1, CIN 2, CIN 3, and CIN 2,3 were, respectively, 32.5%, 43.2%, 53.2%, and 47.8%. The statistical study showed significant differences between CIN 1 versus CIN 2, CIN 1 versus CIN 3, and CIN 1 versus CIN 2,3. For CIN 1, the mean Ki-67 labeling index was 32.8% when the lesion disappeared and was 34.6% for persisting lesions. There was no statistically significant difference. CONCLUSIONS: Ki-67 labeling index did not predict persisting CIN 1.