Quantitative cytopathology of endometrial lesions

Abstract

Morphometric and multivariate statistical methods were used to discriminate endometrial carcinoma from benign cells in cytologic studies. Clumps of epithelial cells that appeared most diagnostically relevant were selected from aspirated samples of 70 endometrial cancer patients. The cells' cytologic character was reduced to a combination of five quantitative parameters--nuclear size, degree of anisokaryosis, nuclear from index, homogeneity of nuclear chromatin texture, and regularity of nuclear arrangement. The 5-variate cluster analysis demonstrated that the 70 cases could be classified into three definite groups: Group A (17 cases) was characterized by cells of small nuclear size, slight anisokaryosis, homogeneous chromatin texture, and regular nuclear arrangement; Group C (12 cases) by cells of large nuclear size, marked anisokaryosis, heterogeneous chromatin texture, and irregular nuclear arrangement; and Group B (41 cases) by cells of intermediate parameter values. Group C was derived from 10 cases of adenocarcinoma and 2 of atypical hyperplasia, while Groups A and B were not derived from any cases of malignancy. The computer-assisted morphometric statistical method can objectively classify the endometrial cells into malignant and benign, with improved validity and reproducibility. The cytopathologic finding, if detected by this method, may serve as a surrogate endpoint biomarker.