Computerized nuclear morphometry for the prediction of inguinal lymph nodes metastases in squamous cell carcinoma of...

This study evaluated the sensitivity and specificity of computerized morphometry in predicting lymph nodes metastases (LNM) in patients with squamous cell carcinoma (SCC) of the vulva. Histologic samples obtained from 20 consecutive cases of SCC of the vulva with positive inguinal LNM were morphometrically assessed and compared with samples from 20 consecutive cases of vulvar SCC negative for LNM. Computerized morphometry was performed on tumor cells and on adjacent nonneoplastic epithelial cells located 2-4 mm from the tumor margins. Computerized morphometric variables of tumor cell nuclei in patients with negative LNM significantly differed from those in patients with positive LNM. Morphometric differences in nuclear size and contour regularity were detected when comparing the nonneoplastic nuclei adjacent to the tumor of both groups. Multivariate analysis showed that the only independent predictors of LNM were the depth of the invasion (P= 0.005) and the mean nuclear roundness of the nonneoplastic nuclei adjacent to the tumors (P= 0.008). Using these variables, a discriminant score revealed a sensitivity of 90% and a specificity of 86.4% for predicting LNM in SCC of the vulva. Our data suggest that cells from the primary tumors with LNM differ morphometrically from primary tumors with no LNM. In addition, normal epithelial cells adjacent to the tumor express morphometric changes between the two groups. The results of our study justify the need for a prospective study of a larger number of patients to evaluate the reproducibility and the clinical use of the data.