AN IMAGE-ANALYSIS STUDY ON NUCLEAR MORPHOLOGY IN METASTASIZED AND NON-METASTASIZED SQUAMOUS-CELL CARCINOMAS OF THE TONGUE

In a retrospective case-control study on 46 metastasized and 34 non-me tastasized primary tongue carcinomas, the nuclear morphology and chrom atin pattern were assessed in 3 mu m thick, formalin-fixed, paraffin-e mbedded, and Feulgen-stained tissue sections of surgical resection spe cimens, by means of high-resolution computer-assisted image analysis. The aim of this study was to disclose differences in karyometric featu res, such as nuclear size-, shape-, and chromatin-pattern features, be tween these groups, with a view to developing a discriminant function that can predict the occurrence of metastasis for the individual patie nt. In addition, the lymph node metastases of 31 patients and the norm al tongue epithelium of 21 patients were also assessed, to study the p ossible differences between these two groups and primary tumours. In t he metastasized tumours, the chromatin was significantly more condense d (P=0 . 01) and exhibited significantly less variation in chromatin c ondensation (P<0 . 001) than in the group of non-metastasized carcinom as. Comparison of lymph node metastases with their primary tumours dis closed only minor differences in chromatin pattern. These findings sug gest that only minor genetic differences exist between primary tongue carcinomas and their metastases. Tumour cells of tongue carcinomas sho wed highly significant differences from cells of normal tongue mucosa for most karyometric features. Logistic regression analysis resulted i n a classifier, based on the circularity of the nucleus (CIRC) and the standard deviation of the chromatin condensation (SD COND), to predic t the occurrence of lymph node metastases. After cross-validation, the percentages of correct classifications in the group of metastasized a nd non-metastasized tumours were 72 and 62 per cent, respectively. The se results are comparable to the classification results obtained from a classifier based on the clinical T-stage, but our karyometric classi fication results show a much more equal distribution between the sensi tivity and specificity. Karyometric features appeared to be more appro priate to predict metastases than biomarkers such as p53, bcl-2, and K i-67. (C) 1998 John Wiley & Sons, Ltd.