The significance of nuclear morphometric variables as prognostic predictors in breast cancer.

The preoperative biopsies from primary breast carcinomas of 504 women were subjected to interactive morphometric analysis of a) the mean nuclear area (NA), b) standard deviation of nuclear area (SDNA), c) mean area of the 10 largest nuclei (NAl0), d) nuclear perimeter (PE), e) standard deviation of nuclear perimeter (SPDE), f) largest nuclear diameter (Dmax) and g) shortest nuclear diameter (Dmin), h) histological grade and i) classical prognostic variables. The above data were correlated with the disease outcome during the mean follow-up period of 11.2 years. Tumor size (p less than 0.0001), morphometric variables (p = 0.0001-0.005) and histological grade (p = 0.03) predicted axillary lymph node metastasis at the time of diagnosis. According to multivariate analysis, tumour size and NA predicted the axillary lymph node metastasis independently. Axillary lymph node status (p less than 0.0001) and histological grade (p = 0.01) predicted the tumour recurrence and recurrence-free survival, whereas the morphometric variables had no significant predictive value. Axillary lymph node status (p less than 0.0001), tumour size (p less than 0.0001), histological grade (p = 0.0012) and morphometric variables (p = 0.003-0.035) predicted the disease-related survival. Of the morphometric variables, NA and the Dmin were the two most important predictors of tumour-related survival in univariate analysis. Dmax had independent prognostic information in multivariate survival analysis. In the same analysis, tumour size and axillary lymph node status were more important predictors. In conclusion, the morphometric variables analysed have independent predictive value in female breast cancer. Their value is, however, inferior to that of the tumour size and axillary lymph node status, but equal to that of the histological grade.