Quantitative investigation of nuclear morphology in glioblastomas and its relation to survival time

To investigate the influence of nuclear morphology of tumor cells on survival time of patients with primary glioblastomas.

Tumor cell nuclei have been measured in paraffin sections from 51 glioblastomas (Ki-67 immunostaining). In each tumor, the region with the highest proliferative activity has been selected for performing the morphometric analysis. Nuclear area, shape variables (roundness factor, Fourier amplitudes) and the proliferation index Ki-67 have been determined. Statistical relationships between these variables have been tested by principal component analysis and Spearman's correlation analysis. The survival time has been tested using Kaplan-Meier analysis. The influence of morphometric variables on survival time and on the time until recurrence has been tested by Cox analysis.

There is a significant correlation of mean value and standard deviation of nuclear area with shape variables. Cox analysis showed a significant influence of the quantitative morphologic variables on survival time for patients with and without complete surgical resection. For patients with complete surgical resection, there was a distinct influence of mean nuclear area on survival time. Patients with incomplete surgical resection of the tumors had a significantly shorter survival. No significant influence of the variables on the time from surgery to recurrence could be shown.

Quantitative morphology of tumor cell nuclei in our set of glioblastoma cases showed a statistically significant relation with the survival time of the patients, indicating the biological significance of nuclear morphology in glioblastomas.