Anatomo-radiological correlation between diffusion tensor imaging and histologic analyses of glial tumors: a preliminary study.

BACKGROUND AND PURPOSE

The challenge of the neurosurgical management of gliomas lies in achieving a maximal resection without persistent functional deficit. Diffusion tensor imaging (DTI) allows non-invasive identification of white matter tracts and their interactions with the tumor. Previous DTI validation studies were compared with intraoperative cortical stimulation, but none was performed based on the tumor anatomopathological analysis. This preliminary study evaluates the correlation between the preoperative subcortical DTI tractography and histology in terms of fiber direction as well as potential tumor-related fiber disruption.

METHODS

Eleven patients harboring glial tumors underwent preoperative DTI images. Correlations were performed between the visual color-coded anisotropy (FA) map analysis and the tumor histology after "en bloc" resection. Thirty-one tumor areas were classified according to the degree of tumor infiltration, the destruction of myelin fibers and neurofilaments, the presence of organized white matter fibers, and their orientation in space.

RESULTS

After histologic comparison, the DTI sensitivity and specificity to predict disrupted fiber tracts were respectively of 89% and 90%. The positive and negative predicted values of DTI were 80% and 95%. The DTI data were in line with the histologic myelin fiber orientation in 90% of patients. In our series, the prevalence of destructed fiber was 31%. Glioblastoma WHO grade IV harbored a higher proportion of destructed white matter tracts. Lower WHO grades were associated with higher preservation of subcortical fiber tracts.

CONCLUSION

This DTI/histology study of "en bloc"-resected gliomas reported a high and reproducible concordance of the visual color-coded FA map with the histologic examination to predict subcortical fiber tract disruption. Our series brought consistency to the DTI data that could be performed routinely for glioma surgery to predict the tumor grade and the postoperative clinical outcomes.