Preoperative quantitative diffusion tensor imaging on the spinal nerve root is a predictor for the surgical outcome of lumbar disc herniation: a prospective study of 117 patients.

Abstract

The diffusion tensor imaging (DTI) parameters (fractional anisotropy [FA] and apparent diffusion coefficient [ADC]) are commonly used to provide quantitative information on tissues. This study aimed to evaluate the predictive value of preoperative DTI of the spinal nerve roots in the surgical outcome of lumbar disc herniation (LDH). A total of 117 LDH patients were included. According to the postoperative improvement rate, the patients were dichotomized into the unfavorable group (N.=35) and favorable group (N.=82). The favorable group had a younger age (P=0.005) and a shorter disease course (P<0.001) than the unfavorable group. The favorable group had higher affected side FA and ADC and lower healthy/affected FA and ADC ratio than the unfavorable group (all P<0.05). Logistic regression analysis showed that younger age, shorter disease course, higher affected side FA and ADC, lower healthy/affected FA and ADC ratio, and lower healthy side ADC were the independent factors associated with positive surgical outcome. ROC analysis showed that the affected side FA had an excellent predictive performance for the surgical outcome (AUC=0.900). The Healthy/affected FA ratio had a good predictive performance (AUC=0.846). The overall predictive accuracy ranged from 0.91 to 0.92. However, ADC had poor predictive performance (AUC ranged from 0.626 to 0.663). These results suggested the preoperative affected side FA value had an excellent predictive performance for the surgical outcome of LDH patients. The LDH patients with a higher preoperative affected side FA value were more likely to have a positive surgical outcome.