Can preoperative cervical spinal diffusion tensor imaging (DTI) indices predict surgical outcomes in patients with...

Anterior cervical discectomy and fusion (ACDF) surgery can effectively prevent disease progression in patients with Hirayama disease (HD) and diffusion tensor imaging (DTI) can quantitatively assess spinal cord function. In this study, we aimed to evaluate the relationship between preoperative spinal DTI indices and the clinical outcomes of patients with HD when treated by ACDF.

We retrospectively analyzed 35 HD patients treated by ACDF. We collated a range of DTI indices, including fractional anisotropy (FA) and apparent diffusion coefficient (ADC) values, prior to surgery with patients in flexion and neutral positions. Patients were divided into improvement (Im) group and non-improvement (Nim) group according to the Odom score, and the difference in surgical outcomes between the two groups was confirmed by quick disabilities of the arm, shoulder and hand (Q-DASH) scores. The DTI indices in the two groups of patients were then compared. Receiver operating characteristic (ROC) curves and area under curve (AUC) were used to evaluate the predictive capability. The correlation between Q-DASH scores and DTI indices was also evaluated.

The FA values in the two groups of patients differed significantly in the cervical flexion position and the different segments were mainly located in the lower cervical spinal cord including the flexion C5/6 (Im group vs. Nim group: 0.501 ± 0.078 vs. 0.362 ± 0.087, P < 0.001) and C6/7 (Im group vs. Nim group: 0.455 ± 0.097 vs. 0.347 ± 0.102, P = 0.003) FA values, the mean FA value for C4/5-C6/7 (Im group vs. Nim group: 0.471 ± 0.067 vs. 0.372 ± 0.078, P < 0.001), mean FA value for C5/6-C6/7 (Im group vs. Nim group: 0.478 ± 0.076 vs. 0.354 ± 0.083, P < 0.001) and mean FA value for the two minimal segments (Im group vs. Nim group: 0.442 ± 0.078 vs. 0.341 ± 0.081, P = 0.001). The ADC values were similar to FA values. The ROC curve for DTI indices in the lower cervical spinal cord had an AUC > 0.7 including: flexion FA value and ADC value for C5/6 (0.877 and 0.931), flexion FA value and ADC value for C6/7 (0.778 and 0.761), flexion mean FA value and ADC value for C4/5-C6/7 (0.846 and 0.859), flexion mean FA value and ADC value for C5/6-C6/7 (0.861 and 0.905), flexion mean FA value and ADC value for the two minimal/maximal segments (0.815 and 0.892). DTI indices including FA value and ADC value were correlated with the preoperative score, final follow-up score and improvement percentage. Flexion ADC value was correlated with improvement score but flexion FA value was not correlated with improvement score.

Preoperative DTI indices of the spinal cord, especially those of the lower cervical spinal cord with patients in the flexion position, can predict the clinical outcome of patients with HD post-surgery. In general, a larger FA value and a smaller ADC value indicate a better surgical outcome.