To the best of our knowledge, this is the first study that conducted both RESOLVE DTI and 3D T2-SPACE imaging to investigate the longitudinal changes of the cisternal segment of the trigeminal nerve, REZ, nuclear zone, the center of PCT, and brain pain-matrix regions in the patients with TN before and after treatment. Several findings presented herein may help understand the etiology and pathophysiology of TN. First, the trigeminal nerve REZ and nuclear zone exhibited more changes than the cisternal segment of the trigeminal nerve and the GMVs of brain pain-matrix regions before and after treatment. Second, the contralateral REZ and nuclear zone were significantly altered after treatment most likely owing to Wallerian degeneration. Third, the AD of the middle part of the cisternal trigeminal nerve, the FA of bilateral nuclear zones, and the MD of the center of PCT significantly changed over time before and after treatment. Fourth, the difference of GMV in the pain-matrix regions between the two hemispheres exhibited similar trends to the VAS before and after treatment. Finally, the DTI indices in the cisternal trigeminal nerve (FA) and nuclear zone (RD and MD) of the affected side were significantly correlated with the GMVs in specific pain-matrix regions (PAG, right S1, and left thalamus) before and after treatment. These findings are further discussed in the following parts of the document.
In the DTI analysis, the results showed that patients with TN exhibited slightly lower AD and FA values with no statistical significance in the affected side of the cisternal segment of the trigeminal nerve before treatment. The significant changes in the AD, RD, MD, and FA values in the affected side after treatment suggested that the treatments significantly altered the microstructural diffusion in the trigeminal nerve. In contrast, the REZ and nuclear zones yielded significant differences in the values of the DTI indices between the two sides before and after treatment, thus indicating that TN led to more tissue alterations in the REZ and nuclear zones than the cisternal trigeminal nerve, most likely owing to the fact that 67% (8/12) of studied patients exhibited NVC at REZ.
Our study also demonstrated that patients with TN not only had lower FA but also had higher AD, RD, and MD values at the affected REZ. Consistent with a prior study [24], our findings suggested that the increased MD and RD may be linked to NVC-induced focal demyelination in the trigeminal nerve REZ [25], neuroinflammatory processes, and/or edema [26] that affected the trigeminal system. Therefore, DTI can detect subtle pathological features at the trigeminal nerve REZ supporting a role for this regional involvement in TN pathophysiology. Moreover, the present study performed the treatments with a focus on the middle part of the cisternal trigeminal nerve in the affected side. Thus, the significant changes of the FA values in the distal and contralateral REZ and nuclear zone suggested that the Wallerian degeneration occurred through the PCT [27, 28], wherein minor changes of the DTI indices were observed after treatment.
The present study employed 3D T2-SPACE images to examine cortical/subcortical brain GMVs based on VBM analysis. A previous study reported that patients with TN had greater GMVs in the thalamus, contralateral S1, amygdala, frontal pole, PAG, primary motor cortex, and basal ganglia and cortical thinning in the orbitofrontal cortex, pregenual ACC, and insula [29], whereas anther study reported that patients with TN showed GMV reductions including the frontal, temporal, and parietal areas, as well as in the left thalamus and right cerebellum [30]. Our study found that patients with TN had significantly lower GMVs in PCC, ACC, insula, amygdala, and S1 in the affected side compared with the contralateral side before treatment. These findings may reflect unique symptoms because TN is characterized by paroxysmal pain triggered by innocuous stimuli or movements and does not involve major sensory losses. However, the differences between the previous findings and ours may be attributed to the mixed trigeminal pain patient groups and the different methodology (T1-VBM versus T2-VBM) used to assess GM changes.
In addition, our results demonstrated that TN led to significant GMV differences between the two hemispheres in some pain-related brain regions (PCC, ACC, insula, amygdala, and S1) and that the therapy helped reduce the GMV difference at 1 month after treatment. However, the GMV differences in some pain-related regions (PCC, insula, amygdala, and S2) became significant again at 6 months after treatment but returned to insignificant outcomes at 12 months after treatment. The sequential changes of GMV in the pain-matrix regions were generally consistent with the VAS in patients with TN. Initially, the pain intensity was measured to e as high as 9.25 ± 0.97 before treatment and was significantly reduced to 0.08 ± 0.29 after 1 month of treatment. However, at 6 months after treatment, two patients who exhibited degraded pain intensity, which resulted in a slight increase in VAS (0.92 ± 2.61), received a second treatment. After the second treatment, one patient had improved symptoms with the VAS reduced from 9 to 2, but the other had deteriorated symptoms with a VAS increased from 2 to 4. As a result, the overall VAS was slightly decreased after 12 months of treatment. Moreover, the correlation analysis revealed significant correlations between the DTI indices of the cisternal trigeminal nerve and nuclear zone and the GMVs of the pain-matrix regions in the affected side before and after treatment. These findings suggested that the changes of GMV in the pain-matrix regions were likely associated with the changes of VAS and that VBM analysis of T2-SPACE was helpful in the noninvasive monitoring and reflection of the pain intensity in patients with TN before and after treatment.
The etiology and pathophysiological mechanisms of TN are still not well understood, and the central contributions in TN are still debated. The most common theory of the classical TN etiology is a peripheral theory that involves the compression of the trigeminal nerve's REZ by blood vessels owing to the surgical experience [5]. Moreover, several studies demonstrated pathological changes and most notably the demyelination of the trigeminal nerves in patients with TN [6, 31–35]. In the present study, despite the fact that minor changes were observed in the DTI indices in the cisternal segment of the trigeminal nerve, no statistical significance was revealed between the affected and contralateral sides before treatment. Nevertheless, the REZ and nuclear zone yielded significant changes in the DTI indices before treatment and more significant changes than the cisternal segment of trigeminal nerve after treatment. This suggested that the REZ and nuclear zone were more sensitive to TN before treatment and had more demyelination after treatment compared with the cisternal segment of the trigeminal nerve.
However, the NVC theory of TN cannot sufficiently explain the disorder because some individuals can develop TN in the absence of NVC, and some individuals with NVC never develop TN. In the present study, 25% (3/12) of the studied patients developed TN without NVC. This suggests that the TN symptoms were not fully attributable to the presence of NVC. It is known that peripheral nerve injury can lead to central nervous system (CNS) plasticity [36, 37] most likely owing to the sustained or repetitive activation of primary afferent fibers and central sensitization [38]. In the present study, our results demonstrated that the changes of DTI indices in the trigeminal nerve were associated with the alterations of GMVs in brain pain-matrix regions before and after treatment. Furthermore, the longitudinal changes of GMV in the pain-matrix regions were generally consistent with the sequential changes of VAS, thus suggesting that the CNS changes in conjunction with trigeminal nerve injury contribute to TN symptoms.
The study is associated with some limitations. First, the small sample size of this study may lead to a low statistical power. Thus, the results ought to be interpreted with care. Second, some patients did not undergo a follow-up MRI scan owing to refusal or dropout. Thus, the incomplete dataset may affect the statistical results. Third, comparisons were not conducted between patients with TN and healthy controls in DTI and VBM analysis because only four age- and sex-matched healthy controls were enrolled in this study. Further investigations with larger groups will be needed to comprehensively compare the differences of cisternal trigeminal nerve and brain pain-matrix regions among healthy subjects and patients with TN before and after treatment.