Nerve matters: Longitudinal microstructural change in the trigeminal nerve is associated with durable pain relief after surgery for trigeminal neuralgia

Abstract

Background

Many patients with medically-refractory trigeminal neuralgia (TN) fail to achieve lasting pain relief following surgery targeting the trigeminal nerve (cranial nerve five; CNV). While some studies using MRI diffusion tensor imaging (DTI) suggest that preoperative CNV microstructure may predict surgical response, the findings remain inconsistent. Furthermore, the relationship between post-surgical CNV microstructural changes and long-term pain relief is not well understood. Using a novel CNV-nerve specific DTI protocol, the present study aimed to determine whether: (1) preoperative CNV microstructure differentiates surgical responders from non-responders and (2) sustained pain relief after surgery is associated with distinct postoperative microstructural changes in CNV.

Methods

We conducted a single-centre, prospective, longitudinal study in TN patients undergoing microvascular decompression (MVD) or percutaneous rhizotomy by balloon compression (BC). Patients underwent preoperative and postoperative (one week, one month, six months and one year) high-resolution DTI scanning of CNV using a novel fluid-attenuated inversion recovery DTI protocol. Healthy controls (HC) were scanned at a single timepoint using the same protocol. CNV microstructure was inferred primarily from fractional anisotropy (FA), supplemented with other diffusion metrics. Responders were defined as patients with immediate and complete pain relief (Barrow Neurological Institute facial pain scale I or IIIa) sustained for at least two years.

Results

Thirty-five TN patients (22 MVD and 13 BC) and 19 HC were studied. There was no difference in FA between HC CNV and affected ipsilateral or unaffected contralateral CNV in TN patients. However, CNV ipsilateral to the painful side of the face showed microstructural alteration in the form of reduced FA compared to the contralateral, unaffected CNV (0.45 vs. 0.49, p = 0.0017). This was largely driven by eventual surgical responders (n = 18, FA ipsilateral 0.45 vs. contralateral 0.49, p = 0.049), whereas non-responders (n = 17) showed no such difference (p = 0.15). Following surgery, responders showed early reduction in ipsilateral CNV FA by one month (0.45 vs. 0.38, p = 0.013), sustained at six months (0.38, p = 0.021) and one year (0.37, p = 0.006). The same pattern was observed for MVD and BC responders. Conversely, non-responders exhibited no significant postoperative CNV FA change. Postoperative pain-free timepoints were associated with significantly lower ipsilateral CNV FA compared to painful states or HC on average (0.39 vs. 0.45 or 0.47, p < 0.0001) and in individual patients experiencing multiple pain recurrences after repeat operations.

Conclusions

Long-term pain relief after TN surgery requires the induction of specific and sustained microstructural changes in the treated CNV, irrespective of surgical modality.

Graphical abstract

This is a visual representation of the abstract.

Keywords

diffusion tensor imaging microvascular decompression MRI percutaneous rhizotomy trigeminal nerve trigeminal neuralgia

Introduction

Trigeminal neuralgia (TN) is a chronic facial pain condition characterized by typically unilateral, shooting or electric shock-like pain attacks, which are usually triggered by innocuous stimuli, radiating through the distribution of the trigeminal nerve (cranial nerve five; CNV). Approximately two-thirds of TN patients have classical TN, where there is neurovascular compression (NVC) with morphological change in the CNV root as it emerges from the pons, at a site often labelled as the root entry zone (REZ).¹ The REZ represents a watershed transition point between centrally- and peripherally-mediated myelination, surmised to render it more susceptible to injury from vascular compression, although the term “REZ” is inexact, being defined variably in the literature and inadequately accounting for individual-to-individual variation in the extent of the centrally-myelinated portion of the CNV root.^2,3 Limited histologic examinations of CNV in classical TN demonstrate demyelination and dysmyelination, although the role of vascular compression as a cause of myelin disruption remains inconclusive.^4–7 Idiopathic TN, by contrast, has a similar clinical presentation to classical TN, but without associated NVC of CNV, and therefore a more uncertain pathophysiology.⁸

Neuroimaging studies using MRI diffusion tensor imaging (DTI) to examine CNV in TN, and classical TN in particular, are plentiful, with a summary provided in Watanabe et al.⁹ The most widely reported DTI measure is fractional anisotropy (FA), with lower FA indirectly reflecting microstructural alterations such as reduced axon density and demyelination.¹⁰ Several reports show reduced FA in TN patients at the proximal portion of the ipsilateral CNV (i.e. on the painful side of the face) compared to the contralateral CNV and healthy controls (HC).^11–26 However, some studies have found contradictory results,²⁷ including the study with the highest-resolution DTI protocol to date (voxel size 1.6 × 1.6 × 1.2 mm³).²⁸ Alterations in other diffusion metrics including mean diffusivity (MD), axial diffusivity (AD) and radial diffusivity (RD) have also been observed in TN, although less reliably.⁹ To some extent, technical challenges in DTI of the cranial nerves, such as poor image spatial resolution for such a small cross-sectional area, multiple tissue interfaces at the skull base causing deleterious image artifacts, and adjacent isotropic, rapidly diffusing cerebrospinal fluid leading to partial volume effects on diffusion indices,²⁹ may account for disparate findings of CNV microstructural change in TN.

Patients with medically-refractory TN are often offered surgery directed at CNV itself. In classical TN, microvascular decompression (MVD) is typically the procedure of choice, and aims to directly relieve NVC, often with excellent long-term pain relief. In addition to MVD, other procedures, more commonly used in idiopathic TN, achieve pain relief through intentional, limited injury to CNV, and include percutaneous rhizotomy or stereotactic radiosurgery.⁸ While some DTI studies suggest that preoperative microstructural characteristics of CNV may predict long-term pain outcome after surgery,^21,30,31 others have not,^11,32 and there remains considerable debate as to whether CNV DTI can be a predictive biomarker of surgical response in TN. Similarly, there are a paucity of data on the postoperative evolution or resolution of microstructural abnormalities in the surgically treated CNV, as well as considerable variability in the timing and overall duration of follow-up imaging across studies.⁹

We recently developed a high-resolution, cerebrospinal fluid-suppressed nerve-specific DTI protocol (fluid-attenuated inversion recovery (FLAIR)-DTI) permitting superior visualization of CNV with tractography that yields more accurate diffusion metrics.²⁹ Armed with this unique protocol, we carried out the most rigorous longitudinal DTI evaluation of CNV to date in TN patients, acquiring DTI data before and at multiple standardized time points after surgery. We hypothesized that preoperative differences in CNV microstructure (as primarily indicated by FA) would distinguish between eventual responders and non-responders to TN surgery, and that durable pain relief is associated with a specific pattern of postoperative microstructural changes in CNV.

Methods

Study participants

This was a prospective, longitudinal, single-centre study of patients undergoing surgical treatment for TN at the University of Alberta Hospital between 2017 and 2024, approved by the Health Research Ethics Board of the University of Alberta (Protocol 00067116).

Inclusion criteria: adults 18–80 years old with medically-refractory classical or idiopathic TN defined using International Classification of Headache Disorders, 3rd edition (ICHD-3) criteria¹; scheduled for surgical treatment by MVD or percutaneous rhizotomy with mechanical balloon compression (BC); and no previous surgical treatment for TN prior to enrollment.

Exclusion criteria: confirmed multiple sclerosis or other secondary cause of TN; formally diagnosed concurrent psychiatric or non-TN chronic pain illness; and any prior non-TN neurosurgical procedures; prior surgical treatment for TN.

Additionally, 19 HC subjects matched to the TN group in mean age and sex distribution, and without chronic pain or psychiatric conditions, were recruited.

Data acquisition and processing

Clinical characteristics and outcome assessment

The following demographic/clinical data were collected: sex; age; duration of TN since diagnosis; side-of-pain; preoperative pain severity measured using a 0–100 visual analog scale (VAS); surgery type (MVD or BC). Additionally, NVC severity scores were determined for each patient based on the scoring system of Sindou et al.³³ : 0 = no neurovascular contact or venous contact alone; 1 = arterial contact with no indentation of nerve root; 2 = arterial contact with displacement and distortion of nerve root; 3 = arterial contact with marked indentation in nerve root. NVC severity scores were derived by a single observer (co-author TS) from preoperative high resolution T2 weighted posterior fossa magnetic resonance imaging (MRI) scans and confirmed (in patients undergoing MVD) by intraoperative findings. A binary classification of NVC was also carried out per ICHD-III criteria¹: “true” NVC was considered present only in patients with a Sindou et al.³³ score of 2 or 3.

TN patients were classified as: responders – documented evidence of immediate, complete, and durable TN pain relief with or without medication for at least two years after surgery (i.e. Barrow Neurological Institute (BNI) facial pain score I or IIIa³⁴); non-responders – TN pain persisting immediately after surgery or early pain recurrence within two years of surgery (BNI II, IIIb, IV or V). We intentionally defined outcome in this manner to compare patients with no pain to those with any pain after surgery.

Timing of participant assessments

All TN patients underwent MRI scanning (acquisition details below) within one month prior to surgery (preoperatively) and then after surgery within 5–12 days (one week), 3–6 weeks (one month), 5–7 months (six months) and 11–14 months (one year). HC subjects underwent a single MRI scanning session. VAS score was recorded at each time point. Not all participants were able to be scanned at all postoperative time points: one week, n = 29; one month, n = 28; six months, n = 18; and one year, n = 16. Note that two-year clinical follow-up data were available for all TN patients.

MRI acquisition

MRI was acquired on a 3T Siemens Prisma (Siemens, Erlangen, Germany) with a 64-channel head radiofrequency coil. Our previously published CNV-specific FLAIR-DTI protocol was used with the following parameters²⁹: 13 slices, 182 × 182 acquisition matrix, 1.2 mm isotropic voxels, b = 1000 s/mm², 20 directions × 5 averages, 10 b₀, TR 3900 ms, TE 68 ms, TI 2300 ms, AP phase-encode direction, GRAPPA R = 2, phase partial Fourier = 6/8, interpolation with zero-filling (× 2) in the axial plane, and total scan time of 7 min and 22 s.

Trigeminal nerve tractography and diffusion metrics

DTI processing was performed using the ExploreDTI toolbox as previously described.^29,35 Deterministic tractography generated a 6.3-mm long CNV segment extending distally from the nerve's first emergence from the pons which included the site of NVC in cases where it was present. Diffusion metrics (FA, MD, AD and RD) were computed across the entire CNV segment.

Statistical analysis

Between-group comparisons

Continuous clinical variables and CNV diffusion metrics (with a focus on FA) were compared between HC and TN as well as responder (R) and non-responder (NR) groups using parametric (Student's t-test) and non-parametric statistical tests (Mann–Whitney U-test). Non-parametric Kruskal–Wallis tests with post-hoc pairwise comparisons were used to compare across three or more groups where appropriate. Categorical variables were compared between groups using chi-squares and Fisher's exact tests where appropriate.

Within-subject comparisons

CNV diffusion metrics were compared within individual subjects between nerves (ipsilateral vs. contralateral) using parametric (paired t-test) and non-parametric (Wilcoxon signed rank) tests where appropriate. Between timepoint comparisons were performed using the Kruskal–Wallis test to better account for patient drop out and sample size variation between timepoints.

Threshold for statistical significance for all comparisons was set at p < 0.05 (two-tailed). All statistical analyses were carried out with Prism, version 10 (GraphPad Software Inc., San Diego, CA, USA).

Results

Clinical characteristics and demographics

All TN patients

Clinical and demographic features of all 35 TN patients and 19 HCs are presented in Table 1. TN and HC groups were well matched in age (56.3 $\pm$ 11.2 years and 54.9 $\pm$ 9.4 years, p = 0.98) and sex distribution (21F/14 M and 10F/9 M, p = 0.70). Average TN duration was 5.6 $\pm$ 4.4 years. “True” NVC was identified in 16/35 TN patients, and average preoperative VAS was 75.1 $\pm$ 27.6 across the entire TN patient group. All patients included in the study underwent first-time surgical treatment by either MVD (n = 22) or BC (n = 13).

Table 1.

Comparison of demographic and clinical characteristics between trigeminal neuralgia (TN) patients and healthy controls (HC), as well as between responders and non-responders (data are the means $\pm$ SD)

	Responders	Non-responders	p-value (two-tailed)	TN	HC	p-value (two-tailed)
Total #	18	17	-	35	19	-
Sex (Female/male)	7/11	14/3	0.015*	21/14	10/9	0.70
Age (years)	60.2 $\pm$ 8.9	49.4 $\pm$ 11.0	0.003**	56.3 $\pm$ 11.2	54.9 $\pm$ 9.4	0.98
Duration of TN (years)	5.0 $\pm$ 3.6	6.3 $\pm$ 5.2	0.38	5.6 $\pm$ 4.4	NA	–
Preoperative VAS (mm)	76.7 $\pm$ 30.1	73.2 $\pm$ 25.2	0.72	75.1 $\pm$ 27.6	NA	–
NVC (yes/no)	10/8	6/11	0.31	16/19	NA	–
NVC severity (0/1/2/3)	4/4/5/5	9/2/4/2	0.26	13/6/9/7	NA	–
Surgery (MVD/BC)	13/5	9/8	0.31	22/13	NA	–

NA = not applicable; NVC = neurovascular compression; NVC severity graded according to Sindou et al.³³ (NVC was considered present if severity score was 2 or 3); MVD: microvascular decompression; BC: balloon compression rhizotomy; *p < 0.05, **p < 0.01.

Response to surgery

There were 18 responders and 17 non-responders to surgery. Females were more likely to be non-responders than responders (NR = 14 female/3 male; R = 7 female/11 male; p = 0.015). Non-responders were younger than responders (49.4 $\pm$ 11.0 vs. 60.2 $\pm$ 8.9 years, p = 0.003), but preoperative duration of TN did not differ between non-responders and responders (6.3 $\pm$ 5.2 vs. 5.0 $\pm$ 3.6 years, p = 0.38). There was no preoperative difference in VAS score (76.7 $\pm$ 30.1 vs. 73.2 $\pm$ 25.2, p = 0.72), NVC severity (p = 0.26) or type of surgery (p = 0.31) between responders and non-responders.

Preoperative CNV microstructure

All TN patients

Preoperative CNV diffusivity metrics for HC (N = 38 nerves, one on each side) and TN patients (N = 35) are shown in Figure 1. There was no difference between preoperative ipsilateral CNV FA in TN versus HC (Ipsi = 0.45; HC = 0.47; p = 0.61), but there were differences in MD (Ipsi = 0.0012 mm²/s; HC = 0.0011 mm²/s; p = 0.0233) and RD (Ipsi = 0.00085 mm²/s; HC = 0.00077 mm²/s; p = 0.0076). In TN patients, preoperative ipsilateral CNV FA was also reduced compared to contralateral CNV FA (Ipsi = 0.45; Contra = 0.49; p = 0.0017), with corresponding elevation in RD (Ipsi = 0.00085 mm²/s; Contra = 0.00079 mm²/s; p = 0.012).

Figure 1.

Preoperative trigeminal nerve (CNV) diffusion metrics. As shown in (a) and (c), there was no difference in preoperative FA or AD between healthy control CNV (HC, 19 nerves on each side, total n = 38) and affected ipsilateral (TN-Ipsi, n = 35) or unaffected contralateral (TN-Contra, n = 35) CNV in TN patients. However, as seen in (b) and (d), MD and RD were significantly higher in the ipsilateral affected CNV compared to HC. Additionally, FA was reduced in the ipsilateral compared to contralateral CNV in TN patients before surgery, (a). This corresponded to a significant increase in ipsilateral CNV RD (d). Individual data points are indicated with white circles. Black lines connect ipsilateral and contralateral nerve metrics within the same patient. *p < 0.05; **p < 0.01. AD = axial diffusivity; CNV = crainial nerve V; FA = fractional anisotropy; MD = mean diffusivity; RD = radial diffusivity; TN = trigeminal neuralgia.

Responders versus non-responders

Preoperative ipsilateral and contralateral CNV FA in surgical responders (n = 18) and non-responders (n = 17) are shown in Figure 2. In responders, ipsilateral CNV FA was reduced compared to contralateral CNV FA (0.45 vs. 0.49, p = 0.049), but no different compared to non-responders (p = 0.54) or HC (p = 0.21). There was also a corresponding increase in ipsilateral CNV RD versus contralateral RD in responders (0.00090 mm²/s vs. 0.00080 mm²/s, p = 0.0244). In responders, ipsilateral MD and RD were also both higher than in HC (MD: 0.0012 mm²/s vs. 0.0011 mm²/s, p = 0.0053; RD: 0.00090 mm²/s vs. 0.00080 mm²/s, p = 0.0027). In non-responders, there was no difference in any ipsilateral CNV diffusivity metrics compared to contralateral CNV or HC.

Figure 2.

Preoperative trigeminal nerve (CNV) diffusion metrics in responders (R, green bars, n = 18), non-responders (NR, red bars, n = 17) and healthy controls (n = 38). (a) There was no difference in mean preoperative FA between ipsilateral and contralateral CNV in non-responders to surgical treatment for TN or compared to healthy controls. However, FA was reduced in the ipsilateral compared to the contralateral CNV of responders but not non-responders. This corresponded to an increased RD in the ipsilateral CNV of responders compared to contralateral CNV (d), alongside significant increases in MD and RD in the ipsilateral CNV of responders compared to healthy controls (b, d). No differences were seen for AD (c). Individual data points are indicated with white circles with black lines connecting ipsilateral and contralateral nerve metrics within the same patient. *p < 0.05; **p < 0.01. AD = axial diffusivity; CNV = crainial nerve V; FA = fractional anisotropy; MD = mean diffusivity; RD = radial diffusivity; TN = trigeminal neuralgia.

Postoperative changes in CNV microstructure

Responders versus non-responders

Longitudinal postoperative ipsilateral CNV FA change in responders and non-responders to TN surgical treatment is displayed in Figure 3. In responders (Figure 3a), there was a significant reduction in ipsilateral CNV FA from baseline by one month (0.45 vs. 0.38, p = 0.013), maintained at six months (0.38, p = 0.021) and one year postoperatively (0.37, p = 0.006). While there was no statistically significant FA drop at the one-week timepoint, a strong trend toward reduction was observed (0.40, p = 0.09). FA changes corresponded to reductions in AD which were significant at the one-week and one-month timepoints, although AD returned back to baseline at six months and one year (see supplementary material, Figure S1). In non-responders (Figure 3b), compared to preoperative FA (pre = 0.46), there was no change in ipsilateral CNV FA at any postoperative timepoints including one week (0.45, p = 0.56), one month (0.45, p = 0.71), six months (0.43, p = 0.31) and one year (0.45, p = 0.73). Similarly, there were no significant changes in MD, AD or RD at any postoperative timepoints in non-responders (see supplementary material, Figure S2).

Figure 3.

Postoperative FA change in ipsilateral (i.e. treated) trigeminal nerve (CNV) for responders (n = 18) (a) and non-responders (n = 17) (b) to surgical treatment for TN. After surgery, mean FA declined significantly from baseline at a group-level in responders (a, green bars) by one month postoperatively, and then remained persistently reduced until one year. Conversely, there was no change in ipsilateral CNV FA in non-responders (b, red bars) at any postoperative time-point. Error bars show the standard error of the mean. *p < 0.05, **p < 0.01, ***p < 0.001. CNV = crainial nerve V; FA = fractional anisotropy; TN = trigeminal neuralgia.

As expected, there was no postoperative change in contralateral (i.e. untreated) CNV FA in either responders or non-responders (see supplementary material, Figure S3).

By surgical treatment type

Postoperative ipsilateral CNV FA change in responders and non-responders undergoing MVD and BC is displayed in Figure 4. Responders to MVD (n = 13) and BC (n = 5) both showed persistent reductions in ipsilateral CNV FA. By contrast, there was no change in ipsilateral CNV FA in surgical non-responders to MVD (n = 9) or BC (n = 8). Illustrative tractographic examples of divergent postoperative trajectories of CNV FA change are shown for individual patients (one responder and one non-responder for each of MVD and BC) in Figure 5.

Figure 4.

Postoperative ipsilateral (treated) CNV FA change in responders and non-responders to (a) microvascular decompression (MVD) and (b) balloon compression (BC) surgery. A similar pattern of postoperative ipsilateral CNV FA change was observed between MVD (a) and BC (b) treatment groups. FA declined after surgery and remained reduced in responders (green lines) to both MVD and BC. However, non-responders (red lines) did not show the same sustained decline in FA. The mean value for each surgical group per time point is displayed. Error bars represent the standard error of the mean. *p < 0.05, **p < 0.01. Group size (n) is indecated for each timepoint. No statistical analyses were carried out for BC group due to small n. CNV = crainial nerve V; FA = fractional anisotropy; MVD = microvascular decompression.

Figure 5.

High-resolution FLAIR-DTI tractography of the ipsilateral trigeminal nerve (CNV) in two individual responders (R) and two non-responders (NR) to microvascular decompression (MVD) and balloon compression (BC) across five time points. Deterministic tractography was performed for each patient at the preoperative, <7 day, 30 day, 180 day and 365 day postoperative time points. MVD R (a) and BC R (b) showed a progressive decline in fractional anisotropy (FA) following surgery, consistent with group-level trends illustrated in Figure 4. By contrast, NRs in both the MVD (c) and BC (d) groups showed sustained or increasing FA values across time. Mean nerve segment FA values at each time point are displayed at the bottom of each image. CNV = crainial nerve V; FLAIR-DTI = fluid-attenuated inversion recovery-diffusion tensor imaging.

Pain-free versus painful states

All postoperative ipsilateral CNV FA values across all timepoints for all patients that were associated with no reported pain (VAS = 0) were combined and compared to timepoints with the presence of any pain (VAS > 0) (Figure 6). Ipsilateral CNV FA was significantly lower at pain-free timepoints compared to painful timepoints (0.39 vs. 0.45, p < 0.0001) and HC (0.39 vs. 0.47, p < 0.0001). There was, however, no difference in ipsilateral CNV FA at painful timepoints compared to HC (0.45 vs. 0.47, p = 0.23).

Figure 6.

Comparison of trigeminal nerve (CNV) FA at painful (n = 45) and non-painful (n = 72) postoperative timepoints. Mean CNV FA is lower on average at pain-free postoperative timepoints compared to painful postoperative timepoints and healthy controls (HC). Error bars show standard error of the mean. ****p < 0.0001. CNV = crainial nerve V; FA = fractional anisotropy.

Divergent microstructural changes following repeat TN surgery

Ipsilateral CNV FA values are shown in Figure 7 for two non-responders who each underwent two separate procedures, but had different postoperative pain profiles. These patients were not included in the original dataset contributing to Figures 1–6. In a patient who experienced temporary pain relief after surgery but then had subsequent pain recurrence (Figure 7A) requiring another surgery, ipsilateral CNV FA was reduced at pain free timepoints but returned to preoperative baseline at painful timepoints (i.e. at pain recurrence). This trend was observed after both consecutive surgeries. However, in a patient who never achieved pain relief with either a first or second surgery (Figure 7B), ipsilateral CNV FA did not show the same extent of reduction at any postoperative timepoints compared to preoperative baseline for either surgery.

Figure 7.

Postoperative trigeminal nerve (CNV) FA change over multiple surgical treatments in two patients. (a) Patient with pain recurrence after initial relief. (b) Patient who never experienced pain relief with surgery (never responder). In the patient with recurrent pain (a), FA was reduced postoperatively at pain-free timepoints (green bars) but returned to preoperative level when pain (red bars) returned. This pattern was consistent for both initial (Sx 1) and repeat (Sx 2) surgical treatment. Postoperative FA, however, was not reduced in the never responder who had no pain relief, across both initial and repeat surgical treatments. A dashed vertical line distinguishes the period for surgical treatment #1 (Sx 1) and surgical treatment # 2 (Sx 2) in both patients. CNV = crainial nerve V; FA = fractional anisotropy.

Discussion

In this single-center, prospective, longitudinal study, we used a novel, high-resolution, CNV-specific FLAIR-DTI protocol²⁹ to evaluate pre- and postoperative CNV microstructure (inferred primarily from FA) in TN patients undergoing first-time surgical treatment. Using this methodologically superior DTI approach, we confirmed that TN patients are characterized by significant reduction in ipsilateral compared to contralateral CNV FA, consistent with loss of nerve microstructural integrity on the affected side. However, more significant alteration of ipsilateral preoperative CNV microstructural integrity was seen in responders who remained pain free after surgery. After surgery of any type, responders but not non-responders consistently showed early and persistent postoperative reduction in ipsilateral CNV FA. Furthermore, FA was markedly reduced at pain-free postoperative timepoints compared to painful postoperative timepoints at the group-level, and at the individual level when we compared patients with different pain response profiles over consecutive repeat surgeries. Collectively, our findings suggest that: (1) CNV microstructure is not only affected by TN, but may also predict response to TN surgery and (2) surgery may need to induce specific and durable microstructural changes in CNV to achieve long-term pain relief.

In our cohort, 51% of surgical TN patients remained pain free (VAS = 0 and BNI I or IIIa) throughout the two-year postoperative follow-up period, and were classified as responders. Most prior studies have considered BNI II (i.e. some pain, no medication) patients as responders^36–38; however, we excluded these patients aiming to dichotomize postoperative response into painful versus pain free states. Based on our definition of response, in line with previous reports, non-responders were more likely to be female and younger at the time of surgery.^39,40 As for NVC, it was identified in only 16 of 35 TN patients; contrary to many older studies, we considered true NVC to be present only when it was associated with nerve deformation or distortion (NVC score 2 or 3) and not simple arterial contact or venous compression (NVC score 1). This binarization aligns with the distinction between classical and idiopathic TN per the latest definition of the ICHD-3.¹ While greater NVC severity has been linked to better postoperative outcomes after MVD for classical TN,^40–43 not surprisingly we found no significant difference in the presence of NVC between surgical responders and non-responders, given our mixed cohort of classical and idiopathic TN patients, treated with both MVD and BC approaches. Exploring postoperative microstructural CNV change in idiopathic versus classical TN is a logical next step for future work.

Our first key finding was of reduced ipsilateral compared to contralateral CNV FA across TN patients, consistent with several prior studies examining diffusion metrics in the cisternal and pontine segments of CNV.⁹ Reduced FA may reflect histopathological findings from autopsy and biopsy studies in TN showing demyelination or aberrant myelination,^4–6 and to a lesser extent axonopathy or axon loss.⁴⁴ Our finding of correspondingly increased RD certainly points to myelin disruption. Notably, we did not observe differences in preoperative ipsilateral CNV FA compared to HC subjects, in contrast to previously published literature.^{12,13,15,17,18,22–24,26,45–47} A key reason may be that previous studies used lower-resolution DTI protocols (e.g. 2 × 2 × 2 mm³ or 1.9 × 1.9 × 3 mm³) susceptible to significant partial volume effects from surrounding cerebrospinal fluid, which can artificially reduce FA²⁹ when nerve volume is reduced, as it is in TN.^9,48 Indeed, the highest-resolution DTI protocol previously used to study TN similarly did not observe reduced CNV FA in TN compared to HC.²⁸ It is also possible that microstructural changes reflected by FA are more marked in classical TN with significant NVC; our mixed population of idiopathic and classical TN patients may therefore have diluted the degree of FA reduction we observed. Finally, we report average FA across a 6.3 mm long cisternal tractographic segment of CNV, which may have been less sensitive to focal FA differences occurring at the exact site of NVC along the proximal root.

Our second key finding is of significant asymmetry between ipsilateral and contralateral CNV FA, prior to surgery, in responders compared to non-responders, driven by relatively greater reduction in ipsilateral CNV FA in responders. This is a subtle finding: as illustrated in Figure 2, non-responders and responders showed a similar pattern of reduced ipsilateral CNV FA, and mean ipsilateral CNV FA itself was not significantly different between responders and non-responders. Several prior studies examining preoperative CNV microstructure have largely shown no predictive ability of CNV FA in relation to surgical response.^11,32,49 One study showed that lower ipsilateral CNV FA compared to HC may portend non-response, although lower FA was reported only at the proximal nerve root (not along the entire nerve) using a low-resolution DTI protocol (0.94 × 0.94 × 3 mm³ voxels) susceptible to partial volume effects.²¹ More recently, Su et al.⁵⁰ using a 2 × 2 × 2 mm³ DTI protocol showed that low FA segments within the ganglion of CNV correlated with pain relief after percutaneous radiofrequency rhizotomy. Taken together, these inconsistent results suggest that the relationship between preoperative CNV microstructure and response to surgery requires further investigation. Moving forward, it will be important to disentangle the potentially confounding effects of NVC, sex, duration of TN prior to surgery, and DTI methodology on CNV FA measurements, which prior studies, as well as our own, have so far been underpowered to do.

Our study provides the most comprehensive picture to date of longitudinal postoperative CNV DTI assessments in TN, beginning as early as 5–12 days (i.e. one week) through to 11–14 months (i.e. one year) after surgery. Successful surgical treatment was accompanied by significant early FA reduction by one month postoperatively, persisting thereafter throughout postoperative follow-up. Importantly, there was no corresponding change in FA in the contralateral CNV over the same period, increasing confidence that this finding is not due to methodological artifacts (e.g. scanner drift). An earlier frequently cited DTI study suggested that CNV FA may recover by approximately six months after successful surgery, speculating that this might represent remyelination or microstructural repair in the treated nerve.⁴⁵ To date, this finding has not been widely replicated. Conversely, several more recent studies report, as we do, decreases in FA in the treated nerve at three months ³¹ and six months^30,51 after stereotactic radiosurgery, or six months after rhizotomy²⁵ in responders. Unexpectedly, we found that the pattern of immediate and persistent postoperative FA reduction was observed in responders across both MVD and BC groups, even though the putative mechanisms of action are different between both surgical approaches and some degree of facial sensory loss is more common with BC.³³ We interpret our findings as consistent with the notion that TN surgery (of any kind) may need to induce a certain degree of microstructural disruption in CNV without which durable pain relief may not be possible. Although we are unaware of any delayed histopathological analyses of the surgically-treated CNV (as opposed to at the time of surgery), the accompanying reduction in AD which we observed suggests that some degree of axonal loss is likely occurring to account for the FA decline in both groups. We speculate that the induction of microstructural changes and some axonal loss by surgery contribute to pain relief through various mechanisms, including the reduction of ephaptic transmission between myelinated A-β and unmyelinated A-δ and C fibers in CNV, reduced hyperexcitability of CNV sensory axons, blunted spontaneous axonal pacemaker activity and downstream resetting of functional alterations at the level of the brain that characterize the TN chronic pain state.^9,44,52

Because many non-responders were actually pain-free at various postoperative timepoints, we compared CNV diffusion metrics grouped into “pain” and “no-pain” states. Interestingly, we observed markedly reduced FA at “no-pain” timepoints, suggesting that microstructural alterations associated with FA reduction are, on average, linked to transient postoperative pain-free states. Furthermore, we analyzed two non-responders with different postoperative pain profiles undergoing repeated surgical treatments for pain recurrence. In the non-responder who experienced temporary pain relief with surgery, we observed reduced FA at pain-free timepoints over consecutive repeat surgeries, followed by an abrupt increase in FA when pain returned. By contrast, no FA reduction was ever seen in the other non-responder who never experienced postoperative pain relief. This provides additional evidence that a surgically-induced microstructural alteration may be required for patients to experience (and sustain) pain relief.

This study is not without limitations. First, while we studied a similar number of participants as in other DTI studies in TN, a larger sample size would have provided additional statistical power to distinguish smaller magnitude inter-group differences. Second, classifying treatment response remains a source of debate in TN,⁵³ and our particular method of binarizing of surgical outcomes into response (i.e. BNI I and IIIa) versus non-response (i.e. BNI II, IIIb, IV and V) is certainly not the only possible approach. We acknowledge that the most stringent method would be to consider as responders only those patients with a postoperative status of BNI I (i.e. pain-free, off medication) because this would eliminate the potentially confounding effects of medication on postoperative pain relief or, theoretically, CNV microstructure. We did choose to include BNI IIIa patients as responders since a small number of patients (n = 2) had immediate and sustained pain relief after BC surgery but were reluctant to discontinue medications due to anxiety about the potential for pain recurrence (known to be higher after percutaneous rhizotomy). Third, because of logistical challenges including postoperative illness, scheduling difficulties through COVID-19 and patient travel considerations, not all 35 TN patients completed every postoperative evaluation. Notably, there was a decrease in follow-up data at more delayed timepoints driven by responders who were more likely to drop out, citing lack of pain and wanting to move on with life.

Conclusions

Using a high-resolution, CNV-specific DTI protocol, we performed a detailed longitudinal evaluation of CNV microstructure in TN patients undergoing surgery. We observed more significant preoperative microstructural alterations (reduced FA) in the affected (i.e. treated) CNV in responders compared to non-responders. Additionally, successful surgery, by any modality, resulted in early and sustained microstructural changes (further reduced FA). Finally, pain-free states after surgery were strongly correlated with altered CNV microstructure (lower FA). It appears that TN surgery needs to induce specific structural changes in the operated CNV in order to be successful, and further that CNV microstructure more generally is a key contributing mechanism underlying durable pain relief in TN. Larger studies using high-resolution DTI are warranted to further understand the temporal course of microstructural changes in CNV after surgery.

Article highlights

High-resolution FLAIR-DTI provides a novel imaging approach to assess trigeminal nerve (CNV) microstructure and its role in treatment response in TN.

Preoperative CNV microstructural differences between responders and non-responders are predictive of surgical response in trigeminal neuralgia.

Successful TN surgery requires early and sustained reductions in CNV microstructure, strongly correlating with pain-free states.

Supplemental Material

sj-docx-1-cep-10.1177_03331024251369827 - Supplemental material for Nerve matters: Longitudinal microstructural change in the trigeminal nerve is associated with durable pain relief after surgery for trigeminal neuralgia

Supplemental material, sj-docx-1-cep-10.1177_03331024251369827 for Nerve matters: Longitudinal microstructural change in the trigeminal nerve is associated with durable pain relief after surgery for trigeminal neuralgia by Hayden J. Danyluk, Abhinav Dhillon, Akshit Ayri, Christian Beaulieu and Tejas Sankar in Cephalalgia

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Ethical statement

This prospective study was approved by the Health Research Ethics Board of the University of Alberta (Protocol #00067116). Written informed consent was obtained from all participants prior to their inclusion in the study.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the Canadian Institutes of Health Research Project Grant #487915, awarded to Principal Investigator Tejas Sankar. Additional support was received from the Backman Family Fund administered through the University of Alberta Hospital Foundation. Hayden Danyluk was supported by the Canadian Institutes of Health Research Canada Graduate Scholarship. Abhinav Dhillon was supported by the Alberta Innovates Summer Studentship Award. Akshit Ayri was supported by the University of Alberta Undergraduate Research Stipend Award. Christian Beaulieu acknowledges the Canada Research Chairs program.

ORCID iDs

Hayden J. Danyluk

Abhinav Dhillon

Tejas Sankar

Supplemental material

Supplemental material for this article is available online.

References

Headache classification committee of the international headache society (IHS) the international classification of headache disorders, 3rd edition. Cephalalgia 2018; 38: 1–211.

Peker

Kurtkaya

Üzün

, et al. Microanatomy of the central myelin-peripheral myelin transition zone of the trigeminal nerve. Neurosurgery 2006; 59: 354–358.

Maarbjerg

Wolfram

Gozalov

, et al. Significance of neurovascular contact in classical trigeminal neuralgia. Brain 2015; 138: 311–319.

Love

Hilton

Coakham

. Central demyelination of the Vth nerve root in trigeminal neuralgia associated with vascular compression. Brain Pathol 1998; 8: 1–11.

Hilton

Love

Gradidge

, et al. Pathological findings associated with trigeminal neuralgia caused by vascular compression. Neurosurgery 1994; 35: 299–303.

Moller

. Vascular compression of cranial nerves: iI: pathophysiology. Neurol Res 1999; 21: 439–443.

Bethamcharla

Reddy

Teich

, et al. Histopathology of the trigeminal ganglion and nerve: a historical review. J Neurosci Res 2023; 101: 1203–1204.

Maarbjerg

Benoliel

. The changing face of trigeminal neuralgia-A narrative review. Headache 2021; 61: 817–837.

Watanabe

Shrivastava

Balchandani

. Advanced neuroimaging of the trigeminal nerve and the whole brain in trigeminal neuralgia: a systematic review. Pain 2025; 166: 282–310.

10.

Beaulieu

. The basis of anisotropic water diffusion in the nervous system - {A} technical review. NMR Biomed 2002; 15: 435–455.

11.

Chen

Yang

Yeh

, et al. Using diffusion tensor imaging to evaluate microstructural changes and outcomes after radiofrequency rhizotomy of trigeminal nerves in patients with trigeminal neuralgia. PLoS ONE 2016; 11: e0167584.

12.

Liu

Butzkueven

, et al. Microstructural abnormalities in the trigeminal nerves of patients with trigeminal neuralgia revealed by multiple diffusion metrics. Eur J Radiol 2013; 82: 783–786.

13.

Desouza

Hodaie

Davis

. Abnormal trigeminal nerve microstructure and brain white matter in idiopathic trigeminal neuralgia. Pain 2014; 155: 37–44.

14.

Lutz

Thon

Stahl

, et al. Microstructural alterations in trigeminal neuralgia determined by diffusion tensor imaging are independent of symptom duration, severity, and type of neurovascular conflict. J Neurosurg 2015; 124: 1–8.

15.

Pang

Sun

Fan

. Correlations between the trigeminal nerve microstructural changes and the trigeminal-pontine angle features. Acta Neurochir (Wien) 2019; 161: 2505–2511.

16.

Kumaran

Gurram

Viswamitra

, et al. Utility of DTI (diffusion tensor imaging) metrics to study microstructural changes of trigeminal nerve in patients with trigeminal neuralgia (TN). Neurol INdia 2022; 70: 270–274.

17.

Wang

, et al. Diffusion tensor and kurtosis imaging reveal microstructural changes in the trigeminal nerves of patients with trigeminal neuralgia. Eur Radiol 2023; 33: 8046–8054.

18.

Luo

Chen

Zhang

, et al. Trigeminal nerve white matter fiber abnormalities in primary trigeminal neuralgia: a diffusion Spectrum imaging study. Front Neurol 2022; 12: doi.org/10.3389/fneur.2021.798969.

19.

Lee

Moon

Tak

, et al. Atrophic changes and diffusion abnormalities of affected trigeminal nerves in trigeminal neuralgia using 7-T MRI. Stereotact Funct Neurosurg 2019; 97: 169–175.

20.

Moon

You

Baek

, et al. 7.0 Tesla MRI tractography in patients with trigeminal neuralgia. Magn Reson Imaging 2018; 54: 265–270.

21.

Hung

PSP

Chen

Davis

, et al. Predicting pain relief: use of pre-surgical trigeminal nerve diffusion metrics in trigeminal neuralgia. NeuroImage: Clinical 2017; 15: 710–718.

22.

Wang

, et al. A preliminary study of diffusion tensor imaging in root entry zone of primary trigeminal neuralgia. Front Neuroanat 2023; 17: 1112662.

23.

Zhang

Mao

Cui

, et al. Diffusion tensor imaging of axonal and myelin changes in classical trigeminal neuralgia. World Neurosurg 2018; 112: e597–e607.

24.

Leal

PRL

Roch

Hermier

, et al. Diffusion tensor imaging abnormalities of the trigeminal nerve root in patients with classical trigeminal neuralgia: a pre- and postoperative comparative study 4 years after microvascular decompression. Acta Neurochir 2019; 161: 1415–1425.

25.

Chen

, et al. Longitudinal alterations of the cisternal segment of trigeminal nerve and brain pain-matrix regions in patients with trigeminal neuralgia before and after treatment. BMC Neurosci 2021; 22: 77.

26.

Qiu

Jiang

, et al. Diffusion tensor imaging reveals microstructural alteration of the trigeminal nerve root in classical trigeminal neuralgia without neurovascular compression and correlation with outcome after internal neurolysis. Magn Reson Imaging 2020; 71: 37–44.

27.

Lee

Chong

Chen

, et al. The timing of stereotactic radiosurgery for medically refractory trigeminal neuralgia: the evidence from diffusion tractography images. Acta Neurochir 2018; 160: 977–986.

28.

Fujiwara

Sasaki

Wada

, et al. High-resolution diffusion tensor imaging for the detection of diffusion abnormalities in the trigeminal nerves of patients with trigeminal neuralgia caused by neurovascular compression. J Neuroimaging 2011; 21: e102–e108.

29.

Danyluk

Sankar

Beaulieu

. High spatial resolution nerve-specific {DTI} protocol outperforms whole-brain {DTI} protocol for imaging the trigeminal nerve in healthy individuals. NMR Biomed 2021; 34: e4427.

30.

Tohyama

Hung

Zhong

, et al. Early postsurgical diffusivity metrics for prognostication of long-term pain relief after Gamma Knife radiosurgery for trigeminal neuralgia. J Neurosurg 2019; 131: 539–548.

31.

Pikis

Bunevicius

Donahue

, et al. Diffusivity metrics three months after upfront gamma knife radiosurgery for trigeminal neuralgia may be correlated with pain relief. World Neurosurg 2021; 153: e220–e2e5.

32.

Chen

Yang

Weng

, et al. Diffusion tensor imaging for assessment of microstructural changes associate with treatment outcome at one-year after radiofrequency Rhizotomy in trigeminal neuralgia. BMC Neurol 2019; 19: 52.

33.

Sindou

Leston

Decullier

, et al. Microvascular decompression for primary trigeminal neuralgia: long-term effectiveness and prognostic factors in a series of 362 consecutive patients with clear-cut neurovascular conflicts who underwent pure decompression. J Neurosurg 2007; 107: 1144–1153.

34.

Han

Shetter

Smith

, et al. Gamma Knife radiosurgery for trigeminal neuralgia: experience at the Barrow Neurological Institute. Int J Radiat Oncol Biol Phys 2000; 47: 1013–1019.

35.

Leemans

Jeurissen

Sijbers

, et al. ExploreDTI: a graphical toolbox for processing, analyzing, and visualizing diffusion {MR} data. 17th Annual Meeting of Intl Soc Mag Reson Med 2009: 3537.

36.

Burchiel

KIMJ

Clarke

Haglund

, et al. Long-term efficacy of microvascular decompression in trigeminal neuralgia. J Neurosurg 1988; 69: 35–38.

37.

Wang

Raygor

Cage

, et al. Prospective comparison of long-term pain relief rates after first-time microvascular decompression and stereotactic radiosurgery for trigeminal neuralgia. J Neurosurg 2018; 128: 68–77.

38.

Raygor

Wang

Ward

, et al. Long-term pain outcomes for recurrent idiopathic trigeminal neuralgia after stereotactic radiosurgery: a prospective comparison of first-time microvascular decompression and repeat stereotactic radiosurgery {Kunal}. J Neurosurg 2018; 131: 1207–1215.

39.

Heinskou

Rochat

Maarbjerg

, et al. Prognostic factors for outcome of microvascular decompression in trigeminal neuralgia: a prospective systematic study using independent assessors. Cephalalgia 2019; 39: 197–208.

40.

Ishaque

Xie

Danyluk

, et al. Comparison of prognostic scoring systems to predict durable pain relief after microvascular decompression for trigeminal neuralgia. World Neurosurg 2022; 157: e432–ee40.

41.

Hardaway

Gustafsson

Holste

, et al. A novel scoring system as a preoperative predictor for pain-free survival after microsurgery for trigeminal neuralgia. J Neurosurg 2020; 132: 217–224.

42.

Panczykowski

Jani

Hughes

, et al. Development and evaluation of a preoperative trigeminal neuralgia scoring system to predict long-term outcome following microvascular decompression. Neurosurgery 2020; 87: 71–79.

43.

Tugend

Joncas

Chang

, et al. Update and Evaluation of a Preoperative Scoring System to Predict Long-Term Outcomes After Microvascular Decompression in Trigeminal Neuralgia. Oper Neurosurg (Hagerstown) 2025. https://pubmed.ncbi.nlm.nih.gov/40391987/.

44.

Devor

Govrin-Lippmann

Rappaport

. Mechanism of trigeminal neuralgia: an ultrastructural analysis of trigeminal root specimens obtained during microvascular decompression surgery. J Neurosurg 2002; 96: 532–543.

45.

DeSouza

Davis

Hodaie

. Reversal of insular and microstructural nerve abnormalities following effective surgical treatment for trigeminal neuralgia. Pain 2015; 156: 1112–1123.

46.

Chai

You

Zhang

, et al. Diffusion tensor imaging of microstructural alterations in the trigeminal nerve due to neurovascular contact/compression. Acta Neurochir 2019; 161: 1407–1413.

47.

Tugend

Joncas

Traylor

, et al. An analysis of diffusion tensor imaging in classical trigeminal neuralgia. Cephalagia 2025; 45: 3331024251322505.

48.

Danyluk

Lee

Wong

, et al. Hippocampal and trigeminal nerve volume predict outcome of surgical treatment for trigeminal neuralgia. Cephalalgia 2020; 40: 586–596.

49.

Lee

Chen

Yang

, et al. Diffusivity parameters of diffusion tensor imaging and apparent diffusion coefficient as imaging markers for predicting the treatment response of patients with trigeminal neuralgia. J Neurosurg 2020; 132: 1993–1999.

50.

Wang

, et al. A novel indicator to predict the outcome of percutaneous stereotactic radiofrequency rhizotomy for trigeminal neuralgia patients: diffusivity metrics of MR-DTI. Sci Rep 2024; 14: 9235.

51.

Hung

PSP

Tohyama

Zhang

, et al. Temporal disconnection between pain relief and trigeminal nerve microstructural changes after Gamma Knife radiosurgery for trigeminal neuralgia. J Neurosurg 2020; 133: 727–735.

52.

Danyluk

Lang

Monchi

, et al. Pre-operative Limbic System Functional Connectivity Distinguishes Responders From Non-responders to Surgical Treatment for Trigeminal Neuralgia. Front Neurol 2021; 12: 716500.

53.

Granzer-Corno

Rana

Dick

, et al. Different scale, different pain? Discordant pain measurements after surgery for trigeminal neuralgia. World Neurosurg 2024; 194: 123481.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.09 MB