Twenty-three TN patients and 20 HC were included in this study between 2017 and 2020 (Table 1). All 23 TN patients were included in the volumetric analysis, while only 19 TN patients were included in the metabolite analysis because of inadequate spectral quality (n = 2) or failure to acquire MRS scans (n = 2).
Clinical characteristics and demographics:
All TN: Clinical and demographic features of all 23 TN patients and 20 HCs are presented in Table 1. TN and HC groups were well matched in age (56.3 10.4 years and 54.9 9.4 years respectively, p = 0.65) and sex distribution (14F/9M and 11F/9M, p = 0.70). Average TN duration was 5.3 3.9 years, with right-sided TN more common than left-sided TN (15R/8L). Pre-operative VAS was 77.6 27.3. This study included virgin surgical procedures for 19/23 TN patients, with MVD most common (16 MVD, 7 BC). All TN patients were on antiepileptic medication at surgery, including carbamazepine/oxcarbazepine (n = 21) and/or gabapentin/pregabalin (n = 12). Three TN patients were also on antidepressant/anxiolytic medication, six were on baclofen, one was taking opioids, and two others were taking cannabis oil.
[insert Table 1]
By response to surgery: In total, there were 17 responders to surgery and 6 non-responders. Non-responders were exclusively female (6F/0M), while responders had a balanced sex distribution (8F/9M). Non-responders were younger than responders (47.4 10.3 years and 58.6 9.7 years respectively, p = 0.016), and showed a trend toward longer duration of TN compared to responders (8.8 4.7 years versus 4.6 3.3 years respectively, p = 0.061). Frequency of surgery type did not differ between outcome groups. The proportion of patients taking baclofen was higher in non-responders than responders (p = 0.021), with no other differences in medication use. Individual TN patient clinical profiles are presented in Table 2.
[insert Table 2]
Native orientation analysis: In HCs, the volume of the left thalamus was larger than the right (8209, 7895-9195 IQR and 7985, 7607-8974 IQR respectively; p < 0.001). There was no difference between left and right thalamus volume across all TN patients together (8077, 7590-8520 IQR and 8138, 7665-8506 IQR respectively; p = 0.88) (data not shown). The volume of the thalamus contralateral to the side-of-pain was larger in TN patients: in left TN patients (LTN) the right thalamus was larger than the left (8308, 8017-8633 IQR and 8111, 7641-8168 IQR respectively; p = 0.008); in right TN patients (RTN) the left thalamus was larger than the right (7895, 7485-8520 IQR and 8258, 7665-8858 IQR respectively; p < 0.001) (Figure 1A). TN was associated with altered interhemispheric asymmetry of thalamic volume: left versus right thalamus interhemispheric %volume difference differed between HC and LTN patients (-3.1, -4.1 to -2.3 IQR; 3.9, 2.8-5.8 IQR; p < 0.001), while there was no difference between HC and RTN patients (3.2, 2.4-4.3 IQR; 2.9, 2.3-4.0 IQR; p = 0.54) or between LTN and RTN (3.9, 2.8-5.8 IQR; 2.9, 2.3-4.0 IQR; p = 0.13) (Figure 1B).
[insert Figure 1]
Ipsilateral orientation analysis: After flipping the brains of left TN patients, we found no differences between average HC thalamus volume (8087, 7763-9084 IQR) and ipsilateral (ips) or contralateral (cont) thalamus volume of responders (ips: 8083, 7639-8348 IQR, p = 0.54; cont: 8365, 7944-8766 IQR, p = 0.48) or non-responders (ips: 7571, 6988-8461 IQR, p = 0.093; cont: 7810, 7239-8695 IQR, p = 0.32). Thalamus volume contralateral to the side-of-pain was larger than ipsilateral to the side-of-pain in both responders (p < 0.001) and non-responders (p < 0.031) (Figure 2A), though interhemispheric %volume difference did not differ between outcome groups (R: 2.9, 2.4-4.4 IQR; NR: 3.3, 2.8-4.6 IQR; p = 0.71) (Figure 2B). We did not observe any significant changes in thalamus volume between pre-operative and 1-week post-operative time points in TN patients regardless of side-of-pain or surgical outcome (data not shown).
[insert Figure 2]
Intracranial volume: Intracranial volume as assessed using the v-scaling factor in FSL’s SIENAX tool did not show any significant differences between any groups across all volumetric comparisons (data not shown).
Shape analysis: There were significant vertex-wise shape differences seen between the thalami of responders and non-responders (Figure 3). Non-responders showed significant contralateral thalamus volume reduction compared to responders in an axially-oriented band spanning the outer thalamic circumference made up of two clusters (peak p = 0.019).
[insert Figure 3]
Native orientation analysis:
There was no pre-operative difference in Cho/Cr between left and right VPM thalamus in HC, RTN, or LTN patients. Overall, across HC and TN patients, there was no difference in mean Cho/Cr (Figure 4A), nor any difference in %inter-hemispheric Cho/Cr difference between LHC and LTN patients (Figure 4B).
In HCs, there was a trend towards left greater than right VPM thalalmus NAA/Cr (1.97, 1.88-2.28 IQR and 1.86, 1.63-2.04 IQR respectively; p = 0.060). There was no difference between left and right NAA/Cr in LTN or RTN patients. Left thalamus NAA/Cr was higher in HC compared to RTN patients (p = 0.050). Additionally, NAA/Cr of the left (ipsilateral) VPM thalamus in LTN patients was greater than the left (contralateral) and right (ipsilateral) VPM thalamus in RTN patients (p = 0.029 and p = 0.010 respectively). There were no other between-group differences (Figure 4C, 4D).
[insert Figure 4]
Ipsilateral orientation analysis: There were no differences between pre-operative ipsilateral and contralateral VPM thalamus Cho/Cr in responders or non-responders. However, pre-operative ipsilateral Cho/Cr in non-responders was reduced compared to HCs (1.02, 0.98-1.13 IQR; p = 0.038). There were no other between-group or within-group between-side thalamus Cho/Cr differences (Figure 5A, 5B).
There were no differences between ipsilateral and contralateral VPM thalamus NAA/Cr in responders or non-responders, and no other between-group or within-group between-side thalamus NAA/Cr differences (Figure 5C, 5D).
[insert Figure 5]
Post-operative metabolite change at 1-week: Pre-operatively, average VPM thalamus Cho/Cr in HCs did not differ compared to pre-operative contralateral Cho/Cr in either responders or non-responders (Figure 6A). Furthermore, pre-operative contralateral Cho/Cr did not differ between responders and non-responders (p = 0.32). One week after surgery, Cho/Cr in non-responders was significantly lower than in responders (0.83, 0.67-0.97 IQR and 1.04, 0.93-1.17 IQR respectively; p = 0.038) and HCs (p = 0.005), and every single non-responder showed a reduction in Cho/Cr irrespective of surgical procedure type (n = 3 MVD and n = 3 BC). Conversely, responders showed no Cho/Cr change with surgery (p = 0.57). (Figure 6A-C).
Mirroring Cho/Cr, pre-operative average VPM thalamus NAA/Cr in HCs did not differ compared to pre-operative contralateral VPM thalamus NAA/Cr in responders or non-responders (Figure 6D). Furthermore, pre-operative contralateral VPM thalamus NAA/Cr did not differ between responders and non-responders (p = 0.77). One week after surgery, NAA/Cr in non-responders was significantly lower than in responders (1.94, 1.83-2.41 IQR and 1.59, 1.37-1.79 IQR respectively; p = 0.038) and HCs (p = 0.005), and every single non-responder showed a reduction in NAA/Cr irrespective of surgical procedure type (n = 3 MVD and n = 3 BC). Conversely, responders showed no NAA/Cr change with surgery (p = 0.20). (Figure 6D-F).
[insert figure 6]