If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Division of Brain, Imaging, and Behaviour - Systems Neuroscience, Krembil Research Institute, Toronto Western Hospital, University Health Network, Ontario, CanadaDepartment of Surgery and Institute of Medical Science, University of Toronto, Ontario, CanadaCollaborative Program in Neuroscience, University of Toronto, Ontario, CanadaTemerty Faculty of Medicine, University of Toronto, Ontario, Canada
Division of Brain, Imaging, and Behaviour - Systems Neuroscience, Krembil Research Institute, Toronto Western Hospital, University Health Network, Ontario, CanadaDepartment of Surgery and Institute of Medical Science, University of Toronto, Ontario, CanadaCollaborative Program in Neuroscience, University of Toronto, Ontario, CanadaTemerty Faculty of Medicine, University of Toronto, Ontario, Canada
Division of Brain, Imaging, and Behaviour - Systems Neuroscience, Krembil Research Institute, Toronto Western Hospital, University Health Network, Ontario, CanadaTemerty Faculty of Medicine, University of Toronto, Ontario, Canada
Temerty Faculty of Medicine, University of Toronto, Ontario, CanadaRadiation Medicine Program, Princess Margaret Hospital and University of Toronto, Toronto, Ontario, Canada
Collaborative Program in Neuroscience, University of Toronto, Ontario, CanadaCentre for Multimodal Sensorimotor and Pain Research, University of Toronto, Ontario, CanadaUniversity of Toronto Centre for the Study of Pain, Toronto, Ontario, CanadaDivision of Clinical & Computational Neuroscience, Krembil Research Institute, Toronto Western Hospital, University Health Network, Ontario Canada
Address reprint requests to Dr. Mojgan Hodaie, Division of Brain, Imaging, and Behaviour - Systems Neuroscience, Toronto Western Hospital, University Health Network, Room 4W W-443, 399 Bathurst Street, Toronto, Ontario, Canada M5T 2S8.
Division of Brain, Imaging, and Behaviour - Systems Neuroscience, Krembil Research Institute, Toronto Western Hospital, University Health Network, Ontario, CanadaDepartment of Surgery and Institute of Medical Science, University of Toronto, Ontario, CanadaCollaborative Program in Neuroscience, University of Toronto, Ontario, CanadaTemerty Faculty of Medicine, University of Toronto, Ontario, CanadaDivision of Neurosurgery, Krembil Brain Institute, Toronto Western Hospital, University Health Network, Ontario, Canada
The hippocampus is bilaterally smaller in trigeminal neuralgia (TN) patients.
•
In patients who had pain relief, hippocampal abnormalities were normalized.
•
Hippocampal changes are driven by CA2/3, CA4, dentate gyrus and hippocampal proper.
•
Females show a more significant change compared to males.
•
TN provides a unique opportunity to study hippocampal changes after pain relief.
Abstract
Chronic pain patients frequently report memory and concentration difficulties. Objective testing in this population points to poor performance on memory and cognitive tests, and increased comorbid anxiety and depression. Recent evidence has suggested convergence between chronic pain and memory deficits onto the hippocampus. The hippocampus consists of heterogenous subfields involved in memory consolidation, behavior regulation, and stress modulation. Despite significant studies outlining hippocampal changes in human and chronic pain animal models, the effect of pain relief on hippocampal abnormalities remains unknown. Trigeminal neuralgia (TN) is a chronic neuropathic pain disorder which is highly amenable to surgical interventions, providing a unique opportunity to investigate the effect of pain relief. This study investigates the effect of pain relief on hippocampal subfields in TN. Anatomical MR images of 61 TN patients were examined before and 6 months after surgery. Treatment responders (n = 47) reported 95% pain relief, whereas non-responders (n = 14) reported 40% change in pain on average. At baseline, patients had smaller hippocampal volumes, compared to controls. After surgery, responders’ hippocampal volumes normalized, largely driven by CA2/3, CA4, and dentate gyrus, which are involved in memory consolidation and neurogenesis. We propose that hippocampal atrophy in TN is pain-driven and successful treatment normalizes such abnormalities.
Perspective
Chronic pain patients have structural abnormalities in the hippocampus and its subfields. Pain relief normalizes these structural abnormalities and impacts patients in a sex-dependent manner.
Bair M.J., Robinson R.L., Katon W., Kroenke K.: Depression and pain comorbidity: A literature review. Arch. Intern. Med. American Medical Association; page 2433–452003.
The hippocampal formation consists of anatomically and functionally distinct subfields including the subiculum, Cornu Ammonis (CA1 – CA4), and dentate gyrus (DG). These subcortical regions play a significant role in memory formation,
Wingenfeld K., Wolf O.T.: Stress, memory, and the hippocampus. In: Szabo K., Hennerici M.G., editors. Hippocampus Clin Neurosci S. Karger AG; page 109–202014.
Similarly, hippocampal volumetric abnormalities have also been reported in other chronic neuropathic pain conditions including trigeminal neuralgia (TN).
TN has several unique features that distinguish it as an ideal model for the study of chronic pain: TN is largely unilateral; is severe in its nature; has stereotypical presentation among patients; and is not associated with other sensory deficits observed in other chronic pain disorders, such as numbness. Utilizing TN as a model, our group recently investigated the impact of chronic neuropathic pain on hippocampal subfields.
However, no previous study has identified reversible hippocampal alterations reported in chronic pain conditions. Given that prior efforts defined an effective treatment as a 20-40% improvement in pain ratings, it remains unknown whether complete pain resolution could affect hippocampal abnormalities. Uniquely, TN can offer valuable insight because it is one of the few pain conditions that are highly amenable to surgical interventions. Surgeries such as Gamma Knife Radiosurgery (GKRS) can result to complete TN pain resolution,
Long-Term Outcomes in the treatment of classical trigeminal neuralgia by gamma knife radiosurgery: A retrospective study in patients with minimum 2-year follow-up.
Regis J., Tuleasca C., Resseguier N., Carron R., Donnet A., Gaudart J., Levivier M.: Long-term safety and effiacy of Gamma Knife surgery in classical trigeminal neuralgia: A 497-patient historical cohort study. J. Neurosurg. 1079–872016.
thereby permitting the investigation of a possible reversal of hippocampal abnormalities following pain relief.
Here, we aim to investigate the structure of the hippocampus and its subfields in TN patients before and after surgery to examine the effect of pain resolution on these structures. We hypothesize that pain relief will lead to an overall increase in hippocampal volume, and normalize previously observed abnormalities.
we hypothesize that pain relief will affect subfields important in neurogenesis, such as DG and CA4. Towards this goal, we plan to investigate structural magnetic resonance images (MRI) of TN patients before and after pain relief. Given that there are sex differences in hippocampal neurogenesis,
Chronic Early Life Stress Alters Developmental and Adult Neurogenesis and Impairs Cognitive Function in Mice. 25. Hippocampus John Wiley and Sons Inc.,
2015: 309-328
we additionally, hypothesize that there will be sex differences after pain relief in TN patients.
Methods
Participants
Ethics
The University Health Network (UHN) Research Ethics Board approved this retrospective TN study. Patient data used in this study were analyzed retrospectively and no active participation. As such, individual patient consent was not required for this retrospective study. The UHN Research Ethics Board approved the recruitment of healthy controls and the image acquisition procedure. An individual written informed consent form was obtained from healthy individuals. All MRI scans were anonymized prior to any imagine analysis.
TN Patients
A total of 61 patients who was treated at Toronto Western Hospital in Canada were included in this study. Patients in this study met the following criteria: I) diagnosis of classical TN according to ICHD-3 criteria; II) GKRS treatment with no prior surgical interventions for TN; III) structural brain MRI prior to and 6 months after GKRS; IV) clinical follow-up 6 months after surgery. Patients with neurodegenerative disorders, TN secondary to multiple sclerosis, stroke, other chronic pain conditions, cranial tumors, and other neurological diseases were excluded from this study.
Images acquisition All T1-weighted images of TN patients were acquired with a 3 Tesla GE Signa HDx MRI scanner (General Electric, Boston, MA) fitted with an 8-channel head coil (fast-spoiled gradient echo, TE = 5.1 ms, TR = 12.0 ms, TI 300 ms, flip angle = 20°, voxel size = 0.86 mm × 0.86 mm × 1.00 mm, 256 × 256 matrix, field of view = 22 cm, 146 slices). It should be noted that TN patients scheduled for GKRS were generally scanned on the day of surgery. These images were used to guide the stereotactic surgery. However, in some cases, due to limited scanner time availability, some scans only captured subcortical regions of interest. In these cases, all subcortical regions, including the entire hippocampus, were captured, and some cortical regions were lost.
Healthy Controls
To assess whether hippocampal abnormalities normalize after pain relief, 61 neurologically healthy individuals from the Cambridge Centre for Ageing and Neuroscience (Cam-CAN) dataset
The Cambridge Centre for Ageing and Neuroscience (Cam-CAN) study protocol: A cross-sectional, lifespan, multidisciplinary examination of healthy cognitive ageing.
were used as age- and sex-matched controls (age difference < 2 years). Exclusion criteria included: (1) depression, and (2) any report of pain. The ipsi- and contralateral sides for healthy individuals were determined based on their matched TN subjects: therefore, if a healthy subject was matched to a right TN patient, the right hemisphere was marked as ipsilateral. We used Cam-CAN dataset instead of healthy controls collected on site, as sex- and age-matched controls could not be recruited. The use of this control group was validated (see below).
Images acquisition All T1-weighted images of subjects from the Cam-CAN dataset were acquired with a 3 Tesla SIEMENS MAGNETOM TrioTim syngo MR B17 32-channel head coil (fast-spoiled gradient echo, TE = 2.99 ms, TR = 2250 ms, TI: 900 ms, flip angle = 9°, voxel size = 1.0 mm isotropic, 256 × 240 matrix, field of view =25.6 cm, 192 slices).
Healthy Controls Validation
To validate our approach to include Cam-CAN subjects in our analyses, we compared 76 healthy controls with T1 anatomical scans collected at Toronto Western Hospital and 76 age- and sex-matched (age difference < 1 year) neurologically healthy individuals from the Cam-CAN database. Both groups underwent the same processing pipeline as performed in the main analyses of this study (see below), including brain segmentation, parcellation, and hippocampal subfield segmentation using FreeSurfer 6.0. The volumetric results were compared at the group level using Wilcoxon test, to investigate their differences, as well as Bayesian Estimation Supersedes the t-test (BEST) test, to investigate their similarities.
All corrected p-values were above 0.05 in the Wilcoxon tests of the hippocampus and its subfields comparing the two groups. Additionally, the 95% Highest Density Interval of a true difference in means included zero in all BEST tests. As such, we showed that the hippocampus its subregions were statistically similar between the examined groups. Therefore, these two datasets, collected with different scanning protocols on different scanners, not only were not statistically different at the group level when comparing volumetric segmentations, but also statistically similar.
Automated Subcortical Segmentation
FreeSurfer 6.0 (https://surfer.nmr.mgh.harvard.edu/) was employed for subcortical segmentation
A Computational Atlas of the Hippocampal Formation Using Ex vivo, Ultra-High Resolution MRI: Application to Adaptive Segmentation of in vivo MRI. 115. Neuroimage Academic Press Inc.,
2015: 117-137
protocol to automatically extract the volumetric values from 12 hippocampal subfields. Given recent evidence suggesting that the hippocampus has functionally meaningful subregions along its longitudinal axis,
All patients underwent GKRS at Toronto Western Hospital, using an Elekta Perfexion system utilizing 4 mm collimators. One single fraction of 80 Gy was delivered to the 100% isodose to the cisternal segment of the symptomatic trigeminal nerve. To minimize the radiation effects, brainstem radiation was restricted to 15 Gy/mm3.
Pain intensity and clinical outcomes were assessed before surgery, and at a 6-month post-treatment follow-up visit. Pain intensity was measured using two instruments: a Numeric Rating Scale (NRS)
The NRS was an 11-point scale, rated between 0 and 10, with the anchors: 0 = no pain and 10 = the worst imaginable pain. The BNI scale comprised five categories of pain for TN: class I–no trigeminal pain, no medication; class II–occasional pain, no medication; class III–some pain, adequately controlled with medication; class IV–some pain, not adequately controlled with medication; and class V–severe pain, no pain relief. In accordance to the previous literature,
patients who achieved >=75% reduction in pain (or complete pain resolution) on the NRS scale and score of I-III on the BNI scale at follow-up were classified as responders, whereas those who achieved <75% pain improvement on the NRS scale and score of IV-V on the BNI scale were classified as non-responders. There were seven patients that fit within the BNI III category but had somewhat less than 75% improvement of their pain with the NRS scale (please refer to Table 1). Each of these subjects was reviewed separately to determine whether they fit the category of responders or non-responders. Those who experienced a lower frequency of attacks after treatment, and had lowered their medication dose were considered as responders. The two non-responders in this group continued to experience frequent attacks and had in fact continued or increased their medication dose after an initial attempt to taper. All TN subjects’ demographics are summarized in Table 1.
Indicates situation in which patients could not tolerate pharmacological treatments.
Non-Resp
Distribution indicates the affected peripheral branches of the trigeminal nerve with (V1: ophthalmic branch; V2: maxillary branch; V3: mandibular branch). NRS corresponds to the 0-10 Numeric Rating Scale (Anchors: 0 = no pain; 10 = worst imaginable pain). BNI is Barrow Neurological Institute scale (class I: no trigeminal pain, no medication; class II: occasional pain, no medication; class III: some pain, adequately controlled with medication; class IV: some pain, not adequately controlled with medication; class V: severe pain, no pain relief). Groups are based on the pain relief and BNI scale 6 months after the surgical intervention (Resp: responders; Non-Resp: non-responders).
Abbreviations: M, male; F, females; CBZ, Carbamazepine; GBP, Gabapentin; BCL, Baclofen; LYC, Lyrica; PGB, pregabalin; TOL, Toradol; CYM, Cymbalta; HMO, Hydromorphone; TCA, Tricyclic Antidepressant; ACV, Anticonvulsant; No med, no medication.
Indicates situation in which patients could not tolerate pharmacological treatments.
† Indicates patients that were reviewed in detail for their treatment response according to criteria outlined in section 2.3.
A unified approach for morphometric and functional data analysis in young, old, and demented adults using automated atlas-based head size normalization: Reliability and validation against manual measurement of total intracranial volume.
Given that some participants did not have whole-brain MRI scans, we normalized the volume of the structures of interest using subcortical grey volume (SGV). As previously described,
A unified approach for morphometric and functional data analysis in young, old, and demented adults using automated atlas-based head size normalization: Reliability and validation against manual measurement of total intracranial volume.
we adjusted whole hippocampal volume using the residual method with the following formula:
Where VOIadj is the adjusted volume of interest, VOI is the output volume from the FreeSurfer pipeline, b is the slope of the linear regression between VOI and on SGV, and the SGVmean is the sample mean of the SGV. As such, this approach normalizes the hippocampal volume by removing the influence of the subcortical grey volume. All reported volumes are adjusted with this method.
To validate our approach to use SGV instead of total intracranial volume (ICV) to correct for individual hippocampal size, we performed a Pearson's correlation analysis between hippocampal size and SGV and ICV separately among 434 healthy subjects.
Volumetric Percent Change After Surgery
To assess sex effects in pain relief, we additionally calculated the percentage of volumetric change in the whole hippocampus using the following formula, in males and females:
Where Hippochange is the volumetric percent change after the surgery compared to the pre-surgery time point.
Statistical Analysis
All statistical analyses were conducted in R 3.5.1.
The non-parametric Wilcoxon test was used in our analyses for data that were not normally distributed. The statistical analyses include comparison of age between TN subjects and healthy controls, using a Student's t-tests; comparing hippocampal volumes pre- and post-surgery using paired Wilcoxon-test; comparing the pre- and post-surgery volume to healthy controls using Mann-Whitney U Test
; assessment of hippocampal volumetric percent change after treatment to pre-treatment time point using Wilcoxon singed-rank test (pre-treatment time point = no change baseline); comparison of sex differences in hippocampal changes after pain relief, using Wilcoxon signed-rank test.
All the reported p-values are corrected for multiple comparisons using Bonferroni's test with statistical analyses determined significant if p < 0.05.
Results
Subject Demographics
A total of 61 classical TN patients was included in this study (36F, 25M). All patients experienced unilateral facial pain (26L, 35R). All patients underwent GKRS as their first surgical treatment. The average pain duration prior to surgical intervention was 7.0 ± 7.8 years (mean ± SD, NRS scale). The average age at the time of surgery was 64.9 ± 12.0 years (F: 66.8 ± 10.1; M: 62.0 ± 14.1). Age was not statistically different between males and females (P = 0.14). The healthy cohort selected from the Cam-CAN dataset was 64.8 ± 12.1 years (36F: 66.8 ± 10.1; 25M: 62.0 ± 14.2). The age was not statistically different between the healthy and TN cohort (P = 0.99). Patient demographic information is detailed in Table 1.
Six months following the surgery, 47 patients were identified as responders (responder group), defined as at least 75% reduction in pain (including complete pain resolution) on the NRS scale and BNI sore of I-III. The average age for responders was 64.8 ± 12.1 (28F: 67.4 ± 9.7; 19M: 61.0 ± 14.4) and they reported 95% ± 11% improvement in pain on average after surgery. The remaining 14 patients were identified as non-responders, defined as less than 75% pain improvement on the NRS scale or BNI score of IV-V. Non-responders were 62.6 ± 11.9 years old on average at the time of surgery (8F: 62.5 ± 11.1; 6M: 62.7 ± 13.9) and reported a 40% ± 25% change in their pain after the surgical intervention.
SGV Strongly Correlates With Hippocampal Volume
A Pearson's correlation analysis among 434 healthy subjects from Cam-CAN dataset showed that the hippocampus is correlated to total intracranial volume (ICV) by RLeft Hippocampus - ICV= 0.477 (p < 0.001), RRight Hippocampus - ICV= 0.467 (p <0.001) and to SGV by RLeft Hippocampus - SGV= 0.780 (p < 0.001), RRight Hippocampus - ICV= 0.771 (p <0.001). As such, our analyses showed that using SGV provides a more robust hippocampal volume normalization for head size than traditionally used ICV method.
Reversal of Structural Hippocampal Abnormalities with Pain Relief
automated protocol delineated the volumetric values for the hippocampus. Hippocampal volume changes after surgery were determined for responders and non-responders, and were evaluated based on pain laterality (i.e., ipsilateral referring to the reported pain and surgical side; see Fig 1). In the responder group, the whole hippocampus volume increased bilaterally (pipsilateral < 0.001; pcontralateral <0.001). In the non-responder group, the volumetric changes in the ipsi- and contralateral side were not statistically significant compared to pre-treatment (pipsilateral = 0.10; pcontralateral = 0.35). Furthermore, correlation analyses between pain reduction and hippocampal size change were not statistically significant (ripsilateral = -0.063 pipsilateral = 0.62, rcontralateral = 0.030 pcontralateral = 0.81). Similarly, correlation analyses between pain duration and volume changes in the responder cohort were not statistically significant (ripsilateral = 0.007 pipsilateral = 0.96, rcontralateral = -0.127 pcontralateral = 0.45).
Figure 1Hippocampal volumes in pre- and 6 months post-surgery in 47 responders and 14 non-responders TN patients. Wilcoxon-Paired Test was used for statistical analyses in pre- vs. post-surgery volumes (* p < 0.05, ** P < 0.01, *** P < 0.001). Lines connect the TN patients pre- and post-surgery. Black bars indicate ipsilateral results, whereas grey bars indicate contralateral results. Green lines indicate a volumetric increase from pre- to post-surgery and red lines indicated a volumetric decrease. Whole hippocampus: showing increased bilateral whole hippocampal volume in TN responders but not the non-responders 6 months after surgical intervention.
Additionally, TN patients were compared to healthy individuals from the Cam-CAN database which showed hippocampus is bilaterally smaller in TN patients (pipsilateral < 0.001; pcontralateral <0.001). As the number of subjects in the non-responder cohort was limited, we focused on investigating the effect of pain relief by analyzing the responder group. Our results showed that the responder cohort has significantly smaller ipsi- and contralateral whole hippocampus compared to healthy individuals (pipsilateral = 0.006 pcontralateral = 0.001) prior to the surgery. However, six months after GKRS and pain resolution, the whole hippocampus was not statistically different between those with TN and healthy individuals (pipsilateral = 0.36 pcontralateral = 0.15). The ipsi- and contralateral hippocampi were not asymmetric in neither TN (prior or after surgery) nor healthy individuals (all p values = 1). Results are summarized in Figure 2.
Figure 2Automated hippocampal segmentation for 47 TN responders and their age- and sex-matched healthy controls. Results indicate bilateral smaller hippocampus in TN patients prior compared to healthy controls prior to the surgery. The volumetric differences are normalized after surgical intervention and pain relief. No asymmetry was observed in ipsi- and contra-lateral side of the hippocampus. All reported volumes are corrected for subcortical volume size (please refer to the method section). All p-values are corrected for multiple comparison using Bonferroni correction test (*P < 0.05, ** P < 0.01, *** P < 0.001).
Hippocampal Subfields' Volume Normalize Following Pain Relief
In the responder cohort, the CA1, CA2/3, CA4, granule cell layer of the dentate gyrus (DG), molecular layer of hippocampus proper (ML HP), subiculum, and whole hippocampal body showed a significant bilateral volumetric increase compared to pre-surgery.
Presurgically, bilateral CA2/3, CA4, DG, and ML HP were smaller in TN patients compared to healthy controls. However, post-surgically, these subfields were no longer statistically different from healthy controls, indicating volumetric normalization after pain resolution. The same effect was observed on the contralateral CA1 and ipsilateral subiculum: they were statistically smaller in TN patients compared to healthy controls pre-surgery, but not post-surgery. Results are summarized in Table 2, Figure 3, and Figure 4.
Table 2Summary of Hippocampal Subfield Changes After Surgical Intervention in the Responder Cohort
Wilcoxon-Paired Test was used for statistical analyses in pre- vs. post-surgery volumes, and Mann-Whitney U Test was used for comparing the pre- and post-surgery volumes to healthy controls. All reported volumes are corrected for subcortical volume. All p-values are corrected for multiple comparison using Bonferroni correction.
Abbreviations, CA, cornu ammonis; ML-HP, molecular layer of hippocampus proper; GC-ML-DG, granule cell and molecular layer of the dentate gyrus; Pre-Sub, pre-subiculum; Para-Sub, para-subiculum.
Figure 3Automated hippocampal subfields segmentation for 47 TN responders and their age- and sex-matched healthy controls using FreeSurfer 6.0 and its developmental package. Hippocampal segmentation for an MRI scan delineated by FreeSurgery 6.0. A-D are coronal views. E and F are sagittal view.
Figure 4Hippocampal subfields change after pain relief. Wilcoxon-Paired Test was used for statistical analyses in pre- vs. post-surgery volumes, and Mann-Whitney U Test was used for comparing the pre- and post-surgery volumes to healthy controls. All reported volumes are corrected for subcortical volume size (please refer to the method section). All p-values are corrected for multiple comparison using Bonferroni correction. (*P < 0.05, ** P < 0.01, *** P < 0.001).
Both male and female TN responders showed a bilateral significant increase in the whole hippocampus volume after pain relief (see Fig 5). Following pain relief, male TN responders showed 1.68% ± 2.53 and 2.19% ± 3.43 increase on average on ipsi- and contralateral side, respectively, compared to pre-surgery time point. The volumetric increase was 4.27% ± 4.60 and 4.34 ± 4.63 in females TN responders on the ipsi- and contralateral sides, respectively. These sex-stratified analyses showed that female TN responders had a greater volumetric increase on average compared to males, which was significantly different on the ipsilateral side (P = 0.031).
Figure 5Hippocampal segmentation stratified by sex in the responder cohort. Subfield volume changes are calculated as: VOIchange = [(VOIpost-treatment – VOIpre-treatment)/VOIpre-treatment] X 100. Wilcoxon test is used for statistical analyses (*P < 0.05, ** P < 0.01, *** P < 0.001). Black bars indicate ipsilateral results, whereas grey bars indicate contralateral results.
yet little are known about the significance or possible dynamic nature of the reported hippocampal abnormalities. TN, a robust, unilateral pain syndrome with a substantial likelihood of surgical pain relief presents an ideal model to study possible pain-related hippocampal abnormalities. In our previous study of the hippocampus in TN, we demonstrated that the hippocampal subfields CA1, CA4, ML HP, and DG
are bilaterally smaller in untreated patients. In the current study, we showed that subregions responsible for neurogenesis and memory consolidation, demonstrate a bilateral increase in volume. Furthermore, our results indicated that the whole hippocampus, its body, CA4, ML HP, DG subfields are normalized to expected healthy control levels after the resolution of pain. We further demonstrated that the normalization or volume-recovery highlights sex-specific hippocampal plasticity in TN with a greater volumetric increase in females compared to males. Our work directly fits with important studies pointing to decreased hippocampal volumes associated with chronic stress,
Our study highlights key subfields that are vulnerable to the effects of chronic pain, and points that successful treatment can lead to recovery and normalization. Additionally, it is plausible to assume that recovery of many symptoms commonly associated with pain, such as difficulty with concentration, decreased focus, and routinely reported mental fatigue
may in fact relate to the recovery and normalization of hippocampal structures.
Bilateral Increase in the Hippocampus and its Subregions Following Pain Resolution
The findings of our current study revealed that hippocampal volume recovers after pain relief in the responder cohort (see Fig 1). However, this increase is not observed in the non-responder cohort, who still suffer from persistent TN attacks. The recovery in responders suggests that pain relief could reverse the volumetric and microstructural abnormalities previously seen in the hippocampus. This is important to ponder when considering older individuals who are prone to cognitive decline characterized chiefly by hippocampus atrophy.
Our findings demonstrated bilateral volumetric increase and normalization in hippocampal subregions including the CA2/3, CA4, ML HP, DG, and subiculum (see Fig. 4). These hippocampal subfields are involved in various roles including behavioral regulation, stress, and shaping chronic pain experience.
As such, our results suggest pain relief which is followed by reduced TN attacks could reverse the changes in the CA3 hippocampal subfield. The hippocampus has complex afferent and efferent connections to diverse brain regions including the entorhinal cortex, cingulate cortex, prefrontal cortex, anterior thalamic nucleus, and hypothalamic mammillary bodies. Both the sensory afferent pathway (through the perforant pathway) and the efferent outputs (through the fornix) directly innervate the CA1 hippocampal subregion.
Witter M.P.: Connectivity of the Hippocampus. In: Cutsuridis V., Graham B., Cobb S., Vida I., editors. Hippocampal Microcircuits Springer New York; page 5–262010.
Here, we show a bilateral increase in CA1 subfield volume after pain relief in the responder cohort. This confirms our hypothesis that pain relief leads to increased neuroplasticity in the hippocampal region.
Neurogenesis Could Explain the Hippocampal Volume Recovery
As demonstrated in Figure 4, CA4 and DG, two hippocampal subfields involved in neuronal plasticity and neurogenesis, bilaterally increased in size, and normalized after pain resolution in TN responders. Grey matter plasticity can be due to changes in both neuronal and non-neuronal cells, including neurogenesis, gliogenesis, synaptogenesis, and other neuronal morphological changes.
Anxiety- and depression-like behavior and impaired neurogenesis evoked by peripheral neuropathy persist following resolution of prolonged tactile hypersensitivity.
In line with these previous findings, the current study and our previous investigation reported significant bilateral volume loss in the CA4, DG, and ML HP in TN patients – all of which considered primary hippocampal subfields involved in AHN.
As such, our current study provides a new line of evidence for hippocampal plasticity in chronic neuropathic conditions such as TN. Although it is not possible to pinpoint which mechanisms – neurogenesis, gliogenesis, or synaptogenesis – have caused the GM normalization, evidence suggests that the volume increase could be partially driven by neurogenesis. However, future studies are required to investigate the specific mechanisms underlying the positive effects of pain relief on AHN.
Axial Segmentation Reveals Hippocampal Body is Affected After Pain Relief
Our results demonstrated that the hippocampal body bilaterally increased in size in the responder cohort and is volumetrically normalized to the level of healthy controls (see Fig 3 and Fig 4). Furthermore, the hippocampal head showed contralateral volumetric normalization after pain relief. These changes are important considering recent studies that showed a functional gradient along the longitudinal axis of the human hippocampus.
In a recent meta-analysis of hippocampal abnormalities in chronic pain, Ayoub et al. reported that the right anterior hippocampus showed consistent abnormalities across studies. Furthermore, they tested the resting-state functional connectivity of this region in a large cohort of chronic low-back pain patients and found reduced connectivity to the medial prefrontal cortex in these patients compared to healthy individuals.
Although the hippocampus is involved in various functions, there is evidence that the anterior hippocampus is involved in gist memory, in other words in encoding contextual cues.
Our results support the idea that the anterior hippocampus may be involved in nociception and can be affected by nociceptive input.
Sex Differences in Response to Pain Relief
Our study demonstrates that the bilateral hippocampal volume increase is seen in both males and females after pain relief (Fig 5). However, females showed a larger increase on average compared to males — especially on the hippocampus ipsilateral to TN pain where volumes were statistically greater than in males. Previous studies have delineated that there are sex differences in the adverse effects of stress on cognitive tasks and neurogenesis.
In prolonged stress conditions, females showed a reduction in neurogenesis compare to males, suggesting stress exposure has a greater impact on females compared to male.
These results suggest that pain relief – which generally closely correlates with reduced stress – has a more significant impact on females than males. Considering that neurogenesis in females is more severely impacted by chronic stress, it is possible that pain relief re-establishes neurogenesis to a greater extent in females than males. As our study is the first investigation of the effect of pain relief on the hippocampus and its subfields, future studies are required to further delineate the mechanisms behind sex responses in pain relief.
Future Directions and Study Limitations
The onset of TN is usually after age 50 and the TN cohort investigated in the current study has an average age of 64.9 ± 12.0 years. Therefore, we were limited by the number of age-matched healthy controls collected on-site and had to utilize the Cam-CAN online dataset. We considered the potential impact of different imaging protocols and scanners, however, we ultimately found little evidence for that.
In the current study, the number of non-responders who did not experience pain relief after the treatment was relatively small. As such, we were statistically limited and had to focus on the responder cohort for subfield and sex-dependent analyses.
The severity of TN pain patients suffering from requires the patients to use pain medications and anticonvulsant drugs including carbamazepine, gabapentin, and pregabalin (please refer to Table 1). Although long-term use of antileptics could affect the hippocampal neuronal circuits,
no direct effect of these medications on the hippocampus has been reported and further investigation is required. It should also be noted that: I) given the severity of TN pain, it is not tenable to investigate a medication free population and TN studies are limited to control for the effect of medications, and II) albeit TN patients reduce their medications after GKRS, the majority still undergo pharmacological therapy and the effect of medications on the hippocampus is still present after surgery. Therefore, the findings in the current study reflect a typical TN population.
TN provides a unique opportunity to study the effect of pain relief on hippocampal abnormalities. However, few other pain conditions can be effectively treated as TN. As such, the findings of this study may be limited to TN patients and future studies are required to explore the hippocampal changes in other chronic pain conditions after treatment. Additionally, our study addresses the hippocampal changes after pain relief delineated by an MRI scan at 6 months post-GKRS. As such, the rate of recovery cannot be accurately calculated and future studies with multiple scans post-surgery are required to address this limitation. Although our study points to hippocampal recovery and possible cognitive function changes as the result, future studies are required to directly investigate this link.
Although our study did not report a statistically significant correlation between the amount of pain reduction and hippocampal size change, we note this is an important point to be investigated in future studies with a larger sample size and multiple scans after the surgery. Similarly, our analyses did not show any significant correlation between pain duration and volume changes. However, we hope future studies with a larger sample size would investigate this question and further explore sex difference and pain duration interactions.
References
Adnan A.
Barnett A.
Moayedi M.
McCormick C.
Cohn M.
McAndrews M.P.
Distinct hippocampal functional networks revealed by tractography-based parcellation.
Bair M.J., Robinson R.L., Katon W., Kroenke K.: Depression and pain comorbidity: A literature review. Arch. Intern. Med. American Medical Association; page 2433–452003.
A unified approach for morphometric and functional data analysis in young, old, and demented adults using automated atlas-based head size normalization: Reliability and validation against manual measurement of total intracranial volume.
Anxiety- and depression-like behavior and impaired neurogenesis evoked by peripheral neuropathy persist following resolution of prolonged tactile hypersensitivity.
A Computational Atlas of the Hippocampal Formation Using Ex vivo, Ultra-High Resolution MRI: Application to Adaptive Segmentation of in vivo MRI. 115. Neuroimage Academic Press Inc.,
2015: 117-137
Long-Term Outcomes in the treatment of classical trigeminal neuralgia by gamma knife radiosurgery: A retrospective study in patients with minimum 2-year follow-up.
Chronic Early Life Stress Alters Developmental and Adult Neurogenesis and Impairs Cognitive Function in Mice. 25. Hippocampus John Wiley and Sons Inc.,
2015: 309-328
Regis J., Tuleasca C., Resseguier N., Carron R., Donnet A., Gaudart J., Levivier M.: Long-term safety and effiacy of Gamma Knife surgery in classical trigeminal neuralgia: A 497-patient historical cohort study. J. Neurosurg. 1079–872016.
The Cambridge Centre for Ageing and Neuroscience (Cam-CAN) study protocol: A cross-sectional, lifespan, multidisciplinary examination of healthy cognitive ageing.
Wingenfeld K., Wolf O.T.: Stress, memory, and the hippocampus. In: Szabo K., Hennerici M.G., editors. Hippocampus Clin Neurosci S. Karger AG; page 109–202014.
Witter M.P.: Connectivity of the Hippocampus. In: Cutsuridis V., Graham B., Cobb S., Vida I., editors. Hippocampal Microcircuits Springer New York; page 5–262010.
Conflict of interest: The authors have no conflict of interest to declare.
Support: A.N. is supported by a Canadian Institutes of Health Research (CIHR) Frederick Banting and Charles Best Canada Graduate Scholarships-Master's, Ontario Graduate Scholarship, and the University of Toronto Centre for the Study of Pain (UTCSP) Scientist Award. P.S.P is supported by a CIHR Frederick Banting and Charles Best Doctoral Research Award [grant number: GSD157876]. M.M. is supported by the Bertha Rosenstadt Endowment held at the University of Toronto's Faculty of Dentistry, has salary support as a UTCSP Scientist, and holds an NSERC Discovery Grant. M.H. holds a CIHR Operating Grant [grant number: MOP130555].
Indicates situation in which patients could not tolerate pharmacological treatments.
00296-0&id=fx1.jpg){kind=link}
00296-0&id=gr1.jpg){kind=link}
00296-0&id=gr2.jpg){kind=link}
00296-0&id=gr3.jpg){kind=link}
00296-0&id=gr4.jpg){kind=link}
00296-0&id=gr5.jpg){kind=link}