Thalamic DBS for Neuropathic Pain After Amputation
Thalamic DBS for Neuropathic Pain After Amputation
Patients were referred by clinicians nationally (in Portugal) to a single-center multidisciplinary team consisting of pain specialists, neurologists, neuropsychologists, and neurosurgeons at Hospital de São João in Porto, Portugal. Neuropsychological evaluation excluded psychiatric disorders and ensured minimal cognitive impairment. Neuropathic medications and other surgical treatments may have been attempted with limited success. Selection criteria included neuropathic pain refractory to medicines for at least 2 years, together with absence of surgical contraindications such as coagulopathy or ventriculomegaly. Informed consent was obtained and the study received local ethical board approval. Patients were counseled for the possibility that they may derive no benefit from ventroposterolateral DBS or not tolerate it well, necessitating its removal.
For surgery, a Leksell stereotactic frame (Elekta Instruments) was applied to the patient's shaved head under local anesthesia. A stereotactic head CT scan with a 2-mm slice thickness was performed, with preoperative T1- and T2-weighted MR images of 2-mm slice thickness from a 1.5-T scanner volumetrically fused to the scan using FrameLink stereotactic calculation software (Medtronic). The Leksell stereotactic arc was then fixed to its frame on the awake, semisitting patient. Intravenous vancomycin and cefotaxime antibiotic prophylaxis and dexamethasone were routinely administered before surgery commenced. After bur hole exposure, a Leksell radiofrequency electrode impedance check was performed to 10 mm above target, then DBS electrode insertion to target, and intraoperative C-arm radiographic target confirmation.
A guiding principle in awake electrode targeting is the established somatotopic organization of the somaesthetic thalamic nuclei. Human microelectrode studies have revealed a mediolateral somatotopy in the contralateral ventral posterior thalamus, with the head of the homunculus medial and the feet lateral. The contralateral ventroposterolateral nucleus of the sensory thalamus was targeted for limb pain and found in the plane 10–13 mm posterior to the midcommissural point, and from 5 mm below to 2 mm above it. The arm area of the ventroposterolateral nucleus of the sensory thalamus was 2–3 mm medial, and leg area 1–2 mm medial, to the internal capsule. Thus, targeting was generally 12–14 mm lateral and 0–2 mm anterior to the posterior commissure. A transfrontal extraventricular trajectory from an entry point on or near the coronal suture to the ipsilateral ventroposterolateral nucleus of the sensory thalamus was planned, avoiding blood vessels and sulci.
Targets were implanted with Medtronic Model 3387 quadripolar electrodes with 0.5-mm contacts spaced 1.5 mm apart. Final electrode position was determined by intraoperative clinical assessment that relied upon subjective reporting by the awake patient rather than microelectrode recording. Tactile stimulation of the painful body part was performed intraoperatively to augment pain and assess response if necessary. At either target during surgery, DBS at lower frequencies (≤ 50 Hz) was analgesic and higher frequencies (> 70 Hz) hyperalgesic. Bipolar stimulation of 5–50 Hz was performed initially, with a pulse width of 200–450 μs and amplitude of 0.5–5 V. Stimulation of the ventroposterolateral nucleus of the sensory thalamus aimed to supplant painful sensation by pleasant paresthesias. Adjustment was primarily somatotopic, with the assessor alert to pyramidal signs suggesting capsular involvement. The electrode was rarely moved from its planned target, but occasionally moved 2 mm at a time lateral, medial, proximal, or distal during awake testing to optimize somatotopic coverage. Once satisfactory targeting had been achieved by microdrive adjustment, the patient's electrode was secured in place with a Stimloc skull fixation device (Medtronic). Electrode leads were externalized parietally via temporary extensions, and electrode position was confirmed by postoperative stereotactic CT, again fused to preoperative MRI. Typical surgery duration was 2–3 hours, including stereotactic planning.
Bedside DBS programming was undertaken during the postoperative period both morning and afternoon, in sessions lasting approximately 30 minutes per patient. Different frequencies from 5 to 50 Hz were used alongside varied bipolar stimulation over different contacts at increasing pulse widths and amplitudes to optimize analgesia without unpleasant sensations in the painful body part. Typically, pleasant paresthesia was elicited and occasionally analgesia without paresthesia, mirroring intraoperative results (Video 1).
Video1. Clip showing a postoperative patient with BPA pain who underwent a trial of externalized DBS, describing pain improvement. Copyright Erlick A. C. Pereira. Published with permission. Click here to view with Windows Media Player. Click here to view with Quicktime.
After 48 hours of postoperative clinical assessment, a decision was made whether to permanently implant the electrodes under general anesthesia. The electrodes were connected to a pulse generator (Medtronic Kinetra or Activa PC) implanted in the chest via new extension leads. During postoperative assessment, each patient recorded visual analog scale (VAS) scores at least twice daily at set times while blinded to DBS settings. Targets were tested using analgesic stimulation parameters to determine which electrode contacts conferred maximum analgesia. Visual analog scale scores were averaged for the trial period and compared with preoperative scores to assess for improvement. The decision to proceed to implantation of the pulse generator was made for each individual patient, guided by the multidisciplinary team.
Patients ideally left the hospital a day after implantation of the pulse generator, with progress followed up by clinic appointments at 1, 3, and 6 months, and then yearly thereafter. Continuous rather than on-demand stimulation was encouraged, but in addition to the ability to switch the DBS on and off at will, patients were usually given control over its voltage only, which was typically limited by the clinician to a maximum amplitude of 4 V.
Quantitative assessment of pain and health-related quality of life were performed 1 month before surgery and postoperatively at 1, 3, and 6 months, and then annually by independent, blinded assessors trained in pain medicine but not involved in caring for the patient and unaware of the details of the neurosurgical treatment given. The VAS to rate pain intensity, the University of Washington Neuropathic Pain Score (UWNPS), and Brief Pain Inventory (BPI) were used. Patients recorded VAS scores twice daily in a pain diary over 14 days. The 28 VAS scores were reviewed to ensure consistency, and the mean was then calculated.
Patients also completed a 36-Item Short-Form Health Survey (SF-36) on quality of life along with the pain questionnaires. The SF-36 responses were regrouped into 8 domains of physical functioning: role (physical), bodily pain, general health, vitality, social functioning, role (emotional), and mental health. Results were scored by online tools (http://www.sf-36.org/demos/sf-36.html). Norm-based scores allowed comparison between studies.
As pain assessments were made as repeated measures throughout the follow-up, preoperative and postoperative scores were compared for each group of patients using a general linear mixed model, with a p value < 0.05 considered statistically significant and p < 0.01 considered highly significant.
Methods
Study Population
Patients were referred by clinicians nationally (in Portugal) to a single-center multidisciplinary team consisting of pain specialists, neurologists, neuropsychologists, and neurosurgeons at Hospital de São João in Porto, Portugal. Neuropsychological evaluation excluded psychiatric disorders and ensured minimal cognitive impairment. Neuropathic medications and other surgical treatments may have been attempted with limited success. Selection criteria included neuropathic pain refractory to medicines for at least 2 years, together with absence of surgical contraindications such as coagulopathy or ventriculomegaly. Informed consent was obtained and the study received local ethical board approval. Patients were counseled for the possibility that they may derive no benefit from ventroposterolateral DBS or not tolerate it well, necessitating its removal.
Deep Brain Stimulation Procedure
For surgery, a Leksell stereotactic frame (Elekta Instruments) was applied to the patient's shaved head under local anesthesia. A stereotactic head CT scan with a 2-mm slice thickness was performed, with preoperative T1- and T2-weighted MR images of 2-mm slice thickness from a 1.5-T scanner volumetrically fused to the scan using FrameLink stereotactic calculation software (Medtronic). The Leksell stereotactic arc was then fixed to its frame on the awake, semisitting patient. Intravenous vancomycin and cefotaxime antibiotic prophylaxis and dexamethasone were routinely administered before surgery commenced. After bur hole exposure, a Leksell radiofrequency electrode impedance check was performed to 10 mm above target, then DBS electrode insertion to target, and intraoperative C-arm radiographic target confirmation.
A guiding principle in awake electrode targeting is the established somatotopic organization of the somaesthetic thalamic nuclei. Human microelectrode studies have revealed a mediolateral somatotopy in the contralateral ventral posterior thalamus, with the head of the homunculus medial and the feet lateral. The contralateral ventroposterolateral nucleus of the sensory thalamus was targeted for limb pain and found in the plane 10–13 mm posterior to the midcommissural point, and from 5 mm below to 2 mm above it. The arm area of the ventroposterolateral nucleus of the sensory thalamus was 2–3 mm medial, and leg area 1–2 mm medial, to the internal capsule. Thus, targeting was generally 12–14 mm lateral and 0–2 mm anterior to the posterior commissure. A transfrontal extraventricular trajectory from an entry point on or near the coronal suture to the ipsilateral ventroposterolateral nucleus of the sensory thalamus was planned, avoiding blood vessels and sulci.
Targets were implanted with Medtronic Model 3387 quadripolar electrodes with 0.5-mm contacts spaced 1.5 mm apart. Final electrode position was determined by intraoperative clinical assessment that relied upon subjective reporting by the awake patient rather than microelectrode recording. Tactile stimulation of the painful body part was performed intraoperatively to augment pain and assess response if necessary. At either target during surgery, DBS at lower frequencies (≤ 50 Hz) was analgesic and higher frequencies (> 70 Hz) hyperalgesic. Bipolar stimulation of 5–50 Hz was performed initially, with a pulse width of 200–450 μs and amplitude of 0.5–5 V. Stimulation of the ventroposterolateral nucleus of the sensory thalamus aimed to supplant painful sensation by pleasant paresthesias. Adjustment was primarily somatotopic, with the assessor alert to pyramidal signs suggesting capsular involvement. The electrode was rarely moved from its planned target, but occasionally moved 2 mm at a time lateral, medial, proximal, or distal during awake testing to optimize somatotopic coverage. Once satisfactory targeting had been achieved by microdrive adjustment, the patient's electrode was secured in place with a Stimloc skull fixation device (Medtronic). Electrode leads were externalized parietally via temporary extensions, and electrode position was confirmed by postoperative stereotactic CT, again fused to preoperative MRI. Typical surgery duration was 2–3 hours, including stereotactic planning.
Bedside DBS programming was undertaken during the postoperative period both morning and afternoon, in sessions lasting approximately 30 minutes per patient. Different frequencies from 5 to 50 Hz were used alongside varied bipolar stimulation over different contacts at increasing pulse widths and amplitudes to optimize analgesia without unpleasant sensations in the painful body part. Typically, pleasant paresthesia was elicited and occasionally analgesia without paresthesia, mirroring intraoperative results (Video 1).
Video1. Clip showing a postoperative patient with BPA pain who underwent a trial of externalized DBS, describing pain improvement. Copyright Erlick A. C. Pereira. Published with permission. Click here to view with Windows Media Player. Click here to view with Quicktime.
After 48 hours of postoperative clinical assessment, a decision was made whether to permanently implant the electrodes under general anesthesia. The electrodes were connected to a pulse generator (Medtronic Kinetra or Activa PC) implanted in the chest via new extension leads. During postoperative assessment, each patient recorded visual analog scale (VAS) scores at least twice daily at set times while blinded to DBS settings. Targets were tested using analgesic stimulation parameters to determine which electrode contacts conferred maximum analgesia. Visual analog scale scores were averaged for the trial period and compared with preoperative scores to assess for improvement. The decision to proceed to implantation of the pulse generator was made for each individual patient, guided by the multidisciplinary team.
Patients ideally left the hospital a day after implantation of the pulse generator, with progress followed up by clinic appointments at 1, 3, and 6 months, and then yearly thereafter. Continuous rather than on-demand stimulation was encouraged, but in addition to the ability to switch the DBS on and off at will, patients were usually given control over its voltage only, which was typically limited by the clinician to a maximum amplitude of 4 V.
Outcome Assessment
Quantitative assessment of pain and health-related quality of life were performed 1 month before surgery and postoperatively at 1, 3, and 6 months, and then annually by independent, blinded assessors trained in pain medicine but not involved in caring for the patient and unaware of the details of the neurosurgical treatment given. The VAS to rate pain intensity, the University of Washington Neuropathic Pain Score (UWNPS), and Brief Pain Inventory (BPI) were used. Patients recorded VAS scores twice daily in a pain diary over 14 days. The 28 VAS scores were reviewed to ensure consistency, and the mean was then calculated.
Patients also completed a 36-Item Short-Form Health Survey (SF-36) on quality of life along with the pain questionnaires. The SF-36 responses were regrouped into 8 domains of physical functioning: role (physical), bodily pain, general health, vitality, social functioning, role (emotional), and mental health. Results were scored by online tools (http://www.sf-36.org/demos/sf-36.html). Norm-based scores allowed comparison between studies.
Statistical Analysis
As pain assessments were made as repeated measures throughout the follow-up, preoperative and postoperative scores were compared for each group of patients using a general linear mixed model, with a p value < 0.05 considered statistically significant and p < 0.01 considered highly significant.
Source...