DBS has been used in a number of targets for pain, although recent literature suggests a focus on the sensory thalamus or VP thalamus, the PAG/PVG, and the ACC. DBS has been used to treat a number of pain diagnoses, including phantom limb pain, brachial plexus injury, stroke, cephalalgias, multiple sclerosis, spine injury, failed low back surgery syndrome, and chronic cluster headaches, as reviewed in several studies.^1,2,3,4,5,6

The use of PAG/PVG targets began from animal research that demonstrated analgesia from stimulation of these areas, with subsequent translation to patients.^7,8,9,10,11 Initial studies suggested that DBS in this region acted via opioidergic mechanisms, supported by a reported effect reversal with naloxone, but later studies challenged these conclusions.^1,12 Current models suggest that the ventral PAG works via influencing coping behavior, whereas the dorsal PAG has sympathomimetic effects, which may also have opioidergic effects.^12,13 Overall, the PAG target has been primarily used for nociceptive pain.^2,14

DBS for sensory thalamic targets, the ventral posterior lateral and medial (VPL/VPM) nuclei, arose from initial work that demonstrated symptomatic relief from ablative procedures.^12,15,16,17 In these patients, pain is thought to result in increased thalamic firing,¹⁷ and DBS to this region likely affects thalamo-corticofugal descending pathways.⁵ Recordings at these sites have demonstrated increased spike density as well as possible neural hyperactivity when these regions are stimulated.¹⁸ Overall, this target has been favored for neuropathic pain.^2,14

DBS to both VP and PAG/PVG targets is also thought to exert effects by either disrupting synchronous pathologic high frequencies or by enhancing low frequencies.^1,12 Studies have demonstrated that at both VP and PAG targets, reduction in pain results from lower frequencies ≤50 Hz, whereas increased sensitivity to pain results from higher frequencies >70 Hz.^4,12 One group, which has published extensively on DBS for pain, suggests that only demonstrable efficacy should guide the decision for implanting an electrode at the PAG, VP, or both sites,^1,17 although their own practice is to target the PAG site initially, except in cases of head and facial pain.¹⁹ Importantly, some patients may have components of both neuropathic and nociceptive pain and may therefore derive benefit from DBS therapy to both PAG and VP sites.²

The ACC has also been reported as a successful DBS target for pain based on the effectiveness of cingulotomy.^{20,21,22,23,24} The ACC is thought to modulate pain processing and is involved in the affective component of pain. As a result, DBS therapy to this target likely affects emotional components to pain rather than nociceptive perception, thereby reducing the “unpleasantness” of pain rather than the pain itself.^1,4,6,25

Other targets that have additionally been reported in the literature include the internal capsule,^26,27 ventral striatum and internal capsule,^28,29 centromedian parafascicular complex and the intra-laminar zone,³⁰ and posterior hypothalamus.^31,32

Initial studies of DBS for pain in the United States were based on two multicenter trials from 1989 to 1995, which were not randomized or case controlled.^1,33 Neither trial met criteria for efficacy, with the first trial defining effective therapy as ≥50% of patients with ≥50% pain relief at 1-year follow-up and the second trial defining efficacy as ≥50% of patients with ≥30% pain relief and decrease in the usage of analgesia medication.³³ Multiple reviews have subsequently commented on the limitations of these trials, which include patient heterogeneity, lack of blinded assessment, and lack of standardization among centers.^1,17 Nevertheless, as a result of these studies, DBS for pain in the United States has been used with an off-label status.³³

Subsequent studies, reviews, and meta-analyses have provided a mixed picture regarding the efficacy of DBS for pain. Many published studies have limited numbers of patients, with few centers reporting larger numbers. Ultimately, long-term efficacy has been reported in ˜83% of patients,¹ with one large case series by a well-published center suggesting an overall long-term efficacy of 67% at their center, with no significant differences in efficacy based on stimulation site.¹⁹

Several meta-analyses have suggested efficacy of DBS for pain. A 2010 meta-analysis examined 13 studies and estimated
that 50% of patients had long-term relief with DBS, with a 61% success rate for nociceptive pain diagnoses and 42% success rate with neuropathic pain diagnoses.² A 2005 meta-analysis of 6 older studies (from 1977 to 1997) had also demonstrated 55% to 70% pain relief at >1-year follow-up.¹³ This meta-analysis suggested good to excellent results with PAG stimulation, with further increases in success when combined with thalamic or internal capsule stimulation, a strategy that has been seen in other studies.³⁴

However, there have been limitations to in-depth analysis of both individual studies and meta-analyses due to factors such as heterogeneity in patient diagnoses and selection criteria, DBS sites, stimulation parameters, and use of unblinded assessments. These limitations have been reflected in recent recommendations from the European Academy of Neurology (EAN) in 2016 and the European Federation of Neurological Societies (EFNS) in 2007.^5,35 The EAN’s meta-analysis of studies from 2006 to 2014 that used DBS for a variety of pain diagnoses determined that there was very low quality of evidence.³⁵ Overall, the EAN gave a recommendation of “inconclusive,” highlighting the further need for additional prospective, randomized, controlled trials. Nevertheless, they noted that the overall pain intensity reduction was close to 50% in their included studies.³⁵ This was an update to the EFNS 2007 study, which had calculated an overall 46% long-term success rate of DBS for pain.⁵ Further investigations will be required to determine whether specific DBS targets are best suited for particular diagnoses, as there has been lack of consensus.^5,35,36

Few studies have also utilized randomized or placebocontrolled experimental designs. Marchand et al.³⁷ studied thalamic stimulation and examined the effects of placebo stimulation, ultimately demonstrating a significant placebo effect from thalamic DBS. Rasche et al.³⁴ included placebo testing during the trial stimulation phase; ultimately, 57% of patients passed this phase with demonstrated analgesic effects and were implanted with a pulse generator. Fontaine et al.³⁸ reported a randomized, placebo-controlled, double-blind trial to study the effects of hypothalamic DBS in cluster headache and found that although there was lack of efficacy during the randomized phase of the trial, patients derived a reduction in their attacks during the open phase. Finally, Lempka et al.²⁸ targeted the ventral striatum/anterior limb of the internal capsule in a prospective, double-blind, randomized, placebo-controlled, crossover study and did not meet its primary endpoint of ≥50% improvement in ≥50% of patients with active DBS compared to placebo. However, there were improvements in other measures such as depression, anxiety, and quality of life. These studies have suggested legitimate effects from DBS, although its effects may not be adequately measured purely by pain intensity ratings.

Overall, there have been persistent limitations to comparing published studies, which have included diverse patient selection criteria and diagnoses, heterogeneous methodology, and diversity of intracranial targets. In addition, certain effects may not have been fully studied or reported. For example, DBS for pain may result in substantial insertional effects.³⁹ As increasingly examined by recent published studies, future investigations may also benefit from assessing patient improvements in areas other than pain, such as mood, anxiety, and quality of life, which may improve to a significant degree.⁴⁰ Although a number of studies have suggested that DBS may be more effective for particular diagnoses, or identified particular targets as being more effective for specific diagnoses, there does not yet appear to be general consensus.^13,17,36

Some studies have demonstrated loss of efficacy over time, with some series noting that up to a half of patients who have success during the trial stimulation period do not have sustained benefit at ≥1 year follow-up, which may be attributed to factors such as scarring around electrode targets, brain plasticity, and inconsistent patient reporting.^17,25 To mitigate these effects, groups have resorted to intensive reprogramming, with authors noting that stimulation parameters may increase with time.^1,41 Strategies for reprogramming include changing pulse width or frequency or allowing for cycled stimulation or “off” periods.¹⁷

Optimal patient selection has been important in ensuring that those who undergo DBS benefit from therapy. A multidisciplinary approach is used to screen potential patients who may benefit from DBS, and neuropsychological testing is a key component, with patients who have psychological etiologies excluded. Patients are also typically medication refractory ≥2 years, but there are no restrictions on whether patients have undergone other surgical procedures.^1,17 Some groups examine patients’ responses to analgesics prior to DBS.¹⁸ Quantitative and qualitative assessment of pain is required, as well as quality of life assessments, which may allow clinicians and patients to best see changes that occur after DBS. Contraindications include medical conditions that would make surgery unsafe or anatomic factors such as ventriculomegaly that would prevent placement of an electrode into the surgical target.

DBS electrodes are implanted stereotactically, using either a frame or frameless approach (Fig. 97.1).² As a result, high-resolution magnetic resonance imaging (MRI) is required to allow for accurate targeting. Surgery is performed with
sedation and local anesthesia; however, some targets, such as the ACC, can be performed under general anesthesia. Targets in the thalamus and midbrain are contralateral to the symptomatic side, whereas the ACC target requires bilateral electrode implantation.¹

For targeting coordinates, the PAG is typically found 2 to 3 mm lateral to the third ventricle at the level of the posterior commissure and 10 mm posterior to the midcommissural point and 0 mm vertical; the VP target is typically found 10 to 13 mm posterior to the midcommissural point, 14 to 18 mm lateral, and from 5 mm below to 2 mm above the midcommissural line (Fig. 97.2).^1,4 To isolate a particular limb, studies have suggested that the arm representation of the VPL is 2 to 3 mm medial to the internal capsule, whereas the leg representation is 1 to 2 mm medial to the internal capsule.⁴ The ACC target is typically 20 to 25 mm posterior to the anterior horns of the lateral ventricles, with the electrode tip near the corpus callosum and the electrode in the cingulum bundle.^4,25

Intraoperative physiologic stimulation is used to refine the final target, with microelectrode recording, microstimulation, and macrostimulation. This step is considered critical in order to obtain adequate coverage of the symptomatic region. For instance, patients undergoing VP stimulation should experience paresthesias in previously painful areas, whereas those undergoing PAG stimulation should note a sensation of warmth or analgesia.^4,13 Once targets are identified, the electrodes are introduced, and the leads are externalized for trial stimulation. Patients undergo postoperative imaging, typically via computed tomography (CT), to look for possible complications.

A trial assessment period typically lasts from 5 to 9 days, during which physicians may test different combinations of stimulation parameters to optimize patient effects.² Different approaches have been taken for assessing effectiveness, with some authors using an “N-of-1 trial” approach, in which a patient undergoes randomized pairs of treatments to elucidate the effects of DBS and placebo.¹⁷ Overall, a large percentage of patients typically pass the trial stimulation period.^36,39,42 Some authors have noted that DBS to the ACC may take an extended period to demonstrate effectiveness.⁴ Stimulation parameters have varied across the literature, with some groups reporting bipolar stimulation parameters in the range of 5 to 50 Hz with pulse widths from 100 to 450 µs and amplitudes from 0.1 to 3 V, although ACC stimulation may require higher settings.^1,17 Afterward, the most effective electrodes are internalized and connected to either an infraclavicular or abdominal pulse generator.

Overall, DBS is considered to be a generally safe procedure, with adverse events occurring in 8% to 9% of cases.³⁵ One of the more serious complications is hemorrhage, which has been reported in 1.9% to 4.1% of cases.^2,43 Device malfunction may also occur. There is a risk of infection, which has been reported in 1.9% to 13.3% of cases.^14,43 Other minor risks include diplopia and visual gaze effects, headache, and nausea.^14,17 The risk of permanent neurologic deficit has ranged from 2% to 3.4%, and mortality is rare, ranging from 0% to 1.6%.²

The use of MCS was initiated after a study by Tsubokawa et al.⁴⁴ in 1991, with the demonstration that chronic stimulation of the motor cortex with epidural electrodes inhibited thalamic hyperactivity and ultimately led to symptom control. Following this study, neuromodulation of the motor cortex was examined as an option for the treatment of medically intractable central and peripheral neuropathic pain.^45,46 Based on these initial studies, MCS has been subsequently made an option for patients who have failed pharmacologic therapy or other stimulation techniques like spinal cord stimulation.⁴⁷