Relevant History of the Topic Including Developments Over the Past Decade

The application of neurostimulation to the management of chronic pain syndromes remains a significant challenge for chronic pain specialists. The origins of deep brain stimulation for intractable pain can be traced back to rodent studies in the 1950s. James Olds and Peter Milner implanted electrodes into various targets in the rodent brain, and found such powerful effects of positive reinforcement with electrodes targeted to the septal region that rodents would self-stimulate for prolonged periods whenever allowed to do so. Concurrent observations from clinical studies in which the septal nuclei, including the diagonal band of Broca, were stimulated in patients with schizophrenia confirmed the translational potential of neuromodulation for pain, as patients receiving septal stimulation experienced serendipitous analgesic side-effects ( ). There followed a period of intense preclinical and clinical investigation designed to elucidate further the mechanisms underpinning the powerful analgesic effects reported.

Patients suffering from intractable pain due to malignant oncological diseases and rheumatoid arthritis provided an ethically justifiable source of trial subjects, and septal region deep brain stimulation (DBS) for cancer pain proved moderately effective in early studies. Septal stimulation fell out of favor, however, despite some studies reporting success rates of 60%, and this led to the investigation of alternative deep brain targets for treatment of intractable pain. In subsequent years a wide range of intracranial targets were investigated, including the internal capsule ( ), the ventral posterolateral (VPL) and ventral posteromedial (VPM) sensory nuclei of the thalamus, the centromedian–parafascicular (CmPf) intralaminar region of the thalamus, the periventricular and periaqueductal grey (PVG/PAG), the nucleus accumbens, and the anterior cingulate cortex.

The first studies of thalamic stimulation in the 1960s were based on the historical hypothesis that stimulation of thalamic sensory nuclei would allow the “proprioceptive or epicritic” component of sensation to balance the “protopathic or nociceptive” pain system. The gate-control theory emerged to displace these classical pain theories and provided a conceptual framework for the investigation of sensory internal capsule and VPM–VPL stimulation for the modulation of deafferentiation and neuropathic pain ( ). Positive results with VPM lesioning supported early studies of VPM–VPL stimulation, which consistently demonstrated acute responses to stimulation followed by tolerance to stimulation and recurrence of pain. Reports of benefit in patients suffering from anesthesia dolorosa, complex regional pain syndrome, arachnoiditis, phantom limb pain, and nerve root avulsion confirmed the utility of thalamic stimulation in a wide range of chronic pain states ( ). However, early success rates of 60%–80% decreased to only 30%–40% in later responses, highlighting the need for further investigation.

Advances in the understanding of the endogenous opioid analgesic system led to the concept of “stimulation-produced analgesia” and stimulation of the lateral PVG or PAG. Successful relief from somatoform or nociceptive pain with PVG/PAG stimulation in humans was first reported in the late 1970s ( ). Despite evidence of tolerance to PVG/PAG stimulation, the 1970s and 1980s saw a rapid rise in the use of DBS for pain with over 1000 cases performed in a 16-year period. Indications for stimulation were multifarious and inconsistent, and included cancer pain, rheumatoid arthritis, complex regional pain syndrome, central poststroke and thalamic pain, spinal cord injury, atypical facial pain and trigeminal neuralgia, anesthesia dolorosa, postherpetic neuralgia, failed back surgery syndrome, brachial plexus injury, phantom limb pain, multiple sclerosis, and genital pain ( ).

This explosion of neurosurgical activity resulted in attention from the United States Food and Drug Administration (FDA), which requested the manufacturers of DBS technology to undertake formal clinical trials to determine the safety and efficacy of DBS for chronic pain.

Review of the Current Knowledge Base of the Topic

Medtronic reported results from a trial using its first-generation model 3380 DBS electrode in 1976. The trial was adversely affected by poor enrolment and high attrition rates. Subsequent analysis of the trial design revealed deficiencies in patient selection criteria resulting in a heterogeneous patient population, as well as inconsistencies in deep brain targets, the number of electrodes used, and stimulation parameters. Heavy criticism of this trial led to a second Medtronic study using a more advanced electrode in 1990 (model 3387). This trial was stopped after recruitment of 50 patients due to poor enrolment rates, high rates of patient withdrawal, and limited evidence of efficacy. Neither trial was able to satisfy the predefined endpoint criterion of more than 50% of patients achieving more than 50% pain relief. Medtronic chose not to seek FDA approval of its DBS system for analgesia, and DBS for pain was designated investigational and “off-label.”

The failure of these trials highlighted the importance of strict patient selection criteria, including psychological screening to exclude patients with psychogenic or factitious disorders, cognitive impairment, and psychiatric disease. These principles formed the basis for the contemporary multidisciplinary approach used in many centers continuing to use DBS as a therapeutic strategy for intractable pain.

Summary of the Safety and Efficacy Data

A paucity of large, randomized, controlled studies in recent years means that the current evidence base is derived in a large part from metaanalyses of multiple small studies. A metaanalysis of 13 studies with long-term outcomes for 1114 patients revealed long-term pain relief in 50% of patients. However, success rates ranged from 19% to 79%, with a reduction in response with increasing length of follow-up. Interestingly, this metaanalysis revealed a trend toward higher long-term pain control in patients with nociceptive pain rather than neuropathic pain ( ). Other important findings from this study were:

• •
  

  56% long-term success for neuropathic pain with VPL stimulation, compared with 0% for nociceptive pain

  
  
• •
  

  59% long-term amelioration of nociceptive pain with PVG stimulation, compared with 23% for neuropathic pain.

This data supported the role of thalamic stimulation for neuropathic pain and PVG/PAG stimulation for nociceptive pain. However, subsequent studies have overturned and refuted these conclusions, reflecting the impossibility of a “one-target-fits-all” approach to DBS for chronic pain.

It was also evident that pain due to nerve root avulsion, peripheral neuropathy, and failed back surgery syndrome responded more predictably to DBS than thalamic pain syndromes. However, the significant limitations of the source data with respect to patient selection, DBS technology, and stimulation parameters must be acknowledged.

Approval of DBS technology for the treatment of movement disorders paved the way for further investigation of “off-label” DBS for pain, with more advanced technology and improved patient selection criteria. However, results from more recent clinical trials have shown wide variation in response to stimulation of thalamic targets and PVG/PAG. A Canadian report of ventrocaudalis (Vc) thalamic (13 patients) and PVG/PAG stimulation (9 patients) reported a high rate of pain relief with electrode insertion alone. Vc stimulation appeared more effective than PVG/PAG in 5 patients achieving long-term pain control ( ). However, a study from the UK reporting on DBS for phantom limb pain reported contrary results, with up to 70% pain relief with PVG stimulation and further effects with combined PVG and sensory thalamic stimulation ( ).

There has also been a resurgence of interest in targeting the thalamic intralaminar CmPf, due to its extensive connectivity with cortical and basal ganglia regions. Preclinical and clinical studies indicate a role in modulating the affective and motivational components of pain, as well as in pain processing.

A prospective analysis of 197 patients assessed for DBS in Oxford, UK, provides contemporary support for the continued use of neuromodulation for patients with intractable pain for whom conventional treatments have failed. This study reported on the long-term outcomes for 85 patients receiving DBS for a range of pathologies, including postamputation pain, poststroke pain, brachial plexus avulsion, and cephalalgias. Patients underwent electrode insertion into either the contralateral PVG or ventral posterior (VP) nucleus of the thalamus, or insertion of dual electrodes into both targets. Of the implanted patients, 66% reported improvements in pain control, with the most significant benefits in patients suffering postamputation and poststroke pain.

Advances in neuroimaging and technology are likely to have contributed to the excellent safety profile of stereotactic neurosurgical procedures including DBS. However, the literature suggests a higher rate of stimulation-related adverse effects with PAG stimulation compared with other targets, including oscillopsias, nystagmus, vertigo, nausea, and anxiety.

The UK’s National Institute for Health and Clinical Excellence (NICE) is responsible for issuing evidence-based guidance on medical and surgical interventions. NICE issued guidance on DBS for refractory chronic pain syndromes (excluding headache) in 2011, and remarked on the heterogeneous treatment protocols and patient groups which comprise the evidence base. This guidance acknowledges the efficacy of the procedure in some patients with refractory pain syndromes selected by a multidisciplinary team specializing in pain management ( ).