Central and Deafferentation Pain Syndromes

Evidence status. Invasive treatment evidence is syndrome-specific and generally low certainty. DREZ lesioning has a defined role for brachial plexus avulsion pain; MCS, DBS, SCS, and DRG stimulation have variable or investigational roles in central and phantom pain.

1.Name the pain components

Document spontaneous burning, electric paroxysms, evoked allodynia, anesthesia dolorosa, stump pain, phantom pain, musculoskeletal pain, visceral pain, spasticity-related pain, and autonomic symptoms separately. Map sensory loss and the painful field. Review timing after the lesion, stability, rehabilitation, prosthetic use, sleep, mood, and medication response.

2.What each modality actually delivers

3.Central post-stroke pain

Central post-stroke pain usually follows a lesion along spinothalamocortical pathways and may appear after a delay. Pooled prevalence is roughly 11% across unselected stroke populations, with higher estimates in selected thalamic, medullary, and somatosensory-pathway cohorts; lesion-specific series should not be generalized to all strokes. Burning, cold dysesthesia, allodynia, and hyperalgesia within an area of sensory deficit are typical, and abnormal spinothalamic sensory testing supports the diagnosis; the lesion need not be confined to the thalamus, and a clinically normal sensory examination does not by itself exclude the diagnosis. First confirm that shoulder pain, spasticity, headache, CRPS, entrapment, and depression are not being collapsed into one diagnosis.

Medication and rehabilitation remain first-line. Noninvasive M1 stimulation is often used to frame the hypothesis before an implant is discussed, and a response is reassuring, but it is not a validated stand-alone predictor of motor cortex stimulation outcome and should not be presented to the patient as a trial. For the rare, severely refractory patient, MCS has the most coherent intracranial rationale: the most recent synthesis, pooling 32 retrospective studies and 330 patients, reports meaningful pain improvement in about 64% after MCS and 62% after DBS at mean follow-up under two years. The authors call the comparison preliminary, and it rests entirely on uncontrolled data. Set that against the only randomized, double-blind, sham-controlled DBS trial in this population, which targeted the ventral striatum and anterior limb of the internal capsule in ten patients and did not meet its primary endpoint, although affective and quality-of-life measures improved. DBS may also target sensory thalamus, posterior limb of internal capsule, PAG/PVG, or ACC in research settings. The honest framing for the patient is that the uncontrolled series look encouraging, the one controlled trial did not, and central pain is among the less predictable DBS phenotypes.

4.Brachial plexus avulsion

Preganglionic root avulsion deafferents dorsal-horn neurons and classically produces continuous burning pain plus electric or crushing paroxysms in an insensate, weak limb. This is the syndrome in which DREZ lesioning has its clearest rationale: the operation treats the hyperactive dorsal-horn generator across the avulsed levels. Paroxysmal pain often responds better than continuous background pain, though the series behind that observation are small and continuous background pain is not by itself a reason to decline the operation.

Quote the outcome as a range across uncontrolled series rather than as an expected result. A pooled literature review of 692 patients reported greater than 50% pain reduction in about 82%, with neurological complications in roughly 22% (motor deficits in about 11% and sensory deficits in about 11%, including both transient and permanent events); other reported complications totaled about 1.9%. Inconsistent reporting prevents a reliable pooled estimate of permanent neurological disability; the 11% categories must not be presented as exclusively transient. Durability is the harder number: a contemporary series of 44 patients followed 31 of them beyond twelve months, at a mean of 41 months, and found the mean visual analogue score fell from 9.0 to 4.1, with long-term results graded good in 39%, fair in 29%, and poor in 32%, and 55% able to stop or reduce analgesics.

Preoperative workup should establish which roots are avulsed and distinguish pre- from postganglionic injury, and it should do so with more than imaging. A Horner syndrome points to T1 or C8 preganglionic involvement. Sensory nerve action potentials that are preserved in a clinically anesthetic territory are the signature of a preganglionic lesion, because the dorsal root ganglion and its peripheral axon survive; this is easily misread as evidence against injury. Denervation of paraspinal, rhomboid, and serratus muscles localizes proximal to the plexus. Pseudomeningocele is a classic but insensitive sign: a substantial share of avulsed roots produce none, so imaging tends to underestimate how many levels are avulsed, and the surgical decision should not rest on imaging alone. Document residual motor and sensory function and review any reconstruction before committing to an irreversible lesion. MCS, SCS, and DBS can be considered in selected cases, but neuromodulation should not obscure the unusually strong generator-target match of DREZ for true root avulsion pain.

Plan for a target you may not be able to see. Cord distortion from adhesive arachnoiditis, atrophy, and gliosis at the avulsed levels is common, and identifying the dorsolateral sulcus is often the hardest step of the operation. Free the cord and roots first, identify the sulcus above and below the avulsed segment where intact rootlets remain, and treat the physiological adjuncts as part of the operation rather than as optional: impedance mapping helps separate injured, gliotic tissue from normal gray and white matter; SSEPs and MEPs monitor pathway integrity when the anatomy is distorted (they do not directly delineate tract boundaries); direct stimulation confirms the functional status of remaining ventral roots; and intraoperative recording of dorsal-horn hyperactivity has been used both to guide lesioning and to identify pain-mediating levels the preoperative plan did not anticipate.

5.Spinal cord injury pain

SCI pain is not one entity. Separate at-level radicular/segmental pain, below-level central pain, musculoskeletal overuse, visceral pain, pressure injury, spasticity, and syrinx or instability. A complete lesion with diffuse below-level burning is biologically different from an incomplete lesion with preserved dorsal-column pathways and a focal painful field.

The operative decision turns on where in the injured cord the pain sits. DREZ lesioning is best supported for at-level, segmental "end-zone" pain (pain in the dermatomes corresponding to the injured segments) and performs poorly for diffuse below-level burning pain, particularly perineosacral pain in a completely anesthetic field. No simple completeness or body-region rule establishes candidacy. Residual motor, sensory, and autonomic function, the segmental pain map, distorted anatomy, and limited syndrome-specific evidence must guide an experienced multidisciplinary team; clinical completeness does not authorize unrestricted cord lesioning.

SCS is most plausible when dorsal-column pathways are sufficiently preserved and pain is segmental or incomplete; its evidence for central SCI pain remains inconclusive. MCS and DBS are last-resort research strategies. Intrathecal baclofen is principally a treatment for severe spasticity. It may reduce associated pain, but should not be presented as an established stand-alone treatment for central neuropathic pain.

6.Phantom limb and residual-limb pain

Phantom limb pain reflects distributed peripheral and central reorganization; residual-limb pain may instead arise from neuroma, ischemia, skin breakdown, bone, socket fit, or joint disease. Treat those remediable generators first, and treat them definitively: for a symptomatic amputation neuroma, a 28-patient randomized trial of targeted muscle reinnervation versus standard neuroma excision and burial found that the one-year intention-to-treat comparisons did not reach significance (phantom pain adjusted P = .06; residual-limb pain P = .15), while a longitudinal analysis favored reinnervation for phantom pain. Targeted muscle reinnervation and the regenerative peripheral nerve interface now belong in the reconstructive conversation before any implant is considered. Mirror therapy, graded motor imagery, medication, rehabilitation, and prosthetic optimization remain foundational.

Among stimulation options, 60-day percutaneous peripheral nerve stimulation has support from a small randomized placebo-controlled amputation-pain trial with follow-up to twelve months, although the twelve-month outcomes were not a concurrent blinded comparison. DRG stimulation, SCS, MCS, and DBS have all been reported but rest largely on case series. The largest systematic review included 76 studies, only seven of them randomized trials, and concluded that the available evidence does not provide robust, reliable results. A focal residual-limb or nerve-territory phenotype is a more coherent target for peripheral or DRG stimulation than diffuse phantom pain. Intracranial stimulation remains a last option, not a default escalation.

7.Match generator, reversibility, and prognosis

Prefer a reversible trial when the generator is uncertain, the pain is distributed, or meaningful preserved function could be harmed by lesioning. Prefer a lesion only when the generator is anatomically specific, the expected benefit is syndrome-supported, and the tradeoff is acceptable. Prognosis matters: an ablative cancer-pain procedure may be rational when benefit is expected to outlast the patient, yet unacceptable for a young patient with decades of risk.

No-operation decisions are active care. They should close diagnostic loops, simplify medications, define rehabilitation goals, and direct the patient toward a longitudinal pain team rather than another device consultation.

8.Measure pain as a life problem

At minimum, record intensity, interference, allodynia, sleep, mood, analgesic burden, use of the limb or prosthesis, mobility, and one patient-selected activity. For DREZ, record paroxysmal and continuous components separately. For affective-circuit stimulation, include quality of life and distress. For every implant, report explant, revision, infection, loss of efficacy, and time to failure.