Central pain states are common sequelae of spinal cord injury and stroke. Spinal cord injury pain has been classified based on the pain type (nociceptive, neuropathic), the pain subtype and location (musculoskeletal, visceral, above level, at level, or below the level of spinal injury), and on the specific structure or pathology, if known. The clinical presentation is variable and pain may be burning, aching, shooting, pricking, or tingling. The pathophysiology of central pain is complex and not well understood and may result from a combination of deafferentation-induced plastic changes in supraspinal regions and abnormal input from pain generators at the level of the spinal injury. Alterations in several neurotransmitters occur, including glutamate, GABA, and norepinephrine. Involvement of the spinothalamocortical pathway is strongly supported by animal models, but the precise pathway in humans is unknown. Preclinical research has helped identify a range of cellular and molecular mechanisms underlying neuronal hyperexcitability, which may serve as potential treatment targets for central pain states. The three components of central pain (steady dysesthetic, intermittent neuralgic, and evoked) must all be treated. In central pain of brain origin steady and evoked components predominate, while in central pain of spinal cord origin steady and neuralgic components predominate. The treatment needs to be individualized and a multidisciplinary approach is recommended. The treatment options include pharmacotherapy, physical therapy, cognitive behavioral therapy, neuromodulation, intrathecal drug therapy, and in some cases surgical management. Poorly controlled central pain carries a high suicide risk, hence psychosocial support is crucial.
Keywordscentral pain, central poststroke pain, complex regional pain syndrome, headache, spinal cord injury
Central pain is a term used to describe the pain associated with a wide range of disorders of the central nervous system (CNS). Whereas the disorders themselves are heterogeneous in nature, there is a great deal of overlap in the central mechanisms precipitating pain and the treatment algorithms. The International Association for the Study of Pain (IASP) defines central pain as “pain initiated or caused by a primary lesion or dysfunction of the CNS.” Classically studied disorders of central pain include poststroke, spinal cord injury (SCI), traumatic brain injury, and multiple sclerosis (MS). Central pain is often refractory to treatments, and complete pain relief is rare. In this chapter, we discuss the clinical presentations, pathophysiology, and therapeutic options for central pain of brain and spinal cord origin.
Epidemiology of Central Pain
The leading cause of central pain originating in the brain is stroke. Poststroke pain according to IASP has several pain subtypes with the most common types including central poststroke pain (CPSP), pain secondary to spasticity, shoulder pain, complex regional pain syndrome (CRPS), and headache. Except for MS patients, classically described central pain disorders (stroke, SCI, etc.) are more commonly seen in men. The elderly are more affected in the cases of poststroke pain, while SCI and MS pain tend to affect younger patients. Poststroke pain is generally thought to be underdiagnosed and undertreated. Recent reports suggest that the prevalence varies widely across studies, ranging from 10% to 55% of stroke victims.
In 1906, two French neurologists first described this poststroke “thalamic pain syndrome,” also known as the “Dejerine-Roussy syndrome” in their honor. This syndrome can account for one third of CPSP cases. The first postmortem studies of Dejerine-Roussy syndrome revealed that many of its victims had extrathalamic lesions, and modern imaging methods have confirmed and extended these findings. These pain-generating lesions extend from the first synapse of the dorsal horn, or trigeminal nuclei, to the cerebral cortex. The predominant etiology is vascular in origin, accounting for 90% of brain central pain (supratentorial 78% and infratentorial 12%) ( Fig. 30.1 ). Poststroke pain is associated more frequently with ischemic stroke than hemorrhagic stroke. Extrathalamic sites are involved in 50%–75% of cases. Chronic poststroke pain more commonly occurs in the presence of thalamic and brainstem lesions. In addition to stroke type and localization, other reported risk factors for development of CPSP include female sex, older age at stroke onset, alcohol use, peripheral vascular disease, statin use, depression, increased muscle tone and spasticity, reduced upper extremity movement, and sensory deficits.
SCI pain has many subtypes of central pain that have been organized under the International Spinal Cord Injury Pain (ISCIP) Classification with Tier 1 organized by pain type (nociceptive, neuropathic, and other pain [i.e., fibromyalgia]); Tier 2 organized by pain subtypes (musculoskeletal pain, visceral pain, other nociceptive pain, SCI-related pain, at-level SCI pain, below-level SCI pain, other neuropathic pain); and Tier 3 organized by primary source if known (i.e., cauda equine lesion, syringomyelia, spinal cord lesion, postthoracotomy pain, spasm-related pain, constipation, pressure ulcer). This new pain taxonomy has been incorporated into the International Spinal Cord Injury Pain Basic Data Sets (ISCIPBDS) for standardization in the collection and reporting of SCI pain worldwide. Central pain of spinal origin is predominantly the result of trauma (see Fig. 30.1 ). Pain can also result from spinal cord tumors and demyelinating lesions. The incidence is reported to vary from 34% to 94% in patients with SCI and about 29% in MS patients.
Central pain is also prevalent in patients with chronic degenerative diseases of the CNS. For example, almost 10% of patients with Parkinson disease may have sensory complications, including pain ; and epilepsy can manifest as painful seizures. Also, in contrast to most pathologic processes affecting the CNS, clinicians cannot predict the development of central pain based on the location of a lesion. Many central pain patients maintain their ability to sense touch, vibration, and joint movements. This supports the belief that central pain involves the spinothalamic tract (STT) and its thalamocortical projections. The highest prevalence of central pain is reported in cases of lesions in the spinal cord, medulla, and ventroposterior part of the thalamus.
The ISCIP Classification was developed from a task force for the IASP ( Table 30.1 ). As described earlier, SCI pain is broadly divided into nociceptive and neuropathic with subclassification into second and third tiers based on the anatomic structures involved, site of pain, and etiology. Nociceptive pain may be musculoskeletal or visceral in nature. The former may be secondary to overuse of certain parts of the body to compensate for regions of paresis or result from secondary changes in bone or joints. Neuropathic pain is usually seen in areas of sensory abnormalities. Neuropathic pain has been subdivided based on region into at-level (radicular or central), above-level, and below-level pain to indicate the presumed site of the lesion responsible for pain generation. Following SCI, it is reported that 91% of patients have pain 2 weeks after injury. This decreased to 64% at 6 months. Neuropathic at-level pain was present in 38% at 2 weeks and remained the same at 6 months. Neuropathic below-level pain occurred in 14% of subjects at 2 weeks and increased to 19% at 6 months. The prevalence and type of pain described following SCI over a 1-year period are shown in Fig. 30.2 . The pain can be spontaneous or stimulus evoked. Longitudinal studies indicate that at-level pain has an early onset, while the below-level pain develops months to years after the spinal injury.
|Broad Type (Tier One)||Broad System (Tier Two)||Specific Structures and Pathology (Tier Three)|
|Nociceptive||Musculoskeletal||Bone, joint, muscle trauma, or inflammation |
Secondary overuse syndromes
|Visceral||Renal calculus, bowel dysfunction, and sphincter dysfunction |
|Neuropathic||Above level||Compressive mononeuropathies |
Complex regional pain syndromes
|At level||Nerve root compression (including cauda equina) |
Spinal cord trauma/ischemia
Dual level cord and root trauma
|Below level||Spinal cord trauma/ischemia|
Central pain disorders are one of several types of “neuropathic pain,” and the definition and diagnosis of neuropathic pain have been discussed at length by researchers throughout the world. The definition outlined by the IASP was “pain caused by a lesion or disease of the somatosensory nervous system.” Some experts now believe that central neuropathic pain should be distinguished from peripheral neuropathic pain. A new grading system for neuropathic pain is shown in Table 30.2 . Although a number of questionnaires and standardized self-report measures have been developed to detect neuropathic pain, the described grading system requires a physical examination.
|Criteria to be evaluated for each patient|
c As part of the neurologic examination, these tests confirm the presence of negative or positive neurologic signs concordant with the distribution of pain. Clinical sensory examination may be supplemented by laboratory and objective tests to uncover subclinical abnormalities.
Central pain states likely result from pathophysiologic changes caused by irritation of, or damage to, central pain pathways. The possible pathophysiologic mechanisms that cause and maintain central pain are complex and not well understood (for reviews, see Finnerup ) Injury to the CNS may result in anatomic, neurochemical, inflammatory, and excitotoxic changes that result in a sensitized and hyperexcitable CNS.
Several neurotransmitters, such as glutamate, gamma-aminobutyric acid (GABA), norepinephrine, serotonin, histamine, and acetylcholine, are involved in the processing of noxious input along the pain pathway. The shift in firing from a rhythmic burst to a single spike is determined by noradrenergic, serotonergic, and cholinergic input to the reticular and relay cells of the thalamus. Similarly, excitatory amino acids, such as glutamate, are released in the region of SCI and may lead to neuronal hyperexcitability. At the spinal cord level, substance P and cholecystokinin (CCK) might play an additional role by influencing the voltage-gated sodium and calcium channels. Potassium channels play a critical role in setting the resting membrane potential and controlling the excitability of neurons.
Central pain in SCI may result from a combination of deafferentation-induced plastic changes in supraspinal areas along with abnormal input from a pain generator in the spinal cord ( Fig. 30.3 ). The changes in the CNS may include neuronal hyperexcitability. In SCI, N -methyl- d -aspartate receptor (NMDA) activation might trigger the intracellular cascade leading to the upregulation of neuronal activity/excitability that results in spontaneous and evoked neuronal hyperactivity/hyperexcitability and causes abnormal pain perception. Changes in voltage-sensitive sodium channels can also contribute to changes in nerve membrane excitability. Other important mechanisms might be a loss of endogenous inhibition, including reduced GABA-ergic, opioid, and monoaminergic inhibition. Wide dynamic range (WDR) neuronal hypersensitivity in excitotoxic or ischemic SCI models reveals changes similar to central sensitization following peripheral nerve injury. Analogous to epilepsy, SCI causes one neuronal population to generate hyperactivity and another to respond to this chaotic activity. It appears that a critical threshold in the size of this population must be reached before a patient will experience spontaneous pain.
Advances in functional imaging have increased our understanding of the changes in the brain that are associated with many pain conditions. Central glutamate levels are known to increase in response to pain in healthy humans, and patients with fibromyalgia are known to have elevated central glutamate levels that directly correlate with response to painful stimuli. Decreased opioid binding has also been demonstrated in the pain-processing regions of patients with poststroke pain. The authors go on to suggest that the imbalance of excitatory to inhibitory mechanisms explain, in part, the reason for central pain. Interestingly, gray matter decreased in patients with chronic pain. Whether the change in gray matter is premorbid or due to degeneration from an insult or inflammation from excitatory neurotransmitters (i.e., glutamate) is not known; however, a rat model of peripheral nerve injury was found to be associated with decreased size of the frontal cortex and comorbid anxiety. Patients with complete SCI are known to have a postinjury reorganization of the somatosensory cortex that correlates with pain intensity, a finding previously demonstrated in amputees.
Alterations of the sensory pathways and impaired descending inhibitory mechanism are associated with many pain conditions, including central pain. Craig and coworkers showed that, under normal conditions, the cool-sensitive pathway in the STT might suppress the forebrain’s response to nociceptive STT activity. Damage to this pathway may thus explain some of the phenomena seen in a central pain state. They hypothesize that, for central pain to occur, a lateral lamina I spinothalamocortical pathway lesion must be sufficiently large to produce contralateral sensory symptoms. This assumes that central pain is a release phenomenon resulting from the disruption of the normal integrative controls of sensory processing. The disruption of thermal sensibility results in a loss of the cold-induced inhibition of pain with a resultant disinhibition of cold-evoked burning pain. Craig and coworkers suggested that the ventromedial posterior (VMPo) nucleus of the thalamus plays a critical role. Investigations in primates, however, strongly support the existence of a spinothalamocortical pathway from lamina I and the deep layers of the dorsal horn to the contralateral ventral posterior lateral (VPL) nucleus, which extends to area 1 of the S1 somatosensory cortex. A similar pathway might activate neurons of the SII cortex, because direct projections connect the VPL and VPI (inferior) to SII and SI to SII.
Lesions in the spinothalamocortical pathways can cause ectopic discharges in various neurons of the spinal cord and brain. Such ectopic neuronal discharges create an illusion of noxious input because of the imbalance between the lateral (inhibitory) and medial (excitatory) STT. This might explain why pain occurs more often in patients with partial lesions than in those with complete cord and thalamic injuries. It appears that severe CNS lesions, with total destruction of ascending sensory systems, do not lead to a central pain syndrome and that mild, moderate, or severe disruption of the anterolateral ascending system, with partial or complete preservation of the dorsal column/medial lemniscus functions, is most frequently associated with central pain syndrome. Furthermore, even during remission, dysesthesias and pain could be triggered by additional afferent input to the large fiber/dorsal column/medial lemniscus system and, once established, might not be abolished by additional deafferentation.
Sensory stimuli act on neural systems that have been modified by previous inputs, and the “memory” of which significantly influences pain behavior. The fact that a memory is not activated by the development of a lesion might explain the long delay in the onset of central pain in some patients. The long-term potentiation that is important for this memory might be mediated by NMDA receptors and their influence on calcium conductance. Thus, central pain frequently develops weeks or months after development of the lesion and is associated with sensory changes involving the spinothalamic pathways, especially changes in temperature perception.
The role of microglia in central pain is an area of great interest. Microglia are the macrophages of the brain and spinal cord that release inflammatory mediators in the event of injury or infection. The activation of microglia and subsequent inflammation is thought to precipitate a cycle of further inflammation and activation of astrocytes. Therapies targeting the cytokines, chemokines, and inflammatory mediators involved in this microglial and astrocytic activation are being increasingly explored.
The neuropathic component of central pain is often reported starting days to weeks after the CNS lesion and presents as a steady dysesthesia or neuralgia and may also have an evoked component. Although many tools and tests have been proposed to diagnose central pain, the wide range of types and presentations of central pain disorders makes no single measure completely sensitive or specific. The quality of the pain may be burning, aching, shooting, pricking, and tingling. The discomfort is generally constant but may wax and wane and often has a deep and/or a superficial component. In a minority of patients, the pain is intermittent and daily. Nonpainful tactile, thermal, vibratory, auditory, visual, olfactory, and visceral stimuli can provoke or exacerbate spontaneous pain. Anxiety and/or fear can also exacerbate symptoms. Some patients with central pain exhibit the most striking symptoms seen in clinical practice. Patients with classic Dejerine-Roussy syndrome have a rapidly regressing hemiparesis and a sensory deficit to touch, temperature, and pain. Allodynia, hyperalgesia, and spontaneous severe paroxysmal pain on the hemiparetic side also often occur. These patients can exhibit hemiataxia, hemiastereognosia, and choreoathetoid movements. Patients with central pain may have any or all these features depending on the location of the underlying lesion. Organic signs on sensory examination of patients with thalamic lesions include the so-called thalamic midline split for sensory loss and pain. The fact that central pain of any cause is accompanied by delayed hyperalgesia supports the hypothesis of a polysynaptic response. Pain intensity for brainstem and suprathalamic lesions are moderate in intensity, averaging 61 and 50 mm, respectively (on a 100-mm visual analog scale), while pain in thalamic lesions can be severe (average = 79 mm). Many standardized measures have been developed to assess the different aspects of neuropathic pain to develop specific treatment algorithms and further research efforts.
Patients with a history of spontaneous or evoked dysesthesia, hyperesthesia, or paresthesia should undergo specific but simple bedside testing, which may help distinguish different phenotypes of central neuropathic pain. A prospective study that utilized the ISCIPBDS framework showed that early hypersensitivity by sensory testing was a predictor for later onset of below-level (not at-level) central neuropathic pain. Sensory testing in the region where the pain is localized usually shows a paradoxic hypoalgesia (decreased sensitivity to painful stimulus). The region where the patient feels the pain often has decreased sensitivity to thermal stimuli, especially to cold. In fact, the intensity of the pain seems to be related to the magnitude of loss of thermal sensation. Testing for disturbed temperature sensation can be accomplished with a cold metal instrument, ice, or ethyl chloride spray. Touch can be tested with cotton wool, while pinprick sensation should be assessed using the contralateral side as a control. Chronic poststroke pain patients have an intact vibration sensation. Patients may exhibit mitempfindung (with sympathy), a phenomenon in which stimulation in one area of the body results in a simultaneous sense of the provoked sensation in another part of the body. These patients may also experience alloesthesia, in which a sensory stimulus on one side of the body is perceived on the other side. A subset of patients who experience burning pain lose the sensation of cold, warmth, and sharpness. In another subgroup of patients who experience shooting/pricking/aching pain, tactile allodynia is predominant. Although some disturbance of sensory function is almost always present on physical examination, clinical findings are few or subtle in many patients. Quantitative sensory testing might reveal side-to-side asymmetries in cooling, warmth, and heat-pain sensation thresholds.
Testing for autonomic dysfunction may be important in patients with SCI. Lesions above the sixth thoracic level (splanchnic outflow) are often associated with autonomic dysreflexia. The dysreflexia is characterized by sudden dramatic increases in blood pressure, high or low heart rate, and headache after sensory input, such as a full bladder. This point becomes especially important in SCI patients having surgery below the level of their lesion, including minor operations of the urinary (i.e., cystoscopy) or gastrointestinal (i.e., colonoscopy) systems where the viscera will be stimulated. Even though patients may lack sensation in the area to which they are having surgery, intense stimulation can precipitate major hemodynamic instability. Complications may include seizures and cerebral hemorrhage.
Experimental Models of Central Pain Secondary to Spinal Cord Injury
Interesting insights about the mechanisms of central pain following SCI and the potential effects of drugs on pain behavior have been gained from experimental models in rats. The Stockholm group led by Wiesenfeld-Hallin developed a model of photochemically induced spinal cord ischemia, while Yezierski et al. developed a model of excitotoxic SCI. Rats with lesions involving both white and gray matter develop instantaneous morphine-resistant tactile allodynia, which responds to the systemic GABA-B agonist baclofen, and this can be prevented by pretreatment with the NMDA antagonist, MK-801. Intrathecal morphine and clonidine reduced the allodynia. Injections of a CCK-B antagonist decreased allodynia.