Central Pain States
Nanna Brix Finnerup
Sharona Ben-Haim
Central neuropathic pain or central pain is defined as “pain caused by a lesion or disease of the central somatosensory nervous system.”1 Central pain is well described in stroke, multiple sclerosis (MS), and spinal cord injury (SCI), with the latter involving both traumatic and disease-related injuries. Other conditions associated with central pain include brain trauma, brain tumors, and possibly pain in epilepsy and Parkinson’s disease. Central pain often becomes chronic and may be disabling with a negative impact on quality of life, mood, sleep, and functioning. Treatment of central pain remains challenging. This chapter describes the diagnosis, clinical characteristics, assessment, mechanisms, and treatment of central pain.
Diagnosis
A grading system for defining the level of certainty as to how likely a given pain condition is neuropathic in nature was first published in 20082 and updated in 2016.3 It follows the classical clinical diagnostic methods used in neurology, in that history, clinical examination, and diagnostic test consecutively add to the level of diagnostic certainty.3 The grading system adapted for central pain is illustrated in Figure 28.1.
Central pain is suspected when the history suggests that the pain is related to a central nervous system (CNS) disorder and not to other causes such as spasms, fractures, inflammation, etc. Different pain questionnaires have been developed as screening tools using clusters of descriptors to identify the presence of neuropathic pain. These include the Neuropathic Pain Questionnaire (NPQ),4 the Leeds Assessment of Neuropathic Symptoms and Signs (LANSS),5 the Douleur Neuropathique en 4 Questions (DN4) questionnaire,6 the painDETECT questionnaire,7 the ID-Pain questionnaire,8 and the Spinal Cord Injury Pain Instrument (SCIPI) developed for patients with SCI.9 These screening tools may be useful in epidemiologic research, but they cannot be used alone for identifying central pain in the individual patient.10,11
The first level of the grading system is possible central pain (see Fig. 28.1). Two criteria need to be fulfilled for possible neuropathic pain: (1) There should be a history of a relevant CNS disorder and pain development at or after the onset of the CNS disorder. Following stroke, SCI, and other acute onset CNS disorders, the pain typically develops within months after the onset, but the onset of central pain may be delayed up to about 1 year after the incident.11,12,13 (2) The pain distribution should be neuroanatomically plausible, which means that the pain should be distributed within the area of the body that is affected by the CNS disorder—either the complete affected area or a smaller or larger proportion.
The next level is probable central pain (see Fig. 28.1). This involves a clinical sensory examination and confirmation of sensory abnormalities in the same neuroanatomical location. Demonstration of sensory loss to one or more sensory modalities such as touch, pinprick, cold, and warmth compatible with the CNS lesion is essential. Positive sensory signs (e.g., dynamic or cold allodynia) should also be compatible with the CNS lesion. Sensory abnormalities are often present on clinical examination, but a more thorough examination using quantitative sensory testing may be necessary.3,14
The final level is definite central pain (see Fig. 28.1). This requires a diagnostic test that confirms the CNS lesion. This often includes a computed tomography (CT) or magnetic resonance imaging (MRI) scan of the brain in patients with stroke, MS, SCI, or other CNS lesion. Heat- or laser-evoked potentials can also be used to demonstrate a lesion or disease of the spinothalamic tract (STT) pathways, and trigeminal reflex recordings may be useful in the diagnosis of secondary trigeminal neuralgia in MS. Reaching the level of definite central pain using these positive criteria in the grading system means that a CNS lesion can explain the pain but does not determine the cause of the pain.3 This poses particular problems in CNS lesions where sensory abnormalities are common regardless of the presence of pain. Therefore, the exclusion of other causes of pain is critical when diagnosing central pain.
The exclusion of other types of pain is challenging given the lack of specific identifiers for distinguishing neuropathic pain from other pain types. Also, many patients experience more than one type of pain, and sometimes, different pain types occur in the same body location making diagnosis even more difficult.15,16,17 A number of differential diagnoses should be considered when diagnosing central pain, including musculoskeletal pain, which is very common following CNS lesions
due to overuse, myofascial shoulder pain, stress fractures, spasticity, dystonia, etc., visceral pain, and peripheral neuropathic pain.16,18,19,20,21,22
due to overuse, myofascial shoulder pain, stress fractures, spasticity, dystonia, etc., visceral pain, and peripheral neuropathic pain.16,18,19,20,21,22
Clinical Characteristics
Central pain is characterized by spontaneous and/or evoked pain in an area with partial or complete sensory loss to one or more modalities. Decreased sensation to thermal or painful stimuli is present in most patients with central pain, whereas decreased sensation to other modalities is less common.23,24,25,26 The location of pain and sensory abnormalities is within an area compatible with the CNS lesion (Fig. 28.2). Spontaneous pain may be ongoing, intermittent, or paroxysmal. Studies in different central pain conditions show that pain descriptors such as hot/burning, shooting, pricking, pins and needles, cold, sharp, squeezing, and aching are common.9,10,11,21,25,27,28,29,30,31,32,33,34,35 Evoked pain may be present as allodynia, which is pain due to a stimulus that does not normally provoke pain, and hyperalgesia, which is increased pain from a stimulus that normally provokes pain.36,37 There may also be aftersensations, which is pain continuing after the stimulation has ceased. Cold allodynia and touch-evoked allodynia are particularly common in central pain.24,27,31,32,34,38 Hyperpathia is an abnormal—often explosive—painful reaction to a stimulus in an area with increased sensory threshold when the stimulus exceeds the threshold. Due to aftersensations, the distinction between spontaneous and evoked pain may be difficult. Spontaneous pain may also be generated by central sensitization mechanisms by which decreased thresholds and temporal summation cause ongoing pain from stimuli common in daily life (e.g., movement, breathing, ambient temperature).39 Nonpainful abnormal sensations are also common.11,31,35,40 These include paresthesia, which are evoked or spontaneous abnormal sensations that are not unpleasant, and dysesthesia, which are unpleasant abnormal sensations.36 Such sensations are described in terms of tingling, numb, cold, pressing, or warm.
Clinical Assessment
In the assessment of central pain, it is relevant to evaluate the intensity, impact, treatment, quality, and temporal aspects of pain as well as physical and emotional functioning and quality of life (Table 28.1).41 The pain intensity of average and worst pain may be evaluated using a numerical rating scale, a visual analog scale (VAS), or a verbal rating scale.42,43,44 Different scales have been developed to assess the multidimensional aspects of pain. One such scale is the Multidimensional Pain Inventory (MPI), which assesses pain and the impact of pain on physical and emotional functioning.45 It has been adapted for use in SCI (MPI-SCI).46 Other scales may be used to assess mood, sleep, resilience, pain catastrophizing, disability, and satisfaction with life.44
The quality of spontaneous and evoked pain may be assessed using open-ended questions or by providing the patients with a list of typical pain descriptors, for example, by using specific neuropathic pain scales such as the Neuropathic Pain Symptom Inventory (NPSI),47 the NPQ,4 and the painDETECT questionnaire.7
A sensory examination is important to identify sensory loss and gain by assessing decreased or increased sensation in the affected area compared with an area not affected by the CNS lesion. Simple bedside tests include assessment of static and dynamic touch using a brush or cotton ball; assessment of pinprick sensation using a stick, pin, or monofilament; assessment of vibration sense with a 128-Hz tuning fork; and assessment of cold and warm sensation using thermo rollers or cold and warm metal or glass objects (see Table 28.1).14,37 Allodynia and hyperalgesia can be quantified by using a numeric rating scale
for evoked pain intensity. Mapping of sensory abnormalities and pain is a crucial part of the diagnosis of central pain to ensure that the distribution is compatible with a CNS lesion (see Fig. 28.2). Detailed quantitative sensory testing can supplement bedside sensory examination, but it is time-consuming and mainly used for research purposes.14
for evoked pain intensity. Mapping of sensory abnormalities and pain is a crucial part of the diagnosis of central pain to ensure that the distribution is compatible with a CNS lesion (see Fig. 28.2). Detailed quantitative sensory testing can supplement bedside sensory examination, but it is time-consuming and mainly used for research purposes.14
TABLE 28.1 Clinical Assessment of Central Pain | ||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
Laboratory tests can be used to confirm the diagnosis of a CNS lesion and specific abnormalities of the pain pathways, including the use of CT and MRI scans. Trigeminal reflexes are used in the diagnosis of trigeminal pain, and contact heat- and laser-evoked potentials are useful for assessing the function of the spinothalamic tract.14
Specific Central Pain Conditions
CENTRAL POSTSTROKE PAIN
Central poststroke pain (CPSP) is pain arising after a cerebrovascular lesion including ischemic and hemorrhagic lesions of the brainstem, thalamus, operculum-insula, and cerebral cortex.48,49 Central pain occurs in 3% to 8% of stroke patients11,18,28,35,50 and is particularly common in patients with infarctions in the lateral medullary infarctions, where the prevalence may be as high as 25%, and in the thalamus, where the prevalence is reported to be around 17% to 18%.12,51,52
CPSP is experienced contralateral to the side of the stroke. It may affect small areas such as the hand or the whole hemibody, which is common in thalamic infarctions.23,27,35,51 In patients with lateral medullary infarctions, the pain distribution may be crossed, involving one side of the body contralateral to the lesion and the other side of the face ipsilateral to the lesion, and periorbital pain is commonly described.12,53
CENTRAL PAIN IN MULTIPLE SCLEROSIS
MS is a gradually progressing demyelinating disease with multiple sites of demyelination in the CNS. MS is an autoimmune disorder triggered by environmental factors in patients with a genetic predisposition. The inflammatory attacks give rise to plaques in the white matter of the brain and spinal cord and as the disease advances, the myelin loss is followed by damage to the axons and neurons with associated grey matter loss.54,55 Central pain is present in about 25% of patients with MS.56
The distribution of central pain is compatible with spinal or brain lesions of the somatosensory nervous system.57,58 It may be described as burning, tingling, pricking, and squeezing pain, sometimes associated with evoked pain.34,56 In addition to this central pain type that is similar to central pain in other conditions, patients with MS may experience two specific types of central pain conditions, namely, secondary trigeminal neuralgia and Lhermitte’s phenomenon.56,59,60,61,62 In patients with trigeminal neuralgia, the MRI scan should document a lesion in the ipsilateral side of the pons along the cause of trigeminal afferents or the recordings of trigeminal reflexes and trigeminal-evoked responses should show increased latency.62 The pain is similar to classical trigeminal neuralgia with brief, severe, paroxysmal pain attacks restricted to one or more division of the trigeminal nerve. It is often unilateral but may be bilateral. The attacks may arise spontaneously or be triggered by movement or mechanical stimuli.61,62 Lhermitte’s phenomenon is an electric shock-like sensation that is transient and short lasting. It is typically provoked by neck movement and is felt spreading down in the back.59,61 It is more common in younger patients and in patients with a progressive or progressive-relapsing cause and is assumed to be caused by ectopic discharges due to demyelination in the dorsal column.60,61,63
CENTRAL PAIN IN SPINAL CORD INJURY
SCI is a term that includes both traumatic and nontraumatic causes. The most common traumatic causes are traffic accidents and falls, and nontraumatic causes include syringomyelia, tumors, ischemia, hemorrhage, arteriovenous malformations, transverse myelitis, and infections.64 Chronic pain is one of the most disabling consequences of an SCI—present in 70% to 80% of patients.13,65,66,67,68 Neuropathic pain is present in about 50% of patients.13,66,69 The onset of central pain may be immediately after the SCI or it may be delayed up to 1 year. A delayed onset and facial pain should alert the physician to the possibility of development of syringomyelia.70 Neuropathic pain following SCI is divided into pain felt at and below the level of lesion.16,66 At-level pain is located within the dermatome of the neurologic level of injury and three dermatomes below the neurologic level. This type of neuropathic pain may be caused by the SCI or a lesion of the nerve roots, and it is often not possible to determine whether it is a peripheral neuropathic pain caused by root lesion or a central pain caused by spinal cord lesion. Below-level pain is pain felt more than three dermatomes below the neurologic level and is a central pain caused by the spinal cord lesion. In patients with syringomyelia, the pain distribution is often segmental.71,72 Different types of pain seen following SCI, including central pain, can be found in Chapter 40, and the readers are referred to that chapter for further details on SCI pain.
OTHER CENTRAL PAIN CONDITIONS
The diagnosis of central pain in brain trauma is often difficult and only few studies exist. The central pain is similar to that of other central pain patients.73,74 The onset may be delayed for several months and often becomes persistent. The pain distribution is often unilateral, corresponding to the side with most severe sensory abnormalities and related to decreased thermal sensitivity. Ongoing pain described as pricking, cold, pressing, wretched, and burning is common, and some patients described touch- and cold-evoked allodynia.73 There are very few reports in the literature on central pain in brain tumors, but it has been described in tumors affecting the thalamus and the parietal cortex.75,76 Central pain following surgery such as dorsal root entry zone lesions, thalamic destructions, and mesencephalic and medullary tractotomies are only infrequently reported because these procedures are rarely done today.77
Pain is a common problem in Parkinson’s disease. The pain is often located in the lower back and lower extremities, and musculoskeletal pain is the most common type.78,79 It is also suggested that some pain conditions in Parkinson’s disease are related to the CNS disease and has been termed central pain.78,79,80 This type of pain is rare and is a poorly localized diffuse pain that is often associated with “off condition,” autonomic symptoms, visceral pain, and improvement with levodopa administration.78,79,81 This type of pain, however, does not entirely fulfill the criteria for
central pain. Even though altered thermal and pain thresholds have been found in Parkinson’s disease, suggesting altered responsiveness to painful stimuli,81,82 lower laser N2/P2 amplitudes are also found in pain-free patients,83 and one study found that the pain thresholds were not different in patients with and without pain.84 However, there is increasing evidence that the basal ganglia and the dopaminergic system play an important role in the gating of nociceptive information and pain modulation,85,86 and it is possible that some pain types in Parkinson’s disease are related to altered sensory processing of nociceptive inputs.80 Epileptic seizures may rarely be reported as pain.87,88 One case report describes a patient with a dysplasia in the posterior insula.89 Spontaneous high-frequency activity and spiking from this area propagated to other areas including the parietal operculum and the midcingulate gyrus. During the attack, the patients reported hemibody pain that started in the left hand and foot.89
central pain. Even though altered thermal and pain thresholds have been found in Parkinson’s disease, suggesting altered responsiveness to painful stimuli,81,82 lower laser N2/P2 amplitudes are also found in pain-free patients,83 and one study found that the pain thresholds were not different in patients with and without pain.84 However, there is increasing evidence that the basal ganglia and the dopaminergic system play an important role in the gating of nociceptive information and pain modulation,85,86 and it is possible that some pain types in Parkinson’s disease are related to altered sensory processing of nociceptive inputs.80 Epileptic seizures may rarely be reported as pain.87,88 One case report describes a patient with a dysplasia in the posterior insula.89 Spontaneous high-frequency activity and spiking from this area propagated to other areas including the parietal operculum and the midcingulate gyrus. During the attack, the patients reported hemibody pain that started in the left hand and foot.89
Preclinical Models
Different models of SCI, stroke, brain trauma, and MS are used in preclinical pain research.90,91,92,93,94,95,96 SCI models include hemisection, contusion, ischemic, electrolytic, and excitotoxic models.97,98,99,100 The models mimic human CNS lesions reasonably well, but the assessment of pain and its translation into human central pain remains a challenge.101
One of the most obvious challenges for all preclinical models of pain is the difficulty in assessing the correlate of ongoing pain. Except for a very few studies that have used conditioned place preference paradigms102 or overgrooming,103 most preclinical studies of central pain exclusively use evoked responses to mechanical, cold, or warm stimuli. This may be valid because evoked pain is common and predicts the development of central pain,31,38,104 but evoked pain-related responses do not always correlate with ongoing pain, and ongoing central pain may exist without evoked pain. A particular challenge in assessing pain-like behavior in models of CNS injury is the frequent coexistence of spasticity making the use of simple withdrawal reflexes unreliable as a pain-like behavioral outcome.101,105,106,107,108 Because withdrawal reflexes are spinally mediated, persist after complete spinal transection, and may be increased as part of the spastic syndrome, methods that depend on brainstem or cortical processing (e.g., operant escape or place escape/avoidance paradigms) are needed.109
Mechanisms
Experimental animal models and studies in humans have provided insights into the pathophysiology of central pain, which involves plasticity in various areas and pathways of the CNS. Sensitization of the CNS plays a major role in central pain and involves neuronal hyperexcitability resulting in an increased response to synaptic inputs, decreased threshold, expansion of receptive fields, and access of low-threshold Aβ mechanoreceptors to pain pathways.110,111 The clinical consequences are allodynia, hyperalgesia, and aftersensations. Central sensitization may also include spontaneous discharges.112 Spontaneous discharges in damaged pain pathways or deafferented rostral neurons may be responsible for spontaneous ongoing or intermittent pain.113 Seemingly, spontaneous pain may also be a result of decreased thresholds in nociceptor excitation and temporal summation of stimulus-evoked pain occurring at physiologic levels from stimuli caused by daily activity such as breathing, movement, and ambient temperature.39 Recent studies have found that sensory hypersensitivity (mechanical and cold allodynia, hyperalgesia, and temporal summation of pain) precedes and predicts later development of central pain (CPSP and below-level SCI pain), supporting a role of neuronal hyperexcitability also for ongoing pain.31,38,104
The functional changes underlying central sensitization involve a range of anatomical, neurochemical, excitotoxic, and inflammatory mechanisms.114,115,116 Changes in ion channels (sodium, calcium, potassium, acid-sensing) and phosphorylation of glutamate receptors may occur at various levels of the CNS,117,118,119,120,121 and altered opioid receptor binding has also been demonstrated in patients with central pain.122,123 Changes in the functions of microglia and astrocytes and release of inflammatory mediators also contribute to chronic pain following CNS injury.90,119,124,125 Loss of inhibition, either through the altered balance of descending inhibitory and facilitatory pathways, loss of interneurons containing glycine or γ-aminobutyric acid (GABA), or via decreased GABAergic inhibitory function through downregulation of the potassium chloride exporter, may be additional mechanisms underlying central sensitization.126,127
Lesions of the STT and STT-thalamocortical pathways play an important but not well-understood role in the mechanisms of central pain. Impairment of pain and temperature sensation and STT injury are hallmarks of central pain,48,128,129 but abnormal STT function is not a sufficient condition as STT lesions are also frequent in pain-free subjects. Approximately 50% of patients with a lesion of the STT after SCI or operculo-insular strokes develop central pain,130 and the question is why STT lesions cause pain in some but not all individuals. Although central pain can occur in patients with complete lesions of the STT,131 it is suggested that central pain is caused by sensitization of partially preserved residual STT neurons in some patients.125,132,133,134 This is consistent with studies showing that microsimulation in the ventral caudal (Vc) sensory nucleus of the thalamus, which receives dense STT terminations, evokes pain in patients with central pain.135,136 It is likely that the role of lesioned versus preserved ascending pathways in central pain depends on the pain phenotype. Patients with evoked pain are more likely to have preserved large- and small-fiber function than patients with spontaneous pain only, as shown in studies using quantitative sensory testing, somatosensory, and laser-evoked potentials and functional MRI (fMRI).26,137,138,139