European Journal of Neurology 2004, 11 (Suppl. 1): 22–30 Post-stroke pain case study: clinical characteristics, therapeutic options and long-term follow-up P. Hansson Neurogenic Pain Unit Multidisciplinary Pain Center, Department of Rehabilitation Medicine, Karolinska Hospital, Stockholm, Sweden Keywords: anticonvulsants, antidepressants, central post-stroke pain, chronic pain, comorbidities, neuropathic pain Central post-stroke pain (CPSP) is a syndrome characterized by sensory disturbances and neuropathic pain. In 40%)60% of CPSP patients, the onset of central pain following a stroke occurs more than 1 month after the stroke, creating a source of diagnostic uncertainty or significant delay in treatment since healthcare providers familiar with CPSP may no longer be caring for the patient when the symptoms occur. In addition to chronic pain, the presence of somatosensory abnormalities is the most important diagnostic corollary of CPSP. Neuropathic or central pain has been estimated to occur in up to 8% of patients after a stroke, and about 18% of stroke patients with a somatosensory disturbance will develop CPSP. Although largely a matter of conjecture, it is generally agreed that damage to spinothalamic sensory pathways plays a significant role in the pathogenesis of CPSP. A comprehensive examination of the patient for sensory deficits is essential before treatment can be initiated. Functional disturbances such as depression, anxiety and sleep disturbances are significant comorbid conditions associated with CPSP; the physician should incorporate an assessment of these potential comorbidities into the examination. Treatment options for CPSP are limited; at present, amitriptyline is the drug of first choice. Other drugs including antidepressants, anticonvulsants, antiarrhythmics, opioids and N-methyl-D-aspartate antagonists may provide relief for some patients who do not respond to amitriptyline. Included in this review is a case study outlining the challenges of managing the patient with CPSP. Case presentation Approximately 12 years ago, TF, a 40-year-old single male, suffered a right-side brain stem infarct, with signs and symptoms of a partial Wallenberg syndrome. Computerized tomography and magnetic resonance imaging results failed to confirm a specific site of infarct. A painful condition confined to the left side developed after 1 week. Symptoms included numbness in the right side of the face and in the left thoracic and abdominal regions (Fig. 1). He also experienced numbness and tingling in the left arm as well as numbness, continuous, deep aching pain, and mechanically evoked allodynia in the left leg, the latter of which was experienced as burning pain following contact with clothing, bed sheets and the wind. The pain was continuous, independent of movement, and varied in intensity from 2 to 7 on a 10-cm visual analog scale. The pain was also worsened by a cold climate and improved by a warm one. A repeat examination at 3 months revealed no motor involvement, but a Correspondence: Per Hansson, Neurogenic Pain Unit Multidisciplinary Pain Center, Department of Rehabilitation Medicine, Karolinska Hospital, Stockholm, Sweden 17176 (tel.: +46 851 775435; fax: +46 851 776641; e-mail: per.hansson@kirurgi.ki.se). Ó 2004 EFNS dysmetria of the right arm. A spectrum of complex sensory abnormalities was still present at the 3-month evaluation, consisting of hypoesthesia to warmth, cold and pinprick, tactile allodynia, and spontaneous pain in the left leg. We diagnosed this individual as having central post-stroke pain (CPSP) after considering the onset, pattern and continuous nature of his pain symptoms relative to stroke symptoms in conjunction with signs of somatosensory dysfunction in the painful areas. CPSP: clinical setting Pain is a well known although often overlooked consequence of stroke (Leijon et al., 1989; Andersen et al., 1995; Segatore, 1996). The source of the pain may be either peripheral or central in origin (Leijon et al., 1989; Widar et al., 2002). A common source of pain after stroke is nociceptive pain localized to the shoulder and resulting from paresis and changed muscular tone and posture (Leijon et al., 1989). A small proportion of stroke patients experience neuropathic pain, which is a direct result of damage to the nervous system itself (Leijon et al., 1989). Dejerine and Roussy’s description of Ôthalamic painÕ is most often cited as the seminal report in the medical literature describing pain in 22 23 Post-stroke pain Number of patients 15 10 5 12 24 36 Months Figure 2 Time to onset of pain in CPSP. n ¼ 27. (Boivie, 1999). Figure 1 Distribution of symptoms. association with the symptoms of stroke. However, the term Ôthalamic syndromeÕ or Ôthalamic painÕ is insufficiently broad to describe pain following a stroke, since lesions at any level of the central nervous system that affect pain signaling may result in neuropathic pain (Bowsher, 1999). The term Ôcentral post-stroke painÕ (CPSP) is therefore more appropriate to describe neuropathic pain following a cerebrovascular accident (Andersen et al., 1995; Segatore, 1996). CPSP is a syndrome characterized by sensory disturbances and neuropathic pain. Similar pain syndromes may accompany trauma to the brain, disease of the brain or spinal cord, syringomyelia/bulbia or multiple sclerosis (Bowsher, 1999). The onset of central pain following a stroke occurs more than 1 month after the stroke in 40%)60% of all patients (Andersen et al., 1995; Bowsher, 1999; Widar et al., 2002). Figure 2 shows the wide range of time to onset of central pain in a small patient cohort. Thus, a common source of diagnostic uncertainty is that symptoms of CPSP regularly occur after a significant passage of time from the precipitating event. In fact, neurologists and others familiar with CPSP may no longer be caring for these patients when their symptoms occur, resulting in misdiagnosis or significant delay before treatment is begun (Bowsher, 1999). Ó 2004 EFNS European Journal of Neurology 11 (Suppl. 1), 22–30 The most important diagnostic corollary of CPSP, in addition to the pain in a stroke-affected body region, is the occurrence of negative and/or positive symptoms and signs in the same body region (Bowsher, 1999). In 1906, Dejerine and Roussy pointed out that sensory disturbances were among the most prominent symptoms in patients with infarcts of the thalamus (Boivie et al., 1989; Andersen et al., 1995). Leijon et al. (1989), in a more recent landmark study of CPSP, confirmed that somatosensory abnormalities are indeed the only unifying characteristic in these patients, along with their unexplained chronic pain. The somatosensory abnormalities associated with CPSP are far from uniform among patients (Bowsher, 1999). Clinical recognition may therefore be facilitated by a more detailed knowledge of the various clinical presentations (see Clinical Features section below). Epidemiology Since most prior studies of CPSP have been based on selected case material, the incidence of CPSP in an unselected stroke population has been difficult to determine precisely (Andersen et al., 1995). Neuropathic or central pain has been estimated to occur in up to 8% of patients after a stroke during a 1-year follow-up (Andersen et al., 1995). About 18% of stroke patients with a somatosensory disturbance, such as changes in sensibility to touch, temperature and pain, go on to develop CPSP (Andersen et al., 1995). In an earlier study of the clinical and physiologic characteristics of central pain, the authors noted that the median age of the stroke patients they studied was 57, suggesting that there may be a significant age difference between CPSP patients and the general stroke population (median age 75) (Bowsher, 1996). However, a more recent prospective study of the evolution of CPSP in a general stroke population found no age difference when comparing CPSP patients to stroke patients who did not develop pain 24 P. Hansson (Andersen et al., 1995). The issue of median age in CPSP patients is subject to ongoing debate. Pathophysiology CPSP is clearly within the category of neuropathic pain syndromes since it is initiated by a lesion to the central nervous system. However, a century after its first description by Dejerine and Roussy, the pathogenesis of CPSP is still largely a matter of conjecture and hypothesis. Several theories have been offered in the recent clinical literature (Boivie, 1999). These include release of inhibition by a lesion of the medial lemniscal pathway, a specific lesion of the spinothalamic pathway that causes a loss of cold-related inhibition of burning pain, and abnormal bursting activity in the thalamus (Boivie, 1999). The results obtained by Boivie et al. (1989) show that some CPSP patients lack signs of a lesion in the medial lemniscal pathway, not supporting a unifying disinhibition hypothesis involving the thalamus. It is generally agreed that damage to the spinothalamiocortical sensory pathways plays a significant role in the pathogenesis of CPSP. Any general hypothesis must therefore explain how a precipitating lesion anywhere along the spinothalamic and thalamocortical pathways for pain signal processing (including spinal lesions, lesions in the thalamus or cortical lesions) may be responsible for the symptoms of CPSP (Boivie et al., 1989). Diagnostic procedures This patient had suffered a right-side brain stem infarct and presented with post-stroke sequelae that developed within 1 week, including spontaneous and tactileevoked pain in his left leg. Burning and aching pain such as this patient described is experienced by many patients with CPSP, as is pricking and lacerating pain (Leijon et al., 1989). Furthermore, patients with CPSP frequently note an increase in pain when exposed to cold, as did this patient (Widar et al., 2002). A comprehensive bedside examination, including probing of somatosensory functions, should be performed. The latter should be explored using an array of instruments, including a camel hair brush or cotton for touch sensations, a cold and warm metallic roller for temperature sensations, and a pin for pain sensations (Hansson, 2002). The use of these different instruments will provide information about important sensory pathways (Verdugo and Ochoa, 1992; Hansson et al., 2001). Frequently observed in patients with CPSP are altered sensation with hypophenomena or hyperphenomena to noxious and non-noxious temperature and mechanical stimuli (Boivie et al., 1989). In order to make a diagnosis of neuropathic pain, it is important to fully delineate the neuroanatomical distribution of positive and negative symptoms and signs. A pain drawing (Fig. 1) that has been filled out by the patient can help delineate the neuroanatomical distribution of spontaneous pain (Hansson et al., 2001). When examining the patient for sensory deficits, after an initial test in an unaffected area, it is recommended to begin testing in the affected area and then move outward until skin that responds normally to stimuli is found. Conversely, when examining the patient for positive sensory signs such as mechanical or thermal allodynia, begin testing in the unaffected area and move into the painful area. This approach will minimize the patient’s exposure to painful stimuli (Hansson et al., 2001). Comorbid conditions and functional disturbances associated with CPSP Although there are no formal studies of the interaction of mood and pain state in CPSP, the experience of chronic pain can lead to depression, anxiety and sleep disturbance. As such, inquiries as to the length and quality of sleep as well as the patient’s mood should be made. CPSP: treatment approaches Once a diagnosis is established, the next clinical challenge is to select an effective treatment. Indeed, clinical trials regarding the treatment of CPSP are generally small, and there are no extensive clinical data to support any of the treatments discussed. Until further research is conducted for existing therapies, or until new agents are developed, the limited options presented below represent our best treatment modalities. Treatment options There is a general consensus that amitriptyline is the drug of first choice, but not all patients report a response. If a response is obtained, intolerable side-effects may prevent continued treatment. Other drugs may provide relief for at least some patients who do not respond to amitriptyline. However, there is no practical information at present to guide initial treatment choice based on presenting pain symptoms or other factors. If initial therapy fails, as it often does, there are as yet no guidelines to rationally direct the clinician stepwise through the alternatives. The following sequence may stand as general guidance, but other reasons may suggest starting at a different point and proceeding in a different order. Ó 2004 EFNS European Journal of Neurology 11 (Suppl. 1), 22–30 Post-stroke pain Antidepressants Tricyclic antidepressants (TCAs) have become an established first choice in many neuropathic pain syndromes. Amitriptyline, in a comparison trial versus carbamazepine, was the first TCA to be tested in CPSP, and it was the first agent of any type to demonstrate efficacy in a placebo-controlled trial (Table 1) (Leijon and Boivie, 1989; Bowsher, 1999). In general, amitriptyline was well tolerated, with only one patient discontinuing treatment (Leijon and Boivie, 1989). Benefit was noted regardless of lesion type ) thalamic, suprathalamic or brain stem ) and appeared to be related to the attainment of a threshold plasma concentration (Leijon and Boivie, 1989). On the basis of this trial and clinical experience, amitriptyline is considered the drug of first choice. The utility of amitriptyline is limited by its well-known and sometimes serious side-effects, including sedation, urinary retention, orthostatic hypotension and cardiac arrhythmia. Antidepressants with a better side-effect profile (i.e. serotonin-specific re-uptake inhibitor category) appear to be significantly less effective in CPSP, but there are no published clinical studies that confirm this lack of benefit (Bowsher, 1999). Anticonvulsants A variety of anticonvulsant agents have been employed in the treatment of CPSP, but other than the trial of carbamazepine versus amitriptyline cited above, there have been few controlled clinical studies until quite recently. Although Leijon and Boivie (1989) found carbamazepine to be clearly less effective than amitriptyline, some patients did benefit from carbamazepine, and more may have done so had higher plasma levels been attained. As with the TCAs, the use of carbamazepine is limited by its side-effect profile and by the need for very gradual dose escalation, especially in the elderly (Leijon and Boivie, 1989). Lamotrigine monotherapy at 200 mg/day has been compared to placebo in a population of 27 CPSP patients and found to be moderately effective (median pain score reduced from 7 to 5 out of 10); 12 patients (44%) experienced a pain score reduction of at least 2 out of 10 and were considered responders (Vestergaard et al., 2001). Side-effects were relatively mild, and except for the occurrence of mild rash (one or two instances), did not limit completion of treatment. However, the potential side-effects of lamotrigine, such as Stevens–Johnsons/Lyells syndrome, call into question its general usefulness in neuropathic pain. A recent case report on the efficacy of gabapentin in a patient with highly resistant symptoms suggests the need for a larger trial of this agent in CPSP, especially since it has proved to be efficacious and well tolerated in Ó 2004 EFNS European Journal of Neurology 11 (Suppl. 1), 22–30 25 other neuropathic pain syndromes (Chen et al., 2002). Given its relative tolerability, safety and lack of interaction with other drugs, a trial of gabapentin in individual patients seems worthwhile in those who have failed to respond to more recognized treatments. Antiarrhythmics A small, double-blind, placebo-controlled trial in 16 patients with CPSP (n ¼ 6) or spinal cord injury (n ¼ 10) examined the effect of intravenous lidocaine (5 mg/kg over 30 min) on both spontaneous ongoing and evoked pains. Lidocaine was found to be significantly superior to placebo in reducing the intensity of spontaneous ongoing pain for up to 45 min postinjection. Side-effects were moderate, consisting mainly of lightheadedness. Lidocaine was also superior to placebo for reducing brush-induced allodynia and mechanical hyperalgesia, but not thermal allodynia and hyperalgesia (Attal et al., 2000). Another study of intravenous lidocaine (5 mg/kg/h for 60–90 min) in 111 patients with chronic, non-malignant pain found that, of the patient population with pain of central nervous system origin (n ¼ 13), only one patient experienced ÔexcellentÕ pain relief (nine patients reported no relief and three patients reported partial relief) (Galer et al., 1993). One small series of nine patients with post-stroke pain showed a response rate of 89% to oral mexiletine, the oral congener of lidocaine, in patients who had previously failed treatment with antidepressants, anticonvulsants or surgery (Awerbuch and Sandyk, 1990), but in this author’s experience mexiletine has proven ineffective in CPSP. Opioids Administration of morphine and other opioids to unselected populations of patients with CPSP and related central pain syndromes has repeatedly yielded disappointing results and given rise to controversy, if not bias, with regard to their use in CPSP (Bowsher, 1995; Attal et al., 2002). In a small study of patients with central pain, Attal et al. (2002) showed that there is a small segment of the CPSP patient population that responds well to long-term oral morphine and that these patients can be identified based on their response to intravenous morphine 60 min after the start of an infusion. There was no significant effect of morphine in reducing ongoing pain when all 15 patients with either post-stroke pain or spinal cord injury were considered as a group. However, seven patients (46%) did respond to intravenous morphine and their pain was improved after 1 month of oral morphine therapy. Side-effects were prevalent among those who continued oral morphine treatment beyond 1 month, and only three patients continued oral morphine after 1 year. Finally, 26 P. Hansson a non-randomized, single-blind, placebo-controlled study of 14 patients with central pain caused by multiple sclerosis found that four patients were responsive to morphine therapy (>50% pain reduction at a mean dose of 41 mg), supporting the findings that central pain is poorly responsive but not totally unresponsive to opioids (Kalman et al., 2002). N-methyl-D-aspartate (NMDA) antagonists A case report presented a patient with CPSP unresponsive to several other therapies who experienced significant improvement after an intravenous trial of ketamine (Vick and Lamer, 2001). This individual patient was subsequently maintained on specially compounded oral ketamine (50 mg three times per day) plus oral diazepam (5 mg three times per day) to allay the dysphoria associated with ketamine administration. Other long-term pain medicines were subsequently withdrawn and the patient has been maintained for at least 9 months with reduced levels of pain and increased functionality (Vick and Lamer, 2001). The problem of a narrow therapeutic window offsets the use of ketamine in the clinical routine setting. Motor cortex stimulation In 1991, Tsubokawa et al. (1991) demonstrated that electrical stimulation of the motor cortex resulted in a reduction of pain in patients with CPSP. Subsequent results of the use of motor cortex stimulation to alleviate pain in patients with central pain have varied, with responses in several small studies ranging upto 100% (Hosobuchi, 1993). Katayama et al. (1998) reported excellent or good pain control in almost 50% of patients with CPSP using motor cortex stimulation, and Placebo (n ¼ 15) A. Mean daily rating of pain Week 1 of trial 5.5 ± 1.5 Week 2 of trial 5.4 ± 1.6 Week 3 of trial 5.3 ± 1.9 Week 4 of trial 5.3 ± 2.0 B. Global assessment of change in pain Improved 1 Unchanged or worsened 14 C. Depression score at end of treatment period Mean 2.6 Range 0–6 Nguyen et al. (1999) obtained a 77% response rate when the positioning of the stimulating electrodes was confirmed using somatosensory evoked potentials. At present, the technique may be recommended in exceptional cases of CPSP where everything else has failed and should be performed only in high profile neurosurgical pain management centres. Optimizing treatment A first step in optimizing treatment would seem to be the prompt trial of amitriptyline, since more clinical experience has been gained with this agent than any other. The choice of a second drug, and whether it is to be used in conjunction with or instead of the TCA, depends on whether any clinical benefit was obtained and the tolerability of any side-effects that may have resulted from the TCA. If there is no benefit or if sideeffects are troublesome, moving on to a separate and distinct drug category, such as the anticonvulsants, seems appropriate. Unfortunately, there is much less evidence upon which to base the selection of a second drug to be substituted for the TCA. So far, correlations between symptom complexes and response to treatment, or between lesion type and response to treatment, have failed to yield any universally applicable rules for selecting among a wide variety of treatment options. Case resolution TF was given low-dose carbamazepine (the drug of choice for many neurologists in 1990) but experienced intolerable side-effects and no pain relief. Amitriptyline was tried early with significant pain relief, but due to Amitriptyline (n ¼ 15) Carbamazepine (n ¼ 14) 4.7 ± 1.3 4.3 ± 1.7a 4.4 ± 1.3b 4.2 ± 1.6b 4.6 ± 1.2 4.4 ± 1.5 4.4 ± 1.4a 4.2 ± 1.7 10b 5b 5 9 2.2 0–8 3.0 0–7 Table 1 Placebo-controlled comparison of amitriptyline and carbamazepine in CPSP a P < 0.01; bP < 0.05 compared to placebo. Results based on 4-week randomized cross-over design of amitriptyline (75 mg/day by day 6), carbamazepine (800 mg/day by day 18) and placebo (double-dummy) in patients with unequivocal stroke episode; seeking remedy for pain that started after the stroke; with pain not related to nociceptive, peripheral neuropathic or psychogenic cause. Adapted with permission from Leijon and Boivie (1989). Ó 2004 EFNS European Journal of Neurology 11 (Suppl. 1), 22–30 Post-stroke pain side-effects (see below) other drugs were tried. Vigabatrin, dextromethorphan and gabapentin all failed to provide pain relief, even at maximum doses. The possible use of opioids was discussed with the patient, but he decided against such a trial. Finally, amitriptyline was re-titrated to 75–100 mg/day (25 mg initially with a 25 mg increase every 7–10 days) and was found to eliminate the deep aching leg pain while reducing the dynamic mechanical allodynia by about 50%. This patient experiences sedation, dizziness and difficulty concentrating at an amitriptyline dose of 75 mg and, accordingly, takes this drug only during particularly painful episodes. TF receives a pension and works halftime as a product manager at a biomedical company. His involvement in golfing has been reduced to the role of referee, since the continuing tactile allodynia prevents him from playing golf. Amitriptyline, while providing pain relief, did not improve any of the patient’s quality of life parameters, in large part because the drug is sedating. CPSP ) clinical features Clinical advances in the treatment of CPSP will ultimately depend on gaining a sound understanding of its etiology and pathogenesis. Three studies published in the last 10–12 years have described larger samples of patients than previously reported (Tables 2–4). The main difference in the study designs is the manner of patient selection. In the study by Leijon et al. (Table 2), patients were referred to a pain clinic or selfselected to attend a pain clinic based on a desire to find relief from their symptoms. The other two studies acquired patients consecutively according to protocol, prospectively in the case of Andersen et al. and retrospectively in the case of Widar et al. (Tables 3 and 4, respectively). The studies differ also in the depth of the reporting on each population sample, that of Leijon et al. being the most exhaustive. Where similar data were collected, however, results of a single study are generally supportive of the others, and no findings of any one study stand in direct contradiction to the findings of the others. One consistent finding of all three studies is that patients with CPSP do not necessarily have a lesion of the thalamus as originally proposed by Dejerine and Roussy (although the lesions affect the thalamus indirectly) (Leijon et al., 1989). In fact, all three studies show that lesions of the thalamus may be found in fewer than one-third of their respective samples (12%)33%). Pain descriptors are also challenging as a guide to identifying patients with CPSP. For instance, a burning Ó 2004 EFNS European Journal of Neurology 11 (Suppl. 1), 22–30 27 Summary and conclusions A pressing issue in CPSP is obtaining clarity of diagnosis. While in many cases symptoms appear promptly after a stroke and are clearly associated with a central cause for pain, substantial numbers of patients do not develop symptoms for many months. This displacement in time together with a general lack of uniformity of symptoms and signs at presentation may delay diagnosis considerably. At present, amitriptyline is the first-line treatment for CPSP. However, not all patients respond to amitriptyline and some may experience intolerable side-effects. There is little consensus regarding second-line therapies, and the choice of treatment is largely empirical if TCAs fail (e.g. anticonvulsants, opioids or, in exceptional cases, motor cortex stimulation). Nevertheless, it is important to work through the few treatment alternatives promptly while giving each an adequate trial. quality appears prominent among the list of terms used by each set of patients to describe their pain; however, it was not the most significant term in all studies and far from a majority of patients described their pain in this fashion. Onset of pain symptoms occurred within a month of stroke in more than half of the patients in each study, and a large majority of these patients experienced pain symptoms within 6 months of a stroke. Patients developing central pain longer than a year after a stroke may simply represent extreme examples in a continuum, or they may have specific types of central nervous system damage that take longer to evolve. Regardless, these Ôslow progressorsÕ represent a diagnostic challenge and, once treatment strategies become apparent, may require different therapeutic approaches than patients whose pain appears more promptly. One particularly striking observation made by Leijon et al. is that all patients experiencing CPSP in their sample also evidenced disturbances of temperature sensibility. This finding was well supported by Andersen et al. and to a lesser degree by Widar et al. who used a less elaborate clinical evaluation scheme. The near universal presence of altered temperature sensibility in CPSP is both a useful diagnostic cue and a strong observation in support of a unified mechanistic hypothesis. Namely, any of a great variety of stroke lesions affecting input to the spino-thalamocortical tract can result in cell death and/or synaptic reorganization, and thereby induce symptoms of CPSP. 28 P. Hansson Table 2 Neurological symptoms and signs, including pain characteristics in CPSP (Leijon et al., 1989) Summary findings d Pain is usually well-located and severe with daily fluctuation and has a variety of qualities. The most common pain quality is burning. It is equally common for pain to be deep or superficial d The only common feature in all CPSP patients is a disturbance of non-noxious and/or noxious temperature sensibility; only about half have abnormalities of touch and vibration d In most cases, stroke lesions are extrathalamic Study population Location of stroke n ¼ 27 (20 male; 7 female) Brain stem 8 Evidence of unequivocal stroke Thalamus 9 Seeking remedy for central post-stroke pain Supratentorial/extrathalamic 6 that was not nociceptive, peripheral Not located 4 neurogenic or psychogenic in origin Pain quality Symptom onset Burning 59% Immediate 15% Aching 30% Within 1st month 37% Pricking 30% 1–3 months 26% Lacerating 26% 5–12 months 11% Shooting 11% 24–34 months 11% Squeezing 11% Throbbing 11% Other 19% Note: Most patients reported more than one kind of pain Somatosensory analysis d Decreased sensibility to innoxious temperatures 100% d No temperature perception between 0 and 50 °C 70% d Decreased sensitivity to noxious cold or heat 93% d Abnormal pinprick sensitivity 96% d Abnormal cold sensitivity 93% d Raised threshold to touch 52% d Raised threshold to vibration 41% d Spontaneous or evoked paresthesia or dysesthesia 89% Table 3 Incidence of CPSP – prospective study (Andersen et al., 1995) Summary findings d Overall incidence of CPSP was 8% (16 of 207) in a consecutive series of stroke patients enrolled in the course of a single year d All CPSP patients except one had dysesthesia or allodynia (induced thermally or mechanically) on sensory testing (94%; 15 of 16) d In most cases of CPSP, stroke lesions were extrathalamic (75%) d Significantly more abnormal sensibility to cold and warm stimuli was observed in patients with CPSP than non-pain patients with somatosensory dysfunction Study population Location of stroke n ¼ 207 (66% male; 36% female) Thalamus 4 (25%) CPSP diagnosed in 16 (8%) Extrathalamic 12 (75%) Consecutive stroke patients admitted in 1 year Examined within 1 week of stroke and again at 1, 6 and 12 months Pain quality Symptom onset Lacerating 50% Within 1st month Aching 25% 1–6 months Burning 19% >6 months Freezing 19% Squeezing 19% Other 13% Note: Most patients reported more than one kind of pain descriptor 63% 19% 19% Ó 2004 EFNS European Journal of Neurology 11 (Suppl. 1), 22–30 Post-stroke pain Table 4 Long-term pain conditions after a stroke (Widar et al., 2002) 29 Summary findings d Pain is usually well-located, fluctuates in intensity and has a variety of qualities. It is reported to be worsened by a cold environment. The most common pain qualities are stabbing, aching, dull and burning d In most cases of CPSP, stroke lesions are extrathalamic Study population Location of stroke n ¼ 43 (30 male; 13 female) Brain stem 5 Evidence of unequivocal stroke Thalamus 5 Pain lasting > 6 months post-stroke Supratentorial/extrathalamic 27 with no other pain condition Not located 6 Pain quality Symptom onset Stabbing, aching, dull and burning Within 1st week 33% Within 1st month 20% 2–6 months 47% Note: Most patients reported more than one kind of pain descriptor Central pain appears in many disguises and therefore requires a meticulous diagnostic work-up. In addition, central pain, regardless of etiology, is notoriously difficult to treat with available strategies and only a small number of patients report significant but usually partial relief. In order to facilitate the development of treatment options we need further insights into pathophysiologic mechanisms of CPSP. References Andersen G, Vestergaard K, Ingeman-Nielsen M, Jensen TS (1995). Incidence of central post-stroke pain. Pain 61:187– 193. Attal N, Gaude V, Brasseur L et al. (2000). Intravenous lidocaine in central pain: a double-blind, placebocontrolled, psychophysical study. Neurology 54:564–574. Attal N, Guirimand F, Brasseur L, Gaude V, Chauvin M, Bouhassira D (2002). Effects of IV morphine in central pain: a randomized placebo-controlled study. Neurology 58:554–563. Awerbuch GI, Sandyk R (1990). Mexiletine for thalamic pain syndrome. Int J Neurosci 55:129–133. Boivie J (1999). Central pain. In: Melzack R, Wall PD, eds. Textbook of Pain, 4th edn. Churchill Livingstone, Edinburgh, pp. 879–914. Boivie J, Leijon G, Johansson I (1989). Central post-stroke pain – a study of the mechanisms through analyses of the sensory abnormalities. Pain 37:173–185. Bowsher D (1995). The management of central post-stroke pain. Postgrad Med J 71:598–604. Bowsher D (1996). Central pain: clinical and physiological characteristics. J Neurol, Neurosurg, Psychiatry 61:62–69. Bowsher D (1999). Central post-stroke (Ôthalamic syndromeÕ) and other central pains. Am J Hospice Palliative Care 16:593–597. Chen B, Stitik TP, Foye PM, Nadler SF, DeLisa JA (2002). Central post-stroke pain syndrome: yet another use for gabapentin? Am J Phys Med Rehabilitation 81:718–720. Ó 2004 EFNS European Journal of Neurology 11 (Suppl. 1), 22–30 Therefore, in-depth characterization of symptoms and signs in relation to treatment outcomes with available strategies must continue with the aim of trying to identify predictors for choice of treatment. The lack of an appropriate animal model of CPSP certainly inhibits a more rapid progress within this area. New treatment strategies are badly needed. Galer BS, Miller KV, Rowbotham MC (1993). Response to intravenous lidocaine infusion differs based on clinical diagnosis and site of nervous system injury. Neurology 43:1233–1235. Hansson P (2002). Neuropathic pain: clinical characteristics and diagnostic workup. Europ J Pain 6:47–50. Hansson P, Lacerenza M, Marchettini P (2001). Aspects of clinical and experimental neuropathic pain: the clinical perspective. In: Hansson PT, Fields HL, Hill RG, Marchettini P, eds. Progress in Pain Research and Management, Vol. 21: Neuropathic Pain: Pathophysiology and Treatment. IASP Press, Seattle, pp. 1–18. Hosobuchi Y (1993). Motor cortical stimulation for control of central deafferentation pain. Adv Neurol 63:215–217. Kalman S, Osterberg A, Sorensen J, Boivie J, Bertler A (2002). Morphine responsiveness in a group of well-defined multiple sclerosis patients: a study with i.v. morphine. Europ J Pain 6:69–80. Katayama Y, Fukaya C, Yamamoto T (1998). Poststroke pain control by chronic motor cortex stimulation: neurological characteristics predicting a favorable response. J Neurosurg 89:585–591. Leijon G, Boivie J (1989). Central post-stroke pain – a controlled trial of amitriptyline and carbamazepine. Pain 36:27–36. Leijon G, Boivie J, Johansson I (1989). Central post-stroke pain – neurological symptoms and pain characteristics. Pain 36:13–25. Meyerson BA, Lindblom U, Linderoth B, Lind G, Herregodts P (1993). Motor cortex stimulation as treatment of trige- 30 P. Hansson minal neuropathic pain. Acta Neurochirurgica (Wien) 58:150–153. Nguyen JP, Lefaucheur JP, Decq P et al. (1999). Chronic motor cortex stimulation in the treatment of central and neuropathic pain. Correlations between clinical, electrophysiological and anatomical data. Pain 82: 245–251. Segatore M (1996). Understanding central post-stroke pain. J Neurosci Nursing 28:28–35. Tsubokawa T, Katayama Y, Yamamoto T, Hirayama T, Koyama S (1991). Chronic motor cortex stimulation for the treatment of central pain. Acta Neurochirurgica (Wien) 52:137–139. Verdugo R, Ochoa JL (1992). Quantitative somatosensory thermotest. A key method for functional evaluation of small calibre afferent channels. Brain 115:893–913. Vestergaard K, Andersen G, Gottrup H, Kristensen BT, Jensen TS (2001). Lamotrigine for central poststroke pain: a randomized controlled trial. Neurology 56:184–190. Vick PG, Lamer TJ (2001). Treatment of central post-stroke pain with oral ketamine. Pain 92:311–313. Widar M, Samuelsson L, Karlsson-Tivenius S, Ahlstrom G (2002). Long-term pain conditions after a stroke. J Rehabilitation Med 34:165–170. Ó 2004 EFNS European Journal of Neurology 11 (Suppl. 1), 22–30