Poststroke pain (PSP) refers to a broader range of clinical conditions leading to pain after stroke, but not restricted to pain of central neuropathic nature (CPSP). PSP affects from 11 to 55% of patients following a CNS vascular event [8-10]. It includes several painful conditions such as CPSP, painful shoulder, lumbar and dorsal pain, complex regional pain syndrome, tension type headache and spasticity-related pain [10-12]. In particular, musculoskeletal pain is prevalent in PSP patients [13], being probably secondary to decreased muscle strength and altered descending modulatory system tonus [14]. Myofascial pain syndrome (MPS) is defined by the occurrence of regional pain and stiffness, limited range of motion in the affected muscle, satellite trigger points and twitch response to palpation of taut bands in the muscles [15,16].

MPS can also serve as a peripheral pain generator and alter the function of descending modulatory pathways as has been recently reported in other chronic pain conditions [17]. Besides the obvious implications of the presence of musculoskeletal pain in PSP on treatment, it may also pose diagnostic difficulties due to the presence of referred pain that commonly accompanies myofascial pain syndromes. Despite its recognition as part of the general PSP diagnostic possibilities, the prevalence of MPS has never been characterized in patients with CPSP. The identification of the co-occurrence of MPS with CPSP pain is of major importance for two reasons: first, using the current diagnostic criteria of neuropathic pain [18], MPS pain can be erroneously considered as neuropathic pain since it also refers to the presence of pain in an area of somatosensory system lesion. Second, and most important, it has been recently proposed that CPSP should be a diagnosis of exclusion, since there would be no pathognomonic feature of the syndrome [19], being reserved for patients with PSP and without other clear nociceptive or peripheral neuropathic pain syndromes. Although apparently sound, this proposal must be tested, since it is not known to which extent CPSP overlaps with other PSP syndromes such as MPS.

Inclusion criteria were adults (> 18years), presenting definite neuropathic pain with sensory deficits in the same topographic area related to a lesion to the somatosensory system [18]. Exclusion criteria were the presence of pain of exclusive nociceptive or peripheral neuropatic origin, the presence of major aphasia or other cognitive deficit impairing the report of the sensory abnormalities and pain characteristics.

Pain onset was insidious in 31 (77.5%) patients and presented during the first three months after stroke in 30 (75.0%) (Table 1). Pain was reported in the whole hemi-body in 22 (55.0%) patients, and had a multifocal distribution in the remainder.

Median pain intensity according to VAS was 10 (5 to 10) and the average pain duration period was 5.73 (±4.39) years. There was no difference in pain intensity among the different lesion site groups. Neuropathic pain was continuous-ongoing in 34 (85.0%) patients and intermittent in the remainder. Burning was the most common descriptor (70%), followed by electric shock-like paroxysms (22.5%). Nineteen (47.5%) patients had more than one descriptor for their neuropathic pain. Main aggravating factors were contact to cold (62.5%), mood swings (52.5%), movement of the painful limb (37.5%), and contact to heat (20%). There was more than one aggravating factor in 33 (82.5%) patients (Table 1). No statistically significant difference was found regarding the descriptors of pain, aggravating factors, and the MPQ scores among the lesion groups. Thermo-sensory abnormalities were universal in the series and are expressed on Table 2.

Painful shoulder syndrome was diagnosed in 4 (10.0%) patients and shoulder-hand syndrome in 1(2.5%). MPS was diagnosed in 27 (67.5%) patients and was more common in the supratentorial extra-thalamic group (P <0.001) (Table 3). It was more frequent in patients with pyramidal deficits (82.35%) than those without it (56.52%), however this difference was not statistically significant (P=0.017). The main muscles affected by MPS in CPSP patients were presented on Table 4. The spatial relationship between thermal deficits, pain and MPS pain area for each patient is illustrated in Figure 1. Autonomic abnormalities were found in 23 (57.5%) cases: Horner’s sign was observed in 6 (15.0%) cases, hypothermia of the upper and lower extremities in 10 (25.0%) patients, hyperemia in 9 (22.5%), edema in 5 (12.5%), hyperhydrosis in 2 (5.0%) and pallor in 2 (5.0%). The mean score of the BDQ was 22.87±11.96. No significant differences were observed among the different stroke location groups and pain quastionaires and scales scores. Importantly, CPSP patients with and without MPS did not differ in pain intensity (VAS), MPQ or BDS scores.

We performed a clinical-radiological evaluation of a group of CPSP patients focusing on the co-occurrence of PSP syndromes other than CPSP, in particular the role of MPS in these patients. CPSP was diagnosed according to the definite revised neuropathic pain criteria [18]. The presence of thermo-sensory deficits in all patients is a clinical hallmark of central pain, described in other series [4,26-28]. We looked for the presence MPS in a standardized fashion. MPS was present in the majority of cases (67.5%), suggesting that “pure” neuropathic pain syndrome is present in the minority of CPSP. Instead, most patients had mixed pain syndromes in which central neuropathic pain was associated with other syndromes (nociceptive). It has been recently proposed that CPSP should be a diagnosis of exclusion in PSP patients [19]. Our data suggest that this proposal might lead to under-diagnose of CPSP in PSP patients with associated MPS. A clear limitation of the design of our study is the lack of a control group. Evaluating the presence of MPS in a PSP population without CPSP would help us to better understand the factors influencing the occurrence of MPS in PSP patients. However, even if PSP patients without CPSP had a high prevalence of MPS, the prevalence of MPS in CPSP would still be quite high (as shown here), suggesting that for nomenclature and definition purposes CPSP should not be a diagnosis of exclusion. Certainly, this assumption must be confirmed by other larger studies, as well as the role of MPS in pain treatment and rehabilitation in these individuals. Furthermore, the identification of a central neuropathic element in a pain of musculoskeletal origin can be difficult and in some cases, several pain types might be present in the same area of the body [19], the findings of our study corroborate to this view. Motor deficits, spasticity, movements disorders and altered central descending pain modulation [14,29] may all induce overload to the muscles and trigger myofascial pain [2,4,5,8,10-12,27]. MPS in its turn can serve as peripheral generator of nociceptive inputs that may alter pain perception, as has been suggested to occur in other chronic pain conditions [17]. MPS was more frequent in supratentorial extra-thalamic (92.9%) and thalamic-capsular (100%) subgroups when compared with thalamic (50.0%) and brain stem (37.5%) stroke groups. This finding could be a consequence of the magnitude of the motor deficits and spasticity; however, we found no association on the presence of pyramidal deficits and the presence of MPS. Many patients in the supratentorial extra-thalamic group presented extensive brain lesions, usually secondary to occlusion of major arteries, resulting in major motor deficits, postural abnormalities, hypertonia and spasticity. This is further supported by the finding that 4 of the 8 patients in the thalamic group had an infarction located in the territory of the thalamic-geniculate artery, encompassing most of the sensory pathways while preserving motor cortical-spinal fibers [30,31].

To our best knowledge, the prevalence of MPS in patients with CPSP has not been described previously. Interestingly, the presence of MPS was not associated with more intense pain or pain associated mood disorders in CPSP patients, however, this can be due to some limitations of the study, such a small number of patients and a ceiling effect related to the high pain intensity of these highly refractory individuals. Also, we did not quantify the intensity of MPS, which could help better understand its relationship to pain intensity and disability. In our sample of CPSP individuals the mean pain duration period was high (5.73years), the pain scores, as well as the depression rates, were elevated, indicating a condition of high chronicity, psychosocial stress and refractoriness to treatment. The limited access of the stroke patients to a physical therapy program, performed in only five (12.5%) patients could have increased the incidence of MPS in this series. Moreover, a comparative analysis of the CPSP series with a control group was not performed. Still, our data suggest that MPS should be viewed as a common comorbid condition co-occurring with the CPSP syndrome complex. Similarly, we found painful shoulder in four (10.0%) patients and shoulder-hand syndrome in one (2.5%). Shoulder-hand syndrome is caused by glenohumeral joint subluxation due to motor paresis and is commonly associated with painful shoulder and is one of the possible presentations of the complex regional pain syndrome in PSP patients [12,32,33]. Autonomic abnormalities were found in 57.5% of our patients. Many CPSP patients present motor deficits and avoid movement of painful parts of the body. Prolonged immobilization may induce sensory, neurovegetative, motor and trophic abnormalities that can worsen pain and induce complex regional pain syndrome, a condition associated with neurovegetative abnormalities. [33,34] Autonomic dysfunction can also be related to the encephalic lesion such as lateral medullary stroke and Wallenbergs Syndrome in 4 (10.0%) patients.