top of page
  • robinapark

Central post-stroke pain. Need to know.

Updated: Sep 12, 2021

Definitions/Diagnostic criteria


CPSP is a diagnosis of exclusion

Onset in the first days following the stroke is rare, 2/3rds of patients will have pain within the first month and nearly all occur within a year

Severe pain on the contralateral side of the body to lesioned hemisphere

Associated sensory changes in the painful region.

In lateral medullary stroke pain can be located in areas of sensory loss which include the ipsilateral face and contralateral body.

Pain can be spontaneous or evoked.

Spontaneous pain in some cases is episodic, whereas in others it may be continuous.

Spontaneous pain is burning (47% to 59%) or aching (30% to 41%), often deep within the painful body area.

Less commonly may be lacerating pain (7% to 26%) or pricking pain (6% to 30%).

When pain is evoked, it can be caused by a nociceptive source that induces pain out of proportion to the stimulus (hyperpathia).

59% of patients had hyperalgesia in the painful region on pinprick testing, 37% had hypoalgesia, and only 4% had a normal pinprick sensation.

Patients with stroke lesions in the thalamus are more likely to report that pinprick sensation is hyperalgesic

Pain can also be evoked by non-nociceptive sources (allodynia), but the actual clinical response to different nonpainful stimuli can vary.

2/3rds have allodynia on examination. These patients will experience pain with limb movement (70%), although some find that movement relieves the pain (19%).

Thermal and tactile stimulation can evoke pain or reduce it as well.

Nearly half of patients will have pain with cold stimuli (48%), but a few find that cold reduces their discomfort (7%).

A warm stimulus was found to increase pain in 22% and reduce it in another 30%. - Frequently, patients will find that emotions increase pain (19%), and 37% achieve relief of pain with rest.

Just over half (52%) of patients with CPSP have no hemiplegia, and only 37% have moderate hemiplegia. Severe hemiplegia is less commonly associated (11%). On the other hand, ataxia is present in 62%, but choreoathetosis is found in only about 4%.

Epidemiology (Treister, Hatch, Cramer & Chang 2017)

Pain from any cause is common after stroke in 14% to 43% in acute and subacute stages

Central post-stroke pain (CPSP) occurs in only about 8% of patients & 9% after thalamic hemorrhage.

Only 5% of these have moderate to severe pain

More frequently musculoskeletal processes, including hemiplegic shoulder pain, arthritis, myofascial disorders, and tendonitis.


Pain originates from the stroke lesion itself. It induces neuroplastic changes that result in disordered central nervous system (CNS) processing within the distributed sensory network

- CPSP can result from a lesion anywhere along the spinothalamic and thalamocortical pathways

- However, it must be recognized that pain does not always result from these lesions

- Currently, it is recognized that cortical, subcortical, thalamic, and lateral brainstem strokes can all cause CPSP, in addition to sensory changes and impaired motor control and weakness, but pain is not always a necessary outcome.

- Pathophysiology of CPSP is unclear but likely maladaptive neuroplastic changes within the CNS that result in aberrant sensory perception.

- Occurs more commonly with ischaemic than haemorrhagic strokes

- CPSP can be considered a kind of “deafferentation phenomenon” because lesions are located within the spinothalamocortical pathway and probably resulting in maladaptive alterations in synaptic facilitation and inhibition.

- Less likely would be a type of “release phenomenon” in which lesions of the lemniscal pathway carrying nonpainful sensation leave the pain pathways and unbalanced pain signals are transmitted to the cortical centers.

- Theories depend on several concepts:

First is that injury to the spinothalamocortical tracts is necessary;

Second, that the pain is a result of late neuroplastic changes within CNS

Third, that pain perception is linked closely to the emotional state of the individual and the behavioral drive that signals a homeostatic imbalance.




No single medication is completely effective. It is likely that CPSP has multiple physiologic causes associated with neuropharmacologic changes at multiple locations.

In any one patient, multiple pain-inducing mechanisms may also be at play. This not only complicates our understanding of the etiology of CPSP but also complicates management as well.

- Non-pharmacological

Physical conditioning, modification of functional task performance, supportive counseling, and behavior modification.

TENS has minimal evidence but can be trialled

- Pharmacological

Pharmacology is not used in isolation, especially since the clinical response to treatment with medications is literally “fifty-fifty.” That is, at best only about half the patients will derive any benefit from the medications prescribed, and those who do will have only about 50% relief of their symptoms.

Medications useful in managing CPSP fall into five categories:

Membrane stabilizers

Aminergic agents

Glutamate antagonists

GABA agonists

N-type calcium channel blockers

When medications are used, it is recommended that one class be tried at a time and if one class fails, a medication from a different class be used for the next trial.

Carbamazepine = modest, but not statistically significant, effect in reducing pain scores in 36% of patients with CPSP versus a 7% reduction during the placebo phase.

Amitriptyline, given at a maximum dose of 75 mg daily, amitriptyline resulted in a reduction in pain scores in 67% of patients versus 7% during the placebo phase ( P < 0.05) and outperformed carbamazepine in comparison.

Duloxetine, at doses of 60 to 120 mg daily in patients with either stroke- or spinal cord–related pain, there was a trend toward improved pain scores on a visual analog scale, but a significant reduction in brush ( P = 0.035) and cold ( P < 0.001) allodynia.

CHECK - Intravenous ketamine, 48% of subjects with CPSP experienced an average 40% reduction in pain, with the effect lasting less than an hour after the infusion ended.

Pregabalin and gabapentin are agents worthy of trial from this class, but they are clinically often limited by side effects in patients with post-stroke pain.

- Interventional

Deep brain stimulation, motor cortex stimulation - both have mixed evidence.


Golden pearls

References / Articles / Resources

  1. Practical Management of Pain, Fifth Edition

  2. Treister, A. K., Hatch, M. N., Cramer, S. C., & Chang, E. Y. (2017). Demystifying Poststroke Pain: From Etiology to Treatment. PM & R : the journal of injury, function, and rehabilitation, 9(1), 63–75.

52 views0 comments

Recent Posts

See All


bottom of page