Monday, March 27, 2017

CRPS: Integrating Concepts of Neuroinflammation Into Integrated Pain Management

Peter A. Moskovitz, MD (left) and Marc S. Cooper, PhD (right)

Peter A. Moskovitz, MD (left) and Marc S. Cooper, PhD (right)

By Debra Nelson-Hogan

Peter Moskovitz, MD, and Mark S. Cooper, PhD, presented a provocative plenary session called Deconstructing Hysteria: Integrating Concepts of Neuroinflammation into Integrated Pain Management. The presentation was a well-crafted conversation between a physician and a neuroscientist on how neuroinflammation may underlie the unusual sensory and motor dysfunctions of Complex Regional Pain Syndrome (CRPS). The speakers also addressed whether medically unexplained symptoms, often diagnosed as hysteria, conversion reaction, or somatoform disorder, are also connected to neuroinflammatory impairments.

Cooper briefly outlined the history of hysteria beginning with Dr. Jean-Martin Charcot and Dr. Sigmund Freud, two neurologists from the late 1800s. Charcot was considered to be one of the founders of modern neurology, and Freud the founder of psychoanalysis.  Early neurological concepts of hysteria were largely constructed at the Salpêtrière insane asylum, where Charcot had his practice. The presumed psychological etiology of hysteria, which developed through the work of Freud and Janet, later became known as psychosomatic illness. These etiological concepts evolved to what we today call somatoform disorders or conversion disorders.

Moskovitz pointed out that Freud’s genius was the realization that psychology is biologically based. If emotions happen in the body (as Walt Whitman told us 150 years ago), physical impairments can be caused by emotions and repressed desires. This is the theoretical interpretation of medically unexplained symptoms, which are often viewed as hysteria, conversion reaction, or somatoform disorders. Moskovitz and Cooper believe that “emotions can influence, in part, the onset and course of physical impairments, but the physical impairments themselves are not literally caused by emotional disorder.” Instead, they believe that stress and injury can lead to cryptic neuroinflammation within the central nervous system (CNS). Individuals with neuroinflammation in the CNS can have altered psycho-physiological couplings, which can produce atypical symptoms, such as unusual linkages between emotional states and sensory-motor processing.

This putative neuroinflammatory etiology creates a differential diagnostic conflict for practitioners who mistakenly think that “If the symptoms and signs don’t fit a specific disease, then, intuitively, the cause must be psychological,” Moskovitz said. This thought process commonly occurs when a physician is unfamiliar with the symptoms and signs of CRPS, and does not consider CNS neuroinflammation as part of the differential diagnosis.

“Everything about CRPS is atypical. You have ’non-anatomic pain,’ ’non-anatomic swelling,’ ’non-physiological sensory disturbances,’ hyperalgesia, allodynia, hyperpathia, ’non-neurological‘ motor impairment. You have the dystonic postures – which don’t make sense from a lesional neurologist’s point of view. And universally, people with CRPS are anxious, and they’re depressed. These patients are frightened to death of their disease, and they are deeply disturbed by it,” Moskovitz said.

Cooper gave several examples of how neuroinflammatory lesions in the nervous system could produce clinical signs that could be easily misdiagnosed as hysterical or psychogenic. He took the audience back to the Salpêtrière for a case study of Blanche Wittman, who had a hypersensitive spot on her abdomen just above her ovary. Cooper noted that “Charcot would press on that trigger point, and the woman’s response was dramatic: a ’hysteria-type‘ reaction where dystonia began to spread through her body.”   He then showed a video of a man in Australia who had the same type reaction and explained that these responses are atypical of people experiencing neuropathic pain, but they do happen.  In fact, Cooper described having experienced a dystonic spasm himself, after a cervical radiculopathy at C8. “One night I rolled on to that irritated nerve and I went into a full body spasm. I lost my ability to vocalize. My intercostals and limbs were locking up. It was part of the whole-body dystonic spasm. I entered a tube of pain. It was a sensory experience that I’ve never experienced, before or since. As I came out of the spasm, I had a dissociative experience. My right arm was completely gone. I could not sense the arm, or move it,” he explained.

Cooper added that because his dystonic spasm was associated with a peripheral nerve injury, his medical diagnosis was far different than it might have been.   Had he been a woman in Paris in the 1880s, of limited means, having these dystonic spasms and sensory disturbances on a recurring basis, most likely he would have been committed to the Salpêtrière asylum as having hysteria. Today, he might have been diagnosed as having a conversion disorder. Cooper believes that pressure on his inflamed cervical nerve triggered a seizure-like event in his spinal cord or brainstem, producing symptoms that have been traditionally viewed as hysteria.

Cooper described published case studies that showed “exaggerated startle responses, hyperekplexia, trismus, the tightening of the jaws. The presumed etiology was brainstem encephalopathy (brainstem inflammation).” In one case, a person was bitten on the right leg by an insect, motor symptoms developed in the leg, then turned into exaggerated startle reactions and body spasms, which were triggered by auditory, visual, or tactile stimuli (1).” Neurologists are now thinking about inflammation in the brainstem as an etiology for unusual neurological symptoms, but mechanistically we have to consider why neuroinflammation in this region would be so powerful.

Cooper also spoke of Harvard researcher Anne Louise Oaklander, MD, PhD, who has developed a rodent model for these postures (2). She passes a needle through a given branch of the sciatic nerve. Lesions through the tibial nerve or the sural nerve result in a fixed dystonia posture in the rodent’s leg, lasting for about two weeks.

“Other studies show that peripheral nerve injuries not only evoke local inflammation at the site of the nerve injury, but they also activate neuroinflammation at the first-order synapse in the dorsal horn,” he continued. “You see the activation of resident immune ― neuroimmune cells ― called microglia. These cells change morphology. They aggregate and they begin self defense and repair programs that alter the neurophysiology of the neurons.”

Cooper noted that several studies show that neuroinflammation can spread great distances through axons to second-order synapses. An injury, whether it is in a peripheral nerve or the spinal cord, causes “microglial activation, ipsilateral and somatotopic to that injury site.   But second-order axons can relay injury signals and release cytokines in targets such as the thalamus, causing a secondary site of neuroinflammation in a supraspinal location (3). Neuroinflammation in the thalamus alters cortical-thalamic signaling, and thalamocortical dysrhythmias become a secondary consequence of this chain of events.” Cooper took us back to Charcot, who believed that the lesions of hysteria were a function of a biochemical or a physiological etiology, but could not be detected because of the lack of appropriate imaging technology.

With regard to the fixed dystonias of CRPS, Cooper has hypothesized that microglial activation in the spinal cord can trigger a retrograde spread of signaling to the motor cortex and establish a second neuroinflammatory site there (Figure 1). “Disinhibition of cortical spinal neurons in the motor cortex could be driving disinhibition of motor neurons in the spinal cord, leading to excessive gain in this motor control neural circuit and, ultimately, fixed dystonia.”


Figure 1. A neuroinflammatory model for the generation of fixed dystonias. (reprinted from Cooper and Clark, 2012). Neuroinflammation is established in the spinal cord, ipsilateral and segmental to the peripheral nerve injury. Retrograde migration of neuroinflammation (red arrow) occurs via corticospinal neurons, leading to a secondary neuroinflammation site in the motor cortex. Loss of chloride-dependent inhibitory tone and/or an increase in excitatory tone in the cortical and spinal neuroinflammatory foci lead to tonic firing of segmental motor neurons. A neuroinflammation-mediated loss of intracortical inhibition in the motor cortex may also be required for the genesis of fixed dystonia. (Reprinted from Cooper MS, Clark VP. Neuroinflammation, neuroautoimmunity, and the co-morbidities of CRPS. J Neuroimmune Pharmacology. 2012 Aug 25 [Epub ahead of print]). Images modified from The Inner Man™, Medical Illustrations Company, LLC.)

Some movement disorders appear to be driven by neuroinflammation, Cooper said. Consider Sydenham’s Chorea, a streptococcal infection followed by the development of an autoimmune response. The body is developing antibodies to fight off strep, but the antibodies that bind to strep also cross-react with neuronal targets in the basal ganglia.

“In this patient, you can see changes in the MRI — you see unilateral caudate, bilateral putamen changes, and also a change in one of the thalami. You find radiographic evidence of structural changes and neuroinflammation. Behaviorally, you observe chorea and emotional instability. This makes sense. There’s dysfunction in the basal ganglia, which is apparently being driven by multiple neuroinflammatory lesions in these deep brain nuclei,” Cooper added.

What about some of the patients that Charcot defined as hysterics? Augustine, one of Charcot’s most famous patients, had about a thousand “hysterical fits” per year. These hysterical fits were characterized by premonitions, wild emotional swings, and dystonic postures. Phenotypically, they fit the diagnosis of temporal lobe epilepsy. “Centers driving temporal lobe epilepsy can be located in the hippocampus,” Cooper said, noting that new evidence has been discovered that neuroinflammation inside the hippocampus may be one of the disorder’s key drivers (4). Whether a patient has a unilateral hemidystonia or a whole-body decorticate position really depends on how much the seizing event is emanating from the hippocampus to the basal ganglia and cortex (5). Cooper suggested that the abnormal postures in Augustine were probably ictal dystonias. But the underlying etiology of temporal lobe epilepsy takes us back to the key issue of neuroinflammation.

Autoimmune Components to CRPS

Cooper cited CRPS research published in Pain demonstrating that 90% of the adults studied had autoantibodies to one of two neurotransmitter receptors, the beta-2 adrenergic receptor and the muscarinic-2 acetylcholine receptor; 55% had autoantibodies to both of those autoantigens (6). These autoantigens are present in the sympathetic chain, spinal cord, peripheral tissues such as the heart, supraspinal sites such as the cerebellum and pons, motor cortex, and peripheral nerves, as well as the thalamus. So potentially the targets, depending on whether the autoantibodies extravasate from the vasculature, whether they get to a site, may be the initiators of a focal neuroinflammatory lesion. Cooper suggested we consider whether CRPS is a multifocal, autoimmune-mediated neuroinflammatory disorder, based on the presence of these autoantibodies and clear evidence of pro-inflammatory cytokine biomarkers in the blood and cerebrospinal fluid.

He noted that Christian Maihöfner and his colleagues have found that the somatosensory map in the brains of CRPS patients is altered (7). Moreover, Candy McCabe and her colleagues determined that CRPS patients have very unusual reactions to various stimuli. “If you show them an ambiguous figure like a duck-rabbit, or the Necker cube, a host of symptoms occur merely from looking at these figures (8). Should these abnormal reactions be interpreted as a psychoactive sign, or is it really psychogenic (caused by a pathology originating in the mind)? Is the cause of the disorder a psychological disturbance, or are we really just seeing sensory-motor integration problems being played out?“ Cooper calls these somatotopic disorders, which arise from distortions of somatotopic maps, and they’re not unique to CRPS patients. They also occur in low-back pain patients (9).

This begs the question: “Do these distortions of the somatosensory map correspond to what we call somatoform disorders?” Cooper compared the way we route neural information to a GPS map of the body. “We have multiple somatotopic maps in our CNS: spinal cord, thalamus, cortex, cerebellum. What happens if these somatotopic maps become distorted? Then, delivery of neural information doesn’t get to the right location. The somatotopic maps are very important, and if they become distorted, they need to be treated through psycho-physiological therapies to become healed.

Cooper then took us to the DSM-IV TR criteria. “We begin to see that these sensory deficits are present in motor and sensory functions.” Even if the symptom of pain is produced by looking at an ambiguous figure, it is not feigned. There’s no way of explaining it without utilizing integrated psycho-physiology and by considering how these pathways might be affected by neuroinflammation. Cooper pointed out that everything about CRPS fits the DSM-IV criteria for somatoform disorder, noting that CRPS also includes somatovisceral dysfunction, unusual rashes, and cognitive impairment. However, Cooper believes that CRPS is better explained as an autoimmune-mediated neuroinflammatory disorder, with many functional overlays that could be misinterpreted as a somatoform disorder. Moskovitz summarized: “Can we explain these ’unexplained medical symptoms‘ through physiological means, or do we still need to view this cluster of symptoms as a psychological disease?”

Moskovitz attributes these diagnostic categorizations to what he calls “the linear computation of medical understanding” that produces documents like the DSM-IV and the ICD-9. “They fit into a linear pattern of putting together symptoms and signs and biomarkers.” But, the practitioner’s mind is non-linear. He said, “As practitioners, we make hypotheses about our patients before we tick off anything on the differential diagnosis checklist. The inter-subjectivity of the practitioner/patient relationship is the first thing that engages. If we didn’t have it, we wouldn’t be healers; we wouldn’t be practitioners worth anything. In the literature, Moskovitz noted, the probable differential diagnosis of “the Thick-File Patient” is somatoform disorder. But perhaps it is something else (e.g., an autoimmune disorder). When practitioners are confused, they’re quick to think of psychological impairment in the patient, not in themselves.

Moskovitz cited a French expression, la belle indifference, which has been long considered a classic sign of conversion disorder. “The patient is indifferent to her symptoms. She exhibits indifference.” Perhaps the doctor is confused, as when Osler thought that a “morceau de papier” (the patient’s list of symptoms or problems) is a sign of “neuresthenia” (neurosis). Another condition characterized by indifference is anosognosia, a neurological condition that occurs in stroke. Moskovitz declared there is no scientific distinction between the neurological condition anosognosia and the psychological condition la belle indifference. “The only apparent difference is that anosognosia is due to a definable neurological lesion, a right-sided CVA, an ischemic or hemorrhagic injury, that affects the deep somatosensory apparatus of second-order mapping, and la belle indifference is the same condition, same symptoms, and same signs in the absence of a definable neurological injury.” This is analogous to the distinction between CRPS Type I and Type II. “Whereas CRPS Type II follows a definable neurological injury, CRPS Type I is the same disease with the same symptoms, the same signs, in the absence of a definable neurological injury.”

Moskovitz noted that expert opinion may not be uniform, but there is an expanding consensus that CRPS Type I does have an underlying neurological injury at its base, whether that injury is initially thermal, chemical, toxic, traumatic, or immune. “Anosognosia is the same, I propose. I’m speculating a little bit,” Moskovitz said. “Anosognosia Type II is, using the analogy from CRPS, the “absence of knowledge of a noxious condition” that follows a definable neurological injury, like right-sided CVA. Anosognosia Type I, Moskovitz’s classification of anosognosia, is the same condition, the same symptoms, the same signs in the absence of a definable neurological injury. Well, I’m going to propose that there is one,” declared Moskovitz. “Dr. Cooper and I both hold to the theory that underlying the impairment of CRPS is probably neuroinflammation. Anosognosia Type I may be a neuroinflammatory insult to the second-order mapping areas of the temporal parietal region, just as CRPS is a neuroinflammatory insult to the nociceptive apparatus,” he said.

Moskovitz concluded: “The science of neuroinflammation and immune-mediated mechanisms of disease holds the greatest promise into the future. Be that as it may, until we validate the biomarkers of CRPS, the ethical obligation of practitioners is to trust the patient’s narrative of illness, the patient’s story of suffering. The science is incomplete. Stay tuned, because I think a synthesis of understanding is coming soon.”

Cooper summarized the concepts discussed in the lecture, and showed a diagram from the work of Dr. Richard Banati, who is a pioneer in the use of PET (positron emission tomography) for imaging neuroinflammation in supraspinal areas. In 2001, Banati and colleagues performed a study with eight chronic pain patients (10). “As far as I know, this is the only time that chronic pain patients have been examined with PET, looking for sites of neuroinflammation in supraspinal areas. In each of the seven patients with unilateral nerve injury, neuroinflammation was found in the contralateral thalamus, not in the ipsilateral thalamus, and this fits the idea that inflammation is propagating through interneurons, to the second-order synapse in the contralateral thalamus through a nociceptive pathway. In the case of the person who had bilateral nerve injury, having both thalami inflamed also makes sense from a somatotopic perspective.   So I’d like to leave you with the idea that neuroinflammation does not necessarily remain confined to an injured peripheral nerve. Nerve injury can induce neuroinflammation ipsilaterally and somatotopically within the cord. And in a minority of cases, neuroinflammation may spread either through descending fibers in a retrograde fashion, or it can move anteriorly through fibers to supraspinal sites,” he explained. “There is a new physiological mechanism to explain why chronic pain may present in the absence of a visible nerve injury in the periphery. Healing may have taken place in the periphery, but things have changed in the supraspinal level. This type of neuroinflammatory lesion may be what Charcot envisioned long ago as the lesion of hysteria. We can now view central neuroinflammatory lesions as drivers for thalamocortical dysrhythmias, drivers for the changes of the somatosensory cortex, as well as dysfunctions that may radiate from those changes.”

Cooper concluded by saying; “Neuroinflammation can spread through the neuraxis via axonal projections. Autoantibodies can ignite neuroinflammation. We know this from Sydenham’s Chorea. We know that from Parkinson’s disease. We suspect this in CRPS. Non-dermatomal pain can be caused by thalamic neuroinflammation and/or body schema distortions. We’re very grateful that Dr. Candy McCabe and Dr. Jenny Lewis came to this conference and provided a day-long workshop on the importance of understanding these distortions and utilizing novel therapies that are psycho-perceptional in nature, such as mirror box therapy and graded motor imagery, designed to heal by resetting this body schema.”

Debra Nelson-Hogan is the Director of Education and Credentialing for the American Academy of Pain Management.


  1. Kellett MW, Humphrey PR, Tedman BM, Steiger MJ. Hyperekplexia and trismus due to brainstem encephalopathy. J Neurol Neurosurg Psychiatry. 1998;65(1):122125.
  2. Oaklander AL, Fields HL. Is reflex sympathetic dystrophy/complex regional pain syndrome type I a small-fiber neuropathy? Ann Neurol. 2009;65(6):629–638.
  3. Saab CY, Hains BC. Remote neuroimmune signaling: A long-range mechanism of nociceptive network plasticity. Cell. 2009;32(2):110–117.
  4. Das A, Wallace GC 4th, Holmes C, et al. Hippocampal tissue of patients with refractory temporal lobe epilepsy is associated with astrocyte activation, inflammation, and altered expression of channels and receptors. Neuroscience. 2012 Sep 18;220:237-246.
  5. Cleto Dal-Cól ML, Bertti P, Terra-Bustamante VC, et al. Is dystonic posturing during temporal lobe epileptic seizures the expression of an endogenous anticonvulsant system? Epilepsy Behav. 2008;12(1):39-48.
  6. Kohr D, Singh P, Tschernatsch M, et al. Autoimmunity against the β2 adrenergic receptor and muscarinic-2 receptor in complex regional pain syndrome. 2011;152(12):2690-2700.
  7. Maihöfner C, Handwerker HO, Neundörfer B, Birklein F. Cortical reorganization during recovery from complex regional pain syndrome. Neurology. 2004;63(4):693701.
  8. Cohen HE, Hall J, Harris N, McCabe CS, Blake DR, Jänig W. Enhanced pain and autonomic responses to ambiguous visual stimuli in chronic Complex Regional Pain Syndrome (CRPS) type I. Eur J Pain. 2012;16(2):182-95.
  9. Bray H, Moseley GL. Disrupted working body schema of the trunk in people with back pain. Br J Sports Med. 2011;45(3):168–173.
  10. Banati RB, Cagnin A, Brooks DJ, et al. Long-term trans-synaptic glial responses in the human thalamus after peripheral nerve injury. Neuro Report. 2001;12(16):34393442.

Originally printed in the American Academy of Pain Management’s quarterly publication, The Pain Practitioner.




Leave a Reply

Your email address will not be published. Required fields are marked *

More CRPS Stories

About The Academy

The American Academy of Pain Management improves the lives of people with Pain by advancing a person-centered, integrative model of pain care through evidence-guided education, credentialing, and advocacy.