Clinical manifestations

Presentation and course

Different types of injury like direct trauma, crush, laceration, surgery, burn, or immobilization have been described to cause peripheral dystonia. In most case reports of peripheral dystonia, the neck and limbs are the most frequently affected anatomical site (35).

Among these, cervical dystonia is the most common posttraumatic dystonia. Posttraumatic cervical dystonia may represent a unique syndrome that is distinct from the idiopathic variant. Truong and colleagues reported six cases of torticollis following neck injury with clinically distinct features (70). The clinical features of posttraumatic cervical dystonia were later expanded by other authors (26; 34). A history of significant neck trauma associated with a motor vehicle accident is reported by about 10% to 20% patients presenting to neurologists with cervical dystonia (12; 69; 54). Interestingly, in some cases, the injury can be relatively mild (70; 22). A positive association between trauma and cervical dystonia was found in a large study of 202 patients with dystonia and 202 controls (14).

Based on the temporal relation between trauma and dystonia, posttraumatic cervical dystonia can be classified as acute onset or delayed onset (70; 26; 69; 22; 54). When the onset is acute (within 4 weeks), the dystonia is characterized by markedly reduced cervical mobility, prominent shoulder elevation with trapezius hypertrophy, absence of involuntary movements, sensory tricks or activation maneuvers, and poor response to botulinum toxin injection. This is in contrast to delayed-onset cervical dystonia (between 3 months and 1 year), which is clinically indistinguishable from nontraumatic idiopathic cervical dystonia (68; 69).

In addition, the torticollis tends to persist during sleep and does not improve during rest or support, in contrast to idiopathic torticollis, which often disappears under these circumstances (70).

Increased frequency of laterocollis and depression in a group of early posttraumatic cervical dystonia (within 4 weeks) was also reported in a case controlled study (54). Frei and colleagues reported eight of nine patients with posttraumatic cervical dystonia with predominant laterocollis (22).

When compared to patients with no antecedent trauma, there were no group differences in the family history of dystonia (54).

Complex regional pain syndrome (CRPS) is also a frequent association in the posttraumatic group and may represent a variant of posttraumatic cervical dystonia (34; 22). In a large study of 121 patients with trauma-induced complex regional pain syndrome, the interval between the trauma and dystonia ranged from immediate (33%) to more than 1 year (25%) (72). Because of the unusual features and possible medicolegal implications, physicians may tend to diagnose this condition as a psychogenic disorder or litigation-oriented behavior.

In a systematic review, 713 cases with peripherally induced movement disorders from 133 different studies were analyzed (73). Dystonia was the most frequent peripherally induced movement disorder (72%). The median time interval from trauma to the onset of dystonia was 21 days. More than one third of patients with peripherally induced movement disorders were also diagnosed with CRPS (36%), and 14% of the patients were diagnosed with a psychogenic movement disorder.

Although cervical dystonia is the most reported focal dystonia resulting from peripheral injury, there is a growing number of reports of focal dystonia involving other anatomical sites, including limbs, oromandibular muscles, lower cranial structures, and eyelids (61; 21; 59; 63; 23). In cases of limb dystonia, the feet are more likely to be involved than the hands, although exact figures are not known. In all these cases, the nature of trauma seems to be less critical to the development of focal dystonia than the location of the insult. Moreover, no clear correlation between the severity of the trauma and subsequent development of dystonia was established in any published reports. In many cases, the dystonia starts locally in the injured region but may later spread to involve adjacent and ipsilateral body parts, eventually crossing to the contralateral side. In cases where focal dystonia involved the foot, hand, and neck, the sites of injuries were reported in the areas of ankle, thumb/elbow, and neck/shoulder, respectively.

One form of peripherally induced dystonia is “runner’s dystonia”. In one series of patients with runner’s dystonia with a mean age of 44.6±10.43 years and a mean duration of symptoms of 7.2±4.44 years who initially noted dystonia of one leg during long-distance running, two of five had injury to the affected leg within one year prior to the onset of the dystonia, raising the possibility of peripherally induced dystonia (76). In another series involving seven cases (four male, three female), several different exercise triggers (cycling, hiking, long-distance running, drumming) were identified (37). The mean age of symptom onset was 53.7±6.1 years. The median symptom duration prior to diagnosis was 4 (9.5) years. Several patients underwent unnecessary procedures prior to being appropriately diagnosed. Over a median of 2 (3.5) years, signs and symptoms progressed to impair walking.

Traumas involving the face or oral and dental structures were reported in cases with peripherally induced oromandibular dystonia (59). The term “edentulous dyskinesia” has been described in patients with misaligned dentures, which may have caused an impairment of proprioception of the oral cavity, resulting in dystonia (39). Ocular lesions preceded the development of blepharospasm in 12% of patients in one series (28). Fixed jaw opening dystonia and dystonias involving the tongue, lips, and neck were reported following dental procedures (63). Charness and colleagues reported 24 patients with task-specific dystonic flexion of the fourth and fifth finger associated with ipsilateral ulnar neuropathy (10). Because patients with dystonia are predisposed to peripheral injury, it is possible that both resulted independently from a common antecedent trauma (10).

Table 2. Comparison of Idiopathic Torsion Dystonia and Peripherally Induced Dystonia

Prognosis and complications

The natural history and long-term outcome of peripherally induced dystonia have not been well studied, although no cases of complete recovery have been reported. Fixed contractures may develop if physical therapy and muscle relaxation treatments are delayed. The natural history and outcome also depend on the underlying etiology. Generally, secondary dystonias have a rather poor prognosis unless the underlying cause can be corrected and treated (30). Most patients suffer psychological stress not only due to their disorders but pending litigations as well (Scarano and 33). A study by Sa and colleagues suggested an important role of psychological factors in the etiology or maintenance of abnormal posture and associated disabilities in these patients (58). Pain, which can be severe, may not be adequately treated, and the requirements of strong analgesics may be perceived by some physicians as drug-seeking behavior. In a case-control study of cervical dystonia following peripheral trauma, depressed mood requiring treatment was a prominent feature, and depression was reported significantly more often in those with preceding trauma (54).

Biological basis

Etiology and pathogenesis

Establishing a causative relationship between peripheral injury and subsequent dystonia is often challenging experimentally, and there is currently no animal model that adequately replicates the clinical syndrome. The central debate revolves around whether the relationship between trauma and subsequent dystonia is causative or purely coincidental. One of the main issues is the latency between trauma and the onset of dystonia, which poses difficulties in reliably supporting a causal relationship. To support the notion that peripheral injury can genuinely cause dystonia, it is necessary to demonstrate that lesions of peripheral sensory pathways can induce extensive changes both proximal and central to the site of damage. Several lines of evidence suggest that peripheral injury can lead to reorganization at cortical, subcortical, and spinal cord levels, resulting in motor dysfunction.

For instance, studies have shown that deafferentation of the spinal cord in humans leads to neuronal and reflex hyperexcitability that extends beyond the affected segment (45). Patients with focal hand dystonia have exhibited abnormal finger representation and overlapping tactile receptive fields on functional imaging studies, indicating possible plastic reorganization (03). Individuals with digit or limb amputation have shown larger motor-evoked potentials and expanded cortical representations that occupy cortical territories previously responsible for the amputated digit or limb (11; 38; 66; 17). Furthermore, studies by Byrnes and colleagues using transcranial magnetic stimulation demonstrated a displacement of the corticomotor map in patients with writer's cramp, which was temporarily altered following botulinum toxin injection (07). Based on this evidence, peripheral deafferentation may induce changes in neural transmission and reorganization of local neural circuitry, leading to increased evoked and spontaneous motor responses. This effect may be mediated through gamma neuron overactivity resulting from disinhibition, ectopic activation, ephaptic transmission, or collateral sprouting (62).

Opponents of the existence of peripherally induced dystonia present three main arguments against the entity: (1) The "denominator problem" refers to the small number of people who develop dystonia despite a large number of patients with trauma (74). However, the low prevalence of dystonia could be attributed to different genetic susceptibility to acquiring the disease, as the etiology of the disorder is likely multifactorial. (2) The "anatomic criterion" highlights the challenge of clearly correlating the site of trauma with dystonia. In a 2007 study, van Rijn and colleagues suggested that once triggered, the central mechanism underlying dystonia could facilitate its spread to other segments without requiring additional trauma (72). (3) The "duration criterion" is another argument challenging the validity of posttraumatic dystonia criteria, as reports exist of dystonia appearing as late as three years after the trauma (05).

Contrary to the historical view of dystonia as a pure motor disorder, clinical and experimental evidence suggests that sensorimotor integration is abnormal in focal dystonia (01). Reduced inhibition at cortical, subcortical, and spinal levels, as well as abnormalities in sensory-motor integration and maladaptive plasticity, have been proposed as the main mechanisms underlying dystonia development (43; 42).

Abnormal sensory input in the etiology of posttraumatic dystonia is illustrated the best by noxious sensory input from complex regional pain syndrome (CRPS). Dystonia is often associated with other sensory and autonomic phenomena as part of complex regional pain syndrome. Complex regional pain syndrome is classified as type I if the trauma is a soft-tissue injury associated with spontaneous pain and allodynia but no obvious nerve injury (previously known as reflex sympathetic dystrophy) or as type II (causalgia) if the nerve injury is present and sensory symptoms are not limited to the territory of the injured nerve. The pathophysiology of complex regional pain syndrome includes several mechanisms: facilitated neurogenic inflammation, pathological sympatho-afferent coupling, and neuroplastic changes in the central nervous system (47). The prevalence of dystonia in complex regional pain syndrome ranges from 5% to 60% in different reports (57).

Maladaptations at different levels–peripheral, cortical, and spinal–have been demonstrated (40). Abnormal fMRI pattern of activation is seen in complex regional pain syndrome patients with dystonia on imaginary hand movements (25). Transcranial magnetic stimulation and magnetoencephalogram suggest presence of motor cortex disinhibition in patient with complex regional pain syndrome. Fixed dystonia associated with complex regional pain syndrome has been attributed to maladaptive neuroplastic changes throughout the neuroaxis (51). These changes are associated with abnormal representation of the dystonic limb in the primary somatosensory cortex (16) and reduced cortical inhibition (02).

Basal ganglia involvement in the pathophysiology of dystonia is supported by a series of imaging studies. Functional connectivity within the sensory motor cortex is impaired as is connectivity with other brain regions (06). Functional imaging studies in children with complex regional pain syndrome have shown increased activity in the affected limb in putamen and globus pallidus and decreased activity in the caudate nucleus in comparison to the unaffected side in the presence of noxious stimulus (44). Interestingly, this abnormal basal ganglia activation persists even after stimulus is removed (04).

Reports of psychosomatic illness and malingering in patients with peripheral dystonia have raised the question of psychological factors in the etiology of peripheral dystonia and CRPS. Schrag and colleagues found in a group of 103 patients with dystonia that 37% had psychogenic dystonia, and 29% fulfilled the criteria for somatization disorder (64). Unlike psychogenic tremor, where tests for distraction and entrainment are often useful, in psychogenic dystonia these tests are not helpful because posture can be maintained in one limb while the opposite limb performs voluntary movements (29). Interestingly, cortical inhibition is impaired in both organic and psychogenic dystonia (19). In a study, Quartarone and colleagues looked at the cortical plasticity using a protocol known as “paired-associative stimulation” in which an electrical stimulus to the median nerve at the wrist is paired with a TMS pulse (56). The authors showed increased plasticity in organic dystonia with normal response in psychogenic dystonia, suggesting that this could be a useful way to differentiate between organic and psychogenic dystonia.

However, from the practical standpoint, clinical examination and established electrophysiological testing (EMG) remain the preferred mode of diagnosis. It should be noted that the presence of psychiatric symptoms like depression and OCD does not rule out an organic cause because movement disorders in general are associated with psychiatric symptoms (40).

Epidemiology

The prevalence of posttraumatic dystonia has been reported to vary from 1% (69) to 25% (22; 54). A retrospective study from the Italian Dystonia Registry identified peripheral trauma as a minor contributor, 0.9%, to idiopathic or acquired dystonia (15). Why do only a small proportion of patients with peripheral injury develop dystonia? If injuries alone cause dystonia, it should be much more prevalent than it is. Peripherally induced movement disorders represented only a small proportion of a high incidence of local limb trauma. This issue, known as the “denominator problem,” is used as the main contra argument for the causality between trauma and dystonia. However, a low prevalence of disease should not be used as an argument against its existence and should make one think that genetic susceptibilities predispose certain individuals to develop dystonia post trauma whereas others do not. Jankovic and van der Linden reported that 18 of 23 patients developed focal dystonia of the body part that was injured within a year after injury (36). They proposed that some patients may have had a preexisting subclinical or very mild movement disorder, which may have contributed to the injury; and the injury may trigger, not cause, the expression of the movement disorder. Furthermore, nine patients had reflex sympathetic dystrophy, now termed complex regional pain syndrome, suggesting the frequent association between peripheral movement disorders and pain. Focal dystonia was also reported as a motor manifestation of reflex sympathetic dystrophy, which may precede sudomotor, vasomotor, or pain symptoms by weeks or months (65). Chronic immobilization, which may induce alterations in cortical sensory representations and remodeling, is also an important risk factor for segmental dystonia (53). In a study of 10 patients whose right arm was immobilized for at least two weeks, MRI showed a decrease in cortical thickness in the left primary motor and somatosensory area, as well as a decrease in fractional anisotropy in the left corticospinal tract (41).

A more recent survey from a population insurance dataset showed that the incidence of dystonia in the trauma group was 2.27 per 100,000 person-years whereas the nontrauma group had only 0.71 per 100,000 (46).

Differential diagnosis

Pseudodystonic postures can also occur after trauma and may be related to peripheral nerve or nonneurologic injury (32). It is important to distinguish pseudodystonic postures from focal true dystonia, as the treatment strategies and outcomes are significantly different. Indeed, many patients with cervical pathology due to local trauma such as atlantoaxial subluxation, focal myositis, or fibrosis may develop abnormal neck posture, which can be incorrectly attributed to true focal dystonia. These features should be recognized by the fixed posture and lack of resolution during sleep and probably represents tearing and invagination of capsular ligaments around the joints. The eighth nerve lesion or lesions in the vestibular pathway may also cause abnormal head posture resembling cervical dystonia. Severe local injury can tear muscles and lead to unopposed contraction by antagonist muscles with resultant posturing of the injured body part. Extreme pain and evidence of soft tissue injury and evidence of late tissue fibrosis are usually the clues to pseudodystonia, and the treatment usually consists of surgical ligation of fibrotic bands with rehabilitation and physical therapy. Treatments for focal dystonia are generally ineffective in cases with pseudodystonia. It is also important to exclude the possibility of psychogenic dystonia and secondary gain. They are usually accompanied by atypical weakness or sensory symptoms, fixed postures, and lack of worsening by action or improvement by sensory tricks.

The movement patterns are usually not stereotyped but, rather, fluctuate.

Psychogenic dystonia should be considered in patients with associated psychological disturbances or in which medico-legal factors play a role. Remittance of symptoms with suggestion, inconsistency, and presence of multiple somatizations suggest the diagnosis of psychogenic movement disorder (29). The fixed painful dystonia after peripheral trauma frequently overlaps with CRPS and psychogenic movement disorders. A summary of the clinical features that differentiate psychogenic dystonia from peripherally induced dystonia and CRPS are summarized in a review of trauma-induced dystonia by Hawley and Weiner (31). Sudden onset, fixed dystonia and acute peripheral trauma have been associated with functional dystonia (24; 18).

Table 3. Clinical Features of Dystonia Types

Diagnostic workup

When the onset of focal dystonia is immediate or within a few days following peripheral injury, there is a clear temporal association, arguing against it being coincidental. The problem, however, arises when the time of peripheral injury has long preceded the onset of dystonia, making a causal relationship difficult to ascertain. It is unclear how distant a peripheral trauma should be from the onset of dystonia for it to be considered causal. Various authors used different maximal cut-off periods between 3 months and 8 years as being acceptable in establishing a causal relation, and these numbers were arbitrary (48; 62; 70; 34; 22). Truong and colleagues initially used very rigid criteria with the time preceding the development of dystonia of only four days to produce a homogenous subset to analyze the syndrome (70). After establishing the critical characteristics of the posttraumatic torticollis syndrome, they used these criteria with a more liberal timeframe preceding the development of dystonia in their subsequent paper and noticed a remarkable uniformity of the syndrome (22). If the typical features are present, a further extension of time preceding the onset of dystonia may be allowed. In order to prove a cause-and-effect relationship after peripheral injury, Jankovic has proposed the following criteria for the diagnosis of peripherally induced movement disorders as follows: (1) the trauma is severe enough to cause local symptoms for at least two weeks or requires medical evaluation within two weeks after trauma, (2) the initial manifestation of the movement disorder is anatomically related to the site of injury, and (3) the onset of movement disorder is within days or months (up to 1 year) after the injury (34). Moreover, the causal relationship should be supported by the absence of other causes capable of producing the same symptoms, presence of complex regional pain syndrome, and poor response to conventional treatment. Despite the above criteria, certain limitations exist, and not all patients who satisfy the above criteria can be confirmed to have a peripherally induced movement disorder. Frei and colleagues also raised the question whether a more liberal outer limit of the time frame of the trauma could be accepted, especially when the typical symptoms are present (22). Kumar and Jog have reviewed the diagnostic criteria and the current controversies in the field of “posttraumatic dystonia” (a term they prefer to “peripherally induced dystonia”) (40).

As mentioned above, in the case of dystonia, electrophysiology is not helpful in the diagnosis of psychogenic condition. Advances suggest that paired-associative stimulation using transcranial magnetic stimulation is able to differentiate them (56). Using blink reflex recovery cycle to test brainstem inhibition in patients with blepharospasm, Schwingenschuh and colleagues showed that patients with clinically defined organic blepharospasm displayed reduced inhibition in the brainstem as expected (67). However, the brainstem inhibition was normal in patients with suspected psychogenic blepharospasm.

Management

The symptomatic treatment of dystonia has markedly improved since the introduction of botulinum toxin and as a result of advanced surgical treatments, including deep brain stimulation. In most cases, the treatment is merely symptomatic, aimed to improve posture and function and to relieve associated pain. In the case of cervical dystonia, many different treatments have been tried over the years, although localized therapy using botulinum toxin injection seems to provide a high rate of response with a low incidence of side effects. Oral medications such as anticholinergic agents, dopamine receptor antagonists, and GABAmimetic agents have been used in a trial-and-error manner and have a low rate of efficacy. Globus pallidus deep brain stimulation may be considered for more complex forms of cervical dystonia, primary dystonia, or when more widespread dystonia is present. Isolated reports of successful deep brain stimulation for peripheral dystonia have been published (09; 50).

Although the above treatments have also been tried in patients with peripherally induced dystonia, the results have been disappointing. Anticholinergics are usually ineffective. Because of widespread muscle involvement, for example all cervical muscles in peripherally induced cervical dystonia in contrast to only limited sets of muscles in cases of idiopathic torticollis, only mild improvement is usually obtained with botulinum toxin injection, although some authors reported satisfactory control of pain and dystonia (70; 26; 32; 69). Sankhla and colleagues reported poor response to various medications, including trihexyphenidyl, baclofen, and clonazepam in cases with peripherally induced oromandibular dystonia (59). Sympathetic blocks, sympathectomies, and guanethidine infusions rarely provide lasting relief of pain, but corticosteroids may be helpful in some patients with associated complex regional pain syndrome (32). In addition to pharmacological and surgical modalities, potential psychological and stress factors should be explored and adequately treated (58). Occupational and physical therapy such as mirror feedback therapy should also be considered (49).

The use of spinal cord stimulation treatment for CRPS was reviewed by the Neuromodulation Appropriateness Consensus Committee and was considered to be efficacious and strongly recommended for implementation, as it was supported by well-designed experimental, clinical, or epidemiological studies (level I) (13). Psychotherapy, cognitive behavioral therapy, rehabilitation, antidepressants, and hypnosis have been used when psychogenic dystonia has been suspected (29).