48 Painful legs and moving toes

In 1971 Spillane et al. described six patients afflicted by persistent pain in the legs and involuntary movements in their toes under the heading of ‘painful legs and moving toes’. This name has stuck and there have subsequently been a number of other reports of this disorder (Nathan 1978, Barrett et al. 1981, Schott 1981, Wulff 1982, Montagna et al. 1983, Schoenen et al. 1984, Bovier et al. 1985, Leger et al. 1985, Gastaut 1986, Averianor et al. 1993, Walters et al. 1993, Dressler et al. 1994, Pla et al. 1996, Tan and Tan 1996, Sanders et al. 1999). It is a rare disorder: out of 4,780 database patients with movement disorders diagnosed at Mayo Clinic, Arizona, from 1996 to 2006, only 14 cases were diagnosed with painful legs and moving toes or a related disorder. Ages ranged from 25 to 84 years (mean, 69 years) (Alvarez et al. 2008).

The condition is characterized by pain in one or both lower limbs associated with persistent involuntary movements of the toes, which sometimes spreads to involve more proximal parts of the limb. In some cases the disorder seems to arise spontaneously without any apparent underlying cause (Spillane et al. 1971, Walters et al. 1993, Dressler et al. 1994), but more commonly it follows damage to the lower spinal cord or cauda equina (Nathan 1978), nerve roots (Spillane et al. 1971, Nathan 1978, Barrett et al. 1981, Wulff 1982, Averianor et al. 1993, Dressler et al. 1994), peripheral nerves (Montagna et al. 1983, Bovier et al. 1985, Leger et al. 1985, Dressler et al. 1994, Pla et al. 1996, Mattos et al. 2002), or musculoskeletal structures of the foot (Schott 1981, Dressler et al. 1994).

Nerve biopsy has been reported as showing Wallerian degeneration or demyelination (Spillane et al. 1971), but available material has been very limited. Nerve conduction studies have been normal or revealed slowing of conduction and reduction in amplitudes, depending on the underlying cause (Spillane et al. 1971, Nathan 1978, Schott 1981, Montagna et al. 1983). Thus, there may be evidence of a proximal, localized, or generalized neuropathic process.

Electromyography (EMG) reveals long duration (0.5–1 second) bursts of activity, which can be comprised of normal motor units (Dressler et al. 1994) or show evidence of chronic partial denervation (Montagna et al. 1983). Although the majority of lesions causing this disorder affect nerve roots or peripheral nerves, the EMG activity is organized in complex sequences, activating flexors, extensors, abductors, and adductors of digits in various patterns. Schoenen et al. (1984) described two types of EMG patterns, the first consisting of short bursts of erratic firing which sometimes affected antagonistic muscles synchronously, and the second showing arrhythmic bursts of longer duration activity alternating between antagonistic muscles. As the EMG discharges in the first group, however, could not be prevented by local nerve block sufficient to paralyze voluntary activity, it seems likely they resulted from neuromyotonia (see Chapter 45) and as all of the second group had involuntary movements at other sites, including orofacial dyskinesia, the exact classification of these patients is uncertain.

Nathan (1978) suggested the coordinated movements indicated integration within the central nervous system and proposed the movements were induced by activation of local spinal circuits of interneurons and neurons, due to spontaneous frequent firing of afferents in the posterior roots. He felt the pain resulted from other impulses passing to higher regions in the nervous system.

Dressler et al. (1994) noted the causalgic nature of the pain and reported some patients with identical involuntary movements but without pain. Similar painless movements have been noted by others (Walters et al. 1993). Dressler et al. 1994 proposed a spectrum of disorders, ranging from causalgia without movements at one extreme to movements without pain at the other, and painful legs and moving toes interposed. They suggested that peripheral nerve injury might lead to reorganization of central processing of afferent sensory information within the nervous system, a factor that has been postulated to underlie causalgia. This concept is in keeping with the hypothesis of Nathan (1978).

As mentioned above, the majority of cases of painful legs and moving toes occur in assocation with lesions of the lower spinal cord, cauda equina, nerve roots, peripheral nerves, or musculoskeletal tissues of the distal lower limb. Pathologies have included direct trauma to the spinal cord and cauda equina (Nathan 1978, Dressler et al. 1994), lumbosacral radiculopathy secondary to spondylosis or disc prolapse (Spillane et al. 1971, Dressler et al. 1994), nerve root compression by Tarlov cyst (Nathan 1978), haemangioma or a lower lumbar vertebra (Dressler et al. 1994), peripheral neuropathy (Montagna et al. 1983, Gastaut 1986, Dressler et al. 1994), tarsal tunnel syndrome with accessory soleus muscle (Pla et al. 1996), fracture (Dressler et al. 1991), orthopaedic surgery (Schott 1981), varicose vein stripping (Dressler et al. 1994), soft tissue trauma (Schott 1981, Dressler et al. 1994), herpes zoster (Nathan 1978, Dressler et al. 1994, Ikeda et al. 2004), and HIV (Pitagoras et al. 1999, Mattos et al. 2002). When surgery or trauma to the limb itself have precipitated the disorder they have usually involved the foot and have otherwise seemed quite uncomplicated. In a small number of cases, there is no apparent precipitant and Dressler et al. (1994) found this to be the case in five out of 18 of their patients.

When there is an antecedent trigger, the first symptoms of painful leg and moving toes may commence weeks, months, or years later. In one case a fall on the back seemed to bring on the symptoms within a day, although there had been a fracture of an ipsilateral metatarsal 8 months earlier (Dressler et al. 1994). The pain and movements may be unilateral or bilateral, more often bilateral (Alvarez et al. 2008). They can commence on one side and spread to the opposite limb, occasionally resolving in the initially affected leg.

Pain is usually the first indication of painful legs and moving toes and it can precede the onset of involuntary movements by a widely varying interval (between days to years) (Spillane et al. 1971, Nathan 1978, Schott 1981, Dressler et al. 1994). In a few cases movements have commenced before the pain (Spillane et al. 1971, Dressler et al. 1994). The pain is usually not limited to a segmental or dermatomal distribution and radiates much more widely. For example, the observation of a painful tongue in a patient with painful legs and moving toes syndrome has been reported (Schwingenschuh and Bhatia 2008). Pain can vary in intensity from a mild discomfort to being extremely severe. Often it is a constant deep pain and may be described as an ache, pressure, bursting, tightness, crushing, pulling, or burning (Spillane et al. 1971). It may be constant or throbbing and can increase in severity and radiate throughout the limb in response to a variety of stimuli. In addition to signs of loss of neurological function caused by any underlying lesion, there may be a variety of other sensory phenomena, including cutaneous hyperpathia, hyperaesthesiae, and allodynia. These features have been considered to be consistent with causalgia (Dressler et al. 1994). Spillane et al. (1971) emphasized that even when this syndrome accompanied compressive radiculopathy the symptoms did not resemble those of pain arising from nerve root irritation and there is no aggravation by sneezing, coughing, defecating, or bending.

The involuntary movements (Fig. 48.1) usually consist of complex sequences involving flexion, extension, abduction, and adduction of the toes in various combinations. There may be sinuous clawing, restraightening, fanning, and circular movements. When they are minimal they may appear as a slow quivering (Spillane et al. 1971). They cannot be imitated. The movements may be continuous or intermittent, but pauses are often quite short. With an effort of will it is usually possible to arrest them briefly, but they recommence again after a few seconds or when the patient's attention is diverted. They may also be altered by a change of posture (Fig. 48.2). At other times, however, they may be precipitated by voluntary contraction of foot muscles (Nathan 1978).

Sometimes involuntary movements can affect more proximal muscles and there can be dorsiflexion and plantar flexion of the ankle and inversion and eversion of the foot (Nathan 1978, Montagna et al. 1983). Even involvement as proximal as the thigh has been recorded (Spillane et al. 1971). Ebersbach et al. 1998 described an unusual case of unilateral painful legs and moving toes syndrome with spread to the ipsilateral hand. However, the semirhythmic muscle activities in the two extremities were different excluding one central pacemaker to account for both.

Similar movements have been described in the fingers with pain in the arm with brachial plexus damage (Verhagen et al. 1985), but this seems extremely rare and neuromyotonia secondary to peripheral nerve damage has not definitely been excluded (see Chapter 45).

In addition to the involuntary movements, patients show the motor and sensory signs associated with spinal, radicular, or peripheral nerve pathology. Occasional cases may also show vascular skin changes with vasodilation or pallor (Nathn 1978) and Sudek's atrophy has been reported (Dressler et al. 1994).

As mentioned above, identical involuntary movements of the toes may develop without pain and these have been idiopathic or associated with similar pathology. In the latter case they may also be accompanied by objective neurological findings (Dressler et al. 1994).

In many patients the involuntary movements seem to be proportionally related to the pain. Thus, when the pain is severe, the movements are more prominent. Abolition of the pain using lumbar sympathetic block, however, is not necessarily associated with disappearance of movements and conversely movements may stop while the pain remains present (Spillane et al. 1971, Nathan 1978). Thus, although the two are linked, one does not seem to be the direct result of the other.

Polysomnographical recordings show that the involuntary movements are considerably reduced during sleep, but are not totally abolished and can be seen during both slow wave and rapid eye movement periods (Montagna et al. 1983). In addition, disruption of sleep pattern is common with increased sleep latency, decreased total sleep, and a diminished percentage of time spent in deep sleep stages (Montagna et al. 1983). Once established, painful legs and moving toes usually persist, although disappearance from a limb has been recorded (Dressler et al. 1994).

A variety of treatments have been attempted without consistent success. These include administration of benzodiazepines, carbamazepine, tricyclic antidepressants, and baclofen (Montagna et al. 1983). Treatment with progabide, a gamma amino butyric acid agonist, has been reported to decrease involuntary movements, relieve the pain, and improve the local circulation (Bouvier et al. 1985) but further studies have to be done. Gabapentin has also been explored (Villarejo et al. 2004, Aizawa 2007). Guieu et al. (1994) claimed to have successfully treated two patients by the administration of adenosine triphosphate (ATP) based on the findings of decreased circulating adenosine levels and again further studies are needed on both of these two agents. In addition, atropine-like drugs, quinine sulphate, propranolol, local cooling, local anaesthetic injections, vibration, transcutaneous electrical stimulation, and acupuncture have not been helpful (Schott 1981). Lumbar sympathetic blockage initially seemed to help some patients (Spillane et al. 1971, Shime and Sugimato 1998), but on subsequent evaluation the results have been disappointing (Nathan 1978). Lumbar epidural block was reported to be helpful in two patients (Okuda et al. 1998) and epidural spinal cord stimulation has been said to be effective in individual patients, but again further studies need to be done (Takahashi et al. 2002, Raina et al. 2007).