| Sign In to gain access to subscriptions and/or personal tools. |
DOI: 10.1016/J.JHSB.2006.04.027 © 2006 The British Society for Surgery of the Hand
Pain Phenomena and Sensory Recovery following Brachial Plexus Avulsion Injury and Surgical RepairsThe Royal National Orthopaedic Hospital, Stanmore, Hammersmith Hospital, London, and St Mary’s Hospital, London, UK Correspondence: Professor Thomas Per Carlstedt, PhD, The Royal National Orthopaedic Hospital, Stanmore, UK.E-mail: thomascarlstedt{at}fsmail.net
Seventy-six patients with severe brachial plexus avulsion injuries were studied using pain questionnaires and quantitative sensory testing. There was significant correlation between pain intensity and the number of roots avulsed prior to surgery (P = 0.0004) and surgical repairs were associated with pain relief. Sensory recovery to thermal stimuli was observed, mainly in the C5 dermatome. Allodynia to mechanical and thermal stimuli was observed in the border zone of affected and unaffected dermatomes in 18% of patients assessed early (<6 months) and 37% patients at later stages. Pain and sensations referred to the original source of afferents occurred at a later stage (>6 months) in 12% of patients and were related to nerve regeneration. By contrast, "wrong-way" referred sensations (e.g. down the affected arm while shaving or drinking cold fluids) were reported by 44% of patients and often occurred early, suggesting CNS plasticity. Understanding sensory mechanisms will help develop new treatments for severe brachial plexus injuries.
Key Words: brachial plexus injury • re-implantation • neuropathic pain • referred sensation • CNS plasticity After dorsal spinal nerve root avulsions from the spinal cord, numbness in the affected limb may be associated with extreme, almost unbearable, intractable pain (Berman et al., 1998; Birch et al., 1998; Wynn Parry, 1980). Chronic pain occurs in about 5% of patients with peripheral nerve injury (Sunderland, 1993) and this also applies, generally, to patients with plexus injuries where the lesion lies distal to the dorsal root ganglion. The majority of patients with spinal cord root avulsion injury report severe chronic pain at some point in the course of their condition (Woolf, 2004), which, in our previous smaller study of different subjects, was related to the number of avulsed roots (Berman et al., 1998). This "avulsion", or deafferentation, pain is characteristically constant and crushing, usually felt in the hand in patients with brachial plexus injury, with intermittent bursts of pain shooting down the arm. It may occur from the time of injury, or within days, and is often intractable, lasting from months to years, or even decades. Clinical observations first indicated that successful surgical repair was associated with relief of avulsion pain, even in patients in whom repair was delayed for years and there was no prospect of motor recovery (Berman et al., 1996). In a prospective study following plexus repair, a close temporal relationship was noted between reduction in avulsion pain and functional recovery (Berman et al., 1998). It was concluded that nerve repair can reduce pain from spinal root avulsions and that the mechanism could involve successful regeneration, with restoration of peripheral inputs (e.g. from muscle), and/or central connections. Sensory stimulation (thermal and mechanical) of the affected limb in patients with brachial plexus injury is often perceived abnormally, or experienced at remote sites, with, or without, phantom sensations. This is thought to be related to reorganisation of the somatosensory system after deafferentation (Condes-Lara et al., 2000; Flor et al., 1995, 2000; Kaas, 2000; Malin and Winkelmuller, 1985; Moore et al., 2000; Ramachandran and Hirstein, 1998). Although pain and sensory phenomena have been studied thoroughly in patients with post-traumatic neuropathic and phantom pain (Kew et al., 1997; Ramachandran, 1992), there have been few studies which investigate the relationship of pain to referred sensation and cortical reorganisation after deafferentation (Flor et al., 1995; Grusser et al., 2001; Knecht et al., 1995). In this study, we have assessed pain intensity and the recovery of the different sensory modalities using pain questionnaires and quantitative sensory testing after surgical repairs of brachial plexus avulsion injuries, the most severe type of traumatic nerve injuries. We have also recorded and elicited sensory and pain phenomena, including skin hypersensitivity and referred sensations. The objective of the study was to achieve a better understanding of sensory recovery and pain phenomena following brachial plexus avulsion injury and surgery. It is hoped that, in the future, this may lead to the development of new treatments for brachial plexus injuries.
Patients Seventy-six patients, 72 male and 4 female, who had sustained brachial plexus injury with at least two spinal nerve root avulsions, were studied. The mean age of the patients at the time of the injury was 28 (SEM1) years. All patients underwent surgery within 2 months of the injury by the same two surgeons (TC, RB). The patients had been referred to the Peripheral Nerve Injury Unit, Royal National Orthopaedic Hospital, London, UK, and the clinical assessments and neurophysiological studies were performed at the Hammersmith Hospital, London, UK. Written consent and local ethics committee approval were obtained. Patients with associated spinal cord or brain injury, injury to the proximal major blood vessels or double level lesions were excluded. Avulsion or intraspinal injury was confirmed by computerized tomography (CT)-myelography, per-operative electrophysiology and direct intraoperative inspection of the exposed brachial plexus and spinal cord (if applicable). The spinal nerves were either "ruptured" outside the intervertebral foramina or injured proximal to the dorsal root ganglia (DRG), either as an intraspinal root rupture or root avulsion from the spinal cord. Ruptured spinal nerves (i.e. distal to the DRG) were grafted. For intraspinal root ruptures or avulsion from the spinal cord (i.e. central to the DRG), nerve transfers or spinal root reimplantation i.e. reconnection by graft directly into the spinal cord (Carlstedt et al., 1995, 2000), were performed.
Pain history
Sensory testing
Allodynia was defined as pain at perception thresholds that were non-noxious in the intact contralateral limb or in control subjects.
Referred sensation and phantom phenomena Referred sensations were provoked by mechanical, thermal and electrical stimuli. For the mechanical and thermal sensations, the methods and equipment described above were used; for the electrical stimulus, a TENS machine (Model TP 120Z, RDG Medical, Surrey, UK) with a pulse width of 200 µs and a burst mode of maximum 9 pulses per burst, 2 bursts per second, was applied.
Statistics
Fifty-four patients had undergone brachial plexus repair with graft and nerve transfer, 14 patients had reimplantation of avulsed nerve roots and eight patients had no repair at all, on account of late referral to the Unit (Table 1).
The mean time of assessment was 4 (SEM 0.5) years after injury. The earliest first assessment was 4 days post-injury and the latest follow-up assessment was made after 40 years. Thirty patients were assessed more than once. It was not possible to assess most patients prior to surgery as many were referred early after injury and surgical repairs were performed urgently. Pre-operatively, their pain levels varied greatly, depending on severity of the brachial plexus and other injuries, and also the different analgesic treatment strategies of referring hospitals.
The correlation of pain scores to number of root lesions
Severity of pain after different surgical repairs
Sensory recovery Seventy-six patients were studied with quantitative sensory testing at different time points after injury and, of these, 23 patients were studied more than once. Sensory testing in affected dermatomes generally showed very poor or no recovery, with poor or wrong localisation. The best recovery was found at the C5 dermatome, for thermal (see Figs 1 and 2), but not mechanical, stimuli. Pin-prick, cotton wool, proprioception, vibration and monofilament perception thresholds generally remained unchanged, particularly below the elbow.
Allodynia Some patients reported allodynia, provoked by either mechanical or thermal stimuli. Patients could suffer from allodynia provoked by a specific stimulus or from a combination of various stimuli. Allodynia to mechanical and thermal stimuli was observed mainly in border zone of affected and unaffected dermatomes in 18% of patients studied at early (<6 months) and 37% patients at later stages. The percentage of patients with allodynia for each modality recorded at the clinic is shown in Fig 3, and for the sites of allodynia in Fig 4. Some patients experienced allodynia to more than one modality, and at more than one site.
Referred sensation The majority of the patients (56%) reported referred sensations at some point after injury. Forty-four per cent of patients denied that they had ever been aware of referred sensations and these could not be provoked by the investigators. While the intensity of referred sensation was not quantified, these generally became less intense over time and, then, either levelled out or completely disappeared. In a minority of patients (20% of those reporting referred sensations) the intensity of referred sensation did not change or became stronger over time. The relative frequency of patients with different referred sensations, and the time-course of the "wrong-way" and "right-way" referred sensations, are shown in Fig 5. Pain and sensations referred to the original source or target of afferents ("right-way") occurred at a later stage (all >6 months after injury), whereas "wrong-way" referred sensations often occurred early (<6 months in the vast majority of patients with "wrong-way" sensation). "Wrong-way" referred sensations, included "pins and needles" sensation down the affected arm or hand, whilst shaving, touching or tapping the ipsilateral cheek and/or lip. In some cases, thermal stimuli, or electrical stimuli delivered by TENS, could reproduce this (Fig 6). Two patients experienced sensations in the affected arm, or hand, when the ipsilateral leg was stimulated. In another two patients, a referred sensation was experienced in the injured limb when they drank fluids. Most of these sensations were usually described as "pins and needles", "fizzing" or a dull touch sensation. Some patients reported that such referred sensations could be unpleasant, with a painful, or gripping, quality. In one patient, coughing produced a squeezing, gripping feeling at the wrist. Some patients were not aware of the existence of the referred sensation until this was provoked by the investigators.
There was no statistically significant difference of VAS (L), VAS (H), VAS (P), McGill (sensory, S) and McGill (total, T) scores in patients with referred sensation and patients without referred sensation (P values using the Mann–Whitney U test were 0.9341, 0.4463, 0.7067, 0.2712 and 0.6287, respectively).
Neuropathic pain associated with brachial plexus avulsion injury can be very severe, persistent and resistant to treatment (Birch et al., 1998). Our previous small study reported that the severity of the pain was related to the number of spinal root avulsions and that there was a temporal relationship between pain and recovery of function following operative repairs (Berman et al., 1998.) This study, performed in a different and larger cohort of patients, confirms the previous finding of a correlation between the severity of pain and the extent of the nerve injury. Patients that had surgical repairs of the injured brachial plexus had less pain than patients who did not undergo surgery. There are distinctive aspects of brachial plexus injuries which may involve both central and distal axons of sensory neurons at different spinal root levels. Peripheral, as well as central, mechanisms, thus, contribute to the generation of neuropathic pain (Attal and Bouhassira, 1999; Campbell, 2001; Coderre et al, 1993; Jensen et al., 2001; Melzack, 1975; Woolf, 2004). There is evidence that pain experienced after brachial plexus avulsion injury parallels the generation of abnormal activity within the dorsal horn of the spinal cord. Following dorsal rhizotomy in cats, increased burst activity is seen in dorsal horn cells as early as 14 days (Loeser and Ward, 1967). Continuous high frequency activity has also been reported in the cat dorsal horn and, significantly, this type of firing is characteristic when the denervation is due to avulsion rather than rhizotomy (Ovelmen-Levitt, 1988), in accord with clinical observations. This may result from ischaemic damage and death of inhibitory interneurons after avulsion. A high proportion of root avulsion patients report at least transient pain relief following dorsal root entry zone (DREZ) lesion (Nashold et al., 1983), implicating this region in avulsion pain. Pain has been shown to improve in some patients after repair of avulsed roots with reimplantation in our previous studies (Carlstedt et al., 2000, 2004). The mechanism of pain relief in these patients and other patients after successful nerve repair is uncertain. There is evidence that pain improvement may precede recovery of cutaneous sensation or muscle function (Berman et al., 1996, 1998). One possible mechanism could be recovery, or regeneration, of proprioceptors in reinnervated muscles, impulses from which may inhibit or "gate", transmission in the dorsal spinal cord. There are other possibilities, which are speculative but supported by animal models. Studies examining the sprouting of adjacent intact afferents into denervated areas after extensive dorsal root rhizotomies have failed to demonstrate convincing growth into the deafferented region, even although abundant "vacant" synaptic sites are created (McMahon and Kett-White, 1991; Rodin and Kruger, 1984; Rodin et al., 1983). However, inducing afferents into a growth, or regeneration, mode by a peripheral conditioning nerve injury at the time of initial lesion produces an expanded central terminal field representation (McMahon and Kett-White, 1991; Molander and Aldskogius, 1988) and these are able to elicit responses in postsynaptic neurons (Molander et al., 1997). The patients who have both avulsions and ruptures at the time of injury, and "nerve injury" during surgical repairs of intact or injured nerves, may have concomitant central sprouting with the peripheral nerve regeneration, thus leading to amelioration of deafferentation. One-third of the patients experienced allodynia, either mechanical, or thermal, or both. The area of allodynia was at the border zone, usually at the border of the T1 and T2 dermatomes, at the back of the elbow. The allodynia was, generally, not bothersome and minor in comparison with the deafferentation pain. Abnormally sensitised peripheral nociceptive fibres can induce secondary changes in anatomic reorganisation in the dorsal horn and central processing, leading to spinal cord hyper- excitability and allodynia (Baron, 2000a, b). The intact nociceptors of the adjacent uninjured spinal nerves may acquire abnormal spontaneous activity and chemical sensitivity may play a role in creating, or maintaining, an abnormal pain state (Campbell, 2001; Wu et al., 2002). However, ongoing activity in nociceptive systems is not essential to produce mechanical allodynia (Baron and Maier, 1995; Baron and Saguer, 1995). The findings are different after brachial plexus injuries in neonates (Anand and Birch, 2002). While recovery of function after spinal root avulsion was demonstrably related to surgery, there were remarkable differences from adults, including excellent restoration of sensory function in children and evidence of exquisite CNS plasticity, i.e. perfect localisation of restored sensation in avulsed spinal root dermatomes, now presumably routed via nerves that had been transferred from a distant spinal region. Remarkably, there was no evidence of long-term chronic pain behaviour or neuropathic syndromes, although pain was reported normally to external stimuli in unaffected regions. We proposed that CNS plasticity (e.g. sprouting or maturation of descending inhibitory tracts) may account for their lack of long-term chronic pain after spinal root avulsion injury. Sensory recovery was poor in most patients, with or without surgical repair. Only the injured C5 dermatome tended to recover to some extent for thermal sensation. This could be attributed to collateral sprouting of fibres from an overlapping dermatome, rather than regeneration through repair (Bear Mark et al., 2001). These clinical findings are, however, not unexpected, and poor recovery of sensibility after nerve injury is well documented in adults (Birch et al., 1998; Lundborg, 2000, 2003; Richter, 1977). Touch stimuli are often wrongly localised after nerve repair, presumably because regenerated axons fail to return to their original target. The underlying mechanisms of improvement of localisation are unclear and may include cerebral cortex reorganisation (Wall and Kaas, 1986). It is now established that the central nervous system can become reorganised after deafferentation, which may explain the "wrong-way" referred sensations (Kaas et al., 1983). The plasticity of somato-sensory cortex has been shown in animal lesion studies (Florence et al., 1998). The reorganisation of the somatosensory system which may underlie referred sensation has been demonstrated by imaging studies (Kew et al., 1997; Moore et al., 2000). The trunk and hand area representations are situated adjacent to each other (Bear et al., 2001; Jain et al., 1998), as are the face and hand areas and there is evidence that dorsal column lesions result in expansion of inputs from the face and arm into the former hand territory (Halligan et al., 1993; Jain et al., 1998; Polley et al., 1999; Ramachandran et al., 1992). Although the reorganisation of the somatosensory cortex by expansion of the adjacent sensory area of the cortex to the deafferented cortex area may explain such referred sensations, it would not as easily explain some of the visceral referred sensations experienced in our patients, or sensation referred to the deafferented area from the lower limb. However, large-scale reorganisation could occur in different levels, including the thalamus (Banati, 2002; Banati et al., 2001; Grusser et al., 2004; Jain et al., 1998, 2000; Merzenich and Jenkins, 1993; Moore et al., 2000; Pons et al.,1991; Turton and Butler, 2001). There are a number of observations with respect to referred sensation that are unexplained. In our study, almost half the patients experienced no referred sensation – some patients were studied decades after the injury and denied any such sensory experience. In some patients, referred sensation gradually faded away while others continued to experience referred sensations indefinitely, presumably reflecting the dynamic nature of plasticity (Garraghty and Kaas, 1992; Knecht et al., 1998). Some patients experienced referred sensation soon after the injury, suggesting unmasking of preexisting connections (Borsook et al., 1998), whereas, in others, it appeared later and could be due to plasticity as a result of intracortical sprouting (Kaas et al., 1983). A curious sensory phenomenon called "Mitempfindung" (sympathy), in which pain evoked by stimuli is perceived in distant body regions, can occur in healthy individuals, and may reflect aberrant central connectivity (Bean, 1981; Evans, 1976; Richter, 1977; Schott, 1988; Sterling, 1973). Successful peripheral nerve regeneration and lack of central plasticity, on the other hand, could explain the normal or "right-way" referral of sensation, e.g. after intercostal nerve transfers (T3, T4 and T5) to the ulnar nerve. Percussion of the nerve (Tinel’s sign) or stimulation of the ulnar nerve territory of the hand leads to referral of sensation to the T3 and T5 dermatomes in the chest wall, usually months after the operation. There was no clear relationship between chronic neuropathic pain and referred sensation in our patients. The latter was, generally, not painful or superimposed on the spontaneous, or ongoing, pain of the affected limb. A strong correlation has been reported between cortical reorganisation and the magnitude of phantom limb pain, whereas no such correlation was found in non-painful phantom phenomena: there may be different neural substrates which are responsible for painful and non-painful phenomena (Flor et al., 1995, 2000; Grusser et al., 2001, 2004;Knecht et al., 1998). In conclusion, this study confirmed that pain intensity after brachial plexus avulsion injury was related to the number of roots avulsed and that successful surgical repair was associated with pain relief. The recovery of the sensory function remains poor, despite technical advances in surgical nerve repair. CNS plasticity appears to be involved in referred sensations, particularly in the early stages after nerve injury. The use of neurotrophic factors may enhance nerve regeneration and novel surgical strategies now under consideration, such as reimplanting avulsed nerve roots into spinal cord in association with olfactory nerve ensheathing cells, may restore sensation and ameliorate pain. To achieve full functional recovery of sensation after deafferentation, including accurate localisation, strategies that address reorganisation of the CNS may be necessary. Our study provides a basis for comparison of outcomes with such novel treatments.
The authors thank the International Spinal Research Trust (ISRT) for their financial support. Received for publication November 22, 2005. Accepted for publication April 27, 2006.
  CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati What's this?
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
Abstract
PATIENTS AND METHODS
