| Sign In to gain access to subscriptions and/or personal tools. |
DOI: 10.1177/1753193408087029
Age-Dependent Development Of Chronic Neuropathic Pain, Allodynia and Sensory Recovery after Upper Limb Nerve Injury in ChildrenFrom the Peripheral Neuropathy Unit, Imperial College London, Hammersmith Hospital, London, UK and the St Andrew’s Centre for Plastic Surgery, Broomfield Hospital, Chelmsford, UK Correspondence: Professor Praveen Anand, Imperial College London, London, UK., E-mail:p.anand{at}imperial.ac.uk.
Forty-nine children with distal upper limb nerve injury were studied at a mean follow-up of 2 years 3 months. Patients who were aged 5 years or younger at the time of nerve injury (15/49) had no chronic neuropathic pain symptoms or allodynia. Patients with allodynia on quantitative sensory testing but no spontaneous pain (8/49) were all older than 5 years and those reporting spontaneous chronic neuropathic pain (5/49) were all older than 12 years at the time of injury. Previous studies of adults with similar nerve injuries report chronic hyperaesthesia in up to 40% of cases. Semmes–Weinstein monofilament testing showed a positive correlation between age at injury and abnormal sensory threshold (r = 0.60, P<0.0001). These findings indicate that young children show better sensory recovery and are less likely to develop long-term chronic neuropathic pain syndromes than adults following nerve injury.
Key Words: neuropathic pain • nerve injury • children • sensory recovery Traumatic peripheral nerve injuries are associated with a range of unpleasant sensations, including spontaneous pain and cold intolerance. In adults, chronic pain following peripheral nerve injury occurs in about 5% of cases (see Sunderland, 1993). The spontaneous pain may be associated with allodynia (painful perception of a non-noxious stimulus) to mechanical or thermal stimuli, dysaesthesia and hyperpathia (Birch and Anand, 2003; Sood and Elliot, 1998). In a follow-up study of digital nerve repair in adults, 40% complained of significant hyperaesthesia for up to 2 years after surgery (Goldie et al., 1992). Although clinicians recognise that these problems rarely occur in children, the literature is almost devoid of information on this subject in this age group. Our previous studies in children with obstetric brachial plexus injuries showed excellent sensory recovery following surgical repair, and lack of long-term chronic neuropathic pain syndromes, in comparison with adults (Anand and Birch, 2002). Sensory recovery after successful surgical repair also appeared to be associated with some reduction of neuropathic pain in adults with brachial plexus injury (Berman et al., 1998). Cold intolerance is a relatively common symptom after peripheral nerve injury, variously described as ‘‘an icy cold feeling that can progress to pain’’ (Engkvist et al., 1985) and as ‘‘cold-associated symptoms’’ (Campbell and Kay, 1998). The incidence in adults varies in different studies – 83% (Irwin et al., 1997), 73% (Campbell and Kay, 1998), 67% (Craigen et al., 1999) and 53% (Lithell et al., 1997). More descriptive studies (Cheng et al., 1998; Faivre et al., 2003) have not conducted quantitative sensory testing (QST) to assess cool allodynia, viz. a painful perception of a non-noxious cool temperature stimulus applied to the affected skin region. Skin contact cooling-evoked pain is not always present in patients with symptomatic cold intolerance, indicating complex mechanisms in the latter, including sympathetic overactivity. Again, there is little information on thermal intolerance and allodynia in children. We have studied 49 children seen consecutively at a mean of 2 years and 3 months after upper limb peripheral nerve injury, to determine the incidence of chronic pain and hypersensitivity. Symptoms of pain and hypersensitivity were recorded according to a system described previously in adults (Hazari and Elliot, 2004; Sood and Elliot, 1998), where pain following peripheral nerve injury is subdivided on clinical grounds. We also performed QST to investigate allodynia and sensory recovery (Anand and Birch, 2002).
Over the period 2003 to 2005, 49 patients, 29 boys and 20 girls, were studied in an out-patient clinic, with informed parental consent and local Research Ethics Committee approval. The age at injury ranged from 1 month to 18 years (mean 8 years 5 months) and the age at follow-up ranged from 2 years 3 months to 25 years (mean 10 years 9 months). The clinical features of the injuries are listed in Table 1. Three of the nerve injuries were as a result of a bone fracture, nine were caused by a knife or sharp metal, 20 were crush injuries and 17 were caused by glass. Three patients’ nerves were divided in the upper arm, two in the forearm, nine at the wrist, six in the palm and 29 were finger injuries. As a result of these injuries, there were 11 median, 12 ulnar, one dorsal branch of the ulnar nerve, two radial and 33 digital nerve injuries.
The type of injury ranged from a full set of median, ulnar and radial nerve injuries in the forearm or arm to crush/avulsion injuries of the digital tips. In the latter, terminal branches of the digital nerves were always divided and no formal nerve repair is carried out, leaving the patient at risk of developing two end-neuromas. After more proximal nerve divisions, nerve coaptation was always performed in the standard way, using interrupted 8/0, 9/0 or 10/0 Ethilon sutures under loupe or microscope magnification.
Assessment of pain and hypersensitivity Sensory testing was performed at a mean follow-up time of 2 years 3 months (range 6 months to 9 years 3 months). Sensory testing was performed using methods previously described in children with obstetric brachial plexus injury (Anand and Birch, 2002). These methods of analysis are summarised below. The normal values for these tests have also been obtained from the same paper (Anand and Birch, 2002). The feasibility and reproducibility of QST in children as young as 3 to 4 years has also previously been demonstrated by Hilz et al. (1996). The full set of QST was possible in 34 children, with more limited QST in 11 children who were under 4 years of age. For these children under 4 years, it was still possible to demonstrate the presence, or absence, of hypersensitivity or allodynia by the application of punctate or cotton wool stimuli as described below, and by phrasing the question in terms understandable to a young child.
Cotton wool and pinprick sensation
Thermal perception thresholds
Monofilament perception threshold Graphs were created and statistical tests were performed using GraphPad Prism version 3.02 for Windows (GraphPad Software, San Diego, California, USA).
Assessment of pain and hypersensitivity The symptoms and results of objective testing are summarised in Table 2. Forty-four of the 49 patients, or their parents, reported no pain-related symptoms. Five complained of pain (age at nerve injury range 12–17 years). Two had ongoing spontaneous basal pain and one of these also reported spikes of spontaneous pain. Three complained of pain on pressure over the site of nerve injury and four complained of pain related to movement. One of these five children also complained of symptoms of cold intolerance.
When the pain scores, clinical examination findings and results of QST were reviewed, it appeared that the children could be divided into three groups (see Fig 1): Group A – those without chronic pain or hypersensitivity (age range 1–15 years, n = 36); Group B – those without chronic pain but with hypersensitivity, or allodynia, on testing (age range 5–18 years, n = 8); and Group C – those with chronic spontaneous pain, with or without hypersensitivity or allodynia (age range 12–17 years, n = 5).
Of the eight patients with hypersensitivity and no spontaneous pain, two were found to be hypersensitive to pinprick, four had cool allodynia and four had warm allodynia, when tested using thermal perception thresholds. Of the five patients with spontaneous pain, one had monofilament allodynia/pinprick hypersensitivity.
Thermal perception thresholds
Monofilament perception threshold Fig 4 shows the age at injury plotted against the lowest monofilament detectable in the distribution of the damaged nerve(s). Twenty-one of 39 (54%) recorded subnormal detection thresholds. The abnormal monofilament values ranged from 0.0794g (monofilament no. 4) to 12.6g (monofilament no. 14), demonstrating a partial return of sensation in all nerves. There was a correlation between increasing age at injury and poorer sensory recovery for the monofilament detection threshold (rs = 0.60, P = < 0.0001). As expected, the recovery was better in the more distal injuries. Injuries distal to the distal wrist crease (n = 25) had normal monofilament tests in 56% of cases and injuries in the finger (n = 21) had normal monofilament tests in 76% of cases. The correlation between increasing age at injury and poorer sensory recovery for monofilament detection threshold was still significant for injuries distal to the distal wrist crease (rs = 0.52, P = 0.0075) and injuries in the finger (rs = 0.44, P = 0.0457).
Traumatic peripheral nerve injuries in adults are associated with a range of chronic unpleasant sensations, including spontaneous pain, allodynia and cold intolerance. This study suggests that very young children are less likely to suffer from these chronic pain phenomena. This difference between adults and children has been demonstrated previously following brachial plexus injury. Berman et al. (1998) described adult patients treated by nerve transfer or grafting for brachial plexus injury. All had pain pre-operatively; postoperatively, 50% showed no pain improvement. Severe chronic pain is also usual in adults who have not undergone surgery for spinal root avulsions (Parry, 1984), often continuing for many years or even decades. In contrast, Anand and Birch (2002) reported the results of brachial plexus injury and repair in neonates in a subsequent paper; no child had any symptoms of long-term chronic neuropathic pain or hypersensitivity. In this study, no patient of 5 years or less at the time of nerve injury appeared to have chronic neuropathic pain or hypersensitivity symptoms or signs. Some patients between the ages of five and 12 years exhibited symptoms and signs of hypersensitivity (allodynia), but did not report chronic spontaneous pain. All the patients with spontaneous chronic neuropathic pain as well as hypersensitivity in this study were 12 or more years of age at the time of nerve injury. A number of authors have reported on the incidence of cold intolerance after injury to the limbs in adults, and specifically after peripheral nerve injury (Irwin et al., 1997). Irwin and his colleagues reported an 83% incidence of this complaint in 398 patients. Campbell and Kay (1998) reported that 73% of patients with hand injuries described cold-related symptoms, of which pain was the most frequent. Craigen et al. (1999) reported that 67% of patients with hand injuries had cold-related symptoms. In a Swedish study of 40 patients between 20 and 67 years of age with digital injuries, Lithell et al. (1997) found that all had reduced tolerance to cold, but only 53% described cold-induced pain. In more temperate Istanbul, 64% of median nerve injury patients complained of similar symptoms postoperatively (Polatkan et al., 1998). The incidence of cold intolerance after digital replantation is equally high (Craigen et al., 1999; Elliot et al., 1997; Gelberman et al., 1978; Tark et al., 1989). However, in all of these studies, assessment was based on a history of symptoms and QST was not performed. There are few reports of cold intolerance in children. Cheng et al. (1998) reported that 40% of 26 children, injured and treated by replantation between the ages of 14 months and 12 years, complained of cold intolerance. In a series of eight children who had distal digital replantation at a mean age of 9 years, Faivre et al. (2003) reported that two complained of cold intolerance and four complained of either dysaethesiae or paraesthesiae in the fingertip. Our findings that only three of 49 children under 18 years of age complained of cold intolerance, and four of 34 tested demonstrated cool allodynia on QST, support the previous observations of a lower incidence of cold intolerance in children. Although young children do develop chronic pain in association with other pathologies such as surgery, cancer and arthritis (Anand and Carr, 1989), the lack of long-term chronic neuropathic pain following peripheral nerve injury described above is in accordance with previous observations in brachial plexus injury (Anand, 1992; Birch and Anand, 2003; Birch et al., 1998). The mechanisms responsible for the differences in the perpetuation of pain at different ages are not understood. A number of possible peripheral and central nervous system mechanisms have been suggested. Sensory conduction velocities in humans have been shown to reach adult values by 5 years of age (Garcia et al., 2000). This coincides with the age at which chronic hypersensitivity was reported in this study. Pain-related behaviour has been studied in rats using the sciatic nerve axotomy model, which leads to autotomy (self-mutilation) (Wall et al., 1979). In rat peripheral nerves, sodium channels show increases and clustering in the postnatal development period of the nervous system (Vabnick et al., 1996; Vabnick and Shrager, 1998; Safronov et al., 1999), and the development to an adult distribution coincides with both autotomy (‘‘self-mutilation’’) after nerve injury (at approximately 3 weeks) and adult rat nerve conduction velocities. Human axon ion channels are essentially similar to other species (Scholz et al., 1993). Yiangou et al. (2000) have shown that injured nerves from human neonates lack a particular sensory neurone-specific sodium channel (SNS/NaN 180 kDa subunit) that is notably present in neuromas from adult patients. Some pain phenomena may, thus, require the presence and redistribution of appropriate sodium channels, which may vary during postnatal development of the peripheral nervous system. Self-mutilation has been described in children with obstetric brachial plexus injury, in 3.9% (McCann et al., 2004) and 4.7% (Al-Qattan, 1999) of cases. However, this is more likely to reflect lack of protective sensation, rather than a response to pain, as it is much more common in children with congenital insensitivity to pain. Our patients with obstetric brachial plexus injuries rarely showed such behaviour, and denied that this was associated with any pain. The parents regarded this as a ‘‘bad habit’’ that led to injury because of lack of nociception and, in all cases, was ameliorated by education and maturity. Of course, this does not preclude the presence of positive phenomena, or neuropathic pain, in pre-verbal children. Central nervous system development, and associated structural or functional re-organisation, could also potentially explain the symptomatic differences after nerve injury between young children and adults. Neuromagnetic source imaging, used to assess cortical reorganisation, has shown low levels of reorganisation in congenital amputees, which is not significantly different from that of traumatic amputees without phantom pain (Flor et al., 1998). In contrast, traumatic amputees with pain do show substantial cortical reorganisation. Our findings support those of previous authors that sensory outcome after nerve injury and/or repair in younger patients is better than that in adults (Efstathopoulos et al., 1995; Polatkan et al., 1998; Weinzweig et al., 2000). Weinzweig et al. (2000) demonstrated better recovery of two-point discrimination (2PD) after digital nerve repair in those below 40 years of age and, in particular, noted that 13 of 15 patients younger than 11 gained what they termed good to excellent results (less than, or equal to, 10 mm 2PD), compared to an average of 11.2 mm in the over 40 years age group. Lundborg and Rosen (2001) have suggested that improved functional outcome in children is based on a functional remodelling of the projected areas of the injured nerve in the somatosensory cortex, and that this learning process is optimal between 5 and 10 years of age. Interestingly, in contrast to the findings in terms of mechanical sensitivity, we found no significant correlation between age and thermal perception thresholds. The reason for this is unclear, but may be related to either differential sensory fibre recovery after nerve injury, or CNS remodelling, in these young patients. There is evidence that sensory recovery may be associated with amelioration of neuropathic pain in adults with brachial plexus injury (Berman et al., 1998). Age-related mechanisms of pain and regeneration after nerve injury deserve further investigation. Understanding these mechanisms may provide new strategies for improving sensory recovery and the treatment of adult neuropathic pain syndromes.
We would like to thank GlaxoSmithKline PLC for supporting D.D.A. and O.T., and the Chelmsford Medical Education and Research Trust for funding for the thermal threshold testing equipment. Received for publication November 22, 2006. Accepted for publication August 9, 2007.
  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
