| Sign In to gain access to subscriptions and/or personal tools. |
DOI: 10.1016/j.jhsb.2006.08.019
Enhanced Sensory Recovery after Median Nerve Epair Using Cortical Audio–Tactile Interaction. A Randomised Multicentre StudyFrom the Department of Hand Surgery, Lund University, Malmö University Hospital, SE-205 02 Malmö, Sweden Correspondence: Birgitta Rosén OT PhD, Department of Hand Surgery, Lund University, Malmö University Hospital, SE-205 02 Malmö, Sweden, Tel.: +46 40 331 934; fax: +46 40 928 855. E-mail:birgitta.rosen{at}hand.mas.lu.se
The "Sensor Glove System" offers an alternate afferent inflow from the hand early after nerve repair in the forearm, mediated through the hearing sense, implying that deprivation of one sense can be compensated by another sense. This sensory "by-pass" was used early after repair of the median nerve with the intention of improving recovery of functional sensibility by maintaining an active sensory map of the hand in the somatosensory cortex during the deafferentation period. In a prospective multicentre clinical study, one group (n = 14) started early after surgery with sensory re-education using the Sensor Glove System and the control group (n = 12) received conventional sensory re-education, starting 3 months postoperatively. The patients were checked regularly during a 1-year period, with focus on recovery of tactile gnosis. After 12, months, tactile gnosis was significantly better in the Sensor Glove System group. This highlights the timing for introduction of training after nerve repair, focusing on the importance of immediate sensory re-learning.
Key Words: nerve repair • sensory re-education • sense substitution • brain plasticity • cross-modality • outcome Recent advances in neuroscience and cognitive science have opened new possibilities for the future to improve sensory recovery after nerve repair, especially with respect to functional sensibility and, specifically, the capacity for identification and discrimination of touch (Rosén et al., 2003). In classical sensory re-education, nothing is done to the denervated hand and the de-afferented brain during the first months after nerve repair. Sensory re-education programmes are started when some perception of touch can be demonstrated in the distal palm, i.e. about 3 months after nerve repair at the wrist level (Dellon et al., 1974; Wynn-Parry and Salter, 1976). The insensate hand and the changes in the corresponding cerebral cortical areas which occur after injury are left unattended, from the sensory relearning point of view, for a time period of several months. Within minutes after a deafferentation injury, such as amputation of an arm or major nerve injury, there is a cortical response with an immediate and long-standing re-organisation of the sensory brain cortex. The silent area, no longer receiving any sensory input, triggers an expansion and invasion from adjacent cortical areas (Kaas et al., 1983; Merzenich and Jenkins, 1993; Wall et al., 2002). This is the initiation of a dynamic interplay in the cortical neural networks, which is influenced by several biological and psychological events during regeneration and reinnervation. The outcome of nerve division then repair, in terms of recovered tactile gnosis, in adults is often disappointing (Allan, 2000; Jaquet et al., 2001; Jerosch-Herold, 1993; Kallio and Vastamäki, 1993; Lundborg et al., 2004). We think that one reason for this is the long initial period of absent sensibility, which allows major functional cortical reorganisation changes to take place as a result of lost sensory input initially and misdirected axonal outgrowth later: the "cortical hand map" is completely changed. The timing for onset of sensory re-education may be of critical importance. We should differentiate between ‘Phase 1’(before any reinnervation has occurred in the hand) and ‘Phase 2’ (when some reinnervation of the hand has occurred). There are good reasons to use strategies to enhance the recovery during both of these two phases and to initiate sensory re-education very early, i.e. in Phase 1, during the first postoperative days. The brain is organised holistically, with an extensive capacity for cross- and multi-modality and there is an ongoing, activity-dependent competition for "brain space" between different sensory inputs. The use of vision to guide the re-training of sensation is the base for classical sensory re-education, but there is a continuous interplay between all of the senses. This multi- and cross-modal activity of the brain is based on multi-sensory neurons that receive more than one type of sensory signals and it has been demonstrated that we are able to extract information from one sensory modality and use it in another by using polymodal association centres (Bavelier and Neville, 2002; Pascual-Leone and Hamilton, 2001; Tanabe et al., 2005). This holistic concept, with functional interdependence of activity between the different areas of the brain, makes a reevaluation of the traditional territorial concept of the brain necessary. It also opens up new possibilities to use the plastic potential of the brain at a very much earlier stage in rehabilitation after nerve repair. Cortical audio–tactile interaction has been reported in animal and human studies (Gobbele et al., 2003; Lutkenhoner et al., 2002) and we have presented a model for alternative sensibility, based on sense substitution, using hearing as a substitute for sensibility (Lundborg et al., 1999). Miniature microphones are mounted in the fingertips of a glove, or attached dorsally with a silicone ring directly onto the finger (Figs 1a and b), in what is called a ‘Sensor Glove System’. The stimuli generated by active touch of various structures (each structure giving a specific friction sound) can be picked up, amplified and transposed to stereophonic acoustic stimuli by this system. Using the Sensor Glove System, it is possible to train the brain to localise different fingers and identify different textures, allowing use of this alternative sensory feedback for activities of daily living. This principle is used to provide the sensory brain cortex with an alternate sensory input at a time when regenerating nerve fibres have not yet reached the peripheral targets. We have recently, with fMRI technique, demonstrated an audio–tactile interaction in persons trained with the Sensor Glove System (Lundborg et al., 2005).
In this controlled, randomised study, subjects were equipped with either a Sensor Glove System within 2 weeks after surgery and underwent a sensory re-learning programme at this time, or received traditional treatment with sensory re-education starting when some evidence of reinnervation was present in the hand.
The study design was a prospective randomised multi-centre study including six hand centres in Sweden (Göteborg, Linköping, Malmö, Stockholm, Uppsala, Örebro), and was approved by the Ethical Committees at Lund, Uppsala, Linköping, Örebro, Göteborg and Stockholm Universities. The study included 30 consecutive patients over 18 years who were less than 2 weeks from a complete, clean-cut transection of the median or combined median/ulnar nerves at the wrist or distal forearm level. All patients had given their approved consent. Communication problems due to language or severe psychiatric problems were exclusion criteria. Using sealed envelopes, patients were randomised to receive therapy treatment postoperatively with either the Sensor Glove System early after injury and surgery or conventional sensory re-education training. Table 1 includes the demographic data of the two study groups.
Twenty-six of the initial 30 patients completed the study, and 14 of these 26 patients (mean age 35 years) used the Sensor Glove System while carrying out conventional rehabilitation following nerve repair, including specific sensory re-education exercises twice daily, from the first postoperative week through the first 3 postoperative months. Twelve patients (mean age 33 years) were randomised to conventional rehabilitation during the first 3 postoperative months. At the 3 months follow-up, all received information and a home programme about conventional sensory re-education (Dellon, 1981).
Sensory re-education procedures
Conventional training
Follow-up and assessment
Assessments were performed according to standardised procedures, and s2PD testing was carried out according to the "Moberg Method" (Moberg, 1990), as described by the ASSH and ASHT (ASHT, 1992; ASSH, 1978). The test is carried out in a descending order, starting with 15 mm, to assess the level at which responses were correct (7 out of 10 correct at just blanching of the skin), and were quantified as 0 to 3 (0 =
Analysis
At the initial assessment, both groups naturally started at zero tactile gnosis (Table 2). The median improvement in STI-testing after 12 months in the patients that had used the Sensor Glove System was 2 (IQR 0–3.25) indicating a true change (score > 1.3 in STI-test) from baseline in tactile gnosis (Rosén, 2003). This compares with the control group that showed a median improvement of 0 (IQR 0–0.75) at this time.
Outcome in the sensory, motor and pain/discomfort domains and the "total score" from the three domains are shown in Table 2. No differences between the groups could be seen after 12 months in "total score" or in the motor domain. Neither were there any differences between the groups in experienced pain/discomfort. Tactile gnosis outcome was specifically addressed in this study and group comparison (Mann–Whitney U-test) at the 12 months follow-up showed significantly better tactile gnosis (P = 0.008), as expressed with STI-testing, in the Sensor Glove System group (Table 2 and Fig 2). Two-point discrimination did not demonstrate any difference between the groups. A clear floor effect (most patients could not discriminate between one and two points at 15 mm distance) was seen in both groups with test result 16 mm or more in 10 cases in the Sensor Glove System group and in nine cases in the conventional training group (Fig 3).
The functional reorganisation of cortex after nerve transsection and repair is a rapid process, involving disappearance of the representation corresponding to the denervated body parts and associated expansion of adjacent cortical territories. In such a situation, it is hoped that activation of cortical hand representation might be effective in maintaining the cortical hand map. It is well known that the premotor cortex can be activated by just imaging a movement, so-called "motor imagery" (Jeannerod, 1994; Jeannerod and Frak, 1999; Kosslyn et al., 2001; Lotze et al., 1999). It has been demonstrated that the pattern of somatosensory activation during motor imagery is very similar to the pattern observed during movement execution (Ehrsson et al., 2003). Also, observing movements activates mirror neurons in the frontal cortex (Celnik et al., 2005; Rizzolatti and Craighero, 2004; Rizzolatti et al., 1996). Mirror neuron areas have also been shown to be involved in understanding the intention of others (Iacoboni et al., 2005). Reading or listening to action words, related to hand movements, may activate hand representational areas in the motor cortex (Hauk et al., 2004; Pulvermuller, 2005). In respect of ‘sensory imagery’ and involvement of primary sensory cortical areas, a few observations have been reported (Yoo et al., 2003). There are also ways to activate the somatosensory cortex by, for example, observing a body part (Keysers et al., 2004) or the hands (Hansson et al., 2005) being touched. Several studies also suggests that the SI and SII cortex is related to the mirror neuron system (Avikainen et al., 2002; Mottonen et al., 2005; Yoo et al., 2003). Another possible principle is activation of the somato-sensory cortex using a mirror placed transversally in front of the patient with the denervated nerve injured hand hidden behind the mirror and the healthy hand being reflected as an illusion of the injured hand (Moseley, 2004; Ramachandran and Hirstein, 1998; Rosén and Lundborg, 2005). Touching the healthy hand in this situation may give an illusion of touching the injured hand. Whatever method is chosen there are good reasons to start sensory re–education after nerve repair much earlier than we do today, with the aim of inhibiting, or at least minimising, the reorganisation process in the somatosensory cortex which is induced by the nerve injury. In the very early phase after nerve injury and repair, all of these principles may constitute potential methods to feed the somatosensory cortex with input from the denervated body part. In this study, we utilised the brain’s capacity for audio–tactile interaction so that acoustic information was used as a substitute for missing tactile information (Lundborg et al., 1999). Cortical audio–tactile interaction has previously been reported in animal and human studies (Gobbele et al., 2003; Lutkenhoner et al., 2002). The principle is based on the crossmodal capacity of the brain, i.e. hearing substitutes for touch. The resemblance in perceptual experience between sound and touch is bridged by the vibratory sense. The Sensor Glove, by using audio–tactile interaction, facilitates relearning once sensation returns to the hand, by maintaining a hypothetically better prepared somato-sensory cortex for the necessary re-learning process. Our hypothesis was confirmed that maintaining activation of the cortical hand maps, i.e. preserving cortical hand representation in Phase 1 after the nerve repair (when there was no sensibility in the hand), would facilitate later recovery of functional sensibility (tactile gnosis). Tactile gnosis is one of the components of importance for the summarised outcome after nerve repair (Rosén and Lundborg, 2000). We, therefore, find it encouraging that the tactile gnosis assessment demonstrated improvement, even when use of the Sensor Glove had no impact on the "total score", which also includes other components such as motor function and pain problems. In the present study, training with the Sensor Glove System started within the first postoperative days, i.e. in a phase of ongoing profound cortical reorganisation, when the cortical hand projection is diminishing, or disappearing, as a result of expanding adjacent cortical areas. A recent fMRI study has shown that acoustic stimuli from the hand, processed by the Sensor Glove System, can activate the somatosensory cortex in healthy individuals who have trained with the equipment (Lundborg et al., 2005) and we hypothesise that this is the case also in nerve injured patients. The findings in this study support such a hypothesis. The Sensor Glove System was used until reinnervation of the hand was obvious. It is not known whether continuing use of the system in Phase 2 would be beneficial. Combination of methods to activate the sensory cortex during Phase 1 should also be considered. The Sensor Glove System may also have a use in patients lacking sensibility after lesions in the central nervous system with disturbed body awareness or due to neurological disease. In respect of nerve injury, our study highlights the need for refinement of sensory re-education programmes, with emphasis on the timing of such programmes in rehabilitation after nerve repair.
This study was supported by grants from the Swedish Medical Research Council, project no 5188, The Swedish Brain Foundation, Torsten och Ragnar Söderbergs Stiftelse, the Faculty of Medicine Lund University and Malmö University Hospital. Special thanks go to the therapists at the Hand Surgery Units in Örebro, Stockholm, Uppsala, Linköping, Göteborg and Malmö who took part in the study. Received for publication January 13, 2006. Accepted for publication August 26, 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
16 mm, 1 = 11–15 mm, 2 = 6–10 mm, 3 = <6 mm) (
