Abstract
We used end-to-side nerve coaptation combined with standard end-to-end neurotisations to treat 11 patients who presented with complete (six cases) or incomplete (five cases) traumatic brachial plexus injuries. All patients were available for functional evaluation at a minimum of 2 years postoperatively. In three patients with shoulder abduction recovery, electromyographical studies (EMG) showed a contribution from the end-to-side neurotisation. In the remaining cases end-to-side neurotisations were unsuccessful. Our study did not demonstrate a reliable role for end-to-side nerve suture in brachial plexus surgery. We believe that at present end-to-side suture must not be a substitute for standard reconstructive techniques in brachial plexus surgery. Occasionally termino-lateral nerve sutures may represent a support to standard reconstructive procedures especially in case of severe injuries when few undamaged donor nerves are available.
INTRODUCTION
Traumatic lesions of the brachial plexus can be very difficult to treat especially in complete plexus lesions. Different techniques can be used for primary reconstructive surgery depending on the type and site of the nerve lesion. Nerve repair using sutures or grafts has been well documented (Shin et al., 2005). In cases of nerve avulsions, nerve transfers (neurotisations) can be used. These techniques sacrifice healthy nerves that are then transposed and adapted to distal damaged stumps in order to recover the functional loss.
Based on experimental studies published by Viterbo et al. (1992), a growing interest has developed in the use of termino-lateral neurorraphy in order to avoid sacrificing of intact donor nerves. This technique was described originally by Letievant in 1873 (Duplay and Reclus, 1893; Letievant, 1873; Papalia et al., 2007). Several basic research studies have disclosed many of the biological mechanisms by which termino-lateral nerve fibres regenerate (Al-Qattan, 2001; Viterbo et al., 1992; Zhang and Fischer, 2002; Zhao et al., 1997). The colonisation of the recipient nerve originates from the axons of the donor nerve by means of terminal sprouting and collateral sprouting process at the level of the last intact node of Ranvier. The development of these regenerating units from the donor nerve is influenced by nerve factors released from Schwann cells of the recipient nerve and from the distal degenerating nerve stump (Isaacson et al., 1992; Lundborg et al., 1994; Matsumoto et al., 1999; Noah et al., 1997; Papalia et al., 2003; Rowan et al., 2000; Tham and Morrison, 1998). Despite the number of experimental studies, clinical application of end-to-side nerve suture still remains restricted to a small number of patients who presented with peripheral nerve lesions. Traumatic brachial plexus injuries are a possible field of application for this type of nerve repair. End-to-side nerve suture may play a role in the surgical strategy for these severe lesions as it offers the advantage of sparing the donor nerve, unlike in traditional neurotisations. The first description of end-to-side nerve sutures for the treatment of brachial plexus injuries was reported by Harris and Low (1903). The authors applied the technique to three patients (two adults and one child with Erb’s palsy) by suturing part of the fifth to the sixth cervical nerve root. The clinical outcome was poor but some signs of recovery were recorded suggesting the potential benefit of the technique. Following this experience, it is only recently that end-to-side nerve reconstructions have been used in the treatment of brachial plexus injuries but the reports in the literature are sparse and contrasting (Mennen, 2003; Pienaar et al., 2004). Therefore we evaluated a series of patients who received end-to-side neurotisations in our department.
MATERIALS AND METHODS
Between January 2001 and December 2005 37 adults were treated surgically for traumatic lesions of the brachial plexus. Of these 11 had end-to-side neurotisations. The decision to perform end-to-side neurotisations was made by the surgeons, based on their personal experience, evaluation of the lesion, and on availability of nerves for reconstruction. All patients were available for the follow-up at a minimum of 2 years (mean 56.8 months, range, 25 to 92.2) postoperatively.
Patients
There were ten men and one woman with an average age of 33 (range 16–55) years. The right plexus was injured in four patients and the left in seven. In all cases the aetiology was a motorcycle or motorvehicle accident. None of the patients was treated as an emergency because in this series we did not see any sharp open lesions and/or vascular injuries associated with brachial plexus injuries. The mean time from injury to surgery was 3.5 (range 2–6) months.
Preoperatively all patients were assessed clinically, with brachial plexus magnetic resonance imaging (MRI) and with electromyography (EMG). The patients all suffered closed injuries of the brachial plexus. In six cases there was a complete lesion of the brachial plexus (C5-T1) with root avulsions. In one case there was a partial lesion (C6-T1) with root avulsions. In the remaining four patients we observed a partial brachial plexus lesion with root ruptures (two cases C5-C6 – two cases C5-C6-C7) (Table 1). Elbow flexion graded at M3 was present preoperatively in two patients (case 2 and 6) due to forearm muscle function. The trapezius was always preserved and the serratus was preserved (M3) in one patient (case 11).
Surgical technique
All 11 patients underwent reconstruction by means of end-to-end neurotisations combined with end-to-side coaptation. For the end-to-side sutures the donor nerves were: the phrenic nerve in eight cases (in four cases with autologous sural nerve interposition, ranging from 4 to 14 cm), the hypoglossal nerve (XII) in one case (with autologous sural nerve interposition), C5 in one case (with autologous sural nerve interposition) and C7 in one case (Table 1). The recipient nerves were the suprascapular nerve (SSC) in three cases, the axillary nerve in two cases and the medial trunk, upper trunk, lower cord, posterior cord, posterior and medial cords and C6 in one case each (Table 1). In these often complex brachial plexus lesions we tried to use as many techniques as possible to restore some function. The first choice was traditional end-to-end neurotisations with end-to-side nerve neurotisations as a supplement. When possible the ‘single donor nerve – single receiving nerve’ principle was applied for the end-to-side sutures. Shoulder abduction and elbow flexion were considered the main goals of our surgery. End-to-side nerve sutures were almost always performed in order to recover shoulder abduction. To this purpose we used, when possible, suprascapular (SSC) nerve neurotisation by means of spinal accessory nerve and end-to-side nerve sutures to restore axillary nerve. Intercostal transfer to the musculocutaneous nerve or Oberlin procedure (Oberlin, 2000) was used as a first choice to restore elbow flexion.
Clinical assessment
Functional mobility was evaluated at shoulder (S) and elbow (E) level using the Medical Research Council (MRC) evaluation scale (Medical Research Council, 1976). The patients were also assessed subjectively as very satisfied, satisfied, somewhat disappointed or very disappointed with the outcome after surgery using normalised SF-36® Health Survey (v1) scores.
RESULTS
Functional results
The outcomes are summarised in Table 1. We observed a successful (M4) shoulder functional recovery (abduction – external rotation) in two cases, partially successful (M3) in four cases and poor or absent (M0-M1-M2) in five cases. Elbow flexion recovered successfully in one case (M4) but was poor or absent in eight cases of the series (M0-M1). The remaining two patients had M3 elbow flexion preoperatively.
Because all but one patient underwent both end-to-side and end-to-end neurotisations, assessment of the success of the end-to-side neurotisations was difficult. When possible, it was performed with electromyographical study analysing the selective action of deltoid and suprascapularis muscle. In some patients (cases 2, 5 and 6) assessment of the effectiveness of end-to-side neurotisations was not possible. When end-to-side coaptation (hypoglossal to SSC) alone was performed in order to obtain shoulder function we observed M3 degree abduction strength (case 3). In two patients (case 5 and 8) with M3-M4 shoulder abduction recovery, electromyographical studies (EMG) showed a possible contribution from the end-to-side coaptation. In the remaining patients end-to-side neurotisations were unsuccessful.
Subjective results
Six patients were very satisfied (two) or satisfied (four) with their result. Five patients were disappointed (three) or very disappointed (two) with their results. The main reason for dissatisfaction was persisting pain and poor movement. The mean physical functioning SF-36® score was 76.8 points (range, 45 to 100), the mean role-physical score was 72.7 points (range, 25 to 100) while mean bodily pain score was 44.7 points (range, 10 to 100).
DISCUSSION
In recent medical literature the first end-to-side neurotisation for a brachial plexus injury was reported by Viterbo, who sutured C5 and C6 roots to the phrenic nerve (Viterbo et al., 1995). Since then further studies have been published, describing heterogeneous results (Amr and Moharram, 2005; Ferraresi et al., 2002; Haninec et al., 2007; Mennen, 2003; Pienaar et al., 2004; Weigel et al., 2008). The largest case study of end-to-side nerve sutures performed to treat different peripheral nerve injuries was published by Mennen (2003). In this study eight patients underwent brachial plexus repair and seven out of eight were followed up between 18 and 36 months postoperatively. The level of brachial plexus injury was not reported and the associated procedures were not described. The results of motor (deltoid and biceps) and sensory recovery, according to the MRC scale, were moderate (four successful cases – M4; one partially successful – M3; two unsuccessful – M0). These findings were not confirmed in a subsequent study performed by Pienaar et al. (2004) on nine patients with eight traumatic and one obstetric incomplete brachial plexus lesion. In two cases there was partial sensory recovery but no useful motor recovery was observed due to end-to-side neurotisation. Recently Haninec et al. (2007) reported their experience with end-to-side nerve suture in a homogeneous series of incomplete brachial plexus injuries. They employed intraplexual donor nerves (ulnar – median – radial) and the axillary nerve as a recipient reporting a motor recovery of the deltoid muscle in 64% of patients (nine out of 14).
Our study did not demonstrate significant benefit from end-to-side neurotisation in primary reconstructive surgery for traumatic closed adult brachial plexus injuries. Although in one patient with a complete brachial plexus injury isolated hypoglossal to SSC end-to-side neurotisation resulted in the recovery of some active shoulder abduction (M3) with a typical lingual synkinesia confirming the procedure as a potential source of regenerating axons in a recipient nerve. This observation is also interesting as it contradicts the results of previous studies in which the hypoglossal nerve seemed to be an unsuitable donor for termino-lateral nerve suture in brachial plexus repair (Ferraresi et al., 2002; Malessy et al., 1999).
The usefulness of end-to-side nerve suture in the repair of brachial plexus injury is still questionable. The potential of the technique has been confirmed by several experimental studies thus opening interesting perspectives for clinical use of this neurorraphy in brachial plexus injuries since the procedure does not require the sacrifice of any donor nerves. In some reports the results are promising but in most cases, such as here, the results are often poor or at best unclear. In our opinion this technique must not be a substitute for standard neurotisations which are more reliable in adult brachial plexus surgery. Occasionally end-to-side coaptation may support standard reconstructive procedures especially in cases of severe brachial plexus injuries when few undamaged donor nerves are available.