Our unit has pursued a policy of using donor nerves from the same limb for grafting. Nerves which have already been affected by the primary injury are selected where possible, thus avoiding any new sensory deficit.

36 of the 41 brachial plexus repairs were available for outcome data collected prospectively over 2 years. Over a nine year period, donor nerves used for the 41 brachial plexus repairs included the lateral cutaneous nerve of the forearm, superficial radial, medial cutaneous of the forearm, ulnar and sural nerves. Patients were grouped into having injured nerve grafts only (A), injured and uninjured nerve grafts (B) and uninjured nerve grafts. The repaired brachial plexus nerves were assessed by measuring the MRC grading of the power of movement of the muscle innervated by that nerve (i.e. elbow flexion for musculocutaneous nerve). These were graded as good (MRC grading 3 or better), fair (MRC grade 1 or 2), or poor (MRC 0).

The greatest success for nerve grafting was elbow flexion with good results in 22 out of 27 assessments. Using Mann-Whitney test, Group A had significantly better results (p=0.025) than group C. However, ignoring the poorer results of shoulder abduction there was no significant difference between all 3 groups of patients.

We conclude that using injured nerve grafts taken distal to the lesion in the brachial plexus is as effective as using nerve material from an uninjured limb.

Our unit has pursued a policy of using donor nerves from the same limb for grafting. Nerves which have already been affected by the primary injury are selected where possible, thus avoiding any new sensory deficit.

Methods

36 of the 41 brachial plexus repairs were available for outcome data collected prospectively over 2 years. Over a nine year period, donor nerves used for the 41 brachial plexus repairs included the lateral cutaneous nerve of the forearm, superficial radial, medial cutaneous of the forearm, ulnar and sural nerves. Patients were grouped into having injured nerve grafts only (A), injured and uninjured nerve grafts (B) and uninjured nerve grafts. The repaired brachial plexus nerves were assessed by measuring the MRC grading of the power of movement of the muscle innervated by that nerve (i.e. elbow flexion for musculocutaneous nerve). These were graded as good (MRC grading 3 or better), fair (MRC grade 1 or 2), or poor (MRC 0).

Results

The greatest success for nerve grafting was elbow flexion with good results in 22 out of 27 assessments. Using Mann-Whitney test, Group A had significantly better results (p=0.025) than group C. However, ignoring the poorer results of shoulder abduction there was no significant difference between all 3 groups of patients.

In patients with DM (Diabetes Mellitus types I & II), primary frozen shoulders tend to be refractory to all forms of treatment. We collected tissue from the joint capsule of shoulder joints from a variety of patients undergoing surgery as follows:

Diabetic Group (DFS): patients with DM who have primary frozen shoulders.

Other patients suffering from primary frozen shoulders (FS)

Control group (NS). Patients undergoing shoulder surgery that does not involve stiffness of the gleno-humeral joint.

Tissue was collected from near to the rotator interval under arthroscopic control. Fibroblast lines were established by serial passage. Thereafter they were exposed to graded concentrations of insulin in vitro for 24 hours and the supernatant retained for assay. Fibroblast lines were analysed from 3 subjects in each group (n=9). Luminex multiplex analysis was performed for MMPs (Matrix Metalloproteinases). TIMP-1 (Tissue Inhibitor of MetalloProteinases) expression. Informed consent was obtained from all subjects.

Results: Production of MMP 1,2,3 and 8 by fibroblast lines were distinct between patient groups. MMP-1 production in DFS (mean 716pg/mL) was significantly reduced compared to FS derived patient cells (mean 972pg/mL) (p=0.0138, Mann-Whitney Test). Moreover, striking differences were observed when fibroblasts from DFS patients were compared with those from NS controls (mean 5898pg/mL) (p< 0.000). Calculating MMP-1/TIMP-1 ratios revealed significantly lower ratios in DFS (2597), or FS (2860) compared with NS (24,326) (p < 0.001). There was no significant difference between ratios of MMP1/TIMP1 in DFS and FS (p=0.977). MMPs 7,9,12 and 13 were not detected in any of the samples.

This is the first time these enzymes have been measured and quantified in cells derived from shoulder tissues. Primary Frozen Shoulders produce less MMPs and have a smaller MMP/TIMP ratio than controls. Similarly the diabetic patient derived cells produce less MMP-1, at an even lower level. These deficiencies in MMP1 production may reflect an altered capacity for local tissue re-modelling. MMP modulation may allow therapeutic intervention in the diabetic and frozen shoulder group of patients.