We searched for prognostic factors which could influence outcome after surgery for traumatic lesion of the axillary nerve.

Forty-five surgical interventions to repair injured axillary nerves were preformed between 1993 and 2000. We analyzed outcome at a mean 56 months (range 15-96). Twenty-five isolated lesions were treated by nerve graft (n = 20), direct suture (n = 2) or neurolysis (n = 3). Four associated axillary and musculocutaneous nerve injuries were treated by axillary graft and musculocutaneous neurolysis (n = 2) or double grafts (n = 2). Eleven injuries involving both the axillary and suprascapular nerve were treated by double nerve graft (n = 4), axillary graft with suprascapular neurolysis (n = 5), or axillary graft with an irreparable suprascapulary injury (n = 2). Five axillary nerve lesions were associated with lesions of the rotary cuff; treatment associated suture of the cuff (n = 3) or reinsertion fixation of the tuberosities (n = 2) prior to nerve repair by axillary graft (n = 4) or neurolysis (n = 1).

For the isolated axillary lesions, results were very good or good for 16/20 nerve grafts, 2/2 direct sutures and 2/3 neurolyses. For the patients with an associated musculocutaneous lesion, shoulder function was considered very good for one; mean elbow flexion strength was 29% (15-50%) of the healthy side. For the eleven axillary and suprascapular injuries, outcome was very good or good for two. Very good or good results were not achieved for any of the five patients with associated cuff lesions. Factors predictive of poor outcome were a preoperative Constant score below 40 points, age over 40 years, time to operation greater than 15 months, and multiple nerve or associated cuff injury.

The favorable prognosis of isolated lesions of the axillary nerve was confirmed. The risk of failure does however persist and is related to late management despite well defined surgical technique.