[Fusion and function after eight scapulothoracic arthrodesis]

Scapula alata resulting from a deficient musculus serratus anterior leads to shoulder instability, pain and loss of elevation. Etiologies include fascioscapulohumeral dystrophy and lesion of the thoracicus longus nerve. Dynamic (muscle transfer) or static (scapulopexia or scapulothoracic arthrodesis) stabilization methods can be proposed. The purpose of this study was to assess fusion and function after eight scapulothoracic arthrodeses performed in five patients. We used cerclage with compression after intercostal avivement to increase the contact surface.

This retrospective study included four men and one woman. Three patients had fascioscapulohumeral dystrophy who underwent bilateral arthrodeses and two patients with post-traumatic injury to one thoracicus longus nerve. All scapulothoracic joints were unstable; six were painful. There was a cosmetic prejudice in all cases. Preoperative function was 71+/-6 degrees antepulsion (range: 60-80 degrees ) and 71+/-7 degrees abduction (range: 60-80 degrees ). Postoperative assessment included: reduction of the scapula alata, gain in motion, Constant score (raw and weighed), subjective assessment of daily activity, complications, respiratory function in patients with double arthrodesis and bone fusion on the CT scan.

Patient follow-up was 21.5 months on average. The scapula alata was reduced in all cases. Mean gain in motion was 39+/-21 degrees antepulsion and 41+/-26 degrees abduction. The raw Constant score was 77.75+/-11.4/100 and the weighted score was 81.5+/-9.1/100. Daily activity was scored 100% in four of five patients. Complications were transient intercostal dysesthesia, mild exercise-induced pain, one pneumothorax which did not require drainage. There was no evidence of an impact on respiratory function. The control scans revealed fusion in five shoulders and nonunion in three.

Scapulothoracic arthrodesis enabled reduction of the scapula alata and improved shoulder function for elevation and daily activities. The gain in motion and Constant score were satisfactory and similar to other results in the literature. The main drawback was deficient external rotation. Pain was mild and resulted from the deafferentation. It resolved six months postoperatively. Secondary exercise-related pain was mild with little impact. Technically, authors have proposed using different ribs for the fusion, depending on the patient's morphology and searching for a good position for the scapula for external rotation. There have been few postoperative complications: we had one pneumothorax and no neurological, vascular or pulmonary complications. Our results are the first reporting CT fusion findings. Indirect signs of nonunion on plain x-rays reported in earlier studies are not fully reliable, since our CT scans demonstrated nonunion in three shoulders. The lack of fusion does not mean poor function, as was also noted by others, since function was improved (mean gain 27 degrees flexion and 25 degrees abduction).