129. RECONSTRUCTION OF MAJOR TISSUE LOSS OF THE HUMERUS AFTER TUMOUR RESECTION: THE INDUCED MEMBRANE TECHNIQUE IN THREE CASES

Abstract

Purpose of the study: Resection of a malignant primary tumour of the proximal humerus implies sacrifice of a large part of the humeral shaft and the periarticular muscles. Reconstruction can be difficult and raises the problem of preserving function. Recent work has demonstrated the pertinence of combining a glenohumeral prosthesis with an allograft. Several complications are nevertheless reported: non-union, allograft resorption, loosening. We report three cases of malignant primary tumours requiring wide resection of the humerus which were treated by reconstruction with a shoulder arthrodesis applying the induced membrane technique.

Material and methods: Three patients (mean age 15 years) presented a malignant primary tumour of the proximal humerus (Ewing sarcoma or osteosarcoma) which was locally extensive but not metastatic. Resection implied resection of 16 cm of the humerus (mean). The same procedure was used for the three patients: first phase: wide resection of the tumour and neighbouring soft tissues which removed the majority of the proximal end of the humerus and the glenohumeral joint, then insertion of a cement spacer; second phase: reconstruction with a shoulder arthrodesis using cancellous grafts positioned inside the induced membrane. Stabilisation was ensured by insertion of a non-vascularised fibula inside the membrane and with a plate fixation on the scapular spine.

Results: Mean follow-up is five years. There has been no local recurrence and no distant spread. The arthrodeses and the reconstructions healed without reoperation within six to eight months. The functional outcomes were not different from those obtained with shoulder arthrodesis with a mean elevation of 90°.

Discussion: There are many advantages of reconstruction with shoulder arthrodesis using the induced membrane technique: possible wide initial resection, more satisfactory carcinological resection, the periarticular muscles are not pertinent after arthrodesis; there is no need for prosthetic elements or an allograft exposing to later complications; the reconstruction time is a simple procedure; elevation remains satisfactory.

Conclusion: This technique should be included in the surgical armamentarium just like vascularised transfers, allografts and massive prostheses. The indication should be reserved for extensive resection.