Page 142 - Reconstructive Techniques in Musculoskeletal Tumor Surgery

Chapter 7
(interquartile range [IQR] 14 to 32) at surgery were included (table 1). A total of 33 patients (87%) had a malignant tumor (predominantly osteosarcoma, n = 20, 53%), ve patients (13%) were treated for a benign but aggressive lesion, mostly giant cell tumors of bone (n = 4, 11%). A further 26 patients (68%) were treated with chemotherapy according to appropriate protocols, two (5%) underwent radiotherapy.
Allografts were harvested during post-mortem tissue donation by our national bone bank. Proximal humeral grafts included tendons of the rotator cu , pectoralis major and latissimus dorsi; allografts of the proximal femur had the tendons of the glutei and iliopsoas attached. Distal femoral and proximal tibial grafts included the knee capsule and all surrounding ligaments. Following retrieval, articular cartilage was covered with gauze soaked in dimethylsulphoxide31 and allografts were stored at -80°C32. Processing of the allografts was performed at either Osteotech (Eatontown, New Jersey) or the Musculoskeletal Transplant Foundation (Edison, New Jersey). Grafts were either not subjected to additional sterilization or irradiated with low-dose gamma radiation (< 25 kGy).
All patients had a biopsy pre-operatively to obtain a histological diagnosis. Resections were planned on conventional radiographs, CT and/or MRI. All patients received prophylactic cephalosporins pre-operatively. During tumor resection, the allograft was thawed in saline with gentamicin and ucloxacillin. Following resection, the graft was cut to t the resected defect freehand, and appropriate structures were prepared to attach corresponding host structures. All osteotomies were fashioned transversely. Muscle ap rotations (n = 9, 23%) were only used where there was poor soft-tissue coverage.
Follow-up routinely included conventional radiographs but if a recurrence was suspected an MRI scan was obtained. Medical les and radiographs of the reconstruction were evaluated to obtain details about patients, tumors, treatment, and reconstructions. Complications and failures were classi ed into types 1 to 5, according to Henderson et al33, 34 (type 1, soft-tissue failure and instability; type 2, graft-host nonunion; type 3, structural failure; type 4, infection; and type 5, tumor progression). Nonunion was de ned as surgical intervention to facilitate union of the allograft-host junction3. Fractures were diagnosed on imaging. If the allograft was removed (partially or completely), or if the reconstruction was converted to an allograft-prosthetic composite or arthrodesis, we considered the reconstruction to be a failure.
140