Page 120 - Reconstructive Techniques in Musculoskeletal Tumor Surgery

Chapter 6
Materials and Methods
To identify patients who were eligible for this nationwide retrospective study, we searched an electronic database of our national bone bank for massive allografts that had been delivered to all four appointed centers of orthopaedic oncology from 1989 to 2012. We then evaluated the diagnosis and procedure information of the patients who had received the grafts, and all of those who had been treated for a primary tumor of a long bone with hemicortical resection and allograft reconstruction were included. The minimum duration of follow-up was 24 months.
Allografts were harvested under sterile conditions during postmortem tissue donation and stored at -80°C afterward21. Grafts were processed by either Osteotech (Eatontown, New Jersey) or the Musculoskeletal Transplant Foundation (Edison, New Jersey) and either not subjected to additional sterilization or sterilized with low-dose gamma radiation (<25 kGy). In most patients, biopsies were performed to obtain a histological diagnosis and the biopsy track was excised in continuity with the tumor. A wedge resection was performed in all patients — in some cases because of an atypical presentation or unclear diagnosis preoperatively. Resections were planned with use of an array of conventional radiographs, magnetic resonance imaging (MRI) scans, and computed tomography (CT) scans. All patients received prophylactic cephalosporins prior to surgery. Allografts were thawed in saline solution with antibiotics during the resection and subsequently cut to t the resected defect. Osteosynthesis was performed if the reconstruction was not considered intrinsically stable.
Medical les were evaluated to obtain characteristics of the patients, tumors, surgery, reconstruction, and treatment. Tumor grade was strati ed into four groups: benign, low-grade malignant, intermediate-grade malignant, and high- grade malignant. Surgical margins were de ned as being adequate (marginal or wide with no tumor cells at the margins)22, questionable (the pathologist in doubt about whether there were tumor cells at the margins), or intralesional. The reconstruction length and the percentage of the cortical circumference that was resected were measured on conventional radiographs in two directions and corrected for magni cation. The extent of cortical resection was classi ed as <25%, 25% to 50%, 51% to 75%, or >75%.
Complications and reasons for failure were classi ed as mechanical (nonunion or fracture), infection, and oncological according to the system described by Henderson et al23. A patient was considered to have had a nonunion if a surgical intervention had been performed to facilitate osseous union7. Fractures were
118