Page 168 - Reconstructive Techniques in Musculoskeletal Tumor Surgery

Chapter 8
Minimum follow-up was 12 months. We excluded patients in whom the allograft was removed or revised within 12 months for reasons other than nonunion.
Intercalary allografts were the preferred method of reconstruction for any patient with a primary bone tumor in whom we considered it possible to resect the tumor with adequate margins while preserving joints. Alternative treatments included vascularized bular autografts, hybrid reconstructions, or intercalary (custom-made) implants. Osteoarticular allografts13, 14 or modular endoprostheses15 were used when (part of ) a joint had to be sacri ced. To minimize bias with regard to the in uence of contact at the allograft-host junction, we chose to only include transverse one-plane osteotomies in reconstructions with plate xation (either plates alone, or in combination with an intramedullary nail); whenever technically feasible, this was the preferred method for cutting and xation of allografts in both centers. A prerequisite for inclusion was the availability of digital radiographs in the anteroposterior and lateral direction taken in the rst 30 days after surgery, because these radiographs were used to assess the degree of contact at the allograft-host junction.
During the periods under study, a total of 208 osteotomies were performed in 104 patients for an intercalary allograft reconstruction of the femur or tibia. We excluded 112 osteotomies (54%): 29 (26%) because the osteotomy could not be assessed on postoperative imaging, 26 (23%) because imaging from the rst postoperative month was not available, 21 (19%) because it was a step-cut or oblique osteotomy, 16 (14%) because other types of osteosynthesis were used, eight (7%) because the reconstruction failed due to other reasons within 12 months after the index procedure, six (5%) because the patient died within 12 months after the index procedure, and four (4%) because the patient was lost to follow-up. This left 96 osteotomies in 57 patients (34 males, 60%) available for analysis. Thirty-seven patients (65%) were operated on in center one, 20 (35%) in center two. Median follow-up was calculated using the reverse Kaplan-Meier method, and was equal to 8.6 years (95% con dence interval [CI], 6.1-11.2).
Surgical technique
Allografts were harvested and processed according to techniques that have been described previously1, 8, 16, 17. The diagnoses were based on preoperative biopsy, and the biopsy tracts were excised in continuity with the tumor. Resections were planned on an array of conventional radiographs, magnetic resonance imaging (MRI), and computed tomography (CT). All patients received prophylactic cephalosporin
166