Page 169 - Reconstructive Techniques in Musculoskeletal Tumor Surgery

antibiotics. During tumor resections, the allografts were thawed in saline solution. Allografts were cut freehanded1 or with use of computer navigated techniques18. Intraoperatively, the surgeon checked that cortical contact could be obtained. In a subset of our patients, a virtual bone bank system was used to select the allograft that best matched the planned resection19. Additional cancellous bone grafting was performed in 11 osteotomies (11%), indications included dissatisfying compression at the osteotomy and suboptimal bone quality at the docking site.
Antibiotics were continued for one to seven days after surgery. Postoperatively, patients were mobilized under supervision of a physical therapist. Routine follow- up included conventional radiographs in two directions. MRI and/or CT scans were obtained in case of (suspected) complications. We recorded patient sex, age at surgery, diagnosis, tumor localization, date of surgery, localization within bone level (diaphyseal or meta-epiphyseal), type of neoadjuvant or adjuvant therapy, total resection length, the use of additional (intramedullary) bone grafts, and muscle aps.
Radiographic classi cation
We introduce a novel classi cation system, in which we classify the degree of
contact into grades 1, 2A, 2B, and 3. Grade 1 was de ned as full contact over the
entire length of the osteotomy in both directions; no radiolucent line was visible.
Grade 2 was de ned as partial contact and was further divided into grades 2A
(≥50% contact) and 2B (<50% contact). Grade 3 was de ned as a lack of cortical 8 contact; a radiolucent line was visible over the entire length of the osteotomy
( gures 1-4).
Figure 1. Anteroposterior X-ray of a proximal femoral allograft. A radiolucent line cannot be identi ed; there is full contact (white arrow): grade 1.
Allograft nonunion
167