Page 175 - Reconstructive Techniques in Musculoskeletal Tumor Surgery

A lack of cortical contact was the most important risk factor for nonunion in our study. None of the osteotomies that demonstrated full contact between the allograft and host bone developed a nonunion. A number of previous studies addressed the in uence of gap size on healing of bone defects. Claes et al showed that primary bone deposition in the metatarsus of sheep occurs in osteotomy gaps of less than 1 mm and that inferior healing occurs in gaps greater than 2 mm24. They concluded that treatment of simple diaphyseal fractures is improved when interfragmentary gaps are prevented.
One option to maximize the contact surface between allogeneic and host
bone is to use step cut osteotomies, which have been associated with a 74%
increase in contact surface as compared with transverse osteotomies23. Although
step-cut osteotomies may be preferable theoretically, transverse osteotomies
are still the technique of choice in our centers, for a number of reasons. First,
transverse osteotomies consist of a single cut and are the least technically demanding. Therefore, the chance of obtaining full contact is higher than with
more complicated step-cut osteotomies. Second, a transverse osteotomy is the
only type of osteotomy in which uniform pressure distribution between can be
obtained23. Third, in contrast to step cut osteotomies, transverse osteotomies do
not require further soft tissue exposure. The limited extent of soft tissue dissection
has been described as a factor that contributes to the chance of initial healing
of allografts23, 25. Fourth, transverse osteotomies are quick and therefore may be 8 associated with a lower risk of infection as compared to more complicated step
cut osteotomies.
Frisoni et al analyzed factors a ecting outcomes of intercalary femoral allografts6. They radiographically reviewed osteotomies to assess contact at the allograft-host junctions, and de ned “good contact” as at least two of the four cortices being separated by a radiolucent line of less than 2 mm. They reported that “good” versus “poor” contact did not in uence the risk of delayed union. However, it may be questioned how one can reliably or reproducibly measure a gap of 1 to 2 mm on radiographs that have not been taken according to a prede ned protocol. In future studies, CT scan images may be used to determine the exact gap size. Because CT images were only available for a small number of patients, we chose to classify the osteotomies in a limited number of categories that could easily and reproducibly be distinguished on conventional radiographs. Indeed, our classi cation system demonstrated good interrater reliability.
Allograft nonunion
173