The use of the free vascularised bone graft for nonunion of the scaphoid: a systematic review

BACKGROUND Fractures of the scaphoid are well known to be problematic especially when complicated by avascular necrosis, nonunion and carpal collapse. Fixation techniques have involved nonvascularised bone grafting; however, in the presence of avascular necrosis, generally poor union rates (47%) occur as identified by a meta-analysis performed by Merrell et al. The introduction of pedicled vascularised bone grafts showed further improvement; however, in the presence of carpal collapse, union rates as low as 50% have been reported by Chang et al. amongst others using the 1,2-intercompartmental supraretinacular artery pedicled graft. The difficulty lies in having a short pedicle with limited manoeuvrability to correct a humpback deformity and insert into the scaphoid cavity. Prior trauma to the soft tissues or distal radius may prohibit the use of pedicled grafts. The aim of this systematic review is to examine the published evidence for the use of free vascularised bone grafts in cases of scaphoid nonunion. METHODS A systematic review was performed with the following defined search strategy on MEDLINE and Google Scholar: ((scaphoid nonunion) OR scaphoid pseudarthrosis) AND bone graft. Articles were reviewed and data compiled into tables for analysis. Statistical analysis was performed with determination of descriptive statistics, and differences between the groups were calculated using categorical variables and chi-square test. A p value of 0.05 or less was considered to be statistically significant. RESULTS Two hundred and sixty-three articles were identified with a total of 12 articles meeting the inclusion criteria. Two hundred and forty-five cases of scaphoid nonunion were identified through the articles included in this systematic review. Fifty-six patients underwent free vascularised bone grafts from the medial femoral condyle with a 100% union rate and correction of humpback deformity, and 188 patients underwent free vascularised bone grafting from the iliac crest with an 87.7% union rate. The difference between the two similar groups was statistically significant (p = 0.006). CONCLUSIONS The promising data suggests that the medial femoral condylar free graft based on the descending genicular vessels can be considered in cases of proximal pole avascular necrosis and humpback deformity or in situations where other flaps are precluded or deemed unlikely to cause union.