Epidemiology of periprosthetic femoral fractures in 5417 revision total hip arthroplasties

AIMS: The goals of this study were to define the risk factors, characteristics, and chronology of fractures in 5417 revision total hip arthroplasties (THAs). PATIENTS AND METHODS: From our hospital's prospectively collected database we identified all patients who had undergone a revision THA between 1969 and 2011 which involved the femoral stem. The patients' medical records and radiographs were examined and the relevant data extracted. Post-operative periprosthetic fractures were classified using the Vancouver system. A total of 5417 revision THAs were identified. RESULTS: There were 668 intra-operative fractures, giving an incidence of 12%. Fractures were three times more common with uncemented stems (19%) than with cemented stems (6%) (p < 0.001). The incidence of intra-operative femoral fracture varied by uncemented stem type: fully-coated (20%); proximally-coated (19%); modular fluted tapered (16%) (p < 0.05). Most fractures occurred during the insertion of the femoral component (35%). One-third involved the diaphysis and 26% were of the calcar: 69% were undisplaced. There were 281 post-operative fractures of the femur (20-year probability = 11%). There was no difference in risk for cemented and uncemented stems. Post-operative fractures were more common in men < 70 years (p = 0.02). Periprosthetic fractures occurred earlier after uncemented revision of the femoral component, but later after a cemented revision. The most common fracture type was a Vancouver B1 (31%). Of all post-operative fractures, 24% underwent open reduction and internal fixation and 15% revision arthroplasty. CONCLUSION: In revision THA, intra-operative fractures occurred three times more often with an uncemented stem. Many were undisplaced diaphyseal fractures treated with cerclage fixation. While the risk of post-operative fracture is similar between uncemented and cemented components, they occur at notably different times depending on the type of stem fixation. TAKE HOME MESSAGE: In revision THA, intra-operative periprosthetic femoral fractures occur three times more often with uncemented stems. Many are non-displaced diaphyseal fractures treated with cerclage fixation. While postoperative fracture risks are equivalent between uncemented and cemented components, they occur at notably different time periods based on stem fixation type.