Periprosthetic femoral fractures--incidence, classification problems and the proposal of a modified classification scheme.

INTRODUCTION: The increasing incidence of periprosthetic fractures correlates directly with the year-after-year increasing frequency of primary joint replacement surgery. The most common fracture localisation is the femur. The undisputed leader in frequency is the fracture that occurs around a total hip arthroplasty. Unfortunately, no general epidemiologic data exist dealing with exact fracture incidence numbers. Furthermore, existing classifications are lacking important information like time point of fracture occurrence, type of the implanted prosthesis and implantation technique (cemented vs. cementless). Additionally, information about mechanical quality of the bone structure and the fracture type are also missing in part. METHODS: We scanned the literature for adequate and widely used classifications in the field of hip and knee arthroplasty. In a next step we analyzed those classification systems in order to find out to what extent they are able to describe the specific aspects of the fracture event. Therefore we compared the existing classifications and presented their most relevant emphasis. Furthermore, we looked at our own patient population to evaluate incidence of fracture occurrence over time and percentage of loosened components. RESULTS: The existing classification systems address themselves specifically to the task of describing fracture localization and to some extent fracture type, or combine these two in order to calculate the possibility of loosening of the implanted prosthesis. Some of the important criteria like mechanical quality of the bone stock, primary implantation technique or time point of the prosthesis loosening (prior to or because of the fracture) remain ignored. The incidence of periprosthetic femur fractures at our department increased approximately 2.5 fold over the past two decades. The risk of suffering from a periprosthetic fracture was substantially higher after THA than after TKA. We observed a loose femoral component of the THA in about 45 % of the cases. Finally, we postulate the application of a modified classification for periprosthetic fractures as an alternative to the already published ones; not only for the femur, but also universally for all joints with an arthroplasty. CONCLUSION: The classification that is introduced in this study allows, in our opinion, a differentiated reflection of the given post-traumatic pathologic changes and enables the description of the fracture itself according to a generally accepted fracture classification scheme.