[Osteosynthesis for periprosthetic supracondylar fracture above a total knee arthroplasty using a locking compression plate].

PURPOSE OF THE STUDY: The frequency of periprosthetic fractures related to total knee arthroplasty is increasing, with a prevalence of 1.3% on the average and with women being affected more often (4 out of 5 patients). Fractures of the distal femur are common, while tibial fractures are rare. Crucial for treatment is to distinguish fractures of the metaphysis above the femoral component, which remains firmly fixed, from those involving the knee joint replacement and component loosening. Supracondylar periprosthetic fractures are almost always managed surgically, using methods of osteosynthesis with an angle condylar or DCS plate, or a short retrograde- inserted supracondylar intramedullary nail. The recent use of implants such as LCPs with angle-stable screws has offered good prospects. This retrospective study presents our first experience with an LCP for treatment of supracondylar periprosthetic fractures of the knee joint.

MATERIAL AND METHODS: Between 2005 and 2008, a total of 13 supracondylar periprosthetic knee fractures were treated by the LCP technique. The patient group included 10 women and three men the average age was 67.4 (range, 56-81) years. The fractures were classified using the system proposed by Su et al. and the AO classification system. According to the Su classification, 12 types I and II fractures and one type III fracture were indicated for osteosynthesis. Based on the AO classification, there were four type 33 A1 fractures, five 33 A2 fractures, three 33 A3 fractures and one 33 C2 initially incorrectly classified as type 33 A3 fracture. The average time between total knee arthroplasty and injury was 6.8 years. In all patients fractures occurred after primary implantation of a cemented condylar total knee replacement without a femoral stem.The fractures were treated by a less invasive technique of LCP implantation within an average of 2.5 days of injury. The patients were followed up until radiographic fracture union, and complications were recorded.

RESULTS: The 13 patients were treated by LCP osteosynthesis through a less invasive approach. One patient had primary spongioplasty, two had spongioplasty after an interval of 7 weeks. One patient died of a disease unrelated to trauma and surgery at 3 months after osteosynthesis. In one patient, osteosynthesis failed with fragment dislocation shortly after the operation. The case analysis showed that the initial indication was marginal and the comminuted zone was too low above the implant, with the fracture line extending to the component. Subsequently, conversion to revision total knee arthroplasty involving a stem was carried out. In nine patients, bone union was achieved in an average of 18 weeks, with radiographic evidence of fracture union. No complications such as wound infection, delayed wound healing or thromboembolic disease were recorded. No bone union failure and pseudoarthrosis development occurred.

DISCUSSION: There are only few reports on the treatment of supracondylar periprosthetic knee fractures and evaluation of its results in the literature, and the groups evaluated are small. In a meta-analysis of cases from the 1981 to 2006 period, Herrera et al. have found only 29 assessable studies with a total of 415 cases, i.e., an average of 14 cases per study. The usual method of treatment was DCS plate osteosynthesis. Complications associated with conventional osteosynthesis techniques, as reported by various authors, may reach up to 30% (pseudoarthrosis development, 9% osteosynthesis failure, 4% necessity of revision surgery, 13% fracture malunion, 47%).Good results have been achieved with a retrograde-inserted intramedullary nail. The use of an LCP has been reported in the literature only occasionally. The classification system described by Rorabeck et al. is most widely used, but the system proposed by Su et al. seems more convenient to us, because fractures are placed in three groups, according to the localisation of a fracture line and its distance from the femoral component, as follows: type 1 fracture, fracture line is proximal to the femoral component type 2 fracture, fracture line starts at the level of a proximal edge of the femoral component and runs proximally type 3 fracture, fracture line extends below the upper end of the femoral component. Type 1 fracture is indicated for a retro- grade-inserted intramedullary nail, type 2 fracture for LCP osteosynthesis, and type 3 fracture for revision total knee arthroplasty. The use of LCPs in the treatment of supracondylar fractures of total knee arthroplasty, with a success rate of 86%, is described by Ricci et al. Other authors also report better outcomes with the use of LISS or LCP methods than with conventional osteosynthesis techniques.

CONCLUSIONS: Osteosynthesis with an angle-stable table LCP is an efficient method suitable also for the treatment of periprosthetic fractures of the distal femur above total knee arthroplasty. It offers all advantages of angle-stable implants. It is more effective for osteoporotic bone than a DCS implant or a condylar plate, because it provides better fixation stability for the distal fragment. However, further studies are needed to compare its efficiency with that of an IM nail.