Purpose: During revision procedures for total knee arthroplasty with reconstruction of the lower femur (TKA after tumour resection) it is difficult to extract the stem from the proximal femur (if noncemented) and spare bone stock. The purpose of this study was to describe and analyse aspects related to the use of a complementary approach for massive trochanterotomy allowing easier access to the centromedullary canal of the femur and thus facilitate extraction of the femoral stem and periprosthetic cement.

Material and methods: This technique was used for five patients between 1991 and 1999. There were four women and one man, aged 18–45 years. The femoral piece was changed in three patients because of a fractured non-loosened implant and in two cases because of loosening. The revision implant was a total reconstruction prosthesis in one case (Link) and a GUEPAR implant in four. Massive trochanterotomy or corticotrochanterotomy was performed in all cases sparing the muscle insertions.

Results: This retrospective analysis was performed at a mean follow-up of five years (3–12). There were no cases of loosening or implant fracture. The trochanteric fragment (or corticotrochanteric fragment) healed normally in all cases. One female patient experienced moderate pain in the sitting position related to the presence of osteosynthesis material in the hip, but no implant removal was necessary.

Conclusion: Complementary trochantotomy facilitated removal of the inferior femoral piece via a direct approach to the summit of the stem allowing direct expulsion with the periprosthetic cement. Direct vertical access to the medullary canal allows good control of the revision prosthesis and limits unnecessary bone loss without creating any particular iatrogenic problem other than longer time for trochanter healing.

Purpose: Treating fractures of the femur in patients with a total hip arthroplasty is a difficult task. The frequency of these fractures is estimated at 1% to 3%. The purpose of this work was to evaluate the long-term clinical and radiological outcome and to search, by type of fracture, for factors predictive of late complications in order to improve indications.

Material and methods: Since 1985, we collected data on 29 patients (17 women and 12 men) mean age 73.3 years who had a fracture of the femur after total hip arthroplasty. By definition, intraoperative fractures were excluded. Two independent operators noted epidemiological, therapeutic and imaging data and classed the fractures using the Vancouver classification. Treatments used were noted and clinical and radiological outcomes at mean follow-up of 24 months (6–140) were analysed.

Results: Falls were the cause of the fractures in 31 patients (84%). Six of these patients had a revision prosthesis. Five fractures were in a zone of unprotected weakness. The fractures were trochanteric (n=9), periprosthetic (n=18), or below the stem (n=2). Nine patients were treated by isolated osteosynthesis and eleven by replacing the prosthesis. The Beals score was used to assess outcome taking into account the stability and the quality of the implant fixation as well as fracture realignment. In patients whose fracture was around the stem, outcome was excellent in two, good in one, and poor in four. For fractures of the lesser trochanter (n=4), outcome was excellent in two and good in two. For fractures below the stem (n=2), outcome was good in both. There was one case of deep infection.

Discussion: Considering the same types of fractures, orthopaedic treatments produced poor results (6/11), unlike prosthesis replacement (2/11) and osteosynthesis (1/9). The number of loosenings was underestimated, leading to failure of orthopaedic treatment.

Conclusion: In the event of fracture of the proximal femur in patients with a total hip arthroplasty, any suspected loosening, particularly of a cemented stem, should lead to prosthesis revision rather than orthopaedic treatment or simple osteosynthesis.