Proximal femoral bone loss is often observed after total hip arthroplasty (THA) and continues to complicate revisions. The purpose of this study was to evaluate the clinical and radiographic outcomes of patients with the unusual finding of proximal femoral cortical hypertrophy after THA or endoprosthesis.

Three patients were identified with femoral stem tip perforation through the posterolateral femoral shaft cortex. Known risk factors for perforation were osteoporosis, distal osteolysis causing late stem migration, and endoprosthesis after previous fracture/internal fixation. Two patients had dual energy x-ray absorptiometry (DXA) of both hips. Follow-up was 16, 17 and 28 years.

All three patients had clinically stable femoral components and none had thigh loading pain. All three patients had dense cortical bone buttressing the undersuface of the implant collar and a dense medial femoral cortex. Proximal femoral DXA regions of interest were 120 and 145% of the contralateral femur in the migration and post-fracture endoprosthesis patients, respectively.

The endoprosthesis patient was revised to a THA 28 years post-operatively due to acetabular erosion. Excess dense cortical bone in the proximal femur was removed with a high speed bur and enabled revision with a primary non-cemented femoral component.

Common factors which appeared to be necessary for a stable implant with proximal femoral densification included collared femoral components, stem alignment more vertical than the femoral neck axis, stem perforating the posterolateral femoral cortex, no inhibition of translation along the stem axis by cement or biological ingrowth, and not having the implant revised despite its unorthodox appearance.

Long-term maintenance of the proximal femoral cortex after THA or endoprosthesis is possible. Paradoxically, a rare subset of the complication of femoral shaft perforation demonstrated by these anomalous cases suggested an alternative approach to the prevention of stress shielding bone loss after THA.

In vitro loading of the proximal femur has improved our understanding of stress shielding after total hip arthroplasty. However, previous load simulators often use simplified loading regimens that may not produce physiologic baseline strains. The purpose of this study was to compare the femoral strain levels produced when using simplified and more complex loading.

A mechanical load simulator was developed which could simultaneously apply a spinal load and nine of eleven available muscle loads to the proximal femur in heelstrike and stair climbing modes. Computer controlled electromechanical actuators were attached to a strain gauged fresh cadaver femur (donor body weight 39 6kg) with metal cables. A spinal load of 668 N (SPL) was applied alone and in combination with individual muscle loads of 267 N to determine the effect of each muscle on femoral strain. The magnitude and direction of the joint reaction force (JRF) was monitored in real time by a three-dimensional force transducer proximal to a metal acetabulum. Anterior, middle and posterior portions of the gluteus medius (ABD), iliotibial band (ITB), short external rotators (SER), vastus lateralis, adductors, rectus femoris, hamstrings, iliopsoas, and gluteus maximus were simulated.

SPL was applied and ABD and ITB were adjusted to produce a JRF magnitude of 2.0 BW. SPL was applied with two combinations of nine muscle loads adjusted in heelstrike mode to produce a JRF magnitude of 2.0 and 2.5 BW and JRF trajectory aligned within one degree of the radiographically determined compression trabecular stream axis.

Both nine-muscle combinations produced lower medial compression strains and substantially lower lateral tension strains than SPL+ABD+ITB in heelstrike and stair climbing. Simplified loading caused a bending moment in the proximal femur resulting in higher strains. Combined loading at 2.5 BW produced compression at 10 of 12 gauges in heelstrike mode and 9 of 12 gauges in stairclimbing.