Long-Term Clinical Consequences of Stress-Shielding after Total Hip Arthroplasty without Cement*

Remodeling of the femur, or so-called stress-shielding, was observed on the two-year postoperative radiographs of forty-eight (23 per cent) of 207 hips that were part of a consecutive, non-selected series of 223 hips that had had a primary arthroplasty with use of the anatomic medullary locking hip system. Three patients (three hips) died within ten years after the arthroplasty, leaving forty-four patients (forty-five hips) who had a minimum of ten years of clinical follow-up. At the time of the latest follow-up, thirty-eight patients (86 per cent) reported that they had either no or mild pain related to the hip, forty-two (95 per cent) had less pain than they had had preoperatively, and forty-one (93 per cent) were satisfied with the results of the arthroplasty. Two patients had a reoperation, but neither procedure involved the femoral component; specifically, one patient had a revision of a loose acetabular component and one had an exchange of a polyethylene liner. No femoral component was associated with clinical or radiographic evidence of loosening. Femoral osteolysis, confined to zones 1 and 7 of Gruen et al., was observed on the ten-year radiographs of four of the thirty-three hips for which such radiographs were available. Stress-shielding (defined as evidence of pronounced femoral bone-remodeling on the two-year radiographs) had not adversely affected the outcome for these four hips by the time of the latest follow-up. The findings regarding postoperative pain, function, and over-all satisfaction for the forty-four patients (forty-five hips) who were included in the present study were similar to those reported for our larger (parent) series of patients who had been managed with the anatomic medullary locking hip system and to those reported for a similar series of patients who were followed for 9.5 years after the insertion of a porous-coated anatomic prosthesis. In addition, the prevalence of acetabular and femoral osteolysis (four [12 per cent] of thirty-three hips) and that of revision of the femoral component (zero [0 per cent] of forty-five hips) were lower than those for our larger (parent) series (fifty-four [39 per cent] of 137 hips and three [1 per cent] of 201 hips, respectively) as well as those for the series of patients who had been managed with the porous-coated anatomic prosthesis (thirty-five [45 per cent] and four [5 per cent] of seventy-eight hips, respectively).