We retrospectively reviewed 107 patients (131 hips) treated by revision THA between August 2000 and December 2004. These included revision of the femoral component in eight hips (6%), revision of the acetabular component in 45 hips (34%), and revision of both components in 78 hips (60%). Among these, femoral osteolysis occurred in 76 hips (58%). These hips were treated as follows: revision of the femoral component in 40 hips (52.7%) when the femoral osteolysis was associated with stem loosening, bone grafting into the osteolytic lesion in eight hips (10.5%) when the femoral stem was well fixed and the osteolytic defect was well contained, and simple curettage of femoral osteolysis when the femoral component was well fixed and the osteolytic lesion was not contained during isolated acetabular revision in 28 hips (25 patients; 36.8%). These 25 patients make up our study group. We followed these patients prospectively at the time of scheduled visits to evaluate femoral osteolysis. In all hips, the femoral and acetabular components were cementless and were judged clinically and radiographically well fixed at the time of revision. The indication for revision in all hips was polyethylene liner failure and resulting extensive osteolysis around the acetabulum. The revision surgery was the first revision for all hips. The underlying diagnosis at the time of primary arthroplasty was osteonecrosis in 19 hips (79%) and osteoarthritis in five (21%). There were 14 men and seven women with a mean age of 47.1 years (range, 26–68 years), a mean height of 164.6 cm (range, 155–177 cm), and a mean weight of 61.9 kg (range, 50–75 kg) at the time of revision. The left hip was involved in six patients and the right hip in 12; both hips were involved in three. The average interval between initial arthroplasty and index acetabular revision was 8.2 years (range, 4.9–12.3 years) Two patients (two hips) died from causes unrelated to surgery, and two patients (two hips) were lost to followup before the end of the minimum 3-year followup period (mean, 4.3 years; range, 3–7.4 years); this left 21 patients (24 hips) as the subjects. These 21 patients were followed a minimum of 3.9 years (mean, 8.3 years; range, 3.9–12.3 years) after the primary arthroplasty. None of the four patients who died or were lost to followup showed progression of femoral osteolysis by the time of their last evaluation (8–23 months). All patients provided informed consent for participation, and our Institutional Review Board approved the protocol.
We used the Harris-Galante™ II porous-coated acetabular cup (Zimmer, Inc, Warsaw, IN) in all primary THAs. The outer diameter of the primary acetabular component averaged 53 mm (range, 46–60 mm). The polyethylene thickness averaged 7.0 mm (range, 3.3–10.3 mm). Twenty femoral components were cementless HG MultiLock™ hip prostheses and four were cementless Harris-Galante™ stems (all manufactured by Zimmer).
In all hips, we revised the shell because of severe metal damage or extensive osteolysis of the acetabulum, although the cup was well fixed. We performed all revision procedures with an anterolateral approach. We revised all acetabular components with larger-diameter cementless components and modular polyethylene liners. The acetabulum was underreamed by 2 mm and press-fit into place. We performed bone grafting for all osteolytic acetabular lesions. The osteolytic lesion around the femoral component was removed with angled curettes, and tissue samples routinely underwent pathologic evaluation. We did not bone graft the defects in the femur to avoid graft in uncontained defects from escaping into the hip. All femoral heads were replaced with 28-mm cobalt-chrome heads. We used the InterOp™ cup (Sulzer Orthopedics, Inc, Austin, TX) in eight hips and the Trilogy® acetabular component (Zimmer) in 16 hips. The Durasul™ (Sulzer Orthopedics) highly cross-linked polyethylene liner was used in eight hips and the Longevity® (Zimmer) highly cross-linked liner was used in 16 hips. The external diameter of the acetabular component averaged 60.8 mm (range, 54–67 mm). The average thickness of the polyethylene liner was 10.6 mm (range, 7.3–14.3 mm).
The postoperative rehabilitation protocol was the same for all patients, who were allowed progressive weightbearing as tolerated on the third day after surgery. Patients did not receive any antiosteolytic medications including bisphosphonate during followup.
We assessed each patient clinically and radiographically before and after revision surgery at 4 weeks, 3 months, 6 months, 12 months, and annually thereafter. We obtained preoperative and postoperative Harris hip scores [12
]. Thigh pain was rated as none, slight, mild, moderate, or severe using the same criteria as for the pain category of the Harris hip score [12
Standard radiographs included an anteroposterior view of the pelvis and anteroposterior and lateral views of the proximal part of the femur. Radiographs taken 4 weeks after index surgery served as the baseline for all subsequent comparisons. We considered an acetabular component radiographically loose when migration had occurred or a circumferential radiolucency of 2 mm could be measured. We measured cup migration on serial radiographs, and a linear change greater than 3 mm or a rotational change 8° or greater was considered migration [19
]. Radiographic evaluations of the socket were done using DeLee and Charnley zones [6
]. Osteolysis was defined as the radiographic appearance of any focal area of bone resorption 2 mm wide or greater that was not evident on the radiograph obtained immediately after surgery [27
]. We measured the size of the area of osteolysis on the anteroposterior radiograph as the greatest diameter of the lesion in the horizontal and vertical axes [18
]. A correction factor for magnification, based on the known diameter and the measured radiographic diameter of the femoral head, was applied to all measured radiographic calculations. All measurements were made by one blinded observer (CHJ). To check reproducibility, a second blinded observer (KJL) also measured the osteolytic lesion in five hips. The intraobserver and interobserver measurements errors of osteolysis size were 2% and 3%, respectively. We identified lesion location using the zones defined by Gruen et al. [11
]. All hips showed focal osteolysis proximal to the lesser trochanter around the femoral stem at the time of the revision, mainly in Gruen Zones 1 and 7. Lesions did not span more than one Gruen zone. We compared the latest followup radiograph and the initial postoperative radiograph after revision to determine progression of the osteolytic lesion. If the osteolysis increased its initial postoperative area by more than 50% or greater than 1 cm2
, we defined the osteolysis as having progressed. If the osteolysis decreased by more than 50% of its immediate postoperative area or greater than 1 cm2
, we defined the osteolysis as having regressed [13
]. Osteolysis with an area change less than 50% and less than 1 cm2
was considered stabilized. Bone graft consolidation around the acetabulum was evaluated using the method of Peters et al. [21
], and linear polyethylene wear was measured with a radiographic digitizer according to the method of Livermore et al. [16
]. Femoral component fixation was graded as radiographic ingrowth, stable fibrous, or unstable according to the criteria of Engh et al. [8
]. We defined instability of the femoral stem as subsidence greater than 3 mm, a change of position, or a continuous radiolucent line wider than 2 mm [9