 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1998 The Journal of Bone and Joint Surgery, Inc. A Femoral Component Inserted without Cement in Total Hip Arthroplasty. A Study of the Tri-Lock Component with an Average Ten-Year Duration of Follow-up* , OMAHA, NEBRASKAInvestigation performed at University of Nebraska Medical Center and Omaha Veterans Administration Medical Center, Omaha
Seventy-four total hip arthroplasties in sixty-six patients were performed, between 1983 and 1986, with use of a Tri-Lock femoral component inserted without cement. This tapered cobalt-chromium component has a fixed head and a circumferential proximal porous coating. Follow-up was conducted with use of a questionnaire, physical examination, and radiographic analysis. At the time of the latest follow-up, fifteen patients (eighteen hips) had died, three patients (four hips) had been lost to follow-up, and one patient (one hip) had refused to participate in the follow-up study; however, the status of fifteen hips at the time of death could be verified. Thus, clinical follow-up data were available for sixty-six of the original seventy-four hips. The average age at the time of the operation was sixty-two years (range, seventeen to eighty-four years), and the average interval between the operation and the latest follow-up evaluation was 10.0 years (range, 8.3 to 11.6 years). The Harris hip score was determined for forty-three hips (forty-one patients) in which the prosthesis was in situ at the time of the latest follow-up. The score was good for thirteen hips and excellent for twenty-eight, so the rate of clinical success was 95 per cent. Two patients had a fair result. One of them had persistent pain and the other had limited motion, but neither had radiographic evidence of loosening of the femoral or acetabular component. All forty-one patients were satisfied with the result. The probability (with standard error) of survival of the femoral component at ten years, with revision as the end point, was 0.95 ± 0.03. The rate of revision of the femoral component because of aseptic loosening was one (2 per cent) of sixty-six. The overall rate of aseptic loosening of the femoral component in the hips that were followed radiographically was two (4 per cent) of forty-seven. Only one (2 per cent) of the forty-seven acetabular cups had evidence of aseptic loosening. There was no radiographic evidence of distal osteolysis around the prostheses that were well fixed. Proximal osteolysis was present in five (11 per cent) of forty-seven hips, but none of the lesions compromised the stability of the prosthesis or the bone and there were no associated fractures. At an average of ten years postoperatively, the Tri-Lock femoral component functioned well overall and patient satisfaction was high.
Loosening of the implant after total hip arthroplasty has been widely reported in association with early prosthetic designs and so-called first-generation techniques of cementing6,7,19,32,36,37. In response to this problem, orthopaedists and engineers have improved the designs of the prostheses8,9 and the techniques of cementing25-27,30,31, and they also have developed mechanisms to promote biological ingrowth1,5,11,13,29. Porous-coated femoral and acetabular components were designed to promote fixation by stable osseous ingrowth in an attempt to increase the longevity of the prosthesis by decreasing the prevalence of mechanical loosening. Few long-term follow-up studies of the results of total hip arthroplasty without cement have been reported in the literature12. Overall, the short and intermediate-term results that were reported previously4,17,20,21,24,28,35,40 were encouraging, although the issue of late loosening of the components remained. In 1983, we began to use the Tri-Lock femoral component (DePuy, Warsaw, Indiana) exclusively in a consecutive series of patients. The early results in a subgroup of these patients were reported previously35. In that study, forty-seven hips in forty-two patients were followed for an average of forty-six months (range, twenty-four to sixty-six months), and forty-one hips (87 per cent) had a good or excellent Harris hip score. The study included both bipolar and total hip arthroplasties. One hip was revised because of infection, and one had radiographic evidence of early loosening. No patient had distal osteolysis. The purpose of the current study was to determine the rates of clinical success and aseptic loosening in patients who had had a total hip arthroplasty without cement. We also sought to identify the prevalence of osteolysis after an average duration of follow-up of ten years (range, 8.3 to 11.6 years).
The Tri-Lock femoral component that was used in this study was made of cobalt-chromium and had a tapered, straight, collarless stem. The stem was narrow in the coronal plane and much broader in the sagittal plane (Figs. 1-A and 1-B). Seven different sizes of stems were available in order to match the size of the femoral canal in the anterior-posterior plane. These stems varied only in the medial-lateral width; all had the same anterior-posterior width in the proximal portion. After the stem had been inserted into the femoral canal, it had an initial medial-to-lateral press-fit in the proximal portion of the canal without distal fill. The proximal five-eighths of the stem was circumferentially porous-coated (average pore size, 150 to 400 micrometers). Each stem had a fixed head with a thirty-two-millimeter diameter. Stems with a fixed head are no longer manufactured, but a modular-head version recently became available for use.
A Lunceford porous-coated acetabular cup (DePuy) was inserted in all hips. The cup was made of cobalt-chromium; it had three pegs for fixation and a fixed polyethylene liner. The acetabulum was reamed to a diameter that was one millimeter smaller than that of the cup, which was placed in a press-fit manner. Between June 1983 and October 1986, seventy-four primary total hip arthroplasties that involved insertion of a Tri-Lock femoral component without cement were performed in sixty-six patients. All patients also had insertion of a press-fit acetabular cup without cement. (In another group of patients, mainly those who had a fracture, a bipolar cup was inserted at the discretion of the surgeon, but these patients were not included in the study.) The diagnosis for which the index procedure was performed included osteoarthrosis in fifty-five hips (74 per cent), avascular necrosis in six (8 per cent), rheumatoid arthritis in six (8 per cent), avascular necrosis secondary to a fracture in two (3 per cent), and childhood hip disease in five (7 per cent). The diagnoses for the latter five hips included a slipped capital femoral epiphysis (two hips) and congenital dysplasia, Perthes disease, and sequelae due to infection (one hip each). Fifty hips (68 per cent) were in men, and twenty-four (32 per cent) were in women. The average age at the time of the operation was sixty-two years (range, seventeen to eighty-four years). All operations were performed by or under the direction of one of us (O. M. J.) and another surgeon, at two local hospitals. A posterior operative approach was used for all procedures. All patients received antibiotics prophylactically both preoperatively and postoperatively, until the drains were removed at twenty-four to forty-eight hours. The patients were allowed to stand on the first or second day after the operation. Toe-touch weight-bearing was allowed for three weeks, followed by partial weight-bearing for three weeks and by full weight-bearing thereafter. Follow-up consisted of radiographic analysis, physical examination, and a patient questionnaire that was developed by The Hip Society18. Patients who were unable to return to our institutions for follow-up had radiographs made at a local physician's office or hospital, and these were then sent to us for review. These patients also completed the questionnaire through a telephone interview. When a patient had died, phone contact was made with a family member to determine the survival of the prosthesis, the overall function, and the satisfaction of the patient with the hip just before death. All clinical and radiographic evaluations were performed by individuals who had not been involved in the index procedure or the care of the patient. Clinical evaluation consisted of completion of a questionnaire by all living patients and physical examination of those who were able to return for follow-up. Pain was rated as none, mild, moderate, or severe; pain also was assessed according to the time of onset, as none, with first steps, only after long walks, with all walking, or at all times. If the patient was employed, his or her occupation was recorded, and work capacity within the most recent three months was graded from 0 to 100 per cent with use of 25 per cent intervals. Each patient graded his or her level of activity, ranging from bedridden or confined to a wheelchair to strenous manual labor. The ability to put on shoes and socks, the ability to ascend and descend stairs, and the ability to walk with or without support also were noted. Finally, patients were asked whether the operation had resulted in a decrease in pain and the need for pain medication and whether they were satisfied with the overall result. Physical examination consisted of range-of-motion testing of both hips; evaluation of limp, the Trendelenberg sign, and abductor lurch; and measurement of limb length, from the anterior superior iliac spine to the medial malleolus. A Harris hip score16 also was determined on the basis of the information that had been obtained. A score of 90 to 100 points was considered excellent; 89 to 80 points, good; 70 to 79 points, fair; and less than 70 points, poor. The components were evaluated on anteroposterior radiographs of the pelvis and anteroposterior and frog-leg lateral radiographs of the affected hip. The femoral component was evaluated quantitatively with regard to varus or valgus position and migration. Subsidence of the component was measured from the medial proximal tip of the beaded surface to the center of the lesser trochanter and from the lateral proximal tip of the beaded surface to the tip of the greater trochanter. A change of five millimeters or more was believed to indicate subsidence17. The porous coating was examined for debonding, and the stem was evaluated for any evidence of breakage. Radiolucent lines were defined as linear, lucent areas surrounding the prosthesis and generally parallel to it, measuring less than two millimeters in the greatest thickness26. Osteolysis was defined as any area of non-linear radiolucency surrounding the prosthesis and measuring two millimeters or more in the greatest diameter26. These areas were identified, measured, and recorded with regard to their location in the anteroposterior and lateral zones described by Gruen et al.15 and by Johnston et al.18. Components with circumferential radiolucency were considered to be radiographically loose. Proximal resorption of bone was measured, in millimeters, as loss of the height or thickness of the calcar, or both. Zones of hypertrophy or resorption of bone of the femoral shaft were noted, as were areas of bone loss and increased trabecular bone. The system of Brooker et al.3 was used to record ectopic ossification. Analysis of the acetabular components included measurement of horizontal and vertical migration relative to the teardrop. Movement of more than two millimeters or tilt of more than 5 degrees was considered to be evidence of loosening. Because accurate radiographic measurement of polyethylene wear in cobalt-chromium acetabular cups and femoral heads is difficult with use of the technique of Livermore et al.22, this parameter was not assessed. Osteolysis was defined as focal areas of radiolucency that measured more than two millimeters in the greatest dimension and were not parallel to the cup; these areas were recorded according to the three zones of DeLee and Charnley10. Components with a circumferential radiolucent line were considered to be radiographically loose. Kaplan-Meier survival analysis was performed to generate two survival curves for the femoral component with use of two different end points: revision and radiographic failure.
Clinical Results The average Harris hip score at the time of the latest follow-up was 91.5 points (range, 76 to 100 points). Preoperative Harris hip scores were not used for comparison as they were not available for all patients. The scores for four hips (three patients) were excluded because of medical problems. Two hips in one patient were excluded because the patient had Alzheimer disease, one hip was excluded because the patient had had a resection arthroplasty on the contralateral side, and one hip was excluded because the patient was bedridden with multiple non-related medical problems. Thus, forty-three hips (forty-one patients) were available for complete analysis. The score was rated as excellent for twenty-eight hips (65 per cent), as good for thirteen (30 per cent), and as fair for two (5 per cent). One of the two patients who had a fair score had no pain; the score was low because of limited mobility. The other patient who had a fair score had moderate pain. Of the forty-three hips (forty-one patients) for which the Harris hip score was calculated, thirty-four (79 per cent) caused no pain and nine (21 per cent) caused some pain. Eight hips (19 per cent) were painful with the first steps that the patient took and then the pain dissipated; the remaining hip (2 per cent) was painful only after the patient walked for thirty minutes. Only one patient (2 per cent) rated the pain as moderate and stated that it had caused him to modify or give up some activities. This patient had neither a loose component nor pain in the thigh, and the source of the pain was not identified. Only four patients (four hips; 9 per cent) had pain in the thigh at the time of the latest follow-up; all four had an excellent or good Harris hip score, and none were limited in their activities by the pain. Thirty-two (78 per cent) of the forty-one patients did not need any type of support for walking: nineteen (46 per cent) could walk without assistance for more than sixty minutes; eight (20 per cent), for thirty-one to sixty minutes; and five (12 per cent), for eleven to thirty minutes. Seven patients (17 per cent) needed to use one cane for long walks, and two (5 per cent) always used one cane. Of the nine patients who needed support, five (12 per cent of the series) could walk for more than sixty minutes and four (10 per cent of the series), for eleven to thirty minutes. Twenty-two patients (54 per cent) walked with a limp. Six patients (15 per cent) had a moderate limp, and sixteen (39 per cent) had a slight limp that could be detected only by the examiner. Of the six patients who walked with a moderate limp, one had radiographic evidence of loosening of the femoral component, two were limited by degenerative changes in the contralateral knee, and two had 1.5 centimeters of shortening of the limb on the side of the arthroplasty. The reason why the sixth patient limped was not evident, but he had an excellent Harris hip score. Of the sixteen patients who had a slight limp, ten limped for a reason unrelated to the arthroplasty. No direct cause was found to explain why the remaining six patients limped; the Harris hip score was excellent for four of these patients, good for one, and fair for one. None of the six patients had pain in the thigh or radiographic evidence of loosening of a component. Twenty-eight patients (68 per cent) were no longer working outside of the home, and thirteen (32 per cent) were still employed outside of the home. Of these thirteen, nine had been able to work at full capacity for the three months before the examination; three, at 75 per cent capacity; and one, at only 25 per cent capacity. None of these patients had a sedentary job that required little time on the feet. Six patients had a job that involved moderate or strenuous labor and seven, a job that required some time on the feet but no heavy lifting or vigorous activity. With the numbers available for study, we detected no significant difference in the Harris hip scores between the patients who were working and those who were not or among those who were working at various capacities. Twenty-seven patients (66 per cent) had no difficulty putting on shoes or socks, thirteen patients (32 per cent) had only slight difficulty, and one patient (2 per cent) had extreme difficulty. Twenty-three patients (56 per cent) were able to ascend and descend stairs in a normal manner, thirteen (32 per cent) needed to use the banister or an assistive device, and five patients (12 per cent) had to put two feet on each step. Twenty-two patients (54 per cent) were able to rise from a chair without any support from the upper extremities; the other nineteen (46 per cent) could rise independently with the aid of the upper extremities. All forty-one patients, including those in whom the implant was considered to be a failure radiographically and those who had a fair Harris hip score, stated that they were satisfied with the result. Forty patients (98 per cent) said that they had increased function compared with the preoperative level and that the level of pain had decreased notably. Thirty-nine patients (95 per cent) had had a decrease in the need for pain medication.
Radiographic Results There were radiolucent lines around eleven (23 per cent) of the forty-seven femoral implants. Radiolucent lines were more common in zones VIII (seven implants; 15 per cent), VII (five implants; 11 per cent), XIV (four implants; 9 per cent), and I (three implants; 6 per cent) of Gruen et al.15 (Figs. 2-A and 2-B). Five hips (11 per cent) had areas of osteolysis (Figs. 3-A and 3-B). Osteolytic lesions were most common in zones VIII (five hips; 11 per cent), XIV (four hips; 9 per cent), I (three hips; 6 per cent), and VII (two hips; 4 per cent). Only one implant had an osteolytic lesion around its distal portion; the lesion developed after the prosthesis had become radiographically loose. None of the proximal lesions compromised the stability of the implant or the bone, and there were no associated fractures. The patient who had the largest focal areas of osteolysis had lesions that measured six millimeters in zone I and eight millimeters in zone VIII. However, this patient had no distal radiolucency and no migration, and the implant was believed to be radiographically stable (Figs. 4-A and 4-B). There was circumferential radiolucency around only one femoral stem (2 per cent), and this implant was considered to have failed according to the radiographic criteria.
Twenty hips (43 per cent) had proximal bone loss. Twelve hips had rounding of the calcar, three had loss of calcar thickness averaging two millimeters, and seven had loss of calcar height averaging four millimeters. Focal areas of resorption were noted in four femoral shafts (9 per cent), all of which had a well fixed implant. Nine hips (19 per cent) had grade-II heterotopic ossification, and one (2 per cent) had grade-III ossification3. No patient had migration or tilt of the acetabular component. Only one patient had an area of definite osteolysis adjacent to the acetabular cup, in zone I10. This also was the only patient who had circumferential radiolucency about the cup, which was considered to be loose. In addition, he had loosening of the femoral component with associated distal osteolysis.
Complications There were three infections, at six, thirty-nine, and 134 months postoperatively. All three infections were treated with a resection arthroplasty without reimplantation. Methicillin-resistant Staphylococcus aureus grew on culture of specimens obtained intraoperatively from the patient who had the infection at 134 months. This patient had had an infection of the shoulder, also with methicillin-resistant Staphylococcus aureus, thirteen months earlier and had been managed with débridement and antibiotics at that time. Before the infection of the hip, he had been satisfied with the result of the arthroplasty. There was one failure due to aseptic loosening. This was in a patient who had previously had a mastectomy for the treatment of breast cancer. Before the operation on the hip, there had been acetabular metastases on that side. At the time of the total hip arthroplasty, specimens obtained from the acetabulum showed no pathological evidence of persistent tumor; however, the patient was managed with high doses of steroids, both preoperatively and postoperatively, for treatment of the breast cancer. She also was being managed for post-menopausal osteoporosis. The femoral component began to subside, distal osteolysis developed, and the component eventually eroded through the lateral cortex of the femur. Revision with insertion of a stem with cement was performed 8.5 years postoperatively. Two patients had two dislocations each. Both patients were managed with closed reduction, and neither needed a revision. Two patients had a urinary-tract infection during their hospital stay and were managed uneventfully. One patient had a persistent hematoma that resolved without operative evacuation. One patient had a documented deep venous thrombosis, although routine screening was not performed in this series.
Statistical Analysis
The primary purpose of the current study was to evaluate the long-term results of total hip arthroplasty with use of the Tri-Lock femoral component without cement. At an average of 10.0 years (range, 8.3 to 11.6 years) after the operation, forty-one (95 per cent) of the forty-three hips had a good or excellent clinical result on the basis of the Harris hip score16. Previously, we had reported a good or excellent clinical result for forty-one (87 per cent) of forty-seven hips at an average of forty-six months35. Therefore, the clinical results remained stable since the time of our first report. In our previous report, we included patients who had a bipolar femoral head as well as those who had a fixed acetabular cup, but we did not include the patients who had a bipolar implant in the present report so that we could evaluate a more homogeneous group of patients. During the study period, forty hips (forty patients) were treated with a bipolar cup mainly because of a fracture, avascular necrosis, or the surgeon's preference. With the numbers available, no significant difference was detected between the two types of implants with regard to the Harris hip scores, the survival of the implant, radiolucency, osteolysis, or revision. Aseptic loosening developed in one (5 per cent) of the twenty-one patients who had a bipolar implant and were alive at the time of follow-up. This rate was slightly higher than the rate of 2 per cent for the patients who had had a total hip arthroplasty; however, with the numbers available, we did not detect a significant difference (p = 0.107). We are aware of only one other long-term study of the results of total hip arthroplasty without cement. In that report, Engh et al.12 reviewed the results of use of the anatomic medullary locking stem (DePuy, Warsaw, Indiana) in 167 patients (174 hips) at an average of eleven years (range, ten to thirteen years) postoperatively. The overall clinical results were remarkably similar to those for our patients; however, Engh et al. did not report a Harris hip score. Of the 147 patients (154 hips) who did not need a reoperation, 138 (94 per cent) were satisfied with the result compared with forty-one (100 per cent) of our patients. In the study by Engh et al., 138 (94 per cent) of their patients had a decrease in pain and a corresponding increase in function postoperatively compared with forty (98 per cent) of the forty-one patients in our series. One hundred and twenty-eight (87 per cent) of the patients in the series of Engh et al. had no or mild pain compared with forty (98 per cent) in our series. Nineteen (13 per cent) of the patients in the study by Engh et al. said that pain limited their activity compared with one (2 per cent) of our patients. Pain in the thigh was noted in association with thirteen (8 per cent) of 154 hips in the series of Engh et al. compared with four (9 per cent) of the forty-three hips in the current series. The fact that a high level of pain relief and function had been maintained in both of these long-term studies supports the early and intermediate-term findings of previous studies of arthroplasty without cement. Pellegrini et al.28 reported that forty-three (84 per cent) of fifty-one patients who had had insertion of a Tri-Lock femoral component had a good or excellent Harris hip score at an average of 6.5 years (range, five to eight years) postoperatively. Heekin et al.17 reported good and excellent results in more than 90 per cent of ninety-one patients at each follow-up interval between two and five years after insertion of a porous-coated anatomic prosthesis (Howmedica, Rutherford, New Jersey). Martell et al.24 found no appreciable decline in the hip scores during the first six years after insertion of Harris-Galante components and reported that 87 per cent of 110 patients had a good or excellent result at an average of sixty-seven months (range, fifty-five to seventy-nine months) postoperatively. Kim and Kim21 reported that 85 per cent of fifty patients had a good or excellent result at a minimum of seven years (range, eighty-four to eighty-nine months) after insertion of an anatomic medullary locking femoral component. The rate of revision of the femoral component because of aseptic loosening in the current series was 2 per cent (one of sixty-six hips). An additional stem was loose radiographically but was still in place at the time of the latest follow-up; thus, the overall rate of aseptic loosening of the femoral component was 3 per cent. The probability of survival of the femoral component (with standard error) at ten years, with use of revision as the end point, was 0.95 ± 0.03. This compares favorably with the 1.5 per cent rate of revision and the 3 per cent rate of loosening of the femoral component noted by Engh et al.12. Those authors reported that the survival of the anatomic medullary locking femoral component (with standard error) at twelve years was 0.97 ± 0.02. The results of the current series also compare favorably with the long-term results of arthroplasties in which the femoral component was inserted with cement. Mulroy and Harris25 reported a rate of revision of the femoral component due to aseptic loosening of 2 per cent (two of 105 hips) and an overall rate of loosening of the femoral component of 3 per cent (three of 105 hips) at an average of eleven years (range, ten to 12.7 years) after use of so-called second-generation cementing techniques. At the time of the fifteen-year follow-up evaluation, the rate of revision of the femoral component due to aseptic loosening was still only 2 per cent (two of 102 hips); however, four (2 per cent) of the original 162 components had been revised and seven (7 per cent) of the 102 surviving components were radiographically loose26. Schulte et al.34 reported that three (3 per cent) of ninety-eight hips had a revision due to aseptic loosening and seven (7 per cent) had aseptic loosening overall at a maximum of twenty years after an arthroplasty in which a Charnley femoral stem had been inserted with use of early cementing techniques. In the current series, no patient had distal osteolysis in association with a stem that was not radiographically loose. The overall prevalence of osteolysis around the femoral stem was 11 per cent (five of forty-seven). These lesions all occurred in the proximal aspect of the femur, in areas directly exposed to the joint space33. The one patient who had osteolysis distal to the porous coating of the femoral stem had had radiographic loosening before any evidence of distal osteolysis was noted. This stem had become loose after the patient sustained an intraoperative fracture of the femur. Pellegrini et al.28 found no evidence of distal osteolysis five to eight years after arthroplasty with the Tri-Lock femoral stem. However, short and intermediate-term follow-up studies of the Harris-Galante14,23,24,38,39, anatomic medullary locking20, and porous-coated anatomic components4,17,21,41 have revealed rates of distal osteolysis ranging from 7 per cent (of 100 hips with a porous-coated anatomic component)17 to 29 per cent (of forty-one hips with a Harris-Galante component)12. Rates of generalized osteolysis ranging from 8 per cent (of 121 hips with a Harris-Galante component)24 to 56 per cent (of fifty-two hips with a porous-coated anatomic component)21 also have been noted. Relative sealing of the femoral canal may occur either by osseous ingrowth or by means of a cement mantle. Schmalzried et al.33 discussed the role of the so-called effective joint space in allowing particles to migrate, leading to focal osteolysis and eventual aseptic loosening. Previous studies have shown a high prevalence of distal femoral osteolysis in association with stems inserted without cement, even those that are well fixed14,17,23,24,38,41. Several clinical and histological studies have demonstrated that the Harris-Galante femoral component, because of its non-circumferential porous coating, readily allows distal migration of particles through fibrous membranes around its smooth surfaces, leading eventually to distal osteolysis2,14,23,24,38,39. Like the Tri-Lock component, both the anatomic medullary locking and the porous-coated anatomic femoral component are circumferentially porous-coated. The anatomic medullary locking component is extensively porous-coated and is designed so that a press-fit is achieved in the distal portion of the canal. The porous-coated anatomic component is proximally porous-coated; it has an anatomical posterior bow proximally, and the diameter of its distal end is much larger than that of the Tri-Lock component. Both of these components, because of their larger distal diameter, are in much closer contact with the femoral isthmic cortex than the Tri-Lock component. This distal fit may predispose the hip to proximal stress-shielding and relatively less proximal bone ingrowth. It also may predispose these two components to greater relative proximal motion during physiological flexion of the femur during gait. The Tri-Lock component is very narrow in the coronal plane; it has a straight stem and no distal press-fit. This stem geometry may lead to less proximal stress-shielding and allow the prosthesis to remain relatively unaffected by flexion of the femur during gait, which in turn may help to maintain the effective sealing of the canal and to prevent distal migration of particles. Overall, the Tri-Lock femoral component inserted without cement performed well in our patients, all of whom were satisfied with the result of the arthroplasty. The prevalence of aseptic loosening was low, and distal osteolysis did not develop in association with any well fixed component. Osteolytic lesions were found in the proximal aspect of the femur, although none of these lesions compromised the stability of the bone or the prosthesis. We are optimistic regarding the future clinical and radiographic success of this component. However, additional follow-up is needed to determine the ultimate course of the proximal osteolytic lesions.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
  CiteULike  Connotea  Del.icio.us  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |