 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1997 The Journal of Bone and Joint Surgery, Inc. Primary Total Hip Arthroplasty with a Porous-Coated Acetabular Component. Seven-to-Ten-Year Results*  , CHICAGO, ILLINOISInvestigation performed at the Department of Orthopaedic Surgery, Rush-Presbyterian-St. Luke's Medical Center, Chicago
One hundred and eighty-four patients who had been managed, between 1984 and 1985, with a total of 204 consecutive primary total hip arthroplasties with insertion of a Harris-Galante type-I acetabular component without cement were prospectively studied. There were eighty-two men (45 per cent) and 102 women (55 per cent). The mean age at the time of the operation was fifty-two years (range, twenty to eighty-four years). One hundred and fifty-seven patients (173 hips) were available for clinical review at a mean of 104 months (range, seventy-eight to 126 months). At this time, the mean preoperative Harris hip score of 52 points (range, 12 to 79 points) had improved to a mean of 90 points (range, 44 to 100 points). Two patients (two hips) had had an exchange of an excessively worn polyethylene liner. One patient (one hip) had had débridement and grafting of an area of massive retroacetabular osteolysis. Two stable acetabular components (1 per cent) had been revised at the time of femoral revision. None of the cups had been revised because of aseptic loosening. Radiographic examination of 150 patients (165 hips) at a mean of 104 months (range, seventy-eight to 126 months) revealed that 156 cups (95 per cent) were stable. Eight cups (5 per cent) were considered to be possibly unstable, with a radiolucent line between the prosthesis and the bone that was one millimeter wide or less in at least four of five zones; two of the eight had a complete radiolucent line in all zones. One component, which had been implanted with a bulk allograft from a femoral head, migrated more than two millimeters in the first two years and then stabilized without complication. Osteolysis was seen in seven acetabula (4 per cent) and was limited to the periphery of the cup in six. Survivorship analysis at ten years revealed that the acetabular component had a 99 per cent chance of survival (95 per cent confidence interval, 0.98 to 1.0) with revision or aseptic loosening as the end point and a 97 per cent chance of survival (95 per cent confidence interval, 0.95 to 1.0) with revision, aseptic loosening, or reoperation because of a problem related to the acetabular component as the end point.
Although contemporary cementing techniques have improved the longevity of femoral components inserted with cement in total hip arthroplasty, there is controversy with regard to the fate of acetabular components inserted with cement. Rates of loosening of such components have been reported to range from 19 to 40 per cent at ten years in series of 100 to 302 hips2,7,16,27. Attempts to minimize loosening, such as use of a metal-backed implant, have failed to improve intermediate-term results6,18. Studies of histological specimens have suggested that osseous growth into porous-coated acetabular components is reliable8,9,17,26, and good clinical results have been demonstrated with these implants in short-term studies21,23. Consequently, many surgeons have begun to use porous-coated prostheses without cement. As the use of these implants has increased, there has been a heightened awareness of osteolysis induced by wear particles. Evidence of osteolysis around acetabular components inserted without cement was originally reported by Santavirta et al. Later, Zicat et al. reported a rate of osteolysis of 18 per cent (thirteen of seventy-one hips) at a minimum of eighty-six months, and Stulberg et al. reported a rate of seven of nineteen hips at a minimum of eighty-four months. In order to provide a satisfactory alternative to implants inserted with cement, porous-coated acetabular components must demonstrate not only a lower rate of loosening over time but also a low rate of complications, such as osteolysis, that may jeopardize the longevity of the implant and make later reconstruction difficult. We present the seven-to-ten-year results of a consecutive series of primary total hip arthroplasties with insertion of a Harris-Galante type-I porous-coated acetabular component (Zimmer, Warsaw, Indiana) without cement. The goal of our review was to investigate the clinical and radiographic performance of this implant at the time of an intermediate-term follow-up.
One hundred and eighty-four patients who were managed, between March 1984 and December 1985, with a total of 204 consecutive primary total hip arthroplasties with insertion of a Harris-Galante type-I porous-coated acetabular component without cement were followed prospectively. During the period of the study, twenty-two patients (twenty-six hips) died of causes unrelated to the operation before they had been followed for the minimum duration of seven years, and five patients (five hips) were lost to follow-up. Thus, 157 patients (173 hips) were available for clinical review, at a mean of 104 months (range, seventy-eight to 126 months). Seven of these patients (eight hips) were not examined radiographically at the time of the review. However, telephone interviews established that the acetabular components had remained in place. The diagnosis was osteoarthrosis for 122 hips (60 per cent), avascular necrosis for fifty-one (25 per cent), rheumatoid arthritis for eighteen (9 per cent), ankylosing spondylitis for five (2 per cent), traumatic osteoarthrosis for five (2 per cent), fracture of the neck of the femur for two (1 per cent), and a remote history of infectious arthritis for one (less than 1 per cent). There were 102 women (55 per cent) and eighty-two men (45 per cent). The mean age at the time of the operation was fifty-two years (range, twenty to eighty-four years). Ten women and ten men had bilateral hip replacement. One hundred and six left hips and ninety-eight right hips were replaced. The mean weight and height of the women was sixty-six kilograms (range, forty-three to 102 kilograms) and 152 centimeters (range, 138 to 183 centimeters). The mean weight and height of the men was eighty-one kilograms (range, forty-seven to 108 kilograms) and 161 centimeters (range, 138 to 190 centimeters). A modular titanium-alloy femoral component with a twenty-eight-millimeter cobalt-chromium head (Harris-Galante; Zimmer) was inserted without cement in 151 hips, a non-modular titanium-alloy femoral component with a twenty-eight-millimeter head (Gustilo-Kyle; Zimmer) was inserted without cement in fourteen, and a femoral component with a twenty-eight-millimeter head (Precoat; Zimmer) was inserted with cement in thirty-nine. The Harris-Galante type-I acetabular component is hemispherical and is made of commercially pure, unalloyed titanium. It is fully coated with sintered titanium fiber-metal. The size of the implanted components ranges, in two-millimeter increments, from forty-six to sixty-six millimeters (Table I). The shell has holes to allow supplemental fixation with 4.5 or 5.1-millimeter titanium-alloy cancellous-bone screws. The modular polyethylene liner is held in place by titanium tines around the periphery of the metal shell.
All operations were performed in an operating room equipped with vertical laminar airflow, and members of the operating team wore body-exhaust suits. A direct lateral approach to the hip was used, according to the technique described by Hardinge. Hemispherical reamers were used to remove all remaining cartilage until a bleeding bed of bone was obtained; no attempt was made to remove the subchondral bone plate. All acetabular components were inserted with a so-called line-to-line technique—that is, the outer diameter of the component to be inserted was equal to the outer diameter of the final reamer used in the preparation of the acetabulum. When the acetabular component was inserted, three to six cancellous-bone screws were placed for supplemental fixation. The metal tines of the acetabular shell were bent inward about one millimeter to enhance fixation of the insert, and then the polyethylene insert was impacted in place. Patients began physical therapy on the second postoperative day. Patients who had a femoral component inserted with cement were allowed to bear full weight as tolerated immediately, but they used two crutches for support for six weeks postoperatively. Patients who had a femoral component inserted without cement were restricted to bearing 50 per cent of their weight with the aid of two crutches for six weeks and then were allowed to bear weight as tolerated with the aid of two crutches for six weeks. Thereafter, a cane alone was used as needed. Clinical and radiographic assessments were carried out at six weeks, three months, six months, and twelve months and at annual intervals thereafter. Clinical evaluation was performed with use of the Harris hip score5. An anteroposterior radiograph of the pelvis was made at each examination. An external device was not used to position patients for radiographs, but radiographs were repeated if they were considered unacceptable for evaluation. The six-week postoperative radiograph was considered the baseline radiograph for all comparisons. The techniques used to evaluate the postoperative radiographs were detailed by Martell et al. Briefly, a modification of the method of DeLee and Charnley was used to assess the location and thickness of radiolucent lines at the interface between the prosthesis and the bone. A total of five zones (A1, A2, B1, B2, and C) were evaluated. Acetabular components that demonstrated a change in position of 2 degrees or more or migration of two millimeters or more were considered unstable. Any acetabular component that did not migrate but was associated with radiolucent lines in at least four of five zones, with at least one zone having a radiolucent line more than two millimeters wide, was designated as probably unstable. Any acetabular component that had not migrated but was associated with a radiolucent line in at least four of five zones, with no zone having a radiolucent line more than two millimeters wide, was designated as possibly unstable (Table II).
Quantitative analysis with a Sigma Scan digitizing system (Jandel Scientific, San Rafael, California) was used to assess possible acetabular migration as described by Martell et al. The diameter of the head of the femoral prosthesis was used as the standard for correction of magnification. Broken screws and radiolucent lines associated with screws were identified. Fragmentation of the fiber-metal pad from the acetabular component also was recorded. The presence of osteolytic lesions in the acetabulum was assessed on anteroposterior radiographs, and the time when any lesions were first seen was noted. The lesions were classified by location as retroacetabular, marginal (intra-articular), or associated with screws. Wear of the polyethylene liner, as seen on the serial anteroposterior radiographs of the pelvis, was measured quantitatively with use of a modification of the technique described by Livermore et al. A transparent template with concentric circles was centered on the radiograph over the head of the femur. Digitized measurements were made of the shortest radius from the center of the femoral head to the outer surface of the acetabular component and of the radius from the center of the femoral head to the outer surface of the acetabular component along a line parallel to the interteardrop line. At the latest follow-up examination, the difference between these two measurements was subtracted from the difference between the measurements made on the initial radiographs in order to determine the total measurable wear of the polyethylene liner. For the hips that had an exchange of the polyethylene liner, the time to reoperation was used as the final follow-up interval for the measurement of wear of the original liner. A Kaplan-Meier survivorship analysis was used to assess the life span of the acetabular component11,19, and all of the patients who had been lost to follow-up were included. Two analyses were performed. In the first analysis, components that were considered to be unstable or probably unstable and those that had had a revision of both the shell and the liner were defined as failed components. In the second analysis, the definition of a failed component was expanded to include components in patients who had had a reoperation for a problem related to the acetabular component, including retroacetabular osteolysis and excessive wear of the polyethylene liner. The Student t test, analysis of variance, and the Pearson test were used to determine if there was a relationship between the rate of polyethylene wear and the variables of age, gender, height, weight, body-mass index (weight in kilograms divided by the square of the height in meters), diagnosis, type of stem (Gustilo-Kyle, Harris-Galante, or Precoat), size of the metal shell, and angle of insertion of the cup. When such a relationship was demonstrated at the level of p < 0.05, that variable was used in a multiple-regression analysis with the rate of wear as the dependent variable.
Clinical Results Three hips (2 per cent) had a reoperation because of a problem related to the acetabular component, but the acetabular shell was retained. In two asymptomatic hips, the polyethylene liner of the acetabular component was exchanged because of excessive wear. In one of these hips, which was revised at eighty-five months, an acetabular component with an outer diameter of forty-six millimeters articulated with the twenty-eight-millimeter titanium head of a non-modular Gustilo-Kyle femoral component. The inclination angle of the cup on insertion was 46 degrees. At the time of the reoperation, this patient had a Harris hip score of 97 points. In the other hip, which was revised at eighty-four months, an acetabular component with an outer diameter of forty-eight millimeters articulated with the twenty-eight-millimeter cobalt-chromium head of a modular Harris-Galante femoral component. The inclination angle of the cup on insertion was 68 degrees. At the time of the reoperation, the Harris hip score for this patient was 100 points. In the third hip that needed a reoperation for a problem involving the acetabular component, débridement and grafting of an area of massive retroacetabular osteolysis was performed at eighty-nine months (Figs. 1-A, 1-B, and 1-C). At the reoperation, the lesion was found in the anterior aspect of the ilium. Direct communication between the joint cavity and the lesion was evident through an osseous lesion in the lateral cortex of the iliac bone. The femoral and acetabular components were stable. The Morse taper of the modular femoral head-and-neck assembly showed extensive corrosion, and the polyethylene liner showed only mild wear. The large, contained defect of the ilium was curetted and packed with autogenous cancellous bone. The polyethylene liner and the cobalt-chromium femoral head were exchanged. Histological analysis of the granuloma showed macrophages, polyethylene debris, and corrosion products10. At the time of the reoperation, the Harris hip score for this patient was 100 points.
In addition to the two hips that had a revision of both the acetabular and the femoral component, twenty-three hips (13 per cent) had a reoperation because of a problem that was unrelated to the acetabular component. Eighteen hips had a revision of the femoral component, three hips had excision of extensive heterotopic bone, one hip had an exchange of the femoral head in order to correct a limb-length discrepancy, and one hip had incision and drainage of a superficial infection. The 157 patients (173 hips) available for clinical review had had a mean preoperative Harris hip score of 52 points (range, 12 to 79 points), which improved to a mean of 90 points (range, 44 to 100 points) at the most recent follow-up examination. One hundred and forty-four hips (83 per cent) had a good or excellent result (80 to 100 points), nine hips (5 per cent) had a fair result (70 to 79 points), and twenty hips (12 per cent) were rated poor (less than 70 points). One hundred and forty-nine hips (86 per cent) caused little or no pain at the most recent follow-up evaluation. The reasons for the twenty poor results (in sixteen patients) were analyzed further. Nine patients (eleven hips) had a problem associated with the femur: four patients (six hips) had osteolysis of the femur (in association with an unstable femoral component in two patients [three hips], with a stable femoral component in one patient [one hip], and with one stable and one unstable femoral component in one patient [two hips]), two patients (two hips) had an unstable femoral component in the absence of osteolysis of the femur, and three patients (three hips) had a problem related to the trochanter (one had a fracture of the greater trochanter; one, a fracture of the lesser trochanter; and one, trochanteric bursitis). Of the remaining patients who had a poor result, three (five hips) had a severe spondyloarthropathy, two (two hips) had a chronic disabling medical problem, one (one hip) had disabling low-back pain, and one (one hip) was limited by severe rheumatoid arthritis. Postoperative complications were recorded for the entire series of 184 patients (204 hips). Four patients (four hips; 2 per cent) had grade-IV heterotopic ossification, according to the classification of Brooker et al. Of those, three (2 per cent) needed a reoperation for excision of the heterotopic bone. Three patients (three hips; 2 per cent) had a dislocation of the femoral component, but a reoperation was not necessary. One component dislocated once and two dislocated more than once. Two patients (two hips; 1 per cent) had a superficial wound infection; one was managed with oral antibiotic therapy and the other had incision and drainage in addition to antibiotic therapy. One patient (one hip; less than 1 per cent) had deep venous thrombosis, and one patient (one hip) had a fatal pulmonary embolism. Peroneal nerve palsy was observed in one patient (one hip) who had severe rheumatoid arthritis. Initially, the patient had a mixed sensory and motor nerve palsy. Motor activity returned, but the patient was left with chronic causalgia. One patient (one hip) who had chronic renal and hepatic disease died in the perioperative period from electrolyte imbalance. There were no deep infections. No neurovascular complications resulted from intrapelvic placement of the screws, despite the fact that postoperative radiographs of 132 hips (65 per cent) demonstrated at least one screw penetrating a minimum of two screw threads beyond the inner table of the pelvis. None of the titanium tines used for fixation of the polyethylene liner failed, and none of the liners dissociated from the metal acetabular component.
Radiographic Results Ninety-eight (59 per cent) of the 165 hips had no radiolucent lines around the acetabular component. Radiolucent lines that were one millimeter wide or less between the acetabular component and bone were common: 24 per cent of the hips had such a line in zone A1; 19 per cent, in zone A2; 7 per cent, in zone B1; 17 per cent, in zone B2; and 19 per cent, in zone C (Fig. 2). A radiolucent line that was more than one millimeter wide but less than two millimeters wide was seen in a single zone in one hip (1 per cent). No hip had a radiolucent line that was more than two millimeters wide.
A radiolucent line associated with a single screw was seen in three hips (2 per cent), at a mean of 81.7 months (range, seventy-six to ninety-one months) after the operation. None of the screws had broken. Limited fragmentation of the fiber-metal pad was seen in three hips (2 per cent), at the periphery of the cup. In two of the three hips, fragmentation of the porous coating was seen within one year after implantation of the component and it progressed only slightly during the follow-up period. In the third hip, fragmentation was first noted at seventy-four months. All three of these components were considered to be stable. Osteolysis of the acetabulum was seen in seven hips (4 per cent): four (3 per cent) of the 130 hips containing a Harris-Galante femoral component, two of the fourteen hips containing a Gustilo-Kyle femoral component, and one (5 per cent) of the twenty-one hips containing a Precoat femoral component. Six (4 per cent) of the hips had limited marginal (intra-articular) osteolysis. The seventh hip (1 per cent) had radiographic evidence of early retroacetabular osteolysis (at sixty-eight months), which advanced extensively until reoperation at eighty-nine months, as described (Figs. 1-A, 1-B, and 1-C). The mean time to the radiographic appearance of the lesions in the seven hips was seventy-seven months (range, forty-two to 104 months). None of the hips had osteolysis associated with the acetabular screws. The radiographs of the thirty-four patients (thirty-nine hips) who had less than seventy-eight months of follow-up also were reviewed. Twenty-three of these patients (twenty-eight hips) had been followed for at least twelve months (mean, fifty months; maximum, seventy-five months). None of them had a revision of the acetabular component. In addition, none of the components were considered to be unstable or probably unstable, and none were associated with a complete radiolucent line between the acetabular component and bone. One acetabular component was considered to be possibly unstable. No acetabular osteolysis or broken screws were found, and no radiolucent line was noted adjacent to any screw. The mean amount of wear of the polyethylene liner at the time of follow-up was 0.98 millimeter (range, zero to 4.51 millimeters). The mean rate of wear was 0.11 millimeter per year (range, zero to 0.59 millimeter per year). Statistical analysis indicated that the rate of wear of the polyethylene liner was associated with the gender of the patient (men, 0.14 ± 0.10 millimeter per year [mean and standard deviation], and women, 0.11 ± 0.10 millimeter per year [p = 0.055, Student t test]), type of stem (Gustilo-Kyle, 0.17 ± 0.18 millimeter per year; Harris-Galante, 0.13 ± 0.10 millimeter per year; and Precoat, 0.06 ± 0.07 millimeter per year [p < 0.001, analysis of variance]), diagnosis (avascular necrosis, 0.15 ± 0.10 millimeter per year; osteoarthrosis, 0.12 ± 0.11 millimeter per year; and rheumatoid arthritis, 0.06 ± 0.07 millimeter per year [p < 0.01, analysis of variance]), and age (negative correlation, p = 0.0001, r2 = 0.16). However, there were strong associations among the parameters of gender, type of stem, diagnosis, and age that precluded meaningful conclusions from multiple-regression analysis. Multiple-regression analysis revealed that only the type of stem was significantly associated with the rate of wear (p < 0.001). No association was found between the rate of wear and the variables of body-mass index, height, weight, size of the metal shell, and angle of insertion of the cup. Patients who had femoral osteolysis had a significantly higher mean rate of wear than those who did not (0.17 ± 0.11 millimeter per year compared with 0.11 ± 0.10 millimeter per year [p < 0.005, Student t test]). With the numbers available, this was not true for the seven patients (seven hips) who had acetabular osteolysis (0.19 ± 0.13 millimeter per year compared with 0.12 ± 0.10 millimeter per year [p = 0.10, Student t test]). A Kaplan-Meier survivorship curve revealed a 99 per cent chance of survival of the acetabular component at 120 months (95 per cent confidence interval, 0.98 to 1.0) when a failed component was defined as one that had been considered to be unstable or probably unstable or as one that had had revision of both the metal shell and the liner. When the definition of a failed component was expanded to include hips that had had a reoperation because of any problem related to the acetabular component (retroacetabular osteolysis or excessive wear of the polyethylene liner), the acetabular component had a 97 per cent chance of survival at 120 months (95 per cent confidence interval, 0.95 to 1.0). Of the twelve hips that were followed for more than 120 months, one was revised (at 123 months). The reoperation, which was performed at another institution, consisted of revision of the femoral component as well as the stable acetabular component. The cup was forty-four millimeters in diameter, and the operating surgeon preferred to replace it with a larger cup in order to provide a thicker polyethylene insert.
The purpose of the present study was to evaluate the intermediate-term results of a consecutive series of primary total hip arthroplasties with insertion of a Harris-Galante porous-coated acetabular component without cement. In order for implants inserted without cement to be adopted as the preferred component for acetabular reconstruction, they must demonstrate better longevity than acetabular components inserted with cement. In addition, they must demonstrate a low rate of complications, such as osteolysis, that can jeopardize a stable implant and lead to failure. The results in our study compare favorably both with those reported for acetabular components inserted with cement and with those reported for acetabular components inserted without cement. Loosening of acetabular components that have been inserted with cement is time-dependent and may be a biological phenomenon24. Rates of loosening of 29 per cent (twenty-nine of 100 hips) at ten years and 25 per cent (twenty-eight of 112 hips) at twelve to fifteen years have been reported previously2,27. Schmalzried and Harris reported the five-year results after eighty-three arthroplasties with a Harris-Galante type-I porous-coated acetabular component implanted without cement21. None of the acetabular components were revised because of loosening, and none were radiographically loose. In the present study, one acetabular component, in an acetabulum that had been reconstructed with a bulk allograft from the femoral head, migrated early. As the cup is fixed by ingrowth of host bone, not allograft bone, the instability of this component is not surprising. A complete radiolucent line of one millimeter or less in width was observed in 1 per cent of the hips in the present report. Although these implants did not migrate, this is a suboptimum interface that probably represents soft-tissue ingrowth. Over-all, the prevalence of possibly unstable sockets was 5 per cent (eight of 165 hips). This compares very favorably with the results, after a similar duration of follow-up, for acetabular components inserted with cement. These low rates of aseptic loosening are not surprising, as a study of this type of socket demonstrated acetabular bone ingrowth in ten of eleven components that had been implanted in primary total hip replacement and subsequently retrieved at autopsy17. Other short-term studies of acetabular implants inserted without cement have demonstrated a low rate of loosening, but the reported rates of osteolysis of nineteen (17 per cent) of 113 hips to seven of nineteen hips have raised concern13,22,25,28. The over-all prevalence of osteolytic lesions (retroacetabular, marginal, or those associated with screws) seen in the present series was 4 per cent (seven) of 165 hips. The difference in the rates of osteolysis among the various designs of the implant may be related to a number of variables, including the diameter of the femoral head, the thickness of the polyethylene insert, the degree of tolerance between the polyethylene insert and the metal shell, the quality of the polyethylene, and the stability of the polyethylene insert9. In the present series, in which only anteroposterior radiographs of the pelvis were analyzed for the presence of osteolysis, periacetabular osteolytic lesions were first recognized at a mean of seventy-seven months postoperatively. As osteolysis is a time-dependent phenomenon, its prevalence may increase with time. Therefore, osteolysis must be recognized as a potential mode of late failure of porous-coated acetabular components inserted without cement. We recognize that the true prevalence of osteolysis may have been somewhat higher given that some lesions can be recognized only on an oblique radiograph21. The mean rate of wear of the polyethylene liner in the present series was 0.11 millimeter per year. This rate was similar to that reported for other total hip replacements with a cobalt-alloy femoral head and an ultra-high molecular weight polyethylene socket9. However, the subset of patients who had a non-modular titanium-alloy femoral head inserted without cement had a higher rate of wear (0.17 millimeter per year), attesting to the relatively poor performance of untreated titanium-alloy bearings. In the present study, the prevalence of femoral osteolysis was associated with the magnitude of wear of the polyethylene liner, whereas the prevalence of acetabular osteolysis was not. This difference may possibly have been due to the small number of patients who had an acetabular osteolytic lesion. In summary, the Harris-Galante type-I acetabular component demonstrated low rates of loosening and osteolysis at a minimum of seven years after the operation. The observed rate of radiographic loosening was less than that reported for acetabular components inserted with cement in other studies that had a similar duration of follow-up2,7,16,27. In addition, the rate of osteolysis compared favorably with that reported for other acetabular implants inserted without cement22,25,28. Although the intermediate-term results with this implant are excellent, the long-term problems with it may be secondary to particulate debris and wear of the polyethylene liner rather than to loss of fixation.
NOTE: The authors thank Elise Matuch and Phyllis Velez for their assistance in the preparation of this manuscript and Professor Sue Leurgans, Director of the Section of Biostatistics, Department of Preventive Medicine, Rush-Presbyterian-St. Luke's Medical Center, for her assistance in the statistical analysis of the data.
*One or more of the authors has received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article. In addition, benefits have been or will be directed to a research fund, foundation, educational institution, or other non-profit organization with which one or more of the authors is associated. Funds were received in total or partial support of the research or clinical study presented in this article. The funding sources were Zimmer, Incorporated, Warsaw, Indiana, and the Musculoskeletal Research Foundation.
  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. |