The long-term success of primary cementless total hip arthroplasties (THA) using the fully porous-coated Anatomic Medullary Locking stem (AML; DePuy, Warsaw, IN) has been reported.10,11 Despite the success of the AML, however, several modifications were incorporated into the design of the second-generation stem with the goals of further improving implant stability, adaptability to variations in anatomy, and stress transfer to periprosthetic bone.
The second-generation porous-coated stem manufactured by DePuy, the Prodigy® stem, is cast cobalt-chromium (CoCr) with circumferential porous coating covering the proximal ⅞ of the component (Fig 1). To improve stability and biomechanics, the Prodigy® stem incorporated four design additions: (1) wider range of neck lengths, (2) varus and neutral neck shaft angles, (3) increased lateral offset, and (4) a 10° anteverted neck. The additional neck lengths and two neck-shaft angles allow the surgeon to accommodate a wider variety of patient anatomies. The increased lateral offset and anteverted neck were designed to improve hip biomechanics by reducing joint reaction force and optimize range of motion (ROM) and stability. Additionally, the Prodigy® stem is produced in three proximal stem dimensions (proximal triangles) to optimize individual patient fit, theoretically helping increase the consistency of bone ingrowth in a larger population. The Prodigy® stem design also attempts to address one of the few concerns reported with the first-generation AML stem: periprosthetic bone loss related to stress shielding in hips implanted with large, rigid stems.5-8 Modulus mismatch between the implant and femur has been thought to be a probable cause of this and of thigh pain.5,6,17 To address these issues, the shape of Prodigy® stems greater than 13.5 mm in diameter was modified by removing metal from its medial side, rendering its mediolateral bending stiffness equivalent to a 13.5-mm stem thereby theoretically increasing load transfer to the femur and decreasing periprosthetic bone loss (and potentially thigh pain).
On the acetabular side, the first-generation, porous-coated, modular AML cup (DePuy) with ACS polyethylene liners (Acetabular Cup System, DePuy) was replaced by the second-generation Duraloc® (DePuy) acetabular component. The Duraloc® cup is a hemispheric titanium fully porous-coated modular component (Fig 2). Design changes made with the Duraloc® attempted to address the first-generation component problems, which reportedly contributed to a 21% incidence of clinical failure at 2-years after arthroplasty.2 Changes included an improved locking mechanism designed not to interfere with liner-shell conformity, dome-loading polyethylene, and a required minimum polyethylene thickness of 6 mm.
We hypothesized that the second-generation Prodigy® femoral component would perform at least as well as the first-generation AML stem with respect to revision rates and survivorship, but design changes in the second-generation femoral component would result in (1) more consistent bone ingrowth, (2) less pain and limp, and (3) less stress shielding. We hypothesized that the design changes incorporated into the second-generation Duraloc® cup would decrease (1) component failure, (2) polyethylene wear, and (3) osteolysis as compared with the first-generation porous-coated modular AML cup with an ACS liner. Additionally, we expected high patient satisfaction and a patient-perceived improvement in hip function with the use of these second-generation porous-coated components.
MATERIALS AND METHODS
We retrospectively reviewed clinical, radiographic, and patient questionnaire data from 157 consecutive, unselected THAs (150 patients) performed with Prodigy® stems and Duraloc® acetabular cups from July 1993 to December 1994. All patients who had primary THAs were considered candidates for these components. There were 90 men (94 hips) and 60 women (63 hips). The average age of the patients at the time of surgery was 56.2 years (range, 20-80 years). The preoperative diagnoses was osteoarthritis in 119 hips (116 patients), osteonecrosis of the femoral head in 13 hips (12 patients), rheumatoid arthritis in one hip (one patient), developmental dysplasia of the hip in 18 hips (15 patients), and fracture in six hips (six patients).
Inclusion criteria for the study were patients with THAs performed with the Prodigy® and Duraloc® components with a minimum 5-year clinical or radiographic followup. Patients whose outcome was known before the 5-year followup also were included in our survivorship analysis. This included patients with a successful outcome who died, or patients who had a failed outcome (that is, revision), before the 5-year followup. Of the 157 THAs in the series, 12 hips (12 patients) with less than 5 years followup were lost and, therefore, not included in our analysis. The remaining 145 hips (138 patients), or 87% of the eligible patients, comprised our study population and were followed up for an average of 6.7 years (range, 0.1-8.2 years). Three patients (three hips) died 0.1 to 4.3 years after the primary surgery, and two patients (two hips) had revision surgeries at 0.1 and 7.2 years, respectively. Excluding these five hips, the average followup for the 140 hips was 6.8 years (range, 5-8.2 years). Of the 143 unrevised hips, 138 had a minimum of 5 years of radiographic followup, 139 had minimum 5 years of patient-assessed outcome data from a patient questionnaire (described below), and 140 had a minimum of 5 years of physician-assessed followup.
Two types of Duraloc® cups were used. The Duraloc® 100 is a hemispheric, titanium alloy porous-coated design, with a central hole to accommodate an insertion device, which subsequently was filled with a hole eliminator (a threaded titanium insert). The Duraloc® 1200 was similar to the Duraloc® 100 except for the presence of 12 holes in the dome of the shell to allow additional screw fixation. The Duraloc® 100 was the default cup choice. If a patient had questionable bone quality, potentially requiring adjunct screw fixation, the Duraloc® 1200 cup was used. In our population, this occurred once; however, no screws actually were used. Duraloc® 100 cups were implanted in 144 of the hips (99%), and the Duraloc® 1200 was implanted in one hip (1%). The following head sizes were used: one 22-mm head (1%), four 26-mm heads (3%); 136 28-mm heads (93%), and four 32-mm heads (3%). Sixty-eight femoral heads were ceramic and 77 were CoCr. All polyethylene liners were sterilized using gamma irradiation in air. Highly crystalline polyethylene liners (Hylamer, DePuy) were used in 87 of 145 hips (60%), and conventional polyethylene liners (Enduron, DePuy) liners were used in 58 of 145 hips (40%).
The posterior approach was used in 126 (87%) hips, and a modified direct lateral approach16 was used in 19 (13%) hips. No formal posterior capsular repair was performed when the posterior approach was used. Routinely, the surgeon reamed the femoral canal with rigid reamers to a diameter 0.5 mm smaller than that of the component diameter. When there was poor bone quality, the canal was reamed to the same size as the component diameter (line to line). The acetabulum was reamed to 1 mm less than the outer shell diameter. The procedures were performed by three surgeons (CAE Sr., CAE Jr., GAE). Trial components were used in all cases. All patients were instructed to maintain 50% weightbearing for 6 weeks postoperatively.
Clinical and radiographic evaluations were performed preoperatively and postoperatively at 6 weeks, 4 months, 6 months, and 1 year with annual or biannual followup thereafter. Radio-graphs included an anteroposterior (AP) view of the pelvis and AP and lateral views of the femur. One independent surgeon (CJC) who was not involved with the primary procedures interpreted the radiographs. Bone ingrowth was assessed according to the criteria of Engh et al,9 and stress shielding was assessed according to the criteria of Engh and Bobyn.5 Cup loosening was defined as migration greater than 5 mm or greater than 5° change in tilt relative to a line defined by the inferior border of the pelvic teardrops. Radiolucent lines in all DeLee and Charnley zones4 were considered suggestive but not definitive of loosening unless migration (as defined above) occurred. Radiographs also were assessed for femoral or acetabular osteolytic lesions greater than 1.5 cm2 as determined by measuring the maximum dimensions of each lesion in mutually perpendicular directions. DeLee and Charnley zones4 were used to categorize lesions adjacent to the acetabular component, whereas Gruen zones12 were used for femoral lesions.
For hips with 5-year AP radiographsof the pelvis and immediately postoperative AP radiographs of the pelvis, two-dimensional head penetration data were assessed using a computer-assisted technique described by Sychterz et al.21 For 111 hips that had at least three postoperative serial AP radiographs (one postoperative and at least two annual followup radio-graphs), true wear rates were calculated according to a linear regression technique described by Sychterz et al.22
Patient satisfaction with the procedure, function, and location and severity of pain were assessed from patient responses to a written, institution-specific questionnaire used by the senior author for the past 15 years as referenced in previous studies.11,14,15 As part of the written questionnaire administered in the office waiting room by a clinical assistant, patients were asked to mark yes or no to “Has your new hip increased your function and daily activity?” and “Are you satisfied with the results of your hip operation?” Patients also were asked to “Indicate the amount of pain you experience normally” with choice of answers as (1) no pain, (2) slight pain, (3) mild pain -no affect on average activity, (4) moderate pain -affects activity somewhat, (5) severe pain, or (6) intolerable pain. However, not all respondents answered all of the questions. Fifteen patients did not rate hip pain, which left 117 respondents. If a patient was unable to return to our institution for followup, we attempted to complete a patient questionnaire by telephone and have standardized radiographs sent for review. Limp was assessed by the operating surgeon as part of the clinical evaluation.
Statistical analyses were performed using SPSS software (SPSS Version 8.0, SPSS Inc, Chicago, IL). Component survivorship was computed using the Kaplan-Meier method with any revision as an endpoint and then with radiographic loosening as an end point. The independent samples t test was used to compare the mean wear rate, liner shelf life, and age between the Hylamer and Enduron liner groups. Probability values less than 0.05 were considered indicative of statistical significance.
With two of the 145 hips (1.4%) in this population revised, Kaplan-Meier survivorship using any revision as an end point showed a 99% survival rate at 5 years postoperatively (95% confidence interval [CI], 98.1-100%). Eight-year survivorship using revision as an end point was 97.2% (95% CI, 92.8-100%). One of the two revised hips had reoperation 6 weeks after the index surgery for recurrent dislocation. In this patient, the stable femoral component and stable metal shell were retained but the neutral polyethylene liner was exchanged for a liner with a 10° lip and the femoral head size was increased to 32 mm. This patient has had no more instances of instability. The second revised hip had reoperation 7 years after the index procedure for unstable fibrous fixation of the stem, subsidence, and subsequent dislocation. Intraoperative assessment of the components revealed a solidly fixed acetabular shell and a grossly loose femoral stem. The stem was revised to a larger Prodigy® component with a 36-mm head, and the polyethylene liner was exchanged.
Radiographic signs of femoral bone ingrowth (proximal cortical bone atrophy, presence of spot welds) were evident in 136 of 138 hips (98.6%) and were absent in two hips (1.4%). Of the two hips with no signs of bone in-growth, one had reoperation at 7 years (as stated above) and the second was graded as loose, had 15 mm of progressive subsidence, but was not symptomatic enough to need revision at the time of this review. With radiographic loosening as an end point, the 5-year survivorship was 99.3% (95% CI, 98-100%), and the 8-year survivorship was 97.2% (95% CI, 92.8-100%).
Seven percent of patients (nine of 123 hips) reported hip pain considered activity limiting (moderate or severe), 5% of patients (seven of 140) were observed by the physician to have a slight limp, and an additional 4% (four of 140) were observed to have a moderate limp.
Eighty three percent of hips (115 hips) had visible plain radiographic evidence of stress shielding (evidence of cortical thinning, increased cortical or trabecular porosity, or decreased cortical or trabecular density). Stress shielding was graded as mild in 63% (88 hips), moderate in 16% (72 hips), and severe in 4% (five hips).
All acetabular components were stable at last followup and no components failed. However, six of the 145 hips (4.1%) experienced dislocation between 3 days and 7 years postoperatively. As previously stated, two of these hips had revision surgery, one for polyethylene exchange and the other for stem revision with polyethylene exchange; neither experienced subsequent dislocation. Additionally, three of the 12 hips lost to followup were known to have experienced at least one dislocation, bringing the number of known dislocated hips to nine of 148 (6.1%). Of the seven hips that were dislocated and not revised, four had only one dislocation, and three had multiple dislocations. All dislocations occurred in hips in which a postero-lateral approach was used. When considering surgical approach, the dislocation rate was 7.0% (nine of 129) for the posterolateral approach and 0% (0 of 19) for the modified direct lateral approach.
The mean true wear rate was 0.10 ± 0.14 mm/year and the median true wear rate was 0.08 mm/year. We found no relationship between the true wear rate and patient age. The mean true wear rate for Hylamer liners was similar to the Enduron liners (0.10 ± 0.15 versus 0.09 ± 0.11 mm/year, respectively, despite the Hylamer liners being implanted into a younger (p < 0.001) group of patients (51.4 versus 62.6 years, respectively).
Acetabular osteolytic lesions greater than 1.5 cm2 were found in three hips (2%) and were located in Charnley and DeLee Zone 2. The mean true wear rate for these hips was 0.18 mm/year. In all three hips, the acetabular component remained stable and the apex hole plug migrated into the lytic lesion (Fig 3). Seven hips (5%) had evidence of femoral osteolytic lesions greater than 1.5 cm2. Three hips had a large lesion in Gruen Zone 1, and the other four hips had a large lesion in Zone 7. The mean true wear rate for the seven hips with femoral lesions greater than 1.5 cm2 was 0.22 mm/year. We saw no evidence of distal osteolysis including the area around the cutout in the stem.
Satisfaction with the outcome of their total hip replacement was reported by 118 of 121 respondents (98%, 125 of 128 hips). Improved hip function was seen in 116 of 120 respondents (97%, 123 of 127 hips).
Our series of THAs is the second-generation of the AML stem coupled with a second-generation porous-coated modular cup. To our knowledge, it is the only series of cementless THAs in which design changes in hip components can be compared with the initial constructs, and in which the operations were performed at one institution.
Our study has several limitations. This retrospective study lacked a control group from the same interval, two different types of polyethylene liners were used, and the components were implanted with two surgical approaches. Despite these potentially confounding factors, however, survival was high, suggesting they did not influence the results in the medium term. Moreover, certain measures, such as radiographic outcome parameters, are prone to interobserver variability, but parameters such as revision rates and survivorship analysis are not dependent on such error. We attempted to eliminate some of the variability associated with radiographic interpretations by using only one surgeon not associated with the primary THA. However, we acknowledge variability in assessing parameters such as stress shielding and osteolysis still exist and, therefore, caution must be used when comparing results among different studies. To reduce variability further, radio-graphic results from the current study were compared only with results of previous studies from our institution as consistent radiographic technique and patient positioning at our institution have made radiographs more comparable across time. Finally, we recognize comparing our results to results of previously published studies precludes any statistical analysis, and as such, we cannot be certain whether differences between our results and historical results are significant. Obviously this limitation applies to any study using historical controls.
Revision rates with the second-generation components compared favorably whereas bone ingrowth seemed superior to that of the first-generation prostheses (Table 1). Our revision rate at an average of 6.7 years was 1.4%. Two hips in our cohort were revised, one for dislocation 6 weeks postoperatively and one for loosening 7 years after the index THA. In a series of 393 primary THAs using the AML stem with similar followup, the femoral revision rate was 1.5%; three stems were revised for loosening, two for stem fracture, and one for infection.7 Additionally, in our study, 98.6% of the hips were bone ingrown at 5 years according to the criteria described by Engh et al.9 In a previous review of AML prostheses, only 78% of the hips achieved bone ingrowth at 2 years after THA.8 In that study, 17% of hips were graded as stable with fibrous fixation, and 5% were graded as loose. Possible reasons for improved osseointegration include the Prodigy® offering a greater selection of stem-size options and a greater area of porous coating than the first-generation component, and a change in implantation technique from these earlier studies. We now under ream by 0.5 mm or ream the medullary canal line-to-line to improve the canal fill. This differs from previously reported series using the AML stem in which 20% to 42% of the stems were undersized.8,11
Pain and limp also seemed to improve with the use of the Prodigy® stem, decreasing by nearly half the amount reported in previous AML series (Table 1). In that series, 7% of the respondents reported activity limiting pain and 9% had a slight or moderate limp. In a previous minimum 2-year followup study of 307 THAs using the AML stem, 14% of the patients had thigh pain and 21% had a physician-detected limp.6 The minimum 10-year results of the same series revealed 13% of patients had activity-limiting pain.11 Results of pain and limp with the Prodigy® stem also compared favorably with results of other first-generation porous-coated stems. A 15% prevalence of thigh pain and 11% prevalence of moderate or severe limp were reported by Heekin et al in a 5-to 7-year followup study of a series of 100 THAs performed with a first-generation proximally porous-coated prosthesis (Porous Coated Anatomic stem; Howmedica, Rutherford, NJ).13 At 8 to 11 years followup, Archibeck et al reviewed 92 THAs using a second-generation circumferential proximally porous-coated prosthesis (Anatomic Hip; Zimmer, Warsaw, IN).1 Eleven of the patients had activity-limiting pain, and 5% had a moderate or severe limp.
Despite hypothesizing that a reduction in mediolateral bending stiffness for large (greater than 13.5 mm in diameter) Prodigy® stems would reduce the incidence of stress shielding, 20% of the hips had moderate or severe stress shielding. In a study of the AML stem, 14% of hips (Table 1) had moderate or severe stress shielding at 5-years after THA.8 The lack of improvement in stress shielding despite reducing stem stiffness suggests stem stiffness may not be the most important variable affecting bone loss after THA. These findings are consistent with those reported by Sychterz and Engh who suggested stress shielding is more influenced by the density of the bone into which a stem is implanted than the stiffness of the stem.20
Our results for the Duraloc® acetabular component were encouraging and represented an improvement from previous short-term and long-term studies of this component and its predecessor (Table 1). In the current study, there were no loose cups or radiographic evidence of cup migration. By contrast, acetabular component failure with the first-generation cup and ACS liner was 21% at a mean followup of 3.6 years.2 Our results also differed from those of a series of Duraloc® 100 cups assessed using Ein Bild Röntgen Analyse (EBRA).18 With cup loosening defined as migration of 1 mm or greater, EBRA analysis judged 48% of Duraloc® cups (30 of 63 cups) to be loose at 2 years followup. The difference in criteria used for cup loosening between our study and the EBRA study could account for the differences in our results. As shown in another study,19 1 mm is within the range of measurement error when considering pelvic rotation differences on radiographs and errors in locating the hip center. Therefore, we think a more reliable measure of loosening occurs after 5 mm of change.
Although 60% of the polyethylene liners included in our study were made of Hylamer, a polyethylene previously shown to have increased wear,23 wear rates in our study were not abnormally high. The mean wear rate at 6.7 years after THA was 0.10 mm/year, with the rate for Hylamer liners (0.10 mm/year) being very similar to Enduron liners (0.09 mm/year). These rates compared well with those in to a study of 127 first-generation cups with ACS liners that had a mean wear of 0.08 mm/year at 6.4 years.22 It is unclear why Hylamer liners in our study did not show increased wear, especially because Hylamer liners were implanted into a statistically younger group of patients. Perhaps the shelf lives of these liners gamma-irradiated in air contributed to their comparable wear rates. The liners we used had relatively short shelf lives. The mean shelf life for the Hylamer liners was 0.6 years and the mean shelf life for Enduron liners was 1.2 years. Typically, deterioration of the mechanical properties of gamma-irradiated in air polyethylene is associated with shelf lives greater than 1 year.3 We will continue to follow up these patients, as a difference in wear rates may begin to appear with longer followups.
In concurrence with the normal mean wear rates and the relatively short followup for assessing the presence of osteolysis, only a small percentage of patients in the current study had radiographic evidence of osteolytic lesions. Five percent of the hips had femoral osteolytic lesions at least 1.5 cm2 with all lesions being confined to the proximal periarticular regions. Osteolysis adjacent to the acetabular component occurred in only 2% of hips. All of these cases had migration of the apex hole eliminator into the lytic cyst. Using the AML construct, patients in a previous study at slightly longer mean followup (8.75 years) had a 12% prevalence of femoral osteolysis and 18% prevalence of acetabular osteolysis.24 After the completion of our series, a Duraloc® cup was developed with a positive stop to prevent the inadvertent over insertion and progressive advancement of the plug. We expect this feature will eliminate migration of the apex hole plug and minimize a potential pathway for particulate debris.
Early followup of the Prodigy-Duraloc® combination in primary cementless THA revealed encouraging clinical and radiologic results. The second-generation Prodigy® component had similar clinical results compared with its predecessor, the AML component. The ability to achieve bone ingrowth with the Prodigy® stem was better than with the first-generation stem. Revision rates and stress shielding also compared favorably with those in other series in which the AML femoral component was used.7,8 The performance of the second-generation Duraloc® ace-tabular component at 5 years after surgery was superior to performance of the previous AML and ACS cup designs. Longer followup is required to provide a more complete understanding of the effects of the modifications made to the Prodigy® and Duraloc® components and to learn if these changes will translate into improved long-term results, particularly regarding polyethylene wear and osteolysis.
We thank C. Anderson Engh, Jr., MD, and Gerard A. Engh, MD, for contributing cases to our study and Christi Sychterz Terefenko for technical contributions and assistance with revision to the manuscript.
1. Archibeck MJ, Berger RA, Jacobs JJ, Quigley LR, Gitelis S, Rosenberg AG, Galante JO. Second-generation cementless total hip arthroplasty: eight to eleven-year results. J Bone Joint Surg Am
2. Bono JV, Sanford L, Toussaint JT. Severe polyethylene wear in total hip arthroplasty: observations from retrieved AML PLUS hip implants with an ACS polyethylene liner. J Arthroplasty
. 1994;9: 119-125.
3. Currier BH, Currier JH, Collier JP, Mayor MB, Scott RD. Shelf life and in vivo duration: impacts on performance of tibial bearings. Clin Orthop Relat Res
4. DeLee JG, Charnley J. Radiologic demarcation of cemented sockets in total hip replacement. Clin Orthop Relat Res
5. Engh CA, Bobyn JD. The influence of stem size and extent of porous coating on femoral bone resorption after primary cementless hip arthroplasty. Clin Orthop Relat Res
6. Engh CA, Bobyn JD, Glassman AH. Porous-coated hip replacement: the factors governing bone ingrowth, stress shielding, and clinical results. J Bone Joint Surg Br
7. Engh CA, Hooten JP, Zettl-Schaffer KF, Ghaffarpour M, McGovern TF, Macalino GE, Zicat BA. Porous-coated total hip replacement. Clin Orthop Relat Res
8. Engh CA, Massin P. Cementless total hip arthroplasty using the Anatomic Medullary Locking stem. Clin Orthop Relat Res
9. Engh CA, Massin P, Suthers KE. Roentgenographic assessment of the biologic fixation of porous-surfaced femoral components. Clin Orthop Relat Res
10. Engh CA Jr, Claus AM, Hopper RH, Engh CA. Long term results using the Anatomic Medullary Locking hip prosthesis. Clin Orthop Relat Res
11. Engh CA Jr, Culpepper WJ2nd, Engh CA. Long-term results of use of the anatomic medullary locking prosthesis in total hip arthroplasty. J Bone Joint Surg Am
12. Gruen TA, McNeice GM, Amstutz HC. Modes of failure of cemented stem-type femoral components: a radiographic analysis of loosening. Clin Orthop Relat Res
13. Heekin RD, Callaghan JJ, Hopkinson WJ, Savory CG, Xenos JS. The porous-coated anatomic total hip prosthesis, inserted without cement. Results after five to seven years in a prospective study. J Bone Joint Surg Am
14. McAuley JP, Culpepper WJ, Engh CA. Total hip arthroplasty: concerns with extensively porous coated femoral components. Clin Orthop Relat Res
15. McAuley JP, Moore KD, Culpepper WJ2nd, Engh CA. Total hip arthroplasty with porous-coated prostheses fixed without cement in patients who are sixty-five years of age or older. J Bone Joint Surg Am
16. Mulliken BD, Rorabeck CA, Bourne RB, Nayak N. A modified direct lateral approach in total hip arthroplasty. J Arthroplasty
17. Skinner HB, Curlin FJ. Decrease pain with lower flexural rigidity of uncemented femoral prostheses. Orthopedics
18. Stockl B, Sandow M, Krismer M, Biedermann R, Wimmer C, Frischhut B. Migration of the Duraloc cup at two years. J Bone Joint Surg Br
19. Sutherland CJ, Wilde AH, Borden LS, Marks KE. A ten-year follow-up of one hundred consecutive Muller curved-stem total hip replacement arthroplasties. J Bone Joint Surg Am
. 1982;64: 970-982.
20. Sychterz CJ, Engh CA. The influence of clinical factors on peri-prosthetic bone remodeling. Clin Orthop Relat Res
. 1996;322: 285-292.
21. Sychterz CJ, Engh CA Jr, Shah N, Engh CA. Radiographic evaluation of penetration by the femoral head into the polyethylene liners over time. J Bone Joint Surg Am
22. Sychterz CJ, Engh CA Jr, Yang A, Engh CA. Analysis of temporal wear patterns of porous-coated acetabular components: distinguishing between true wear and so-called bedding-in. J Bone Joint Surg Am
23. Sychterz CJ, Young AM, McAuley JP, Engh CA. Comparison of head penetration into Hylamer and Enduron polyethylene liners. J Arthroplasty
24. Zicat B, Engh CA, Gokcen E. Patterns of osteolysis around total hip components inserted with and without cement. J Bone Joint Surg Am