The purpose of this study was to compare oxinium versus metal-on-polyethylene wear in two consecutive prospective randomized series of low friction total hip arthroplasty at a minimum 10-year follow-up.

A total of 100 patients with a median age of 60.9 years were randomized to receive either oxinium (50 hips) or metal (50 hips) femoral head. The polyethylene socket was EtO sterilized in the first 50 patients, whereas it was highly cross-linked and remelted (XLPE) in the following 50 patients. The primary criterion for evaluation was linear head penetration measurement using the Martell system by an investigator blinded to the material. Also, a survivorship analysis was performed using wear related loosening revised or not as the end point.

Complete data were available for analysis in 40 hips at a median follow-up of 12.9 years (11 to 14), and in 36 hips at a median follow-up of 12.3 years (10 to 13) in the EtO sterilized and XLPE series, respectively.

In the EtO sterilized series, the mean steady-state wear rate was 0.245 ± 0.080 mm/year in the oxinium group versus 0.186 ± 0.062 mm/year in the metal group (p = 0.009). In the XLPE series, the mean steady-state wear rate was 0.037±0.016 mm/year in the oxinium group versus 0.036±0.015 mm/year in the metal group (p = 0.94). The survival rate at 10 years was 100% in both XLPE series, whereas it was 82.9% (IC 95%, 65–100) and 70.5% (IC95%, 50.1–90.9) in the metal-EtO and oxinium-EtO series, respectively.

This RCT demonstrated that up to 14-year follow-up, wear was significantly reduced when using XLPE, irrespective of the femoral head material. Also, no osteolysis related complication was observed in the XLPE series. In the current study, oxinium femoral heads showed no advantage over metal heads and therefore their continued used should be questioned related to their cost.

The purpose of this study was to compare the effect of femoral head material (delta ceramic versus metal) on polyethylene wear in a consecutive prospective randomized series of low friction total hip arthroplasty.

A total of 110 patients with a mean age of 60.6 ± 9.3 (34–75) years were randomized (power of 90%, alpha of 5%) to receive either a metal (55 hips) or a delta ceramic (55 hips) femoral head. The polyethylene socket was moderately cross-linked (3 Mrads of gamma radiation in nitrogen) and annealed at 130°C in all hips. All other parameters were identical in both groups. The primary criterion for evaluation was linear head penetration measurement using the Martell system, performed by an investigator blinded to the material of the femoral head. Creep and steady state wear values were calculated.

At the minimum of 3-year follow-up, complete data were available for analysis in 38 hips at a median follow-up of 4.4 years (3.0 to 5.7), and in 42 hips at a median follow-up of 4.0 years (3.0 to 5.4) in the metal and delta ceramic groups, respectively. The mean creep, measured as the linear head penetration at one year follow-up, was 0.42 ± 1.0 mm in the metal group versus 0.30 ± 0.81 mm in the delta ceramic group (Mann and Whitney test, p = 0.56). The mean steady state penetration rate from one year onwards measured 0.17 ± 0.44 mm/year (median 0.072) in the metal group versus 0.074 ± 0.44 mm/year (median 0.072) in the delta ceramic group (Mann and Whitney test, p = 0.48). No case of delta ceramic femoral head fracture was recorded, and no hip had signs of periprosthetic osteolysis.

This study demonstrated that up to 5-year follow-up, delta ceramic femoral head did not significantly influence creep neither wear of a contemporary annealed polyethylene. Longer follow-up is necessary to further evaluate the potential clinical benefits of delta ceramic.

Purpose of the study: The purpose of this prospective study was to evaluated the risk of fracture of 22.2 mm Delta ceramic heads.

Material and methods: A preclinical study was performed on twenty 22.2 mm Delta ceramic femoral heads with a medium neck with 20 22.2 mm Delta ceramic femoral heads with a short neck. A V40TM cone was used in all cases with a 5 40 angle. In vitro tests consisted in the assessment of the fracture force under static pressure before and after stress tests, and with a static force shock test (Charpy model) simulating a microseparation during subluxation phenomena. Between April 2007 and April 200, a consecutive series of 55 composite Delta heads were used in 55 patients undergoing cemented total hip arthroplasty (THA). A polyethylene cup was sterilised under vacuum at 3 Mrads with a post-radiation temper (Duration^®), and a stainless steel femoral piece with a highly polished surface and a V40TM Morse cone (Legend^®). This series issued from a randomised prospective study designed to compare wear with 22.2mm Delta ceramic heads with stainless steel heads with the same diameter. Inclusion criteria were age < 75 years, degenerative disease on naive hip and patient residing in France. Mean patient age was 59.2±6.9 years (range 44–70).

Results: The resistance of the 22.2 mm heads was significantly less during the static tests before and after the stress tests. Nevertheless, the mean resistance was higher than the FDA recommendations of 46K. The tests simulating a microseparation showed a significantly superior resistance for the 22.2 mm heads. All patients had from 1 to 2 years follow-up (usual delay for 80% of in vivo ceramic fractures). There were no cases of femoral head fracture in this series. There were no cases of early wear at this same follow-up and no case of femoral or acetabular osteolysis.

Discussion and Conclusion: The results of this study indicate that the resistance of the 22.2mm ceramic Delta heads is very much superior to the recommendations for in vitro tests. In this series, the risk of fracture in vivo remained nil to two years follow-up. The pertinence of this ceramic implant for decreasing polyethylene wear in vivo is under evaluation.

Purpose of the study: The purpose of this prospective randomised study was to evaluate the wear of cemented polyethylene cups as a function of the material of the femoral head: oxinium versus metal.

Material and methods: This series included 50 primary arthroplasties implanted from January 2006 to May 2006 in 50 patients (27 women and 23 men), mean age 60.6±11.4 years (21–75). The same femoral piece made of highly polished M30NW stainless steel with a quadrangular section was used for all implants. Similarly, all patients had a polyethylene cup sterilised with ethylene oxide (CMK21, Smith and Nephew). The femoral head was made of stainless steel for 25 hips and oxinium for 25 hips. The major outcome was penetration of femoral head into the cup (associated with true wear and creep) measured at minimum two years follow-up using the Martell method modified according to the recommendations of the author for an all-polyethylene cup. Patient-related and technique-related factors were studied. Non-parametric tests were used for the statistical analysis.

Results: There was no significant difference between the two groups regarding preoperative data. Two patients died, one was lost to follow-up; for three patients, the radiographs were excluded by the software. The analysis thus included 44 hips with a median follow-up of 2.01 years (1.9–2.3), 22 in the Oxinium™ group and 22 in the metal group. The median penetration rate was 0.16 mm/year in the Oxinium^® group versus 0.19 mm/year in the metal group (Mann-Whitney, p=0.46). Annual volumetric penetration in the Oxinium™ group was comparable with that of the metal group (Mann-Whitney, p=0.76). conversely, using the radiograph taken at one year as the reference value (true wear), wear was 0.066 mm/y in the Oxinium™ group versus 0.19mm/y in the metal group (Mann-Whitney, p=0.38).

Discussion and Conclusion: The results of this series indicate that using an oxinium femoral head reduces polyethylene wear. Mid-term results appear to be necessary to confirm these findings.

Introduction: The aim of this study was to evaluate the in vivo migration patterns using EBRA-FCA of a consecutive series of polished femoral components cemented line-toline.

Materials and Methods: Between January 1988 and December 1989, 164 primary total hip arthroplasties were performed in 155 patients. The mean age at the time of the index arthroplasty was 63.8 ± 11.6 years. A single design prosthesis was used combining an all-polyethylene socket and a 22.2 mm femoral head. The mono-block double tapered (5.2°) femoral component made of 316-L stainless steel had a highly polished surface (Ra = 0.04 μm) and a quadrangular section(Kerboull^® MKIII, Stryker). The femoral preparation included removal of diaphyseal cancellous bone to obtain primary rotational stability of the stem prior to the line-to-line cementation. For each patients, all available AP radiographs of the pelvis were digitized (Vidar Sierra Plus, Vidar System Corporation, Herdon, Virginia) and linked to an IMB-compatible computer. The EBRAFCA software is a validated method designed to assess migration of a femoral component through comparable pairs of radiographs.

Results: At the minimum 15-year follow-up, 73 patients (77 hips) were still alive and had not been revised at a mean of 17.3 ± 0.8 years (15–18 years), 8 patients (8 hips) had been revised for high polyethylene wear associated with periacetabular osteolysis, 66 patients (69 hips) were deceased, and 8 patients (10 hips) were lost to follow-up. Among the 8 revision procedures, the femoral component was loose in 3 hips. A total of 1689 radiographs (mean 10.3 per hip) were digitized. At the last follow-up, the mean subsidence of the entire series was 0.63 ± 0.49 mm (median of 0.61 mm; range 0 to 1.94 mm). When using a 1.5 mm threshold (accuracy of the EBRAFCA method) for subsidence, 4 of the 142 stems with adequate EBRA-FCA data were considered to have migrated. Using a threshold of 2 mm for subsidence, none of the 142 stems were considered to have migrated. The patterns of migration were calculated every 2 years giving 9 intervals. The evolution of mean subsidence during the whole follow-up period remained below 1.5 mm.

Discussion: Mean subsidence of this quadrangular highly polished femoral component remained below the accuracy of the method (± 1.5 mm) throughout the entire follow-up period. Of the 142 hips analyzed, only four (2.8%) had subsided of more than 1.5 mm and none more than 2 mm. This study demonstrates that contrary to other cemented femoral components that have provided excellent survival in the long term frequently associated with stem subsidence, a highly polished cemented double tapered femoral component with a quadrangular cross-section cemented line to line does not subside up to 18-year follow-up.

Introduction: Although a number of methods have been described to treat recurrent dislocation following total hip arthroplasty, this complication remains a challenging problem. The purpose of this retrospective study was to evaluate the minimum 2-year outcome in a consecutive series patients treated with a cemented tripolar unconstrained acetabular component for recurrent dislocation.

Patients and Methods: Fifty-one patients presenting with recurrent dislocation following primary or revision total hip arthroplasty in the absence of an identifiable curable cause were treated with a cemented tripolar unconstrained acetabular component. There were thirty-nine females and twelve males with a mean age at the time of the index procedure of 71.3 years. A single acetabular component design was used consisting of a stainless steel outer shell with grooves for cement fixation with a highly polished inner surface. This shell articulated with a mobile intermediate component with an opening diameter smaller than the 22.2-mm femoral head. No locking ring or other mean of constraint was associated.

Results: Of the fifty-one patients, forty-seven have had complete clinical and radiological evaluation data at a mean follow-up of 31.2 months (twenty-four to 56.3 months). The cemented unconstrained tripolar acetabular component restored complete stability of the hip in forty-nine patients (96%). The mean Merle d’Aubigné hip score was 15.8 ± 2.0 at the latest follow-up. Radiographic analysis revealed no or radiolucent lines less than 1 mm thick located in a single acetabular zone in forty-three of forty-seven hips (91.5%). The cumulative survival rate of the acetabular component at 36 months using revision for dislocation and/or mechanical failure as the end point was 93.3 ± 4.6% (95% confidence interval, 84.4% to 100%).

Conclusion: A cemented tripolar unconstrained acetabular component was highly effective in the treatment of recurrent dislocation with none of the complications associated with constrained devices. However, because longer term follow-up is needed to warrant that dislocation and loosening rates will not increase, the use of such a device should be limited to strict indications.

Introduction: The purpose of this study was to report on the minimal 10-year followup results of a prospective randomized and a historical series of low friction cemented hip arthroplasties according to the surface finish of the femoral implant.

Patients and Methods: The prospective randomized series included 284 patients (310 hips) with a mean age of 64.1 years. Among these 310 hips, the femoral component had a highly polished surface (Ra = 0.04 micron, MKIII, Stryker) in 165 hips, and a matte surface finish (Ra = 1.7 microns, CMK3, Vector Orthopedique) in the remaining 145 hips. The historical series that was operated by the sames surgeons according to the same surgical technique included 111 patients (123 hips) with a satin finish femoral component (Ra = 0.9 micron, CMK2, Sanortho). Clinical results were rated according to the Merle d’Aubigne hip score. Radiologic analysis was performed according to the criteria of Barrack et al. for the definition of loosening. Moreover, a survival analysis according to the actuarial method was conducted.

Results: At the minimum 10-year follow-up evaluation, 43 patients (48 hips) were lost to follow-up (0.3 to 8.7 years), 80 patients (83 hips) were deceased (0.1 to 13.6 years), 26 patients had revision of either or both components (0.9 to 15.9 years), and 246 patients (276 hips) were alive and had not been revised after a mean 12.3 ± 1.9 years (10 to 16 years). Radiologic loosening of the femoral component, including revision, occured in one, four, and 15 hips for the polished, stain, and matte groups, respectively. The survival rate at 13 years of the femoral component, using loosening as the end point, was 97.3 ± 2.6% (95% CI, 92.2 to 100%), 97.1 ± 2.1% (95% CI, 93 to 100%), and 78.9 ± 5.8% (95% CI, 67.6 to 90.3%) for polished, satin, and matte stems, respectively.

Discussion: This study demonstrated that cement fixation of a femoral component was more reliable in the long term with a polished or satin surface finish. Based upon our results and the review of the literature, we recommend abondoning the use of cemented stems with a surface roughness greater than 1 micron.

Purpose of the study: The purpose of this prospective study was to compare insert wear radiographically at minimum three years comparing crosslinked (Duration, Howmedia) versus regular (Stratec Medical) polyethylene.

Material and methods: We used two types of UHMWPE for a series of 140 total hip arthroplasties (THA) (February 1999 – August 2000). Duration polyethylene (Howmedica) was used for 63 implants. This UHMWPE undergoes gamma ray treatment under nitrogen which leads to a stabilization phase under temperature and time conditions enabling preferential formation of cross links between the molecular chains. Vecteur Orthopedic polyethylene implants exposed to gamma radiation under air were used for 62 implants. Eighty-seven patients were reviewed at minimum three years radiological follow-up. This series included 60 women and 27 men, mean age 64 years (range 32–93 years). The trans-trochanteric approach was used in all cases for insertion of a Charnley-Kerboull cemented femoral stem. All cups were cemented (abnormal implant position was an exclusion criterion). Cups were assigned randomly with a permutation table. Wear measurements were made by graphic construction comparing the immediate postoperative and last follow-up anteroposterior pelvis x-rays.

Results: Follow-up was at least 36 months for 87 patients. Mean follow-up for these 87 patients was 49.9 months. Mean wear was 0.32 mm for crosslinked poly-ethylene and 0.35 mm for regular polyethylene. Five patients presented wear ≥ 1 cm for crosslinked poly-ethylene and there were nine patients with ≥ 1 cm for regular polyethylene.

Discussion: In vitro, crosslinked polyethylene exhibits significantly improved mechanical qualities for wear resistance. Several factors can account for differences in polyethylene wear: the size and composition of the prosthetic head, and patient gender, age, activity level and BMI. These factors were comparable in our two groups. It does however appear that even with crosslinked poly-ethylene, after four years implantation, wear is limited, close to the detection level of the measurement method.

Conclusion: The measurement method used in this prospective randomized study of polyethylene wear comparing crosslinked versus regular polyethylene showed no difference between the two groups at four years follow-up.

Purpose of the study: The purpose of this propsective randomized study was to evaluate wear resistance of cemented polyethylene cups with the same design but regular or crosslinked polyethylene.

Material and methods: This series included 144 first-intention arthroplasties implanted between July 2000 and July 2002 in 137 patients (92 females, 45 males), mean age 66.2 years (range 16–88 years). The same femoral piece ws used in all patients with a 22.2 mm head. The cup was made of highly crosslinked polyethylene (Durasul™, Zimmer) for 83 hips and regular polyethylene (Duration™, Stryker) for 61 hips). The main outcome criterion was penetration of the femoral head into the cup at two years minimum follow-up (associating true wear and creep) measured with the Martell method modified by the cup manufacturer’s recommendations. Influence of patient-related factors and surgery-related factors was assessed. Non-parametric statistical tests were applied.

Results: The two groups of patients were not significantly different preoperatively. Two patients in the Durasul™ group died at 14 and 31 months follow-up. Median follow-up for the 137 survivors (142 hips) was 28.8 months (range 3–48.8 months) in the Durasul™ group and 24.8 months (ragne 0–47.9) in the Duration™ group. Among these 142 hips, 66 in the Durasul™ group and 51 in the Duration ™ group were reviewed clinically and radiographically with at least two years follow-up. The median penetration was 0.104 mm/yr in the Durasul™ group and 0.242 mm/yr in the Duration™ group (Mann-Whitney test, p=0.01, power 78%). There was a highly significant negative correlation between follow-up time and wear in the Duration™ group (Spearman r = −0.4, p=0.005) but no significant correlation in the Dursul™ group (Spearman r = −0.2, p=0.06).

Discussion and conclusion: The results of this series indicate a notable reduction in wear for highly cross-linked polyethylene cups, with no specific material-related complication. Mid-term results would however be necessary to validate these findings.

Introduction: The purpose of this retrospective study was to report on the minimal 5-year follow-up results of a consecutive series of cemented total hip arthroplasties following acetabular fracture.

Materials and methods: Between January 1980 and December 1995, fifty-three total hip arthroplasties were performed in 53 patients (16 females and 37 males). The mean age of the patients at the time of the index arthroplasty was 53.1 years (range, 24–84 years). The initial fracture concerned one wall in 18 patients, one column in 7, and both columns in 6. It was a complex fracture in 11 patients, and was unknown in the remaining 11 patients. Twenty-three of the 53 fractures had had a non-operative treatment, while 30 had had a surgical treatment. The mean time between the fracture and the arthroplasty was 16.4 10.8 years. All prostheses were of Charnley-Kerboull design, combining a 22.2-mm femoral head and an all-polyethylene socket. Both components were cemented.

Results: At the minimum 5-year follow-up evaluation, 35 patients were still alive and had not been revised at a mean of 12.4 3.8 years (range, 7–21 years), 6 patients had been revised, 5 patients had died from unrelated causes, and 7 patients were lost to follow-up. The mean Merle d’Aubigne hip score was 16.7 1.3 at the latest follow-up. Revision was performed for high polyethylene wear associated with periprosthetic osteolysis in 5 hips at a mean of 10.3 years, and for deep sepsis in one. The survival rate of the whole series at 15 years, using revision for any reason as the end-point, was 79.2 9.7 % (95% confidence interval, 60.3 to 98.2%). The survival rate at 15 years, using radiologic loosening as the end-point, was 94.7 % (95% confidence interval, 84.7 to 100%) for hips of which fracture had been treated non-operatively, versus 75.5 13.0% (95% confidence interval, 49.9 to 100%) for hips of which fracture had been surgically treated. The difference was not significant with the numbers available (log-rank test, p = 0.44).

Discussion and conclusion: The results of this series indicated that the mechanical failure rate of total hip arthroplasty following acetabular fracture was high in the mid- to long-term. The young age of the patients, the predominantly male cohort, and the modifications of the acetabulum structure due to the fracture could account for this phenomenon.

Purpose: The purpose of this retrospective study was to evaluate the clinical and radiological results of Charnley-Keroboull total hip arthroplasty performed in patients aged less than 50 years. We searched for factors which might affect wear and sustained fixation.

Material and methods: Among the 2,804 arthroplasties performed in patients aged less than 50 years between 1975 and 1995, we selected randomly 287 (10% of the annual operations). These prostheses were implanted in 222 patients (144 women and 78 men), mean age 40.1±8 years (15–50). All of the arthroplasties were inserted via a transtrochanteric approach. Charnley-Kerboull implants were cemented in all patients using a metal polyethylene bearing. Functional outcome was assessed with the Postel-Merle-d’Aubigné score. Cup wear was measured with the Chevrot technique. The actuarial method was used to calculate prosthesis survival.

Results: At last follow-up, 155 patients (210 hips) were living and had not had a revision procedure at mean 16.1±4.6 years, 23 patients (25 hips) required revision of the acetabular or femoral element, ten patients (10 hips) had died, and 34 patients (42 hips) were lost to follow-up. The mean preoperative functional score was 9.6±2.5 (9–15) versus 17.2±0.8 (9–18) at last follow-up (Wilcoxon rank test p< 0.001). For the acetabular element, there was certain loosening in 15 hips and possible loosening in 24. For the femoral element, loosening was certain for 12 implants and possible for four. Twentyfive hips required revision, including 17 for aseptic loosening. Mean wear was 0.12±0.21 mm (0–2.23). Among the 287 hips, 196 had wear measured at less than 0.1 mm/yr (mean 0.02 mm/yr). Mean overall implant survival, defining revision as failure, was 85.4±5.0% at twenty years (95%CI 78.4–92.4). Among the factors tested, only abnormally rapid wear (> 0.1 mm/yr) was predictive of failure.

Discussion: The results of this series allow us to conclude that total hip arthroplasty using a Charnley-Kerboull implant remains the best solution for young patients in terms of implant survival.

Purpose: Certain authors have indicated that the primum movens of loosening of cemented femoral stems is related to the decohesion of the cement implant interface. Under such conditions, improvement in this interface was attempted with the development of a rough surface for the femoral piece. The purpose of this retrospective analysis was to evaluate results a minimum ten years follow-up in a consecutive series of total hip arthroplasties comparing types of femoral implant surfaces.

Material and methods: Between January 1988 and December 1989, 311 THA were implanted in 286 patients aged 63.6±11.8 years (26–91). All implantations were performed via the transtrochanteric approach by two senior surgeons. Two types of femoral implant were used: 166 non-polished pieces with a round section Ra = 3 mm (CMK3, Vector Orthopédique) and 145 polished pieces with a quadrangular section Ra = 0.4 mm (MKIII, Stryker Howmedica). The preoperative data were comparable for the two groups. Functional results were assessed with the Postel Merle d’Aubigné score (PMA). Actuarial survival curves were plotted.

Results: At minimum ten years follow-up, 187 patients (204 hips) were living and had not required revision at mean follow-up of 11.7±2.5 years (10–14), 15 patients (15 hips) had revision of the acetabular and/or femoral element, 54 patients (58 hips) had died, and 30 patients (34 hips) were lost to follow-up. The mean preoperative function score was 11.2±2.5 (4–16) versus 17.5±0.5 (10–18) at last follow-up (Wilcoxon rank test, p< 0.001). Cumulative survival at 13 years, taking radiographic loosening of the femoral piece as the endpoint, was 78.9±5.8% (95%CI 67.6–90.3%) for unpolished implants versus 97.3±2.6 (95%CI 92.2–100) for polished implants. The difference was significantly different (p< 0.001).

Discussion: The results of this analysis indicate that radiographic survival of unpolished cemented femoral pieces is significant inferior compared with polished pieces. The increased adherence of the femoral cement is probably the cause of increased shear stress at the bone-cement interface. The respective influence of section and surface remain to be determined.

Purpose: The purpose of this retrospective study was to evaluate at minimal 5 years follow-up outcome in a consecutive series of total hip arthroplasties implanted for fracture of the acetabulum.

Material and methods: The series included 53 arthroplasties implanted between January 1980 and December 1995 in 53 patients, 16 women and 37 men, mean age 53.1 years (24–84). The initial fracture involved the acetabular wall in 18 patients, one column in seven and two columns in six. It was a complex fracture in eleven cases and classification was unknown in the eleven other hips. Orthopaedic treatment was used for 23 patients and surgery for 30. Mean time between fracture and arthroplasty was 16.4±10.8 years. Cemented Charnley-Kerboull implants with a metal-polyethylene bearing were used. The Postel-Merle-d’Aubigné (PMA) score was used to assess functional outcome. The actuarial survival was determined.

Results: At five years minimum follow-up, 33 patients were alive and had not undergone revision at mean follow-up of 12.4±3.8 years (7–21). Six patients had had revision of the acetabular and/or femoral element, five patients had died, and seven were lost to follow-up. Revisions were required for cup wear associated with periacetabular osteolysis. The mean preoperative functional score was 10.6±2.5 versus 16.2±2.8 (8–18) at last follow-up (Wilcoxon rank test, p< 0.0001). Cumulative survival, taking revision as failure, was 90.3±6.5% at 15 years (95%CI 77.6–100%) for hips treated orthopaedically versus 66.5±14.5 (95%CI 38.1–94.9%) for hips treated surgically. The difference for the analyzable hips was not significant (logrank, p=0.69).

Discussion: The results of this series confirm that the long-term risk of mechanical failure of total hip arthroplasty for fracture of the acetabulum is high.

Purpose: Periprosthetic ossification is a frequent complication of total hip arthroplasty and can have a major functional impact. Non-steroidal anti-inflammatory drugs (NSAID) can provide effective prevention but with a risk of morbidity. The purpose of this work was to evaluate the efficacy of an anti-Cox2 agent, cele-coxib, for this indication.

Material and methods: Total hip arthroplasty was performed in 42 patients with a relative (gastrointestinal) contraindication for the use of NSAID. These patients were given celecoxib (Celebrxy(r)) 200 mg bid starting the day before the operation and continuing for at least five days. A control group of 42 age- (±3 yr) and sex-matched patients who underwent surgery for the same indication performed by a surgeon with equivalent experience was also established. The control patients were given ketoprofen (Profénidy(r)) 50 mg qid for two days then 150 mg bid for five days. The approach, implant, and other adjuvant treatments were equivalent between the two groups. Ossifications were analysed on the follow-up films taken at least three months after surgery. The Brooker classification was used. The exact Fisher test was used for the statistical analysis.

Results: The two groups each included 31 women and eleven men, mean age being the same in the two groups (67.12 yrs). Mean follow-up was very similar (8.44 vs 8.6 months). Aetiologies were: primary degenerative hip (n=30), degenerative hip disease after dysplasia (n=9), sequela of infantile arthritis (n=1), revision total hip prosthesis (n=2). Two patients in each group interrupted their treatment between day 2 and 4 because of intolerance. There were no cases of significant haematoma in either group. No ossification > grade 2 was observed. The overall rate of ossification was 42.5% in the control group versus 48.6% in the celecoxib group. The rate of grade 2 ossifications was 8% in the cele-coxib group versus 12% in the control group. These rates were not significantly different (Fisher’s exact test= 0.6).

Discussion: In this study, celecoxib and ketoprofen were found to have equivalent efficacy for the prevention periprosthetic ossification. This is an interesting perspective in the probable hypothesis of less morbidity with anti-Cox 2 antiinflammatory drugs used in combination with an antalgesia protocol.

Aims: The purpose of this retrospective study was to report on the treatment of migrated ununited greater trochanter following total hip arthroplasty. Methods: Between January 1986 and December 1999, 72 non-unions of the greater trochanter in 71 patients were treated using a trochanteric claw plate. The mean age of the patients was 66 ± 11 years. The average time to re-operation was 8 months. Fixation of the non-united greater trochanter was performed using a claw plate only in 47 hips, and the association of frontal wires with a claw plate in the remaining 25 hips. The main criterion for evaluation was the consolidation of the greater trochanter judged as follows: bony consolidation (no pain, no Trendelenboug gait, radiologic fusion); fibrous consolidation (moderate pain, no Trendelenboug gait, radiologic fusion difficult to assess); and non-union (Trendelenboug gait and/or absence of radiologic fusion). Results: The average follow-up of the series was 4 years (1 to 14 years). The mean d’Aubigné score significantly increased from 13.5 preoperatively to 15.9 at last follow-up (paired signs test, p < 0.0001). Bony consolidation was obtained in 51 hips, fibrous in 9, whereas repeat non-union occurred in 12. The only predictive factor for union was the use frontal wires in association with a claw plate that provided 87.5% of unions and no failure (Chi square test, p = 0.006). Conclusions: This study indicated that non-union of the greater trochanter following total hip arthroplasty can be successfully treated with frontal wires in conjunction with a trochanteric claw plate.

Aims: The purpose of this prospective randomized study was to report on the minimal 10-year follow-up results of a consecutive series of cemented total hip arthroplasty according to the surface finish of the femoral implant. Methods: Between January 1988 and December 1989, 311 total hip arthroplasties were performed in 286 patients. Implants were of Charnley Kerboull design combining a 22.2-mm femoral head and an all-polyethylene socket. The cemented femoral implant was available in two configurations: polished (Ra = 0.4 μm) and matte (Ra = 3 μm). The polished stem (MK III, Stryker Howmedica) was implanted in 166 hips, and the matte stem (CMK 3, Vecteur Orthopedic) in the remaining 145. The mean age of the patients was 63.6 years. Results: At the minimum 10-year follow-up evaluation, 187 patients (204 hips) were still alive and had not been revised at a mean of 11.7 years (10–14 years), 15 patients (15 hips) had been revised, 54 patients (58 hips) had died from unrelated causes, and 30 patients (34 hips) were lost to follow-up. The mean d’Aubigné hip score was 17.7 ± 0.3 at the latest follow-up. The survival rate at 13 years, using radiologic loosening as the end-point, was 97.3 ± 2.6% (95% confidence interval, 92.2 to 100%) for polished stems, versus 78.9 ± 5.8% (95% confidence interval, 67.6 to 90.3%) for matte stems (log-rank test, p = 0.0001). Conclusions: This study demonstrated that cemented fixation of femoral stems was significantly more reliable in the long-term with a polished surface.

Aims: The aim of this open prospective study was to evaluate the minimum 2-year follow-up outcome of a consecutive series of low friction total hip arthroplasties combining zirconia on polyethylene. Methods: Between January 1997 and June 1999 fifty-five total hip arthroplasties were performed in 51 patients. The mean age was 52.2 ± 12 years. The 22.2-mm femoral head made of zirconia ceramic was secured to the femoral component through a Morse taper that had an angle of 11°25 for 27 hips and 5°40 for 33 hips. All prostheses were of Charnley-Kerboull design. Clinical results were evaluated according to the Merle d’Aubigné hip score. Wear of the acetabular component and periprosthetic osteolysis was measured on serial radiographs of the pelvis. Results: The mean follow-up of the series was 32 months (24 to 48 months). No patient was lost to follow-up. The mean functional hip score significantly increased from 12.2 ± 2.6 preoperatively to 17.8 ± 0.2 at the latest follow-up (paired Student’s t test, p < 0.0001). None of the acetabular or femoral component had migrated. Wear of the socket was always undetectable on plain radiographs. However, lytic endosteal lesions of the calcar were observed in 19 of the 55 arthroplasties (34.5%). These lytic lesions appeared between the first and second postoperative year. Conclusions: Early calcar osteolysis observed in this study can be related to either run-in wear or to preliminary severe wear. The authors do not recommend further use of zirconia ceramic until long-term follow-up studies are available.

Aims: The purpose of this retrospective study was to report on the minimum 10-year follow-up results of a consecutive series of cemented low friction total hip arthroplasties performed for avascular necrosis of the femoral head. Methods: One hundred and twenty-two THAs in 96 patients were performed between January 1980 and December 1990. All prostheses were of Charnley- Kerboull design, combining a 22.2-mm femoral head and an all-polyethylene socket. Both components were cemented. The mean age of the series was 50.8 ± 13.3 years (21–85 years). Eighty hips were graded Ficat III and 42 hips were graded Ficat IV. Results: At the minimum 10-year follow-up evaluation, 59 patients (75 hips) were still alive and had not been revised at a mean of 13.9 years (10–21 years), 7 patients (7 hips) had been revised, 20 patients (24 hips) had died from unrelated causes, and 10 patients (16 hips) were lost to follow-up. The mean dñAubignŽ hip score was 17 ± 1 at the latest follow-up. The mean wear rate for unrevised hips was 0.07± 0.06 mm per year. Revision was performed for polyethylene wear associated with periprosthetic osteolysis in 6 hips and for deep sepsis in one. Three hips had recurrent dislocations. The survival rate at 15 years, using revision for any reason as the end-point, was 88.5% (95% conþdence interval, 80.2 to 96.9%). Conclusions: This series indicated that Charnley Kerboull low friction total hip arthroplasty for avascular necrosis could provide satisfactory long-term clinical and radiologic results.

Purpose: Nonunion of the trochanter after total hip arthroplasty using the transtrochanteric approach is a serious complication. Incidence is estimated at 3%. Failure after simple cerclage reaches 40%. A plate was therefore designed specifically for fixation of trochanteric fractures. The purpose of this retrospective analysis was to assess outcome in a continuous series of trochanteric fractures treated with this plate.

Material and methods: The series included 72 nonunions in 71 patients treated between 1986 and 1999. Mean age was 66 years. Most of the arthroplasties had been performed for primary degenerative joint disease or for hip dysplasia. The time interval between artrhoplasty and treatment of the nonunion was eight months on the average. The trochanter was fixed with a plate alone in 47 hips and in combination with a frontal wire in 25. The primary outcome criterion was trochanter healing scored as: union (pain free hip, radiological fusion and stability), nonunion (lack of radiological fusion and/or presence of an instability), and doubtful union (moderate pain, no instability, radiological fusion difficult to affirm).

Results: Mean follow-up was 47 months (range 12 – 14). Mean functional score at last follow-up was 15/9 compared with 13.5 preoperatively (paired test, p < 10–4) with 51 unions, 12 nonunions and 9 doubtful unions. Among the factors studied, trochanter fixation technique was the only factor predictive of outcome. Results were as follows for plate fixation alone: union 62%, doubtful union 13% and failure 25%. For hips with wire and plate fixation, the results were union 87.5%, doubtful union 12/5%, failure 0%. The difference was highly significant (chi-square, p = 0.006) in favour of wire-plate fixation of trochanteric fractures.

Discussion and conclusion: This study leads to the conclusion that the treatment of choice for trochanteric nonunion is wire-plate fixation which provides more than 90% good and very good results.

Purpose: The purpose of this retrospective study was to assess clinical and radiological outcome at ten years follow-up at least in a continuous series of total hip arthroplasties performed in patients with aseptic necrosis of the femoral head.

Material and methods: The series included 122 arthroplasties performed between January 1980 and December 1996 in 96 patients (26 women and 70 men), aged 50.8±13.3 years (range 21–85 years). The underlying cause was essential necrosis in 40.6%, cortocosteroid threapy in 19.8%, chronic alcoholism for 17%, trauma for 12%, and another medical cause for 10%. The Ficat and Arlet classification was grade IV 80 hips, grade III 42 hips. All arthroplasties were performed via the transtrochanteric approach. Cemented Charnley-Kerboull implants were used with a metal-polyethylene couple in all cases. The Merle d’Aubigné score was used to assess functional outcome. Cup wear was measured according to Livermore. Actuarial survival curves were calculated.

Results: Follow-up was greater than ten years for all patients (mean 13±2.6 years, range 10–21 years). At last follow-up 59 patients (75 hips) were living and had not undergone revision and seven patients (seven hips) had undergone revision of the acetabular and/or femoral component. Twenty patients (24 hips) had died and ten patients (16 hips) were lost to follow-up. Revision procedures were performed for periacetabular osteolysis in one hip, infection in one hip, and acetabular loosening in five hips. Three patients experienced luxation or subluxation. The mean preoperative function score was 11 (range 5 – 16) compared with 17 (14–18) at last follow-up (p < 0.05). Clinical outcome was good or excellent in 95% of the patients and fair or poor in 5%. Mean overall cup wear was 0.965 mm (0–5) for non revised hips. Lucent lines were found around eleven acetabular implants and six femoral implants. Cumulative survival, considering revision as failure, was 88.5% at 15 years (95% confidence interval 80.2 – 96.9%).

Discussion and conclusion: The results of this series lead to the conclusion that low-friction total hip arthroplasty is the treatment of choice for advanced stage femoral head necrosis.