A longitudinal study was done assessing the latest radiographs available in a series of collarless Corail uncemented stems which have reported survival rates of 95% at 20 years. Parameters scored included the degree of “Ghosting” or delamination, the Gruen, the stem fit in the femoral canal and the degree of calcar resorption. Patient and implant demographics were noted. At 3 years the loosening rate was 23% reaching 64% at 6 years after the index procedure. It was present in all age groups, with a peak in the 24 to 49 year age group. Males and females had the same occurrence. When it was present this was always in zone 1. It was present in 6% of patients in zone 7, but then always associated with zone 1 changes. High and standard off-set stems had the same loosening rates. The looser fit of the stem the higher the incidence of loosening. There was no correlation to the type of bearing surface or the degree of calcar resorption. Those patients with a BMI of 25 – 35 had lower loosening rates compared to those with higher or lower BMI's. We postulate that cancellous bone on-growth onto the hydroxyapatite coating associated with loading flexural micro-motion leads to hydroxyapatite being pulled off the smooth stem substrate in zone 1. Progressive delamination of the hydroxyapatite then occurs. The triple-tapered design though imparts continued stability. We report high loosening rates in the Corail stem and suggest a mechanism for its development.
The study was started in 2004 to determine the best bearing surface in the long term, and to measure the metal ion levels generated by each of the bearing surfaces. We present the latest updated results. A prospective randomised study was started in 2004 to compare the wear characteristics of Ceramic on X linked Polyethylene (C.O.P.), Ceramic on Ceramic (C.O.C), Ceramic on Metal (C.O.M.) and Metal on Metal (M.O.M) bearings. The level of Cobalt and Chrome ions in red blood cells have been documented at serial intervals, using the ICP – MS method. Aside from the bearing surfaces the rest of the implant is standard, using a Pinnacle Cup, Corail Stem and 28mm heads. 256 Cases were enrolled on the study. To date 71 cases have been lost due to death (26), revision (9) and lost to follow up (36), leaving us with 185 for follow-up. An even spread of cases in each bearing surface are still available for follow up, viz. 46 C.O.P, 48 C.O.C., 44 C.O.M. and 47 M.O.M. Average follow up is currently 4.8 years, ranging up to 9 years.Purpose Of Study
Material and methods
Metal ion levels are used to track the performance of metal containing bearings in hip replacement patients. Changes in whole blood metal ion levels are indicators of wear rates in these bearings. Normal metal ion levels are variable, and range widely. Changes in these “non-bearing” levels over a period of time may influence the monitoring of these bearings. As part of a prospective randomised trial of different bearing surfaces, whole blood metal ion levels were monitored. This included four cohorts of patients, namely ceramic-on-ceramic (CoC), ceramic-on-polyethylene (CoP), ceramic-on-ceramic (CoC) and ceramic-on-metal (CoM). Serial whole blood metal ion levels in the non-metal bearings, namely CoC and CoP, were analyzed. There was no consistency in these whole blood metal ion levels. Intra-patient variations in these levels over a period of time could not be due to bearing-produced metal ions as these were all metal free, and thus not the source of any endogenous ions. These intra-patient variations may reflect changes in exogenous exposure to these metal ions, fluctuations in these patient's metabolic functions or production of metal ions from non-bearing prosthetic sources.Methods.
Results.
To look at the difference in metal ion levels using Ceramic on Metal (COM) and Metal on Metal (MOM) bearings in Total Hip Replacements, comparing the results between well placed and poorly placed cups. Metal ion levels using the ICP-MS method of assay have been studied as part of a prospective randomised trial between 4 different bearing combinations. The ion levels have been raised in the COM and MOM bearings. Metal ions are not raised in the Ceramic on Ceramic (COC) and Ceramic on X linked Poly(COP) bearings, showing that no other material in the study prosthesis aside from the bearing surfaces contribute to raised ion levels. Cup inclination and cup anteversion has been measured in all patients.Purpose of Study:
Methods:
Monitoring the performance of hip replacements post-operatively is tedious and costly, necessitating radiological examinations as well as other specialized examinations such as whole blood metal ion levels. In an effort to control escalating costs, we conducted an ethically approved clinical trial to assess the efficacy of basic acoustic monitoring equipment to asses these implants. An electronic stethoscope was successfully used to record sounds from the hips of participants with different bearing surfaces. The sounds were recorded while conducting a standardized movement sequence. A 5th order Savitzky-Golay filter with a window width of 21 points was used to remove background noise. The recordings were also listened to by ear and three primary classes of sounds were identified. Frequency components contained in the classes were identified using spectrograms and Welch power density spectra. The sounds were correlated with different patient factors including component positioning, BMI and length of time that the implant was in situ. The skewness and kurtosis of the power spectra were calculated and found to be different for each class. Further frequency analysis was conducted with the aid of the discrete wavelet transform. This met with some success as different frequency levels were found in each sound class. All bearing surfaces produced some noise. The most sounds were produced by the ceramic-on-metal group, even though not in the audible range, and those participants with a body mass index in the obese range. Sounds were also detected in the ceramic-on-polyethylene implants. However, no consistent links between these factors and the sounds produced could be identified. Specifically, the lack of correlation between sound occurrence and length of implantation indicates that this technique is not useful in predicting possible failures or future complications in real time.Method.
Results.
Multiple measurements have been described for the assessment of developmental dysplasia of the hip (DDH). In particular, the centre edge angle (CEA) has been described by Wiberg to assess the position of the femoral head in relation to the acetabular edge in patients over the age of five years. The purpose of this study is twofold. Firstly to assess the reliability of all measurements available in the literature and secondly to evaluate whether or not the CEA can be reliably measured below five years of age. Eighty seven patients were included for assessment. Radiographs were measured within six months of spica cast/Batchelor cast removal, depending on whether closed or open reduction was performed. A web based computer programme was used to store the radiographs electronically and with the help of an electronic template the following measurements were recorded: CEA, AI, centre head distance discrepancy ratio (CHDDR), Smith's c/b and h/b ratios. Three readers recorded measurements at two intervals, to determine intra and inter reader reliability.Purpose of Study
Methods
The search for the ideal bearing surface in Total Hip Replacements continues. The current ‘best’ materials are felt to be combinations of metal, ceramics and cross-linked polyethylene. Laboratory studies suggest that ceramic-on-metal articulations may provide distinct advantages. This study aims to identify the best bearing surface combination with the lowest adverse side effect profile. Between February 2004 and September 2007, 164 hips were replaced in 142 patients. 39% were male and 69% were female. The average age at surgery was 53 years (17-72 years). Follow-up assessment included radiographs, the Harris Hip Score and whole blood samples for metal ion levels. Complications to date included 3 hips which needed femoral revision because of surgery related factors, and 3 cases of sepsis of which 1 settled and 2 needed revision. One hip needed revision of head and liner to a larger bearing size for recurrent dislocations, and is no longer being followed up for blood metal ions. Post-operative whole blood metal ion levels were compared to pre-operative levels to determine the increase or decrease in metal ion levels. There were no changes in those patients with ceramic-on-ceramic and ceramic-on-polyethylene articulations. Moderately raised whole blood metal ion levels were noted at 3 months in the ceramic-on-metal group, while the metal-on-metal group show the greatest increase. This study agrees with laboratory bearing surface wear studies demonstrating lower wear rates in the ceramic-on-metal group compared to the metal-on-metal group. With concerns related to high blood metal ion levels in metal-on-metal articulations, ceramic-on-metal bearing surfaces may well become a bearing surface of choice in the future, but progress needs to be monitored in the longer term.
Ceramic-on-metal (COM) bearings have shown reduced wear and friction compared with metal-on-metal (MOM) bearings in-vitro. Lower wear has been attributed to a reduction in corrosive wear, smoother surfaces, improved lubrication and differential hardness reducing adhesive wear. Clinical studies have also shown reduced metal ion levels in-vivo compared with MOM bearings. The aim of this study was to examine two explanted COM bearings (one head and cup, one head only), and to assess the effect of in-vivo changes on the wear performance of the COM bearings by comparing the wear of the explanted bearings with three new COM implants in a hip wear simulator. Two 28mm diameter COM bearings were provided for analysis. These were visually examined and surface profilometry was performed using a 2-D contacting profilometer (Form Talysurf, Taylor Hobson, UK). Scanning electron microscopy was used to image the regions of transfer on the ceramic heads, and EDX to assess the transfer composition (Philips XL30 ESEM). Hip simulator testing was conducted for 2 million cycles (Mc) comparing the explanted bearings with three new 28mm COM bearings. Tests were performed in a Prosim simulator (SimSol, UK), which applied a twin peak loading cycle, with a peak load of 3kN. Flexion-extension of − 15 to 30 degrees was applied to the head and internal-external rotation of +/− 10 degrees was applied to the cup, components were mounted in the anatomical position. The lubricant was 25% (v/v) calf serum supplemented with 0.03% (w/v) sodium azide and was changed approximately every 0.33Mc. Wear was measured gravimetrically at 0.5, 1 and 2 Mc. Regions of material transfer, identified on both ceramic explant heads, were shown to be CoCr material by EDX analysis, suggesting metallic transfer from the metal cup. Profilometry traces across metallic transfer showed comparable surface roughness measurements compared to unworn material. The overall mean wear rate for the new COM bearings at 2Mc was 0.047 ± 0.06mm3/Mc. The mean wear rate for the explanted head articulated with a new cup was slightly lower at 0.034mm3/Mc. The mean wear rate for the explanted head and cup was highest at 0.15mm3/Mc. It was noted that the explanted head/cup had higher bedding in wear compared with the other bearings, but still significantly less than a new MOM bearing (mean bedding-in wear rate 2.03 ± 2.59 mm3/Mc). The steady-state wear was comparable with the new bearings. As the orientation of these implants in-vivo was unknown, it is proposed that the elevated wear during bedding-in of the explanted head/cup bearing may be due to the alignment of the components. The wear rates of the explanted ceramic head against a new cup were comparable with the new bearings, suggesting that the presence of metallic transfer on the ceramic head does not adversely affect the wear behaviour of COM bearings.
The combination of a ceramic head articulating against a metal acetabular liner (CoM) has shown reduced metal ion levels compared with a metal-on-metal bearing (MoM) in hip simulator studies. A randomized prospective clinical trial was undertaken using CoM and MoM bearings in an otherwise identical total hip procedure. The initial clinical results were encouraging. This report comprises a further review of metal ion data. Patients received identical components with the exception of the bearing surface material but all were 28mm diameter. All components were supplied by DePuy International Ltd. Patients were assessed pre-operatively, 3m, 12m and >
24m (median 32m). Whole blood samples were collected at regular follow-ups, frozen and analysed in batches using high resolution Inductively Coupled Plasma – Mass Spectrometry (ICP-MS). All recruited patients are included irrespective of outcome. However some patients failed to attend specific follow-ups and some contaminated samples had to be discarded. Statistical significance was analyzed using a non-parametric comparison (Mann-Whitney test). After 3m and 12m implantation there were between 21 and 24 patients available for analysis in both the CoM and MoM cohort and after >
24m point 10 and 9 respectively. There were four outliers (either Cr or Co >
10ug/l) in both the CoM and MoM groups. In common with previous studies (with the exception of two marginal outliers), these were related to component position. They were implanted with either a cup abduction angle of >
55°, an anteversion angle of >
30° or both. Other studies with the same design of component have reported no significant outliers. The median Cr and the Co levels are lower with the CoM bearing compared with the MoM at all measurements points following implantation. The median background (pre-operative) levels for the combined CoM and MoM group were Cr: 0.22ug/l and Co: 0.49ug/l. These were significantly different (p=0.006). In the CoM group, the median 12m Cr and Co values were 0.43ug/l and 0.72ug/l respectively. The comparable values for MoM are 0.68ug/l and 0.83ug/l. Increases in metal ion levels from pre-operative levels are used as the primary ion level outcome in this study because the background level will comprise of the order of 30–50% of the overall value. The increase in Cr for CoM and MoM from pre-op levels to 12m significantly different for Cr (p=0.015). It has a lower significance for combined metal ion levels (p=0.029). This difference in not significant for Co (p=0.195). In agreement with predictions from hip simulator studies, CoM bearings in this study produced lower levels of metal ions than comparable MoM bearings at all time points. However the difference is less than that predicted in the laboratory and is much more pronounced with Cr than with Co.
Recurrent patellar dislocation is a relatively common disorder in young patients. Historically, treatment options have been based on the underlying disorder predisposing the patient to the dislocation. This has resulted in various soft tissue reefing procedures, patella tendon realignment procedures and boney realignment procedures. Further research has shown that the medial patello-femoral ligament (MPFL) is the primary restraint to lateral patella subluxation and dislocation. Many authors have published their successful treatment of recurrent patella dislocation by reconstruction of the medial patellofemoral ligament. The most widely used is autologous semitendinosis tendon grafts, as well as synthetic materials, and MPFL reconstructions may be combined with boney procedures. Varieties of fixation techniques have been described involving both the patella and femoral sides. We present a technique of MPFL reconstruction using the autologous ipsilateral quadriceps tendon. Our technique avoids the morbidity associated with semitendinosis graft harvesting and the drill holes in, and potential resulting fracture of, the patella. The technique is also simple and is associated with decreased procedure costs. We present the technique and a series of six patients (seven knees) with follow up ranging from eight months to nine years. The average age of patients at the time of surgery 16 to 28 years (mean = 20 years). There have been no redislocations. The median Kujala patellofemoral knee score at follow up was 97 out of 100 (Range 69–100). The results compare very favorably to published results using other techniques. Our technique of reconstructing the MPFL is reliable, produces good results using an objective knee score, and is cost effective. Seventy staff members participated from a potential pool of approximately one hundred staff on duty at the time. Of the seventy staff who participated in this research project a total of three staff members were within 50 mls of the correct amount for each of the three samples. Overall staff were very poor at estimating blood loss. Staff working in the operating theatre, no matter what their affiliation or years of experience, are not accurate when estimating blood loss spilt into a patients bed. A tool that aids in blood loss estimation is a valuable addition to the theatre resource manual.
Dislocations remain a significant problem, especially after revision hip surgery. Revision of components, particularly in elderly patients with co-morbidities, can be fraught with complications. The surgeon’s options are sometimes restricted, particularly when the acetabular and femoral components are well fixed. Increased head lengths are often utilised to increase tissue tension, and thus improve stability. As a niche solution we have designed a low cost modular femoral neck extender. They are manufactured from medical grade Cobalt-Chrome, conforming to ISO 200, CE mark and EN46001 standards. Available in three incremental lengths and with different connecting Morse tapers, increases in effective neck lengths of up to 49 mms can be achieved. When both the original acetabular and femoral components are well orientated, the resultant increased tissue tension imparts stability to the hip. We present a series of five patients where we have used a femoral neck extender to achieve stability of a total hip replacement. Four patients had had multiple previous dislocations. One patient was unstable at the time of revision surgery because of a high hip centre. The average age of the patients was 72 years, and the number of previous dislocations averaged four. The average follow-up after surgery was 22 months. No patients have redislocated their hips. We present our novel femoral neck extenders as an elegant and cost effective solution to convert an unstable hip to a stable hip, especially when the patient has well fixed and orientated components not in themselves requiring revision.
A randomised prospective study of four bearing surfaces in hip replacements is being conducted. The primary objective is to identify the best long term bearing surf ace clinically and radiographically, and metal ion levels have been measured in all cases. Patients have been randomised to the four bearing surfaces viz. Ceramic-on-XLinked Polyethelene, Ceramic-on-Ceramic, Metal-on-Metal and Ceramic-on-Metal. Pre-operative blood samples and follow-up blood samples for metal ion analysis using ICP-MS method have been taken in all patients. As at February 2008 187 patients have been recruited, and metal ion levels at one year are available in 52 patients. Metal ion levels are not increased with Ceramic-on-XLPE or Ceramic-on-Ceramic bearings. At one year follow-up the metal ion levels in Ceramic-on–Metal bearings is half that of Metal-on-Metal bearings using mean levels, and one third using median levels. Of note is that chromium levels in Ceramic-on-Metal bearings is the least elevated. Due to the laboratory evidence that ceramic-on-metal bearings have the best surf ace wear characteristics with no head stripe wear on a ceramic head, and the laboratory and clinic al evidence of lower metal ion levels, Ceramic-on-Metal hip replacements could be one of the bearing surfaces of the future.
Recurrent patellar dislocation is a relatively common disorder in young patients. Historically, treatment options have been based on the underlying disorder predisposing the patient to the dislocation. This has resulted in various soft tissue reefing procedures, patella tendon realignment procedures and boney realignment procedures. Further research has shown that the medial patellofemoral ligament (MPFL) is the primary restraint to lateral patella subluxation and dislocation. Many authors have published their successful treatment of recurrent patella dislocation by reconstruction of the medial patellofemoral ligament. The most widely used is autologous semitendinosis tendon grafts, as well as synthetic materials, and MPFL reconstructions may be combined with boney procedures. Varieties of fixation techniques have been described involving both the patella and femoral sides. We present a technique of MPFL reconstruction using the autologous ipsilateral quadriceps tendon. Our technique avoids the morbidity associated with semitendinosis graft harvesting and the drill holes in, and potential resulting fracture of, the patella. The technique is also simple and is associated with decreased procedure costs. We present the technique and a series of 6 patients (7 knees) with follow up ranging from 8 months to 9 years. The average age of patients at the time of surgery 16–28 years (mean = 20years). There have been no redis-locations. The median Kujala patellofemoral knee score at follow up was 97 out of 100 (Range 69–100). The results compare very favourably to published results using other techniques. Our technique of reconstructing the MPFL is reliable, produces good results using an objective knee score, and is cost effective.
Dislocations remain a significant problem, especially after revision hip surgery. Revision of components, particularly in elderly patients with co-morbidities, can be fraught with complications. The surgeon’s options are sometimes restricted, particularly when the acetabular and femoral components are well fixed. Increased head lengths are often utilized to increase tissue tension, and thus improve stability. As a niche solution we have designed a low cost modular femoral neck extender. They are manufactured from medical grade Cobalt-Chrome, conforming to ISO 200, CE mark and EN46001 standards. Available in 3 incremental lengths and with different connecting Morse tapers, increases in effective neck lengths of up to 49 mm can be achieved. When both the original acetabular and femoral components are well orientated, the resultant increased tissue tension imparts stability to the hip. We present a series of 5 patients where we have used a femoral neck extender to achieve stability of a total hip replacement. 4 patients had had multiple previous dislocations. 1 patient was unstable at the time of revision surgery because of a high hip centre. The average age of the patients was 72 years, and the number of previous dislocations averaged 4. The average follow-up after surgery was 22 months. No patients have redislocated their hips. We present our novel femoral neck extenders as an elegant and cost effective solution to convert an unstable hip to a stable hip, especially when the patient has well fixed and orientated components not in themselves requiring revision.
Whole blood metal ion levels remain a concern in those patients undergoing total hip replacement with metal bearing surfaces. The determination of baseline reference levels are essential if useful information can be gleaned from in vivo studies of functioning implants. We set out to prospectively determine chromium and cobalt metal ion concentrations in patients undergoing total hip replacement to determine reference levels of these metal ions. 100 patients with normal renal function, no occupational or environmental exposure to cobalt and chromium, and an absence of implanted metals were recruited into the study. Metal ion levels were determined using two different assay methods. Both ICP-MS (Inductively Coupled Plasma Mass Spectroscopy) and GFAAS (Graphite Furnace Atomic Adsorption Spectroscopy) are well recognized analytical techniques for the quantification of trace elements. Levels were correlated with gender, age and place of residence. There was considerable variability in whole blood metal ion levels, with the ICPMS being more sensitive and consistent than the GFAAS method. Direct comparison of concentration levels determined by the two methods revealed no significant correlation. There was no correlation with age, gender and place of residence Our findings would favour the use of the ICP-MS to determine reference levels and as a baseline for metal ion surveillance pre-operatively in patients undergoing metal-on-metal total hip replacements. We also determined that changes in whole blood metal ion levels are more significant than actual levels in patients who have undergone total hip replacement.
A randomised prospective study of 4 bearing surfaces in hip replacements is being conducted. The primary objective is to identify the best long term bearing surface clinically and radiographically, and metal ion levels have been measured in all cases. Patieents have been randomised to the 4 bearing surfaces viz. Ceramic on XLinked Poly, Ceramic on Ceramic, Metal on Metal and Ceramic on Metal. Pre operative blood samples and follow up blood samples for metal ion analysis using the ICPMS method have been taken in all patients. As at February 2008 187 patients have been recruited, and metal ion levels at 1 year are available in 52 patients. Metal ion levels are not increased with Ceramic on XLinked Poly or Ceramic on Ceramic bearings. At 1 year follow up the metal ion levels in Ceramic on Metal bearings is half that of Metal on Metal bearings using the mean levels, and one third using the madian levels. Of note is that the chromium levels in Ceramic on Metal bearings is the least elevated. Due to laboratory evidence that Ceramic on Metal bearings have the best surface wear characteristics with no head stripe wear, and laboratory and clinical evidence of lower metal ion blood levels, Ceramic on Metal hip replacements could be a bearing surface of the future.
Whole blood metal ion levels remain a concern in those patients undergoing total hip replacement with metal bearing surfaces. The determination of baseline reference levels are essential if useful information can be gleaned from in vivo studies of functioning implants. We set out to prospectively determine chromium and cobalt metal ion concentrations in patients undergoing total hip replacement to determine reference levels of these metal ions. 100 patients with normal renal function, no occupational or environmental exposure to cobalt and chromium, and an absence of implanted metals were recruited into the study. Metal ion levels were determined using two different assay methods. Both ICP-MS (Inductively Coupled Plasma Mass Spectroscopy) and GFAAS (Graphite Furnace Atomic Adsorption Spectroscopy) are well recognised analytical techniques for the quantification of trace elements. Levels were correlated with gender, age and place of residence. There was considerable variability in whole blood metal ion levels, with the ICP-MS being more sensitive and consistent than the GFAAS method. Direct comparison of concentration levels determined by the two methods revealed no significant correlation. There was no correlation with age, gender and place of residence. Our findings would favour the use of the ICP-MS to determine reference levels and as a baseline for metalion surveillance pre-operatively in patients undergoing metal-on-metal total hip replacements. We also determined that changes in whole blood metal ion levels are more significant than actual levels in patients who have undergone total hip replacement.
This study reports on ceramic-on-metal (CoM) bearings in total hip replacement. Whole blood metal ion levels were measured. The median increase in chromium and cobalt at 12 months was 0.08 μg/1 and 0.22 μg/1, respectively, in CoM bearings. Comparable values for metal-on-metal (MoM) were 0.48 μg/1 and 0.32 μg/1. The chromium levels were significantly lower in CoM than in MoM bearings (p = 0.02). The cobalt levels were lower, but the difference was not significant. Examination of two explanted ceramic heads revealed areas of thin metal transfer. CoM bearings (one explanted head and acetabular component, one explanted head and new acetabular component, and three new heads and acetabular components) were tested in a hip joint simulator. The explanted head and acetabular component had higher bedding-in. However, after one million cycles all the wear rates were the same and an order of magnitude less than that reported for MoM bearings. There were four outliers in each clinical group, primarily related to component malposition.
The search for the ideal bearing surfaces to be used in Total Hip Replacement continues. The current “best” materials are felt to be various combinations of metal, ceramics and cross-linked polyethylene. Laboratory studies suggest that ceramic-on-metal articulations may provide distinct advantages. This study aims to identify the best combination with the lowest side effect profile. In February 2004 a prospective randomised trial on different bearing surfaces was started. The combinations selected were ceramic-on-cross-linked polyethylene, ceramic-on-ceramic, metal-on-metal and ceramic-on-metal. Institutional ethics clearance was obtained. In all patients uncemented femoral stems are used, with an uncemented porocoated acetabular shell. A uniform 28mm femoral head size was selected. Blood samples have been taken to measure the metal ion concentrations in all patients. These are measured pre operatively, and repeated at follow up visits at 3 months and 1 year, with further follow up at 3,5 and 10 years post operatively. Whole blood ion levels are measured using a graphite furnace atomic absorption spectrometer. Between February 2004 and April 2006 one hundred and ten hips have undergone total hip replacement. There are 105 patients (5 bilateral). 40% are males and 60% female. The average age at operation is 52 years (17 to 72). 49% hips are left and 51% right. Follow up includes blood samples and the Harris Hip Score. Complications to date have been surgeon related, with three femoral components needing early revision for technical reasons. This has not affected the bearing surfaces. Radiological and clinical assessment shows no difference between the different bearing surface groups. Post operative whole blood metal ion levels are compared to the patient’s pre operative level. To date there is no increase in the metal ion levels for the ceramic-on-cross linked polyethylene and ceramic-on-ceramic articulations. The ceramic-on-metal group is providing moderately raised metal ion levels, and the highest metal ion levels are in the metal-on-metal articulation group. At one year, the ceramic-on-metal group demonstrates a drop to close to pre-operative levels and these are still significantly lower than the metal-on-metal group. This confirms laboratory studies on the ceramic-on-metal articulation, which demonstrate significantly lower wear than comparable metal-on-metal articulations. The high level of metal ions in the latter groups has always been of concern. This study demonstrates a lower blood level of metal ions in the ceramic-on-metal group. If the in vivo wear rate in this group continues to replicate the laboratory wear studies, this articulation becomes a very attractive bearing surface in younger active patients, and may well become a bearing surface of choice in the future.
The authors evaluate the incidence, patterns and causative factors of avascular necrosis (AVN) in patients with developmental dysplasia of the hip (DDH) and to follow up these patients to determine what their long term functional and radiological outcome is. All patients treated for DDH by the same consultant with the subsequent development of AVN were assessed. Outcome was assessed by grading the AVN using the Kalamchi and McEwan classification at final follow up. A group of 250 hips with DDH were treated over a 16 year period and reviewed. All hips that developed AVN were studied. AVN was seen in 15% of hips treated with closed reduction and 62% of hips after open reduction–32% of the hips treated in the open reduction group were treated elsewhere and subsequently referred. If use of a Pavlik harness fails, children with DDH should be treated with pre reduction traction, closed reduction and spica cast after the age of 4 months. In the surgical group a capsulorrhaphy should be avoided. Poor radiological outcome at final follow up was not necessarily equivalent to a poor clinical outcome.
The search for the ideal bearing surfaces to be used in Total Hip Replacement continues. The current “best” materials are felt to be various combinations of metal, ceramics and cross linked polyethylene. This study aims to identify the best combination with the lowest side effect profile. In February 2004 a prospective randomised trial on different bearing surfaces was started. The combinations selected were ceramic on cross linked polyethylene, ceramic on ceramic, metal on metal and ceramic on metal. Institutional ethics clearance was obtained. In all patients uncemented femoral stems are used, and an uncemented porocoated acetabular shell. 28mm Head size was selected. Blood samples have been taken to measure the metal ion concentrations in all patients. These are measured pre operatively, and repeated at intended follow up visits at 3 months, 1, 3, 5 and 10 years post operative using a graphite furnace atomic absorption spectrometer. Between February 2004 and 2006 seventy hips have undergone total hip replacement. There are 85 patients (11 bilateral). 40% are males and 60% female. The average age at operation is 52 years (17 to 72). 46% Hips are left and 54% right. Follow up includes blood samples and the Harris Hip Score. Complications to date have been surgeon related, with three femoral components needing early revision for technical reasons. This has not affected the bearing surfaces. Ten patients have hetero-topic ossification. Cup inclination averages at 48 degrees (32 degrees to 69 degrees). Post operative blood metal ion levels are compared to the patient’s pre-operative level. To date there is no increase in the metal ion levels for the ceramic/cross linked poly ethylene and ceramic/ceramic articulations. The ceramic metal group is providing intermediate raised metal ion levels, and the highest metal ion levels are in the metal on metal articulation group. In the laboratory the ceramic on metal articulation demonstrates the least wear of all the groups studied, with metal on metal second. The high level of metal ions in the latter groups has always been of concern. This study demonstrates a lower blood level of metal ions in the ceramic on metal group. If the in vivo wear rate in this group is as good as the laboratory wear, it becomes a very attractive bearing surface in younger active patients.
The aim of this study was to asses the results of total hip replacements using the Elite Plus femoral stem. During the period 1995 to 2000, 212 total hip replacements were done using the Elite Plus femoral stem. These were followed up prospectively. The cohort of patients included 11 with bilateral hip replacements. 38% of patients were male and 62% were female. The average age at surgery was 61 years, with 18% being younger that 50 years at the time of surgery. All hip replacements were done using the same surgical and cementing techniques. Both cemented and uncemented cups were used in this cohort of patients. 2 patients died peri-operatively, and 22 hips were lost to follow-up. 6 hips have been revised, with 1 revision being due to sepsis and 5 due to loosening. A further 4 hips have radiographic evidence of early loosening, and 1 other hip has developed late sepsis. None of these 5 has yet been revised. Our survivorship at an average of 9 years is 97%. The survivorship of total hip replacements using the Elite Plus femoral stem in our unit is 97% at an average of 9 years. This compares very well with the results reported in other series. We do note though that there are 5 hips that may need revision, and this would bring the survivorship down to 94%. We feel that our good results are due to careful attention to surgical and cementing techniques, and this may explain our improved results compared to previous reports.
Good short-term results with Mt Blanc uncemented acetabular cups have been previously reported. However, in the medium term, we have observed acetabular loosening related to large granulomatous lytic lesions. To determine the cause of the polyethylene load causing the granulomatous lytic lesions, we subjected six explanted Mt Blanc acetabular cups to retrieval analysis. We also reviewed the literature on polyethylene locking mechanisms in uncemented metal-backed cups and on the deformability of metal-backed cups. We subjected the retrieved cups to stereo-photographic analysis and to dye penetration and surface scanning electron microscopy techniques. We demonstrated severe polyethylene wear and particle generation on the back surface of the polyethylene insert. This was due both to two-body sliding wear, as characterised by surface deformation and delamination of the polyethylene, and to three-body abrasive wear, as characterised by surface roughness and embedded titanium particles. The literature confirmed that the locking mechanism of the Mt Blanc cup was particularly poor and the deformability greater than in other cups tested. This confirmed the wear patterns on the back-surface of the polyethylene liner. We caution against the use of uncemented cups that have poor locking mechanisms for the polyethylene liners and those that deform excessively. The combination of poor locking mechanisms and titanium shells is especially dangerous.
The aim of this study was to measure polyethylene wear in uncemented metal-backed cups and compare it with cemented ultra-high molecular weight (UHMW) polyethylene cups in a controlled double-blind study. The study group was made up of 91 patients aged 50 to 70 years undergoing THR for unilateral OA of the hip between February 1995 and July 2002. The male to female ratio was 40:60. In all patients, a cemented stem and 28-mm ceramic head was inserted, using a third-generation cementing technique and UHMW polyethylene. Patients were randomly allocated to receive either a cemented or uncemented acetabular cup. Eight patients were lost to follow-up. Cemented cups were used in 28 patients (mean age 64 years). The mean thickness of the polyethylene was 9.6 mm (7.5 to 12.5). The mean liner thickness in the metal-backed cups was 8.9 mm (7 to 12.2). In measuring wear, baseline 3-month postoperative radiographs were compared with the most recent follow-up radiographs and Martell software was used. The polyethylene in metal-backed cups had a mean wear rate of 0.49 mm at 4.7 years, with a mean annual wear rate of 0.12 mm. The cemented polyethylene cups had a mean wear rate of 0.45 mm at 5.3 years, with a mean annual wear rate of 0.11 mm. The study is ongoing. Currently we conclude that there is no significant difference in the annual wear rate of polyethylene in uncemented metal-backed cups and cemented cups.
Osteolysis and subsequent mechanical loosening often occurs in hip arthroplasties using polyethylene-on-ceramic (POC) bearings. This has prompted an ongoing search for alternative bearing surfaces. Ceramic-on-ceramic (COC) and metal-on-metal (MOM) prostheses are widely used, with good clinical results. Using hip simulator studies, we compared ceramic-on-metal (COM) and MOM prostheses. We found COM pairings had 100-fold lower wear rates than MOM. The wear particles from both articulations were oval to round in shape and in the nanometer size range, with the COM producing smaller particles than the MOM. In both pairings, particle size decreased as the bearings bedded in. The volumetric particle loads were far smaller with COM bearing-surfaces than in currently-used MOM prostheses. These findings have encouraged us to investigate the use of these novel bearing surfaces. Ethical approval has been obtained, and a prospective randomised clinical trial comparing POC, MOM, COC and COM bearing surfaces has started.
The assessment of large allografts in acetabular reconstruction surgery is notoriously difficult. Because of their invasive natures, methods such as tetracycline-labeled histological examination are not recommended. Radio-isotope studies are unreliable in assessing the degree of incorporation because labeled tissues remain hot for extended periods. CT scans are impractical because of the scatter generated by the metallic prosthetic components. We used DEXA to assess the quality of large acetabular bone grafts immediately after surgery and at regular intervals thereafter. Software programmes were used to subtract the prosthetic components and give values for the remaining bony structure. DEXA results confirmed the initial adequacy of our grafting techniques. Serial scans showed the response of the grafts to both revascularisation and loading. Increases in high-load areas were higher than in low-load areas, reflecting the response of live bone to in vivo stresses. There were also changes reflecting the ongoing revascularisation of the grafts. These findings were born out by radiographs. DEXA is useful in assessing the incorporation and biological responses of large allografts in revision arthroplasty.
Over five years, 85 low-cost primary total arthroplasties (Eortopal Bulteamex) were done at a referral hospital. These were followed up for a mean of 48 months (minimum of 18 months). There were 11 revisions (13%), with four (4.7%) necessary for aseptic loosening, two (2.3%) for recurrent dislocations, four (4.7%) for sepsis and one (1.3%) for a periprosthetic fracture. When these results were compared with the Trent Regional Arthroplasty Register, the revision rate was noted to be four times higher than in the Trent study, with aseptic revisions being twice as high and infection rates three times higher. Dislocation rates were half those in the Trent study. We concluded that our lower dislocation rate probably reflected the quality of our surgery. Our higher sepsis rate was probably related to the hospital environment, and the high aseptic loosening rate due to the quality of the ‘low-cost’ prosthesis. We conclude that to be cost-efficient, ‘low-cost’ pros-theses must be of good quality and that the hospital environment must be optimal. This study highlights the need for an Arthroplasty Register in South Africa.
We have previously reported on early lytic lesions occurring when collared titanium prostheses are used. Previous finite element analysis studies (FEAs) showed that lytic lesions of the calcar were due to concentration of polyethylene wear particles under the collar by a ‘pumping action’. Further follow-up of these calcar lytic lesions showed that their rate of increase in size progressively slowed down. Further FEAs were performed to determine why this was so. An FEA mesh construct was developed, incorporating the new parameters of no contact between the collar and the calcar bone. A mechanical model to determine displacement parameters was also developed. These FEA studies demonstrated that the pumping action of the collar became less efficient as the size of the lytic lesions increased. This led to less concentration of polyethylene particles under the collar and fewer granulomatous reactions. The change in the proximal prosthesis-cement-bone construct may lead to cement mantle deterioration and earlier failure. We still recommend caution when a collared prosthesis is used, and the material and geometry of the prosthesis remain important.
Complex acetabular defects after failed total hip arthroplasty (THA) remain a major challenge in revision surgery. We managed 29 patients, of whom 27 had type-III and two type-IV defects (AAOS classification).The mean age of the 16 men and 13 women was 68 years (22 to 96). Use of a modular uncemented acetabular revision system allowed us accurately to position the construct, and then optimise the orientation of the polyethylene liner in respect of stability in the reduced hip. The modularity of the system allowed good access to do an impaction bone graft to restore the defects in the bone stock. Our follow-up ranged from 2 to 25 months. The orientation of the acetabular construct was measured radiologically and was at 50°. Our complications included four dislocations, two transient nerve palsies, one deep infection, four deep venous thromboses and one death from a pulmonary embolism. We conclude that the use of a modular acetabular reconstruction system is promising in these extremely difficult cases.
We managed three elderly patients who had central fracture dislocations with early total hip arthroplasty (THA), using anteprotrusio supports. Bone grafting was used to re-establish acetabular bone stock. Intraoperatively and postoperatively, these patients had no more complications than did patients undergoing THA for hip fractures. However, the surgical times were longer than for routine THA and blood replacement was slightly higher. Patients were mobilised early and aggressively. All became independent walkers and regained good range of movement. Radiologically the acetabular/pelvic fractures united and good bone-implant interfaces were established. There was no excessive heterotrophic bone formation. We regard THA in the management of acetabular fractures in the elderly as a reasonable approach, enabling patients to mobilise early and keeping morbidity to an acceptable level.
In June 1999 we instituted a prospective, randomised, double blind study to determine whether in the use of the Profix® total knee replacement system the addition of screws improved the fixation of the tibial base plate. There were 145 arthroplasties in 137 patients, 77% of them female. Their mean age was 66 years. In 86% of the patients there was osteoarthritis and in 14% an inflammatory arthritis. On a random basis, supplementary screws were inserted through the base plate into the tibia in 49% of arthroplasties and the remaining 51% were not fixed. In both groups the postoperative femoral angle measured 95% and at 12-month evaluation the tibial angle measured 89°, giving a total valgus angle of 6°. To date no tibial base plates have come loose. It is felt that the large central titanium peg plays a major role in assisting fixation of the base plate to the tibia and that supplementary screws are unnecessary. There is a small cost saving when screws are not used. We postulate that one may in the future minimise polyethylene debris tracking and osteolysis.
Anterior knee pain after total knee arthroplasty (TKA) occurs in 5% to 30% of patients whether or not the patella has been resurfaced. We retrospectively reviewed our patients, none of whom underwent patellar resurfacing. Only 2% had anterior knee pain, none requiring revision surgery. Our follow-up was between two and five years. We paid particular attention to removing osteophytes and conducting a thorough peripatellar synovectomy and a circumpatellar cautery denervation. Our results compare favourably to those in the literature, whether or not the patellae were resurfaced. We conclude that patellar resurfacing in TKA is unnecessary when careful attention is paid to the peripatellar tissues.
Complex acetabular defects after failed total hip arthroplasty (THA) remain a major challenge in revision surgery. We managed 29 patients, of whom 27 had type-III and two type-IV defects (AAOS classification). The mean age of the 16 men and 13 women was 68 years (22 to 96). Use of a modular uncemented acetabular revision system allowed us accurately to position the construct, and then optimise the orientation of the polyethylene liner in respect of stability in the reduced hip. The modularity of the system allowed good access to do an impaction bone graft to restore the defects in the bone stock. Our follow-up ranged from 2 to 25 months. The orientation of the acetabular construct was measured radiologically and was at 50°. Our complications included four dislocations, two transient nerve palsies, one deep infection, four deep vein thromboses and one death from pulmonary embolism. We conclude that the use of a modular acetabular reconstruction system is promising in these extremely difficult cases.
Resorption of the calcar below the collar of a titanium femoral prosthesis was observed. Biopsies of these lesions showed concentrations of polyethylene. We assessed the size of the resorption and correlated this with the size of the femoral prosthesis and the time since implantation. The age and the weight of the patient were also linked to the size of the prosthesis. We conducted a finite element analysis (FEA) of the femoral component-femur complex in both the loaded and unloaded situation. The FEA study demonstrated changing pressure under the collar that can be translated into microbending motions, with the degree of the movement dependent on the size of the prosthesis, the material of the prosthesis and the weight of the patient. We hypothesise that the existence of a ‘polyethylene pump’ due to the bending movements of the collared prosthesis concentrates polyethylene particles under the collar. We therefore postulated that the calcar resorption is due to the polyethylene granulomatous lesions, resulting from the micromotion of the collar of the prosthesis.
We have introduced a radiological scoring system to assess our technical competence in hip replacement surgery. We have also used it to assess the progress of the registrars in our training programme. This scoring method involves several parameters, including positioning of the components and the quality of interfaces. We compared our results before and after the introduction of this scoring system, and found that the quality of our surgery had improved. We conclude that an objective scoring system is valuable as a training aid, as well as in maintaining standards in our unit.
Over five years, 85 low-cost primary total arthroplasties (Eortopal Bulteamex) were done at a referral hospital. These were followed up for a mean of 48 months (minimum of 18 months). There were 11 revisions (13%), with four (4.7%) necessary for aseptic loosening, two (2.3%) for recurrent dislocations, four (4.7%) for sepsis and one (1.3%) for a periprosthetic fracture. When these results were compared with the Trent Regional Arthroplasty Register, the revision rate was noted to be four times higher than in the Trent study, with aseptic revisions being twice as high and infection rates three times higher. Dislocation rates were half those in the Trent study. We concluded that our lower dislocation rate probably reflected the quality of our surgery. Our higher sepsis rate was probably related to the hospital environment, and the high aseptic loosening rate due to the quality of the ‘low-cost’ prosthesis. We conclude that to be cost-efficient, ‘low-cost’ prostheses must be of good quality and that the hospital environment must be optimal. This study highlights the need for an Arthroplasty Register in South Africa.
Between February 2000 and August 2002, 60 Oxford unicompartment knee replacements were done on 51 patients, nine of whom had bilateral surgery. The mean age of patients, 82% of whom were women, was 66 years (45 to 83). Primary osteoarthritis was the pathology in 97% and post-traumatic arthritis in 3%. The mean range of movement increased from 113° preoperatively to 120° at the most recent follow-up. Complications included one case of deep vein thrombosis, one patient with bilateral tibial component loosening and three patients with loose cement particles in the joint. Most patients have no pain, but some have mild or occasional pain. One patient with bilateral unicompartmental replacements now has lateral knee pain. Unicompartment knee replacements are an alternative to total knee replacements, but there is a significant learning curve, particularly with regard to cementing techniques. Attention needs to be paid to removing all loose cement from the joint. Patient selection is critical. The complication rate remains low, however, and the results seem satisfactory.
In June 1999 a randomised double-blind study on tibial base plate fixation was started to determine whether, when using the Profix® total knee replacement, the addition of screws improved the fixation of the tibial base plate. To date 138 total knee arthroplasties (THAs) have been performed, 119 of which were available for study. Selected randomly, 56 patients had supplementary screws inserted through the base plate and 63 did not. After a minimum of 12 months follow-up, 27 patients had some radiolucent lines at the prosthesis-bone interface on the tibial component. These lines occurred in 14 cases with supplementary screws and 13 without screws. In one patient without supplementary screws, the tibial tray had subsided into the tibia. Statistically there was no apparent difference between the two groups. The early results of our study raise questions about the value of supplementing base plate fixation in uncemented THAs, especially considering the additional cost of the screws and their potentially detrimental consequences.
Severe central facture dislocations of the hip in the elderly can be catastrophic. Conservative treatment yields poor results with stiff painful hips. Reasonable hip function may be achieved with multiple surgical procedures and extended periods of immobilisation, but morbidity and mortality remain high. We managed three elderly patients who had central fracture dislocations with early total hip arthroplasty (THA), using anteprotrusio supports. Bone grafting was used to re-establish acetabular bone stock. Intraoperatively and postoperatively, these patients had no more complications than did patients undergoing THA for hip fractures. However, the surgical times were longer than for routine THA and blood replacement was slightly higher. Patients were mobilised early and aggressively. All became independent walkers and regained good range of movement. Radiologically the acetabular/pelvic fractures united and good bone-implant interfaces were established. There was no excessive heterotrophic bone formation. We regard THA in the management of acetabular fractures in the elderly as a reasonable approach, enabling patients to mobilise early and keeping morbidity to an acceptable level.
Primary total hip replacements are routine procedures with good outcomes. To ensure uniformly good results it is important that a thorough preoperative assessment of the patient is made. The prosthesis best suited to the patient and the pathology must be carefully selected and the optimal surgical technique must take into account patient, pathology and prosthesis parameters. We discuss patients’ problems such as morbid obesity, the different arthritides and neuromotor abnormalities. Acetabular problems, including dysplastic acetabula and acetabula protrusio, are dealt with in detail. We examine post-traumatic hip pathologies, including retained fracture implants, nonunions and ankyloses. On the femoral side, dysplastic femurs, post-traumatic malunions and post-osteotomies are dealt with.
We have introduced a radiological scoring system to assess our technical competence in hip replacement surgery and the progress of the registrars in our training programme. The scoring method involves several parameters, including positioning of the components and the quality of interfaces. We compared our results before and after the introduction of this scoring system, and found that the quality of our surgery had improved. We conclude that an objective scoring system is valuable as a training aid, as well as in maintaining standards in our unit.
Resorption of the calcar below the collar of a titanium femoral prosthesis was observed. Biopsies of these lesions showed concentrations of polyethylene. We assessed the size of the resorption, correlating it with the size of the femoral prosthesis and the time since implantation. The age and the weight of the patient was also linked to the size of the prosthesis. We conducted a finite element analysis (FEA) of the femoral component-femur complex in both the loaded and unloaded situation. FEA demonstrated changing pressure under the collar. This can be translated into micro-bending, with the degree of movement dependent on the size of the prosthesis, the material of the prosthesis and the weight of the patient. We postulate the existence of a ‘polyethylene pump’ owing to the bending motion of the collared prosthesis, and that calcar resorption is due to the resultant polyethylene granulomatous lesions .
In a radiological study, we evaluated the outcome of the Chiari osteotomy as a primary method of femoral head containment in a distinct group of patients with Perthes’ disease. Even when Salter’s prerequisites are met, the results of a Salter osteotomy are known to be poor in this particular group of patients. At the time of operation, the mean age of the 13 patients who underwent Chiari osteotomy was 9 years 4 months. The osteotomy was performed early in the disease process, before femoral head deformity had occurred. The hip was considered to be at risk because of the relatively late onset of the disease. Measurements were made on the preoperative and latest follow-up X-rays, and on the preoperative arthrogram. Patients were followed up for a mean of 3 years 4 months. On the preoperative arthrogram there was no femoral head deformity or hinging on abduction. At the time of surgery, it was too early to assign a hip reliably to a particular lateral or Herring lateral pillar group. However, during the follow-up period, 12 of the hips manifested as Catterall group IV and one as Catterall group II. Further, 11 hips advanced to become lateral pillar type B, and two to become lateral pillar type C. At follow-up, nine hips could be reliably graded according to the Stulberg classification: five were Stulberg type II and four Stulberg type III. It was clear that none of the remaining hips would be Stulberg type IV or V. The Chiari osteotomy achieves a congruent hip in a specific group of patients where a poor outcome would otherwise be anticipated.
Many authors believe that patellar resurfacing decreases the incidence of anterior knee pain. We analysed the results of 98 of our own patients (103 knees). Over the past two years, we performed total knee arthroplasty on 23 men (one bilateral) and 75 women (four bilateral). None of the patellae were replaced, but we carefully removed osteophytes, debrided the patella, and performed a thorough peripatellar synovectomy and circumpatellar cautery denervation. All patients on our database were telephoned, and those who reported pain were examined independently by a consultant and a registrar. If there was any controversy about the site of the pain, a third surgeon saw the patient. No patient had pain severe enough to warrant revision surgery. Only two (2%) had anterior knee pain, and in neither of them was it marked. We do not know whether our favourable results are attributable to the peripatellar synovectomy and/or circumpatellar cautery denervation. While we concede that a control study of patients who have undergone patellar resurfacing might be necessary before a firm conclusion can be drawn, we question whether, with such a low incidence of patellar problems, this is ethically justifiable.
Among elderly Caucasians, fractures of the femoral neck are a common cause of disability. Intertrochanteric and intra-capsular fractures occur equally often, and both are about three times more common in women than in men. Risk factors include neurological impairment, malnutrition, impaired vision, malignancy and decreased activity. We found that in black South Africans femoral neck fractures occur equally often in men and in women. Intracapsular fractures are comparatively rare, occurring in one of every eight female patients and one of every 3.5 male patients. Further, we found that in both black men and black women the femoral neck was consistently and significantly shorter than in whites. These results suggest that a short femoral neck may offer protection not only against intracapsular fracture, but also possibly against fracture of the femoral neck in general. In addition, greater cortical thickness in black people probably offers further protection.
A nine-year-old girl presented with a four-year history of progressive bowing of the left tibia. She had been seen in our clinic three and four years earlier, when no treatment was advised. She had been complaining of mild pain in the left leg for one month but was otherwise not very perturbed about her deformed leg. Examination showed bowing of the left tibia, no leg length discrepancy, no limp, and a normal left knee and ankle. Radiological examination showed features of osteo-fibrous dysplasia of the left tibia, with eccentric expansion of the cortex, intracortical osteolysis, marginal sclerosis encroaching on the medulla and diffuse lesional calcification. No other bones were involved. Because of the progressive bowing of the tibia and the mild pain, the recommended loose observation of the patient was abandoned. Daily treatment with 30 mg IVI pamidronate for three days resolved the pain. One month after the latest presentation the lesion was biopsied. Histological examination confirmed the diagnosis of osteofibrous dysplasia of the tibia. There was no evidence of adamantinoma. The literature on this rare bone lesion in childhood supports the use of open biopsy if the deformity becomes painful, the bowing is progressive and the patient presents after the age of nine years. Important differential diagnoses include fibrous dysplasia and adamantinoma.
Between January 1990 and October 2000, 108 total hip arthroplasties using a cemented polished titanium stem and a cemented ultra-high molecular weight polyethylene (UHMWP) cup were performed at our hospital. Because during routine follow-up visits we noted instances of resorption of the calcar, we decided to assess whether this was a problem. We were able to assess 85 of the original 108 hips. Calcar resorption was observed in 43 hips. The extent of calcar resorption varied from 1 mm to over 15 mm. In one patient a biopsy showed typical polyethylene granuloma. Because there is a risk of long-term failure, we concluded that it is inadvisable to use a cemented polished titanium stem when UHMWP is one of the bearing surfaces.