With better understanding of the factors influencing the performance of metal-on-metal bearings in total hip arthroplasty and with improved manufacturing technology, metal-on-metal bearings have become a promising option for active patients1-4. Nonetheless, several concerns still exist regarding the potential consequences of prolonged exposure to increased metal ion levels5,6, such as the issues of hypersensitivity7, carcinogenicity8, and adverse reaction to metal debris, which is also termed a pseudotumor9,10. The science of tribology, mathematical models, and joint simulator studies suggest that large-diameter components with small clearance reduce ion release from articulations11,12. This assumption led to the introduction of metal-on-metal total hip resurfacing and metal-on-metal total hip arthroplasty with large-diameter heads (i.e., head diameter >36 mm). Both of these types of hip arthroplasty systems provide improved stability and low-wear articulating surfaces, which is particularly attractive for use in patients with high-demand activity levels.
Measurement of metal ion release in patients who underwent metal-on-metal total hip resurfacing has shown increased levels13-15 or similar ion levels as compared with the ion levels in patients who underwent metal-on-metal total hip arthroplasty in which a 28-mm femoral head was used2,16,17. With the Durom total hip arthroplasty prosthesis with a large-diameter head (Zimmer, Warsaw, Indiana), significantly higher cobalt ion levels were reported in comparison with the Durom metal-on-metal total hip resurfacing prosthesis (Zimmer)18,19. Since both Zimmer systems—the metal-on-metal total hip resurfacing prosthesis and the total hip arthroplasty prosthesis with a large-diameter femoral head—presented the same bearing characteristics, wear and corrosion at the junction between the femoral neck and the adapter sleeve and open femoral head design were suspected to be responsible for the additional load of metal ion release.
Although a total hip arthroplasty prosthesis with a metal-on-metal bearing and a large-diameter femoral head is more stable than a total hip arthroplasty prosthesis with a 28-mm femoral head and is less challenging for the surgeon to perform than metal-on-metal total hip resurfacing is, those advantages should be weighed against the risk of exposure to higher metal ion levels from modular junctions. The aims of this study were: (1) to compare chromium, cobalt, and titanium ion concentrations in the whole blood of patients who received any of four types of total hip arthroplasty implants with a large-diameter femoral head; (2) to assess the progression of ion levels over time; and (3) to identify factors influencing metal ion concentrations.
Between April 2006 and November 2008, patients with degenerative hip joint disease who were scheduled to undergo unilateral total hip arthroplasty with use of a large-diameter femoral head were recruited from two hospitals. Patients were not considered for our study if they met one of the following exclusion criteria: pregnancy, known cutaneous metal allergy, renal insufficiency, the presence of another metal implant in the body, or not being willing to comply with the planned protocol. All patients gave written consent for inclusion in the study. The study was approved by our institutional scientific and ethics committees. A total of 173 patients fulfilled the inclusion criteria. Twenty-four patients were excluded for the following reasons: seven decided not to participate, five had simultaneous bilateral surgery, three were scheduled to receive a different type of total hip arthroplasty implant with a large-diameter femoral head than the types included in this study, two chose to have metal-on-metal total hip resurfacing, and seven underwent conventional hip arthroplasty because acetabular preparation did not allow for use of a monoblock acetabular cup. Five patients were excluded after surgery for the following reasons: one had an intraoperative greater trochanteric fracture fixed with a cable grip plate, one had early cup revision for malposition, one declined participation after sustaining cardiac problems, one died from sepsis, and one refused to participate because of postoperative limb-length inequality. This left 144 patients for analysis. Seven patients were lost to follow-up.
Patients received one of four different total hip arthroplasty implants (Biomet, DePuy, Smith & Nephew, or Zimmer) with a large-diameter femoral head (Fig. 1). The Biomet modular M2a-Magnum (Biomet, Warsaw, Indiana) femoral heads were implanted with uncemented titanium Taperloc (Biomet) femoral stems with a proximal titanium plasma spray coating. M2a-Magnum femoral heads are opened to receive a large adapter sleeve that closes the opening of the femoral head. The M2a-Magnum uncemented acetabular components are coated with titanium through a vacuum-sealed, titanium porous plasma spray process. The femoral head and acetabular component are both made of cast high-carbon-content cobalt-chromium alloy. Surface roughness was <25 nm, sphericity deviation was lower than 5 µm, and radial clearance was approximately 76 µm to 152 µm. The DePuy ASR XL (DePuy, Warsaw, Indiana) modular femoral heads and adapter sleeves that were used had uncemented titanium Tri-Lock (DePuy) femoral stems that were proximally coated with titanium beads. The uncemented ASR acetabular component is coated with cobalt-chromium beads fixed by heat treatment. The femoral head and acetabular component were both made of cast high-carbon-content cobalt-chromium alloy. Surface roughness was <5 nm, sphericity deviation was approximately 5 µm, and radial clearance was approximately 50 to 80 µm, depending on the cup diameter. The ASR large-diameter-head total hip arthroplasty system was recalled by DePuy in 2010. The Smith & Nephew BHR (Smith & Nephew Orthopaedics, Warwick, United Kingdom) modular femoral heads and adapter sleeves were implanted with uncemented titanium Anthology (Smith & Nephew) femoral stems featuring proximal titanium porous coating. The outer surface of the BHR acetabular component consists of as-cast cobalt-chromium beads covered with hydroxyapatite. The femoral head and acetabular component are both made of as-cast, high-carbon-content cobalt-chromium alloy. Surface roughness was <20 nm, sphericity deviation was <10 µm, and radial clearance was approximately 150 µm. The Zimmer Metasul modular femoral heads and adapter sleeves were used in conjunction with grit-blasted, uncemented titanium CLS Spotorno (Zimmer) femoral stems. The Durom uncemented acetabular component is coated with titanium through a vacuum-sealed, titanium porous plasma spray process. This implant was the worldwide version, which differs from the U.S. version in that it has a thinner porous titanium coating applied with use of a different process. The femoral head and acetabular component are both made of wrought-forged, high-carbon-content cobalt-chromium alloy. Surface roughness was <6 nm, sphericity deviation was <10 µm, and radial clearance was approximately 75 µm.
Photographs of the different large-diameter-head total hip arthroplasty implants used in this study: the Biomet Magnum (top right), the DePuy ASR (bottom right), and the Zimmer Durom (bottom left). An image was not available for the Smith & Nephew system. Ti = titanium, THA = total hip arthroplasty, CrCo = cobalt-chromium, and LDH = large-diameter femoral head.
The adapter sleeves of the Zimmer, DePuy, and Smith & Nephew groups were similar, consisting of a hollow cone (inner taper matching femoral stem taper and outer taper matching femoral head taper) made of cobalt-chromium and a thickness of approximately 3 mm (Fig. 1). For the Biomet implant, the adapter sleeve consisted of a large disk with a central tapered hole (inner taper matching the Type-1 femoral stem taper and outer taper matching the large opening of the femoral head [2° taper] made of titanium alloy). For some manufacturers, the contact area of the adapter sleeve with the femoral stem taper or the femoral head taper varies according to the design and adapter sleeve length and the design of the femoral stem taper.
To facilitate the logistics of implant inventory and instrumentation, we performed eight to twelve implantations of each large-diameter-head total hip arthroplasty type alternately until the planned number of patients was enrolled. All procedures were performed with the patient in the lateral decubitus position and by five hip surgeons who had considerable experience with uncemented hip arthroplasty and monoblock acetabular component implantation. A posterior surgical approach was used in 126 hips, and a modified Hardinge approach was used in eighteen hips. The surgical procedure proposed by each manufacturer was followed. In particular, for all cases, the adapter sleeve was inserted in the femoral head on the back table by the nurse and impacted with two sharp mallet blows. The surgeon impacted the femoral head on a clean, dry prosthetic femoral neck by multiple blows with a 1-kg mallet.
The primary outcome measures were the postoperative venous whole-blood concentrations (measured in µg/L) of chromium, cobalt, and titanium. Our standardized methods of blood collection and sample analysis with a high-resolution, sector-field, inductively coupled plasma mass spectrometer (Element 2 HR-ICP-MS; Finnigan MAT, Bremen, Germany) by a blinded independent laboratory (Trace Elements Laboratory, London Laboratory Services Group, London, Ontario, Canada) have been described previously16. Other outcome measures included the preoperative and two-year postoperative Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC)20 scores and the two-year postoperative activity levels, as evaluated with use of the University of California at Los Angeles (UCLA) activity score21. Radiographs of the pelvis, made postoperatively and at the time of the last follow-up, were analyzed for implant position and signs of loosening22. The arc of cover of the femoral head was calculated according to the method of De Haan et al.23.
Source of Funding
An unrestricted research grant was received from Zimmer.
Statistical Analysis
All statistical analyses were performed with SPSS, version 15.0 (SPSS, Chicago, Illinois). All 144 patients were included in the analyses. Continuous variables are presented as the mean plus the standard deviation, except for the metal ion concentrations, which are reported as median, mean, and range. Categorical variables are reported as frequency and percentage. Our methods were as follows: (1) To compare the patient population of each large-diameter-head total hip arthroplasty type, we conducted chi-square tests for categorical variables (i.e., side, sex, diagnosis, and sleeve length), analyses of variance for normally distributed continuous variables (i.e., age, preoperative WOMAC score, femoral head size, acetabular cup inclination, arc of cover of femoral head, and postoperative UCLA score), and Kruskal-Wallis tests to compare continuous variables that did not follow a normal distribution (i.e., body-mass index and postoperative WOMAC score). (2) Preoperative metal ion concentrations were analyzed with Kruskal-Wallis tests, as data were not normally distributed. Preoperative whole-blood titanium levels were found to be a significant predictor of postoperative titanium levels for implants from all four manufacturers. To account for the preoperative ion concentration, we subtracted the preoperative values from the metal ion concentrations obtained at each follow-up period and analyzed the difference between postoperative and preoperative data with use of Kruskal-Wallis tests to detect if differences existed between the four implant types for each metal ion at each follow-up point. Post hoc Mann-Whitney 2×2 tests were performed when the Kruskal-Wallis test showed a significant difference. (3) To account for non-normality of the data, the Friedman repeated-measures analysis of variance (nonparametric test for repeated measures) and the Wilcoxon nonparametric test were assessed if there were differences in ion concentrations between follow-up periods. (4) Spearman correlations evaluated if there was a relationship between chromium, cobalt, and titanium ions at twelve and twenty-four months and weight, height, and arc of cover, whereas the Mann-Whitney test was used for sex, sleeve length (short [—4 mm, —3 mm, —1 mm, 0 mm, and +2 mm] or long [+3 mm, +4 mm, +5 mm, +6 mm, and +8 mm]), head diameter (=49 mm or =50 mm), and acetabular cup abduction (<55° or =55°). Statistical significance was set at p < 0.05.
Demographic, implant-related, clinical, and radiographic data of all study groups are reported in the Appendix. Table I presents the whole-blood metal ion levels (in µg/L) in each study group at each follow-up period. For chromium ion levels, no significant differences were found between the four groups at all follow-up time periods. For cobalt ion levels, a significant difference was found between the groups at three, six, twelve, and twenty-four months, with the Zimmer implant showing the highest levels and the Biomet implant the lowest (p = 0.027, <0.001, 0.007, and 0.001 at three, six, twelve, and twenty-four months, respectively). Despite showing the lowest cobalt levels at each follow-up, very extreme cobalt levels were observed in two patients of the Biomet group (Fig. 2) (11.6 µg/L at three months and 16.5 µg/L at twelve months in one patient and 13.0 µg/L at twenty-four months in one patient). We noted an interaction between sex and type of large-diameter-head total hip arthroplasty system and also between head diameter (=49 mm or =50 mm) and type of large-diameter total hip arthroplasty system. In other words, cobalt levels were different between manufacturer types only in male patients or in patients who received a femoral head that was =50 mm in diameter. Since femoral head diameter and sex are closely related, it was not possible to determine which—male sex or a femoral head diameter of =50 mm—had the greatest influence on cobalt levels. At twenty-four months, 2×2 Mann-Whitney tests were thus performed in male patients. Significantly lower cobalt levels were found in the Biomet group (versus the Zimmer group [p = 0.001] and versus the Smith & Nephew group [p = 0.002]) and higher cobalt levels were observed in the Zimmer group (versus the Biomet group [p = 0.001] and the DePuy group [p = 0.015]) (Fig. 2). As for the effect of time on cobalt ion levels, only the Smith & Nephew group continued to demonstrate a further increase in cobalt levels between twelve and twenty-four months (p = 0.023). At all follow-up periods, titanium ion levels measured in the Zimmer group were significantly higher than in the other groups (2×2 Mann-Whitney tests performed in male patients at twenty-four months: p < 0.001 versus the Smith & Nephew group, p < 0.001 versus the Biomet group, and p = 0.001 versus the DePuy group) (Fig. 3). The Biomet large-diameter-head total hip arthroplasty group showed significantly increasing titanium levels between six and twelve months (p = 0.013), and a trend was observed between twelve and twenty-four months (p = 0.09). When all patients were pooled together, no significant correlations were seen at all follow-up evaluations between cobalt, chromium, or titanium ion levels and height, weight, or arc of cover. At twelve months, a significant difference was found between longer versus shorter sleeve lengths in the Zimmer (3.56 vs. 1.77 µg/L, p = 0.007) and DePuy (1.87 vs. 0.94 µg/L, p = 0.035) groups. However, this difference was no longer significant at twenty-four months. With all four groups pooled together, cup inclination of <55° or =55° (1.12 vs. 2.18 µg/L, respectively, p = 0.077) showed a difference in the twenty-four-month cobalt level that almost reached statistical significance. In the Zimmer group, patients with the open femoral head design (head diameter >50 mm) showed higher cobalt levels at twelve months (2.8 µg/L) compared with patients with closed femoral head design (1.9 µg/L), but this was not statistically different (p = 0.067).
Box plot chart showing cobalt ion levels stratified by manufacturers and follow-up periods. Box lengths represent the interquartile range (first to third quartiles). The line in the center of the boxes shows the median value. The lower end of the vertical bar below each box represents the lowest non-outlier value. Similarly, the upper end of the vertical bar above each box represents the highest non-outlier value. Data indicated by the circles are outliers (being more than 1.5 to 3.0 times the interquartile range over the third quartile), and data flagged by the small stars are extreme values (more than three times the interquartile range over the third quartile). Data represented by the large stars (and the corresponding values) were placed at the 10 µg/L level arbitrarily to facilitate visualization of the box plot chart at a greater scale. Each mark indicates one patient.
Box plot chart showing titanium ion levels stratified by manufacturers and follow-up periods. Box lengths represent the interquartile range (first to third quartiles). The line in the center of the boxes shows the median value. The lower end of the vertical bar below each box represents the lowest non-outlier value. Similarly, the upper end of the vertical bar above each box represents the highest non-outlier value. Data indicated by the circles are outliers (being more than 1.5 to 3.0 times the interquartile range over the third quartile), and data flagged by the small stars are extreme values (more than three times the interquartile range over the third quartile). Each mark indicates one patient.
At the time of the last follow-up, no cases of clinical or radiographic loosening of the acetabular cup or femoral component were found. One patient in the Zimmer group required revision surgery for persisting groin pain and a mass in the lateral aspect of the buttock. A work-up for infectious disease was negative. Radiographs demonstrated osteolysis of the femoral calcar and lesser trochanter (Fig. 4-A). Whole-blood cobalt ion levels were 1.02, 5.06, and 3.78 µg/L at six, twelve, and twenty-four months, respectively. Preoperative ultrasound-guided sampling of the intra-articular effusion revealed chromium, cobalt, and titanium levels of 10,396, 8188, and 273 µg/L, respectively. Intraoperative findings included a massive, creamy-like effusion in the posterolateral aspect of the hip, minimal necrosis in the short rotator and gluteus minimus muscles, and capsular necrosis. Metallic deposits were observed at the base of the prosthetic femoral neck and inside the adapter sleeve (Figs. 4-B and 4-C).
Anteroposterior radiographs and 1.5 Tesla T2-weighted stir axial magnetic resonance image showing signs of a reaction to metal debris in a patient who presented with pain after undergoing a Zimmer large-diameter-head total hip arthroplasty. When compared with the immediate postoperative radiograph (left), the three-year postoperative radiograph (center) shows the development of osteolysis of the calcar and lesser trochanter (arrow), and the magnetic resonance image (right) shows the presence of massive joint effusion extending around the greater trochanter (arrow).
Figs. 4-B and 4-C After disengaging the well-fixed femoral head from the neck of the femoral stem, black metallic debris was observed intraoperatively at the base of the taper (Fig. 4-B) and inside the adapter sleeve (Fig. 4-C).
Metal-on-metal bearings in femoral heads that have a diameter of 28 or 32 mm have produced few adverse reactions, low ion release, and excellent clinical results with more than twenty-five years of use24-26. First intended for the revision of failed hip resurfacings, large-diameter-head total hip arthroplasty systems became more popular as primary procedures, with the advantage of increased stability and reported beneficial wear characteristics with subsequent lower metal ion release. The goal of this study was to compare the liberation of metal ions, as measured in the whole blood of patients who received one of four different metal-on-metal large-diameter-head total hip arthroplasty systems. We found significant differences in cobalt and titanium ion levels between the various large-diameter-head total hip arthroplasty systems, and of concern was the observation of metal deposits inside and outside the taper adapter of one patient with a Zimmer large-diameter-head total hip arthroplasty implant.
Our study has limitations. In vivo measurements of whole-blood ion levels do not always reflect the local ion load about the implant and do not allow the discrimination of ion release from wear of articulating surfaces or other sources. It thus only represents an estimation of the total ion load liberated from the implants. A direct comparison of each large-diameter-head total hip arthroplasty system with its metal-on-metal total hip resurfacing counterpart would have allowed estimation of the ion load generated at bearing surfaces compared with ions released from modular junctions or passive corrosion of exposed surfaces. Although this study is the largest prospective report on metal ion levels in patients with large-diameter-head total hip arthroplasty, the limited number of subjects per group and the presence of missing values most likely reduced the potential for reaching significance in some analyses. In addition, a randomized study would have resulted in a more optimal protocol design but would have been logistically difficult to conduct. Nonetheless, patient characteristics were similar in all groups (see Appendix). Furthermore, our study allowed valid comparisons of whole-blood ion levels between study groups because the blood was collected prospectively by experienced nurses at multiple follow-up periods according to a standardized technique, and the analyses were performed blindly in an independent laboratory with use of high-resolution, sector-field, inductively coupled plasma mass spectrometry with low detection limits.
As shown in Table I and illustrated in Figure 2, we observed different cobalt ion levels in the four large-diameter-head total hip arthroplasty groups at all follow-up times. At twenty-four months, the Zimmer large-diameter head total hip arthroplasty group demonstrated a mean cobalt level that was four times greater than that of the Biomet implant group (2.68 vs. 0.65 µg/L, respectively; p = 0.001). No difference in chromium ion levels was found between groups. Chromium levels are much less predictable than cobalt levels, and serum or whole-blood cobalt levels are therefore most often relied on for the evaluation of metal ion release after total hip arthroplasty27,28. Type of metal, manufacturing process, design characteristics, and type of coating may all play a role in ion release, along with the design of modular junctions. We have previously reported the significant contribution of the junction between the cobalt-chromium-alloy adapter sleeve and the titanium femoral stem neck of the Zimmer large-diameter-head total hip arthroplasty system to whole-blood cobalt ion levels19. In that previous study, we observed cobalt levels of 0.7 µg/L after Durom metal-on-metal total hip resurfacing and 2.2 µg/L after Durom large-diameter-head total hip arthroplasty at twelve months postoperatively, which represents a threefold increase in cobalt release in the large-diameter-head total hip arthroplasty group compared with the metal-on-metal total hip resurfacing counterpart (p < 0.001). Modular junctions in total hip arthroplasty can cause significant metal ion release by means of wear or corrosion when the protective oxidized metal surface is disrupted by fretting or micromotion29,30, and the magnitude of the corrosion processes is related to the number and quality of metallic junctions29-31. Garbuz et al. stopped a randomized study comparing the Durom large-diameter-head total hip arthroplasty system to the Durom metal-on-metal total hip resurfacing system because of concern over high metal ion levels in the large-diameter-head total hip arthroplasty group18. Moreover, clinical signs of wear and deformation of the cobalt-chromium adapter sleeve of the Durom large-diameter-head total hip arthroplasty system along with metallic deposits from corrosion were seen in one patient of this study as well as in seven other patients who underwent revision arthroplasty in our hospital (eight of 538 cases = 1.5%). This would support the impression that the junction of the adapter sleeve and the femoral stem may be a weak link in the Zimmer large-diameter-head total hip arthroplasty systems. More research is needed to confirm or refute this assumption, but wear and fretting corrosion at the junction between the adapter sleeve and the femoral stem neck may enter a continuous cycle that could produce more micromotion, promoting further metal ion release locally and into the blood. Significant damage to the femoral stem taper might also become problematic, as a well-fixed femoral stem may need revision. As in this study, local metal-ion load about the hip joint was found to be directly related to the occurrence of adverse tissue reactions10,32. Svensson et al. and others have reported similar problems9,33,34. With regard to the DePuy ASR XL system, although it was recalled in 2010 because of a high revision rate, no loosening or revision of these components occurred in our study. Despite the use of a cobalt-chromium adapter sleeve, the cobalt levels associated with the DePuy implant at the twenty-four month follow-up period were significantly lower than those associated with the Zimmer device (1.78 versus 2.99 µg/L, respectively [p = 0.015]). In our opinion, the bearing surface and/or cobalt-chromium adapter sleeve or other features of the ASR XL system seem to possess better design characteristics than those of the Zimmer device. The mean cobalt levels for the type of DePuy implant used in our study were lower than 3.26 µg/L (range, 1.1 to 32 µg/L) at forty-one months (range, ten to fifty-seven months) after surgery, as measured by Langton et al.10. The Smith & Nephew large-diameter-head total hip arthroplasty system was the only system that demonstrated an increasing cobalt level at the twenty-four-month follow-up point and longer follow-up is needed to see if cobalt levels will stabilize with time as was reported with the Smith & Nephew hip-resurfacing system14,35.
In the present study, the Biomet group showed the lowest cobalt levels at each follow-up point. Nevertheless, the Biomet metal-on-metal articulation and/or the junction between the adapter sleeve and the femoral head can have similar problems as shown by the extremely high cobalt levels in two patients. Although better wear characteristics of the Biomet bearing surface or other design characteristics of the Biomet large-diameter-head total hip arthroplasty implants may be responsible for the low levels of cobalt ions, one may hypothesize that the design of the taper adapter sleeve is likely one important factor to explain the lower ion levels in the Biomet design. The Biomet large-diameter-head total hip arthroplasty system is the only system that possesses an adapter sleeve made of titanium, and it possesses design characteristics that differ substantially from other large-diameter-head total hip arthroplasty systems. The junction between the titanium adapter sleeve and the titanium stem neck of the Biomet large-diameter-head total hip arthroplasty system cannot participate in cobalt ion release. Interestingly, even with a titanium acetabular coating and a titanium modular sleeve, the Biomet system did not show higher titanium levels when compared with the levels in the other groups. However, the titanium levels in the Biomet group were seen to be significantly increasing in the six-month to twelve-month follow-up period (p = 0.013), with a trend toward further increases observed in the twelve to twenty-four-month period (p = 0.09). The titanium adapter sleeve of the Biomet system may produce more titanium ions with time. Conversely, it may, with time, undergo cold-welding with the titanium stem neck and reduce the potential for fretting corrosion and titanium ion release. Moreover, the Biomet adapter sleeve is large and thick, which may reduce its deformation under load, hence decreasing the potential for micromotion and fretting corrosion, both at the junction with the stem and the femoral head. At the twenty-four-month follow-up point, titanium ion levels were significantly higher in association with the Zimmer large-diameter-head total hip arthroplasty group as compared with the other three groups (p < 0.001). We have previously shown that the titanium plasma spray coating of the Durom acetabular cup produces high titanium ion levels16,36. On the other hand, the titanium plasma spray on the outer surface of the acetabular cup of the Biomet system may be more stable.
Why would ion release be greater with metal-on-metal large-diameter-head total hip arthroplasty systems than with smaller metal-on-metal articulations? We have previously observed increased ion levels in head sizes of 50 mm or more of the Zimmer large-diameter-head total hip arthroplasty system and hypothesized that the open femoral head design in femoral heads with a diameter of 50 mm or more could be responsible for these levels due to passive corrosion of the increased surface area of exposed metal36. We now believe that the open femoral head design is unlikely to be the only factor involved, if it is involved at all. The increased head diameter of a metal-on-metal bearing produces more friction at the articulating surface, which increases the rotational moment at the junction between the adapter sleeve and the stem neck. The friction moment also depends on other factors besides head diameter, such as clearance, area in contact between articulating surfaces, and surface roughness. Moreover, longer sleeve length may reduce the contact area with the stem neck. In this study, higher cobalt levels were measured in patients in the long sleeve group compared with the short adapter sleeve group of the Zimmer (p = 0.007) and DePuy (p = 0.035) systems at the twelve-month follow-up point, but these differences became insignificant at twenty-four-month follow-up. Finally, the suction effect of large metal-on-metal articulations may further increase constraints at the junction between the adapter sleeve and the stem neck. All of these factors are minimized with a smaller femoral head diameter, which may explain the lower metal ion load as well as the rarity of adverse reaction to metal debris with the 28 and 32-mm metal-on-metal articulations. Mechanical demands on the adapter sleeve-stem neck junction may also increase in younger, active patients for whom large-diameter-head total hip arthroplasty is frequently required. Factors that may provide a stronger connection between the adapter sleeve and the stem neck could include greater surface contact area, material favoring cold-welding with the stem neck, and a design that avoids use of an adapter sleeve or modularity altogether. Older taper designs (long, without a slimmer neck) as in the Biomet Taperloc femoral stem may be favorable compared with newer, shorter trunnions with a slender neck (e.g., the Anthology design by Smith & Nephew and the CLS design by Zimmer). Surgeons may also play a role by cleaning and drying the stem taper37 and correctly impacting the adapter sleeve with heavy blows in the axis of the stem neck. It is unknown if modular large heads without a sleeve may also sustain the same problem at the femoral head stem-neck junction. Until a better understanding of the impact of increasing head size on the taper junction in large-diameter-head total hip arthroplasty is gained, surgeons should be cautious about using the largest head size in a metal-on-metal large-diameter-head total hip arthroplasty system unless using it in implants without a modular junction, such as metal-on-metal total hip resurfacing implants.
We have shown that large-diameter-head total hip arthroplasty systems vary in terms of metal ion release and among several design and bearing characteristics; It appears that the adapter sleeve is playing a more significant role in ion release than anticipated. Therefore, the high levels of metal ions that have been observed in association with some total hip arthroplasty implants with metal-on-metal bearings should not be attributed solely to the bearing surfaces. Increasing the femoral head size in hip arthroplasty implants is beneficial in many respects, especially with regard to joint stability and avoidance of component impingement. However, current technologies and/or currently available metal-on-metal large-diameter-head total hip arthroplasty designs may not yet allow for the widespread use of large modular heads with metal-on-metal articulation in total hip arthroplasty, especially in young and active patients who generate high loads at the hip joint. Further research is needed to better understand the favorable design characteristics, such as modular junctions, in metal-on-metal large-diameter-head total hip arthroplasty. Furthermore, in addition to laboratory testing, new implant designs should undergo a thorough clinical evaluation before they are allowed to be used in a widespread manner.