Metal Release in Patients Who Have Had a Primary Total Hip Arthroplasty

Transcription

1 Copyright ] 998 by The Journal of Bone and Joint Surgery, Incorporated Metal Release in Patients Who Have Had a Primary Total Hip Arthroplasty A PROSPECTIVE, CONTROLLED, LONGITUDINAL STUDY* BY JOSHUA J. JACOBS, M.D.f, ANASTASIA K. SKIPOR, M.S.t, LESLIE M. PATTERSON, R.N.f, NADIM J. HALLAB, PH.D.t, WAYNE G. PAPROSKY, M.D.f, CHICAGO, JONATHAN BLACK, PH.D.*, KING OF PRUSSIA, PENNSYLVANIA, AND JORGE O. GALANTE, M.D., D.SC.f, CHICAGO, ILLINOIS Investigation performed at Rush-Presbyterian-St. Luke's Medical Center, Chicago ABSTRACT. There is an increasing recognition that, in the long term, total joint replacement may be associated with adverse local and remote tissue responses that are mediated by the degradation products of prosthetic materials. Particular interest has centered on the metaldegradation products of total joint replacements because of the known toxicities of the metal elements that make up the alloys used in the implants. We measured the concentrations of titanium, aluminum, cobalt, and chromium in the serum and the concentration of chromium in the urine of seventy-five patients during a three-year prospective, longitudinal study. Twenty patients had had a so-called hybrid total hip replacement (insertion of a modular cobalt-alloy femoral stem and head with cement and a titanium acetabular cup without cement), fifteen had had insertion of an extensively porous-coated cobalt-alloy stem with a cobalt-alloy head and a titaniumalloy socket without cement, and twenty had had insertion of a proximally porous-coated titanium-alloy stem with a cobalt-alloy head and a titanium socket without cement. The remaining twenty patients did not have an implant and served as controls. The results of our study showed that, thirty-six months postoperatively, patients who have a well functioning prosthesis with components containing titanium have as much as a threefold increase in the concentration of titanium in the serum and those who have a well functioning prosthesis with cobalt-alloy components *One or more of the authors has received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article. In addition, benefits have been or will be directed to a research fund or foundation, educational institution, or other non-profit organization with which one or more of the authors is associated. Funds were received in total or partial support of the research or clinical study presented in this article. The funding sources were Grant AR39310 from the National Institute of Arthritis and Musculoskeletal and Skin Diseases; Zimmer, Incorporated, Warsaw, Indiana; and the Rush Arthritis and Orthopedics Institute, Chicago, Illinois. ( Department of Orthopedic Surgery, Rush Arthritis and Orthopedics Institute, Rush Medical College, Rush-Presbyterian-St. Luke's Medical Center, 1653 West Congress Parkway, Chicago, Illinois address for Dr. Jacobs: jacobs@ortho4.pro.rpslmc.edu. timn Biomaterials, 409 Dorothy Drive, King of Prussia, Pennsylvania have as much as a fivefold and an eightfold increase in the concentrations of chromium in the serum and urine, respectively. The predominant source of the disseminated chromium-degradation products is probably the modular head-neck junction and may be a function of the geometry of the coupling. Passive dissolution of extensively porous-coated cobalt-alloy stems was not found to be a dominant mode of metal release. CLINICAL RELEVANCE: Increased concentrations of circulating metal-degradation products derived from orthopaedic implants may have deleterious biological effects over the long term that warrant investigation. This is a particularly timely concern because of recent clinical trends, including the reintroduction of metalon-metal bearing surfaces and the increasing popularity of extensively porous-coated devices with large surface areas of exposed metal. Accurate monitoring of the concentrations of metal in the serum and urine after total hip replacement also can provide insights into the mechanisms of metal release. Our findings suggest that fretting corrosion at the head-neck coupling is an important source of metal release that can lead to increased concentrations of chromium in the serum. Determinations of the concentrations of metal in the serum and urine may be useful in the diagnosis of patients who are symptomatic after a total joint replacement as increased levels are indicative of at least one mode of mechanical dysfunction (for example, fretting corrosion) of the device. Total joint arthroplasty has emerged as one of the success stories of modern orthopaedics. In most patients, the short and intermediate-term (two to seven-year) results have been excellent when appropriate prosthetic designs and operative techniques have been used. However, as longer-term results have become available and as newer prosthetic designs that are intended for use in younger, more active individuals have been introduced, several problems that warrant concern and careful investigation have come to light. Among these problems is the increasing recognition that, in the long term, total joint replacements may be associated with adverse local VOL. 80-A, NO. 10, OCTOBER

2 1448 J. J. JACOBS ET AL. TABLE I CONCENTRATIONS OF TITANIUM IN THE SERUM* Preop. or Baseline At 12 Mos. At 36 Mos. Group 1 (20 patients)t 1.60 ( ) ( ) ( ) 11 Group 2 (15 patients)t 2.93 ( ) ( ) ( ) 7 Group 3 (20 patients)t 1.40 ( ) ( ) ( ) 6 Group 4 (20 patients)t 1.70 ( ) ( ) ( ) 19 *The values for the mean and the range are given as nanograms per milliliter (parts per billion). Concentrations below of 2.11 nanograms per milliliter (parts per billion) were assigned a value that was one-half of, according to convention. tthe patients in Group 1 had a so-called hybrid total hip replacement (a modular cobalt-alloy femoral stem and head inserted with cement and a titanium acetabular cup inserted without cement); those in Group 2, an extensively porous-coated modular cobalt-alloy stem with a cobalt-alloy head and a titanium-alloy socket, inserted without cement; those in Group 3, a proximally porous-coated modular titanium-alloy stem with a cobalt-alloy head and a titanium socket, inserted without cement; and those in Group 4 (controls), no implant. and remote tissue responses 1 ' 627, These adverse effects are mediated by the degradation products of prosthetic materials, which may be present as particles of wear debris, colloidal organometallic complexes (specifically or non-specifically bound), free metal ions, or inorganic metal salts or oxides; they may also be present in an organic storage form such as hemosiderin. Much of the focus on the long-term biocompatibility of implant materials has centered on the metal components of joint replacements because of their tendency to undergo electrochemical corrosion resulting in the formation of (at least transiently) chemically active degradation products' 9. The issue of metal release from prosthetic devices has taken on an increasing sense of urgency, for several reasons. These reasons include (1) the recognition, in a variety of clinical settings, of the deposition of extensive metal particulate debris within local and remote tissues that have enormous specific surface areas available for electrochemical interaction with the surrounding tissue fluids ; (2) the reintroduction of metal-on-metal articulating devices because of the recognition of deleterious local effects (periprosthetic osteolysis) of polyethylene wear debris from metal and ultra-high molecular weight polyethylene wear couples ,3638 ; (3) the recent observation of severe mechanically assisted crevice-corrosion processes in modular and multiple-part metal joint-replacement components associated with extensive local deposition of corrosion products in the face of newer designs incorporating increasing amounts of modularity; (4) the increasing popularity of porous-coated devices, with large surface areas, that are designed to be inserted without cement, particularly in younger patients who have a postoperative life expectancy of more than twenty years; and (5) the concern that metal-degradation products can stimulate the development of osteolysis, both directly, through macrophage activation, and indirectly, by accelerating polyethylene wear by means of a threebody mechanism 21. The purpose of the current study was to measure the concentrations of metal in the serum and urine of patients who had had a primary total hip arthroplasty. The concentrations of titanium, aluminum, cobalt, and chromium in the serum as well as the concentration of chromium in the urine were determined for three groups of patients and a group of controls. Investigations such as this one are important initial steps in our understanding of the bioreactivity and bioavailability of metal-degradation products generated by total hip-replacement components. Study Groups Materials and Methods Seventy-five patients were prospectively enrolled in this study. Twenty did not have an implant or systemic disease, and they served as controls. The remaining fifty-five had had a unilateral primary total hip replacement for the treatment of osteoarthrosis; all had a well functioning prosthesis (a Harris hip score 17 that was good [80 to 89 points] or excellent [90 to 100 points]), and none had radiographic evidence of loosening or osteolysis at the follow-up intervals of twelve and thirty-six months. All fifty-five patients had a titanium or THE JOURNAL OF BONE AND JOINT SURGERY

3 METAL RELEASE IN PATIENTS WHO HAVE HAD A PRIMARY TOTAL HIP ARTHROPLASTY 1449 titanium-alloy acetabular component. No metal implant other than the total hip replacement was used in any patient. The fifty-five patients were divided into three groups on the basis of the composition and the method of fixation of the femoral stem. Group 1 comprised twenty patients who had had a so-called hybrid total hip replacement. The mean age of the nine men and eleven women at the time of the operation was sixty-six years (range, forty-one to seventy-seven years). The femoral component consisted of a modular cobalt-alloy stem with a cobalt-alloy head and had been inserted with cement. Six Iowa and fourteen Precoat femoral components (both types manufactured by Zimmer, Warsaw, Indiana) were inserted. The acetabular component, which had been inserted without cement, consisted of a commercially pure titanium shell with a diffusion-bonded, commercially pure, titanium fiber-metal porous-coated surface (Harris-Galante II; Zimmer). The acetabular component was secured to the pelvis with a varying number of Ti-6A1-4V self-tapping screws (generally two or three). A so-called snap-fit ultra-high molecular weight polyethylene liner served as the bearing surface. Group 2 comprised fifteen patients who had had insertion of a cobalt-alloy femoral component with an extensively porous-coated surface and a titanium-alloy acetabular component; both components had been inserted without cement. The mean age of the seven men and eight women at the time of the operation was sixty-two years (range, forty-two to eighty-two years). Eight AML, three Solution, and four Prodigy femoral components (all three types manufactured by DePuy, Warsaw, Indiana) were inserted. The surgeon considered the quality of the bone stock and the geometry of the femur in determining which type of modular stem was best for each patient. The acetabular component consisted of a titanium-alloy shell with a beaded commercially pure titanium porous coating. Twelve Duraloc and three Solution acetabular components (both types manufactured by De Puy) were inserted. The twelve Duraloc shells were secured to the pelvis with three porous-coated titanium-alloy spikes, and the three Solution shells were secured with two titanium-alloy screws. The articulating couple consisted of a modular cobalt-alloy femoral head and an ultra-high molecular weight polyethylene liner. Group 3 comprised twenty patients who had had insertion of a titanium-alloy femoral stem with a cobalt-alloy femoral head and an acetabular component that was identical to that used in Group 1; both the femoral and the acetabular components had been inserted without cement. The mean age of the ten men and ten women at the time of the operation was fifty-four years (range, thirty-eight to sixty-seven years). The femoral stem was made of Ti-6A1-4V with a commercially pure titanium fiber-metal porous-coated surface that had been diffusion-bonded onto the proximal aspect of the device. Twelve Anatomic femoral components, seven Harris-Galante Multilock components, and one Harris-Galante porous-coated (HGP) component (all three types manufactured by Zimmer) were inserted. The articulating couple consisted of a modular cobalt-alloy femoral head and an ultrahigh molecular weight polyethylene liner. Group 4 consisted of twenty patients who did not have an implant or systemic disease; they served as controls. The mean age of the ten men and ten women at the time of enrollment in the study was fifty-nine years (range, thirty-five to seventy-five years). All patients provided informed consent. The study was approved by the Human Investigation Committees at Rush-Presbyterian-St. Luke's Medical Center and Central DuPage Hospital. All patients who met the criteria for selection (unilateral osteoarthrosis of the hip and no other metal implant in the body) were eligible for enrollment in the study. The levels of metal in the serum were measured in all patients at each of the three testing intervals. All prostheses had been inserted between 1989 and Twenty-five patients dropped out of the study. Specifically, eight patients dropped out because they had another total joint replacement; six, because they had missed at least one follow-up visit; five, because they were not willing to continue to participate; two, because of mechanical failure necessitating revision of the prosthesis; and one each, because of death, renal failure requiring dialysis, occupational exposure to metals, and recurrent instability. CD Q. CL Z O I- 2 Lil O Z o Group 1 mam Group 2 i H Group 3 M Group 4 1/2 D.L. "Different than controls (p< 0.03) MONTHS FIG. 1 Bar graph showing the concentrations of titanium in the serum, in nanograms per milliliter (parts per billion), as a function of time in the four groups that were studied. At twelve and thirty-six months postoperatively, the levels in Group 3 had increased 1.8-fold and 3.0-fold, respectively, compared with the preoperative concentrations. Group 3 also had a 1.7-fold increase at thirty-six months compared with the level at twelve months. Intergroup comparisons at thirty-six months revealed increases of 1.9-fold, 2.2-fold, and 3.4-fold in Groups 1, 2, and 3, respectively, compared with the level in the controls (Group 4). Group 1 had a so-called hybrid total hip replacement (a modular cobalt-alloy femoral stem and head inserted with cement and a titanium acetabular cup inserted without cement); Group 2 had an extensively porous-coated modular cobalt-alloy stem with a cobalt-alloy head and a titanium-alloy socket, inserted without cement; Group 3 had a proximally porous-coated modular titanium-alloy stem with a cobalt-alloy head and a titanium socket, inserted without cement; and Group 4 (controls) had no implant. D.L. = detection limit. VOL. 80-A, NO. 10, OCTOBER 1998

4 J. J. JACOBS ET AL. TABLE II CONCENTRATIONS OF ALUMINUM IN THE SERUM* Preop. or Baseline At 12 Mos. At 36 Mos. Group 1 (20 patients)t Mean concentration 2.17± Group 2 (15 patients)t Mean concentration ( ) Group 3 (20 patients)t Mean concentration ( ) Group 4 (20 patients)t Mean concentration 1.65 ( ) 1.25 ( ) *The values for the mean and the 95 per cent confidence interval (for the groups and time-periods in which all values were above the detection limit) or for the mean and the range (values in parentheses) are given as nanograms per milliliter (parts per billion). Concentrations below the detection limit of 0.35 nanogram per milliliter (part per billion) were assigned a value that was one-half of, according to convention. tthe patients in Group 1 had a so-called hybrid total hip replacement (a modular cobalt-alloy femoral stem and head inserted with cement and a titanium acetabular cup inserted without cement); those in Group 2, an extensively porous-coated modular cobalt-alloy stem with a cobalt-alloy head and a titanium-alloy socket, inserted without cement; those in Group 3, a proximally porous-coated modular titanium-alloy stem with a cobalt-alloy head and a titanium socket, inserted without cement; and those in Group 4 (controls), no implant. Collection of Specimens A twenty-four-hour sample of urine and a contemporaneous sample of blood were collected from each subject immediately preoperatively (or at baseline for the controls) and then twelve and thirty-six months later. All vessels and utensils that were used for the collection of specimens were verified to be free of metal contamination as described previously 22. Blood samples were obtained with use of siliconized butterfly needles in triplicate polypropylene syringes. The first syringe was used to rinse the system, and metal-ion analysis was performed on the contents of the second and third syringes. The blood was separated and frozen at -80 degrees Celsius as serum and clot fractions. The urine samples also were stored at -80 degrees Celsius. (Additional details of the collection process were reported previously 22.) Metal-Ion Analysis The concentrations of titanium, aluminum, cobalt, and chromium in the serum and the concentration of chromium in the urine were measured with graphite-furnace Zeeman atomic absorption spectrophotometry as described previously 3 ' All samples were tested in triplicate. The detection limits, in nanograms per milliliter (parts per billion), were 2.11 for the concentration of titanium in the serum, 0.35 for the concentration of aluminum in the serum, 0.30 for the concentration of cobalt in the serum, 0.03 for the concentration of chromium in the serum, and for the concentration of chromium in the urine. Statistical Analysis The data are reported as the mean for each group at each timeinterval. Concentrations below were approximated as one-half of to facilitate calculation of the means, which are used to provide a relative scale for the reader. Intergroup comparisons were made independent of the means with use of the Kruskal-Wallis non-parametric analysis of variance as the groups had left-censored data (values less than the limit of detection with graphite-furnace Zeeman atomic absorption spectrophotometry). The Wilcoxon-Mann-Whitney test was used if the Kruskal-Wallis test had revealed significant differences at the p < 0.05 level. Intragroup longitudinal comparisons were made with use of the Friedman test followed by the sign test when the former test indicated a level of significance of p < All values were ranked with use of these non-parametric tests. Values below were counted as ties and then ranked at the same level. Correlations were established with use of the Spearman rank-order correlation test. Results This was not a randomized series; there was a selection bias with regard to the type of implant as well as the method of fixation, both of which were chosen according to the preference and the practice pattern of the surgeon. At one center (Rush-Presbyterian-St. Luke's Medical Center), surgeons tended to insert the prosthesis without cement only in younger, higherdemand patients, whereas at the other center (Central DuPage Hospital), they tended to insert the device without cement in all primary total hip replacements. Therefore, although the groups were closely matched with regard to gender, there were significant differences in age among them. Specifically, the patients in Group 3 had a younger mean age than those in both Group 1 (sixty-six compared with fifty-four years; p < 0.003, Kruskal-Wallis test, and p < , Wilcoxon- Mann-Whitney test) and Group 2 (sixty-two compared THE JOURNAL OF BONE AND JOINT SURGERY

5 METAL RELEASE IN PATIENTS WHO HAVE HAD A PRIMARY TOTAL HIP ARTHROPLASTY 1451 TABLE III CONCENTRATIONS OF COBALT IN THE SERUM* Preop. or Baseline At 12 Mos. At 36 Mos. Group 1 (20 patients)t 0.16 ( ) ( ) ( ) 14 Group 2 (15 patients)t 0.25 ( ) ( ) ( ) 13 Group 3 (20 patients)t 0.18 ( ) ( ) ( ) 15 Group 4 (20 patients)t All concentrations undetectable 0.56 ( ) ( ) 19 The values for the mean and the range are given as nanograms per milliliter (parts per billion). Concentrations below of 0.30 nanogram per milliliter (part per billion) were assigned a value that was one-half of, according to convention. fthe patients in Group 1 had a so-called hybrid total hip replacement (a modular cobalt-alloy femoral stem and head inserted with cement and a titanium acetabular cup inserted without cement); those in Group 2, an extensively porous-coated modular cobalt-alloy stem with a cobaltalloy head and a titanium-alloy socket, inserted without cement; those in Group 3, a proximally porous-coated modular titanium-alloy stem with a cobalt-alloy head and a titanium socket, inserted without cement; and those in Group 4 (controls), no implant. with fifty-four years; p < 0.003, Kruskal-Wallis test, and p < 0.03, Wilcoxon-Mann-Whitney test). Furthermore, the patients in Group 1 had an older mean age than the controls in Group 4 (sixty-six compared with fiftynine years; p < 0.003, Kruskal-Wallis test, and p < 0.03, Wilcoxon-Mann-Whitney test). Concentrations of Titanium in the Serum (Table I and Fig. 1) Most of the concentrations of titanium in the serum of the control subjects (Group 4) were below the detection limit of 2.11 nanograms per milliliter (parts per billion) at all three testing intervals. However, in Groups 1 and 3, in which an implant had been inserted, the number of patients with a non-detectable level decreased at the twelve and thirty-six-month intervals compared with the levels measured preoperatively. This was particularly true in Group 3. Intragroup comparisons revealed that the concentrations were significantly increased (1.8-fold and 3.0- fold) at twelve and thirty-six months, compared with the preoperative levels, in the patients who had a titanium-alloy femoral component and a titanium acetabular component that had been inserted without cement (Group 3) (p < 0.01, Friedman and sign tests). There also was a 1.7-fold increase in Group 3 at thirtysix months compared with the level at twelve months (p < 0.005, Friedman test, and p < 0.05, sign test). With the numbers available, no significant differences between the preoperative and postoperative concentrations were detected in Groups 1, 2, or 4. Intergroup comparisons revealed that the concentration at thirty-six months was significantly increased (1.9-fold, 2.2-fold, and 3.4-fold) in Groups 1, 2, and 3, respectively, compared with the concentration in Group 4 (controls) (p < 0.001, Kruskal-Wallis test, and p < 0.005, Wilcoxon-Mann-Whitney test). Furthermore, the thirty-six-month level was 1.9-fold greater in Group 3 than in Group 1 (p < 0.001, Kruskal-Wallis test, and p < 0.03, Wilcoxon-Mann-Whitney test). Concentrations of Aluminum in the Serum (Table II) Very few of the levels of aluminum in the serum were below of 0.35 nanogram per milliliter (part per billion). Intragroup comparisons revealed no significant increases in any group. Similarly, intergroup comparisons showed no significant differences at either the twelve or the thirty-six-month testing interval. Concentrations of Cobalt in the Serum (Table III) Most patients had levels of cobalt in the serum that were below of 0.30 nanogram per milliliter (part per billion). Neither intragroup nor intergroup comparisons revealed any significant differences, with the numbers available. One patient (in Group 3) was excluded from the study because of late recurrent instability of the hip prosthesis. This patient had the highest level of cobalt VOL. 80-A, NO. 10, OCTOBER 1998

6 1452 J. J. JACOBS ET AL m Q Q. ^ 0.35 O Uj 0.15 g 0.10 O I 0.00 Group 1 rnmm Group 2 Eggga Group 3 I^H Group 4 D.L. 'Different than controls (p<0.03) FIG. 2 (p<0.02) MONTHS (p<0.03)- Bar graph showing the concentrations of chromium in the serum, in nanograms per milliliter (parts per billion), as a function of time in the four groups that were studied. At thirty-six months postoperatively, the level in Group 3 had increased 4.5-fold compared with the preoperative concentration. Intergroup comparisons revealed increases of 3.0-fold, 1.8-fold, and 5.4-fold in Groups 1, 2, and 3, respectively, compared with the level in the controls (Group 4). In addition, the level in Group 1 had increased 1.7-fold compared with that in Group 2. Group 1 had a so-called hybrid total hip replacement (a modular cobalt-alloy femoral stem and head inserted with cement and a titanium acetabular cup inserted without cement); Group 2 had an extensively porous-coated modular cobalt-alloy stem with a cobalt-alloy head and a titanium-alloy socket, inserted without cement; Group 3 had a proximally porous-coated modular titanium-alloy stem with a cobalt-alloy head and a titanium socket, inserted without cement; and Group 4 (controls) had no implant. D.L. = detection limit. 36 in the serum (13.1 nanograms per milliliter at twelve months; this decreased to 1.4 nanograms per milliliter at thirty-six months, which was among the highest levels at this follow-up interval), chromium in the serum (4.2 nanograms per milliliter at twelve months and 1.5 nanograms per milliliter at thirty-six months), and chromium in the urine (2.69 nanograms per milliliter at twelve months and 2.26 nanograms per milliliter at thirty-six months) of the patients who were studied. These concentrations were one to two orders of magnitude higher than those in the control subjects and than the preoperative concentrations. Concentrations of Chromium in the Serum (Table IV and Fig. 2) Most of the levels of chromium in the serum were higher than of 0.03 nanogram per milliliter (part per billion). Intragroup comparisons revealed a 4.5-fold increase in Group 3 at thirty-six months compared with the preoperative concentration (p < 0.02, Friedman and sign tests). No other increases were noted on intragroup comparisons. In contrast, intergroup comparisons demonstrated significant (3.0-fold, 1.8-fold, and 5.4-fold) increases at thirty-six months in Groups 1, 2, and 3, respectively, compared with the levels in the controls (p < 0.002, Kruskal-Wallis test, and p < 0.02, Wilcoxon-Mann- Whitney test). Furthermore, the patients who had had a so-called hybrid total hip replacement (Group 1) had a 1.7-fold increase at thirty-six months compared with those who had a cobalt-alloy stem that had been inserted without cement (Group 2) (p < 0.002, Kruskal- Wallis test, and p < 0.03, Wilcoxon-Mann-Whitney test). Concentrations of Chromium in the Urine (Table V and Fig. 3) Most of the concentrations of chromium in the urine were higher than of nanogram per milliliter (part per billion). Intragroup comparisons revealed no significant differences, with the numbers available. However, intergroup comparisons demonstrated a 3.0-fold increase at twelve months in Group 1 compared with the level in the controls (p < 0.05, Kruskal-Wallis test, and p < 0.005, Wilcoxon-Mann- Whitney test). Furthermore, at thirty-six months, there were increases of 3.2-fold, 2.8-fold, and 7.5-fold in Groups 1,2, and 3, respectively, compared with the level in the controls (p < 0.002, Kruskal-Wallis test, and p < 0.04, Wilcoxon-Mann-Whitney test). There was a strong correlation between the concentrations of chromium in the serum and those in the urine in the patients who had an implant, at both twelve months (r = 0.58, p < 0.001, Spearman rank-order correlation test) and thirtysix months (r = 0.74, p < 0.001, Spearman rank-order correlation test). Discussion To our knowledge, this is the largest prospective, controlled study of the concentrations of metaldegradation products from total joint-replacement components. There is a considerable body of literature concerning the levels of chromium and cobalt in the THE JOURNAL OF BONE AND JOINT SURGERY

7 METAL RELEASE IN PATIENTS WHO HAVE HAD A PRIMARY TOTAL HIP ARTHROPLASTY 1453 TABLE IV CONCENTRATIONS OF CHROMIUM IN THE SERUM* Preop. or Baseline At 12 Mos. At 36 Mos. Group 1 (20 patients)t 0.06 ( ) ( ) ( ) 1 Group 2 (15 patients)t 0.08 ( ) ( ) ( ) 2 Group 3 (20 patients)t 0.06 ( ) ( ) ( ) 5 Group 4 (20 patients)t 0.06 ( ) ( ) ( ) 11 *The values for the mean and the range are given as nanograms per milliliter (parts per billion). Concentrations below of 0.03 nanogram per milliliter (part per billion) were assigned a value that was one-half of, according to convention. tthe patients in Group 1 had a so-called hybrid total hip replacement (a modular cobalt-alloy femoral stem and head inserted with cement and a titanium acetabular cup inserted without cement); those in Group 2, an extensively porous-coated modular cobalt-alloy stem with a cobaltalloy head and a titanium-alloy socket, inserted without cement; those in Group 3, a proximally porous-coated modular titanium-alloy stem with a cobalt-alloy head and a titanium socket, inserted without cement; and those in Group 4 (controls), no implant. serum and urine after total joint replacement 35-8 ' but relatively few studies of the levels of titanium and aluminum"' ' 42. Many of these investigations have been hampered by technical limitations of the analytical instruments, inadequate precautions against contamination, suboptimum study design, or combinations of these factors. Furthermore, it is difficult to compare the results from different laboratories as different techniques and protocols are used. The current prospective, longitudinal study, which included contemporaneous controls who did not have an implant, circumvented many of the limitations just mentioned. First, more subtle differences between groups could be detected. Detection of differences between groups has been difficult in previous studies because of the variations among individuals, and among samples from the same individuals, with regard to the concentrations of metal in the serum and urine 35. The intergroup comparisons were particularly able to reveal differences as the sample was larger than that used in the intragroup comparisons. Second, implants with different compositions, designs, and methods of fixation were studied with use of an identical protocol, facilitating comparisons among groups and inferences about the mechanisms of metal release. The normal serum levels of the metals that were studied in the current series have been reported previously as one to ten nanograms per milliliter (parts per billion) for aluminum, 0.15 nanogram per milliliter for chromium, and 0.1 to 0.2 nanogram per milliliter for cobalt 49. We previously reported the normal level of titanium as being less than 4.1 nanograms per milliliter 22. In the current study, the concentrations of aluminum and cobalt in the serum of the control subjects (Group 4) and the preoperative concentrations in Groups 1, 2, and 3 were consistent with the normal values that have been reported previously. The normal levels of chromium and titanium in the present study were slightly lower than those that have been reported previously. This is probably indicative of improvements in the methods of graphite-furnace Zeeman atomic absorption spectrophotometry that have minimized contamination and interferences. The patients in the current study who had a well functioning titanium-alloy prosthesis that had been inserted without cement had a 3.4-fold increase in the concentration of titanium in the serum, compared with the control value, at thirty-six months after the operation. This finding is in contrast to that in our previous report, a retrospective study in which increases in the levels of titanium in the serum were detected only in patients who had a loose total hip prosthesis and were scheduled to have a revision 22. No such increases were seen in two groups of patients who had had a clinically successful primary total hip replacement. Dorr et al. also reported an increase in the level of titanium in the serum only in association with loose implants". This underscores the problem of detecting subtle differences in the concentrations of trace metals in the nanogramsper-milliliter (parts-per-billion) range. VOL. 80-A, NO. 10, OCTOBER 1998

8 1454 J. J. JACOBS ET AL. Group 1 (20 patients)t Group 2 (15 patients)t Group 3 (20 patients)t Group 4 (20 patients)t TABLE V CONCENTRATIONS OF CHROMIUM IN THE URINE* Preop. or Baseline 0.12 ( ) ( ) ( ) ( ) 8 At 12 Mos ( ) ( ) ( ) ( ) 6 At 36 Mos ( ) ( ) ( ) ( ) 12 *The values for the mean and the range are given as nanograms per milliliter (parts per billion). Concentrations below of nanogram per milliliter (part per billion) were assigned a value that was one-half of, according to convention. tthe patients in Group 1 had a so-called hybrid total hip replacement (a modular cobalt-alloy femoral stem and head inserted with cement and a titanium acetabular cup inserted without cement); those in Group 2, an extensively porous-coated modular cobalt-alloy stem with a cobaltalloy head and a titanium-alloy socket, inserted without cement; those in Group 3, a proximally porous-coated modular titanium-alloy stem with a cobalt-alloy head and a titanium socket, inserted without cement; and those in Group 4 (controls), no implant. In the present study, more modest but significant increases in the levels of titanium were seen in the patients who had a titanium or titanium-alloy acetabular component and a cobalt-alloy femoral stem (Groups 1 and 2), compared with the level in the controls, at thirtysix months after the operation. Together, these findings suggest that the acetabular and femoral components both are sources of systemically disseminated titaniumdegradation products. At thirty-six months postoperatively, the increases in the concentrations of titanium in the serum are probably accompanied by an accumulation in remote tissue deposits 46. In contrast to the levels of titanium, the levels of aluminum in the serum were not increased in the current study. This may be due to the facts that (1) there is a relatively small amount of aluminum (6 per cent) in the alloy; (2) in Group 3, only the screws and the substrate of the femoral stem were made from the alloy, whereas the shell and the porous coating were made from commercially pure titanium (similarly, in Group 1, only the screws and, in Group 2, only the shell, spikes, and screws were made from the alloy); (3) the degradation of titanium-alloy implants occurs primarily through abrasion of the oxide layer, which is relatively poor in aluminum; and (4) the pathways of aluminum transport and excretion may differ from those of titanium. Because the present study did not include fecal analysis, the possible contributory role of biliary excretion cannot be determined. The levels of cobalt in the serum were universally low in the current study; most of the values were below. However, in other populations of patients, in whom the generation of metal-degradation products would be expected to be greater than that in patients who have a well functioning primary total hip prosthesis, we observed increases in the levels of cobalt in the serum in association with both metal-on-metal devices 29 and failed implants demonstrating corrosion at the modular junctions 30. Increases in the concentrations of chromium in the serum and urine of patients who had a well functioning total hip prosthesis also were noted in the current study. Initially, we had hypothesized that patients who had an extensively porous-coated stem that had been inserted without cement would have higher concentrations of chromium because of the larger surface area available for passive dissolution. Thus, the finding that the patients in Group 3, in whom a titanium-alloy stem and a titanium socket had been inserted without cement, had the highest concentrations of chromium in the serum and urine was unexpected. The only source of chromium in these patients was the articulating surface or the bore of the modular cobalt-alloy head. On the basis of our previous work demonstrating a high prevalence of corrosion at head-neck interfaces 14 and increased levels of cobalt in the serum and of chromium in the urine that were significantly related to the magnitude of corrosion at the head-neck junction in patients who had had a revision total hip replacement (p < 0.05) 30, we surmise that the source of the chromium in the bloodstream THE JOURNAL OF BONE AND JOINT SURGERY

9 METAL RELEASE IN PATIENTS WHO HAVE HAD A PRIMARY TOTAL HIP ARTHROPLASTY 1455 and urine was primarily the bore of the femoral head. Wear and corrosion of the articular surface also may have contributed to some extent to the increases that we observed. Another unexpected finding in the current study was that patients who had a cobalt-alloy femoral stem that had been inserted with cement (Group 1) had significantly higher concentrations of chromium in the serum (p < 0.03) than did those who had an extensively porous-coated stem that had been inserted without cement (Group 2). This observation led us to perhaps the most salient finding of this study: that passive dissolution from the porous surfaces of well fixed implants was not a dominant mode of metal release. This finding further implicates the head-neck junction as the major source of chromium-degradation products as presumably the intraosseous portion of a cobalt-alloy stem that is well fixed with cement would not be a source of either fretting debris or passive dissolution. However, the possibility that some fretting occurs between the stem and the cement mantle, contributing to metal release, cannot be precluded. Thus, on the basis of the results of this study, it appears that fretting corrosion of the modular junction of the head and neck may be a dominant mode of chromium release in patients who have a well functioning primary total hip prosthesis. Although modular implants have advantages in terms of cost and versatility, there is a growing concern about degradation at modular connections. The observation that the highest levels of chromium were associated with mixed-metal head-neck junctions (a cobalt-alloy head and a titanium-alloy stem) may lead to the conclusion that galvanic effects are operant. However, there were other confounding differences between the groups that may be more determinative. In particular, the patients in Group 3, who had a mixedmetal femoral component, were younger and presumably more active than those in Groups 1 and 2, in whom the head-neck junction and the stem were made of similar metals. The activity level could be expected to affect the magnitude of fretting corrosion at the head-neck junction. The most likely explanation for the lower levels of chromium in Group 2 are the differences in the geometry of the taper of the femoral component and the methods of fabrication (including tolerances). The geometries of the tapers in Groups 1 and 3 were identical (the so-called 6-degree taper), whereas the tapers in Group 2 had a larger diameter (12/14 for the Prodigy components and 14/16 for the Solution and AML components). The taper geometry, surface finish, and fabrication methods are believed to play important roles in the frettingcorrosion process' 14 ' 54. There is currently a trend to avoid the manufacture of smaller-diameter tapers for the femoral component. Our previous work has shown that, even in the absence of increases in the concentrations of metal in the serum and urine, deposition and accumulation of metal can occur locally and in remote organ stores in association with a well functioning device 26. Concerns regarding the release and distribution of metal-degradation products are justified on the basis of previously reported potential toxicities of titanium, aluminum, cobalt, and chromium. Cobalt and chromium are known essential trace metals 49 ; depletion in the body consistently results in a physiological deficiency syn Q Q Q. ~ 0.35 Q 0.30 Group 1 Group 2 Group 3 Group 4 D.L. "Different than controls (p<0.03) FIG. 3 MONTHS Bar graph showing the concentrations of chromium in the urine, in nanograms per milliliter (parts per billion), as a function of time in the four groups that were studied. At twelve months postoperatively, the concentration in Group 1 had increased 3.0-fold compared with that in the controls (Group 4). At thirty-six months, there were increases of 3.2-fold, 2.8-fold, and 7.5-fold in Groups 1,2, and 3, respectively, compared with the level in Group 4. GroUp 1 had a so-called hybrid total hip replacement (a modular cobalt-alloy femoral stem and head inserted with cement and a titanium acetabular cup inserted without cement); Group 2 had an extensively porous-coated modular cobalt-alloy stem with a cobalt-alloy head and a titanium-alloy socket, inserted without cement; Group 3 had a proximally porous-coated modular titanium-alloy stem with a cobalt-alloy head and a titanium socket, inserted without cement; and Group 4 (controls) had no implant. D.L. = detection limit. 36 VOL. 80-A, NO. 10, OCTOBER 1998

10 1456 J. J. JACOBS ET AL. drome, and specific repletion reverses the abnormalities. Cobalt, a cofactor in vitamin B12, and chromium, an essential cofactor in the interaction between insulin and its receptor, are toxic in high enough concentrations in vivo 2. Chronic serum levels of aluminum of greater than 190 nanograms per milliliter (parts per billion) have been etiologically linked with a number of clinical disorders, including vitamin-d-resistant osteomalacia, hypochronic anemia, and global dementia 2 """. Hexavalent chromium, a possible implant-degradation product, was characterized as a class-i human carcinogen by the International Agency for Research on Cancer in Other metals used in orthopaedic devices, which have shown carcinogenicity in laboratory animals but have not been investigated by the International Agency for Research on Cancer, include aluminum, cobalt, and titanium Increased levels of cobalt can induce polycythemia and testicular toxicity and can interfere with DNA repair 253. Broad reviews of the toxicology of these elements are available ' 52. Toxicity may occur by means of metabolic alterations, alterations in the interactions between the host and a parasite, immunological interactions of metal moieties because of their ability to act as haptens (specific immunological activation) or antichemotactic agents (non-specific immunological suppression) 71534, or chemical carcinogenesis 204 "- 43. Despite the long-standing recognition of the toxicity of some of these elements in in vitro and animal models, no causal relationship has been definitively established in patients (either individually or in groups) who have had a total hip replacement. This may be due in large part to the difficulty of observation as most symptoms that are attributable to systemic and remote toxicity can be expected to occur with finite frequency in any population of orthopaedic patients. Thus, the identification of implant-referable disease processes at sites remote from the implant depends on the availability of epidemiological studies comparing large populations of patients who have and have not had a joint replacement or on the ability to perform tests on patients before and after removal of the device. There is currently a dearth of epidemiological follow-up data on potential remote and systemic effects related to joint-replacement procedures. We are conducting a parallel postmortem retrieval program in which many of the patients from Groups 1,2, and 3 in the current study are enrolled 26. The results of this postmortem program may shed some light on this situation over time. Studies of the bioavailability of metal species that have been released from prosthetic materials are urgently needed so that the clinical importance of increased concentrations of metal in the serum and urine can be determined. Central to these studies is the speciation of the metal moieties present in body fluids and tissue stores. Many of the metals used in implants have valence and ligand-dependent toxicities in mammalian systems' 3. It will be an enormous challenge to perform these studies, given the technical complexities of working with concentrations in the nanograms-per-milliliter (parts-per-billion) range. Current technological tools (graphite-furnace Zeeman atomic absorption spectrophotometry and inductively coupled plasma-mass spectrometry) can measure only the concentrations of the elements; they cannot provide any information on their chemical form or their biological activity. The literature delineating the chemical form of the degradation products of metal joint-replacement prostheses is currently quite limited. Specific toxicological experiments on relevant chemical species that have been identified in bioavailability studies can be conducted to delineate the specific toxicities of the degradation products in animal models and in cell and tissue cultures. This will ultimately allow the risks to patients who are to have a joint-replacement procedure to be determined. Methods for the comprehensive monitoring and assessment of concentrations of metal in the serum and urine after total joint replacement also may evolve sufficiently to play a diagnostic role. Our database currently consists of more than 400 measurements of levels of titanium in serum, some of which were obtained as long as eight years after the replacement. These data demonstrate that concentrations of titanium in serum that are greater than eight nanograms per milliliter (parts per billion) approximately two times the normal level are almost always associated with one of several clinical findings: (1) aseptic loosening of the implant 24, (2) the presence of multiple implants, (3) accelerated wear of the countersurface in the presence of a carbonfiber-reinforced polyethylene articulating surface 28, (4) renal insufficiency, or (5) accelerated wear due to delamination of the polyethylene from a metal-backed component, resulting in unintended metal-on-metal articulation 28. In one recent case, observation of an increased level of titanium in the serum was the first demonstrable indication of failure of a metal-backed patellar component in an otherwise asymptomatic patient (unpublished data). We also have noted very high concentrations of chromium in the serum and urine and of cobalt in the serum of a patient who had impending failure of a metal-on-metal surface replacement 29. In the present study, the highest levels of cobalt and chromium in the serum and of chromium in the urine were in a patient who had late recurrent instability of the hip prosthesis. We believe that the high levels of metal in this patient were indicators of increased corrosion at the head-neck junction; this belief is based on our previously published studies 1430 as well as on recent clinical observations of five such instances of late instability of a total hip replacement that was associated with an atypical intense periprosthetic tissue reaction (characterized by extensive necrosis) in the presence of severe corrosion at the head-neck junction 39. Findings such as these point to the potential diagnostic usefulness of determinations THE JOURNAL OF BONE AND JOINT SURGERY

11 METAL RELEASE IN PATIENTS WHO HAVE HAD A PRIMARY TOTAL HIP ARTHROPLASTY 1457 of the levels of metal in the serum in certain clinical in whom the device has been in place for two to seven settings. As our database continues to expand, the diag- years. However, as the numbers of patients with a nostic usefulness of these tests will be further delineated, device in place for ten to twenty-five years increase The benefits of total hip replacement are real and rapidly, researchers and clinicians should undertake indramatic. However, our clinical experience suggests vestigations to clarify the toxicological importance of that this benefit is not without risk. The worldwide clin- orthopaedic implant-degradation products, ical picture suggests that the toxicological risks, both j i r NOT* The authors express their appreciation to Aaron G. Rosenberg. M.D., Mitchell U. systemic and at remote sites, are low tor most patients sheinko P,M.D., and stcvengiteiis.m.d., for allowing us to include their patients in this study. References 1. Agins, H. J.; Alcock, N. W.; Bansal, M.; Salvati, E. A.; Wilson, P. D., Jr.; Pellicci, P. M.; and Bullough, P. G.: Metallic wear in failed titanium-alloy total hip replacements. A histological and quantitative analysis. 7. Bone and Joint Surg., 70-A: , March Angle, C. R.: Organ-specific therapeutic intervention. In Metal Toxicology, edited by R. A. Goyer, C. D. Klaassen, and M. P. Waalkes. San Diego, Academic Press, Bartolozzi, A., and Black, J.: Chromium concentrations in serum, blood clot and urine from patients following total hip arthroplasty. Biomaterials, 6: 2-8, Bhattacharyya, M. H.; Wilson, A. K.; Silbergeld, E. K.; Watson, L.; and Jeffery, E.: Metal-induced osteotoxicities. In Metal Toxicology, edited by R. A. Goyer, C. D. Klaassen, and M. P. Waalkes. San Diego, Academic Press, Black, J.; Maitin, E. C; Gelman, H.; and Morris, D. M.: Serum concentrations of chromium, cobalt and nickel after total hip replacement: a six month study. Biomaterials, 4: , Black, J.; Sherk, H.; Bonini, J.; Rostoker, W. R.; Schajowicz, F.; and Galante, J. O.: Metallosis associated with a stable titanium-alloy femoral component in total hip replacement. A case report../. Bone and Joint Surg., 72-A: , Jan Bravo, I.; Carvalho, G. S.; Barbosa, M. A.; and de Sousa, M.: Differential effects of eight metal ions on lymphocyte differentiation antigens in vitro./ Biomed. Mater. Res., 24: , Brien, W. W.; Salvati, E. A.; Betts, E; Bullough, P.; Wright, T.; Rimnac, C; Buly, R.; and Garvin, K.: Metal levels in cemented total hip arthroplasty. A comparison of well-fixed and loose implants. Clin. Orthop., 276: 66-74, Brown, S. A.; Flemming, C. A.; Kawalec, J. S.; Placko, H. E.; Vassaux, C; Merritt, K.; Payer, J. H.; and Kraay, M. J.: Fretting corrosion accelerates crevice corrosion of modular hip tapers. /. Appl. Biomater., 6:19-26, Collier, J. P.; Surprenant, V. A.; Jensen, R. E.; Mayor, M. B.; and Surprenant, H. P.: Corrosion between the components of modular femoral hip prostheses. J. Bone and Joint Surg, 74-B(4): , Dorr, L. D.; Bloebaum, R.; Emmanual, J.; and Meldrum, R.: Histologic, biochemical, and ion analysis of tissue and fluids retrieved during total hip arthroplasty. Clin. Orthop., 261: 82-95, Elinder, C. G., and Friberg, L.: Cobalt. In Handbook of the Toxicology of Metals, edited by L. Friberg, G. F. Nordberg, and V. B. Vouk. Ed. 2, vol. 2, pp Amsterdam, Elsevier, Friberg, L.; Nordberg, G. F.; and Vouk, V. B. [editors]: Handbook of the Toxicology of Metals. Volume 1: General Aspects. Amsterdam, Elsevier, Gilbert, J. L.; Buckley, C. A.; and Jacobs, J. J.: In vivo corrosion of modular hip prosthesis components in mixed and similar metal combinations. The effect of crevice, stress, motion, and alloy coupling./ Biomed. Mater. Res., 27: , Gillespie, W. J.; Frampton, C. M. A.; Henderson, R. J.; and Ryan, P. M.: The incidence of cancer following total hip replacement./ Bone and Joint Surg., 70-B(4): , Gitelman, H. J. [editor]: Aluminum and Health: A Critical Review. New York, Dekker, Harris, W. H.: Traumatic arthritis of the hip after dislocation and acetabular fractures: treatment by mold arthroplasty. An end-result study using a new method of result evaluation. J. Bone and Joint Surg., 51 -A: , June Harris, W. H.: The problem is osteolysis. Clin. Orthop., 311: 46-53, Jacobs, J. J.; Gilbert, J. L.; and Urban, R. M.: Corrosion of metallic implants. In Advances in Operative Orthopedics, edited by R. N. Stauffer. Vol. 2, pp St. Louis, C. V. Mosby, Jacobs, J. J.; Rosenbaum, D. H.; Hay, R. M.; Gitelis, S.; and Black, J.: Early sarcomatous degeneration near a cementless hip replacement. A case report and review. / Bone and Joint Surg., 74-B(5): , Jacobs, J. J.; Shanbhag, A.; Giant, T. T.; Black, J.; and Galante, J. O.: Wear debris in total joint replacements. J. Am. Acad. Orthop. Surg., 2: , Jacobs, J. J.; Skipor, A. K.; Black, J.; Urban, R. M.; and Galante, J. O.: Release and excretion of metal in patients who have a total hip-replacement component made of titanium-base alloy. / Bone and Joint Surg., 73-A: , Dec Jacobs, J. J.; Urban, R. M.; Schajowicz, E; Gavrilovic, J.; and Galante, J. O.: Particulate associated endosteal osteolysis in titanium-base alloy cementless total hip replacement. In Particulate Debris from Medical Implants: Mechanisms of Formation and Biological Consequences, ASTM STP1I44, pp Edited by K. R. St. John. Philadelphia, American Society for Testing and Materials, Jacobs, J. J.; Skipor, A. K.; Black, J.; Manion, L.; Urban, R. M.; and Galante, J. O.: Metal release in patients with loose titanium alloy total hip replacements. In Transactions of the Fourth World Biomaterials Congress, p Berlin, European Society for Biomaterials, Jacobs, J. J.; Skipor, A. K.; Black, J.; Patterson, L. M.; Paprosky, W. G.; and Galante, J. O.: A 3-year prospective study of serum titanium levels in patients with primary total hip-replacements. In Medical Applications of Titanium and Its Alloys: The Material and Biological Issues, ASTM STP 1272, pp Edited by S. A. Brown and J. E. Lemons. West Conshohocken, Pennsylvania, American Society for Testing and Materials, Jacobs, J. J.; Skipor, A. K.; Urban, R. M.; Black, J.; Manion, L. M.; and Galante, J. O.: Transport of metal degradation products of titanium alloy total hip replacements to reticuloendothelial organs. An autopsy study. Trans. Soc. Biomater., 17: , Jacobs, J. J.; Urban, R. M.; Wall, J.; Black, J.; Reid, J. D.; and Veneman, L.: Unusual foreign-body reaction to a failed total knee replacement: simulation of a sarcoma clinically and a sarcoid histologically. A case report./ Bone and Joint Surg., 77-A: , March VOL. 80-A, NO. 10, OCTOBER 1998