http://www.aboutjoints.com
Urinary Excretion Levels of Metal Ions in
Patients Undergoing Total Hip Replacement with a
Porous-Coated Prosthesis: Preliminary Results

In: Quantitative Characterization and Performance of Porous Implants for Hard Tissue Applications, ASTM STP 953

by Lynne C. Jones,1 David S. Hungerford,2 Robert V. Kenna,3
Guy Braem,4 and Virginia Grant5

1Research associate, Orthopaedic Surgery, Johns Hopkins University School of Medicine, Baltimore, MD 21239.

2Professor, Orthopaedic Surgery, Johns Hopkins University School of Medicine, and
Chief, Division of Arthritis Surgery, Good Samaritan Hospital, Baltimore, MD 21239.

3Research associate, Orthopaedic Surgery, Good Samaritan Hospital, Baltimore, MD 21239.

4Research technician, Johns Hopkins University School of Medicine, Baltimore, MD 21239.

5Staff chemist, Chesapeake Bay Institute of the Johns Hopkins University School of Medicine, Baltimore, MD 21239.

REFERENCE: Jones, L. C.. Hungerford, D. S., Kenna, R. V., Braem, G., and Grant, V., "Urinary Excretion Levels of Metal Ions in Patients Undergoing Total Hip Replacement with a Porous-Coated Prosthesis: Preliminary Results," Quantitative Characterization and Performance of Porous Implants for Hard Tissue Applications, ASTM STP 953, J. E. Lemons, Ed., American Society for Testing and Materials, Philadelphia, 1987, pp. 151—162.

ABSTRACT: Porous-coated prostheses implanted without bone cement are currently being evaluated for use in patients undergoing total joint replacement (TJR). One parameter under study is the potential release of metal ions from these prostheses. In order to determine if there is a systemic increase in cobalt, chromium, or nickel levels within the body subsequent to total joint replacement with a porous-coated prosthesis, 24-h urine specimens were collected from patients prior to and subsequent to TJR with a PCA total hip prosthesis. Metal ion analysis was achieved using flameless atomic absorption spectroscopy. Increases in urinary cobalt and nickel excretion were detected in several patients at six months and in most patients at one year after surgery. However, these differences were not statistically significant. No differences between the preoperative and postoperative time periods (one week, six months, and twelve months) were detected for urinary levels of chromium. Although the metal ion levels for all of the patients studied appear to be in the range handled by the body’s systemic compensatory mechanisms, which adjust levels of trace elements, continued follow-up is needed to determine the patterns and the long-term significance of metal ion release.

KEY WORDS: porous implants, biocompatibility, corrosion, total hip replacement, porous-coated prostheses, flameless atomic absorption, metal ion release

The issue of biocompatibility of orthopedic implants, although not new, has come recently to the fore in orthopedic surgery with the introduction of porous-coated prostheses, which can be used in the cementless application of total joint arthroplasty. Although all metallic implants are subjected to corrosive forces within the body, these forces may have a larger impact on porous-surfaced prostheses. Porous-coated implants have an increased total surface area, as well as intimate contact with bone. It has been suggested that these two factors may contribute to an accumulation of corrosion products retained in the body, which, if given sufficient quantities and time, may lead to potential immunological, toxicological, or carcinogenic effects [1—6]. Due to the limited clinical experience with porous-coated prostheses, there is a paucity of information available pertaining to metal ion release from these implants in patients.

The principal objective of our investigation was to determine if there were increased levels of the metal ions which constitute the metallic components retained by patients undergoing total hip arthroplasty. Analysis of metal ion levels in urine samples has been shown to give an accurate assessment of the total body load [7—13J. A long-term study was instituted in which 24-h urine samples from patients undergoing total hip replacement (THR) with a PCA total hip prosthesis were collected at various time intervals prior to and subsequent to surgery and were analyzed for changes in total metal ion content. Levels of the metals that comprise this prosthesis—cobalt, chromium, and nickel—were determined using flameless atomic absorption spectroscopy. The following is a report of our preliminary findings for up to twelve months after surgery.

Materials and Methods

The study population included 30 patients who were scheduled for cementless total hip replacement with a PCA total hip system (Howmedica, Inc., Rutherford, NJ). Participation in this clinical study, approved by the Institution Review Board, was on a volunteer basis; consent was obtained only after the study had been fully explained to each individual. At the time of this writing, information was available for various time periods up to one year for 17 patients. The following descriptive information was collected from each patient: occupation, location of residence, alcohol consumption, and level of tobacco smoking. All patients underwent cementless implantation of a PCA total hip prosthesis, an implant with an ultrahigh-molecular-weight polyethylene (UHMWPE) surface articulating with a cobalt-chromium alloy surface.

Twenty-four hour urine samples were collected at the following time periods: preoperatively, and postoperatively at one week, six months, and twelve months. [Urine specimens are continuing to be collected for each of these patients on an annual basis.] Urine collection was achieved using sterile Medi-Flex specimen containers (Tn-State Hospital Supply Corp., Howell, MI) and Medi 24-h urine containers (Medi, Inc., Holbrook, MA). The patients were asked not to ingest any vitamins (particularly B-complex vitamins) 48 h prior to collection of urine. They were also asked not to contaminate the collection vessels by rinsing or wiping the surfaces. At the time of receipt, the total 24-h urine output was measured and recorded. The samples were then agitated, and two 50-mL aliquots were drawn into sterile Corning centrifuge tubes (Corning, Corning, NY). These tubes were stored in a Revco ultralow-temperature freezer (West Columbia, SC) at -70°C until the time of analysis. An additional 10-mL sample was taken and used for analysis of specific gravity.

Calibration standards were evaluated in order to determine the accuracy of the methodology for each of the ions under analysis. Each of the standards was prepared from reference solutions purchased from the Fisher Scientific Co. (Fairlawn, NJ). Solutions ranging from 5 to 50 parts per billion (ppb) were analyzed. The diluent used was ultrapure deionized water (Millipore Milli-Q System, Bedford, MA). In addition, Standard Reference Material (SRM) No. 2670 (National Bureau of Standards, Washington, DC) was analyzed for normal and elevated chromium and nickel levels in urine. Both calibration evaluations yielded reproducible, valid results.

At the time of analysis, all the urine samples were removed from the freezer and placed in a warm-water bath until thawed. The tubes were then centrifuged at 200 rpm for 10 mm. If a precipitate had formed, determinations of the volume of the supernatant and the recipitate were made. A concentration of approximately 60-pt of ultrapure nitric acid (Ultrex grade, J. T. Baker Chemical Co., Philipsburg, NJ) for each 10 mL of supernatant was added to the pellet only. The dissolved pellet was then returned to the supernatant.

Analytical determinations of the cobalt, chromium, and nickel content of the urine samples were made using the "method of additions" [14]. This involves analysis of three aliquots per specimen under study: these aliquots were (1) undiluted, (2) spiked with 2 ppb cobalt and 2 ppb chromium, and (3) spiked with 10 ppb nickel. All the specimens were analyzed using a Perkin-Elmer Model 4000 atomic absorption spectrometer with a Model HGA 500 lectrothermal source (Perkin-Elmer, Norwalk, CT). This technique included the use of pyrolyzed graphite furnace tubes and deuterium background correction. The analytic program was optimized for each phase (drying, ashing, and atomization) and for each metal wavelengths and slit widths) [14,15].

Care was taken to minimize all possible sources of contamination. This included subjecting all containers, laboratory glassware, and equipment to acid washing with 10% nitric acid. In addition, all triplicate values were reviewed to detect possible decreases in analytic efficiency, as well as periodic sampling error or contamination. If the three values were inconsistent, the specimen was reanalyzed.

Statistical analysis of the results was achieved using analysis of variance techniques for elated measurements, and the comparisons were made using linear contrast methods. All methods of analysis stressed the value of comparing the variability of the data for an individual patient, as well as the variability between patients.

Results

At this time, we are reporting only the preliminary findings obtained. As this is a very early stage of a long-term investigation, the information available involves only a fraction of the 30 patients to be included (17 evaluated preoperatively, 14 at six months postoperatively and 4 at one year postoperatively). For the 17 patients included in the preoperative evaluation, the diagnosis at presentation was primarily osteoarthritis (n = 11), although a few patients had diagnoses of avascular necrosis of bone (n = 3) and rheumatoid arthritis (n = 3). This group included 11 males and 6 females. The age of the patients ranged from 14 to 77 years with a mean of 58 years. Most of the patients resided in Maryland (n = 13), primarily in Baltimore and its suburbs (n = 7); the others lived in New York (n = 2), Pennsylvania (n = 1), and Illinois (n = 1). The majority of the patients were not smokers or excessive alcohol consumers. All but 1 patient had jobs in which occupational exposure is not known to occur; 1 patient was a machinist.

All values for urinary metal ion excretion are reported as the mean plus or minus the standard error. All the statistical analyses weighted the comparison of each individual to himself. As no significant differences were seen in the specific gravities of the samples for each individual patient, no conversion of the results from micrograms per litre was attempted.

Preoperative and one-week-postoperative specimens were collected and analyzed for 17 of the study patients (Fig. 1). For these patients, the average urinary levels of the preoperative samples were 0.95 ± 0.28 µg/L for cobalt, 1.41 ± 0.24 µg/L for chromium, and 4.31 ± 1.14 µg/L for nickel. These averages and variances are similar to those reported by previous investigators [10—13,16,17]. Postoperatively (at one week), values of 1.08 ± 0.28 µg/L for cobalt, 2.85 ± 1.34 µg/L for chromium, and 11.8 ± 6.74 µg/L for nickel were determined.

FIG. 1—Mean control values for urinary metal ion levels for 17 patients scheduled to undergo total hip replacement. The bar lines indicate the 95% confidence limits.

Although increases were seen for each metal ion, these increases were not statistically significant. The urinary chromium and nickel levels for one patient were significantly elevated one week postoperatively (17.3 and 85.2 µg/L, respectively). As these values are considerably outside the range seen for the other patients, it was suggested that these results may have been due to contamination of the specimen. As is shown in Table 1, this patient had normal levels at the six- and twelve-month time periods. The average values, if this patient is excluded, are 1.53 ± 0.27 µg/L for chromium and 5.08 ± 1.00 µg/L for nickel. Three patients underwent subsequent THR for the contralateral hip and those results are not included in the six- and twelve-month results reported herein.

TABLE 1 - Urinary metal ion rsults determined for four patients followed up to one year subsequent to total hip replacement.a
Postoperative
Patient No.
Metal Ion
Preoperative
1 Week
6 Months
1 Year
1
 cobalt chromium nickel

0.52
1.9
0

1.40
17.3b
85.2b
0.26
1.5
7.4
5.13
0.56
17.33
2
 cobalt chromium nickel
0.49
0.8
5.2
1.01
3.4
0
1.56
4.7
29.2
3.69
1.23
10.6
3
 cobalt chromium nickel
0.15
0.7
0
2.89
1.4
12.1
0.56
1.1
11.1
1.57
2.35
17.78
4
 cobalt chromium nickel
1.54
3.7
14.29
0.92
0.56
6.94
2.32
1.18
12.42
1.20
0.65
8.6
Mean + standard error cobalt chromium nickel
0.68 + 0.30
1.78 + 0.70
4.87 + 2.44
1.56 + 0.46
5.67 + 3.92
26.06 + 19.87
1.18 + 0.47
2.12 + 0.86
15.03 + 4.84
2.90 + 0.92
1.20 + 0.41
13.58 + 2.33
aThe values are in micrograms per litre.
bPossible contamination is not ruled out. The mean nickel value for one week postoperatively is 6.35 + 3.04 µg/L when this value is deleted.
The mean chromium value is 1.79 + 0.84 µg/L when excluding this value.

Figure 2 graphically illustrates the results for these patients evaluated at six months (n = 14). At six months, a slight increase was detected for cobalt (1.57 ± 0.50 µg/L) and chromium (1.52 ± 0.27 µg/L), while a larger average increase was found for nickel (11.43 ± 2.92 µg/L). This elevation in nickel ion level is primarily a reflection of the elevated values found for two patients (29.2 and 41.0 µg/L). One of the patients returned to normal levels (10.6 µg/L) by one year. The one-year follow-up for the other patient is not available at this time. However, using statistical methods that allow comparison of the variance within one patient to that of all patients (that is, ANOVA for related measurements), it was determined that these differences were not statistically significant. No relationships between the elevated values and age, occupation, alcohol consumption, or cigarette smoking were detected.

At this time, results are available for only 4 patients for the one-year follow-up period (Table 1). This group included 2 males and 2 females with an average age of 68 years (the range was 61 to 74). None of the patients was a smoker and their alcohol consumption ranged from abstinence to one drink per day. No change in the level of urinary chromium (1.20 ± 0.41 µg/L) was detected. Cobalt was observed to increase in 3 of the 4 patients, with an average value of 2.90 ± 0.92 µg/L (Fig. 3). With respect to nickel, an average value of 13.58 ± 2.33 µg/L was seen, also reflecting an increase in 3 of the 4 patients. Although these data reflect a threefold increase for cobalt and nickel in comparison with the preoperative values, this increase was not found to be statistically significant.

Fig. 2 - Six-month follow-up results for 14 THR patients, demonstrating no increase in urinary cobalt or chromium content, in comparison with preoperative levels, and an elevation in nickel content due, in part, to large increases observed in 2 patients. Excluding these 2 patients, the average urinary nickel level was 7.49 + 1.16 µg/L at six months. The bar lines encompass the 95% confidence limits.

Fig. 3 - One-year results for 5 THR patients, showing increases in urinary cobalt and nickel, in comparison with preoperative levels, for 3 of the 4. No changes were seen in the urinary chromium levels over time. The 95% confidence limits are shown by the bar lines.

Discussion

Although a variety of metals have been inserted into patients since the late 1700s, there remains a paucity of information available on the biological effects of these implants on the human body. Evaluation of retrieved specimens, as well as experimental findings, indicates that metallic implants exhibit corrosive behavior [18—21]. With the advent of metal-on-metal prostheses, accumulation of wear debris surrounding the joint has also been observed [22,23]. The development of ultrahigh-molecular-weight polyethylene (UHMWPE) for plastic-to-metal articulating surfaces has reduced the quantities of metallic products observed in the evaluated joints [24]. However, an unacceptable rate of loosening of the components of the prosthesis in patients with total knee and total hip arthroplasty has been seen with the use of methacrylate cement for implant fixation. It is for this reason that porous-coated prostheses, which allow biologic ingrowth for implant fixation, were developed. Renewed concerns about the toxicity and carcinogenicity of implant material have been primarily related to the increased surface area of the prosthesis, the intimate contact between the bone and the metal, and the technology involved in the sintering of the beads. It is important to relate these concerns to the clinical situation.

The issue of biocompatibility of orthopedic implants involves the study of potential local and systemic effects of the metal ions released from the surfaces of each prosthetic component. In the laboratory, corrosion products of Co-Cr alloy prostheses have demonstrated cytotoxic or carcinogenic potential [1—5]. Concerns pertaining to cytotoxicity are usually in reference to the local reaction of the biologic tissue in juxtaposition with the implant material. Not as well studied are the possible toxic effects on the metabolic functions of other tissues in the body. Carcinogenesis may be either a local or a systemic effect. The evaluation of the possible local carcinogenic effect is dependent on the collection of retrieved specimens from deceased patients. Although this project has been initiated at our institute, the results are not available for report. In order to assess, indirectly, the level of corrosion and, directly, the level of systemic response to potential corrosion products of orthopedic implants, serum and urine samples have been monitored by previous investigators [14,25—28]. Both methods of biological monitoring have been validated for use as indicators of exposure to cobalt, chromium, and nickel 17—13]. Collection and analysis of urine samples offer the advantages of ease of collection, minimum discomfort to the patient, and a minimum of matrix complications. For these reasons, we enlisted patient volunteers for the evaluation of 24-h urine specimens subsequent to THR with a PCA total hip system. There is an increased risk of specimen contamination during collection of urine samples; however, care was taken to minimize possible contamination by supplying uncontaminated containers to each patient and by increasing the awareness of each patient of possible sources of contamination.

Animal studies have indicated that there is an increase in urinary metal ion levels subsequent to implantation of a metal substrate. Using Sprague-Dawley rats, Wapner et al. evaluated urine chromium levels in animals implanted with Co-Cr microspheres [ASTM Specification for Cast Cobalt-Chromium-Molybdenum Alloy for Surgical Implant Applications (F 75-82)] of several surface areas [29]. They saw a significant increase in urine chromium only in the animals implanted with spheres having 300 times the basic ratio of surface area to body weight. This increase was most dramatic at 10 days, returning to slightly elevated levels at 100 days. Woodman et al. found a significant elevation of urinary nickel levels in rabbits implanted with porous cast Co-Cr alloy at six months (63.01 ± 1.94 ng/ mL) in comparison with solid implants (39.23 ± 1.23) and controls (28.51 ± 1.52) [30]. However, limitations in the experimental design and in animal models make it difficult to assess the clinical relevancy of these experiments. Metallic microspheres implanted loosely in bone are not an adequate representation of the clinical situation. Animals, which have diets, metabolisms, and kinematics that are different from those of humans may not sufficiently model the human situation. As porous-coated prostheses are being implanted in humans, biological monitoring of humans appears to be the best alternative.

Only a few laboratories have attempted to analyze the concentrations of metal ions in urine in patients subsequent to arthroplasty in which metallic components have been used. Coleman et al. evaluated the metal concentrations in the urine, blood, and hair of patients with hip arthroplasties [25]. They found that, preoperatively, the average value for urinary cobalt excretion was 0.5 µg/L, with 0.4 µg/L for urinary chromium excretion. In patients with metal-to-metal THR prostheses, Coleman et al. detected increases in urinary cobalt and chromium levels (24.0 and 6.2 µg/L, respectively), with a very high increase in cobalt (73.0 µg/L) and chromium (26.0 µg/L) in one patient. This is in contrast to their findings for patients with metal-to-plastic THR prostheses (0.7 µg/L of cobalt, 1.2 µg/L of chromium). Jones et al. found an increase in urinary levels ranging from 20 to 55 µg/L in hypersensitive individuals with loosened metal-to-metal THR prostheses [31]. In a study that evaluated the urinary content of cobalt in patients with cementless porous and nonporous Austin-Moore prostheses, Jorgenson et at. found no significant difference between the two groups [26]. However, the level in both groups appeared to be slightly elevated when compared with those values accepted as the normal range [25]. In a retrospective study of patients undergoing total knee replacement, we found no significant difference in urinary cobalt or chromium in patients with cementless as opposed to cemented PCA total knee prostheses in comparison with controls [32]. The results of this study lend support to the possibility of increases in urinary cobalt excretion in some patients by six and twelve months. We were unable to detect any difference in urinary chromium excretion between our preoperative and postoperative specimens, regardless of the time that had elapsed. However, none of our patient results for urinary cobalt or chromium were in the range of the elevations detected by Coleman et al. or Jones et al. for metal-to-metal prostheses [25,31].

Urinary nickel levels for patients after THR have not been documented by previous investigators. However, our six-month and one-year averages for urinary nickel are within the range established by Adams et al. for normal values in a comparison of the results of seven laboratories (10.5 ± 5.1 µg/L) [16]. Despite this, urinary nickel levels at six months for two patients were in the range seen in some individuals exposed occupationally (5.0 to 36.0 µg/L)—a high-risk group for carcinogenesis [16]. However, in one of the two patients this level returned to the preoperative level by one year. The one-year result for the other patient is not available at this time.

The difficulty of documenting a statistically significant increase in urinary cobalt and nickel levels may be due to the relatively small number of patients included. It is also a reflection, however, of the variability of response from individual to individual. The finding of elevations in 3 of the 4 patients studied at one year lends support to the importance of continuing investigation.

The biologic effect of a systemic increase in cobalt and nickel at the levels we have seen is unknown. Cobalt supplementation in patients at levels of 20 to 50 mg/day may result in polycythemia, transient hyperglycemia, and hyperplasia of bone marrow [33]. Increases of nickel in the diet have lead to dermatitis and hypersensitization [34]. Allergy to nickel has been documented by Deutman et al. to occur in 5.8% of patients prior to THR [35]. In addition, these authors also suggest that THR with a nickel-containing prosthesis may trigger hypersensitivity reactions in some patients. There is a lack of information on abnormal accumulation of cobalt, chromium, and nickel in specific tissues throughout the body. Accumulations of metal ions in the soft tissue and bone surrounding metal-on-metal prostheses have been reported [22,23]. No information is available clinically as to the level of exposure of the bone cells in juxtaposition with porous implants. However, Woodman et al. were unable to detect increases in metal ion levels in cortical bone surrounding solid implants of several alloy types in an animal model [36]. It is clear that analysis of various tissues at autopsy is warranted in order to determine the full extent of the consequences of metal ion release.

The PCA total hip prosthesis is made of 63.0% cobalt, 28.0% chromium, and 0.7% nickel. The increase of nickel in some patients despite the low amounts present in the composition of the prosthesis, is not easily explained. The differences in the changes of excretion of these metal ions may be due to differences in (1) corrosion processes, (2) solubility coefficients, (3) binding to proteins, or (4) excretion mechanisms. Additional in vitro and in vivo research is necessary to elucidate this enigma.

Our results suggest that there is an increase in metal ions released from porous-coated total hip implants. However, the location and mechanism of release is not known. In order to distinguish whether the increase is a consequence of wear of the articulating surfaces or a result of increased corrosion of the porous surfaces, a comparison of our results with those obtained from cemented porous-coated prostheses, where the implant design is the same but the porous surface is not exposed to bone, is necessary.

The response of the body to porous-coated cobalt-chromium prostheses needs to be defined. Because of the limited number of clinical trials, little is known about the variables affecting biologic ingrowth and possible metal ion release from these prostheses. The rate of metal ion release is not known, which thereby clouds the issue of whether the patient will be subjected to a level of metal ion concentration that could lead to potential carcinogenic or toxicologic effects. Two approaches to this problem include the monitoring of patients for tumor formation and the monitoring of urinary metal ion levels. We found no statistically significant increases in urinary metal ion levels in any of the patient groups studied. Although increases in metal ion levels have been detected in some of the patients with cementless porous-coated total hip prostheses, the levels seen are not sufficient to cause immediate alarm. Recent reports of tumor formation juxtaposed to Co-Cr implants suggest that these metals may create a carcinogenic environment in some patients [37,38]. The prostheses of the two patients cited were both of the metal-to-metal articulating type. This leads to the generation of extremely fine particulate debris, itself a chronic irritant, as well as exposure to a surface area of metal many times that of the rigid implant. The reporting of two cases that indirectly point to carcinogenesis out of hundreds of thousands of cases suggests that this potential exists for only a very small minority. Because of the seriousness of this issue, however, it is important that these studies be continued.

Conclusions

We have reported our preliminary findings in a long-term clinical study of the urinary excretion of metal ions subsequent to THR with a cementless porous-coated prosthesis. Based on the study results available to date, there were no statistically significant increases in urinary cobalt, chromium, or nickel levels postoperatively (1, 26, and 52 weeks) in comparison with preoperative levels. However, increases were detected in several patients by six months and in 3 of the 4 patients studied at one year for urinary cobalt and nickel. This suggests that corrosion of the implant may be occurring and that it can be detected in some patients one year postoperatively. We would like to emphasize that this is a preliminary report of our results. Only with longer term follow-up and an increased number of patients recruited will these results be validated and the implications of these findings determined.

Acknowledgments

We wish to express our appreciation to A. Hester and K. Connor for their technical assistance. We also wish to thank J. M. Frazier for his guidance and the use of his laboratory during the study.

  1. Heath, J. C., "The Production of Malignant Tumors by Cobalt in the Rat," British Journal of Cancer, Vol. 10, 1956, pp. 668—673.
  2. Hueper, W. C., "Experimental Studies in Metal Carcinogenesis: I. Nickel Cancers in Rats," Texas Report of Biological Medicine, Vol. 10, 1952, pp. 167—186.
  3. Hueper, W. C., "Experimental Studies in Metal Carcinogenesis: VII. Tissue Reactions to Parenterally Introduced Powdered Metallic Chromium and Chromite Ore," Journal of the National Cancer Institute, Vol. 16, 1955, pp. 447—470.
  4. Memoli, V. A., Woodman, J. L., Urban, R. M., and Galante, J. 0.. "Malignant Neoplasms Associated with Orthopaedic Implant Materials," Transactions of the Orihopaedic Research Society, Vol. 7. 1982, P. 164.
  5. Rae, T., Journal of Bone and Joint Surgery, Vol. 57B, 1975, pp. 444—450.
  6. Evans, E. M., Freeman, M. A. R., and Vernon-Roberts, V., Journal of Bone and Joint Surgery, Vol. 56B, 1974, pp. 626—642.
  7. McNeely, M. D., Nechay, M. W., and Sunderman, F. W., Jr., Clinical Chemistry, Vol. 18, 1972, pp. 992—995.
  8. Schroeder, H. A., Nason, A. P., and Tipton, 1. H., "Essential Trace Metals in Man: Cobalt," Journal of Chronic Diseases, Vol. 20, 1967, pp. 869—890.
  9. Mitman, F. W., Wolf, W. R., Kelsay, J. L., and Prather, E. S., "Urinary Chromium Levels of Nine Young Women Eating Freely Chosen Diets," Journal of Nutrition, Vol. 105, 1975, pp. 64—68.
  10. Hambridge, K. M., American Journal of Clinical Nutrition, Vol. 27, 1974, pp. 505—514.
  11. Sjogren, B., Hedstrom, L., and Ulfvarson, U., International Archives of Occupational and Environmental Health, Vol. 51, 1983, pp. 347—354.
  12. Pednx, A., Pellet, F., Vincent, M., Dc Gaudemaris, R., and Mallion, J. M., Toxicological European Research, Vol. 5, 1983, pp. 233—240.
  13. Aitio, A., IARC Science Publications, Vol. 53, 1984, pp. 497—505.
  14. Black, J., Maitin, E. C., Getman, H., and Morris, D. M., Biomaterials, Vol. 4, 1983, pp. 160—164.
  15. Guthrie, B. E., Wolf, W. R., and Veillon, C., Analytical Chemistry, Vol. 50, 1978, pp. 1900—1902.
  16. Adams, D. B., Brown, S. S., Sunderman, F. W., Jr., and Zachariasen, H., Clinical Chemistry, Vol. 24, 1978, pp. 862—867.
  17. Underwood, E. J., "Nickel," Trace Elements in Human and Animal Nutrition, 4th ed., Academic Press, New York, 1977, Chapter 6, pp. 159—169.
  18. Cahoon, J. R., Journal of Biomedical Materials Research, Vol. 7, 1973, pp. 375—383.
  19. Rose, R. M., Schiller, A. L., and Radin, E. L., Journal of Bone and Joint Surgery, Vol. 54A, 1972, pp. 854—862.
  20. Cohen, J., Journal of Bone and Joint Surgery, Vol. 44A, 1962, pp. 307—316.
  21. Oron, U. and Alter, A., Clinical Orthopaedics and Related Research, Vol. 185, 1984, pp. 295—300.
  22. Winter, C. D., "Tissue Reactions to Metallic Wear and Corrosion Products in Human Patients," Prosthesis and Tissue: The Interface Problem, Wiley, New York, 1974, pp. 11—26.
  23. Michel, R., Hofmann, J., Loer, F., and Zilkens, J., "Trace Element Burdening of Human Tissues Due to the Corrosion of Hip-Joint Prostheses Made of Cobalt-Chromium Alloys," Archives of Orthopaedic and Traumatic Surgery, Vol. 103, 1984, pp. 85—95.
  24. Swanson, S. A. V., Freeman, M. A. R., and Heath, J. C., Journal of Bone and Joint Surgery, Vol. 55B, 1973, pp. 759—773.
  25. Coleman, R. F., Herrington, T., and Scales, J. T., "Concentration of Wear Products in Hair, Blood, and Urine After Total Hip Replacement," British Medical Journal, Vol. 1, 1973, pp. 527—529.
  26. Jorgensen, T. J., Munno, F., Mitchell, T. 0., and Hungerford, D., Clinical Orthopaedics and Related Research, Vol. 176, 1983, pp. 124—126.
  27. Pazzaglia, V. E., Minoia, C., Ceciliani, L., and Riccardi, C., Acta Orthopaedica Scandinavica, Vol. 54, 1983, pp. 574—579.
  28. Linden, J. V., Hopfer, S. M., Grossling, H. R., and Sunderman, F. W., Jr., Annals of Clinical and Laboratory Science, Vol. 15, 1985, pp. 459—464.
  29. Wapner, K. L., Black, J., and Morris, D., "Chromium Release by Cast Co-Cr Alloy in Vivo: Ionic Valence and its Implications for Morbidity," Transactions of the Orthopaedic Research Society, Vol. 8, 1983, p. 240.
  30. Woodman, J. L., Urban, R. M., Lim, K., and Galante, J. 0., "Cobalt, Chromium and Nickel Release from Porous Coated Cast Cobalt Chromium Alloy," Transactions of the Orthopaedic Research Society, Vol. 9, 1984, p. 150.
  31. Jones, D. A., Lucas, H. K., O’Driscoll, M., Price, C. H. G., and Wibberley, B.,Journal of Bone and Joint Surgery, Vol. SiB, 1975, pp. 289—296.
  32. Jones, L. C., "Biocompatibility of Implant Materials" Total Hip Arthroplasty: A New Approach, D. S. Hungerford, A. Hedley, E. Habermann, L. Borden, and R. V. Kenna, Eds., University Park Press, Baltimore, 1984, pp. 108—121.
  33. Carlberger, G., Kinetics and Distribution of Radioactive Cobalt Administered to the Mammalian Body, Karolinska Institutet, Stockholm, 1961.
  34. Norseth, T., IARC Science Publications, Vol. 53, 1984, pp. 395—401.
  35. Deutman, R., Mulder, T. J., Brian, R., and Nater, J. P., Journal of Bone and Joint Surgery, Vol. 59A, 1977, pp. 862—865.
  36. Woodman, J., Shinn, W., Urban, R., and Galante, J., Journal of Biomedical Materials Research, VoL. 18, 1984, pp. 463—466.
  37. Penman, H. G. and Ring, P. A., Journal of Bone and Joint Surgery, Vol. 66B, 1984, pp. 632—634.
  38. Swann, M., Journal of Bone and Joint Surgery, Vol. 66B, 1984, pp. 629—631.

DISCUSSION

Z. Glaser1 (written discussion)—In regard to the high nickel ion content in the urine of patients after implantation of porous prostheses, is there a possibility that the source of the nickel could be another implant or a dental bridge? Is the body possibly excreting nickel from tissue or cells other than the implant site?

L. C. Jones, D. S. Hungerford, R. V. Kenna, G. Braem, and V. Grant (authors’ closure)— As in vivo corrosion of metallic orthopedic implants occurs at a very low rate, other factors that may contribute to temporal changes in urinary metal ion levels have been sought. In this study, only patients undergoing a primary surgery were included (that is, no revisions). None of the patients had any other orthopedic implants. It is somewhat unlikely that corrosion and wear of dental implants contributed to the increase demonstrated in some individuals. Wear debris from dental implants would primarily be digested. Cobalt, chromium, and nickel are poorly absorbed by the intestines. Therefore, the contribution of this potential source would probably not have a significant impact on the total urinary excretion of these metal ions. No relationship between the epidemiological data gathered (age, sex, occupation, health, and so forth) and the urinary measurements was detected. However, this may be partly due to the low numbers of patients evaluated at this time. It is possible that nickel is released from the cells at the implant site due to trauma. Increases in circulating and excretory nickel levels have been correlated with other types of trauma, including myocardial infarction, acute stroke, and severe burns. The fact that increases in nickel have been observed in the areas surrounding implants, which do not contain nickel, lends support to this hypothesis. However, this hypothesis remains to be tested. A likely source of metal ion release is the debris from implantation of the components. This might vary from patient to patient, depending on the extent of lavage of the surgical site and the effectiveness of the individual’s body in removing this debris from the joint cavity and implant interface.