A sixty-four-year-old woman underwent an uncomplicated, cementless total hip arthroplasty for the treatment of osteoarthritis of the left hip. During this procedure, she received a press-fit titanium acetabular cup (Pinnacle; DePuy, Warsaw, Indiana), a modular cobalt-chromium metal liner (Ultamet; DePuy), and a proximally fixed titanium femoral stem (Summit; DePuy) with a 36-mm cobalt-chromium femoral head. She had no other metallic implants in place. After the total hip arthroplasty, the patient was a physically active person who jogged on a weekly basis and played tennis in an A-level league. The combination of these activities equated to an average of four to five workouts each week. The patient was a nonsmoker and drank alcohol only seldomly, at social events. She was 1.66 m tall, weighed 59.4 kg, and had a body mass index of 21.6. Preoperatively, the renal function had been normal: the level of serum creatinine had been 0.8 mg/dL, the level of blood urea nitrogen had been 11 mg/dL, and the serum electrolyte levels had been normal. This patient is reported on at forty-two months postoperatively in this case report; she was also reported on at thirteen months postoperatively in our previous study of seven patients12.
Study Protocol
We used the two-week-long test protocol of Heisel et al.12 to evaluate serum levels of cobalt and chromium ions and urine levels of chromium during periods of three different activity levels: a week of low-intensity activity, a day of acute high-intensity activity in which a treadmill test was administered, and, finally, a week of high-intensity activity. In addition, we collected all urine produced during the two-week protocol. Patient activity was quantitatively assessed with use of a computerized two-dimensional accelerometer (StepWatch Activity Monitor; Cyma, Seattle, Washington) worn on the ankle11-13. The accelerometer recorded the movement of the lower extremity in real time from the moment the patient arose in the morning until she retired to bed in the evening. This activity data were extrapolated and reported as cycles per year.
Serum samples were obtained for the measurement of cobalt and chromium levels; urine samples were collected for the measurement of chromium levels only. Urine cobalt was not tested because the detection limit for cobalt with our method is higher than that for chromium. A total of fourteen blood samples were taken; two on day 1, eight on day 8 (the day of the treadmill test), two on day 9, and two on day 15. All needles, the intravenous lines that were used on the day of the treadmill test, and the utensils that were used for blood collection were tested before their use to ensure that they were free of metal contamination. Blood samples were collected in polypropylene syringes after the first 5 mL of obtained blood was discarded. The initial flow of blood effectively flushed the tubing, removing detachable contaminants.
Urine was collected in twelve-hour batches that included a morning volume (7 am to 7 pm) and an evening volume (7 pm to 7 am). On the first day, two blood samples were collected and the first twelve-hour urine collection was begun to measure baseline levels of cobalt and chromium in the serum and chromium in the urine. Also on this day, a urinalysis and routine blood tests (a complete blood count, a comprehensive blood chemistry profile, and an evaluation of the erythrocyte sedimentation rate, the prothrombin time, and the international normalized ratio) were also obtained to assess health and renal function. The patient then started the week of low-intensity activity. During that week, the patient was instructed to be as inactive as possible, avoiding exercise, long walks, and sports activities.
On day 8, the patient returned to the clinic and had an intravenous line inserted. Two blood samples were collected to determine the serum levels of cobalt and chromium. The patient then started the treadmill test (acute high-intensity activity), which included walking and/or running on the treadmill as fast as her physical condition allowed for sixty minutes. The speed and speed changes during the test were recorded. Venous blood samples were obtained every fifteen minutes during the treadmill test and at ninety and 120 minutes after the start of the test (i.e., samples taken at fifteen, thirty, forty-five, sixty, ninety, and 120 minutes). Additionally, two blood samples were obtained on the morning after the treadmill test.
A week of high-intensity activity followed the treadmill test. The patient was instructed to be physically active and to exercise as much as possible. At the end of the high-intensity activity week, two blood samples were collected to determine the serum levels of cobalt and chromium and routine blood and urine samples were collected to assess health and renal function. A total of fourteen blood samples and twenty-eight twelve-hour urine samples were obtained during the two-week protocol. A flowchart of the experimental protocol is available in the Appendix.
Trace Metal Analysis
The collected blood was prepared in a centrifuge, and the serum was separated from the clotted fraction with use of a previously reported, validated technique12. The contents of the urine containers for each twelve-hour period were mixed thoroughly, and three 6-mL samples were taken for analysis. The concentrations of cobalt and chromium in serum and the concentration of chromium in urine were measured with graphite-furnace Zeeman atomic absorption spectrophotometry12,14. The detection limits, in parts per billion (ppb), were 0.3 for cobalt in serum, 0.03 for chromium in serum, and 0.015 for chromium in urine. Analyses were performed in the same laboratory and with the same methods as described by Heisel et al.12.
Statistical Methods
Data were presented descriptively as the mean and the standard deviation of the evaluated values. The influence of independent variables on chromium ion excretion and concentration in twenty-four-hour urine collections was assessed with univariate regression analysis.
Activity
The activity of the patient was recorded and extrapolated as a mean (and standard deviation) of 3.10 ± 0.85 million cycles per year during the week of low-intensity activity, 5.72 million cycles per year on the day of acute high-intensity activity during the treadmill test (average speed, 6.6 km/h; 5.6, 6.1, 6.6, and 7.9 km/h for the first, second, third, and fourth fifteen-minute period), and 4.12 ± 0.78 million cycles per year during the week of high-intensity activity. This represents a mean increase in activity of 33% during the week of high-intensity activity and an increase of 85% during the treadmill day compared with the week of low-intensity activity. There was no substantial difference in the activity levels for the patient between the measurements reported in the first study by Heisel et al. (low-intensity activity, 2.73 ± 0.59 million cycles; high-intensity activity, 3.87 ± 0.30 million cycles) and the measurements reported in this study12.
Serum Cobalt
Figure 1 displays all of the serum cobalt levels over the course of a two-week period for both the current study and the previous study. The mean serum cobalt level during the present study was 0.82 ± 0.02 ppb for the week of low-intensity activity, 0.74 ± 0.09 ppb for the treadmill day, and 0.75 ± 0.09 ppb for the week of high-intensity activity. There was no substantial difference in cobalt levels between the two weeks. However, the mean serum cobalt level forty-two months postoperatively (0.77 ± 0.08 ppb) was 25% lower than the level at thirteen months postoperatively (1.02 ± 0.09 ppb)12.
Serum Chromium
Figure 2 displays all of the serum chromium levels over the course of a two-week period for both the current study and the previous study. The mean serum chromium level during the present study was 0.80 ± 0.04 ppb for the week of low-intensity activity, 0.77 ± 0.04 ppb for the treadmill day, and 0.79 ± 0.07 for the week of high-intensity activity. There was no substantial difference in chromium levels between the two weeks. The mean serum chromium level forty-two months postoperatively (0.79 ± 0.06 ppb) was 47% lower than the mean level thirteen months postoperatively (1.5 ± 0.05 ppb)12.
Urine Volume
A total of 37.4 L of urine was collected in twenty-eight containers. The urine volumes in each twelve-hour collection period were variable. During the low-intensity activity week, the mean daily urine volume was 2.70 ± 0.54 L (1.52 ± 0.27 L daytime and 1.18 ± 0.45 L nighttime). During the high-intensity activity week, the mean daily urine volume was 2.64 ± 0.39 L (1.28 ± 0.25 L daytime and 1.36 ± 0.25 L nighttime). There was no substantial difference in the daily urine excretion between the high and low-intensity activity weeks (Fig. 3).
Urine Chromium
A total mass of 16.72 µg of chromium was excreted. During the low-intensity activity week, the mean daily chromium excretion was 1.14 ± 0.26 µg (mean concentration: 0.86 ± 0.14 ppb); 0.73 ± 0.10 µg in the daytime and 0.41 ± 0.19 µg in the nighttime. In the high-intensity activity week, the mean daily chromium excretion was 1.24 ± 0.24 µg (mean concentration: 0.95 ± 0.13 ppb); 0.57 ± 0.16 µg in the daytime and 0.67 ± 0.11 in the nighttime. There was no substantial difference in daily chromium excretion between the high and low-intensity activity weeks (Fig. 3).
Correlations
Chromium urinary excretion was positively and significantly correlated to 24-hour urine volume (r = 0.75; p = 0.02) (Fig. 4). The highest twenty-four-hour urinary chromium excretion was 1.685 µg (3.5 L of urine) on day five (during the low-intensity activity week), and the lowest twenty-four-hour chromium excretion was 0.830 µg (2.1 L of urine) at day 8, following the treadmill test.
No correlation was found between chromium concentration and 24-hour urine volume (r = -0.09; p = 0.75). No correlations were found between patient activity and 24-hour chromium excretion (r = -0.21; p = 0.46) or urine chromium concentration (r = 0.01; p = 0.99).
Homeostatic mechanisms regulate the concentrations of chromium and cobalt in the body15. Cobalt is almost completely excreted in urine within twenty-four to forty-eight hours16, whereas chromium is not fully excreted in urine and can accumulate in the cell walls of tissue and red blood cells15. In the normal population, mean urinary chromium levels range between 0.228 and 0.624 ppb17,18. Individuals with occupational exposure can have urinary chromium values of 1.2 ± 1.4 ppb19. Urinary chromium levels are not substantially influenced by gender, age, consumption of alcohol, or smoking18.
In this study, urinary chromium excretion was strongly correlated to urine volume but independent of patient activity and serum ion concentration. This suggests that renal excretion is the key determinant of serum metal ion levels. This finding is consistent with reports of dramatic elevations of serum cobalt and chromium ion levels in patients with decreased or absent renal function and a metal-on-metal bearing hip prosthesis20. This result also accounts for the transient increases in serum cobalt and chromium ion levels that have been reported in association with marathon running21. During a marathon, as during other extreme and lengthy physical challenges, the production of urine is minimal and fluid is lost mostly through sweat and respiration. Because there is little urinary ion excretion under these conditions, serum cobalt and chromium levels rise. With adequate rehydration and urine production, serum ion levels decrease12.
The patient reported on in this case report was specifically selected because she was known to be doing very well clinically, had been documented as being very physically active, and was willing to comply with the two-week urine-collection protocol. The accelerometer data at forty-two months after surgery were in good agreement with those obtained at thirteen months postoperatively. Despite the fact that our patient was continuously very active between evaluation times, the serum cobalt and chromium levels were substantially lower (25% and 47%, respectively) at the time of this study than they were at the time of the previous study at thirteen months postoperatively. These findings are consistent with the results of wear simulation tests of metal-on-metal bearings, which showed that little additional wear occurs following a run-in phase of variable duration22.
We performed this study to explore the hypothesis that if ion production increases with activity, then the serum ion concentration will rise, unless urine ion excretion increases or deposition and accumulation of metal degradation products are occurring in any tissues. We did not observe increases in urinary excretion. We are unaware of any method to determine whether additional ion deposition is occurring in any tissue as a function of patient activity.
Despite substantial variations in patient activity, serum ion levels and daily urinary ion excretion were essentially constant. A likely explanation for these observations is that, in a patient with a well-functioning prosthesis, ion production is essentially constant and independent of activity. The steady-state wear of the bearing, at least with regard to the activities of the patient under investigation, does not appear to be a measurable source of ion production. Other potentially important sources include the dissolution of wear particles generated during the run-in phase of the bearing4,12 and corrosion of the bearing and nonbearing surfaces.
Due to the substantial practical issues and compliance issues associated with collecting all urine for two weeks, we have data from only one subject, but we believe that our patient is representative of many who are currently receiving metal-on-metal bearings in that she is fit and physically active. Although this patient, who had normal renal function and a well-functioning metal-on-metal bearing, did not demonstrate higher ion levels in association with higher activity levels, other patients with a metal-on-metal bearing hip prosthesis may have higher ion levels as a result of increased production due to less satisfactory tribology or corrosion, reduced excretion as a result of decreased renal efficiency or other physiologic variables, or a combination of tribologic and physiologic factors.
Note: The authors thank Anastasia K. Skipor, MS, for technical support.