Mild or moderate renal insufficiency does not increase circulating levels of cobalt and chromium in patients with metal-on-metal hip arthroplasty O. Lainiala, A. Reito, P. Jämsä, A. Eskelinen ---- Published 31 August 2017 The Journal of Bone & Joint Surgery Abstract Aims To determine whether there is any association between glomerular filtration rate (GFR) and blood cobalt (Co) and chromium (Cr) levels in patients with metal-on-metal (MoM) hip arthroplasty. Patients and Methods We identified 179 patients with a unilateral 36 mm diameter head as part of a stemmed Summit-Pinnacle MoM hip arthroplasty. GFR was calculated using the Modification of Diet in Renal Disease formula. Results Normal renal function (GFR ≥ 90 ml/min/1.73 m2) was seen in 74 patients. Mild renal insufficiency (GFR 60 to 89 ml/min/1.73 m2) was seen in 90 patients and moderate renal insufficiency (GFR 30 to 59 ml/min/1.73 m2) in 15. There was no statistical difference in Co and Cr levels between patients with normal renal function and those with mild or moderate renal insufficiency. No correlation was seen between creatinine and blood metal ion levels or between GFR and blood metal ion levels. Linear regression analysis did not show any association between Co and Cr levels and GFR. Conclusion We did not find any association between GFR and blood metal ion levels. Consequently, the accumulation of Co and Cr in blood due to renal insufficiency does not seem to be a major risk in patients with mild or moderate renal insufficiency. Metal-on-metal (MoM) total hip arthroplasty (THA) and hip resurfacing are associated with elevated blood cobalt (Co) and chromium (Cr) levels from increased wear of the bearing surfaces and damage to the trunnion-taper interface.1,2 A small number of studies have reported elevated metal ion levels in relation to poor kidney function.3,4 It is known that the excretion of Co through the kidneys is the more straightforward process. Cr has a greater tendency to bind to protein, meaning that Cr excretion is not as predictable as it is for Co.5,6 When MoM implants were still widely used, it was suggested that careful consideration should be given to patients with renal insufficiency before they underwent such surgery, and some authors listed renal insufficiency as an absolute contraindication for MoM arthroplasty.7-9 However, the relationship between renal function and blood metal ion levels has not been established in a large study. Concerns about the nephrotoxicity of Co and Cr have been reported, but no evidence of increased renal damage was seen in two studies with follow-up of ten years.10,11 Adverse neurological and cardiovascular effects potentially due to Co and Cr exposure have been described in case reports. In the worst case, cardiomyopathy led to the death of a patient.12,13 This study set out to determine whether the renal function of patients with a MoM THA is related to whole blood levels of Co and Cr. We asked: is there a difference in median Co or Cr levels between patients with normal renal function and mild, moderate, severe or end-stage renal insufficiency?; 2) is there a correlation between blood metal ion levels and plasma creatinine concentration or glomerular filtration rate (GFR)?; and does GFR predict the blood Co and Cr after adjustment for other variables affecting the levels of blood metal ions? Patients and Methods This was a retrospective cross-sectional study for which institutional review board approval had been obtained. To determine the relationship between renal function and blood metal ion levels, we sought to find as homogeneous as possible a study population in terms of implant properties. Component head size is one of the most commonly mentioned variables affecting blood levels of Co and Cr.14 Consequently, we chose to study a cohort of patients who had undergone a Pinnacle (DePuy, Warsaw, Indiana), MoM THA as Pinnacle has a fixed 36 mm diameter head, thereby eliminating head size as confounding variable. Between December 2002 and January 2011, 368 patients underwent 429 Pinnacle MoM implants at our hospital. Patients with any type of contralateral hip implant were excluded. These included 61 patients with bilateral Pinnacle MoM (122 implants) and 81 patients with any other type of contralateral hip implant, leaving 226 patients with unilateral Pinnacle MoM THAs for study. Because arthroplasty stems differ in terms of wear and clinical performance,15,16 we included only the 207 patients with the most frequently used Summit stem (DePuy). Ten patients with a Corail stem (DePuy) and nine patients with an S-ROM stem (DePuy) were excluded (Fig. 1). Blood samples Blood samples for Co and Cr analyses were acquired from the antecubital vein with a 21-gauge needle connected to a Vacutainer system (Becton, Dickinson and Company, Franklin Lakes, New Jersey) and trace element tubes containing sodium ethylenediaminetetraacetic acid. The first 10 ml was discarded to avoid Co and Cr contamination from the needle. Samples were analysed in an independent institution. Standard operating procedures were established for measurements using dynamic reaction cell inductively-coupled plasma (quadrupole) mass spectrometry (Agilent 7500 cx, Agilent Technologies, Santa Clara, California). Blood Co or Cr levels over 7 μg/L were considered elevated.17 Measurements were carried out between November 2010 and July 2013. GFR GFR was calculated using the Modified Diet in Renal Disease (MDRD) formula:18 GFR (mL/min/1.73 m2) = 186 × (plasma-creatinine [µg/L]/88.4)-1.154 × (age)-0.203 × (0.742 if female) × (1.212 if African American). The last pre-operative plasma creatinine values of each patient were retrieved from the hospital database. In four patients, the creatinine value measured on the first post-operative day had to be used, as pre-operative values appeared to be outliers, differing from the patient’s general creatinine trend. For post-operative creatinine values, we sought, in the first instance, to retrieve values that had been measured less than one year before the measurement of blood Co and Cr. If such a value was not available, we used a value that had been measured at a maximum of one year after the blood Co and Cr measurements. If neither was available, we used a value that had been measured over a year after the metal ion measurement, with the stipulation that a creatinine value of similar magnitude had to be measured before the measurement was used for analyses. With this proviso, we considered it reasonable to assume that renal function had remained similar between the measurements. Values measured more than a year before metal ion measurements were not used for analysis, as worsening of renal function could not have been identified. The trend in creatinine values for each patient was analysed individually to make sure that values included in the study were not measured during an episode of acute kidney disease, for example pyelonephritis. Implant orientation The orientation of the implant is known to affect the blood levels of Co and Cr.19 The inclination of the acetabular component was measured from an anteroposterior plain radiograph using the plane of ischial tuberosities as reference. The anteversion was determined using a previously described mathematical method.20 Statistical analysis As the distribution of blood Co and Cr is right-skewed, medians and interquartile ranges (IQR) are reported and tests for non-Gaussian variables were used. The distribution of creatinine and GFR are normal, therefore means with ranges are reported. The Mann-Whitney U test was used to test differences in blood Co and Cr levels between groups with different stage of renal function. The chi-squared test was used to test differences in percentage of patients with elevated blood metal ion levels. Spearman’s correlation coefficient was used to analyse any correlation of blood metal ion levels with creatinine and GFR. The most commonly cited variables affecting the blood Co and Cr levels are bilaterality,21 implant type (THA/resurfacing) and brand,22 implant orientation19 and gender of the patient.23 As we only included patients with unilateral Summit-Pinnacle MoM THA with fixed 36 mm head size, the only co-variables left are GFR; gender; time of exposure (time from surgery to metal ion measurement) and implant orientation (inclination and anteversion). Further, as gender is included in the MDRD formula, the only co-variates included in multivariable analyses were GFR, time of exposure and implant orientation. Linear regression adjusted for GFR, acetabular component inclination angle, anteversion angle and time from implantation to metal ion measurement (time of exposure) was used to predict the risk for increasing blood levels of Co or Cr. Log10-transformation was used for blood Co and Cr to achieve normality. Linear regression adjusted for blood metal ion concentration and exposure time was used to predict the change in GFR (difference between pre-operative and post-operative value). A model which took into account the interaction between blood metal level and time of exposure was used: a p-value < 0.05 was considered statistically significant. Results Blood Co and Cr measurements were available for 198 patients (96%). Of these, at least one blood creatinine measurement was available for 188 patients (95%). In eight patients, the last creatinine value was measured more than one year before the blood Co and Cr measurements, so they were excluded. One patient with end-stage renal insufficiency was excluded as he was undergoing dialysis. Thus, 179 patients with a unilateral 36 mm Summit-Pinnacle MoM THA and blood Co, Cr and creatinine measurements were available for analysis. There were 88 (49%) men and 91 (51%) women. Their mean age at the time of post-operative creatinine measurement was 69.2 years (47 to 84). The mean time from surgery to post-operative creatinine measurement was 6.2 years (1.9 to 11.5). The mean post-operative creatinine value was 75.4 µmol/L (39 to 173) and the mean GFR 87.1 mL/min/1.73 m2 (35.8 to 176.1). The median time from pre-operative creatinine measurement to surgery was 28 days (IQR 18 to 41). The mean pre-operative creatinine value was 70.6 µmol/L (38 to 120) and the mean pre-operative GFR 93.3 mL/min/1.73 m2 (54.7 to 159.0). The mean increase in creatinine value was 4.8 µmol/L (-26 to 85), and the mean decrease in GFR was 6.2 mL/min/1.73 m2 (-62.9 to 32.5). Based on post-operative creatinine measurement, there were 74 patients (41%) with normal renal function (GFR ≥ 90 ml/min/1.73m2), 90 patients (50%) with mild renal insufficiency (GFR 60 to 89 ml/min/1.73m2) and 15 (8%) with moderate renal insufficiency (GFR 30 to 59 ml/min/1.73m2). The post-operative creatinine measurements were carried out at a median 1.6 months before the blood metal ion measurements (IQR 5.8 months before to 0.4 months after). The median Co value was 1.5 μg/L (IQR 0.7 to 4.2) and Cr 1.1 μg/L (IQR 0.7 to 1.8). A total of 20 (11%) patients had Co > 7 μg/L and five (3%) patients had Cr > 7 μg/L (Table I). All patients with an elevated Cr also had an elevated Co. The mean inclination angle was 46.9° (26.7° to 66.9°) and anteversion 22.6° (0° to 43.6°). There was no statistical difference in median Co and Cr levels, or in the percentage of patients with Co or Cr levels > 7 μg/L, between patients with normal renal function and those with mild or moderate renal insufficiency (Table I). No correlation was seen between post-operative plasma creatinine and blood metal ion levels (Co: r = 0.029, p = 0.704, Cr: r = -0.081, p = 0.278), or between GFR and blood metal ion levels (Co: r = -0.107, p = 0.153, Cr: r = -0.071, p = 0.344) (Figs 2 and 3). On linear regression analysis, GFR did not predict blood Co or Cr levels (Tables II and III). Furthermore, blood Co and Cr levels did not predict the change of GFR from pre-operative to post-operative in linear regression analysis adjusted for time of exposure (Table IV). Discussion We did not find any association between GFR and blood metal ion levels, so systemic accumulation of Co and Cr due to their insufficient excretion by the kidneys does not seem to be a major problem in patients with mild or moderate renal insufficiency. Conversely, it seems unlikely that levels of Co and Cr of similar magnitude to those seen in this study would cause significant renal damage. There were no patients in our cohort with severe renal insufficiency, however, we take the view that the relationship between GFR and blood metal ion levels should be studied more closely in patients with severe or even end-stage renal insufficiency. Our study has some weaknesses. First, as a retrospective study, creatinine and blood metal ion levels were not measured at the same time. However, we excluded patients in whom we could not reliably exclude the potential development of chronic renal insufficiency. As the chronic changes in renal function usually occur over several years and acute changes were excluded by selecting the creatinine values as described in the ‘Patients and Methods’ section, we consider it reasonable to say that the interval between creatinine and blood metal ion measurements is not likely to affect our results. Secondly, we did not include comorbidities, occupational exposure and dietary habits in our analysis, as this was a retrospective analysis and these details were not systematically registered. Thirdly, there were no patients with severe renal insufficiency: the proportion of patients with moderate renal insufficiency were significantly smaller than patients with normal GFR or mild renal insufficiency. This mainly reflects the fact that renal insufficiency was considered as contraindication for MoM hip arthroplasty and MoM implants were only used in patients with adequate renal function. We believe that this cohort is a good representation of a typical group of patients with MoM hip arthroplasties. Poor renal function was commonly listed as a contraindication for MoM hip arthroplasty when the MoM arthroplasties were still widely used,7,9 but there are only a few studies about association of renal function and blood metal ion levels. Hur et al3 compared blood levels of Co and Cr in five patients with chronic renal failure (all had GFR < 25, data not shown in the original article, but we calculated it using the MDRD formula from the creatinine concentration, age and gender data provided in their article) and six patients without renal failure. They concluded that patients with renal failure have 100-times higher levels of Co (12.5 µg/L versus 0.1 µg/L). However, the higher mean value among patients with renal failure is explained by one patient with a Co of 51.6 µg/L and another with a of Co 8.0 µg/L (this patient had bilateral MoM hip implants). If a median value had been used, as should be done with skewed data, the difference would have been much smaller (2.3 µg/L versus 0.0 µg/L).3 Brodner et al4 described two patients with renal failure who had serum Co levels that were more than 100-times normal. Concerns about the nephrotoxicity of Co and Cr have been also reported. Marker et al11 did not see a significant decrease in GFR in 98 patients with a Metasul (Zimmer, Winterthur, Switzerland) MoM hip arthroplasty at ten-year follow-up. Median Co levels at the ten-year follow-up visit were 0.75 µg/L (0.3 to 50.1) and Cr 0.95 µg/L (0.3 to 58.6). Corradi et al10 did not note any difference in renal markers between patients who had undergone MoM hip resurfacing ten years previously and healthy age and gender-matched volunteers. In that study, the median Co was 1.8 μg/L (0.3 to 9.0) and Cr 1.5 μg/L (0.2 to 4.6). Fabi et al24 presented a case of a 40-year-old male patient, who had a metal-on-polyethylene hip, in which the polyethylene liner had worn out and wear-through of the titanium acetabular component was seen. In that patient, the GFR was 35 (moderate renal insufficiency, calculated with MDRD formula from data in the article) at the time of revision, which increased to 51 (moderate renal insufficiency) after revision of the implant. The serum titanium level was 456.9 µg/L, Co 1.6 µg/L and Cr 0.9 µg/L.24 Based on our results, the accumulation of Co and Cr due to renal insufficiency in patients with mildly or moderately reduced renal function seems unlikely. Therefore, the increased wear of a bearing surface or damage to the trunnion-taper interface are much more likely to be the reason for accumulation of Co and Cr in the blood. It also seems unlikely that normal or moderately elevated blood Co and Cr concentrations would cause renal failure. However, the significance of severe and end-stage renal insufficiency in the accumulation of Co and Cr, and on the other hand, the possible nephrotoxicity of extremely elevated (hundreds μg/L) Co and Cr levels need further research. Until this information is available, these patients may require closer follow-up than those with normal renal function. Take home message: - Mild or moderate renal insufficiency is unlikely to cause accumulation of Co and Cr. - Co and Cr concentrations equal to those seen in this study are unlikely to cause significant renal damage. - More research is needed about the effects of severe and end-stage renal insufficiency on Co and Cr levels, and about the effect of ultra-high Co and Cr levels on renal function.