original article
http://www.kidney-international.org & 2007 International Society of Nephrology
see commentary on page 376
Mortality effect of coronary calcification and phosphate binder choice in incident hemodialysis patients GA Block1, P Raggi2, A Bellasi3, L Kooienga4 and DM Spiegel4 1
Clinical Research Division, Denver Nephrology, Denver, CO, USA; 2Department of Cardiology, Emory University, Atlanta, GA, USA; Department of Nephrology, San Paolo Hospital, University of Milano, Italy and 4Department of Nephrology, University of Colorado Health Sciences Center, Denver, CO, USA
3
The risk of death in hemodialysis patients treated with calcium-containing phosphate binders or sevelamer is not known. We assessed all-cause mortality in 127 patients new to hemodialysis assigned to calcium-containing binders or sevelamer after a median follow-up of 44 months from randomization. This was a predetermined secondary end point of a randomized clinical trial designed to assess progression of coronary artery calcium (CAC) scores in the two treatment arms. Thirty-four deaths occurred during the follow-up period: 23 in subjects randomized to calciumcontaining phosphate binders and 11 in subjects randomized to sevelamer. Baseline CAC score was a significant predictor of mortality after adjustment for age, race, gender, and diabetes with increased mortality proportional to baseline score (P ¼ 0.002). Mortality was borderline significantly lower in subjects randomized to sevelamer (5.3/100 patient years, confidence interval (CI) (2.2–8.5) compared to those randomized to calcium-containing binders (10.6/100 patient years, CI 6.3–14.9) (P ¼ 0.05). The greater risk of death for patients treated with calcium-containing phosphate binders persisted after full multivariable adjustment (P ¼ 0.016, hazard ratio 3.1, CI 1.23–7.61). In subjects new to hemodialysis baseline CAC score was a significant predictor of all-cause mortality. Treatment with sevelamer was associated with a significant survival benefit as compared to the use of calcium-containing phosphate binders. Kidney International (2007) 71, 438–441. doi:10.1038/sj.ki.5002059; published online 3 January 2007 KEYWORDS: phosphate binders; sevelamer; hemodialysis; coronary artery calcification; mortality
Correspondence: GA Block, Clinical Research Division, Denver Nephrology, CCRI, 130 Rampart Way Suite 300B, Denver, CO 80230, USA. E-mail:
[email protected] Received 4 September 2006; revised 27 October 2006; accepted 7 November 2006; published online 3 January 2007 438
It has been well documented that patients undergoing hemodialysis demonstrate extensive coronary artery calcification (CAC) compared to patients with normal kidney function.1–5 Observational and experimental data suggest that the risk of CAC is related to abnormalities in mineral metabolism and that there is a direct relationship between serum phosphorus, serum calcium, and serum calcium phosphorus product (Ca P) with increased mortality.6–11 In recent trials, investigators have reported that CAC progression can be attenuated in hemodialysis subjects treated with sevelamer as opposed to those treated with calcium-containing phosphate binders.1–3 However, it is still unknown whether phosphate binder choice affects clinical outcomes. Accordingly, for the purpose of this study, we examined the relationship between phosphate binder choice (calciumcontaining phosphate binders versus sevelamer), CAC assessed by electron beam computed tomography (EBCT), and mortality, in a cohort of incident hemodialysis subjects. Assessment of all-cause mortality was a secondary end point of a randomized trial previously reported whose primary end point was change in CAC at 18 months.2 RESULTS
The baseline clinical and laboratory characteristics of the 127 subjects included in this analysis are shown in Table 1. There were no significant baseline differences between the sevelamer and calcium-containing binder groups. At the end of follow-up, 34 total deaths were recorded: 11 in the sevelamer group and 23 in the calcium binder group. Baseline CAC score was a significant predictor of mortality as shown in Figure 1. Subjects with no evidence of CAC had a significantly lower mortality rate (3.3/100 patient years, confidence interval (CI) 0.4–6.1) as compared to subjects with a CAC score 1–400 (mortality rate 7.0/100 patient years, CI 2.7–11.4) and those with a CAC score 4400 (mortality rate 14.7/100 patient years, CI 8.1–21.4) (P ¼ 0.002). After adjustment of baseline CAC score for age, race, gender, and diabetes, the presence of a baseline CAC score 4400 remained significantly associated with increased mortality (hazard ratio ¼ 4.5, P ¼ 0.016, CI 1.33–15.14). Kidney International (2007) 71, 438–441
original article
GA Block et al.: Effect of coronary calcification and phosphate binder
Table 1 | Baseline characteristics of study subjects
Hypertension (%) Diabetes mellitus (%) Congestive Heart Failure (%) ASCVDa (%) Dyslipidemia (%) Smoking (active or former, %) ‘Statin’ medication use (%) Ace Inhibitor medication use (%) Cholesterol total (mg/dl) Cholesterol LDL (mg/dl) Hemoglobin (g/l) C-reactive protein (mg/l) Albumin (g/dl) Calcium (corrected)b (mg/dl) Phosphorus (mg/dl) Ca P (corrected) (mg2/dl2) Intact PTH Baseline CACS (mean)
Calcium salts (n=67)
P-value
56714.8 58
58714.8 64
NS NS
43 27
39 37
NS
95 60 13 37 30 66 36 50 159743 73734 11.671.8 6.776.4 3.570.6 9.370.9 5.271.6 48714 3557317 61671436
98 56 18 22 33 58 30 55 158734 75729 12.072.1 8.2711.3 3.670.5 9.370.8 5.371.4 49713 3587360 58671190
NS NS NS 0.08 NS NS NS NS NS NS NS NS NS NS NS NS NS NS
ASCVD, atherosclerotic cardiovascular disease; CACS, coronary artery calcium score; Ca P, calcium phosphorus product; LDL, low-density lipoprotein; NS, nonsignificant; PTH, parathyroid hormone. a ASCVD: atherosclerotic cardiovascular disease (history of previous myocardial infarction, angina pectoris, angioplasty with/without stent, coronary artery bypass graft, angiographic evidence of atherosclerotic disease, cerebrovascular disease, claudication, lower extremity interventions for atherosclerosis, aortic aneurysm). b Corrected calcium: serum calcium adjusted for serum albumin: total measured calcium + 0.8 (4.0 albumin g/dl).
Survival distribution function
Race (%) Caucasian Black
Sevelamer (n=60)
0.75 P=0.002 0.50 CAC=0 CAC1-400 CAC.400
0.25
0.00 0
6
12 18
24
48
No. at risk CCS = 0 46 CCS < 400 42 CCS . 400 39
30 36 42 Months
42 41 37
42 40 35
39 36 31
34 32 26
18 14 15
54
60
66
4 1 4
Figure 1 | Adjusted survival by baseline CAC score. Multivariable adjusted (age, race, gender, diabetes) association between baseline CAC score and survival. P-value represents significance across all three groups.
1.00 Survival distribution function
Age (mean7s.d.) Male (%)
1.00
P =0.016 0.75
0.50
0.25 Calcium Sevelamer
0.00
CAC scores were similarly distributed at baseline in both treatment groups (CAC ¼ 0: calcium 33%, sevelamer 40%; CAC 1–400: calcium 36%, sevelamer 30%; CAC 4400: calcium 31%, sevelamer 30%). Mortality was borderline significantly higher in subjects treated with calcium binders (10.6/100 patient years, CI 6.3–14.9) compared to those randomized to sevelamer (5.3/100 patient years, CI 2.2–8.5, P ¼ 0.05). Factors associated with increased mortality in univariate analysis included phosphate binder assignment (P ¼ 0.06), baseline CAC score (P ¼ 0.003), baseline history of atherosclerotic cardiovascular disease (P ¼ 0.02), and baseline C-reactive protein (P ¼ 0.06). After multivariable adjustment for age, race, gender, diabetes, albumin, Kt/V, history of atherosclerotic cardiovascular disease, C-reactive protein, and baseline CAC score, a significant increase in mortality was observed for calciumtreated subjects (P ¼ 0.02, hazard ratio 3.1, CI 1.23–7.61) (Figure 2). DISCUSSION
This analysis represents the secondary end point of a randomized clinical trial and is notable for describing that baseline CAC score and choice of phosphate binder (calcium Kidney International (2007) 71, 438–441
No. at risk Calcium Sevelamer
0
6
67 60
63 57
12 18
60 57
24
30 36 42 Months
48
55 51
45 47
22 25
54
60
66
5 4
Figure 2 | Adjusted survival by phosphate binder assignment. Multivariable adjusted (age, race, gender, diabetes, history of atherosclerotic cardiovascular disease, C-reactive protein, albumin, Kt/V, and baseline CAC score) association between phosphate binder treatment assignment (calcium versus sevelamer) and survival.
versus sevelamer) are independent predictors of mortality in subjects who are new to hemodialysis. It is well recognized that vascular calcification is common and progressive in patients with advanced chronic kidney disease. Several lines of evidence suggest that cardiovascular calcification is a marker of poor prognosis in chronic kidney disease stage 5, but only one study, so far, has addressed the risk of mortality in patients that underwent quantitative EBCT imaging for CAC.12 The current trial confirms the finding that the severity of CAC at the time of initiation of hemodialysis is an important predictor of long-term survival. Even after adjustment for age, race, gender, and diabetes, a baseline CAC 4400 was associated with a greater than fourfold increase in mortality. 439
original article
Three previous randomized clinical trials have shown that the use of a calcium-containing phosphate binder is associated with a rapidly progressive increase in the extent of CAC, whereas the use of sevelamer is associated with modest, if any, increase in CAC score.1–3 The current analysis extends those findings and shows that the use of sevelamer confers a significant survival benefit as compared to the use of calciumcontaining phosphate binders. After adjustment for age, race, gender, diabetes, and baseline CAC score, the use of calciumcontaining phosphate binders was associated with a doubling of mortality. These results provide compelling evidence that phosphate binder selection and mineral metabolism are linked with mortality in subjects new to hemodialysis. Of note, these data also demonstrate improved survival in subjects with no CAC at the time of hemodialysis initiation. As shown in the parent study,2 these subjects did not accrue substantial calcification during the first 18 months of dialysis, independent of phosphate binder treatment, and this appears to have translated into a long-term survival benefit. This observation suggests that a proportion of hemodialysis subjects may be initially protected from progressive arterial calcification via mechanisms which remain unclear but which are associated with a favorable prognosis. It remains to be elucidated whether active intervention to prevent or attenuate vascular calcification in patients with earlier stages of chronic kidney disease would ultimately impact long-term outcomes once renal replacement therapy is initiated. There are a few limitations to our analysis. This incident population was modestly distinct from the generalized incident population with respect to younger age and a higher prevalence of diabetes. This analysis was carried out on an intent-to-treat basis using the randomized binder as the treatment assignment group. Patients were randomized to phosphate binder therapy and followed for 18 months on the prescribed binder, but after that time physicians were free to change phosphate binder therapy at their discretion. Misclassification bias, however, would reduce, rather than enhance, the likelihood of finding a significant difference between randomized treatment groups. The tendency to minimize any difference in outcome would be true whether patients switched from sevelamer to calcium or from calcium to sevelamer. Subjects treated with sevelamer consistently experience a reduction in total and low-density lipoprotein cholesterol.13 Although there has been recent evidence that low-density lipoprotein reduction with statin therapy does not have a demonstrable effect on survival in diabetic patients undergoing dialysis,14 it remains possible that the lipid lowering effects of sevelamer may have an affect on all-cause mortality. Lastly, it is worth noting that our results are not, on the surface, consistent with the recent report (available in abstract form only) from a large randomized trial comparing all-cause mortality in dialysis subjects given either sevelamer or calcium-containing phosphate binders (dialysis clinical outcomes revisited). We believe this is likely attributable to significant differences in study design. This prospective randomized trial was conducted in patients new to 440
GA Block et al.: Effect of coronary calcification and phosphate binder
hemodialysis and followed for survival status over a median follow-up of 44 months, whereas the dialysis clinical outcomes revisited trial enrolled only subjects prevalent on long-term dialysis (average 38 months) who had Medicare as a primary payer and who were followed for a mean follow-up of 20 months from randomization. It is noteworthy that in the dialysis clinical outcomes revisited trial, those subjects who were followed for at least 2 years did demonstrate a survival benefit when treated with sevelamer, particularly in the elderly subgroup (those who are likely to be the most heavily calcified). Thus, the different population characteristics, the shorter duration of follow-up and the inability to discriminate those subjects who were more calcified from those who were not calcified are likely to explain the differences in results from our own report. In summary, we have shown that the presence and severity of CAC is a predictor of all-cause mortality in incident hemodialysis patients and that the use of sevelamer as a phosphate binder is associated with decreased mortality. These data lend support to the Kidney Disease Outcomes Quality Initiative guidelines regarding the preferential use of sevelamer in the presence of cardiovascular calcification.15 These results further question the advisability of using calcium-based phosphate binders as first line therapy. MATERIALS AND METHODS Between September 2000 and December 2002, 129 adult subjects new to hemodialysis were randomized in blocks of 10 and stratified by diabetic status to receive either calcium-containing phosphate binders or sevelamer hydrochloride. The parent study was an 18month randomized, open-label clinical trial assessing the effect of phosphate binders on progression of CAC detected by EBCT.2 During the study period, subjects remained on their assigned phosphate binder. After the final scan, subjects were given phosphate binders at the discretion of their primary nephrologist. Of the 129 subjects, two subjects from the parent study were subsequently determined to have non-qualifying baseline EBCTs and were excluded from this analysis. The remaining 127 patients comprised the final study population. For all subjects, verification of vital status was conducted through 31 December 2005 and assessed by chart analysis and physician verification (GB, DS, PR) as well as confirmation with the Social Security Death Index (http://ssdi. rootsweb.com). The median follow-up time was 44 months. LABORATORY AND IMAGING PROCEDURES
Routine biochemical measurements were obtained at baseline and performed by the facilities usual laboratory (DaVita Laboratories, Gambro Laboratories). EBCT imaging was performed using a C-150 scanner (GE-Imatron, San Francisco, CA, USA) according to a standard protocol as described previously.3 CAC score quantification was performed with the Agatston method.16 STATISTICAL ANALYSIS
One hundred and twenty-seven subjects who underwent a baseline EBCT scan were included in the follow-up analysis for occurrence of all-cause mortality. Comparison of baseline Kidney International (2007) 71, 438–441
original article
GA Block et al.: Effect of coronary calcification and phosphate binder
characteristics was done using the Fischer exact test for categorical variables and the t-test for continuous variables. Kaplan–Meier survival curves were generated and a log-rank test was used to compare treatment group differences in time to death. All outcomes other than death, including subjects lost to follow up, subjects who underwent kidney transplant, and subjects still alive at the end of the study period (31 December 2005), were censored. Mortality incidence rates (number of deaths divided by patient-years of follow-up) and their 95% CIs were calculated for each treatment group. Each of the above analyses were performed separately for the entire population as well as for subgroups based on baseline CAC score. After verifying that the proportionality assumption of the Cox proportional hazards regression model was valid, mortality risk was assessed including baseline covariates known to be predictive of mortality (age, race, gender, diabetes, albumin, and Kt/V) and those variables found on univariate analysis to have a Po0.10 (phosphate binder assignment, baseline CAC score, history of atherosclerotic cardiovascular disease and C-reactive protein). All P-values are two tailed. Pp0.05 were considered statistically significant. Analyses were conducted using SAS (version 8.2; Cary, NC, USA) and GraphPad Prism software (version 4.0; San Diego, CA, USA).
2.
3.
4.
5.
6.
7.
8.
9. 10.
11. 12.
ACKNOWLEDGMENTS
The parent study was supported financially by Genzyme, Corp. All aspects of the design, conduct, and analysis of the trial, including ownership of the data were under the sole authority of the authors without any restrictions on publication. Conflict of interest: GB is the Advisor and consultant to Genzyme, Corp., PR is the Advisor and consultant to Genzyme Corp., and DS is the Speakers Bureau Genzyme Corp.
13.
14.
15.
REFERENCES 1. Chertow GM, Burke SK, Raggi P. Sevelamer attenuates the progression of coronary and aortic calcification in hemodialysis patients. Kidney Int 2002; 62: 245–252.
Kidney International (2007) 71, 438–441
16.
Block GA, Spiegel DM, Ehrlich J et al. Effects of sevelamer and calcium on coronary artery calcification in patients new to hemodialysis. Kidney Int 2005; 68: 1815–1824. Braun J, Asmus H-G, Holzer H et al. Long term comparison of a calcium free phosphate binder and calcium carbonate-phosphorus metabolism and cardiovascular calcification. Clin Nephrol 2004; 62: 104–115. Braun J, Oldendorf M, Moshage W. Electron beam computed tomography in the evaluation of cardiac calcifications in chronic dialysis patients. Am J Kidney Dis 1996; 27: 394–401. Goodman WG, Goldin J, Kuizon BD et al. Coronary artery calcification in young adults with end stage renal disease who are undergoing dialysis. N Engl J Med 2000; 342: 1478–1483. Young EW, Akiba T, Albert J et al. Magnitude and impact of abnormal mineral metabolism in hemodialysis patients in the dialysis outcomes and practice patterns study (DOPPS). Am J Kidney Dis 2004; 44: S34–S38. Young EW, Albert JM, Satayathum S et al. Predictors and consequences of altered mineral metabolism: the dialysis outcomes and practice patterns study. Kidney Int 2005; 67: 1179–1187. Block GA, Klassen PS, Lazarus JM et al. Mineral metabolism, mortality, and morbidity in maintenance hemodialysis. J Am Soc Nephrol 2004; 15: 2208–2218. Goodman WG, London GM, Amann K et al. Vascular calcification in chronic kidney disease. Am J Kidney Dis 2004; 43: 572–579. Reynolds JL, Joannides AJ, Skepper JN et al. Human vascular smooth muscle cells undergo vesicle mediated calcification in response to changes in extracellular calcium and phosphate concentrations: a potential mechanism for accelerated vascular calcification in ESRD. J Am Soc Nephrol 2004; 15: 2857–2867. Shanahan CM. Mechanisms of vascular calcification in renal disease. Clin Nephrol 2005; 63: 146–157. Matsuoka M, Iseki K, Tamashiro M et al. Impact of high coronary artery calcification score (CACS) on survival in patients on chronic hemodialysis. Clin Exp Nephrol 2004; 8: 54–58. Chertow GM, Raggi P, McCarthy J et al. The effects of sevelamer and calcium acetate on proxies of atherosclerotic and arteriosclerotic vascular disease in hemodialysis patients. Am J Nephrol 2003; 23: 307–314. Wanner C, Krane V, Marz W et al. Atorvastatin in patients with type 2 diabetes mellitus undergoing hemodialysis. N Engl J Med 2005; 353: 238–249. National Kidney Foundation. Clinical practice guidelines for bone metabolism and disease in chronic kidney disease. Am J Kidney Dis 2003; 42: 1–201. Agatston AS, Janowitz WR, Hildner F et al. Quantification of coronary artery calcium using ultra fast computed tomography. J Am Coll Cardiol 1990; 15: 827–832.
441
