- Email: [email protected]
HMG-coenzyme a reductase inhibitor use is associated with mortality reduction in hemodialysis patients
HMG-Coenzyme A Reductase Inhibitor Use Is Associated With Mortality Reduction in Hemodialysis Patients Nancy A. Mason, PharmD, George R. Bailie, PharmD, PhD, Sudtida Satayathum, MS, Jennifer L. Bragg-Gresham, MS, Takashi Akiba, MD, PhD, Tadao Akizawa, MD, PhD, Christian Combe, MD, Hugh C. Rayner, MD, Akira Saito, MD, Brenda W. Gillespie, PhD, and Eric W. Young, MD ● Background: Cardiovascular disease is the most common cause of mortality in patients with end-stage renal disease. Cardiovascular benefits of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins) have been clearly established in the general population, but not in dialysis patients. This study examined statin prescription patterns and assessed the relationship between statin prescription and clinical outcomes in hemodialysis (HD) patients. Methods: Data were analyzed from the Dialysis Outcomes and Practice Patterns Study, a prospective observational study of HD patients randomly selected from representative dialysis facilities in France, Germany, Italy, Spain, the United Kingdom, Japan, and the United States. Predictors of statin prescription were investigated by means of logistic regression. Cox regression models tested the association between statin prescription and risk for mortality and cardiac events, with adjustments for common demographic factors and comorbid conditions. Results: Statins were prescribed for 11.8% of HD patients overall. Most facilities (81.2%) prescribed statins to less than 20% of their patients. Patients prescribed statins had a 31% lower relative risk for death compared with those not prescribed statins (P < 0.0001). Statins were associated with a 23% lower cardiac mortality risk (P ⴝ 0.03) and a 44% lower noncardiac mortality risk (P < 0.0001). At a facility level, prescribing statins was associated with lower overall mortality rate, with a 5% lower risk for every 10% increase in number of patients prescribed statins within the facility (P ⴝ 0.02). Conclusion: Statin prescription is associated with reduced mortality in HD patients, providing additional support for the value of statin therapy in this patient group. Am J Kidney Dis 45: 119 –126. © 2004 by the National Kidney Foundation, Inc. INDEX WORDS: Cardiovascular disease; cholesterol; hemodialysis (HD); 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors; hyperlipidemia; statins.
C
ARDIOVASCULAR DISEASE is the most common cause of mortality in patients with end-stage renal disease (ESRD), accounting for more than 40% of deaths in this population.1 Many factors contribute to this high rate of cardiac mortality, including traditional risk factors typically identified in the general population and renal-specific risk factors. Traditional risk factors include age, sex, family history of cardiovascular disease, hypertension (HTN), smoking, diabetes mellitus, dyslipidemia, obesity, and lack of physical activity.2 Additional renal failure– related risk factors potentially present in hemodialysis (HD) patients include HTN associated with volume overload, anemia, disordered metabolism of calcium and phosphorus, hyperparathyroidism, accumulation of metabolic wastes, chronic inflammation, and hypercatabolism.3 Dialysis treatment confers a tendency toward developing cardiac disease through the possibility of inadequate dialysis and bioincompatibility of HD membranes.3 As with the general population, some patients with ESRD have elevated low-density lipoprotein (LDL) cholesterol levels, but a nontraditional lipoprotein pattern also is seen in these
patients. This pattern is characterized by normal or low levels of LDL, a low level of high-density lipoprotein (HDL), a high triglyceride level, and
From the College of Pharmacy; Department of Biostatistics, School of Public Health; Division of Nephrology, Department of Veterans Affairs Medical Center, University of Michigan; Nephrology Pharmacy Associates; University Renal Research and Education Association, Ann Arbor, MI; Renal Research Institute Inc, New York; Albany College of Pharmacy, Albany, NY; Tokyo Women’s Medical University, Tokyo; Wakayama Medical University, Wakayama; Tokai University, Kanagawa, Japan; Hôpital St André, Bordeaux, France; and Birmingham Heartlands Hospital, Birmingham, UK. Received April 5, 2004; accepted in revised form September 1, 2004. The Dialysis Outcomes and Practice Patterns Study is supported by research grants from Amgen Inc and Kirin Brewery Ltd without restrictions on publications. This project was funded in part by Renal Research Institute Inc. Presented as an abstract at the 2001 ASN/ISN World Congress of Nephrology, San Francisco, CA, October 2001. Address reprint requests to Nancy A. Mason, PharmD, University of Michigan College of Pharmacy, 428 Church St, Ann Arbor, MI 48109-1065. E-mail: [email protected] © 2004 by the National Kidney Foundation, Inc. 0272-6386/04/4501-0013$30.00/0 doi:10.1053/j.ajkd.2004.09.025
American Journal of Kidney Diseases, Vol 45, No 1 (January), 2005: pp 119 –126
119
120
MASON ET AL
elevated non-HDL cholesterol levels, with elevated levels of lipoprotein remnants (very-LDL and intermediate-density lipoprotein) and small dense LDL particles that are considered highly atherogenic.4 Unfortunately, no randomized controlled studies show that treating dyslipidemias reduces the incidence of atherosclerotic cardiovascular disease in dialysis patients. Based on observational data and the strength of evidence supporting the benefits of treatment in the general population, the National Kidney Foundation– Kidney Disease Outcomes Quality Initiative (K/ DOQI) work group, responsible for developing clinical practice guidelines for managing dyslipidemias in patients with chronic kidney disease, recommended the use of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins) as first-line drug therapy for high LDL and nonHDL cholesterol levels in patients with chronic kidney disease.5 This study examines statin prescription patterns and associated outcomes in a large international HD patient population, using data from the Dialysis Outcomes and Practice Patterns Study (DOPPS I). METHODS DOPPS I is a large, prospective, longitudinal, and observational study of HD prescription practices and outcomes of patients randomly selected from representative dialysis facilities in 3 continents (5 countries in Europe: France, Germany, Italy, Spain, and the United Kingdom; Japan; and the United States). A stratified random sample of dialysis facilities was selected in each country. A census of prevalent HD patients older than 17 years was used to randomly select 20 to 40 patients from each selected facility. This study has been described previously.6,7 Data have been longitudinally collected since 1996 for the United States (148 facilities), 1998 for Europe (101 facilities), and 1999 for Japan (65 facilities). Follow-up continued through mid-2001. Some patients were lost to death, transplantation, change in treatment modality, withdrawal from dialysis therapy, recovery of renal function, or transfer to another facility. These patients were replaced on an ongoing basis. Briefly, the major goal of the DOPPS is to investigate the effect of HD practice patterns on patient outcomes. Primary study end points for the DOPPS are mortality, hospitalization, vascular access outcomes, and quality of life. Data are collected for each participating patient by using survey questionnaires administered by a study coordinator at each facility. Data include patient demographic characteristics, medical history, laboratory values, dialysis unit practices and outcomes, and prescribed over-the-counter and prescription drugs. The sample size was planned to give adequate
statistical power to analyze the associations between practice patterns and the 4 outcomes noted. Prescribed statin treatment practice was evaluated using DOPPS I data from the United States, Europe, and Japan. All medications listed in the medical record (both prescription and nonprescription) were recorded for each patient at entry into the study and at 4-month follow-up intervals and coded into the appropriate drug class, such as statins. Actual consumption of prescribed medicines was not recorded.
Statistical Methods For analysis, data included 9,846 patients in the United States, 4,591 patients in Europe, and 2,784 patients in Japan. Descriptive statistics for statin prescription and total cholesterol levels were computed based on a prevalent crosssection of patients (n ⫽ 7,365) between January and May 2000. Survey sampling weights were used to adjust for the cluster sampling of patients within facilities. Logistic regression was used to investigate predictors of statin prescription, based on a single record per person between January and May 2000. Covariate effects are presented as adjusted odds ratios (ORs). Potential predictors investigated include age, sex, race, years with ESRD, summary comorbid conditions (coronary artery disease [CAD], congestive heart failure, cardiovascular disease, HTN, peripheral vascular disease, and diabetes), and physician attitude toward hyperlipidemia treatment. Indicators were used to account for missing variables. Models took into account facility clustering with generalized estimating equations. Statin prescription also was investigated as a predictor of all-cause mortality, cardiac event, and cardiac mortality. Cox regression models, with study entry as the time origin and a time-dependent covariate indicating statin prescription, were used to estimate the relative risk (with versus without statins) for all-cause mortality, cardiac event, and cardiac mortality in 3 patient groups: all patients (n ⫽ 17,221), patients with CAD (n ⫽ 6,637), and patients with HTN (n ⫽ 12,980). All Cox regression models were stratified by country and adjusted for age, sex, race, body mass index, predialysis serum concentrations of albumin and hemoglobin, years with ESRD, normalized protein catabolic rate, and 15 summary comorbid conditions present at study entry (CAD, congestive heart failure, other cardiovascular disease, HTN, cerebrovascular disease, peripheral vascular disease, diabetes mellitus, lung disease, cancer [excluding skin], human immunodeficiency virus/acquired immunodeficiency syndrome, gastrointestinal bleeding, neurological disease, psychiatric disease, recurrent cellulitis, and dyspnea). Smoking was not significant after adjusting for comorbidities and was not included in the models. Facility clustering effects were addressed with robust SE estimates based on the sandwich estimator.8 The association between variation in facility-level statin prescription and mortality was examined by using Cox regression. Facility-level statin prescription is defined as the percentage of patients in the facility prescribed a statin at study entry. All statistical analyses were performed using SAS software, version 8.2 (SAS Institute, Cary, NC).
STATINS IN HEMODIALYSIS
121
Fig 1. Percentage of patients with a statin prescription among a prevalent cross-section of hemodialysis patients (n ⴝ 7,365) by country and date of reported prescription.
RESULTS
1). Of patients with a serum cholesterol level of 200 mg/dL or greater (ⱖ5.17 mmol/L), 15.7% overall were prescribed statins, whereas 17.5% of those with CAD and 22.0% of those with a history of myocardial infarction (MI) were prescribed these agents. There was considerable intercountry variability in the relative proportion of dialysis patients with each high-risk comorbid condition (high cholesterol level, CAD, or MI) and percentage of statin prescription in these specific subpopulations. However, in all countries, less than 28% of patients in any given high-risk category were treated with statins. Among patients treated with a statin in early 2000, simvastatin was prescribed for the majority of patients (58.4% to 72.3%) in France, Italy, and Spain. At the same time, pravastatin was the
In a prevalent cross-section of HD patients (n ⫽ 7,365), prescription of statins varied depending on the patient’s country of residence and date of reported prescription (Fig 1). The United States, France, and Germany reported greater prescription of statins than the other 4 countries, with the proportion of patients prescribed a statin increasing steadily during the study period for all countries except Spain. Other European countries and Japan had lower frequencies of statin prescription and tended to have smaller increases in the prescription of these agents over time. Taking all countries together, in early 2000 (the most recent year for which comparative data for all countries were available), statins were prescribed for 11.8% of HD patients overall (Table
Table 1. Percentage of Patients With High Cholesterol Levels, CAD, or MI and Percentage Prescribed Statins (January 2000 to May 2000) Patients With Condition Receiving Statins (%)
Patients With (%)
Country
Statin Prescription
Cholesterol ⱖ200 mg/dL*
CAD History
MI History
Cholesterol ⱖ200 mg/dL*
CAD History
MI History
France Germany Italy Japan Spain United Kingdom United States Overall
15.7 11.9 3.5 7.1 4.9 8.1 16.6 11.8
43.5 59.2 33.5 15.0 34.1 37.2 18.3 20.8
28.4 42.2 19.8 18.5 22.4 30.2 46.1 33.4
9.9 13.6 6.1 5.1 6.1 9.0 15.8 10.8
12.9 2.7 5.0 11.9 5.8 3.9 27.5 15.7
24.0 14.5 4.4 6.2 5.3 10.5 22.3 17.5
26.4 20.0 7.6 5.0 4.2 15.4 27.5 22.0
*Equivalent to 5.17 mmol/L.
122
MASON ET AL
Fig 2. Distribution of facility-level statin prescription (January 2000 to May 2000).
most commonly prescribed agent in Germany and Japan (51.4% and 58.1%, respectively). Atorvastatin was the most commonly prescribed agent in the United States (44.9%), with simvastatin the second most common (32.7%). The United Kingdom reported nearly equal prescription of pravastatin (48.5%) and simvastatin (49.7%). Analysis of statin prescription patterns within facilities showed that, from January 2000 to May 2000, a large majority of facilities (81.2%) prescribed statins to 20% or less of their patients (Fig 2). Some 16.5% reported no patients administered statins in their facilities, whereas a small
proportion (⬍1%) of facilities prescribed statins to more than 50% of their patients. Several demographic factors and comorbid conditions were examined to determine their association with statin prescription at study entry (Table 2). Based on multivariate analysis, patients were significantly less likely to be prescribed a statin if they were older (OR, 0.89/10year increase in age; P ⫽ 0.01), male (OR, 0.74; P ⫽ 0.04), or had been on dialysis therapy for a longer period (OR, 0.95/1-y increase in time on dialysis therapy; P ⫽ 0.001). There was a greater likelihood of a patient being prescribed a statin in
Table 2. Association of Demographic Factors and Comorbid Conditions With Statin Prescription at Study Entry Mean or Percentage Characteristic
Demographics Age (/10-y increase) Black (v other) Male (v female) Time with ESRD (/1-y increase) Comorbidities (yes v no) Coronary artery disease Congestive heart failure Other cardiovascular disease Hypertension Peripheral vascular disease Diabetes Facility characteristics Hospital (v nonhospital based) Aggressive hyperlipidemia therapy†
Statin
No Statin
OR*
P
61.6 13.4 45.0 4.6
59.7 17.9 57.5 5.4
0.89 0.65 0.74 0.95
0.01 0.12 0.04 0.001
46.5 32.2 34.4 78.9 28.5 43.4
35.3 29.3 33.1 73.1 20.8 32.4
1.55 0.84 0.8 1.1 1.52 1.39
0.01 0.47 0.21 0.56 0.03 0.05
55.7 62.3
54.8 50.9
0.83 1.84
0.31 0.0006
NOTE. Mean or percentage and 3-hydroxy-3 methylglutamyl coenzyme A inhibitor use was based on a point-prevalent sample from January 2000 to May 2000 (n ⫽ 8,615). Models (not shown) also examined factors of income, education, facility size, physician-patient contact time, and other comorbid conditions. *OR adjusted for all variables listed in the table, as well as for country. †Facilities with medical directors who agreed with the statement, “Hyperlipidemia is aggressively sought and treated in our patients.”
STATINS IN HEMODIALYSIS
123
Table 3. Association of Mortality and Cardiac Events With Statin Prescription Outcome
All-cause mortality All patients Patients with CAD Patients with HTN Cardiac event* All patients Patients with CAD Patients with HTN Cardiac mortality† All patients Patients with CAD Patients with HTN Noncardiac mortality All patients Patients with CAD Patients with HTN
Hazard Ratio
P
95% Confidence Interval
0.69 0.70 0.70
⬍0.0001 ⬍0.0001 ⬍0.0001
0.60-0.79 0.60-0.83 0.61-0.80
0.78 0.83 0.72
0.03 0.14 0.01
0.62-0.98 0.64-1.06 0.57-0.93
0.77 0.78 0.71
0.03 0.05 0.01
0.61-0.97 0.60-1.00 0.56-0.91
0.56 0.53 0.59
⬍0.0001 ⬍0.0001 ⬍0.0001
0.46-0.69 0.40-0.69 0.47-0.73
NOTE. Hazard ratios were estimated using Cox regression models with a time-dependent covariate indicating statin prescription. All models were adjusted for age, sex, race, body mass index, serum albumin level, hemoglobin level, time with ESRD, normalized protein catabolic rate, and summary comorbid conditions. *Cardiac event indicates patient death or hospitalization because of coronary disease. †Cardiac mortality indicates patient death from coronary disease.
the presence of CAD (OR, 1.55; P ⫽ 0.01), peripheral vascular disease (OR, 1.52; P ⫽ 0.03), or diabetes (OR, 1.39; P ⫽ 0.05). There also was an increased likelihood of statin use in facilities with medical directors who agreed with the statement, “Hyperlipidemia is aggressively sought and treated in our patients” (OR, 1.84; P ⫽ 0.0006). Other factors, such as income, education, facility size, physician-patient contact time, and other comorbid conditions, were not significant in explaining the variation in statin prescription.
Fig 3. Adjusted survival curve for all-cause mortality for patients prescribed versus not prescribed statins. Estimates were calculated from a Cox regression model adjusted for age, sex, race, body mass index, serum albumin level, hemoglobin level, time with ESRD, normalized protein catabolic rate, and summary comorbid conditions. Statin use was modeled as a time-varying covariate, although the curves presented here represent no statin use versus continuous statin use.
Examination of all-cause mortality data showed that patients prescribed statins had a 31% lower relative risk for death compared with those not prescribed statins (P ⬍ 0.0001) after adjusting for demographics and comorbidities (Table 3; Fig 3). A significantly lower mortality risk (30%; P ⬍ 0.0001) extended to each subgroup with CAD and with HTN. Risk for death or hospitalization because of coronary disease also was significantly lower among the HD population as a whole and the subgroup with HTN. Statins were associated with a 23% lower cardiac mortal-
124
MASON ET AL Table 4. Association of Mortality and Cardiac Events With Facility Statin Prescription
Outcome (/10% increase in facility statin prescription)
All-cause mortality All patients Patients with CAD Patients with HTN Cardiac event* All patients Patients with CAD Patients with HTN Cardiac mortality† All patients Patients with CAD Patients with HTN Noncardiac mortality All patients Patients with CAD Patients with HTN
Hazard Ratio
P
95% Confidence Interval
0.95 0.94 0.95
0.02 0.05 0.03
0.90-0.99 0.89-1.00 0.90-0.99
0.93 0.93 0.92
0.09 0.19 0.07
0.86-1.01 0.85-1.03 0.84-1.01
0.91 0.91 0.90
0.02 0.07 0.01
0.84-0.99 0.82-1.01 0.82-0.98
0.97 0.95 0.96
0.40 0.23 0.25
0.91-1.04 0.87-1.04 0.89-1.03
NOTE. Hazard ratios were estimated using Cox regression models with a time-dependent covariate indicating facility statin prescription. All models adjusted for age, sex, race, body mass index, serum albumin level, hemoglobin level, time with ESRD, normalized protein catabolic rate, and summary comorbid conditions. Facility statin prescription was defined as the percentage of patients prescribed a statin in the facility. *Cardiac event indicates patient death or hospitalization caused by coronary disease. †Cardiac mortality indicates patient death from coronary disease.
ity risk (P ⫽ 0.03) and 44% lower noncardiac mortality risk (P ⬍ 0.0001) in this patient population. The lower all-cause mortality associated with statin use was found in all 7 countries included in the analysis. The magnitude of the lower risk ranged from 19% to 83%. No significant difference in the statin effect among countries was found (interaction P ⫽ 0.32). The association between facility statin prescription and outcomes also was examined (Table 4). The practice of prescribing statins was associated with a lower overall mortality rate for all patients, with a 5% lower risk for every 10% increase in number of patients prescribed statins in the facility (P ⫽ 0.02). Risk for death or hospitalization because of coronary disease also was significantly lower in the HD population as a whole. At the facility level, statins were associated with a 9% lower cardiac mortality risk for every 10% increase in number of patients prescribed statins in the facility (P ⫽ 0.02), whereas noncardiac mortality was not significantly affected. DISCUSSION
Analysis of a prevalent cross-section of 7,365 HD patients showed that statin prescription was
associated with 31% lower all-cause mortality and 23% lower risk for cardiac mortality in HD patients. The lower mortality risk found in this study supports earlier observations from 3,700 incident dialysis patients in the US Renal Data System Dialysis Morbidity and Mortality Study Wave 2 from 1996 to 1998. In that study, Seliger et al9 found 32% lower all-cause mortality and 36% lower cardiovascular mortality. The magnitude of the lower mortality risk observed in these 2 studies is similar to that seen in large, randomized, controlled trials of the general population10-12 and studies including patient cohorts with impaired renal function.13-16 Observational data are subject to some inherent limitations, including treatment bias. One potential explanation for these results is that patients could have been selected for statin treatment because they were expected to survive long enough to benefit. For instance, in clinical practice, it is possible that within the population of patients with CAD, for whom outcomes will be adjusted similarly, only those judged likely to survive long enough to benefit may be prescribed statins. Hence, there would be bias within the CAD group for statins to be administered to those with a better prognosis. However, the facil-
STATINS IN HEMODIALYSIS
ity-level association would less likely be affected by this patient-level effect and therefore supports the conclusion that this bias is not a major factor. Moreover, it is at least as likely that statin therapy would be administered preferentially to patients with more severe coronary disease, potentially creating bias in the opposite direction. Statin prescription also was associated with a remarkable 44% lower risk for noncardiac mortality. The potential protective mechanisms associated with statins have been reviewed in detail elsewhere and probably reflect a combination of factors in addition to normalization of lipid profile. If the survival benefit of statin use were caused entirely by normalization of lipid levels, then adjusting for cholesterol level would reduce or eliminate the significance of statins in a statistical model. However, with our data, adjusting for lipid levels did not reduce the effect of statins. In addition, in a model including cholesterol level but omitting statins, only very low cholesterol levels (⬍140 mg/dL [⬍3.63 mmol/ L]) were associated with increased mortality risk (relative risk, 1.36; P ⬍ 0.0001). Low cholesterol levels in dialysis patients likely are caused by systemic inflammation and malnutrition, conditions associated with increased mortality.17 It appears that normalizing lipid levels is not the primary mechanism of action for statins in this population. Emerging data suggest that statins possess antiproliferative, anti-inflammatory, antioxidant, anticoagulant, and immunosuppressive properties and improve endothelial function and blood pressure.9,18 It is possible that these effects could be of specific benefit in the clinical setting of dialysis and could contribute to noncardiac mortality reduction. Despite a high prevalence of cardiovascular risk factors in dialysis patients, statins are prescribed infrequently. Only 11.8% of dialysis patients were prescribed statins, whereas the K/DOQI work group estimated that more than 60% of dialysis patients require hyperlipidemia treatment.5 Of patients with seemingly clear indications for statin therapy (cholesterol ⬎200 mg/dL [⬎5.17 mmol/L] and a history of CAD or MI), only 15.7% to 22% were prescribed the recommended therapy. Even in the United States, France, and Germany (the countries with the highest use), less than 28% of these at-risk patients were prescribed statins, a
125
finding also reported by other investigators.9,19,20 The overall prescription pattern appeared to reflect the facility-practice pattern with regard to statin prescription. One factor contributing to low statin prescription may have been the lack of evidence supporting the benefits of statins in dialysis patients, especially at the time of this data collection in early 2000. The K/DOQI guidelines on dyslipidemia management (published April 2003) note the absence of large, randomized, controlled, interventional studies that associate dyslipidemias with cardiovascular disease in patients with chronic kidney disease.5 Evidence of the benefits of statin treatment derives mainly from studies of the general population and small controlled trials of simvastatin in HD patients21,22 and simvastatin and atorvastatin in dyslipidemic patients,21,23 as well as the observational study noted earlier.9 Although evidence is sparse, the proven benefits of statin therapy in the general population and the extremely high cardiovascular risk present in dialysis patients has prompted the current recommendations for statin use in dialysis patients.5,18 Statin prescription patterns also may be influenced by concerns about the financial burden and compliance issues that accompany the addition of medications to the regimens of dialysis patients. These patients bear the financial and emotional burdens of complex medication regimens, dietary restrictions, and the demands of dialysis. Although these concerns are valid, clinical and financial benefits of lower mortality (and presumably morbidity) risk associated with statin prescription may outweigh the cost and inconvenience of this therapy. The statin drug of choice varied greatly among the countries represented in this study. Simvastatin and/or pravastatin were used commonly in all countries, whereas atorvastatin was used only in the United States. Some of this variation likely was caused by drug availability, relative cost, health system policies, and drug marketing strategies in each country. Statin product choice also may have reflected preferential use of agents with less renal elimination. Atorvastatin and pravastatin do not accumulate in patients with renal failure and thus require no dose adjustment, whereas a dose reduction of 50% is recommended for the other agents in dialysis patients.5
126
MASON ET AL
In conclusion, statin prescription was associated with lower mortality in this large international study of HD patients, providing additional support for the value of statin therapy in this patient group. Although dialysis patients are at high risk for cardiovascular complications, very few are prescribed this therapy. This observational study suggests that more frequent use of statins may lead to significant improvements in dialysis patient outcomes. Ongoing randomized trials, such as A Study to Evaluate the Use of Rosuvastatin in Subjects On Regular Hemodialysis,24 Study of Heart and Renal Protection,25 and the Die Deutsche Diabetes Dialyse Studie,26 will provide additional insight into the clinical value of these agents in dialysis patients. ACKNOWLEDGMENT The authors acknowledge the contributions of Curtis A. Johnson, PharmD, and Wendy L. St. Peter, PharmD, to this project.
REFERENCES 1. US Renal Data System: USRDS 2002 Annual Data Report. The National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, 2002 2. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults: Executive Summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 285:2486-2497, 2001 3. Locatelli F, Bommer J, London GM, et al: Cardiovascular disease determinants in chronic renal failure: Clinical approach and treatment. Nephrol Dial Transplant 16:459-468, 2001 4. Wanner C, Krane V, Metzger T, Quaschning T: Lipid changes and statins in chronic renal insufficiency and dialysis. J Nephrol 14:S76-S80, 2001 (suppl 4) 5. National Kidney Foundation: K/DOQI Clinical Practice Guidelines for Managing Dyslipidemias in Chronic Kidney Disease. Am J Kidney Dis 41:S1-S92, 2003 (suppl 3) 6. Young EW, Goodkin DA, Mapes DL, et al: The Dialysis Outcomes and Practice Patterns Study (DOPPS): An international hemodialysis study. Kidney Int Suppl 74:S74-S81, 2000 7. Goodkin DA, Mapes DL, Held PJ: The Dialysis Outcomes and Practice Patterns Study (DOPPS): How can we improve the care of hemodialysis patients? Semin Dial 14:157-159, 2001 8. Klein JP, Moeschberger ML: Survival Analysis Techniques for Censored and Truncated Data. New York, NY, Springer, 1997, p 417 9. Seliger SL, Weiss NS, Gillen DL, et al: HMG-CoA reductase inhibitors are associated with reduced mortality in ESRD patients. Kidney Int 61:297-304, 2002 10. Scandinavian Simvastatin Survival Study Group: Randomised trial of cholesterol lowering in 4444 patients with
coronary heart disease: The Scandinavian Simvastatin Survival Study (4S). Lancet 344:1383-1389, 1994 11. Sacks FM, Pfeffer MA, Moye LA, et al: The effect of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels. N Engl J Med 335:1001-1009, 1996 12. Shepherd J, Cobbe SM, Ford I, et al: Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia. N Engl J Med 333:1301-1307, 1995 13. Heart Protection Study Collaborative Group: MRC/ BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: A randomised placebo-controlled trial. Lancet 360:7-22, 2002 14. Tonelli M, Sacks FM, Kiberd B, et al: Pravastatin for secondary prevention of cardiovascular events in persons with mild chronic renal insufficiency. Ann Intern Med 138: 98-104, 2003 15. Sever PS, Dahløf B, Poulter NR, et al: Prevention of coronary and stroke events with atorvastatin in hypertensive patients who have average or lower-than-average cholesterol concentrations, in the Anglo-Scandinavian Cardiac Outcomes Trial-Lipid Lowering Arm (ASCOT-LLA): A multicentre randomised controlled trial. Lancet 361:1149-1158, 2003 16. Holdaas H, Fellström B, Jardine AG, et al: Effect of fluvastatin on cardiac outcomes in renal transplant recipients: A multicentre, randomized, placebo controlled trial. Lancet 361:2024-2031, 2003 17. Liu Y, Coresh J, Eustace JA, et al: Association between cholesterol level and mortality in dialysis patients. JAMA 291:451-459, 2004 18. Seliger SL, Stehman-Breen CO. Are HMG-CoA reductase inhibitors underutilized in dialysis patients? Semin Dial 16:179-185, 2003 19. Longenecker JC, Coresh J, Powe NR, et al: Traditional cardiovascular disease risk factors in dialysis patients compared with the general population: The CHOICE study. J Am Soc Nephrol 13:1918-1927, 2002 20. Harris K, Thomas M, Short C, Moore R: Assessment of the efficiency of treatment of dyslipidaemia in renal outpatients. J Nephrol 15:263-269, 2002 21. Saltissi D, Morgan C, Rigby RJ, Westhuyzen J: Safety and efficacy of simvastatin in hypercholesterolemic patients undergoing chronic renal dialysis. Am J Kidney Dis 39:283-290, 2002 22. Chang JW, Yang WS, Min WK, et al: Effects of simvastatin on high-sensitivity C-reactive protein and serum albumin in hemodialysis patients. Am J Kidney Dis 39:12131217, 2002 23. Harris KP, Wheeler DC, Chong CC, Atorvastatin in CAPD Study Investigators: A placebo-controlled trial examining atorvastatin in dyslipidemic patients undergoing CAPD. Kidney Int 61:1469-1474, 2002 24. Fellstrom B, Holdaas H, Jardine A: Why do we need a statin trial in hemodialysis patients? Kidney Int Suppl 84: S204-S206, 2003 25. Baigent C, Landry M: Study of Heart and Renal Protection (SHARP). Kidney Int Suppl 84:S207-S210, 2003 26. Wanner C, Krane V, Ruf G, et al, for Die Deutsche Diabetes Dialyse Studie Investigators: Rationale and design of a trial improving outcome of type 2 diabetics on hemodialysis. Kidney Int Suppl 71:S222-S226, 1999
C
ARDIOVASCULAR DISEASE is the most common cause of mortality in patients with end-stage renal disease (ESRD), accounting for more than 40% of deaths in this population.1 Many factors contribute to this high rate of cardiac mortality, including traditional risk factors typically identified in the general population and renal-specific risk factors. Traditional risk factors include age, sex, family history of cardiovascular disease, hypertension (HTN), smoking, diabetes mellitus, dyslipidemia, obesity, and lack of physical activity.2 Additional renal failure– related risk factors potentially present in hemodialysis (HD) patients include HTN associated with volume overload, anemia, disordered metabolism of calcium and phosphorus, hyperparathyroidism, accumulation of metabolic wastes, chronic inflammation, and hypercatabolism.3 Dialysis treatment confers a tendency toward developing cardiac disease through the possibility of inadequate dialysis and bioincompatibility of HD membranes.3 As with the general population, some patients with ESRD have elevated low-density lipoprotein (LDL) cholesterol levels, but a nontraditional lipoprotein pattern also is seen in these
patients. This pattern is characterized by normal or low levels of LDL, a low level of high-density lipoprotein (HDL), a high triglyceride level, and
From the College of Pharmacy; Department of Biostatistics, School of Public Health; Division of Nephrology, Department of Veterans Affairs Medical Center, University of Michigan; Nephrology Pharmacy Associates; University Renal Research and Education Association, Ann Arbor, MI; Renal Research Institute Inc, New York; Albany College of Pharmacy, Albany, NY; Tokyo Women’s Medical University, Tokyo; Wakayama Medical University, Wakayama; Tokai University, Kanagawa, Japan; Hôpital St André, Bordeaux, France; and Birmingham Heartlands Hospital, Birmingham, UK. Received April 5, 2004; accepted in revised form September 1, 2004. The Dialysis Outcomes and Practice Patterns Study is supported by research grants from Amgen Inc and Kirin Brewery Ltd without restrictions on publications. This project was funded in part by Renal Research Institute Inc. Presented as an abstract at the 2001 ASN/ISN World Congress of Nephrology, San Francisco, CA, October 2001. Address reprint requests to Nancy A. Mason, PharmD, University of Michigan College of Pharmacy, 428 Church St, Ann Arbor, MI 48109-1065. E-mail: [email protected] © 2004 by the National Kidney Foundation, Inc. 0272-6386/04/4501-0013$30.00/0 doi:10.1053/j.ajkd.2004.09.025
American Journal of Kidney Diseases, Vol 45, No 1 (January), 2005: pp 119 –126
119
120
MASON ET AL
elevated non-HDL cholesterol levels, with elevated levels of lipoprotein remnants (very-LDL and intermediate-density lipoprotein) and small dense LDL particles that are considered highly atherogenic.4 Unfortunately, no randomized controlled studies show that treating dyslipidemias reduces the incidence of atherosclerotic cardiovascular disease in dialysis patients. Based on observational data and the strength of evidence supporting the benefits of treatment in the general population, the National Kidney Foundation– Kidney Disease Outcomes Quality Initiative (K/ DOQI) work group, responsible for developing clinical practice guidelines for managing dyslipidemias in patients with chronic kidney disease, recommended the use of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins) as first-line drug therapy for high LDL and nonHDL cholesterol levels in patients with chronic kidney disease.5 This study examines statin prescription patterns and associated outcomes in a large international HD patient population, using data from the Dialysis Outcomes and Practice Patterns Study (DOPPS I). METHODS DOPPS I is a large, prospective, longitudinal, and observational study of HD prescription practices and outcomes of patients randomly selected from representative dialysis facilities in 3 continents (5 countries in Europe: France, Germany, Italy, Spain, and the United Kingdom; Japan; and the United States). A stratified random sample of dialysis facilities was selected in each country. A census of prevalent HD patients older than 17 years was used to randomly select 20 to 40 patients from each selected facility. This study has been described previously.6,7 Data have been longitudinally collected since 1996 for the United States (148 facilities), 1998 for Europe (101 facilities), and 1999 for Japan (65 facilities). Follow-up continued through mid-2001. Some patients were lost to death, transplantation, change in treatment modality, withdrawal from dialysis therapy, recovery of renal function, or transfer to another facility. These patients were replaced on an ongoing basis. Briefly, the major goal of the DOPPS is to investigate the effect of HD practice patterns on patient outcomes. Primary study end points for the DOPPS are mortality, hospitalization, vascular access outcomes, and quality of life. Data are collected for each participating patient by using survey questionnaires administered by a study coordinator at each facility. Data include patient demographic characteristics, medical history, laboratory values, dialysis unit practices and outcomes, and prescribed over-the-counter and prescription drugs. The sample size was planned to give adequate
statistical power to analyze the associations between practice patterns and the 4 outcomes noted. Prescribed statin treatment practice was evaluated using DOPPS I data from the United States, Europe, and Japan. All medications listed in the medical record (both prescription and nonprescription) were recorded for each patient at entry into the study and at 4-month follow-up intervals and coded into the appropriate drug class, such as statins. Actual consumption of prescribed medicines was not recorded.
Statistical Methods For analysis, data included 9,846 patients in the United States, 4,591 patients in Europe, and 2,784 patients in Japan. Descriptive statistics for statin prescription and total cholesterol levels were computed based on a prevalent crosssection of patients (n ⫽ 7,365) between January and May 2000. Survey sampling weights were used to adjust for the cluster sampling of patients within facilities. Logistic regression was used to investigate predictors of statin prescription, based on a single record per person between January and May 2000. Covariate effects are presented as adjusted odds ratios (ORs). Potential predictors investigated include age, sex, race, years with ESRD, summary comorbid conditions (coronary artery disease [CAD], congestive heart failure, cardiovascular disease, HTN, peripheral vascular disease, and diabetes), and physician attitude toward hyperlipidemia treatment. Indicators were used to account for missing variables. Models took into account facility clustering with generalized estimating equations. Statin prescription also was investigated as a predictor of all-cause mortality, cardiac event, and cardiac mortality. Cox regression models, with study entry as the time origin and a time-dependent covariate indicating statin prescription, were used to estimate the relative risk (with versus without statins) for all-cause mortality, cardiac event, and cardiac mortality in 3 patient groups: all patients (n ⫽ 17,221), patients with CAD (n ⫽ 6,637), and patients with HTN (n ⫽ 12,980). All Cox regression models were stratified by country and adjusted for age, sex, race, body mass index, predialysis serum concentrations of albumin and hemoglobin, years with ESRD, normalized protein catabolic rate, and 15 summary comorbid conditions present at study entry (CAD, congestive heart failure, other cardiovascular disease, HTN, cerebrovascular disease, peripheral vascular disease, diabetes mellitus, lung disease, cancer [excluding skin], human immunodeficiency virus/acquired immunodeficiency syndrome, gastrointestinal bleeding, neurological disease, psychiatric disease, recurrent cellulitis, and dyspnea). Smoking was not significant after adjusting for comorbidities and was not included in the models. Facility clustering effects were addressed with robust SE estimates based on the sandwich estimator.8 The association between variation in facility-level statin prescription and mortality was examined by using Cox regression. Facility-level statin prescription is defined as the percentage of patients in the facility prescribed a statin at study entry. All statistical analyses were performed using SAS software, version 8.2 (SAS Institute, Cary, NC).
STATINS IN HEMODIALYSIS
121
Fig 1. Percentage of patients with a statin prescription among a prevalent cross-section of hemodialysis patients (n ⴝ 7,365) by country and date of reported prescription.
RESULTS
1). Of patients with a serum cholesterol level of 200 mg/dL or greater (ⱖ5.17 mmol/L), 15.7% overall were prescribed statins, whereas 17.5% of those with CAD and 22.0% of those with a history of myocardial infarction (MI) were prescribed these agents. There was considerable intercountry variability in the relative proportion of dialysis patients with each high-risk comorbid condition (high cholesterol level, CAD, or MI) and percentage of statin prescription in these specific subpopulations. However, in all countries, less than 28% of patients in any given high-risk category were treated with statins. Among patients treated with a statin in early 2000, simvastatin was prescribed for the majority of patients (58.4% to 72.3%) in France, Italy, and Spain. At the same time, pravastatin was the
In a prevalent cross-section of HD patients (n ⫽ 7,365), prescription of statins varied depending on the patient’s country of residence and date of reported prescription (Fig 1). The United States, France, and Germany reported greater prescription of statins than the other 4 countries, with the proportion of patients prescribed a statin increasing steadily during the study period for all countries except Spain. Other European countries and Japan had lower frequencies of statin prescription and tended to have smaller increases in the prescription of these agents over time. Taking all countries together, in early 2000 (the most recent year for which comparative data for all countries were available), statins were prescribed for 11.8% of HD patients overall (Table
Table 1. Percentage of Patients With High Cholesterol Levels, CAD, or MI and Percentage Prescribed Statins (January 2000 to May 2000) Patients With Condition Receiving Statins (%)
Patients With (%)
Country
Statin Prescription
Cholesterol ⱖ200 mg/dL*
CAD History
MI History
Cholesterol ⱖ200 mg/dL*
CAD History
MI History
France Germany Italy Japan Spain United Kingdom United States Overall
15.7 11.9 3.5 7.1 4.9 8.1 16.6 11.8
43.5 59.2 33.5 15.0 34.1 37.2 18.3 20.8
28.4 42.2 19.8 18.5 22.4 30.2 46.1 33.4
9.9 13.6 6.1 5.1 6.1 9.0 15.8 10.8
12.9 2.7 5.0 11.9 5.8 3.9 27.5 15.7
24.0 14.5 4.4 6.2 5.3 10.5 22.3 17.5
26.4 20.0 7.6 5.0 4.2 15.4 27.5 22.0
*Equivalent to 5.17 mmol/L.
122
MASON ET AL
Fig 2. Distribution of facility-level statin prescription (January 2000 to May 2000).
most commonly prescribed agent in Germany and Japan (51.4% and 58.1%, respectively). Atorvastatin was the most commonly prescribed agent in the United States (44.9%), with simvastatin the second most common (32.7%). The United Kingdom reported nearly equal prescription of pravastatin (48.5%) and simvastatin (49.7%). Analysis of statin prescription patterns within facilities showed that, from January 2000 to May 2000, a large majority of facilities (81.2%) prescribed statins to 20% or less of their patients (Fig 2). Some 16.5% reported no patients administered statins in their facilities, whereas a small
proportion (⬍1%) of facilities prescribed statins to more than 50% of their patients. Several demographic factors and comorbid conditions were examined to determine their association with statin prescription at study entry (Table 2). Based on multivariate analysis, patients were significantly less likely to be prescribed a statin if they were older (OR, 0.89/10year increase in age; P ⫽ 0.01), male (OR, 0.74; P ⫽ 0.04), or had been on dialysis therapy for a longer period (OR, 0.95/1-y increase in time on dialysis therapy; P ⫽ 0.001). There was a greater likelihood of a patient being prescribed a statin in
Table 2. Association of Demographic Factors and Comorbid Conditions With Statin Prescription at Study Entry Mean or Percentage Characteristic
Demographics Age (/10-y increase) Black (v other) Male (v female) Time with ESRD (/1-y increase) Comorbidities (yes v no) Coronary artery disease Congestive heart failure Other cardiovascular disease Hypertension Peripheral vascular disease Diabetes Facility characteristics Hospital (v nonhospital based) Aggressive hyperlipidemia therapy†
Statin
No Statin
OR*
P
61.6 13.4 45.0 4.6
59.7 17.9 57.5 5.4
0.89 0.65 0.74 0.95
0.01 0.12 0.04 0.001
46.5 32.2 34.4 78.9 28.5 43.4
35.3 29.3 33.1 73.1 20.8 32.4
1.55 0.84 0.8 1.1 1.52 1.39
0.01 0.47 0.21 0.56 0.03 0.05
55.7 62.3
54.8 50.9
0.83 1.84
0.31 0.0006
NOTE. Mean or percentage and 3-hydroxy-3 methylglutamyl coenzyme A inhibitor use was based on a point-prevalent sample from January 2000 to May 2000 (n ⫽ 8,615). Models (not shown) also examined factors of income, education, facility size, physician-patient contact time, and other comorbid conditions. *OR adjusted for all variables listed in the table, as well as for country. †Facilities with medical directors who agreed with the statement, “Hyperlipidemia is aggressively sought and treated in our patients.”
STATINS IN HEMODIALYSIS
123
Table 3. Association of Mortality and Cardiac Events With Statin Prescription Outcome
All-cause mortality All patients Patients with CAD Patients with HTN Cardiac event* All patients Patients with CAD Patients with HTN Cardiac mortality† All patients Patients with CAD Patients with HTN Noncardiac mortality All patients Patients with CAD Patients with HTN
Hazard Ratio
P
95% Confidence Interval
0.69 0.70 0.70
⬍0.0001 ⬍0.0001 ⬍0.0001
0.60-0.79 0.60-0.83 0.61-0.80
0.78 0.83 0.72
0.03 0.14 0.01
0.62-0.98 0.64-1.06 0.57-0.93
0.77 0.78 0.71
0.03 0.05 0.01
0.61-0.97 0.60-1.00 0.56-0.91
0.56 0.53 0.59
⬍0.0001 ⬍0.0001 ⬍0.0001
0.46-0.69 0.40-0.69 0.47-0.73
NOTE. Hazard ratios were estimated using Cox regression models with a time-dependent covariate indicating statin prescription. All models were adjusted for age, sex, race, body mass index, serum albumin level, hemoglobin level, time with ESRD, normalized protein catabolic rate, and summary comorbid conditions. *Cardiac event indicates patient death or hospitalization because of coronary disease. †Cardiac mortality indicates patient death from coronary disease.
the presence of CAD (OR, 1.55; P ⫽ 0.01), peripheral vascular disease (OR, 1.52; P ⫽ 0.03), or diabetes (OR, 1.39; P ⫽ 0.05). There also was an increased likelihood of statin use in facilities with medical directors who agreed with the statement, “Hyperlipidemia is aggressively sought and treated in our patients” (OR, 1.84; P ⫽ 0.0006). Other factors, such as income, education, facility size, physician-patient contact time, and other comorbid conditions, were not significant in explaining the variation in statin prescription.
Fig 3. Adjusted survival curve for all-cause mortality for patients prescribed versus not prescribed statins. Estimates were calculated from a Cox regression model adjusted for age, sex, race, body mass index, serum albumin level, hemoglobin level, time with ESRD, normalized protein catabolic rate, and summary comorbid conditions. Statin use was modeled as a time-varying covariate, although the curves presented here represent no statin use versus continuous statin use.
Examination of all-cause mortality data showed that patients prescribed statins had a 31% lower relative risk for death compared with those not prescribed statins (P ⬍ 0.0001) after adjusting for demographics and comorbidities (Table 3; Fig 3). A significantly lower mortality risk (30%; P ⬍ 0.0001) extended to each subgroup with CAD and with HTN. Risk for death or hospitalization because of coronary disease also was significantly lower among the HD population as a whole and the subgroup with HTN. Statins were associated with a 23% lower cardiac mortal-
124
MASON ET AL Table 4. Association of Mortality and Cardiac Events With Facility Statin Prescription
Outcome (/10% increase in facility statin prescription)
All-cause mortality All patients Patients with CAD Patients with HTN Cardiac event* All patients Patients with CAD Patients with HTN Cardiac mortality† All patients Patients with CAD Patients with HTN Noncardiac mortality All patients Patients with CAD Patients with HTN
Hazard Ratio
P
95% Confidence Interval
0.95 0.94 0.95
0.02 0.05 0.03
0.90-0.99 0.89-1.00 0.90-0.99
0.93 0.93 0.92
0.09 0.19 0.07
0.86-1.01 0.85-1.03 0.84-1.01
0.91 0.91 0.90
0.02 0.07 0.01
0.84-0.99 0.82-1.01 0.82-0.98
0.97 0.95 0.96
0.40 0.23 0.25
0.91-1.04 0.87-1.04 0.89-1.03
NOTE. Hazard ratios were estimated using Cox regression models with a time-dependent covariate indicating facility statin prescription. All models adjusted for age, sex, race, body mass index, serum albumin level, hemoglobin level, time with ESRD, normalized protein catabolic rate, and summary comorbid conditions. Facility statin prescription was defined as the percentage of patients prescribed a statin in the facility. *Cardiac event indicates patient death or hospitalization caused by coronary disease. †Cardiac mortality indicates patient death from coronary disease.
ity risk (P ⫽ 0.03) and 44% lower noncardiac mortality risk (P ⬍ 0.0001) in this patient population. The lower all-cause mortality associated with statin use was found in all 7 countries included in the analysis. The magnitude of the lower risk ranged from 19% to 83%. No significant difference in the statin effect among countries was found (interaction P ⫽ 0.32). The association between facility statin prescription and outcomes also was examined (Table 4). The practice of prescribing statins was associated with a lower overall mortality rate for all patients, with a 5% lower risk for every 10% increase in number of patients prescribed statins in the facility (P ⫽ 0.02). Risk for death or hospitalization because of coronary disease also was significantly lower in the HD population as a whole. At the facility level, statins were associated with a 9% lower cardiac mortality risk for every 10% increase in number of patients prescribed statins in the facility (P ⫽ 0.02), whereas noncardiac mortality was not significantly affected. DISCUSSION
Analysis of a prevalent cross-section of 7,365 HD patients showed that statin prescription was
associated with 31% lower all-cause mortality and 23% lower risk for cardiac mortality in HD patients. The lower mortality risk found in this study supports earlier observations from 3,700 incident dialysis patients in the US Renal Data System Dialysis Morbidity and Mortality Study Wave 2 from 1996 to 1998. In that study, Seliger et al9 found 32% lower all-cause mortality and 36% lower cardiovascular mortality. The magnitude of the lower mortality risk observed in these 2 studies is similar to that seen in large, randomized, controlled trials of the general population10-12 and studies including patient cohorts with impaired renal function.13-16 Observational data are subject to some inherent limitations, including treatment bias. One potential explanation for these results is that patients could have been selected for statin treatment because they were expected to survive long enough to benefit. For instance, in clinical practice, it is possible that within the population of patients with CAD, for whom outcomes will be adjusted similarly, only those judged likely to survive long enough to benefit may be prescribed statins. Hence, there would be bias within the CAD group for statins to be administered to those with a better prognosis. However, the facil-
STATINS IN HEMODIALYSIS
ity-level association would less likely be affected by this patient-level effect and therefore supports the conclusion that this bias is not a major factor. Moreover, it is at least as likely that statin therapy would be administered preferentially to patients with more severe coronary disease, potentially creating bias in the opposite direction. Statin prescription also was associated with a remarkable 44% lower risk for noncardiac mortality. The potential protective mechanisms associated with statins have been reviewed in detail elsewhere and probably reflect a combination of factors in addition to normalization of lipid profile. If the survival benefit of statin use were caused entirely by normalization of lipid levels, then adjusting for cholesterol level would reduce or eliminate the significance of statins in a statistical model. However, with our data, adjusting for lipid levels did not reduce the effect of statins. In addition, in a model including cholesterol level but omitting statins, only very low cholesterol levels (⬍140 mg/dL [⬍3.63 mmol/ L]) were associated with increased mortality risk (relative risk, 1.36; P ⬍ 0.0001). Low cholesterol levels in dialysis patients likely are caused by systemic inflammation and malnutrition, conditions associated with increased mortality.17 It appears that normalizing lipid levels is not the primary mechanism of action for statins in this population. Emerging data suggest that statins possess antiproliferative, anti-inflammatory, antioxidant, anticoagulant, and immunosuppressive properties and improve endothelial function and blood pressure.9,18 It is possible that these effects could be of specific benefit in the clinical setting of dialysis and could contribute to noncardiac mortality reduction. Despite a high prevalence of cardiovascular risk factors in dialysis patients, statins are prescribed infrequently. Only 11.8% of dialysis patients were prescribed statins, whereas the K/DOQI work group estimated that more than 60% of dialysis patients require hyperlipidemia treatment.5 Of patients with seemingly clear indications for statin therapy (cholesterol ⬎200 mg/dL [⬎5.17 mmol/L] and a history of CAD or MI), only 15.7% to 22% were prescribed the recommended therapy. Even in the United States, France, and Germany (the countries with the highest use), less than 28% of these at-risk patients were prescribed statins, a
125
finding also reported by other investigators.9,19,20 The overall prescription pattern appeared to reflect the facility-practice pattern with regard to statin prescription. One factor contributing to low statin prescription may have been the lack of evidence supporting the benefits of statins in dialysis patients, especially at the time of this data collection in early 2000. The K/DOQI guidelines on dyslipidemia management (published April 2003) note the absence of large, randomized, controlled, interventional studies that associate dyslipidemias with cardiovascular disease in patients with chronic kidney disease.5 Evidence of the benefits of statin treatment derives mainly from studies of the general population and small controlled trials of simvastatin in HD patients21,22 and simvastatin and atorvastatin in dyslipidemic patients,21,23 as well as the observational study noted earlier.9 Although evidence is sparse, the proven benefits of statin therapy in the general population and the extremely high cardiovascular risk present in dialysis patients has prompted the current recommendations for statin use in dialysis patients.5,18 Statin prescription patterns also may be influenced by concerns about the financial burden and compliance issues that accompany the addition of medications to the regimens of dialysis patients. These patients bear the financial and emotional burdens of complex medication regimens, dietary restrictions, and the demands of dialysis. Although these concerns are valid, clinical and financial benefits of lower mortality (and presumably morbidity) risk associated with statin prescription may outweigh the cost and inconvenience of this therapy. The statin drug of choice varied greatly among the countries represented in this study. Simvastatin and/or pravastatin were used commonly in all countries, whereas atorvastatin was used only in the United States. Some of this variation likely was caused by drug availability, relative cost, health system policies, and drug marketing strategies in each country. Statin product choice also may have reflected preferential use of agents with less renal elimination. Atorvastatin and pravastatin do not accumulate in patients with renal failure and thus require no dose adjustment, whereas a dose reduction of 50% is recommended for the other agents in dialysis patients.5
126
MASON ET AL
In conclusion, statin prescription was associated with lower mortality in this large international study of HD patients, providing additional support for the value of statin therapy in this patient group. Although dialysis patients are at high risk for cardiovascular complications, very few are prescribed this therapy. This observational study suggests that more frequent use of statins may lead to significant improvements in dialysis patient outcomes. Ongoing randomized trials, such as A Study to Evaluate the Use of Rosuvastatin in Subjects On Regular Hemodialysis,24 Study of Heart and Renal Protection,25 and the Die Deutsche Diabetes Dialyse Studie,26 will provide additional insight into the clinical value of these agents in dialysis patients. ACKNOWLEDGMENT The authors acknowledge the contributions of Curtis A. Johnson, PharmD, and Wendy L. St. Peter, PharmD, to this project.
REFERENCES 1. US Renal Data System: USRDS 2002 Annual Data Report. The National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, 2002 2. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults: Executive Summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 285:2486-2497, 2001 3. Locatelli F, Bommer J, London GM, et al: Cardiovascular disease determinants in chronic renal failure: Clinical approach and treatment. Nephrol Dial Transplant 16:459-468, 2001 4. Wanner C, Krane V, Metzger T, Quaschning T: Lipid changes and statins in chronic renal insufficiency and dialysis. J Nephrol 14:S76-S80, 2001 (suppl 4) 5. National Kidney Foundation: K/DOQI Clinical Practice Guidelines for Managing Dyslipidemias in Chronic Kidney Disease. Am J Kidney Dis 41:S1-S92, 2003 (suppl 3) 6. Young EW, Goodkin DA, Mapes DL, et al: The Dialysis Outcomes and Practice Patterns Study (DOPPS): An international hemodialysis study. Kidney Int Suppl 74:S74-S81, 2000 7. Goodkin DA, Mapes DL, Held PJ: The Dialysis Outcomes and Practice Patterns Study (DOPPS): How can we improve the care of hemodialysis patients? Semin Dial 14:157-159, 2001 8. Klein JP, Moeschberger ML: Survival Analysis Techniques for Censored and Truncated Data. New York, NY, Springer, 1997, p 417 9. Seliger SL, Weiss NS, Gillen DL, et al: HMG-CoA reductase inhibitors are associated with reduced mortality in ESRD patients. Kidney Int 61:297-304, 2002 10. Scandinavian Simvastatin Survival Study Group: Randomised trial of cholesterol lowering in 4444 patients with
coronary heart disease: The Scandinavian Simvastatin Survival Study (4S). Lancet 344:1383-1389, 1994 11. Sacks FM, Pfeffer MA, Moye LA, et al: The effect of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels. N Engl J Med 335:1001-1009, 1996 12. Shepherd J, Cobbe SM, Ford I, et al: Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia. N Engl J Med 333:1301-1307, 1995 13. Heart Protection Study Collaborative Group: MRC/ BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: A randomised placebo-controlled trial. Lancet 360:7-22, 2002 14. Tonelli M, Sacks FM, Kiberd B, et al: Pravastatin for secondary prevention of cardiovascular events in persons with mild chronic renal insufficiency. Ann Intern Med 138: 98-104, 2003 15. Sever PS, Dahløf B, Poulter NR, et al: Prevention of coronary and stroke events with atorvastatin in hypertensive patients who have average or lower-than-average cholesterol concentrations, in the Anglo-Scandinavian Cardiac Outcomes Trial-Lipid Lowering Arm (ASCOT-LLA): A multicentre randomised controlled trial. Lancet 361:1149-1158, 2003 16. Holdaas H, Fellström B, Jardine AG, et al: Effect of fluvastatin on cardiac outcomes in renal transplant recipients: A multicentre, randomized, placebo controlled trial. Lancet 361:2024-2031, 2003 17. Liu Y, Coresh J, Eustace JA, et al: Association between cholesterol level and mortality in dialysis patients. JAMA 291:451-459, 2004 18. Seliger SL, Stehman-Breen CO. Are HMG-CoA reductase inhibitors underutilized in dialysis patients? Semin Dial 16:179-185, 2003 19. Longenecker JC, Coresh J, Powe NR, et al: Traditional cardiovascular disease risk factors in dialysis patients compared with the general population: The CHOICE study. J Am Soc Nephrol 13:1918-1927, 2002 20. Harris K, Thomas M, Short C, Moore R: Assessment of the efficiency of treatment of dyslipidaemia in renal outpatients. J Nephrol 15:263-269, 2002 21. Saltissi D, Morgan C, Rigby RJ, Westhuyzen J: Safety and efficacy of simvastatin in hypercholesterolemic patients undergoing chronic renal dialysis. Am J Kidney Dis 39:283-290, 2002 22. Chang JW, Yang WS, Min WK, et al: Effects of simvastatin on high-sensitivity C-reactive protein and serum albumin in hemodialysis patients. Am J Kidney Dis 39:12131217, 2002 23. Harris KP, Wheeler DC, Chong CC, Atorvastatin in CAPD Study Investigators: A placebo-controlled trial examining atorvastatin in dyslipidemic patients undergoing CAPD. Kidney Int 61:1469-1474, 2002 24. Fellstrom B, Holdaas H, Jardine A: Why do we need a statin trial in hemodialysis patients? Kidney Int Suppl 84: S204-S206, 2003 25. Baigent C, Landry M: Study of Heart and Renal Protection (SHARP). Kidney Int Suppl 84:S207-S210, 2003 26. Wanner C, Krane V, Ruf G, et al, for Die Deutsche Diabetes Dialyse Studie Investigators: Rationale and design of a trial improving outcome of type 2 diabetics on hemodialysis. Kidney Int Suppl 71:S222-S226, 1999