Anemia: An Emerging Cardiovascular Risk Factor and Endocrine Disorder

Traditional and Nontraditional Cardiovascular Risk Factors

Marc A. Pfeffer, MD, PhD (Har-vard Medical School) observed that many clinical studies, some as old as 50 years and others much more recent, have illuminated the traditional risk factors associated with cardiovascular disease. The Framingham Heart Study, for example, has demonstrated that no single risk factor, such as hyperlipidemia or diabetes, is predictive, giving rise to the concept of an individual’s total risk-factor profile.

The Survival and Ventricular Enlargement (SAVE) trial indicated that recurrent myocardial infarction (MI) and the need for cardiac revascularization were both reduced in patients treated with the angiotensin converting enzyme (ACE) blocker captopril. This suggests that ACE inhibitors either have an anti-ischemic effect or that they possess the ability to modify the atherosclerotic process in MI survivors (Rutherford JD et al. Circulation. 1994;90(4):1731).

The international Valsartan in Acute Myocardial Infarction (VALIANT) trial evaluated the effectiveness of angiotensin receptor blocker (ARB) in preventing recurrent MI in 14,703 survivors of MI with left ventricular systolic dysfunction. Valsartan was determined to be as efficacious as captopril for preventing recurrent MI (Pfeffer MA et al. N Engl J Med. 2003;349:1893). Within the study population, approximately 3,400 individuals had previously-diagnosed diabetes, while an additional 580 individuals had been newly diagnosed with diabetes at the time of randomization. The newly-diagnosed diabetic patients had higher morbidity and mortality rates compared with both the nondiabetic and previously-diagnosed patients (Aguilar D et al. Circulation. 2004; in press).

The placebo-controlled and randomized Candesartan in Heart Failure Assessment of Reduction in Mortality and Morbidity (CHARM-Alternative) trial evaluated the ARB candesartan in 7,028 patients with chronic heart failure. Candesartan was demonstrated both to be well tolerated and to reduce cardiovascular morbidity and mortality in patients with reduced ejection fraction intolerant to an ACE inhibitor (Granger CB et al. Lancet. 2003;362 (9386):772). Candesartan reduced hospital admissions in the CHARM-Preserved trial of 3,023 patients with chronic heart failure and left ventricular ejection fractions higher than 40% (Yusuf S et al. Lancet. 2003;363(9386): 777). The CHARM-Added trial then demonstrated that treatment with candesartan in combination with ACE inhibition or other relevant treatments can result in a clinically important reduction in cardiovascular events in a similar population (McMurray JJ et al. Lancet. 2003;362(9386):767).

The association of progressive renal disease with cardiovascular risk is also well established. Any degree of renal insufficiency markedly increases the risk of death up to the point at which a 30-year-old patient on dialysis is at higher risk for death than a healthy person at 85 years of age. Prior to the need for dialysis, there is a step-wise risk of death that parallels increases in creatinine clearance. Thus any treatment that improves creatinine clearance may reduce cardiovascular risk. In the SAVE study, ACE inhibition was associated with a marked reduction in creatinine clearance and significant clinical benefit. The VALIANT trial demonstrated a similar benefit from angiotensin receptor blockade.

The presence of multiple risk factors greatly increases an individual’s probability of cardiovascular morbidity and mortality. The combination of diabetes, anemia, and chronic renal disease is especially predictive of risk, with approximately a 50-fold increase. It has been established that patients with diabetes who have normal hematocrits appear to fare better than those with low hematocrits. Renal failure requiring dialysis in combination with anemia greatly increases the risk for cardiovascular death. Because of the additive effects of risk factors on long-term survival, aggressive multifactorial intervention should be considered as a means of cardiovascular prophylaxis. To evaluate this hypothesis, the recent Steno-2 (Copenhagen) study of patients with type-2 diabetes and microalbuminuria randomized patients to standard treatments (n=80) or to aggressive therapy (N=80). Medication use was targeted to hyperglycemia, hypertension, dyslipidemia, and microalbuminuria while providing secondary prevention of cardiovascular disease with aspirin. Target-driven, long-term, intensified intervention aimed at multiple risk factors reduced the risk of cardiovascular disease, nephropathy, retinopathy, and autonomic neuropathy by 50% (Gaede P et al. N Engl J Med. 2330;348(5):383).

Recent investigations have established a link between C-reactive protein (CRP), a marker of systemic inflammation and a nontraditional factor, and risk for cardiovascular disease. Recent findings in 27,939 healthy women in the Women’s Health Study indicate that CRP is a stronger predictor of risk than low-density lipoprotein (LDL) cholesterol elevation, that it predicts risk for cardiovascular disease in individuals without overt hyperlipidemia, and that it adds prognostic information to risk scoring and LDL cholesterol categories and to the diagnosis of metabolic syndrome (Ridker PM. N Engl J Med. 2002; 347:1557). At present a large-scale, randomized clinical trial entitled Justification for the Use of Statins in Primary Prevention: An Intervention Trial Evaluating Rosuvastatin (JUPITER) is gathering data to evaluate the efficacy of long-term statin therapy for reducing incident cardiovascular disease in subjects with elevated plasma CRP concentrations who do not meet current criteria for initiation of lipid-lowering drug intervention (Ridker PM. Circulation. 2003;2292; Libby P, Ridker PM. Am J Med. 2004;116[suppl 6A]:S9).

Anemia and Diabetes: Prevalence and Predictors

Merlin C. Thomas, MBChB, PhD (Melbourne, Australia) opened his presentation by noting that “diabetes is the leading cause of end-stage renal disease and is, therefore, the leading cause of renal anemia, which has been under-recognized as a risk factor for cardiovascular morbidity and mortality.” Moreover, patients with diabetes with renal impairment are almost twice as likely to have anemia as nondiabetic patients with the same level of renal function (Thomas M et al. Diabetes Care. 2003;26:1164). In addition to occurring more commonly in individuals with diabetes, anemia occurs earlier in patients with diabetes and renal disease. These findings pose the question of whether or not anemia is one of the causes of increased cardiovascular risk in patients with diabetes.

In order to explore this question, Dr. Thomas and colleagues conducted a cross-sectional study of 1,125 patients with type 2 diabetes of at least 10 years’ duration. Approximately one-third of subjects had albuminuria, one-third had moderately impaired renal function, and 40% had macrovascular disease. Using the WHO definition of anemia (male and female hemoglobin levels of 13 gm/dL and 12 gm/dL, respectively), 23% of the study population had anemia (Thomas M et al. Nephrol Dial Transplant. 2004;19:1792). Using the Kidney Disease Outcomes Quality Initiative (K/DOQI) guidelines definition of anemia (hemoglobin less than 11 gm/dL), 8% of the study population had anemia.

Renal function correlates with hemoglobin levels in diabetes. As the glomerular filtration rate (GFR) falls, the hemoglobin level also drops (Thomas MC et al. Diabetes Care. 2003;26:1164). Age has also been correlated with declines in hemoglobin in diabetic patients, because with aging there is an increased prevalence of renal impairment and of anemia. However, because individuals over 80 years of age have relatively little muscle mass, creatinine may remain in the normal range even as GFR decreases. Thus when data were corrected for patient age in the Australian study, age was not associated with an in-creased prevalence of anemia but with renal impairment. Therefore, the anemia of aging is due largely to a progressive age-related decline in renal function. In contrast, there is a direct correlation between declining GFR and the prevalence of anemia.

Individuals with normal GFR and creatinine clearance and even hyperfiltration may still be at risk for anemia if they have albuminuria, as patients with the highest levels of albuminuria may progress to renal impairment. In the Australian study, those patients with albuminuria were at the greatest risk for anemia. The prevalence of anemia among patients with both macroalbuminuria and renal impairment by laboratory values was approximately 60% even though few of them were progressing to end-stage renal disease.

The other important risk factor for low hemoglobin levels is decreasing iron stores, but it is not the reason for the high prevalence of anemia in diabetes. In fact, the rate of iron deficiency among patients with diabetes is equal to or less than that seen in the general population (Thomas MC et al. Diabetes Care. 2003;26:1164). The rate of iron excess is significantly higher in patients with diabetes despite the higher prevalence of anemia (Thomas MC et al. Diabetes Med 2004;21:798). The Australian study also attempted to evaluate which patients were likely to become anemic. Although ACE inhibitors, ARBs, and peroxisome-proliferator-activated receptors have all been implicated in this apparent paradox, no association was uncovered. Furthermore, there was a weak negative effect between HbA1c and the risk for developing anemia. It appears, therefore, that abnormal glycemic control is not responsible for the high prevalence of anemia in the type 2 diabetic population. Similar findings resulted from a study of patients with type 1 diabetes, in which anemia in 21% of patients was predicted by the same risk factors (Thomas MC et al. J Clin Endocrinol Metab. 2004).

If abnormal glucose metabolism is not at the root of anemia in diabetes, how can it be explained? Data point increasingly to erythropoietin. The dominant hypothesis is that as oxygen is delivered to the renal tissue, a sensing apparatus in the peritubal capillaries measures the flow of oxygen. A reduction in oxygen flow normally results in increased release of erythropoietin which, in turn, signals the bone marrow to accelerate the production of erythrocytes. In the kidneys of patients with diabetes, however, this stimulus-response sequence is absent. Thus anemia in diabetes is predominantly a renal disorder involving impaired signaling of erythropoietin release in response to decreased hemoglobin-borne oxygen flow.

Clinical Trials Evaluating Anemia Treatment in Patients at Risk for Cardiovascular Disease

Mark E. Cooper, MD, PhD (Melbourne, Australia) discussed anemia’s critical relationship with cardiovascular disease by presenting a variety of clinical trials evaluating anemia treatments in patients at risk for cardiovascular disease. The Reduction in Endpoints in Non-insulin Dependent Diabetes Mellitus with the Angiotensin II Antagonist Losartan (RENAAL) trial looked at 1,513 patients with type 2 diabetes and renal disease to evaluate various risk factors’ predictive effect on the progression of renal disease (Keane WE et al. Kidney Int. 2003;63:1499). The trial found that four risk factors were independent predictors of worsening renal function: proteinuria, serum creatinine, serum albuminuria, and hemoglobin. These findings suggest that low hemoglobin should now be considered a conventional risk factor.

As Figure 1 indicates, there is a linear relationship between the amount of hemoglobin loss and the time to first event of the trial’s principal endpoints: (i) doubling of serum creatinine, (ii) onset of end-stage renal disease or (iii) death. Ma and colleagues evaluated death rates for diabetic and nondiabetic subjects with anemia measured by hematocrit. They found that decreased hematocrit in diabetic subjects, regardless of the severity of decline, resulted in approximately a 40% increase in mortality (Ma JZ et al. JASN. 1999; 10:610). The Keane and Ma studies add compelling evidence that the risk of mortality from anemia is greatly amplified in diabetes, and that the combination of diabetes and anemia is a very strong predictor of mortality. Diabetes alone can increase the risk of death by 40% to 50%; the addition of anemia appears to impart an additive risk (Collins. Adv Stud in Med. 2003; 3:S14).

Echocardiography can provide evidence regarding the relationship between diabetes with anemia and cardiac mortality. One study thus far unpublished observed a wide range of cardiac anatomical abnormalities in diabetic patients including elevation in the left atrial area and increase in both left ventricular mass and the interventricular septum. These changes are indicative of both diastolic and systolic dysfunction. In patients with both anemia and diabetes, the investigators observed a higher prevalence of diastolic dysfunction in addition to simultaneous diastolic and systolic dysfunction. Patients with anemia also had a corresponding mean elevation of almost 50% in CRP and significant elevations in plasma concentrations of brain naturetic peptide (BNP), a sensitive marker of cardiac dysfunction that predicts mortality in several diseases including diabetes. In this study, BNP concentrations were not decreased when corrected for traditional factors associated with BNP elevation.

A survey of unselected men with diabetes evaluated the relationship between anemia and atherosclerotic disease. The investigators observed a direct relationship between decreasing hemoglobin and coronary heart disease, cerebral vascular disease, and peripheral vascular disease. The correlation remained evident when all forms of atherosclerotic disease were aggregated, with a prevalence of approximately 90% at hemoglobin levels below 11 gm/dL (Austin Medical Center Survey. Diabetes Care. 2003: 26:114).

When looked at from the other direction—the prevalence of anemia in patients with vascular disease rather than the prevalence of vascular disease in patients with anemia – the outcome is similar. Regardless of the form of vascular disease, there is at least a 50% increase in the prevalence of anemia (Thomas MC et al. Nephrol Dial Transplant. 2004;19:1792). The same study presented evidence that patients with either type 1 or type 2 diabetes and hemoglobin concentrations below 12 gm/dL have a two-fold increase in risk for background diabetic retinopathy and a five-fold increase in risk for prolif-
erative retinopathy. Comparison of retinopathy rates between diabetic patients with and without anemia in this study indicated a relative risk of 2.0 in the anemic cohort, confirming a strong association between anemia and progressive retinal disease. One hypothesis for this association is that anemia-related hypoxia may be a potent stimulus of the vascular endothelial growth factor (VEGF), one of the growth factors implicated in retinopathy.

The most compelling evidence to date linking comorbid anemia and diabetes with cardiovascular disease emerged from an Israeli study involving 84 patients with type 2 diabetes, anemia, chronic renal disease, and treatment-resistant congestive heart failure (Silverberg DS et al. Nephrol Dial Transplant. 2003;18:141). All patients continued on their current treatments for congestive heart failure and were treated additionally with erythropoietin and iron for anemia for one year. During the trial, the mean hemoglobin level rose from 10.5 gm/dL to 13.1 gm/dL. Treating anemia resulted in a dramatic and significant reduction in hospital admissions and in congestive heart failure as measured by New York Heart Association (NYHA) class. There was also a significant improvement in left ventricular ejection fraction. Figure 2 summarizes the study’s findings.

The evidence available to date is indicative of a potential improvement in cardiovascular disease with anemia correction in patients with diabetes. As yet, however, there have been no randomized and controlled trials to demonstrate this conclusively. Fortunately, two trials currently in progress may provide a basis for an answer. The European Anemia Correction in Diabetes (ACORD) trial randomizes 160 patients to standard treatment or to aggressive erythropoietin therapy. The primary endpoint is change in left ventricular mass after 15 months. Assuming that the evaluation is favorable, it is uncertain that this single endpoint would be a basis for the routine use of erythropoietin. More endpoint studies will be needed before deciding definitively. The Trial to Reduce Cardiovascular Events with Aranesp® Therapy (TREAT) is designed to evaluate the impact of anemia treatment with darbepoetin alfa on mortality and cardiovascular events in 4,000 patients with type 2 diabetes and chronic kidney disease. This is a placebo-controlled trial in which the treatment target is a hemoglobin concentration of 13 gm/dL. The primary endpoint is all-cause mortality which, in the diabetic population, is more than 80% cardiovascular. The secondary endpoint is nonfatal cardiac events. More than the ACORD trial, the breadth of endpoints of TREAT may provide sufficient evidence to determine whether or not erythropoietin therapy should become standard for diabetic patients with anemia.