Cardiovascular Disease in People with HIV

Introduction

In the past few years disturbing reports have begun to accumulate about young, relatively healthy HIV positive men who have developed heart disease or suffered heart attacks. Researchers and health-care providers treating people with HIV are taking these reports seriously, and increasing attention has been devoted to cardiovascular disease at recent medical conferences.

It remains to be determined whether these cardiovascular manifestations are attributable to HIV infection itself, to antiretroviral drug therapy, to a high prevalence among HIV positive people of traditional risk factors that pose a danger for the population at large -- such as tobacco smoking, older age, male sex, and family history -- or to some other not-yet-known factors.

Heart disease in people with HIV remains poorly understood and continues to generate considerable fear and controversy. But there is good reason to believe that common-sense risk reduction strategies can help delay or prevent cardiovascular problems such as atherosclerosis, heart attacks, and strokes. (For an explanation of cardiovascular conditions, see the sidebar.)

What's the Problem?

By 1998 health-care providers had begun to report on small numbers of young HIV positive people who had experienced coronary heart disease. Many researchers and advocates came to suspect that such events might be related to the elevated levels of blood fats -- in particular cholesterol and triglycerides -- that were being observed in people taking highly active antiretroviral therapy (HAART), especially the then-new protease inhibitor (PI) drugs. Numerous studies were undertaken in an attempt to understand the phenomenon.

Conflicting Results

Max David, MD, of the University of Cincinnati College of Medicine and colleagues reported in the January 1, 2002 issue of Clinical Infectious Diseases that duration of antiretroviral therapy was associated with ischemic heart disease in people with HIV. The researchers compared data from 16 HIV positive people with cardiovascular disease and 32 HIV positive people without heart disease. The group with cardiovascular disease had a lower mean CD4 cell nadir (lowest ever count) and longer exposure to nucleoside reverse transcriptase (NRTI) drugs; both factors may be markers for a longer duration of HIV infection. However, use of PI drugs was similar in both groups.

In contrast, M. Mary-Krause, MD, of INSERM in Paris and colleagues reported at the 8th Conference on Retroviruses and Opportunistic Infections (the Retrovirus conference) in February 2001 that in a French cohort, myocardial infarction rates increased with duration of PI use. Fifty-four heart attacks occurred among 19,795 HIV positive men who had taken PIs. The incidence rate was 8.9 per 10,000 person-years among those taking PIs for less than 18 months, 19.2 among those taking PIs for 18-29 months, and 34.7 among those taking the drugs for more than 30 months (although there were few people in the latter group); the expected incidence rate in the general population (matched for age and sex) is 10.8 per 10,000 person-years. Unfortunately, this study did not control for additional cardiovascular risk factors.

Michel Duong, MD, of Hôpital du Bocage in Dijon, France, and colleagues presented a study of heart disease risk factors at the 41st Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC) in December 2001. The researchers analyzed coronary heart disease risk factors in a group of 99 HIV positive people using HAART who had no known history of heart disease. Participants underwent an exercise stress test to detect silent myocardial ischemia. (In a stress test, a subject exercises on a treadmill while electrocardiogram and blood pressure tests are done.) The researchers found no statistically significant differences in type or duration of antiretroviral therapy, length of HIV infection, viral load, or CD4 cell count between those who had positive stress tests (that is, those with evidence of myocardial ischemia) and those with negative stress tests. Evidence of ischemia was statistically associated with older age, obesity, and elevated cholesterol level.

Retrovirus Conference Clarifies Little

The 9th Retrovirus conference in February 2002 did little to clarify matters, with different research teams presenting conflicting data. Scott Holmberg, MD, from the Centers for Disease Control and Prevention (CDC) and colleagues followed 5,676 HIV positive people at nine clinics (the HOPS cohort) from 1993 to 2001. Slightly more than half the group used PI-based regimens. Although uncommon overall, 13 of 3,013 participants taking PIs suffered myocardial infarctions during the study period, compared with just two of 2,663 participants not using PIs. Although people taking PIs were more than five times as likely to have a heart attack, no increased risk was observed for angina or strokes.

On the other hand, a study by Samuel Bozzette, MD, of the University of California at San Diego and colleagues detected what appears to be a slight decline in heart disease and strokes since the advent of HAART. Dr. Bozzette's team examined the medical records of 36,766 HIV positive patients at U.S. Veterans Administration medical centers over the same eight-year period. During this time there were 1,800 hospital admissions and 500 deaths due to cardiovascular or cerebrovascular disease. Since 1997 rates of hospital admissions and deaths due to heart attack or stroke have fallen compared to the rates prior to the availability of PIs; over the course of the study period, heart attack and stroke admission and mortality rates declined 10-20%. The researchers concluded that their findings did not support an association between PI, NRTI, or non-nucleoside reverse transcriptase inhibitor (NNRTI) use and excess cardiovascular or cerebrovascular problems. However, the results of this study may reflect the fact that by 1997 health-care providers were prescribing statins and other lipid-lowering therapies for HIV positive people with high blood fat levels.

A smaller study by Daniel Klein, MD, and Leo Hurley, MPH, of Kaiser Permanente in Oakland, California, also did not detect an increase in cardiovascular events among 4,159 HIV positive men in the Kaiser health-care system since the introduction of PIs. Between 1996 and 2001 there were 72 hospitalizations for coronary heart disease, including 47 myocardial infarctions. Rates of cardiovascular problems did not differ significantly between those taking and not taking PIs, or between those taking and not taking any antiretroviral therapy. However, the researchers did find that cardiovascular events were almost twice as likely among HIV positive participants compared with HIV negative participants (6.5 vs 3.8 events per 1,000 person-years). The researchers found "no effect of treatment type" on heart disease, and suggested that the higher rate in people with HIV may be due to chronic infection or other cofactors.

Carl Fichtenbaum, MD, and colleagues from the Cincinnati College of Medicine in Ohio evaluated 111 HIV positive people and 25 HIV negative controls. They found that the ten-year risk of progression to coronary heart disease was 4% among HIV positive participants compared with 1% among HIV negative participants. Additionally, among the HIV positive participants, heart disease risk was 6% for those treated with PIs compared with 3% for PI-naive participants.

Finally, results presented at the XIV International AIDS Conference held in Barcelona, Spain, in July 2002 offer perhaps the strongest evidence to date of an association between PIs and cardiovascular disease. Giorgio Barbarini, MD, of University La Sapienza in Rome and colleagues studied 1,200 HIV positive persons, half receiving regimens containing a PI and half taking PI-sparing combinations. After three years there were 23 new-onset cardiovascular events (12 heart attacks, 11 cases of angina) in the PI arm compared with two cases in the PI-sparing arm. The factors most highly associated with heart disease were elevated cholesterol and triglyceride levels, lipodystrophy, increased levels of fibrinogen (a protein that promotes blood clotting), and smoking; notably, all participants in this study were smokers.

Why Does It Happen?

Cardiovascular Risk Factors

Studies of the general population have provided a good understanding of the biology of cardiovascular disease and risk factors for problems such as atherosclerosis, heart attacks, and strokes. Prominent among these is the Framingham Heart Study, a 50-year longitudinal study begun in 1948 in Framingham, Massachusetts, that eventually involved some 10,000 people over two generations. Using population data, researchers have developed calculations to show how various factors affect heart disease risk. (See sidebar for a list of cardiovascular risk factors.)

Some heart disease risk factors are related to demographics. Because atherosclerosis develops over many years, older people (men over age 45 and women over age 55) are most likely to experience adverse cardiovascular events. In general, coronary heart disease is uncommon in young people unless they have a genetic predisposition or multiple risk factors. However, atherosclerosis may begin to develop in young adults well before obvious events such as ischemia or heart attacks occur. Men are more likely to develop heart disease than women. Estrogen has long been thought to play a protective role in women; however, the value of hormone replacement therapy to reduce cardiovascular disease in postmenopausal women is now disputed.

Other factors are related to lifestyle. Tobacco smoking is one of the clearest cardiovascular risk factors. In the Caerphilly Heart Study, which followed more than 2,500 men in a Welsh town between 1979 and 1983, smokers had a 2.3-fold increased risk of heart disease. Studies also have shown that cocaine use may cause more rapid cardiovascular disease progression, and the drug is known to trigger heart attacks.

Diet also plays a role. A diet high in saturated fat and/or cholesterol may predispose a person to atherosclerosis, while a diet high in sodium promotes high blood pressure in some people. Obesity is associated with a greater risk of heart disease, especially excess abdominal fat (the so-called "pot belly" or "spare tire"). Sedentary people are more likely to develop cardiovascular disease; conversely, exercise has a protective effect.

High levels of blood fats -- notably total cholesterol, low-density lipoprotein (LDL) cholesterol, and triglycerides -- are a major cardiovascular risk factor. In the Caerphilly study, people with a total cholesterol level above 200 mg/dL had a 1.5-fold increased risk of heart disease, while those with a triglyceride level above 175 mg/dL had a 1.8-fold increased risk. LDL cholesterol contributes to the formation of plaques on artery walls (atherosclerosis). In contrast, high-density lipoprotein (HDL) cholesterol protects against heart disease; in the Caerphilly study, HDL levels below 40 mg/dL were associated with a 1.4-fold increased risk. (See "Blood Lipids" for more information.)

Other metabolic factors associated with increased cardiovascular risk include high blood glucose levels, insulin resistance (reduced sensitivity of cells to the action of insulin), and diabetes; in the Caerphilly study, people with diabetes had a 2.3-fold increased risk of heart disease. Hypertension also is known to predispose people to heart disease; those with blood pressure over 140/90 had a 1.5-fold increased risk in the Caerphilly study.

Finally, studies have shown that various biochemical factors including C-reactive protein (a chemical that signals inflammation), fibrinogen, plasminogen activator inhibitor (a chemical that interferes with tissue plasminogen activator, an enzyme that dissolves blood clots), and homocysteine (an amino acid) may play a role in the development of cardiovascular disease.

Some experts believe that the emergence of heart disease in people with HIV is due to a high prevalence of traditional risk factors. A majority of people with HIV and AIDS in North America and Europe are male, and as effective treatment delays HIV disease progression, an increasing proportion are approaching the age at which cardiovascular events typically occur in the general population. To combat wasting, some people with HIV consume high-fat diets, and others are sedentary due to fatigue and illness. In addition, several studies have shown that a higher percentage of HIV positive people smoke tobacco compared with the general population; for example, 50% of those in Rainer Weber's DAD database (described below) were smokers. All of these risk factors may interact with metabolic side effects of antiretroviral therapy to increase cardiovascular risk.

HIV Infection and Heart Disease

In the early years of the AIDS epidemic, many people were afflicted with life-threatening opportunistic illesses (OIs), and health-care providers and people with HIV tended to focus on immediate rather than long-term manifestations of the disease.

Even before the advent of HAART, it was clear that HIV positive people were prone to fat metabolism disorders. People with HIV -- especially those with advanced immune deficiency -- often were observed to have high triglyceride levels, particularly if they had wasting syndrome. But they tended to have low cholesterol levels, both LDL and HDL. Although HAART has been shown to increase total and LDL cholesterol levels (discussed below), currently approved antiretroviral drugs are not known to decrease HDL levels. Therefore, some degree of increased cardiovascular risk likely exists in HIV positive people regardless of treatment status.

Some experts believe that chronic HIV infection itself increases heart disease risk. For example, certain cytokines -- such as interferon-alpha and tumor necrosis factor -- that may be present at high levels in people with HIV appear to interfere with lipid metabolism. Chronic inflammation may contribute to atherosclerosis. In addition, HIV infection has been associated with heart muscle damage (cardiomyopathy). Also, people with compromised immune systems are more likely to develop heart infections, and injection drug users are particularly prone to endocarditis.

Finally, many people with HIV have taken androgens and anabolic steroids to combat conditions such as wasting and abnormal body fat redistribution. These hormones can increase blood viscosity (thickness), and may interfere with blood flow and increase the risk of high blood pressure.

Lipodystrophy and Metabolic Abnormalities

Beginning in 1998 -- two years after the advent of HAART -- people with HIV and their health-care providers began to report unusual metabolic problems that appeared to be linked to the new drug combinations. Among these were body fat redistribution, fat accumulation (e.g., in the abdomen, back of the neck, and breasts), loss of fat in the face and extremities, elevated blood fat levels (hyperlipidemia), elevated blood glucose levels (hyperglycemia), and insulin resistance (a potential precursor to diabetes). These manifestations are sometimes collectively referred to as "lipodystrophy syndrome," although experts have not agreed upon an exact definition.

To date lipodystrophy syndrome has most strongly been associated with PIs, but metabolic side effects also have been linked to NRTIs and NNRTIs. Fat abnormalities have even been seen in HIV positive people who have never taken antiretroviral therapy, lending weight to the theory that HIV infection itself contributes to such manifestations. Many experts believe that metabolic changes in people with HIV are due to a combination of HIV infection, immune system recovery, antiretroviral side effects, and perhaps other not-yet-determined factors. (For more on lipodystrophy see "Body Fat Changes: More Than Lipodystrophy," BETA, January 1999; and "HAART Attack: Metabolic Disorders During Long-Term Antiretroviral Therapy," BETA, April 1999.)

Altered body fat metabolism may contribute in several ways to increased cardiovascular risk. Some people with lipodystrophy syndrome develop accumulations of fat around the midsection (the so-called "protease paunch") and visceral abdominal fat surrounding the internal organs. Studies of the general population have shown that abdominal fat accumulation (indicated by a waist-to-hip ratio greater than 1) is a risk factor for heart disease.

Likewise, high blood glucose levels, insulin resistance, and diabetes are known risk factors for cardiovascular disease among the population at large. Different studies have detected varying increased rates of insulin resistance in people using PIs, although it is not yet known whether this is clinically relevant.

Hyperlipidemia

Among the side effects associated with antiretroviral therapy, high blood lipids are the most likely to lead to an increased rate of heart disease among HIV positive people. Because general population studies have shown that high total cholesterol and triglyceride levels are associated with increased risk, when hyperlipidemia began to be seen in people taking HAART, health-care providers and advocates worried that this might be a warning sign of an impending increase in cardiovascular problems among people with HIV.

Hyperlipidemia is quite common among people taking antiretroviral drugs; some studies have found rates higher than 50% in people taking first-generation PIs. Total cholesterol levels as high as 1,000 mg/dL have been seen. (The National Cholesterol Education Program defines total cholesterol levels above 240 mg/dL as presenting a high risk for cardiovascular disease.) High blood fat levels have been most clearly associated with PIs, but also are seen to a lesser extent in some people taking NNRTIs. PIs are believed to cause hyperlipidemia by interfering with proteins that regulate fat metabolism. All approved PI drugs have been linked to hyperlipidemia, with ritonavir (Norvir), saquinavir (Fortovase), and lopinavir/ritonavir (Kaletra) being the worst offenders.

In an attempt to better understand the relationship between antiretroviral therapy and heart disease, Rainer Weber, MD, of the University Hospital in Zurich, Switzerland, and colleagues looked at a database containing information on over 20,000 HIV positive participants in 11 large prospective cohort studies in North America, Europe, and Australia (the DAD study); results were presented at the December 2001 ICAAC. Elevated total cholesterol levels were least common in people who had never taken antiretroviral drugs. Increasing rates of elevated total cholesterol were observed in people who had taken only NRTIs, those who had taken NRTIs plus NNRTIs, and those who had taken NRTIs plus PIs; those who had taken drugs from all three classes had the highest incidence of elevated total cholesterol (44%). Elevated total cholesterol levels were associated with higher CD4 cell counts and lower viral loads. Elevated triglyceride levels also were most often seen in people who had taken all three classes of drugs. In addition, 25% of participants had decreased HDL cholesterol. The same research team presented results at the February 2002 Retrovirus conference showing that while people taking either PI-based or NNRTI-based regimens had elevated total cholesterol levels, those taking NNRTI-based regimens had comparatively higher HDL levels, potentially protecting them against cardiovascular problems. Other studies have suggested that efavirenz (Sustiva) may increase HDL.

Switching and Substituting

Concern about the adverse metabolic side effects associated with HAART has led to some new approaches to anti-HIV therapy. Perhaps the most sweeping change is a shift away from early treatment in asymptomatic people with HIV. As reflected in the most recent treatment guidelines from the U.S. Department of Health and Human Services, antiretroviral therapy is now recommended for asymptomatic HIV positive people with CD4 cell counts below 350 cells/mm³ (vs 500 cells/mm³ in the previous guidelines) and viral loads above 55,000 copies/mL by PCR or bDNA (vs 10,000 copies/mL previously).

Researchers also have explored the possibility of structured intermittent therapy or structured treatment interruption, in which anti-HIV therapy is taken on a cyclical basis with careful monitoring. While small, early studies have shown that blood fat levels decrease during periods off drugs, it is not known whether rising and falling lipid levels present less cardiovascular risk than consistently elevated levels.

Because different antiretroviral drug classes and different drugs within a class affect fat metabolism to different degrees, health-care providers have various options for attempting to minimize hyperlipidemia while maximizing therapeutic benefit. Numerous researchers have conducted "switch studies" in which protease-sparing regimens or new PIs with presumably milder side effects are substituted for older PIs.

An entire session at the February Retrovirus conference was devoted to switch studies. For example, C. Fisac from Hospital Bellvitge in Barcelona found decreased lipid levels after six months in 92 participants who switched from a protease-based regimen to one containing either abacavir (Ziagen, an NRTI), efavirenz (an NNRTI), or nevirapine (Viramune, also an NNRTI). Total cholesterol decreased in all three groups, as did LDL cholesterol. HDL cholesterol increased significantly in the efavirenz and nevirapine groups, but decreased significantly in the abacavir group. Triglyceride levels decreased by 29.5% in the nevirapine group (a statistically significant decline), 9.9% in the abacavir group, and 4% in the nevirapine group.

Similar results for a larger group of 460 participants after one year on a new regimen were presented as a late-breaker at the same conference by Jose Maria Gatell, MD, also from the Barcelona team. This study -- the largest switch study to date -- showed that protease-sparing regimens were well tolerated and maintained good virological suppression; after 12 months viral load remained undetectable in 78% of the nevirapine group, 77% of the abacavir group, and 74% of the efavirenz group (using an intent-to-treat analysis). Although the rate of viral control was lower in the abacavir group, this was balanced by fewer people discontinuing the drug due to side effects.

In early studies the PI amprenavir (Agenerase) appeared to produce milder blood lipid increases, but this has not been consistently borne out in more recent, larger trials. Researchers are now holding out hope for the newest PI, atazanavir (Zrivada). Studies to date suggest that atazanavir does not increase LDL or triglyceride levels as much as other PIs, but does increase HDL cholesterol, leading to a more favorable LDL-to-HDL ratio. At the February Retrovirus conference Peter Piliero, MD, of Albany Medical College in New York and colleagues reported on two clinical trials of atazanavir in treatment-naive participants. After 48 weeks, the lipid elevations seen in those taking nelfinavir were not seen in those receiving atazanavir. The researchers went so far as to conclude, "this suggests that atazanavir may reduce the risk of cardiovascular events in this population." (For more information on atazanavir, see "The HIV/AIDS Drug Pipeline: A Status Report.")

Arterial Dysfunction

In addition to hyperlipidemia, arterial dysfunction is another potential cause of increased cardiovascular risk in people with HIV. Impaired function of the endothelial cells lining blood vessels is known to play a role in the development of atherosclerosis. Previous studies have shown that PI-based regimens are associated with endothelial dysfunction. For example, James Sosman, MD, of the University of Wisconsin and colleagues reported results of a study of arterial function at the November 2001 meeting of the American Heart Association. They found impaired arterial vasodilation (expansion) as measured by ultrasonography in 21 of 28 HIV positive people using PIs.

At the February Retrovirus conference Michael Dubé of Indiana University in Indianapolis and colleagues presented data showing that indinavir (Crixivan) led to endothelial dysfunction in six HIV negative men (five of them smokers) as determined by measurements of blood flow in the legs. After four weeks of indinavir, blood flow was significantly impaired compared to baseline. The drug appeared to interfere with arterial production of nitrous oxide, which acts as a vasodilator (an agent that expands blood vessels and increases blood flow). The participants in this study did not experience significant lipid elevations, suggesting that arterial dysfunction is independent of high blood fat levels. The researchers speculated that insulin resistance, a side effect of indinavir, may play a role in endothelial dysfunction. Other research teams, however, have failed to find an association between HAART and endothelial dysfunction.

Other Biological Mechanisms

High blood pressure (especially systolic pressure) is a known independent risk factor for heart disease. Data are inconclusive as to whether HIV infection itself is associated with hypertension. However, there is evidence that some antiretroviral drugs increase the risk of high blood pressure. For example, Ross Hewitt, MD, from the State University of New York at Buffalo and colleagues analyzed data from 445 people treated with PI-based regimens. After 200 days of therapy, 32% of those taking indinavir had developed high blood pressure (140/90 or higher) compared with 19% of those taking nelfinavir and 18% of those not using PIs. After 600 days, the percentages were 53%, 36%, and 34%, respectively.

Fred Sattler, MD, from the University of Southern California at Los Angeles and colleagues compared data from 42 HIV positive individuals taking HAART who had lipodystrophy, 42 HIV positive people taking HAART who did not have lipodystrophy, and 13 HIV negative controls. After an average of 21 months, hypertension (over 140 diastolic, over 90 systolic, or both) was seen in 74% of those with lipodystrophy and in 48% of HIV positive people without lipodystrophy. Participants with an increased waist-to-hip ratio were significantly more likely to have high blood pressure. The authors suggested that elevated blood pressure may be linked to the constellation of metabolic disorders seen in people with HIV. They concluded that the occurrence of hypertension "portends an increased risk for myocardial infarction, stroke, renal failure, and peripheral arterial disease," and recommended early identification and management.

Keith Henry, MD, of the University of Minnesota and colleagues reported at the February Retrovirus conference that HIV positive participants using indinavir were more likely than the general population to have high-risk levels of C-reactive protein, a chemical marker of chronic inflammation that is associated with increased cardiovascular risk.

Finally, Linda Bausserman, PhD, from Brown University in Providence, Rhode Island, and colleagues reported that HIV positive women had significantly elevated levels of cell adhesion molecules compared with HIV negative women, although there were no statistical differences in terms of type of antiretroviral therapy. Cell adhesion molecules are believed to play a role in the development of atheroscelerosis, and are associated with higher heart attack and stroke risk.