Heart disease and HAART: what's the risk?

The risk that lipid elevations associated with protease inhibitor therapy may lead to an increased risk of heart disease in the future has caused much concern among doctors and patients, but there is still little evidence to show the degree of increased risk that can be expected.

The Second International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV, which took place in Toronto between September 13-15 heard several important presentations on what happens to the blood vessels of people receiving PI treatment, and what the long-term impact of raised cholesterol levels might be in people taking long-term HAART.

Blood vessels change on PI-based HAART

Several changes in blood vessels have been reported in individuals receiving protease inhibitor treatment:

• Protease inhibitor treatment is associated with endothelial dysfunction. The endothelium is the layer of the blood vessel which dilates to allow blood to flow; when its function becomes sluggish there is an increase risk of high blood pressure and arterial blockages.

• The intima media is thicker in protease inhibitor treated individuals. Increased intima meda thickness is a surrogate marker for the risk of cardiovascular events.

James Stein of the University of Wisconsin reported on a comparison of 22 PI-treated individuals and 15 HIV-positive individuals receiving non-PI containing regimens matched for age, CD4 count, NRTI treatment and coronary risk factors other than cholesterol and triglyceride levels. Flow mediated vasodilation was measured, and PI treatment was associated with reduced flow.

Renato Maserati of the University of Pavia measured intima media thickness in 27 PI treated individuals with greater than 18 months PI experience, and compared them with 16 individuals receiving non-PI containing regimens and a control group of 15 age-matched HIV-negative individuals. Intima media thickness was significantly higher in the PI-treated group, although Dr Graeme Moyle of the Chelsea and Westminster Hospital in London pointed out that many of the elevations were within the normal range and that patient numbers in this study may not have been large enough to discount observer error when conducting the assessement.

Frank Goebel of Munich University Poliklinik reported on endothelial dysfunction in 11 PI recipients with cholesterol levels above 300mg/dL compared with four HIV-positive individuals with cholesterol below 200mg/dl and 10 HIV-negative individuals with cholesterol below 200mg/dl. Unfortunately the three groups were not age matched: in particular, the HIV-negative control group were more than half the age of the PI recipients, who averaged 48 years old! However, the assessments of blood flow, blood pressure, coronary vascular resistance and calcification of the arteries were age adjusted, and the researchers concluded that reductions in cardiovascular performance were associated with direct endothelial function rather than calcification of the arteries. Prof. Goebel noted that hypercholesterolemia can cause endothelial dysfunction within one day, and is apparent even before any changes in the appearance of the endothelium are detectable. This implies that cholesterol levels are an adequate surrogate for endothelial function and that sophisticated ultrasound tests should be not be necessary.

There were no reports of direct cellular changes within the cardiovascular system caused by protease inhibitors, and in the absence of any evidence to the contrary, the elevations in cholesterol and triglycerides along with other metabolic changes seen on HAART seem to be the causes of the changes reported above. These studies did not look at cholesterol changes associated with efavirenz treatment, where elevations in cholesterol have also been seen (albeit not so large and not for so long).

What’s the latest thinking on long term risk?

What impact will such changes have on people taking long term PI treatment? The first point to bear in mind is that the risk of coronary heart disease can be influenced by changing levels of cholesterol. If cholesterol levels are reduced, the risk of heart disease goes down. So, people who have persistently high levels of cholesterol will have a greater risk than those who experience a brief elevation to a very high level, followed by a decline.

However, there are other risk factors beside PI treatment to consider, especially as people with HIV live longer.

“Age is to CHD what the CD4 count is to HIV infection – the risk takes off in men at 45 years and approximately ten years later in women. Men are six times more likely to develop CHD at the age of forty, and the gap widens as age increases” says Matthias Egger of the Swiss HIV Cohort, who is now working at Bristol University on a major study of heart disease.

He has compared data from Andrew Carr’s 1999 lipodystrophy study (published in The Lancet) with the Caerphilly Coronary Heart Disease cohort, based on 2512 men recruited between 1979 and 1983 at ages 45-59 and followed since then.

He found that insulin resistance is similar in treatment-naïve HIV-positive individuals and men in the Caerphilly cohort aged under 50, and older men in the Caerphilly cohort have similar insulin resistance levels to men with lipodystrophy, so age is an important confounding factor in any study of CHD and lipodystrophy.

Nevertheless, untreated lipodystrophy increases the relative risk of CHD by 1.5-4-fold when the impact of age is factored in, and the risk is slightly higher in women.

Next, Matthias Egger used data from the Swiss HIV cohort to analyse what impact HAART has on the risk of dying within three years, and asked how many patients would need to start treatment at particular CD4 and viral load levels in order for one death in that group to be averted.

“If the progression risk is very high the number of patients that need to be treated in order to avert one death is extremely low, but when you get into the CD4 350-500 range with viral load of 10,000 copies, HAART reduces the progression risk from 2% to 0%, so you would need to treat 50 patients to avert one death, so 49 people would be exposed to the risk of lipodystrophy for every death averted.”

If the risk of developing lipodystrophy is high over a three to five year period, that in turn will have an impact on the chances of developing heart disease as a consequence of elevated cholesterol.

Using the Framingham data set (a very large study of the lifetime risk of heart disease in a US town), Egger concluded that amongst 50 year old male smokers with cholesterol elevations, the number who would need to be treated for five years in order to produce one heart attack would be 13 individuals. In contrast, 30 year old non-smoking males would be at considerably lower risk, with the need for 71 individuals to be treated to produce one heart attack (if cholesterol levels went untreated).

In considering when to start, how do you factor in what risk you have of developing lipodystrophy?

The Lipodystrophy Workshop heard little new information on predictors of lipodystrophy. Beyond the initial choice of therapy, where protease inhibitors clearly increase the risk, higher baseline triglycerides and cholesterol are associated with subsequent increase in the risk of metabolic abnormalities on HAART. High blood pressure may also be a risk factor. Further discussion of the risks associated with HAART and heart disease can be found elsewhere on this site in Body fat and metabolic changes while on treatment: Heart disease and HAART.