Most protease inhibitors (PIs) have been associated with abnormal lipid profiles and insulin resistance, whereas better lipid profiles (such as an increase in high-density lipoprotein [HDL]) are seen with non-nucleoside reverse transcriptase inhibitors (NNRTIs). One can thus speculate that patients taking PIs are more likely to develop atherosclerosis from hyperlipidemia. This study by Pierone et al tries to answer that question.

Patients were enrolled in this prospective trial if they had been on nevirapine (NVP, Viramune), efavirenz (EFV, Sustiva, Stocrin) or PIs for more than 3 years. There were 40 patients in each arm. Patients with established coronary artery disease were excluded. The metabolic parameters that were studied included fasting lipid profile, C-reactive protein, homocysteine and lipoprotein-a, all of which have been shown to correlate with cardiovascular disease in the general population.

Patients also underwent the following tests:

1. Carotid artery ultrasound to determine carotid artery intima-media thickness (CIMT), which is a validated surrogate marker for the presence and progression of cardiovascular disease.

2. Multislice computerized tomography scan to determine coronary artery calcification. The calcification scores reflect the amount of plaque in the coronary artery.

3. Brachial artery reactivity (BAR) test, which measures endothelial function based on changes in brachial artery diameter when the blood pressure cuff is inflated to induce arm ischemia. The less change in diameter, the more atherosclerosis is present.

Surprisingly, the baseline lipid profiles (see Table) revealed significantly higher total cholesterol and low-density lipoprotein (LDL) in the NNRTI arms, but there were no differences between the 2 NNRTIs.

There were no significant differences in baseline C-reactive protein, homocysteine and lipoprotein-a levels among the 3 arms. Because the 2 NNRTI arms had the same baseline characteristics, the data were combined into a single NNRTI group for comparison to PIs. At baseline, the percentage changes in the BAR test results were not different between the PI and NNRTI groups, nor were the CIMT results. More patients in the NNRTI group had no calcium deposits in their coronary arteries (coronary artery calcium [CAC] = 0) than in the PI group (74% vs. 52%). Patients on PIs were also 9.5 times more likely to have more calcification (CAC >100) than patients on NNRTIs (22% vs. 7%).

At 1-year follow-up, CAC had not significantly changed between the 2 groups, although Dr. Pierone explained that it could take longer than 1 year to develop more calcification. However, CIMT had strikingly increased in both the NNRTI and PI arms, despite good lipid levels, with a trend toward a faster increase in the PI arm.

In summary, patients on PIs are more likely to have atherosclerosis, as evidenced by greater coronary artery calcification at baseline and an increase in carotid artery thickness over the year. What this study does not explain is the pathogenesis. We cannot explain the increased likelihood of atherosclerosis on the basis of abnormal lipids and/or the inflammatory process, as all other cardiovascular disease surrogate markers are the same in both the PI and NNRTI groups. Larger longitudinal studies across different cohorts are needed.