Difficulties in Understanding the Metabolic Complications of AIDS

With the introduction of effective therapy for HIV infection, a variety of changes was reported both in the distribution of fat and in levels of plasma metabolites.

The abnormalities of fat redistribution include the appearance of buffalo hump, increased abdominal size, breasts, and visceral adipose tissue (VAT), and decreased fat in the face, arms, legs, and buttocks.

The changes in plasma metabolites include increased levels of triglycerides, low-density lipoprotein (LDL) and total cholesterol, glucose, and insulin; the latter two indicate insulin resistance. Low levels of high-density lipoprotein (HDL) cholesterol were also found.

Some of these changes, particularly increased VAT, hypertriglyceridemia, low HDL cholesterol levels, and insulin resistance resemble metabolic syndrome or “Syndrome X,” an insulin-resistance syndrome described in HIV-negative individuals that is associated with increased cardiovascular risk.

Likewise, severe peripheral lipoatrophy syndromes are accompanied by hypertriglyceridemia and insulin resistance. Therefore, it has been presumed by some that each of these findings is part of a single syndrome, which was initially attributed to HIV protease inhibitors (PIs).

The present article reviews evidence that multiple metabolic disturbances are occurring, with the phenotype being the sum of the individual components that occur.

It must be remembered that changes in lipid and, to a lesser extent, glucose metabolism were found in HIV infection before the HAART era.

With the onset of HIV PI therapy, triglyceride levels increased, predominantly because of increases in VLDL levels, with some increase of triglycerides in LDL cholesterol. Likewise, PI therapy induced increases in cholesterol because of increases in both VLDL and LDL levels.

It is of note that LDL levels, which were low before HAART, are, on average, restored to normal by PI therapy. But the increase in VLDL levels was more striking. Patients with genetic disorders predisposing them to hyperlipidemia are most severely affected.  PI therapy also leads to small increases in glucose levels and to significant increases in plasma insulin levels that are indicative of insulin resistance.

Furthermore, not all of the insulin resistance in HIV infection can be attributed to PIs. The use of nucleoside reverse-transcriptase inhibitors has also been associated with insulin resistance.

These data left a dilemma as to which aspects of the metabolic changes are due to the direct effects of PIs themselves and which to the restoration of health and/or immune reconstitution.

 One remarkable finding was that the treatment of HIV-negative subjects with indinavir led to no increase in lipoprotein levels, which suggests that some of the increases in lipoprotein levels may not be direct effects of the drugs.

In another study, ritonavir was given to HIV-negative subjects, and lipoprotein levels were analyzed. Unlike indinavir, ritonavir induced increases in levels of VLDL cholesterol and triglycerides in HIV-negative subjects. However, like indinavir, ritonavir did not increase LDL cholesterol levels.

These data suggest that some effects may be specific to drugs within an individual class, such as the induction of hypertriglyceridemia, and that other findings may represent restoration to health, such as a reversal of the decrease in LDL cholesterol. Several studies of HIV infection have solidified this concept.

Another group found that ritonavir significantly increased triglyceride levels in their patients with HIV infection, but indinavir and nelfinavir did not. Although other groups have found increases in triglyceride levels due to indinavir and nelfinavir, it is now clear that the magnitude of increase is much greater for ritonavir. Thus, there are drug-specific effects within the PI class on VLDL triglyceride levels.

Furthermore, LDL cholesterol level increases have been seen in HIV-infected patients who are given non nucleoside reverse-transcriptase inhibitor (NNRTI) therapy as part of HAART. The increase in LDL induced by nevirapine was similar to that induced by indinavir in this randomized trial.

Thus, the data on the effects of PIs in HIV-negative subjects and the effects of PIs and NNRTIs in HIV-infected patients would strongly suggest that the increases in LDL cholesterol represent a restoration of health. However, unlike PI therapy, NNRTI-based HAART leads to increases in HDL levels.

Although both central lipohypertrophy ("buffalo hump," increased abdominal, and increased VAT) and peripheral lipoatrophy (in the cheeks, arms, legs, and buttocks) can be found in the same subjects, there are no data that have suggested that they are statistically linked.

Indeed, there are data that have demonstrated that, on average, patients with HIV infection do not have increased central adiposity. Rather, they have preserved VAT but less limb fat. Furthermore, the examination of many patients with severe facial and limb lipoatrophy revealed decreased fat levels in the neck. Yet most (but not all) patients with "buffalo hump" have large increases in neck fat. A single metabolic process cannot decrease and increase fat in the same area.

 It is well recognized that increased visceral fat is accompanied by hypertriglyceridemia and insulin resistance in patients without HIV infection. It is also well recognized that severe lipoatrophy in patients without HIV infection leads to hypertriglyceridemia and insulin resistance. In patients with HIV infection, the presence of either increased VAT or severe decreases in SAT are also likely to promote these metabolic disturbances.

In summary, multiple factors including individual drugs, classes of drugs, fat distribution, the restoration of health, and, possibly, immune reactivation likely contribute to the changes in metabolism and fat distribution that are seen in HIV infection. The underlying genetics of individuals may also play a role. Some or all of these factors may work together and worsen changes in lipid and glucose metabolism. Understanding each of these contributors and how they interact is essential to understanding the effects of therapy for HIV infection.

09/12/03

Reference
C Grunfeld and P Tien.  Difficulties in Understanding the Metabolic Complications of Acquired Immune Deficiency Syndrome. Clinical Infectious Diseases  37: S43-S46. September 1, 2003.