The 8th Annual Conference on Retroviruses and Opportunistic Infections (Retrovirus) met in Chicago from February 4-8, 2001. Many of the lectures and symposia are available for viewing on the conference's website (www.retroconference.org). Posters and abstracts, referenced and numbered in parentheses in this report, are also available on the website.

The Phenomena Formerly Known as Lipodystrophy

The phenomena formerly known as lipodystrophy were re-dubbed "metabolic complications" at the Retrovirus conference. The long-term side effects included under this title were:

• Elevated levels of lactate in the blood (lactic acidemia, lactic acidosis)

• Tingling, numbness, or pain in the arms or legs (peripheral neuropathy)

• Fat loss in the arms, legs, or face (lipoatrophy)

• Insulin resistance or high blood sugar and/or diabetes

• Fat accumulation inside the abdomen (visceral adiposity)

• Increased fat under the skin of the back or breasts

• Abnormal blood cholesterol or triglycerides levels (dyslipidemia)

• Decreased bone density and other bone disorders (osteopenia, avascular necrosis)

One of the reasons that these side effects are being separated is because they appear to have distinct sets of causes and correlations. Many of these problems (neuropathy, fat loss, decreased bone density) have been observed in HIV-positive people who have never taken therapy, and seem to be caused, at least in part, by HIV infection itself. Others (lactic acidemia, drug-related neuropathy, and fat loss in the face, arms, or legs) seem to be associated with damage to an energy-producing part of human cells called mitochondria, possibly correlated with NRTI treatment. Treatment with PIs is being investigated as a contributing factor to the alterations in fat and sugar metabolism that may lead to high cholesterol and triglycerides, insulin resistance, and high blood sugar (glucose).

Because of the growing appreciation of long-term side effects, over 60 abstracts were devoted to the subject. Most studies focused on defining and finding correlations or causes rather than on treatment. Some of the factors that were examined as possible correlations, other than HIV and anti-HIV medications, were age, physical inactivity, gender, race or ethnicity, and CD4 count.

Incidence Data from MACS

Abstract 538 presented data on the incidence of fat maldistribution in men-who-have-sex-with-men (MSM) from the Multicenter AIDS Cohort Study (MACS). MACS is a long-term, observational cohort study of 5,000 MSM, both HIV-positive and HIV-negative. The study looked at the incidence of fat loss versus fat accumulation in a subset of 868 participants, (62% were HIV-positive and 38% were HIV-negative) using a survey and measurements. Because 73% of the HIV-positive men had taken "HAART" (highly active antiretroviral therapy with at least a 3-drug regimen, most including a PI) and 14% had never taken any anti-HIV medications, they also looked at differences in incidence based on anti-HIV therapy.

The results were that 25 to 35% of all the HIV-positive men had fat loss in the face, arms, or legs (lipoatrophy), compared to only 2% of the HIV-negative men. There was no statistically significant difference attributed to HIV, however, in central fat accumulation (35% of HIV-positive men versus 26% of HIV-negative men). When they looked at people who had a mixture of fat loss and fat accumulation symptoms, the differences were striking: 40% of the HIV-positive group had mixed symptoms, compared to only 1 to 2% of the HIV-negative group. They then looked at the differences according to the type of anti-HIV therapy in this group of people with mixed symptoms. Among people who had no anti-HIV therapy, only 1 to 2% had moderate or severe mixed symptoms. For those who had taken mono or dual NRTI therapy, 8% had mixed moderate or severe symptoms, and 20% of those who had taken HAART fell into this category.

The MACS researchers also looked at changes in blood lipids (cholesterol and triglycerides). They found that a low level of "good cholesterol" (HDL less than 35mg/dl) was associated with being HIV-positive, but did not vary with treatment history. In contrast, a high level of triglycerides (above 400mg/dl) was associated only with HAART. Another interesting finding from this data set was that the incidence of peripheral fat loss and central fat accumulation rose steadily during the first 2 years of HAART, but appeared to stabilize after that. Although this study was not randomized, it did make good use of HIV-negative and untreated HIV-positive control groups to distinguish among side effects due to HIV, those due to treatment, and those due to aging.

Lipoatrophy

Abstract 539 presented results from a small sub-study examining the differing effects of two NRTIs, Zerit and AZT, on both lipoatrophy and central fat accumulation. Both of the drugs were taken with Crixivan and Epivir for 30 months. None of the participants had ever used a PI, but a number of participants, evenly distributed between the groups, had already had mono or dual NRTI therapy with AZT, Videx or Hivid.

Similar to the MACS study, this study saw no differences between groups in central fat accumulation. The results did show a striking difference in fat atrophy. The type of fat loss, or lipoatrophy, most commonly found in people with HIV is a loss of subcutaneous (directly under the skin) fat in the face, arms, and legs. Using skin-fold measurements and physician and patient surveys, this study found that the percentage of people reporting fat loss in the face, arms, or legs after 30 months on therapy was twice as high in the group taking Zerit compared to the group taking AZT.

Looking at Lactate Levels

Many studies concerning correlations between blood lactate levels, mitochondrial damage, and other metabolic side effects were also presented. Lactate is a normal by-product of human cells breaking down fats and sugars. Elevations in lactate levels, known as lactic acidemia, have been observed in people with HIV, especially in those taking NRTI therapy. It is also possible for a person to get "lactic acidosis," which indicates severely elevated lactate levels above 5 to 10 mmol/l and the presence of symptoms. Lactic acidosis can be a very serious, although rare, condition that often involves symptoms such as unexplained severe fatigue, nausea, and/or abdominal pain.

A summarizing lecture was given by researcher Andrew Carr of Australia. Looking at several studies presented at Retrovirus and previously, he estimated that approximately 20% of HIV-positive people on NRTI medications had mild elevations in lactate, 2 to 5 mmol/l, with no accompanying symptoms. He emphasized that routine measurements of lactate levels are not recommended, however, because these mild elevations do not appear to predict whether or not a person will progress to lactic acidosis or have symptoms.

Treatment for Peripheral Neuropathy

One of the few abstracts looking at treatment for metabolic side effects examined the use of aspirin for treatment of neuropathy (Abstract 601). Rather than having people take aspirin orally, these investigators crushed up 375mg of aspirin, dissolved it in 7ml diethyl ether, and applied it to the arms and legs were pain occurred. This was a rigorous, double-blind, placebo-controlled cross-over study. Half of the participants applied the aspirin mixture topically, three times a day for two weeks, then, after a brief wash-out period, they switched to the placebo version (just the ether with no aspirin). The other half started with placebo and then switched to aspirin. Neither the participants nor the researchers knew which mixture was being used at any time. Throughout the study the participants assessed their level of pain relief with a survey known as the Brief Pain Inventory (BPI). When the results were unblinded, they showed that the aspirin mixture relieved pain 30% better than baseline measurements and 30% better than the placebo. In fact, some of the participants who were in the group that started with aspirin dropped out of the study when they switched to placebo because they were unwilling to give up the relief they found.

Decreases in Bone Health

Data on decreases in bone health continues to be compiled. Although the interest in HIV and bone health is fairly new, it appears that these effects have been present for some time and are just now beginning to be noticed. One of the most common effects seen and reported on is osteopenia. Osteopenia is a decrease in bone mineral density. In its most severe form, this loss of bone mineral density can lead to osteoporosis and structural fragility in the bones. Studies presented at Retrovirus estimated the prevalence of osteopenia to be 21 to 45% among people with HIV (Abstracts 626 and 629). The estimated prevalence of osteoporosis, 2 to 3%, was low, but still significantly higher than the rates for HIV-negative people of the same sex and age (Abstract 628).

Several studies also looked at a condition known as avascular necrosis (AVN). Avascular necrosis refers to the death of bone tissue, usually in the hip, due to a decrease in blood supply. Again, this condition is also very rare, but does appear to occur to a higher degree in people with HIV. A study from Johns Hopkins University (Abstract 637) found that the incidence rate for AVN of the hip among people with HIV was 48% higher than that of the general population (1.9 per 1000 person-years compared to 0.04 per 1000 person-years). The majority of bone health studies seemed to support the view that this is HIV related, rather than medication-related, although it is not yet entirely clear. One study (Abstract 631) did correlate osteopenia with elevated lactate levels, and possibly NRTI therapy. As with lactate levels, however, routine monitoring of bone mineral density is not yet recommended because the conditions are rare and their clinical significance is still not known. Like many of these metabolic side effects, the impact of these findings may take on more significance as the population of people living with HIV ages.

Increases in Cardiovascular Disease Risk

Anti-HIV medications, particularly PIs, are known to increase total cholesterol, "bad" (LDL) cholesterol, and triglycerides. Before anti-HIV medications were available, it was shown that people with HIV already had low levels of "good" (HDL) cholesterol. All of these conditions increase a person's risk of cardiovascular disease. Two large studies presented at Retrovirus tried to ascertain if this increase in risk actually translates to higher levels of cardiac problems in people living with HIV.

A retrospective study in French hospitals (Abstract 657) looked at the incidence of heart attacks in people with HIV and with varying lengths of exposure to PIs. They then compared these incidences to those found in the general French population. Of 19,795 HIV-positive men who had taken a PI, there were 54 heart attacks during an 18-month period. The researchers calculated the relative risk of heart attack for people with less than 18 months of PI use, with 18 to 29 months of PI use, and with over 30 months of PI use, compared to the general population. They found that the risk was no higher for those on PIs for less than 18 months, but did increase with longer duration of PI use (1.7 times higher for those on PIs for 18 to 29 months, and 3.1 times higher for those on PIs longer than 30 months). While the findings in this study were compelling, no examination of differences in other cardiac risk factors was done, so the conclusion that PI use accounted for all the difference cannot be made.

Another large study was presented using data on patients at Kaiser Permanente of Northern California (Abstract 655). This is an ongoing, prospective observational study of coronary heart disease that began in 1996 and is following 4,500 HIV-positive and 41,000 age- and sex-matched HIV-negative people. This study found that the overall risk of hospitalization for coronary heart disease was 1.6 times higher among HIV-positive people, but that no differences in risk were attributable to PI use. These researchers did attempt to find differences in underlying cardiac risk factors by surveying the medical records of 264 HIV-positive people and 710 HIV-negative people. They found that people who were HIV-positive did tend to have higher cholesterol, but had lower blood pressure. They found no difference in rates of tobacco use or diabetes, both of which are significant risk factors.

What Can Be Done?

The growing list of known metabolic complications of HIV disease can seem overwhelming, especially because so little is known about the exact causes. The good news, however, is that a lot is known about the treatment of these conditions when they are not HIV-related. This means that there are many proven strategies that HIV-positive people can use to decrease their underlying risk. For instance, bones can be strengthened by maintaining good activity levels, supplementing calcium and vitamin D when needed, and avoiding the use of cigarettes and corticosteroids. It may also be useful to monitor and supplement hormone levels, since hypogonadism (low testosterone or estrogen levels) is a known risk factor for osteoporosis. A healthy weight and a diet including low-cholesterol, low-fat, and low-sugar foods can help control cholesterol, triglyceride and blood sugar levels. In addition, cardiac risk can be lowered by stopping smoking, exercising and using cholesterol- and triglyceride-lowering drugs, if needed.

Choosing Anti-HIV Therapy

Although very few of the symptoms show direct causation by specific anti-HIV therapies, many studies have looked at the effect of stopping or switching medications. In a few cases, stopping a given therapy is definitely warranted. One of these cases is symptomatic lactic acidosis. Lactic acidosis can require that a person interrupt the use of NRTIs and allow lactate levels to return to normal. It is also common to avoid certain PIs to attempt to control cholesterol levels. Many studies presented at Retrovirus also looked at switching from a PI to an NNRTI to alleviate metabolic symptoms (Abstracts 668 - 673).

These studies looked at people who had successful viral suppression (to undetectable viral loads) with a PI-based regimen, and then switched their PI for an NNRTI. In the vast majority of these studies, those who switched maintained viral suppression, suggesting that there is no negative health effect. However, few studies showed a clinically significant effect on cholesterol levels, triglycerides, or body-shape changes, such as facial and limb fat loss and central fat accumulation.