HIV drug therapy works against HIV. This is clear from the dramatic drop in deaths due to AIDS after the use of combination therapy was widely adopted in 1996. But these drugs also act on the body and its metabolism in a number of strange and hard-to-pin-down ways.
In some people -- but not all -- blood lipids soar to alarming levels that are associated with a higher risk of cardiovascular disease. Yet the rates of heart attack and heart disease in people taking HIV drugs continues to run about the same as those for people without HIV. Antiretroviral therapy may make for some worrisome lab tests, but stopping smoking rather than stopping meds remains a better bet for keeping a healthy heart.
In some people -- but not all -- fat deposits in the arms, legs and, most visibly, in the face, melt away, leaving people who experience this fat depletion syndrome bearing the tribal scars of facial wasting. When the fat drains from the face, it may be very hard to recover. One option for treating this condition is to undergo an expensive restoration process of soft tissue augmentation using a filler material injected under the skin of the cheeks.
And in some people -- but not all -- insulin tolerance goes out of whack, which may or may not lead to diabetes. (We won't even discuss bone formation problems.)
If you notice a pattern here, it's that what seems to be a definable lipodystrophy syndrome at the 35,000-foot level, becomes a wilderness of associations and anecdotes once you're on the ground. Inconsistent methods, unvalidated measurements and tiny studies make simply defining this complication contentious as researchers continue to tease out the links between suspected drugs, genetics and HIV itself. So far, no smoking gun has been found and none is expected to turn up. This is, as they say, a multifactorial problem. Scientists looking for answers to this riddle are developing new and more specific lab tests as well as larger and more reliable clinical trials. And clinicians are looking at everything from nuke-sparing regimens to prophylactic glitazones to forestall or minimize the consequences of lipodystrophy. In the meantime, managing the drug-related toxicities of HIV therapy remains an art.
Workers in this ad hoc field recently met in San Diego at the 4th International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV to trade notes and argue over epidemiology and favorite pathogenesis theories. For those seeking more than a tantalizing intellectual exercise, there was little to write home about.
But the weight may be starting to shift on one lingering dispute. For a long time, whenever talk turned to speculation about which drugs were considered prime suspects for causing lipodystrophy, the name of one culprit always came up: Zerit (d4T or stavudine) is a popular nucleoside analog often combined with ddI or 3TC as part of a three-drug combination. The federal HIV treatment guidelines list d4T among its "strongly recommended" choices for first line therapy. Zerit was linked early on with the nerve damage that causes peripheral neuropathy in some people. More recently it's been indicted for contributing to some of the other unpleasant toxicities of ART, such as facial wasting. Of course, a lot of the charges were stimulated by studies and presentations funded by GlaxoSmithKline -- makers of AZT and 3TC -- who'd love to see this competitor sent away. And most of the counter-arguments and obfuscation originated with Bristol Myers Squibb, the makers of d4T. So, although many people were resolutely convinced of d4T's guilt, controversy held consensus at bay. But this contest may finally be coming to a head -- if not a verdict -- as the evidence against d4T solidifies. After data presented this year at the annual Retrovirus Conference, at the International AIDS Conference in Barcelona, and now at the Lipodystrophy Workshop and the annual ICAAC conference that followed it a few days later, the strength of the associations between d4T and fat wasting are becoming too strong to ignore. As one observer put it, "You can't really say it's just a Glaxo thing anymore."
The protease inhibitors were the first drug class to take the blame for fat redistribution and lipodystrophy and evidence continues to mount against some of them. But nucleoside analogs have also received scrutiny for their potential to deplete the energy-producing capacity of cells.
The prevailing theory of how nucleoside analogs contribute to fat wasting and elevated lipids revolves around a pathogenic piñata called mitochondrial toxicity. Cellular respiration is the name for a long series of chemical steps in which cells process glucose and oxygen into a useable form of energy. If any of these steps are slowed down or blocked, the whole process backs up and a kind of overflow system starts to consume glucose without oxygen, which dumps a byproduct called lactate into the blood. As an extreme example of what can go wrong with this system, a poison like cyanide can stop cellular respiration dead in its tracks -- with rapidly fatal results. But the toxicities thought to be caused by nucleoside analogs are far subtler -- so subtle that not all researchers can agree on how, when, where or even if they are happening.
This chain of energy-producing steps takes place mostly within tiny organelles carried inside nearly every cell in the body. These cells-within-cells are called mitochondria, and they possess their own set of DNA instructions to make some of the proteins they need to perform the job of respiration. It's thought that nucleoside analogs -- some more than others -- can, over time, affect the quantity or quality of several of these crucial proteins by interfering with the replication of mitochondrial DNA. The result can be the generation of excess lactate and an energy deficit in affected tissues; in nerve tissue, neuropathy can be a consequence. Some scientists think a different form of mitochondrial toxicity can trigger a cell to die directly through apoptosis. That could be what's happening to fat cells. Obviously more and better research is needed -- fast.
Researchers have been attempting to find a way to directly measure mitochondrial toxicity by quantifying the amount of mitochondrial DNA that can be recovered from cells. Some of this progress was shown at the Lipodystrophy workshop. But serious objections have been lodged against current versions of this technique because they are liable to contamination from other sources of mitochondrial DNA. Next year, the data from mitochondrial toxicity assays may be more convincing, but for now, they're not ready for prime time.
One blood abnormality associated with ART, and with nucleoside analogs in particular, is an increase in blood lactate levels. Slightly elevated lactate levels are not noticeable, but as lactate levels increase, symptoms such as nausea, abdominal pain and distension may become apparent, causing tolerability problems. Lactate levels over 5 mmol/L in the presence of symptoms can indicate a rapidly progressing, life-threatening illness called lactic acidosis. All of the nucleoside analogs have been implicated in this rare lactic acidosis syndrome and every person taking them should know the warning signs: nausea, abdominal pain, fatigue or muscle weakness.
But the meaning of moderately elevated lactate levels, called hyperlactatemia, is not certain, although many people believe nucleoside analog-associated damage to mitochondrial DNA is a cause. Older studies of blood lactates are difficult to interpret because of widely varying methods used to collect and analyze the samples. Adoption of a strict new protocol for the standardized collection and processing of blood lactate samples will hopefully help dispel the fog that has surrounded the meaning of these lab values.
One of the first tipoffs that metabolic abnormalities are afoot is when blood lipid levels start to climb. The key movers are total cholesterol and triglycerides. Rising cholesterol raises eyebrows because of its association with cardiovascular disease -- cholesterol can clog the arteries feeding the heart and set off a heart attack. But the underlying metabolic problems causing the cholesterol rise and the role ART plays are still mysteries. The same is true of the rise in triglyceride levels experienced by some people on ART. Sustained high triglycerides may be a predictor for developing pancreatitis or diabetes -- and increasing rates of diabetes in people on ART is a growing problem. How and why this is happening isn't certain. Still, most clinicians would feel more comfortable if their patients could achieve decent viral suppression without the abnormal labs.
A number of studies looking at lipid abnormalities have pointed the finger at protease inhibitors and at ritonavir in particular as causative factors. Recently a few more studies have shown that nucleoside analogs add their own bump to cholesterol and triglycerides. And a few high quality studies are starting to examine the associations with actual fat loss and recovery.
Andrew Hill told the Lipodystrophy Workshop that there is a "Tower of Babel" when it comes to reporting drug toxicities in clinical trials, with at least seven different ways of reporting abnormal lab values. Graded levels of toxicity are far less sensitive than using continuous values such as changes in area-under-the-curve. Drug company trials can either hide or accentuate toxicities depending on the method chosen to report events. Often these studies just add to the noise.
Until recently, most of the research pointing the finger at Zerit used methods that were easy to dismiss, such as small sample sizes, no controls, and uneven reporting criteria. But this latest wave of evidence is not so easily ignored.
A study presented in Barcelona (Gilead 903, sponsored by Gilead Sciences, Inc., the makers of tenofovir) compared d4T with tenofovir in 600 people over 48 weeks. All volunteers received efavirenz and 3TC. It was their first experience with HIV treatment. By the end of the nearly year-long study, viral load reductions were about the same in both groups. But there was a striking difference in the cholesterol and triglyceride levels between those who took tenofovir and those who took d4T. Triglycerides were unchanged from baseline in the tenofovir group but went up by 74 mg/dl in the d4T group. Cholesterol also went up by 24 mg/dl in the tenofovir group but increased by twice that amount in the d4T group.
More data from Gilead 903 on the lipid profile of tenofovir versus d4T was presented at ICAAC by Joel Gallant of Johns Hopkins University. This trial is scheduled to run for three years -- which counts as long-term compared to most studies, but is little more than a honeymoon in terms of how long most people will be taking these drugs. So far -- one year in -- only lab toxicities have been observed. Gallant made a point of stressing the need for longer-term studies to help accumulate data that might finally correlate lab values with clinical outcomes.
Earlier in the year, a Glaxo-funded study (ESS 40002) showed similar findings at the Retrovirus Conference. As a part of this trial, 111 treatment-naive patients received nelfinavir plus 3TC and either AZT or d4T as the third drug in their regimens. With nelfinavir in the mix, some increases in cholesterol and triglycerides would be expected. But while people receiving AZT had cholesterol increases averaging 32 mg/dl, those on d4T had average increases of 45 mg/dl. For triglycerides, the results were even more striking with the AZT group seeing a bump of 31 mg/dl while the d4T group saw an average increase of 69 mg/dl -- over twice as much.
In addition to looking at cholesterol and triglycerides, the Glaxo-funded ESS 40002 study also recorded lactate levels at baseline and after 48 weeks. While people taking AZT experienced an average lactate increase of 2.3 mmol/L, those on d4T had increases over three times higher, at 7.9 mmol/L -- well into the danger region.
Another Glaxo-funded study presented at ICAAC (TARHEEL) looked at what happened when treatment-experienced patients who had been on a regimen containing d4T and were experiencing lipodystrophy switched to either AZT or abacavir. A subset of 16 people with elevated lactates stopped all treatment until blood values showed a return to safer levels. At that time treatment was restarted without d4T. At 48 weeks lactate levels remained controlled within a normal range. Although this small substudy lacks a control arm for comparison, the results give guidance to a clinician seeking to normalize hyperlactatemia in a patient.
Laboratory abnormalities are intriguing for scientists and alarming for clinicians but they are likely to be bewildering to people on HIV treatment. It's often the changes in the way they look in the mirror that matters foremost. Several studies are now starting to connect the dots by evaluating changes in fat deposits under the skin over time. For fat losses in the arms and legs, a technique called DEXA is reliable and well accepted. For increases in fat around the organs, a CT scan is a more acceptable -- and more expensive -- method. For fat loss in the face, patient self-evaluation or photography of facial fat wasting tell the story, but these, unfortunately, do not carry the same weight as more objective readings such as DEXA.
The report that's gotten everyone's attention comes from a large trial that compared nelfinavir to efavirenz at the same time it compared AZT/3TC with d4T/ddI in people starting their first ART regimen. The trial, called ACTG 384, was conducted by the federally funded AIDS Clinical Trials Group. Results were first presented in July, 2002 in Barcelona and an additional analysis of fat loss was presented at the Lipodystrophy Workshop.
The fat loss substudy of ACTG 384, (ACTG 5005s) measured subcutaneous fat by DEXA in the arms and legs of 157 volunteers at baseline. The results from comparisons at weeks 48, 64 and 80 were presented. Overall, people in both the AZT/3TC group and the d4T/ddI group experienced a brief increase in subcutaneous fat during the first weeks after starting treatment. This may be due to a general improvement in constitutional health that follows the initial suppression of viral load. But by week 48, the average amount of subcutaneous fat in the arms and legs had declined by 7.5 percent in those on d4T/ddI, and was starting to drop, but still above baseline, in those on AZT/3TC. By week 80, the d4T/ddI group was down about 15 percent on average, while the AZT/3TC had slipped by about 7 percent.
When the same group was analyzed by whether they were on nelfinavir or efavirenz, not surprisingly, a greater loss of limb fat was associated with being on the protease inhibitor. At week 80, those on nelfinavir plus nukes were down by 18 percent, while those on efavirenz had sustained a 10 percent fat loss on average. A breakout of results by individual three-drug combinations was not shown and the role of d4T relative to ddI in causing fat wasting cannot be determined from this data.
The TARHEEL study that looked at lactates also measured changes in subcutaneous fat by DEXA after the switch was made from d4T. Of the 118 people with lipodystrophy who enrolled, over 80 percent had been on d4T for over two years. Three-quarters of the volunteers switched to abacavir and the rest to AZT. At 48 weeks, the biggest improvement was noted in the arms, with those on abacavir experiencing an average 37 percent increase in subcutaneous fat and those on AZT a 17 percent gain. Fat gains to the legs were less profound, with abacavir associated with a 15 percent average increase and AZT with about half that. However, a comparison of CT scans of visceral fat at baseline and at 48 weeks did not show a significant change. As with the lactate data, this is a before-and-after study with no concurrent control, and other factors may have played a role in these effects. Yet with the weight of evidence now implicating d4T in accelerated subcutaneous fat loss, when is reasonable doubt overruled by unreasonable risk?
It's likely that all of the nucleoside analogs contribute to mitochondrial toxicity -- some in some cells more than others; some sooner than others. But d4T seems to offer a fast track to fat wasting for far too many people who start taking it. Of course, not everyone is affected -- that's the quandary. Some people still have fat and happy cheeks after years on d4T. If their viral load is suppressed and they tolerate their drugs they'll probably see no reason to switch. But people choosing a regimen today have a different decision to face -- and there's no way to predict how things will go. If Zerit accelerates fat wasting, then there may be better first choices for a nucleoside analog.
Many people on treatment and their doctors pronounced Zerit guilty long ago. The next hearing should come when the committee that writes the federal treatment guidelines meets again to redraw their recommendations.
Back to the GMHC Treatment Issues October 2002 contents page.