icon-folder.gif   Conference Reports for NATAP  
  4th Intl Lipodystrophy Workshop
San Diego at Coronado Beach, Sept 22-25, 2002
Back grey_arrow_rt.gif
Body Composition Changes Observed in ACTG 384
Reported by Jules Levin
  This was a study in treatment naive patients. The main findings reported by the study investigators were: ddI/d4T based regimens result in greater loss of limb fat than do AZT/3TC regimens. And, NFV based regimens result in greater loss of limb fat than do EFV based regimens, when combined with NRTIs. Actual percent changes discussed below.
In Barcelona the ACTG reported results from the main study of ACTG 384. You can read those results in a NATAP report from that conference which was written by David Margolis, MD at this link:
At this Lipodystrophy Workshop, Michael Dube reported preliminary results of the sub-study of 384, which looks at body composition changes for study patients. Dube used DEXA scans to evaluate the changes in body composition after patients started the study drugs. DEXAs were done before starting study drugs and at different timepoints thereafter.
The role of different antiretroviral drugs in causing body fat changes is controversial. Although studies have been conducted, we are unsure how much individual drugs contribute to body changes. It remains uncertain how much is contributed by drugs and other risk factors including the immune dysfunction and dysregulation caused by HIV, and by HAART. Other studies have been limited by the effects of prior therapies used before the study making it difficult to identify causation.
A5005s is the metabolic substudy of 384. 980 patients enrolled in 384 and 330 in A5005s. 156 of the 330 had DEXA scans when entering 384, to evaluate body composition before starting drugs in 384. 127 patients had DEXAs at week 48 of the study, and 107 patients had DEXA scans at week 64. The study analysis is ITT (intent-to-treat), which means all patients enrolled were counted in the analysis even if a patient discontinued.
In the main study, patients received either AZT/3TC or d4T/ddI. In addition patients received either nelfinavir (PI), efavirenz (NNRTI) or nelfinavir+efavirenz. After an initial failure of the regimen patients started with they were switched to a second but different regimen. This is a bit complicated but the body composition data is easier to follow.
Dube reported that baseline characteristics (gender, race/ethnicity, age, BMI, CD4 count, and HIV RNA) were similar across treatment groups. At baseline the 156 patients who had DEXAs were on average 36 yrs old, had 260 CD4s, just over 100,000 viral load (5.1 log), BMI 24.1 kg/m2). The baseline characteristics for patients in AZT/3TC and d4T/ddI arms were also similar: 35 yrs, mostly male, CD4 250, 5.4-5.1 log viral load, BMI about 23, amount of limb fat & trunk fat, and 50-60% were white.
On average Limb Fat Declined More on D4T/ddI than AZT/3TC This section discusses the median percent change in limb fat, so not everyone had these changes but these were the average changes. After week 16, the limb fat started to decline in patients in both groups (AZT/3TC and d4T/ddI). In the first 16 weeks limb fat increased, and this could be due to improved health as patients with lowest CD4 count & highest viral load were the ones experiencing increases in limb fat initially. But that trend in increased limb fat stopped at week 16 and limb fat decreased. Limb fat decreased for patients receiving AZT/3TC and for patients receiving d4T/ddI, but the patients receiving d4T/ddI had significantly greater decreased limb fat than patients receiving AZT/3TC at weeks 48, 64 and 80. Due to the increase during the first 16 weeks the patients receiving AZT/3TC saw their limb fat decline to baseline limb fat levels by week 64, and they declined to less than the baseline level at week 80. Patients receiving d4T/ddI declined to baseline limb fat levels by week 32 and contnued declining after that. Whites and non-whites experienced similar changes in limb fat. Dube reported that by week 48 limb fat declined 7.5% from baseline for the patients receiving d4T/ddI. But limb fat continued to decline after 48 weeks and it appeared to decline about 15% from baseline at week 80 for patients vreceiving d4T/ddI. Patients receiving AZT/3TC appeared to experience limb fat decline of about 7% at week 80 compared to baseline.
Trunk Fat
The changes in trunk fat (trunk is part of body from shoulders to waist) were similar between the AZT/3TC & d4T/ddI groups. Trunk fat increased, similar to what occurred with limb fat, during the first 32 weeks for patients receiving AZT/3TC and d4T/ddI but started to decline equally in both groups after 32 weeks. By week 80 trunk fat was the same in patients whether they received AZT/3TC or d4T/ddI, and was about 5% greater than before starting study therapy. But the study did not look at regimen specific trunk fat change. In other words, they did not look at trunk fat change in patients receiving efavirenz plus AZT/3TC and compare it to patients receiving nelfinavir plus AZT/3TC or nelfinavir plus d4T/ddI.
Nelfinavir vs Efavirenz
Limb fat increased in the first 16 weeks in patients receiving nelfinavir and efavirenz. But after week 16 limb fat declined more in the nelfinavir patients, and the difference between the two groups (NFV vs EFV) was significant at week 80. But limb fat declined for patients whether they received nelfinavir or efavirenz, compared to what their limb fat was before starting therapy. Limb fat appeared to decline about 10% from baseline to week 80 for patients receiving efavirenz plus nukes compared to about 18% for patients receiving nelfinavir (p=0.019). This finding is interesting and peaked interest at the Workshop, because although limb fat loss is generally perceived to be associated mostly with nucleosides this finding underscores that it may also be associated with PI use. The association of limb fat loss with PI use received some attention at this Workshop. There is some evidence from studies that although limb fat appears mostly associated with nuke use it can be associated also with PI use. And there is a little evidence that PI use in combination with nukes may cause a synergistic limb fat loss compared to PI sparing regimens.
Trunk fat increased on average by about 8% in the efavirenz patients but after 80 weeks trunk fat in the nelfinavir patients was about the same as when they started therapy. But again trunk fat increase for both efavirenz and nelfinavir patients in the first 32 weeks, but declined in both groups after that. The decline was steeper & bigger for the patients receiving nelfinavir than for patients receiving efavirenz. So by week 80 patients receiving efavirenz had on average experienced about an 8% increase in trunk fat but it appeared as though trunk fat was declining. For patients receiving nelfinavir trunk fat was at week 80 about the same as it was before starting therapy but appeared to still be declining. 80 weeks is a long time for a study but in real life for patients 80 weeks is a relatively short time as many patients started HAART 7 years ago. So based on the decline slopes of trunk fat here, trunk fat may be much less after several years on therapy. In fact, in Carl Grunfeld's preliminary findings from the FRAM study which he reported at Barcelona, he reported that fat loss in the trunk is a problem for patients; and he feels fat wasting throughout the body including the trunk is in general greater than appreciated.
These results are preliminary and have not yet been correlated with metabolic changes (cholesterol, triglycerides, glucose), so when this is done we should have interesting information. Dube emphasized that the results observed for the combination of d4T/ddI cannot be generalized to other d4T regimens such as d4T/3TC.
Dube's Summary
--Limb fat increased early with both NRTI regimens and PI or NNRTI regimens
--At week 48, 64, and 80, randomization to ddI/d4T resulted in greater percent of limb fat loss
--At week 80, randomization to nelfinavir resulted in greater percent loss of limb fat
--No between-groups differences occurred with respect to changes in trunk fat; this not my observation from viewing graphs he presented as I saw less trunk fat in nelfinavir patients than in efavirenz patients by week 48 and this trend continued
--Changes in limb fat correlated positively with changes in trunk fat
--trunk & limb fat changes were comparable between whites and non-whites
--ddI/d4T based regimens result in greater loss of limb fat than do AZT/3TC regimens
--NFV based regimens result in greater loss of limb fat than do EFV based regimens, when combined with NRTIs
--white race may not necessarily predispose to greater loss of limb fat. Prior to this data most observations were that whites tended to experience more fat loss than African-Americans and perhaps Hispanics. I'm not convinced by this one study that whites do not experience more fat loss.