The fat gains experienced after going on antiretroviral therapy may not depend as much on the specific regimen as it depends on whether one had a high viral load before starting treatment, according to a study presented at CROI 2015, in Seattle, Washington.
The study, A5260s, presented by Grace McComsey, M.D., of the Case Western Reserve University, looked for differences in body composition changes in various body compartments (peripheral fat and subcutaneous adipose tissue [SAT], as well as central fat and lean body mass) in people living with HIV initiating treatment on one of two ritonavir (Norvir)-boosted protease inhibitors (atazanavir [Reyataz] or darunavir [Prezista]) or an integrase-based regimen over 96 weeks.
However, the study didn't find any fat gain differences based on treatment regimen. Instead, the researchers found associations with other factors that were present at baseline (when individuals went on treatment). Specifically, having a high viral load when starting treatment was most consistently associated with body composition changes after starting antiretroviral therapy.
"I think it was amazing -- regardless of the regimen that you look at, the fat gains were so much more, if you start with very high viral load, and if you adjust for the inflammation markers, viral load remained by itself associated with higher gains," said McComsey.
Even though there has been a decline in the incidence of lipoatrophy (fat wasting) with the use of nucleoside or nucleotide analogue drugs that are more mitochondria-friendly, increased body fat (central adiposity or lipohypertrophy) continues to occur in people initiating antiretroviral therapy.
The pathogenesis of lipohypertrophy remains unclear. Initially, it was blamed on protease inhibitors, however, studies showed that switching off of protease inhibitors to either non-nucleoside reverse transcriptase inhibitors (NNRTIs), such as efavirenz (Sustiva, Stocrin), or more recently to integrase inhibitors, did not significantly improve central fat. Although, in at least one study, the use of ritonavir-boosted atazanavir was associated with greater gains in both peripheral and central fat compared to the use of efavirenz. However, there had not been any well-powered studies comparing changes in body composition in people initiating protease inhibitors versus those starting on integrase inhibitors.
Therefore, McComsey and colleagues conducted A5260s, a metabolic and cardiovascular sub-study of A5257, which had enrolled more than 1,800 treatment-naive individuals living with HIV with a viral load of at least 1,000 copies/mL. The study had randomized individuals to three NNRTI-sparing regimens: tenofovir/emtricitabine (Truvada) with either atazanavir/ritonavir, darunavir/ritonavir, or raltegravir (Isentress). The participants were also stratified according to baseline viral load ≥ or < 100,000 copies/mL, or having a low or high Framingham 10-year risk score of coronary heart disease.
The sub-study included 328 participants, and had specific exclusion criteria: no known cardiovascular disease, no diabetes and no use of lipid-lowering agents. At baseline and at week 96, the sub-study participants underwent a DXA scan as well as a computed tomography (CT) of their visceral and subcutaneous abdominal fat.
At baseline, measurements were taken of leptin, adiponectin (which are both involved in fat metabolism and regulation) and selected inflammatory markers and monocyte activation markers (IL-6, hs-CRP, D-dimer, sCD14, sCD163).
The baseline characteristics of the study participants were very well balanced between the arms. The median age was 36; 90% of the participants were male and 40% were white non-Hispanic. About one third of participants were in the high viral load strata. The body mass index (BMI) median was 25, and the body composition at baseline was very similar to or in line with other U.S. treatment initiation studies in the era of starting antiretroviral therapy with a higher CD4 count, according to McComsey.
There were two measures of peripheral fat: limb fat by DXA and subcutaneous adipose tissue (SAT) by CT.
- Limb fat change at week 96: All three study arms led to significant increases from baseline to week 96, but there were no significant differences between the three arms.
- atazanavir/ritonavir: 11%
- raltegravir: 20%
- darunavir/ritonavir: 14%
- SAT change at week 96: There were significant increases from baseline in all three study arms with no significant differences between regimens.
- atazanavir/ritonavir: 23%
- raltegravir: 25%
- darunavir/ritonavir: 20%
- Proportion of participants with >10% and >20% limb fat loss at week 96: Overall, 16% to 18% lost at least 10%; and 6% to 8% lost 20% or more of their baseline limb fat with no significant differences between the arms.
There were two measures of central fat: trunk fat by DXA and visceral adipose tissue (VAT) by CT.
- Trunk fat at week 96: There were significant increases in all three arms from zero to 96 weeks, but no significant differences between the study arms.
- atazanavir/ritonavir: 16%
- raltegravir: 29%
- darunavir/ritonavir: 21%
- VAT change at week 96: About a 30% increase from zero to 96 weeks in all three study arms with no significant differences.
- atazanavir/ritonavir: 31%
- raltegravir: 33%
- darunavir/ritonavir: 29%
Lean Body Mass
There were significant increases in lean body mass as well, of 1.2% to 2%, with a trend toward greater gains in lean mass on atazanavir/ritonavir. BMI increased in all the arms by about 3%, but there was no significant difference between arms. There was also no change in VAT to total adipose tissue ratio in either arm or between the arms.
The changes in central fat correlated well with changes in peripheral fat, and all these fat changes correlated well with the lean body mass changes, as well.
Baseline Associations With Changes in Body Composition
The researchers also looked into the association of parameters at baseline with the changes in body composition. Most strikingly, for both SAT and VAT, they found that the people who started in the high viral load strata gained two or three times the amount of fat than people who started in the lower viral load strata, regardless of the treatment regimen. It was notable though that both lower baseline leptin and higher adiponectin were also associated with greater gains in fat.
Demographics, baseline BMI, CD4 or inflammatory markers were not associated with increases in VAT, though higher IL-6 levels at baseline were associated with greater SAT gains in the study.
In contrast, increases in lean body mass were associated with higher IL-6 and D-dimer as well as lower CD4 count and higher HIV viral load at baseline.
"A 30% gain in fat after two years -- that's really bad. It is a short duration of treatment!" McComsey said, regarding the association of fat gain with high baseline viral loads. She stressed the need for more research to determine the cause of lipohypertrophy in people living with HIV.
Other studies presented at CROI 2015, in a thematic discussion titled "Fat Without Borders," also addressed the topic. The SECOND-LINE RCT DXA sub-study found that limb fat gain isn't limited to first-line regimens -- it was observed in both boosted-protease-inhibitor-based and raltegravir-based regimens. Again, there was no significant difference in regimens (including no difference between the regimens that spared or contained nucleoside analogues). This trial, which was conducted in Argentina, India, Malaysia, South Africa and Thailand, did find more demographic differences -- with more limb fat loss in African participants than among Asian participants, and with women being more likely to gain limb fat.
A second study by Kristine M. Erlandson et al, which looked at fat gain after treatment initiation in various countries, found that incremental increases in BMI to the point of being overweight or obese were associated with increases in markers of inflammation over time. Erlandson suggested that unhealthy weight gain could cause inflammation.
Other factors may also lead to both the increase in inflammation and the increase in weight. One possibility is that the cause is similar to what appears to be causing cardiovascular disease -- chronic inflammation caused by both ongoing viral replication and by the microbial translocation (the leaking of gut flora into systemic circulation). Damage to the gut mucosa is a consequence of HIV infection, and having had a higher viral load could lead to more gut damage and more inflammation.
Another possibility is suggested by McComsey's other finding -- that both lower baseline leptin and higher adiponectin were also associated with greater gains in fat. Low leptin and high adiponectin levels are often found among stressed, starving individuals -- and alter the metabolism in ways that tell the body to store calories as fat. HIV replication does alter metabolism; it is quite possible that high viral replication makes the body think that it is starving -- and when antiretroviral therapy removes most of the HIV, it may be very difficult to reset that programming.
Theo Smart is an HIV activist and medical writer with more than 20 years of experience. You can follow him on Twitter @theosmart.