icon-folder.gif   Conference Reports for NATAP  
  9th Conference on Retroviruses and Opportunistic Infections
Seattle, Washington, February, 2002
Back grey_arrow_rt.gif
Switching Antiretroviral Therapy for Lipoatrophy
David Alain Wohl, MD - University of North Carolina AIDS Treatment and Research Center
  Thousands of HIV-infected persons have been subjected to observation, surveying, randomization, DEXA, CT, inch-pinching and measurements of various body parts with non-stretch tape, yet a full understanding of the links between antiretroviral therapies and fat redistribution, particularly fat wasting, remains elusive. Some data do implicate nucleoside analogues as contributing to lipoatrophy - with some agents worse than others - but there may be additive or synergistic effects of concomitant use of protease inhibitors. (editorial note: additional potential contributing factors to lipoatrophy include dysfunction of the immune system due to HIV and immune reconsitution from HAART, insulin resistance, elevated cholesterol and triglycerides, aging, genetic predisposition, and perhaps others). It was only last year at the 8th CROI that we were warned that altering HIV therapies based solely on metabolic considerations is tricky business and was premature. The rationale for this concern was that since antiretrovirals are almost always used in combination finding the culprit agent in the mix that once removed would lead to the resolution of the complication in question has proved difficult.
One of the most straightforward ways in which to examine the metabolic effects of a particularly suspect agent is to do what seems most obvious - remove it and replace it with an agent from a different antiretroviral class considered less likely to produce the complication and see if the problem improves. Such studies have demonstrated improvement in lipids and glucose related parameters, almost always following a switch from a protease inhibitor. Body shape changes, however, are slower to change and more difficult to assess. Despite these challenges, three switch studies presented at the 9th CROI have tackled lipoatrophy and each reported early improvements in fat wasting following the switch from d4T, or less commonly ZDV, to abacavir (ABC). ABC is an obvious candidate to study since it has been found in the laboratory to be less toxic to mitochondria and therefore, if you are an adherent to the Mitochondria-Metabolic Complications Connection Theory, less likely to contribute to fat changes.
In the largest study (Grace McComsey, et al Abstr 701-T), sponsored by ABC maker GlaxoSmithKline (GSK), 118 subjects (83% male, 65% white) receiving d4T with HIV RNA levels below 400 copies/mL were switched from d4T to either ABC (if ZDV experienced or intolerant) or to ZDV/3TC fixed dose combination (if ZDV naive). All subjects had to have lipoatrophy by clinician or subject report, or symptoms of hyperlactatemia with lactate levels elevated above 2.2 mmol/L, or a lactate level of more than 3.3 mmol/L. Only 16 subjects entered with elevated lactate, therefore, the majority of subjects had lipoatrophy. Of those enrolled, 86 (73%) switched d4T to ABC. More than 80% of subjects had been on d4T for over 2 years. Total body DEXA scans were obtained at baseline, week 12 and week 24 along with a single slice CT scan of the abdomen, anthropometric measurements and a self-reported body image survey. At six months, there was an increase in fat by 25% in the arms, 6% in the legs and 9% in the trunk as determined by DEXA. Despite the encouraging DEXA results, the percentage of subjects reporting positive body shape changes at week 24 for the face, legs, arms and buttocks were 27%, 21%, 21% and 11%, respectively. Approximately, 60% reported no change in their arms and half of subjects reported no facial fat changes at six months. When asked about the seemingly muted self-reported improvements, Dr. McComsey noted that at baseline many subjects reported recent worsening of fat changes and that during the study fewer patients reported advancing lipoatrophy and a greater proportion described either no change or improvement in fat wasting. This suggests the switch may have exerted primarily a stabilizing effect (i.e. preventing fat wasting from worsening), at least at 24 weeks following the switch.
The switch was well tolerated. Only one subject experienced virologic failure but there were 5 cases of ABC hypersensitivity. Lactate levels decreased in both subjects with and without elevated lactate levels at baseline. The study continues to collect data up to 48 weeks and these results should be presented soon.
Another study investigating nucleoside switching conducted in Australia was reported by Andrew Carr, et al (Abstr 32). In this randomized study, 111 subjects with clinician confirmed moderate to severe lipoatrophy while receiving d4T or ZDV, a viral load less than 400 copies/mL and no prior ABC experience were assigned to replace only their d4T or ZDV with ABC versus continuation of all pre-study therapy. d4T was a pre-entry nucleoside in 84% of the subjects. Almost all the subjects were male. About half the subjects were receiving a protease inhibitor and the remainder a non-nucleoside reverse transcriptase inhibitor. At baseline, a surprisingly high proportion of subjects, 85%, had truncal fat accumulation. Median d4T treatment duration was approximately 3 years.
At 24 weeks there were significant increases in arm, leg and buttock fat as demonstrated by DEXA and CT scanning in the ABC group compared to the control subjects who did not switch. When arms and legs were considered together, there was approximately a 10% increase in fat from baseline in the ABC receiving subjects compared to a 2% gain in the control subjects (p=0.016). Arms and legs in this study had similar increases in fat - in contrast to the GSK study that demonstrated a greater fat gain in the upper compared to lower extremities. Subcutaneous but not intra abdominal fat was also seen to increase following switch to ABC. Unfortunately, there was little change in self-reported lipoatrophy. Five subjects (10%) experienced ABC hypersensitivity reactions and there were two heart attacks among control subjects, one of which was fatal. Not surprisingly, lactate levels decreased following the change to ABC. Virologic suppression was more likely to be maintained among those who switched compared to those who continued pre-study entry therapy (98% versus 82%).
Lastly, in another, more complicated Australian switch study M. John, et al (Abstr 700-T) sought to examine whether the switch from both a protease inhibitor and a nucleoside improves or prevents lipoatrophy. The study randomized 40 subjects with HIV RNA levels below 400 copies/mL who were being treated with either: d4T/3TC/indinavir or d4T/3TC/nelfinavir or ZDV/3TC/indinavir to continue to this therapy or switch their regimen to ZDV/3TC/ABC. DEXA determined changes in fat of the arms and legs at weeks 24 and 48 were the primary endpoints of the study. Modest changes in limb fat were described in the switch arm compared to the continued therapy arm. At week 48, an increase of 0.2 kg (~5%) of fat was observed in the upper extremity (arm) among those who switched compared to an increase in fat in controls of 0.06 (p=0.03). The arm normally contains about 7 kg of fat, therefore these changes are not likely to stimulate the purchase of a new wardrobe, although at baseline the fat content of the arms and legs of the subjects were each below 4 kg. Similar results were observed in fat changes of the leg with an increase in fat of 0.16 kg in the switch subjects compared to continued fat loss of 0.15 kg in the continued therapy group (P=0.03).
Importantly, those who switched from both a protease inhibitor and d4T had the best gains in arm fat. One subject who switched experienced virologic failure and there were three ABC hypersensitivity reactions. Intra-abdominal fat and total cholesterol, LDL, HDL and lactate levels did not change from baseline to week 48.
In summary these studies indicate that a switch of nucleosides specifically, d4T to ABC, leads to increases in limb fat. However, the changes in fat seen were modest and, in general, were not as well appreciated by the subjects as they were by the DEXA machines. It is important to recognize, though, that the majority of the data presented were obtained after only 24 weeks following the switch in therapy. Greater reversal of lipoatrophy may be noted with longer follow-up. Additionally, these studies, while not demonstrating massive return of fat over 24 weeks, do suggest that the switch may forestall further wasting of fat, at least as reported by the subject. It is notable that the Carr study did not demonstrate significant worsening of lipoatrophy among those not switching but the McComsey and John studies did. The rate and extent to which fat wasting continues likely has much to do with whether the subject has already îmaxed-out╣ the degree to which they have loss fat.
Unfortunately, there were limited data regarding reversal of facial fat loss - one of them most disturbing aspects of lipoatrophy for many patients. Measurements of the changes in the fat of the face are not as well developed as they are for other parts of the body. However, self-report may be the most important determinant of success of interventions designed to ameliorate this problematic complication.
Each of the studies presented emphasized the substitution of a nucleoside for ABC. Whether other alterations in the composition of antiretroviral therapies can lead to changes in peripheral fat is unclear. The AIDS Clinical Trials Group is conducting a study comparing the switch from d4T or ZDV to ABC versus the complete elimination of nucleosides and initiation of a protease inhibitor and non-nucleoside based regimen among persons with lipoatrophy. This study, ACTG 5110, is open now and provides a unique opportunity to add tremendous insight into a vexing and concerning problem.
Momentum is building for rational switches in therapy to reduce or reverse lipoatrophy. These studies indicate we may be on the right track when we eliminate certain nucleosides in favor of others. The gains after only 6 months are admittedly modest but as one person living with lipoatrophy shared with me, "A 10% increase in fat, its better than losing 10%." Further, these studies have helped to zero in on the agents most likely to be associated with fat loss during therapy and which can be avoided when the composition of initial therapy is being devised.
A last word about another study that is relevant to the switch story. Martinez and colleagues in Spain (Abstr LB17) also conducted a large study involving antiretroviral switching that was not motivated by metabolic complications but which provides important information for those choosing to rid a successful regimen of a protease inhibitor. The study involved 460 patients receiving a protease inhibitor along with two nucleosides who had plasma HIV RNA levels below 200 copies/mL for at least 6 month. These subjects were randomized to stop their protease inhibitor and switch to efavirenz (EFV) or nevirapine (NVP) or abacavir (ABC) - the three popular PI-sparing agents. (It is interesting to note that an almost identical study being conducted by yours truly and the ACTG in the U.S. closed early due to almost complete lack of enrollment.)
After a year following the switch, virologic suppression was not significantly different when analyzed by intent-to-treat (74%, 78% and 77% for EFV, NVP and ABC, respectively). By on-treatment analysis, there seemed to be a trend for less virologic control among those assigned ABC with 94% of the EFV and NVP randomized subjects suppressed compared to 87% of those on ABC (p=0.6). Much of the virologic failures in all the arms could be explained by previous nuceloside treatment prior to the regimen on which subjects entered the study. ABC was better tolerated than the non-nucleoside agents studied. Additionally, the proportion of subjects with cholesterol levels below 240 mg/dL was much lower in the ABC arm (6%) compared to the non-nuke arms (22-24%), however, this may be misleading as the levels of HDL (so called, "good cholesterol") rose in the EFV and NVP arms but decreased in the ABC arm.
This first head-to-head study of switching to these PI-sparing antiretrovirals indicates that the switch to any of the three agents incurs similar risks of losing viral suppression. This is important news to those with PI-related metabolic complications eyeing a switch.