icon star paper   Articles  
Back grey_arrow_rt.gif
 
 
Switching to Abacavir or Trizivir for Lipodystrophy & Lipids
 
 
  "A 48-Week, Randomized, Open-Label Comparison of Three Abacavir-Based Substitution Approaches in the Management of Dyslipidemia and Peripheral Lipoatrophy"
 
AIDS Journal of Acquired Immune Deficiency Syndromes May 1, 2003; 33(1):22-28. G. J. Moyle; C. Baldwin; B. Langroudi; S. Mandalia; B. G. Gazzard.
 
Abstract: The mechanisms by which dyslipidemia and lipoatrophy develop during antiretroviral therapy are not clear. No treatment of lipoatrophy is currently established. This was an open-label randomized study of HIV-positive individuals on a first-line therapy containing stavudine (d4T) with either a protease inhibitor (PI) or nonnucleoside reverse transcriptase inhibitor (NNRTI) and with hypercholesterolemia (defined as total cholesterol >5.2 mmol/L or >180 mg/dL) and/or lipoatrophy and with a viral load of <50 copies/mL. Patients switched d4T to abacavir (ABC) (group 1), a PI or NNRTI to ABC (group 2), or d4T and PI or NNRTI to ABC plus AZT (group 3). Patients were followed-up with fasting blood levels, dual-energy X-ray absorptiometry (DXA), and computed tomography (CT) scans for 48 weeks.
 
Patients were eligible for inclusion if they (1) were HIV-positive and on a first-line therapy containing d4T with either a PI or an NNRTI, (2) had hypercholesterolemia (defined as total cholesterol >5.2 mmol/L or >180 mg/dL) and/or clinical lipoatrophy, and (3) had a viral load <50 copies/mL. Lipoatrophy was assessed as described in the entry criteria for the HIV lipodystrophy case definition study, involving patient- and physician-agreed moderate or severe lipoatrophy at one or more sites.
 
Thirty patients were included, with 27 completing 48 weeks of therapy. One ABC hypersensitivity reaction was the only serious adverse event. All patients' viral loads remained at <50 copies/mL. CD4 cell counts rose in groups 2 and 3 but fell modestly in group 1. Total and low-density lipoprotein cholesterol improved significantly in groups 2 and 3. Triglycerides fell significantly in group 2. In contrast, total, arm, and leg fat mass (by DXA) rose significantly in group 1 but fell modestly in groups 2 and 3. Visceral adiposity (by CT scan) was unaffected in all groups.
 
Authors concluded that abacavir represents a virologically effective replacement for d4T, PI, or NNRTI in persons on successful first-line therapy. Replacement of a PI or NNRTI with ABC leads to modest improvement in both cholesterol and triglycerides. Replacement of d4T with ABC leads to modest improvements in fat mass.
 
Author Reporting
 
Patients were followed-up every 12 weeks for 48 weeks. Ten patients were receiving NNRTIs (7 received efavirenz and 3 received nevirapine), and 20 were receiving PIs (11 received nelfinavir, 6 received indinavir, and 3 received ritonavir/saquinavir). Lamivudine (3TC) was the second nucleoside in 21 patients, and didanosine (ddI) was the second nucleoside in the remaining patients. Use of 3TC and ddI was evenly distributed across groups.
 
Twenty-seven patients completed 48 weeks. Two patients withdrew for personal reasons before week 12 (groups 1 and 2), and 1 patient (3%) discontinued because of ABC hypersensitivity (group 1). No other serious or severe adverse events were reported.
 
All patients retained virologic control with <50 copies/mL through week 48. The median CD4 cell count fell in group 1 from 546 to 499 cells/mm3 and rose in group 2 (from 581 to 710 cells/mm3) and group 3 (from 518 to 610 cells/mm3); however, these differences were not statistically significant.
 
The mean cholesterol level fell significantly from baseline to 48 weeks for the whole group, with an average reduction of 0.89 mmol/L. The changes in total cholesterol were driven by declines in LDL cholesterol, which fell significantly by an average of 0.84 mmol/L for the total group. Groups 2 and 3 experienced significant declines in total and LDL cholesterol, whereas in group 1, total and LDL cholesterol rose nonsignificantly. HDL was unchanged across all groups. Triglycerides decreased significantly in group 2 and nonsignificantly in group 3 but rose nonsignificantly in group 1.
 
Insulin levels rose nonsignificantly when d4T was replaced by ABC (group 1) but tended to fall in arms where the PI or NNRTI (groups 2 and 3) was replaced. Fasting glucose remained unchanged in all groups. The 2-hour glucose readings from the glucose tolerance test fell nonsignificantly in all groups. There were no changes in lactate or testosterone levels throughout the study.
 
Dual X-ray absorptiometry scan data demonstrated significant increases in both arm and leg fat mass in individuals substituting d4T for ABC (group 1). Mean arm and leg fat mass increased by 0.57 kg and 1.08 kg, respectively, from baseline to 48 weeks (increases of 41% and 52%, respectively) but were unchanged or tended to decline in group 2 (declines of 22% and 11%, respectively) and group 3 (declines of 4% and 8%, respectively). These differences were not evident at week 24. Lean mass did not change in any of the groups. Total abdominal adiposity as assessed by CT scanning increased in all groups, and visceral fat mass increased in groups 1 and 2 but fell in group 3. All observed changes were modest 10% at maximum) and not statistically significant.
 
DISCUSSION by Authors
 
This study evaluated three possible approaches to manage lipid, insulin, and morphologic changes in persons on d4T plus PI-based or NNRTI-based first-line antiretroviral regimens. With all approaches, ABC maintained virologic control and significant differences in CD4 cell count were not observed. Previous studies using ABC or ABC plus AZT-based substitutions have indicated that virologic control is maintained in first-line therapy patients but less reliably in individuals with prior incomplete viral suppression of thymidine-based therapy or known archived nucleoside analogue resistance mutations. Pretreatment CD4 count and viral load do not appear to affect the efficacy of ABC-based regimens when ABC is used as a substitution agent in individuals with full viral suppression on first-line therapy.
 
For the management of hypercholesterolemia, total and LDL cholesterol improved significantly when the PI or NNRTI was replaced by ABC. These data are consistent with several previous reports of this substitution approach. Data from a randomized study comparing PI replacement with ABC, nevirapine, or efavirenz found declines in total cholesterol to be greatest in the group switched to ABC. Our data suggest that replacement of d4T with ABC in these circumstances may lead to modest rises in total cholesterol. As with our study, a similar nonsignificant rise in total cholesterol was observed in a larger 24-week randomized study of this approach. These small rises in cholesterol and triglycerides are not of sufficient magnitude to influence cardiovascular risk meaningfully.
 
None of the substitution approaches led to significant improvements in HDL cholesterol, triglycerides, insulin, or lactate over 48 weeks, although Arms 2 and 3, which involved the replacement of a PI or NNRTI with ABC, indicated trends to improvement in triglycerides and insulin consistent with larger studies.
 
Fat mass in both arms and legs improved significantly over 48 weeks when d4T was replaced by ABC. This benefit was not observed when the d4T-based regimen was replaced with one containing AZT, however. Replacement of a PI or NNRTI with ABC did not influence fat mass over 48 weeks. Baseline limb fat mass (combined total of arm and leg fat mass by DXA) was similar in our study to that of patients in a randomized 24-week evaluation of ABC substitution of thymidine analogues. In that study, limb fat mass rose by a total of 11% 24 weeks after substitution compared with a 2% rise in the continuation group. In our study, no benefits were observed by 24 weeks, but after 48 weeks, a substantial increase in limb fat (relative change of 52% in legs and 41% in arms) had occurred.
 
The mechanism by which fat mass increases after d4T substitution is not clear. In our study, inclusion of the alternative thymidine analogue AZT in the new regimen resulted in no accrual of the fat mass benefit, suggesting that thymidine analogues may be contributory to continuation of the syndrome. Lactate values, potentially a marker of mitochondrial toxicity, were affected similarly and nonsignificantly by all substitution approaches, suggesting that this may not be the mechanism involved.
 
Most intriguingly, lipids and insulin levels rose in the group where fat mass recovered but fell in the populations where fat mass did not recover. This suggests that these factors are incompletely (at best) linked. PIs are known to increase lipids and insulin resistance through a number of demonstrated mechanisms, and the removal of PIs is generally associated with some improvement in dyslipidemia. Studies of ABC-based regimens not including thymidine analogues as first-line therapy are underway but have not reported lipid data. Rises in lipids with initial d4T-based therapy are greater than those seen with AZT or tenofovir. Oxidation of lipids is normal in persons on dual nucleoside analogue therapy with or without PI therapy, suggesting that mitochondrial dysfunction is not the mechanism for this observation. Of note, the groups also differ with regard to CD4 count changes. As with the Mitox study, switching to ABC was associated with a modest decline in the median CD4 count (-4 cells/mm3 over 24 weeks in the Mitox study and -47 cells/mm3 over 48 weeks in this study), whereas CD4 values rose (by +51 cells/mm3 at week 24 in the Mitox study and by +129 (group 2) and +92 (group 3) at week 48 in this study). This suggests a possible association between immune recovery mechanisms and fat loss/recovery, which warrants further investigation beyond the scope of our study.
 
In summary, replacement of d4T, PI, or NNRTI in first-line therapy patients with ABC maintains virologic control. Replacing d4T with ABC leads to improvements in limb fat mass, but these improvements are not accompanied by improvements in fasting lipids. Replacement of a PI or NNRTI with ABC is associated with improved fasting cholesterol (with trends to better triglycerides and insulin levels) but without benefits to fat mass.
 
Financial support was provided from St. Stephen's AIDS Trust, with additional funding from Glaxo SmithKline, UK.
 
 
 
 
  icon paper stack View Older Articles   Back to Top   www.natap.org