Replacing stavudine by abacavir reduces lactate levels and may improve lipoatrophy

Stavudine has been associated with lipoatrophy and hyperlactatemia, and its removal might improve morphological and metabolic changes. In a prospective case-control study, serum lactate levels and lipoatrophy were examined after replacing stavudine by abacavir in patients on successful antiretroviral therapy. This intervention was safe and provided significant reductions in lactate levels over 6 and 12 months. A trend towards an improvement of lipoatrophy was also noted.

Lipodystrophy is one of the most disappointing adverse events related to the use of highly active antiretroviral therapy (HAART). Its pathogenesis and management are challenging issues for which there are still no clear answers [^1,2]. Both protease inhibitors (PI) [³] and nucleoside analogues (NA) [⁴] seem to be involved in its pathogenesis, whereas little is known about the role played by non-nucleoside analogues (NNA) [⁵]. Although definitions differ widely, lipodystrophy typically refers to both fat accumulation and fat wasting or lipoatrophy [²]. Lipoatrophy has been associated mainly with NA, and might result from mitochondrial damage [^6,7]. The inhibition of the mitochondrial DNA polymerase-γ by NA leads to mitochondrial DNA depletion, which subsequently enhances anaerobic glycolysis and leads to hyperlactatemia. In fact, the measurement of serum lactate levels has been proposed for the screening of mitochondrial toxicity in HIV-positive individuals receiving NA [⁸]. Dideoxinucleosides are the most potent inhibitors of mitochondrial DNA polymerase-γ and, not surprisingly, have been associated with the highest rates of hyperlactatemia [⁹]. More recently, cross-sectional studies have strongly implicated stavudine in the development of both lipoatrophy and hyperlactatemia [^10-13] and have found a positive correlation between lipoatrophy and lactate levels [⁹].

Once lipoatrophy has developed, its reversion is quite difficult, although different approaches may be taken. A replacement of antiretroviral agents associated with the greatest risk of lipoatrophy by others with a lower risk has been proposed, and preliminary studies have provided encouraging early evidence with regard to the reversibility of both fat wasting and hyperlactatemia using this approach [^11-13]. In a recent randomized study [¹⁴], a mild improvement in lipoatrophy was noticed after 24 weeks of switching stavudine or zidovudine to abacavir; however, no reductions in lactate levels were noted.

We present here the results of a large prospective study in which serum lactate levels and lipoatrophy were both assessed at 48 weeks in two groups of patients, one in which stavudine was replaced by abacavir and another in which patients remained on their original stavudine-containing regimen.

For the purpose of this analysis, only 112 of the original 146 HIV-infected patients originally included in the study were selected. Individuals who failed therapy, were lost to follow-up, or developed abacavir hypersensitivity or intolerance were not considered. All patients were asymptomatic at the time being recruited into the study, and complained of lipoatrophy, although being with a regimen that included stavudine for longer than 12 months. Subcutaneous fat loss was self-reported and confirmed by physical examination. All patients had plasma HIV-RNA levels below 50 copies/ml and CD4 cell counts above 200 cells/mm³ at study entry.

In 49 patients (cases), stavudine had been replaced by abacavir, and this regimen had been maintained for 12 months. The remaining 63 patients (controls) continued under the same stavudine-containing regimen. In both groups, no other changes in the rest of the antiretroviral agents were performed during the study period.

After 6 and 12 months of follow-up, serum lactate levels, plasma HIV-RNA levels and CD4 lymphocyte counts were assessed. In a subgroup of subjects (22 cases and 12 controls), lipodystrophy was further assessed by electrical bioimpedance (measuring total body fat and lean body mass) and anthropometry (quantifying skinfolds and perimeters) at baseline and at 12 months.

Fisher's exact tests or chi-square tests were used, when appropriate, for the comparison of categorical variables. A two-tailed t-test was used to compare continuous variables. Changes from baseline at each assessment were tested by the two-sample t-tests. Only P values below 0.05 were considered significant.

Baseline demographics, immunological, virological and metabolic parameters were well balanced among the treatment arms (Table 1). There were no significant differences between groups regarding concomitant antiretroviral therapy (cases didanosine in 50%, lamivudine in 25%, NNA in 82%, and PI in 18%; and controls didanosine in 40%, lamivudine in 37%, NNA in 82%, and PI in 21%). Mean serum lactate levels in both groups were slightly elevated at baseline (2.5 ± 1 and 2.4 ± 0.7 mmol/l in cases and controls, respectively). Median plasma HIV-RNA values and CD4 cell counts remained constant throughout the study period, without significant differences between treatment arms.

Table 1

Significant reductions in serum lactate levels were recorded in cases with respect to controls and with respect to baseline (Fig. 1). This decline in serum lactate levels was already manifest at 6 months, although it became more apparent at 12 months. A slight but significant reduction in serum lactate levels was found in controls only at 12 months, which could be explained by body compensatory mechanisms developed over time [¹⁵].

Fig. 1

In the subgroup of 34 patients in whom anthropometry and bioelectrical impedance were performed, no significant differences were observed comparing both treatment arms nor with respect to baseline (Table 1). However, there was a trend towards higher perimeter and skinfold values at 12 months in cases and to lower values in controls (Table 1). Likewise, cases showed a trend towards fat gain, whereas controls showed a significant decrease in total body fat and percentage of body fat. Body weight remained unchanged in both groups.

Our results confirm recent findings of a benefit on peripheral fat as early as 24 weeks after replacing stavudine by abacavir. In our study this benefit was extended over 48 weeks. Moreover, we demonstrated a progressive reduction in lactate levels using this intervention. Given the positive correlation between hyperlactatemia and lipoatrophy, sustained reductions in lactate may translate into further improvements in fat wasting. Larger studies with longer follow-ups are warranted to confirm our findings and to assess whether a complete reversal of lipoatrophy may be achieved in some individuals. In the meantime, and considering the significant fat loss recorded at 12 months in those of our patients who continued on stavudine, the association between stavudine and lipoatrophy should no longer be ignored.

In conclusion, lipoatrophy is a difficult to manage adverse event of antiretroviral therapy, and as long as NA remain an essential part of HAART, replacing stavudine by abacavir may permit a reduction in lactate levels and may slightly improve fat wasting in patients with stable HIV disease.