Effect of reducing the dose of stavudine on body composition, bone density, and markers of mitochondrial toxicity in HIV-infected subjects: a randomized, controlled study

BACKGROUND: Stavudine is widely used in developing countries. Lipoatrophy and mitochondrial toxicity have been linked to stavudine use, but it is unclear whether switching to a lower dose can reduce these toxicities while maintaining human immunodeficiency virus (HIV) suppression. METHODS: HIV-infected subjects receiving standard-dose stavudine with undetectable HIV type 1 RNA for < or =6 months were randomized (ratio, 3:2) to receive one-half of the stavudine dose (switch arm) or to maintain the dose (continuation arm) while continuing to receive all other prescribed antiretrovirals. The following measurements were obtained at baseline and week 48: fasting lactate, pyruvate, and lipid levels; results of whole-body dual-energy x-ray absorptiometry; and mitochondrial DNA (mtDNA) measurements in fat and peripheral blood mononuclear cells. Change from baseline to week 48 was compared within and between groups. RESULTS: Twenty-four patients (79% of whom were men and 79% of whom were African American; median age, 45 years) were enrolled in the study, 15 were enrolled in the switch arm, and 9 were enrolled in the continuation arm. The median duration of stavudine treatment was 55 months (range, 21-126 months). The median CD4 cell count was 558 cells/mm(3) (range, 207-1698 cells/mm(3)). At baseline, the study arms had similar demographic characteristics and laboratory indices, except for body mass index, total lean body mass, and triglyceride levels (all of which were higher in the switch arm). Three patients (2 in the switch arm) discontinued the study because of study-unrelated reasons. CD4 cell counts remained unchanged. At 48 weeks, 6 patients (4 [27%] in the switch arm and 2 [22%] in the continuation arm) had detectable HIV RNA levels (median, 972 copies/mL; range, 60-49,400 copies/mL). All patients with detectable HIV RNA levels reported significant lapses in treatment adherence; none exhibited mutations in HIV genotype. After the treatment switch, significant changes from study entry to week 48 were noted only for lactate level (median change, -0.27 mmol/L; range, -1.2 to 0.25 mmol/L; P = .02) and fat mtDNA (median change, 40 copies/cell; range, -49 to 261 copies/cell; P = .02). In the continuation arm, a significant loss of bone mineral density was seen at week 48 (median change, -1.7%; range, -6.3% to 0.8%; P = .02). The only significant between-group difference was the change in bone mineral density from baseline (P = .003). CONCLUSIONS: Reducing stavudine dose by one-half increased fat mtDNA and decreased lactate levels, suggesting improvement in mitochondrial indices while preserving HIV suppression in subjects who maintained adherence. A significant loss of bone mineral density was seen in patients receiving standard-dose stavudine but not in those receiving low-dose stavudine. These results suggest that switching to low-dose stavudine may improve mitochondrial indices while maintaining virological suppression.