Background: Lactic acidosis (LA) is a potentially life-threatening complication of antiretroviral (ARV) therapy. Few randomized prospective studies have compared LA between different ARV regimens.
Methods: Characterization of cases of LA (serum lactate >5 mmol/l and arterial pH<7.35 or bicarbonate <20 mmol/l) and symptomatic hyperlactataemia (SH; serum lactate >2.2 mmol/l and symptoms) was made in a randomized open-label 2×2 factorial study of stavudine/lamivudine (d4T/3TC)-based versus didanosine/zidovudine-based therapy and lopinavir/ritonavir-based versus efavirenz (EFV)-based therapy in 1,771 HIV-infected adults initiating therapy between 2004 and 2008.
Results: The LA incident rate was 3.5/1,000 person-years (95% CI 1.8-5.9), and for combined LA/SH was 11.0/1,000 person-years (95% CI 7.9-14.9). There were two deaths (15% mortality) among 13 LA cases; all 11 survivors experienced symptom resolution and started new ARV regimens. LA cases were more likely to be female (OR 7.19, 95% CI 1.84-40.75; P=0.001) and had a higher body mass index (BMI; P<0.0001) compared with non-cases. There was no increase in LA according to ARV regimen, age or CD4(+) T-cell count at randomization. When combined, LA/SH cases (n=41) were more often female (OR 4.76, 95% CI 2.36-10.08; P<0.0001), had increased BMI (P<0.0001), were more likely to be assigned d4T/3TC (OR 3.17, 95% CI 1.50-7.28; P=0.001) and were more likely to be assigned EFV (OR 2.18, 95% CI 1.08-4.61; P=0.026).
Conclusions: Female sex and increased BMI were associated with severe LA in this large randomized trial of first-line ARV in South Africa. While female sex, increased BMI and d4T are previously described risk factors for the development of clinically significant lactate elevations, the independent risk associated with EFV is a novel observation warranting further investigation.