EFV concentrations were satisfactorily described by a one-compartment model with first-order elimination. This model has already been used for adults (4
) to describe the EFV concentration-time course. The following observations support the model's use.
s, and AUCs were consistent with those of previous studies: 1.64 mg/liter, 3.71 mg/liter, and 65.2 mg/liter·h, respectively, in our study compared to 1.18 to 1.45 mg/liter, 4.09 to 5.52 mg/liter, and 60 to 63.6 mg/liter·h in previous studies (7
) (Table ).
(ii) The CL/F
(0.21 liter/h/kg) was consistent with those of previous studies: 0.19 liter/h/kg for 50 children (7
) and 0.30 liter/h/kg for 33 children (21
(iii) The CL/F
(per kilogram of body weight) decreases significantly with age, from 2.5 to 15 years. This is consistent with an increase in hepatic activity (including the metabolism of EFV by CYP2B6) between the ages of 1 and 4 years, exceeding adult levels (9
Three different modeling approaches, i.e., fitted, fixed to an exponent of 0.75, and fixed to an exponent of 1 for body weight on CL/F, could have been used, since the same objective function values were obtained and the age effect was significant in all three cases. CL was finally handled as linearly related to body weight, mainly because using this general approach, the body weight exponent on CL was estimated as 1.13 (closer to 1 than to 0.75) and because doses are given linearly with body weight (in milligrams per kilogram of body weight).
Different relationships between concentrations and efficacy were used in order to adapt dosages for children. In the Pediatric AIDS Clinical Trials Group study (18
), the dose of EFV was adjusted if the AUC from 0 to 24 h was out of the target interval of 190 to 380 μmol/liter·h (corresponding to 60 to 120 mg/liter·h), representing the range from the 50th percentile to twice the 50th percentile of such values in adults receiving 600 mg of EFV per day. Other studies adapt a criterion derived from the concentration-efficacy/toxicity relationship established for adults: for concentrations measured between 8 and 20 h after drug intake, virologic failure was significantly higher in patients with EFV concentrations of <1 mg/liter and central nervous system toxicity was three times more frequent in patients with EFV concentrations of >4 mg/liter (14
). These concentration-effect relationships were not confirmed in children. Only one study showed a concentration-efficacy link in children: 50% of children who had EFV AUC values less than 49 mg/liter·h reached less than 400 copies/ml of HIV RNA by week 8 compared with 86% of children who had AUC values greater than 49 mg/liter·h (P
= 0.01) (7
). In our study, a significantly higher percentage of children had viral load decreases greater than 2 log10
copies/ml after 3 months of treatment when the AUC was greater than 51 mg/liter·h (97% [32/33] versus 71% [10/14]; P
= 0.02) at 2 weeks of treatment. This confirms that the EFV AUC that improves treatment efficacy is around 50 mg/liter·h. The target Cmin
could also be related to efficacy: 97% of children with a Cmin
of >1.1 mg/liter had a viral load decrease greater than 2 log10
copies/ml at month 3 compared to 69% of children with a Cmin
of <1.1 mg/liter (P
= 0.02). The relationship between EFV plasma concentrations and efficacy did not exclude the influence of other factors (such as exposure to other antiretroviral drugs or the genotype of the virus). In this study, seven children had Cmin
s above 4 mg/liter, but none had adverse events related to EFV administration, so no relationship could be established between EFV concentrations and toxicity.
For adults and children, efficacy could be related to both AUC and Cmin
, and studies were not adequate to determine whether Cmin
or AUC is best correlated with efficacy. Theoretically, Cmin
may be more important to provide continuous suppression of viral replication; thus, we considered Cmin
the target pharmacokinetic parameter in our study. In all studies, a high percentage of children fell below the target concentrations (7
). Similarly, with the recommended dose of EFV administered to 46 out of 48 children in the present study, 19% of children had Cmin
s of <1 mg/liter. Among them, 89% weighed less than 15 kg, so we concluded that children weighing 11 to 15 kg were more likely to have insufficient EFV concentrations. In our study, all children weighing less than 15 kg received 200 mg of EFV, the recommended dose for children weighing 13 to 15 kg; there is no recommendation for children weighing less than 13 kg. Two strategies to increase Cmin
s in these youngest children could be considered: increase the once-daily dose or shorten the dose interval to every 12 h. The latter alternative was not considered, although it could avoid increasing EFV exposure, because the major aim of this trial was to test a once-daily combination to increase compliance. According to simulations, to optimize the percentage of children with Cmin
s between 1.1 and 4 mg/liter (related to efficacy and toxicity, respectively), children should receive the following once-daily EFV dose: 25 mg/kg from 2 to 6 years, 15 mg/kg from 6 to 10 years, and 10 mg/kg from 10 to 15 years. These assumptions should be prospectively confirmed, because conclusions were drawn on the basis of only 48 children and no data on toxicity with these increased doses are available.