Long-term pharmacokinetics of amprenavir in combination with delavirdine in HIV-infected children

Five heavily pretreated HIV-infected children were put on amprenavir and delavirdine plus two nucleoside reverse transcriptase inhibitors to boost amprenavir levels and to use the antiretroviral activity of a non-nucleoside reverse transcriptase inhibitor. No data are available about this combination in children. It was well tolerated, and the median reduction in viral load was 1.5 log after 18 months. Delavirdine boosted amprenavir trough levels more than 10-fold, and delavirdine trough levels remained several fold above susceptible HIV strains.

Combination antiretroviral therapy is well established in the respective therapeutic guidelines for HIV-infected children. However, the efficacy and toxicity of only some nucleoside reverse transcriptase inhibitors (NRTI) [^1,2] and the protease inhibitors (PI) nelfinavir [³] and lopinavir/ritonavir [⁴] were investigated in a greater number of paediatric patients. In children with repeated drug failure as a result of the emergence of resistant viruses the best studied drugs in this age group might not work anymore. Therefore, paediatricians are forced to use drugs that have been evaluated solely in adults. Here we report the course of HIV-infected children with multiple drug failure, in whom genotypic resistance analysis demonstrated susceptibility against the PI amprenavir and non-nucleoside reverse transcriptase inhibitors (NNRTI). Efavirenz was investigated in a greater number of children [⁵], and for nevirapine only a small observational study has been reported [⁶]. However, we demonstrated earlier that efavirenz dramatically reduces the plasma concentrations of amprenavir in children [⁷] and nevirapine has the same potential (product Information on viramune). In contrast, delavirdine inhibits the cytochome P450 enzyme [⁸] and increases the plasma concentrations of PIs [⁹] in adults. Experience with delavirdine in children is sparse up to now [¹⁰]. Nevertheless, we decided to treat our patients with a combination of delavirdine and amprenavir and two additional NRTI. At the time of the start of the study the boosting of amprenavir with ritonavir was uncommon. In addition, we also determined the long-term pharmacokinetics of amprenavir and delavirdine in five HIV-infected children over a period of 18 months.

Five HIV-infected children (8-14 years old) were included. The median age was 9 years (range 5.5-11 years). The median time of antiretroviral pre-treatment was 37 months (range 34-50 months) and the median number of previous antiretroviral drugs was four NRTI (range 4-5) and three PI (range 1-4). Patients were treated with antiretroviral combinations including amprenavir, delavirdine and two NRTI (3 × abacavir/stavudine and 2 × abacavir/lamivudine). The doses of amprenavir and delavirdine were set at 40 mg/kg bodyweight per day twice a day and were adjusted for weight gain (product Information in agenerase) [¹⁰]. For the NRTI the recommended dosages were administered. The median viral load at beginning of the new regimen was 5.1 log₁₀ (range 4.3-5.2 log₁₀). The median absolute CD4 cell count was 699 cells/μl (range 666-1878 cells/μl) and the median relative CD4 cells were 28% (range 16-31%).

A pharmacokinetic analysis was performed 4 weeks, 12 months and 18 months after changing to the new regimen. Blood samples were drawn after fasting overnight at timepoints 0, 2, and 4 h after directly observed intake of the drugs. The level at timepoint 0 h was considered to be the trough level. The time of the last intake of drugs the day before was recorded, and was between 12 and 13.5 h before drawing the first blood sample (trough level). Amprenavir and delavirdine levels were determined by LC-tandem mass spectrometry [¹¹]. The lower limit of quantification was 78 ng/ml.

The area under the time concentration curve from 0 to 4 h (AUC_0-4) was calculated by the trapezoidal rule.

All five patients were treated over the whole study period of 18 months. Adherence was reassured every visit. None of the patients suffered from liver or renal disease during the time of the study. Three out of five children developed a mild transient rash for a few days 1-2 weeks after the start of the treatment, which was the only side-effect of the current regimen.

At the first timepoint the median amprenavir trough level was 2503 ng/ml (range 1931-5194 ng/ml). After 12 and 18 months the trough level decreased markedly but stabilized at a median of 595 ng/ml (range 464-1777 ng/ml). Delavirdine trough levels varied over a wide range during the study period (range 289-5213 ng/ml).

The results of the AUC for all patients are shown in Fig. 1. The AUC_0-4 and trough levels of amprenavir decreased 66 and 78% after 12 months, and 62 and 75% after 18 months in comparison with initial values, and stabilized thereafter. In contrast, the mean AUC_0-4 and trough levels of delavirdine were only 18 and 14% lower after 12 months and approximately 1.5-fold and threefold higher after 18 months. This tendency was observed for both drugs in every single patient.

Fig. 1

The median reduction in viral load was 2.5 log₁₀ (0.6-2.9 log₁₀) after 3 months, 1.6 log₁₀ (-0.5-3.4 log₁₀) after 12 months and 1.5 log₁₀ (-0.5-3.4 log₁₀) after 18 months of therapy. Accordingly, the median absolute CD4 cell counts were 818 cells/μl (548-1471 cells/μl), 960 cells/μl (674-1818 cells/μl) and 655 cells/μl (501-1559 cells/μl) at 3, 12 and 18 months.

The combination of amprenavir and delavirdine with two NRTI was well tolerated, and none of the patients suffered from an AIDS-defining event. Furthermore, we observed that delavirdine boosted amprenavir trough levels more than 10-fold above the trough levels observed in amprenavir monotherapy (product information on agenerase). The plasma levels of delavirdine varied widely, but trough levels were always several fold above susceptible HIV strains, and peak levels were approximately 10% of the maximal measured levels documented in adult studies (product information on rescriptor). This combination seems to be antiretrovirally effective and safe. Moreover, adherence and tolerance was excellent. None of the children complained of the pill burden. Overtime amprenavir levels decreased after 12 months, possibly as a result of the induction of cytochrome enzymes, but stabilized thereafter still above the recommended trough levels. To the best of our knowledge, no report has so far been published on children about the combination of PIs such as amprenavir and delavirdine. As efavirenz, besides lowering amprenavir blood levels [⁷], might cause central nervous system toxicity (product information on sustiva) and nevirapine might cause liver toxicity (product information on viramune), there might be an indication for this drug in children. It is therefore justified to evaluate the use of delavirdine as a 'booster drug' for PI in a greater number of HIV-infected children.

Acknowledgements

The authors are indebted to all patients participating in this study, and would like to thank G. Strotmann and S. Schlieben for excellent cooperation in collecting the plasma samples.