Correspondence: Dr Thomas N. Kakuda, Janssen Research and Development, Titusville, NJ 08560-0200, USA. Tel: 001 609 730 7528; e-mail: email@example.com
Etravirine is a substrate and inducer of cytochrome P450 (CYP) 3A and a substrate and inhibitor of CYP2C9 and CYPC2C19. Darunavir/ritonavir is a substrate and inhibitor of CYP3A. Artemether and lumefantrine are primarily metabolized by CYP3A; artemether is also metabolized to a lesser extent by CYP2B6, CYP2C9 and CYP2C19. Artemether has an active metabolite, dihydroartemisinin. The objective was to investigate pharmacokinetic interactions between darunavir/ritonavir or etravirine and arthemether/lumefrantrine.
This single-centre, randomized, two-way, two-period cross-over study included 33 healthy volunteers. In panel 1, 17 healthy volunteers received two treatments (A and B) in random order, with a washout period of 4 weeks between treatments: treatment A: artemether/lumefantrine 80/480 mg alone, in a 3-day course; treatment B: etravirine 200 mg twice a day (bid) for 21 days with artemether/lumefantrine 80/480 mg from day 8 (a 3-day treatment course). In panel 2, another 16 healthy volunteers received two treatments, similar to those in panel 1 but instead of etravirine, darunavir/ritonavir 600/100 mg bid was given.
Overall, 28 of the 33 volunteers completed the study. Co-administration of etravirine reduced the area under the plasma concentration–time curve (AUC) of artemether [by 38%; 90% confidence interval (CI) 0.48–0.80], dihydroartemisinin (by 15%; 90% CI 0.75–0.97) and lumefantrine (by 13%; 90% CI 0.77–0.98) at steady state. Co-administration of darunavir/ritonavir reduced the AUC of artemether (by 16%; 90% CI 0.69–1.02) and dihydroartemisinin (by 18%; 90% CI 0.74–0.91) but increased lumefantrine (2.75-fold; 90% CI 2.46–3.08) at steady state. Co-administration of artemether/lumefantrine had no effect on etravirine, darunavir or ritonavir AUC. No drug-related serious adverse events were reported during the study.
Co-administration of etravirine with artemether/lumefantrine may lower the antimalarial activity of artemether and should therefore be used with caution. Darunavir/ritonavir can be co-administered with artemether/lumefantrine without dose adjustment but should be used with caution.
In 2010, an estimated 1.8 million people died from HIV/AIDS, while 1.1 million people died from malaria [1, 2]. The region with the highest prevalence of HIV/AIDS – sub-Saharan Africa – also has the highest prevalence of malaria . HIV infection increases the risk of malarial infection and HIV-infected individuals with severe immunodepression are also at increased risk of severe malaria and death .
The fixed dose combination treatment of artemether/lumefantrine is recommended for acute treatment of uncomplicated malaria caused by Plasmodium falciparum [5, 6]. The fast-acting artemether component lowers the parasite burden, while the more slowly eliminated lumefantrine component eliminates residual malarial parasites .
First-line treatment for HIV infection normally includes two nucleoside/nucleotide reverse transcriptase inhibitors (NRTIs), with either a nonnucleoside reverse transcriptase inhibitor (NNRTI) or a ritonavir-boosted HIV protease inhibitor (PI/r) [8, 9]. Second- or third-line treatment can include combinations of NRTIs with PIs such as darunavir/ritonavir .
There is the potential for drug–drug interactions when using artemether/lumefantrine with antiretrovirals . Artemether and lumefantrine are primarily metabolized by cytochrome P450 (CYP) 3A; artemether is also metabolized to a lesser extent by CYP2B6, CYP2C9 and CYP2C19. Artemether has an active metabolite, dihydroartemisinin, which is converted to inactive metabolites by glucuronidation . In addition, artemether is known to auto-induce its own metabolism.
Use of the NNRTIs nevirapine or efavirenz has been shown to lower the pharmacokinetic exposures of artemether [12-16]. Reduced levels of artemether from this drug interaction may lower the antimalarial efficacy of this treatment . Use of PI/r may inhibit CYP3A metabolism. Three studies have shown increases in pharmacokinetic exposures of lumefantrine when combined with lopinavir/ritonavir [16-18]. Increased lumefantrine exposure might be associated with the QT prolongation .
In this study, the drug–drug interaction between artemether/lumefantrine and two antiretrovirals – the NNRTI etravirine and the PI/r darunavir/ritonavir – was investigated. Etravirine has been evaluated mainly in treatment-experienced patients, at the dose of 200 mg twice daily (bid) . Etravirine is a substrate and inducer of CYP3A and a substrate and inhibitor of CYP2C9 and CYPC2C19; it also inhibits P-glycoprotein . Darunavir/ritonavir has been evaluated in treatment-experienced patients at the dose of 600/100 mg bid [22-25]. Darunavir/ritonavir is a substrate and inhibitor of CYP3A. Based on these effects on metabolism, interactions between etravirine or darunavir/ritonavir and artemether/lumefantrine were expected.
The primary objectives of the study were (1) to determine the effects of etravirine or darunavir/ritonavir on the pharmacokinetics of artemether and its metabolite dihydroartemisinin after single and multiple doses; and lumefantrine after multiple doses in healthy volunteers and (2) to determine the effects of artemether/lumefantrine on the steady-state pharmacokinetics of etravirine or iso and darunavir/ritonavir.
This randomized, two-way, two-period, cross-over study in healthy volunteers was conducted at a single centre in Warsaw, Poland. All volunteers signed written informed consent forms before study procedures were initiated. The protocol was approved by the local Ethics Committee.
Volunteers were divided between two panels: panel 1 evaluated artemether/lumefantrine with or without etravirine, while panel 2 evaluated artemether/lumefantrine with or without darunavir/ritonavir (Fig. 1).
All patients in panels 1 and 2 received two treatments (A and B) with a washout period of 4 weeks between treatments. Treatment A was artemether/lumefantrine administered over 3 days [six doses of four tablets (20/120 mg)] at 0, 8, 24, 34, 48 and 60 h for both panels 1 and 2.
In panel 1, treatment B was etravirine 200 mg bid for 21 days with artemether/lumefantrine 80/480 mg from day 8 (3-day treatment course).
In panel 2, treatment B was darunavir/ritonavir 600/100 mg bid for 21 days with artemether/lumefantrine 80/480 mg from day 8 (3-day treatment course). The first four subjects were sequenced to treatment B before randomization as QT prolongation was an anticipated concern. After careful evaluation of these subjects with no apparent concerns, randomization commenced.
All treatments were administered with food (i.e. within 10 min of completing a standardized meal) as food increases the absorption of artemether, darunavir, etravirine and lumefantrine.
The pharmacokinetics of darunavir/ritonavir and etravirine over 12 h were assessed at day 8 and day 11.
The pharmacokinetics of artemether and dihydroartemisinin over 8 h were assessed after the first intake and > 72 h after the last intake of artemether/lumefantrine; the pharmacokinetics of lumefantrine over 264 h were assessed after the last intake of artemether/lumefantrine. Plasma concentrations of artemether, dihydroartemisinin and lumefantrine were evaluated using an in-house method. Plasma concentrations of etravirine or darunavir/ritonavir, as applicable, were assayed using validated liquid chromatography tandem mass spectrometry methods .
Adverse events were reported by the subjects for the duration of the study. Special attention was paid to those subjects who discontinued for an adverse event, or who experienced a serious adverse event. Blood samples were collected for evaluation of serum chemistry and haematology at baseline and during the study. Electrocardiograms (ECGs) were recorded in triplicate (in the supine position and after at least 5 min of rest). During the co-administration of darunavir/ritonavir and artemether/lumefantrine, subjects were closely monitored for prolongation of the Fridericia-corrected QT interval (QTcF) .
Pharmacokinetic parameters [maximum plasma concentration (Cmax), minimum plasma concentration (Cmin) and area under the plasma concentration–time curve (AUC)] of each drug were obtained using noncompartmental analysis with extravascular input (WinNonlin Professional 4.1; Pharsight, Mountain View, CA). AUC was determined using the linear-linear trapezoidal rule. The pharmacokinetic parameters were then compared using a linear mixed effects model, using log-transformed data and controlling for treatment, sequence and period. A 90% confidence interval was constructed around the difference between least squared means of test and reference. The results are presented as least square mean (LSM) ratios and associated 90% confidence intervals (CIs), with the data transformed back to the original scale.
A total of 33 subjects were enrolled in the study and 28 (84.8%) completed the study. Seventeen subjects were randomly assigned to panel 1 (etravirine) and 16 to panel 2 (darunavir/ritonavir). Of the five subjects (15.2%) who discontinued the study, there were three in panel 1 and two in panel 2. The reasons for discontinuation were adverse events (three subjects) and withdrawal of consent (two subjects).
All subjects were healthy Caucasian nonsmoking men. The median age was 27 years (range 21–54 years). No relevant differences in age or body mass index were observed between the randomization groups in either of the panels.
Panel 1: artemether/lumefantrine with or without etravirine
Table 1 and Figure 2 show the summary pharmacokinetic results from the etravirine part of the study. After single-dose co-administration, etravirine lowered the AUC8h of artemether by 52% and that of dihydroartemisinin by 13% (Fig. 2a,b). After multiple doses, co-administration of etravirine lowered the AUClast of artemether by 38% and that of dihydroartemisinin by 15% and the AUC264h of lumefantrine by 13%. The mean AUC12h of etravirine was 8492 ng/h/mL when dosed alone, vs. 9522 ng/h/mL when co-administered with artemether/lumefantrine, with no significant difference in exposure between the two treatments (LSM ratio 1.10; 90% CI 1.06–1.15).
Table 1. Mean (standard deviation) area under the plasma concentration–time curve (AUC) of artemether (ART) and dihydroartemisinin (DHA) after single and multiple doses and lumefantrine (LUM) after multiple doses, with and without etravirine (ETR)
ART/LUM + ETR (test)
LSM ratio (90% CI)
CI, confidence interval; LSM, least square mean.
Artemether AUC8h (ng⋅h/mL)
Dihydroartemisinin AUC8h (ng⋅h/mL)
Artemether AUClast (ng⋅h/mL)
Dihydroartemisinin AUClast (ng⋅h/mL)
Lumefantrine AUC264h (μg⋅h/mL)
In the etravirine part of the study there was one serious adverse event (grade 3 tendon rupture, judged to be unrelated to study medication) and two subjects permanently discontinued study medication because of adverse events: one with grade 2 enterocolitis during treatment with artemether/lumefantrine alone and one with grade 1 rash during combined treatment with etravirine and artemether/lumefantrine. No relevant differences in the incidence of adverse events were observed between the treatments of the study.
Panel 2: artemether/lumefantrine with or without darunavir/ritonavir
Table 2 and Figure 3 show the summary pharmacokinetic results from the darunavir/ritonavir part of the study. Results were consistent between the single-dose and multiple-dose analyses. In the multiple-dose analysis, co-administration of darunavir/ritonavir lowered the AUC8h of artemether by 16% and that of dihydroartemisinin by 18% and increased the AUC264h of lumefantrine 2.75-fold. The mean AUC12h of darunavir was 52526 ng/h/mL when dosed alone, vs. 51271 ng/h/mL when co-administered with artemether/lumefantrine, with no significant difference in exposure between the two treatments (LSM ratio 0.96; 90% CI 0.90–1.03). The mean AUC12h of ritonavir was 6218 ng/h/mL when dosed alone, vs. 5498 ng/h/mL when co-administered with artemether/lumefantrine, with no significant difference in exposure between the two treatments (LSM ratio 0.90; 90% CI 0.80–1.01).
Table 2. Mean (standard deviation) area under the plasma concentration–time curve (AUC) of artemether (ART), dihydroartemisinin (DHA) and lumefantrine (LUM) after multiple doses, with and without darunavir/ritonavir (DRV/r)
ART/LUM + DRV/r (test)
LSM ratio (90% CI)
CI, confidence interval; LSM, least square mean.
Artemether AUClast (ng⋅h/mL)
Dihydroartemisinin AUClast (ng⋅h/mL)
Lumefantrine AUC264h (μg⋅h/mL)
In the darunavir/ritonavir part of the study there were no serious adverse events; one subject permanently discontinued study medication because of a grade 4 elevation in liver enzymes during the darunavir/ritonavir alone treatment. No relevant differences in the incidence of adverse events were observed between the treatments of the study.
Given the predicted effects of darunavir/ritonavir on lumefantrine levels, the ECG results were analysed in detail. Median changes in ECG and vital signs parameters were generally minor and not considered to be clinically relevant. One subject had a QTcF > 450 ms during co-administration of darunavir/ritonavir and artemether/lumefantrine. No QTcF change from baseline greater than 60 ms was observed. Overall, one subject experienced grade 1 tachycardia during administration of darunavir/ritonavir alone. No other ECG abnormalities were reported as adverse events during the course of the study. Vital signs-related adverse events were observed in one subject (hypertension) during co-administration of darunavir/ritonavir and artemether/lumefantrine. No other vital signs-related adverse events were reported.
In this study of healthy volunteers, co-administration of etravirine (at steady state) reduced the exposure of artemether, lowering the Cmax and AUC8h by 4% and 52%, respectively, after a single dose of artemether/lumefantrine. Dihydroartemisinin Cmax was slightly increased (1.05-fold) and AUC8h slightly decreased (15%) after single-dose administration. After multiple (i.e. six)-dose administration of artemether/lumefantrine, etravirine (at steady state) decreased artemether Cmin, Cmax and AUClast by 18%, 28% and 38%, respectively, and decreased those of dihydroartemisinin by 17%, 16% and 15%, respectively.
Artemether is known to undergo auto-induction, and this was observed in this trial when assessing the pharmacokinetics of artemether and dihydroartemisinin in the absence of either darunavir/ritonavir or etravirine. After multiple-dose intake of artemether/lumefantrine alone or in the presence of etravirine, Cmin and Cmax of lumefantrine were comparable; however, the AUC264h of lumefantrine was 13% lower after the combined intake of artemether/lumefantrine and etravirine compared with multiple-dose intake of artemether/lumefantrine alone.
The effects of the NNRTI etravirine on lowering levels of artemether, dihydroartemisinin and lumefantrine in this study are consistent with CYP3A induction by etravirine. Studies of other NNRTIs also known to induce CYP3A show similar findings. In a study of HIV-infected individuals, co-administration of nevirapine lowered the AUCs of artemether and dihydroartemisinin by 72% and 45%, respectively, with no effects on lumefantrine [12, 16]. In two studies (one in healthy volunteers and the other in HIV-infected individuals), co-administration of efavirenz lowered the AUC of artemether by 51–77%, that of dihydroartemisinin by 46–75% and that of lumefantrine by 21–25% [13, 15].
The results from the darunavir/ritonavir part of this study are also consistent with similar studies of the PI lopinavir/ritonavir (400/100 mg bid), which have also shown rises in the AUC of lumefantrine [16-18].
The 600/100 mg bid dose of darunavir/ritonavir was used in this study as this dose is expected to maximize the interaction. Darunavir/ritonavir is also approved for use in certain patients with a dose of 800/100 mg once daily . Although this dose was not formally evaluated, it is expected that the magnitude of interaction will be similar if not less as a result of the lower daily dose of ritonavir.
In vitro studies suggest that HIV PIs may have activity against Plasmodium parasites , and could potentiate the efficacy of antimalarial drugs . There is also evidence for antimalarial activity of etravirine in vitro . However, it is not clear whether this efficacy is consistent across these antiretrovirals, and whether antimalarial efficacy would be shown at the drug levels found in vivo.
A recent randomized study compared treatment with nevirapine vs. lopinavir/ritonavir in HIV-infected children in sub-Saharan Africa. The patients in the nevirapine arm had a significantly higher risk of developing malaria . The more favourable drug interaction profile of the lopinavir/ritonavir arm and potential antimalarial activity of this PI may have contributed to this effect. This analysis needs to be repeated in other studies comparing NNRTIs with PIs in countries with a high incidence of malaria.
When artemether/lumefantrine is co-administered with darunavir/ritonavir, the rises in lumefantrine AUC are a potential safety concern, given the potential for QT prolongation. It is difficult to draw definitive conclusions on the safety of this combination from this drug–drug interaction study, with the small sample size and use of healthy volunteers. Pharmacovigilance is required in HIV-infected people taking this combination, to evaluate the safety implications. The World Health Organization guidelines suggest that lumefantrine is generally well tolerated for the treatment of malaria, and does not carry a significant risk of cardiac toxicity . However, this guideline is for overall malaria treatment, and does not concentrate on people coinfected with HIV, and taking antiretrovirals which increase lumefantrine concentrations.
In conclusion, based on the results of this drug–drug interaction study, etravirine or darunavir/ritonavir can be co-administered with artemether/lumefantrine without dose adjustment. However co-administration of darunavir/ritonavir and artemether/lumefantrine is not recommended with other drugs that may cause QT prolongation. No effects of artemether/lumefantrine on etravirine, darunavir or ritonavir pharmacokinetics were observed.
We thank the volunteers who participated in the study, the staff at MTZ Clinical Research in Warsaw, Poland who conducted the study, and the data management, statistical and clinical teams at Janssen and SGS BioSciences for processing, analysis and reporting of the results. We also thank Dr Andrew Hill for assistance with preparation of this study manuscript.
Conflicts of interest: All the authors (TK, RD, YD and PM) are employees of Janssen, which developed the drugs darunavir and etravirine.
Funding source: This study was sponsored by Janssen EMEA Medical Affairs, Beerse, Belgium.