The interaction between artemether-lumefantrine and lopinavir/ritonavir-based antiretroviral therapy in HIV-1 infected patients

Subjects and study design

We conducted a sequential, two-period, adaptive design, open-label, pharmacokinetic
and safety drug-drug interaction study at the Groote Schuur Hospital Clinical Pharmacology
Research Ward in Cape Town, South Africa.

HIV-infected adults (18 years of age or older) with CD4+ lymphocyte counts greater
than 200 cells/?L were enrolled. Participants enrolled were stable on treatment with
lopinavir-based ART for a minimum of six weeks. They were compared with a group of
patients who were antiretroviral (ARV)-naïve and not yet eligible for ART, according
to the South African National HIV Treatment Guidelines at the time 16], 17]. The participants were otherwise well adults without renal disease and were not geriatric,
underweight, overweight or obese 18].

Exclusion criteria for safety reasons were a current diagnosis of malaria, known hypersensitivity
to artemether or lumefantrine, pregnancy (as confirmed by a serum Beta-HCG test),
breast-feeding, or clinically relevant hepatic or renal dysfunction. In addition,
those with a pre-existing (or family history of) prolonged QT interval, cardiac dysrhythmia,
electrolyte disturbances or taking any drugs known to prolong the QT interval, were
excluded. Exclusion criteria for potential confounding of the pharmacokinetic parameters
included participants using other drugs known to interact via the CYP450 enzyme system,
current smokers, or alcohol users who would not abstain from alcohol intake for the
trial duration. Caffeine, grapefruit juice or strenuous exercises were not permitted
from 24 h before and during study admission.

Ethics, consent and permissions

Patients provided written informed consent prior to enrollment. Regulatory approval
was received from the University of Cape Town Research Ethics Committee and the South
African Medicines Control Council (Clinical Trial Registration number NCT00869700).
The procedures followed were in accordance with the Good Clinical Practice Guidelines,
including the Helsinki Declaration of 1975, as revised in 2008.

Dosing and pharmacokinetic sample collection

As there was a safety concern about increases in lumefantrine concentrations secondary
to inhibition by ritonavir and lopinavir, patients on lopinavir-based treatment were
admitted for a single dose of artemether-lumefantrine (80 mg/480 mg) in a dose-finding
safety phase. Pharmacokinetic and safety results were analysed and reviewed by the
Data Safety Monitoring Board prior to approval of the adapted dose used in the multiple-dosing
phase. The ARV-naïve participants took part in the multiple-dosing phase only, when
the recommended adult 80 mg/480 mg artemether-lumefantrine dose was given at 0, 8,
24, 36, 48 and 60 h 12] .

In both study groups, all doses were administered with 40 mL of soya milk (0.8 g fat)
and a meal containing a minimum of 6 g of fat within one hour of each dose, with the
exception of dose 2 (at 8 h) when only soya milk accompanied the dose.

Participants were admitted for rich pharmacokinetic sampling (until 72 h after the
first artemether-lumefantrine dose). Subsequent samples were collected on an outpatient
basis until day 21. Venous blood samples were collected into heparinised (LH PST II)
BD Vacutainer® tubes. The blood tubes were pre-chilled on ice for 10 min; all samples
were again chilled before being centrifuged at 4 °C for 10 min at 2000 g. The resulting
plasma was stored at ?80 °C within 30 min of the blood draw. Pharmacokinetic assays
were done within four months of sample collection.

For the Phase 1 (single-dose) pharmacokinetic profile: Plasma concentrations of lumefantrine
were assayed at pre-dose (0 h), 0.5,1, 1.5, 2, 3, 4, 5, 6, 8, 14, 24, 36, 48, 60,
72, 96, 120, 144, 168, 336 and 504 h post and artemether/dihydroartemisinin concentrations
were assayed at pre-dose (0 h), 0.5, 1, 1.5, 2, 3, 4, 5, 6, 8, and 24 h after the
first artemether-lumefantrine dose.

For the Phase 2 (full-treatment dose) pharmacokinetic profile: Plasma concentrations
of lumefantrine were assayed at pre-dose (0 h), 0.5,1, 1.5, 2, 3, 4, 5, 6, 8, 14,
24, 30, 36, 42, 48, 54, 60, 61.5, 62, 63, 64, 65, 66, 68, 70, 72, 96, 120, 144, 168,
336 and 504 h, and artemether/dihydroartemisinin concentrations were assayed at pre-dose
(0 h), 0.5, 1, 1.5, 2, 3, 4, 5, 6, 8, 24, 60, 61.5, 62, 63, 64, 65, 66, 68, 70 and
72 h post-dose.

Pharmacokinetic assays

Concentrations of lumefantrine, artemether, and dihydroartemisinin were determined
by the Division of Clinical Pharmacology Laboratory, University of Cape Town using
validated liquid chromatography tandem mass spectrometer (LC-MS/MS) assays as described
previously 16].

Safety data collection

A clinical evaluation and full-blood count, renal function tests, liver enzymes, lactate
and glucose blood tests were performed at screening and at the final safety visit
21 days after the first artemether-lumefantrine dose in both the single-dose and multiple-dose
phases of the study. CD4+ lymphocyte counts and HIV-1 viral loads and serum pregnancy
tests (in all women) as well as urine tests for drugs of abuse (amphetamines, benzodiazepines
and opiates) were performed at screening. Twelve-lead single electrocardiograms (ECGs)
were performed at screening, pre-dose and at the expected time of maximal lumefantrine
plasma concentration (68 h post-dose) 19]. An independent cardiologist assessed all ECGs and the QT interval was corrected
using the Fridericia formula 20]. Adverse events were solicited throughout the study, starting on completion of screening
and recording the onset, duration, severity, relationship to study drug and need for
treatment 21], 22]. These were classified using MedDRA preferred terms. Some participants in the lopinavir
group were also included in a methods sub-study evaluating more intensive methods
for eliciting adverse event data from participants including checklists, in-depth
interviews and focus group discussions 22].

Statistical methods

The sample size was calculated to demonstrate a 2-fold change in lumefantrine exposure
(day 7 concentration or AUC), i.e., such that the 90 % confidence intervals (CIs)
for geometric mean ratios lie outside the interval 0.5 to 2.0 with a power of 80 %
19]. Thirteen participants were required in each group and a total of 18 participants
were recruited for each arm to accommodate potential dropouts 16].

Data analysis and pharmacokinetic modelling (non-compartmental) were performed using
Stata 13 (StataCorp, College Station, Texas). Concentrations below the limits of quantification
were considered missing. Area under the concentration-time curve (AUC
_0-?
) was calculated using the trapezoidal rule. Elimination half-life was calculated
as ln(2) ? ?z, where ?z is the first order rate constant associated with the terminal
(log-linear) portion of the curve, estimated by linear regression of time vs. log
concentration, using the default of last three data points.

In order to predict a safe dose for administration in Phase 2, the lumefantrine concentration-time
data (0–8 h) from our single-dose safety phase (Phase 1) were compared with those
in 18 ARV-naïve subjects included in our prior antimalarial-antiretroviral drug interaction
study using geometric mean ratios 16]. In the latter study the subjects completed a full course of artemether-lumefantrine
using the same schedule as in the multiple-dose phase.

Determinants of lumefantrine day-7 concentrations, AUC
_(0-inf)
and C
_max
values were explored using linear regression of the log transformed values, with results
reported as geometric mean ratios (GMR). The Spearman rank correlation test was used
to test the correlation between lumefantrine day-7 concentrations and AUC
_(0-inf)
. Logistic regression was used to explore the determinants of day-7 lumefantrine concentrations
below the reported therapeutic concentration (280 ng/mL). Continuous and categorical
covariates were compared between groups at baseline using Kruskal-Wallis and Chi-squared
tests, respectively. Kruskall Wallis tests were also used for simple comparisons of
the day-7 lumefantrine concentrations, AUC and C
_max
values between groups. In order to account for the repeated measures per subject,
particularly given previously reported auto-induction effects with the artemisinins,
mixed-effect regression models were used to assess the possible impact of dose-occasion
on artemether and dihydroartemisinin exposure, where the responses were log-transformed
AUC and C
_max
values.

Secondary safety endpoints included frequency and severity of adverse events, changes
in haematological, serum biochemical and urinalysis parameters, and vital signs between
screening and follow-up. The risk of adverse drug reactions was compared between treatment
groups using logistic regression. ECG parameters (PR-, QRS-, RR- and QT-intervals)
were compared within groups between screening and the predicted lumefantrine T
_max
using the Wilcoxon signed rank test, while the Wilcoxon rank sum test was used to
compare these between groups and within period, and their correlation with lumefantrine
concentrations was assessed using the Spearman Rank correlation test.