Kamunkhwala Gausi (1, 6), Kamija S. Phiri (3), Feiko ter Kuile (4), Michael Boele van Hensbroek (4), Michael Essan (3), Dianne Terlouw (2, 4), Eva Maria Hodel (2, 4) and Lawrence Kazembe (1)
(1) Chancellor College, Zomba, Malawi; (2) Malawi-Liverpool-Wellcome Trust Clinical Research Programme, Blantyre, Malawi; (3) College of Medicine, Blantyre, Malawi; (4) Liverpool School of Tropical Medicine, Liverpool, United Kingdom (5) university of Cape town, South Africa.
Objectives: Oral artemether-lumefantrine (AL) is often administered shortly after the completion of parenteral quinine in the treatment of hospitalized children with severe malaria in sub-Saharan Africa. However, quinine is associated with QT-prolongation on the electrocardiogram and lumefantrine has similarities to halofantrine, an antimalarial known for QT-prolongation. Little is known whether this effect is increased by sequential administration of lumefantrine preceded by quinine. Cardiotoxicity of antimalarial medicines cannot be easily tracked in areas where antimalarial drugs are often used (e.g. Sub-Saharan Africa), as these are often resource poor settings where ECG monitoring might not be available and identification of high risk patients might therefore be challenging. The aim of the study was to develop methods to characterize the cardiac safety (using QT/QTc) of artemether-lumefantrine when preceded by quinine in treating malarial cases among children in Malawi and identify risk factors of QT-prolongation
Methods: Secondary data from a clinical trial on “intermittent preventive therapy post-discharge (IPTpd)” was analysed. The trial recruited 133 children with severe malaria anaemia aged 4–59 months. Electrocardiograph assessments were conducted 12 hours after the last quinine dose which was immediately prior to the first AL dose (0 hour), and again 6 hours and 62 hours later. The trial had two arms: those administered 6 quinine doses and those administered 5 quinine doses. A Fixed effect model(FEM) and a linear mixed-effects model (LMEM) were derived and adopted to analyse the effect on the QT-interval of AL preceded by quinine. The LMEM was based on simulated lumefantrine concentration over time profiles parameter from a similar study in Uganda. Model estimation was based on Stochastic Approximation of Expectation Maximization (SAEM) and covariate selection was based on stepwise forward and backward elimination, while Akaike Information Criteria (AIC) was used to compare the models.
Results: Administering quinine followed by LA lead to a linear regression-corrected QT (QTc)of > 500 ms in 5 children administered 6 quinine doses. The trend of the QTc across the three time points was that most patients started with an elevated QTc (mean of 464.8 ms) due to the effect of quinine, but the mean QTc decreased at 6 hours to 461.16 ms (95%-CI: 457.63 – 464.69 ms) then starts to rise at 62 hours 462.58 (95%-CI: 458.43 – 465.28 ms). The participants administered 6 quinine doses (n =50) had a pronounced increase in QTc between 6 to 62 hours compared to the 5 quinine doses (n = 71). The parameter estimates and corresponding 95% confidence interval from LMEM were 465 (462, 468) ms for the intercept and -0.35 (-0.748, -0.072) ms for slope indicating that even with an increase in concentration of LA, QTc decreases with time. The model identified 4 covariates that were affecting QT-interval of the patients: sex, number of quinine dose, haemoglobin and feeding condition. LMEM model offered the best fit to data with the lowest AIC = 2570.797.
Conclusions: The co-administration of LA did not maintain the high QTc values caused by quinine, but female patients receiving 6 quinine doses were at high risk of QT-prolongation. Similarly, anaemic undernourished patients are also at a relatively high risk of QT-prolongation.
References:
[1] Bonate,P. 2011. Springer.
[2] Djimde, A. 2011. Antimicrobial Agents and Chemotherapy, 55(9), 3994–3999.
[3] Phiri, K. 2012. The Lancet infectious diseases, 12(3), 191–200.
[4] World Health Organization (WHO). 2016. Tech. Rep. Geneva.
Reference: PAGE 27 (2018) Abstr 8656 [www.page-meeting.org/?abstract=8656]
Poster: Drug/Disease Modelling - Safety