Does acute malnutrition in young children increase the risk of treatment failure following artemisinin-based combination therapy? A WWARN individual patient data meta-analysis

Background: The geographical, demographic, and socioeconomic distributions of malaria and malnutrition largely overlap. It remains unknown whether malnutrition affects the efficacy of WHO-recommended artemisinin-based combination therapies (ACTs). A previous systematic review was inconclusive as data were sparse and heterogeneous, indicating that other methodological approaches, such as individual patient data meta-analysis, should be considered. The objective of this study was to conduct such a meta-analysis to assess the effect of malnutrition (wasting and stunting) on treatment outcomes in children younger than 5 years treated with an ACT for uncomplicated falciparum malaria.

Methods: We conducted a meta-analysis of individual patient data from studies identified through a systematic review of literature published between 1980 and 2018 in PubMed, Global Health, and Cochrane Libraries (PROSPERO CRD42017056934) and inspection of the WorldWide Antimalarial Resistance Network (WWARN) repository for ACT efficacy studies, including children younger than 5 years with uncomplicated falciparum malaria. The association of either acute (wasting) or chronic (stunting) malnutrition with day 42 PCR-adjusted risk of recrudescence (ie, return of the same infection) or reinfection after therapy was investigated using Cox regression, and with day 2 parasite positivity using logistic regression.

Findings: Data were included from all 36 studies targeted, 31 from Africa. Of 11 301 eligible children in 75 study sites, 11·5% were wasted (weight-for-height Z score [WHZ] <-2), and 31·8% were stunted (height-for-age Z score [HAZ] <-2). Decrease in WHZ was associated with increased risk of day 2 positivity (adjusted odds ratio 1·12, 95% CI 1·05-1·18 per unit; p=0·0002), treatment failure (adjusted hazard ratio [AHR] 1·14, 95% CI 1·02-1·26, p=0·016), and reinfection after therapy (AHR 1·09, 1·04-1·13, p=0·0003). Children with milder wasting (WHZ -2 to -1) also had a higher risk of recrudescence (AHR 1·85, 1·29-2·65, p=0·0008 vs WHZ ≥0). Stunting was not associated with reduced ACT efficacy.

Interpretation: Children younger than 5 years with acute malnutrition and presenting with uncomplicated falciparum malaria were at higher risk of delayed parasite clearance, ACT treatment failure, and reinfections. Stunting was more prevalent, but not associated with changes in ACT efficacy. Acute malnutrition is known to impact medicine absorption and metabolism. Further study to inform dose optimisation of ACTs in wasted children is urgently needed.

Funding: Bill & Melinda Gates Foundation.

Translation: For the French translation of the abstract see Supplementary Materials section.