ACT Africa Baseline Study Group

ACT Africa Baseline Study Group

A pooled analysis of early parasitological response to Artemisinin Combination Therapy (ACT) in Africa.

Update and overview

The Study Group pooled 83 studies (29,493 patients) conducted between 1999 and 2012; including 15 unpublished studies and 68 published studies. Out of these 83 studies, 34 were conducted in West Africa (n=10,676), 31 in East Africa (n=8,331), four in Central Africa (n=631), four studies in Southern Africa (n=666), and the remaining ten studies were multi-regional (n=9,211).

46% (n=13,664) of patients from the studies were treated with artemether-lumefantrine, 17% (n=4,907) with the fixed dose combination of AS-AQ (ASAQ-FDC), 13% (n=3,925) with the non-fixed dose combination of AS-AQ in loose formulation (ASAQ-Loose NFDC), 9% (n=2,505) with the non-fixed dose combination of AS-AQ in co-blister formulation (ASAQ-Co-blistered NFDC), and 15% (n=4,492) with dihydroartemisinin-piperaquine (DP).

The Study Group formed in October 2011, with an open invitation to join and share data. Research groups with relevant data sets were contacted between October and November 2011. Potential participants met at the December 2011 ASTMH Annual Meeting to discuss governance and publication policy. The study group closed in March 2013. The latest results were presented at the BMC Challenges in malaria research conference in September 2014 in Oxford, UK; and at the RSTMH Biannual Meeting September 2014 in Oxford, UK. The results: Clinical determinants of early parasitological response to ACTs in African patients with uncomplicated falciparum malaria: a literature review and meta-analysis of individual patient data were published in BMC Medicine in September 2015.

Current guidelines recommend in-depth assessment for the presence of malaria drug resistance in a specific geographical region if malaria parasites are still present in the blood of over 10% of patients three days after treatment. The final results of the study suggest that the recommended threshold should be decreased from 10% to 5% for African populations. Lowering the percentage could mean that monitoring mechanisms would be able to identify areas at high risk of resistance early on, so that appropriate intervention measures could be taken.


Slow clinical and parasitological response after artemisinin therapy for uncomplicated falciparum malaria has emerged in Western Cambodia and may have spread to other sites in the Mekong region. Further spread of these parasites poses a major global public health threat, with the greatest potential impact in sub-Saharan Africa where disease burdens are greatest and systems for resistance monitoring and containment are weakest. Presently, we do not fully understand the mechanisms underlying the delay in early parasite clearance. Temporal analysis in several Southeast Asian sites shows prolonged clearance, but this has not yet been reported in Africa.

It is critical to define the extent to which slow parasitological response to artemisinin therapy may have spread, so that control strategies can be devised accordingly. However, clinical studies are difficult to compare over time and between sites since both host (e.g. age and immunity) and parasite (e.g. baseline parasite density) or partner drug factors can significantly influence parasite clearance.  

  • Define the spectrum of early parasitological response following ACT in African clinical trials, and the cofactors affecting early and late treatment outcome
  • Investigate geographical and temporal trends in early parasitological outcome which, after controlling for confounding factors, may indicate declining artemisinin efficacy

To address these objectives, we are conducting a retrospective analysis of parasitological response in the first three days after ACT in Africa.

Studies in Asian sites are underway to define the longer parasite clearance phenotype following artemisinin therapy. These have generally adopted artesunate monotherapy with required frequent sampling of parasitaemia during the first 3 days of treatment. This protocol is more difficult to conduct than previous therapeutic efficacy studies and would require significant revision of standard protocols which may not be easily adaptable to outpatient care, particularly in children in Africa. As a first step, we intend to conduct a pooled analysis of individual patients’ early parasitological response after ACT treatment in Africa. 

Essential inclusion criteria for data sets
  • Clinical trials including artemisinin combination therapy or artemisinin monotherapy;
  • Individual patient data including:
    • Minimum of daily sampling of parasitaemia in the first 3 days or at least on days 2 and 3 following treatment
    • Details of treatment regimens including dosage based on weight or age 
    • Baseline characteristics of the patient, including age and sex; and,
  • Ethical clearance of study protocol.
Desirable criteria (not required for inclusion)
  • Time of administration of medicines and collection of blood smears
  • Gametocyte measurement at day 0 and during follow up
  • Haemoglobin (Hb) estimation at day 0 and follow up
Primary end point
  • Proportion of patients remaining parasitaemic on days 1, 2, and 3, and relationship to day 28 and/or day 42 treatment outcomes
Secondary end points
  • Presence and clearance of gametocytes on days 3, 7, 14 and 28
  • Hb and/or haematocrit changes from baseline to days 3, 7, 14 and 28
Data standardisation and analysis

After upload to the WWARN Data Repository, WWARN standardised the data sets according to the WWARN Clinical Data Management and Statistical Analysis Plan and pooled them into a single database of quality-assured individual patient data. Analyses (subject to Study Group approval) aimed to include:  

  • Proportion of patients remaining parasitaemic at days 1, 2 and 3
  • Host, parasite and/or treatment factors that influence the early parasitological response for the following WHO-recommended first-line treatment regimens in Africa: artemether-lumefantrine, dihydroartemisinin-piperaquine, SP-artesunate, and amodiaquine-artesunate
  • Variation and trends of early parasitological response across location, after controlling for known factors; and
  • Variation and trends of recurrent parasitaemia and recrudescence for each of the recommended ACTs at days 28 and/or 42, stratified by region and transmission intensity

Analyses aimed to identify baseline responses for tested parameters to recommended drug regimens in East, West, Central, Northern, Southern or Horn of Africa. This information will help investigators to plan for targeted prospective clinical trials with rich parasitological sampling. The latter will help to discern the spectrum of clearance times and will also inform the design and development of new study protocols.

Study Group governance

The Study Group comprised participating investigators who contributed relevant data sets to the pooled analysis. Data sets remain the property of the investigator. The Study Group collectively made decisions with respect to including additional studies, data analysis and plans for publication, in line with the WWARN Publication Policy. The Study Group identified one or two people to coordinate activities including data analysis and drafting of publications and reports for group review. It was proposed that Ambrose Talisuna led the Study Group ad interim. The management and coordination structure of this Study Group consisted of the group leaders: Ambrose Talisuna, Louis Penali and Umberto D’Alessandro (epidemiologists), and Kasia Stepniewska and Prabin Dahal (statisticians). In addition there was an advisory committee composed of: Carol Sibley, Philippe Guerin, Ric Price, Grant Dorsey, Hali dou Tinto, Folarin Onikepe and Sodiomon Sirima, and others to be added from other regions in Africa. The data management procedures were in line with WWARN procedures. 

Once the data was uploaded into the WWARN Data Repository, it was curated and standardised into one format using the WWARN Clinical Data Management and Statistical Analysis Plan and pooled into a single database of quality-assured individual patient data. The WWARN statistician(s) were responsible for statistical analyses. 

For further information, email Prabin Dahal