Artemisinin derivatives, such as artesunate, artemether and dihydroartemisinin, are the most parasiticidal of all the antimalarial drugs against Plasmodium falciparum 1. Their pharmacokinetic (rapid absorption, short half life ~1 hour) and pharmacodynamic (high parasite reduction ratio) properties mitigate against the development of resistance. In vivo resistance to this drug class has not yet been described 2. However, several challenges remain, including the limited availability of GMP quality drugs, the emergence of fake drugs, and their optimal deployment in malaria-endemic countries. Artemisinin-based drugs are easily available in several malaria-endemic countries and are sold as mono- or combination therapies. They are manufactured in several countries, packaged differently and contain package inserts of varying quality and accuracy. WHO currently recommends malaria-endemic countries to use three days of an artemisinin-based combination treatment (ACT) e.g. artemether/lumefantrine (A/L), artesunate (4 mg/kg/day) plus an effective, longer acting partner drug, such as amodiaquine or mefloquine 3. If used alone, seven days of treatment of an artemisinin derivative is necessary because shorter courses result in unacceptably high rates of recrudescence. Most recently, WHO has called drug manufacturers to switch the production of artemisinin monotherapies towards combined therapies because of the risk of developing drug resistance. Artesunate monotherapy of ≤ 5 days duration in South-East Asian and African patients is associated with a failure rate of up to 50% and the four doses regimen of artemether/lumefantrine has resulted in failure rates of between 15 to 18% in African children, Thai adults and non immune travellers 4567891011. Recrudescent falciparum malaria is associated with a risk of developing severe malaria, especially in non-immune individuals, such as travellers or expatriate workers.

Probable recrudescent P. falciparum malaria was recently diagnosed in three non-immune Europeans at the Geneva University Hospital. These three patients were initially treated in sub-Saharan Africa with artemisinin-based drugs. This motivated a thorough examination of package inserts of artemisinin-based drugs available in Africa. The potential association between treatment failure and inadequate information on drug leaflets are discussed in this study.

Packages of artemisinin-containing therapies currently sold in Africa were obtained from returning patients, WHO Roll Back Malaria, and the Internet. The list of twenty-five specialities presented here is not exhaustive (Table 1), but can reasonably be considered as representative of the current African market.

The majority (10/17) of manufacturers of artemisinin monotherapies recommend five days of treatment instead of the WHO recommended seven days. Only three drug packages had paediatric dosing and none had specific recommendations for non-immune patients e.g. that 5 days of treatment is likely to be inadequate. Some packages contained less tablets than indicated in the package insert. The package inserts included in combination treatments were generally better, with doses both for children (7/8) and adults and with sufficient tablets to complete a therapeutic course. Of note was the use of the lower dose of mefloquine (15 mg/kg) rather than the higher dose (25 mg/kg) that is used in multidrug resistant areas of SE Asia.

Sub-optimal antimalarial therapy results in increased morbidity and mortality of individuals living in malaria endemic areas, as well as travellers and expatriate workers 1314. Whilst well-established risk factors include poor compliance, fake drugs, limited training of drug vendors, high cost and drug resistance 151617, little attention has focused on the number of packaged tablets and the quality of information in package inserts. Both are crucial elements in the successful treatment of malaria. Many patients in endemic countries buy antimalarial drugs without prior medical consultation from local markets, small shops, pharmacies and mobile street vendors. Most would probably assume that one drug package contains a therapeutic course. Travellers and field workers probably have the same assumptions as local residents. Clearly, drugs that are sold already packaged with insufficient tablets to make up a therapeutic course carry a risk of treatment failure, which is increased further, if not all of the packaged drugs are taken. Drugs at parasiticidal concentrations need to act on at least four asexual erythrocytic cycles to eliminate all asexual malaria parasites; thus, a minimum of six days of therapeutic drug concentrations is required 1. Published data document failure rates of 50% with three days of artesunate in Africans, 50% with five days of artesunate in Vietnamese patients, and 18% with the for four doses regimen of artemether/lumefantrine in European travellers 579.

This study shows that a substantial number of artemisinin-based drugs currently sold in Africa do not respect current WHO recommendations in term of treatment duration in case of monotherapy and many lack appropriate information regarding children dosage. Packages containing fewer pills than indicated in the manufacturers leaflet were also found. Concerning the use of low dose of mefloquine (15 mg/kg) in combination treatment, this may not be a problem in the short term in Africa where mefloquine as monotherapy or combined with artesunate has shown high efficacy in recent trials. Nevertheless, great care should be taken in monitoring cure rate, as this lower dose may potentially favour the development of mefloquine resistance 181920.

Travellers, NGO and other workers should be aware not only of the hazards of contracting malaria, but also how to obtain optimal treatment. This is especially relevant for those who may be located in remote rural areas. Drug manufacturers should also play their part in reducing the risks of treatment failure. It is suggested that all currently available artemisinin-based monotherapies should be packaged for adults and children with a complete therapeutic course of seven days of treatment, that package inserts should be improved e.g. warning that a full course of treatment must be taken even if patients feel well. In view of recent WHO recommendations, manufacturers should shift their production in favour of combined therapies rather than monotherapy to avoid the development of resistance for what is often called the 'last effective weapon against falciparum malaria'. Written information should avoid proposing different doses by trying to differentiate patients based on their presumed malaria immune status, as this may be confusing both for patients and prescribers. Similarly, packaging for combinations should be explicit, easy to use and contain the requisite number of tablets. Travellers should be aware that artemether/lumefantrine is known as Riamet^® in temperate countries and CoArtem^® in malaria-endemic countries. Adult Riamet^® contains 24 tablets (six dose regimen) but adult CoArtem^® contains only 16 tablets (4 dose regimen). Setting international standards on package inserts might be a useful step that could be championed by a relevant international body. The training of local drug sellers and community health workers in malaria endemic countries is another strategy that could be adopted.

YJ: collected and analysed data and wrote the manuscript

FC: clinically managed the patients and co-supervised the study and the writing of the manuscript

LL: co-supervised the study and revised the manuscript

WT: collected data and co-supervised the writing of the manuscript