Artemether-lumefantrine is one of the most widely used ACT for treatment of uncomplicated P. falciparum malaria in sub-Saharan Africa. Previous four-dose regime was associated with 15% treatment failure, whereas, the now recommended six-dose regime showed, in a recent pooled analysis, a 28-day PCR-corrected parasitological cure rate of 97% in malaria endemic areas .
The effect of anti-malarial drugs in patients treated in non-endemic areas is less studied, but can be highly informative since patients are non-immune and not subject to re-exposure. Treatment failures with artemether-lumefantrine have been reported in two travellers after visiting Sierra Leone and Congo . Since lumefantrine is highly lipophilic and bioavailability depends on concurrent food intake , these failures have, in the absence of pharmacological assessments, been speculated to be due to sub-optimal lumefantrine concentrations. Also different sets of parasite populations have been suggested to have caused the recurrent episodes.
Here, a late treatment failure of artemether-lumefantrine is described in a traveller despite adequate plasma concentrations of lumefantrine. Genotyping showed the same msp2, pfcrt and mdr1 patterns, confirming parasite recrudescence. Despite that the patient reported nausea (without vomiting) and not eating the first day, the lumefantrine concentration (498 ng/ml) on day 6 corresponded to what has been estimated to be sufficient to prevent recrudescence, i.e. day 7 concentration >280 ng/mL [8, 9]. Artemether and dihydroartemisinin were, as expected due to their short half-lives, below the level of detection on day 6. A recent publication points to desbutyl lumefantrine possibly being a better predictor for treatment outcome than lumefantrine , and the desbutyl lumefantrine concentration in the traveller (2.96 ng/ml) was in the same but lower range of that in children with successful outcome (mean 15.5 range 0.6-58.2 ng/ml). Little is known about the pharmacokinetics of desbutyl lumefantrine and the exact relevance of a therapeutic threshold for the metabolite needs further studies.
Reduced in vivo sensitivity to artemisinin has so far only been reported from Cambodia as reduced parasite clearance time . Reduced in vitro susceptibility to artemisinin has also been described in a traveller returning from South East Asia . Here, the prompt reduction of parasitaemia suggests good effect of artemether/dihydroartemisinin. The late treatment failure might, however, suggest reduced sensitivity to lumefantrine since the concentrations were well above therapeutic cut-off. This is further supported by the finding of pfcrt 76 K and pfmdr1 86 N alleles that have been associated with a five-fold increase of lumefantrine inhibitory concentration in vitro (IC50) and also found to be selected for after treatment with artemether-lumefantrine in vivo in Kenya  and Tanzania . Not finding multiple copies of pfmdr1, also associated with reduced sensitivity to lumefantrine, might argue against this. However, multiple copies of pfmdr1 have only rarely been described in Africa. In vitro susceptibility testing could have confirmed resistance however parasites were not available for culture.
The patient was administered several antibiotics and paracetamol at the time of artemether-lumefantrine treatment. Some reduced effect (higher recrudescence rates) have been seen when co-medicated with quinine, a drug with the same biotransformation pathway as lumefantrine (CYP 3A4) . Moreover, ciprofloxacin which has been shown to have anti-malarial effect  was administered for one week. Although the present co-medication is not expected to have affected concentrations, interactions with other drugs might need further investigations.
In the current, as well as previous treatment failures in travellers, symptoms reappeared on day 14–24 after starting artemether-lumefantrine, reflecting late clinical failure [6, 7]. Interestingly, the symptoms were more pronounced and the parasite densities higher at recrudescence . Although this is probably due to patients waiting longer before seeking care (not believing it to be malaria again), milder symptoms might have been expected considering potential “strain-specific” immunity.
Rapid reduction of parasite biomass and symptoms together with few side effects has forwarded artemether-lumefantrine as an attractive alternative to mefloquine and atovaquone-proguanil for treatment of uncomplicated P. falciparum malaria in travellers. Although several studies show that artemether-lumefantrine does not have full efficacy, the drug combination has meet WHO > 95% cure rates , and is recommended as six-dose regimen in all areas irrespective of levels of drug resistance or host immunity.
The treatment failure described here might have been due to reduced sensitivity to lumefantrine but suboptimal concentrations of desbutyl lumefantrine and a missed dose cannot be fully ruled out even if the lumefantrine concentrations were adequate. Drugs chosen for treatment of P. falciparum malaria, especially in non-immune highly vulnerable individuals such as children and travellers, should preferably be efficacious enough to allow for single missed doses or uneven drug absorption. In endemic areas, partial immunity is likely to contribute to the effect of antimalarial drugs and thus overestimate cure rates. In line with this, decreasing effect of artemether-lumefantrine was indeed seen in an area of decreasing malaria transmission and immunity .
The present case forwards the need to further monitor the effect of artemether-lumefantrine treatment of P. falciparum malaria. Patients should be well informed to seek prompt care in the event of fever the weeks after completing artemether-lumefantrine treatment.