This trial in three separate malaria areas of Afghanistan confirmed that both chloroquine and dihydroartemisinin-piperaquine are well-tolerated and broadly efficacious treatments for vivax malaria. Initial parasite clearance was significantly faster after dihydroartemisinin-piperaquine, consistent with the pharmacodynamic properties of artemisinin-containing therapies observed in P. falciparum  and P. vivax [9, 24]. Nevertheless 97% of patients in the chloroquine arm were parasite free and afebrile by day 2, and there were no recurrences in either arm before day 28, suggesting that P. vivax remains sensitive to chloroquine in the study area, which included locations both north and south of the Hindu Kush divide of central Afghanistan. These data extend the geographic range of efficacy monitoring beyond Eastern Afghanistan where chloroquine has been shown previously to be efficacious in clearing P. vivax parasitaemia [9, 10]. These findings stand in marked contrast to data from Indonesia where P. vivax chloroquine-resistance manifests as frequent failure to clear parasites by day 4 [6, 29–31] and recurrence rates by day 28 approximate 50% [29–31] or higher .
Over the 56-day follow-up period dihydroartemisinin-piperaquine proved superior to chloroquine in terms of the prespecified survival analysis of recurrence rates. Approximately 9% of cases had recurrent P. vivax infections in the chloroquine arm, compared to 3% in the dihydroartemisinin-piperaquine arm, with all failures occurring on or after day 35. Recurrences of P. vivax may consist of a mixture of relapses from liver hypnozoites, recrudescences of the erythrocytic infection (due to inadequate drug levels or resistance), and reinfections acquired from additional innoculations. It is not possible with current methodologies to distinguish reliably between these possibilities [32, 33]. However, in studies undertaken in a neighbouring region of Pakistan, radical treatment with primaquine following chloroquine prevented the majority of recurrent infections over the course of 9-12 months, indicating that most recurrences following chloroquine monotherapy are likely to be relapses from liver hypnozoites [34–36]. In this study, primaquine was not given, and none of the treatments administered has activity against P. vivax hypnozoites. The rapid clearance of parasites in the chloroquine group, and the fact that failures were not seen before day 28, suggest that recrudescences associated with chloroquine-resistance did not contribute significantly to the number of recurrences. For these reasons, the majority of recurrences observed in this study are likely to have been relapses.
The pattern of recurrences is consistent with the known properties of the drugs under study. Chloroquine is effective in preventing early relapses (<28 days) of vivax malaria in chloroquine-sensitive vivax malaria [4, 9, 10, 36–39]. This is because chloroquine is eliminated very slowly (terminal half-life one to two months), and provides post-treatment prophylaxis against relapse for 3-4 weeks after treatment . Piperaquine also has a very long terminal elimination half- . In an area where chloroquine-resistant P. vivax is prevalent DHA-PPQ has been shown in several comparative trials to prevent recurrences of P. vivax over a six-week period more effectively than ACT containing partner drugs with shorter half-lives [5, 14, 15]. In these studies, the recurrence rate (typically measured over 42 days) can be correlated with the half-life of the partner drug; for example, recurrences are fewer in patients treated with dihydroartemisinin-piperaquine compared to artemether-lumefantrine [5, 14] and artesunate-amodiaquine , presumably because the lumefantrine and amodiaquine components have shorter terminal elimination half-lives (approximately 10 and 18 days respectively) than piperaquine.
In the only previous trial directly comparing dihydroartemisinin-piperaquine to chloroquine (combined with sulfadoxine-pyrimethamine, CQ-SP) for P. vivax infection in Papua New Guinea, high levels of chloroquine-resistance were present, as evidenced by prolonged parasite clearance times in the CQ-SP arm. There was an 87% rate of recurrence in the first 42 days after CQ-SP compared to 30.6% in the dihydroartemisinin-piperaquine arm , a difference that is likely to reflect both failure to clear initial parasitaemia (leading to recrudescence) as well as failure to suppress relapse. This study provides the first comparison of chloroquine and dihydroartemisinin-piperaquine in chloroquine-sensitive P. vivax, and confirms the longer lasting post-treatment prophylactic effect of dihydroartemisinin-piperaquine compared to chloroquine in this setting. The observed proportion of chloroquine-treated patients with recurrence over 56 days is less than in a study undertaken in the Jalalabad Malaria Reference Centre in 2004  (where more than 40% of patients treated with chloroquine had recurrence from days 28-42 after treatment), for reasons that are unclear. Ancillary studies would help to shed further light on these issues, including pharmacological assessment of chloroquine and piperaquine levels in blood during the follow-up period. This would also allow the question of chloroquine resistance to be excluded more definitively.
This study was not powered to undertake detailed assessments of factors associated with recurrence. In the univariate analysis (although not the multivariate), younger age was found to be a risk factor for recurrence, a finding consistent with previous studies examining this effect over longer follow-up periods [35, 37]. Possible explanations for this include pharmacokinetic effects [39, 41], reduced immunity compared to adults and higher burdens of hypnozoites. Patients recruited in Maimana had a significantly higher rate of failure than those from Jalalabad in the univariate analysis although the effect disappeared in the multivariate analysis, suggesting that other factors may have contributed to this finding. The data do not provide substantive support for the existence of chloroquine resistance at Maimana or any of the study sites. The finding of a strong association between lower baseline haemoglobin and recurrence of P. vivax infection was not anticipated and requires further investigation.
It is possible that the action of dihydroartemisinin-piperaquine is simply to delay relapses of vivax malaria rather than prevent them; in other words the survival curves for recurrence with dihydroartemisinin-piperaquine and chloroquine will eventually merge. Studies with longer-term follow-up would be needed to determine if this is the case and if dihydroartemisinin-piperaquine offers tangible benefits in terms of long-term health of the individual (preventing relapse and associated anaemia) and possibly the community (by reducing transmission). Any health benefits have to be balanced against the relative cost of each treatment course, which is currently in favour of chloroquine use. These issues also have to be considered in the context of the challenges of treating patients with radical therapy in this region [35, 36].