Background
The biological features of vivax malaria provide major challenges for pre-elimination and elimination programmes in areas co-endemic for
Plasmodium falciparum and
Plasmodium vivax [
1]. These features include the early appearance of gametocytes, a high proportion of asymptomatic or chronic carriers [
2] and the parasite latent form hypnozoites that produce relapses. Relapses make a major contribution to the global
P. vivax burden [
3]. It was estimated that relapses constituted 76–90% and 79% of total infections in Papua New Guinea and Thailand, respectively [
4]. To eliminate this reservoir of latent infections, simple effective radical treatment of vivax using 8-aminoquinolines, such as primaquine or the recently approved analogue tafenoquine, is required.
In most countries with endemic
P. vivax, the preferred first-line radical treatment for
P. vivax remains chloroquine, combined with 7- to 14-day primaquine regimens, and this has barely changed since the 1950s [
5]. Chloroquine is no longer recommended for the case management of falciparum malaria due to the spread of parasite resistance [
6], resulting in the use of different treatment regimens for
P. vivax and
P. falciparum in areas where these species are co-endemic. This is programmatically complicated. The use of a single regimen to treat all species of malaria would simplify malaria treatment guidelines [
7]. Artemisinin-based combination therapy (ACT) is emerging as the best option in this context, particularly in settings where there are concerns about chloroquine-resistant
P. vivax [
8]. ACT is effective against the blood stage of
P. vivax [
9], but must be co-administered with primaquine to eliminate
P. vivax hypnozoites [
8,
10]. Short-course primaquine regimens are preferable as they have been proven to have an efficacy not inferior to the standard 14 days’ regimens [
11‐
14]. However, there is a relative lack of data on the safety, pharmacodynamics and pharmacokinetics of ACT when provided in combination with daily primaquine regimens for the radical cure of
P. vivax [
15]. Safety is a major concern when deploying new treatments, but there are also concerns that primaquine/ACT drug interactions may reduce the overall regimen efficacy by inhibiting CYP2D6 or reducing synergistic effect of the current regimen. There are uncertainties about the best partner ACT, as drugs with a longer half-life may prevent early relapses.
Characterization of parasites and patients’ drug metabolism is needed to ascertain the pharmacodynamics, pharmacokinetics and therapeutic success of these regimens. The main malaria clinical trials outcomes are parasitological clearance and recurrence rate [
16]. Vivax recurrence includes: (i) recrudescence of parasites that have been previously cleared and microscopically undetectable; (ii) re-infection from another mosquito bite; and, (iii) relapses, i.e., activation of hypnozoites, genetically related (homologous) or unrelated (heterologous). The primaquine metabolite that is active against human hypnozoite is unknown [
17], but the metabolism of primaquine to its active metabolite is dependent on the cytochrome P450 enzyme CYP2D6 [
17‐
19]. Low
CYP2D6 activity results in slow metabolism of primaquine to the active metabolite. CYP2D6 activity may be a proxy of primaquine’s active metabolite exposure and a risk factor for relapse among primaquine recipients [
20‐
22]. Similarly, the cytochrome P450 enzymes CYP2C8 was investigated as it is known to participate in the metabolism of chloroquine [
23].
A previous randomized clinical trial in Brazil of the treatment of uncomplicated vivax malaria compared the safety and efficacy of the fixed-dose ACT artemether–lumefantrine and artesunate–mefloquine against the standard treatment with chloroquine, all three in combination with short-course primaquine (0.5 mg/kg/day for 7–9 days) [
24]. This current study investigated the pharmacokinetics of the long-acting ACT component (schizonticidal drugs) with concomitant primaquine upon the safety and efficacy of these three treatment regimens and the influence of genetic variability of parasite and host, including the frequency of mutations in CYP2D6 gene over relapse rate.
Discussion
This study showed by genotyping polymorphic loci of
P. vivax that relapse due to related parasites was the most frequent cause of recurrence by day 63 in all three treatment arms of the previous trial [
32], where 13–16% of infections recurred by day 63. The current study also showed that the treatment failure rate in patients with reduced inferred CYP2D6 activity (26%) was higher than in the general population (≅ 11%) [
33]. Moreover, 8 out of 9 recurrences among patients with low CYP2D6 activity were related relapses compared to 18 out of 25 with inferred normal CYP2D6 activity. Low CYP2D6 activity and presumptive low exposure to the active metabolite of primaquine or disruption of the PQ-enzyme interactions might influence related relapse rates in Brazil, although further studies are need to elucidate this effect. Likewise, the impaired activity of CYP2C8 may result in a lower exposure to desethylchloroquine, the chloroquine active metabolite. The PK parameters and clinical outcomes of the single patient with impaired CYP2C8 in CQ arm did not differ from the overall population.
Other causes of low exposure to primaquine are also potential risk factors for relapses, such as low adherence or impaired bioavailability. Similarly, others factors which affect the success of anti-infective therapeutics may influence the responses to the radical treatment, such as differences in the biology of the parasite, the immune status of the patients and the density of latent hypnozoites [
34].
This was the first time that some of the ACT combinations, namely ASMQ, were evaluated for the cure of
P. vivax in clinical trial conditions, and thus in combination with daily primaquine (0.5 mg/kg/day for 7–9 days) regimens. The correlation between drug exposure to the long-acting components of the ACT and the risk of recurrence and AEs was also assessed. Potential interactions between primaquine and the ACT, and also the safety of these regimens, have not been extensively assessed [
15]. Drug interactions such as lumefantrine inhibiting CYP2D6 [
35] may reduce the overall regimen efficacy; on the other hand, drugs with a longer half-life may prevent early relapses. This study could not demonstrate a significant associations between the PK parameters of the long half-life drugs and the risk of recurrence, or the risk of relapse due to either homologous and heterologous relapses or time to recurrence (Additional files
3,
4,
5), although the numbers of relapses were small. Patients with longer chloroquine elimination half-life estimates were more likely to report pruritus. Transient, mild to moderate pruritus is a well-known adverse effect of chloroquine [
36] and a threat to treatment adherence. A smaller drop in haemoglobin by day 14 was associated with higher CQ exposure (AUC), which may reflect better therapeutic efficacy achieved with higher concentrations of CQ [
37].
This study has several limitations. The blood sampling schedule was designed to evaluate the blood levels of the long half-life drugs; the pharmacokinetic data allowed the prediction of the drug exposures up to 63 days post-treatment with up to 8 sample points available for each patient. It did not allow a proper modelling of primaquine levels and limited the calculation of the elimination half-life of lumefantrine. The absence of desethylchloroquine blood levels measurement is another study limitation. A trial designed with 6 months follow-up would have been able to evaluate relapses with more accuracy, as the median time to vivax recurrence in Brazil is 71 days [
13]. The characterization of parasites and drug metabolism were conducted only in 35 patients with treatment failure limiting the comparisons. Genotyping vivax parasites to infer relapse frequencies also presents limitations. Genotyping in vivax does not allow differentiation between new infections (reinfections) or activation of unrelated (heterologous) hypnozoites. In this study, only recurrences with homologous and similar parasites were considered relapses. However, relapses are often heterologous activation of hypnozoites [
31,
38]. The recrudescence of sub-microscopic parasite population [
39] is also a biologically plausible explanation for homologous parasites in two samples. Future use of more sensitive parasite detection strategies, such as ultrasensitive PCR of all consecutive samples of the patients who failed could elucidate these results.
Conclusion
The genotyping of polymorphic loci of P. vivax showed that relapse due to genetically related parasites was the most frequent cause of recurrence in all three treatment arms. The high proportion of CYP2D6 genetic polymorphisms among patients with recurrent infections suggests that impaired primaquine metabolism might influence the related relapse rates in Brazil among patients receiving primaquine for radical cure, further studies are needed to confirm this finding. The three ACT regimens were very effective, and there was no association between drug exposure levels of the long-acting components of the ACT and risk of relapse. The ACT was well tolerated overall. These results provided further reassurance about the safety of the combined use of ACT and short-course primaquine (0.5 mg/kg/day for 7–9 days) to treat uncomplicated malaria vivax in Brazil.
Acknowledgements
The authors thank all the patients who participated in this study, and CEPEM, FMT-HVD, VPPLR/Fiocruz and SEFAR/Fiocruz, teams. The authors thank the Program for Technological Development in Tools for Health-PDTIS-FIOCRUZ for the use of the DNA sequencing (RPT01E) and Real-Time PCR (RPT09D) facilities. CAB and TNS thank CNPq for the research scholarship support. ACRS thanks the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) for scholarship support (Finance Code 001). DFR thanks Fapemig for scholarship support. MVL is a CNPq fellow. The Brazilian National Council of Research by Science Without Border programme (233618/2014-7) granted funds to AD scholarship.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.