Fixed-Dose Artesunate–Amodiaquine Combination vs Chloroquine for Treatment of Uncomplicated Blood Stage P. vivax Infection in the Brazilian Amazon: An Open-Label Randomized, Controlled Trial

Summary In the Brazilian Amazon, the artesunate–amodiaquine combination was more effective in preventing Plasmodium vivax recurrence. With a favorable safety profile, this antimalarial treatment proved to be a good first-line alternative. Chloroquine resistance is probably underestimated in the area.

The combination of artesunate-amodiaquine (ASAQ) against P. falciparum has been adopted in many countries and is one of the World Health Organization (WHO) prequalified antimalarials. The ASAQ fixed-dose combination (FDC) tested in this study (artesunate-amodiaquine Winthrop, or ASAQ Winthrop) was codeveloped by DNDi and Sanofi. This formulation was shown to result in better compliance and reduced risk of emergence of P. falciparum resistance when compared with loose-dose combinations and coblisters [21,22]. We conducted this clinical trial in order to evaluate the efficacy and safety of the ASAQ FDC compared with that of CQ against uncomplicated P. vivax infection.

Study Site
Our study was undertaken at Fundação de Medicina Tropical Dr. Heitor Vieira Dourado (FMT-HVD), Manaus, in the Brazilian Western Amazon region. Malaria transmission is restricted to rural areas, and P. vivax causes around 90% of malaria cases [23]. Antimalarials are only provided in health units after diagnostic confirmation of infection. National guidelines recommend CQ (total dose of 25 mg/kg over 3 days) and PQ (0.5 mg/kg/day over 7 days) for P. vivax infection treatment and artemether-lumefantrine or artesunate-mefloquine for P. falciparum [24]. Studies showed therapeutic failure rates as high as 10% with CQ monotherapy for P. vivax in 2005 [17].

Study Design
This was an open-label, randomized, noninferiority, controlled trial that compared the efficacy and safety of ASAQ (Winthrop; Sanofi, Morocco) and CQ (Farmanguinhos, Brazil) for treatment of uncomplicated P. vivax blood stage infection. The study was based on the WHO antimalarial drug efficacy protocols modified for P. vivax [25,26]. Patients were followed up for 42 days to assess drug efficacy and safety.
The ethics review board of FMT-HVD (0426/2011) and the National Brazilian Committee of Ethics (16532/2011) approved the study. All adult patients and legal guardians of children provided written informed consent. The protocol is available in the Supplementary Materials.

Patients
Individuals aged >6 months with a body weight >5 kg with slide-confirmed P. vivax monoinfection, parasite density between 250 and 100 000 parasites/µL, and axillary temperature ≥37.5°C or history of fever in the last 48 hours were considered eligible. Exclusion criteria included pregnant or breast-feeding women; plan of leaving the study area in the following 42 days; known hypersensitivity to 1 of the investigational products; blurred vision suggesting retinopathy; presence of at least 1 severe criterion of malaria; known severe concomitant or underlying disease; and women of childbearing potential unwilling to use an effective contraceptive method during the study.

Procedures, Randomization, and Concealment
Enrolled patients were randomly assigned to 1 of the treatment arms by a study nurse through a computer-generated list that linked each patient's study number to an opaque zone, thus, concealing the assigned treatment arm. Neither the study doctors nor the laboratory personnel responsible for reading the slides and the polymerase chain reaction (PCR) test had access to patients' allocation during the study.
ASAQ was administered as a fixed-dose combination with 3 oral formulations (25/67.5 mg, 50/135 mg, 100/270 mg) and CQ as 150-mg tablets. Dosing was based on established weight bands for each drug (Supplementary Table 1). Study nurses supervised all dosing at enrollment and 24 and 48 hours after. When necessary, tablets were dissolved in water and administered orally using a syringe to ensure proper dosing to children. Patients were followed on an outpatient basis. PQ was withheld until day 42 or at day of recurrence. Recurrences during follow-up were treated with fixed-dose artesunate-mefloquine combination (Farmanguinhos, Brazil). The dose was administered a second time if vomiting occurred within 30 minutes.
Visits for clinical and microscopic assessments were scheduled on days 1, 2, 3, 7, 14, 28, and 42 after enrollment. Laboratory blood assessments (total blood cell count, bilirubin, glycemia, aspartate aminotransferase, alanine aminotransferase [ALT], and creatinine) were performed at enrollment and repeated on day 7 and day 28. Electrocardiograms were performed at enrollment, day 2, and day 28 for all participants aged >10 years and evaluated by a cardiologist. Blood samples were collected on day 7 and day of recurrence for CQ and desetylchloroquine (DCQ) blood assays, measured using high-performance liquid chromatography as previously described [27]. The sum of CQ and DCQ blood levels was used to classify parasites as sensitive or resistance based on the 100-ng/mL cutoff level, as previously established [12,26,28]. Technical problems with sample storage and collection, consisting of insufficient volume of blood for the analyses, prevented measurement of day 7 blood levels of AQ.
Two experienced technicians examined microscopic blood smears. Parasite density was calculated using the white cell count from the nearest total blood cell count (usually from enrollment, day 7, or day 28) [29].

Molecular Analyses
Real-time PCR was performed on filter-paper samples from enrollment or recurrences to confirm P. vivax monoinfection according to described protocols [30]. For the genotyping procedures for comparing day 0 and recurrence samples, 3 highly polymorphic gene regions were chosen according to diversity and discriminatory potential in the region, namely, msp1F3, MS2, and MS8 [31][32][33][34]. As there is no standard recommendation for classifying P. vivax recurrences due to difficulties distinguishing reinfection, recrudescence, and relapse [35,36], we adapted the classification method recommended by WHO for P. falciparum [25]. For each recurrence, samples were classified as homologous if at least 1 allele for each loci investigated was detected in both paired samples and as heterologous if all alleles for a given marker were different. A recurrent homologous P. vivax episode could derive either from a recrudescent-resistant parasite or from a dormant liver stage. For this study, we chose to treat these events as treatment failures after correction. Further details on the molecular characterization are provided in the Supplementary Materials.

Statistical Analyses and Sample Calculation
We assumed a 90% efficacy rate for CQ within 28 days [17] and 95% efficacy for ASAQ [37]. The noninferiority margin was defined at 5%, considering alpha at 2.5% for the unilateral test and power of 90%, a sample size of 145 individuals per arm, increased to 190 after adding 15% loss to follow-up and 10% risk of mixed infections.
The primary endpoint was genotype-adjusted adequate cure and parasitological response (ACPR) at day 28 for the per-protocol (PP) population. Secondary endpoints, performed in both the PP and intention-to-treat (ITT) populations, included crude ACPR on day 28 and ACPR on day 14 and day 42, prevalence and incidence of positive asexual parasitemia and gametocyte carriage, hemoglobin changes from baseline, and incidence of adverse events (AEs).
Noninferiority was demonstrated if the lower bound of the 95% confidence interval (CI) of the risk difference between the 2 arms was greater than −0.05 in the PP population populations. The CIs were calculated using normal approximation, with continuity correction in the Wald limits applied in the case of empty cells [38]. If noninferiority was shown, a 2-sided Fisher exact test to assess superiority was performed in the ITT population. This analysis was performed despite not being described in the protocol-following recommendation [39] and due to the large difference observed. Both the PP and ITT are presented. Time to recurrence was assessed by survival curves using the Kaplan-Meier estimator and log-rank tests. The proportion of patients with negative blood slides for both asexual and sexual parasites and fever clearance was analyzed at days 1, 2, and 3. A general estimating equation (GEE) model [40] was used to compare person-week gametocyte densities during follow-up; this consisted of a univariate analysis with robust standard errors, with the number of gametocyte-positive slides per individual per week as the outcome.
Anemia was graded according to WHO references [41]. Neutropenia and elevation of ALT were considered AEs of special interest [42][43][44]. An independent data monitoring committee comprised of 3 renowned experts reviewed the study protocol and evaluated the results of efficacy and safety analyses. An additional analysis was performed only for the CQ arm to investigate a possible association between delayed parasite clearance (>72 hours) as the dependent variable and risk of recurrence, the outcome variable, using logistic regression.

RESULTS
Patients were enrolled between January 2012 and June 2013 ( Figure 1). Most patients were adult males and resided in the nonactive transmission Manaus urban area (Table 1).
Noninferiority was demonstrated (97.5% CI lower bound = 2.74) as genotype-adjusted ACPR rates at day 28 of 100% (165/165) for ASAQ and 93.6% (161/172) for CQ. Although superiority of ASAQ was not demonstrated on day 28 in the ITT population (P = .345), it was demonstrated in the PP population and on day 42 for both populations (Tables 2 and 3; Figure 2). The difference was larger between day 28 and day 42.
Parasitemia clearance was faster, and the proportion of patients clearing fever at day 1 was higher for ASAQ (Tables 2 and 3).   (Table 4). Severe AEs were only observed in the ASAQ arm, none with sequelae. DISCUSSION We observed high efficacy of the ASAQ FDC against CQ-resistant P. vivax with respect to ACPR and other efficacy outcomes. This is the first study to compare the ASAQ combination with chloroquine against P. vivax infection and the first trial to compare an ACT with CQ in Latin America [1].
A remarkable result of this study is the high efficacy of ASAQ at day 42 (for which all recurrences occurred after day 40). Interpretation of failures occurring after day 28 is complicated by the possibility of recurrences being relapses or reinfections [9,25]. Despite this caveat, we believe the observed late recurrences Genotyping detailed data available in Supplementary Table S1.
Abbreviation: ACPR, adequate cure and parasitological response. a Losses to follow-up and samples with genotype assessment..
indicate the existence of low-grade resistance to CQ, which would suppress but not completely eliminate circulating parasites [12]. This is based on 2 important assumptions: very low risk of reinfection, as most patients (90.5%) resided in urban areas with no active malaria transmission [23], and negligible contribution of relapses, which usually occur after day 35 in the region [32,45,46]. In the absence of PQ administration, the drug's elimination half-life is an important property to be considered against P. vivax due to the post-treatment prophylactic effect [9]. Its influence is illustrated by 2 trials that compared the ASAQ loose-dose combination to dihydroartemisinin-piperaquine, an ACT with a longer half-life. Lower efficacy against late recurrences of ASAQ [43] was not confirmed when PQ was coadministered [44]. This rationale is supported by the higher efficacy of CQ in preventing late recurrences compared with drugs with shorter half-lives (arthemeterlumefantrine) in areas of low prevalence of CQR [47]. Thus, our results provide strong evidence of a higher-than-expected CQR prevalence because, even though CQ and its metabolites' half-lives are longer, there was a large number of recurrences in the CQ arm.
The genotype-adjusted ACPR rates were measured in an attempt to gain a more rigorous case definition of CQR. Genotyping has been applied in P. vivax efficacy studies with interesting results [35,36]. However, unlike in P. falciparum, there are no well-defined criteria to differentiate recrudescence from reinfection and relapse [25]. PCR correction will not necessarily lead to more precise estimation of efficacy in areas of low endemicity, where homologous hypnozoites are more likely [34].
ASAQ was more effective in clearing patent gametocytemia. CQ may take up to 4 days for gametocyte clearance [48]. Studies have found an association between prolonged gametocytemia and asexual stage recurrence [49]. It is not clear if use of an ACT has an effect on reducing transmission.
Hb reduction at day 7 was higher in patients who received ASAQ. Possible explanations include higher inflammatory response and/or oxidative stress associated with treatment, suppression of the blood marrow by artesunate, and pitting [50,51]. The risk of severe anemia after ACT is considered low [52]; indeed, there was no severe anemia in our study.
The safety and tolerability of ASAQ FDC were consistent with what was found in previous studies that used this combination in the approved indication "treatment of uncomplicated P. falciparum malaria" [42]. Sinus bradycardia occurred without symptoms or clinical complications, and there was no difference regarding the corrected QT inverval (QTc).
A limitation of our study was the small number of children included. This population is at higher risk of treatment failure [53,54], for which age-stratified sampling would be important. There was a higher loss to follow-up in the ASAQ arm, which had an impact on the efficacy assessment in the ITT population. An examination of the notification database did not reveal a higher rate of failure among losses. As we were not able to assess the pharmacokinetic profile of amodiaquine, evaluation of the effect of a longer elimination half-life in this population on efficacy against late recurrences was not possible. By not coadministering PQ, we were able to estimate CQR, which at 11.5%, considering concurrent blood drug levels criterion, is very high. This is important, as patients receiving CQ are at higher risk of recurring episodes from resistant parasites that arise from either recrudescence or relapse. These findings also suggest that the usual methods for measuring CQ efficacy are prone to underestimate the true CQR rates. Most studies of P. vivax efficacy restrict follow-up to 28 days in order to minimize the contribution of relapses, considering CQ elimination half-life. We believe this can result in underestimation of resistance, which, in our study, could be as high as 30% based on the risk difference against an efficacious comparator. This difference highlights the need to develop better designs and strategies for measuring antimalarial efficacy against P. vivax.

CONCLUSIONS
This is the first trial to compare an ACT with CQ as blood schizontocidal against P. vivax in Latin America. We demonstrated that CQR is underestimated in the region and that a change in treatment policy should be considered. Recently Plasmodium vivax has received increased attention, requiring development of new strategies and tools [55]. The choice of an effective schizontocidal drug, therefore, cannot be underestimated. Although CQ is still the first-line therapy in most endemic countries, resistance is a growing problem, probably underestimated because of coadministration with PQ and misclassification of recrudescence as relapse. The decision to substitute CQ for an ACT as the drug of choice for treating P. vivax infections is complex and should be based on different factors, including the prevalence of CQ-resistant parasites; the current treatment adopted for P. falciparum; efficacy of the partner drug; interaction of the ACT with PQ; and the potential impact of reducing transmission [13]. In order to establish unified treatment recommendations for P. falciparum and P. vivax, it is necessary to provide a more effective option and reduce the risk of treating P. falciparum with CQ due to misdiagnosis [50]. We believe that CQ should be replaced by a more efficacious  alternative in this region and expect policymakers to consider these findings in order to define the best strategies to accelerate the path toward malaria eradication.

Supplementary Data
Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the author to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the author, so questions or comments should be addressed to the author.

Notes
Funding. This study was funded by Sanofi. The study sponsor participated in the development the protocol, interpreted the data, helped write the report, and performed monitoring.