Chemoprevention offers a promising strategy for prevention of malaria in African children. However, the optimal chemoprevention drug and dosing strategy is unclear in areas of year-round transmission and resistance to many antimalarial drugs. To compare three available regimens, we conducted an open-label randomized controlled trial of chemoprevention in Ugandan children. Methods and Findings: This study was conducted between June 28, 2010, and September 25, 2013. 400 infants were enrolled and 393 randomized at 6 mo of age to no chemoprevention, monthly sulfadoxine-pyrimethamine (SP), daily trimethoprim-sulfamethoxazole (TS), or monthly dihydroartemisinin-piperaquine (DP). Study drugs were administered at home without supervision. Piperaquine (PQ) levels were used as a measure of compliance in the DP arm. Participants were given insecticide-treated bednets, and caregivers were encouraged to bring their child to a study clinic whenever they were ill. Chemoprevention was stopped at 24 mo of age, and participants followed-up an additional year. Primary outcome was the incidence of malaria during the intervention period. During the intervention, the incidence of malaria in the no chemoprevention arm was 6.95 episodes per person-year at risk. Protective efficacy was 58% (95% CI, 45%-67%, p,0.001) for DP, 28% (95% CI, 7%-44%, p = 0.01) for TS, and 7% for SP (95% CI, 219% to 28%, p = 0.57). PQ levels were below the detection limit 52% of the time when malaria was diagnosed in the DP arm, suggesting non-adherence. There were no differences between the study arms in the incidence of serious adverse events during the intervention and the incidence of malaria during the 1-y period after the intervention was stopped. Conclusions: For preventing malaria in children living in an area of high transmission intensity, monthly DP was the most efficacious and safe, although adherence may pose a problem. Monthly SP and daily TS may not be appropriate in areas with high transmission intensity and frequent resistance to antifolates.
Ethical approval was obtained from the Uganda National Council for Science and Technology, the Makerere University School of Medicine Research and Ethics Committee, and the University of California, San Francisco, Committee on Human Research. We performed a randomized, controlled, open-label trial from June 28, 2010, to September 25, 2013, comparing the efficacy and safety of three regimens versus no therapy for the prevention of malaria in Tororo District, eastern Uganda, an area with intense year-round malaria transmission and an entomological inoculation rate estimated at 562 infectious bites per person-year in 2002 [7]. Convenience sampling was used to enroll a cohort of 400 infants of 4–5 mo of age from the Tororo District Hospital Maternal and Child Health Clinic between June 28, 2010, and February 24, 2011. Eligibility criteria included the following: (1) born to HIV-uninfected mothers, (2) residency within 30 km of the study clinic with no intention of moving outside the study area, (3) agreement to come to the study clinic for any illness and to avoid medications outside the study protocol, (4) provision of informed consent by parent/guardian, (5) no history of allergy or sensitivity to any study drugs, (6) absence of active medical problem requiring in-patient evaluation or chronic medical conditions requiring frequent attention, and (7) absence of clinically significant electrocardiogram (ECG) abnormalities, family history of long QT syndrome, and current use of drugs that prolong the QTc interval. Only one eligible child was enrolled per household, and at enrollment each household was given two long-lasting ITNs. The purposes and benefits of sleeping under the ITNs were explained to the primary caregivers, and ITN use was encouraged throughout the study. A randomization list using permuted variable-sized blocks of 4, 8, and 12 was computer generated by a member of the study not directly involved in patient care. Study participants were randomized to their assigned treatment group at 6 mo of age using premade, consecutively numbered, sealed envelopes. Treatment allocation was performed by nurses not involved with patient care. Study drugs were dosed as follows: TS (co-trimoxazole, Kampala Pharmaceutical Industries, Uganda), single dose once daily; SP (Kamsidar, Kampala Pharmaceutical Industries, Uganda), single dose each month; and DP (Duo-Cotexin, Beijing Holley-Cotec Pharmaceuticals, China), once daily for three consecutive days each month; each drug was provided for administration at home according to weight-based guidelines. All study drugs were in compliance with local Good Manufacturing Practices. At the time of treatment allocation, parents/guardians were given a 2-mo supply of drugs and a diary with dates for dosing and check-offs to indicate administration. Parents/guardians were instructed to re-administer drugs if children vomited within 30 min of administration and to bring children to the clinic if they vomited again. During each visit to the study clinic, parent/guardians were questioned about study drug use and resupplied to maintain a 2-mo supply. Participants received all of their medical care at a designated study clinic open every day. Parents/guardians were encouraged to bring their children to the clinic any time they were ill. Children who presented with a documented fever (tympanic temperature ≥38.0°C) or history of fever in the previous 24 h had blood obtained by finger prick for a thick blood smear. If the smear was positive, the patient was diagnosed with malaria, and a complete blood count and thin blood smear for parasite speciation were performed. Episodes of uncomplicated malaria were treated with artemether-lumefantrine (AL), the recommended first-line treatment in Uganda. AL was administered twice a day for 3 d, with the first daily dose given directly observed in the clinic and the second daily dose administered at home. Episodes of complicated malaria (severe malaria or danger signs) [8] or treatment failures occurring within 14 d of prior therapy were treated with quinine. Routine evaluations, including thick blood smears and assessment of use of ITNs and adherence to study drugs, were done monthly. Complete blood count and glucose and alanine aminotransferase levels were assessed every 4 mo. ECGs were performed monthly in every fifth study participant randomized to DP if they could be brought to the clinic on the day they took their third dose in a given month. Adverse events were assessed and graded according to severity (mild, moderate, severe, life-threatening) using standardized criteria at every clinic visit. A serious adverse event was defined as any adverse experience that resulted in death, life-threatening experience, participant hospitalization, persistent or significant disability or incapacity, or specific medical or surgical intervention to prevent serious outcome. Diagnosis of incident episodes of non-malarial illnesses, including diarrheal illnesses and respiratory tract infections, were based on a prespecified list of diagnostic criteria developed by the study team. Medications with antimalarial activity were avoided for the treatment of non-malarial illnesses when possible. Antihelmintics, iron sulfate, and vitamin A were prescribed following Integrated Management of Childhood Illnesses guidelines. Chemoprevention was stopped at 24 mo of age, and study participants were followed-up one additional year until they reached 36 mo of age. Study participants were prematurely withdrawn from the study for (1) movement out of the study area, (2) failure to be seen in the study clinic for >60 consecutive days, (3) withdrawal of informed consent, or (4) inability to comply with the study schedule and procedures. Thick and thin blood smears were stained with 2% Giemsa for 30 min. Parasite density was estimated by counting the number of asexual parasites per 200 white blood cells and assuming a white blood cell count of 8,000 per microliter. A smear was deemed negative if no parasites were seen in 100 high-powered fields. Microscopy quality control included rereading all blood smears and resolution of any discrepancies by a third microscopist. Piperaquine (PQ) drug levels were measured from capillary blood collected on filter paper on the day malaria was diagnosed among study participants randomized to monthly DP. Levels of drugs from the other treatment arms were not measured because of resource constraints. Briefly, dried blood spots were punched from Whatman cards and extracted with 100 µl of a 1∶1 mixture of 20% trichloroacetic acid and acetonitrile. Nevirapine-d5 was used as the internal standard because its retention time (1.17 min) was the closest to that of PQ (1.21 min) among several compounds screened. The extract (10 µl) was directly injected onto a liquid chromatography–tandem mass spectrometry system API5000. Separation was achieved on a PFP column (2.1×50 mm, 3 µm) eluted with aqueous 0.14% trifluoroacetic acid, 10 mM ammonium formate, and acetonitrile. The PQ assay demonstrated a lower limit of quantitation of 10 ng/ml with a calibration range of 10–100 ng/ml. Inter- and intra-day accuracy ranged from 97.6% to 106% and from 93.9% to 112%, respectively, and inter- and intra-day variation ranged from 7.4% to 12% and from 4.4% to 14%, respectively. This study was part of a protocol that included two randomized controlled trials to evaluate the protective efficacy and safety of antimalarial chemopreventive regimens in distinct populations of HIV-unexposed (HIV-uninfected children born to HIV-uninfected mothers) and HIV-exposed (HIV-uninfected children born to HIV-infected mothers) children. Here we only present the findings from the trial conducted in HIV-unexposed children. The other results will be presented elsewhere. The study presented here was designed to test the hypotheses that chemoprevention lowers the incidence of malaria compared to no chemoprevention, and that the optimal chemoprevention regimen is DP. The sample size was calculated to detect at least a 32% lower incidence of malaria in the DP arm compared to that in the TS arm. We assumed that the incidence of malaria would be 1.85 episodes per person-year with TS chemoprevention based on a prior cohort study in the same area [3], and thus we calculated that we would need to enroll 100 participants in each arm to detect our targeted protective efficacy with 80% power at 95% significance (two-sided), allowing for 10% loss to follow-up. Data were double-entered and verified in Microsoft Access, and statistical analyses were performed using Stata, version 12 (StataCorp). All analyses used a modified intention-to-treat approach, including all study participants randomized to therapy and including all follow-up time until the study participant reached a study end point or early study termination, regardless of whether the intervention was stopped due to an adverse event. Descriptive statistics included means and standard deviations for continuous variables and proportions for categorical variables. The primary outcome was the incidence of malaria, defined as the number of incident episodes per time at risk, during the period the intervention was given (6–24 mo of age). Treatments within 14 d of a prior episode were not considered incident events. Time at risk was from the day following the initiation of study drugs to the last day of observation, minus 14 d after each treatment for malaria. Secondary outcomes included the incidence of complicated malaria, all-cause hospitalizations, diarrheal illnesses, respiratory tract infections, and serious adverse events or adverse events of moderate or greater severity (grade 3–4); the prevalence of moderate–severe anemia (hemoglobin<8 gm/dl) measured at the time of each episode of malaria and at the time of routine testing done every 4 mo; and the prevalence of parasitemia and gametocytemia measured at the time of monthly routine blood smears in study participants who were asymptomatic. Response to antimalarial therapy among children diagnosed with symptomatic malaria was also a secondary outcome, and will be presented separately. To assess the impact of chemoprevention on the development of naturally acquired immunity, the incidence of malaria, complicated malaria, and hospitalizations was compared between treatment arms for children 24–36 mo of age after the intervention was stopped. Incidence outcomes were compared using a negative binomial regression model, and prevalence outcomes and measure of use of ITNs at the time of monthly assessments were compared using generalized estimating equations with adjustment for repeated measures in the same study participant. For all analyses, only the assigned study arm was included as a covariate, with the exception of a multivariable analysis performed for the primary outcome of malaria incidence, which also included other covariates that were thought to be potential independent risk factors for malaria. Measures of association were expressed as protective efficacy (PE = 1 minus the incident rate ratio or prevalence ratio) during the intervention and incident rate ratios after the intervention was stopped. p<0.05 was considered statistically significant.