Corresponding Author Abdoulaye A. Djimde, Malaria Research Training Center, Department of Epidemiology of Parasitic Diseases, Faculty of Medicine, Pharmacy and Odonto–Stomatology, University of Bamako, Mali. E-mail: email@example.com
Objective To assess the efficacy of intermittent preventive treatment (IPT) against malaria in school-aged children.
Methods This was an open randomized controlled trial of seasonal IPT among school children (IPTsc) aged 6–13 years in Kollé, Mali. The study began in September 2007 and completed follow-up in May 2008. Students were randomized to one of three study arms: Sulfadoxine–pyrimethamine plus artesunate (SP/AS), amodiaquine plus artesunate (AQ/AS) or vitamin C. All students received two full treatment doses, given 2 months apart during the season of high transmission from September to December. Groups were compared with respect to incidence of clinical malaria, asymptomatic parasitemia and haemoglobin concentration.
Results A total of 296 students were randomized, and retention in the study was 99.3%. Clinical malaria incidence in the SP/AS and AQ/AS arms was reduced by 66.6% and 46.5%, respectively, vs. vitamin C (P < 0.001). There were fewer clinic visits for any cause among the children receiving SP/AS or AQ/AS (P = 0.024). The prevalence of asymptomatic parasitemia was fivefold higher in the vitamin C arm than either SP/AS or AQ/AS at each post-treatment evaluation (P < 0.001). At the end of the transmission period, children treated with IPT had lower rates of anaemia (SP/AS, 17.7%; AQ/AS, 16.0%; vitamin C, 29.6%; P = 0.039).
Conclusion IPT among school children reduced the rates of clinical malaria, all-cause acute clinic visits, asymptomatic parasitemia and anaemia among school-aged children.
Le traitement préventif intermittent utilisant la thérapie combinée à base d’artémisinine réduit la morbidité de la malaria chez les écoliers au Mali
Objectif: Evaluer l’efficacité du traitement préventif intermittent (TPI) contre la malaria chez les écoliers.
Méthodes: Essai randomisé contrôlé ouvert du TPI saisonnier chez les écoliers âgés de 6 à 13 ans à Kollé au Mali. L’étude a débuté en septembre 2007 et le suivi terminé en mai 2008. Les élèves ont été randomisés à l’un des trois bras d’étude: sulfadoxine-pyriméthamine plus artésunate (SP/AS), amodiaquine plus artésunate (AQ/AS) ou vitamine C. Tous les élèves ont reçu deux doses complètes de traitement, administrées avec un intervalle de deux mois pendant la saison de haute transmission allant de septembre à décembre. Les groupes ont été comparés selon l’incidence de malaria clinique, la parasitémie asymptomatique et la concentration d’hémoglobine.
Résultats: 296 élèves ont été randomisés et la rétention dans l’étude était de 99,3%. L’incidence de malaria clinique dans les bras SP/AS et AQ/AS a été réduite de 66,6% et 46,5% respectivement, par rapport au bras vitamine C (p < 0,001). Il y avait moins de visites cliniques pour une quelconque raison chez les enfants recevant SP/AS ou AQ/AS (p = 0,024). La prévalence de la parasitémie asymptomatique était 5 fois plus élevée dans le bras vitamine C que dans le bras SP/AS ou AQ/AS à chaque évaluation post-traitement (p < 0,001). À la fin de la période de transmission, les enfants traités selon le TPI avaient des taux plus faibles d’anémie (SP/AS 17,7%, AQ/AS 16,0%, vitamine C 29,6%, p = 0,039).
Conclusion: Le TPI chez les écoliers réduit le taux de malaria clinique, toutes les causes de visite clinique aiguë, la parasitémie asymptomatique et l’anémie chez les enfants d’âge scolaire.
El tratamiento preventivo intermitente utilizando terapia de combinación basada en la artemisina, reduce la morbilidad por malaria entre niños en edad escolar en Mali
Objetivo: Evaluar la eficacia del tratamiento preventivo intermitente (TPI) frente a la malaria en niños escolares.
Métodos. Ensayo abierto, aleatorizado y controlado de TPI estacional entre niños escolares (TPIesc) con edades entre los 6 y 13 años en Kollé, Mali. El estudio comenzó en Septiembre 2007 y se completó el seguimiento en Mayo 2008. Los estudiantes fueron aleatorizados a uno de los tres brazos del estudio: Sulfadoxina-pirimetamina más artesunato (SP/AS), amodiaquina más artesunato (AQ/AS), o vitamina C. Todos los estudiantes recibieron dos dosis completas de tratamiento, dadas con dos meses de separación durante la época de trasmisión alta entre Septiembre y Diciembre. Los grupos fueron comparados con respecto a incidencia de malaria clínica, parasitemia asintomática y concentración de hemoglobina.
Resultados. 296 estudiantes fueron aleatorizados, y la retención en el estudio fue del 99.3%. La incidencia de malaria clínica en los brazos de SP/AS y AQ/AS fue reducida en un 66.6% y 46.5%, respectivamente, versus vitamina C (p <0.001). Hubo menos visitas clínicas por cualquier razón entre los niños que recibieron SP/AS o AQ/AS (p =0.024). La prevalencia de parasitemia asintomática fue >5-veces más alta en el brazo de vitamina C que en el de SP/AS o AQ/AS en cada evaluación post-tratamiento (p < 0.001). Al final del periodo de transmisión, los niños tratados con PTI tenían una menor tasa de anemia (SP/AS 17.7%, AQ/AS 16.0%, vitamina C 29.6%; p =0.039).
Conclusiones. El TPIesc redujo las tasas de malaria clínica, visitas clínicas agudas por cualquier causa, la parasitemia asintomática, y la anemia entre niños escolares.
In Mali, malaria accounts for 36% of medical consultations in school-aged children during peak transmission season (Ministry of Health, Republic of Mali 2002), and is a major cause of school absenteeism (Trape et al. 1993). Asymptomatic malaria infection also contributes to the high rate of childhood anaemia (Kurtzhals et al. 1999), which has been associated with worse cognitive outcomes (Hare 2004; Engle et al. 2007). The enormous number of symptomatic malaria infections continues to result in widespread absenteeism at work and school, hindering development and worsening poverty (Castillo-Riquelme et al. 2008; Teklehaimanot & Mejia 2008). Resistance to antimalarials has undermined the efficacy of previously reliable drugs such as chloroquine and sulfadoxine–pyrimethamine (SP), resulting in widespread reliance on newer artemisinin-based combination therapies (ACTs) (Mutabingwa 2005; Ogbonna & Uneke 2008). Evaluation and implementation of prevention strategies, including those using new classes of antimalarial drugs, are urgently needed (Breman et al. 2004).
Intermittent preventive treatment (IPT) of malaria is one strategy aimed at protecting at-risk populations from both clinical malaria and the sequelae of asymptomatic parasitemia (Manzi et al. 2008). IPT consists of giving full therapeutic doses to populations at risk whether or not they are infected. Widely used in pregnant women in malaria endemic zones (Briand et al. 2007; ter Kuile et al. 2007), IPT has also been studied extensively for use in infants throughout Africa (Greenwood 2007). However, the utility of IPT in school-aged children (IPTsc), i.e. children aged six and over has yet to be thoroughly evaluated. While school-aged children are less likely to die from malaria than younger children, they remain at significant risk for acute illness from malaria, anaemia, worse school performance and school absenteeism (Hare 2004). The efficacy of IPT to reduce malaria morbidity was demonstrated in a study of children up to 10 years old in Kambila, Mali, in which there was a 40% reduction in the incidence of malaria in children receiving IPT with SP (Dicko et al. 2008). More recently, a randomized study of IPT in school-aged children using SP in combination with amodiaquine showed a reduction in anaemia with improvements of cognitive function tests (Clarke et al. 2008). We conducted a randomized trial of IPTsc using two ACTs to reduce malaria morbidity in older children living in a West African community.
Study setting and population
The trial was conducted from September 2007 to January 2008, in Kollé, Mali, a rural village of 3000 inhabitants situated 57 km to the south-west of Bamako (the capital city of Mali). Other than the clinic established for this study, there is no medical facility in Kollé. The study clinic was staffed 24 h/day, 7 days/week. Based on our own epidemiological surveillance data, malaria is hyperendemic during the short transmission season from September to December. The prevalence of Plasmodium falciparum in the dry season is between 40% and 50%; this jumps to 70–85% during the rainy season. Inclusion criteria were: matriculation in the village school, age 6–13 years, absence of severe acute illness, ability to attend follow-up visits, written and expressed informed consent/assent from student, parent and teacher, and no history of allergy to study medications. Those not meeting all inclusion criteria or with a history of chronic disease were excluded from the study.
Study design and randomization
At the start of the study period, we obtained approval from the village elders at a public meeting. Over a 5-day period, a village crier was used to recruit interested subjects. The study was explained to children accompanied by a parent or guardian in the local language. After informed consent, the students were assigned a computer generated random number which linked them to one of the three study arms, Arm A (SP/AS): SP 25/1.25 mg/kg in a single dose plus artesunate 4 mg/kg once per day for 3 days; Arm B (AQ/AS): amodiaquine 10 mg/kg/day plus artesunate 4 mg/kg/day, each given once per day for 3 days; Arm C: vitamin C 250 mg tablet given once per day for 3 days. Vitamin C was chosen as the control medication as it is not known to have an effect on parasite growth. Researchers were aware at the time of initial treatment dose the arm to which each student was assigned.
Our previous studies in the site indicated that the incidence of malaria infection in the control arm would be 75%. Seeking to detect a 33% reduction in malaria infection with a power of 95%, an alpha risk of 5% and allowing 10% for lost to follow-up, we needed 97 children in each arm for a total sample size of 291 children.
Follow-up study visits were performed monthly through January 2008, at which the history, physical examination and finger-prick blood collection were repeated, and students with signs and symptoms of malaria were treated accordingly. Students received a second course of study medication in November, 2 months after the initial course. We amended the original protocol to allow for an additional study follow-up visit in May 2008, after preliminary data analysis showed widening differences in haemoglobin and asymptomatic parasitemia in the medication arms vs. Vitamin C. In addition to the active monthly follow-up from September–January and in May, participants were instructed that they should come to the clinic for any illness.
All unscheduled visits by participants to the clinic during the study period from September to January were documented. Children presenting with signs and symptoms of malaria underwent thick and thin smears and haemoglobin testing. Clinical malaria was defined by parasitemia detected in the setting of physical examination findings of axillary temperature >37.5 °C and/or complaints of headache, vomiting or joint pain. Because our recent studies in this area showed that both AQ/AS and SP/AS had efficacy rates above 95% (A. A. Djimde, H. Maiga, A. H. Beavogui, I. Sagara, C.P.O. Sangare, M. Tekete, A. Dara, Z. Traoré, C. N'Dong, O. B. Traore, S. Dama, H. Niangaly, N. Diallo, O. K. Doumbo, unpublished data) children with clinical malaria were treated either with the same IPT medication used in their study arm, or if in the vitamin C arm, SP was used. Cases of severe malaria, defined by World Health Organization (WHO) criteria (WHO 2000), were treated with IV quinine. Clinicians provided supportive care and treatment for all other diagnoses according to the national guidelines.
Treatment groups were compared with respect to the incidence of clinical malaria, number of clinic visits for any acute illness, the prevalence of anaemia (haemoglobin <11.0 g/dl), mean haemoglobin concentration and the presence of parasitemia. The effect of IPT on clinical malaria incidence was also analysed with stratification by age groups of 2-year increments. Continuous variables of parasitemia and haemoglobin levels were analysed across groups using the one-way anova. Categorical data were evaluated with the Pearson’s chi-square analysis. The Kaplan–Meier survival analysis was calculated based on the time to first episode of clinical malaria; second episodes of malaria in the same patient were excluded from the analysis. Annual incidence was not calculated as passive case detection was terminated at the end of the transmission season in January 2008. Analyses were performed using SPSS software (11.0, SPSS, Inc.).
Of a total of 475 students aged 6–13 years in the village, 305 were recruited and 296 were enrolled. There was a total loss to follow-up of 2 (0.7%) students at the end of the initial trial period in January 2008; there were an additional 10 students lost to follow-up at the study visit added in May 2008 (Fig. 1). One student was excluded from school for being less than 6 years old, and the other refused follow-up at the second treatment dose; both students were from the SP/AS arm. The final analysis consisted of 294 students; 96 in the SP/AS group, 100 in the AQ/AS group and 98 in the control group (Fig. 1). Baseline characteristics including age, haemoglobin concentration and parasitemia were similar between the study arms (Table 1).
At the end of the initial follow-up in January, children treated with IPT were significantly less likely to be anaemic (SP/AS, 17.7%; AQ/AS, 16.0%; vitamin C, 29.6%; P = 0.039). There were no cases of severe anaemia (haemoglobin <5.0 g/dl). At the January follow-up, average haemoglobin was slightly higher in the treatment groups than in the control group [SP/AS, 12.3 mg/dl (95% CI: 12.0–12.6); AQ/AS, 12.2 mg/dl (95% CI: 11.9–12.5); vitamin C, 11.7 mg/dl (95% CI: 11.4–12.0); P = 0.05]. By May 2008, this difference had narrowed and was not statistically significant [SP, 12.0 mg/dl (95% CI: 11.8–12.3 mg/dl]; AQ/AS 12.0 mg/dl (95% CI: 11.8–12.2 mg/dl; vitamin C, 11.8 mg/dl (95% CI: 11.6–12.1 mg/dl)]. The vitamin C arm continued to have higher rates of anaemia, though these differences were not statistically significant.
The initial prevalence of parasitemia was <10% in all groups (Table 1 and Fig. 2). In October, one month after the first treatment, parasitemia was observed in a significantly larger proportion of children in the control group than in those who received IPT (SP/AS, 3.1%; AQ/AS, 16.0%; Vitamin C, 41.8%; P < 0.001). At the November evaluation, the proportion of children with parasitemia in the three groups converged (Fig. 2). The second treatment was given in November, after which the level of parasitemia dropped and remained low in the medication arms, while the vitamin C arm continued to have nearly one third of subjects with asymptomatic parasitemia slides positive. In January 2008, the parasitemia by treatment group was 6.3%, 8.0% and 32.7% for SP/AS, AQ/AS and vitamin C, respectively (P < 0.001). At the final study follow-up in May 2008, the rates remained relatively unchanged with SP/AS and AQ/AS groups having 6.6% and 6.2% parasitemia, vs. 34.4% in the vitamin C group (P < 0.001).
There were a total of 146 unscheduled clinic visits by 114 different study participants. Of these cases, 103 (70.5%) had malaria confirmed by thick smear. There were no cases of severe malaria as defined by the WHO guideline (World Health Organization (WHO) 2000). The control arm had the highest number of episodes of clinical malaria with 55 confirmed cases, compared to 18 in the SP/AS arm and 30 in the AQ/AS arm (P <0.001). In the control arm, seven children had two episodes of clinical malaria and one had three episodes. One child in each of the SP/AS and AQ/AS arms had a second episode of symptomatic malaria (Table 2). In total, 46% (95% CI: 35–55%) of children had one or more episode of malaria in the vitamin C group compared to 18.7% (95% CI: 9.5–25%) and 29% (20–38%) in the SP/AS and AQ/AS arms, respectively (P < 0.001). The number of cases of clinical malaria fell with age in all three arms (Table 3).
Table 2. Intermittent preventive treatment reduces the number of clinic visits for acute illness and cases of malaria for 5-month follow-up
*SP/AS, sulfadoxine–pyrimethamine 1-day + artesunate 3-day; AQ/AS, amodiaquine 3-day + artesunate 3-day; vit C, vitamin C control.
†P-values obtained using chi-square test.
All-cause clinic visits
Total malaria episodes
Children with 1 episode of malaria
Children with >1 episode of malaria
Table 3. Age stratified incidence of clinical malaria shows that intermittent preventive treatment reduces infection rates across age groups
*P < 0.05; **P < 0.005; P-values obtained using chi-square test comparing treatment arm to vitamin C.
Age 6 and 7 years
Age 8 and 9 years
Age 10 and 11 years
Age 12 and 13 years
The seasonal incidence of symptomatic malaria (infections acquired during the 140-day period from September through January) was calculated to be 488 cases/1000 child-years, 782 cases/1000 child-years and 1463 cases/1000 child-years for SP/AS, AQ/AS and vitamin C, respectively (Fig. 3). Based on these figures, SP/AS reduced malaria incidence by 66.6% and AQ/AS reduced incidence by 46.5% compared to vitamin C. The number needed to treat to prevent one episode of clinical malaria was estimated to be 3.5 (95% CI: 2–6) for SP/AS and 5.9 for AQ/AS (95% CI: 3–28).
The most common side effects reported were headache, abdominal pain and respiratory symptoms. Across all groups, headache was the most commonly reported side effect for both rounds of treatment. The rate of headache was 8.3% and 4.2% for SP/AS, 13% and 14% for AQ/AS and 10.2% and 7.1% for vitamin C for the first and second treatment courses, respectively. The second most common complaint was abdominal pain with rates of 7.3% and 1.0% for SP/AS, 2.0% and 4.0% for AQ/AS and 0.0% and 3.1% for vitamin C for the first and second courses, respectively. Fever was reported only in the vitamin C group. Other reported events of vomiting, diarrhoea and respiratory complaints of cough or wheeze occurred in fewer than 5% of students. No students elected to quit the study because of these events.
In this large randomized trial of school-aged Malian children, IPTsc with either SP/AS or AQ/AS was superior to vitamin C in reducing malaria incidence, all-cause clinic visits, anaemia and asymptomatic parasitemia. This study was the first to evaluate IPTsc using ACT, the current standard of care for treatment of clinical malaria infection. There was no apparent difference between the ACT arms with respect to parasitemia or anaemia. However, fewer episodes of clinical malaria were observed with SP/AS than with AQ/AS. These findings and those of Dicko et al. (2008) and Clarke et al. (2008) show the potential of IPTsc to improve school attendance and cognitive development by preventing acute illness and reducing anaemia.
Protection against anaemia and asymptomatic parasitemia appeared to be short-lived for both ACT regimens. Although each treatment arm was superior to vitamin C 1 month after the first dose, this difference was lost by 2 months, suggesting that protection wanes between 4 and 8 weeks post-treatment. This time frame is consistent with the findings of an IPT trial in Ghanaian infants, in which protection waned between weeks 5 and 6 post-treatment (Cairns et al. 2008). A shorter interval between IPT courses may therefore be optimal. The benefits of highly effective IPT schedules must be balanced; however, against the possible negative impact on the development of protective immunity (Ross et al. 2008). Indeed, striking a balance between minimizing malaria infection and attempting to retain the development of natural immunity should be a consideration of any IPT intervention. While there are little data on delayed immunity linked to IPT use, results from a Kenyan trial monitoring the use of insecticide-treated bed nets (ITBNs) noted more infants under ITBN entered their second year of life without previous exposure to P. falciparum than control infants (Snow et al. 1996). However, when the benefits of reduced morbidity and mortality due to malaria were weighed against prolonging exposure driven immunity, research has consistently supported the use of ITBNs (Snow & Marsh 2002). To date, there is no evidence that IPT results in a delay in the development of natural immunity. Future studies are required to address this possibility.
Several other questions remain unanswered regarding the use of IPT as a public health intervention in school children. Whether the use of IPT as a public health intervention will be cost-effective and logistically feasible in school-aged children requires more study. Linking IPT to school health programs, as suggested by Clarke et al. (2008), may be an efficient strategy for older children, as opposed to infants in whom IPT has been integrated into immunization visit schedules (Chandramohan et al. 2007). It is also unclear which drugs are best-suited for use as IPT. The majority of studies of IPT have used SP alone, as the drug allows for a single treatment dose while providing lasting protection of up to 6 weeks (Cairns et al. 2008; Clarke et al. 2008; Dicko et al. 2008). ACT regimens may be more effective for IPT than SP alone, but this has not been demonstrated. Resistance to SP is rapidly increasing throughout much of Africa (Laufer et al. 2007; Talisuna et al. 2007), which may reduce its effectiveness for use in IPT. Furthermore, there is some evidence that IPT with SP may encourage the development of resistance (Sokhna et al. 2008), although data are conflicting (Greenwood 2007). While selection of resistance to artemisinin has been reported (Wongsrichanalai & Meshnick 2008), it remains to be determined whether IPT using ACT will affect the development of drug resistance and the future efficacy of these critically important drugs.
Limitations of the study include its open design and the use of vitamin C rather than placebo, which were due to limited financial support. We do not know of any reports of Vitamin C influencing the Plasmodium lifecycle. Although Vitamin C has been noted to minimally improve clinical outcomes in malaria (Wintergerst et al. 2006), we do not feel this adversely influenced the observed results. We cannot rule out the possibility that knowledge of arm assignment may have influenced use of clinic visits, potentially affecting the assessment of clinical malaria episodes. However, it was our impression from interviews of participants and their families that children in all arms visited the clinic for any symptoms of illness. The observed differences in parasitemia and anaemia at scheduled visits are less subject to bias. Although protection from clinical malaria and reduced anaemia are expected to result in improved school attendance and performance (Fernando et al. 2003; Clarke et al. 2008), these were not directly measured in this study.
In summary, IPT in school-aged children using ACT resulted in reduced illness and anaemia. These drugs were well-tolerated and have the potential to reduce the enormous burden of school absenteeism and adverse cognitive outcomes associated with malaria infection. More studies are warranted to identify optimal regimens with respect to specific drugs and the timing of treatments in areas of differing malaria transmission rates.
The authors thank the population of Kolle, School teachers and all students of Kolle for their participation to this study. The authors also thank Mr Abdramane Traore, Mr Fassoro Camara, Mrs Oumou Sow, Mr Toumani Camara and Mr Drissa Camara for technical support.
Financial support was provided by University of Bamako, Mali for the clinical trial. Breanna Barger was supported by Fogarty International Center grant # TW007988 to Vanderbilt University. Dr Abdoulaye Djimde is an European and Developing Countries Clinical Trial Partnership Senior Fellow and a Howard Hughes Medical Institution International Scholar. The National Institutes of Health (1 KL2 RR 025015-01) supports Dr Soren Gantt.