The pathogenesis of post-malaria anaemia is multifactorial. Iron supplementation remains the mainstay of management of moderate and severe anaemia; however the management of mild anaemia (Hb 80-110g/l) is problematic as population supplementation studies of children in malaria endemic areas demonstrate adverse effects in children with mild anaemia. We hypothesize that the anti-inflammatory, anti-malarial and anti-macrophageal iron loading effects of chloroquine could make it a useful drug in the management of mild post malaria anaemia. To test this hypothesis, we plan to randomize children (aged 12 months to 6 years) with post malaria anaemia (Hb 70-110g/l) to receive a standard anti-malarial treatment, co-artemether . All children with parasitologic cure after three days on treatment will be randomised to receive either weekly chloroquine or weekly placebo starting from day 10 till day 90. By comparing the curve of haemoglobin change between day 3 and day 30 in the placebo arms of the two groups, we will test the effect of chloroquine vs. ACT treatment on macrophageal iron loading and release in acute clinical malaria. By comparing the haemoglobin change between day 3 and day 90 between the weekly chloroquine arms and the weekly placebo arms we will test the longer-term anti-inflammatory and anti- malarial effects of weekly chloroquine prophylaxis. In addition to the primary endpoint, we plan to assess potential mechanisms of action by determining parasite clearance, peripheral cytokine production and iron flux

Pathogenesis of Malaria Anaemia

Although the pathogenesis of malaria anaemia is not completely understood (1), mechanisms that have been proposed include immune and non-immune-mediated hemolysis of parasitized and non-parasitized red blood cells, bone marrow dysfunction and iron delocalisation (2-4). Plasmodium falciparum infection is associated with changes in the cell membranes of infected and uninfected erythrocytes causing alterations in membrane permeability and susceptibility to early hemolysis. It is also associated with bone marrow dysfunction probably due to malaria-induced abnormalities in erythroid progenitors, CFU-e and BFU-e (5,6). Jason et a (l7) have shown that malaria exerts a pro-inflammatory immune response in children and is associated with release of cytokines which act at various levels of the erythropoietic pathway to cause anaemia. These cytokines include human tumour necrosis factor (TNF)(8), interleukin (IL)-12, IL-10, IL-6 and IL-4 (7,9), and interferon-γ (IFN-γ) (9). In addition to the absolute concentrations of the cytokines, the balance between opposing anti- and pro-inflammatory responses may determine the clinical characteristics of malaria, including the level of anaemia (7) e.g. high TNF/IL-10 ratio may contribute to the reversible bone marrow suppression seen in malaria patients (10,11). Several investigators have reported the persistence of a fall in haemoglobin after successful treatment of malaria1 (2). Although the precise mechanism for this persistent anaemia is unclear, it is likely to be multifactorial. Studies conducted among Gambian children1 (3), French adults (14), and Thai adults (15) show that cytokine levels remain elevated for one to four weeks after a successful treatment for malaria; and Camacho et al (12) reported more than 80% of Thai adults with malaria had persisting anaemia on days 7, 14, and 21 after successful treatment, while 55% were still anaemic on day 28. On day 28, 46% of these subjects had hypoproliferative erythropoiesis while 7% had ineffective erythropoiesis (12). Although the study did not explore the mechanisms behind these findings, it is likely that persisting elevation of inflammatory cytokines reported by earlier workers (13-15) could be contributory. Greenberg and colleagues (16) have suggested that unusually strong and prolonged Th-1 response in conjunction with an inadequately developed Th-2 response may contribute to persistent anaemia after clearance of parasitaemia. Other workers have reported that persistent microscopically undetectable parasitaemia following successful treatment with an antimalarial was associated with protracted bone marrow suppression (17).

Macrophages and Malaria Anaemia Macrophages are responsible for the removal of senescent red blood cells from the body. They process these erythrocytes to release the heme iron which is then transported to the peripheral tissues including the bone marrow. The mononuclear phagocyte system plays two major roles in iron metabolism (18) - iron recycling from senescent erythrocytes and serving as a large storage depot for excess iron. This macrophageal recycling accounts for most of the 20-24mg of iron required daily for haemoglobin production. The cytokines released as part of the body's response to infection with Plasmodium cause the macrophages to release oxygen and nitrogen radicals leading to oxidative damage to parasitized and non-parasitized red blood cells, enhancing their removal from the circulation and contributing to anaemia (19). In addition, malaria-induced inflammation impedes release of iron from the macrophage-monocyte system hampering the supply of iron to the erythropoietic tissues. This causes considerable delocalisation of iron within the macrophage/monocyte system which makes iron unavailable to the marrow cells for erythropoiesis (18). The clinical practice of administration of iron for malaria anaemia is a result of the observation that serum iron is often low in such patients, as demonstrated in several animal and human studies (20,21). That the hypoferraemia observed in malaria could be largely due to iron sequestration in the macrophages is supported by reports of decreased peripheral iron levels despite normal or increased bone marrow iron (22,23).

Iron supplementation versus macrophageal iron mobilization The recent controversy regarding routine iron supplementation of children in malaria endemic areas with high prevalence of anaemia has further highlighted the need for alternative therapeutic regimens in children with malaria-associated anaemia. The genesis of the 'iron supplementation controversy' arose from several reports of increased morbidity and mortality from malaria and other infections in children from malaria endemic regions supplemented with iron. Recent evidence (reviewed by Prentice et al, Position paper for WHO Expert Consultation, Lyon, June 2006) suggests that the risk of adverse outcomes is less in more anaemic children. However, a growing body of evidence points to the poor rationale for giving iron supplements to children with mild malaria anaemia. First, during acute malaria, there is reduced iron absorption (Doherty et al. submitted, AJCN, 2006) and at least initially, erythropoietic iron supply is likely to come from reticuloendothelial macrophages rather than iron supplements. Secondly, the hypoferreamia associated with malaria has been shown to be due to iron delocalisation rather than absolute iron lack, reviewed in 18. Finally, a number of studies have found little or no benefit of giving iron supplements to children with malaria anaemia compared to other alternative regimens (24-27). While the management of moderate to severe malaria anaemia (Hb <80g/L) is not in contention, the management of mild anaemia remains an enigma because iron supplementation in these children, apart from providing questionable benefit, might in fact be harmful. Results from a large clinical trial of routine iron supplementation of children in an area of both iron deficiency and malaria transmission in Tanzania showed that among 24,076 children recruited those who received iron and folic acid, with or without zinc, were significantly more likely to die or experience adverse events than children who did not receive iron and folic acid (28).

Presently, there are no clear guidelines on the management of children with mild malaria anaemia. Mild anaemia in malaria endemic areas is likely due to either or both malaria and iron deficiency but distinguishing the iron delocalization of malaria from iron deficiency is difficult. It is important to optimize a child's iron nutrition to promote optimal cognitive development (29); however iron supplementation of this group is potentially dangerous hence the urgent need for alternative management strategies for a very common clinical scenario in Africa. Such interventions should take into consideration the complex pathogenesis of malaria anaemia including the mechanisms of iron flux and macrophageal iron delocalization during Plasmodium falciparum infection. It is likely that reduction of malaria-induced macrophageal iron sequestration and inflammation will enhance erythropoietic recovery post-malaria. We hypothesize that the anti-inflammatory, anti-macrophageal iron loading and anti-malarial effects of chloroquine could make it a useful drug in the management of mild post malaria anaemia by reducing macrophageal iron sequestration and interrupting the malaria-induced inflammatory process.

Chloroquine as a macrophageal iron mobilization agent Although resistance has reduced its effectiveness in the prevention and treatment of malaria, the non-antimalarial pharmacological properties of chloroquine make it a potentially useful therapeutic agent for other conditions. Chloroquine has antipyretic and anti-inflammatory properties (30-32). Chloroquine exerts a steroid-sparing effect (33), and inhibits the replication of a number of viruses such as HSV-1 virus (34), HIV-1 and several AIDS related opportunistic microorganisms (35,36). By inhibiting phospholipase A2 and tumour necrosis factor, chloroquine acts as an immunomodulator (37,38); and also acts as a lysosome-stabilizing agent. Clinically, chloroquine is used as a second line anti-inflammatory drug in chronic conditions like rheumatoid arthritis (39).

Chloroquine and iron metabolism Although the role of chloroquine in iron metabolism is still poorly understood, it is likely that many of the effects of chloroquine result from the interference with intracellular free iron. Chloroquine, a weak base, accumulates in acid intracellular compartments increasing the intracellular pH. Legssyer and co-workers have shown that chloroquine significantly reduces incorporation of iron into the liver (20% reduction), spleen (20%) and alveolar (400%) macrophages of rats loaded in vivo with iron dextran40.

Chloroquine and post malaria anaemia Chloroquine (CQ) may likely have three effects at different time points during the erythropoietic response to malaria. Firstly it may block the acute incorporation of iron into reticuloendothelial macrophages during clinical malaria associated with haemolysis and iron delocalization. Secondly it may have an anti-inflammatory effect. Increased serum levels of TNF-α, IFN-γ and nitric oxide depress erythropoiesis via bone marrow depression, dyserythropoiesis and erythrophagocytosis. Continuing inflammation after a malarial event may contribute to the slow resolution of anaemia (13-15) and chloroquine's anti-inflammatory effect might be a useful adjunctive therapy to continue to utilize after its initial antimalarial effect. Lastly chloroquine will have a continuing direct anti-malarial effect to both clear microscopically undetectable persistent infection and prevent further episodes until haematological recovery is optimized.

The anti-anaemic effects of chloroquine have been reported by a number of clinical studies. Salihu and colleagues41 reported a significant anti-anaemia effect of chloroquine given weekly to pregnant women in Cameroon compared to women not given any prophylaxis, even after controlling for possible confounders. Other studies among pregnant women in Cameroon (42), Burkina Faso (42), Uganda (43) and Thailand (44) all showed significant benefit of weekly chloroquine on maternal haemoglobin levels compared to controls. Although these studies were carried out on pregnant women, it is likely that similar benefits will occur in children.

Aim of Study and Hypothesis to be tested In exploring the effect of chloroquine on post malaria anaemia, we hypothesize that post-malarial CQ improves erythropoietic recovery after standard malarial treatment. We further hypothesize that post-malarial CQ improves erythropoietic recovery after co-artemether treatment, by a mechanism other than its anti-malarial effect in controlling residual parasitaemia. To test these hypotheses, we plan to randomize children (aged 12 months to 6 years) with acute malaria to receive either standard anti-malarial treatment (chloroquine plus sulphadoxine/pyrimethamine) or artemisinine combination therapy. Three days after commencement of antimalarial treatment, the children in each of the two arms, whose parasites have been cleared, will be randomised to receive either weekly chloroquine or weekly placebo. By comparing the curve of haemoglobin change between day 3 and day 30 in the placebo arms of the two groups, we will test the effect of chloroquine vs. ACT treatment on macrophageal iron loading and release in acute clinical malaria. By comparing the haemoglobin change between day 3 and day 90 between the weekly chloroquine arms and the weekly placebo arms we will test the longer-term anti-inflammatory and anti- malarial effects of weekly chloroquine prophylaxis. In addition to the primary endpoint (haemoglobin change), we plan to assess potential mechanisms of action by determining parasite clearance by PCR detection, peripheral cytokine production (& markers of inflammation), and indicators of monocyte iron loading and iron flux.

**Additional information: During the course of the study, the Gambian Government changed the antimalarial drug policy making artemisinine the first line antimalarial and discontinuing the use of chloroquine. We were therefore forced to alter the protocol to remove the initial chloroquine treatment.

Inclusion Criteria:

All children aged 12 months to 6 years in the 13 study villages will be enrolled in the study and followed up for the duration of the study. The inclusion criteria for randomization will be:

1. Children aged 12 months to 6 years; and
2. History of fever in the preceding 48 hours or a measured temperature > 37.5oC plus asexual forms of P. falciparum in the peripheral blood film of 500/μl or above; and
3. Hb <110g/l and >69g/l (Our choice of the upper limit of moderate anaemia (70 - 79g/l) is to enable us assess the response to our interventions of severer forms of anaemia while at the same time reducing the risk of adverse events which might occur with lower levels of Hb).

Exclusion Criteria:

1. Refusal of parent or guardian to give consent to the child's participation in the study
2. Inability of the subjects to take oral medications
3. Presence of features of severe malaria as defined by WHO50, with the exception of anaemia and parasite density
4. Children who have urgent need for blood transfusion as indicated by the presence of tachypnoea, tachycardia & gallop rhythm, tender hepatomegaly
5. Children with known haemoglobinopathy
6. Children with a weight for height Z score below -3SD of WHO/NCHS standard
7. Enrolment in another research project