Intermittent preventive treatment against malaria: an update

Abstract

Intermittent preventive treatment (IPT) against malaria is a malaria control strategy aimed at reducing the burden of malaria in certain high-risk groups, namely pregnant women and children. Three strategies – IPT in pregnancy (IPTp), infants (IPTi) and children (IPTc) – are reviewed here focusing on the mechanism of action, choice of drugs available, controversies and future research. Drugs for IPT need to be co-formulated, long acting, safe and preferably administered as a single dose. There is no obvious replacement for sulfadoxine–pyrimethamine, the most commonly utilized drug combination. All strategies face similar problems of rising drug resistance, falling malaria transmission and a policy shift from controlling disease to malaria elimination and eradication. IPT is an accepted form of malaria control, but to date only IPTp has been adopted as policy.

Keywords::

• anemia
• children
• infants
• intermittent preventive therapy
• malaria
• pregnancy
• school children
• seasonal

Malaria is a major cause of morbidity and mortality worldwide. Globally, it is estimated that some 2.37 billion people live at risk of Plasmodium falciparum infection, the most virulent of the human malaria parasites [1]. P. falciparum malaria causes 250 million clinical episodes and 850,000 deaths every year, the overwhelming majority of which occur in sub-Saharan Africa [2–4]. Malaria transmission has fallen recently in some parts of Africa, leading to localized reductions in the malaria burden [5–10], but the incidence remains high in many other areas.

Global malaria control strategies have changed over time. In the 1950s the focus of the WHO was on eradication of malaria, although formal campaigns never included Africa. By the mid-1960s, eradication efforts were seen to have failed. The strategy then moved towards malaria control, prioritizing the reduction of disease burden in children and pregnant women. With historically high levels of funding in recent years, elimination and eradication of malaria have, again, become stated objectives of the WHO. These different goals require distinct operational approaches. Specifically, eliminating or eradicating transmission calls for reducing carriage of malaria in all individuals, whereas reducing disease burden necessitates a focus on treating clinical disease or preventing infections in high-risk groups. Intermittent preventive treatment (IPT) is one approach to reduce parasite carriage and disease burden amongst the most vulnerable individuals.

Intermittent preventive treatment with sulfadoxine–pyrimethamine (SP) was first investigated in pregnant women (IPT in pregnant women [IPTp]) as an alternative to weekly chloroquine (CQ) chemoprophylaxis, a preventive strategy challenged by poor compliance [11,12] and widespread CQ resistance [13]. SP was viewed as a safe alternative that could be directly administered during antenatal consultations. Relatively easy and cheap to implement, IPTp was quickly recommended by the WHO as policy for areas of high malaria transmission [14], although adoption of the policy has taken some years. The success of IPTp led to adaptation of this approach in infants and children. Regular chemoprophylaxis was known to reduce malaria and increase hemoglobin concentrations in children less than 5 years of age. However, children who received regular chemoprophylaxis experienced an increase in malaria attacks the year after prophylaxis was discontinued [15–17]. Although poorly understood, it appears that continuous chemoprophylaxis hinders the development of host immunity in young children, causing a ‘rebound effect’. Researchers hypothesized that SP provided to infants less often during scheduled vaccinations might clear asymptomatic infections and prevent anemia whilst interfering less with immune development. The first trial of IPT in infants (IPTi) showed an impressive protective effect of 59% against clinical malaria and 50% protection against severe anemia [18]. This prompted further investigation of IPTi as a control tool, and development of a research agenda to consider the possibility of extending IPT to protect children outside infancy.

The use of IPT in children has since been studied using a variety of delivery methods. In recent years IPT research in children has been formalized into two broad themes: IPT in young children (IPTc) and IPT in school children (IPTsc). The objectives of these two strategies are different. IPTc attempts to prevent malaria episodes and other acute consequences of infection in young children who lack acquired immunity. To date, IPTc studies have predominantly involved monthly administration of antimalarial drugs to children less than 5 years of age in areas with seasonal malaria transmission [19]. Monthly SP has also been shown to be effective for management of anemia when given with iron supplementation [20–23]. Meanwhile, the primary objective of IPTsc is to reduce the consequences of chronic infection in older children who harbor asymptomatic parasitemia. Reducing anemia was the main benefit, although clearance of long-term infections may also improve cognitive abilities and school performance [24].

Although only IPTp is recommended by the WHO, IPT as a principle has become an accepted form of malaria control. It is thought to be preferable to continuous chemoprophylaxis for several reasons. First, fewer doses of drugs and fewer health-service contacts are required for implementation. Second, because IPT is normally given under supervision, rather than by self-administration, high compliance and correct dosing is more likely than with long-term chemoprophylaxis. Fewer treatments and correct doses may also increase tolerability. Third, as drugs are given intermittently and only to high-risk groups, the selection pressure that drives resistance may be minimized. Finally, because blood-stage malaria is not prevented between each treatment, IPT may be less likely to interfere with the development of antimalarial immunity when used in infants and children.

This review examines the purpose, mechanism, drugs used, and controversies and new directions for each of the IPT strategies.

IPT in pregnancy

Purpose

Each year 50 million pregnancies are at risk of malaria infection, the majority of these pregnancies occurring in sub-Saharan Africa [25]. Malaria in pregnancy is associated with intrauterine growth retardation (IUGR) [26,27], preterm birth [28,29], low birthweight (LBW) [30–32] and maternal anemia [33–35]. The purpose of IPTp is to reduce the risk of LBW and maternal anemia by clearing asymptomatic peripheral and placental parasitemia whilst providing pregnant women intermittent protection against malaria infection between antenatal consultations. The WHO recommends administration of two to three courses of SP (three tablets total, 500 mg sulfadoxine and 25 mg pyrimethamine per tablet) after fetal quickening, with each course given no less than 1 month apart and all prior to the last month of pregnancy [36]. IPTp has been adopted as national policy in 37 countries worldwide, 33 of which are in the sub-Saharan region [2].

Mechanism

A characteristic unique to P. falciparum is that infected erythrocytes adhere to chondroitin sulphate A (CSA) molecules and sequester along the endothelial and syncytiotrophoblast cells of the human placenta [37]. For this reason, screening of peripheral blood may fail to detect infection of the placenta. Parasitemia triggers an inflammatory response, elevating levels of TNF-α and IL-10 [38], but the mechanisms by which malaria initiates parturition are poorly understood. Chronic infection is associated with IUGR, whilst acute infection is more commonly the cause of miscarriage or stillbirth [39]. Placental insufficiency may compromise the exchange of vital nutrients between mother and fetus [40,41]. However, placental malaria does not appear to interfere with the transport of vitamin A [42], iron [43], folacin or cobalamin [44]. It is possible that parasitic invasion of the trophoblast has an effect similar to that of preeclampsia, decreasing placental circulation by undermining normal remodeling of spiral arteries into dilated utero-placental vessels [45]. Although not observed in other studies [46,47], a recent trial found an increase in the risk of hypertension among primigravidae who also had chronic placental malaria. Hypertension could have been caused by the fetal response to placental inflammation [48].

In high-transmission settings, multigravidae exhibit anti-adhesion antibodies that target CSA-binding parasites [49] and are better able to control parasite densities compared with paucigravidae [37]. Placental and CSA-selected parasites selectively transcribe the var2csa gene [50–52]. Because var2csa-specific IgG is only found in women and is a positively associated parity [53], a pre-erythrocytic vaccine that induces protection and is nonstrain-specific would be an appropriate candidate for vaccine development [54].

Infection with HIV removes the parity-specific pattern of malaria risk in pregnant women [55] and coinfection with malaria increases the risk of maternal anemia and LBW by as much as 35% [56]. Monthly IPTp with SP, however, has been shown to achieve similar protection in pregnant women who are HIV-positive as two courses taken by pregnant HIV-negative women [56].

Resistance to SP

Despite the considerable loss of parasite sensitivity in recent years, IPTp with SP remains efficacious in areas where SP treatment failure rates at day 14 among children under 5 years of age are as high as 40% [57]. However, because SP is no longer the first-line treatment for uncomplicated malaria in the region, a similar analysis would ideally establish the prevalence threshold of molecular markers specific to SP resistance above which IPTp with SP should no longer be used. The rapid reduction in SP use has reduced drug pressure for selection of resistant mutants but it is unknown if and how quickly this might translate into renewed parasite sensitivity. If there is a high fitness cost for SP resistance mutations in the absence of SP exposure, then drug sensitivity could return quickly as witnessed in Malawi with CQ. High in vitro sensitivity to CQ emerged within just 5 years of suspending it as the first-line treatment [58].

Establishing a new proxy threshold for policymakers would be important since there may be some locales where SP resistance is extremely high and the use of IPTp may actually increase parasitemia among pregnant women [59]. A prospective birth cohort of 880 Tanzanian women in an area of high resistance to SP showed that 112 (12.7%) had placental malaria at delivery. Data on IPTp exposure was available for 104 patients, of which 17 (16.4%) with placental infection had received no IPTp, 77 (74.0%) were given IPTp early in pregnancy, and ten (9.6%) had been given IPTp recently. Mean parasitemia across treatment groups was 6.7% with the lowest rate, 1.9%, among women who had not been administered IPTp, in contrast to 6.4% in women with early IPTp, and 14.1% among recent recipients. The difference in placental parasitemia between women exposed and unexposed to IPTp was highly significant (adjusted difference: 4.9%; p = 0.003) [59]. Nearby SP resistance levels in children under 5 years of age were very high with day 28 failures in excess of 80% associated with near saturation of the dhfr/dhps quintuple mutation and the appearance of a third mutation in the parasite dhps gene at codon 581 [60]. Pyrimethamine may be culpable. Women in Ghana were given prophylactic pyrimethamine during pregnancy, only to have parasite loads increase [61]. These findings are consistent with a model where the most highly resistant parasites repopulate more rapidly than less fit parasite populations when placed under drug pressure [62].

Despite these observations, the benefits of IPTp with SP continue to outweigh the risks and there are no recommendations to stop using SP in settings of high SP resistance. High priority must be given to identifying safe and affordable alternatives to SP [54,57].

Drugs for IPTp

Currently the WHO policy only recommends SP for IPTp. Owing to rising concerns about continuing efficacy of this drug combination, alternatives are being sought. Several reviews have described potential alternatives to SP [58,63–66] with mefloquine and azithromycin (AZ)-based combinations being the leading candidates under investigation. Having never been used before in the sub-Saharan region on any scale, mefloquine is likely to have good protective efficacy. A recent randomized clinical trial in Benin (n = 1601) showed mefloquine to be more efficacious than SP in preventing placental malaria (prevalence: 1.7 vs 4.4% of women; p = 0.005) and clinical malaria (incidence: 26 cases per 10,000 person-months vs 68 cases per 10,000 person-months; p = 0.007) [67]. Mefloquine also has appeal because it can be administered like SP, as a single observed-dose during antenatal consultations, although the dose of 15 mg per kg bodyweight may not be well tolerated. The Benin study found that 78% of subjects assigned to the mefloquine group experienced vomiting, dizziness, tiredness and/or nausea whilst one subject had severe neuropsychiatric symptoms. The authors suggest a split-dose regimen of 10 mg/kg, followed by 5 mg/kg given over 2 days, or 6–8 h apart may be better tolerated [67]. A Cochrane review of drugs for preventing malaria in travelers noted 22 published case reports of deaths, including five suicides, associated with mefloquine use at normal dosages. No other drugs reviewed had resulted in death when used as prescribed [68].

Azithromycin-based combinations may also be used in IPTp, offering antimalarial and antibacterial protection. Doses of AZ between 500 and 2000 mg have been used in all trimesters of human pregnancy for the treatment of upper and lower respiratory tract infections, skin diseases, Chlamydia trachomatis, mycoplasma and group B streptococci infections. Meta-analysis among pregnant women with C. trachomatis infection found AZ to be associated with significantly fewer gastrointestinal adverse events than erythromycin (odds ratio [OR]: 0.11; 95% CI: 0.07–0.18) and significantly fewer total adverse events (OR: 0.11; 95% CI: 0.07–0.18) [69]. However, AZ in a placebo-controlled trial produced three times more gastrointestinal effects among HIV-positive patients, 67 of 85 (78.9%), compared with 24 of 89 (27.5%) subjects who received placebo [70]. Case reports suggest that HIV-positive patients may also experience temporary ototoxicity with AZ use [71]. No evidence of teratogenicity has been observed in animal models with quantities equivalent to four-times a human treatment dose [72–74].

The potency of AZ, a translation inhibitor, is greatest against the progeny of malaria parasites that inherit a nonfunctioning apicoplast following drug exposure. This manifests as ‘delayed death’ among malaria parasites. AZ requires a 3-day course of daily doses of 1 g with a fast-acting partner drug that is pharmacologically compatible [58]. A smaller dose, shorter course or an inappropriate partner drug can be expected to produce suboptimal results. AZ plus CQ has an additive-to-synergistic effect in vitro against P. falciparum whilst AZ–mefloquine and AZ–pyronaridine combinations are both additive [75]. By contrast, AZ and artesunate (AS) were found to have antagonistic effects when used against fresh P. falciparum samples [75], a confirmation of earlier observations in which clones and culture-adapted malaria parasites were used [76]. In vivo antagonism between AZ and AS may explain the poor performance of the AZ–AS combination in a recent pediatric trial that suspended recruitment early [77].

In vivo synergy between AZ and CQ has been demonstrated in India against P. falciparum infection. After receiving 1 g of AZ for 3 days, 19% of subjects (three of 16) had eradicated parasites by day 28. Of subjects who received a daily dose of 600 mg CQ for 3 days, 27% (four of 15) were parasite-free at day 28. When coadministered, however, AZ and CQ produced 97% parasitological cure through to day 28 [78].

Azithromycin could be combined with piperaquine (PQ) to improve adherence; PQ is at least as effective as, and better tolerated than, CQ [79]. AZ plus SP is another option for IPTp in areas where SP resistance is not high; this combination would avoid the side effect of pruritus found with CQ.

Controversies & new directions

Changing transmission

Several malaria control tools have been scaled-up in recent years, altering local transmission rates. Indoor residual spraying (IRS) has provided protection for 9% of the general population at risk of infection, whilst 31% of households owned at least one insecticide-treated net (ITN) in 2008, compared with 17% in 2006, with ownership exceeding 50% in 13 countries [2]. An estimated 20% of pregnant women received two courses of IPTp with SP between 2007 and 2008, ranging from 3% in Angola to 66% in Zambia [2].

Rwanda reports that malaria transmission has reduced to such a point that parasitemia may no longer affect birthweight [80]. In the Benin study previously described, the superiority of mefloquine over SP in preventing placental and clinical infection did not translate into a statistically significant reduction in the rate of LBW: 8% of women who received mefloquine (n = 735) gave birth to LBW newborns, whilst 9.8% of those administered SP had LBW infants (difference: -1.8%; 95% CI: -4.8–1.1) [67]. At enrollment, 53% of women in both treatment groups said they had slept under an ITN the previous night. Thus, if ITN use continued throughout the trial, SP and mefloquine may have encountered fewer placental parasites upon which to act, diluting the protective effect that either treatment could have exerted on birthweight. Similar interaction between ITNs and IPTp may explain, in part, the results of a randomized placebo-controlled trial in Mozambique. With all participants having received a new ITN at enrollment, two antenatal courses of SP had no protective effect against LBW (relative risk [RR]: 0.99; 95% CI: 0.70–1.39) or overall placental infection (p = 0.964) [81]. SP did demonstrate protection against malaria parasitemia at 8 weeks postpartum (RR: 0.52; 95% CI: 0.27–0.99; p = 0.044). It is unknown if this resulted in protection over a longer time frame, but this observation is important since the maternal risk of infection in the early postpartum period is higher than at any time during pregnancy [82].

In areas where there may be high ITN or IRS coverage resulting in the risk of malaria infection being greatly declined, the protection of SP against LBW may be limited to the effect of sulfadoxine against several common reproductive tract and sexually transmitted infections (RTI/STIs). Sulfadoxine is part of the sulfonamide family, compounds which have been used since the 1930s for their antibacterial properties. Sulfadoxine may be active against Neisseria gonorrhoeae and C. trachomatis since sulfamethoxazole is curative against both [83–85]. Sulfadoxine may also inhibit the effects of bacterial vaginosis. In vitro testing of three sulfonamides, sulfacetamide, sulfathiazole and sulfabenzamide, has shown strong growth inhibition of Gardnerella vaginalis, the most abundant microbial flora found in bacterial vaginosis [86]. AZ-based combinations that contain 3 g of AZ are most likely to provide optimal protection against RTI/STIs whilst clearing P. falciparum. Trials that employ fewer than 3 g may fail to detect these benefits [87,88].

Increasing coverage of IPTp with SP should be encouraged while safe and affordable alternatives are identified, with the exception being localities where extreme SP resistance has been observed. Changing transmission patterns may also reduce parity-acquired immunity in endemic areas, thus making previously immune multigravidae vulnerable to placental parasitemia. Reducing transmission rates raises the question of whether a ‘malaria only’ replacement for SP, such as mefloquine, can be justified if an AZ-based combination is capable of offering similar protection against the occasional P. falciparum infection whilst also offering broader benefits against RTI/STIs.

Future research

Comprised of 40 partner institutions in 28 countries around the world, a collaborative approach to answering questions on IPTp has been undertaken under the auspices of the Malaria in Pregnancy (MiP) Consortium [201]. In sub-Saharan Africa, the MiP Consortium is focused on: identifying at least one safe and effective alternative to SP for use in IPTp in the region with an emphasis on mefloquine and AZ-based combinations; assessing the protective effect of providing seasonal IPTp; and determining the optimal use of IPTp where ITN protection is simultaneously used. Additional studies are being conducted by other investigators with similar objectives. Table 1 summarizes key IPTp trials involving mefloquine or AZ that have already been conducted as well as planned studies that have been registered with ClinicalTrials.gov.

There has been some concern that AZ use might increase the prevalence of AZ- and erythromycin-resistant pneumococci. Trachoma eradication campaigns, having used AZ among targeted children in Australia and Nepal, found that treatment did select for macrolide-resistant pneumococcal strains in the nasopharynx [89,90] and conjunctiva [91]. Selection, however, was transient [89,92]. This needs to be demonstrated in IPTp trials and routinely monitored if an AZ-based combination is adopted as policy.

Delivery strategies need to be investigated specific to mefloquine and AZ-based combinations. Since mefloquine will likely require a split-dose to be well tolerated and AZ-based combinations rely upon 3-day regimens, only the first day of treatment will be observed for certain. Operational research is needed to determine if home visits through an ‘active delivery’ strategy with community health workers, community reproductive health workers and traditional birth attendants are needed to achieve high levels of adherence. A nonrandomized community trial in Uganda found that IPTp provided through a similar network was able to reach 67.5% (1404 of 2081) of pregnant women with two courses of SP whist the formal health system was only able to achieve similar coverage for 39.9% (281 of 704) [93]. Another study of active delivery in Uganda yielded even more impressive results. Community-directed drug distributors of ivermectin for onchocerciasis control also provided at least two courses of IPTp with SP to 88.5% of pregnant women (401 of 473). Health facilities passively reached 52.3% (237 of 453) of participants [94].

Given the combined antimalarial and antimicrobial protection that AZ-based combinations are likely to provide, ethnographic studies should assess whether knowledge of such broad-spectrum protection might affect the motivation of pregnant women to remain compliant and seek at least two courses of IPTp. The combination of AZ–CQ has a unique advantage in that it can be safely administered in any trimester. Knowing this may increase the acceptability of treatment from the perspectives of user and provider. An ethnographic study in Malawi found a common view that the bitterness of CQ meant that it could induce abortion. As a result, sugar-coated CQ tablets were tested and drug use increased by 64% [12]. Similar approaches would need to be evaluated for the AZ–CQ combination.

Where malaria transmission has fallen and IPTp as routine practice is questionable, new strategies should be sought. Tagbor and colleagues have found that intermittent screening for malaria using a rapid diagnostic test (RDT) as part of a focussed antenatal care package and treating only those women found positive for malaria (intermittent screening and treatment [IST]) was as effective as SP IPTp in Ghana [Tagbor et al., Manuscript Submitted]. Since the currently available RDTs are not sensitive enough to detect placental malaria there are still doubts about the usefulness of the IST approach. Given that an effective artemisinin-combination treatment can be used as part of IST and that the drug pressure will be minimized by this approach, it merits further study in a range of epidemiological settings.

IPT in infants

Purpose

The purpose of IPTi is to reduce the incidence of malaria and anemia in infants. As with IPTp, treatment doses of antimalarial drugs are given at times of routine contact with health services. In the case of IPTi, this is at times of vaccination. There is evidence that where SP resistance is not high, IPTi using SP reduces clinical cases of malaria in infants [95]. A meta-analysis of six SP IPTi studies showed a protective efficacy against clinical malaria of 30.3% (95% CI: 19.8–39.4) [96]. However, this analysis did not take into account potential effect modifiers of differences in levels of parasite resistance, different dosing strategies (number of courses of SP IPTi given and SP dose size) and different lengths of follow-up, nor did it include a trial that showed no protective efficacy of SP IPTi in an area of high resistance to SP [97], which was published alongside the meta-analysis. Only three out of seven SP IPTi trials demonstrated a significant effect of IPTi on anemia (Table 2), however, the meta-analysis of six trials showed a protective effect of 21.3% (95% CI: 2–32 ) against moderate anemia [96].

Mechanism

The major effect of IPTi appears to be mediated through chemoprophylaxis. Three studies focusing on the immediate period after an IPTi dose support this theory [98–100]. These studies show an immediate protective benefit from the IPTi dose that wanes over time, and is consistent with the half-lives of the drugs used and drug resistance in each setting. The meta-analysis of six IPTi trials has similar findings of the protection offered by IPTi occurring in the first 35 days after an IPTi treatment [96]. A trial comparing the long-acting drug mefloquine with the short-acting drug chlorproguanil-dapsone (CD) showed mefloquine to be highly efficacious for IPTi while CD was ineffective [97]. The lack of efficacy of CD for IPTi is likely to be attributable to its short half-life.

Drugs for IPTi

Sulfadoxine–pyrimethamine has been used to good effect in five out of seven IPTi trials (Table 2). The components of SP are slowly eliminated with estimates of the plasma elimination half-lives being 5–11 days for sulfadoxine and 3–5 days for pyrimethamine [101–106]. There are currently no data on pharmacokinetics in infants. A consequence of the elimination profile of SP is that levels of the drug sufficient to inhibit parasite growth persist for a considerable time after treatment, providing a period of post-treatment prophylaxis during which successful development of a blood-stage infection is prevented. Watkins et al. estimated this protection may occur as many as 60 days after treatment for fully sensitive P. falciparum[104]. Given that all the IPTi study sites had some level of SP resistance, one would anticipate that the protection given in practice by each IPT dose would be shorter than 2 months. The protection has been shown to last for approximately 6 weeks in Navrongo, Ghana, a low-resistance setting and approximately 3 weeks in high resistance settings in Kilimanjaro, Tanzania [98,99].

Two studies have evaluated alternative drugs to SP. Both studied CD while one studied mefloquine in Tanzania [97] and the other SP–AS and amodiaquine (AQ)–AS in Kenya [107]. There is an ongoing study using SP–AQ and SP–AS in Papua New Guinea. The results of these studies are shown in Table 2. Both alternatives may have problems with implementation. Mefloquine was shown to be highly efficacious at reducing cases of malaria in the first year of life (protective efficacy: 38%; 95% CI: 8–56), however, it has doubtful tolerability as it caused immediate vomiting in 8% of recipients and irritability during the first 2 days postdose in 15% [97]. SP–AS (protective efficacy: 25.7%; 95% CI: 6–41%) and AQ–AS (protective efficacy: 25.9%; 95% CI: 7–41%) were efficacious in Kenya [107]. The combination of SP–AS may face opposition due to the rising resistance of SP regionally and fears of using AS as what is effectively a monotherapy.

Controversies & new directions

WHO guidance: is it possible to implement IPTi?

The WHO recommends adoption of SP IPTi as a control strategy in areas with moderate-to-high transmission (annual entomological inoculation rates >10) and not high parasite resistance to SP. However, precise cut-offs of parasite resistance such as levels of in vivo efficacy or molecular markers of parasite resistance were not defined. More information is needed on the relationship between the prevalence of molecular markers (dhfr and dhps) and the duration of protection provided by SP IPTi. It is also acknowledged that the implementation of IPTi should not detract from efforts to scale-up access to artemisinin combination therapy (ACT) for early treatment, and to ITNs and IRS as preventive measures.

IPTi has been one of the most comprehensively studied malaria control strategies prior to formal implementation [108]. The strategy has been shown to be highly cost effective [109,110], acceptable to populations in several different countries in sub-Saharan Africa [111–113] and implementable at scale [114–116]. However, the caveats of the WHO policy recommendation will raise difficulties for policymakers as countries have within-country variation in transmission and in levels of resistance to SP. These difficulties are exemplified in East and Southern Africa where coastal areas have seen a substantial reduction in malaria transmission [7,117], dry central areas and highland interiors are low transmission [118] and resistance to SP is highest in Africa [60]. For example, in a recently published IPTi trial in Tanzania, in a low-transmission setting, malaria was so uncommon that the study only recorded four episodes of malaria over a 1-year period in 284 infants enrolled in the placebo group [97]. It has also been suggested that in areas where there is largely seasonal malaria, seasonal mass drug administration would provide larger public health benefits than IPTi when linked to vaccination [119]. A web-based tool has been developed to help policymakers [202], which estimates the cases averted in different transmission and seasonal settings. The growth of SP resistance in the last decade due to increased drug use has also cast doubt on the IPTi strategy, as resistant alleles to SP have spread widely across Eastern and Southern Africa and are beginning to circulate in West Africa [60,120]. Indeed, both SP IPTi trials with the best [18] and worst [97] efficacy took place in Tanzania, 7 years apart. It is likely that the WHO will recommend that SP IPTi should not be implemented where the prevalence of the resistant mutation at codon 540 of the dhps gene is greater than 50%. If this is adopted as a policy recommendation by the WHO then it is unlikely that SP IPTi can be implemented in many countries in East Africa as the prevalence of this marker is already above 50% [203].

High protective efficacy of the early SP IPTi trial

The major interest in SP IPTi was caused by the high efficacy found in the first study [18], which also showed a protective effect lasting into the second year of life [121]. The subsequent studies showed protective efficacies varying from 0 to 32%. Three explanations for this discrepancy exist. First, there was a rapid reduction in transmission intensity during the trial period. Two independent modeling exercises support this theory [122,123] with evidence that there was falling transmission at the time of the study [124]. Second, is that SP was acting as a ‘leaky vaccine’. This term describes the hypothesis that a partially effective drug combined with high ITN coverage may have led to attenuated blood-stage infections enabling immunity to develop without leading to clinical disease [125], this theory supported by experimental infections in human volunteers, which found that deliberate ultra-low dose infections were more immunogenic than those acquired naturally [126]. However, this effect was not found in other settings. Third, the finding could be a consequence of simple imbalances in the randomization, for instance, the placebo group were more exposed than the intervention group. While sufficient imbalances in randomization to produce such effects are unlikely a priori, given the result of the trial (high and sustained efficacy that was not replicated elsewhere), imbalance in randomization becomes a plausible explanation, and requires no complex biological or epidemiological hypothesis.

Future research

Alternative drugs to SP for IPTi are urgently needed and need to be long acting. AQ on its own [127] or in combination with AS [107] and mefloquine [97] have been shown to be effective at preventing malaria in infants. Mefloquine has issues of tolerability, whilst AQ and AS are both 3-day courses. Although no trials for new drugs for IPTi are underway, the likeliest new drug candidate is PQ. This drug needs to be fully assessed in randomized controlled trials and will probably need to be partnered with another long-acting antimalarial. Where possible, drugs should be pediatric formulations, designed to be heat stable with a long shelf life, easily transportable and include pharmacological studies as little is known about the pharmacokinetics of antimalarial drugs in this age group.

IPT in children

Purpose

The age distribution of severe and nonsevere malaria varies in different transmission settings [128–131]. Age-specific prevalence of asymptomatic infection, with its associated risk of anemia, also differs [132,133]. As a result, the age-groups that would benefit most from IPT will vary in different epidemiological settings. Even in contexts where infancy is the single year of life with the greatest malaria morbidity, the majority of the lifetime malaria burden lies outside this period. Consequently, IPT focused only on infants would achieve suboptimal protection in the majority of transmission settings.

The purpose of IPTc is thus to prevent clinical malaria and malaria-associated anemia in young children as well as in infants. Delivery of IPT to children following infancy requires the development of delivery systems other than the Expanded Program on Immunization (EPI). This could be challenging in some settings. A major advantage of delivery outside the EPI is that the timing of doses is not dictated by vaccination contacts and IPT can be given at the time of year when the malaria burden is greatest. Given that protection from IPT is restricted to the period when drugs have not been eliminated from the body, this may help to maximize efficacy [98]. Many IPTc studies have taken place in areas with seasonal malaria transmission because in such settings the majority of the malaria burden lies outside infancy and many doses given through the EPI would be given at a time of very low malaria risk [129,134–137]. Delivery of IPT to older children may be appropriate wherever the disease burden extends outside infancy; thus IPT in children should also be considered for areas of perennial transmission.

There have been relatively few studies of IPTsc at present. Under high perennial transmission in Kenya, the main benefits of IPTsc were a reduction in anemia, along with improvements in attention and certain cognitive functions [24]. By contrast, a recently published study from the Côte d’Ivoire compared three interventions, IPT, iron supplementation and anthelmintics over a 6-month period. The study noted that it was only antihelminth treatment that reduced the prevalence of anemia in 6–14-year-old children, but explained that the limited impact of IPT may have been due to an extended dry season in the year of study [138]. However, in a site with lower seasonal transmission in Mali, IPT in schoolchildren aged 6–13 years provided substantial protection against malaria episodes in addition to protection against anemia [139]. In low-moderate transmission sites, IPTsc can therefore be considered as an extended form of IPTc with an alternative delivery strategy.

Mechanism

The mechanism of IPTc has not been investigated directly, but due to the pharmacodynamics of the drugs used it is likely to work in a manner similar to IPTi. The importance of clearing existing infections is likely to increase with age because, in many settings, age-specific prevalence rises during childhood. In high transmission settings, children older than 1 year of age often carry asymptomatic infections, with parasite prevalence greater than 80% [118]. However, it is likely that prevention of new infections will remain an important component of IPT protection because longer-acting drugs are more effective than short-acting drugs for IPTc [140]. Where IPTc with long-acting drugs is given regularly throughout a short transmission season, IPTc can be considered a form of seasonal chemoprophylaxis [19]. Conversely, in areas of high transmission, the mechanism by which IPTsc works is more likely to be clearance of parasitemia, allowing hematological recovery rather than prophylaxis.

Drugs for IPTc

Sulfadoxine–pyrimethamine has been used in many of the trials of IPTc as its safety and tolerability are well established for treatment in children [141,142] and IPTi [96]. In West Africa, where most IPTc trials have taken place, the efficacy of SP remains reasonable, particularly in combination with another efficacious drug [143,144]. However, with the rising resistance to SP in other areas of Africa [3], alternative drugs need to be developed for IPTc. Long-acting drugs with an extended period of prophylaxis require fewer administrations to cover the same period. Single-dose regimens would allow directly observed therapy. There is a need to balance long periods of prophylactic protection against efficacy, safety, tolerability and the potential of drugs to select for resistance [101,145,146]. Slowly eliminated antimalarial drugs offer the largest ‘window of selection’ for drug-resistant parasites [147]. This is of particular importance in perennial transmission settings, or areas with a long transmission season where children would still be exposed to infection whilst there are sub-parasiticidal drug concentrations.

Combining drugs could reduce the likelihood for resistance to develop, particularly if the component antimalarial drugs are eliminated at a similar rate [148]. In IPTc studies SP has been used in combinations with a variety of antimalarial drugs, including AS, AQ and PQ. SP–AQ was shown to be more efficacious than SP–AS [140]. SP–PQ is at least as protective as SP–AQ, and due to the long action of PQ may be even more protective [149]. Other combinations that have been investigated for IPTc are AQ–AS [Tagbor et al., Manuscript in Preparation] and dihydroartemisinin–PQ [149], although the usefulness of artemisinins in IPT is likely to be limited due to their short half-life. Artemisinins may be more useful if IPTsc in high transmission areas becomes common practice, as clearance of existing infections that lead to anemia may be more important than preventing new infections. The results of recent IPTc studies are shown in Table 3.

Controversies & new directions

IPT & ITNs

Use of ITNs is highly efficacious in protecting against malaria [150,151]. It is therefore important to consider the extra benefit provided by IPT to ITN users (and vice versa) given that there may be some overlap in the protection each intervention provides. Previous studies showed a clear additional benefit of chemoprophylaxis with dapsone–pyrimethamine [152,153] in ITN users. A multisite trial has been undertaken in Burkina Faso and Mali to determine the efficacy of IPT in children sleeping under an ITN [204]. Preliminary results suggest that IPTc adds substantially to the benefit of ITNs, although this requires further investigation, particularly in areas where use of ITNs is high among all age groups and where the ‘mass effect’ of ITN is most beneficial.

Rebound

One of the original aims of IPT was to avoid the potential negative effect of chemoprophylaxis on the development of immunity, which may lead to increased morbidity, the rebound effect. High levels of protection have been achieved in a number of studies of seasonal IPT in children, and the potential to improve IPT further by combining it with other interventions such as ITNs, mean that these concerns remain. It is essential to clarify if high levels of protection from IPTc lead to increased morbidity and mortality in later life, and weigh these against the benefits of averted disease. Long-term follow-up of ITN trials was undertaken for this reason and found no rebound effect [154–156]. However, ITNs also act to reduce transmission and it may be this effect that prevented rebound [123]. Despite mixed results of the individual trials, a pooled analysis has concluded that rebound after IPTi is unlikely to outweigh the protection provided [96] and no IPTc studies to date have found conclusive evidence of rebound. However, if implemented in practice, IPTc may result in children receiving a high level of protection against malaria for the first 5 years of life. No trials have lasted this long so there is no robust evidence on whether rebound morbidity will be seen or not. A recently published modeling analysis suggested that even substantial rebound effects of IPTc would still lead to a positive effect overall [157]. In a detailed analysis of follow-up from a year-long chemoprophylaxis trial [16], Aponte and colleagues found that although there was an increase in malaria episodes in the intervention group there were fewer severe episodes compared with the placebo group [17]. To clarify if this is the case in practice, studies need to follow children after a full-length intervention period (e.g., 5 years, rather than a single year) to determine the effect of longer-term IPT. Studies therefore have to be postimplementation Phase IV studies.

Future research

Drug resistance

As for other IPT strategies, a replacement is urgently needed for SP in some areas of Africa. Of the existing available long-acting drugs used for treatment, mefloquine, AQ and PQ are all sufficiently long-acting, but all belong to the same class of drugs (4-aminoquinolines). This raises the possibility of cross-resistance and, potentially, similar tolerability issues for particular children with all of these drugs. Other long-acting drugs in the pipeline are napthoquine, tafenoquine (also effective against gametocytes, liver stages and Plasmodium vivax) and pyronaridine. In recent years, there has been an emphasis on the development of short-half-life drugs for the treatment of malaria in order to avoid rapid development of resistance. However, now there is a need for the development of long-acting drugs specifically for preventive use in IPT.

Delivery strategies

New delivery systems need to be developed for IPTc, partly to allow seasonal targeting of IPT and partly to allow delivery to older children, that is, outside the context of the EPI. Delivery via community-based health workers has been used successfully in Senegal, The Gambia and Ghana to deliver seasonal IPT in children. Other community-based delivery systems should be investigated including the possibility of using other established mass administration programs such as vitamin A supplementation, onchocerciasis control, EPI catch-up or antihelmintics [158,159]. IPTc coverage may be higher when using community health workers than if distributed by EPI clinics [160]. The same effect was seen in a clinical trial in The Gambia [Bojang et al., Unpublished Data] [205]. Experience from Tivaouane district, Senegal, where IPTc has been piloted since 2006 suggests that coverage of seasonal IPTc delivered by trained members of the community is high and equitable [161]. However, because coverage and equity are likely to be affected by socio–cultural factors related to recipients, in addition to the delivery mechanism itself, further research will be needed in specific settings.

Discussion

At present, IPTp with SP is the only strategy that is WHO policy and is widely implemented. SP IPTi has passed through two WHO technical expert groups and has been recommended to be put forward as WHO policy. No IPTc strategy has started the process of WHO policy development. All strategies face three major issues; namely, the usefulness of mass therapy when the incidence of malaria is falling, rising levels of drug resistance to the most commonly used agent, SP, and how to deliver the strategy to those most at risk.

Falling transmission

Successful malaria control with the widespread distribution and use of long-lasting ITNs and improved access to prompt and effective treatment for malaria with ACT has reduced transmission of malaria. This in turn will change the age pattern of the disease burden, increasing the average age of clinical cases [162]. This will affect the age range that IPT programs should be designed to cover. The impact of IPT in infants and young children will decrease as transmission continues to fall, and it may be necessary to extend IPT to older children and even to adults. If major transmission reductions are achieved in the future there will be a lower limit of transmission beyond which giving IPTc to a wide range of age groups is not cost effective or acceptable. However, malaria transmission can be highly focal within a small area [163], particularly where transmission has recently declined [164]. Thus, it may be necessary to tailor malaria control interventions including IPT to different settings found in nearby localities.

The impact of IPT strategies on malaria transmission has been investigated by mathematical modeling studies but not in practice. IPTp and IPTi only treat a small proportion of those people likely to carry parasites and therefore these strategies are unlikely to play a major role in malaria elimination efforts [165]. However, IPTc may make an important contribution towards reducing transmission [157], especially in programs targeting malaria in older children, as these age groups represent a large part of the biomass of malaria that is carried in a population, and may be responsible for a considerable proportion of overall malaria transmission. Adding an artemisinin into the regimen may improve the role of IPT in reducing transmission as they are cidal to mature gametocytes [166,167]. Research agendas should include measures of transmission in their outcomes.

In the face of reducing transmission, countries are advised to continue providing IPTp where IPTp delivery is scaled-up and delivery systems are working, particularly if SP is succeeded by an antimalarial-antibiotic combination. In the absence of a var2csa vaccine, the profile of disease burden is likely to include more multigravidae than ever as transmission rates decrease, underscoring the importance of continued IPTp administration in the antenatal package.

Resistance

The spread of SP resistance in recent years has been explained by its increased use as a first-line treatment for uncomplicated malaria [120]. Now that most countries have switched to ACT as first-line therapies and as coverage of these improves, one would expect selection pressure for SP resistance to reduce. However, SP is still commonly used in the community without reference to formal healthcare systems, which is particularly problematic because dosage may be inadequate and the quality of drugs substandard [168,169]. How much drug pressure is exerted by IPT strategies is unknown. Modeling exercises predict little effect for IPTi due to the small number of individuals treated [146,170], but IPTc may well exert a much larger effect. Thus it is preferable to use a drug regimen different from that which is used for treatment of clinical episodes of malaria for IPT in order to retain the efficacy of drugs for treatment and IPT. In particular, artemisinins should be conserved for treatment because they have little role in prophylaxis, and artemisinin tolerance has now been demonstrated in Southeast Asia [171].

Delivery & equity of access

Active delivery strategies with community-based workers would greatly improve IPT coverage rates that currently rely on passive clinic attendance. Such outreach may improve coverage, equity and the ability to target the most exposed individuals who suffer with the largest malaria burden. Malaria disproportionately affects the rural poor who typically have limited access to formal healthcare due to opportunity and financial costs [169]. Coverage and access to IPT will have a major impact on whether the encouraging results from trials of the different approaches are borne out in practice. IPT strategies will have a limited impact despite high coverage if they only reach individuals who are at low risk of malaria, for example, those living in urban areas. By targeting IPT geographically and adding alternative delivery strategies to clinic attendance, it may be possible to reach the individuals who are most at risk regardless of whether they have access to, or use, formal health facilities. Future research for all forms of IPT should therefore consider which delivery strategy achieves the highest coverage of the most highly exposed individuals, whether IPT can contribute to reducing health inequalities, and how sustainable these approaches are outside of research settings.

Conclusion

Despite the advances in vector control, personal protection and vaccines, drug therapy remains the mainstay of malaria control [172]. Prompt diagnosis combined with access to effective and affordable antimalarial drugs appear to have been successful in reducing morbidity and mortality from malaria [8]. However, while coverage of these interventions remains poor and inequitable [169], other strategies are required to help alleviate the burden of malaria. The use of drugs to prevent malaria and clear existing infections in high-risk groups thus continues to be an attractive control strategy. IPT strategies are simple, appear relatively easy to implement and are efficacious in moderate and highly endemic areas. However, IPT strategies will have to be adapted depending on the levels of transmission and drug resistance. IPTp may have to change to include antibacterial treatment, IPTi may struggle to be implemented and IPTc may need to include a broader age range to maximize the impact on the malaria burden. Developing effective delivery strategies is key to ensuring that the intended benefits of IPT are achievable in practice.

Expert commentary

With historic levels of funding for the fight against malaria, the WHO is working towards the global elimination and eradication of malaria. This is a departure from malaria control measures over the past 50 years which focused on reducing the burden of malaria by treating clinical disease or preventing infections in high-risk groups. Eliminating or eradicating malaria transmission necessitates reducing, and eventually clearing, malaria carriage in all individuals.

The IPT of malaria was developed to mitigate the burden of disease among the most vulnerable groups, namely pregnant women and children. SP has been the most widely used antimalarial for IPT strategies to date, but a replacement drug will be necessary in many areas of Africa due to SP resistance. IPT strategies will also need to be amended as malaria transmission falls due to scale-up of other control efforts, including ACT for treatment, ITNs and IRS of insecticides.

In the case of IPTp, the only widely implemented IPT strategy, two leading candidates have emerged to replace SP and are now being tested in clinical trials: mefloquine and AZ-based combinations. Mefloquine is an antimalarial drug that appears to be safe in the second and third trimesters, but poor tolerability must be resolved before it would be acceptable to policymakers. Split-dosing and the use of microsphere formulations may reduce adverse events. AZ-based combinations are attractive for several reasons; AZ has antibacterial properties that treat most of the common RTIs; AZ is safe in all trimesters of pregnancy; and AZ appears to have a synergistic effect with CQ against P. falciparum. As both AZ and CQ are safe in any trimester of pregnancy, health facilities can provide IPTp at any time. This is important since current guidelines are based on operational convenience and drug safety rather than the natural course of disease; maternal parasite densities peak between 9 and 16 weeks in pregnancy, tapering until term [30,31].

While the burden of malaria is declining in selected areas of the sub-Saharan region, this is not the case for RTIs and STIs. Thus, the contribution of RTI/STIs to adverse birth outcomes will rise as the risk of malaria in pregnancy declines, reinforcing the value of using an AZ-based combination for IPTp. Careful monitoring of pneumococci sensitivity to antibiotics, particularly macrolides, would be critical if AZ is used in IPTp. With reducing transmission, exposure to parasite antigens will decrease resulting in reduced immunity to malaria and specifically the effects of placental parasitemia. For this reason, ministries of health should be encouraged to continue providing IPTp no matter what the replacement for SP is whilst malaria transmission continues. If mefloquine replaces SP and transmission rates continue to decline, then it might be more appropriate to provide IST for malaria; AZ-based combinations, given their broad-spectrum protection against bacterial infections, may be more appropriate for an IPTp strategy that is maintained while transmission rates decline.

For IPTc and IPTi, there is no obvious current replacement for SP. These strategies may be restricted to areas where SP remains effective until sufficient efficacy and safety data is available on alternative regimens. Long-acting drugs will be required because the chief benefit of IPT in infants and young children appears to be chemoprophylaxis. There is evidence that two long-acting drugs are more efficacious than a long-acting antimalarial with an artemisinin [140], but the effect of such an approach on development of drug resistance requires further investigation. As with IPTp, the replacement for SP should not be simultaneously used for treatment of uncomplicated malaria so that IPT strategies do not contribute to resistance of first-line treatments. Seasonal, campaign-style delivery of IPT could possibly rotate different therapies so as to minimize the opportunity for resistance selection. The potential for rebound morbidity after long-term IPTc needs to be monitored and balanced against the malaria episodes that would be averted in childhood and the effects of chronic malaria-attributable anemia.

Under conditions of moderate-to-high transmission, the largest reservoir of malaria parasites proportionate to their population can be found among children under 5 years of age. Thus, IPT that targets young children as well as infants (IPTc) may make substantial reductions in the overall malaria burden. In the future, the continued use and scale-up of various types of IPT may assist ongoing malaria elimination and eradication efforts using other tools. A wider age range may need to be included in IPTc programs as transmission is reduced. The impact on transmission of the IPT strategies in their current form has not been studied. Artemisinin drugs may be useful in IPT for reducing transmission due to their effects on mature gametocytes, however, in current IPT strategies, artemisinins have little role as they do not provide prophylaxis. As malaria control efforts progress towards elimination, alternative mass treatment strategies to IPT may need to be developed and employed to maximize the impact on malaria transmission.

Five-year view

Malaria epidemiology appears to be rapidly changing. In areas where surveillance is established and it is possible to monitor changes in transmission, assess risk and measure antimalarial drug resistance, it is likely that malaria control will become more tailored to the locality. For example, in areas where transmission has fallen, IPTp will be tailored to IST and mass drug administration in the form of IPT will be given to areas at high risk of malaria. In many countries, surveillance systems are weak and it is likely that IPTp will continue to be scaled-up and SP replaced by mefloquine or AZ–CQ combinations once evidence of efficacy has been produced. IPTi may be adopted as policy by the WHO but may struggle to be implemented due to heterogeneity of malaria risk and levels of drug resistance to SP. IPTc is likely to be implemented in countries with seasonal malaria transmission where there is limited SP resistance and will use combinations of long-acting antimalarial drugs such as SP–AQ. Mass treatment strategies such as IPT will need to be further developed to maximize their potential role in malaria elimination.

Table 1. Intermittent preventive treatment in pregnancy trials using mefloquine, azithromycin-based combinations, or intermittent screening and treatment with sulfadoxine–pyrimethamine alternative.

Trial location	Seasonality	Drugs used	Subject profile and sample size	End points	Ref.
Ouidah, Benin	Seasonal	SP IPTp MQ IPTp	All parities enrolled at 16–28 weeks of gestation (n = 1601)	Primary: LBW: MQ = 8% (59 of 735); SP = 9.8% (72 of 730); difference between mean LBW = 1.8% (95% CI: -4.8–1.1%) Secondary: • Placental parasitemia: MQ = 1.7%; SP = 4.4%; p = 0.005 • Clinical malaria: MQ = 26 cases/10,000 person-months vs SP = 68 cases/10,000 person-months; p = 0.007 • Maternal anemia (<10 g/dl): MQ = 16% vs SP = 20%; p = 0.09 • Adverse events: MQ = 78 vs 32%; p < 10^-3	[67]
Blantyre, Malawi	Perennial with seasonal peak	SP IPTp AZ + SP IPTp AS + SP IPTp	All parities enrolled at 14–26 weeks of gestation (n = 141)	Primary: Tolerability: SP = 7; AZ + SP = 7; AS + SP = 12 Pregnancy outcomes: SP = 5; AZ + SP = 5; AS + SP = 7 Secondary: Peripheral parasitemia: SP = 10 (30.3%); AZ + SP = 9 (27.3%); AS + SP = 5 (14.3%); SP vs (AZ + SP): p = 0.50; (SP vs AS) + SP: p = 0.10	[88]
Mangoch, Malawi	Perennial with seasonal peak	CQ MQ IPTp	All parities enrolled (n = 3380)	Primary: Treatment failure: CQ vs MQ; OR = 30.9 vs 11.1; p < 10^-6 Secondary: Peripheral, placental or cord parasitemia: CQ vs MQ; OR = 8.7, 7.4 and 4.1; p < 10^-6	[28]
Thai–Burma border	Seasonal	Placebo MQ IPTp	All parities enrolled at >20 weeks of gestation (n = 339)	Primary: Protective efficacy: MQ = 86% (95% CI: 59–94)	[173]
Juaben, Ghana	Seasonal	IST with RDT and SP Rx + ITN AQ + AS IPTp SP IPTp + ITN	All parities enrolled at 16–20 weeks of gestation (n = 3330)	Primary: Severe anemia (Hb <8 g/dl) Secondary: LBW, maternal anemia (Hb <11 g/dl), placental parasitemia	[206]
Cotonou and Porto-Novo, Benin	Seasonal	CTX MQ IPTp	All parities HIV-positive enrolled at 16–20 weeks of gestation (n = 492)	Primary: Placental parasitemia Secondary: Mean parasite density of placenta at delivery, LBW, maternal anemia (Hb <11 g/dl)	[207]
Benin Gabon, Kenya, Tanzania, Mozambique	Seasonal and perennial with seasonal peak	Trial 1: SP IPTp + ITN MQ IPTp + ITN MQ (split) IPTp + ITN Trial 2: CTX + placebo + ITN CTX + MQ IPTp + ITN	Trial 1: HIV-negative women (n = 4260) Trial 2: HIV-positive women (n = 1070)	Primary end points: Trial 1: LBW Trial 2: Peripheral parasitemia Secondary end points: Trial 1: Placental parasitemia; moderate maternal anemia Trial 2: Placental parasitemia; LBW	[208]
Mangochi, Malawi		SP IPTp (twice) SP IPTp (4-week intervals) SP IPTp (4-week intervals) + AZ (twice)	All parities enrolled at 14–26 weeks of gestation (n = 1320)	Primary: Preterm births Secondary: LBW, duration of gestation, placental and peripheral parasitemia, RTI/STIs	[209]

AQ: Amodiaquine; AS: Artesunate; AZ: Azithromycin; CQ: Chloroquine; CTX: Cotrimoxazole; Hb: Hemoglobin; IPTp: Intermittent preventive treatment in pregnancy; IST: Intermittent screening and treatment; ITN: Insecticide-treated net; LBW: Low birthweight; MQ: Mefloquine; OR: Odds ratio; RDT: Rapid diagnostic test; RTI: Reproductive tract infection; Rx: Treatment; SP: Sulfadoxine–pyrimethamine; STI: Sexually transmitted infection.

Table 2. Published results of trials in intermittent preventive treatment in infants.

Trial location	EIR, seasonality	Incidence per person-year (placebo)	Age at doses (months)	Protective efficacy (%) (95% confidence interval)	Ref.
Ifakara, Tanzania	30, perennial	0.43	2, 3, 9	Malaria: 62 (44–74) Anemia: 50 (8–73)	[18]
Navrongo, Ghana	418, seasonal	1.0	3, 4, 9, 12	Malaria: 25 (14–34) Anemia: 35 (11–53) (both to 15 months of age)	[174]
Manhiça, Mozambique	38, perennial	0.55	3, 4, 9	Malaria: 23 (2–39) Anemia: 13 (-17–35)	[175]
Lambaréné, Gabon	50, perennial with seasonal peaks	0.22	3, 9, 15	Malaria: 17 (-24–44) Anemia: 22 (-1–40) (both to 18 months of age)	[176]
Tamale, Ghana	Perennial with seasonal peaks	0.88	3, 9, 15	Malaria: 23 (12–32) Anemia: 31 (3–51) (both to 18 months of age)	[61]
Kumasi, Ghana	400, perennial with seasonal peak	1.29	3, 9, 15	Malaria: 20 (11–29) Anemia: 7 (-8–20) (both to 18 months of age)	[177]
Kilimanjaro, Tanzania	Two sites, moderate transmission site EIR 10 perennial with seasonal peaks, low transmission EIR <1, seasonal	Moderate transmission site 0.41 Low transmission site 0.018	2, 3, 9	Malaria: SP: -6.7 (-45.9–22.0) MQ: 38.1 (11.8–56.5) CD: 10.8 (-24.6–36.1) Anemia: SP: -16 (-57–14) MQ: -6 (-44–22) CD: -13 (-53–17)	[97]
Kisumu, Kenya	7, perennial peaks	0.98	2, 3, 9	Malaria: SP–AS: 25.7 (6–41) AQ AS: 25.9 (7–41) CD: 16.3 (-5–33) Anemia: SP–AS: 27.5 (-7–51) AQ AS: 23.1 (-12–47) CD: 11.4 (-29–39)	[107]

AQ: Amodiaquine; AS: Artesunate; CD: Chlorproguanil-dapsone; EIR: Entomological inoculation rate (infected bites per person per year); MQ: Mefloquine; SP: Sulfadoxine–pyrimethamine.

Table 3. Trials of intermittent preventive treatment in children.

Trial location (year)	EIR, seasonality	Age group	Drugs used	Delivery schedule	Protective efficacy (%) (95% confidence interval)	Ref.
Niakhar, Senegal (2002)	10, seasonal	≤5 years	SP + AS	Monthly, 3 months	Malaria: 86 (80–90)	[178]
Niakhar, Senegal (2004)	10, seasonal	≤5 years	SP + AQ SP + AS AQ + AS	Monthly, 3 months	Malaria: 93^† (SP–AQ)	[140]
Kambila, Mali (2002–2003)	Seasonal	≤10 years	SP	Bimonthly, 2 months	Malaria: 40 (25–51)	[179]
Hohoe, Ghana (2005–2006)	65, seasonal	<5 years	SP AQ + AS AQ + AS	Bimonthly Monthly Bimonthly	Malaria: 24 (14–33) Anemia: 30 (6–49) Malaria: 69 (63–74) Anemia: 45 (25–60) Malaria: 17 (6–27) Anemia: 32 (7–50)	[180]
Keur Soce, Senegal (2007)	<10^‡, seasonal	<5 years	SP–PIP DHA–PIP SP–AQ	Monthly	SP–PIP lowest cumulative incidence^§ Both SP–PIP and DHA–PIP noninferior to SP–AQ	[149]
Bondo, Kenya (2005–2006)	Very high^¶, perennial	5–18 years	SP–AQ	Three doses, 4-month interval	Anemia: RR = 0.52 (0.29–0.93)	[24]
Kolle, Mali (2007–2008)	High, seasonal	6–13 years	SP–AS AQ–AS	Two doses, 2-month interval	Malaria: 66.6 Malaria: 46.5	[139]

^†No placebo group for comparison; estimate of protective efficacy for SP–AQ is based on hazard ratio compared with SP–AS (0.50 [95% CI: 0.31–0.81]) in 2004 and the protective efficacy of SP–AS in the previous Niakhar study in 2002. There was no significant differences between SP–AS and AQ–AS in this study. Cumulative incidence in SP–AQ group was significantly lower than in other groups.

^‡EIR not measured, but malaria incidence in a comparison area was lower than in the Niakhar study in 2002.

^§Noninferiority trial with no placebo group for comparison, point estimate of cumulative incidence lowest in the SP–PIP group.

^¶EIR not reported.

AQ: Amodiaquine; AS: Artesunate; CI: Confidence interval; DHA: Dihydroartemisinin; EIR: Entomological inoculation rate; PIP: Piperaquine; RR: Relative risk; SP: Sulfadoxine–pyrimethamine.

Key issues

• • Intermittent preventive treatment (IPT) against malaria is a strategy to reduce disease burden in vulnerable groups, namely in pregnant women and children.

• • IPT in pregnancy (IPTp) with sulfadoxine–pyrimethamine (SP) has been shown to improve maternal and fetal outcomes where resistance to SP (day 14 end points) is less than 40% and is recommended by the WHO as a control tool.

• • With growing levels of resistance to SP and falling transmission levels, new drugs for IPTp and new strategies such as intermittent screening and treatment are under evaluation.

• • Extending the role of IPTp to protect against malaria and treat reproductive tract infections by using azithromycin combinations may prolong the strategies usefulness in settings of low malaria transmission.

• • IPT in infants (IPTi) using SP has been shown to reduce cases of malaria and moderate anemia in areas where SP resistance is not high. IPTi is under review by the WHO as a tool for malaria control.

• • The main mechanism of IPTi is through prophylaxis.

• • Replacement drugs for SP IPTi have been investigated but all have problems with tolerability or delivery.

• • IPT in children (IPTc) appears to be highly efficacious, especially when given in the seasonal setting. Long-acting drug combinations have been tested and are efficacious. When used in combination with insecticide-treated nets there is additional protection.

• • Broadening the age range covered by IPTc may make the strategy a useful tool for elimination of malaria by reducing transmission.

• • All strategies face problems of drug resistance and their usefulness in settings of decreasing transmission.

• • Delivery strategies for IPT need to be assessed to see if those most at risk from malaria receive the intervention, and if the strategies are sustainable out of the research context.

Financial & competing interests disclosure

Roly D Gosling and Daniel Chandramohan are salaried by their institutions, Matthew E Cairns is funded by the MRC UK and R Matthew Chico is funded from a grant from Pfizer on a trial of azithromycin and chloroquine for malaria in pregnancy. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.