The global pipeline of new medicines for the control and elimination of malaria

Over the past decade, there has been a transformation in the portfolio of medicines to combat malaria. New fixed-dose artemisinin combination therapy is available, with four different types having received approval from Stringent Regulatory Authorities or the World Health Organization (WHO). However, there is still scope for improvement. The Malaria Eradication Research agenda identified several gaps in the current portfolio. Simpler regimens, such as a single-dose cure are needed, compared with the current three-day treatment. In addition, new medicines that prevent transmission and also relapse are needed, but with better safety profiles than current medicines. There is also a big opportunity for new medicines to prevent reinfection and to provide chemoprotection. This study reviews the global portfolio of new medicines in development against malaria, as of the summer of 2012. Cell-based phenotypic screening, and ‘fast followers’ of clinically validated classes, mean that there are now many new classes of molecules starting in clinical development, especially for the blood stages of malaria. There remain significant gaps for medicines blocking transmission, preventing relapse, and long-duration molecules for chemoprotection. The nascent pipeline of new medicines is significantly stronger than five years ago. However, there are still risks ahead in clinical development and sustainable funding of clinical studies is vital if this early promise is going to be delivered.


Background
Recent years have seen a transformation in the landscape of malaria drug discovery and development. A review of publicly available data from PubMed, clinicaltrials.gov as well as proprietary databases, such as Thomson Pharma, revealed that as of March 2012, over 50 projects were in progress worldwide. These include a wide variety of molecules in the window between late discovery (within 12 months of starting regulatory preclinical studies) and launch. Twenty-seven of these are in formal regulatory studies and beyond. In addition, in the past three years, five new medicines have either been approved by Stringent Regulatory Authorities [those that adhere to the guidelines of the International Commission for Harmonisation (ICH), or prequalified by the World Health Organization (WHO)].
The portfolio of new medicines contains several generations of products. The oldest group reflects the requirement for fixed-dose, artemisinin-combination therapy (ACT), driven by the need to protect artemisinin from the emergence and spread of resistance. This therapy is based around active molecules that were first identified over 30 years ago. The central focus for the development process of this therapy has been ensuring that the historical data available match current international standards of safety and efficacy.
The call for the eradication of malaria in 2007 led to a new research agenda [1] laying out priorities with broader horizons. As well as the focus on case management of uncomplicated and severe malaria, there is now a need for medicines that prevent transmission, stop relapse and provide chemoprotection against new infections. These characteristics are described by Target Product Profiles (TPPs), which give a description of the ideal medicine and the minimum acceptable profile. TPPs provide drug discoverers with a common standard of the unmet clinical needs in malaria control and eradication, thereby allowing a better focus. TPPs were developed with input from countries, field-based clinicians, and discovery and development teams [2], and are refined every two years [3].

Methodology
This review aims to be as complete as possible, and not simply to be a review of projects supported by Medicines for Malaria Venture (MMV). To obtain such a global view, the data were obtained from publicly available sources, such as Medline, clinicaltrials.gov, and company websites, supported by commercial databases, such as Thomson Pharma (www.thomson-pharma.com). There is unfortunately insufficient space to cite every single source. The data were updated in March-April 2012, although it is appreciated that the picture is continually evolving.
Controlling malaria: artemisinin combination therapy as first-line treatment Today, first-line medicines against malaria are fixeddose artemisinin combination therapy (ACT). These medicines are assumed to be active against the blood stages of all of the major forms of Plasmodium which infect humans: falciparum, vivax, malariae, ovale and knowlesi. Fixed-dose combination therapy has the advantage over co-blistered presentations in that it eliminates the potential for monotherapy [4], which is to be avoided since it risks the emergence and selection of resistant parasites [5]. Six of these have been reviewed by regulatory authorities throughout the world (see Table 1, Figures 1 and 2). Artemether-lumefantrine (Coartem W and Coartem W Dispersible from Novartis), artesunate-amodiaquine (Coarsucam™/Artesunate Amodiaquine-Winthrop W from Sanofi), pyronaridine-artesunate (Pyramax W from Shin Poong Pharmaceuticals) and mefloquine-artesunate from Cipla/DNDi-have been prequalified by WHO [6]. Their launch has had a dramatic impact on the number of courses of treatment available to malaria patients. There has been rapid growth from 62.3 million treatments in 2006 to 159.7 million in 2010 (see Figure 2) although, still, not all medicines reach the patients who need them [7]. Considerable progress has been made on price, with the costs of an adult course of treatment falling to $1.00-$1.40. For infants the price can be as low as $0.30 [8], although in 2011 prices rose due to perceived artemisinin shortages [9].
The two major artemisinin combination therapies are: (a) Artemether-lumefantrine (Coartem W and the dispersible paediatric formulation Coartem W Dispersible from Novartis). Over 500 million treatments of these medicines have been delivered since the initial launch in 2002, of which over 100 million are now the dispersible form specifically designed for children. Several generic versions of this medicine have now been produced, including those prequalified by WHO. (b) Amodiaquine-artesunate (Coarsucam™ and Artesunate Amodiaquine Winthrop W , from Sanofi). This was initially approved in Morocco, where it is manufactured and was prequalified by WHO in 2008.
There are four other types of ACT prescribed for the treatment of uncomplicated malaria, which are currently at different stages of registration and approval.
(a) Dihydroartemisinin (DHA)-piperaquine (Eurartesim W from Sigma-tau) was approved by the European Medicines Agency (EMA) in October 2011 [43] and is included in the Malaria Treatment Guidelines this year [44]. New data showing twoyear stability are now available. Prequalification by WHO and submission in key disease-endemic countries is expected in 2012. Holley-Cotec produce another version of DHA-piperaquine (Duo-Cotecxin W ) [45], available in many countries, which is expected to be submitted soon for prequalification. (b) Pyronaridine-artesunate (Pyramax W from Shin Poong Pharmaceuticals) was approved by the Korean Food & Drugs Administration (KFDA) in August 2011 [46], and was approved by the EMA in February 2012, under Article 58 [47], where opinion is given on whether a medicine is suitable for use in countries where the disease is endemic. This avoids the obligation to market the medicine in Europe, and is a decision made in conjunction with WHO, which has now prequalified the product. (c) Artesunate-mefloquine (ASMQ) is a fixed-dose combination produced by Cephalon/Mepha as a paediatric formulation for commercial markets in Africa, and by the Drugs for Neglected Diseases initiative (DNDi) in partnership with Farmanguinhos for use in Brazil. From 2012, the fixed-dose combination will be manufactured and registered by Cipla in India, which will accelerate uptake [48]. WHO prequalification was obtained in September 2012. Currently the market price of mefloquine (over $1,000/kg) makes this combination the most expensive ACT, however work is ongoing to lower the cost of manufacturing. A cheaper synthesis of mefloquine has been developed by Development Chemicals in collaboration with DNDi and MMV. This allows a price similar to other fixed-dose ACT. (d) Artemisinin-naphthoquine (ARCO W , Kunming, China) is available in Africa as a one-day treatment.
There are relatively few data available on the  efficacy and safety of naphthoquine [49], and the product has not yet been submitted for approval either to a Stringent Regulatory Authority or to WHO. The adult dose of artemisinin is high (15-20 mg/kg) [50], reflecting the poor bioavailability of the parent molecule.
Since over 85% of malaria patients are under five years old, development of child-friendly paediatric formulations remains the priority. Coartem W Dispersible (dispersible artemether-lumefantrine) is the first example of a taste-masked dispersible ACT, and was developed in a collaboration between MMV and Novartis [51]. Two other paediatric formulations are in development: a granule formulation of pyronaridineartesunate is to be submitted in early 2013 [52], and a dispersible formulation of DHA-piperaquine is planned for submission later in the same year.  Table 1.
Having more than one ACT available is an advantage. Each medicine will have a different impact, depending on the endemicity of the disease, the likelihood of re-infection, diet, co-medications, and the balance between the different forms of malaria: falciparum, vivax, malariae, ovale or knowlesi. These factors are summarized in Table 2. Resistance to partner drugs has already occurred in some countries, as has been demonstrated for amodiaquine [53], and this remains a threat for other therapy partners in many countries. Clinical signs of decreasing effectiveness of partners have been described, and loss of susceptibility to artemisinin has been reported in the Thai-Cambodia border regions [54,55] as well as more recently in the Thai-Myanmar border region [56]. Having multiple first-line therapies in a country may help reduce the spread of resistance [57,58].
Pyronaridine-artesunate is currently the only ACT with regulatory approval for activity against P. vivax, although it is widely assumed from field experience that other fixed-dose combinations will be equally active against P. malariae and the blood stages of P. vivax and P. ovale [31,32].
What challenges remain for the next generation of medicines? There is a continual threat of the emergence of resistance, both to artemisinin and the partner medicine. This will require new classes of medicines. In addition to this there are four areas of focus for drug discovery. Firstly, in the context of malaria eradication, there is a need for medicines that can be administered as a single dose, which will allow direct monitoring of administration and improve compliance. These should have activity against all existing resistant strains of parasite. Secondly, new medicines are needed that kill gametocytes, and thus prevent transmission. Thirdly, there is a need for medicines which prevent relapses of P. vivax. Finally there is a need for molecules with longer half-lives to give chemoprophylaxis or long-term protection against re-infection [58]. Figure 3 highlights the global portfolio of anti-malarial medicines in development organized by development stage (as of March 2012). Figure 4 shows the global portfolio of antimalarial medicines organized by therapeutic type, as discussed below.

New treatments for severe malaria
In 2010, a study in African children with severe malaria [61] showed a significant reduction of mortality using intravenous or intramuscular injections of artesunate in place of quinine, confirming earlier results in adult patients from Asia [62,63]. WHO now recommends artesunate as a firstline therapy in severe malaria [44]. Artesun W (a for injection), manufactured by Guilin Pharmaceuticals (see Table 3), was the first to be prequalified by WHO in 2010 [6]. The cost of approximately $1.00 per treatment is justified by the better outcome compared with quinine [64]. As with ACT, it is important that there are multiple and reliable production sources, otherwise countries are reluctant to change treatments. IPCA manufactures artesunate for injection for India (see Table 4), and is planning to submit for prequalification.
Other therapies for severe malaria are in the pipeline. Most are adjunctive therapy, given on top of a schizonticide to reduce the sequelae of severe malaria, such as neurological damage. The track record of most potential  adjunct therapy has not been compelling, despite promising results in rodent studies. Adjunct approaches have generally shown little benefit. Recent reports on molecules such as N-acetyl cysteine [123] and pentoxifylline [124] failed to show any significant benefit. One interesting example of adjunct therapy is DF-02 (sevuparin sodium), a heparin derivative with no anti-coagulation activity that blocks rosetting (a key event in severe malaria [68]), which has just completed Phase I trials. The only other molecule specifically being developed for this indication is a TLR-9 antagonist (E6446) from Eisai [125,126]. One success has been sublingual sucrose, which overcomes hypoglycaemia and, in a pilot study, resulted in a significant reduction in mortality [79]. The only new schizonticide specifically being proposed for severe malaria is SAR97276 (albitiazolium bromide), a choline antagonist [71,127]. It is being positioned in severe malaria because of its poor oral bioavailability. Phase II trials have shown higher doses are likely to be needed, at least for uncomplicated malaria [128]. However, given the size of the clinical trials required to show at least noninferiority regarding mortality compared to artesunate, no new product is likely to be developed for  [59], and Thomson Pharma. Compounds have been defined as 'on hold' when no significant progress along the development process has been made publicly available in the last 12 months. Natural products are defined as Herbal Medicinal Products undergoing testing in malaria patients in GCP quality studies, using standardized extracts. Updates of this figure are available on a quarterly basis [60]).
severe malaria unless artesunate becomes compromised by resistance in many countries.
The remaining medicines for severe malaria are all formulations of artemisinin derivatives. An intramuscular artemether is included in WHO's treatment guidelines and is produced by Kunming and Sanofi, amongst others (Table 3). No prequalified drug is available yet though, and there is a question mark over neurotoxicity as seen in preclinical trial animal species, but never confirmed in humans [129]. A sublingual spray of artemether from Eastland Medical Systems Ltd/Proto Pharma is being developed, but there is the risk that this product is most likely to be used as monotherapy treatment of uncomplicated malaria, in contradiction to WHO guidance. Finally, the UNICEF-UNDP-World Bank-WHO Special Programme for research and training in Tropical Diseases (WHO-TDR) is developing artesunate suppositories as pre-referral treatment for patients with severe malaria, enabling them to receive at least some treatment prior to arrival at hospital. Clinical studies involving 12,068 patients showed that this increased the probability of survival of patients more than six hours away from hospital and for the youngest patients (100 mg dose) [66]. This product still needs to be approved by a Stringent Regulatory Authority and/or WHO prequalification, and will require a partner for manufacture and distribution. Adequate distribution of these suppositories to health centres more than six hours away from hospital will pose a significant challenge.

New products in pregnancy
In areas of high transmission, such as sub-Saharan Africa, malaria in pregnancy is a key cause of maternal, perinatal and neonatal morbidity. Case management of malaria seems possible with ACT in the second and third trimesters. Use of artesunate in the first trimester is contraindicated because of the side effects seen in preclinical safety models [130], although these effects have not yet been seen in the registries of patients who were accidentally treated with ACT during pregnancy [10]. The correct dosing of ACT in pregnancy needs to be better defined, as studies show that current doses may be inadequate [131][132][133]. As well as treatment, medicines may be used for intermittent preventive treatment in pregnancy (IPTp) to reduce infections and improve pregnancy outcomes. Options for therapy include azithromycin-chloroquine, mefloquine and also DHApiperaquine (Table 4). Azithromycin has poor activity against malaria, but synergizes with chloroquine, such that the combination kills even chloroquine-resistant strains [75]. Similar clinical synergy has been seen between azithromycin and quinine [74]. This combination is antibacterial, potentially reducing the risk for newborns and mothers from sexually transmitted bacterial infections [134]. Studies are on-going with another        The other medicine commonly discussed for IPTp is mefloquine, which shows significant clinical benefit [135]. Two complications are side effects and cost. Mefloquine causes nausea and neuropsychiatric disorders amongst Caucasian volunteer cohorts [136], but studies show that side effects are balanced by the positive influence of the drug [37]. On-going efforts sponsored by MMV have optimized mefloquine synthesis, with the potential to lower the cost-per-treatment more than twofold to less than $400/kg. The use of ACT in IPTp is complicated by concerns over the use of artemisinin in pregnancy, and also the concern over using the same medicines for prophylaxis and treatment.

Next generation endoperoxides
The world demand for artemisinin products is large. Treating all malaria cases could require in excess of 200 tonnes of artemisinin per year. The price of artemisinin from plants has fluctuated widely (between $300/kg and $1,500/kg), and material can take up to two years from order to delivery. This creates a problem for low-cost therapy. Three solutions have been proposed: first, the use of higher-yielding seeds [137]; second, using yeast for scalable production in bioreactors [138], where Sanofi plans first commercial tonne-scale production in 2012; and third, to make synthetic endoperoxides. MMV established a project with the University of Nebraska, the Swiss Tropical and Public Health Institute, and Monash University to develop synthetic endoperoxides in 2000. The first clinical product was OZ277 (now called Rbx11160 or arterolane). This showed activity in Phase IIa trials [139] in uncomplicated falciparum malaria. However, the clinical activity was not as good as artemisinin, showing adequate clinical and parasitic response at day 28 (ACPR28) of 60-70% after seven days' treatment, compared with a 95% response with artesunate. The plasma exposure was non-linear above 100 mg, and OZ277/Rbx11160 was unstable in infected blood (presumably due to free ferrous iron) [84]. Despite this, Ranbaxy have completed a Phase III study of OZ277/Rbx11160 (150 mg) and piperaquine (750 mg), and the product has been approved in India. This is the first medicine to be developed in India. The lower efficacy of OZ277/ Rbx11160 compared to artesunate [84] may increase the pressure for resistance against piperaquine, which is already being reported in Cambodia [140]. MMV's nextgeneration clinical candidate, OZ439, has superior pharmacokinetics: exposure is dose-proportional, similar in patients and volunteers, and plasma concentrations remain above the mean parasiticidal concentration for more than 72 hours after a single dose, suggesting it could be part of a single-dose cure [84]. Phase IIa studies in uncomplicated P. vivax and P. falciparum malaria have now been completed. The next stage is to investigate drug interactions in volunteers with potential partner drugs.
Three other synthetic endoperoxides have been in preclinical development (Table 4). RKA 182 is a tetraoxane, from the Liverpool School of Tropical Medicine and the University of Liverpool, with a simpler, more symmetrical molecule that may have a lower cost-per-treatment. Preclinical testing showed it to be superior to OZ277 [107], and further work is ongoing to identify a molecule with pharmacokinetics similar to OZ349. CDRI 99/411 [141] from the Central Drug Research Institute in India was taken into Phase I by IPCA but the project is currently on hold. A trioxalane (a six-membered endoperoxide ring), fused with a 4-aminoquinoline, was produced by Palumed (PA1103/SAR116242) [118], but was abandoned in preclinical development. Finally, the semisynthetic endoperoxide, artemisone, showed good activity in Phase II studies in 2009 [121]. Potential advantages of artemisone include lower dose and potential activity in artemisinin-resistant malaria (given the additional thiomorpholino ring). Initially the compound was developed by Bayer, however the company is no longer involved with this molecule and a new partner is needed before clinical studies can recommence.

Preventing the relapse of Plasmodium vivax malaria
The current gold standard for preventing relapse in P. vivax or P. ovale is primaquine, an 8-aminoquinoline [142]. There are two major issues with its use: 14 days' treatment is required, reducing compliance to close to zero [143], and there is an elevated risk of haemolysis in patients with glucose-6-phosphate dehydrogenase (G6PD) deficiency [142], present in 10-20% of the population in malaria-endemic areas [144].
The only molecule currently in clinical development for preventing relapses is tafenoquine (WR 238605), another 8-aminoquinoline originally developed by the Walter Reed Army Institute of Research. Tafenoquine has a longer half-life in man, with potential as a singledose treatment [144][145][146]. The clinical programme is being partnered by GSK and MMV and has now started recruiting patients for a Phase II dose-finding study. Initial results on both the relative efficacy and safety in G6PDdeficient subjects, compared with chloroquine alone or compared to primaquine, are expected towards the middle of 2013. NPC1161B is a related 8-aminoquinoline from the University of Mississippi [147], active against relapses in primate models. At this stage, it is difficult to find any superiority of NPC1161B over tafenoquine, apart from its effect on insect stages [148], highlighting the need for a preclinical model for G6PD-dependent erythrocyte deformation and haemolysis.

Other products in clinical development
The remaining projects in development fall into several distinct groups: (a) Next-generation aminoquinolines. There are a number of 4-aminoquinolines or amino-alcohols in development. These are positioned as the nextgeneration molecules after the current ACTs. As there are already six artemisinin combination products with some form of regulatory dossier, and all have ACPR28 >95%, this makes the hurdles extremely high for such new molecules. A prime advantage would be to have molecules with better safety profiles, but it is almost impossible to predict this pre-clinically. A lower clinical dose could be an advantage (total doses of 1,920 mg lumefantrine, 1,620 mg amodiaquine or 2,880 mg piperaquine are required over three days for adults [44]), especially if this could be given as a single dose. Phase II studies have been completed on ferroquine (from Sanofi) in combination with artesunate, where total doses as low as 300 mg have shown clinical activity (see Table 4). Naphthoquine (from Kunming) is used in a dose of only 400 mg in the combination with artemisinin, and both of these compounds show some promise. Two other aminoquinolines are on hold: AQ-13 is a modified chloroquine, showing similar exposure in Phase I [147], but which is not sufficiently differentiated. N-tertiary-butyl isoquine is a modified amodiaquine from the University of Liverpool and the Liverpool School of Tropical Medicine [115,116], which has not been progressed since the end of Phase I studies. (b) Antibiotic combinations. In addition to the azithromycin-chloroquine and sulphamethoxazoletrimethoprim combination discussed above in the pregnancy section, two other antibiotic-containing combinations are being explored (see Table 4). The 1-deoxy-D-xylulose 5-phosphate (DOXP) inhibitor, fosmidomycin, has been developed with the lincosamide antibiotic, clindamycin, by Jomaa Pharma GmbH (NCT00217451). Fosmidomycin appears to have a rapid parasite-killing activity in adult patients, but not in children, although a full time-course has not been published, and the dose of 3,600 mg/day may ultimately be problematic [149]. Antibiotic combinations could be considered as fallback therapy in the event that artemisinin resistance becomes a significant clinical issue. Under these circumstances, the standard of care could become seven days quinine plus antibiotic, and therefore the barrier to success would be lower. However, as new classes of molecules enter clinical development from discovery, the relative importance of these antibiotics decreases. The major challenge of effectively managing a portfolio of anti-malarial medicines is to balance the investment in old classes of molecules with that in new molecules with unknown risks and opportunities. (c) New compounds against molecular targets. Two compounds against molecular targets have also moved into development. The enzyme dihydroorotate dehydrogenase (DHODH) is known to be essential for the survival of the parasite. A new inhibitor, DSM1, was identified from highthroughput screening at the University of Texas, and the three-dimensional structure of the enzymeinhibitor complex has been resolved [103]. An improved version, DSM265 has recently entered preclinical development. A next-generation inhibitor of dihydrofolate reductase, P218, entered preclinical development, and work has been focused on allaying fears about the propensity to select for pre-existing mutations in dihydrofolate reductase (DHFR) from the widespread use of pyrimethamine [150]. BCX4945, an inhibitor of purine nucleoside phosphorylase developed for other indications, has been suggested for use in malaria [110]. However, the compound showed no activity in mouse models. (d) New compounds from cellular screens. Despite these success stories, screening 'validated' molecular targets has not been particularly fruitful over the past few years, an experience shared with antibacterial drug discovery [151]. Screening directly against the whole parasite has, therefore, been prioritized. Over the past few years, six million compounds (from over 20 companies and university groups) have been screened against the parasite in the intra-erythrocytic stages. This has resulted in almost 30,000 (0.5%) compounds identified with activity at micromolar levels [152][153][154]. This rate is higher than predicted, and also is higher than that seen when screening molecular targets. The first compound to be developed from the parasitefocused screening is NITD609, a spiroindolone developed by the Novartis Institute for Tropical Diseases in Singapore as part of a collaboration with the Swiss Tropical and Public Health Institute and the Dutch Biomedical Primate Research Center [101]. This is now in Phase IIa clinical studies. The move from screen to Phase IIa within five years is a tremendous achievement for a new class of molecules. Target identification in parallel to lead optimization suggested the P-type cation-transporter ATPase 4 (PfATP4), which interestingly had been characterized, but not prioritized as a target [155]. The second compound from this collaboration is GNF156, an imidazolopiperazine identified by the Genome Foundation of Novartis in San Diego as part of the same collaborative network. It exhibits in vitro potency against blood, liver schizont and gametocyte stages of Plasmodium, (but not the hypnozoites), is orally efficacious in mouse malaria models [156], and has recently entered human volunteer studies. Several other molecules coming from similar screens are entering development. Novel oxaborolecontaining compounds synthesized by Anacor have been identified by screening their library against Plasmodium. Genzyme identified an aminoindole in a project with the Broad Institute in Boston [106], which is particularly interesting since no resistant parasite strains have yet been identified. Merck identified MK4815, also based on cellular assays, which underwent preclinical evaluation with MMV and has not progressed further due to a narrow safety window. Finally, the Swiss biotechnology company Actelion identified an anti-malarial from a parasite screen of a highly focused set of molecules, and this compound is also in Phase I clinical trials. (e) Other compounds. Methylene blue has long been known to have an activity against malaria [99], presumably by changing the redox potential of the infected cell. Phase IIa studies have been carried out with other compounds in combination with chloroquine. The results are interesting, but the combination therapy is less active than ACT [98]. Recent suggestions of activity against the gametocytes may require a re-evaluation of its clinical utility.

Natural products
There is a substantial interest in natural products and the possibility for their role as new anti-malarials. This is not a new departure: three of the mainstays of malaria treatment come from natural products: quinine, lapachol (which led to atovaquone) and artemisinin. The Chinese method dao-xing-ni-shi -'acting in the reversed order' was used by Chen Guofu [157] to investigate herbal extracts used by a population as their anti-malarial treatment. However, many of the extracts known to 'cure' malaria have never been observed in situations where (i) it is clear that the patients had malaria (defined by WHO as fever and parasites); (ii) how much of the extract they consumed; and (iii) whether they had malaria seven or 14 days later. In addition, given that most African adults have some sort of immune protection against the malaria parasite, a placebo rate of around 70% is possible, confounding the studies. The identification of new active natural products in malaria has been disappointing [157]. There have been relatively few new compounds identified by purifying individual natural product molecules from extracts and screening them individually against the parasite. This has led to a call for using molecules in combination [159], or for a return to more observational clinical studies [160].
To date, analysis of the literature has found few natural product extracts in development where the human data are clear. One is an extract of Nauclea pobeguinii, a plant from the Democratic Republic of Congo, which showed a parasite and clinical response of 90.3% at day 14 in 65 patients in Phase II [161]. Although this is below the WHO threshold, it is a much better result than would be seen with either a quinine or artemisinin extract. The second natural product is an extract of Argemone mexicana (Mexican poppy) from Mali, which shows a cure rate of 89% at day 28 in Phase II studies [162]. In both cases, the challenge is to produce standardized extracts for clinical studies. Eventually standardized extracts could become treatments by themselves, similar to the Indian Ayush-64, the Ghanian Phyto-laria, and the standard treatment (Argemone mexicana) in Mali [158]. Moreover, if the active ingredient was identified, this could be developed into a standardized form. There is a strong likelihood that the active ingredient will be a metabolite of the original extract, as these extracts are relatively weak in in vitro assays. Therefore, one priority is to analyse the human metabolism of these interesting extracts.

Significant challenges ahead: changing priorities for the eradication agenda
The announcement of a malaria eradication agenda and its subsequent elaboration by various working groups has helped identify priorities for the future [1]. Beyond that, new medicines would ideally fit several TPPs (for more details please see [3]).
(a) Single-dose cure. Not only would changing from a three-day course of therapy to a single dose decrease the cost of treatment but also it would enable directly observed administration, making it far less easy for patients to take an incomplete course of treatment. Medicines that are safe enough to be given as a single dose are needed, that stay at the site of action for at least two parasite life cycles. These medicines need to be active against all species of malaria, especially drug-resistant strains. In addition, it must be clear that the compounds do not easily select for resistant phenotypes in strains that have not yet emerged. The difficulty of studying artemisinin resistance is the current lack of stable cell assays and molecular markers, which make it almost impossible to characterize drug action. (b) Transmission-blocking. Preventing the parasite from being transmitted back to the mosquito in a blood meal would break the cycle of transmission.
Compounds are being sought that are active against the gametocyte stage, particularly stage five. In the next few years, it is hoped that high-throughput screening will be available to screen larger collections of molecules specifically for this task.
Understanding the activity of compounds against P. vivax gametocytes is further complicated by the lack of stable culture methods, and also because clinically they appear early in the infectiontransmission may have happened before a patient seeks medical attention. (c) Hypnozoite treatment and relapse prevention. For P. vivax and P. ovale, an additional hurdle is to prevent relapse of the hypnozoite or dormant liver form. Attempts are underway to develop biological assays for efficacy, and first successes have been achieved [163]. The availability of an assay means that it is now possible to profile new chemical series to see whether they have activity against both the blood and the liver stages. This has led to some success in hits-to-leads chemistry, which shows activity against hypnozoites. (d) Chemoprevention. The most advanced vaccine against malaria is the RTS,S vaccine (currently planned for launch in 2015 for children, which caused a 50% reduction in incidence of acute falciparum infections in young children in initial Phase III studies [164][165][166]). There remains a significant need for a molecule which can protect the adult population from infection. As malaria becomes less common in Africa, the natural immunity within the population will fall, and the continent's population would be as at risk of malaria infection as western travellers are today. New protective medicines will need long half-lives. Depot formulations are possible, but then the medicines must be potent with a human daily dose of less than 10 mg, and preferably less than 1 mg.
Not all molecules meet all criteria above, as new malaria medicines will be a combination of several active ingredients. Understanding the activity of the compounds against each stage in the Plasmodium life cycle (a Malaria Lifecycle Fingerprint) is important as early as possible in the development process to allow a focused development strategy and prioritization between molecules.
In addition, three factors are critical for the development of new medicines. First, safety and tolerability; since these medicines will be used in areas where pharmacovigilance systems are not always completely effective. Second, an appropriate speed of action; for blood stage medicines, the patients have become used to the speed of action of artemisinin. Third, medicines need to be potent; many of the current medicines are active at adult doses approaching 1 g per day. If the next-generation of medicines are more potent, this can greatly impact the cost of treatment. Cost reductions will be much more likely to come from doses that are lower than current therapy, rather than from simpler, cheaper molecules.

Conclusions
The pipeline of new molecules targeting malaria is much richer now than it was two years ago [167]. Two significant events have occurred: first, new medicines have moved down the pipeline, specifically the prequalification of artesunate for severe malaria and the stringent regulatory approval of DHA-piperaquine and pyronaridineartesunate as new fixed-dose anti-malarials. Second, and arguably more interesting, is the number of new molecules and new classes of molecules that are entering the pipeline. There are at least seven new compound families that have been discovered in the past five years. This is a rich portfolio and reflects the commitment of the field as a whole. Nonetheless, from a portfolio point of view there are still gaps. The chance that a new molecule entering Phase I studies will make it to registration is still around 20% for anti-infectives, and for a new combination two molecules would be needed. In addition, new TPPs have emerged from the malaria eradication agenda and compounds need to be measured against these, focusing on transmission-blocking and anti-relapse potential. With the sustained commitment of all of partnersdonors, scientists, academics and the pharmaceutical industrythe next decade will be a very exciting time to be in clinical development of new therapeutics for malaria.

Competing interests
The authors declare that they have no competing interests, beyond the fact that MMV is involved in supporting the development of some of these medicines. This paper was designed as an objective review of the entire worldwide portfolio of medicines in development, and the authors apologise in advance to all colleagues where space constraints prevented a fuller description of the molecules.