Although the treatments for human African trypanosomiasis (HAT), leishmaniasis and Chagas disease (CD) still rely on drugs developed several decades ago, there has been significant progress in the identification, development and use of novel drugs and formulations. Notably, there are now two drugs in clinical trial for HAT, fexinidazole and acoziborole; the liposomal amphotericin B formulation AmBisome has become an essential tool for both treatment and control of visceral leishmaniasis; and antifungal triazoles, posoconazole and ravuconazole, together with fexinidazole, have reached clinical trials for CD. Several other novel and diverse candidates are moving through the pipeline; sustained funding for their clinical development will now be the key to bring new safe, oral, shorter-course treatments to the clinic.
Treatments for the three human diseases such as human African trypanosomiasis (HAT), leishmaniasis and Chagas disease (CD), caused, respectively, by trypanosomatid parasites Trypanosoma brucei spp., Leishmania spp. and Trypanosoma cruzi, have, for decades, been based on drugs with the limitations of variable efficacy, toxicity, high cost and long treatment courses. The causative Trypanosoma and Leishmania pathogens share many unique cellular, biochemical and molecular characteristics, from the kinetoplast (a large catenated mass of DNA in the mitochondrion) and the glycosome (a single membrane-bound organelle enclosing the glycolytic pathway), to unique metabolic pathways, for example thiol metabolism. Despite these unique characteristics and genomes known since 2004, it has been a challenge to identify novel compounds with the potential to be developed into drugs, a challenge accentuated by the nature of the infection and disease pathology — for example, in HAT, the key pathogen is in the central nervous system (CNS), behind the blood–brain barrier; in Chagas, the pathogen is in many tissues possibly in a dormant phase; and in leishmaniasis, there is a diversity of species with infection of different tissues and differences in disease manifestation .
However, significant advances over the past decade have propelled drug R&D for these diseases (See Table 1 for the current R&D pipeline and Figure 1 for examples of new chemical entities in development for these diseases). Firstly, the awareness and advocacy around Neglected Tropical Diseases has brought additional funding, government and international organization commitment and the establishment of dedicated organizations like the Drugs for Neglected Diseases initiative (DNDi). Secondly, advances in in vitro assay methodology enabling screening of large compound libraries, genetically modified parasites for in vitro and in vivo imaging and the integration of pharmacokinetics have greatly improved the models available for drug discovery and development. Thirdly, the re-engagement of the pharma/biotech sector has brought access to large compound libraries (to the point where several million compounds have now been screened against these parasites), the expertize and focus needed to steer the development of a novel compound to a drug and a link to new partnerships for treatment and control.
Human African trypanosomiasis
The past decade has seen extraordinary progress against HAT. The incidence of the disease has declined steadily to a point where just over 2000 diagnosed cases were reported in 2016 (www.who.int/trypanosomiasis_african/en/). The disease, of which there are two forms, caused by Trypanosoma brucei rhodesiense (an acute form of the disease endemic in Eastern and Southern Africa) and T. b. gambiense (endemic in west and central Africa, representing over 95% of all reported cases today), is classically considered in two stages. The first stage involving infection of blood and lymph precedes a second stage involving parasite invasion of the CNS. Typically, the rhodesiense form kills within months of infection, while the gambiense infection takes several years to reach its fatal climax. The second stage, where drugs must cross the blood–brain barrier in sufficient quantities to kill parasites resident there, has offered substantial challenges to developing new molecules to cure the disease. The increasing number of HAT cases in the latter part of the 20th century triggered a series of key interventions. An improved 10-day regimen for melarsoprol, based on pharmacokinetic data , for example, heralded a new attitude to the development and administration of drugs. This was followed by large-scale, albeit ultimately unsuccessful, trials for an orally available drug for stage 1 disease (pafuramidine) [3,4] Pafuramidine is an orally available methoxy prodrug of the potently trypanocidal diamidine, furamidine. The potency of the diamidines has been clear for many years and another diamidine, pentamidine, remains the drug of choice for stage 1 gambiense disease. The prospect of replacing pentamidine with an orally available analog was of considerable interest; hence, the discovery of unanticipated renal toxicity was a disappointment.
In another advance, the World Health Organization secured a gratis supply of eflornithine and this facilitated a shift away from toxic melarsoprol to the introduction of a nifurtimox–eflornithine combination therapy . DNDi, founded in 2003, had HAT as a primary target disease. Since then, DNDi has taken the nitroimidazole drug, fexinidazole, through to phase 3 clinical trials , to be registered for the treatment of stage 2 HAT in 2018. Fexinidazole is orally bioavailable and although the drug is rapidly metabolized to sulfone and sulfoxide derivatives, these are also trypanocidal. Moreover, it was shown that fexinidazole and its metabolites can penetrate the CNS with activity in stage 2 mouse models, with any toxic side effects falling within acceptable limits . The likely dosing will remain challenging, requiring a 10-day administration with added foodstuffs to assure oral bioavailability.
Screens for new drugs against T. brucei are relatively straightforward with high-throughput in vitro screening against bloodstream culture forms offering a means of identifying early hits. Rodent models for stage 1 and stage 2 (CNS involvement) diseases have also been further developed to determine novel compound activity. The classical stage 2 efficacy model, which required 6 months of post-treatment follow-up, has been superseded with an in vivo imaging approach that offers a reliable readout of efficacy in just one-third of the time .
An orally available drug for stage 2 HAT has emerged from a development program under DNDi leadership in partnership with two biotechs, SCYNEXIS and Anacor. SAR around the benzoxaborole class generated compounds with activity against the trypanosomes  and good brain accumulation, leading ultimately to the development of acoziborole (SCYX7158) . Efficacy, PK and toxicity data in rodents all indicated excellent activity with the prospect of single oral dose administration. The compound has proved safe in phase 1 toxicity testing and phase 2 trials to treat stage 2 disease are underway.
Elsewhere, at GSK Tres Cantos, a series of leads against bloodstream-form trypanosomes as well as Leishmania donovani and T. cruzi have been identified , and the Novartis Institute for Tropical Diseases (NITD) has also engaged in a successful program of screening a large compound library for trypanocidal leads. This work was initiated with a screen of 3 million compounds by the Genomics Institute of the Novartis Research Foundation (GNF) against all three kinetoplastid pathogens. Hits were pursued by the NITD and some external researchers. The biological relatedness between T. brucei, T. cruzi and Leishmania was exploited in seeking the mode of action of a key compound, derived from an initial azabenzoxazole hit (GNF5343) . A series of 3000 derivatives, seeking activity in the macrophage model of Leishmania infection, was pursued and to seek MOA, resistance to one compound, GNF6702, was selected in T. cruzi. Whole-genome sequencing revealed mutations to the β4 subunit of the proteasome, and expressing the mutated T. cruzi gene in T. brucei-induced resistance, pointing to this protein as the drug's target in both species. The compound cured a stage 2 mouse model using 100 mg/kg once per day by oral gavage for 7 days.
The Novartis screen also revealed a diverse series of scaffolds which offered hits for development to leads . In addition to the proteasome inhibitors, for example, a series of 2-aryl-benzothiazol amines showed good potency and a urea derivative, [(S)-2-(3,4-difluorophenyl)-5-(3-fluoro-N-pyrrolidylamido)benzothiazole], with improved metabolic stability, and in vivo oral dosing could cure mice of both stage 1 and stage 2 infections (20 doses of 50 mg/kg given every 12 h for 10 days cured the stage 2 model) . Other scaffolds have been pursued focusing on potency alongside PK and metabolic stability characteristics, without yet yielding stage 2 cures . Screening of validated drug targets too has enabled the discovery of new chemicals with good trypanocidal activity, although, to date, it has proved more difficult to develop those hits into molecules with PK parameters enabling sufficient brain penetration to cure models of stage 2 disease. Campaigns around N-myristoyltransferase, S-adenosylmethionine decarboxylase and methionine tRNA synthetase are all examples of targets against which trypanocidal compounds have been developed and can cure stage 1 models of disease, but where further work is required to obtain a cure once the CNS is infected [16–18].
There are two main clinical manifestations of leishmaniasis — visceral leishmaniasis (VL) which is potentially fatal and endemic in the Indian subcontinent, East Africa and Brazil, and cutaneous leishmaniasis (CL), a disfiguring disease which normally self-cures within 3–18 months, distributed mainly throughout central and south America, the middle east, and north and east Africa. Estimates of the incidence of VL and CL are 0.5 and 2–3 million, respectively . Currently, there are many drugs and treatments for leishmaniasis, all with limitations. VL drugs that are effective in the Indian subcontinent, the liposomal formulation of amphotericin B (AmBisome), miltefosine and paromomycin, have proved to be significantly less effective against VL in East Africa, whereas the reverse is true for pentavalent antimonials (e.g. Pentostam, SSG) due to drug resistance in India. For HIV VL co-infections and post kala-azar dermal leishmaniasis (PKDL), there are inadequate treatments. For CL, options are even more limited (antimonials remain predominant) and R&D efforts have been minimal. For CL, there have been two other notable approaches with use of topical formulations, for example with paromomycin , as well as use of immunomodulators (to accelerate self-cure) as adjunct therapy (for example, imiquimod and pentoxyfylline) .
Three points about the biology of the Leishmania parasite are germane to antileishmanial drug R&D — (a) the drug must be able to access and accumulate in the phagolysosomal compartment of the macrophage where the Leishmania amastigote survives and multiplies; (b) the two major forms, VL and CL, will probably require drugs with different pharmacokinetic properties and pharmaceutical formulations; and (c) there are 20 species of Leishmania that cause VL and CL with some significant differences in drug sensitivity. These must be considered as we drive for safe, oral, short-course drugs (10 days) that are effective for the treatment of VL, in all endemic regions, and are also effective for the treatment of CL (see target product profiles that detail the requirements of treatments at www.dndi.org/diseases-projects/target-product-profiles).
VL drug discovery has benefited from both rational and empirical approaches with advances in molecular biology and structural biology leading to the elaboration of validated targets and inhibitors and high-throughput and high-content screening of large libraries, leading to the identification of novel chemical series as well as the identification of novel targets, for example, the proteasome (ibid), and a lead compound GNF6702 [1,12,13,22]. Programs that have integrated screening and medicinal chemistry have produced lead compounds from novel chemical entities, such as the oxaboroles, nitroimidazoles and aminopyrazoles, and preclinical candidates in the development pipeline (www.dndi.org). Collaborations between the biotech/pharma sector and academia and public–private partnerships, for example between GSK and Dundee University, have led to promising drug candidates. Other research in academia has identified novel natural products and the potential of drug repurposing, with examples being the antidepressant imipramine, antimalarial tafenoquine, anti-TB drug delamanid and anti-HAT drug fexinidazole ; a subsequent clinical trial with fexinidazole for VL was not successful, emphasizing the need for ‘purpose-designed’ drugs.
For CL, there has always been the hope that the oral drugs developed for VL will also offer potential for treatment. Recent studies in a rodent model with oxaboroles, nitroimidazoles and aminopyrazoles (www.dndi.org) are substantiating this hope. Other studies in rodent models have shown how infection has an impact on drug skin permeation (important for topicals)  and the accumulation of drug formulations , also emphasizing the importance of using disease models in building PK–PD analysis. Liposomal amphotericin B is a proven treatment for a complex form of CL and mucocutaneous leishmaniasis. The first guidelines for CL clinical trials have recently been established ; it is now time to put them into practice with novel drugs.
Economic and social trends in the last few decades have had a major impact on CD epidemiology . Because of migration, CD patients are increasingly found in non-endemic countries, making CD a global public health problem not limited just to Latin America [28–30]. The success of vector control programs has led to a switch of the disease transmission mode, with congenital transmission making up a larger proportion of the total case transmission than was previously the case [31,32]. A range of incentives to promote both the involvement of various stakeholders in R&D (e.g. FDA Priority Review Voucher) and access to treatment (e.g. Chagas Coalition) [33–37] have emerged. Current trypanocidal treatment options are limited to two old nitroheterocyclic drugs, benznidazole and nifurtimox, both of which cause substantial side effects. Following the failure of azoles, posaconazole and fos-revuconazole, at the doses/exposure and duration tested in clinical trials for CD, current research is aimed at moving forward drugs with new mechanisms of action while assessing different regimens of currently available drugs [38–40]. The CD R&D landscape has changed dramatically over the past 10 years . New tools and collaborative approaches have revolutionized the field, and there is now a new approach to identifying novel and improved drugs against T. cruzi  although many challenges remain .
A major hurdle in the development of new drugs for CD is the absence of a test that can assess treatment efficacy in a timely fashion. Demonstrating efficacy relies on serological tests that show disappearance of T. cruzi antibodies (seroconversion). Although relatively quick in children, seroconversion can take decades to occur in adults. Unfortunately, surrogate markers of cure have not yet been identified [44,45]. The FDA recently approved benznidazole for use, but only in children aged 2–12 years , due to the inability to assess efficacy in adults, emphasizing the need for surrogate biomarker(s) of cure, particularly in adults.
Defining cure in CD patients remains controversial as parasitological cure has not yet been shown to be essential in preventing progression of the disease; some investigators argue that drugs which reduce parasite load can also be considered for further development . Debate is still ongoing regarding tropism of the parasite in the human body during the chronic disease phase. Potential reservoir sites and silent non-replicating T. cruzi have been proposed as strategies allowing parasites to evade host immune surveillance. T. cruzi strain variation across the geographical range of CD also influences response to different therapies . Although there is a widespread belief that parasite persistence is needed for symptomatic disease to occur, there is an indication, at least in preclinical murine models, that transient organ infection could be sufficient to induce CD symptoms . Translating findings from rodent models of disease, where pharmacokinetic and pharmacodynamics criteria, as well as parasite distribution to different organs, may be at variance to the situation in humans, remains a major impediment in CD drug development .
New compounds entering clinical trials for CD are few and far between. The original phase II proof-of-concept clinical trial of fexinidazole was interrupted due to safety and tolerability issues . This compound is still being pursued in spite of its being activated by the same parasite nitroreductase responsible for activation of both benznidazole and nifurtimox, since it may be active within a reduced administration regimen and thus provide better compliance .
There is a clear knowledge gap and the need for continued research in order to better understand CD and the determinants for its progression . New potential markers of cure and new targets or pathways inhibited by lead compounds have been identified recently using techniques described earlier [14,54,55]. Although phenotypic screening has identified several new leads for development against HAT, the model systems for CD have been less readily translated, and revisiting target-based screening may accelerate identification of new chemical space and compounds of interest with anti-T. cruzi activity.
A better understanding of the host/parasite interaction in CD through more focused research in a collaborative environment with sustainable funding is the only way forward for developing the best suited drugs for this disease.
Conclusion — where next?
The successful identification of several series of novel selective small molecules by screening several million compounds, mainly from libraries created by industry, against T. brucei, L. donovani and T. cruzi, has demonstrated the utility of this paradigm. A better understanding of the impact of disease pathology on PK–PD relationships, in both animal models of infection and humans, should now be a focus to shape the bridges between discovery to preclinical development and preclinical to clinical development . For all these diseases, the drug R&D process cannot be separated from improvement in biomarkers/diagnostics where clarity around diagnostics for CNS infection in HAT, and disease progression for Chagas (indeterminate to chronic) and VL (asymptomatic to symptomatic for the elimination program in the Indian subcontinent) is essential. With vaccines still a distant hope, the essential elements of rapid diagnosis and rapid treatment will be a mainstay of disease control in the next decade.
Current drugs and treatments for human African trypanosomiasis (HAT), leishmaniasis and Chagas disease (CD) have significant limitations.
Over the past decade a major effort, including academia, industry and public–private partnerships, to screen compound libraries and to improve assays and models has resulted in the identification of novel lead compounds.
Several new drugs, both novel and repurposed compounds, have reached clinical trials for HAT and CD, whereas for leishmaniasis a liposomal formulation of an established drug has led to a transformation of treatment.
A new treatment for HAT, fexinidazole, will be registered in 2018 and several novel compounds for all three diseases will be progressed though preclinical development.
Structures of several compounds in development for the three diseases.
|Preclinical development||Clinical development|
|Preclinical development||Clinical development|
central nervous system
Drugs for Neglected Diseases initiative
Genomics Institute of the Novartis Research Foundation
human African trypanosomiasis
mode of action
Novartis Institute for Tropical Diseases
post kala-azar dermal leishmaniasis
structure activity relationships
M.P.B. is part of the Wellcome Centre for Molecular Parasitology funded by a core grant from the Wellcome Trust [104111/Z/14/Z].
Views in the present study represent those of the authors but not necessarily those of their respective institutions.
Simon Croft is a member of the Scientific Advisory Committee of DNDi and has served as a consultant for discovery and development of drugs for these diseases for Novartis and GSK. Michael Barrett has collaborative research projects with GSK and Novartis. Eric Chatelain has no competing interests associated with this article.