For reprint orders, please contact:
[email protected]
Future Microbiology
Review
Drug resistance in human African trypanosomiasis Michael P Barrett†1, Isabel M Vincent1, Richard JS Burchmore1, Anne JN Kazibwe2 & Enock Matovu2 Wellcome Trust Centre for Molecular Parasitology, Institute of Infection, Immunity & Inflammation, College of Medical Veterinary and Life Sciences, University of Glasgow, Glasgow, G12 8QQ, Scotland 2 Makerere University School of Veterinary Medicine, Kampala, Uganda † Author for correspondence: Tel.: +44 141 330 6904 n Fax: +44 141 330 4600 n
[email protected] 1
Human African trypanosomiasis or ‘sleeping sickness’ is a neglected tropical disease caused by the parasite Trypanosoma brucei. A decade of intense international cooperation has brought the incidence to fewer than 10,000 reported cases per annum with anti-trypanosomal drugs, particularly against stage 2 disease where the CNS is involved, being central to control. Treatment failures with melarsoprol started to appear in the 1990s and their incidence has risen sharply in many foci. Loss of plasma membrane transporters involved in drug uptake, particularly the P2 aminopurine transporter and also a transporter termed the high affinity pentamidine transporter, relate to melarsoprol resistance selected in the laboratory. The same two transporters are also responsible for the uptake of the stage 1 drug pentamidine and, to varying extents, other diamidines. However, reports of treatment failures with pentamidine have been rare from the field. Eflornithine (difluoromethylornithine) has replaced melarsoprol as first-line treatment in many regions. However, a need for protracted and complicated drug dosing regimens slowed widespread implementation of eflornithine monotherapy. A combination of eflornithine with nifurtimox substantially decreases the required dose and duration of eflornithine administration and this nifurtimox-eflornithine combination therapy has enjoyed rapid implementation. Unfortunately, selection of resistance to eflornithine in the laboratory is relatively easy (through loss of an amino acid transporter believed to be involved in its uptake), as is selection of resistance to nifurtimox. The first anecdotal reports of treatment failures with eflornithine monotherapy are emerging from some foci. The possibility that parasites resistant to melarsoprol on the one hand, and eflornithine on the other, are present in the field indicates that genes capable of conferring drug resistance to both drugs are in circulation. If new drugs, that act in ways that will not render them susceptible to resistance mechanisms already in circulation do not appear soon, there is also a risk that the current downward trend in Human African trypanosomiasis prevalence will be reversed and, as has happened in the past, the disease will become resurgent, only this time in a form that resists available drugs.
Human African trypanosomiasis (HAT), also known as sleeping sickness in its second stage when the central nervous system is involved, is caused by subspecies of the protozoan parasite Trypanosoma brucei spp. [1] . Many of the basic parameters of the disease have been reviewed recently in Future Microbiology along with recent advances in development of novel agents to treat the disease [2] . There are two forms of HAT caused by two separate subspecies. T. b. gambiense prevalent in Central and West Africa causes a chronic form classically taking around 18 months to progress through the first (hemolymphatic) stage to the second (neurological) stage and then another 18 months to death. In Eastern and Southern Africa, T. b. rhodesiense causes an acute disease with shortened first and second stage, leading to death within weeks to 10.2217/FMB.11.88 © 2011 Future Medicine Ltd
months of infection. Recent evidence, however, indicates that avirulent T. b. gambiense and less acutely virulent forms of T. b. rhodesiense might also exist [3] . Vaccines will not be effective against HAT since the parasites have the ability to sequentially change a surface glycoprotein coat that covers the entire surface. Although attempts to destroy the tsetse fly vectors that transmit the disease have been used effectively in control programs, their continent-wide application is logistically challenging [4] . As a consequence, chemotherapy has been central to efforts to manage HAT and has also been central to efforts to move to a situation of disease elimination [5] . However, relatively few drugs are available and they have been used for between 20 and 80 years. The development of resistance to the Future Microbiol. (2011) 6(9), 1037–1047
Keywords drug resistance n eflornithine transporter n HAPT1 n human African trypanosomiasis n nitroreductase n P2 transporter n sleeping sickness n treatment failure
n
part of
ISSN 1746-0913
1037
Review
Barrett, Vincent, Burchmore, Kazibwe & Matovu
current drugs would be calamitous in terms of sustained management of the disease and would reverse the current downward trend in prevalence. In this review we discuss mechanisms of resistance to currently used drugs as determined in the laboratory and also discuss the potential emergence of resistance in the field. Treatment failure & drug resistance in human African trypanosomiasis
Like all microbes, trypanosomes have the ability to evolve resistance, through genetic mutations, to drugs used in their treatment (Figure 1) . However, it has proven difficult to link instances of parasite resistance to treatment failure in the field. The difficulties associated with harvesting of parasites, especially T. b. gambiense, from infected patients, to allow studies into parasite genotype and phenotype, exacerbate this problem. A recent report that patient cerebrospinal fluid (CSF) from HAT patients yielded parasites capable of robust in vitro and in vivo (rodent)
growth if they were first cultivated over fibroblast feeder layers [6] might offer improvements in our ability to study field isolates. A number of other important issues need to be considered with regard to treatment failure that do not necessarily relate to parasite drug resistance. Difficulties in drug administration
The distribution of HAT drugs today is controlled directly by the WHO. In response to requests from endemic country National Programs, or other health workers involved in HAT clinics, drugs are shipped according to a well-regulated schedule with no involvement of the private sector. Problems involving inferior quality or counterfeit drugs (as has been shown to be of significance for malaria [7]) is not an issue for HAT, although it has been a problem in veterinary trypanosomiasis [8] . However, treatments for HAT are notoriously difficult to administer (reviewed in [2,9]), with all current registered drugs requiring protracted parenteral administration (oral nifurtimox being available
M-TSH PM P
M
N
TbAT1
HAPT1 TbAAT6
LAPT1
ODC
M-TSH
P
E
TbMRPA
P
N*
NTR
TSH
Orn
Put
Spd
GSH-Spd GSH
K
S Future Microbiol. © Future Science Group (2011)
Figure 1. The known mechanisms of drug resistance in Trypanosoma brucei. Routes of uptake for each of the currently used trypanocidal drugs are shown. Also, where known, cellular targets are marked, including ODC for eflornithine and the NTR that metabolizes nifurtimox into its active form (N*). Pentamidine binds to kinetoplast DNA (K), although it is not clear that this binding is responsible for its mode of action. Resistance to eflornithine relates to loss of TbAAT6. Resistance to melarsoprol (or its active metabolite melarsen oxide) relates to loss of the TbAT1 (P2) and HAPT1 transporters. Alternatively, upregulation of TbMRPA and can cause resistance when melarsen oxide-trypanothione conjugates are pumped from the cell. Pentamidine resistance also relates to loss of TbAT1 and HAPT1 transporters. Suramin enters by receptor-mediated endocytosis at the flagellar pocket and resistance has been shown to relate to changes in the endocytic pathway. Nifurtimox resistance can come about when the nitroreductase activity involved in its activation is diminished. E: Eflornithine; M: Melarsoprol; N: Nifurtimox; NTR: Nitroreductase; ODC: Ornithine decarboxylase; P: Pentamidine; S: Suramin.
1038
Future Microbiol. (2011) 6(9)
future science group
Drug resistance in human African trypanosomiasis
through the WHO list of essential medicines without being licensed). For stage 1 disease pentamidine is given, usually by intramuscular injection, with once-daily 4 mg kg-1 dosing for 7 days. Suramin is usually given by slow intravenous injection every 3–7 days, to a total of five dosings over a 4-week period. For stage 2 disease, melarsoprol is administered intravenously in a 3.6% solution in propylene glycol, typically every day for 10 days (although earlier schedules involved inclusion of interrupted courses lasting a month). Eflornithine as monotherapy, is given over 14 days with four 100 mg kg-1 intravenous infusions per day. When used in the nifurtimox-eflornithine combination therapy (which is occurring with increasing frequency) eflornithine doses are halved (200 mg/kg twice per day for 7 days) with nifurtimox being given orally three times a day for 10 days. The protocols required to sustain these recommended administration parameters are complex and can inevitably be compromised in some instances, leading to treatment failure (and possibly even promoting selection of resistance). Pharmacokinetics, pharmacodynamics & difficulties in disease staging
The general efficacy of the current anti-trypanosomal drugs is variable within populations. This can be due to trypanocidal potency of the drugs themselves or due to variability in pharmacokinetic behavior in different people. Pentamidine is highly potent, yielding IC50 values in the order of 1–10 nM in a typical 3 day in vitro drug sensitivity assay. Melarsoprol and suramin also show pronounced activity in these assays killing trypanosomes in the low nanomolar range. By contrast, nifurtimox has an IC50 of around 5 µM in these assays and eflornithine only acts in the tens of µM range [2,9] . The time that parasites must be exposed to drug in order for them to be certain to die is also critical, hence the drug-clearance rate in vivo is of great importance. Pharmacokinetic properties of the antitrypanosomal drugs are, therefore, also highly relevant to their activity. Ultimately, in order to be active, parasites must be exposed to drug at lethal concentrations for sufficient time to induce a trypanocidal effect. In stage 2 HAT, when drugs must reach parasites in the CNS and other privileged body compartments, pharmacokinetics become even more important. Since different drugs have different abilities to distribute to these compartments, correct diagnosis with respect to whether the disease is at stage 1 or at stage 2 is critical, but not straightforward. future science group
Review
Pentamidine displays extensive tissue retention and binds serum proteins, which contributes to a large volume of distribution and long terminal half-life. However, the drug is extensively metabolized (rates of metabolism can vary between individuals) and it crosses the blood–brain barrier at low levels [10] . For suramin, nearly all of the injected drug (>99%) is protein-bound in plasma, which gives it a long terminal half-life (41–78 days) with free drug being available in sufficient quantities to affect parasites due to equilibria in its protein binding. The charged nature of suramin precludes delivery beyond the blood–brain barrier [11] . Melarsoprol is metabolized rapidly to melarsen oxide, which is the active molecule in vivo. Clearance of the active metabolite is also relatively fast (a half-life of 3.5 h). Accumulation across the blood–brain barrier appears to be relatively modest (only 1–2% of maximum plasma levels in CSF [9]). Within the brain, therefore, the drug accumulates to concentrations relatively close to the minimum needed to exert activity. Thus minor shifts in the susceptibility of parasites to the drug can affect efficacy profoundly, and this could have implications in treatment failure where a relatively low drop in sensitivity to drug could render parasites in CSF nonsusceptible to the levels accumulating in this compartment. Eflornithine’s permeation into rodent brain has recently been reported to be low [12] . CSF to plasma ratios between 0.1 and 0.9 have been reported in humans [9] , indicating possible variability in brain permeation among humans and differences between its ability to enter rodent versus human brains. The mean half-life in plasma following intravenous injection of eflornithine is only in the order of 3 h. In recent years, studies have indicated that the use of white cell count in CSF can act as a surrogate marker for stage 2 infection, and most national programs use this in staging the disease, given the difficulty associated with finding trypanosomes in CSF [1] . In some instances, 5 white cells per µl of CSF is taken as the cut-off point, while in others it is 20. Given the severe toxicity associated with melarsoprol, recommendations were made to use pentamidine to treat the ‘early–late’ stage of the disease in patients without trypanosomes in CSF and with fewer than 20 white cells per µl. However, Lejon et al. [13] showed that in cases of >10 white cells per µl there was an increased risk of treatment failure with pentamidine, and Balasagram et al. [14] showed that patients in which white cell counts were
