- Email: [email protected]
Clinical atovaquone-proguanil resistance of Plasmodium falciparum associated with cytochrome b codon 268 mutations
Microbes and Infection 8 (2006) 2599e2604 www.elsevier.com/locate/micinf
Original article
Clinical atovaquone-proguanil resistance of Plasmodium falciparum associated with cytochrome b codon 268 mutations Lise Musset a,b,*, Olivier Bouchaud c, Sophie Matheron d, Laurent Massias e, Jacques Le Bras a,b a Laboratoire de biologie animale et parasitaire, EA209, Universite´ Paris Descartes, 4 av de l’observatoire, 75006 Paris, France Centre National de Re´fe´rence Paludisme, APHP, Hoˆpital Bichat-Claude Bernard, 46 rue Henri Huchard, 75877 Paris Cedex 18, France c Unite´ de maladies infectieuses et tropicales, APHP, Hoˆpital Avicenne, 125 rue de Stalingrad, 93000 Bobigny, France d Unite´ de maladies infectieuses et tropicales, APHP, Hoˆpital Bichat-Claude Bernard, 46 rue Henri Huchard, 75877 Paris Cedex 18, France e Laboratoire de toxicologie, APHP, Hoˆpital Bichat-Claude Bernard, 46 rue Henri Huchard, 75877 Paris Cedex 18, France b
Received 7 July 2006; accepted 12 July 2006 Available online 10 August 2006
Abstract Plasmodium falciparum resistance to atovaquone-proguanil has so far been associated with Y268S or Y268N mutations in cytochrome b, although these changes were identified in only seven of the 11 treatment failures. Here, we describe 10 new cases of atovaquone-proguanil treatment failures among which the parasite resistance was confirmed in six cases, either by identifying correct plasma drug concentrations or by observing in vitro atovaquone resistance. Resistance was consistently associated with codon 268 mutations (Y268S or a previously unidentified mutation, Y268C). Notably, mutations were not detected before the treatment but only after the drug exposure. Ó 2006 Elsevier Masson SAS. All rights reserved. Keywords: Plasmodium falciparum; Chemotherapy; Resistance; Cytochrome b; Atovaquone; Proguanil; Malarone
1. Introduction In 1996, the atovaquone-proguanil (AP) combination was registered in North America and Europe for the prophylaxis and treatment of malaria. While this safe and efficient combination is increasingly used in developed countries, its high cost precludes a large use in endemic countries. Atovaquone, a ubiquinone analogue binding to cytochrome b of plasmodial mitochondria, inhibits electron transfer of the respiratory chain
Abbreviations: AP, atovaquone-proguanil; pfcytb, cytochrome b gene; pfdhfr, dihydrofolate reductase gene; IC50, inhibitory concentration 50%; C, cysteine; pfmsp, merozoite surface protein gene; N, asparagine; PCR, polymerase chain reaction; S, serine; Y, tyrosine; WHO, World Health Organization. * Corresponding author. Laboratoire de Parasitologie-Mycologie, Hoˆpital Bichat-Claude Bernard, 46 rue Henri Huchard, 75877 Paris Cedex 18, France. Tel.: þ33 140 257 899; fax: þ33 140 256 763. E-mail address: [email protected] (L. Musset). 1286-4579/$ - see front matter Ó 2006 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.micinf.2006.07.011
[1]. Proguanil is mainly considered as a pro-drug of cycloguanil, an inhibitor of the dihydrofolate reductase also involved in pyrimidine biosynthesis. However, in combination with atovaquone, proguanil, by itself, lowers the effective concentration at which the former collapses the mitochondrial membrane potential [2]. The proguanil target and the detailed mechanism implicated in AP synergy remain unknown. Since the introduction of AP combination, 11 cases of treatment failures have been published in travelers returning from Africa [3e7]. Seven failures exhibited a modification of codon 268 of the cytochrome b gene ( pfcytb268), mostly from tyrosine (Y) to serine (S) and four failures were reported without any pfcytb mutation. Thus, the usefulness of pfcytb268 mutations for predicting Plasmodium falciparum AP resistance has been questioned [8]. The limitation for understanding the failure mechanisms may also have been related to an insufficient description of the cases, impairing discrimination between parasite resistance and poor drug absorption. Atovaquone is a very
2600
L. Musset et al. / Microbes and Infection 8 (2006) 2599e2604
lipophilic drug and its low bioavailability is 4-fold increased (reaching 23%) with simultaneous intake of fatty food. This poor absorption of atovaquone may induce treatment failures without parasite resistance. With this in mind, this study was designed to investigate putative causes of 10 additional AP treatment failures identified over a 3-year period.
2. Materials and methods
2.2.3. Pfdhfr genotyping The three major pfdhfr mutations (at positions 51, 59, and 108) associated with cycloguanil resistance were studied with a restriction method. 2.2.4. Parasite population analysis Parasite diversity within isolates was determined by a fragment-analysis method of pfmsp1 and pfmsp2 polymorphisms. This method allows for the detection of all genotypes accounting for more than 2% of the whole parasite population [12].
2.1. Patients Between January 2003 and September 2005, six cases of AP treatment failures were identified among 298 patients suffering from uncomplicated P. falciparum malaria and treated with AP (four tablets daily for 3 days). Most patients, being 9e75 years of age, had returned from Central or West African countries. Eighty percent were living in France but were native from the country of infection. In our hospital, a voluntary monitoring of malaria treatment efficacy is systematically proposed. Based on the WHO standard protocol, follow-up plans clinical and parasitological examinations between Day 0 (i.e. first day of treatment) and Day 28 (generally, Day 3, Day 7 and Day 28). Early and late treatment failures were defined following WHO criteria [9]. Informed consent was not required for this study as the following procedures are part of the French national recommendations for the care and surveillance of malaria [10]. As our laboratory is the national reference centre for malaria, we were informed of four additional cases of late AP treatment failure. In agreement with national consensus, all patients with failure were re-treated with quinine.
2.2.5. Drug measurements Determinations of drug concentrations in plasma were performed by reverse phase high performance liquid chromatography. The lower limits of quantification were 1.4 mM, 0.03 mM and 0.04 mM for atovaquone, proguanil and cycloguanil, respectively. Results were compared to effective atovaquone plasma concentrations in malaria treatment: between 3 and 20 mM, 0.6 and 18 mM, 0.3 and 2.2 mM on Day 3, Day 8 and Day 21, respectively [13]. 2.3. Classification of treatment failures To define a reliable allele of AP resistance, treatment failures were classified into three categories: (i) Failures in absence of AP resistance: incorrect plasma drug dosages associated with Day-of-failure parasites in vitro susceptible to atovaquone, (ii) Failures caused by AP resistance: correct drug dosages or Day-of-failure parasites resistant to atovaquone, and (iii) Indeterminate. 3. Results
2.2. Investigation of treatment failures For all treatment failure patients, atovaquone in vitro susceptibility testing, pfcytb and pfdhfr genotypings, and merozoite surface proteins 1 and 2 ( pfmsp1 and pfmsp2) polymorphism analyses were performed on Day-0 and Day-of-failure isolates. Measurements of atovaquone, proguanil and cycloguanil concentrations were determined from plasma to assess for correct drug absorption and compliance.
2.2.1. Atovaquone in vitro susceptibility testing An in vitro isotopic test was used to determine the atovaquone inhibitory concentration 50% (IC50). The in vitro atovaquone resistance threshold is between 40 and 1900 nM [11].
2.2.2. Cytochrome b genotyping The entire pfcytb gene was analysed by sequencing [11]. Position of pfcytb268 was further investigated using a nested polymerase chain reaction followed by a restriction full-length polymorphism method.
Of the 10 analysed atovaquone-proguanil treatment failures, parasite resistance was present, absent and indeterminate in five, three and two cases, respectively (Table 1). In absence of travel between Day 0 and Day-of-failure, reinfection was excluded in all the patients. In patients 1, 2 and 3, the atovaquone in vitro phenotype excluded parasite resistance, with susceptible values below the in vitro minimum resistance threshold of 40 nM. Day 3 drug level measurements confirmed poor absorption or bad compliance. In these cases, no pfcytb polymorphism was identified in parasites isolated before and after the treatment failure. In patients 4e8, drug level measurements or the atovaquone in vitro phenotype confirmed parasite resistance with susceptible values far above the resistance threshold. All these cases were associated with triple mutant pfdhfr genotype and carried a pfcytb268 mutation, these being Y268S (n ¼ 3) or a novel mutation Y268C (n ¼ 2). In addition, a change S299N was observed in Day-0 and Day-of-failure parasites from patient 4. Pre-treatment isolates were available for all cases except Day-0 sample for patient 5. They showed the parasites to be pfcytb wild-type on Day 0, and until Day 7 for patient 4 and Day 10 for patient 8.
Table 1 Characterization of atovaquone-proguanil (AP) falciparum malaria treatment failures
Age/sex/weight (kg) Country of infestation Chemoprophylaxis Day 0a
Day-of-failure
Plasma drug level (mM) pfdhfr genotypeb pfcytb genotypec pfmsp genotypesd
Failure caused by AP resistance
Patient 1
Patient 2
Patient 3
Patient 4
40/M/70 Mali
17/F/55 Ivory Coast
22/F/60 Mali
38/F/86 Burkina Faso
51/M/77 Burkina Faso and/or Mali CP underdosed CP underdosed CP underdosed CP correct Fever Headaches Fever and shiver Fever and asthenia
None Fever, diarrhoea and vomiting Parasitaemia 0.002% 0.3% Dosage Correct Correct Remark D1: vomiting Intakes with low fatty food Symptoms
Day Symptoms Parasitaemia Atov. IC50 (nM) Atovaquone Proguanil Cycloguanil D0 Dfailure D0 Dfailure D0 1
Dfailure
1
552 (1)
552 (1)
2
716 (1)
767 (1)
55/M/83 Ivory Coast
14/F/52 Ivory Coast
28/M/76 Mali
CP correct Fever
None Fever
None Fever
CP correct Fever and shiver
23/F/>100 Burkina Faso and/or Senegal CP correct Fever
4% Correct Hospitalised patient
0.15% Correct (D23)g D0: 4 pills D10: no parasite 39 Fever
0.2% 2.8% Correct Correct D2: vomiting Hospitalised patient
5% 0.25% Unsuccessful 10,400
1.1% 1.5% Unsuccessful Unsuccessful
D26 1.9 <0.03 <0.04 Triple mutant Triple mutant wt Cys268 552 (0.8), 603 (0.2) 735 (0.4), 755 (0.6) 552 (1)
D8 of 4.1 second <0.03 treatment <0.04 Triple mutant Triple mutant wti Ser268 611 (1)
D3
wt wt wt wt 600 (1)
682 (1)
855 (1)
611 (1)
600 (1)
1.4 ND ND Triple mutant Triple mutant wt Ser268 582 (0.8), 592 (0.2) 682 (0.8), 864 (0.2) 582 (1)
755 (1)
682 (1)
855 (1)
864 (1)
0.47% 17,000
0.04% Unsuccessful
5% 8230
D3
13.4 1.8 0.2 Triple mutant Triple mutant Asn299h Cys268, Asn299 632 (1)
D2 of second 10.1 ND treatment 2.1 ND 0.1 ND ND Triple mutant Triple mutant Triple mutant ND wt Ser268 Ser268 ND 572 (1)
714 (1)
ND
632 (1)
638 (1)
714 (1)
717 (1)
D3
767 (1)
36/M/100 Guinea
25 Fever
D3
716 (1)
Patient 10
22 Fever
1% 1.49
2
Patient 9
13% Correct Self-extended with 2 pills/ day for 6 days 26 Fever
0.5% 9.89
wt wt wt wt 552 (1)
Patient 8
Unknown Correct 2 AP treatmentsf (D0 and D21)
11 Fever and diarrhoea 0.75% 7.87
<1.4 <0.03 <0.04 Triple mutant Triple mutant wt wt 552 (1)
Patient 7
0.35% Correct D7: no parasite
7 None
3 1.1 0.05
Indeterminate Patient 6e
0.007% Incorrect 2 pills/day for 6 days
3 Fever
D3
Patient 5
<1.4 <0.03 <0.04 Triple mutant Triple mutant wt wt 552 (0.8), 641 (0.2) 716 (0.4), 772 (0.6) 552 (0.1), 641 (0.9) 716 (0.9), 772 (0.1)
836 (0.8), 986 (0.2) 572 (1) 956 (0.9), 986 (0.1)
26 Fever, shiver
3 Asthenia
<1.4 0.04 1.6
28 None
D3
L. Musset et al. / Microbes and Infection 8 (2006) 2599e2604
Treatment
Failure in absence of AP resistance
Criteria of classification of the failures were in bold, CP: chloroquineeproguanil, wt: wild-type, and ND: not done. a Day of the diagnosis and initiation of the treatment. b Dihydrofolate reductase genotype. c Cytochrome b genotype. d Polymorphisms concerning merozoite surface proteins 1 and 2: each clone was characterized by the size of amplification products in base pair, in brackets its proportion in the parasite population. e Case shortly reported [11]. f Day 0: first AP treatment prescribed by a practitioner, symptoms returned on Day 21 leading to an ineffective self AP treatment. g A single AP dose on Day 0, consequently, parasite recrudescence on Day 23 was treated with a new standard AP treatment under supervision, symptoms returned 13 days later, on Day 39. h Same on Day 7. i Same on Day 10. 2601
2602
L. Musset et al. / Microbes and Infection 8 (2006) 2599e2604
No conclusion could be drawn on parasite resistance with results from patients 9 and 10. In the former, undetectable plasma drug levels were identified but atovaquone in vitro tests were unsuccessful, thus parasite resistance could not be excluded. Nevertheless, the short time (3 days) that elapsed before parasite recrudescence and the presence of the Y268 codon in pfcytb suggested the treatment failure in the absence of AP parasite resistance. Observations from the latter show that AP failure is not systematically associated with symptoms on Day 28. Its residual plasma concentration of atovaquone on Day 3 was at the limit of detection. As the intakes were supervised, the low level of drug could have resulted from the patient’s weight higher than 100 kg. In fact, compared to a patient of 70 kg, oral clearance and volume of distribution of atovaquone were increased by 40% [13]. Nevertheless, even if this low level of atovaquone contributed to the failure, the presence of Y268S mutation on Day 28 suggested parasite resistance.
4. Discussion In imported uncomplicated P. falciparum malaria, AP has rapidly become the first line antimalarial drug in most European infectious diseases wards. Epidemiological monitoring of P. falciparum resistance to AP has thus become essential and requires reliable molecular markers. Since the emergence of AP resistant parasites, Y268S and Y268N PfCYTb changes have been proposed as a molecular marker of P. falciparum AP resistance. However, they were not identified in all treatment failures. In the current study, pfcytb codon 268 mutations were unambiguously associated with AP parasite resistance, two cases harbouring a new change, Y268C. The previously reported Y268N mutation was not observed in this series [3]. Our Y268S or Y268C parasites exhibited an important level of resistance with atovaquone IC50 values around 10,000 nM, far above the resistance threshold of 40 nM. The higher IC50 value was obtained with Y268C mutation (plus
Fig. 1. Pfmsp2 genotypic composition of Plasmodium falciparum parasites isolated from patient 6 and patient 10 before treatment and upon parasite recrudescence. Genotypes of each isolate were visualized by electropherograms. White peaks represent the internal lane standard (470e1021 base pairs) and grey peaks represent genotypes. Each genotype was characterized by the size of the msp2 amplification product and a quantitative estimation of its proportion in the parasite population was derived from the value yielded by the area under the curve of the corresponding peak (value in Table 1). In these two cases, the atovaquone-proguanil treatment failure was associated with parasite resistance.
L. Musset et al. / Microbes and Infection 8 (2006) 2599e2604
S299N). This is in accordance with biochemical data suggesting a role for cysteine in impairment of atovaquone binding. In fact, yeast transfection studies demonstrated that tyrosine changes by a nucleophilic residue, cysteine or serine, suppressed cytochrome beatovaquone interactions and conferred atovaquone resistance [14]. Involvement of weak hydrogen bonds from the aromatic side chain of tyrosine (fully conserved in all cytochrome b) is recently suggested as crucial for positioning ubiquinol in the active site of cytochrome b [15]. It is unlikely that change S299N will be implicated in the high resistance level as this position is in the inner mitochondrial membrane, far from the atovaquone binding site. We were unable to detect pfcytb mutations on codon 268 in pre-treatment isolates. Two hypotheses could be suggested. Firstly, pfcytb268 was present but in the minority of pre-treatment isolates making it undetectable for classical genotyping methods. Secondly, pfcytb268 mutation has appeared within patient from a parasite population initially pfcytb wild-type. Pfmsp1 and pfmsp2 genotypings being more sensitive to detect a minority genotype, all Day-0 and Day-of-failure isolates were further analysed with these methods. Profiles suggest that mutation appeared after Day-0 as pfmsp1 and pfmsp2 genotypes identified after recrudescence were also present in Day-0 isolates in sufficient amount (18e100%) for PCR detection of the pfcytb268 mutations (Fig. 1). In this event, profile from patient 6 could be the result of two genetic modifications, the acquisition of the pfcytb268 mutation followed by a switch generating two distinct pfmsp2 genotypes. Three previous cases of AP treatment failure with simultaneous genotyping of Day-0 and Day-of-failure parasites have been reported [4,5,7]. Y268S was identified before treatment in one case, while in the other two, a change in pfcytb genotype was observed. In these latter two cases, an identical pfmsp1 genotype composition was found in Day-0 and Day-of-failure isolates. Even though a previous study demonstrated the absence of pfdhfr triple mutations effect on the intrinsic activity of proguanil, in this study, all Day-0 and Day-of-failure isolates from treatment failures associated with AP resistance were pfdhfr triple mutant, except for patients 1 and 9 [16]. The first explanation could be odds, taking in account the elevated proportion of isolates bearing triple mutations in West and Central Africa (>50%). The second one could be resistant to cycloguanil as a requirement for the expression of AP resistance. Not considering treatment failures reported from external hospitals, two cases of AP treatment failures were identified as associated with parasite resistance among 298 patients treated between 2003 and 2005. Several patients were not followed for 28 days, and our calculations predict the actual prevalence of AP resistance in this population to be equal to or higher than 0.08%. Treatment failures associated with parasite resistance, occurring after Day 20, sometimes without symptoms, highlight the importance of monitoring AP treatment efficacy for at least 28 days. Finally, pfcytb codon 268, with changes Y268S and Y268C, is associated with AP resistance. Nevertheless, mutations have not been detectable before exposure to the drug. Even if pfmsp genotyping suggest an
2603
emergence of mutations, within patient, from parasites initially wild-type, improvements of the sensitivity of existing genotypic methods will be useful to confirm or not the absence of mutation on Day-0. Acknowledgements We thank all collaborating centres for their participation in collecting materials and data. This work was supported by the French Ministry of Health (grant to the National Reference Centre). LM is the recipient of a PhD grant from the French Ministry of Education and Research. We thank David Fidock and Je´roˆme Clain for helpful discussions and suggestions. References [1] M. Fry, M. Pudney, Site of action of the antimalarial hydroxynaphthoquinone, 2-[trans-4-(4 0 -chlorophenyl)cyclohexyl]-3-hydroxy-1,4-naphthoquinone (566C80), Biochem. Pharmacol. 43 (1992) 1545e1553. [2] I.K. Srivastava, A.B. Vaidya, A mechanism for the synergistic antimalarial action of atovaquone and proguanil, Antimicrob. Agents Chemother. 43 (1999) 1334e1339. [3] Q.L. Fivelman, G.A. Butcher, I.S. Adagu, D.C. Warhurst, G. Pasvol, Malarone treatment failure and in vitro confirmation of resistance of Plasmodium falciparum isolate from Lagos, Nigeria [Online]. Available from: Malar. J. 1 http://www.malariajournal.com/content/1/1/1 (2002) (accessed 11.01.05). [4] E. Schwartz, S. Bujanover, K.C. Kain, Genetic confirmation of atovaquone-proguanil-resistant Plasmodium falciparum malaria acquired by a nonimmune traveler to East Africa, Clin. Infect. Dis. 37 (2003) 450e451. [5] A. Fa¨rnert, J. Lindberg, P. Gil, G. Swedberg, Y. Berqvit, M.M. Thapar, N. Lindegardh, S. Berezcky, A. Bjo¨rkman, Evidence of Plasmodium falciparum malaria resistant to atovaquone and proguanil hydrochloride: case reports, BMJ 326 (2003) 628e629. [6] O. Wichmann, N. Muehlberger, T. Jelinek, M. Alifrangis, G. PeyerlHoffmann, M. Mu¨hlen, M.P. Grobusch, J. Gascon, A. Matteelli, H. Laferl, Z. Bisoffi, S. Ehrhardt, J. Cuadros, C. Hatz, I. Gjørup, P. McWhinney, J. Beran, S. da Cunha, M. Schulze, H. Kollaritsch, P. Kern, G. Fry, J. Richter, The European Network on Surveillance of Imported Infectious Diseases, Screening for mutations related to atovaquone/proguanil resistance in treatment failures and other imported isolates of Plasmodium falciparum in Europe, J. Infect. Dis. 190 (2004) 1541e1546. [7] S. Kuhn, M.J. Gill, K.C. Kain, Emergence of atovaquone-proguanil resistance during treatment of Plasmodium falciparum malaria acquired by a non-immune north American traveller to west Africa, Am. J. Trop. Med. Hyg. 72 (2005) 407e409. [8] S.R. Meshnick, B. Trumpower, Multiple cytochrome b mutations may cause atovaquone resistance (letter), J. Infect. Dis. 191 (2005) 822. [9] World Health Organization, Assessment and monitoring of antimalarial drug efficacy for the treatment of uncomplicated falciparum malaria [Online]. Available from: http://www.who.int/malaria/includes_en/ whomalariapublications19982004.htm (2003) (accessed 09.05.05). [10] Management and prevention of imported Plasmodium falciparum malaria. The 12th Consensus Conference of Anti-infectious. Therapy of the French-speaking Society of Infectious Diseases, 14 April 1999, Arch. Pediatr. 7 (2000) 201e208. [11] L. Musset, B. Pradines, D. Parzy, R. Durand, P. Bigot, J. Le Bras, Apparent absence of atovaquone/proguanil resistance in 477 Plasmodium falciparum isolates from untreated French travellers, J. Antimicrob. Chemother. 57 (2006) 110e115. [12] S. Jafari, J. Le Bras, O. Bouchaud, R. Durand, Plasmodium falciparum clonal population dynamics during malaria treatment, J. Infect. Dis. 189 (2004) 195e203.
2604
L. Musset et al. / Microbes and Infection 8 (2006) 2599e2604
[13] Z. Hussein, J. Eaves, D.B. Hutchinson, C.J. Canfield, Population pharmacokinetics of atovaquone in patients with acute malaria caused by Plasmodium falciparum, Clin. Pharmacol. Ther. 61 (1997) 518e530. [14] J.J. Kessl, H.K. Ha, A.K. Merritt, B.B. Lange, P. Hill, B. Meunier, S.R. Meshnick, B.L. Trumpower, Cytochrome b mutations that modify the ubiquinol binding pocket of the cytochrome bc1 complex and confer anti-malarial drug resistance in Saccharomyces cerevisiae, J. Biol. Chem. 280 (2005) 17142e17148.
[15] H. Palsdottir, C.G. Lojero, B.L. Trumpower, C. Hunte, Structure of the yeast cytochrome bc1 complex with a hydroxyquinone anion Q0 site inhibitor bound, J. Biol. Chem. 278 (2003) 31303e 31311. [16] D.A. Fidock, T.E. Wellems, Transformation with human dihydrofolate reductase renders malaria parasites insensitive to WR99210 but does not affect the intrinsic activity of proguanil, Proc. Natl. Acad. Sci. U.S.A. 94 (1997) 10931e10936.
Original article
Clinical atovaquone-proguanil resistance of Plasmodium falciparum associated with cytochrome b codon 268 mutations Lise Musset a,b,*, Olivier Bouchaud c, Sophie Matheron d, Laurent Massias e, Jacques Le Bras a,b a Laboratoire de biologie animale et parasitaire, EA209, Universite´ Paris Descartes, 4 av de l’observatoire, 75006 Paris, France Centre National de Re´fe´rence Paludisme, APHP, Hoˆpital Bichat-Claude Bernard, 46 rue Henri Huchard, 75877 Paris Cedex 18, France c Unite´ de maladies infectieuses et tropicales, APHP, Hoˆpital Avicenne, 125 rue de Stalingrad, 93000 Bobigny, France d Unite´ de maladies infectieuses et tropicales, APHP, Hoˆpital Bichat-Claude Bernard, 46 rue Henri Huchard, 75877 Paris Cedex 18, France e Laboratoire de toxicologie, APHP, Hoˆpital Bichat-Claude Bernard, 46 rue Henri Huchard, 75877 Paris Cedex 18, France b
Received 7 July 2006; accepted 12 July 2006 Available online 10 August 2006
Abstract Plasmodium falciparum resistance to atovaquone-proguanil has so far been associated with Y268S or Y268N mutations in cytochrome b, although these changes were identified in only seven of the 11 treatment failures. Here, we describe 10 new cases of atovaquone-proguanil treatment failures among which the parasite resistance was confirmed in six cases, either by identifying correct plasma drug concentrations or by observing in vitro atovaquone resistance. Resistance was consistently associated with codon 268 mutations (Y268S or a previously unidentified mutation, Y268C). Notably, mutations were not detected before the treatment but only after the drug exposure. Ó 2006 Elsevier Masson SAS. All rights reserved. Keywords: Plasmodium falciparum; Chemotherapy; Resistance; Cytochrome b; Atovaquone; Proguanil; Malarone
1. Introduction In 1996, the atovaquone-proguanil (AP) combination was registered in North America and Europe for the prophylaxis and treatment of malaria. While this safe and efficient combination is increasingly used in developed countries, its high cost precludes a large use in endemic countries. Atovaquone, a ubiquinone analogue binding to cytochrome b of plasmodial mitochondria, inhibits electron transfer of the respiratory chain
Abbreviations: AP, atovaquone-proguanil; pfcytb, cytochrome b gene; pfdhfr, dihydrofolate reductase gene; IC50, inhibitory concentration 50%; C, cysteine; pfmsp, merozoite surface protein gene; N, asparagine; PCR, polymerase chain reaction; S, serine; Y, tyrosine; WHO, World Health Organization. * Corresponding author. Laboratoire de Parasitologie-Mycologie, Hoˆpital Bichat-Claude Bernard, 46 rue Henri Huchard, 75877 Paris Cedex 18, France. Tel.: þ33 140 257 899; fax: þ33 140 256 763. E-mail address: [email protected] (L. Musset). 1286-4579/$ - see front matter Ó 2006 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.micinf.2006.07.011
[1]. Proguanil is mainly considered as a pro-drug of cycloguanil, an inhibitor of the dihydrofolate reductase also involved in pyrimidine biosynthesis. However, in combination with atovaquone, proguanil, by itself, lowers the effective concentration at which the former collapses the mitochondrial membrane potential [2]. The proguanil target and the detailed mechanism implicated in AP synergy remain unknown. Since the introduction of AP combination, 11 cases of treatment failures have been published in travelers returning from Africa [3e7]. Seven failures exhibited a modification of codon 268 of the cytochrome b gene ( pfcytb268), mostly from tyrosine (Y) to serine (S) and four failures were reported without any pfcytb mutation. Thus, the usefulness of pfcytb268 mutations for predicting Plasmodium falciparum AP resistance has been questioned [8]. The limitation for understanding the failure mechanisms may also have been related to an insufficient description of the cases, impairing discrimination between parasite resistance and poor drug absorption. Atovaquone is a very
2600
L. Musset et al. / Microbes and Infection 8 (2006) 2599e2604
lipophilic drug and its low bioavailability is 4-fold increased (reaching 23%) with simultaneous intake of fatty food. This poor absorption of atovaquone may induce treatment failures without parasite resistance. With this in mind, this study was designed to investigate putative causes of 10 additional AP treatment failures identified over a 3-year period.
2. Materials and methods
2.2.3. Pfdhfr genotyping The three major pfdhfr mutations (at positions 51, 59, and 108) associated with cycloguanil resistance were studied with a restriction method. 2.2.4. Parasite population analysis Parasite diversity within isolates was determined by a fragment-analysis method of pfmsp1 and pfmsp2 polymorphisms. This method allows for the detection of all genotypes accounting for more than 2% of the whole parasite population [12].
2.1. Patients Between January 2003 and September 2005, six cases of AP treatment failures were identified among 298 patients suffering from uncomplicated P. falciparum malaria and treated with AP (four tablets daily for 3 days). Most patients, being 9e75 years of age, had returned from Central or West African countries. Eighty percent were living in France but were native from the country of infection. In our hospital, a voluntary monitoring of malaria treatment efficacy is systematically proposed. Based on the WHO standard protocol, follow-up plans clinical and parasitological examinations between Day 0 (i.e. first day of treatment) and Day 28 (generally, Day 3, Day 7 and Day 28). Early and late treatment failures were defined following WHO criteria [9]. Informed consent was not required for this study as the following procedures are part of the French national recommendations for the care and surveillance of malaria [10]. As our laboratory is the national reference centre for malaria, we were informed of four additional cases of late AP treatment failure. In agreement with national consensus, all patients with failure were re-treated with quinine.
2.2.5. Drug measurements Determinations of drug concentrations in plasma were performed by reverse phase high performance liquid chromatography. The lower limits of quantification were 1.4 mM, 0.03 mM and 0.04 mM for atovaquone, proguanil and cycloguanil, respectively. Results were compared to effective atovaquone plasma concentrations in malaria treatment: between 3 and 20 mM, 0.6 and 18 mM, 0.3 and 2.2 mM on Day 3, Day 8 and Day 21, respectively [13]. 2.3. Classification of treatment failures To define a reliable allele of AP resistance, treatment failures were classified into three categories: (i) Failures in absence of AP resistance: incorrect plasma drug dosages associated with Day-of-failure parasites in vitro susceptible to atovaquone, (ii) Failures caused by AP resistance: correct drug dosages or Day-of-failure parasites resistant to atovaquone, and (iii) Indeterminate. 3. Results
2.2. Investigation of treatment failures For all treatment failure patients, atovaquone in vitro susceptibility testing, pfcytb and pfdhfr genotypings, and merozoite surface proteins 1 and 2 ( pfmsp1 and pfmsp2) polymorphism analyses were performed on Day-0 and Day-of-failure isolates. Measurements of atovaquone, proguanil and cycloguanil concentrations were determined from plasma to assess for correct drug absorption and compliance.
2.2.1. Atovaquone in vitro susceptibility testing An in vitro isotopic test was used to determine the atovaquone inhibitory concentration 50% (IC50). The in vitro atovaquone resistance threshold is between 40 and 1900 nM [11].
2.2.2. Cytochrome b genotyping The entire pfcytb gene was analysed by sequencing [11]. Position of pfcytb268 was further investigated using a nested polymerase chain reaction followed by a restriction full-length polymorphism method.
Of the 10 analysed atovaquone-proguanil treatment failures, parasite resistance was present, absent and indeterminate in five, three and two cases, respectively (Table 1). In absence of travel between Day 0 and Day-of-failure, reinfection was excluded in all the patients. In patients 1, 2 and 3, the atovaquone in vitro phenotype excluded parasite resistance, with susceptible values below the in vitro minimum resistance threshold of 40 nM. Day 3 drug level measurements confirmed poor absorption or bad compliance. In these cases, no pfcytb polymorphism was identified in parasites isolated before and after the treatment failure. In patients 4e8, drug level measurements or the atovaquone in vitro phenotype confirmed parasite resistance with susceptible values far above the resistance threshold. All these cases were associated with triple mutant pfdhfr genotype and carried a pfcytb268 mutation, these being Y268S (n ¼ 3) or a novel mutation Y268C (n ¼ 2). In addition, a change S299N was observed in Day-0 and Day-of-failure parasites from patient 4. Pre-treatment isolates were available for all cases except Day-0 sample for patient 5. They showed the parasites to be pfcytb wild-type on Day 0, and until Day 7 for patient 4 and Day 10 for patient 8.
Table 1 Characterization of atovaquone-proguanil (AP) falciparum malaria treatment failures
Age/sex/weight (kg) Country of infestation Chemoprophylaxis Day 0a
Day-of-failure
Plasma drug level (mM) pfdhfr genotypeb pfcytb genotypec pfmsp genotypesd
Failure caused by AP resistance
Patient 1
Patient 2
Patient 3
Patient 4
40/M/70 Mali
17/F/55 Ivory Coast
22/F/60 Mali
38/F/86 Burkina Faso
51/M/77 Burkina Faso and/or Mali CP underdosed CP underdosed CP underdosed CP correct Fever Headaches Fever and shiver Fever and asthenia
None Fever, diarrhoea and vomiting Parasitaemia 0.002% 0.3% Dosage Correct Correct Remark D1: vomiting Intakes with low fatty food Symptoms
Day Symptoms Parasitaemia Atov. IC50 (nM) Atovaquone Proguanil Cycloguanil D0 Dfailure D0 Dfailure D0 1
Dfailure
1
552 (1)
552 (1)
2
716 (1)
767 (1)
55/M/83 Ivory Coast
14/F/52 Ivory Coast
28/M/76 Mali
CP correct Fever
None Fever
None Fever
CP correct Fever and shiver
23/F/>100 Burkina Faso and/or Senegal CP correct Fever
4% Correct Hospitalised patient
0.15% Correct (D23)g D0: 4 pills D10: no parasite 39 Fever
0.2% 2.8% Correct Correct D2: vomiting Hospitalised patient
5% 0.25% Unsuccessful 10,400
1.1% 1.5% Unsuccessful Unsuccessful
D26 1.9 <0.03 <0.04 Triple mutant Triple mutant wt Cys268 552 (0.8), 603 (0.2) 735 (0.4), 755 (0.6) 552 (1)
D8 of 4.1 second <0.03 treatment <0.04 Triple mutant Triple mutant wti Ser268 611 (1)
D3
wt wt wt wt 600 (1)
682 (1)
855 (1)
611 (1)
600 (1)
1.4 ND ND Triple mutant Triple mutant wt Ser268 582 (0.8), 592 (0.2) 682 (0.8), 864 (0.2) 582 (1)
755 (1)
682 (1)
855 (1)
864 (1)
0.47% 17,000
0.04% Unsuccessful
5% 8230
D3
13.4 1.8 0.2 Triple mutant Triple mutant Asn299h Cys268, Asn299 632 (1)
D2 of second 10.1 ND treatment 2.1 ND 0.1 ND ND Triple mutant Triple mutant Triple mutant ND wt Ser268 Ser268 ND 572 (1)
714 (1)
ND
632 (1)
638 (1)
714 (1)
717 (1)
D3
767 (1)
36/M/100 Guinea
25 Fever
D3
716 (1)
Patient 10
22 Fever
1% 1.49
2
Patient 9
13% Correct Self-extended with 2 pills/ day for 6 days 26 Fever
0.5% 9.89
wt wt wt wt 552 (1)
Patient 8
Unknown Correct 2 AP treatmentsf (D0 and D21)
11 Fever and diarrhoea 0.75% 7.87
<1.4 <0.03 <0.04 Triple mutant Triple mutant wt wt 552 (1)
Patient 7
0.35% Correct D7: no parasite
7 None
3 1.1 0.05
Indeterminate Patient 6e
0.007% Incorrect 2 pills/day for 6 days
3 Fever
D3
Patient 5
<1.4 <0.03 <0.04 Triple mutant Triple mutant wt wt 552 (0.8), 641 (0.2) 716 (0.4), 772 (0.6) 552 (0.1), 641 (0.9) 716 (0.9), 772 (0.1)
836 (0.8), 986 (0.2) 572 (1) 956 (0.9), 986 (0.1)
26 Fever, shiver
3 Asthenia
<1.4 0.04 1.6
28 None
D3
L. Musset et al. / Microbes and Infection 8 (2006) 2599e2604
Treatment
Failure in absence of AP resistance
Criteria of classification of the failures were in bold, CP: chloroquineeproguanil, wt: wild-type, and ND: not done. a Day of the diagnosis and initiation of the treatment. b Dihydrofolate reductase genotype. c Cytochrome b genotype. d Polymorphisms concerning merozoite surface proteins 1 and 2: each clone was characterized by the size of amplification products in base pair, in brackets its proportion in the parasite population. e Case shortly reported [11]. f Day 0: first AP treatment prescribed by a practitioner, symptoms returned on Day 21 leading to an ineffective self AP treatment. g A single AP dose on Day 0, consequently, parasite recrudescence on Day 23 was treated with a new standard AP treatment under supervision, symptoms returned 13 days later, on Day 39. h Same on Day 7. i Same on Day 10. 2601
2602
L. Musset et al. / Microbes and Infection 8 (2006) 2599e2604
No conclusion could be drawn on parasite resistance with results from patients 9 and 10. In the former, undetectable plasma drug levels were identified but atovaquone in vitro tests were unsuccessful, thus parasite resistance could not be excluded. Nevertheless, the short time (3 days) that elapsed before parasite recrudescence and the presence of the Y268 codon in pfcytb suggested the treatment failure in the absence of AP parasite resistance. Observations from the latter show that AP failure is not systematically associated with symptoms on Day 28. Its residual plasma concentration of atovaquone on Day 3 was at the limit of detection. As the intakes were supervised, the low level of drug could have resulted from the patient’s weight higher than 100 kg. In fact, compared to a patient of 70 kg, oral clearance and volume of distribution of atovaquone were increased by 40% [13]. Nevertheless, even if this low level of atovaquone contributed to the failure, the presence of Y268S mutation on Day 28 suggested parasite resistance.
4. Discussion In imported uncomplicated P. falciparum malaria, AP has rapidly become the first line antimalarial drug in most European infectious diseases wards. Epidemiological monitoring of P. falciparum resistance to AP has thus become essential and requires reliable molecular markers. Since the emergence of AP resistant parasites, Y268S and Y268N PfCYTb changes have been proposed as a molecular marker of P. falciparum AP resistance. However, they were not identified in all treatment failures. In the current study, pfcytb codon 268 mutations were unambiguously associated with AP parasite resistance, two cases harbouring a new change, Y268C. The previously reported Y268N mutation was not observed in this series [3]. Our Y268S or Y268C parasites exhibited an important level of resistance with atovaquone IC50 values around 10,000 nM, far above the resistance threshold of 40 nM. The higher IC50 value was obtained with Y268C mutation (plus
Fig. 1. Pfmsp2 genotypic composition of Plasmodium falciparum parasites isolated from patient 6 and patient 10 before treatment and upon parasite recrudescence. Genotypes of each isolate were visualized by electropherograms. White peaks represent the internal lane standard (470e1021 base pairs) and grey peaks represent genotypes. Each genotype was characterized by the size of the msp2 amplification product and a quantitative estimation of its proportion in the parasite population was derived from the value yielded by the area under the curve of the corresponding peak (value in Table 1). In these two cases, the atovaquone-proguanil treatment failure was associated with parasite resistance.
L. Musset et al. / Microbes and Infection 8 (2006) 2599e2604
S299N). This is in accordance with biochemical data suggesting a role for cysteine in impairment of atovaquone binding. In fact, yeast transfection studies demonstrated that tyrosine changes by a nucleophilic residue, cysteine or serine, suppressed cytochrome beatovaquone interactions and conferred atovaquone resistance [14]. Involvement of weak hydrogen bonds from the aromatic side chain of tyrosine (fully conserved in all cytochrome b) is recently suggested as crucial for positioning ubiquinol in the active site of cytochrome b [15]. It is unlikely that change S299N will be implicated in the high resistance level as this position is in the inner mitochondrial membrane, far from the atovaquone binding site. We were unable to detect pfcytb mutations on codon 268 in pre-treatment isolates. Two hypotheses could be suggested. Firstly, pfcytb268 was present but in the minority of pre-treatment isolates making it undetectable for classical genotyping methods. Secondly, pfcytb268 mutation has appeared within patient from a parasite population initially pfcytb wild-type. Pfmsp1 and pfmsp2 genotypings being more sensitive to detect a minority genotype, all Day-0 and Day-of-failure isolates were further analysed with these methods. Profiles suggest that mutation appeared after Day-0 as pfmsp1 and pfmsp2 genotypes identified after recrudescence were also present in Day-0 isolates in sufficient amount (18e100%) for PCR detection of the pfcytb268 mutations (Fig. 1). In this event, profile from patient 6 could be the result of two genetic modifications, the acquisition of the pfcytb268 mutation followed by a switch generating two distinct pfmsp2 genotypes. Three previous cases of AP treatment failure with simultaneous genotyping of Day-0 and Day-of-failure parasites have been reported [4,5,7]. Y268S was identified before treatment in one case, while in the other two, a change in pfcytb genotype was observed. In these latter two cases, an identical pfmsp1 genotype composition was found in Day-0 and Day-of-failure isolates. Even though a previous study demonstrated the absence of pfdhfr triple mutations effect on the intrinsic activity of proguanil, in this study, all Day-0 and Day-of-failure isolates from treatment failures associated with AP resistance were pfdhfr triple mutant, except for patients 1 and 9 [16]. The first explanation could be odds, taking in account the elevated proportion of isolates bearing triple mutations in West and Central Africa (>50%). The second one could be resistant to cycloguanil as a requirement for the expression of AP resistance. Not considering treatment failures reported from external hospitals, two cases of AP treatment failures were identified as associated with parasite resistance among 298 patients treated between 2003 and 2005. Several patients were not followed for 28 days, and our calculations predict the actual prevalence of AP resistance in this population to be equal to or higher than 0.08%. Treatment failures associated with parasite resistance, occurring after Day 20, sometimes without symptoms, highlight the importance of monitoring AP treatment efficacy for at least 28 days. Finally, pfcytb codon 268, with changes Y268S and Y268C, is associated with AP resistance. Nevertheless, mutations have not been detectable before exposure to the drug. Even if pfmsp genotyping suggest an
2603
emergence of mutations, within patient, from parasites initially wild-type, improvements of the sensitivity of existing genotypic methods will be useful to confirm or not the absence of mutation on Day-0. Acknowledgements We thank all collaborating centres for their participation in collecting materials and data. This work was supported by the French Ministry of Health (grant to the National Reference Centre). LM is the recipient of a PhD grant from the French Ministry of Education and Research. We thank David Fidock and Je´roˆme Clain for helpful discussions and suggestions. References [1] M. Fry, M. Pudney, Site of action of the antimalarial hydroxynaphthoquinone, 2-[trans-4-(4 0 -chlorophenyl)cyclohexyl]-3-hydroxy-1,4-naphthoquinone (566C80), Biochem. Pharmacol. 43 (1992) 1545e1553. [2] I.K. Srivastava, A.B. Vaidya, A mechanism for the synergistic antimalarial action of atovaquone and proguanil, Antimicrob. Agents Chemother. 43 (1999) 1334e1339. [3] Q.L. Fivelman, G.A. Butcher, I.S. Adagu, D.C. Warhurst, G. Pasvol, Malarone treatment failure and in vitro confirmation of resistance of Plasmodium falciparum isolate from Lagos, Nigeria [Online]. Available from: Malar. J. 1 http://www.malariajournal.com/content/1/1/1 (2002) (accessed 11.01.05). [4] E. Schwartz, S. Bujanover, K.C. Kain, Genetic confirmation of atovaquone-proguanil-resistant Plasmodium falciparum malaria acquired by a nonimmune traveler to East Africa, Clin. Infect. Dis. 37 (2003) 450e451. [5] A. Fa¨rnert, J. Lindberg, P. Gil, G. Swedberg, Y. Berqvit, M.M. Thapar, N. Lindegardh, S. Berezcky, A. Bjo¨rkman, Evidence of Plasmodium falciparum malaria resistant to atovaquone and proguanil hydrochloride: case reports, BMJ 326 (2003) 628e629. [6] O. Wichmann, N. Muehlberger, T. Jelinek, M. Alifrangis, G. PeyerlHoffmann, M. Mu¨hlen, M.P. Grobusch, J. Gascon, A. Matteelli, H. Laferl, Z. Bisoffi, S. Ehrhardt, J. Cuadros, C. Hatz, I. Gjørup, P. McWhinney, J. Beran, S. da Cunha, M. Schulze, H. Kollaritsch, P. Kern, G. Fry, J. Richter, The European Network on Surveillance of Imported Infectious Diseases, Screening for mutations related to atovaquone/proguanil resistance in treatment failures and other imported isolates of Plasmodium falciparum in Europe, J. Infect. Dis. 190 (2004) 1541e1546. [7] S. Kuhn, M.J. Gill, K.C. Kain, Emergence of atovaquone-proguanil resistance during treatment of Plasmodium falciparum malaria acquired by a non-immune north American traveller to west Africa, Am. J. Trop. Med. Hyg. 72 (2005) 407e409. [8] S.R. Meshnick, B. Trumpower, Multiple cytochrome b mutations may cause atovaquone resistance (letter), J. Infect. Dis. 191 (2005) 822. [9] World Health Organization, Assessment and monitoring of antimalarial drug efficacy for the treatment of uncomplicated falciparum malaria [Online]. Available from: http://www.who.int/malaria/includes_en/ whomalariapublications19982004.htm (2003) (accessed 09.05.05). [10] Management and prevention of imported Plasmodium falciparum malaria. The 12th Consensus Conference of Anti-infectious. Therapy of the French-speaking Society of Infectious Diseases, 14 April 1999, Arch. Pediatr. 7 (2000) 201e208. [11] L. Musset, B. Pradines, D. Parzy, R. Durand, P. Bigot, J. Le Bras, Apparent absence of atovaquone/proguanil resistance in 477 Plasmodium falciparum isolates from untreated French travellers, J. Antimicrob. Chemother. 57 (2006) 110e115. [12] S. Jafari, J. Le Bras, O. Bouchaud, R. Durand, Plasmodium falciparum clonal population dynamics during malaria treatment, J. Infect. Dis. 189 (2004) 195e203.
2604
L. Musset et al. / Microbes and Infection 8 (2006) 2599e2604
[13] Z. Hussein, J. Eaves, D.B. Hutchinson, C.J. Canfield, Population pharmacokinetics of atovaquone in patients with acute malaria caused by Plasmodium falciparum, Clin. Pharmacol. Ther. 61 (1997) 518e530. [14] J.J. Kessl, H.K. Ha, A.K. Merritt, B.B. Lange, P. Hill, B. Meunier, S.R. Meshnick, B.L. Trumpower, Cytochrome b mutations that modify the ubiquinol binding pocket of the cytochrome bc1 complex and confer anti-malarial drug resistance in Saccharomyces cerevisiae, J. Biol. Chem. 280 (2005) 17142e17148.
[15] H. Palsdottir, C.G. Lojero, B.L. Trumpower, C. Hunte, Structure of the yeast cytochrome bc1 complex with a hydroxyquinone anion Q0 site inhibitor bound, J. Biol. Chem. 278 (2003) 31303e 31311. [16] D.A. Fidock, T.E. Wellems, Transformation with human dihydrofolate reductase renders malaria parasites insensitive to WR99210 but does not affect the intrinsic activity of proguanil, Proc. Natl. Acad. Sci. U.S.A. 94 (1997) 10931e10936.