High resistance of Plasmodium falciparum to sulphadoxine/pyrimethamine in Northern Tanzania and the emergence of dhps resistance mutation at codon 581

Background: Sulphadoxine-pyrimethamine (SP) a widely used treatment for uncomplicated malaria and recommended for intermittent preventive treatment of malaria in pregnancy, is being investigated for intermittent preventive treatment of malaria in infants (IPTi). High levels of drug resistance to SP have been reported from north-eastern Tanzania associated with mutations in parasite genes. This study compared the in vivo efficacy of SP in symptomatic 6-59 month children with uncomplicated malaria and in asymptomatic 2-10 month old infants. Methodology and Principal Findings: An open label single arm (SP) standard 28 day in vivo WHO antimalarial efficacy protocol was used in 6 to 59 months old symptomatic children and a modified protocol used in 2 to 10 months old asymptomatic infants. Enrolment was stopped early (87 in the symptomatic and 25 in the asymptomatic studies) due to the high failure rate. Molecular markers were examined for recrudescence, re-infection and markers of drug resistance and a review of literature of studies looking for the 581G dhps mutation was carried out. In symptomatic children PCR-corrected early treatment failure was 38.8% (95% CI 26.8-50.8) and total failures by day 28 were 82.2% (95% CI 72.5-92.0). There was no significant difference in treatment failures between asymptomatic and symptomatic children. 96% of samples carried parasites with mutations at codons 51, 59 and 108 in the dhfr gene and 63% carried a double mutation at codons 437 and 540. 55% carried a third mutation with the addition of a mutation at codon 581 in the dhps gene. This triple: triple haplotype maybe associated with earlier treatment failure. Conclusion: In northern Tanzania SP is a failed drug for treatment and its utility for prophylaxis is doubtful. The study found a new combination of parasite mutations that maybe associated with increased and earlier failure. © 2009 Gesase et al. The protocol for this trial and supporting CONSORT checklist are available as supporting information; see Protocol S1 and Checklist S1. The study was conducted in Hale Health Centre, Tanga Region, situated 32 km north of Muheza where SP resistance was first observed in Tanzania. The district experiences perennial, holoendemic malaria although, in recent years, transmission appears to have declined substantially. The entomological inoculation rate from the neighbouring district of Muheza was 148 infectious bites per person per year in 2000[25]. The study site was chosen due to its proximity to the site of a clinical trial of IPTi comparing SP, chlorproguanil-dapsone and mefloquine. The study protocol was approved by the Ethics Review Committees of the National Institute of Medical Research of Tanzania and the London School of Hygiene and Tropical Medicine and was registered as a clinical trial with the National Institute of Health (Clinicaltrials.gov identifier {“type”:”clinical-trial”,”attrs”:{“text”:”NCT00361114″,”term_id”:”NCT00361114″}}NCT00361114). The protocol included an arm for chlorproguanil/dapsone in the 6–59 month study after the SP arm was completed, however it was not possible to procure the drug and recently the drug has been withdrawn from clinical development[26]. All children between the ages 6 and 59 months who attended Hale Health Centre during July–August 2006 with a fever or history of fever during the study period were screened for malaria using a rapid diagnostic test (RDT) (Paracheck, Orchid Biomedical Systems, Verna, India). Children with a positive RDT had a thick blood smear read and those with a positive blood smear were referred to the study clinician. Study inclusion criteria were: (1) weight of ≥4.5 kgs, (2) not -enrolled in the IPTi trial, (3) absence of severe malnutrition (weight-for-height <3 standard deviations from the norm), (4) slide-confirmed infection with P. falciparum only with an initial parasite density of between 2,000 and 200,000 asexual parasites per microliter, (5) absence of general danger signs (inability to drink or breastfeed; vomiting; recent history of convulsions; lethargy or unconsciousness; inability to sit or stand up) or other signs of severe and complicated falciparum malaria according to WHO definitions, (6) measured axillary temperature ≥37.5°C, (7) ability to attend stipulated follow-up visits, (8) informed consent provided by parent/guardian; (9) absence of history of hypersensitivity reactions to SP and (10) no prior antimalarial use in the preceding 2 weeks. Consent was obtained from caretakers of 2–10 month old infants who attended the Maternal Child Health (MCH) clinic for immunization or weighing, for screening for P. falciparum infection. From those consented for screening, finger prick blood was obtained for the rapid diagnostic test, thick and thin blood smear preparation and filter paper samples for molecular studies. Children who had a positive blood slide were further assessed for their eligibility for inclusion into the study and enrolled in the drug sensitivity study after obtaining an informed consent. The eligibility criteria were the same as for the study of symptomatic children except that there was no history of fever in the last 48 hours, that measured axillary temperature should be less than 37.5°C and that the presence of P. falciparum parasitaemia at any density was acceptable. In order to detect a 15% difference in adequate parasite clearance by day 28 between symptomatic 6–59 month old children and asymptomatic 2–10 month old children with 80% power at the 5% significance level using a ratio of 2 symptomatic cases to 1 asymptomatic case, we estimated that 292 symptomatic children and 146 asymptomatic infants would be required. Children were treated with SP (Fansidar®, Roche, France) by weight (1 tablet containing 500 mg sulfadoxine and 25 mg pyrimethamine; ½ tablet for weights 4.5–10 kg, 1 tablet for 11–20 kg, 1½ tablets for 21–30 kg). The content and solubility of the SP tablets were confirmed by solubility testing and high performance liquid chromatography at the London School of Hygiene and Tropical Medicine. The study drugs achieved the expected concentrations of SP in solution when compared to controls (Fansidar®, Roche, France) purchased in the UK. SP was given under observation by study staff and children were observed for at least 1 hour after treatment. If a child vomited within 30 minutes of receiving the drug, the full dose was repeated. if a child vomited between 30 minutes and an hour, half the dose was repeated. If a study child vomited the study medication twice, the study child was given rescue treatment and excluded from the study. Rescue treatment for uncomplicated malaria was artemether- lumefantrine (Coartem®, Novartis, Basel, Swizterland) and for severe malaria was parenteral quinine. Children in the symptomatic study were seen at the clinic on days 1, 2, 3, 7, 14, 21, and 28 after treatment and home visits were made for those who failed to report. Parents were encouraged to bring any child who became ill between specified visits to the clinic where they were evaluated and treated by a study clinician thoughout the study period. Malaria blood films and filter paper samples were obtained from children in the symptomatic group at all active and passive follow-up time points. Blood samples were not collected on days 1, 2, 3, and 21 from the asymptomatic infants if the infant remained well. However, blood samples were collected at any time if the infant became symptomatic. The primary end point, parasitological failure by day 28 was defined as (1) development of danger signs or severe malaria, (2) parasitaemia on Day 2 that was higher than that on Day 0, (3) parasitaemia on Day 3 ≥25% of the count on Day 0, (4) parasitaemia on or after Day 4. Failures were further divided into early treatment failures (within day 3 post treatment), late clinical failures (recorded fever plus parasitaemia from day 4 to day 28 post treatment), and late parasitological failure (parasitaemia at day 14 or day 28 post treatment in the absence of fever). In the symptomatic study children with parasitaemia on or after day 4 were treated with rescue treatment if they became symptomatic or until they reached day 28 when all parasitaemic children were treated. In the asymptomatic study any child on or after day 4 with parasitaemia was treated with rescue treatment. The literature search for reports of the A581G mutation up to October 2007 was done using the National Library of Medicine search engines, Pubmed and Medline. The following terms were included in the search queries: dhfr, dhps, sulphadoxine, sulfadoxine, pyrimethamine, Fansidar, Africa, prevalence, malaria and resistance. To be included in the review, articles had to include analysis of codon 581 of the dhps gene in isolates collected from study sites in Africa. Blood smears were stained with 20% Giemsa for 20 minutes and read by two independent microscopists for speciation, and quantification. Parasite density was estimated by counting parasites against 200 White Blood Cells (WBC). A blood smear was considered negative if no asexual forms were seen after observing 500 WBC. Discordant results (33% difference in quantification or positive/negative results) were read by a third microscopist; agreement between any two micoscopists and the average parasite density were deemed to be the correct finding. Parasite counts were adjusted assuming a standard WBC of 8000 per microlitre. DNA was extracted from bloodspots dried on filter papers by soaking overnight in 1 mL. of 0.5% saponin-1× phosphate buffered saline (PBS). The segment was then washed twice in 1 ml of PBS and boiled for 8 min in 100 µL PCR quality water with 50 µL 20% chelex suspension (pH 9.5). dhfr and dhps were PCR amplified and point mutations at codons 51, 59, 108 and 164 of the dhfr gene and codons 436, 437, 540, 581, and 613 of the dhps gene were genotyped using a dotblot methodology previously described by Pearce et al[27]. The probed blots were visualised through alkaline phosphatase-catalysed breakdown of the flourogenic substrate (ECF) (GE Healthcare, Buckinghamshire, UK) and the chemifluorescent signal scanned on a TYPHOON Trio® Phosphoimager (GE Healthcare, Buckinghamshire, UK). The stringency and specificity of the hybridisation process was confirmed by inspection of a series of four controls of known single genotype variant sequence. All blots with non-specifically bound probes were stripped and re-probed. A sequence variant was considered to be present in the PCR product when the intensity of signal was higher than that of the background. The presence, absence, and relative abundance of hybridisation signal was recorded for every probe at each locus. A sample was considered to have a single haplotype when only one sequence variant was found at each locus. Where alternative sequences were present in the same these were designated as a mixed genotype infection. Where mixture was detected at one locus only we inferred a mixture of 2 haplotypes which varied only by that codon. We tested for the presence of mutations in addition to those at codon 436, 437, 540, 581, and 613 of dhps by direct sequencing. PCR products were purified using ExoSAP-IT® (USB Corporation, Cleveland, Ohio, USA). Cycle sequencing was performed using Applied Biosystems BigDye V 3.1 and samples loaded on the ABI-3730 capillary system. Sequence reads were checked by eye and edited using the Seqman (DNAstar Inc., Madison, WI, USA). The presence of SNPs was confirmed by reads through both forward and reverse strands. Recrudescent and new infections were differentiated first by typing size and sequence of the highly polymorphic repeat region of MSP2[28] and then by typing repeat length polymorphism at the PfPK2 microsatellite marker[29]. The size polymorphism of PCR amplified FC27 and IC1 fragments of MSP2, was determined by agarose gel electrophoresis, stained with SYBR® Green 1 (Invitrogen™ Ltd, Paisley, UK) and scanned on a TYPHOON Trio® phosphoimager (GE Healthcare, Buckinghamshire, UK). The gel image was analysed using ‘Imagequant software™’(Molecular Dynamics, Foster City, CA, USA) and fragment sizes were calibrated to known fragment sizes in HyperladderIV (Bioline™, London, UK) which was run in duplicate on every gel. Pre- and post-treatment samples for each patient were compared according to sequence and size of the PCR amplified MSP2 fragments. Recrudescent infections were characterized as having at least one identical allele present in both pre and post treatment samples. Matching alleles were defined as those for which the analysis software estimated the sizes to be within 15 bp of each other. Samples where no alleles matched in the pre and post treatment were classified as new infections. Pre-and post treatment sample pairs which were classified as having a recrudescent infection according to MSP2 matches were then compared at the Pfpk2 microsatellite locus. The Pfpk2 microsattelite repeat was pcr amplified using the protocol described in Anderson et al[29] and fragments were run on an ABi 3730 DNA analyzer (Applied Biosystems, Foster City, USA) with LIZ-500 size standard and analyzed using Genemapper software (Applied Biosystems, Foster City, USA). Any pairs of pre and post treatment samples which matched at MSP2 but did not match at PfPK2 were re-classified as reinfections. If either pre or post sample failed to amplify, they were classified as undetermined. Multiplicity of infection (MoI) was assessed by examining the numbers of alleles detected at MSP2 and pfPK2. Where the number of alleles at these two loci differed, the higher of the two values was used since this is the minimum number co-infecting genotypes which can explain the observed diversity. Data were double entered into an Access (Microsoft Corps, Seattle, USA) data base and analyzed in STATA 9.0 (Stata Corps,Texas, USA). Crude and PCR-corrected rates (excluding new infections and indeterminate PCR) were estimated. A survival analysis[30] was carried out by censoring children at the time of a PCR-corrected new infection or undetermined treatment failure and loss to follow up. Logistic regression was used to determine factors associated with treatment failures (cases) compared to non-failures (controls). The following variables identified a priori were included in the logistic regression model: age, parasite density, study population (asymptomatic or symptomatic) and molecular markers. In addition, any factor that was significant at the 10% level in the crude analysis was included in the model. Further survival and regression analyses were carried out in order to examine the relationship with time to failure and the presence of the mutation 581G in the dhps gene.