High prevalence of mefloquine-resistant falciparum malaria in eastern Thailand.

High prevalence of mefloquine-resistant falciparum malaria in eastern Thailand A.L. Fontanet,l B.D. Johnston,2 A.M. Walker,1 W. Rooney,3 K. Thimasarn,4 D. Sturchler,5 M. Macdonald,6 M. Hours,6 & D.F. Wirth1 In order to assess the risk and predictors of mefloquine resistance we monitored a cohort of 113 patients in eastern Thailand who had been treated for uncomplicated falciparum malaria with a single dose of 15 mg/kg of the drug and followed up for 42 days. The overall treatment failure rate at day 42 was 59.1% (95% confidence interval (CI) = 50%, 68%) with only 2.7% of the patients being lost to follow-up. There were 6.4% RIII, 20.9% RII, 31.8% RI, and 40.9% sensitive responses, based on a modified WHO classifi- cation. A low haemoglobin level on the day of treatment and diarrhoea during the first two days after treat- ment were independent predictors of treatment failure. These findings remained statistically significant in a Cox proportional hazards model, after controlling for other baseline characteristics and adverse effects. Although a history of digestive disorders prior to treatment was associated with diarrhoea on day 2 (P = 0.024), it was in itself not a predictor of treatment failure (adjusted hazard ratio = 1.16; 95% Cl = 0.35, 2.14). A total of 60 patients with an R response were hospitalized for 7 days to receive supervised treat- ment with quinine-tetracycline. Only three had a positive thick smear for asexual forms of Plasmodium falciparum 14 days later, and quinine-tetracycline therefore remains a good alternative treatment for mefloquine-resistant falciparum malaria. Introduction Since 1984 mefloquine has been used to treat uncom- plicated falciparum malaria in Thailand because of its activity against multidrug-resistant falciparum strains and its efficacy in a single dose. The cure rates obtained with mefloquine in the initial field trials of the drug were close to 100% (1-4). Recent studies have revealed, however, that these rates have declined to 70% in eastern and western Thailand (5, 6). Mefloquine was initially given in combination with sulfadoxine and pyrimethamine, which permit- ted its use at a dose of 15 mg/kg body weight (7). However, in July 1990, treatment centres dealing with Khmer displaced persons along the eastern border of Thailand switched from use of the triple combination to that of mefloquine alone. Epidemiology Department, Harvard School of Public Health, 667 Huntington Avenue, Boston, MA 02115, USA. Requests for reprints should be sent to Dr Fontanet at this address. 2 Registered Nurse, Thailand. 3 WHO Laboratory Specialist,, Bangkok, Thailand. 4 Malaria Division, Ministry of Public Health, Bangkok, Thailand. 5 F. Hoffmann-La Roche, Basle, Switzerland. 6 United Nations Border Relief Operations, Aranyaprathet, Thai- land. Reprint No. 5396 Since 1990 a dramatic increase in the incidence of malaria has been reported by the United Nations Border- Relief Operations (UNBRO) involved in supervising the Khmer displaced persons camps (see Fig. 1). This increase may be related to the greater mobility and exposure to malaria-infested areas of the population living in and around the camps caused by the changing military and political situations. Emergence of mefloquine resistance may also have Fig. 1. Monthly incidence of malaria (all types) in Site 8, 1988-91. (Source: UNBRO) 2000 1500~~~~~~~~~~~~~~~~~~~~~00 1500-Z >l 0 JFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJ 1988 1989 1990 19911 Months Bulletin of the World Health Organization, 71 (3/4): 377-383 (1993) © World Health Organization 1993 377 A.L. Fontanet et al. contributed to this phenomenon, and treatment fail- ures have been reported by the medical staff respon- sible for refugee care. We therefore initiated an in vivo test of mefloquine efficacy to estimate the drug failure rate at the current dose of 15 mg/kg. Patients and methods Study site The study was carried out in Site 8, a Khmer dis- placed persons camp on the Thai-Cambodian border. The camp has a population of 45 000, 40% of whom are under 10 years of age; it is also used as a referral hospital for patients for Cambodia. Accord- ing to data from the outpatient department and the hospital, malaria is the leading cause of morbidity and mortality among the camp's inhabitants. The monthly malarial incidence typically varies be- tween 50 and 2000 cases, depending on the season. Most of the cases (90%) are due to Plasmodium fal- ciparum, with the remainder being P. vivax infec- tions. Uncomplicated falciparum malaria cases are treated with 15 mg/kg of mefloquine (Lariam, Roche) in a single dose; severe cases are treated with intravenous quinine for up to 7 days, which is replaced with oral quinine and tetracycline when the patient is able to swallow tablets. Mosquito collections made over the period 1983-85 showed that areas close to the camp were heavily infested with both of the region's malaria vectors-Anopheles dirus and A. minimus. In Site 8 itself it was very rare to encounter either vector in the mosquito light-trap collections until recently, when an occasional A. dirus was collected and A. minimus became much commoner. Study procedure Inclusion of patients. Patients treated with meflo- quine for uncomplicated falciparum malaria at the outpatient department of the camp in August and September 1991 were eligible for inclusion in the study. The following were contraindications for meflo- quine treatment: cerebral symptoms; haemoglobin level <7 g/dl; high parasitaemia (>10 parasites per thick film field); and pregnancy. Treatment consisted of administration of 15 mg/kg of mefloquine in a single dose for patients weighing <40 kg; those whose weight was >40 kg were given three tablets (750 mg) in a single dose. Immediately after treatment, the informed consent to participate in the study was obtained from the patients with the help of a translator. Patients were recruited for the study if they agreed to remain available for follow-up for 42 days and if they were willing to have further blood samples taken. The patients had just received therapy when their consent for follow-up was sought and were aware that their subsequent care did not depend on participating in the study. This study design was approved by the ethics committees of both UNBRO and Harvard School of Public Health. All the thick smears were examined by an expert microscopist from the Thai Malaria Division, and patients were retained for the study if they had >1600 asexual forms of P. falciparum per,l of blood and no mixed falciparum and vivax infections. Finally, patients were included in the study only if they had had no history of treatment with mefloquine or quinine during the previous month. Routine clinical information was collected while the patients were observed for 1 hour after treatment, in order to record accurately whether or not they vomited. Follow-up. Study subjects were asked to return to the outpatient department on days 2, 3, 7, 14, 21, 28 and 42. At each visit any malaria symptoms and side- effects were recorded, and a thick blood smear was stained with Giemsa. Patients were told to report to the outpatient department if they experienced any fever or symptoms of malaria outside the regular follow-up days. Treatment responses were labelled RI, RII, RIII or sensitive, using a slightly modified WHO classifi- cation (RI: all patients with a negative slide on day 7, but which became positive again for asexual forms of P. falciparum between day 8 and day 42; RII: patients with a slide still positive for asexual forms of P. falciparum on day 7; RIII: patients whose para- sitaemia on day 2 had decreased by <75% of that on day 0; and sensitive: patients who remained free of parasitaemia after day 6). R is used subsequently in this article to indicate either an RI, RII or RIII res- ponse. Any positive slide on day 7 was confirmed by examination of a new slide on day 9. Patients with an R response were considered to be symptomatic if they had either a recorded temperature of > 37.5 °C or complained of the reappearance of at least two of the following symptoms: fever, headache, and joint pain. All patients with R responses were treated oral- ly in hospital for 7 days with quinine (30 mg/kg/day) and tetracycline (25 mg/kg/day). A final thick smear was performed 14 days after the first dose of quinine had been administered to confirm the cure. Patients with an RIII response were treated on day 2 only if they were symptomatic at that time; otherwise, a daily smear was taken until the parasitaemia cleared. Patients with an RII response were treated only if they were symptomatic or if the control slide at day 9 was positive. WHO Bulletin OMS. Vol 71 1993378 Mefloquine-resistant falciparum malaria in eastern Thailand A total of 5 ml of venous blood was collected from adult patients who had an R response before day 14, after their further consent had been obtained. Serum was separated from the blood by centrifuga- tion, kept on ice, and stored at < 4 'C. A hand centri- fuge was used and separation of the serum may therefore have been incomplete for samples from some patients. For each of these patients, a control was selected in the cohort from a patient of the same age (±5 years), sex, weight (±5 kg), day of treatment, and who had a negative smear. The permission of the control patient was obtained before 5 ml of venous blood was collected to determine the drug level in serum. The control patient's status was maintained even if he/she exhibited an R response later in the follow-up. Serum levels of mefloquine were deter- mined using high-performance liquid chromatogra- phy (HPLC). Statistical analysis. Differences in proportion were analysed using x2 and Fisher's exact tests, and means were compared using Student's t-test. A multivariate analysis was performed using linear regression, logistic regression, and Cox proportional hazards models. Table 1: Baseline characteristics subjects No. of males No. of females No. in age group: <14 years .15 years Mean age (years) No. of documented mefloquine treatments in past year:8 0 1-2 3 No information No. with history of vomiting or diarrhoea prior to treatment Mean haemoglobin level (g/dl) Geometric mean parasitaemia (per,u) Mean mefloquine dose (mg/kg) of the 113 study 95 18 8 105 32 ± 11.3 53/105 (50)b 38/106 (36) 15/106 (14) 7 33/113 (29.2) 10.8 ± 1.6 33 000 14.32 ± 1.92 Results Patients' baseline characteristics A total of 113 patients were enrolled in the study in August and September 1991; their baseline character- istics are summarized in Table 1. The small number of children (8/113) in the cohort could have arisen because of the severity of their disease, leading to intravenous or oral treatment with quinine, and hence making them ineligible for the study. A history of vomiting or diarrhoea prior to treat- ment was common (29.2% of patients) and was signifi- cantly associated with parasitaemias >100 000 asexual forms of P. falciparum per gl of blood (P = 0.007). The relationship between haemoglobin level at admission and the other variables was examined using tabular analyses and linear regression pro- cedures. Women and recent camp arrivals had lower haemoglobin levels than the other study sub- jects (P = 0.002 and P = 0.01, respectively). Al- though there was little association between haemo- globin level and the number of mefloquine treatments in the previous year, there was evidence that individ- uals who had received the drug within the previous 120 days had lower haemoglobin levels (P = 0.016). Adverse effects A total of 6 of the 113 patients (5%) vomited within 1 hour of receiving mefloquine; four were treated a Shown in the right-hand column are the number of subjects. b Figures in parentheses are percentages. again because the drug appeared to have been regurgi- tated in the vomitus. On day 2, dizziness was the commonest com- plaint of the patients (79%). Vomiting and diarrhoea occurred almost equally frequently (20% and 25%, resp.) and both were significantly associated with a history of digestive disorders prior to treatment (P <0.001 and P = 0.024, resp.) (Table 2). There was no association between the dose of mefloquine received by the patients (mean, 14.32 ± 1.92 mg/kg) and the occurrence of side-effects, even after controlling for age, sex, and initial parasitaemia using a logistic regression model. Table 2: Association between digestive disorders before and on day 2 after treatment with mefloquine Vomiting Diarrhoea Digestive on day 2:a on day 2:b disorders Yes No Total Yes No Total Yes 13 20 33 13 20 33 No 8 66 74 14 60 74 Total 21 86 107 27 80 107 a P= 0.001 (X2 test).b p= 0.024 (x2 test). WHO Bulletin OMS. Vol 71 1993 379 A.L. Fontanet et al. Mefloquine failure Only 3 of the 113 patients (2.7%) did not complete the follow-up. Information provided by their families or neighbours indicated that they were not sympto- matic when they left the cohort. These patients were excluded from subsequent analyses. The overall parasitological failure rate among the remaining patients was 59% (65/110) (95% confidence interval (CI) = 50, 68%). On the basis of a modified WHO classification, there were 7 RIII (6.4%), 23 RII (20.9%), 35 RI (31.8%), and 45 sensi- tive cases (40.9%). Over the whole follow-up period, 34 of the 43 patients (79%) with a positive slide on day 3 showed an R response. A total of 35 of the 66 patients (53%) with a negative slide on day 3 were sensitive cases. All the patients with a positive slide on day 7 were checked again on day 9 and, with the exception of two, remained positive (25/27). A total of 35 of 56 patients (63%) with an R response were symptomatic when treatment with quinine was initiated. The levels of parasitaemia were not significantly different for symptomatic and asymptomatic patients. We analysed the following baseline characteris- tics and recorded side-effects as predictors of an R response using a Cox proportional hazards model: age; gender; number of malaria episodes in the past year; number of documented mefloquine treatments in the past year; number of days since the last meflo- quine treatment; history of recent arrival in the camp; history of a trip outside the camp during the past month; number of days of illness prior to the consultation; history of digestive disorders prior to treatment; fever at the consultation; parasitaemia on day 0; haemoglobin level on day 0; mefloquine dose; diarrhoea on day 2; vomiting on day 2; and dizziness on day 2. There were two independent and important pre- dictors of an R response: the haemoglobin level at day 0; and diarrhoea in the first two days after treat- ment with mefloquine. The probability of resistance decreased with increasing haemoglobin levels and was higher for persons who experienced diarrhoea on day 2 (21/27 = 78%) than for those who did not (42/80 = 53%). Table 3 shows the results of a Cox proportional hazards model that incorporates these two predictors together with age, which had a non- significant protective effect. Patients who had digestive disorders before treatment with mefloquine did not appear to have a higher risk of an R response. The hazard ratio asso- ciated with prior digestive disorders, estimated using a model that incorporated the predictors listed in Table 3, with a term for prior digestive troubles, was 1.16 (95% CI: 0.35, 2.14). Table 3: Predictors of mefloquine resistance Identified in the study Adjusted Predictor hazard ratioa Age (years) 0.98 (0.96, 1.00)b Haemoglobin level on 0.80c day 0 (g/dl) (0.67, 0.96) Diarrhoea on day 2 1.92d (1.09, 3.39) a For age and haemoglobin level the adjusted hazard ratio esti- mates the amount by which the hazard of failure is multiplied for an increase of one unit in these variables. b Figures in parentheses are the 95% confidence intervals. c P = 0.016. d p = 0.024. Using a similar model, we found that a positive malaria smear on day 3 was significantly associated with treatment failure after controlling for the other variables (adjusted hazard ratio = 2.22 (95% CI: 1.27, 3.87)). Serum mefloquine concentrations The serum mefloquine concentrations of 10 patients with an R response during the first 14 days of follow- up were compared with those of 10 controls, match- ed for age, sex, weight, and day of treatment. The mefloquine levels were almost identical for four of the pairs; for four pairs the levels were higher in the controls; and for two pairs were higher in the cases. Differences in mefloquine serum concentrations were much larger and in favour of the controls for the two pairs from whom blood was taken early in the follow-up period (day 3). A total of 10 of the 13 patients with recrudescence between day 7 and day 14 had a mefloquine level >500 ng/ml at recrudes- cence; the mean serum mefloquine concentrations of these 13 patients was 646 ng/ml. For 26 patients there was no correlation between the haemoglobin level on day 0 and the mefloquine serum concentrations on day 3 to day 18. Samples of blood were taken on day 3 from two patients who had diarrhoea on day 2. On day 3 the serum mefloquine concentrations for these patients were much lower than those of three patients who did not have diarrhoea (227 and 499 ng/ml compared with 1455, 1693, and 2922 ng/ml). Subsequent course of patients showing resistance Three of the seven RIII cases were symptomatic and were treated intravenously on day 3 with quinine. The remaining four cases were followed up without further therapy until recrudescence occurred (day 21 (two), day 23, and day 28). WHO Bulletin OMS. Vol 71 1993380 Mefloquine-resistant falciparum malaria in eastern Thailand A total of 60 patients were treated again orally with quinine (30 mg/kg/day) + tetracycline (25 mg/ kg/day) for 7 days in hospital; a control malaria smear was obtained for all of them (14 days after the first day of quinine for 44 patients and between day 14 and day 42 for the others). Only three were positive for P. falciparum and one for P. vivax. Fal- ciparum cases were re-treated with mefloquine (25 mg/kg). Of the five patients who did not want to be hos- pitalized, three received a single dose of 25 mg/kg of mefloquine; their control smears on day 14 were negative. The other two were mistakenly given only 15 mg/kg of mefloquine; one of them who had a positive control smear was successfully treated with quinine-tetracycline over 7 days. Discussion Resistance against many widely used antimalarial drugs has begun to emerge in south-east Asia (8). In vitro resistance to mefloquine was described in eastem Thailand even prior to the introduction of the drug as a first-line treatment for falciparum ma- laria (9). However, the first trial of mefloquine on the Thai-Cambodian border reported an excellent cure rate (3). Only recently has the susceptibility of some falciparum strains in this area decreased to such an extent that even adequate mefloquine concen- trations in blood are insufficient to clear parasitaemia (10). This finding, together with a recent increase in the number of malaria cases in some Khmer refu- gee camps, led us to initiate the in vivo test of meflo- quine efficacy that we have reported' in this article. The overall risk of treatment failure in our follow- up of 110 patients was 59%. The proportion of patients who dropped out was only 2.7% and is insufficient to explain the high risk of failure as an artefact due to the elimination of healthy patients from the cohort (i.e., if healthy patients are lost to follow-up, the proportion of treatment failures is falsely increased). Some patients may have been reinfected during the follow-up; however, the early treatment failures, i.e., RII and RIII responses (which cannot be re-infections), accounted for a large pro- portion of the total. Such a high level of clinical resistance among adult males, whose immunity is likely to be greater than that of women and children, is particularly disturbing. Doses of 25 mg/kg of mefloquine, which are designed for the treatment of non-immunes, could be adopted for semi-immunes in areas of multiresistance to increase cure rates. A clinical trial of mefloquine at doses of 15 mg/kg and 25 mg/kg is in progress in the study area. However, we are concerned that patients with an R response had a mefloquine serum concentration > 500 ng/ml, i.e., a level that should be associated with successful treatment (11). Digestive disorders prior to treatment were com- mon and significantly associated with higher parasi- taemia. Patients with such disorders were more likely to report vomiting or diarrhoea on day 2. However, they were not at increased risk of failure, as shown by the Cox proportional hazards model (adjusted hazard ratio = 1.16; 95% CI = 0.35, 2.14). This indi- cates that disease-induced diarrhoea may not de- crease mefloquine absorption enough to impair the drug's efficacy but that drug-induced diarrhoea appears to do so. The mefloquine serum concentra- tions on day 3 for the two patients with diarrhoea were very low indeed. The inverse relationship between haemoglobin level and the risk of treatment failure is interesting. We initially hypothesized that people infected with a resistant strain would develop anaemia through repetitive malaria attacks, and it appears that recent prior therapy with mefloquine was associated with lower haemoglobin levels. Alternatively, mefloquine may be more rapidly cleared from the blood of patients with anaemia since the drug is less able to fix on their red blood cells (12). It should be noted, however, that serum mefloquine concentrations did not correlate with initial haemoglobin levels for the 26 patients for whom both were measured; the hy- pothesized increased rates of mefloquine elimina- tion in anaemic patients therefore remain speculative. Quinine-tetracycline was still effective against mefloquine-resistant falciparum strains (there were only three recrudescences on day 14 among the 60 re-treated patients). However, at the time when they received quinine-tetracycline, some of the patients may still have had mefloquine in their blood, since its half-life is long. This should therefore not be con- sidered as an evaluation of the effect of quinine- tetracycline alone on mefloquine-resistant strains. Mefloquine has been a convenient treatment for malaria in the refugee camps on the Thai-Cambodian border since its single-dose administration can easily be controlled. Over the past 5 years, high-grade resistance has developed, however, and other drug regimens are now urgently needed. Quinine- tetracycline remains a valuable alternative for the treatment of mefloquine-resistant falciparum malaria, but the more widespread use of quinine should not be allowed to jeopardize, by creating drug resistance, its intravenous use as a life-saving drug for severe malaria. Acknowledgements We thank the Khmer staff in Site 8 camp's hospital for participating in the study; Mr Kianarong Kawpasit for WHO Bulletin OMS. Vol 71 1993 381 A.L. Fontanet et al. analysing the slides; the staff of the Christian Alliance Mission for collaborating in managing the patients; and the United Nations Border Relief Operations for technical assistance. Dr Y. Bergqvist is thanked for determining the blood levels of mefloquine by high-performance liquid chromatography. Dr A. Spielman and Dr U. Brinkmann are thanked for their advice on the study design and interpretation of the findings. Finally, Ms N. Jain and Mr F. Hopkins are thanked for their contributions during the preparation stages of the study. The study was supported by grants from the Pharmaco- epidemiology Training and Research Fund, Hoffman- La Roche, and the Julian Mack Funds. A.L. Fontanet is a recipient of a Lavoisier and a Fondation MWrieux scholarship and B.D. Johnston, of a Rotary Foundation scholarship. Resume Forte pr6valence du paludisme A falciparum r6sistant & la m6floquine dans I'est de la Thailande Une cohorte de sujets traites par la mefloquine dans l'est de la Thallande a ete suivie afin d'eva- luer le risque et les facteurs predictifs de resistan- ce a ce medicament. A cette fin, nous avons traite 113 malades atteints de paludisme a falciparum sans complication avec 15 mg/kg de mefloquine en dose unique, et nous les avons suivis pendant 42 jours. Au jour 42, le taux global d'echec etait de 59,1% (intervalle de confiance (IC) a 95%: 50- 68%), 2,7% seulement des malades ayant ete perdus de vue. Selon une classification OMS modifiee, nous avons observ6 6,4% de cas de resistance Rill, 20,9% de resistance Rhl, 31,8% de resistance RI et 40,9% de sujets sensibles. Au jour 2, les malades se plaignaient princi- palement de vertiges (79%) tandis que 20% souf- fraient de vomissements et 25% de diarrhee. Deux variables independantes permettaient de predire l'echec du traitement: un faible taux d'hemoglobine le jour du traitement, et la surve- nue d'une diarrhee dans les deux jours suivant l'administration du medicament. Ces observations ont conserve leur signification statistique dans un modele de risque proportionnel de Cox aprbs cor- rection des autres parametres de base et des effets ind6sirables. Pour expliquer la relation inverse entre le taux d'hemoglobine et le risque d'echec therapeutique, nous avons suppose que les sujets infectes par une souche resistante de Plasmodium falciparum pr6sentaient une anemie due a des crises repetees de paludisme. Bien que l'association soit faible entre le taux d'hemoglobi- ne et le nombre de traitements par la mefloquine que les sujets avaient requ au cours de l'annde precedente, les malades les plus recemment trai- tes semblaient avoir les taux d'hemoglobine les plus faibles. Nous avons 6galement recherche s'il existait un taux accru d'elimination de la mefloqui- ne chez les sujets anemiques. Mais les taux seriques de mefloquine n'etaient pas correles au taux initial d'hemoglobine chez les 26 sujets chez lesquels ces deux taux ont ete mesures. Des ant6c6dents de troubles digestifs avant le traite- ment etaient associes a la survenue d'une diar- rhee au jour 2 (P = 0,024), mais ne constituaient pas en eux-memes un facteur predictif de l'echec du traitement (rapport de risque corrige = 1,16; IC 95% = 0,35-2,14). Cela laisse a penser que la diarrhee due a une maladie pourrait ne pas redui- re I'absorption de la mefloquine suffisamment pour nuire a son efficacite, alors que la diarrh6e iatrogene pourrait le faire. Au total, 60 malades pr6sentant une resistan- ce au medicament ont ete hospitalises pendant 7 jours et ont requ un traitement par la quinine et une tetracycline. Seuls trois d'entre eux avaient un frottis en goutte epaisse positif pour les formes asexuees de P. falciparum au bout de 14 jours. Le traitement associant la quinine a la tetracycline reste une bonne alternative chez les sujets atteints de paludisme a falciparum r6sistant a la mefloquine dans la zone etudi6e. References 1. Doberstyn, E.B. et al. Single-dose therapy of falci- parum malaria with mefloquine or pyrimethamine- sulfadoxine. Bulletin of the World Health Organiza- tion, 57: 275-279 (1979). 2. Harinasuta, T. et al. Phase-l1 clinical trial of meflo- quine in patients with chloroquine-resistant falcipa- rum malaria in Thailand. Bulletin of the World Health Organization, 61: 299-305 (1983). 3. Meek, S.R. et al. Treatment of falciparum malaria with quinine and tetracycline or combined mefloquine- sulfadoxine-pyrimethamine on the Thai-Kampuchean border. American journal of tropical medicine and hygiene, 35: 246-250 (1986). 4. Nosten, F. et al. Malaria on the Thai-Burmese border: treatment of 5192 patients with mefloquine- sulfadoxine-pyrimethamine combination. Bulletin of the World Health Organization, 65: 891-896 (1987). 5. Boudreau, E.F. et al. Mefloquine kinetics in cured and recrudescent patients with acute falciparum malaria and in healthy volunteers. Clinical and phar- macological therapeutics, 48: 399-409 (1990). 6. Nosten, F. et al. Mefloquine-resistant falciparum malaria on the Thai-Burmese border. Lancet, 337: 1140-1143 (1991). 7. Malikul, S. The current situation of the anti-malarial programme in Thailand. South-east Asian journal of tro- pical medicine and public health, 19: 355-359 (1988). 382 WHO Bulletin OMS. Vol 71 1993 Mefloquine-resistant falciparum malaria in eastern Thailand 8. Peters, W. The problem of drug resistance in ma- laria. Parasitology, 90: 705-715 (1985). 9. Boudreau, E.F. et al. Type-Il mefloquine resistance in Thailand. Lancet, 1: 1355 (1982). 10. Childs, G.E. et al. In vitro and clinical correlates of mefloquine resistance of plasmodium falciparum in eastern Thailand. American journal of tropical medi- cine and hygiene, 44: 553-559 (1991). 11. Slutsker, L.M. et al. Mefloquine therapy for Plasmodium falciparum malaria in children under 5 years of age in Malawi: in vivo/in vitro efficacy and correlation of drug concentration with parasitological outcome. Bulletin of the World Health Organization, 68: 53-59 (1990). 12. Chevii, R. & Fitch, C.D. The antimalarial drug mefloquine binds to membrane phospholipids. Anti- microbial agents and chemotherapy, 21: 581-586 (1982). WHO Bulletin OMS. Vol 71 1993 383