In vivo drug resistance of falciparum malaria in mining areas of Venezuela
Dr A. Aché, Residencias ‘La Macareña’, Calle Principal, No. 2. El Mácaro, Estado Aragua, Venezuela. E-mail: firstname.lastname@example.org
The Lot Quality Assurance Double-Sampling Plan (LQADSP) technique was used in three areas, Maripa, Kilómetro 88 and Ikabaru, to assess the efficacy of antimalarials used routinely by the VenezuelanMalaria Programme. The use of chloroquine (25 mg/kg), chloroquine (40 mg/kg) and the combination of sulfadoxine (500 mg) and pyrimethamine (25 mg) registered treatment failures above the threshold level of 25% in Maripa and Kilómertro 88. In Ikabaru the use of chloroquine (40 mg/kg) did not surpass that quality level and could possibly be less than 10%. Quinine (30 mg/kg) was totally effective in curing patients in all three areas. The use of this technique seems adequate for rapid field evaluations and in this case for providing appropriate information to assist this health programme. However, whilst being an ideal technique for surveying areas in which considerable variation may exist among lots and particularly for Plasmodium falciparum infections in these areas, repeated surveys should be carried out in the same areas over time to monitor changes in the susceptibility of this parasite to first-, second- and third-line drugs. In that way, national drug policies can be modified adequately.
Since Maberti made the first field observation on simultaneous resistance of Plasmodium falciparum to pyrimethamine and chloroquine, in the State of Trujillo, western Venezuela, in July 1959 (Maberti 1960; Bruce-Chwatt 1970; Peters 1970), many years elapsed before this problem was again mentioned in the early 1970s. In those years few reports of treatment failures with chloroquine were made by the regional malaria programme managers in the States of Amazonas, Bolívar and Delta Amacuro, southern Venezuela, where only P. falciparum infections occur. These failures were detected in local diamond miners and the migrant population, who seasonally prospected at the mining site of Guaniamo in Bolívar. In 1975, treatment failure was confirmed for chloroquine (Godoy et al. 1975); later on, local strains of P. falciparum were shown to be resistant to sulfa and pyrimethamine in that same state (Godoy et al. 1977). Clinical failures with chloroquine were treated by incrementing the dosage in accordance to programme's therapeutic regimens. However, even as reports increased of such events from remote and scattered areas, they reflected temporal patterns with higher numbers of treatment failures during the rainy seasons, that wore off during the dry season.
To this day, chloroquine remains the recommended first-line drug because adequate response is still observed in P.falciparum prone areas and also Plasmodium vivax infections amount to 60–65% of the cumulative annual incidence cases. The combination of sulfadoxine-pyrimethamine is second-line and quinine the third-line treatment. In the first half of the 1990s, small in vitro studies carried out in Amazonas had confirmed therapeutic failures with the use of chloroquine, amodiaquine and mefloquine (Maynadie et al. 1989; Magris & Riggione 1994). To respond to growing concerns about the evaluation of the efficacy of malaria treatments in Venezuela and South America, this in vivo field study was carried out in three sites, in the State of Bolívar, following a standard World Heath Organization (WHO)/PAHO (Pan-American Health Organization) protocol for the assessment of antimalarials for uncomplicated P. falciparum in the Americas [WHO/MAL (Malaria) 1996; Organización Mundial de la Salud/Organización Panamericana de la Salud (OMS/OPS) 1998]. This paper reports that the proportion of parasitological failures for chloroquine and the combination of sulfadoxine-pyrimethamine, both of which were higher than a prefixed cut off point of 25% inonly two of three chosen locations. No clinical failureswere registered with the use of quinine in any of these areas.
In order to update knowledge on the therapeutic efficacy of drug regimens being used by the national Malaria Programme the application of this protocol was undertaken with the ultimate objective of ascertaining continued usefulness or awareness of failures in routine treatments of uncomplicated falciparum malaria (WHO/MAL 1996; OMS/OPS 1998).
Materials and methods
Three municipalities in the State of Bolívar were chosen because they normally exhibited the highest registries for malaria during 1996–1998. The Annual Parasite Incidence (API) for 1999 did not vary greatly from previous years between them: Sifontes (106‰), Sucre (103‰) and Gran Sabana (74‰); and the proportion of malaria infections caused by P. falciparum was 43.3%, 49.1% and 26.1%, respectively. The annual malaria-specific death rate for the state during these years averaged 0.09/1000 inhabitants; all confirmed by thick blood smears and autopsies. In Sifontes, south-east of Bolívar, the selected site was Kilómetro 88 (Km 88) (Figure 1) because it serves as a central point to accrue active and passive malaria cases amongst gold miners from at least 10 nearby mining sites; as well as Pemones, members of an indigenous ethnic group and the influx of Amerindians because of its proximity to the Republic of Guyana (average 10–12 h of jungle walk). However, Km 88 provided adequate logistics (electricity for microscopes and was 172 km from the nearest hospital for the processing of haematological samples). In the municipality of Sucre, Maripa was the second site chosen because it also serves as a central point for the detection of active and passive malaria cases for agricultural settings and for miners reaching its port via the Caura river from nearby mining areas. It also provided electricity for thick smear examinations. The third site is Ikabaru, municipality of Gran Sabana, on the border to Brazil, which serves as a confluence point for many mining areas nearby. Scattered Pemon communities are to be found amongst these mining areas and there is a heavy influx of Brazilian miners. As this site did not have continuous electrical service, blood samples were sent daily by air to the hospital of Santa Elena de Uairén during the study period. Thick-blood smears were all examined in Ikabaru.
The standard WHO/PAHO protocol for the assessment of antimalarials for uncomplicated P. falciparum in the Americas (WHO/MAL 1996; OMS/OPS 1998) consists of recording essential patient information, clinical assessment, body temperature, parasitaemia, body weight on Day 0 (prior to treatment), supervised treatment with the stipulated drug, clinical assessment with examination of body temperature on Days 1, 2, 3, 7, 14, 21, and 28, and parasitological examination (thick smear) on Days 3, 7, 14, 21 and 28. It has the purpose of determining the practical efficacy of particular drug regimens offered by national programmes for uncomplicated falciparum malaria. In this particular case, the first- (chloroquine), second- (sulfadoxine-pyrimethamine) and third-line drugs (quinine), offered by the Venezuelan Malaria Programme (VMP), were each tested separately and sequentially in three areas.
Overall classification of therapeutic responses are to be recorded in three categories: early treatment failures (ETF), late treatment failures (LTF) and adequate clinical response (ACR) as follows (WHO/MAL 1996; OMS/OPS 1998). Early treatment failure (ETF): Development of danger signs of severe malaria on Days 1, 2 or 3, in the presence of parasitaemia; or parasitaemia on Day 2 > Day 0; or parasitaemia on Day 3 μ 25% of count of Day 0. Late treatment failure (LTF): Development of danger signs or severe malaria in the presence of parasitaemia on Day 3; or earlier return of patients because of the development of danger signs in the presence of parasitaemia; or presence of parasitaemia (same parasite) on Days 7, 14, 21 or 28. Adequate clinical response (ACR): Absence of criteria of ETF and LTR, as well as parasitaemia during follow-up period.
This protocol was carried out under the responsibility of qualified medical personnel whose first responsibility was the welfare of the patients enrolled in the test, with the verbal approval of the Venezuelan Ministry of Health and the Venezuelan Malaria Programme as part of regular surveillance activities. It was conducted in accordance to WHO Guidelines for Good Clinical Practice (WHO 1995). All patients enrolled were required to sign an informed consent letter (Annex 7, WHO/MAL 1996; Anexo 7, OMS/OPS 1998).
Only individual incident malaria cases to P. falciparum were enrolled randomly and unblinded for each treatment, as they were detected by active and passive surveillance, and after having accepted voluntarily to participate in the trials. Patients once infected and treated with one drug were not re-enrolled for a different treatment. Treatment failures to one drug were registered as such and when given an alternative successive treatment it was not recorded as an adequate clinical response. Enrolment ceased once the critical values established for treatment failures were overcome in the first or second stage of the sampling technique (Table 1). Patients who were not enrolled in the trials were given the same drugs in line with the VMP drug policy and according to their clinical status.
Table 1. Table for calculating the minimal sample size according to the two stage Lot Quality Assurance Method with confidence level of 95% and power 80%
|0.050|| 21|| 1|| 16|| 0|
|0.075|| 29|| 3|| 16|| 0|
|0.100|| 42|| 5|| 16|| 0|
|0.125|| 63|| 10|| 16|| 0|
|0.150|| 103|| 18|| 26|| 1|
|0.175|| 190|| 37|| 48|| 5|
|0.200|| 441|| 95||110||17|
Subjects were enrolled in Maripa and Ikabaru during the peak falciparum incidence period which occurred during July to September in 1999 and in Km 88 during that corresponding period in 2000. All patients underwent a pre-treatment examination in which they were checked for fever, parasitaemia (thick smear) and haematocrit/haemoglobin and given a full clinical examination. Inclusion criteria were: age older than 6 months; mono-infection with P. falciparum with a parasitaemia in the range of 500–5000 asexual parasites per μl; absence of fever in the 72 h before examination; axillary temperature < 39.5 °C; no general danger signs or signs of severe and complicated falciparum malaria according to definition given by WHO; ability to come for the stipulated follow-up visits, and easy access to the health facility; absence of febrile conditions caused by diseases other than malaria; absence of severe malnutrition; Informed consent of patient/guardian; absence of history of hypersensitivity reactions to sulphonamides or any other drugs; absence of skin conditions which could increase the risk of severe adverse reactions to the scheduled drug; haemoglobin value above 5.0 g/dl or haematocrit above 15%.
Microscopic blood examination
Preparation and staining of the blood slides followed the procedures outlined in Basic Malaria Microscopy, Part I (WHO 1991). Two thick smears were taken: one for adequate parasitaemia enrolment which required at least one parasite for every 12–16 white blood cells, corresponding to about 500 asexual parasites per μl; a second smear was used to calculate the parasite density. Parasitaemia was measured counting the number of asexual parasites against a mean count of 6000 leucocytes per μl. This mean count was taken as more appropriate for human populations in the Americas and was introduced as a modification of WHO's protocol for the assessment of antimalarials for uncomplicated P. falciparum in the Americas (OMS/OPS 1998). The number of asexual parasites was counted against 500 leucocytes using hand tally counters; once started a field was always counted to the end. Therefore, usually final leucocyte counts would be over 500. If 500 parasites were counted without reaching 500 leucocytes, the count was stopped after completing the reading of the last field, and parasitaemia was calculated according to the following formula:
In addition, 100 fields of the second thick film were examined for the exclusion of mixed infections. The presence of P. falciparum gametocytes was noted but did not figure in the evaluation of the test.
Haematological assessment was done by measuring haematocrit (Hto) and haemoglobin (Hb). The haematocrit was measured by means of the microhaematocrit method described by Lévy-Lambert (1974) and haemoglobin levels were determined by colour scales using the card test of Stott & Lewis (1995). Haematocrit and haemoglobin measurements were not possible in the Maripa area due to a lack of laboratory facilities. Nevertheless it was considered important to carry out the therapeutic tests to enable the VMP to have information on drug failures in that area.
Drugs and formulations
The drugs employed for therapeutic efficacy testing were purchased at Pharma® (Milan, Italy). The following drugs and formulationswere used for the trials, each one jointly with Primaquine (0.75 mg base/kg), in accordance to specific VMP drug policy: choroquine as first-line treatment, tablets of 150 mg base as phosphate; sulfadoxine-pyrimethamine as second-line treatment, tablets 500 mg S + 25 mg P; quinine as third-line treatment, tablets of 500 mg salt as sulphate.
Two oral treatments with chloroquine were given: (1) 25 mg/kg body weight daily for 3 days (2) 40 mg/kg body weight daily for 4 days. Sulfadoxine-pyrimetamine was given orally as single dose equivalent to 1.25 mg pyrimethamine/kg body weight. Quinine was given orally at a dose of 30 mg/kg body weight/day, divided into three doses, at 8-h intervals and for 7 days.
In the Ikabaru area it was decided to use chloroquine only at a dose of 40 mg/kg body weight daily for 4 days since malaria drugs were freely sold by foreign miners and it was assumed that chloroquine resistance may exist, although there had been no previous reports, and in view of limited transportation by light aircrafts to Santa Elena.
A lot is defined as the population area under active and passive surveillance by the VMP and which is assigned to a health centre from which malaria control activities are carried out continuously.
The Lot Quality Assurance Double-Sampling Plan (LQADSP) was used in all selected sites (Lemeshow & Taber 1991) and sample sizes for all three lots depended on the following parameters: Po = 0.25 – upper threshold level of 25% of clinical failures as a starting point for these trials by the VMP; beyond which replacement of the drug under study is deemed necessary as suggested by WHO/PAHO protocols (WHO/MAL 1996; OMS/OPS 1998); Pa = 0.10 – lower threshold level of 10% of clinical failures below which it would be more acceptable to continue the utilization of the present drug; α = 0.05 and the probability of concluding that a chosen lot has a low incidence of clinical failures when, in fact, it has a high level (type 1 error); β– probability of concluding that a chosen lot has a high incidence of clinical failures when, in fact, it has a low level (type II error), 1 –β the desired power of the test = 80% to enable the proper detection of sites with clinical failures. These pre-fixed parameters for a two-stage sampling accounted for 16 patients (n1 = 16) tobe enrolled in the first stage for a critical value equal to zero (d1 = 0) and a total of 42 patients (n1 + n2 = 42) tobeaccrued in the second stage for a critical value of 5 (d2 = 5) for each three lots selected for the study. Sampling was stopped once d* + 1 cases were observed for any of the treatments tested (Table 1).
The sampling frame for the three lots of Maripa, Ikabaru and Km 88 consisted of five, seven and four communities, respectively, and all having similar coverage of control methods by the VMP. In order to assure that a random selection of subjects in these sampling units covered fairly all communities, patients were selected by a systematic stratified sample in accordance to the weighted proportion of falciparum malaria cases, registered in each community, in 3 years previous to this study.
Since sample sizes used in LQADSP are too small to provide confidence intervals or any ‘lot’ specific parameter estimate, analyses were based on the hypothesis that the observed number of treatment failures in the samples, selected from each of the three sites, will be less than or equal to a specific critical value (d1 = 0) in the first stage to conclude that the actual proportion of clinical failures is significantly less than 25% (Po) in each lot for the corresponding site. If the observed number of clinical failures in the first stage sample (n1 = 16) is greater than 5 (= d2), it will be concluded that the actual proportion of clinical failures in the population of the selected sites is not significantly less than Po and conclude that the proportion of treatment failures is above 25%. On the other hand, if the observed number of clinical failures in the first stage is larger than (d1 = 0), but less than (d2 = 5), enrolment will proceed to the second stage of sampling. In the second stage, we continue sampling until either d2 + 1 clinical failures are registered, indicating that treatment failures are above 25%, or if out of the total 42 patients (n1 + n2) being enrolled without d2 clinical failures being exceeded as an indication of a low proportion of treatment failures (< 10%) (Pa).
Two lots, Maripa and Km 88 (Tables 2 and 3), proved unacceptable for three administered treatments: chloroquine (25 mg/kg), chloroquine (40 mg/kg) and the combination of sulfadoxine 500 mg and pyrimethamine 25 mg. Ineach case, surpassing the fixed upper threshold level of 25% of clinical failures once the critical value of d2 + 1 (n = 6) was recorded. Only early treatment failures were registered for chloroquine (25 mg/kg) and enrolment was stopped immediately once the critical value was overcome, whilst a combination of early treatment failures and late treatment failures were recorded with the administration of chloroquine (40 mg/kg) and the combination of sulfadoxine 500 mg and pyrimethamine 25 mg. Four late treatment failures in each lot (n = 8) with the administration of chloroquine (40 mg/kg) is an indication that the increased dose delayed the appearance of recrudescences; but therapeutic efficacy is still unacceptable at the upper quality level (Po = 0.25).
Table 2. Maripa: Efficacy of antimalarials for uncomplicated falciparum infections
|Chloroquine|| 6||ã = 33.0|| 4 Males|| 6 ETF||ã = 1860.0|
| 25 mg/kg|| ||SD = 24.4|| 2 Females|| ||SD = 594.1|
|Chloroquine||23||ã = 26.7||18 Males|| 2 ETF, 4 LTF,||ã = 1929.6|
| 40 mg/kg|| ||SD = 19.8|| 5 Females||17 ACR||SD = 962.8|
|Sulfadoxine 500 mg||26||ã = 22.0||17 Males|| 3 ETF, 3 LTF,||ã = 1799.2|
| + Pyrimethamine 25 mg|| ||SD = 20.6|| 9 Females||20 ACR||SD = 763.6|
|Quinine||16||ã = 28.5||11 Males||16 ACR||ã = 2195.0|
| 30 mg/kg|| ||SD = 18.1|| 5 Females|| ||SD = 968.7|
Table 3. Kilómetro 88: efficacy of antimalarials for uncomplicated falciparum infections
|Chloroquine 25 mg/kg|| 6||ã = 31.3|| 4 Males|| 6 ETF||ã = 2880.0||ã = 12.0||ã = 39.5|
|SD = 8.2|| 2 Females|| ||SD = 872.6||SD = 1.0||SD = 3.0|
|Chloroquine 40 mg/kg||13||ã = 28.6||11 Males|| 2 ETF||ã = 2707.6||ã = 11.4||ã = 37.6|
|SD = 7.7|| 2 Females|| 4 LTF |
|SD = 974.3||SD = 1.4||SD = 4.4|
|Sulfadoxine 500 mg +||11||ã = 27.0|| 8 Males|| 4 ETF||ã = 2404.5||ã = 11.2||ã = 36.9|
| Pyrimethamine 25 mg|| ||SD = 8.5|| 3 Females|| 2 LTF |
|SD = 707.3||SD = 1.2||SD = 3.3|
|Quinine||16||ã = 30.4||13 Males||16 ACR||ã = 2260.3||ã = 11.4||ã = 35.1|
| 30 mg/kg|| ||SD = 8.5|| 3 Females|| ||SD = 1052.7||SD = 1.4||SD = 5.0|
Only Ikabaru (Table 4) proved an acceptable lot for the administration of chloroquine (40 mg/kg) and the combination of sulfadoxine 500 mg-pyrimethamine 25 mg, in accordance to the pre-set quality levels (Po = 0.25) with treatment failures probably less than 10% (Pa = 0.10), and recording adequate clinical response in all patients enrolled.
Table 4. Ikabaru: efficacy of antimalarials for uncomplicated falciparum infections
|Chloroquine||16||ã = 30.0||15 Males||16 ACR||ã = 2190.7||ã = 11.4||ã = 37.7|
| 40 mg/kg|| ||SD = 11.3|| 1 Female|| ||SD = 1110.2||SD = 1.4||SD = 3.8|
|Sulfadoxine 500 mg +||16||ã = 28.7||11 Males||16 ACR||ã = 2999.8||ã = 11.3||ã = 37.0|
| Pyrimethamine 25 mg|| ||SD = 19.0|| 5 Females|| ||SD = 1182.5||SD = 1.2||SD = 4.0|
|Quinine||16||ã = 27.3|| 9 Males||16 ACR||ã = 3065.6||ã = 11.0||ã = 36.4|
| 30 mg/kg|| ||SD = 7.3|| 7 Females|| ||SD = 1644.2||SD = 1.4||SD = 5.6|
All three lots tested as acceptable for quinine (30 mg/kg) and recorded adequate clinical responses in all patients. Treatment failures are probably less than 10% (Pa = 0.10).
Higher parasite densities were recorded in the two mining lots of Km 88 and Ikabaru, perhaps because of the fact that miners usually delay seeking medical assistance and infections have more time to advance. Also, mainly young males are infected with P. falciparum, even more so in mining areas, because they are the predominant gender and as indication of a higher occupational risk for malaria and P. falciparum infections.
During the study period no patients showed general danger signs such as vomiting, lethargic or unconscious state or recent histories of convulsions. No patients presented signs of complicated malaria because in all cases of failure treatments quinine was administered. Neither were there any losses to follow-up during the study period.
In this study the LQADS technique proved to be quick, cheap and useful for monitoring the efficacy of antimalarials by field workers and physicians of a malaria programme in areas with a high grade of human circulation and migration. It serves as an ideal tool for continuous surveying of these same areas, because variation of therapeutic failures will be seen amongst the three lots over time, as to enable adequate updating of the national drug policy. The use of random stratified samples allowed the selection of subjects that covered all sampling units within the three chosen lots and provided valuable information about first-, second- and third-line drugs being used by the programme within these areas.
Administration of chloroquine as well as the combination of sulfadoxine and pyrimethamine amounted to clinical failures above the prefixed quality level of 25% in two of the chosen lots, Maripa and Km 88. This information serves as an indication that they may become areas of multidrug-resistant falciparum malaria in the near future. Failures were seen much earlier with the dose of chloroquine (25/kg). Only in Ikabaru chloroquine (40 mg/kg) and the combination of sulfadoxine-pyrimethamine were seen as effective. Finally, no therapeutic failures were registered in the three lots with Quinine (30 mg/kg). It seems clear that these therapeutical failures are indicating an important and qualitative extension of chloroquine and sulfadoxine-pyrimethamine resistant strains in the State of Bolívar. The degree of that extension still has to be delimited as well as pre-fixed quality levels for different epidemiological situations.
In an effort to delay the development and spread of resistance to antimalarial drugs it seems reasonable that monotherapy should be continued by the VMP, given the circumstances of adequate public and private financial support, quick procurement and controlled free distribution of drugs by the Ministry of Health, prompt diagnosis and treatment, as well as vector control through extensive or focal space and house-spraying, use of impregnated bednets or hammocks and environmental manipulation with the use of Eucalyptus plants. We are proposing that the national antimalarial treatment policy be based on the use of chloroquine (40 mg/kg) as the first-line drug until a pre-fixed level of just over 50% of treatment failures has been reached so as not to deplete the option of combining this drug with other antimalarials in the future. An affordable chloroquine replacement for use as a first-line drug in these areas would be amodiaquine. It was used with good results between 1985 and 1992 until is was suspended on account of reported adverse effects. As the second-line treatment, sulfadoxine-pyrimethamine combined with amodiaquine until the observed fraction of treated uncomplicated cases prevented from evolving into severe and complicated malaria is dependent on the use of quinine. Eventually, we foresee quinine replacing those drugs as a second-line treatment. In time, we expect third-line treatments and perhaps even first- and second-line treatments to depend more on combination therapy with several drugs. Artesunate plus mefloquine could be a viable option, although costly and requiring more supervision. Other combinations will be tried over the coming years to assure therapeutic efficacy, safety, continued consumer compliance and cost effectiveness so that proper selections can be made by the VMP. Finally, there is awareness that perhaps faster sequential changes in drug policy will be unavoidable over the next decade in our country, but at the same time we are hopeful that utmost care will be taken to determine the relative merits of these combination therapies for different epidemiological conditions; exerting the necessary precautions to avoid the exhaustion of some old and new antimalarials and ensuring that the spread of falciparum resistance will not be faced solely with the use of drugs, but also with the help of on-going vector control measures and community participation.
We thank Ornelio Rojas for the elaboration of the map showing the trial area.