RESEARCH

Efficacy of mebendazole and levamisole alone or in combination against intestinal nematode infections after repeated targeted mebendazole treatment in Zanzibar

Efficacité du mébendazole et du lévamisole seuls ou en association contre les nématodoses intestinales après traitement périodique ciblé par le mébendazole à Zanzibar

Eficacia del mebendazol y el levamisol, solos o combinados, contra las infecciones intestinales por nematodos después del tratamiento repetido focalizado con mebendazol en Zanzíbar

M. Albonico^I; Q. Bickle^II; M. Ramsan^III; A. Montresor^IV; L. Savioli^IV; M. Taylor^II

^IIvo de Carneri Foundation, Milan, Italy
^IIDepartment of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, England. Correspondence should be addressed to Dr Bickle (email: quentin.bickle@lshtm.ac.uk)
^IIIPublic Health Laboratory Ivo de Carneri, Pemba Island, Zanzibar, United Republic of Tanzania
^IVParasitic Diseases and Vector Control, Communicable Diseases, World Health Organization, Geneva, Switzerland

ABSTRACT

Objective: To evaluate the efficacy of and resistance to mebendazole (500 mg) and levamisole (40 or 80 mg), alone or in combination, for the treatment of Ascaris lumbricoides, Trichuris trichiura and hookworm infections on Pemba Island — an area exposed to periodic school-based mebendazole treatment since 1994.
Methods: A randomized, placebo-controlled trial was carried out in 914 children enrolled from the first and fifth grades of primary schools. Stool samples collected at baseline and 21 days after treatment were examined by the Kato–Katz technique to assess the prevalence and intensity of helminth infection.
Findings: Efficacies of mebendazole and levamisole as single treatments against intestinal nematode infections were comparable with those in previous trials, but mebendazole treatment of hookworm infections gave significantly lower cure (7.6%) and egg reduction (52.1%) rates than reported in a study undertaken before the beginning of periodic chemotherapy (cure rate, 22.4%; egg reduction rate, 82.4%). Combined treatment with mebendazole and levamisole had a significantly higher efficacy against hookworm infections (cure rate, 26.1%; egg reduction rate, 88.7%) than either drug given alone. No difference in mebendazole efficacy was found in children who had been treated repeatedly compared with those who had not been treated previously.
Conclusion: The overall efficacy of mebendazole against hookworm infections after periodic chemotherapy is reduced. The efficacy of benzimidazoles in chemotherapy-based control programmes should be monitored closely. Combined treatment with mebendazole and levamisole may be useful as a tool to delay the development of benzimidazole resistance.

Keywords: Ascariasis/drug therapy; /Trichuriasis/drug therapy; Ascaris lumbricoides/drug effects; Trichuris/drug effects; Mebendazole/pharmacology/administration and dosage; Levamisole/pharmacology/administration and dosage; Placebos; Drug therapy, Combination; Treatment outcome; Ancylostoma/drug effects; Necator americanus/drug effects; Drug resistance; Randomized controlled trials; Comparative study; United Republic of Tanzania (source: MeSH, NLM).

RÉSUMÉ

Objectif: Evaluer l'efficacité du mébendazole (500 mg) et du lévamisole (40 ou 80 mg) et la résistance à ces composés administrés seuls ou en association pour le traitement des infections à Ascaris lumbricoides, Trichuris trichiura et les ankylostomes sur l'île de Pemba, une région où sont effectués depuis 1994 des traitements périodiques par le mébendazole en milieu scolaire.
Méthodes: Un essai contrôlé randomisé contre placebo a été réalisé sur 914 enfants des premier et cinquième niveaux de l'enseignement primaire. Des échantillons de selles recueillis au début de l'étude et 21 jours après le traitement ont été examinés par la technique de Kato-Katz afin d'évaluer la prévalence et l'intensité des helminthiases.
Résultats: L'efficacité du mébendazole et du lévamisole en traitement unique des nématodoses intestinales était comparable à celle observée lors de précédents essais, mais le traitement des ankylostomiases par le mébendazole donnait des taux de guérison (7,6 %) et de réduction du nombre d'œufs (52,1 %) significativement plus faibles que lors d'une étude réalisée avant le début de la chimiothérapie périodique (taux de guérison : 22,4 % ; taux de réduction du nombre d'œufs : 82,4 %). Le traitement associé par le mébendazole et le lévamisole était significativement plus efficace contre les ankylostomiases (taux de guérison : 26,1 % ; taux de réduction du nombre d'œufs : 88,7 %) qu'un traitement par chacun des médicaments administrés séparément. Aucune différence d'efficacité du mébendazole n'a été observée entre les enfants ayant reçu une chimiothérapie périodique et ceux qui n'avaient jamais été traités auparavant.
Conclusion: Il semble d'après les résultats que l'efficacité globale du mébendazole contre les ankylostomiases après chimiothérapie périodique soit diminuée. L'efficacité des benzimidazolés dans le cadre des programmes de lutte reposant sur la chimiothérapie doit être étroitement surveillée. Un traitement associant le mébendazole et le lévamisole peut être utile pour retarder le développement de la résistance à cette catégorie de médicaments.

Mots clés: Ascaridiase larvaire/chimiothérapie; Trichocéphalose/chimiothérapie; Ascaris lumbricoides/action des produits chimiques; Trichuris/action des produits chimiques; Mébendazole/pharmacologie/administration et posologie; Lévamisole/pharmacologie/administration et posologie; Placebo; Polychimiothérapie; Evaluation résultats traitement; Ankylostome/action des produits chimiques; Necator americanus/action des produits chimiques; Résistance aux médicaments; Essai clinique randomisé; Etude comparative; République-Unie de Tanzanie (source: MeSH, INSERM).

RESUMEN

Objetivo: Evaluar la eficacia del mebendazol (500 mg) y el levamisol (40 u 80 mg), solos o combinados, y la resistencia a ellos como tratamiento de las infecciones por Ascaris lumbricoides y Trichuris trichiura y de la anquilostomiasis en la Isla de Pemba, una zona en la que desde 1994 se llevan a cabo periódicamente tratamientos con mebendazol en las escuelas.
Métodos: Se llevó a cabo un ensayo aleatorizado y controlado mediante placebo en 914 niños inscritos en alguno de los cinco primeros años de escuela primaria. Las muestras de heces obtenidas en la situación de partida y a los 21 días de tratamiento fueron analizadas mediante la técnica de Kato-Katz a fin de evaluar la prevalencia y la intensidad de las infecciones por helmintos.
Resultados: La eficacia del mebendazol y el levamisol como tratamientos únicos de las infecciones intestinales por nematodos fue comparable a la observada en ensayos anteriores, pero la administración de mebendazol contra las infecciones por anquilostoma logró unas tasas de curación (7,6%) y de reducción del número de huevos (52,1%) significativamente inferiores a las notificadas en un estudio emprendido antes del comienzo de la antibioticoterapia periódica (tasa de curación, 22,4%; tasa de reducción del número de huevos, 82,4%). El tratamiento combinado con mebendazol y levamisol tuvo una eficacia significativamente mayor contra las infecciones por anquilostoma (tasa de curación, 26,1%; tasa de reducción del número de huevos, 88,7%) que cualquiera de los medicamentos por separado. No se observó ninguna diferencia en cuanto a la eficacia del mebendazol entre los niños tratados repetidamente y los que no habían sido tratados anteriormente.
Conclusión: La eficacia global del mebendazol contra la anquilostomiasis se ve reducida después de la antibioticoterapia periódica. Debería vigilarse estrechamente la eficacia de los benzimidazoles en los programas de control basados en la antibioticoterapia. El tratamiento combinado con mebendazol y levamisol podría ser una valiosa arma para retrasar la aparición de resistencia a los benzimidazoles.

Palabras clave: Ascariasis/quimioterapia; Tricuriasis/quimioterapia; Ascaris lumbricoides/efectos de drogas; Trichuris/efectos de drogas; Mebendazol/farmacología/administración y dosificación; Levamisol/farmacología/administración y dosificación; Placebos; Quimioterapia combinada; Resultado del tratamiento; Ancylostoma/efectos de drogas; Necator americanus/efectos de drogas; Resistencia a las drogas; Ensayos controlados aleatorios; Estudio comparativo; República Unida de Tanzanía (fuente: DeCS, BIREME).

Introduction

Public health programmes to control morbidity associated with soil-transmitted helminth infections depend mainly on the delivery of anthelminthic drugs to primary-school children (1). In theory, four single-dose drugs are available (albendazole, levamisole, mebendazole, and pyrantel); in practice, however, most control programmes only use the benzimidazoles (albendazole and mebendazole) because they are given as a single-dose tablet and children do not need to be weighed (2). Aside from reducing the load of worms, benzimidazole treatment also improves the nutritional status and cognitive development of children infected with Ascaris lumbricoides, Trichuris trichiura, and hookworms and reduces hookworm-associated anaemia in children and in women of childbearing age (3–9). A number of studies have shown the short- and long-term benefits of periodic treatment with a single dose of 500 mg mebendazole in endemic areas (8, 10, 11).

Levamisole also is used in helminth control programmes with good results, although it is less effective than mebendazole against T. trichiura and Necator americanus infections. Comparative trials of both drugs, alone or in combination, are few, however, and none have compared single-dose administration of 500 mg mebendazole and 40 or 80 mg levamisole (12–16).

At present, concern exists about the possible emergence of drug resistance to the anthelminthic compounds used to control intestinal nematodes in humans, particularly given the well-documented and widespread problem of anthelminthic resistance in livestock as a consequence of frequent periodic mass treatments (17). The efficacy of combined treatments that use anthelminthics with differing modes of action (e.g. mebendazole plus levamisole) need to be assessed, both to explore possible additive or synergistic effects and to identify a combination that could delay the occurrence of anthelminthic drug resistance to each class of drug (18).

Comparison of the efficacy of mebendazole in the treatment of A. lumbricoides, T. trichiura, and hookworm infection in an area where the school-age population has been exposed to the drug for the last five years with data collected before the beginning of the control programme and with other efficacy data collected elsewhere can give valuable information on possible changes in drug susceptibility of worms.

The primary purpose of this trial was to evaluate the efficacy of single-dose mebendazole in an area where this drug has been used widely for periodic chemotherapy targeted at schoolchildren. A secondary objective was to assess and compare the efficacy of mebendazole with that of levamisole given as one or two 40-mg tablets and the efficacy of the two drugs given in combination.

Materials and methods

Study area and study population

The study was carried out on Pemba Island, the smaller of the two islands of Zanzibar, United Republic of Tanzania. Important features of the island have been described in detail elsewhere (19, 20). Intestinal helminth infections affect most of the population and cause a heavy burden of disease in children and in women of childbearing age (21). The National Helminth Control Programme was initiated in Zanzibar in June 1994, and since then, mebendazole 500 mg has been given to schoolchildren as a single dose every 4–6 months (21).

The study was conducted in August 1999 among children enrolled in the first grade (Standard 1) and fifth grade (Standard 5) of 10 public schools on Pemba Island. The schools were chosen randomly from the 72 schools on the island. Children at Standard 1 had not yet been treated with mebendazole in school, whereas children at Standard 5 had been exposed to 15 rounds of mebendazole treatment.

Children were excluded from the trial if they did not have parental or guardian permission to participate, did not provide a stool sample, had significant comorbidities (e.g. severe diarrhoea, severe anaemia, or high fever), or had recently transferred to the school from an area outside Zanzibar.

Study design

The study was a randomized, placebo-controlled trial. Before children were enrolled in the study, parents or guardians of children in the selected schools were given a comprehensive explanation of the risk and benefits of the trial, and verbal consent was sought.

Children enrolled in the study were assigned randomly to one of four treatment groups to receive 500 mg mebendazole (Janssen, Belgium), 40 or 80 mg levamisole (Zeneca, UK), 500 mg mebendazole plus 40 or 80 mg levamisole, or placebo. Placebo pills resembled mebendazole in colour, size, taste, and shape.

On the day before the scheduled treatment date, children eligible to participate in the trial were given a container in which to bring a fresh stool sample the next day. On the day of treatment, the stools were collected and the children's weights were recorded. Randomization was blocked on weight, and a computer-generated programme was used to create two randomized treatment lists: one for children who weighed 15–20 kg, who were to receive one tablet of 40 mg levamisole, and another for children who weighed 21–60 kg, who were to receive two tablets (80 mg). One tablet of mebendazole 500 mg and one of placebo was given irrespective of body weight. Treatments given were placed in sealed, opaque envelopes and were coded with a number. Children were identified by these numbers only throughout the study.

Twenty-one days after treatment, all children were revisited to collect a further stool sample. Any child who failed to bring a stool sample was followed for up to 24 days. Parents and children were instructed to report to the teacher and refer to the nearest health centre with any severe adverse effects that occurred in the week after treatment. After completion of the study, children in the placebo group and children positive after the follow-up survey were treated with mebendazole 500 mg.

The study was approved by the Zanzibar Health Research Council, and by the ethical committees of WHO and the London School of Hygiene and Tropical Medicine.

Parasitology

After both surveys, egg counts in stool samples were assessed within six hours of the sample being produced at the Public Health Laboratory Ivo de Carneri. All laboratory investigations were blinded, so the technicians who examined the slides were unaware of the treatment the patients received. Stools were analysed using the Kato–Katz technique according to WHO guidelines (22). The slides were examined within one hour of preparation to avoid overclarification of hookworm eggs. Comparison of egg counts before and after treatment allowed calculation of the cure rate and the egg reduction rate. A random sample of 10% of the smears prepared for the Kato– Katz technique was read by two different technicians to evaluate the accuracy of the diagnosis and the precision of the egg counts. Slides were re-examined if the quality control showed a >10% difference in egg counts.

To describe the egg count distribution in the study population before and after treatment, intensity of infection was classified as "light", "moderate", or "heavy" on the basis of faecal egg counts. As no heavy infections with A. lumbricoides were present at baseline when analysed according to WHO cut-off points, categories for A. lumbricoides intensity of moderate and heavy infection were revised to include a meaningful number of children for comparison (Table 1). Categories for intensity of T. trichiura and hookworm infection are according to WHO guidelines (Table 1) (23).

Statistical analysis

Data were entered and analysed with EpiInfo software. Cure rates were calculated as the percentage of children with egg counts >0 before treatment who had negative egg counts after treatment. The percentage reduction in prevalence was calculated as [(N⁺/n) – (N₂₁⁺/n)]/(N⁺/n), where N⁺ = the number of positive children at baseline, N₂₁⁺ = the number of positive children 21 days after treatment, and n = the total number of children with samples from both day 0 and day 21. Both cure rates and percentage reductions in prevalence were calculated. As the sensitivity of the Kato–Katz method could be influenced by the intensity of infection, however, reductions in prevalence were comparable only when pre-treatment intensities of infection were similar.

The percentage reduction in eggs induced by treatment was estimated as 100[1 – exp(–D)]%], where D = å (log_e(E₂₁) – log_e(E₀))/n, E₂₁ = the egg count at 21 days (for each individual), E₀ = the count before treatment, and n = the number of children. Confidence intervals were based on the inter-individual variation in the differences in the logarithms of the egg counts. Proportions were compared with standard c² tests. Geometric mean egg counts were estimated as exp[å(log_e(c+1))/n] – 1, where c = the count (eggs per gram) for a particular individual and n = the total number of samples. Geometric means were compared with analyses of variance by ANOVA if Bartlett's test of heterogenicity indicated homogeneity of variances and by the Kruskal–Wallis test if Bartlett's test was significant at the 5% level.

Results

Analysis of baseline data

Of 1137 children examined at the baseline, 904 (79.5%) returned stool samples at follow-up. Loss at follow-up was high compared with that in other surveys in Pembian schoolchildren because of an unexpected closure of schools on Pemba Island one week before the scheduled 21-day follow-up. The schools were closed for three weeks to allow children to help harvest an unusually large crop of cloves. The research staff thus walked from hut to hut through each village to trace missing children. Subsequent analysis showed that older, male children were more likely to be lost to follow-up, presumably because of their increased involvement in clove picking (12.2 vs 11.6 years, P<0.001; 54.5% boys vs 47.0% girls, P<0.05) (Table 2). Children subsequently lost at follow-up had similar baseline prevalence and intensity of A. lumbricoides, T. trichiura, and hookworm infections to children who were followed up.

Results are presented on the cohort of 904 children who returned a stool sample for the baseline and post-treatment surveys (Table 3). Their mean (SD; range) age was 11.5 (2.4; 7–18) years, and 45% were boys. The four groups were homogeneous at baseline for age and helminth infections, but differed in the proportion of males to females (Table 2). The prevalence of any helminth infection was 99.7%, with 38.3% and 54.9% of children harbouring double and triple infections, respectively. Table 3 gives the prevalence and intensity of A. lumbricoides, T. trichiura, and hookworms at baseline and 21 days after treatment in the four treatment groups.

Overall drug efficacy

Follow-up egg counts, cure rates, reductions in prevalence and egg reduction rates for the three nematode infections were statistically significantly better with all of the drug regimens compared with those at baseline, except for cure rates for hookworm infections with mebendazole and for T. trichiura infections with levamisole (although in both cases the mean egg counts were reduced substantially). Compared with placebo, all drug treatments produced significantly higher cure rates and egg reduction rates and lower prevalences at follow-up, except for the egg reduction rate for levamisole in T. trichiura infections (Table 3).

Both drugs had very high efficacy (98.5% and 99.1% egg reduction rates for levamisole and mebendazole, respectively) against A. lumbricoides. Mebendazole alone and in combination with levamisole had better efficacy than levamisole alone for T. trichiura infection (81% and 85% vs 41.5% egg reduction rates, P<0.001). Levamisole treatment produced a marginally significant reduction in prevalence of hookworm infection, which was greater than the reduction seen with mebendazole (8.9% vs 3.6%, P<0.05); the combination had better efficacy in reducing prevalence than either drug alone (23.6%, P<0.001). The egg reduction rate for hookworm infection was 88.7% for the combined treatment, but significantly less for either drug alone (61.3% for levamisole and 52.1% for mebendazole, P<0.001).

Figure 1 shows the distribution of intensities of infections, expressed as eggs per gram faeces, before and after treatment for A. lumbricoides, T. trichiura, and hookworm infections. Higher efficacy of treatment is indicated by an increase in negative and light infections and a decrease in moderate and heavy infections 21 days after treatment. Stratified analysis by intensity was performed on the cure rates and egg reduction rates for each species of helminth. For any treatment, lower cure rates were achieved in children with heavy infections compared with light infections: A. lumbricoides 93.3% vs 97.0% (P = 0.07), T. trichiura 8.3% vs 19.3% (P = 0.1), and hookworm 8.4% vs 16.3% (P<0.05). On the other hand, higher egg reduction rates were achieved in children with heavy compared with light infections for T. trichiura (93.8% vs 81.4%, P<0.05) and hookworm (91.3% vs 73.2%, P<0.01). Efficacy of treatment was not influenced by whether A. lumbricoides, T. trichiura, or hookworm infections presented as single or multiple infections. Although adverse effects were not investigated actively, no adverse events were reported after any single or combined treatment in the week following the administration of anthelminthics.

Drug efficacy in Standard 1 and Standard 5 children

Overall, 446 children were in Standard 1 classes and had a mean (SD) age of 9.5 (1.5) years, and 458 children were in Standard 5 classes and had a mean (SD) age of 13.4 (1.3) years. Boys accounted for 49.8% and 40.4% of children from Standard 1 and Standard 5 classes, respectively. Tables 4 and 5 give the prevalences and intensities of A. lumbricoides, T. trichiura, and hookworm infections at baseline and 21 days after treatment in the four treatment groups in Standard 1 and Standard 5 children, respectively. The baseline mean intensities of infection for all three species of helminths were significantly higher in the Standard 1 children than in the Standard 5 children (P<0.001). In addition, Standard 1 children at baseline had higher prevalences of moderate and heavy infection for each helminth than Standard 5 children — (A. lumbricoides: moderate 14.1% vs 9.2%, heavy 18.8% vs 9.0%; T. trichiura: moderate 45.3% vs 22.3%, heavy 2.5% vs 0.2%; and hookworm: moderate 12.1% vs 6.3%, heavy 11.4% vs 5.0%).

In both Standard 1 and Standard 5 children, but particularly in Standard 1 children, mebendazole and its combination with levamisole were more effective than levamisole in T. trichiura infection (the efficacy of levamisole did not differ significantly from that of placebo). In Standard 1 children, the effect of mebendazole against hookworm infection was similar to that of placebo, although its reduction of intensity of infection approached statistical significance (egg reduction rate 37%, P = 0.06), while levamisole was more effective (egg reduction rate 66.3%, P<0.001) and the combination was the best treatment (egg reduction rate 92.9%, P<0.001). In Standard 5 children, the combination was still the better option (although to a lesser extent than in Standard 1 children) than either drug alone for the treatment of hookworm infection (egg reduction rate 81.9% vs 56.3% for levamisole and 60.4% for mebendazole, P<0.05). Overall, none of cure rate, percentage reduction in prevalence, or egg reduction rate differed significantly between Standard 1 and Standard 5 children for any helminth infection. Furthermore, no evidence was seen of a reduced cure or egg reduction rate for mebendazole relative to levamisole in Standard 5 children compared with Standard 1 children.

Discussion

This study confirms the extremely high prevalence of intestinal nematode infections in Zanzibar despite the periodic chemotherapy control programme that started in 1994. The objective of this programme, however, was reduction in intensity rather than prevalence of infection. In this respect, the significantly lower egg counts at baseline in Standard 5 children who had received 15 rounds of treatment with mebendazole compared with children in Standard 1 classes, who had never been treated, are encouraging. A possible explanation for this finding could be an intrinsic decline in intensity with increasing age. This is not supported by a previous study on untreated children in the same population, however, which showed that between the ages of 9 (Standard 1) and 13 years (Standard 5), the mean intensity of ascariasis in fact increased, the intensity of hookworm infections remained stable, and the intensity of trichuriasis slightly decreased, although to a much lesser extent than between Standard 1 and Standard 5 children in the present study (21). The benefit of periodic chemotherapy is shown further in the significant reduction of moderate and heavy helminth infections — those most associated with morbidity — in Standard 5 children. From the public health perspective, it seems, therefore, that the helminth control programme is effective in reducing intensity, although the low efficacy of treatment of hookworm infections found in this trial is worrying.

High, and similar, levels of efficacy in the treatment of A. lumbricoides infection were achieved using mebendazole, levamisole, and their combination. The efficacy of each treatment was equally excellent in light and heavy A. lumbricoides infections, with egg reduction rates approaching 100%. This result is consistent with previous efficacy trials and is also similar — for mebendazole — to the results of an efficacy study carried out in Pemba in 1993 before the mebendazole-based control programme was started (14, 24–27).

Drug efficacies in T. trichiura infection were also similar to those seen in previous trials, as treatment with mebendazole reduced the prevalence of infection, while levamisole had a much smaller effect (14, 26–29). Reductions in egg counts were significant for mebendazole and for levamisole, although mebendazole and the combined treatment were both significantly better than levamisole alone. When compared with the efficacy trial performed in Pemba before the control programme started, the mebendazole cure and egg reduction rates were not significantly different (24). The poor efficacy of levamisole against T. trichiura has already been reported, although Ismail et al. showed an egg reduction rate of 73% (2, 14). The finding that egg reduction rates were significantly higher in both T. trichiura and hookworm "heavy" infections compared with "light" infections is consistent with trials with albendazole (29, 30).

Striking features of this study were the low cure (7.6%) and egg reduction (52.1%) rates when mebendazole was used against hookworms. Mebendazole efficacy was lower when compared with that in published studies and, more specifically, compared with that in the trial carried out in Pemba before exposure to periodic treatment, in which mebendazole had a cure rate of 22.4% and an egg reduction rate of 82.4% (14, 24–27). Furthermore, in a recent trial comparable with our trial in the neighbouring island of Mafia, in which there is no school-based deworming programme and use of benzimidazoles has been very limited, mebendazole efficacies were similar (cure rate 31.3%; egg reduction rate 78.1%) to the pre-treatment values on Pemba and very much higher than the current efficacy in Pemba (31).

A number of potential confounding factors were ruled out by the fact that both trials were carried out in schoolchildren by the same staff and used the same methods and drug (mebendazole 500 mg from Janssen). It is notable that the cure rates and egg reduction rates for Ascaris and Trichuris were comparable between the two trials. Interestingly, levamisole also showed a lower efficacy against hookworms in the present study than in previous studies (32–34). A possible explanation is that levamisole has been reported to be more effective against A. duodenale than N. americanus, but the latter pathogen is the most prevalent species in Pemba (20). If mebendazole also was less effective against N. americanus, repeated use of mebendazole in Pemba could have created an increased Necator:Ancylostoma ratio, which would have resulted in an overall lower efficacy against hookworms. Data from recent studies suggest that the Necator:Ancylostoma ratio is indeed higher now than it was before periodic anthelminthic treatment (Albonico, unpublished data, 2000) (20).

An alternative explanation for the apparent reduced efficacy of mebendazole compared with historical controls is the selection of drug resistance by the repeated treatment regimen. If a proportion of drug-resistant worms was present in the worm population in Pemba, we would have expected resistant worms to have accumulated in children after multiple rounds of treatment and reinfection and to have represented a higher proportion of worms than that found in Standard 1 children. In this case, a lower drug efficacy would have been expected in the Standard 5 children. In fact, the efficacy of mebendazole in Standard 5 children who had received multiple (15) doses of mebendazole was not lower than in Standard 1 children who had never been treated; this suggests that a mebendazole-resistant population of worms had not accumulated in the repeatedly treated Standard 5 children. Nevertheless, the possibility that drug resistance is emerging in Pemba cannot be excluded. In Mali, failure of mebendazole in the treatment of human hookworm infections has been reported (35). In that case, however, the population had not been exposed to periodic treatment with mebendazole, although mebendazole had been available in that community for some years; pre-exposure efficacy data were not available for comparison and the sample of children studied was rather small. These factors suggest a need for caution when interpreting these results. A subsequent drug efficacy study in Mali (36) showed reasonably good efficacy of mebendazole and even better results from albendazole, which show a lack of significant resistance in human hookworms to benzimidazoles in that area. Poor efficacy of pyrantel against hookworms has been found in north-west Australia, but again it is not known if this represents a decline in efficacy because of emerging resistance (37).

The efficacy of the combined administration of mebendazole 500 mg and levamisole 40 or 80 mg was evaluated for the first time in this study and showed higher efficacy than either drug alone against hookworm infections. This is a promising result, because the combined administration of two different anthelminthic drugs could be used as the treatment of choice in this context. In addition, the adoption of combination therapy with drugs with distinct modes of action when used at early stages has been shown to delay the onset of anthelminthic drug resistance (18).

Although this study clearly shows that the efficacy of mebendazole in the treatment of hookworm infections in Pemba Island schoolchildren is lower than in previous studies, demonstration of anthelminthic resistance is still lacking. A need exists, therefore, for alternative methods by which to address this question, such as the in vitro egg hatch assay — a technique widely used in veterinary medicine (38) and recently adapted to human hookworms (39). This would require comparisons with strains of hookworms that had not been exposed to treatment. The early detection of drug resistance is of the utmost importance, because the in vivo and in vitro tests currently available in the veterinary field detect resistance when the proportion of worms carrying the drug resistance allele in a population already has reached >25% (40). The development of methods for the early detection of benzimidazole resistance in human nematodes would be of great value, therefore, and the polymerase chain reaction methods developed by veterinary scientists need to be evaluated for their applicability in human nematodes (41).

Acknowledgements

Special thanks to Professor Peter Smith for his valuable advice on statistical analysis, and for the revision of the manuscript. The authors thank also Dr Nerissa Kohen and Victoria Wright for their contribution towards the data collection in Pemba Island. We are grateful to the staff of the Helminth Control Programme, as well as the staff of the Public Health Laboratory Ivo de Carneri in Pemba Island, whose dedication and enthusiasm made this study possible. We acknowledge the donation of placebo and mebendazole from Pharmamed (Malta) and Janssen (Belgium) and of levamisole from Zeneca (UK). This study was supported generously by Parasitic Diseases and Vector Control, Division of Communicable Diseases Control, Prevention and Eradication, World Health Organization, Geneva.

Conflicts of interest: none declared.

References

1. World Health Organization. Report of the WHO informal consultation on the use of chemotherapy for the control of morbidity due to soil-transmitted nematodes in humans. Division of Control of Tropical Diseases. Geneva: World Health Organization; 1996. WHO document WHO/CTD/SIP.96.2.        

2. World Health Organization. WHO model prescribing information. Drugs used in parasitic diseases. Geneva: World Health Organization; 1995.        

3. Stephenson LS, Latham MC, Adams EJ, Kinoti SN, Pertet A. Physical fitness, growth and appetite of Kenyan school boys with hookworm, Trichuris trichiura and Ascaris lumbricoides infections are improved four months after a single dose of albendazole. Journal of Nutrition 1993;123:1036-46.         

4. Stephenson LS, Latham MC, Adams EJ, Kinoti SN, Pertet A. Weight gain of Kenyan schoolchildren infected with hookworm, Trichuris trichiura and Ascaris lumbricoides is improved following once- or twice-yearly treatment with albendazole. Journal of Nutrition 1993;126:656-65.        

5. Stoltzfus RJ, Albonico M, Tielsch JM, Chwaya HM, Savioli L. School-based deworming program yields small improvement in growth of Zanzibar schoolchildren. Journal of Nutrition 1997;127:2187-93.         

6. Nokes C, Bundy DAP. Does helminth infection affect mental processing and educational achievement? Parasitology Today 1994;10:14-8.        

7. Simeon DT, Grantham-McGregor SM, Callender JE, Wong MS. Treatment of Trichuris trichiura improves growth, spelling scores and school attendance in some children. Journal of Nutrition 1995;125:1875-83.        

8. Stoltzfus RJ, Albonico M, Chwaya HM, Tielsch JM, Schulze KJ, Savioli L. Effects of the Zanzibar school-based deworming program on iron status of children. American Journal of Clinical Nutrition 1998;68:179-86.         

9. Atukorala TMS, de Silva LDR, Dechering WHJC, Dassenaeike de CTS, Perera RS. Evaluation of effectiveness of iron-folate supplementation and anthelminthic therapy against anaemia in pregnancy — a study in the plantation sector of Sri Lanka. American Journal of Clinical Nutrition 1994;60:286-92.        

10. Albonico M, Shamlaye N, Shamlaye C, Savioli L. The control of intestinal parasitic infections in the Seychelles: a comprehensive and sustainable approach. Bulletin of the World Health Organization 1996;74:577-86.        

11. Sorensen E, Ismail M, Hettiarachchi I, Dassenaieke de CTS, Amarasinghe DKC. The prevalence and control of soil-transmitted nematode infections among children and women in the plantations in Sri Lanka. Ceylon Medical Journal 1996;41:37-41.        

12. Asaolu SO, Holland CV, Crompton DWT. Community control of A. lumbricoides in rural Oyo State, Nigeria: mass, targeted and selective treatment with levamisole. Parasitology 1991;103:291-8.        

13. Holland CV, Crompton DW, Taren DL, Nesheim MC, Sanjur D, Barbeau I, et al. Ascaris lumbricoides infection in pre-schoolchildren from Chiriqui Province, Panama. Parasitology 1987;3:615-22.        

14. Ismail MM, Premaratne UN, Suraweera MGW. Comparative efficacy of single dose anthelmintics in relation to intensity of geohelminth infections. Ceylon Medical Journal 1991;36:162-7.        

15. Al Saffar G, Al Saleem M, Bakhous IJ. L-tetramisole in the treatment of ancylostomiasis. Transactions of the Royal Society of Tropical Medicine and Hygiene 1971;65:836-7.        

16. Kan SP, Chua SH. The efficacy of single doses of mebendazole plus levamisole in the treatment of Trichuris trichiura and Ascaris lumbricoides infections. Yokogawa M, editor. Collected papers on the control of soil-transmitted helminthiases, volume III. Tokyo: APCO; 1986. p. 82-5.        

17. Geerts S, Coles GC, Gryseels B. Anthelmintic resistance in human helminths: learning from the problem with worm control in livestock. Parasitology Today 1997;13:149-51.        

18. Barnes EH, Dobson RJ, Barger IA. Worm control and anthelmintic resistance: adventures with a model. Parasitology Today 1995;11:56-63.        

19. Savioli L, Dixon H, Kisumku UM, Mott KE. Control of morbidity due to Schistosoma haematobium on Pemba island; programme organization and management. Tropical Medicine and Parasitology 1989;40:189-94.        

20. Albonico M, Stoltzfus RJ, Tielsch JM, Chwaya HM, Ercole E, Cancrini G. Epidemiological evidence of a differential effect of hookworm species, Ancylostoma duodenale and Necator americanus, on iron status of children. International Journal of Epidemiology 1998;27:530-7.        

21. Renganathan E, Ercole E, Albonico M, De Gregorio G, Alawi KS, Kisumku UM, et al. Evolution of operational research and development of a national control strategy against intestinal helminths in Pemba Island, 1988-92. Bulletin of the World Health Organization 1995;73:183-90.        

22. Ash, LR, Orihel TC, Savioli L. Bench aids for the diagnosis of intestinal parasites. Geneva: World Health Organization; 1994.        

23. Montresor A, Crompton DWT, Hall A, Bundy DAP, Savioli L. Guidelines for the evaluation of soil-transmitted helminthiasis and schistosomiasis at community level. Geneva: World Health Organization; 1998. WHO document WHO/CDS/SIP/98.1 (available from the Programme Parasitic Diseases and Vector Control, World Health Organization, CH 1211 Geneva 27, Switzerland).        

24. Albonico M, Smith PG, Hall A, Chwaya HM, Alawi KS, Savioli L. A randomised controlled trial comparing Mebendazole and Albendazole against Ascaris, Trichuris, and the hookworms. Transactions of the Royal Society of Tropical Medicine and Hygiene 1994;88:585-9.        

25. Abadi K. Single dose mebendazole therapy for soil-transmitted nematodes. American Journal of Tropical Medicine and Hygiene 1985;34:129-33.        

26. Albonico M, Renganathan E, Bosman A, Uledi MK, Kassim SA, Savioli L. Efficacy of a single dose of mebendazole on prevalence and intensity of soil-transmitted nematodes in Zanzibar. Tropical and Geographical Medicine 1994;46:142-6.        

27. Bennet A, Guyatt H. Reducing intestinal nematode infection: efficacy of albendazole and mebendazole. Parasitology Today 2000;16:71-4.        

28. Miller M. Use of levamisole in parasitic infections. Drugs 1980;19:122-30.        

29. Pene P, Vincentelli JM, Soula G, Bourderioux Ch, Rossignol JF. Le Zentel (albendazole) dans le traitement des nematodes intestinales. Etude multicentrique en Afrique de l'Ouest. (Zentel (albendazole) for the treatment of intestinal nematodes. Multicentric study in West Africa). Médecine d'Afrique noire 1981;28:483-5. In French.        

30. Rossignol JF, Maissonneuve H. Albendazole: placebo controlled study in 870 patients with intestinal helminths. Transactions of the Royal Society of Tropical Medicine and Hygiene 1983;77:707-11.        

31. Albonico M, Ramsan M, Wright V, Khatib J, Haji HJ, Taylor M, et al. Soil-transmitted nematode infections and mebendazole treatment in Mafia Island school-children. Annals of Tropical Medicine and Parasitology 2002;96:717-26.        

32. Farid Z, Bassily S, Miner WF, Hassan A, Laughli LW. Comparative single-dose treatment of hookworm and roundworm infections with levamisole, pyrantel and bephenium. Journal of Tropical Medicine and Hygiene 1977;80:107-8.         

33. Huys J, van den Berghe G, Freyens P, Kayihigi J. Treatment of ancylostomiasis with levamisole. African Journal of Medicine and Medical Science 1976;5:75-7.         

34. Lucas AO, Oduntan SO. Treatment of hookworm infection and other parasites with l-tetramisole (Ketrax). Annals of Tropical Medicine and Parasitology 1972;66:391-8.        

35. De Clercq D, Sacko M, Behnke J, Gilbert F, Dorny P, Vercruysse J. Failure of mebendazole in treatment of human hookworm infections in the southern region of Mali. American Journal of Tropical Medicine and Hygiene 1997;57:25-30.         

36. Sacko M, De Clercq D, Behnke J, Gilbert FS, Dorny P, Vercruysse J. Comparison of the efficacy of mebendazole, albendazole and pyrantel in treatment of human hookworm infections in the Southern Region of Mali, West Africa. Transactions of the Royal Society of Tropical Medicine and Hygiene 1999;93:195-203.         

37. Reynoldson JA, Behnke JM, Pallant LJ, Macnish MG, Gilbert F, Giles S, et al. Failure of pyrantel in treatment of human hookworm infections (Ancylostoma duodenale) in the Kimberly region of North West Australia. Acta Tropica 1997;68:301-12.        

38. Coles GC, Bauer C, Borgsteede FH, Geerts S, Klei TR, Taylor MA, et al. World Association for the Advancement of Veterinary Parasitology (WAAVP) methods for the detection of anthelmintic resistance in nematodes of veterinary importance. Veterinary Parasitology 1992;44:35-44.        

39. Albonico M, Wright V, Ramsan M, Haji HJ, Taylor M, Savioli L, et al. Development and evaluation of the egg hatch assay (EHA) for detection of anthelminthic drug resistance by human hookworms. In: Abstract from SOIPA XXII Congress, Torino, June 2002. Parassitologia 2002;44 Suppl 1:4.        

40. Martin PJ, Anderson N, Jarret RG. Detecting benzimidazole resistance with faecal egg count reduction tests and in vitro assays. Australian Veterinary Journal 1989;66:236-40.        

41. Roos HM, Kwa MSG, Grant WN. New genetic and practical implications of selection for anthelmintic resistance in parasitic nematodes. Parasitology Today 1995;11:148-50.