INTRODUCTION
Fascioliasis, caused by infection with the trematode flukes Fasciola hepatica or Fasciola gigantica, is an important cause of morbidity in both humans and animals. The human burden of disease is increasing; estimates for the total number of current human infections vary from 2.5 to 17 million, and there are reports of a rise in the number of infections in at least 51 countries with important public-health and economic implications.1,2 Effective treatment and control measures are hampered by limited understanding of the human disease phenotypes and few effective drugs. The current treatment of choice for human fascioliasis is triclabendazole. Triclabendazole seems to be well-tolerated but is expensive, the human formulation is not widely available, and resistance in animal isolates is well described.3–11 Other drugs used in the past have included bithionol and nitazoxanide; there are no published randomized controlled trials for the former, and the latter has an efficacy rate on the order of 36% in adults (intention to treat, ITT).12
The case definitions for fascioliasis used in published clinical trials have usually relied upon demonstration of Fasciola ova in feces, because such determination is highly specific and easily measurable. However, this definition underestimates the disease burden because egg shedding is an intermittent process, and both symptomatic and asymptomatic patients—with acute or chronic infection—may have negative stool results.2,13 Patients with significant morbidity will be excluded by this definition. This may be relevant, in that the drugs recommended for treatment may be more or less effective against differently aged or developed flukes, which in turn may correlate with the presence of ova in stool.14
Fascioliasis is a significant problem in central Vietnam, with seropositivity rates of 8% in some areas (T. T. Hien, unpublished data). Recently, there has been evidence of the effectiveness of the anti-malarial drug artesunate in the treatment of other fluke diseases, notably Schistosoma mansoni and Schistosoma japonicum.15–19 The mechanism of action is not clear, but animal experiments have demonstrated damage to the fluke tegument.18 Artesunate is cheap, and extensive data from studies in severe malaria demonstrate its good safety profile.20 Vietnamese physicians have extensive experience in the use of artesunate through its use in the treatment of malaria; the first randomized controlled trials of the artemisinin derivatives in malaria were carried out in Vietnam. The Vietnamese National Malaria Guidelines now recommend artesunate as the first-line treatment, and when used in combination the drug now forms the basis of the WHO’s global antimalarial drug strategy. More recently, there are in vitro and in vivo data demonstrating antifasciola effects of artemether and artesunate.18 We conducted a randomized controlled pilot study to compare artesunate with triclabendazole in the treatment of symptomatic human fascioliasis in Vietnam. Our patients present with abdominal pain, fever, and abdominal pain and have ultrasound scans showing parenchymal liver lesions consistent with acute fascioliasis. Thus the primary endpoint chosen was resolution of abdominal pain, because this is the most prominent symptom suffered by patients with fascioliasis in our practice. The “complete response rate,” meaning resolution of all symptoms, normalization of eosinophil count, and improvement in ultrasound appearances, formed the secondary endpoints. Presence of Fasciola ova in stool was not part of the case definition.
MATERIALS AND METHODS
Study design.
This prospective randomized open-label pilot study compared oral triclabendazole with oral artesunate for the treatment of human fascioliasis. This trial was registered on-line with the International Standardised Randomised Controlled Trial Number ISRCTN75869075 at http://www.controlled-trials.com/ISRCTN75869075/fasciola.
Study subjects.
Any patient of age > 8 years and meeting the case definition for fascioliasis was eligible for study entry (see Table 1). Serological testing was performed using a Fasciola-specific ELISA based upon a purified protein antigen.21 Patients were treated as hospital in-patients for 10 days and received directly observed therapy. Patients either self-referred or were referred by physicians from Qui Nhon Province in Central Vietnam, an area endemic for human fascioliasis. Study recruitment took place from February 2004 to December 2005.
Ethical approval and consent.
The study received ethical approval from the Ethical Committee of the Hospital for Tropical Diseases, Ho Chi Minh City, Vietnam. Witnessed written or verbal informed consent was obtained from all patients prior to study entry.
Randomization.
A random number sequence was generated using Microsoft Excel by a member of the Clinical Trials Unit, Hospital for Tropical Diseases, who was not involved in any other aspect of the study. The randomization sequence was concealed from the physicians allocating treatment and managing the patients, prior to patient enrolment. Treatment allocations were kept in sequentially numbered, sealed, opaque envelopes, which were opened in strict numerical order only upon enrollment of the patient to the study after all inclusion and exclusion criteria had been checked.
Study site.
The study took place at the Hospital for Tropical Diseases, Ho Chi Minh City, Vietnam. The Hospital for Tropical Diseases (HTD), Ho Chi Minh City, is a 500-bed infectious-diseases hospital serving the local community and is the tertiary referral center for infectious diseases for all of southern Vietnam.
Study medication administration and follow-up.
Patients were randomly allocated in a ratio of 1:1 to receive either 2 doses of triclabendazole (Egaten, Novartis, Basel, Switzerland), 10 mg/kg, 12 hours apart with food, or artesunate (Guilin Pharmaceutical Factory No. 1, Guilin, China), 4 mg/kg body weight per day, once daily for 10 days. All study subjects were treated as in-patients for 10 days and had directly observed therapy.
Endpoints.
The primary endpoint was resolution of abdominal pain at hospital discharge (day 10). The secondary endpoint was the complete response rate at 3 months post-treatment defined as all 3 of 1) resolution of symptoms; 2) normalization of eosinophil count; and 3) improvement in ultrasound appearances. These 3 parameters were also evaluated individually 3 months post-treatment, on ITT and per-protocol (PP) bases.
Study investigations.
All patients had a full history and physical examination according to the study proforma. Complete blood count including absolute eosinophil count and percentage, alanine-amino transferase (ALT), aspartate-amino transferase (AST), serology using a Fasciola-specific ELISA, and liver ultrasound scan were performed at study entry, hospital discharge, and 3 months post-treatment.
Power calculation.
Assuming a cure rate of 70% in the triclabendazole arm, a sample size of 100 patients would have 80% power to detect a difference of 27% between the treatment arms, α = 0.05.
Statistical analysis.
Baseline characteristics were described using frequency and proportion for categorical variables and median, interquartile range and range for continuous variables. The primary endpoint (abdominal pain) at discharge was analyzed using Fisher’s exact test. Other findings at hospital discharge were compared using analysis of covariance (ANCOVA) for continuous endpoints and Fisher’s exact test for discrete endpoints.
The secondary endpoint (complete response rate) was analyzed using Fisher’s exact test. The relative risk for complete response and its 95% confidence intervals were calculated. The secondary endpoint was examined on both an ITT and PP bases, where patients lost to follow-up were considered as treatment failures.
RESULTS
Baseline characteristics.
Fifty patients were randomized to receive triclabendazole and 50 patients to receive artesunate treatment. The 2 groups were broadly similar at baseline (Table 2). The median age was 29.5 years, and there was a preponderance of female patients (62%). The median duration of illness was the same for both groups, but there was a broader distribution in duration of illness in the triclabendazole group (interquartile range 7–365 days versus 7–90 days, TBZ versus ATS); 35% of patients had elevated AST at baseline, and 10% of patients had elevated ALT (9% had both). No patients had jaundice at presentation.
Efficacy.
Analysis of the primary endpoint showed a significant difference in favor of artesunate, with 100% of patients being free of abdominal pain at discharge, compared with 88% of the patients receiving triclabendazole (Fisher’s exact test, P = 0.027). However, there was a lower rate of fever clearance in the artesunate group compared with the triclabendazole group (see Table 3). There were no differences in liver enzymes, absolute or percentage eosinophil counts, Fasciola antibody titer levels, or ultrasonographic changes between the 2 groups at hospital discharge.
The secondary endpoint was assessed at 3 months post-treatment (Tables 4 and 5). Both ITT and PP analyses revealed a significantly higher proportion of patients receiving triclabendazole to have a complete response (10% versus 36%, artesunate versus triclabendazole, ITT, P = 0.004, relative risk 0.28, 95% confidence interval 0.11–0.69). Individual analysis of rates of eosinophilia normalization, absence of clinical symptoms and improvement in ultrasound scan appearances revealed that patients receiving artesunate had significantly lower rates of peripheral blood eosinophilia normalization and clinical symptom improvement, but there was no significant difference in the proportion of patients with improved ultrasonographic findings.
DISCUSSION
This study has demonstrated a possible role for artesunate in the treatment of fascioliasis in humans. Initial symptom control, measured at hospital discharge, was better in patients treated with artesunate compared with triclabendazole. This contrasts with the situation at 3 months where, using a strict definition of complete response, there was a poorer response in patients receiving artesunate than triclabendazole. In fact, the complete response rate at 3 months was disappointing for both treatments. Interestingly, there were no differences in the rates of resolution or improvement in liver lesions measured by ultrasonography between treatment groups at 3 months at 66% for the artesunate group versus 70% for the triclabendazole group. The difference in outcome at 3 months is predominantly explained by a difference in the rates of normalization of eosinophil count between triclabendazole and artesunate (although there was a difference in symptoms that reached statistical significance). The background presence of other enteric parasites in our population is high (for example, hookworm ova positivity approaches 65%), and the difference in eosinophilia resolution might represent differences in other antihelminthic activity of artesunate and triclabendazole, although triclabendazole is notable for its narrow spectrum.22
Clinical research in fascioliasis is hampered by the connected problems of defining clinical disease and defining a satisfactory response to treatment. Most studies use the presence of Fasciola eggs in feces as an essential criterion for disease definition.3,6,12,13,23 This is attractive because the technique is relatively simple, the presence of eggs is highly specific for infection, and there is a logical corresponding marker of cure, namely, stool egg negativity. Other markers of parasitic infection, such as an elevated peripheral eosinophil count, are less useful in parasite-endemic areas where there may be many infections contributing to eosinophilia, rendering it less useful as a marker of successful treatment. However, the use of stool egg positivity is problematic. Egg shedding in fascioliasis is an intermittent event, and stool examination may be repeatedly negative in both acute and chronic disease.2,13,24 Relying on positive coprological examination for fascioliasis diagnosis in clinical trials may have important consequences. On the one hand, it may lead to overestimates of drug efficacy; stool egg positivity may be associated with a better response to treatment because it may correlate with the age of the fluke, or its anatomic location, which may be associated with increased susceptibility to treatment. The dose-finding studies with triclabendazole suggest this could be the case, because the cure rate in ova-negative cases was significantly lower than in ova-positive cases.7 In addition, there have been few placebo-controlled trials in fascioliasis, but there are data that suggest the spontaneous stool clearance rate may be up to 30%.12 Thus it may be unnecessary to treat all asymptomatic egg shedders.
However, the most important effect of using an ova-based disease definition is that a large number of people with clinically significant disease will be excluded from clinical trials, resulting in a lack of data to guide treatment in these patients. In our experience, this is the largest group of patients with fascioliasis. Our patients have significant morbidity with fever and abdominal pain, high peripheral eosinophilia counts, and ultrasonography showing parenchymal liver disease. We rarely demonstrate Fasciola ova in stool but are confident of the diagnosis in our patients because of positive serology, isolation of ectopic flukes from patients, ultrasonographic appearances, a response rate to treatment consistent with the published literature, and the failure to demonstrate ova from other fluke species in stool (for example, Clonorchis spp.). The low rate of ova isolation from stool in our patients may represent the stage of the fascioliasis epidemic in Vietnam, which may be younger than in South America, where infection is believed to occur predominantly in children, and where egg-positivity rates appear to be higher.25 We suggest that our patients probably have acute fascioliasis, and the low cure rate in our patients suggests that triclabendazole may be less effective in acute fascioliasis.
An effect of artesunate in human fascioliasis is plausible. Artesunate has been shown to be a safe and extremely effective antimalarial drug. Although the mechanism of action remains controversial, its antiparasitic effects are not limited to Plasmodium species. There is both in vitro and in vivo evidence of a useful effect in other fluke diseases, including schistosomiasis, and more recently there is in vitro evidence of activity against other flukes, including F. hepatica and Clonorchis sinensis.18,26 Artesunate’s tolerability is well documented in numerous clinical trials in malaria. Its short half-life may protect it against development of resistance, which is already well described for triclabendazole. The use of artesunate in fascioliasis should be examined in further clinical trials, including dose-finding studies and its use as both alternative and adjunctive treatment. Investigation of other artemesinin derivatives, with longer half-lives than artesunate, such as artemether, may also be fruitful. Future intervention studies should aim to include patients with symptomatic fascioliasis and move away from stool egg positivity as an essential criterion in the case definition. In parallel with this, there is a place for further detailed descriptive studies of human fascioliasis to better define the disease phenotypes and natural history.
Case definition for fascioliasis
| All of | Clinical symptoms consistent with fascioliasis, including but not restricted to:
|
| Ultrasound scan of the liver consistent with human fascioliasis | |
| Positive Fasciola serology | |
| Eosinophilia > 400,000 cells/mL | |
| Living in an endemic area for fascioliasis |
Baseline characteristics of patients in study
| Characteristics | Statistic | Artesunate (no. of patients = 50) | Triclabendazole (no. of patients = 50) |
|---|---|---|---|
| Age (years) | Median (IQR) | 29.5 (24–36) | 29.5 (22–37) |
| Range | 9–65 | 13–74 | |
| Sex | No. of female/total cases (%) | 34/50 (68) | 28/50 (56) |
| Days of illness (days) | Median (IQR) | 30 (21–30) | 30 (30–60) |
| Range | 7–90 | 7–365 | |
| Abdominal pain | No. of symptomatic cases/total cases (%) | 50/50 (100) | 49/50 (98) |
| Fever | No. of symptomatic cases/total cases (%) | 14/50 (28) | 20/50 (40) |
| Itch | No. of symptomatic cases/total cases (%) | 4/50 (8) | 6/50 (12) |
| Eosinophil (%) | Median (IQR) | 30 (23–43) | 34 (25–46) |
| Range | 8–67 | 11–69 | |
| Eosinophil count (cells/μL) | Median (IQR) | 3,250 (2,072–5,240) | 3,354 (2,320–6,380) |
| > 400 counts/μL | Range | 756–10,400 | 942–12,900 |
| No. of cases/total cases (%) | 50/50 (100) | 50/50 (100) | |
| Serum Ab (ELISA) | Median (IQR) | 6,400 (5,920–10,880) | 6,400 (5,184–7,616) |
| Range | 0–12,800 | 0–12,800 | |
| ALT (UI/L) | Median (IQR) | 25 (21–32) | 27 (21–35) |
| Range | 14–75 | 15–65 | |
| AST (UI/L) | Median (IQR) | 32 (24–40) | 32.5 (24–43.5) |
| Range | 12–105 | 16–91 |
Clinical outcomes and laboratory findings at hospital discharge
| Outcome | Artesunate (no. of patients = 50) | Triclabendazole (no. of patients = 50) | P value* |
|---|---|---|---|
| * For each variable, ANCOVA test was performed, adjusting for the same baseline covariate. | |||
| † Fisher’s exact test. | |||
| ‡ Four cases missing in triclabendazole and 5 in artesunate groups. | |||
| § Two cases missing in triclabendazole and 4 in artesunate groups. | |||
| ¶ Four cases missing in triclabendazole and 3 in artesunate groups. | |||
| || Two cases missing in triclabendazole and 4 in artesunate groups. | |||
| Abdominal pain-free: | |||
| No. of cases (%) | 50/50 (100) | 44/50 (88) | 0.027† |
| Relative risk of being pain free (95% CI) | 1.14 (1.03–1.26) | ||
| Fever: no. of cases (%) | 8/50 (16) | 0/50 (0) | 0.006† |
| ALT‡ (UI/L) | |||
| Median (IQR) | 30 (24–34) | 28 (24–35) | 0.95 |
| Range | 18–89 | 14–80 | |
| AST ‡ | |||
| Median (IQR) | 42 (30–56) | 36.5 (27–52) | 0.63 |
| Range | 20–170 | 19–180 | |
| Eosinophil§ (%) | |||
| Median (IQR) | 25 (14–38) | 32.5 (22.5–41) | 0.15 |
| Range | 0–60 | 1–69 | |
| Eosinophil§ count (cells/μL) | |||
| Median (IQR) | 2,014 (1,010–3,770) | 3,150 (1,870–4,720) | 0.22 |
| Range | 0–9,710 | 554–23,200 | |
| > 400 counts/μL: no. of cases/total cases (%) | 43/46 (93.5) | 48/48 (100) | 0.11† |
| Serum Ab ¶ (ELISA) | |||
| Median (IQR) | 5,152 (3,872–6,400) | 5,168 (3,200–6,400) | 0.93 |
| Range | 1,600–12,800 | 0–14,400 | |
| Liver lesion detected by ultrasound|| | 0.93† | ||
| Improved: no. of cases (%) | 5/46 (10.9) | 5/48 (10.4) | |
| Unchanged: no. of cases (%) | 36/46 (78.3) | 39/48 (81.3) | |
| Worsened: no. of cases (%) | 5/46 (10.9) | 4/48 (8.3) | |
Endpoints at 3 months receiving study treatments by ITT analysis
| No. of cases with endpoint (%) | ||||
|---|---|---|---|---|
| Endpoint | In triclabendazole group | In artesunate group | Relative risk (95% CI of RR)* | P value † |
| * Treatment with triclabendazole as reference group. | ||||
| † Exact test. | ||||
| No clinical symptoms (abdominal pain, fever, jaundice, etc.) | 46/50 (92) | 38/50 (76) | 0.83 (0.69–0.98) | 0.05 |
| Eosinophil ≤ 400 counts/μL | 21/50 (42) | 8/50 (16) | 0.38 (0.19–0.78) | 0.008 |
| Improvement in ultrasound scan appearances | 35/50 (70) | 33/50 (66) | 0.94 (0.72–1.23) | 0.83 |
| Complete response | 18/50 (36) | 5/50 (10) | 0.28 (0.11–0.69) | 0.004 |
Endpoints at 3 months receiving study treatments by PP analysis
| No. of cases with endpoint (%) | ||||
|---|---|---|---|---|
| Endpoint | In triclabendazole group | In artesunate group | Relative risk (95% CI of RR)* | P value † |
| * Treatment with triclabendazole as reference group. | ||||
| † Exact test. | ||||
| No clinical symptoms (abdominal pain, fever, jaundice, etc.) | 46/46 (100) | 38/42 (90.5) | 0.90 (0.82–1.0) | 0.048 |
| Eosinophil ≤ 400 counts/μL | 21/46 (45.7) | 8/42 (19.0) | 0.42 (0.21–0.84) | 0.012 |
| Improvement in ultrasound appearances | 35/46 (76.1) | 33/42 (78.6) | 1.03 (0.82–1.29) | 0.80 |
| Complete response | 18/46 (39.1) | 5/42 (11.9) | 0.30 (0.12–0.75) | 0.007 |
Address correspondence to Jeremy N. Day, Oxford University Clinical Research Unit, Hospital for Tropical Diseases, Ho Chi Minh City, Vietnam. E-mail: jday@oucru.org
Authors’ addresses: Tran Tinh Hien, Ng Thanh Truong, Nguyen Thi Dung, Nguyen Thi Hue, and Tran Kim Dung, Hospital for Tropical Diseases, 190 Ben Ham Tu, Quan 5, Ho Chi Minh City, Vietnam. Nguyen Hoang Minh and Hoang Dinh Dat, Binh Dinh Provincial Hospital, Qui Nhon City, Vietnam. Phung Quoc Tuan, James I. Campbell, Jeremy J. Farrar, and Jeremy N. Day, Oxford University Clinical Research Unit, Hospital for Tropical Diseases, 190 Ben Ham Tu, Quan 5, Ho Chi Minh City, Vietnam, Telephone: +84 8923 7954, Fax: +84 8923 8904, E-mail: jday@oucru.org.
Financial support: The study was funded by the Wellcome Trust, United Kingdom. Triclabendazole (Egaten) was provided free as an unrestricted gift from Novartis International AG, Basel, Switzerland.
Disclaimer: The supporting bodies had no role in study design, running of the trial, result collection, analysis, or authorship of this work.
REFERENCES
- 1↑
Esteban JG, Gonzalez C, Curtale F, Munoz-Antoli C, Valero MA, Bargues MD, el-Sayed M, el-Wakeel AA, Abdel-Wahab Y, Montresor A, Engels D, Savioli L, Mas-Coma S, 2003. Hyperendemic fascioliasis associated with schistosomiasis in villages in the Nile Delta of Egypt. Am J Trop Med Hyg 69 :429–437.
- 2↑
Mas-Coma S, Bargues MD, Esteban JG, 1999. Human fasciolosis. Dalton JP, ed. Human Fasciolosis. Wallingford, United Kingdom: CAB International Publishing, 411–434.
- 3↑
Apt W, Aguilera X, Vega F, Miranda C, Zulantay I, Perez C, Gabor M, Apt P, 1995. Treatment of human chronic fascioliasis with triclabendazole: drug efficacy and serologic response. Am J Trop Med Hyg 52 :532–535.
- 4
el-Karaksy H, Hassanein B, Okasha S, Behairy B, Gadallah I, 1999. Human fascioliasis in Egyptian children: successful treatment with triclabendazole. J Trop Pediatr 45 :135–138.
- 5
Mas-Coma S, Bargues MD, Valero MA, 2005. Fascioliasis and other plant-borne trematode zoonoses. Int J Parasitol 35 :1255–1278.
- 6↑
Millan JC, Mull R, Freise S, Richter J, 2000. The efficacy and tolerability of triclabendazole in Cuban patients with latent and chronic Fasciola hepatica infection. Am J Trop Med Hyg 63 :264–269.
- 7↑
Talaie H, Emami H, Yadegarinia D, Nava-Ocampo AA, Massoud J, Azmoudeh M, Mas-Coma S, 2004. Randomized trial of a single, double and triple dose of 10 mg/kg of a human formulation of triclabendazole in patients with fascioliasis. Clin Exp Pharmacol Physiol 31 :777–782.
- 8
Alvarez-Sanchez MA, Mainar-Jaime RC, Perez-Garcia J, Rojo-Vazquez FA, 2006. Resistance of Fasciola hepatica to triclabendazole and albendazole in sheep in Spain. Vet Rec 159 :424–425.
- 9
Gaasenbeek CP, Moll L, Cornelissen JB, Vellema P, Borgsteede FH, 2001. An experimental study on triclabendazole resistance of Fasciola hepatica in sheep. Vet Parasitol 95 :37–43.
- 10
Moll L, Gaasenbeek CP, Vellema P, Borgsteede FH, 2000. Resistance of Fasciola hepatica against triclabendazole in cattle and sheep in The Netherlands. Vet Parasitol 91 :153–158.
- 11↑
Overend DJ, Bowen FL, 1995. Resistance of Fasciola hepatica to triclabendazole. Aust Vet J 72 :275–276.
- 12↑
Favennec L, Jave Ortiz J, Gargala G, Lopez Chegne N, Ayoub A, Rossignol JF, 2003. Double-blind, randomized, placebo-controlled study of nitazoxanide in the treatment of fascioliasis in adults and children from northern Peru. Aliment Pharmacol Ther 17 :265–270.
- 13↑
Espinoza JR, Maco V, Marcos L, Saez S, Neyra V, Terashima A, Samalvides F, Gotuzzo E, Chavarry E, Huaman MC, Bargues MD, Valero MA, Mas-Coma S, 2007. Evaluation of Fas2-ELISA for the serological detection of Fasciola hepatica infection in humans. Am J Trop Med Hyg 76 :977–982.
- 14↑
Mottier L, Alvarez L, Fairweather I, Lanusse C, 2006. Resistance-induced changes in triclabendazole transport in Fasciola hepatica: ivermectin reversal effect. J Parasitol 92 :1355–1360.
- 15↑
Borrmann S, Szlezak N, Faucher JF, Matsiegui PB, Neubauer R, Binder RK, Lell B, Kremsner PG, 2001. Artesunate and praziquantel for the treatment of Schistosoma haematobium infections: a double-blind, randomized, placebo-controlled study. J Infect Dis 184 :1363–1366.
- 16
De Clercq D, Vercruysse J, Kongs A, Verle P, Dompnier JP, Faye PC, 2002. Efficacy of artesunate and praziquantel in Schistosoma haematobium infected schoolchildren. Acta Trop 82 :61–66.
- 17
Inyang-Etoh PC, Ejezie GC, Useh MF, Inyang-Etoh EC, 2004. Efficacy of artesunate in the treatment of urinary schistosomiasis, in an endemic community in Nigeria. Ann Trop Med Parasitol 98 :491–499.
- 18↑
Keiser J, Shu-Hua X, Tanner M, Utzinger J, 2006. Artesunate and artemether are effective fasciolicides in the rat model and in vitro. J Antimicrob Chemother 57 :1139–1145.
- 19↑
Utzinger J, Chollet J, Tu Z, Xiao S, Tanner M, 2002. Comparative study of the effects of artemether and artesunate on juvenile and adult Schistosoma mansoni in experimentally infected mice. Trans R Soc Trop Med Hyg 96 :318–323.
- 20↑
Dondorp A, Nosten F, Stepniewska K, Day N, White N, 2005. Artesunate versus quinine for treatment of severe falciparum malaria: a randomised trial. Lancet 366 :717–725.
- 22↑
Flohr C, Tuyen LN, Lewis S, Quinnell R, Minh TT, Liem HT, Campbell J, Pritchard D, Hien TT, Farrar J, Williams H, Britton J, 2006. Poor sanitation and helminth infection protect against skin sensitization in Vietnamese children: a cross-sectional study. J Allergy Clin Immunol 118 :1305–1311.
- 23↑
Curtale F, Hassanein YA, Savioli L, 2005. Control of human fascioliasis by selective chemotherapy: design, cost and effect of the first public health, school-based intervention implemented in endemic areas of the Nile Delta, Egypt. Trans R Soc Trop Med Hyg 99 :599–609.
- 24↑
Richter J, Freise S, Mull R, Millan JC, 1999. Fascioliasis: sonographic abnormalities of the biliary tract and evolution after treatment with triclabendazole. Trop Med Int Health 4 :774–781.
- 25↑
Mas-Coma MS, Esteban JG, Bargues MD, 1999. Epidemiology of human fascioliasis: a review and proposed new classification. Bull World Health Organ 77 :340–346.
- 26↑
Keiser J, Shu-Hua X, Jian X, Zhen-San C, Odermatt P, Tesana S, Tanner M, Utzinger J, 2006. Effect of artesunate and artemether against Clonorchis sinensis and Opisthorchis viverrini in rodent models. Int J Antimicrob Agents 28 :370–373.