Urinary schistosomiasis is caused by the blood fluke, Schistosoma haematobium. The disease, which causes chronic ill-health, is endemic in most African and Eastern Mediterranean countries (Chitsulo 2000; Engels 2002; Steinmann 2006). It is especially important in poor, rural areas where attempts to alleviate poverty also promote water resources development that may increase transmission and hence exacerbate the disease burden (Danso-Appiah 2004; Fenwick 2006b; Steinmann 2006). In some areas of sub-Saharan Africa there is an overlap in distribution with S. mansoni resulting in mixed infections (WHO 2002). The two parasites infect about 131 million people (Davis 2003) and are associated with considerable morbidity and even mortality (van der Werf 2003). A recent meta-analysis suggested that the burden due to schistosomiasis has been significantly underestimated, since disability weights might be two to 15 times higher than previously estimated (King 2005). The social and economic burden of schistosomiasis is thought to be even greater (WHO 2002).
Mode of infection
The infection is acquired through contact with freshwater infested with the infective cercariae shed from the intermediate host snail (Bulinus spp.). Once cercariae have penetrated the human skin, the parasites develop into the adult worm within, on average, 63 to 65 days (Smith 1976; Ghandour 1978), and the worms usually migrate to the blood vessels draining the bladder where they reside and produce large numbers of eggs. On average, adult worm pairs live for three to five years, but some can live up to 30 years with the reproduction potential of one schistosome pair estimated to be up to 600 billion schistosomes (Gryseels 2006). The eggs of S. haematobium have a terminal spine and must traverse the bladder tissues towards the lumen of the bladder and urinary tract for elimination via urine. In the process, a considerable number become trapped in the bladder walls and surrounding tissues to initiate immune-induced inflammatory reactions, which subsequently lead to morbidity. It is important to note that eggs trapped in the tissues cause disease rather than the worms themselves.
Symptoms and effects
The disease can present as chronic, which is most common, or acute. Haematuria (blood in urine) and dysuria (painful urination) are the main early symptoms of the disease. For most people who are regularly exposed, the severity of disease depends upon the intensity of infection. Mostly individuals with few schistosome worms, and especially adults, remain asymptomatic, although about 80% of infected children show early symptoms and signs of disease (Mott 1983; Olds 2000). Late-stage complications are insidious and include calcification of the bladder wall, bladder stones, and secondary bacterial infection (Jordan 1993). Tissue damage caused by trapped eggs can lead to diffuse or localized wall thickening of the bladder and the distal ureter hydronephrosis or hydroureter, which may eventually lead to kidney failure (Kardorff 2001; WHO 2002; van der Werf 2003).
Elevated urine albumin levels and reported pain upon micturition by children have a strong correlation with S. haematobium infection (Rollinson 2005). An important long-term consequence of infection is squamous cell carcinoma of the bladder (Jordan 1993; King 2005; Shiff 2006). A recent review points out that bladder carcinoma is the seventh most common cancer worldwide in men and that the highest incidence rate among men is found in Egypt (37.1 per 100,000 person-years) (Murta-Nascimento 2007), which might be related to S. haematobium infection and morbidity (Jordan 2000). Eggs produced in venous blood vessels elsewhere such as the vertebral column, and resulting in granuloma formation, may cause spinal cord compression and neurological complications. Severe chronic disease occurs later in life following the infection, and many deaths are rarely acknowledged to be due to schistosomiasis because there is hardly any recognition of the link between infection in early life and later development of severe disease.
Sustained heavy infection leads to iron deficiency anaemia and other nutritional deficiencies, especially in children (Awasthi 2003; King 2005). The disease often results in retarded growth, reduced physical activity, and impaired cognitive function in children (Stephenson 1993; Nokes 1999; PCD 1999; Jukes 2002; WHO 2002).
Parasitological diagnosis by microscopy of urine for parasite eggs is the most practical and widely used method for identifying infected individuals (Hassan 1994). Egg output in urinary schistosomiasis can be influenced by several factors, such as time of collection of urine (peak egg excretion occurs around noon), day-to-day variations, seasonal variations, and environmental conditions (Braun-Munzinger 1992). Therefore negative results following microscopic examination of a single urine specimen, as with a single stool for intestinal schistosomiasis, are not reliable, particularly in areas characterized by low intensities of infection (de Vlas 1992). Indeed, measurement of prevalence and intensities of infection by egg count has shortcomings (Gryseels 1996; de Vlas 1997; Utzinger 2001b). Egg count is quantified using a nucleopore membrane by urine filtration of a standard 10 mL volume of urine. Reagent strips for detecting blood in the urine (haematuria), and recently, monoclonal antibody-based dipstick tests for detecting schistosome-specific by-products are used to diagnose the disease (Bosompem 1997; Bosompem 2004). Clinically, the disease is diagnosed by reported terminal blood after urination or by inspecting urine for haematuria. Diagnosis on the basis of presence of blood in urine is less reliable in adults (RUSG 1995; Ansell 1997). This is because blood in the urine of an adult may be due to causes other than urinary schistosomiasis. Ultrasound was introduced in the 1970s to detect schistosomal pathology first in the hospital and then in field studies (Hatz 2001). It is a safe, rapid, non-invasive, and relatively inexpensive technique for assessing bladder or urinary tract pathology both in the hospital and in community surveys (Hatz 1990).
Disease control strategies
There is no effective antischistosomal vaccine (Gryseels 2000; Fenwick 2006a), although significant progress has been made in recent years (McManus 2008). Therefore, schistosomiasis control programmes have the primary objective of reducing the burden of disease. Four main control strategies have been employed with varying success.
- Health education to promote good hygiene and sanitation, especially among school-aged children and caregivers. It discourages practices such as bathing in streams and indiscriminate disposal of refuse that tend to increase risk of the infection. The ultimate goal is to decrease the number of eggs reaching and contaminating the environment, particularly freshwater bodies. However, the long-term impact of health education on the transmission of schistosomiasis in rural traditional communities is questionable (Kloos 1995; Sow 2003).
- Water supply and sanitation to reduce frequency of water contact for most domestic activities such as fetching water for drinking, washing clothing, or bathing in streams and ponds; and access to adequate sanitation to avoid environmental contamination with parasite eggs.
- Control of the intermediate host snail by environmental management such as removal of vegetation around banks of streams and lining irrigation canals with concrete slabs (Steinmann 2006); and treating infested water bodies with molluscicide to destroy the intermediate host snail. The important role environmental management as part of an integrated control approach has played in conquering S. japonicum in China has been emphasized (Utzinger 2005).
- Morbidity control by chemotherapy of the human population aims to reduce disease burden and thereby transmission. Past control measures focused largely on reducing or interrupting transmission, but such measures have not been sustainable due to high cost and operational difficulties (WHO 2002). The advent of safe, efficacious, and inexpensive drugs shifted the emphasis to morbidity control in areas of high disease burden, endorsed by the World Health Organization (WHO) in the mid-1980s (WHO 1993; WHO 2002), while in low-burden areas the emphasis is to interrupt transmission of the infection. Although chemotherapy has emerged as the most cost-effective control strategy because of availability of inexpensive drugs, it has been suggested that in most endemic areas addition of preventive measures focusing on clean water, adequate sanitation, and health education to complement chemotherapy is necessary to achieve long-term sustainable schistosomiasis control (Utzinger 2001a; Singer 2007).
Chemotherapy is targeted especially at school-aged children (Magnussen 2001; WHO 2002; Savioli 2004). The assumption is that reducing the worm burden in childhood, when infection intensity is highest, will prevent most long-term complications occurring later in adulthood.
Several drugs have been used or tried for the treatment of urinary schistosomiasis and later abandoned because of poor effect or adverse events: antimonials, niridazole, lucanthone, hycanthone, oltipraz, cyclosporin A, levamisole, and oxamniquine; see Cioli 1995 for a comprehensive review.
Current treatment options are limited to praziquantel and metrifonate.
- Praziquantel. Praziquantel is the only drug on the WHO Model List of Essential Medicines for treating S. haematobium. This broad-spectrum antischistosomal drug is effective against all Schistosoma species, although it is refractory against immature parasites (Sabah 1986). Praziquantel is administered orally at a standard dose of 40 mg/kg body weight. The most common adverse effects are gastrointestinal, including abdominal pain, nausea, vomiting and diarrhoea, and are usually mild and last less than 24 hours.
- Metrifonate. Metrifonate was introduced as a drug for humans in the 1960s (Snellen 1981) and has been used extensively to treat urinary schistosomiasis. The standard dose of 7.5 to 10 mg/kg given three times at 14-day intervals has been used extensively and is mostly well tolerated (Forsyth 1967; Davis 1969; Rugemalila 1981; Feldmeier 1987). Adverse effects are mainly as a result of cholinergic stimulation and include fatigue, muscular weakness, tremor, sweating, salivation, fainting, abdominal colic, diarrhoea, nausea, vomiting, and bronchospasm. Its use has been limited after a suggestion that it was inferior clinically, economically, and operationally to praziquantel (Feldmeier 1999). Subsequently, metrifonate was withdrawn from the WHO Model List of Essential Medicines (Cioli 2000; Utzinger 2004).
Other drugs have potential as treatment options for urinary schistosomiasis, such as artemisinin derivatives, albendazole, and amoscanate. Albendazole is often administered together with praziquantel for simultaneous control of schistosomiasis and soil-transmitted helminthiasis.
- Artemisinins. The antischistosomal activity of the artemisinins, such as artesunate and artemether, was discovered in the early 1980s (Le 1982; Le 1983). The artemisinins are active against the liver stages (immature) worms, while the invasive stages and adult worms are less susceptible to the drugs. Adverse effects are minor and last for less than 24 hours. Artemisinin monotherapy may not be beneficial due to stage-specific activity, but combination with existing drugs effective against other stages (eg praziquantel) may improve therapeutic efficacy.
- Albendazole. Albendazole is indicated for the treatment of a variety of worm infestations. In recent years it has often been co-administered with praziquantel with the goal of simultaneously controlling schistosomiasis and soil-transmitted helminthiasis (Friis 2003; Zhang 2007). Albendazole is administered orally (usually as single 400 mg dose), and reported adverse effects include gastrointestinal upsets, headaches, and dizziness, while rash, fever, elevated liver enzymes, and hair loss occur less frequently. There have been reports of elevated liver enzymes, headaches, loss of hair, low levels of white blood cells (neutropenia), fever, and itching if taken at higher doses and/or for a long period of time.
- Amoscanate. Amoscanate is a broad-spectrum anthelminthic drug that exhibits activity against all major human schistosome parasites (Striebel 1976), other systemic parasites (eg filariae), and gastrointestinal nematodes (eg hookworms). It has been tested extensively in China using the locally produced equivalent called 'nithiocyaminum' (Bueding 1976; Striebel 1976). Toxicity in experimental animals was quite low, and mutagenicity tests in bacteria gave negative results; however, mutagenic metabolites were detected in urine of mammals given amoscanate (Batzinger 1977). It was abandoned because of concerns over liver toxicity and availability of better drugs, such as praziquantel (Cioli 1995). It is possible that amoscanate may represent a unique, broad-spectrum schistosomicide with the appropriate structural modifications to decrease liver toxicity (Cioli 1995).
Combinations of antischistosomal drugs have also been tested with the aim of improving therapeutic efficacy.
- Artemisinin derivatives (artesunate or artemether) plus praziquantel. This combination is suggested because artesunate and artemether are effective against immature worms, and artemether has shown in mouse models to prevent infection. Combining artesunate or artemether with praziquantel, which is effective against adult worms, may improve therapeutic efficacy.
- Metrifonate plus praziquantel. The rationale for this combination is that both drugs are independently effective against S. haematobium and that their targets of action in the parasite are not linked. Combination may improve therapeutic efficacy by offering mutual protection to each drug, and it may also slow or prevent the development of resistance.
- Albendazole plus praziquantel. Albendazole has broad activity, and it has been suggested that combining with praziquantel may help improve therapeutic efficacy. This combination has not been tested widely.
Praziquantel is virtually the only drug currently available for clinical management and control of urinary schistosomiasis. The sharp reduction in price of praziquantel has stalled advancement of other potential control options, such as vaccines, new drugs, and diagnostics (Utzinger 2007). It is noteworthy that pressure on praziquantel is growing, following the policy adopted at the 54
To evaluate antischistosomal drugs, used alone or in combination, for treating urinary schistosomiasis. Specifically:
- Praziquantel, metrifonate, and artemisinin derivatives versus placebo; and to assess the appropriate dose for each from randomized comparisons by dose.
- Praziquantel versus metrifonate.
- Praziquantel plus other drugs (eg metrifonate, albendazole, or artemisinins) versus praziquantel alone.
Other relevant drugs or comparisons will be included in the future if they help address relevant safety, efficacy, or policy questions.
Criteria for considering studies for this review
Types of studies
Randomized and quasi-randomized controlled trials.
Types of participants
Individuals infected with S. haematobium diagnosed either microscopically for the presence of S. haematobium eggs in a standard filtrate of 10 mL of urine or by haematuria in endemic areas.
Types of interventions
Praziquantel, metrifonate, artemisinin derivatives, or albendazole alone or in combination versus placebo or different doses of same drug; or other relevant antischistosomal drugs.
Types of outcome measures
Parasitological failure, defined as treated individuals who remained positive for eggs in the urine at follow up (distinguishing between one to three and three to 12 months post-treatment). Egg reduction rate (one to three or three to 12 months post-treatment).
- Reduction in the percentage of people with a heavy infection (currently defined as ≥ 50 eggs/10 mL urine (WHO 2002).
- Clearance of haematuria.
- Measures of anaemia (mean haemoglobin; proportion of participants anaemic).
Functional indices (measured by standardized replicable techniques)
Resolution of bladder or urinary tract pathology, as measured by ultrasound, by standard international classification (CWG 1992; Richter 1996), or other standardized methods. Physical growth, including weight-for-age, height-for-age, weight-for-height, upper mid-arm circumference, and triceps skinfold thickness. Physical fitness. Cognitive function and educational achievement.
- Serious (fatal, life-threatening, requiring hospitalization, or discontinuation of treatment).
Search methods for identification of studies
We attempted to identify all relevant trials regardless of language or publication status (published, unpublished, in press, and ongoing).
We searched the following databases using the search terms and strategy described in Table 1: Cochrane Infectious Diseases Group Specialized Register (August 2007); Cochrane Central Register of Controlled Trials (CENTRAL), published in The Cochrane Library (2007, Issue 3); MEDLINE (1966 to August 2007); EMBASE (1974 to August 2007); and LILACS (1982 to August 2007). We also searched the metaRegister of Controlled Trials (mRCT) using 'Schistosoma haematobium' as the search term (August 2007).
Researchers and organizations
We contacted individual researchers working in the field and experts from the UNICEF/UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases (TDR) for unpublished data and information on ongoing trials.
We checked the reference lists of all studies identified by the above methods.
Data collection and analysis
Selection of studies
Anthony Danso-Appiah (ADA), with assistance from Vittoria Lutje, the Cochrane Infectious Diseases Group (CIDG) Information Retrieval Specialist, searched the literature and retrieved studies. ADA screened the results to identify potentially relevant trials and assessed the eligibility of trials for inclusion in the review using an eligibility form based on the inclusion criteria; Paul Garner (PG) verified these procedures. ADA scrutinized each trial to ensure it has been included only once. If different parts of the same data were reported in different publications, ADA identified them and linked the data to the parent study. ADA attempted to contact the authors of potentially relevant trials for clarification if eligibility was unclear and listed all potential studies excluded along with the reason for exclusion in the Characteristics of excluded studies.
Data extraction and management
ADA extracted data of trial characteristics such as methods, participants, interventions, and outcomes. ADA recorded the data on standard forms, which PG cross-checked. ADA and PG resolved discrepancies through discussion and contacted Jianping Liu (JPL), Piero Olliaro (PO), and Jürg Utzinger (JU) on technical issues. Data were double-entered and cross-checked to make sure there were no errors. ADA scrutinized each trial to identify multiple publications from a single data set and attempted to contact trial authors for clarification, or insufficient or missing data. ADA extracted the number of participants randomized and number analysed in each treatment group, which allowed us assess the most appropriate type of analysis to carry out and to calculate the percentage loss to follow up. For dichotomous outcomes, ADA recorded the number of participants experiencing the event in each group of the trial. For continuous outcomes summarized using geometric mean, ADA extracted means and their standard deviations on the log scale when provided. If the data were provided as arithmetic mean, ADA extracted the means for each group and their standard deviations (SD), standard error (SE), or confidence interval (CI), where possible.
Stratified data were extracted according to the stratifications and follow-up times. Most included trials defined intensity of infection by egg count as light, moderate, and heavy (instead of according to WHO 2002), and we based the treatment failure rate on these categories. We extracted information such as brand of drug used, dose, participant age, diagnostic criteria, endemicity, whether the trial was hospital- or community-based, and whether there had been simultaneous application of other control measures during the trial (eg health education or use of molluscides). To allow assessment of the interdependence between observations in a trial, we extracted data on repeated follow ups and number of communities involved in each trial. Data on haematuria from King 2002 were extracted from graphs.
Assessment of risk of bias in included studies
ADA examined design issues relating to internal validity, and PG checked the assessment. Generation of allocation sequence was described as adequate if the method used indicated that the resulting sequences were unpredictable, unclear if trial was randomized but method not described, inadequate if sequences could be predicted, or not described (Jüni 2001). Allocation concealment was described as adequate if methods used prevented prior knowledge of investigators enrolling participants and participants of treatment assignment, inadequate if participants and investigators enrolling participants could foresee upcoming assignment, or not described (Jüni 2001). ADA noted who was blinded to the interventions, such as the participants, care providers, or outcome assessors. The inclusion of all randomized participants in the main analysis was assessed as adequate if more than 90% were included in the analysis, inadequate if 90% or less, or unclear. Given that these cut-offs are arbitrary and subject to sample size for a given study, ADA also reported actual percentages. ADA reported the overall number randomized and the number included in the review for trials not using all the trial arms in the analysis.
Review Manager 4.2 was used for the statistical analyses and dichotomous outcomes (failure rates) were presented as risk ratios (RR) with 95% confidence intervals (CI). To minimize selection bias and the effect of participant attrition, we calculated the proportion of parasitological failure from the total number of participants at follow up and conducted per protocol analysis. We considered RR to be more appropriate because event rates were high. We intended to analyse by intention-to-treat, but this was not possible due to the lack of information in some trial reports. Continuous data were presented as weighted mean differences (WMD) with their standard deviation (SD) or standard error (SE). Egg counts were reported mostly as percentage reduction in geometric mean with rates of reduction over 90% across trials irrespective of background drug or dose. Because treatment effects were obvious in terms of egg excretion, we decided to report them in a table instead of combining in a meta-analysis.
The effects were obvious in comparisons against placebo; therefore we restricted the analysis to the two primary outcomes, three secondary outcomes, and adverse events. We expressed them by number-needed-to-treat (NNT), where possible, and related this to background endemicity.
The impact of follow-up time on cure rate has been elucidated and interpreted from the analysis of available research data; short follow-up times give better treatment effect in terms of parasitological cure than long follow-up times of same background drug and endemicity (Danso-Appiah 2002). To account for this, we analysed treatment failure based on two follow-up categories as short (one to three months) and long (three to 12 months), and also according to dose.
Where data were sufficient we conducted sensitivity analyses to assess the robustness of the results to the quality components. We tested for heterogeneity using the chi-squared and I
Description of studies
Twenty-four trials (6315 participants), reported in 35 published articles, met the inclusion criteria (see Characteristics of included studies); none were cluster-randomized. Four articles were published from the same trial data (King 1988), and another three from the same study (Stephenson 1989). Wilkins 1987a reported two trials, but we included one (Nyamari trial named Wilkins 1987a) and excluded the other (Simote trial named Wilkins 1987b) because the latter did not randomize the participants. Nineteen trials were excluded from the review (see Characteristics of excluded studies).
Of the 24 trials included in the analysis, 20 evaluated praziquantel (eight specified Biltricide (Bayer)). Nine trials assessed metrifonate (three specified Bilarcil (Bayer)). Three trials assessed the combination of praziquantel with albendazole, and one trial assessed praziquantel plus artesunate. For the two primary outcomes, 21 trials reported cure rate or failure rate, and 20 reported egg reduction rate. Nine trials reported adverse events. There was lack of uniformity in diagnostic criteria ( Table 2) and classification of intensity of infection across trials ( Table 3). The WHO classifies the intensity of infection as light (1 to 49 eggs/10 mL urine) or heavy (≥ 50 eggs/10 mL urine) (WHO 2002). However, the trials used different classifications for light infection (eg 1 to 5, 1 to 29, 60 to 249, and 250 to 500 eggs/10 mL urine). Moderate and heavy infections were classified the same way with often considerable overlaps between intensity categories.
Trial setting and participants
The trials were conducted in Africa: nine in East Africa (six in Kenya and three in Tanzania); five in Southern Africa; four in the Horn of Africa (three in Sudan and one in Somalia); four in West Africa; and two in Central Africa. Nineteen trials were conducted in the 1980s, shortly after praziquantel was introduced in the market, one in the early 1990s, and three in the new millennium. Twenty-two trials involved children aged up to 15 years; the other two trials recruited only boys (Doehring 1985; Befidi-Mengue 1992). Four trials recruited children with mixed infection of S. haematobium and S. mansoni (Jewsbury 1977; Doehring 1985; Kardaman 1985; Taylor 1988). Participants were identified in community surveys in all except two trials that recruited patients attending hospital (Davis 1981) or a combination of patients attending hospital and participants detected during a field survey (Omer 1981).
Risk of bias in included studies
The methods used to generate the allocation sequence were adequate in the 11 trials that used computer-generated numbers, random-number tables, randomized cards, permutation table, or randomized block design. One trial used sequential allocation (inadequate; Pugh 1983), and the methods used to generate the allocation sequence were unclear in 12 trials. Only three trials used adequate methods to conceal allocation (Aden Abdi 1989; Olds 1999; Borrmann 2001); the methods were unclear in the remaining 21 trials. Eight trials employed blinding and described who was blinded (six were double-blind and two single-blind); the remaining were unclear. For follow up at one to three months, 17 trials included 90% or more participants in the analysis (adequate), and two trials were unclear. For follow up at three to 12 months, 12 trials included 90% or more participants in the analysis (adequate) and five trials were unclear.
Effects of interventions
1. Metrifonate versus placebo
Jewsbury 1977 measured parasitological failure at one to three months and showed a marked effect in favour of metrifonate (RR 0.42, 95% CI 0.27 to 0.64; 64 participants, Analysis 1.1), but loss to follow up was high (44%). The effect also favoured metrifonate when failure was measured at three to 12 months in Jewsbury 1977, Stephenson 1985, and Stephenson 1989 (RR 0.53, 95% CI 0.29 to 0.95; 680 participants, Analysis 1.1), although there was significant heterogeneity.
Loss to follow up was still high in Jewsbury 1977, but less marked in the other two trials (Stephenson 1985; Stephenson 1989). In terms of differences in failure rates, there seemed to be an association with the level of endemicity: Jewsbury 1977 and Stephenson 1989 (high endemicity) led to higher rates of failure at three to 12 months than Stephenson 1985 (low endemicity), but the lower dose used in Stephenson 1989 may confound the observed higher failure rate. There was no obvious association of failure with age (all trials included children of up to 15 years) or follow up (all three trials measured failure at eight months).
Egg reduction rate
All four trials measured this at three to 12 months and demonstrated that metrifonate reduced egg excretion by over 90%. The placebo groups ranged from a 5.5% decrease to a 66.2% increase ( Table 5).
Two trials, Stephenson 1985 and Stephenson 1989, showed that participants in the metrifonate group had higher levels of mean haemoglobin than those in the placebo group (RR 0.30, 95% CI 0.28 to 0.32; 607 participants, Analysis 1.2).
Jewsbury 1977 assessed adverse events and recorded none.
2. Praziquantel versus placebo
Praziquantel (40 mg/kg x 1 oral) was superior to placebo at one to three months' follow up (RR 0.39, 95% CI 0.27 to 0.55; 534 participants, 4 trials, Analysis 2.1) and at three to 12 months (RR 0.23, 95% CI 0.14 to 0.39; 433 participants, 3 trials, Analysis 2.1). There was significant heterogeneity in the meta-analysis, possibly due to loss to follow up, which was high in McMahon 1979 (31.6% and 36.9% for short and long follow-up times, respectively), less than 10% for Stephenson 1989, Olds 1999, and Borrmann 2001, and unreported in Taylor 1988.
Egg reduction rate
Praziquantel had egg reduction rates of over 98% (geometric mean) in four trials and a 95% rate in Befidi-Mengue 1992, and these were greater than those achieved with the placebo (5.3% to 64%). Doehring 1985 reported a median reduction rate of 98.7% in the praziquantel group and 48.6% in the placebo group. The trials used different dosing schedules, but there was no clear relationship between the egg reduction rates and dosing schedules ( Table 5).
Olds 1999 recorded 15% excess of mild to moderate adverse events with praziquantel compared with placebo, and Borrmann 2001 reported combined events across comparison groups (127 mild and 6 moderate events); see Table 6. Neither trial recorded serious adverse events.
3. Artesunate versus placebo
One trial, Borrmann 2001, which had two months' follow up, made this comparison.
There was no obvious benefit with artesunate (118 participants, Analysis 3.1).
Egg reduction rate
There was no significant difference in the egg reduction rate at two months' follow up (ERR
There was no clear difference between artesunate and placebo at two months (65% versus 53%).
Adverse events were reported as combined events (127 mild and six moderate events, Table 6) and not by comparison group. No serious adverse events were reported.
4. Praziquantel plus artesunate versus placebo
One trial with two months' follow up made this comparison (Borrmann 2001).
There was a clear difference between the combination and placebo for failure rates at two months (RR 0.24, 95% CI 0.15 to 0.38; 118 participants, Analysis 4.1).
Egg reduction rate
The egg reduction rate was high for the combination compared with placebo (ERR
The urine erythrocyte counts were similar for the combination and placebo (65% versus 53%).
There were 127 mild and six moderate adverse events reported, but they were not separated by intervention group ( Table 6).
5. Praziquantel plus albendazole versus placebo
Praziquantel plus albendazole significantly reduced parasitological failures compared to placebo (RR 0.45, 95% CI 0.35 to 0.59; 471 participants, 3 trials, Analysis 5.1). Jinabhai 2001, which was conducted in a low-endemic area, showed a better effect compared with Beasley 1999 (moderate and high endemicities) or Olds 1999 (very high endemicity).
Egg reduction rate
6. Metrifonate versus praziquantel
Some early studies investigated a single dose of 10 mg/kg metrifonate (the standard dose is 7.5 to 10 mg/kg three times at 14-day intervals) with the standard single dose of 40 mg/kg praziquantel. Although the single metrifonate dose was inferior in three trials measuring failure at one to 12 months, the 95% CI were too wide for statistical significance (RR 2.31, 95% CI 0.91 to 5.82; 462 participants, Figure 1), due to significant heterogeneity between the trials (I
|Figure 1. Metrifonate (different regimens) vs praziquantel (30 mg/kg or 40 mg/kg, single dose): Parasitological failure.|
There was no significant difference in failure when metrifonate (10 mg/kg three times at 14-day intervals) was compared with praziquantel (30 mg/kg) in a small trial involving 54 participants (McMahon 1983, Analysis 6.1). The metrifonate regimen was then changed to three doses of 10 mg/kg every four months for one year), and this resulted in effects similar to the standard 40 mg/kg of praziquantel (Figure 1).
Effect on light and heavy infections
One trial reported a subgroup analysis that showed that there was no significant difference between metrifonate (10 mg/kg every four months for one year) and praziquantel (40 mg/kg) curing light infections (626 participants, 1 trial, Analysis 7.1), but that this metrifonate dose was better at controlling heavy infections (615 participants, Analysis 7.2). Given that the subgroup was stratified after randomization, care should be taken in interpreting these results.
Egg reduction rate
Both metrifonate (two and three doses of 10 mg/kg) and praziquantel (single dose 40 mg/kg) led to reductions in egg excretion of over 98% in two trials (McMahon 1983; Doehring 1985), while in three trials a single dose of metrifonate (10 mg/kg) also resulted in an egg reduction of over 90% (Pugh 1983; Wilkins 1987a; Stephenson 1989) ( Table 5).
Stephenson 1989 showed that participants in the metrifonate group had greater mean haemoglobin levels than those in the praziquantel group (RR 0.19, 95% CI 0.17 to 0.21; 208 participants, Analysis 6.2).
McMahon 1983 (54 participants) reported similar minor adverse events between metrifonate (10 mg/kg) and praziquantel (30 mg/kg), except for abdominal pain and vomiting, which occurred more frequently in the metrifonate group than the praziquantel group (40% versus 13% and 8% versus 0%). No serious adverse events were reported. Wilkins 1987a (184 participants) compared metrifonate (10 mg/kg x 1) versus praziquantel (40 mg/kg x 1) and reported no serious adverse event. Commonly reported adverse events for the combination treatment included headache, weakness, dizziness, nausea/vomiting, diarrhoea, abdominal pain, general malaise, and fever. Among these events, abdominal pain, general malaise, and fever were reported more frequently in those treated with praziquantel than metrifonate.
7. Metrifonate regimens: 5 mg/kg x 3, given in one day versus 7.5 mg/kg x 3, given fortnightly
One trial with 201 participants made this comparison (Aden Abdi 1989).
There was no significant difference in parasitological failure (201 participants, Analysis 8.1).
Egg reduction rate
Egg reduction rate (geometric mean) was 96% for the one-day regimen versus 97% for the fortnightly regimen ( Table 5).
There was little difference in the percentage of mild adverse events reported for the fortnightly regimen (7%) versus the one-day regimen (9%) ( Table 6).
8. Metrifonate (10 mg/kg x 1) plus praziquantel (10 mg/kg) versus praziquantel (40 mg/kg)
Wilkins 1987a showed that the combination was inferior to praziquantel at reducing parasitological failure (72 participants, Analysis 9.1). The same trial reported an egg reduction rate of over 90% for the combination therapy ( Table 5).
9. Metrifonate (10 mg/kg x 1) versus metrifonate (10 mg/kg x 1) plus praziquantel (10 mg/kg)
10. Artesunate plus praziquantel versus praziquantel alone
Borrmann 2001 showed no statistically significant difference between the combination and single treatment for parasitological failure (177 participants, Analysis 11.1). There was no obvious difference in egg reduction rates (ERR
11. Different metrifonate doses
Rey 1984 compared three doses with one and two doses of 10 mg/kg metrifonate. There was no significant difference in the number of parasitological failure between two and three doses at one month and four months ( Analysis 12.1). There were fewer parasitological failures with the three-dose regimen over the one-dose regimen at one month's follow up (RR 2.75, 95% CI 1.29 to 5.85; 93 participants) and four months' follow up (RR 1.52, 95% CI 1.03 to 2.25; 111 participants, Figure 2).
|Figure 2. Metrifonate (10 mg/kg x 1) vs metrifonate (10 mg/kg x 3): Parasitological failure.|
12. Different praziquantel doses versus standard dose (40 mg/kg x 1 oral)
There was no significant difference between the standard dose and 20 mg/kg x 2 (4 trials, Figure 3), 30 mg/kg (6 trials, Figure 4), and 20 mg/kg dose (2 trials, Figure 5); these results were similar for follow up at one, three, and six months.
|Figure 3. Praziquantel (2 x 20 mg/kg) vs praziquantel (standard 40 mg/kg): Parasitological failure.|
|Figure 4. Praziquantel (30 mg/kg) vs praziquantel (standard 40 mg/kg): Parasitological failure.|
|Figure 5. Praziquantel (20 mg/kg) vs praziquantel (standard 40 mg/kg): Parasitological failure.|
Losses to follow up were generally high in some trials, but these did not differ across treatment and control groups within a single trial. There was no significant heterogeneity between the trials, and background endemicities did not seem to play a role; all trial sites had high endemicities except the trial by Davis 1981 (not specified). Examining for a differential effect between heavy and moderate or light infections with 30 mg/kg versus 40 mg/kg, a subgroup analysis of one small trial did not demonstrate a difference (116 participants, King 1989, Analysis 13.5). Here caution should be exercised in the interpretation of the data since the subgroup was selected after randomization.
Egg reduction rate
Five trials all showed no apparent differences in egg reduction rate (geometric mean); all had greater than 95% reduction in both arms, except for Oyediran 1981 in which the 30 mg/kg dose gave an 85.7% reduction compared with 97.7% for the standard dose ( Table 5).
Two trials measured haematuria (King 1989; King 2002). King 1989 (117 participants) showed no difference in the rate of clearance between 30 mg/kg x 1 and the standard 40 mg/kg x 1 dose at three months (100% versus 99%). However, King 2002 (200 participants) showed a clear difference at six weeks' follow up between 20 mg/kg x 1 and the standard 40 mg/kg x 1 (40% versus 63%).
Davis 1981 recorded similar numbers of mild adverse events for each dose: 19%, 29%, and 17% for 30, 40, and 20 mg/kg x 2, respectively. Kardaman 1985 reported slightly higher rates with 20 mg/kg x 2 than the single dose of 40 mg/kg, but no numbers were reported. Neither trial reported serious adverse events ( Table 6). Oyediran 1981 reported combined adverse events across 40, 30, and 20 mg/kg and recorded only two moderately severe events (umbilical pain). No serious adverse events were recorded.
Most of the 24 included trials were conducted many years ago, mostly in the 1970s and 1980s, and thus the standards of methodological quality did not reach the high standards that we would expect from trials carried out today; for example, only four out of the 24 trials used adequate methods to conceal allocation. However, effect sizes are so marked that it is unlikely that methodological quality will have caused such substantive biases to interfere with the marked effects and differences reported.
Both metrifonate and praziquantel showed good effects, but no trial compared the standard dose of each drug in a head-to-head comparison; instead trials compared different doses of each. Given that no trial compared the standard dose of metrifonate (7.5 to 10 mg/kg 3 times at 14-day intervals) with that of praziquantel (40 mg/kg) in a head-to-head assessment, discussion of adherence to treatment from currently available data is limited. However, the failure rate with the recommended standard dose of metrifonate (7.5 to 10 mg/kg 3 times at 14-day intervals) is 19% to 48%, while that of praziquantel (single 40 mg/kg oral dose) is 0% to 37% at one to three months' follow up. A dose of 7.5 mg/kg metrifonate produced more failures than 10 mg/kg, both doses administered three times at 14-day intervals. There appears to be no difference in effects of metrifonate 10 mg/kg given every four months for one year and the standard dose of praziquantel (40 mg/kg), but this may not be conclusive as the evidence came from only one trial (King 1988). Metrifonate (10 mg/kg 3 times at 14-day intervals) showed a similar effect to praziquantel (30 mg/kg). Public health programmes often recommend multiple-dose regimens, such as for metrifonate (3 doses of 7.5 to 10 mg/kg administered once every 14 days or every 4 months), but these are difficult to implement and might compromise overall compliance.
Both metrifonate and praziquantel showed high degrees of uncertainty around their effect estimates as shown by the wide confidence intervals. The small numbers in some of the trials may explain the levels of uncertainty. In this review we have analysed data mainly around infectivity and assumed statistical significance to be equal to clinical significance because it is not likely that small differences in effect of drugs being evaluated can mean large risks or clinical effects.
A single dose of 20 or 30 mg/kg of praziquantel was similarly efficacious compared to the standard dose of 40 mg/kg in terms of all outcomes measured in this review. Given current emphasis on controlling morbidity in high burden areas and morbidity, especially in children, is associated with the number of eggs in an individual (WHO 2002), this finding suggests lower doses of praziquantel may be effective in morbidity control. However, these results should be considered with caution. While it is true that parasite load (expressed by egg counts) is an important factor in both morbidity for the individual patient and environmental contamination (WHO 2002), a sub-curative dose may unduly put the drug under selective pressure and favour parasite resistance (Doenhoff 1998). Pharmacokinetic data of different doses of praziquantel are few and old, and have been obtained in healthy volunteers rather than in patients with schistosomiasis (Leopold 1978). An exponential increase was found in the area under the curve (AUC) with the praziquantel dose in the range of 5 to 50 mg/kg, with a six-fold increase from 20 to 50 mg/kg (Leopold 1978). However, these data do not come from infected patients, and hence cannot be extrapolated so easily. The artemisinins, best known for their use as antimalarial drugs, have been found to be effective against immature schistosomes in laboratory studies (Utzinger 2001a; Utzinger 2001c; Utzinger 2002). However, results from one low-quality trial show that artesunate is not effective when used alone or when combined with praziquantel. This may, to some extent, be explained by the fact that mature worms are less sensitive to the artemisinins (Utzinger 2007).
It has been suggested that there is a significant infection-associated loss of performance in a person with schistosomiasis that can be improved through antischistosomal treatment (Bergquist 2005; King 2005). This would necessitate any comprehensive assessment of antischistosomal drugs to include outcomes of subtle disease such as resolution of bladder or urinary tract pathology, growth, physical fitness, cognitive function, and educational achievement. Most trials did not investigate these outcome measures because the focus tended to be on measures of infectivity. However, we may include functional outcome measures in future updates if trials provide comprehensive data.
The rationale behind the widely spaced dosing interval of metrifonate treatment derives from its long-lasting effect on red blood cells and plasma cholinesterases (Plestina 1972). However, the clinical significance of this effect and why adverse events disappear during the first 12 to 24 hours but the recovery of the enzymes takes more than four to six weeks is not known (Plestina 1972). Safety studies have shown no serious adverse events in patients treated with 5 to 10 mg/kg metrifonate daily for six to 12 days (Snellen 1981), and various reviews of metrifonate's toxicology and pharmacology during its extensive use for urinary schistosomiasis in the 1970s concluded that it had very few adverse events (Holmstedt 1978). Also, metrifonate is currently used in Alzheimer's disease, which requires a high dose and extended regimen, and a systematic review has concluded an overall good tolerability with only mild to moderate adverse events (López-Arrieta 2006). In the current review, although adverse events were generally poorly assessed in the few trials measuring this, no trial recorded a serious adverse event, and no significant differences in the number and type of adverse events between metrifonate and praziquantel were recorded, except for abdominal pain where greater numbers of participants in the metrifonate group were reported with this adverse event.
Implications for practice
Both praziquantel and metrifonate are efficacious (with few adverse events) for treating urinary schistosomiasis, but metrifonate requires multiple administrations and hence is operationally less convenient and more costly in community-based control programmes. However, leaving praziquantel as the only antischistosomal drug raises considerable concern in case resistance develops against this drug. We suggest metrifonate be reconsidered for the WHO Model List of Essential Medicines.
Implications for research
Well-designed trials are required to investigate the following areas.
We would like to express our sincere gratitude to Prof Paul Garner; without his support and guidance this review would not have been possible. We also thank Prof Rashida Barakat, Dr Lester Chitsulo, and Prof Donato Cioli for critically reading this review and for providing useful comments and suggestions, Sarah Donegan for statistical advice, and Gill Gyte for drafting the plain language summary.
This document is funded by the UK Department for International Development (DFID) for the benefit of developing countries. The views expressed are not necessarily those of DFID. J Utzinger acknowledges financial support from the Swiss National Science Foundation (project no. PPOOB-102883 and PPOOB-119-129). P Olliaro is a staff member of the WHO; the authors alone are responsible for the views expressed in this publication and they do not necessarily represent the decisions, policy, or views of the WHO.
N Squires prepared the original version of this review (Squires 1997) with support from the North West Regional Health Authority, UK, and the European Commission (Directorate General XII), Belgium.
Data and analyses
- Top of page
- Authors' conclusions
- Data and analyses
- What's new
- Contributions of authors
- Declarations of interest
- Sources of support
- Index terms
Last assessed as up-to-date: 15 October 2007.
Protocol first published: Issue 1, 1996
Review first published: Issue 2, 1997
Contributions of authors
Anthony Danso-Appiah developed the protocol and carried out the systematic review; this included assessing methodological quality, analysing and interpreting the data, and drafting the manuscript. Jürg Utzinger, Jianping Liu, and Piero Olliaro assisted in the interpretation of the results and revising the text. All authors helped with revisions following the referees' comments.
Declarations of interest
Sources of support
- Liverpool School of Tropical Medicine, UK.
- Department for International Development, UK.
- Swiss National Science Foundation (project no. PPOOB-102883), Switzerland.
Medical Subject Headings (MeSH)
MeSH check words
* Indicates the major publication for the study