Efficacy and safety of single-dose 40 mg/kg oral praziquantel in the treatment of schistosomiasis in preschool-age versus school-age children: An individual participant data meta-analysis

There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

Abstract

Background

Better knowledge of the efficacy and safety of single-dose 40 mg/kg oral praziquantel in preschool-age children is required, should preventive chemotherapy programs for schistosomiasis be expanded to include this age group.

Methodology

We analyzed individual participant-level data from 16 studies (13 single-arm or cohort studies and three randomized trials), amounting to 683 preschool-age children (aged <6 years) and 2,010 school-age children (aged 6–14 years). Children had a documented Schistosoma mansoni or S. haematobium infection, were treated with single 40 mg/kg oral praziquantel, and assessed between 21 and 60 days post-treatment. Efficacy was expressed as arithmetic mean and individual egg reduction rate (ERR) and meta-analyzed using general linear models and mixed models. Safety was summarized using reported adverse events (AEs).

Principal findings

Preschool-age children had significantly lower baseline Schistosoma egg counts and more losses to follow-up compared to school-age children. No difference in efficacy was found between preschool- and school-age children using a general linear model of individual-participant ERR with baseline log-transformed egg count as covariate and study, age, and sex as fixed variables, and a mixed model with a random effect on the study. Safety was reported in only four studies (n = 1,128 individuals); few AEs were reported in preschool-age children 4 and 24 hours post-treatment as well as at follow-up. Three severe but not serious AEs were recorded in school-age children during follow-up.

Conclusions/significance

There is no indication that single-dose 40 mg/kg oral praziquantel would be less efficacious and less safe in preschool-age children compared to school-age children, with the caveat that only few randomized comparisons exist between the two age groups. Preventive chemotherapy might therefore be extended to preschool-age children, with proper monitoring of its efficacy and safety.

Author summary

Schistosomiasis is a diseases caused by helminths (parasitic worms) which affects the intestinal and urogenital systems. In areas where schistosomiasis is endemic, the disease is controlled by the large scale distributing of praziquantel, primarily targeting school-age children. Younger children (preschool-age) too might be affected by schistosomiasis, but are currently not receiving praziquantel within treatment campaigns. Instead, preschool-age children are treated on a case-by-case basis because the current praziquantel formulation is not adapted to young children. Questions have also been raised as to whether the standard dose of 40 mg/kg given once is effective in preschool-age children. To answer this question, we collected individual-participant data from a series of studies in which 40 mg/kg of praziquantel had been given to children with intestinal or urinary schistosomiasis, and compared its efficacy and tolerability across age-groups. Since few direct comparisons had been made, we used statistical tools to make these comparisons. We found no evidence that treatment is less efficacious in preschool- than in school-age children and conclude that 40 mg/kg praziquantel may be given to preschool-age children in large-scale programs. When this happens, efficacy and tolerability will have to be closely monitored.

Related collections

Most cited references 33

Record: found
Abstract: found
Article: found

Is Open Access

Clinical Efficacy and Tolerability of Praziquantel for Intestinal and Urinary Schistosomiasis—A Meta-analysis of Comparative and Non-comparative Clinical Trials

Julien Zwang, Piero L. Olliaro (2014)

Introduction Some 779 million people are estimated to live in areas with varying levels of risk of contracting schistosomiasis [1]. The control and treatment of all forms of schistosomiasis is currently based on a single drug, praziquantel (PZQ). The World Health Organization (WHO) recommends that, in areas where the prevalence of infection is sufficiently high not to warrant individual diagnosis, a single dose of 40 mg/kg PZQ be distributed for preventive chemotherapy to either entire communities (through mass treatment) or school-aged children; or, where transmission is low, to be used to treat individuals with demonstrated infection [2]. Of note, while school-aged children are the main target of interventions, also younger children (preschool-aged) are now recognized as a vulnerable population [3], but data for this age group are limited. PZQ has been available for human use for over three decades, and distributed systematically through preventive chemotherapy from 2006. The cumulative number of treatments has been growing since. Some 34 million have received PZQ in 2010, and seven times more (235 millions) are projected for 2018 [4]; WHO has set target for 75% of the at-risk population to be under regular preventive chemotherapy [5]. With expanding use comes the need to monitor how PZQ performs in different areas, doses, over time and against different Schistosoma species. Two Cochrane systematic reviews have analyzed randomized controlled trials of anti-schistosomiasis treatments for S. haematobium [6] and S. mansoni [7]. A broader, aggregated data meta-analysis including non-comparative studies which did not qualify for the Cochrane reviews was undertaken here to help define more fully the efficacy and safety profile of PZQ across all species causing urinary and intestinal schistosomiasis, including mixed infections. The data generated from this meta-analysis was also intended to be used to help design future clinical investigations, in particular in young children treated with a new paediatric formulation currently developed by a public-private consortium [8]. Efficacy outcomes were measured for the different age-groups and doses and compared between various doses and to other drugs. Similarly, the tolerability profile of PZQ was assessed as incidence of adverse events (AE) and compared between various doses and to other drugs. Methods Data collection Published studies were identified by the Cochrane collaboration through electronic searches from January 1, 1990, up to November 2012 of MEDLINE, EMBASE, LILACS, the Cochrane Infectious Diseases Group's trials register and the Cochrane Central Register of Controlled Trials (CENTRAL) using the search term ‘praziquantel’ published in English, French or Portuguese. To qualify for inclusions, patients with a microscopic confirmation of schistosomiasis infection were to be on PZQ mono-therapy at any dosage and dosing regimen, using any formulation and brand; and could be either non-comparative or comparative (randomized controlled trial, quasi-randomized trials). In order to exclude the confounding effect of reinfections, efficacy analysis was restricted to the first 8 weeks post-treatment; hence, otherwise eligible studies with an endpoint beyond 8 weeks were not included in the final analysis. Statistical analysis The aggregated data (as reported in the publications) by species (S. haematobium, S. mansoni, S. japonicum, or mixed infections) were extracted from eligible studies of the 273 comparative and non-comparative clinical trials identified through the systematic review. Attrition bias refers to systematic differences between the number of patients at enrolment and at endpoint; it is measured as the number of patients not assessed out of the number of patients enrolled, and is considered high when greater than 10%. Cure rates (CR, defined as the conversion from a positive test pre-treatment to a negative test up to 8 weeks post-treatment) were calculated as provided in the articles. The confidence intervals for the CR were set at 95% (95%CI). The eggs reduction rate (ERR) was defined as the proportional reduction in the mean eggs per gram post-treatment vs. pre-treatment, calculated using geometric or arithmetic means and reported separately depending on how provided in the article. For both outcomes, the endpoint or time of assessment was divided in two groups: within a month (3 to 4 weeks) and between one and two months (5 to 8 weeks). The Spearman test was used to assess the bivariate correlations between the PZQ dose and CR or ERR in all treatment arms of comparative and non-comparative studies. Tolerability was assessed by calculating the incidence of adverse events (AE) defined as any sign or symptom occurring after the start of treatment (drug intake), irrespective of whether that sign or symptom was present at baseline or not, of its severity and drug-event relationship. The mean incidence was presented for the PZQ 40 mg/kg treatment groups excluding PZQ 40 mg/kg syrup, and Levo-PZQ 20 mg/kg. Most of the publications did not report the brand name and only two studies compared directly two different brands (Biltricide and Distocide) of PZQ. The 95%CI for the mean CR, ERR, AE were calculated using a bootstrap resampling method with a maximum of 1000 replicates [9]. For randomized controlled studies assessing the efficacy (CR) and tolerability (AE) of PZQ vs. other drugs, placebo, or comparing different PZQ dosing regimens, risk ratios with 95% confidence intervals (RR, 95%CI), meta-analysis with random effect on the study/site was used and pooled RR presented using the DerSimonian and Laird procedure for random effects models [10]. Heterogeneity was expressed as I2 [11]. CR and ERR were log-transformed in multivariate meta-regression to assess the PZQ dose-effect (continuous in mg/kg), along with age (continuous in year), endpoint (continuous in week) and date (continuous in year) with a random intercept for each study/site when the Lagrange multiplier (LM) test was significant to account for heterogeneity. Graphical displays of comparisons (PZQ vs. comparator groups) and heterogeneity for CR and ERR were illustrated using Forest plots [12]. Age groups were categorized as (i) preschool-aged children (<6 years old), (ii) school-aged children (6–19 years old), (iii) adults (20 years old or more), or (iv) all ages if age-specific data could not be extracted. The sample size (number of subjects by site), the endpoint (weeks), the intensity of infection at baseline (egg counts before treatment) were presented according to the Schistosoma species. Data were analyzed using Stata v11 (Stata Corp.). The PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) statement [13] was used as a guide in the reporting of this study. Results Study characteristics Of the 273 published studies identified by systematic search of the literature, 92 were PZQ treatment trials; of these, 37 studies had an endpoint for efficacy beyond 2 months and were excluded, leaving 55 studies (41 comparative, 14 non-comparative) with an endpoint within 8 weeks [14]–[68]. The first eligible study was published in 1979, and half of the studies were conducted by 1998. The studies enrolled a total of 19,499 subjects in 189 treatment arms. The median study size was 206 patients (range 43–1,540). Attrition was acceptable (9%, n = 17,718), leaving 91% of the subjects with efficacy outcomes at the time of the study endpoint; of these, 42% were assessed within 4 weeks in 30 studies, and 58% between 5–8 weeks in 25 studies. More subjects were assessed between 5–8 weeks for S. mansoni (65%) and S. japonicum (71%), while more were assessed on week 3–4 for S. haematobium in 56% of cases and mixed S. mansoni/haematobium infections in 71% of cases. PZQ contributed to 74% (n = 13,048) and comparator drug or placebo to 26% (n = 4,670) of all subjects with outcomes. Of the subjects treated with PZQ at doses comprised between 10 and 60 mg/kg, PZQ 40 mg/kg was the most frequent dose (56%, n = 9,990). CR was assessed in 17,017 subjects (of whom 12,273 (69%) treated with PZQ) and ERR in 13,007 subjects (77%, n = 10,023 on PZQ) (Figure 1). 10.1371/journal.pntd.0003286.g001 Figure 1 Flow chart of the number of studies and patients screened and eligible for the efficacy analyses of cure rate (CR) and egg reduction rates (ERR). Of the 41 comparative studies identified, 19 directly compared different doses and schedules of PZQ; 7 compared praziquantel with artesunate combined with praziquantel, sulfalene, sulfamethoxypyrazine/pyrimethamine, or mefloquine; 6 with artesunate alone; 3 with metrifonate; 1 with nitrifonate and 1 with metrifonate+nitrifonate; 6 with oxamniquine, 4 with oltipraz, 1 with albendazole, 1 with mefloquine, and 3 with PZQ in combination with artemether, albendazole, or metrifonate. Studies were conducted in 24 countries and 82 sites. The largest population was from the WHO AFRO region (13,251 subjects, 68%), followed by EMRO (Egypt, Sudan and Saudi Arabia, 23%). The largest groups were S. mansoni subjects enrolled in Egypt (n = 2,606, 13.4%), Kenya (11.9%), Sudan (9.0%) and Uganda (6.6%) (Table S1). The risk of attrition bias was low in 57%, high in 38% and not assessable in 5%. The risk was high in 45% (10/22) of the community-based studies and 31% (10/32) of the school-based studies. The assessment of bias and the main characteristics of these studies are summarized in Table 1 and Table S2. 10.1371/journal.pntd.0003286.t001 Table 1 Main characteristics of the studies included – S. haematobium (sh), S. japonicum (sj) and mixed infections S. haematobium-S. intercalatum (si). Publication Country End point Species Age range Total (n) PZQ referent Comparator Screening method dose n drug n sample(n)*day(n) Bornmann 2001 Gabon 2 sh 5–13 296 pzq40 89 other 207 1*1 Burchard 1984 Gabon 2 sh 5–14 138 pzq60 65 other 73 1*1 dup Davis 1981 Zambia 1 sh 7–17 151 pzq40 45 pzq20-60 106 1*3 de Clercq 2002 Senegal 2 sh 7–14 267 pzq40 133 other 134 1*2 Inyang-Etoh 2008 Nigeria 2 sh 4–20 174 pzq40 42 pzq40 + other 220 2*2 Keiser 2010 Ivory coast 1 sh 8–16 83 pzq40 26 other 57 2*1 King 2002 Kenya 2 sh 4–23 200 pzq40 101 pzq20 99 1*2 Latham 1990 Kenya 2 sh 7–15 48 pzq40 16 other 32 1*1 McMahon 1983 Tanzania 2 sh 1–60* 77 pzq30 30 other 47 1*2 McMahon 1979 Tanzania 1 sh 7–15 125 pzq40 65 pzq30 and placebo 60 3*3 Midzi 2008 Zimbabwe 2 sh 2–19 624 pzq40 624 1*3 N'goran 2003 Ivory coast 1 sh 5–15 354 pzq40 354 1*1 dup Olds 1999 Kenya 2 sh 6–19 380 pzq40 95 other 285 2*1 Oyideran 1981 Nigeria 1 sh 7–13 82 pzq40 40 pzq30 and placebo 42 1*3 Rey 1983 Niger 1 sh 15–19 188 pzq40 54 pzq30 and other 134 2*2 Sissoko 2009 Mali 1 sh 6–15 781 pzq40 389 other 392 1*1 dup Tchuente 2004 Cameroon 1 sh school 515 pzq40 515 1*2 Wilkins 1987 Gambia 1 sh 5–17 619 pzq40 143 pzq10-20 and other 476 1*1 Kern 1984 Gabon 2 sh + si 10–17 158 pzq60 77 other 81 1*2 Belizario 2007 Philippines 1 sj 10–19 203 pzq40 102 pzq60 and other 101 2*2 Hou 2008 China 2 sj 10–60 196 pzq60 55 other 141 2*1 tri Olds 1999 Phillipines, China 2 sj 5–16 793 pzq40 203 other and placebo 590 2*1 Olliaro 2011 Philippines 1 sj 7–12 200 pzq40 101 pzq60 99 1*2 dup Abu elyazed 1998 Egypt 2 sm 5–50 939 pzq40 551 pzq60 388 3*3 Barakat 2005 Egypt 1 sm 5–39 83 pzq40 38 other 45 1*2 Berhe 1999 Ethiopia 2 sm 5–17 541 pzq40 541 1*1 Botros 2005 Egypt 2 sm 7–73 271 Pzq40 165 other 106 1*3 daSilva 1986 Brazil 1 sm 14–60* 94 pzq55 48 other 46 3*1 Declerq 2000 Senegal 2 sm 6–61 156 pzq40 39 other 117 1*1 dup Declerq tmih 2000 Senegal 2 sm 1–50 110 pzq40 36 other 74 1*1 Degu 2002 Ethiopia 2 sm 10–14 148 pzq40 148 1*1 Friis 1988 Botswana 2 sm school 81 pzq40 81 1*1 dup Ghandour 1995 Saudi Arabia 1 sm 1–50 170 pzq40 170 na Gryseels 1987 Burundi 2 sm <20 and ≥20 1138 pzq40 272 pzq20-30 other 866 1*1 dup Guisse 1987 Senegal 1 sm 5–15 130 pzq40 67 pzq60 63 2*2 Homeida 1989 Sudan 2 sm 1–60* 806 pzq40 400 pzq40 brand2 406 1*1 Ismail 1994 Egypt 2 sm 6–18 463 pzq40 463 1*1 Kabatereine 2003 Uganda 2 sm 5–50* 482 pzq40 482 3*1 Kardaman 1983 Sudan 1 sm 5–60* 388 pzq40 388 2*1 Massoud 1984 Egypt 1 sm school 179 pzq40 59 pzq10-20 120 1*1 McMahon 1981 Tanzania 1 sm 1–60* 91 pzq40 49 pzq50 42 1*3 Metwally 1995 Egypt 1 sm 8–16 366 pzq40 149 pzq20 217 3*3 tri Mohamed 2009 Sudan 1 sm 8–17 92 pzq40 46 other 46 1*2 Navaratnam 2012 Uganda 1 sm 1–5 297 pzq40 149 syrup pzq40 148 3*1 Obonyo 2010 Kenya 1 sm 7–12 212 pzq40 106 other 106 1*1 dup Olds 1999 Kenya 2 sm 6–19 367 pzq40 82 other placebo 285 2*1 Olliaro 2011 Brazil 1 sm 10–19 190 pzq40 96 pzq60 94 1*2 dup Olliaro 2011 Mauritania 1 sm 10–19 185 pzq40 92 pzq60 93 1*2 dup Olliaro 2011 Tanzania 1 sm 10–19 244 pzq40 119 pzq60 125 1*2 dup Raso 2004 Ivory coast 2 sm 1–60* 161 pzq40 161 3*1 Simonsen 1990 Ethiopia 1 sm 5–14 206 pzq40 206 2*1 Sousa-Figueiredo 2012 Uganda 1 sm 1–7 369 pzq40 369 1*2 Stelma 1997 Senegal 2 sm 5–75 86 pzq40 44 other 42 2*2 Taddese 1988 Ethiopia 1 sm 17–52 194 pzq40 99 other 95 1*1 Teesdale 1984 Malawi 1 sm 9–15 69 pzq40 18 other 51 4*1 Thiongo'o 2002 Kenya 2 sm 5–17 1018 pzq40 526 pzq60 and other 492 1*3 dup Utzinger 2000 Ivory coast 1 sm 6–14 194 pzq60 194 1*4 El Tayeb 1988 Sudan 1 sm+sh 7–12 111 pzq40 54 other 57 1*2 Kardaman 1983 Sudan 1 sm+sh 5–60* 43 pzq40 43 2*1 Kardaman 1985 Sudan 2 sm+sh 7–11 211 pzq40 211 1*1 Taylor 1988 Zimbabwe 1 sm+sh 10–15 373 pzq40 77 pzq10-20-30 and placebo 296 3*1 The largest number of subjects with efficacy outcomes was for a S. mansoni infection (57.8%), followed by S. haematobium (29.3%), S. japonicum (7.9%) and mixed infections (5%). Most of the subjects were school-aged children (63.8%); preschool-aged children accounted for 2.9%, adults 4.7%, and subjects of all ages 28.6% (Table S3). Two studies including all age's subjects also specified age categories, including schoolchildren [36], [66]. Laboratory diagnosis To diagnose and quantify the infection, the trials on S. haematobium used the filtration method with up to two specimens in duplicates over three days except in one study using reagent strips, while trials on S. mansoni used the Kato-Katz technique with up to three specimens over three days in triplicates (Table S4). Egg counts were reported using different approaches (number of specimens and tests) for the different intestinal or urinary schistosomiasis species for 13,135 subjects. The mean egg count before treatment was 910 (95%CI 369–1642) and 251 (95%CI 201–307) eggs per gram of feces for S. mansoni, in studies using arithmetic or geometric means, respectively; 178 (95%CI 95–274) eggs per gram of feces for S. japonicum; and 125 (95%CI 60–196) and 137 (95%CI 70–226) eggs per mL of urine for S. haematobium, for arithmetic and geometric means, respectively. Efficacy In subjects treated with PZQ, the efficacy of PZQ in any species (n = 13,105) was measured in 508 (4%) preschool, 7,776 (59%) school-aged children, 428 (3%) adults, and 4,393 (34%) subjects of all ages. The number of treatment arms with different doses of PZQ varied greatly; the 40 mg/kg dose was by far the most common (66%, 77/117), followed by the 60 mg/kg dose (14%, 16/117). All doses were not tested on each and every species or age groups. The only dose administered in preschool-aged children was 40 mg/kg for S. mansoni; school-aged children received doses ranging 10–60 mg/kg (72% were on 40 mg/kg); adults received 20–40 mg/kg; studies on all-age subjects administered doses ranging 20–60 mg/kg (76% were on 30 mg/kg). Cure rates (CR) with PZQ Mean dose-specific CRs with 95%CIs by species are presented in Figure 2. CRs for any dose of PZQ appeared to be highest in S. japonicum infections (40 and 60 mg/kg); and were higher in S. haematobium, mixed S. haematobium/intercalatum and S. mansoni infections than in pure and mixed S. mansoni/haematobium infections. 10.1371/journal.pntd.0003286.g002 Figure 2 Forest plot of praziquantel (PZQ) cure rates with 95% CIs by species and dose (all age groups). sh, S. haematobium; si, S. intercalatum; sj, S. japonicum; sm, S. mansoni. The recommended dose of 40 mg/kg achieved CRs of 94.7% (95% CI 92.2–98.0) for S. japonicum, while it was 77.1% (95% CI 68.4–85.1) for S. haematobium, 76.7% (95% CI 71.9–81.2) for S. mansoni, and 63.5% (95% CI 48.2–77.0) for mixed S. haematobium and S. mansoni infections. Dose-effect analysis There was a significant relationship (Spearman test) between the CRs in subjects treated for S. mansoni and the PZQ dose: from 26.2% with PZQ 10 mg/kg to 84.6% with PZQ 60 mg/kg (r = 0.434, p = 0.001), as well as for mixed S. mansoni + S. haematobium infections (r = 0.764, p = 0.001) but not for S. haematobium (r = 0.019, p = 0.923) nor for S. japonicum (r = 0.396, p = 0.437). Endpoint analysis PZQ 40 mg/kg CRs assessed on week 3–4 were 82.7% (95%CI 70.3–92.9) and on week 5–8 were 69.9% (95%CI 58–78.7) for S. haematobium, and were 79.6% (95%CI 72.8–85.7) and 73.9% (95%CI 67.1–80.6) for S. mansoni, respectively. Although a direct comparison is not possible, 95%CIs overlap for both species. CR with PZQ 40 mg/kg vs. comparators The RR (95%CI) of CR between praziquantel 40 mg/kg and other PZQ regimens, placebo or other treatments are presented in Figure 3 for S. haematobium and Figures 4 and 5 for S. mansoni. 10.1371/journal.pntd.0003286.g003 Figure 3 Forest plot of relative risks of cure rates with 95%CIs, S. haematobium, PZQ 40 mg/kg vs. comparators. as, artesunate; ol, oltipraz; pzq, praziquantel; sp, sulfadoxine-pyrimethamine; met, metrifonate, mq, mefloquine; p, placebo; comp, comparator; unit next to the drug: dose in mg/kg; RR, risk ratio; I2 (Higgins' I squared) is calculated for pooled subgroups as = 100%*(Q - df)/Q, where Q is Cochran's heterogeneity statistic and df the degrees of freedom. 10.1371/journal.pntd.0003286.g004 Figure 4 Forest plot of risk ratios of cure rates with 95%CIs, S. mansoni, PZQ 40 mg/kg vs. other PZQ regimens. comp, comparator; ci, confidence interval; pzq, praziquantel; unit next to the drug: dose in mg/kg; RR, risk ratio; I2 (Higgins' I squared) is calculated for pooled subgroups as = 100%×(Q - df)/Q, where Q is Cochran's heterogeneity statistic and df the degrees of freedom. 10.1371/journal.pntd.0003286.g005 Figure 5 Forest plot of risk ratio of cure rates with 95%CIs, S. mansoni, PZQ 40 mg/kg vs. other regimens. as, artesunate; ox, oxamniquine; pzq, praziquantel; sp, sulfadoxine-pyrimethamine; mq, mefloquine; comp, comparator; ci, confidence interval; unit next to the drug: dose in mg/kg; RR, risk ratio; I2 (Higgins' I squared) is calculated for pooled subgroups as = 100%×(Q - df)/Q, where Q is Cochran's heterogeneity statistic and df the degrees of freedom. Using meta-analysis regression model with random effect on the sites, the CR for treating S. haematobium with praziquantel 40 mg/kg was higher than praziquantel 20 mg/kg (RR = 0.71, 95%CI 0.56–0.90, p = 0.004) and not different from praziquantel 30 mg/kg (p = 0.575); PZQ 40 mg/kg had higher CR than artesunate alone (RR = 0.55, 95%CI 0.36–0.83, p = 0.005) or in combinations, mefloquine alone, and metrifonate 10 mg/kg (RR = 0.15, 95%CI 0.04–0.58, p = 0.001). On S. mansoni, using similar methods, the CR of PZQ 40 mg/kg was higher than PZQ 20 mg/kg (RR = 0.65, 95%CI 0.59–0.72, p = 0.001), PZQ 30 mg/kg (RR = 0.89, 95%CI 0.75–0.95, p = 0.004), and not different from higher doses (50 mg/kg, p = 0.544; 60 mg/kg, p = 0.477); the CR for PZQ 40 mg/kg was significantly higher than artesunate and combinations, and myrrh (p = 0.001 for all comparisons); not different from oxamniquine 15, 20, 30 mg/kg; slightly lower than oxamniquine 40 mg/kg (RR = 1.09, 95%CI 1.01–0.18, p = 0.034), but not significantly different from oxamniquine 50 mg/kg (RR = 1.65, 95%CI 0.99–2.75, p = 0.056). On S. japonicum, using similar methods, the CR of PZQ 40 mg/kg was not different from PZQ 60 mg/kg (RR 1.02, 95%CI 0.97–1.07, p = 0.461), and higher than placebo (p = 0.001). On mixed S. haematobium and mansoni, the CR of PZQ 40 mg/kg was not significantly higher from lower PZQ dose (10 mg/kg: RR 0.15, p = 0.060; 20 mg/kg RR 0.63, p = 0.135; 30 mg/kg RR 0.86, p = 0.278). Eggs reduction rate (ERR) The ERR was measured for 13,007 subjects in 126 study/sites. ERR by species and PZQ dose from non-comparative and comparative trials are presented in Figure 6. 10.1371/journal.pntd.0003286.g006 Figure 6 PZQ egg reduction rates with 95% CIs by species and dose. ci, confidence interval; sh, S. haematobium; si, S. intercalatum; sj, S. japonicum; sm, S. mansoni. The mean ERR was over 90% in subjects of any age treated with PZQ doses greater than 10 mg/kg for S. haematobium and 87% or more for S. mansoni and 89% or more for S. mansoni/haematobium mixed infections (40 mg/kg); for S. japonicum, the ERR was ∼95% (40 and 60 mg/kg). There was no significant relationship (Spearman test) between the ERRs in subjects treated with any PZQ dose and species: S. mansoni (r = −0.126, p = 0.370), S. haematobium (r = 0.057, p = 0.786), as well as for S. japonicum (r = 0.236, p = 0.764). With PZQ 40 mg/kg, the ERR assessed was 94.6% (95%CI 89.9–98.0) on week 3–4 and 93.4% (95%CI 83.2–100) on week 5–8 for S. haematobium, for S. mansoni, the ERR was 87.4% (95%CI 82.7–91.5) and 72.0% (89.0%, 95%CI 83.7–94.2) respectively. More details on efficacy rates by age groups and dose are given in Table 2. 10.1371/journal.pntd.0003286.t002 Table 2 Cure rates (CR) and egg reduction rates (ERR) with 95% confidence intervals (95%CI) calculated by boot-strapping by age-group, species and praziquantel dose. Age group Species PZQ dose (mg/kg) Cure rate (CR) Egg reduction rate (ERR) CR Lower 95%CI Upper 95%CI N patients N treatment arms % assessed within 1 month Endpoint (median week) ERR Lower 95%CI Upper 95%CI N patients N treatment arms % assessed within 1 month Endpoint (median week) preschool sm 40 69.0% 56.4% 81.7% 414 2 100% 4 85.6% 82.2% 89.0% 414 2 100% 4 school-aged sh 10 87.0% 77.6% 97.6% 38 1 100% 4 20 98.1% 98.1% 98.1% 53 1 100% 4 98.4% 94.4% 99.9% 35 1 100% 4 30 71.0% 71.0% 71.0% 31 1 100% 4 92.6% 85.7% 99.6% 50 2 100% 4 40 76.6% 67.8% 85.2% 2490 18 56% 4 93.9% 89.1% 98.7% 1856 18 67% 4 60 82.5% 75.0% 90.0% 65 2 0% 6 sh + si 60 85.6% 76.0% 91.0% 77 3 0% 5 sj 40 94.7% 92.2% 98.0% 406 3 67% 3 95.0% 90.1% 99.9% 203 2 100% 3 60 97.5% 97.0% 98.0% 200 2 100% 3 95.4% 91.0% 99.9% 200 2 100% 3 sm 10 26.2% 26.2% 26.2% 61 1 100% 4 20 40.3% 24.9% 58.0% 376 4 75% 4 91.7% 86.7% 97.3% 100 1 0% 6 30 63.0% 63.0% 63.0% 187 1 0% 6 96.1% 93.5% 98.8% 187 1 0% 6 40 74.6% 68.3% 80.6% 2340 19 47% 5 89.1% 83.3% 94.2% 1856 18 43% 5 60 78.6% 67.8% 90.6% 667 5 60% 4 84.2% 73.5% 94.2% 667 5 60% 4 sm+sh 10 11.1% 4.2% 18.1% 73 2 100% 4 20 38.3% 36.7% 40.0% 61 2 100% 4 30 51.8% 31.4% 72.2% 72 2 100% 4 40 67.6% 52.4% 81.2% 342 5 60% 4 98.0% 87.1% 99.7% 54 1 100% 4 adult sh 30 97.4% 97.4% 97.4% 39 1 100% 4 40 94.4% 94.4% 94.4% 54 1 100% 4 sm 20 55.0% 55.0% 55.0% 53 1 0% 6 91.2% 84.7% 98.2% 53 1 0% 6 30 87.0% 87.0% 87.0% 102 1 0% 6 98.0% 95.5% 99.9% 102 1 0% 6 40 94.2% 91.0% 96.0% 180 3 67% 4 77.0% 63.0% 98.2% 180 3 67% 4 all ages sh 20 50.0% 50.0% 50.0% 99 1 0% 6 95.0% 90.9% 99.3% 99 1 0% 6 30 87.0% 87.0% 87.0% 30 1 0% 8 99.0% 95.5% 99.9% 30 1 0% 8 40 70.0% 70.0% 70.0% 101 1 0% 6 98.0% 95.3% 99.9% 101 1 0% 6 sj 60 96.4% 96.4% 96.4% 55 1 0% 6 sm 40 76.7% 67.7% 84.5% 3443 20 40% 5 85.5% 79.0% 91.4% 3443 20 40% 5 50 88.1% 88.1% 88.1% 42 1 100% 4 98.7% 95.4% 99.9% 42 1 100% 4 55 79.2% 79.2% 79.2% 48 1 100% 4 93.5% 87.0% 99.9% 48 1 100% 4 60 94.4% 92.5% 95.9% 575 3 67% 3 83.5% 68.7% 92.0% 575 3 67% 3 sm+sh 40 43.2% 43.2% 43.2% 37 1 100% 4 89.0% 80.6% 98.3% 37 1 100% 4 Legend: sh, S. haematobium; si, S. intercalatum; sj, S. japonicum; sm, S. mansoni. Brand analysis Both brands (Biltricide and Distocide) of PZQ 40 mg/kg were effective in reducing infection intensity (ERR was 99.5% for both groups)[31]; similarly, there was no difference in CR with either 40 mg/kg (pooled RR 0.99, 95%CI 0.96–1.03, p = 0.745)[31], [46], or 20 mg/kg (RR 0.85, 95%CI 0.54–1.31, p = 0.453)[31]. Tolerability Adverse events (AEs) Of the 273 published studies identified, signs and symptoms recorded within 48 hours of treatment were reported in 12,435 subjects enrolled in 40 studies: 25 studies from the efficacy analysis, contributing to 75% of the subjects assessed for tolerability (n = 9,151) and 15 additional studies, (n = 3,284) [69]–[83], meaning that 45% of the studies eligible for the efficacy meta-analysis reported on tolerability. Ninety-six (96) treatment arms were analyzed of which 64 were PZQ administered from 20 to 80 mg/kg. Most of the recorded AEs were gastro-intestinal, neurological and dermatological (Figure S1). On average the incidence of subjects experiencing at least one AE was 56.9% (95%CI 47.4–67.9) in twelve studies reporting this tolerability outcome and treating 2,027 subjects with PZQ 40 mg/kg (all brands). The incidence of specific AEs ranged from 2.3% for urticaria to 31.1% for abdominal pain (Table 3) – detailed below. 10.1371/journal.pntd.0003286.t003 Table 3 Adverse event incidence, praziquantel 40 mg/kg. Adverse event Number Incidence 95%CI Bootstrap Studies Patients (%) Lower bound Upper bound Any adverse event 13 2272 56.0 45.2 66.4 Abdominal pain 30 6212 31.1 22.0 39.0 Muscle pain 2 129 29.2 10.0 48.0 Joint pain 3 642 25.7 7.4 59.0 Dizziness 30 6328 13.9 9.1 19.0 Headache 27 5642 13.7 9.1 18.0 Diarrhea 27 5790 12.7 8.0 17.0 Fatigue 10 2279 11.6 5.4 18.0 nausea 22 5508 10.6 6.9 14.0 Itching/rash 15 2885 10.4 3.9 19.0 Weakness 5 882 10.0 3.7 17.0 Haematuria 3 727 9.6 0.0 23.0 Vertigo 3 304 8.7 3.8 14.0 Vomiting 26 5339 7.2 4.8 9.7 Legend. ci, confidence interval. AEs with PZQ 40 mg/k vs. comparators In comparative studies, and using meta-regression with random effect on the study/site, subjects treated with PZQ 40 mg/kg were at lower risk for any AE compared to PZQ 60 mg/kg (RR 0.73, 95%CI 0.59–0.90, p = 0.003), oxamniquine 25 mg/kg (RR 0.63, 95%CI 0.50–0.78, p = 0.001), metrifonate 3*10 mg/kg (RR 0.73, 95%CI 0.55–0.98, p = 0.036), while they were at higher risk compared to L-PZQ (RR 1.31, 95%CI 1.05–1.63, p = 0.018) and AS+SP (RR 2.26, 95%CI 1.50–3.41, p = 0.004); there was no difference between 40 mg/kg and PZQ doses (20 mg/kg, 30 mg/kg, 2*20 mg/kg), metrifonate 10 mg/kg, metrifonate 10 mg/kg + niridazole 25 mg/kg. When different brands were compared, Biltricide had more AEs than Distocide (RR 1.50, 95%CI 1.31–1.72, p = 0.001). The most frequent AEs are listed below by decreasing frequency in PZQ 40 mg/kg recipients. The incidence of abdominal pain was 31.8% (95%CI 24.4–39.9) in 6,495 subjects treated with PZQ 40 mg/kg in 30 treatment arms. Subject treated with PZQ 40 mg/kg were at higher risks for abdominal pain than PZQ 20 mg/kg (RR = 1.80, 95%CI 1.31–2.48, p = 0.001), metrifonate 10 mg/kg (RR = 1.50, 95%CI 1.21–1.86, p = 0.001), AS+SP (RR = 3.32, 95%CI 1.70–6.49, p = 0.001); while there was no significant difference between PZQ 40 mg/kg and PZQ at various dose (60 mg/kg, 30 mg/kg, 2*20 mg/kg, 2*25 mg/kg, 2*15 mg/kg, 2*35 mg/kg, 2*30 mg/kg, syrup 40 mg/kg), L-PZQ, mefloquine, AS, ASMQ, ASSP, metrifonate 30 mg/kg, nirifonate 150 mg/kg, metrifonate 10mg/kg + nirifonate 250 mg/kg. Divergent results were found when PZQ was compared to oxamniquine: in a study, subjects treated with PZQ 40 mg/kg were at lower risks (RR = 0.48, 95%CI 0.28–0.83, p = 0.001) than oxamniquine 25 mg/kg, while in another study they were at higher risks compared to oxamniquine at 15, 20, 30, 40 mg/kg (p<0.05). Subjects treated with Biltricide were at higher risk of abdominal pain than those treated with Distocide (RR = 2.34, 95%CI 1.74–3.14, p = 0.001). Muscle pain was reported in 29.2% (95%CI 10.0–48.0) of the 129 subjects receiving PZQ 40 mg/kg at two study/sites and not different from PZQ 2*30 mg/kg. No difference was detected either in two other studies comparing PZQ 55 mg/kg and oxamniquine 15 mg/kg. Joint pain was reported in 20.2% (95%CI 4.9–42.3) of the 1,097 subjects enrolled in four PZQ 40 mg/kg treatment arms. In comparative studies no difference was detected with metrifonate 10 mg/kg and oxamniquine 25 mg/kg; subjects treated with PZQ 40 mg/kg Distocide (3.7%) were at lower risk compared to PZQ 40 mg/kg Biltricide brand (7.4%, RR 0.50, 95%CI 0.28–0.89, p = 0.018). Headache was reported in 13.6% (95%CI 9.3–18.6) of the 5,958 PZQ 40 mg/kg recipients enrolled in 27 treatment arms. Subjects treated with PZQ 40 mg/kg were at lower risks than those on oxamniquine 20 mg/kg (RR 0.31, 95%CI 0.11–0.89, p = 0.020), oxamniquine 2*15 mg/kg (RR 9.00, 95%CI 1.18–68.42, p = 0.034) while no difference was detected in other studies vs. other dose of PZQ (from 20 up to 60 mg/kg, syrup 40 mg/kg or L-PZQ), artesunate and combinations, mefloquine, niridazole, or metrifonate. The incidence of diarrhea was 12.9% (95%CI 8.6–17.9) in 6,106 PZQ 40 mg/kg recipients enrolled in 27 treatment arms. Subjects treated with PZQ 40 mg/kg were at higher risks compared to PZQ 2*30 mg/kg (RR 14.10, 95%CI 1.92–103.68, p = 0.009) and oxamniquine 40 mg/kg (RR 0.03, 95%CI 0.01–0.19, p = 0.001). The risk was also higher with Biltricide than Distocide (RR 2.28, 95%CI 1.46–3.56, p = 0.001), while there was no difference between PZQ 2*20 mg/kg (6%) and PZQ 2*15mg/kg (1%) and PZQ 2*25 mg/kg (5%), or between PZQ 40 mg/kg tablet and syrup formulation, or between 40 mg/kg and other PZQ doses, artesunate combinations, metrifonate 10 mg/kg, mefloquine, and other oxamniquine doses. The incidence of dizziness was 11.9% (95%CI 7.9–16.2) in 5,522 PZQ 40 mg/kg recipients enrolled in 26 treatment arms. Subjects treated with PZQ 40 mg/kg were at lower risks than oxamniquine at any dose: 20 mg/kg (RR 0.31, 95%CI 0.21–0.48, p = 0.001), 25 mg/kg (RR 0.56, 95%CI 0.37–0.84, p = 0.005), 30 mg/kg (RR 0.21, 95%CI 0.14–0.32, p = 0.001), 40 mg/kg (RR 0.19, 95%CI 0.13–0.27, p = 0.001), while they were at higher risks compared to metrifonate 10 mg/kg (RR 1.60, 95%CI 1.06–2.43, p = 0.001); there was no difference between the different dose of PZQ treatment and syrup, L-PZQ, oltripaz 2*15 mg/kg, metrifonate 30 mg/kg, or niridazole 150 mg/kg. Nausea was reported in 10.6% (95%CI 6.8–14.9) in 5,824 PZQ 40 mg/kg subjects in 22 treatment arms. Subjects treated with PZQ 40 mg/kg were at higher risks for nausea compared to 2*30 mg/kg (RR 2.47, 95%CI 1.18–5.16, p = 0.001) and L-PZQ (RR 4.50, 95%CI 1.56–12.96, p = 0.001), while there was no difference between the different dose of PZQ (20, 25, 30, 2*15, 40, 2*25, 60, 2*35, 80 mg/kg), brands and formulations, artesunate and combinations, oltripaz 2*15 mg/kg, metrifonate 10, 30 mg/kg, oxamniquine (15, 20, 25, 30, 40 mg/kg), niridazole 150 mg/kg. The incidence of itching/rash was 9.8% (95%CI 3.8–18.2) in 3,340 PZQ 40 mg/kg recipients in 16 treatment arms. Splitting the dose (PZQ 2*20 mg/kg) decreased the risk of itching/rash (RR 0.03, 95%CI 1.01–0.52, p = 0.016) in one study; no difference was detected between PZQ brands, tablets vs. syrup, and between PZQ 40 mg/kg and 2*30, 2*25 mg/kg, metrifonate 10 and 30 mg/kg, oxamniquine at various doses (20, 25, 30, 50 mg/kg), artesunate combinations and niridazole 150 mg/kg. Fatigue was reported in 9.6% (95%CI 4.0–16.3) of the 2,595 PZQ 40 mg/kg recipients in 10 arms. Subjects treated with PZQ 40 mg/kg were at lower risks compared to oxamniquine 25 mg/kg (RR 0.17, 95%CI 0.05–0.58, p = 0.005) while there was no difference between PZQ 40 mg/kg compared to other doses of PZQ (2*15, 2*20, 2*25, 2*30 mg/kg), between PZQ brands, formulations, L-PZQ, oxamniquine 15 mg/kg, and metrifonate 10 mg/kg. The incidence of vomiting was 7.9% (95%CI 5.2–10.9) in 5,722 PZQ 40 mg/kg recipients enrolled in 27 treatment arms. Subjects treated with PZQ 40 mg/kg were at lower risks compared to 60 mg/kg (RR 0.44, 95%CI 0.26–0.72, p = 0.001) but at higher risk than PZQ 2*30 mg/kg (RR 2.51, 95%CI 1.26–4.97, p = 0.008); the risk was higher with Biltricide than Distocide (RR 3.53, 95%CI 1.88–6.63, p = 0.001). There was no difference between tablets and syrup, and between 40 mg/kg and other doses (20, 30, 2*20 mg/kg), L-PZQ, AS, ASSP, ASMQ, oxamniquine (15, 20, 25, 30 mg/kg) or metrifonate 3*10 mg/kg. Discussion This is, to our knowledge, the largest collection of PZQ treatment trials analyzed so far, with over 14,000 subjects receiving the drug at different doses. This population is much larger, but intrinsically less homogenous, than that of the two available Cochrane systematic reviews [6], [7]. This study complements the Cochrane systematic reviews by broadening the number of studies analyzed for efficacy as well as tolerability, and by allowing a side-by-side analysis of all Schistosoma species, including co-infections. The overall conclusions of these reviews are generally concordant, despite some minor differences, which are mostly related to the different criteria for including/excluding studies in either analyses. Provided basic methodological standards are guaranteed, relaxing eligibility criteria for meta-analysis to allow for both comparative and non-comparative trials can broaden the database and complement more classical meta-analysis of randomised controlled trials. This and the Cochrane meta-analyses point to the lack of methodological standardization of the studies analyzed: the number of subjects per study, age groups, species and PZQ doses varied greatly across the studies; different methodologies were used to detect eggs in excreta (number of samples taken; number of tests/sample) and to quantify efficacy (CR, ERR with arithmetic or geometric means). While studies extend over three decades and a range of countries, trends over time cannot be reliably derived. In 38% of the studies, patient attrition was greater than 10%, and this more in community-based (45%) than school-based studies (31%). Tolerability was unevenly assessed and reported. Statistical models have helped in deriving trends but cannot compensate for the lack of direct comparisons for dose and age effects. On the other hand, the large number of subjects allows generalizable conclusions, despite the inherent limitations of aggregated-data meta-analyses. The main findings of this meta-analysis are that: (1) Schistosoma species appear to respond differently to PZQ, with S. japonicum having the highest and mixed S. mansoni/haematobium infections the lowest response rates, both in terms of CR and ERR; (2) a dose-response trend was apparent for CR in S. mansoni and mixed S. mansoni/haematobium infections, but not S. haematobium or S. japonicum. No significant trend was apparent for ERR, the currently preferred outcome measure [4] for any of the Schistosoma species; (3) age did not appear to influence treatment outcomes. However, this should be interpreted with caution as the age groups enrolled were generally broad and details by age are generally not provided in the papers; furthermore, preschool-aged children are minimally represented in this population, received only 40 mg/kg, and only for S. mansoni; (4) a single praziquantel dose of 40 mg/kg appears a reasonable compromise for all species and ages, although in a proportion of cases efficacy may be lower than expected. The most studied groups were school-aged children (64% of all subjects), S. mansoni infections (58%) and the PZQ dose of 40 mg/kg (56%); 68% of subjects were in the WHO AFRO region (where the prevalence of the infection is highest). Preschool-aged children accounted for only ∼3% of the total population (meaning that information on younger children is limited, and that conclusions on age-related efficacy and safety may change when more data accumulate in this age group). It should also be noted that community-based studies (which generally enroll subjects of all ages) tend to have more drop-outs than school-based studies. Overall, the CR achieved with the WHO-recommended dose of 40 mg/kg was highest for S. japonicum (94.7%, 95%CI 92.2–98.0), followed by S. haematobium (77.1%, 95% CI 68.4–85.1) for S. haematobium, and S. mansoni (76.7%, 95% CI 71.9–81.2), and mixed S. haematobium and S. mansoni infections (63.5%, 95%CI 48.2–77.0). Recent WHO Standard Operating Procedures recommend that control programs should further investigate drug performance in populations where the ERR is found to be lower than 90% [4]. The average ERR obtained in school-aged children with the dose of 40 mg/kg was 95% for S. japonicum, 94% for S. haematobium, and 89% for S. mansoni. Since these values are derived from a collection of studies, the fact that the lower bound of the 95%CI (obtained by bootstrapping) was 90% for S. japonicum, 89% for S. haematobium, and 81% for S. mansoni means that a proportion of these sites might warrant further assessment. However, it is difficult to compare results obtained with a variety of diagnostic approaches and using different calculations (geometric or arithmetic means) to a ‘reference for drug efficacy’ that is based on a single examination of a single specimen and is expressed as geometric mean. Praziquantel efficacy may be influenced by a variety of factors, which could not be explored in detail using aggregated data and meta-analysis methods. Pre-treatment intensity of infection is one, which could not be fully accounted for by having it as covariate in the model, primarily because of the diversity of the diagnostic and calculation approaches used, and because it is reported as group mean (individual subject data meta-analyses are better suited to address this issue). The WHO-recommended dose of 40 mg/kg compared favorably to all other PZQ regimens and other treatments tested. The dose-response curve appears to be flat for S. haematobium and to plateau at 40 mg/kg for S. mansoni. This must be due to different species susceptibility, because this happens in spite of exposure to praziquantel increasing overproportionally with the dose (the first-pass-metabolism in the liver being dose-dependent with regard to capacity) [84]. Similar to the Cochrane review [7], oxamniquine at 40 and 50 mg/kg appears to be an effective, but less well tolerated, alternative limited however to S. mansoni, and no longer available in the WHO AFRO and EMRO regions. We provide an extensive report of safety findings. Tolerability was variably assessed in 12,435 subjects enrolled in 40 studies. Safety information was provided for 45% (25/55) of the studies included in the efficacy meta-analysis; we identified an additional 15 studies with safety information which were not included in the efficacy analysis. Reporting on safety was highly variable, and we cannot confidently conclude whether the absence of a given AE in a certain study means that it did not occur or it was not investigated. Lastly, the frequencies reported must be taken separately for each individual AE (they should not be accumulative), as some subjects might have experienced more than one AE. From comparative studies, risk seems not to change with the PZQ dose overall, although there are indications that higher doses may induce more events in some cases (e.g. 60 mg/kg had more of any AE and vomiting than 40 mg/kg; more abdominal pain with 40 mg/kg than 20 mg/kg), and that a split dose (20 mg/kg twice) may be better tolerated (in particular fewer cases of itching/rash). A direct comparison of two brands of praziquantel (Biltricide and Distocide) found the former to cause more AEs (abdominal pain, fever, diarrhea, headache, vomiting). The reason could be related to higher blood levels: when the pharmacokinetics of these two brands given at 40 mg/kg to healthy volunteers was compared, Biltricide peak concentration (Cmax) was 1.9 times higher (mean 1.281 vs. 0.685 µg/ml) and the area under the concentration curve (AUC) was 1.7 (mean 3550 vs. 2133 ng/h/ml) times higher than Distocide [47]. One limitation of the tolerability analysis relates to the diversity of the definitions of ‘events’ and methods to express incidence rates across studies. Therefore we opted for a permissive definition allowing for any sign or symptom occurring after treatment, acknowledging that this may indeed overestimate the real contribution of the treatment to the occurrence of events. Moreover information on signs or symptoms and their severity before treatment was only collected in a few studies so that it was not possible to detect which events were treatment-emergent. The adoption of more standardized methodologies in clinical studies would facilitate meta-analyses and strengthen the quality of evidence, as already pointed out for urinary schistosomiasis [85]; some of these questions can be answered more adequately only through an individual-subject data meta-analysis. Box 1 This is the largest collection of trials on praziquantel for treating urinary and intestinal schistosomiasis which has been meta-analysed for efficacy and safety Provided basic methodological standards are guaranteed, relaxing eligibility criteria for meta-analysis to allow for both comparative and non-comparative trials can broaden the database and complement more classical meta-analysis of randomised controlled trials Results support World Health Organization recommendations and are consistent with Cochrane systematic reviews Box 2 Steinmann P, Keiser J, Bos R, Tanner M, Utzinger J (2006) Schistosomiasis and water resources development: systematic review, meta-analysis, and estimates of people at risk. Lancet Infect Dis 6: 411–25. World Health Organization (2013) Assessing The Efficacy Of Anthelminthic Drugs Against Schistosomiasis And Soil-Transmitted Helminthiases. Available: http://apps.who.int/iris/bitstream/10665/79019/1/9789241564557_eng.pdf. Accessed 2/1/2014. DerSimonian R, Laird N (1986) Meta-analysis in clinical trials. Control Clin Trials 7: 177–88. Pocock SJ, Travison TG, Wruck LM (2008) How to interpret figures in reports of clinical trials. BMJ 336: 1166–9. Kramer CV, Zhang F, Sinclair D, Olliaro PL (2014) Drugs for treating urinary schistosomiasis. Cochrane Database Syst Rev 8: CD000053. Supporting Information Checklist S1 Prisma checklist. (PDF) Click here for additional data file. Figure S1 Flow chart of the number of studies and patients screened and eligible for the safety analysis. (BMP) Click here for additional data file. Table S1 Number of patients enrolled by country and species (all treatment arms). (PDF) Click here for additional data file. Table S2 Study design of the included publications: risks of bias and attrition. (PDF) Click here for additional data file. Table S3 Classing of studies by age groups based on reported age ranges. (PDF) Click here for additional data file. Table S4 Diagnostic approaches used in the studies (number of study/sites). (PDF) Click here for additional data file.

0 comments Cited 70 times – based on 0 reviews      Review now

Bookmark

Record: found
Abstract: found
Article: found

Is Open Access

Efficacy and safety of praziquantel 40 mg/kg in preschool-aged and school-aged children: a meta-analysis

Julien Zwang, Piero L. Olliaro (2017)

Background Children carry most of the schistosomiasis burden. While school-aged children are the principal target group of preventive chemotherapy with praziquantel, limited information on efficacy and safety exists for preschool-aged children. Methods Here, we conducted a meta-analysis of clinical trials of praziquantel for treating children with any form of schistosomiasis. Efficacy was reported as cure rate (CR) and egg reduction rates (ERR); statistical corrections were applied based on methodological disparities across trials to derive the predicted geometrical mean ERR (pERRgm). Safety was reported as frequencies of adverse events. Results Forty-seven comparative and non-comparative studies were identified, enrolling 15,549 children of whom 14,340 (92%) were assessed between 3 and 8 weeks post-treatment with praziquantel 40 mg/kg (the WHO-recommended treatment, n = 8,380, 56%) or comparators (n = 5,960, 44%). The median age was 10 years (range 1–19), 11% (n = 1,694) were preschool-aged. The CR and pERRgm with praziquantel 40 mg/kg were respectively: S. haematobium, 73.6% (95% CI: 63.5–81.40, 25 study arms) and 94.7% (95% CI: 92.7–96.4); S. mansoni, 76.4% (95% CI: 71.5–81.0, 34 arms) and 95.3% (95% CI: 94.2–96.2); S. mansoni/S. haematobium, 67.6% (95% CI: 54.1–80.7, 5 arms) and 93.4% (95% CI: 89.9–96.2); S. japonicum, 94.7% (95% CI: 92.2–98.0) and 98.7% (95% CI: 98.3–99.2). Mixed-effect multivariate analysis found no significant difference between preschool- and school-aged children for CR or pERRgm in S. haematobium (P = 0.309 and P = 0.490, respectively) or S. mansoni (P = 0.982 and P = 0.895) after controlling for time of assessment, formulation, intensity of infection and detection method. Praziquantel was reportedly safe at all ages, with only mild reported adverse events which cleared rapidly after treatment. Conclusions Praziquantel 40 mg/kg was effective at reducing infection intensity in all Schistosoma species without differences between preschool- and school-aged children. However, conclusions should be tempered because of the limited number of preschool-aged children enrolled, disparities in study procedures and limited information made available in publications, as well as the current imperfect test-of-cure. Also, although reportedly well-tolerated, safety was inconsistently assessed. Studies in target groups, individual-data meta-analysis and standardised methodologies are needed for more robust evidence-base. Electronic supplementary material The online version of this article (doi:10.1186/s13071-016-1958-7) contains supplementary material, which is available to authorized users.

0 comments Cited 44 times – based on 0 reviews      Review now

Bookmark

Record: found
Abstract: found
Article: found

Is Open Access

Schistosoma mansoni Infections in Young Children: When Are Schistosome Antigens in Urine, Eggs in Stool and Antibodies to Eggs First Detectable?

J. Stothard, Jose Sousa-Figuereido, Martha Betson … (2011)

Introduction Throughout the last decade several large-scale preventive chemotherapy campaigns, waged against neglected tropical diseases, have progressively scaled up operations to reach nationwide coverage levels in Uganda [1], [2]. For control of intestinal schistosomiasis, as caused by Schistosoma mansoni infection, an active monitoring and surveillance programme, set within the national control programme (NCP), has provided important disease-specific information, assessing the impact of treatment upon the recipient population, as well as, re-alignment of original control objectives first set forth in 2003 [3], [4]. Following WHO guidelines, mass-drug administration of praziquantel (PZQ) is typically focused towards treatment of school-aged children (≥6 years) and adults who reside within disease endemic regions [5], [6]. PZQ is provided free of charge by the NCP and analysis of school and(or) community treatment registers has shown that several million people have received at least one annual treatment of PZQ within the last five years [1], [7]. Although this represents a considerable achievement, targeted epidemiological surveys have revealed that coverage is incomplete as in certain areas, e.g. shoreline environments of Lakes Victoria and Albert, large numbers of preschool-aged children (≤5 years) and infants (≤1 years) are infected with S. mansoni and have been largely overlooked by the treatment campaign [8], [9], [10]. To ensure that this unfortunate health inequality does not persist the treatment needs of younger children are being assessed and we have recently called for formal inclusion of these young children within the Ugandan NCP [11]. It can be safely assumed, for example, that mass-treatment initiatives are vital in most in shoreline villages where infections can be common. Given the geographical focality of schistosomiasis and itinerancy of lakeshore communities, however, an important future challenge for the NCP is collection of sufficient disease-specific information to better tailor local drug needs and set parameters for subsequent programme monitoring [12], [13]. Attention will therefore focus upon those sections of villages where young children are frequently bathed in freshly drawn lake water or are within range of regular ambulation to the lake margins. Owing to the unique natural history and developmental biology of schistosomes within the mammalian host [14], accurate identification of infected cases is challenging [15], even more so in the younger child where the founding worm population has only recently established and begun to mature. Before female worms develop their full egg-laying capacity, sporadic deposition of eggs may take place with a proportion of these being voided into the bowel lumen and ejected in faeces whilst the remainder become trapped within the host's tissues [16]. Interacting with this are also the beginnings of the child's innate and adaptive immune responses to excretory-secretory products of the worms themselves, as well as these responses being primed or modulated by maternally induced effects, for example, during pregnancy and(or) breastfeeding [17], [18], [19], [20]. It is also of particular note that the child's immune system is in a maturing flux of recognition between self- and non-self epitopes [21] and the efficacy of PZQ, which is poor against immature worms of S. mansoni [22], is only starting to be explored in this ageclass [11]. From a general diagnostic perspective as existing tools are sub-optimal, improvement of methods and techniques for detection of intestinal schistosomiasis continues [15] but in the context of the younger child, it is not yet clear which of the present methods, or combinations thereof, is either most appropriate or applicable for routine use within the NCP. We therefore report on a field-based study which attempted to determine the age of first infection in very young children with available techniques and also estimate, as accurately as possible, the general prevalence of intestinal schistosomiasis within this ageclass from a typical lakeshore community. The performance of methods that detect schistosome - antigens in urine, antibodies to egg antigens in serum and eggs in stool - was compared. For ease of comparison, our methods are subsequently referred to as: an antigen detection method (ADM), an indirect egg detection method (IEDM) and a direct egg detection method (DEDM), respectively. Materials and Methods This field study was carried out in April 2009 in Bugoigo on Lake Albert (GPS co-ordinates, N 01° 54′.481″, E 31° 24′.597″), a fishing village impoverished both in terms of sanitation and hygiene that has been the location of several previous research/control studies on intestinal schistosomiasis [23], [24], [25], [26]. Prevalence of infection within local school-aged children has been continuously high (>50%) despite annual chemotherapy [27] and infections in infants and preschool children were first formally recorded in July 2007 [11]. Study location and participants Owing to itinerancy, the exact number of inhabitants in Bugoigo is not precisely known but is likely in the region of several thousand. The village contains up to three thousand traditional hut dwellings which stretch 3–4 km along the lakeshore and up to 1–2 km inland. Sanitation and hygiene in this village is minimal with few potable water sources and insufficient pit latrines. Household water is typically drawn directly from the lake at specific collection points and then taken back to each homestead in plastic jerry cans for subsequent domestic use. These lakeshore margins, like elsewhere on Lake Albert, provide conducive aquatic habitats for Biomphalaria spp., the intermediate snail hosts of S. mansoni, and can be found throughout the year, although infected snails vary in numbers seasonally [28], [29]. The immediate and longer-term objectives of this study were explained to the local community mobiliser who identified a total of 134 mothers that were willing to participate, bringing up to two of their infants/preschool children (≤5 years of age), and attend the two-day clinic commencing on the following day. After obtaining written informed consent from each mother on her own behalf and on behalf of her child(ren), urine, stool and fingerprick blood samples were obtained from all participants on the first day of the clinic. Mothers were then asked a suite of detailed questions recording their demography and water contact behaviours (the questionnaire is available upon request to the corresponding author). After receipt of the second-day stool (and urine sample), all participants, regardless of their infection status, were treated for schistosomiasis and soil-transmitted helminthiasis with PZQ (40 mg/kg) (CIPLA, Mumbai, UK) and 400 mg albendazole (GSK, Uxbridge, UK) under medical supervision in conditions typical of mass-drug administration [30]. For smaller children, a chewable albendazole half-tablet (200 mg) was given and PZQ tablets were first crushed in orange juice before being administrated by spoon-feeding by their mother under supervision. The diagnostic findings for schistosomiasis here are reported for the children only. Schistosome antigens in urine (ADM) Each child's urine sample was visually inspected for macro-haematuria/turbidity and a random sample was tested for micro-haematuria with Hemastix (Bayer, UK) to exclude the possibility of urinary schistosomiasis or other active urinary tract infections. A 50 µl aliquot was then tested for the presence of schistosome circulating cathodic antigen (CCA) using a commercially available lateral flow immuno-chromatographic urine dipstick (Rapid Medical Diagnostics, Pretoria, RSA) originally developed in Holland [31]. On a subset of 90 children, urine-CCA tests were performed in duplicate to assess variation between dipsticks. To facilitate better recording of the visual intensity of the CCA reaction band within the test zone, results were visually graded against a reference chart for: trace, single (+), double (++) and triple (+++) positive reactions [32]. When creating binomial variables to depict infection status according to CCA, two variations were taken into account: the first considering trace results as negative infection status and the second considering trace results as positive infection status. The urine CCA reagent strip is referred to as an ADM (antigen detection method) from now on. Antibodies to soluble egg antigens (IEDM) A commercially available ELISA kit (IVD Inc.; Carlsbad, USA) was used to test for host antibodies (IgG/M) to soluble egg antigens (SEA) according to manufacturer's instructions. Approximately 75 µl of finger-prick blood was taken from each child and serum was harvested, then diluted 1∶40 with specimen dilution buffer before loading a total of 100 µl into each ELISA microwell [11]. Positive and negative control sera were included on each batch of testing. Upon completion, each ELISA plate was placed on a white card and the colour within each microwell (ranging from colourless to yellow) was recorded by visual inspection. Positive reactions were classified either as trace (faint yellow), single (+, light yellow), double (++, yellow) or triple (+++, dark yellow) upon visual comparison with the control sera. The SEA-ELISA is referred to as an IEDM (indirect egg detection method) from now on. Direct egg-detection methods in stool (DEDM) Three parasitological methods Kato-Katz, percoll and FLOTAC, henceforth referred to as direct egg detection methods (DEDMs), were attempted on each stool specimen to visualise eggs. However, owing to the differing amounts of stool required for each technique, it was not always possible to assemble a complete data set for every child with each of these three methods. Duplicate Kato-Katz (K-K) thick smears (41.7mg) were made from first and second day stool samples (N = 242 children) [33]. The four faecal smears were each examined under the microscope at x100, schistosome eggs were counted and later expressed as eggs per gram (epg) of faeces. Infection intensity was classified as light (1–100 epg), medium (101–400 epg) and heavy (>400 epg) infections according to WHO guidelines [5]. The methodology of Eberl [34] using sedimentation of schistosome eggs by centrifugation through a solution of percoll (Percoll 77237 (1.130 g/ml), Fluka, Sigma-Aldrich Chemie GmbH, Switzerland) was also implemented on-site to visualize eggs (N = 96 children on first day stool). The egg-floatation procedure known as FLOTAC [35] was performed off-site back in Kampala on a formalin-fixed stool specimen archive (N = 191 children taken from the second day stool) whereby schistosome eggs are collected by floatation centrifugation through a solution of zinc sulphate at specific gravity of 1.35. Data handling and statistical analyses Data were collected from each individual using pro-forma data sheets, which were then transferred into electronic format using Microsoft Excel. The data thus collated were analysed using MS Excel and R statistical package version 2.8.0 [36]. For prevalence data and diagnostic parameters, 95% confidence intervals (CI95) were estimated using the exact method [37]. Prevalence comparisons were performed using (one-tailed) Fisher's exact modification of the 2×2 chi-squared test [38]. For infection intensity values, the arithmetic mean of positive cases was chosen as the measure of central tendency. Data from the FLOTAC and percoll methods were analysed by combining with K-K results and revising the diagnostic criterion so individuals were considered positive if an egg was detected by at least one DEDM. The diagnostic performances of the ADM (including and excluding trace reactions as a positive diagnosis) and IEDM were tested qualitatively as a rapid diagnostic for intestinal schistosomiasis, considering DEDMs as the ‘gold-standard’. Additionally, a third ‘gold standard’ was created using data from the ADM (including and excluding trace reactions as positive diagnoses) against which to test IEDM data (N = 242). Diagnostic sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated according to the different ‘gold standards’ [38]. The diagnostic powers of ADM and IEDM were calculated using all individuals, and then segregated by sex or age (≤3 years of age versus >3 years of age). P-values 399 epg) 0.4 0.0–2.3 5 years ADM (CCA incl. trace) 242 42.6 36.3–49.1 NA 6 months + reaction 17.8 13.2–23.2 9 months ++ reaction 11.6 7.8–16.3 11 months +++ reaction 1.7 0.5–4.2 2 years IEDM (SEA-ELISA) 242 45.9 39.5–52.4 NA 6 months + reaction 15.3 11.0–20.5 6 months ++ reaction 19.4 14.6–25.0 1 year +++ reaction 11.2 7.5–15.8 9 months DEDMs (all * ) 242 24.4 19.1–30.3 NA NA IEDM+DEDM +ADM 242 47.5 41.1–54.0 NA NA $: Age of first positive. *Note not all 242 children were examined with percoll and FLOTAC (see methodology). Ages of becoming first positive The age of first positive (AFP) for each method is presented in Table 1. For DEDM, the youngest child with eggs in stool was 9 months old, with medium and heavy infections found at 3 and 5 years of age, respectively. For ADM, trace reactions, single, double and triple positives were found in an ascending series of 6 months, 9 months, 11 months and 2 years of age, respectively. For IEDM, trace reactions began at 5 months of age while single, double and triple positive reactions were found in children as young as 6 months, 1 year and 9 months old, respectively. All tests concur on a mean age of first infection within the third year of life. ADM detected infections slightly ahead of I/DEDMs (3.2 years v. 3.4 years v. 3.7 years, respectively). The order of this temporal series is largely concordant with an absolute minimum age of becoming first positive. Cross-tabulations of diagnostic scores In the absence of a genuine ‘gold standard’ where the infection status of each child is precisely known, it is necessary to explore relationships between diagnostic scores and infection intensities empirically, and to cross-tabulate diagnostic permutations by investigation. There was negligible variation in diagnostic performance of all protocols tested when classifying the data according to sex and age (data not shown) and general trends were reported from now on. Plotting the relationship between ADM and DEDM revealed some immediate trends, Fig. 2. Whilst there were children positive for ADM who were egg-negative, as the epg increases there was a corresponding increase in the proportion of positive ADM tests and once medium/heavy intensity infections were reached, all ADM tests were clearly positives, see Fig. 2A. Plotting the faecal epg of each child against the intensity of the corresponding ADM test further revealed this positive association, see Fig. 2B. 10.1371/journal.pntd.0000938.g002 Figure 2 Comparisons of ADM and DEDM. Figure 2A Rectangular bar chart representing egg infection intensity classifications (as calculated by Kato-Katz) versus reaction intensity of the ADM (visual strength of the CCA urine dipstick test band). Figure 2B Boxplot of the egg faecal epg against ADM reaction intensity shows a positive increasing association. Considering the relationship between IEDM and DEDM revealed similar trends, see Fig. 3. Despite some children being positive for IEDM while being egg-negative, as the faecal epg increases there was a corresponding increase in the IEDM reaction strength, with all medium/heavy intensity infections diagnosed as clear strong positives (Fig. 3A & B). 10.1371/journal.pntd.0000938.g003 Figure 3 Comparisons of IEDM and DEDM. Figure 3A Rectangular bar chart representing egg infection intensity classifications (as calculated by Kato-Katz) versus reaction intensity of the IEDM (visual strength of the SEA-ELISA test well). Figure 3B Boxplot of the egg faecal epg against IEDM reaction intensity shows a positive increasing association. The relationship between ADM and IDEM was less clear-cut. Children who were ADM negative or trace had a median negative (or trace) IEDM reaction, but the proportionate increase of ADM positives with rising IEDM designations of positive (+) or strong positives (++/+++) was not as great as that seen with DEDM. For example, nearly 40% of children who were IEDM strong positive elicited a negative ADM reaction, Fig. 4A. As the intensity of the ADM result stepped up towards double and triple positive reactions, this typically corresponded better with increasing IEDM classifications, Fig. 4B. 10.1371/journal.pntd.0000938.g004 Figure 4 Comparisons of ADM and IEDM. Figure 4A Rectangular bar chart representing ADM intensity classifications (visual strength of the CCA urine dipstick test band) versus reaction intensity of the IEDM (visual strength of the SEA-ELISA test well). Figure 4B Boxplot of the IEDM reaction intensity against ADM shows a positive increasing association but the relationship is less clear-cut than that shown in Figures 2B & 3B . Using available data it was possible to conduct an exploration of diagnostic performances of the ADM and IEDM versus DEDM and against each other (Table 2). First, considering an ADM trace reaction to be an infection negative and comparing with positive diagnosis by at least one of the DEDMs, the ADM had a sensitivity of 59.3%, specificity of 95.6%, PPV of 81.4% and NPV of 87.9%. When considering an ADM trace reaction to be a positive infection, and comparing to diagnosis by at least one of the DEDMs, ADM had a sensitivity of 81.4%, specificity of 69.9%, PPV of 46.6% and NPV of 92.1%. 10.1371/journal.pntd.0000938.t002 Table 2 Comparison of diagnostic scores by ADM and IEDM using DEDM (all) as ‘gold standard’. Diagnostic test Diagnostic target N Sensitivity Specificity PPV NPV (%,CI 95) (%,CI 95) (%,CI 95) (%,CI 95) ADM (excl. trace) DEDM 242 59.3 95.6 81.4 87.9 (45.8–71.9) (91.6–98.1) (66.6–91.6) (82.6–92.1) ADM (incl. trace) DEDM 242 81.4 69.9 46.6 92.1 (69.1–96.3) (62.7–76.5 (36.7–56.7) (86.3–96.0) IEDM DEDM 242 93.2 69.4 49.5 96.9 (83.5–98.1) (62.2–76.0) (40.0–59.2) (92.4–99.2) ADM (excl. trace) 242 86 62.8 33.3 95.4 (72.1–94.7) (56.0–69.5) (24.7–42.9) (90.3–98.3) ADM (incl. trace) 242 60.2 64.7 55.9 68.7 (50.1–69.7) (56.2–72.7) (46.1–65.3) (60.0–76.5) The IEDM when compared to diagnosis by DEDM, demonstrated a sensitivity of 93.2%, specificity of 69.4%, PPV of 49.5% and NPV of 96.9%. For details on the performance of the ADM or IEDM compared to diagnosis by all DEDMs (using a subset of the data), and for CI95 around each value, see Table 2. Discussion With limited access to safe water sources, and high levels of local transmission of S. mansoni, conditions in Bugoigo are particularly conducive for young children to acquire S. mansoni infections, and from a very early age. Approximately half of our children had intestinal schistosomiasis. As might be expected, regardless of techniques used, there was an obvious positive association between increasing diagnostic patency of infection with increasing age of the child. Presumably this was resultant from a progressive temporal accumulation of antigens, eggs and antibodies. Congruence between diagnostic methods became most apparent in children between 3¼–3¾ years of age, broadly consistent with the overall mean age of infected children within our sample. Prevalence of intestinal schistosomiasis in children under 3 years of age, however, was 35.5% (CI95 27.9–43.8%) and other studies have also revealed that schistosomiasis in very young children can be common [9], [11], [39]. While egg excretions of these children were of ‘light’ intensity, such infected children will not normally receive praziquantel treatment until they have either entered primary school or if the NCP now formally extends its treatment remit to include this ageclass. Thus an infected child could therefore wait up to 3–4 years before receiving first treatment, and with this may have already entered a more ‘chronic’ stage of disease [40], [41], [42], [43]. For example, earlier clinical and ultrasound studies in Uganda in children aged 6 and above, have shown significant hepatosplenomegaly (i.e. putative morbidity from intestinal schistosomiasis), and while they have not yet developed pipe-stem liver fibrosis, up to 15% can have diffusely echogenic livers with pocketed foci, typical of image pattern B (‘the starry sky’ classification) [3], [27]. Without medication, it is likely that these preschool children will progress towards ‘moderate’ infection intensities before they become of school age. This might better explain the observations of Balen et al. that many adolescent Ugandan children have surprisingly severe intestinal schistosomiasis [44]. Although it is not yet proven that infection in very early childhood leads to heightened morbidity in later childhood and adolescence, this scenario appears plausible. From animal models, it is known that only a fraction of penetrating cercariae successfully migrate to, and later mature in, the hepatic portal system. After adult worms reach full fecundity, schistosome eggs can be found in stool around 6 weeks after cercarial exposure and it is commonly held that females of S. mansoni produce up to 100–300 eggs per day, although many fail to be voided into the faeces [16]. Given the insensitivity of DEDMs in stool [34], it is not surprising that false negatives are inferred and the low egg-detection threshold(s) likely contribute to the longer apparent lag of 7–8 months between infection and egg-patency apparent between experimental schistosomiasis and the situation in the field. Moreover, it should be noted that the relationship between excreted eggs in stool and worm burdens is not always straightforward [45] and that infected laboratory animals are typically exposed with a single substantive dosing of cercariae. By contrast, and in this natural setting, exposure and infection is likely a more gradual process, i.e. the so-called trickle infection dynamic [46], and our children are at least two orders of magnitude greater in body size than most animal models. An age of first infection? While some children were patently infected during the first year of life, others were not. Thus a sub-set of children exists with increased infection risk factors which we explain by the following synopsis. As children are born throughout the year, in a largely asynchronous fashion, whilst their initial age of first exposure to unsafe water might be broadly similar (i.e. within first few months of life as mothers begin to bathe them in jerry-can collected water or in the lake directly) their accumulated infection risk will not be equivalent owing the seasonality of local transmission factors and their particular timeframe of exposure within it contingent upon their mother's infant bathing and domestic water drawing practices [11]. Estimating this accumulated risk of infection reliably over the seasonal time frame of potential exposure is problematic as day-to-day variations within water collection times, its storage and actual domestic use (within each household) introduce many stochastic processes. Estimating cumulative infection risk is therefore easily confounded but an ad hoc investigation of infection risk associated with jerry-can collected water in June 2009, however, has confirmed that sentinel laboratory-bred mice could become infected to freshly drawn water [28]. Seasonal patterns, which operate in umbrella fashion over and above these specific-exposure patterns, no doubt effect this asynchronous age of first infection. Thus there will be no ‘absolute age’ of first infection but rather a ‘range of ages’ depending upon these intricate covariates of exposure. Only after a child has passed through sufficient ‘windows of exposure’, their probability of infection rises to an eventual certainty, after which, it is incumbent on the diagnostic tools to capture their parasitological status as accurately as possible. Comparison of diagnostic scores From first appearances the ADM looks to best capture and identify infections in early stage, especially when we consider trace results as putative infection positives. A contentious issue in the use of the CCA reagent strip has been the interpretation of the exact diagnosis of this ‘trace’ result which can be confounded by non-specific inflammatory factors or breast-feeding [32], [47]. Interpretation of ‘trace’ is more contentious when surveying children under three years of age, where worm burdens are presumably lower than what might be expected in their school-aged counterparts. Interestingly, the percentage change in prevalence estimated according to the CCA reagent strip when excluding and including trace results as a positive diagnosis is significantly larger in the very young children (≤3 years of age) –10.1% v. 36.2% (+358%) – than in those aged four and five years of age –30.1% v. 52.7% (+175%) which is fitting with our understanding of increasing worm burdens through time. Thus we postulate that using ‘trace’ as positive firmly points towards a future use of the urine CCA-dipstick as an early indicator of infections which are as yet to become egg- or antibody-patent. It is particularly notable that the prevalence based on the ADM, when considering trace as positive, is very close to that of IEDM (Fig. 1), yet the diagnostic performance with it was not particularly congruent (see Fig. 3 and Table 2) so we still have an incomplete understanding of this infection progression. The dynamics of other ADM have been explored elsewhere in the context of recently acquired infection but not in very young children [48]. The ADM showed very promising diagnostic performance and robust field performance with high sensitivity and NPV scores (83.9% and 85.3%, respectively) when we considered trace results as a positive diagnoses and very high specificity and PPV scores (95.5% and 90.5%, respectively) when we considered trace results as negative diagnoses. This bimodal use of the test criteria could be advantageous from a control perspective. For instance, if a confident estimate of the suspected occurrence of infections within a population is needed, one should consider trace results as positives. On the other hand, to monitor the prevalence of ‘actual’ infection, or rather more easily identify those who do not, one should consider trace results as negatives. The former would be important if treatments were to be given out en masse as triggered by exceeding an aggregated local prevalence threshold while the latter would be important if treatment were to be withheld in an individual patient setting on the basis of test and treat. Towards promotion of safe water With the gradual rise of infection prevalence in older children (over and above our asynchronous infection hypothesis), this trend must represent the spread of several risk factors, rather incipiently, across our cohort. Aggregation of infections in schistosomiasis is well-known [49] but it would be interesting to establish why approximately half of our study cohort had no evidence of infection despite living within the same village. As we were insufficiently aware of the exact locations of sampled households within the village, this could simply represent a cryptic spatial micro-patterning (i.e. these children who live slightly further away from the lake have less contact with viable cercariae) so we are now undertaking fine scale mapping of these individual households with GPS units. If, however, other causal factors could be identified and, perhaps more importantly, were these amenable to manipulation, it could lead to future infection mitigation measures. Presently, within the NCP there are no health education materials targeted towards these mothers and their young children. More importantly and in terms of policy realignment of the NCP, a useful formal recommendation would be to initiate cross-sectorial activities with water and sanitation NGOs to improve immediately the domestic water quality at Bugoigo and elsewhere along the Lake Albert shoreline. Rather than focusing upon expensive infrastructure development, it could be achieved by introduction of simple water storage or modification measures. For example, as schistosome cercariae are an ephemeral larval stage, freshly drawn water can be rendered harmless for schistosomiasis by simple resting for 24 hrs, by crude filtration or by introduction of mild disinfectants [16]. Thus without initiating a better dialogue with these women of children bearing age through better public health education, mothers will remain sadly ignorant of the risks that making use of this unsafe water has for themselves and that of the future health of their child [11]. In this dialogue, the NCP should be receptive to explore which infection mitigation measures are best feasible and, by this token, help to provide safe water for domestic use which is well-received, implementable and effective. Epidemiological indicators and treatment needs It is evident that infants and preschool children in Bugoigo, and other similar lakeshore villages of Uganda [12], are living in need of treatment. However, addressing how these children could be best identified is not yet clear, as are epidemiological parameters which should be collected for estimating treatment needs and also impact assessment. For example, should mass-treatment of all infants/preschool children take place when a sub-sample of an equivalent age range has been proven to be infected, or should treatment be allocated based at an individual level using the result of a diagnostic test in a ‘test and treat’ setting? It is outside the remit of this present paper to make a cost-effectiveness calculation but a clear drawback of the IEDM method is that, whilst initially useful to establish if a child is infected (and there is no evidence in these data to suggest a passive maternal transfer of antibodies has been confounding), monitoring this parameter after treatment will be largely uninformative owing to residual antibody titres remaining after infection has putatively cleared. Thus IEDM is only useful at baseline but as an initial estimate of infection prevalence could be powerfully applied in identification and selection of villages, or sentinel locations, to first define the extent of the problem at intervention baseline. This of course assumes the majority of examined children can mount an antibody response and is not confounded by high levels of immune-suppression, by HIV for example, which is likely high in these fishing villages. In contrast, both ADM and DEDMs have the potential ability to better track the dynamics of worm populations after treatment [22], [34] but the insensitivity of DEDMs is of particular concern. Put simply, numerous adult worms may reside within the host but are yet not depositing sufficient eggs to be visualised in stool on the day of sampling. Thus through lack of alternatives a pragmatic way forward would be to focus upon more widespread application of the ADM. The advantages of the ADM have been discussed elsewhere in the context of programmatic monitoring [14] but the future challenge will be for the NCP to meet the financial costs of using these rapid diagnostic tests in scale-up of operations. This is particularly true if these are to be used in a ‘test and treat’ setting when large numbers of tests would be utilized [14]. Given the low price of PZQ treatment, to maintain an affordable diagnosis versus treatment differential, a rational strategy would be to examine a sub-set of children and if local prevalence exceeded a given threshold, mass-treatment is advised. Such a strategy is presently within the resources available to the NCP but best sample sizes and prevalence thresholds remain to be determined. Supporting Information Checklist S1 STROBE Checklist (0.10 MB DOC) Click here for additional data file.

0 comments Cited 43 times – based on 0 reviews      Review now

Bookmark

All references

Author and article information

Contributors

Piero L. Olliaro: Role: ConceptualizationRole: InvestigationRole: MethodologyRole: Project administrationRole: ResourcesRole: SupervisionRole: Writing – original draftRole: Writing – review & editing

Jean T. Coulibaly: Role: InvestigationRole: Writing – review & editing

Amadou Garba: Role: InvestigationRole: Writing – review & editing

Christine Halleux: Role: Project administrationRole: Writing – review & editing

Jennifer Keiser: Role: InvestigationRole: ResourcesRole: Writing – review & editing

Charles H. King: Role: InvestigationRole: ResourcesRole: Writing – review & editing

Francisca Mutapi: Role: InvestigationRole: ResourcesRole: Writing – review & editing

Eliézer K. N’Goran: Role: InvestigationRole: ResourcesRole: Writing – review & editing

Giovanna Raso: Role: InvestigationRole: ResourcesRole: Writing – review & editing

Alexandra U. Scherrer: Role: InvestigationRole: ResourcesRole: Writing – review & editing

José Carlos Sousa-Figueiredo: Role: InvestigationRole: ResourcesRole: Writing – review & editing

Katarina Stete: Role: InvestigationRole: ResourcesRole: Writing – review & editing

Jürg Utzinger: Role: InvestigationRole: ResourcesRole: ValidationRole: Writing – review & editing

Michel T. Vaillant:

ORCID: http://orcid.org/0000-0003-4714-8128

Role: ConceptualizationRole: Data curationRole: Formal analysisRole: MethodologyRole: SoftwareRole: ValidationRole: Writing – original draftRole: Writing – review & editing

Antonio Montresor: Role: Editor

Journal

Journal ID (nlm-ta): PLoS Negl Trop Dis

Journal ID (iso-abbrev): PLoS Negl Trop Dis

Journal ID (publisher-id): plos

Journal ID (pmc): plosntds

Title: PLoS Neglected Tropical Diseases

Publisher: Public Library of Science (San Francisco, CA USA )

ISSN (Print): 1935-2727

ISSN (Electronic): 1935-2735

Publication date (Electronic): 22 June 2020

Publication date Collection: June 2020

Volume: 14

Issue: 6

Electronic Location Identifier: e0008277

Affiliations

[1 ] Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University of Oxford, Oxford, United Kingdom

[2 ] Swiss Tropical and Public Health Institute, Basel, Switzerland

[3 ] University of Basel, Basel, Switzerland

[4 ] Unité de Formation et de Recherche Biosciences, Université Félix Houphouët-Boigny, Abidjan, Côte d’Ivoire

[5 ] Centre Suisse de Recherches Scientifiques en Côte d’Ivoire, Abidjan, Côte d’Ivoire

[6 ] Department of Control of Neglected Tropical Diseases, World Health Organization, Geneva, Switzerland

[7 ] UNICEF/UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases (TDR), World Health Organization, Geneva, Switzerland

[8 ] Center for Global Health and Diseases, Case Western Reserve University, Cleveland, Ohio, United States of America

[9 ] Schistosomiasis Consortium for Operational Research and Evaluation, University of Georgia, Athens, Georgia, United States of America

[10 ] NIHR Global Health Research Unit Tackling Infections to Benefit Africa (TIBA), Ashworth Laboratories, University of Edinburgh, Edinburgh, United Kingdom

[11 ] Institute of Immunology and Infection Research, Centre for Immunity, Infection and Evolution, School of Biological Sciences, Ashworth Laboratories, University of Edinburgh, Edinburgh, United Kingdom

[12 ] Division of Infectious Diseases and Hospital Epidemiology, University Hospital Zurich, University of Zurich, Zurich, Switzerland

[13 ] Department of Life Sciences, Natural History Museum, Wolfson Wellcome Biomedical Laboratories, London, United Kingdom

[14 ] Centro de Investigação em Saúde de Angola, Hospital Provincial, Bengo, Angola

[15 ] Division of Infectious Diseases, Department of Medicine II, University Medical Center Freiburg, University of Freiburg, Freiburg, Germany

[16 ] Centre of Competences for Methodology and Statistics, Luxembourg Institute of Health, Strassen, Luxembourg

World Health Organization, SWITZERLAND

Author notes

The authors have declared that no competing interests exist.

* E-mail: michel.vaillant@ 123456lih.lu

Author information

Michel T. Vaillant http://orcid.org/0000-0003-4714-8128

Article

Publisher ID: PNTD-D-19-02157

DOI: 10.1371/journal.pntd.0008277

PMC ID: 7360067

PubMed ID: 32569275

SO-VID: c5fcd69b-7ee9-49b9-a2f0-353ed5edeb27

License:

This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

History

Date received : 27 December 2019

Date accepted : 8 April 2020

Page count

Figures: 10, Tables: 8, Pages: 23

Funding

The Special Programme for Research and Training in Tropical Diseases (TDR) supported this work and the clinical trial reported in Olliaro et al. (2013) with core funding. The Luxembourg Institute of Health supported the analyses with core funding. The data from Zimbabwe was collected during a research project funded by the World Health Organization (FM) and Thrasher Research Fund (FM). This research was commissioned in part by the National Institute for Health Research (NIHR) Global Health Research programme (16/136/33) using UK aid from the UK Government (FM). JK and JTC are grateful to the European Research Council for financial support (ERC-2013-CoG 614739-A_HERO). CHK is supported by the Schistosomiasis Consortium for Operational Research and Evaluation (SCORE), funded by the University of Georgia Research Foundation through a grant from the Bill & Melinda Gates Foundation. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Custom metadata

PLOS Publication Stage vor-update-to-uncorrected-proof

Publication Update 2020-07-14

Data Availability Data were not produced by the current Meta-Analysis and only the original authors could make them available on an individual basis. These data are held in the repository of the Infectious Diseases Data Observatory ( IDDO.org). IDDO promotes data sharing and data re-use to generate new evidence that improves health and understanding of disease. Requests to access data can be submitted by email to dataccess@ 123456iddo.org via the Data Access Application Form available at IDDO.org/accessing-data. If eligible, requests will be reviewed by the IDDO Data Access Committee to ensure that use of data protects the interests of the participants and researchers according to the IDDO principles of data sharing (see https://www.iddo.org/data-sharing/accessing-data).

Efficacy and safety of single-dose 40 mg/kg oral praziquantel in the treatment of schistosomiasis in preschool-age versus school-age children: An individual participant data meta-analysis

Read this article at

Abstract

Background

Methodology

Principal findings

Conclusions/significance

Author summary

Related collections

SDG 3: Good Health and Well-Being

Most cited references 33

Clinical Efficacy and Tolerability of Praziquantel for Intestinal and Urinary Schistosomiasis—A Meta-analysis of Comparative and Non-comparative Clinical Trials

Efficacy and safety of praziquantel 40 mg/kg in preschool-aged and school-aged children: a meta-analysis

Schistosoma mansoni Infections in Young Children: When Are Schistosome Antigens in Urine, Eggs in Stool and Antibodies to Eggs First Detectable?

Author and article information

Contributors

Journal

Affiliations

Author notes

Author information

Article

History

Page count

Funding

Categories

Custom metadata

Comments

Comment on this article

Similar content 252

Cited by 5

Most referenced authors 486