Safety of praziquantel in persons with and without schistosomiasis: systematic review and meta-analysis

Millions of praziquantel doses have been delivered in schistosomiasis endemic populations through preventive chemotherapy. However, no comprehensive assessment of short and long-term safety has been conducted. This systematic review assessed safety of praziquantel in persons with and without schistosome infections who received praziquantel treatment. MethodsWe identified relevant studies (published, unpublished, in press or preprint) that assessed safety of praziquantel without language restriction. We searched MEDLINE, EMBASE, CINAHL, and LILACS from 1978 to 31st October 2021, using well-formulated and piloted search strategy. We also searched the Cochrane Infectious Diseases Group Specialized Register, CENTRAL (The Cochrane Library 2021), mRCT, Google Scholar, Hinari and Africa Journals Online. References of relevant studies were checked and experts were contacted for additional studies. One author searched and managed the search output. Two authors selected studies, extracted data and assessed quality of the included studies for risk of bias. Activities at all stages were checked independently by a third reviewer. Discrepancies were resolved through discussion among the authors. Data were analysed with RevMan v5.4 and STATA v17. Binary outcomes were reported as risk ratio using random-effects model and continuous outcomes as mean difference, all presented with their 95% confidence intervals. P-value was set at 0.05. Heterogeneity was assessed using I2-static and where possible sensitivity analysis was conducted. When pooling of data was not possible, we presented data in a narrative synthesis and as tables. Main resultsThe search retrieved 3202 studies of which 134 met the inclusion criteria; 94 (70.1%) were conducted in Africa, 17 (12.7%) in Asia, 15 (11.2%) in the Americas (14 from Brazil), 4 (3.0%) in the Middle East and 3 (2.2%) in Europe. Praziquantel mostly resulted in mild-to-moderate and transient adverse events, however, majority of the included studies had design issues, including very short follow-up times (mostly few hours) for assessing incidence of adverse events. Less than <10% of the studies reported severe or serious adverse events. The subgroup analyses of twenty studies comparing school age children (SAC) and adults, and involved over one million participants found no difference in the nature of adverse events, but SAC experienced higher incidence than adults: headache (RR 3.07, 95% CI 2.32 to 4.06, twenty studies, I2 = 98%, p < 0.00001), dizziness (RR 1.80, 95% CI 1.36 to 2.37, p = 0.0001), vomiting (RR 2.43, 95% CI 1.87 to 3.14, I2 = 98%; p < 0.00001), four time for abdominal pain (RR 3.97, 95% CI 3.09 to 5.10, I2 = 96%, p < 0.00001), nausea (RR 1.67, 95% CI 1.32 to 2.12, I2 = 97%, p < 0.0001), general discomfort (RR 1.32, 95% CI 1.03 to 1.68, I2 = 97%, p < 0.00001), fever (RR 4.78, 95% CI 3.04 to 7.52, I2 = 98%, p < 0.00001), diarrhoea (RR 1.41, 95% CI 1.12 to 1.78, I2 = 92%, p < 0.00001), itching (RR 2.42, 95% CI 1.58 to 3.70, I2 = 93%, p <0.0001) and breathing difficulty (RR 2.46, 95% CI 1.41 to 4.29, I2 = 92%, p = 0.002). There was no statistically significant difference in incidence of swelling. Some of the studies that assessed safety in pregnant women reported serious events including miscarriages, foetal deaths and congenital anomalies, but the evidence is incoclusive given the limited numbers. Some studies reported praziquantel-related visual adverse events, but evidence is limited and remains inconclusive. There was paucity of data on long term adverse events, and events in co-morbidity, polypharmacy, co-infection with taeniasis. Generally, adverse events research in this area lacked methodological rigour. ConclusionsThe evidence generated from this review involving millions of people and millions of doses from different geographic locations with mostly mild-to-moderate and transient adverse events shows praziquantel is safe. However, given that the primary studies included in the review had design issues, including over 95% assessing adverse events over very short follow-up times, means serious long-term adverse events would have been missed. Also, the fact that some pregnant women who received praziquantel experienced serious events including miscarriages, foetal deaths and congenital anomalies calls for caution in the inclusion of pregnant women, particularly in their first trimester, in preventive chemotherapy campaigns. Additionally, the studies that reported severe visual adverse events raise safety concerns. Praziquantel is now offered repeatedly in endemic communities and the fact that in some settings up to 90% of those without infection could be offered the drug and the fact that there was no study that compared safety between infected and non-infected recipients, warrants further research. Evidence on safety in pregnant women and their foetuses, co-morbidity, polypharmacy, co-infection with taeniasis, as well as co-administration with drugs used in other preventive chemotherapy programmes, remain inconclusive and further research with long follow-up that should include blood chemistry analysis to provide additional evidence on long term safety, is warranted. This systematic review has exposed the lack of methodological rigour in adverse events studies and recommends future research should use robust and standardized design, methods, conduct and reporting.