Abstract
BACKGROUND Portal hypertension commonly accompanies advanced liver disease and often gives rise to life-threatening complications, including bleeding (haemorrhage) from oesophageal and gastrointestinal varices. Variceal bleeding commonly occurs in children with chronic liver disease or portal vein obstruction. Therefore, prevention is important. Primary prophylaxis of variceal bleeding in adults is the established standard of care because of the results of numerous randomised clinical trials demonstrating the efficacy of non-selective beta-blockers or endoscopic variceal ligation in decreasing the incidence of variceal bleeding. In children, band ligation, beta-blockers, and sclerotherapy have been proposed as alternatives for primary prophylaxis of oesophageal variceal bleeding. However, it is unknown whether those treatments are of benefit or harm when used for primary prophylaxis in children. OBJECTIVES To assess the benefits and harms of sclerotherapy compared with sham or no intervention for primary prophylaxis of oesophageal variceal bleeding in children with chronic liver disease or portal vein thrombosis. SEARCH METHODS We searched The Cochrane Hepato-Biliary Group Controlled Trials Register, CENTRAL, PubMed, Embase Elsevier, and two other registers in February 2019. We scrutinised the reference lists of the retrieved publications, and performed a manual search of the main paediatric gastroenterology and hepatology conference (NASPGHAN and ESPGHAN) abstracts from January 2008 to December 2018. We searched four registries for ongoing clinical trials. There were no language or document type restrictions. SELECTION CRITERIA We included randomised clinical trials irrespective of blinding, language, or publication status assessing sclerotherapy versus sham or no intervention for primary prophylaxis of oesophageal variceal bleeding in children with chronic liver disease or portal vein thrombosis. DATA COLLECTION AND ANALYSIS We used standard Cochrane methodology to perform this systematic review. We used the intention-to-treat principle to analyse outcome data, and GRADE to assess the certainty of evidence per outcome. MAIN RESULTS We found only one randomised clinical trial that fulfilled our inclusion criteria. The trial was at high risk of bias. The trial included 108 Brazilian children with median age of 4.3 years (range 11 months to 13 years). Fifty-six children were randomised to prophylactic sclerotherapy (ethanolamine oleate 2%) and 52 children to no intervention (control). Children were followed up for a median of 4.5 years. Eight children (six from the sclerotherapy group versus two from the control group) dropped out before the end of the trial. The follow-up was from 18 months to eight years. Mortality was 16% (9/56 children) in the sclerotherapy group versus 15% (8/52 children) in the control group (risk ration (RR) 1.04, 95% confidence interval (CI) 0.44 to 2.50; very low-certainty evidence). Upper gastrointestinal bleeding occurred in 21% (12/56) of the children in the sclerotherapy group versus 46% (24/52) in the control group (RR 0.46, 95% CI 0.26 to 0.83; very low-certainty evidence). There were more children with congestive hypertensive gastropathy in the sclerotherapy group than in the control group (14% (8/56) versus 6% (3/52); RR 2.48, 95% CI 0.69 to 8.84; very low-certainty evidence). The incidence of gastric varices was similar between the sclerotherapy group and the control group (11% (6/56) versus 10% (5/52); RR 1.11, 95% CI 0.36 to 3.43; very low-certainty evidence). The incidence of bleeding from gastric varices was higher in the sclerotherapy group than in the control group (4% (3/56) versus 0% (0/52); RR 6.51, 95% CI 0.34 to 123.06; very low-certainty evidence). The study did not assess health-related quality of life. Oesophageal variceal bleeding occurred in 5% (3/56) of the children in the sclerotherapy group versus 40% (21/52) of the children in the control group (RR 0.13, 95% CI 0.04 to 0.42; very low-certainty evidence). The most prevalent complications (defined as non-serious) were pain and fever after the procedure, which promptly resolved with analgesics. However, numerical data on the frequency of these adverse events and their occurrences in the two groups were lacking. No funding information was provided. We found no ongoing trials. AUTHORS' CONCLUSIONS The evidence, obtained from one randomised clinical trial at high risk of bias, is very uncertain on whether sclerotherapy has an influence on mortality and if it may decrease first upper gastrointestinal or oesophageal variceal bleeding in children. The evidence is very uncertain on whether sclerotherapy has an influence on congestive hypertensive gastropathy, incidence on gastric varices, and incidence of bleeding from gastric varices. Health-related quality of life was not measured. There were no serious events caused by sclerotherapy, and analysis of non-serious adverse events could not be performed due to lack of numerical data. The GRADE assessment of each outcome showed a very low-certainty evidence. The results of the trial need to be interpreted with caution. Larger randomised clinical trials, following the SPIRIT and CONSORT statements, assessing the benefits and harms of sclerotherapy compared with sham or no intervention for primary prophylaxis of oesophageal variceal bleeding in children with chronic liver disease or portal vein thrombosis are needed. The trials should include important clinical outcomes such as death, failure to control bleeding, and adverse events.
