Antiemetic medications for preventing chemotherapy-induced nausea and vomiting in children: a systematic review and Bayesian network meta-analysis

Abstract

Purpose Children continue to experience chemotherapy-induced nausea and vomiting (CINV), despite effective antiemetic medications. Recommendations in clinical practice guidelines are underpinned by narrative syntheses and meta-analyses that compare only two treatments. This means not all antiemetics have been compared to one another, and estimates remain imprecise. We apply network meta-analysis (NMA) to overcome these limitations by comparing multiple treatments simultaneously. Methods A systematic review identified and critically appraised RCTs comparing antiemetics recommended and licensed for the prevention of CINV in children. Bayesian NMA compared and ranked antiemetic effectiveness for the outcomes complete (CR) and partial response (PR) in the acute, delayed, and overall phases, nausea, and decreased food intake. Antiemetics given with and without dexamethasone were compared in separate networks as their underlying populations differed. Results Sixteen RCTs (3115 patients receiving moderately (MEC) or highly emetogenic chemotherapy (HEC)) were included. When given with dexamethasone, NK1 antagonists with ondansetron ranked highest for CR and PR in the acute and overall phases, PR in the delayed phase, and decreased food intake. Post hoc analysis shows further a benefit of adding olanzapine to regimens of aprepitant and ondansetron. Ondansetron ranked lower than palonosetron, for CR in the delayed and overall phases, and ondansetron was less effective than palonosetron for nausea prevention. Rankings for other regimens, including those given without dexamethasone, were uncertain or inconsistent across outcomes. Conclusions Our findings serve to support the current recommendations of olanzapine (when given with aprepitant and ondansetron) and NK1 antagonists’ regimens receiving HEC, but note that evidence of a significant difference in relative benefit, between patients receiving MEC and HEC, does not yet exist. Recommendations for palonosetron as the preferred 5HT3 antagonists may be extended, particularly, to those who are at high risk of nausea.