Meta-analysis of neurokinin-1 receptor antagonists (NK-1 RA) for chemotherapy-induced nausea and vomiting (CINV)

Abstract

8047 Background: CINV remains a problem for many patients treated with highly emetogenic chemotherapy. NK1 RAs are antiemetics that could improve CINV prophylaxis. We evaluated the effectiveness of NK1 RAs to completely control acute and delayed CINV. Methods: We used an optimized search strategy to identify RCTs between 1990–2003 comparing NK1 RAs with control therapy. Three reviewers independently collected data trials, resolving disagreements by consensus. The endpoints were complete response (total freedom from CINV, no rescue drugs) measured as risk ratios (RR), and numbers-needed-to-treat for benefit (NNTB). Acute and delayed phases were analyzed separately. Sensitivity analyses included heterogeneity and publication bias tests. Results: The search yielded seven RCTs (n = 1568 evaluable patients). All trials had good methodological quality, only one trial describing checks to ensure blinding. All patients were chemotherapy-nave and received cisplatin 50–100 mg/m2. Pharmaceutical companies sponsored all trials. For acute CINV, NK1 RAs used alone or with standard therapy were not better than control; for delayed CINV, the treatment effect favored NK1 RAs over control (Table). There was significant heterogeneity in acute CINV due to sample size. Delayed CINV was homogeneous, and outlying studies did not adversely affect results. We found no evidence of publication bias. Conclusions: NK1 RAs added to standard therapy significantly increased complete response rates from delayed CINV, with no significant effect on acute CINV. Results for acute CINV need cautious interpretation because of between-trial inconsistency. Results in delayed CINV are more robust, but we cannot exclude sponsor bias. Future RCTs funded independently may determine to what extent results of this meta-analysis were sponsor-biased. No significant financial relationships to disclose.