RESUMEN
The use of ibrutinib is hampered by major bleeding events and atrial fibrillation. Speculating whether randomized controlled trials might underestimate the risk of adverse events in clinical practice, we conducted a systematic review and meta-analysis studying patients treated in any setting and indication. We systematically searched the literature using MEDLINE and EMBASE databases for case series, cohort studies, or randomized controlled trials and retrieved all data in parallel. Proportions of patients with adverse events were pooled in relevant subgroups using the binominal distribution and Freeman-Tukey double arcsine transformation. Among 2'537 records screened, 85 were finally included, comprising 7'317 patients. Methodological quality according to the Newcastle-Ottawa Scale was rated as moderate to poor with regard to bleeding events and atrial fibrillation; 106 studies were excluded because of missing data at all. Reported events varied substantially between 0 % and 78 % (any bleedings), 0 % and 25 % (major bleedings), and 0 % and 38 % (new-onset atrial fibrillation). Pooled estimates were 28 % (95 % confidence interval 22 %, 34 %), 3 % (2 %, 4 %), and 8 % respectively (7 %, 10 %). The risk of events was higher in studies with an older population, high ibrutinib dosage, thrombocytopenia, antithrombotic treatment, and retrospective studies. In conclusions, reporting of bleeding events and atrial fibrillation varied substantially among studies. These observations, in combination with the estimates obtained, suggest a relevant risk in clinical practice.