Effect of tranexamic acid for acute spontaneous intracerebral haemorrhage: a systematic review and individual patient data meta-analysis.

Abstract

Spontaneous intracerebral haemorrhage (ICH) has high rates of mortality with no proven haemostatic treatment. We conducted the first systematic review and individual patient data meta-analysis (IPDMA) to assess the effect of tranexamic acid on outcomes in spontaneous ICH. We searched databases such as MEDLINE, EMBASE and CENTRAL for randomised controlled trials comparing intravenous tranexamic acid to placebo in adults with spontaneous ICH treated within 12 hours. The primary outcome was 90-day functional status (modified Rankin scale). Secondary outcomes included early mortality and haematoma expansion. One-stage random-effects analysis used generalised linear mixed models. Risk of bias was assessed using the Cochrane Risk of Bias (RoB) 2 tool. We screened 1131 records; nine trials (3194 participants) met inclusion and five trials (2860 participants; 90%) provided individual patient data. ROB was low in all trials. At 90 days, 53.2% of patients receiving tranexamic acid had a worse functional outcome compared with 53.6% in the placebo (adjusted common OR 0.93, 95% CI 0.81 to 1.07). Mortality at day 7 and haematoma expansion were significantly reduced with tranexamic acid (adjusted OR 0.70, 95% CI 0.51 to 0.95; and OR 0.81, 95% CI 0.68 to 0.97). No between-trial heterogeneity was observed. This systematic review and IPDMA of tranexamic acid for spontaneous ICH found no improvement in 90-day functional outcomes. However, small yet significant reductions in early mortality and haematoma expansion were observed. The early mortality benefit and favourable safety profile support further research into ultra-early treatment and as part of ICH care bundles for selected patient populations. CRD42017054978.