Risk of intracranial haemorrhage with bolus versus infusion thrombolytic therapy: a meta-analysis

Background Although thrombolytic therapy given by bolus injection seems to be as effective as infusion over 60–90 min, no single trial has been adequately powered to detect clinically important safety differences between the two strategies. We did a meta-analysis to find out whether bolus administration of thrombolytics is associated with an increased frequency of intracranial haemorrhage. Methods We identified randomised trials comparing bolus with infusion thrombolytic therapy by a search of electronic databases, reference lists, and conference proceedings. Odds ratios for primary intracranial haemorrhage, nonhaemorrhagic stroke, mortality, and reinfarction were calculated for each trial and were combined in a fixed-effects model. Findings Seven trials, involving a total of 103 972 patients, met our inclusion criteria. Bolus treatment was associated with an increased risk of intracranial haemorrhage compared with infusion (0·8 vs 0·6%; odds ratio 1·25 [95% Cl 1·08–1·45]; p=0·003). The increased risk was most striking in trials comparing bolus with infusion administration of the same agent (1·75 [1·32–2·33], p=0·0001), but was also evident in trials comparing a newer-generation bolus agent with standard infusion therapy (1·25 [1·03–1·50]; p=0·02). The rates of non-haemorrhagic stroke (0·94 [0·81–1·09]), 30-day mortality (1·01 [0·97–1·06]), or reinfarction (1·04 [0·97–1·11]) did not differ between the two strategies. Interpretation The increased risk of bolus thrombolytic treatment seems to be primarily associated with the method of administration rather than properties of the agents. Although the increased risk of intracranial haemorrhage is small, physicians should balance this risk against the putative benefits of easier administration with no difference in mortality or reinfarction.