Thrombolysis for stroke: clinical judgment at its apogee

Abstract

In the Correspondence section of today's Lancet, we publish a selection of letters challenging the report by Jonathan Emberson and colleagues on the effect of treatment delay, age, and stroke severity on the effects of thrombolysis with alteplase for acute ischaemic stroke. Emberson and colleagues concluded from their meta-analysis of 6756 patients that despite early increases in fatal intracranial haemorrhage, alteplase improves the overall likelihood of a good stroke outcome at 3–6 months when delivered within 4·5 h of the initial stroke symptoms, with earlier treatment increasing proportional benefit, irrespective of age or stroke severity. In view of the clinical heterogeneity across trials, should these trial data have been pooled, asks David Newman. Why did the authors not assess sex, question Valeria Caso and Patrik Michel, since studies have suggested that women could benefit from thrombolysis treatment more than men do. Peter Appelros and Andreas Terént dispute the suggestion that an increased risk of early death from intracranial haemorrhage in patients receiving alteplase is offset by an increase in disability-free survival. They comment that “for every 3·6 patients rescued from dependency at 6 months, there is an excess of one death in the acute phase”. For Mattias Brunström and Bo Carlberg, the benefit observed in patients receiving alteplase between 3 and 4·5 h after first symptoms “is a matter of interpretation”. They estimate that 19 patients need to receive thrombolysis for one to benefit. “For each patient saved to a good stroke outcome, another has a symptomatic intracranial bleeding.” This finding leads Brunström and Carlberg to conclude that thrombolysis between 3 and 4·5 h after stroke should be given “only when strongly indicated”. Will the correspondents be convinced by the authors’ reply and their attempt to remove uncertainties about the balance of benefit and risk when deciding to give alteplase to patients? Differences of opinion will surely remain. No matter how many primary trial data one has, and no matter how robust a meta-analysis is, ultimately the decision to give thrombolysis is one of clinical judgment, not absolute statistical certainty. Thrombolysis in acute strokeIn their meta-analysis of intravenous thrombolysis with alteplase for acute ischaemic stroke, Jonathan Emberson and colleagues (Nov 29, p 1929)1 concluded that the increased risk of early death from intracranial haemorrhage in patients taking alteplase was offset by an increase in disability-free survival. This interpretation gives the impression that the high level of mortality in the acute phase tends to diminish with time, but that is not the case. In the latest Cochrane analysis of thrombolysis for acute ischaemic stroke,2 thrombolysis gave no survival advantage between 7 days after thrombolysis and the end of follow-up. Full-Text PDF