Skip to main content

03-07-2011 | Stroke | Article

Symptomatic ICH has limited impact on prognosis after stroke thrombolysis


Free abstract

MedWire News: Symptomatic intracerebral hemorrhage (ICH) contributes to poor outcomes in patients who undergo stroke thrombolysis, but is less important than factors such as age and stroke severity, say researchers.

Daniel Strbian (Helsinki University Central Hospital, Finland) and colleagues examined the 3-month functional outcomes of 985 stroke patients who underwent intravenous thrombolysis between 1995 and 2008.

As reported in the journal Neurology, the team looked at the effect of symptomatic ICH using three different definitions.

The Safe Implementation of Thrombolysis in Stroke Monitoring Study definition is parenchymal hemorrhage type 2 within 36 hours of thrombolysis combined with a deterioration of at least 4 points on the National Health Institutes of Health Stroke Scale (NIHSS) or death.

Symptomatic ICH by this definition occurred in 2.1% of the cohort, and increased patients' risk for poor outcomes (modified Rankin Scale 3-6) 1.7-fold and for death 4.8-fold, after accounting for confounders including age, stroke severity, and onset-to-treatment time.

The European Cooperative Acute Stroke Study (ECASS)-II definition is any intracranial hemorrhage combined with at least a 4-point deterioration in NIHSS score that was likely caused by the hemorrhage.

This occurred in 7.0% of the cohort, and was associated with a 1.6-fold and 3.8-fold increase in the risk for poor outcomes and death, respectively.

The National Institute of Neurological Disorders and Stroke trial definition is the appearance of hemorrhage on imaging at 24 hours or 7-10 days after stroke combined with any decline in neurologic status.

This occurred in 9.4% of patients, and raised their risk for poor outcomes 1.6-fold and for dying 3.4-fold.

All three definitions improved prediction of poor outcomes and death when added to a model including age, stroke severity, onset-to-treatment time, baseline glucose levels, prior disability, and the presence of a hyperdense artery sign or early infarct signs on baseline imaging.

The best result was achieved with the ECASS-II definition, which improved prediction of very poor outcomes (modified Rankin Scale 5-6) by 11.3% and of death by 9.3%, according to integrated discrimination improvement analysis.

However, according to area under the receiver operating characteristic curve analysis, the original model successfully distinguished between over 80% of patients with and without these outcomes. This indicates that many patients with symptomatic ICH would have had poor outcomes anyway, say the researchers.

In an accompanying editorial, Andrew Southerland (University of Virginia Health Systems, Charlottesville, USA) and colleagues noted that "decision support tools regarding the thrombolysis risk-to-benefit ratio in individual patients could be valuable."

They pointed out that fear of symptomatic ICH is the primary reason given for withholding thrombolysis. "If we could reassure patients and caregivers of a substantial net benefit in individual cases (as opposed to population estimates) then perhaps the collective comfort with [recombinant tissue plasminogen activator] in our physician community might follow," they said.

But they cautioned: "Only after rigorous external validation and consistent performance at a high level could such a tool be introduced into clinical decision-making for acute stroke patients."

MedWire ( is an independent clinical news service provided by Springer Healthcare Limited. © Springer Healthcare Ltd; 2011

By Eleanor McDermid