Thrombolysis After 3 Hours in Stroke Needs Review?

March 20, 2015

Current guidelines on giving thrombolysis 3 to 4.5 hours after stroke onset is based on uncertain evidence, and there needs to be "urgent reconsideration" of the available data to guide policy decisions, a new review concludes.

The review, published in the British Medical Journal (BMJ) on March 17, is authored by a team led by Brian Alper, MD, EBSCO Information Services, Ipswich, Massachusetts.

"From analyzing all the available data, tPA [tissue plasminogen activator] after 3 hours for stroke patients may not be of any benefit but has a definite risk of fatal bleeding," Dr Alper told Medscape Medical News.

He pointed out that most guidelines on tPA in stroke recommend its use up to 4.5 hours after symptom onset, including those by the American Heart Association/American Stroke Association (AHA/ASA), which give tPA a Class 1 recommendation in this time window, but he says the data do not justify such recommendations.

"Other societies give weaker recommendations but the latest guidelines of the Canadian Association of Emergency Physicians give a weak recommendation against use of tPA after 3 hours. They are recognizing the uncertainty of the benefit and the greater consistency in harm," he added.

"Unless and until there are data showing unequivocal benefits to outweigh known harms, we believe that there should not be any strong recommendation or encouragement for the use of alteplase beyond 3 hours after stroke," Dr Alper and colleagues conclude in the BMJ paper.

To Medscape Medical News, he added: 'We want information in the guidelines to be based on data. We want everyone making policy on this to recheck and understand the published data, and ideally we would like the raw data from the key studies to be made available so that we and others can do deeper analyses."

But the lead investigators of the main thrombolysis trials disagree strongly with the review, saying it is too "simplistic," and the AHA/ASA said its stands by its guidelines.

Uncertainties in Key Results

Dr Alper and colleagues' concerns revolve around uncertainties in the results of both trials of tPA in the 3- to 4.5-hour time window in stroke.

They note that the ECASS-3 trial suggested benefit in this time period, but there were baseline differences between the two groups: namely that the placebo group had a higher rate of previous stroke. "In an analysis of only those patients who had not had a previous stroke, the results are no longer significant, and the overall difference between tPA and placebo was much smaller," Dr Alper said.

He added that the other trial — IST-3 — actually shows a trend toward harm with tPA given after 3 hours.

"This was reported to be nonsignificant but the researchers used 99% confidence intervals for this analysis, which was not in the statistical analysis plan," he points out. "It is traditional to use 95% confidence intervals for such analyses, and when this is done this result becomes significant for harm."

Dr Alper and colleagues also dispute conclusions reached by a Cochrane review and a meta-analysis of individual-patient data, both published in 2014 and suggesting benefit of tPA in the 3- to 4.5-hour window.

"Neither of these papers gave a clear accurate summary for tPA given between 3 and 4.5 hours which is supported by the data. The trouble is that the data reported in the trial publications does not always match the conclusions reached," Dr Alper claimed. "Then there are new reviews of the trials which summarize the conclusions, which can compound any misunderstanding or errors made in translation."

ECASS-3, IST-3 Investigators Respond

Professor Werner Hacke, MD, University of Heidlberg, Germany, who headed up the ECASS-3 trial, dismissed this "travesty of an article" as having been written by people "who have never seen a patient with an acute stroke and follow a personal crusade in order to withhold the only proven and working treatment for stroke."

IST-3 lead investigator, Peter Sandercock, MD, University of Edinburgh, United Kingdom, said Dr Alper and colleagues are taking too simplistic a view.

"This review is taking a very black-and-white view of the data but it doesn't work like that," he commented to Medscape Medical News. "It isn't the case that thrombolysis works up to 3 hours but not at 1 minute past 3 hours. There are many other factors that need to be taken into account."

Dr Sandercock acknowledged uncertainties around the data. "We are the first to recognize that the data for the 3- to 4.5-hour window is uncertain. We are not dealing with cardiology thrombolysis trials in 60,000 patients. We simply don't have data like that in stroke. There is certainly room for debate."

But he insists that he and his colleagues are doing everything possible to establish which stroke patients benefit from thrombolysis and that they are not trying to hide anything.

"The IST-3 trial was an investigator-initiated trial funded by government and scientific societies. There was no funding from the pharmaceutical industry. We do not have a bias. We are trying to maximize transparency. We are just trying to establish who does and does not benefit from thrombolysis. We are producing publications as fast as we can and we have had the data independently reviewed by the Oxford thrombolysis trialists — a very highly respected group. We are happy to share the data. If other groups want to analyze our data, they need to submit a formal request to us for specific analyses."

In response to Dr Alper's specific concerns about the IST-3 trial, Dr Sandercock pointed out that the primary endpoint was a modified Rankin scale score of 0 to 2, and this did show a trend toward harm with thrombolysis in the 3- to 4.5-hour window.

"But this time window was just one subgroup," he added, "and you have to be very careful about overinterpreting subgroup results, which is why we decided to use 99% confidence intervals for these analyses — so that people didn't make major decisions based on one subgroup analysis."

He further noted that if the same data are analyzed in an ordinal shift analysis, there is a trend toward benefit.

Dr Sandercock says the totality of the evidence does suggest benefit.

"The meta-analysis of all the data suggests that for an averagely severe stroke (National Institutes of Health Stroke Scale 11) for every 100 patients treated with thrombolysis within 3 hours, 10 more will have a good recovery. In the 3- to 4.5-hour window this benefit is reduced to 5."

But he stressed that other factors also need to be taken into account. "For example, if the patient has had a very mild stroke the benefit is less and so you might not want to expose them to the bleeding risk."

He added that the Stroke Thrombolysis Trialists' Collaborative Group is trying to clarify these factors and come up with an easy-to-understand risk-benefit calculator as to which patients may benefit from thrombolysis for stroke. "But this is a very difficult task."

AHA/ASA Defend Their Guidelines

The AHA/ASA also responded to the BMJ review paper. In a statement issued to Medscape Medical News, it said their guidelines were based on "all of the emerging science and the existing body of literature together, as synthesized by experts."

The AHA/ASA noted: "Our existing guidance on use of tPA between 3-4.5h after a stroke holds a Level of Evidence 'B' rating, which indicates further research might either support or reduce our confidence in the available evidence," adding that, "We are always eager to see new relevant data to inform our recommendations." 

The AHA/ASA also makes the following points:

  • The BMJ review does not specify the particular patient population that the AHA/ASA guidelines refer to. "Our guidance on use of tPA between 3-4.5h after stroke is limited to a specific patient population. This patient population includes patients under age 80, without extremely severe neurologic deficits, not taking anticoagulant drugs, not showing extensive brain injury on imaging, and without a history of both prior stroke and diabetes. These patients were the population of the ECASS 3 trial upon which we based the recommendation. The BMJ article relies on findings from IST 3 but does not make clear that the reported IST 3 findings do not specifically address this group of patients identified in our guidelines. Fully 55% of patients in IST 3 were over the age of 80, and are not relevant to the AHA/ASA recommendation."

  • The BMJ article discusses only mortality from hemorrhagic transformation, not all-cause mortality. The AHA/ASA guideline committee considered all-cause mortality for the risk-benefit analysis for use of tPA in the 3- to 4.5-hour window. The committee also considered the risks to patients if tPA is not administered within the 3- to 4.5-hour window, which include anticipated significant disability and even stroke-mediated death.

  • Despite limitations in the BMJ article, the AHA/ASA agrees with its call to make all trial data available for transparent and independent analysis. The current guideline's Level of Evidence B rating for the 3- to 4.5-hour window means that we believe that additional data and data analyses are desirable.

Dr Alper and his coauthors have disclosed no relevant financial relationships.

BMJ. Published online March 17, 2015. Article


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.