Time to Find Out if Mild Strokes Benefit From Thrombolysis

Susan Jeffrey

February 24, 2011

February 24, 2011 (Los Angeles, California) — New studies are raising the question of whether patients with mild strokes may still reap benefit from tissue plasminogen activator (tPA).

In 1 report, Pooja Khatri, MD, associate professor of neurology and director of acute stroke at the University of Cincinnati Academic Health Center, Ohio, and colleagues used the Greater Cincinnati/Northern Kentucky (GC/NK) epidemiologic study to estimate the number of stroke patients excluded from treatment only by the mildness of their stroke.

Dr. Pooja Khatri

If it were assumed that tPA is as effective in a mild as in a serious stroke, up to 3700 fewer patients would be disabled, potentially saving $200 million annually in disability costs, according to their calculations.

"I think it's time to have a big trial of this large population to really understand how those risks and benefits play out," Dr. Khatri said. A randomized trial in these patients with mild stroke is now being planned.

Their findings were presented here at the American Heart Association/American Stroke Association International Stroke Conference 2011.

Emergent Mild Stoke

The most common reason for not treating stroke patients who arrive early to the emergency department and are otherwise eligible for tPA is mild deficits, Dr. Khatri said. "This represents over half of strokes in the United States," she noted.

However, a growing literature shows that patients perceived to have mild strokes, defined as a National Institutes of Health Stroke Scale (NIHSS) score of 5 or less, actually have substantial disability at 90 days. Results of 2 previous cohort studies, for example, showed about 30% of patients with an NIHSS score of 5 or less at discharge had modified Rankin Scale (mRS) scores of 2 to 6 at 3 months, indicating measurable disability.

In this study, the researchers used the GC/NK cohort to estimate how many patients with mild stroke arrive at the hospital within the previously recommended 3.5-hour treatment window and to project, if tPA was assumed to be at least as useful in this group as in patients with more severe strokes, how many fewer patients would be disabled.

During 2005, 150 stroke patients in the region who were apparently eligible for tPA arrived within 3.5 hours of symptom onset but were not treated, "possibly appropriately, we really don't know," Dr. Khatri noted. The researchers then adjusted this incidence estimate by using US census data for age, race, and sex. They extrapolated the estimate to the 2010 US population and concluded that 43,180 mild strokes eligible for tPA occur annually.

Excluding patients with baseline disability (about 37%) and assuming a treatment effect of tPA or other effective treatments of 8% to 13%, they estimated that 2176 to 3761 fewer of these patients would be disabled each year after their mild stroke.

If disability was assumed to be moderate (that is, an mRS score of 2), and they "conservatively" estimate a lifetime cost of $100,000, then more than $200 million in disability expenditure could potentially be saved each year, the authors write.

Dr. Khatri and colleagues are seeking funding from the National Institutes of Health for a placebo-controlled trial of tPA in patients with mild stroke. The trial will be called Potential of rtPA for Ischemic Strokes with Mild Symptoms (PRISMS).

"We've got a group of collaborators in the United States and Canada, and we're planning a 2-part phase 3 trial, with about 1500 patients, where they're going to be randomized to tPA vs placebo if they have NIHSS of 5 or less, and don't appear clearly disabled to the physician," she said.

"So we're in the early stages, and we at this point have moved forward with planning the trial."

"Controversial" Exclusion Criteria

Among those collaborating on that trial is Steven R. Levine, MD, from University Hospital Downstate Stroke Center and the State University of New York Health Science Center, Brooklyn, who is a member of the scientific advisory board. Dr. Levine was also an investigator on the National Institute of Neurological Disorders and Stroke (NINDS) study that first established the efficacy of tPA in moderate to severe stroke.

Exclusion criteria for that original trial actually called for excluding only patients with either 4 specific stroke syndromes (isolated dysarthria, facial weakness, ataxia, or sensory loss), or "major symptoms that are rapidly improving by the time of randomization," with emphasis on "major" symptoms, he said in an interview with Medscape Medical News.

The intention, he said, was to exclude patients with "Lazarus" effect — that is, essentially full recovery after major symptoms — in whom benefit of tPA treatment was unlikely to outweigh the at-that-time undefined risk.

Yet RISS over the next 15 years has become one of the most common, subjective, poorly defined reasons for excluding otherwise eligible patients for tPA because it's left qualitative.

However, when the NINDS trial was published in The New England Journal of Medicine in 1995, the word "major" was omitted from the methods section during editing, Dr. Levine said.

"We excluded 13% of the screened patients because of RISS [rapidly improving stroke symptoms] in the trial," he said. "As a result, the package insert for Activase said that tPA was not recommended for rapidly improving stroke symptoms. Yet RISS over the next 15 years has become one of the most common, subjective, poorly defined reasons for excluding otherwise eligible patients for tPA because it's left qualitative."

In a paper presented here, Dr. Levine and colleagues used data from the original NINDS rt-PA Stroke Trial database to look at various definitions of RISS used in the overall cohort, and specifically at outcomes in each treatment arm for those with and those without RISS.

They calculated RISS as baseline NIHSS score minus the score at 2 hours, and used 3 definitions of RISS: improvement of 4 or more points, of 25% or more, or of 50% or more.

Depending on the definition, RISS occurred in 29%, 34%, and 17% of 624 participants in the trial. Independent of the definition, RISS was associated with better functional outcomes and a lower risk for death vs no RISS. But even among those with RISS, poor functional outcome was seen in 39% to 59% of patients, again depending on the definition.

After adjustment for initial NIHSS score and baseline mRS score, a significant favorable treatment effect for tPA was seen for patients with RISS by all 3 definitions.

Table. tPA Treatment Effect in RISS Patients (Using 1 of 3 Definitions)

Definition Odds Ratio (95% CI) P Value
> 4 NIHSS points 1.83 (1.22 - 2.75) ≤ .004
> 25% improvment 2.02 (1.35 - 3.02) ≤ .004
> 50% improvement 1.99 (1.33 - 2.97) ≤ .004

CI = confidence interval

"So it suggests that people with RISS, independent of how you define it, still generally have good outcomes more frequently when treated with tPA compared to placebo," Dr. Levine concluded. "Understanding which patients with improvement prior to a treatment decision benefit from tPA still remains an important challenge."

It is hoped that the PRISMS trial will provide some answers about this group of patients, he said. "The issue is ethically, can people have equipoise? If you have somebody with a minor stroke, there are a lot of places that still treat them and there are some that don't. And yet the [NINDS] trial included anybody with any measurable deficit on the stroke scale, even 1 point … and they still overall did better with drug than not."

At least if a trial shows they do better with [tPA], maybe we can get doctors to treat them, even though theoretically they should be.

Dr. Khatri points out that this subgroup was small — fewer than 50 patients. Dr. Levine said, "But she's pulling them out because people aren't treating them. At least if a trial shows they do better with [tPA], maybe we can get doctors to treat them, even though theoretically they should be."

Outcomes "Not Uniformly Good"

Finally, Eric E. Smith, MD, MPH, from the University of Calgary, Alberta, Canada, and colleagues used data from the American Heart Association's Get With The Guidelines-Stroke (GWTG-Stroke) registry to examine outcomes for those presenting with mild or rapidly improving stroke symptoms.

Again, they found that a high proportion of these patients who were excluded from tPA treatment solely on the basis of mild symptoms nevertheless had poor outcomes.

Among 93,517 patients arriving within 2 hours of symptom onset in 1 of 1092 GWTG-Stroke hospitals, 29,200 or 31.2% were excluded from tPA treatment because of mild or improving symptoms; the median initial NIHSS score was 2.

Despite their mild symptoms, "Discharge outcomes were not uniformly good," Dr. Smith and colleagues write: 28.3% could not be discharged home but went instead to inpatient rehabilitation or skilled-nursing facilities; 1% died. Almost 30% could not ambulate without assistance at hospital discharge.

"A controlled trial of reperfusion therapy in this population may be warranted," they conclude.

A Wake-Up Call

Dr. Steven Greenberg from Harvard Medical School moderated a press conference here at which Dr. Khatri's research was presented. He pointed out that to most patients, there is no such thing as a mild stroke.

"What I think this abstract does really well is act as a kind of a wake-up call to us that either from a cost standpoint as Dr. Khatri analyzed it, or from a personal day to day life standpoint, you could almost argue there's no such thing as a mild stroke," he told Medscape Medical News.

Less disabled patients may also be at lower risk for bleeding from tPA, he added.

So we're not weighing a lesser benefit against just as big a risk, we're weighing a lesser benefit against a lesser risk.

"We know that the amount of injured brain is proportionate to the risk of bleeding from tPA, so the smallest strokes have the lowest risk typically," Dr. Greenberg said.

"So we're not weighing a lesser benefit against just as big a risk, we're weighing a lesser benefit against a lesser risk. In the aggregate, the balance may say, yes we should be treating them just as much as more severe strokes, and for individuals or society, it may have just as big an impact."

Don't Be "Lulled to Sleep"

Philip B. Gorelick, MD, MPH, John S. Garvin Professor, head of the Department of Neurology and Rehabilitation, and director of stroke research at the University of Illinois College of Medicine, Chicago, pointed out that even a mild stroke or transient ischemic attack should prompt early and urgent diagnostic evaluation to determine the cause.

"The most important message is to define what's causing it and then apply the appropriate treatment, whether that's using an antiplatelet agent or an anticoagulant, going on to have a carotid endarterectomy or angioplasty and stent or clot retrieval as indicated, that's the key," Dr. Gorelick told Medscape Medical News. "So don't be lulled to sleep by a mild stroke presentation."

"Whether or not tPA is efficacious in this situation we don't know for sure — we need further validation studies on a large scale to prove or disprove that from a safety standpoint because the stroke damage is presumably small and the vascular injury is presumably more limited in these patients," he added. "It's likely that the treatment will be safe, but that remains to be proven."

The study by Khatri and colleagues funded by the National Institutes of Health/National Institute of Neurological Disorders and Stroke. Dr. Khatri reports that she has received a research grant from the National Institutes of Health for the IMS III trial and is receiving honoraria for American Academy of Neurology Acute Stroke Management Seminar, Grand Rounds Lectures. Dr. Levine reports no conflicts of interest. Dr. Smith reports he has served as a consultant/advisory board member for Genentech. Disclosures for coauthors appear in the abstracts.

American Heart Association/American Stroke Association International Stroke Conference 2011: Abstracts #75, 207, 145. Presented February 9, 10, 11, 2011.


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.