Investigators reported that there was a four-fold increase in the use of tPA for ischemic stroke at California Pacific Medical Center after applying less rigid eligibility criteria.
TORONTO—Current NINDS eligibility criteria for treatment with recombinant tissue plasminogen activator (tPA) are too strict, according to researchers who found that the use of more relaxed criteria at their institution over a four-year period substantially increased the number of patients undergoing treatment — with outcomes as good as or better than those in clinical trials.
The researchers re-evaluated many common IV tPA treatment criteria and reduced eligibility criteria to include patients who had disabling neurological symptoms caused by acute cerebral ischemia, with time from symptom onset of less than 3 to 4.5 hours.
Use of the new criteria resulted in about a four-fold increase in the use of tPA at California Pacific Medical Center (CPMC), reported Jack C. Rose, MD, associate director of Neurocritical Care at CPMC in San Francisco. Importantly, this occurred without an increase in bleeding rates, he added.
He presented results of the Simplified Management of Acute Stroke Using Revised Treatment (SMART) criteria on behalf of lead researcher David Tong, MD, medical director of CPMC Comprehensive Stroke Care Center, at the AAN annual meeting here in April.
The NINDS guidelines are largely based on the results of the NINDS tPA study, published in 1995 in the New England Journal of Medicine (NEJM), which led to the approval of the drug for the treatment of acute stroke in 1996.
“The premise of SMART is that most if not all of the original very stringent NINDS IV tPA clinical trial treatment criteria for acute stroke are not evidence-based and therefore are probably unnecessary,” Dr. Rose said.
“They were created a priori, without any data on whether they were appropriate, and in fact are highly conservative. In addition, research over the past 15 years has found that many additional patients, such as those ages 80 and older or who have mild symptoms, can be treated safely and effectively,” he said.
Despite this, most physicians still use the same criteria as the original NINDS trial, Dr. Rose said. In fact, some centers use even stricter criteria in terms of the age cutoff, he said. As a result, tPA use remains very low, at 1 to 3 percent, according to recent reports, he added.
For the study, CPMC researchers re-evaluated the common IV tPA treatment criteria and eliminated those that were not evidence based.
“We were left with essentially only one inclusion criterion for treatment: having an ischemic stroke within 3 to 4.5 hours. We used the less than three-hour time window until the ECASS 3 results [Eastern Cooperative Acute Stroke Study 3, published in 2008 in the NEJM], established a 4.5-hour window [as being safe] in 2008,” Dr. Rose said.
The only exclusion criterion was symptomatic intracranial hemorrhage. All other patients were eligible for treatment.
“Asymptomatic hemorrhage such as chronic subdural hematoma, for example, would not be a contraindication to treatment,” Dr. Rose said. That eliminated about a dozen commonly used exclusion criteria, he said. (See “Excluding the Exclusions.”)
“Notable tPA eligible patients using SMART criteria [who would traditionally be excluded] are those ages 80 and older,” Dr. Rose said, adding that 13 percent of the center's patients are over age 90.
Also, patients with mild stroke are included; “as long as the stroke is disabling, there is no NIHSS (National Institutes of Health Stroke Scale) score cutoff,” he said.
“Additionally, we treat patients with severe stroke; improving stroke, as long as it is still disabling; recent stroke or trauma, and seizures,” Dr. Rose said.
For the study, the researchers applied the new criteria to a prospective consecutive case series of 195 patients who came to their center from July 2006 to Dec. 2009. This represents 25 percent of all ischemic stroke patients evaluated at the center regardless of time window, and 87 percent treated within 4.5 hours of symptom onset, Dr. Rose said.
Of the total, 138 (71 percent) patients received IV tPA alone and are included in this analysis.
The outcomes of the SMART protocol patients were equal to or better than those reported in the NINDS tPA study, the SITS MOST (Safe Implementation of Thrombolysis in Stroke-Monitoring Study), published in The Lancet in 2007, and ECASS-3, Dr. Rose said.
Results showed that 59 percent of patients treated using the SMART criteria had good outcomes with no residual deficit, defined as modified Rankin scores of 0 to 1. In contrast, an average 46 percent of treated patients in the tPA clinical trials and an average 33 percent in the placebo arms had good outcomes.
The symptomatic hemorrhage rate in SMART patients was 1.5 percent, which Dr. Rose said was similar to the 2 percent hemorrhage rate in the placebo arms of the clinical trials, he said.
The mortality rate in the SMART group was low at 7 percent, compared with an average 18 percent in the placebo arms of the clinical trials.
Had NINDS criteria been used, 42 percent of SMART patients would have been excluded due to age 80 or older, and 24 percent would have been excluded due to mild stroke, with an NIHSS score of less than 5, he said.
Overall, 89 percent of SMART patients had at least one common exclusion to IV tPA treatment, and the average number of exclusions was 1.4 per patient.
The thrombolysis rate at CPMC over the course of the study was 19 percent to 30 percent, compared with national rates of 5 percent or less during that period and 7 percent at CPMC in 2004, before the SMART criteria were established.
The bottom line, Dr. Rose, said, is that under SMART criteria, about 20 to 30 percent of all patients can receive thrombolysis.
“Guideline makers should reassess the old NINDS guidelines, and at least acknowledge that many excluded patients are potentially treatable with good safety and outcomes,” Dr. Rose said.
Other researchers congratulated the CPMC team on its positive experience, but expressed concern about the risk of intracerebral hemorrhage.
“This study is very important,” commented James C. Grotta, MD, chairman of neurology at the University of Texas Medical School at Houston and director of the Stroke Program at Memorial-Hermann Texas Medical Center.
“It demonstrates that we have to start thinking outside box and not be so rigid [about tPA criteria],” said Dr. Grotta, who was a member of the NINDS tPA Stroke Study Group.
He said that tPA is the only drug in neurology where inclusion and exclusion criteria from a clinical trial are included in the package insert.
Dr. Grotta said after tPA was approved, a few early reports of complications, particularly bleeding, when tPA criteria were not followed led to a “state of paranoia” on the part of neurologists and especially emergency medicine specialists.
There is no question that when tPA is used incorrectly, it is dangerous. So if you're just starting out, you need to stick to the rigid criteria until you are familiar with the drug — how to give it, its risks, and so on.
“But once you gain expertise, a lot of the criteria don't make sense,” Dr. Grotta said. At Memorial-Hermann, “where we arguably use more tPA than anywhere, we've been amending the criteria for years, and with good success,” he continued.
“Each year as we give more and more tPA, our hemorrhage rates actually go down,” he said, noting it was 2 percent last year.
As with the SMART researchers, Dr. Grotta said his team considers some of the exclusion and inclusion criteria “rather arbitrary.”For example, there should not be an age cutoff, he said. Despite what many people think, the NINDS trial included patients over 80, he said.
Routinely excluding patients with minor or rapidly improving stroke is the “biggest” mistake, he said. Over 50 percent of these patients will end up disabled if you don't treat. There is no evidence of worse outcomes in experienced centers where they treat patients over 70, patients with mild or improving stroke, or within an extended time window. Additionally, there is no reason why patients who have had a myocardial infarct in the prior six weeks should routinely be excluded from tPA therapy, he said. Similarly, “as long as you are sure a patient has had a stroke, seizure is not a contraindication,” Dr. Grotta said.
But he disagreed with the SMART team regarding the inclusion of patients on anticoagulation. “If a patient is on anticoagulation, we tend to be conservative. If their INR [International Normalized Ratio] is elevated, we don't treat. I think you're asking for disaster,” he said.
His team also does not treat patients with uncontrolled blood pressure. “You must get systolic BP down to 180 [mm Hg] and keep it there,” Dr. Grotta said.
Ralph L. Sacco, MD, chairman of neurology at the University of Miami and president-elect of the American Heart Association, said: “There are obvious advantages to simplifying inclusion and exclusion criteria for the use of tPA in acute stroke.”
But other research had shown that even though “you can get good outcomes, there is an increase in bleeding,” he said.
Dr. Sacco noted that this was an experience at one institution and there may have been some application of the well ingrained NINDS inclusion/exclusion criteria by the doctors there that wasn't captured in the study.
“There would have to be more testing [of the SMART criteria] in community hospitals before one would want to deviate from the current recommendations,” Dr. Sacco said. •
• Wahlgren N, Ahmed N, Vanhooren G, et al, for the SITS-MOST investigators. Thrombolysis with alteplase for acute ischaemic stroke in the Safe Implementation of Thrombolysis in Stroke-Monitoring Study (SITS-MOST): An observational study. 2007;369(9558):275–282.