Does the widespread use of the thrombolytic tissue plasminogen activator (t-PA) produce more benefit or harm in patients who experience an acute stroke?

Bob: If you are like me, you have been bombarded with television, radio, and medical advertisements stressing the need for patients with symptoms of acute stroke to seek medical attention for immediate therapy for their so-called “brain attack.” The treatment they are referring to is the administration of t-PA within three hours of the onset of an acute ischemic stroke. Because t-PA is a new therapy for acute stroke, this prompts the question: What has been the nationwide experience with it?

What does this article say?

Bob: This study reviewed the available data on 248,964 patients admitted to U.S. emergency departments from 1999 to 2002 with the diagnosis of an acute ischemic stroke. Approximately 1 percent (2,594 patients) received t-PA; their in-hospital mortality rate was 11.4 percent. The in-hospital mortality rate for patients who did not receive t-PA was 6.8 percent. Simply stated, nearly 5 percent more patients receiving t-PA died compared with those who did not receive this therapy for their stroke!

Should we believe this study?

Bob: This study is a retrospective, observational study that was not randomized. Therefore, you have to ask if the patients who received t-PA had more severe strokes than those who did not. If so, the increased mortality rate with t-PA would be expected. However, observational studies are useful snapshots of the effect of a drug or therapy in the community setting. When studies of new drugs or interventions are first done, they are performed in controlled environments, often with clear protocols and personnel present to make certain there are no deviations. This determines the efficacy of a drug or intervention. However, what happens (i.e., the effect) when a drug or intervention is rolled out into the community setting, where there is no researcher or oversight committee to check on what is being done? Studies such as this one help answer that question.

Andrea: It's a subtle but important point. The efficacy of a drug in the research setting may be very different from the effectiveness in the community. I remember the difference by word length: effectiveness has more letters and refers to use in a large population.

Mark: One possible explanation for the increased mortality rate with t-PA is the lack of compliance with the intricate protocol required with this therapy. In a review of 63 patients in which t-PA was given at 16 Con-necticut hospitals, the protocol was violated in two thirds of the patients. What was the outcome? One of every four patients (25 percent) who received t-PA died.¹

Andrea: Or perhaps another reason for the increased mortality rate with t-PA is the inappropriate administration of the drug to patients who had a “stroke mimic.” Even facilities with fancy stroke teams will misdiagnose a patient as having an acute stroke when it is really something else. This was first pointed out in a 1995 study that reviewed the final outcomes of 411 patients who were diagnosed with an acute stroke in the emergency department by a stroke team. The study showed that, at discharge, nearly one in five patients had a diagnosis other than acute stroke (e.g., seizure, systemic infection, brain tumor, metabolic disturbance).²

Bob: Whatever the reason, I believe this increased mortality rate is a real phenomenon. A German stroke registry noted that, in patients given t-PA for acute ischemic stroke, the in-hospital mortality rate was 11.7 percent, compared with 4.5 percent in patients not given t-PA.³

Mark: We often talk about the number needed to treat (NNT) and number needed to harm (NNH). We have to put in perspective what the harm can be. Not to sound understated, but death is a major harm—we are not talking about diarrhea or dyspepsia here. In fairness, what is the benefit that t-PA provides to patients who have had an acute ischemic stroke (i.e., the NNT)?

Bob: Only one randomized trial has ever demonstrated any benefit. The National Institute of Neurological Disorders and Stroke (NINDS) study randomized 333 patients with acute ischemic stroke; one half received t-PA and one half received placebo.⁴ At three months, 50 percent of patients who received t-PA had minimal or no disability compared with 38 percent who received placebo. This 12 percent difference translates into an NNT of eight. In the NINDS trial, there was no increase in mortality rates, but the rate of intracerebral hemorrhage was 6.4 percent in patients receiving t-PA and 0.6 percent in patients receiving placebo (NNH = 17).⁴

What should the family physician do?

Bob: When I am faced with a patient who is having an acute ischemic stroke and who meets criteria for t-PA therapy, I try my best to spell out the risks and benefits to the patient and the family. This includes telling them that one in eight patients are helped at three months, one in 17 are harmed, and that there has been a documented increase in death in patients who have received t-PA therapy outside of research trials. When I am asked what I would do if it was my own family member, I answer honestly: I would not give this therapy.

Andrea: What is fascinating is the competition between hospitals to “capture” stroke patients by publicizing themselves as stroke centers, the hallmark of which is the ability to give t-PA therapy.

Mark: At my institution, we religiously follow the NINDS protocol. We have a stroke team that includes an academic neurologist, and we have neurosurgery and interventional radiology on site 24 hours a day. If you do choose to use t-PA for stroke, your institution must be prepared to handle any adverse outcomes.