It took trials and studies on 60,000 heart attack patients to work out that clot-busters prevent deaths in heart attacks, but they only worked on one type of heart attack (which is only about five percent of all heart attacks). We took the time to get it right, refining study after study, until we were sure that introducing a potentially dangerous drug had benefits that outweighed the harm.
Every type of clot-buster that was studied worked, and the earlier it was given, the better. Interestingly, the one that caused the most brain bleeding was tPA.
Studies have been done on clot-busters in stroke. Strokes are also caused by a blockage to a blood vessel in the brain, but there are important differences between strokes and heart attacks:
* The cause of the blockage is never worked out about 30 percent of the time.
* Two very different types of blood clots cause strokes: new blood clots (like in heart attacks) and old blood clots called embolic clots (e.g., an old blood clot that has formed in the heart and travelled to the brain where it blocked a blood vessel).
* This difference between new and old blood clots is very important because clot-busters do not work on old blood clots.
* How many strokes are caused by new blood clots? No one knows for sure because we can't work out what kind of clots (or other causes) are causing the stroke. The best guess is that it is a little less than half.
“It is important to note that the efficacy of thrombolytic drugs depends on the age of the clot. Older clots have more fibrin cross-linking and are more compacted; therefore, older clots are more difficult to dissolve. For treating acute myocardial infarction, the thrombolytic drugs should ideally be given within the first 2 hours. Beyond that time, the efficacy diminishes and higher doses are generally required to achieve desired lysis.” (Cardiovascular Pharmacology Concepts; http://bit.ly/1a5xt1D.)
Twelve important studies have been done: 10 were negative (they showed that clot-busting did not work in stroke), and four of these had to be stopped early because of harm (they were killing people). They have only studied this in about a sixth of the number of stroke patients compared with the heart attack studies. This is a completely different situation from the heart attack studies, where it worked in every study and with every clot-busting drug they tried.
The two positive stroke studies (i.e., they suggest a statistical benefit) have been widely criticized in the medical literature because of their scientific flaws. One of these looked at giving tPA within three hours of the stroke. Interestingly, this study proved that there was no benefit in the first 24 hours, which again is different from the heart attack studies. The other study looked at giving tPA from 3 to 4.5 hours after the stroke.
This second study also contains some major errors. The biggest flaw was an imbalance in baseline stroke severity. The placebo group (those who didn't get tPA) were sicker, with more severe stroke; the tPA group was not as sick and had much more mild strokes. Everyone understands that people with milder strokes do better than those with more severe stroke. Obviously, the study favored tPA because the tPA group had milder strokes, which will do better anyway.
This study was re-analyzed later. If tPA really worked to improve the outcomes in sudden stroke, then there should be a much larger improvement in outcome compared with the placebo group. But there was no difference.
The studies show that tPA causes brain bleeding in about six percent of patients. Brain bleeding is the other type of stroke. It is strange that some people would advocate a treatment for stroke that can cause a (worse) stroke. Brain bleeding is severe, and almost half the people die if this happens.
Why does brain bleeding happen with a clot-buster? Doctors think of clot-busters as being like Drano, the plumbing product used to unblock pipes. You have a damaged blood vessel inside a damaged part of the brain in strokes. It's no surprise that putting “Drano” in there sometimes dissolves through the blood vessel and bleeds into the brain.
The largest study on clot-busting in stroke was published in 2012. This again was a negative study that showed tPA did not work, and it confirmed the increase in early deaths with tPA. Interestingly, the time to give the drug from the stroke studies is completely different from the heart attack studies. After studying 60,000 heart attack patients, every study showed that the earlier the clot-buster was given, the better. The majority of the stroke studies do not show this effect. The reason for this is unclear, but may be because the patients who get to the hospital earlier are somehow different from the ones who arrive later.
One must always remember, however, that once the blood vessel is blocked, the brain cells will be dead in about three minutes. Giving a clot-buster later shouldn't have any effect, and this may be what we're seeing. It's also worth remembering that heart muscle cells and brain cells are very different.
Researchers have used meta-analysis to “prove” that clot-busting works. The simple summary of the problem with this is summed up by the phrase: garbage in, garbage out. Badly done studies when put into a meta-analysis do not magically become good studies. There is at least one review of this that concludes “there is no consistent or proven benefit” to clot-busters in stroke. (The NNT; http://bit.ly/NNTthrombolytics.)
What is the problem here? Why is there controversy in the medical world? Shouldn't it be black and white? Either tPA works, or it does not. Easy.
Sadly, it is not so easy. Statistical analysis of any medical treatment is rarely black and white. It is about the interpretation of the research. This is where the controversy has occurred here.
So why is tPA for stroke being pushed? Because we really want it to work.
Unfortunately, many of those pushing this treatment have conflicts of interest with the manufacturer of the drug. (BMJ 2013;346:f3830, http://bit.ly/17rQvju; BMJ 2013;347:f5535, http://bit.ly/1apXsjG.)
How do we resolve this problem? We think that patients and society in general expect that an offered medical treatment will have passed the test of science. The test of science requires elimination of bias, healthy scientific debate, and replication to see if a study that suggested a statistical benefit can produce that result again, which would make it more believable. Time and time again in medical research, we find that when larger more decisive studies are done, the initial exciting result becomes a disappointing one: it doesn't work. This replication has not happened for stroke, but needs to if we want to maintain our trust in medicine.
Has anyone else raised concerns about this? Recently, the British Medical Journal published a debate and poll on this very subject. (BMJ 4 September 2013; http://bit.ly/172AnTK.) Other experts have also expressed similar concerns. These resources from SMARTEM.org offer further reading: “Delusions of Benefit in the International Stroke Trial” (http://bit.ly/1bEhUwt) and “The Guideline, The Science, and The Gap” (http://bit.ly/1aTVM4u).
Is there anything else that can be done if I have a stroke? We know that having your care in a stroke unit is much more powerful than any drug. And taking part in stroke research will help us answer these questions.© 2014 by Lippincott Williams & Wilkins