Back in 2011, there was an article in Annals of Emergency Medicine discussing what a fantastic job we were doing in diagnosing stroke and avoiding administering tPA to “stroke mimics”. They reported a rate of 1.4% administration to stroke mimics – none of whom had bleeds. The problem I pointed out, both on my blog and in a response letter to Annals, was that the authors invented a new category called “neuroimaging negative” acute stroke – which was probably actually all stroke mimics. This would have changed the rate of tPA administration to stroke mimics from 1.4% to 29.3%. The authors, having financial conflict of interest with the manufacturers of tPA, disagreed.
This study, part of the “Lesion Evolution in Stroke and Ischemia On Neuroimaging” project, evaluated the progression of lesions on MRI following tPA administration. These authors found 231 patients with acute stroke who were initially screened by MRI prior to tPA administration and had evidence of infarction on diffusion weighted imaging. They found that, following tPA administration, only 2 patients had resolution of an MRI DWI lesion. They therefore conclude that “Patients with a stroke are unlikely to have complete DWI lesion reversal within 24 hours after IV tPA treatment,” and patients with no DWI lesion following tPA administration should be considered to have a diagnosis other than acute stroke.
Thus, this confirms my conclusion that the 27.9% of patients from the prior study with “neuroimaging negative” acute stroke ought to universally be considered to have had a diagnosis other than acute stroke. The reality is that we are likely treating an ever-greater number of acute ischemic strokes – and further efforts to push Emergency Physicians to treat additional patients more quickly are certainly going to expose additional patients to avoidable harms.
“Negative Diffusion-Weighted Imaging After Intravenous Tissue-Type Plasminogen Activator Is Rare and Unlikely to Indicate Averted Infarction”
http://www.ncbi.nlm.nih.gov/pubmed/23572476
3 thoughts on ““Neuroimaging Negative” Strokes Are A Lie”
Comments are closed.
Ryan
I must say you do a great job with your blog and wish you continued success.
This is one of some of the most overlooked studies on TPA. There was another study in Annals that talked about this. I will try to find it. What it says is that if you have a stroke mimic its probably safe to get TPA.
If you are truly having a stroke it much more "unsafe."
The way I understand it is this…..If you have a stroke mimic than your brain tissue and "brain blood vessels" are probably pretty sturdy/intact and much less likely to bleed.
If you have a stroke than you are more likley to have dead necrotic brain tissue and injured blood vessels and hence more likely to bleed. Seems pretty obvious here, as I was taught in med school that brain tissue dies within minutes.
Just seems intuitive that giving a clot busting drug into damaged tissue and blood vessels is dangerous. If you give a clot busting drug to a stroke mimic they should be much less likely to bleed.
To extrapolate it even further if you give a TPA to a stroke mimic they are more likely by definition of it being a mimic more likely to have a neg neuroimaging and do better……you could also falsely include them as a true CVA that got better with TPA.
My conclusion is this (I hope you can sense my sarcasm).
If you are having a stroke mimic you are more likely to not bleed with TPA and do better.
If you are having a true stroke you are more likely to bleed and probably do worse.
I have followed the pro and neg TPA arguments for a while and still am not convinced it provides benefit.
Steve Goodfriend
Hi Steve –
Thanks for the well-wishes; much appreciated.
Folks who receive tPA but do not have an acute stroke are certainly less likely to experience intracranial hemorrhage; we know this from the extensive literature regarding thrombolytics for myocardial infarction and massive/submassive PE. When the brain isn't undergoing coagulative necrosis, it's a happier brain.
The key from this new stroke article is that they blow up the concept of "neuroimaging negative stroke." In addition to this Annals article that classifies 27.9% of their patient as this "averted stroke", there are two other studies that looked at this:
Chernyshev et al. "Safety of tPA in stroke mimics and neuroimaging-negative cerebral ischemia"
Uchino et al. "Transient Ischemic Attack after Tissue Plasminogen Activator: Aborted Stroke or Unnecessary Stroke Therapy?"
The Chernyshev found 21% patients to not have an infarct on follow-up imaging, and the Uchino study found 16% not to have an infarct on follow-up. These rates – 16%, 21%, 29% – are likely the true incidence of tPA administration to TIAs and other stroke mimics, not the single digit numbers being touted based on this concept of "averted stroke" and "neuroimaging-negative cerebral ischemia".
There is a problem. you seem to rely on the presence of diffusion weighted imaging to confirm stroke at the hyperacute phase. DWI isn't infallible, particulary early on in the posterior fossa. Of the 1167 patients who were included in this study only 267 were included due to positive lesion. we don't know how many had a final diagnosis of stroke in the remaining group who were treated.
if you put that into perspective, due the lack of availability of pre-treatment mri how would you conclude that these patients never had an ischemic event just based on dwi negativity at follow up? additionaly they excluded people with PWI abnormality with negative dwi pretreatment. there was no advanced vascular imaging to suggest presence of clots/occlusion in this subgroup. unless we have evidence of all of these our practice shouldn't change and a final diagnosis of stroke should remain clinical till we have bullet proof tests for stroke, averted or otherwise.