It isn’t news to anyone in the Emergency Medicine community that tPA isn’t as effective as its efficacy trials suggested, and its overuse is driven by “quality” measures and medicolegal concerns more than any true belief in its usefulness. However, it remains rare in the Neurology literature to challenge the primacy of tPA – it is much more frequent to see articles promoting and/or defending its expanded use.
This small retrospective series looks at a registry of stroke patients eligible for alteplase who received a CT perfusion study as part of their initial evaluation. As criteria for review, the CT perfusion lesion needed to be <15mL in calculated volume. Their final cohort included 366 patients with mostly mild-to-moderate strokes (NIHSS median 8 in each), and a little over half were treated with alteplase, while the remainder were not. As a retrospective and confounded study, the level of evidence is weak, but the untreated population had significantly better outcomes (mRS 0-1 in 57% vs. 69%), and avoided such complications as parenchymal hemorrhage.
The authors conclude:
“we suggest that neither CTA nor standard clinical/NCCT assessment can appropriately define a relatively large sub-group of patients who are clinically eligible for alteplase, yet appear to have no benefit from treatment.”
Yes, if the volume of acutely injured tissue is quite small, the potential benefit of any therapy has an obvious ceiling – even before considering the viability of the affected tissue or the potential effectiveness of reperfusion. But the key point here is one I’ve made, most recently at #smaccDUB: we can better individualize care, and avoid costs and risks, with more information.
Thanks to Robert Goulden for sending this in!
“Too good to treat? Ischemic stroke patients with small CT perfusion lesions may not benefit from thrombolysis”
http://onlinelibrary.wiley.com/doi/10.1002/ana.24714/abstract
Neurologist here. I'm truly interested in where this great divide between neuro and ER comes from. We must all be seeing the same cases, right? All the strokes will typically land in an ER somewhere. So, our experience is roughly equivalent. The evidence is the same. We're both smart groups of physicians. What's going on? Neurologists slowly shake their heads when this comes up, and say, "They don't follow the patients on the ward, so all they see is the few bad bleeds without seeing the recoveries that happen three days later." The ER docs I've heard shake their heads, and say, "They just block out the bad cases." Which is it? We can't both be right about this. Either tPA is over-rated or it is not. I'm an neurologist, so you can guess where my thoughts are. I'm not trying to be argumentative, I'm truly curious how we both came to such drastically different conclusions based on what seems to be the same evidence and clinical experience. We have to work to together to help out patients, and having this great a disparity in perception of treatment outcomes on a very common condition ultimately will confuse patients and families at best, and perhaps lead to delayed or overplayed treatment. I honestly don't think given the level of evidence in stroke trials nowadays, a trial with several hundred patients will change either side's mind. I also highly doubt anyone will repeat the NINDS or any other pivotal tPA trial at this point either. How then do we get along? Any thoughts?
It's hard to say the best way forwards. If it were a simple resolution, it wouldn't be going on 20+ years with continued controversy. The problem, as you point out, is foundational – the strength of evidence is not high and lends itself to divergent interpretation. Either you feel as though you need to protect patients by erring on the side of caution with unproven therapy, or you place faith in the evidence as probably beneficial and follow through.
The reasons why many physicians remain skeptical regarding the evidence regarding tPA are well-documented. The effect size, except in NINDS, is small – and, frequently, smaller than the threshold for statistical significance. ATLANTIS has been conveniently forgotten. Effectiveness is always lower than efficacy, so the generalizability of the evidence to the bedside is limited. Conflicts of interest with Genetech/Boehringer Ingelheim are prolific. IST-3 is embraced by many as a positive trial, despite being negative for the primary outcome and even its secondary effect sizes probably magnified by its obvious biases. We could go on and on.
That said, we've also seen it work. And, based on what we know from advanced imaging, proper patient selection, and successful reperfusion, for certain patients there are likely profound differences in outcomes.
If you're looking for a way forward – while discounting the possibility of ever obtaining new, randomized, blinded, placebo-controlled evidence regarding tPA – then we need unambiguous trials looking at perfusion-based patient selection to better individualize therapy and shrink the target population to those with the maximal risk/benefit ratio. No one is helped by ineffective thrombolysis, and the costs are substantial. Our next steps should be focused on demonstrating sub-populations of acute stroke that have the greatest potential for benefit, and the side effect of these investigations will be to identify sub-populations with no benefit (as suggested by the trial cited above).