First, sorry for the extended silence. The holidays, a very busy stroke service, illness, and most recently a vacation (a real one–not an International Stroke Conference quasi-vacation) combined to put blogging on the back burner. Now that news from ISC is making waves, however, it’s time to go back online.
I want to extend a warm welcome to our distinguished alumnus, Dr. Georgios Manousakis, who has recently joined our little community and, after reading the three NEJM articles about which below, poses the question “Is this the end of endovascular stroke therapy?”
As many of you know by now, NEJM just published online, in concert with their presentations at ISC, three important trials of acute neuroendovascular stroke treatment, all negative. The first, IMS III, was a randomized study of IV tPA alone vs. IV tPA followed by endovascular treatment. The combined treatment was no more effective than tPA alone. In figure 2 of the manuscript and in figure 3 (p. 27) of the supplementary appendix are a variety of subgroup analyses, none of which showed a statistically significant benefit of combined therapy over IV tPA alone. There were non-statistically significant trends toward benefit of endovascular care in subjects receiving IV tPA within 2 hours of symptom onset/last known well, undergoing groin puncture within 90 minutes of tPA bolus, and having a baseline NIHSS score >20.
The second trial was SYNTHESIS, a randomized study of IV tPA vs. endovascular treatment in subjects presenting early enough to be eligible for IV tPA (within 4.5 hours). There were no significant differences in outcome, and in this study, subgroup analysis did not suggest a time-to-treatment interaction.
The third trial was MR RESCUE. This study addressed not only the question of whether endovascular treatment is better than medical treatment (which included IV tPA for those eligible), but also (actually, primarily) whether selection of patients by MR or CT perfusion imaging identifies those most amenable to endovascular treatment. In this study, subjects with a favorable imaging profile fared better in terms of infarct size and functional disability irrespective of whether they received endovascular or standard care. Endovascular intervention did not result in better outcomes overall, or in those subjects with favorable imaging.
So, returning to Dr. Manousakis’s question: Is this the end of endovascular stroke treatment? Short answer: I think not the end, but likely a curtailment. First, a cop-out: Remember that most endovascular stroke treatment focuses on intracranial aneurysms, for which there is much better evidence of its effectiveness. Much additional neuroendovascular care is for AVMs, the scientifically best treatment for which we know next to nothing, but for which there is strong consensus that embolization plays a role.
Regarding the role of endovascular treatment for acute ischemic stroke, you can expect a deluge of commentary to follow this triple whammy of negative studies. Dr. Marc Chimowitz wrote the editorial accompanying NEJM’s online publication, and I’ll let the big guns argue over the finer points of trial design.
I believe that there likely is a population of acute ischemic stroke patients who do stand to benefit from endovascular treatment–we’ve all seen the occasional case of astonishing recovery after a thrombus is plucked from someone’s M1 or basilar artery. One problem, which I think is solvable, is how to identify the patients who stand to benefit. I believe that eventually, we’ll be able to refine our selection criteria based on time from onset, clinical severity, and vascular and parenchymal imaging.
The corollary problem is how to do that quickly. Right now, the time it takes to acquire, process, and interpret multimodal neuroimaging studies is likely attenuating any procedural benefit. It’s like Heisenberg’s uncertainty principle for stroke: Once you know the patient’s precise neurovascular status, it becomes impossible to help him. Either the imaging selection criteria must remain simple (say, non-contrast head CT with ASPECTS score along with CT angiography) or the technology must progress such that more information can be garnered more quickly than is currently possible. Example here.
But from a public health perspective, there’s a bigger problem: There are 800,000 ischemic strokes each year in the U.S. How many of those patients will have just the right combination of clinical deficits, timing from onset, large vessel status, and parenchymal status to qualify them for endovascular care? How many of them will live close enough to a medical center capable of delivering such care? Does it make sense to develop the infrastructure necessary to ensure that the entire population has such access if only a tiny fraction of stroke patients will even qualify for it? In my view, this is the real issue. There likely is a small fraction of stroke patients who stand to benefit from endovascular intervention, but the public health impact of this particular mode of treatment is likely to be negligible.
In conclusion, my prediction is that endovascular care will remain very important in the treatment of aneurysms and AVMs. It will play a continued, but minor, role in the treatment of acute ischemic stroke but for now, in light of the studies above, it should remain confined to the realm of clinical trials.
And don’t get me started on carotid disease . . .