Wednesday, February 13, 2013

Mechanical Embolectomy Kills People


Interestingly, it especially killed people who were going to have a favorable outcome with standard care.

This is MR-RESCUE, a trial evaluating the efficacy of endovascular mechanical thrombectomy for acute ischemic stroke.  Patients were eligible for this trial up to 8 hours from stroke onset, and most were enrolled because they were outside the window for tPA – or received tPA but failed to recanalize.  One of the special features of this study was using emergent MRI to identify a patient subgroup that had a potentially viable "penumbra" of brain tissue surrounding the original infarct.  The imaging hypothesis in this study was that patients would particularly benefit from mechanical intervention in the presence of a penumbra such as this.

However, they were wrong.  Oddly, the authors reported their primary outcome differences in mean mRS.  As discussed on the last blog post, mean and median mRS aren't used in stroke trials because the profound disability/living death/death numbers at the bottom of the scale don't represent the clinically relevant treatment effects.  Regardless, they failed to show benefit of mechanical embolectomy.

Overall, patients simply did poorly.  This is a fine example of the exquisite relationship between NIHSS and outcomes, as the median NIHSS in this trial was 17, less than 20% of the patients had good outcomes (mRS 0-2), and 21% died with in 90 days.  Looking at Figure 2, it's clear the penumbra was an excellent prognostic feature – until the mechanical embolectomy occurred.  Then, mortality jumps from 9% to 16%, and the favorable mRS drops from 23% to 15%.

These authors used the MERCI and Penumbra systems.  You might already be familiar with the MERCI retriever from earlier, negative trials with significant device complications.  Someday we might see the last of it – but, I'm guessing, where there's already money sunk into a device, there's more patient harms to come.

"A Trial of Imaging Selection and Endovascular Treatment for Ischemic Stroke"

No comments:

Post a Comment

Comments on posts > 10 days old will be moderated; blame spammers.