Maddening Lack of Perspective in Endovascular Stroke Therapy

Despite three negative trials in the New England Journal of Medicine last year, accompanied by an editorial calling for a moratorium on reimbursement for endovascular therapy for acute stroke, proponents of this therapy forge ahead.  The negative trials, at the very least, have encouraged researchers to recognize – where tPA advocates fail to see – that patient selection is key, and not all candidates eligible for endovascular therapy are likely to benefit.

However, these authors, despite their well-intentioned effort to better target endovascular interventions, completely miss the mark.  This is a retrospective review of patients undergoing endovascular therapy between 2008 and 2012, stratified by selection protocols.  In 2010, the institution switched from an “all-comers” strategy to an MRI diffusion-weighted imaging-guided protocol.  The hypothesis: patients selected for endovascular therapy on the basis MRI-guided infarct characteristics would display better outcomes than the “all-comers” selection cohort.

And, their registry provides them the necessary substrate.  Between 2008 and 2010, 85 patients with median NIHSS 17 underwent endovascular therapy with dismal outcomes:  9.5% mRS 0-2 at 30 days and 47.6% mortality.  After implementation of the imaging-based protocol, only half of eligible patients qualified – and of the 92 patients with median NIHSS 15.5 who eventually underwent endovascular therapy under the new protocol, 23.9% had mRS 0-2 and mortality dropped to 19.7%.  Victory!  So say these authors.

However, there’s another cohort visible only by its absence in their tabular data – the post-protocol group that did not undergo endovascular therapy.  These 87 patients, with a median NIHSS ~16.5, had imaging characteristics thought by these authors to not qualify for endovascular therapy.  For this group, 23.1% had mRS 0-2 and 30% mortality.  These outcomes are not statistically different from the endovascular group.

The authors do not in any manner address this inconsistency in their data set – and the seemingly favorable intervention of lack-of-intervention.  Rather than the authors’ conclusion imaging could guide therapy – the better conclusion is that endovascular therapy was frighteningly harmful to many in the pre-protocol phase, and simply reducing the number of patients undergoing unnecessary intervention improved global outcomes.  And, for just added perspective: the placebo group in NINDS, with median NIHSS 15, had 26% mRS 0-1 and 21% mortality at 3 months – and this is before the modern post-stroke care that does more to improve outcomes than tPA could ever claim.

The madness continues.

“Addition of Hyperacute MRI Aids in Patient Selection, Decreasing the Use of Endovascular Stroke Therapy”
http://stroke.ahajournals.org/content/early/2014/01/09/STROKEAHA.113.003880.short