Nirali Vora, MD

Athappan G, Chacko P, Patvardhan E, Gajulapalli RD, Tuzcu EM, and Kapadia SR. Late Stroke: Comparison of Percutaneous Coronary Intervention Versus Coronary Artery Bypass Grafting in Patients With Multivessel Disease and Unprotected Left Main Disease: A Meta-Analysis and Review of Literature. Stroke. 2013

Quiz: A patient has left main CAD and can be treated with percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG). Don’t worry, I’m not asking you to choose a procedure since our cardiology colleagues will take care of that. But, you should know the stroke risk with each procedure.

  1. Which procedure carries a higher risk of stroke? CABG per SYNTAX trial 2.2% vs. 0.6%, p=0.03, at 1 year. 
  2. Does the lower stroke rate in PCI catch up to CABG in the long-term?Maybe. SYNTAX 5 year follow up 3.7% vs. 2.4%, p=0.09 at 5 years.

Notice how the PCI rate of stroke went up (0.6 -> 2.4)? The focus of this newly published meta-analysis was to look at a larger sample size and determine if the stroke rates really do equalize between the two interventions after ≥1 year follow up. The authors identified 80,000 patients with more than half of the patients were followed 2+ years out.

The PCI group persistently had a significantly lower cumulative stroke rate than the CABG group, even when looking at 1 vs. 2 vs. 3 vs. 4 vs. 5 years out. In the 5+ years group, there was no significant difference but there were only ~1600 patients. There was no significant difference among diabetics, but there was a trend favoring PCI. In looking at RCTs only, PCI did better in the 30 day perioperative period, but then had no difference with CABG in terms of stroke at 1 year. The authors suggest that the increased rates of stroke in PCI in the later period may be do to post-revascularization atrial fibrillation; or in my opinion, the ongoing vascular risks factors for the patient population overall.

Bottom line for me: there are more strokes after CABG than PCI, even >1 year post-op. However, determining the optimal coronary intervention will be patient dependent and incorporate all-cause mortality, recurrent MI, target revascularization, and cost effectiveness in addition to stroke.