Abstract 179: Patient Selection for Endovascular Stroke Therapy: Can MRA substitute for PWI?
Background: The DEFUSE 2 study demonstrated that in acute ischemic stroke patients, the perfusion-diffusion mismatch profile, termed Target Mismatch (TMM), is associated with favorable clinical outcomes following endovascular reperfusion; no association was present for patients without TMM. Perfusion imaging,however, may not be feasible in various clinical settings. We examined whether a combination of magnetic resonance angiography (MRA) and DWI could be used as an alternative to PWI/DWI.
Methods: We assessed agreement between the MRA/DWI mismatch and TMM profiles. MRA/DWI Mismatch was defined as an ICA or MCA-M1 occlusion with a DWI volume <25 ml, or a distal MCA occlusion with a DWI volume <15 ml. The TMM was defined as a PWI (Tmax >6 sec) / DWI >1.8, DWI <70 ml and PWI (Tmax >10 sec) <100 mL. Favorable Clinical Response was defined as an improvement of ≥8 points on the NIHSS between day 0 and day 30, or an NIHSS score of ≤1 at day 30. The response to reperfusion was assessed in patients with MRA/DWI mismatch and no MRA/DWI mismatch, and similarly, in patients with TMM and no TMM.
Results: In DEFUSE 2, 58 of 94 patients had MRA/DWI mismatch (61.7%) and 76 had TMM (80.9%). The MRA/DWI mismatch model correctly identified 71% of the TMM patients (54/76). It had a high positive predictive value for identifying TMM (54/58; 93.1%), but relatively low sensitivity (54/76; 71%). There was a differential response to reperfusion according to a patient’s TMM status (p=0.03), but not according to their MRA/DWI mismatch status (p=0.5; see figure).
Conclusions: The MRA/DWI mismatch model has high specificity for identifying TMM but lacks sensitivity.Consequently, patient selection based on MRA/DWI criteria may exclude a substantial proportion of patients from treatment who have the potential to respond favorably to reperfusion.
- © 2012 by American Heart Association, Inc.