Background: DEFUSE 2 demonstrated that patients with Target mismatch on MRI had favorable clinical response to reperfusion. This study compares the accuracy of early follow-up perfusion imaging vs. end of procedure angiographic TICI scores for predicting functional outcome.
Methods: Prospectively enrolled ischemic stroke patients underwent a baseline MRI followed by endovascular therapy within 12 hours of symptom onset. Patients with the Target mismatch profile, with both end of procedure TICI score and early follow-up perfusion imaging were included. Reperfusion was defined as % reduction in perfusion lesion (Tmax >6s) volume between baseline and early follow-up. Discordance between TICI and perfusion imaging was defined as TICI 0-2a with >50% reperfusion or TICI 2b-3 with <50% reperfusion.
Results: 66 patients were included; 12 (18%) TICI 0-1, 19 (29%) TICI 2A, 23 (35%) TICI 2B, and 12 (18%) TICI 3. More favorable TICI scores were associated with greater reperfusion (p<0.001), but 33% had discordance between TICI and reperfusion imaging: 55% of TICI 0-2a patients and 14% of TICI 2b-3 patients. MRI reperfusion ≥90% had a more favorable mRS distribution than 50-90% (p=0.001) or <50% (p=0.004). TICI 2b-3 scores had a more favorable mRS distribution than TICI 2a (p=0.032) or TICI 0-1 (p=0.010) (Figure). mRS 0-2 was achieved in 72% with >90% MRI reperfusion vs 60% of TICI 2b-3 (p=0.42). The sensitivity of reperfusion ≥90% for predicting mRS 0-2 was not significantly different from TICI 2b-3 (0.58 vs. 0.68) p=0.45; but specificity was higher for reperfusion (0.80 vs. 0.60) p=0.039.
Conclusions: While there is a strong association between end of procedure TICI and MRI reperfusion, one third of the DEFUSE 2 patients had discordant results. This may be due to additional reperfusion following the procedure, or subsequent thromboembolism. Reperfusion ≥90% on perfusion imaging had a higher specificity than a TICI 2b-3 score for predicting good outcome in Target mismatch patients.