DWI-FLAIR mismatch guided thrombolysis in patients without large-vessel occlusion: real-world data from a comprehensive stroke centre.

Introduction: A significant proportion of ischaemic stroke patients present with unknown symptom onset time. DWI-FLAIR mismatch on MRI can help to identify those eligible for thrombolysis. We set out to analyse the short-term efficacy and safety of thrombolysis in a real-world setting. Methods: A retrospective single-centre observational study was conducted. We collected data between January 2017 and April 2020. Patients with a large vessel occlusion (LVO) were excluded. Outcomes were compared between thrombolysed patients and those who did not receive alteplase due to lack of DWI-FLAIR mismatch or other contraindications. We analysed baseline and discharge NIHSS scores for efficacy and defined good outcome as any neurological improvement (ANI) on the NIHSS. In terms of safety, the presence and severity of intracerebral haemorrhage on follow-up imaging was analysed, and mortality at 90 days assessed. Results: Seventy-one patients were included in this study, of whom 29 received thrombolysis. Significantly more patients had ANI in the thrombolysed group (OR, 3.16; 95% CI, 1.178-8.479; p = 0.020). In a multivariable logistic regression analysis, only thrombolysis correlated with ANI (OR, 3.051; 95% CI, 1.135-8.206; p = 0.027). Two thrombolysed patients suffered intracerebral haemorrhage (6.90%), of whom one was symptomatic and eventually fatal. We did not find a significant difference in 90-day mortality between the two groups (OR, 0.81, 95% CI, 0.134-4.856; p = 1.000). Conclusions: Our real-world data demonstrate that thrombolysis based on DWI-FLAIR mismatch in patients without LVO has an early beneficial effect. The rate of intracerebral haemorrhage was similar to this complication reported in large thrombolysis trials with known onset times.