Maria Marburg ¹ Linda F. Rudolf ¹ Christine Matthis ¹ Alexander Neumann ¹ Alexander Neumann ¹ Robert Schulz ² Björn Machner ³ Peter Schramm ¹ Georg Royl ¹ Philipp J. Koch ^1*

• ¹ University Hospital Schleswig-Holstein, Campus Lübeck, Lübeck, Germany
• ² University Medical Center Hamburg-Eppendorf, Hamburg, Hamburg, Germany
• ³ Schön Klinik Neustadt, Neustadt, Rhineland-Palatinate, Germany

Despite profitable group effects on functional outcomes after mechanical thrombectomy (MT) in large vessel occlusions (LVO), many patients with successful reperfusion show a non-favorable long-term outcome, highlighting the necessity to identify potential biomarkers foreseeing outcome variability. In this regard, the role of perfusion mismatch imaging for outcome variability in the early time window within 6 hours after symptom onset is a matter of debate. We reached out to investigate under which conditions early perfusion mismatch imaging accounts for variability in functional outcome after mechanical thrombectomy. In a retrospective single-center study, we examined 190 consecutive patients with LVO admitted to the Medical Center Lübeck within 6 hours after symptom onset, all of whom underwent MT. Perfusion mismatch was quantified by applying the Alberta Stroke Program Early CT score (ASPECTS) on CT-measured cerebral blood flow (CBF-ASPECTS) and subtracting it from an ASPECTS application on cerebral blood volume (CBV-ASPECTS), i.e., ASPECTS mismatch. Using multivariable ordinal regression models, associations between ASPECTS mismatch and modified Rankin Scale (mRS) after 90 days were assessed. Further, the interaction between ASPECTS mismatch and the core lesion volume was calculated to evaluate conditional associations. ASPECTS mismatch did not correlate with functional outcome when corrected for multiple influencing covariables. However, interactions between ASPECTS mismatch and CBV-ASPECTS (OR: 1.12 (1.06-1.18), p-value<0.001), as well as NCCT-ASPECTS (OR: 1.15 (1.06-1.25), p-value<0.001), did show a significant association with functional outcome. Model comparisons revealed that profoundly in patients with large core lesion volumes (CBV-ASPECTS < 6 or NCCT-ASPECTS < 6), perfusion mismatch showed a negative correlation with mRS. Perfusion mismatch imaging within the first 6 hours of symptom onset provides valuable insights into the outcome variability of LVO stroke patients receiving thrombectomy, but only in patients with large ischemic core lesions.