The study population consisted of consecutive AIS patients registered in the Austrian Stroke Unit Registry (ASUR) and the Acute STroke Registry and Analysis of Lausanne (ASTRAL) database. ASUR is a nationwide prospective registry of the Austrian stroke unit network founded by the Federal Ministry of Health. Anonymized data on baseline characteristics, risk factors and etiology, acute management, and functional outcome at discharge and at 3 months are registered for all patients admitted to one of currently 38 stroke units in Austria. Data collection and clinical ratings are performed by experienced stroke neurologists using standardized definitions of variables and scores. To ensure high data quality, immediate electronic data entry is performed. The web-based database includes online plausibility checks and a help function. Biannual educational meetings are also held to guarantee uniform data documentation. A detailed methodological description has been published previously [11]. On the other hand, ASTRAL is a single center-based cohort of all ischemic stroke patients admitted within 24 hours of stroke onset at the stroke center and/or intensive care unit of the Centre Hospitalier Universitaire Vaudois (CHUV) [12]. This registry is approved by its institution as a clinical and research registry and follows institutional regulations. Data on pre-stroke conditions, cerebrovascular risk factors, imaging and laboratory findings, stroke mechanism, acute management, functional outcome, and recurrences are collected up to 12 months. Data are collected online in a prespecified format by experienced and formed medical personnel. The web-based database also includes online plausibility checks and a help function. There is no uniform policy in ASUR-participating centers on whether patients with preexisting disability mRS ≥2 should be offered MT. At the CHUV, MT was offered standardly when pre-stroke mRS was ≤2 until 2017, and ≤3 since 2018. However, the final decision to offer MT remained at the discretion of the treating stroke unit physician at the time of stroke.

For the current analysis, we extracted from both databases those patients with AIS, treated between January 1, 2015 and December 31, 2019, aged ≥18 years, who had a pre-stroke mRS [13] ≥3, and potentially treatable LVO. Demographic and clinical data collection included age, sex, pre-stroke mRS, stroke severity, history of atrial fibrillation, arterial hypertension, and diabetes mellitus. Stroke severity on admission was measured using the National Institutes of Health Stroke Scale (NIHSS) [14]. LVO was categorized based on treatable occlusions in vessel segments M1 and M2 of the middle cerebral artery, extracranial and/or intracranial internal carotid artery, P1 segment of the posterior cerebral artery, and basilar artery. Patients were further categorized based on treatment received: MT, IVT, both, or none. Patients not receiving MT were considered as BMT.

Primary efficacy endpoint was defined with regard to pre-stroke disability as no or negative difference between pre-stroke mRS and mRS at follow-up at 3 months. Secondary efficacy endpoint was defined as a NIHSS difference of ≥8 points (early neurological improvement) between admission and at 24-48 hours. Further endpoint included mortality at 3 months. Safety endpoint was defined as a symptomatic intracerebral hemorrhage (sICH) according to the Thrombolysis with Alteplase 3 to 4.5 Hours after Acute Ischemic Stroke (ECASS3) study criteria [15].

We summarized the categorical variables as absolute numbers and percentages and continuous variables as means and standard deviation; quantitatively skewed variables were summarized as medians and interquartile range. Categorical variables were compared using the chi-square test, Fisher’s exact test, or Kruskal-Wallis test, whereas continuous variables were compared using the Student’s t-test or Mann-Whitney U test. Multivariable regression models were used to examine the adjusted effects of MT on outcomes. Standard demographic and clinical factors known to influence outcome, including age, sex, admission NIHSS, pre-stroke mRS, hypertension, diabetes mellitus, atrial fibrillation, occlusion site, IVT use, and center, were used as covariates. We performed propensity score matching based on a logistic regression model controlling for the covariates as mentioned above, which matched the patients with most similar regression scores against each other, as implemented in the R package “MatchIt.” Statistical was analyses were performed using the IBM SPSS statistical software version 27 (IBM Co., Armonk, NY, USA), as well as the R statistical software version 4.0.2 (R Foundation for Statistical Computing, Vienna, Austria), with GLM function from the MASS package and “rpart” function from the “rpart” package.

The ASUR is part of a government quality assessment program for nationwide stroke care and is financed by the Federal Ministry of Health. Data entry is mandatory for stroke units. All data are anonymized and centrally administered by the Gesundheit Österreich GmbH, which is the national research and planning institute for health care and a competence and funding center of health promotion. All scientific analyses included in this study were approved and supervised by a national academic review board. Based on this setting, informed consent for patients enrolled in the registry was waived. In ASTRAL, all data came from routine clinical and radiological management and were anonymized before analysis following the principles of the Swiss Human Research Ordinance. Therefore, there was no need for ethical commission approval or patient consent according to the Swiss Human Research Act and the applicable data protection legislation.

Anonymized patient data are administered by the Gesundheit Österreich GmbH; inquiries on data availability are reviewed and granted by an academic review board. The anonymized data from ASTRAL supporting the findings of this study are available from the authors upon reasonable request and after signing a data transfer and use agreement.

Within the study period, 2,111 patients with AIS were entered in the ASTRAL registry. Of those, 347 has pre-stroke mRS 3-5; 170 patients with potentially treatable LVO and pre-stroke mRS 3-5 were included in the study. In the same period, 8,664 AIS patients with pre-stroke mRS 3-5 were included in the ASUR registry. Of those, 292 with potentially treatable LVO and pre-stroke mRS 3-5 were included in the study. Overall, we included 462 patients in the analysis: 292 from the ASUR registry (111 MT, 181 BMT) and 170 from the ASTRAL registry (64 MT, 106 BMT). Altogether, 175 patients underwent MT and 287 received BMT. Patients undergoing MT were younger, had slightly higher admission stroke severity, and lower pre-stroke disability. Thrombolysis in cerebral infarction scores of 2b/3 in patients undergoing MT were achieved in 142 (81.1%) patients. IVT was administered in 75 (42.9%) patients in the MT group and 100 (34.8%) patients in the BMT group (P=0.100). Onsetto-door times and door-to-needle delays did not differ significantly between the two groups (Table 1).

At 3 months after stroke, 47 (26.9%) patients in the MT group returned to their pre-stroke mRS, compared to 58 (20.2%) in the BMT group (P=0.100) (Table 2). After adjusting for age, sex, admission NIHSS, pre-stroke mRS, hypertension, diabetes mellitus, atrial fibrillation, occlusion site, IVT, and center, MT was associated with higher odds of returning to the pre-stroke mRS at 3 months (adjusted odds ratio [aOR], 2.5; 95% confidence interval [CI], 1.39 to 4.47; P=0.002) as compared to BMT (Table 3). Early neurological improvement NIHSS ≥8 points occurred in 50 (30.3%) patients undergoing MT compared to 26 (11.8%) receiving BMT (P<0.001) (Table 2). After adjustment, MT was associated with early neurological improvement NIHSS ≥8 points (aOR, 2.62; 95% CI, 1.41 to 4.88; P=0.002) (Table 3). Death at 3 months occurred in 76 (43.4%) patients in the MT group compared to 180 (62.7%) in BMT group (P<0.001) (Table 2). After adjustment, MT was inversely associated with death at 3 months (aOR, 0.29; 95% CI, 0.18 to 0.49; P<0.001) (Table 3).

As defined by the ECASS3 criteria, sICH was present in seven (4%) patients in the MT group and in six (2.1%) patients in the BMT group (P=0.250) (Table 2).

Using propensity score matching, 168 MT and 168 BMT patients were selected. Except for age, there were no significant differences between the groups (Table 4).

At 3 months after stroke, 47 (28.0%) patients in the MT group returned to their pre-stroke mRS, compared to 32 (19.0%) in the BMT group (P=0.040) (Table 5). Detailed distribution of mRS at 3 months categorized by MT versus BMT is shown in Figure 1. After adjustment, MT was associated with higher odds of returning to the pre-stroke mRS at 3 months (aOR, 2.54; 95% CI, 1.16 to 5.57; P=0.020) compared with the BMT (Table 5). Early neurological improvement NIHSS ≥8 points occurred in 50 (29.8%) patients undergoing MT compared to 18 (10.7%) receiving BMT (P<0.001) (Table 5). After adjustment, MT was associated with early neurological improvement NIHSS ≥8 points (aOR, 2.72; 95% CI, 1.26 to 5.92; P=0.010) (Table 6). Death at 3 months occurred in 71 (42.3%) patients in the MT group compared to 107 (63.7%) in BMT group (P<0.001) (Table 5). After adjustment, MT was inversely associated with death at 3 months (aOR, 0.27; 95% CI, 0.15 to 0.51; P<0.001) (Table 6).

As defined by the ECASS3 criteria, sICH was present in seven (4.2%) patients in the MT group and in four (2.4%) patients in the BMT group (P=0.270) (Table 5).