The TRUST trial was a prospective, open-label, randomized, blinded-endpoint, and multicenter study aimed at assessing the efficacy and safety of intravenous UK administered within 6 h of symptom onset, in accordance with Chinese stroke guidelines [11]. The rationale, design, and methods of the trial have been described previously [17], in addition, the initial and final protocols were shown in the Additional File1 The protocol received approval from the ethics committees of the First Affiliated Hospital of Zhengzhou University and of all participating sites. The study was conducted at 25 hospitals (Additional File 2: Table S1) in China from October 4, 2020, to November 11, 2022. Patients or their representatives provided informed consent before enrollment.

Patients were eligible if they had an acute ischemic stroke and a National Institutes of Health Stroke Scale (NIHSS) score of 0–5, age 18–80 years, treatment initiation within 6 h of symptom onset, first stroke occurrence, or patients with previous stroke of a pre-stroke modified Rankin scale (mRS) score ≤ 1. Head computed tomography (CT) imaging or magnetic resonance imaging (MRI) was required, for patients with suspected clinical neurologic events. Patients with lesions larger than 1/3 middle cerebral artery territory, intracranial hemorrhage, complications, or other contraindications were excluded. We enrolled all ischemic stroke patients with NIHSS ≤ 5 without distinguishing disabling and non-disabling in present research. Considering that there is a certain subjectivity of patients and clinical physicians in evaluating disabling and non-disabling. The protocol provides further details on the inclusion and exclusion criteria [17].

Randomization sequences were generated by an independent statistician using a permuted block randomization schema with a block size of 4 stratified by center. Patients were randomly assigned in a 1:1 ratio to receive either UK intravenous thrombolysis or guideline-based best medicine treatment by a web-based randomization system.

The intravenous thrombolysis group was administered 1,000,000 U UK dissolved in 100 mL of saline via continuous intravenous infusion for 30 min, and the treatment was given within 6 h of stroke onset. The guideline-based best medicine treatment group received antithrombotic treatments, as implemented by the local investigator. The best medication treatment was initiated immediately after randomization. Post-treatment clinical management followed established protocols and clinical guidelines [17]. The study required follow-up neuroimaging (CT imaging or MRI as per institutional standard of care) within 22 to 36 h post-randomization.

The primary outcome was an excellent functional outcome, defined as the mRS score of 0–1 at 90 days. The secondary efficacy outcomes included new clinical vascular event (vascular death, hemorrhagic stroke, ischemic stroke, or myocardial infarction), functional disability (assessed by the 6-level ordinal mRS score, combining levels 5 and 6), general health-related quality of life (EuroQoL group EQ-5D) score, range: 0 (death) to 1 (perfect health), and Barthel Index score ≥ 95 (range: 0 [totally dependent] to 100 [independent]) at 90 days.

The primary safety outcome was symptomatic intracranial hemorrhage (sICH) within 36 h post-randomization, based on the definition of the National Institute of Neurological Disorders and Stroke (NINDS) criteria: clinical progression accompanied by the imaging evidence of intracranial hemorrhage within 36 h of intravenous thrombolysis [18]. Additional safety endpoints included 90-day all-cause mortality, adverse events, and severe adverse events. A blinded central adjudication committee confirmed all reported efficacy and safety endpoints.

We estimated the proportion of the primary efficacy outcome based on the findings from the Clopidogrel with Aspirin in Acute Minor Stroke or Transient Ischemic Attack (CHANCE) study [19]. The primary outcome was estimated at 81.5% for the control group and 88% for the UK group. To detect a 6.5% difference in the primary outcome between the intervention and control group with 80% power and a two-sided alpha of 0.05, accounting for a 5% loss to follow-up, 1002 subjects were required.

Independent academic statisticians analyzed the trial according to the statistical analysis plan and the protocol detailed in Supplement 1. The primary endpoint analysis was conducted on the intention-to-treat population using a generalized linear model (GLM), adjusting for pre-specified covariates: age, gender, baseline NIHSS score, systolic blood pressure, and myocardial infarction. Adjusted odds ratios (ORs), risk ratios (RRs), and their 95% confidence intervals (CIs) were derived from logistic regression and log-binomial regression models, respectively. To evaluate the impact of crossovers between the intervention and control group, we analyzed both the per-protocol population (patients treated according to the randomization assignment [excluding crossovers] and did not violate the inclusion or exclusion criteria or had significant protocol deviations) and the as-treated population (based on the actual treatment received during the study, regardless of the initially assigned randomized treatment). Additionally, results from treating assignment as an instrumental variable were reported as sensitivity analyses to handle the effect of treatment-switching in this study. The secondary binary outcomes were analyzed using the GLM model, as done for the primary endpoint analysis. For the analysis of the ordinal mRS score at 90 days, an ordinal logistic regression model was utilized to calculate common OR and 95% CI. The GLM with a normal distribution and identity link function was employed to estimate the adjusted mean difference and 95% CI for secondary continuous outcomes.

Subgroup analyses were conducted based on age (< 65 or ≥ 65 years), gender (women or men), the time from symptom onset to treatment (≤ 4.5 or > 4.5 h), IIb/IIIa inhibitors prescription (yes or no), the history of hypertension, diabetes mellitus, dyslipidemia, and prior use of antiplatelet or anticoagulant medication (yes or no). The interaction term between the subgroup variable and treatment group in the model was used to assess the homogeneity of the treatment effect.

Given the risk of type I error inflation from multiple comparisons, the subgroup and secondary outcome analysis findings should be considered exploratory. All data analyses were conducted using SAS (version 9.4). Two-sided p values less than 0.05 were deemed statistically significant.