Tirofiban vs standard antiplatelet therapy in acute branch atheromatous disease-related stroke: a multicenter, randomized, open-label, blinded-endpoint, parallel-group, phase III clinical trial

Summary

Background

Tirofiban is sometimes used empirically for branch atheromatous disease (BAD)-related stroke, but its benefit in this population remains unclear. This study aimed to assess the efficacy and safety of intravenous tirofiban administered within 48 h of onset in patients with acute BAD-related stroke.

Methods

This multicentre, randomized, open-label, blinded-endpoint clinical trial enrolled patients with BAD-related stroke within 48 h of onset from 25 hospitals in China between 9 November 2023 and 22 November 2025. Eligible patients were randomly assigned to the tirofiban group (intravenously, 0.4 μg/kg per minute for 30 min, followed by a 0.1 μg/kg per minute for 47.5 h) or standard antiplatelet therapy group (oral aspirin with or without clopidogrel) for 48 h. The primary endpoint was an excellent functional outcome, defined as a score of 0 or 1 on the modified Rankin Scale at 90 days. Safety outcomes included major bleeding within 90 days. This trial was registered on ClinicalTrials.gov (NCT06037889).

Findings

Among 516 patients randomized, 513 were included in the primary analysis (258 with tirofiban and 255 with standard antiplatelet therapy). An excellent functional outcome was observed in 181 of 258 patients (70.2%) in the tirofiban group and 178 of 255 (69.8%) in the standard antiplatelet therapy group (risk ratio, 1.01 [95% CI, 0.90–1.13]; P = 0.93). Major bleeding occurred in 0 of 259 participants in the tirofiban-exposed safety population and 1 of 245 (0.4%) in the standard-antiplatelet-exposed safety population.

Interpretation

Among patients with acute BAD-related stroke, intravenous tirofiban administered within 48 h of onset without concomitant oral antiplatelet therapy was not superior to standard antiplatelet therapy for achieving an excellent functional outcome at 90 days. These findings should not be extrapolated to substantially earlier treatment or to other tirofiban-based antiplatelet strategies.

Funding

National High Level Hospital Clinical Research Funding, Capital's Funds for Health Improvement and Research, CAMS Innovation Fund for Medical Sciences (CIFMS), Noncommunicable Chronic Diseases-National Science and Technology Major Project, and China Shijiazhuang Pharmaceutical Company (CSPC) NBP Pharmaceurical Co., Ltd.

Research in context.

Evidence before this study

We searched PubMed for clinical studies published in English from January 1, 2000 to July 1, 2026, using the search terms (“branch atheromatous disease” OR “ Penetrating artery territory infarction” OR “Ischemic stroke”) AND (“tirofiban” OR “glycoprotein IIb/IIIa inhibitors”). Branch atheromatous disease (BAD) represents a distinct subtype of acute ischemic stroke that is notoriously prone to early neurological deterioration (END) and associated with a high rate of severe disability, but shows low recurrence risk. Currently, there are no specific, evidence-based pharmacological treatments tailored for acute BAD-related stroke. Recent randomized controlled trials (RCTs) have demonstrated that tirofiban can safely improve functional outcomes in specific acute ischemic stroke populations, including patients with disabling non-cardioembolic stroke, and those receiving it as an adjunctive therapy following intravenous thrombolysis or endovascular treatment. Given its therapeutic benefits in these related clinical contexts, tirofiban is increasingly being administered empirically in clinical practice for patients with acute BAD-related stroke. However, whether this empirical approach is definitively effective and safe remains uncertain in BAD, as high-quality RCT evidence specifically targeting this distinct subtype is currently lacking.

Added value of this study

This is the first adequately powered, multicenter RCT assessing intravenous tirofiban vs standard antiplatelet therapy for acute BAD-related stroke. We demonstrated that tirofiban administered within 48 h of onset did not show superiority over standard antiplatelet therapy for 90-day functional outcomes. Major bleeding events were rare, precluding precise between-group safety comparisons. These findings fill a crucial evidence gap by providing randomized evidence on a commonly used empirical treatment strategy and defining its clinical effect under the regimen and treatment window evaluated in this trial.

Implications of all the available evidence

Current evidence does not support the routine empirical use of intravenous tirofiban monotherapy over standard oral antiplatelets for BAD-related stroke presenting within 48 h of onset. These findings should not be interpreted as evidence against all tirofiban-based treatment strategies, particularly because most patients were randomized relatively late after stroke onset, many END events had already occurred before randomization, and concomitant oral antiplatelet therapy was prohibited during tirofiban infusion. Future trials should explore whether ultra-early intervention, active END, or alternative pathophysiological targets can yield actual clinical benefits for this subtype.

Introduction

Branch atheromatous disease (BAD)-related stroke is a distinct and clinically consequential subtype of acute ischemic stroke, accounting for approximately 9.1%–20.2% of cases.1, 2, 3 First described by Caplan as an atherosclerotic process involving the origin of penetrating arteries, BAD differs pathophysiologically from lacunar infarction due to lipohyalinosis and is not well captured in the conventional TOAST (Trial of ORG 10172 in Acute Stroke Treatment) classification system.4, 5, 6 With advances in magnetic resonance imaging, BAD is increasingly recognized as a single subcortical infarction extending along a perforator territory without severe parent artery stenosis. Clinically, BAD-related stroke is notable for its high rate of early neurological deterioration (END) and unfavorable outcome.^1,5,7 Despite often presenting with initially mild deficits, more than 30% of patients remain disabled at follow-up, underscoring the progressive nature of perforator-territory ischemia rather than a benign small-vessel syndrome.^1,5,7 In contrast, the annual risk of recurrent stroke in BAD is relatively low (less than 3.8%).⁷ This high disability-low recurrence profile suggests that preventing END and improving functional outcome, rather than recurrent ischemic events, may be a more clinically meaningful treatment target in this population. Several mechanisms have been proposed to explain END, including perforator hypoperfusion, local thrombosis, and inflammation. However, no definitive pathway has been established, complicating the development of targeted therapies.^4,6 Importantly, intravenous alteplase has not consistently prevented END or improved functional outcomes in this population, and most patients lack a target for mechanical thrombectomy.⁸ Consequently, optimal acute management remains undefined,⁹ as current treatment strategies are largely empirical and no guideline-endorsed therapy specifically addresses BAD-related stroke.¹⁰

Tirofiban, a reversible glycoprotein IIb/IIIa receptor antagonist, provides rapid and potent inhibition of platelet aggregation and has been investigated across several acute ischemic stroke settings.11, 12, 13, 14 Prior studies suggest that tirofiban may reduce END and improve functional outcomes in select populations, including patients without large-artery occlusion and those treated adjunctively after intravenous thrombolysis.11, 12, 13 Current evidence is largely derived from heterogeneous stroke cohorts and was not designed to evaluate etiologic subtypes with distinct early clinical trajectories. Given that BAD is characterized by a presumed atherosclerotic mechanism at the perforator origin and a high risk of early progression, tirofiban has been adopted empirically in clinical practice for BAD despite the absence of randomized evidence.^4,13 Whether early administration of tirofiban improves functional outcomes in patients with BAD-related stroke remains uncertain.

Therefore, we sought to determine whether tirofiban administered within 48 h of stroke onset improves 90-day functional outcomes compared with aspirin with or without clopidogrel in patients with acute BAD-related stroke.

Methods

Study design

The Efficacy and Safety of Tirofiban for patients with Branch Atheromatous Disease (BRANT) study was a multicentre, randomized, open-label, blinded-endpoint, parallel-group, phase III clinical trial conducted at 25 hospitals in China. The trial was prospectively registered with ClinicalTrials.gov (NCT06037889), and its protocol and statistical analysis plan (SAP) are available in Appendix (pp 51–201). Details of the study design have been published.¹⁵ The trial protocol was approved by the institutional review board of Peking Union Medical College Hospital (No I-23PJ1242) and all participating centers. Written informed consent was obtained from all patients before randomization.

The trial was designed and conducted by the steering committee, with safety monitored by an independent data and safety monitoring board (DSMB). An independent clinical research organization (Pharmaron [Chengdu] Clinical Services Co., Ltd.) monitored trial quality. An independent clinical events committee (CEC) adjudicated all clinical and safety events. It was performed in accordance with the ethical principles of the Declaration of Helsinki and the International Council for Harmonization guidelines for Good Clinical Practice. This article followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guidelines.

Participants

Eligible patients were adults aged 18–75 years who presented within 48 h of stroke onset and met predefined MRI-based diagnostic criteria for BAD.¹⁵ BAD was defined as a single deep subcortical infarction in the territory of a penetrating artery (lenticulostriate artery [LSA] or paramedian pontine artery [PPA]) on diffusion-weighted imaging, without more than 50% stenosis of the corresponding parent artery (middle cerebral artery or basilar artery) on vascular imaging. Key exclusion criteria included transient ischemic attack (TIA), intracranial hemorrhage or other nonischaemic intracranial pathology, significant extracranial or intracranial large-artery stenosis (≥50%), cardioembolic sources, receipt or planned receipt of intravenous thrombolysis or endovascular therapy, pre-stroke modified Rankin Scale (mRS) score of 2 or higher, prior intracranial hemorrhage, recent use of tirofiban, and other major conditions conferring high bleeding risk. A full list of inclusion and exclusion criteria is provided in Appendix (pp 1–2), and MRI-based eligibility was confirmed through a centralized imaging review process (Appendix p 3).

Randomization and masking

Eligible patients were randomly assigned in a 1:1 ratio to receive either intravenous tirofiban or guideline-based antiplatelet therapy using a centralized, web-based randomization system. Randomization was performed with dynamic block allocation. No stratification by study center or baseline characteristics was applied. The trial was conducted in an open-label manner. However, trial statistician, members of the independent clinical event committee, and outcome assessors, who were not involved in clinical care and had no access to treatment records, were blinded to treatment allocation.

Procedures

Patients assigned to the tirofiban group received intravenous tirofiban initiated immediately after randomization and continued for 48 h. Tirofiban was administered as a loading infusion of 0.4 μg/kg per minute for 30 min, followed by a maintenance infusion of 0.1 μg/kg per minute for the remainder of the 48-h treatment period.

Patients in the control group received guideline-based antiplatelet therapy for 48 h, consisting of either aspirin (150–300 mg daily) or aspirin (100 mg daily) plus clopidogrel (75 mg daily), at the discretion of the treating physician and in accordance with current stroke guidelines.¹⁶

After completion of the 48-h study treatment period, subsequent antithrombotic therapy was prescribed according to guideline recommendations and recorded for all participants. Post-48-h antithrombotic exposure through 90-day follow-up, including DAPT use and duration, was summarized.

Outcomes

The primary endpoint was an excellent functional outcome at 90 days, defined as an mRS score of 0–1. The mRS score was assessed at 90 days by trained evaluators blinded to treatment allocation using a structured interview conducted either in person or by telephone (Appendix p 4). The primary outcome was prespecified in the protocol and SAP and remained unchanged. The ClinicalTrials.gov entry was corrected before database lock to fix an administrative error during registration, with no impact on study (Appendix pp 5–7).

Secondary efficacy outcomes included excellent functional outcome at 7 days, ordinal mRS scores at 7 and 90 days, National Institutes of Health Stroke Scale (NIHSS) scores at 7 and 90 days, and the Barthel Index score at 90 days. END was defined as an increase of 4 or more points in the NIHSS score or an increase of 2 or more points in the NIHSS motor score within 7 days after randomization. Additional secondary outcomes included ischemic stroke, stroke (ischemic or haemorrhagic), TIA, and a composite of new-onset stroke, myocardial infarction, or all-cause death within 90 days.

Safety outcomes were major bleeding, as defined by the PLATO criteria,^17,18 serious adverse events, adverse events and all-cause death within 90 days. Prespecified secondary outcomes also included changes in bleeding-related laboratory markers at 48 h. Intracranial hemorrhage was recorded as part of safety monitoring.

All clinical events were adjudicated by an independent CEC blinded to treatment allocation.

Sample size calculation

The sample size was calculated based on the primary endpoint of excellent functional outcome at 90 days (mRS 0–1). The expected percentage of excellent functional outcome was 62% in control group and 74% in the tirofiban group, based on prior studies.¹⁵ With a 2-sided α level of 0.05 and 80% power to detect this absolute difference of 12 percentage points, 234 participants were required per group. Assuming a 10% loss to follow-up, the total planned sample size was 516 participants. Detailed assumptions and calculation methods are provided in the trial protocol (Appendix pp 51−153).

Statistical analysis

All efficacy analyses were conducted in the full analysis set according to the modified intention-to-treat (mITT) principle, including all randomized patients who did not withdraw consent and had at least one valid efficacy evaluation. Efficacy analyses were repeated in the per-protocol set including randomized patients without major protocol violations. The safety population included those who had undergone randomization and actually received at least a 30-min loading dose of tirofiban in the tirofiban group and at least one dose of prespecified antiplatelet therapy in the control group, where the safety analyses were conducted.

The primary endpoint, excellent functional outcome at 90 days (mRS 0–1), was compared between groups using the chi-square test, and risk ratios (RRs) with 95% confidence intervals (CIs) were calculated as a measurement of treatment difference. In addition, adjusted RRs were estimated using stabilized inverse probability of treatment weighting (S-IPTW). Missing values were imputed using the last observation carried forward method. Sensitivity analyses were conducted using complete-case analysis, missing values in the primary outcome imputed using worst-case and worst-best scenarios, multiple imputation, and controlling for center effect.

Secondary binary outcomes were analyzed using methods similar to those used for the primary endpoint. Ordinal outcomes, including mRS scores (0–6), were analyzed using generalized odds ratio (GenOR). Continuous outcomes were analyzed using linear regression models or the win ratio method, as appropriate based on data distribution. Time-to-event outcomes were analyzed using the Kaplan–Meier method and compared using log-rank test, with hazard ratios (HRs) and 95% CIs estimated using Cox proportional hazard models. Adjusted analyses for secondary outcomes were performed using the S-IPTW approach.

Prespecified subgroup analyses of the primary endpoint were conducted using interaction terms in regression models. There were 6 prespecified subgroup variables including age, sex, hypertension, diabetes, NIHSS score at randomization, and time from onset to randomization.

All safety analyses were conducted in the safety population. Binary safety endpoints were analyzed using chi-square or Fisher exact tests, and time-to-event safety outcomes were analyzed using Cox proportional hazards models. Adverse events were summarized by system organ class and preferred terms. In addition, the linear regression model or the win ratio method was used to analyze continuous safety outcomes, as appropriate based on data distribution. Detailed statistical analyses are described in the SAP (Appendix pp 154−201).

All statistical tests were 2-sided, with P values less than 0.05 considered statistically significant. No interim analysis was planned. Findings for secondary outcomes and subgroup analyses were considered exploratory. Analyses were performed using SAS software, version 9.4 (SAS Institute Inc) and R software (4.4.1).

Role of the funding source

The funders of this study had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. The manufacturer supplied the study drug but had no access to accumulating or final trial data.

Results

Patient characteristics

From November 9, 2023, to November 22, 2025, a total of 516 patients were randomized (Appendix pp 8–9). Three patients (all in the standard antiplatelet therapy group) were excluded due to withdrawal of consent. The full analysis set included 513 patients (258 patients in the tirofiban group and 255 in the standard antiplatelet therapy group) (Fig. 1). Two patients were lost to 90-day follow-up. After excluding patients with severe protocol violations, 482 patients (244 in the tirofiban group and 238 in the standard antiplatelet therapy group) were included in per-protocol analysis. Our trial was completed on February 26, 2026.

Fig. 1.

Patient Flow in the BRANT Randomized Clinical Trial. BAD, branch atheromatous disease; DWI, Diffusion weighted imaging; ECG, Electrocardiograph; eGFR, estimated glomerular filtration rate.

Baseline demographic and clinical characteristics were generally well balanced between treatment groups (Table 1; Appendix pp 10–12). The median age was 60 years (IQR: 54–67), with 35.5% female participants. The median NIHSS score at randomization was 4 (IQR: 3–5). The median time from stroke onset to randomization was 30.7 h (IQR: 22.7–40.6). The proportions of lesions in the LSA and PPA territories were 49.5% and 50.5%, respectively.

Table 1.

Baseline characteristics of patients.

Characteristic	Tirofiban (n = 258)	Standard antiplatelet therapy (n = 255)
Age, median (IQR), y	60 (53–67)	60 (54–68)
Female, No. (%)	91 (35.3)	91 (35.7)
Han Chinese ethnic group, No. (%)	251 (97.3)	245 (96.1)
Medical history, No. (%)
Hypertension	167 (64.7)	160 (62.7)
Diabetes	86 (33.3)	66 (25.9)
Hyperlipidemia	25 (9.7)	28 (11.0)
Prior ischemic stroke or TIA	33 (12.8)	37 (14.5)
Coronary heart disease	17 (6.6)	19 (7.5)
Smoking, ever	112 (43.4)	95 (37.3)
Regular drinking	67 (26.0)	65 (25.5)
Body mass index, median (IQR), kg/m2	24.9 (22.9–27.6)	24.9 (22.9–27.1)
NIHSS score at randomization, median (IQR)a	4 (3–5)	4 (3–5)
≤ 3	122 (47.3)	113 (44.3)
≥ 4	136 (52.7)	142 (55.7)
Barthel index score randomization, median (IQR)b	80 (60–95)	75 (60–95)
mRS score at randomization, median (IQR)c	2 (1–3)	2 (1–3)
mRS score before stroke, No. (%)c
0	228 (88.4)	227 (89.0)
1	30 (11.6)	28 (11.0)
Systolic blood pressure, median (IQR), mmHg	142 (134–157)	146 (134–158)
>140	144 (55.8)	156 (61.2)
Diastolic blood pressure, median (IQR), mmHg	87 (79–94)	86 (78–94)
>90	81 (31.4)	85 (33.3)
END prior to randomization, No. (%)	24 (9.3)	20 (7.8)
Glucose at hospital arrival, median (IQR), mmol/Ld	6.7 (5.6–9.1)	6.6 (5.6–9.1)
Culprit artery of ischemic lesion, No. (%)
Lenticulostriate artery	130 (50.4)	124 (48.6)
Paramedian pontine artery	128 (49.6)	131 (51.4)
Category of onset of stroke, No. (%)
Known onset time	223 (86.4)	216 (84.7)
Unknown time	35 (13.6)	39 (15.3)
Reasons for no thrombolysis, No. (%)
Out of the therapeutic window	247 (95.7)	245 (96.1)
Presence of contraindications	1 (0.4)	3 (1.2)
Patient refusal	10 (3.9)	7 (2.7)
Prior use of antiplatelet drugs, No. (%)e	20 (7.8)	20 (7.8)
Prior use of anticoagulants, No. (%)f	0 (0.0)	1 (0.4)
Use of antiplatelet drugs from onset to randomization, No. (%)	170 (65.9)	185 (72.5)
Mono	45 (26.5)	37 (20.0)
Dual	125 (73.5)	148 (80.0)
Use of argatroban from onset to randomization, No. (%)	2 (0.8)	3 (1.2)
Onset to MRI time, median (IQR), h	21.9 (12.5–28.7)	19.5 (10.1–28.7)
Onset to randomization time, median (IQR), h	31.2 (24.0–41.2)	29.2 (21.8–39.7)

END, early neurological deterioration; IQR, interquartile range; TIA, transient ischemic attack.

^aScores on the National Institutes of Health Stroke Scale (NIHSS) range from 0 to 42, with higher scores indicating a more severe neurological deficit.

^bScores on Barthel Index range from 0 to 100, with higher scores indicating better functional independence.

^cScores on the modified Rankin Scale range from 0 to 6, with higher scores indicating more severe disability.

^dGlucose measurements at hospital admission were missing for 13 patients.

^eUse of antiplatelet drugs within one month before stroke onset.

^fUse of anticoagulants within one month before stroke onset.

Within 48 h after randomization, 85.5% of patients in the control group received dual antiplatelet therapy, with the remainder received single antiplatelet therapy (Appendix p 13); all patients in the tirofiban group were prohibited from receiving any other antiplatelet drugs.

After the first 48 h of randomization, dual antiplatelet therapy (DAPT) was used in 199 of 258 (77.1%) participants assigned to tirofiban and 211 of 255 (82.7%) assigned to standard antiplatelet therapy; the median DAPT duration was 20 days (IQR: 17–25) and 21 days (20–29), respectively (Appendix p 14).

Primary outcome

At 90 days, 181 of 258 patients (70.2%) in the tirofiban group achieved an excellent functional outcome (mRS 0–1), compared with 178 of 255 (69.8%) in the standard antiplatelet therapy group (unadjusted risk ratio [RR] 1.01, 95% CI, 0.90–1.13; risk difference, 0.35 [−7.58 to 8.28]; P = 0.93) (Table 2 and Fig. 2). The analyses of the primary outcome, adjusted for prespecified covariates and in the per-protocol set, yielded similar results (Table 2; Appendix pp 17–19).

Table 2.

Efficacy and safety outcomes.

Outcome	Tirofiban (n = 258)	Standard antiplatelet therapya (n = 255)	Treatment effect metric	Unadjusted		Adjustedb
				Treatment difference (95% CI)	P value	Treatment difference (95% CI)	P value
Primary efficacy outcome
mRS score 0–1 at 90 d, No. (%)c	181 (70.2)	178 (69.8)	Risk ratio	1.01 (0.90–1.13)	0.93	1.00 (0.89–1.12)	0.94
Secondary efficacy outcomes
mRS score at 90 d, median (IQR)c	1.0 (0.0–2.0)	1.0 (0.0–2.0)	Generalized odds ratio	1.10 (0.84–1.43)	0.50	1.08 (0.82–1.40)	0.59
END within 7 d, No. (%)d	21/257 (8.2)	16/255 (6.3)	Risk ratio	1.30 (0.70–2.44)	0.41	1.36 (0.73–2.54)	0.33
Barthel index score at 90 d, median (IQR)e	100 (95–100)	100 (95–100)	Win ratio	0.92 (0.67–1.26)	0.60	0.92 (0.67–1.25)	0.58
NIHSS score at 7 d, median (IQR)f	2.0 (1.0–4.0)	2.0 (1.0–4.0)	Win ratio	1.05 (0.84–1.32)	0.65	1.05 (0.83–1.32)	0.68
Ischemic stroke within 90 d, No. (%)	5 (1.9)	4 (1.6)	Hazard ratio	1.23 (0.33–4.56)	0.76	1.23 (0.32–4.73)	0.77
Stroke within 90 d, No. (%)g	5 (1.9)	5 (2.0)	Hazard ratio	0.98 (0.28–3.39)	0.98	0.97 (0.27–3.46)	0.97
TIA within 90 d, No. (%)	0 (0.0)	0 (0.0)	Hazard ratio	NE	NE	NE	NE
Composite event within 90 d, No. (%)h	5 (1.9)	5 (2.0)	Hazard ratio	0.98 (0.28–3.39)	0.98	0.97 (0.27–3.46)	0.97
Safety outcomes	n = 261	n = 245
Major bleeding within 90 d, No. (%)i	0/259 (0.0)	1/245 (0.4)	Risk ratio	NE	0.30	NE	0.30
Major bleeding within 7 d, No. (%)	0/261 (0.0)	0/245 (0.0)	Risk ratio	NE	NE	NE	NE
SAEs within 90 d, No. (%)j	3/259 (1.2)	3/245 (1.2)	Risk ratio	0.95 (0.19–4.64)	0.95	0.83 (0.16–4.30)	0.82
AEs within 90 d, No. (%)k	25/260 (9.6)	15/245 (6.1)	Risk ratio	1.57 (0.85–2.91)	0.15	1.53 (0.83–2.83)	0.17
All-cause death within 90 d, No. (%)	0/261 (0.0)	1/245 (0.4)	Hazard ratio	NE	1.00	NE	1.00

AE, Adverse event; SAE, Serious adverse event; END, early neurological deterioration; IQR, interquartile range; NE, not estimable; TIA, transient ischemic attack.

^aThe use of mono and dual antiplatelet agents in standard antiplatelet therapy are provided in Appendix p 13. The overall percentage of dual antiplatelet agents was 85.5% (218/255).

^bAdjusted for age, sex, hypertension, diabetes mellitus, and time from onset to randomization.

^cScores on the modified Rankin Scale (mRS) range from 0 to 6, with higher scores indicating more severe disability. The mRS measurement at 7 days was missing for 1 patient.

^dThe END measurement at 7 days was missing for 1 patient.

^eScores on Barthel Index range from 0 to 100, with higher scores indicating better functional independence. The Barthel Index measurements at 90 days were missing for 2 patients. Baseline Barthel Index was included as an additional covariate in both the unadjusted and adjusted win ratio analysis.

^fScores on the National Institutes of Health Stroke Scale (NIHSS) range from 0 to 42, with higher scores indicating a more severe neurological deficit. The measurements of NIHSS were missing for 1 patient at 7 days. Baseline NIHSS score was included as an additional covariate in both unadjusted and adjusted win ratio analysis.

^gStroke included ischemic and haemorrhagic stroke.

^hComposite endpoint included new-onset stroke, myocardial infarction, and all-cause death.

ⁱMajor bleeding was defined according to PLATelet Inhibition and Patient Outcomes (PLATO).^17,18 Data on major bleeding measurements at 90 days were missing for 2 patients.

^jA summary and details of serious adverse events are provided in Appendix p 28. Data on serious adverse events at 90 days were missing for 2 patients.

^kA summary of adverse events is provided in Appendix p 29. Data on adverse events at 90 days were missing for 1 patient.

Fig. 2.

Distribution of mRS scores at 90 days, No. (%). Scores of the modified Rankin Scale (mRS) range from 0 to 6, with 0 indicating no symptoms and 6 indicating death. A score of 1, indicates no clinically meaningful disability; 2, slight disability; 3, moderate disability; 4, moderately severe disability; 5, severe disability. In the standard antiplatelet therapy group, 0.8% of patients had an mRS score of 5 and 0.4% had an mRS score of 6.

Secondary efficacy outcomes

The rate of post-randomization END was 8.2% in the tirofiban group and 6.3% in the standard antiplatelet therapy group (unadjusted RR, 1.30, 95% CI, 0.70–2.44) (Table 2). The GenOR for the ordinal distribution of mRS scores at 90 days was 1.10 (95% CI, 0.84–1.43). Other secondary outcomes including excellent functional outcome at 7 days, NIHSS score, Barthel index score, TIA, ischemic stroke, stroke, and composite event are summarized in Appendix (pp 20–23), with unadjusted and adjusted estimates and corresponding 95% CIs. Similar results were found in the per-protocol analysis (Appendix pp 17–18).

Subgroup, and sensitivity analyses

A prespecified subgroup analysis showed no treatment effect heterogeneity between the tirofiban and standard antiplatelet therapy groups across six baseline characteristics (Fig. 3; Appendix p 24). In sensitivity analyses, both the complete-case and multiple-imputation analyses yielded an unadjusted RR of 1.01 (95% CI, 0.90–1.13) and an adjusted RR of 1.00 (95% CI, 0.89–1.12), consistent with the primary analysis (Appendix p 25). In a post-hoc exploratory analysis according to antitreatment received during the randomized 48-h period, the P value for interaction was 0.06 after adjustment (Appendix p 26).

Fig. 3.

Subgroup Analysis of Modified Rankin Scale Score of 0 or 1 at 90 Days. ∗ Scores on the National Institutes of Health Stroke Scale (NIHSS) range from 0 to 42, with higher scores indicating more severe neurological deficit.

Adverse events

Major bleeding occurred in 0 of 259 patients in the tirofiban-exposed safety population and 1 of 245 (0.4%) in standard-antiplatelet-exposed safety population (Table 2). No intracranial hemorrhage occurred in the tirofiban-exposed safety population. All-cause mortality occurred in 1 of 245 patients (0.4%) in the standard-antiplatelet-exposed safety population and 0 of 261 (0%) in the tirofiban-exposed safety population (Table 2; Appendix p 27). The rate of patients with adverse events, and serious adverse events were similar between the two groups (Table 2; Appendix pp 28–29). Laboratory safety outcomes were exploratory and had substantial outcome-specific missingness (Appendix pp 30–31). Major safety outcomes were similar in a supplementary analysis according to mITT population (Appendix p 32).

Discussion

Among patients with acute BAD-related stroke, intravenous tirofiban administered within 48 h of stroke onset did not result in a significant improvement in functional outcomes at 90 days compared with standard antiplatelet therapy. Major bleeding was rare (0 vs 1 event among participants with available safety assessments), which precludes a precise comparative safety conclusion.

Previous studies evaluating tirofiban in acute ischemic stroke populations have yielded mixed results.^11,13,19 For example, the RESCUE BT2 trial, which focused on patients with ischemic stroke without medium or large vessel occlusion, included 73% with penetrating artery disease, along with other stroke etiologies such as artery-to-artery embolism. The overall cohort had a higher median baseline NIHSS score than our study (9 vs 4, respectively). This study found that early tirofiban improved the proportion of patients achieving excellent functional outcomes at 90 days (from 22.2% to 29.1%).¹³ However, it remains unclear whether patients with BAD-related stroke benefited from this treatment. Given the clinical significance of BAD-related stroke, several observational studies have explored the potential benefit of tirofiban on functional outcomes. However, the results remain inconsistent.^5,20 Our study add evidence specific to BAD-related stroke, while showing that superiority was not demonstrated for the regimen and timing tested here.

Similarly, the TREND trial found that tirofiban reduced the incidence of END in patients with non-cardioembolic ischemic strokes.¹² However, 58% of participants in TREND had large artery atherosclerosis or unknown stroke etiology, which may have contributed to the positive results in their trial. In contrast, our study did not show a clear reduction in END with tirofiban, highlighting differences in treatment effects across stroke subtypes.¹² Additionally, the STRATEGY trial, which focused on BAD-related stroke, reported that tirofiban failed to reduce the incidence of END (14.6% vs 16.1%). The percentage of mRS score of 0–1 at 90 days was 70.3% vs 64.7% in the tirofiban plus aspirin and aspirin-only groups, respectively.²¹ Notably, our control group had a higher proportion of patients receiving dual antiplatelet therapy (218/255, 85.5%), which may have contributed to better functional outcomes compared with STRATEGY.⁹ However, this high background use of DAPT might attenuate any incremental benefit of tirofiban, as the control group may have already achieved substantial platelet inhibition, making it harder to detect a treatment effect. The post-hoc treatment-received analysis cannot resolve this question because single antiplatelet therapy (SAPT) vs DAPT was not randomized and the SAPT subgroup was small.

While observational studies have suggested that tirofiban may improve outcomes in strokes caused by atherosclerosis,^22,23 these findings have been challenged by RESCUE BT trial and BAD-study.^5,19 Tirofiban before endovascular thrombectomy failed to improve outcomes in large artery atherosclerotic stroke, aligning with our findings in BAD-related stroke.¹⁹ The heterogeneous results of prior tirofiban trials suggest that its efficacy may be contingent on stroke etiology and pathophysiology. The treatment strategy tested here differs from tirofiban-based approaches in which oral DAPT or SAPT overlaps with the tirofiban infusion. Consequently, the present findings should not be interpreted as evidence that tirofiban lacks efficacy across all BAD-related stroke populations, treatment timings, or concomitant antiplatelet regimens.

Interestingly, the recurrence rate of stroke in our cohort was relatively low (9/513, 1.8%), consistent with other BAD cohorts.^7,24 This may be explained by the unique pathophysiology of BAD-related stroke, primarily driven by atherosclerotic plaques at perforator artery origins without significant parent artery stenosis.⁹ These findings align with previous studies reporting lower recurrence rates in minor stroke.²⁵ The overall END rate in our cohort was 15.8% (81/513, comprising 44 events before and 37 after randomization), higher than in other mild to moderate stroke cohorts (4.8%–13.2%), and END was a strong predictor of 90-day disability.^5,12,26 However, the exploratory analysis did not establish whether tirofiban reduced END risk in our cohort. These results suggest that effective strategies to reduce END in BAD-related stroke remain to be explored. Initiation within 24 h may improve prognosis; however, this requires further validation in BAD-related stroke.^12,27

With regards to safety events, the incidence of bleeding events was lower than that in previous tirofiban trials, and no intracranial hemorrhage occurred. This may be attributed to the relatively low median NIHSS score in our cohort, and an age restriction of 75 years or younger in our trial.^12,13,28 However, the very small number of major bleeding events yields limited precision and should not be interpreted as establishing equivalent or comparable safety between treatment strategies.

Our study has several limitations. First, although the trial was open-label, outcome assessment, including the primary endpoint and adverse events, was performed by blinded evaluators to minimize potential bias. Second, more than half of the END occurred before randomization, limiting the potential impact of tirofiban. The 48-h treatment window may not have been optimal for assessing tirofiban's full potential, particularly in BAD-related stroke, where earlier and prolonged intervention may be more effective.¹² Whether tirofiban reduced the incidence of END after randomization remained uncertain, given the exploratory nature of this secondary analysis.¹⁰ Additionally, requiring MRI within 48 h ensured consistent diagnosis but may have excluded patients without access to MRI, limiting generalizability, especially in resource-limited settings. Lastly, our study excluded patients with parent artery stenosis greater than 50%, as these patients are typically classified under large vessel atherosclerosis. However, some of these patients may have had mixed pathology involving BAD, and further studies are needed to explore whether tirofiban could be beneficial in this population.

Conclusions

Among patients with acute BAD-related stroke enrolled within 48 h of onset, most of whom had mild stroke severity, the intravenous tirofiban regimen without concomitant oral antiplatelet therapy, did not demonstrate superiority over standard antiplatelet therapy for achieving an excellent functional outcome at 90 days. The findings do not establish equivalence and should not be generalized to substantially earlier initiation or to other tirofiban-based antiplatelet strategies.

Contributors

SL and HH contributed equally as joint first authors. SL, DW, TN, and JN conceived and designed the study. SL, HH, DW, and JN drafted the initial manuscript. SL, HH, BL, MZ, LHW, QY, PZ, WZ, FZ, YT, XLW, HL, LC, GC, RW, AD, ZZ, XNW, QL, JZ, JW, LW, XL, HC, JM, YZ, LZ, and YS were involved in acquisition of data. SL, LZ, YZ, BP, TN, and JN critically revised the manuscript for important intellectual content. DW, and DZ conducted statistical analyses. SL, HH, DW, DZ, and JN contributed to data interpretation. SL, DW, TN, and JN supervised the study. DW and DZ accessed and verified the data. SL, HH, DW, DZ, and JN had access to the raw data. All authors had reviewed the manuscript and approved the submission for publication. JN had final responsibility for the decision to submit the manuscript for publication. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.

Data sharing statement

Request regarding the de-identified data of this study should be made to Prof. Jun Ni at pumchnijun@163.com upon reasonable request.

Declaration of interests

T. Nguyen reports Associate Editor of stroke, Advisory board for Bayer, Route 92, Medtronic. No other disclosures were reported.