Exploring the Efficacy and Safety of Argatroban as an Adjunct to Antiplatelet Therapy in Ischemic Stroke Patients: A Systematic Review and Meta-analysis

Abstract

Acute ischemic stroke (AIS) is a major cause of disability and mortality worldwide. While antiplatelet therapy is standard for secondary prevention, many patients still experience early neurological deterioration (END). Argatroban, a direct thrombin inhibitor, can potentially limit thrombus progression and improve AIS’s functional outcomes. This meta-analysis assessed the efficacy and safety of argatroban in combination with single (SAPT) or dual antiplatelet therapy (DAPT) compared to antiplatelets alone. Following PRISMA guidelines, a systematic search of PubMed, Scopus, and Web of Science was conducted until January 2025. Randomized controlled trials (RCTs) and cohort studies evaluating argatroban plus antiplatelets versus antiplatelets alone in AIS patients were included. The primary outcome was a 90-day modified Rankin Score (mRS) of 0–2. Secondary outcomes included mRS 0–1 and mRS 3–5 at 90 days, END, and National Institutes of Health Stroke Scale (NIHSS) improvement, stroke recurrence, intracranial hemorrhage (ICH), symptomatic intracranial hemorrhage (sICH), and mortality. We used the mean difference (MD) for continuous variables and odds ratio (OR) for dichotomous ones at 95% confidence intervals (CI) and a P-value of 0.05. A total of 14 studies (four RCTs and 10 cohort studies) were included. Compared to antiplatelets alone, argatroban significantly improved functional outcomes, increasing the incidence of mRS 0–2 (OR = 1.36 [95%CI: 1.05, 1.76, P = 0.02]) and mRS 0–1 (OR = 1.54 [95%CI: 1.08, 2.2, P = 0.02]) while reducing END (OR = 0.42 [95%CI: 0.21, 0.85, P = 0.02]). Argatroban was also associated with greater NIHSS score improvement (MD =  − 0.52 [95%CI: − 0.89, − 0.15, P = 0.005]). No significant differences were observed in mRS 3–5, stroke recurrence, ICH, sICH, or mortality. Subgroup analysis indicated that argatroban combined with DAPT showed the greatest benefits. Argatroban combined with antiplatelet therapy improves functional recovery and reduces END without increasing bleeding risks. These findings support its use, particularly with DAPT, in mild to moderate AIS management. Further large-scale RCTs are needed to optimize dosing strategies and patient selection.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12975-025-01357-x.

Introduction

Acute ischemic stroke (AIS) is a prevalent and frequently catastrophic neurological incident, ranking as the second highest cause of mortality globally [1]. Current targeted treatment strategies encompass artery recanalization and antiplatelet therapy. Nonetheless, there is a deficiency of treatment alternatives within the initial 4.5–48 h following the onset of stroke in patients who have not received reperfusion therapy [2].

Early neurological deterioration (END) occurring within 48 h after acute ischemic stroke (AIS) onset is observed in approximately 14% of patients receiving thrombolysis, although reported rates range from 8 to 28% [3]. END consistently predicts unfavorable outcomes [3]. In addition to direct causes like intracranial hemorrhage (ICH) and malignant edema, the mechanism of END mostly remains ambiguous. In practical practice, medical treatment for inexplicable END may include plasma volume expansion, induced hypertension, and enhanced antithrombotic therapy; however, none have been formally validated to date. A majority of individuals with END continue to encounter a substantial deterioration of their illness despite antiplatelet therapy [4]. This may be ascribed to various potential causes, including thrombus extension, hemodynamic impairment, and inadequate response to antiplatelet medication. In theory, the early administration of anticoagulants may limit thrombus expansion in an intracerebral artery, minimizing the volume of infarcted cerebral tissue and lowering the risks of disability and mortality. It is important to highlight that anticoagulants may pose a risk of hemorrhage. Research has demonstrated that while anticoagulants inhibit stroke progression, the advantage of reduced recurrence is counterbalanced by a comparable rise in ICH [5]. Early use of warfarin in cases of transient ischemic attack (TIA) or mild stroke was correlated with an elevated risk of ICH [6]. The American Heart Association/American Stroke Association (AHA/ASA) guidelines do not endorse immediate anticoagulation in patients with AIS to reduce early recurrent stroke, stop neurological deterioration, and enhance outcomes [2].

In recent decades, direct thrombin inhibitors (DTIs) have been created, demonstrating efficacy comparable to warfarin but being associated with markedly fewer bleeding problems [7]. Among the DTIs, Argatroban can be administered intravenously at an early stage, directly inhibiting free and clot-bound thrombin and thrombin-induced activities, including platelet aggregation and endothelin-1 release. It is deemed safe and feasible for treating atherosclerosis-induced AIS [8]. Furthermore, while the phase IIb ARTSS-2 trial [9] suggested that adjunctive argatroban could offer a net clinical benefit when combined with recombinant tissue plasminogen activator (r-tPA) for AIS, the more recent MOST study [10] found no reduction in poststroke disability with adjunctive argatroban and noted increased mortality—thus tempering the initial optimism raised by ARTSS-2.

Therefore, we conducted the current systematic review and meta-analysis to investigate the efficacy and safety of using argatroban with antiplatelet therapy, whether single antiplatelet therapy (SAPT) or dual antiplatelet therapy (DAPT) in AIS patients.

Methods

This systematic review and meta-analysis was carried out in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [11] and the Cochrane Handbook for Systematic Reviews of Interventions [12] (Supplementary file, PRISMA Checklist).

Searching Process

We searched PubMed, Scopus, and Web of Science to find all the relevant articles that could be included in the present systematic review and meta-analysis. This search was conducted from inception till January 2025 using the Medical Subject Heading (MeSH) terms of two keywords: “Argatroban” AND “Stroke”. Searching was done using title and abstract on PubMed, keywords, title, and abstract on Scopus and Web of Science. The resulting articles from the search process were delivered to EndNote to remove duplicates and start the screening process.

Eligibility Criteria and Screening

The articles that resulted after duplicate removal were uploaded to Rayyan for screening. We included studies in our review if they satisfied the following criteria:

1. Population (P): patients with stroke

2. Intervention (I): administration of argatroban with antiplatelets

3. Comparator (C): antiplatelets alone

4. Outcomes (O):

   1. Primary outcomes: modified Rankin Score (mRS) 0–2
   2. Secondary outcomes: mRS 0–1 and mRS 3–5 at 90 days, END, National Institutes of Health Stroke Scale (NIHSS) change from baseline to discharge, ICH, symptomatic intracranial hemorrhage (sICH), and mortality

5. Study design: randomized controlled trials (RCTs) and cohort studies

Studies were divided according to the comparison groups 1, 2, 3, and others. Comparison group 1 included argatroban + SAPT or DAPT compared with SAPT or DAPT, comparison group 2 included argatroban + DAPT or SAPT compared with DAPT, comparison group 3 included argatroban + SAPT compared with SAPT and others which included different antiplatelets (ozagrel and eptifibatide) compared with argatroban.

Data Extraction

Using Microsoft Excel sheets, two authors independently extracted data from the articles included. We extracted the baseline and summary of the included studies, including study design, population, intervention, comparison group, doses, timing of intervention, sample size, age, and gender. Outcomes including mRS 0–1, mRS 0–2, mRS 3–5, END, NIHSS change from baseline, ICH, sICH, and mortality were also extracted. The sICH was characterized as parenchymal hemorrhage, hemorrhagic infarction, or hemorrhagic transformation identified with critical thinking, accompanied by neurologic impairment, defined as an elevation of 4 points in the overall NIHSS score [8, 10, 13].

Quality and Risk of Bias Assessment

Two authors conducted evaluations of quality and risk of bias, reconciling any differences through consensus or conversation with the senior author. The New Castle Ottawa Scale (NOS), which assigns a star rating from 0 to 9 for each study, was utilized to assess the quality of observational cohort studies. Each question may be assigned a rating of one or zero stars, except the comparison question, which may get a maximum of two stars. A study rated 1–3 stars is considered low quality; 4–6 stars reflect intermediate quality; and 7–9 stars denote high quality [14]. We utilized the Cochrane risk-of-bias tool (Rob-2), which consists of five domains, each encompassing a set of questions. The results are then consolidated using a flowchart to determine three bias levels: low risk, some concerns, or high risk [15].

Data Analysis and Synthesis

Statistical analyses were conducted using Review Manager v.5.3. We estimated the mean difference (MD) and 95%CI for the continuous outcome (NIHSS change). If continuous variables were expressed as a median and interquartile range, the mean and standard deviation were computed based on the median, interquartile range, and sample size, as described by Hozo et al. [16]. Similarly, using odds ratio (OR) and 95%CI for dichotomous outcomes of (mRs 0–1, mRs 0–2, mRs 3–5, END, mortality, ICH, sICH and stroke recurrence). We used the random-effects model if heterogeneity was found, as it accommodates a larger standard error in the pooled estimate, making it suitable for inconsistent or controversial estimates. I² was used for heterogeneity assessment as 0–30% was considered low heterogeneity, 30–70% was moderate heterogeneity, and 70–100% was considered high heterogeneity. A P-value of 0.05 was used to assess the statistical significance of heterogeneity. We used pairwise deletion for missing variables. Subgroup analysis was done based on different comparison groups: (1) argatroban plus SAPT or DAPT versus SAPT or DAPT, (2) argatroban plus SAPT or DAPT versus DAPT alone, (3) argatroban plus SAPT versus SAPT alone, and (4 or others) argatroban versus other antiplatelets (ozagrel, eptifibatide). Sensitivity analysis using leave-one-out was used for heterogeneous outcomes by removing a study at a time and observing the effect on heterogeneity.

Results

Search Results and Study Selection

The search process of the databases resulted in a total of 236 articles, including 119 duplicates. After these articles’ title and abstract screening, 18 were eligible for full-text screening. This process resulted in the exclusion of four studies to include the rest of the 14 studies [1, 8, 10, 13, 17–26] in the current systematic review and meta-analysis (Fig. 1).

Fig. 1.

PRISMA flow diagram of searching and screening

Baseline Characteristics

We identified four RCTs and 10 cohort studies. The populations differed in some studies regarding the types of stroke involving mainly ischemic stroke. Four studies included a comparison of argatroban + SAPT or DAPT against SAPT or DAPT alone (comparison group 1) [13, 18–20], five studies included argatroban + SAPT or DAPT compared with DAPT alone (comparison group 2) [17, 21–24], three studies included argatroban + SAPT vs SAPT alone (comparison group 3) [1, 8, 25], and two studies included a comparison between argatroban and other antiplatelets such as ozagrel and eptifibatide [10, 26]. The SAPT included aspirin or clopidogrel and DAPT included aspirin with clopidogrel. The dose of argatroban that was most commonly used was 60 mg/day during the first 2 days and then twice a day (20 mg/day) for 5 days, while that of clopidogrel and aspirin was clopidogrel at an initial dose of 300 mg, followed by 75 mg daily plus 100 mg aspirin daily. Ozagrel was used as an 80 mg drip infusion over 2 h twice daily, in the morning and the evening, while Eptifibatide was used as a bolus dose of 135 µg/kg followed by a 2-h infusion (0.75 µg/kg/min). The timing of interventions varied across the studies, ranging from 3 to 72 h from stroke onset (Table 1).

Table 1.

Baseline characteristics of the included studies

Study ID	Design	Population	Dose		Timing of intervention	Thrombolysis and/or mechanical thrombectomy	Sample size		Age, mean (SD)		Male, n (%)		Early neurological deterioration definition
			Argatroban	Control			Argatroban	Control	Argatroban	Control	Argatroban	Control
Oguro 2018 [26]	Retrospective cohort	Acute ischemic stroke	Argatroban 120 mg/day infusion over 2 h twice daily in the morning and the evening for 2 days, followed by 20 mg/d for 5 days	Ozagrel (80 mg) drips over 2 h twice daily, in the morning and the evening	Within 48 h of stroke onset	No	353	160	76 (66–83)	74 (65–82)	210 (59.5)	99 (61.9)	NR
Wada 2016 [19]	Retrospective cohort	Ischemic stroke	Argatroban 60 mg/day for the first 2 days, followed by twice daily infusion, at a dose of 10 mg over 3 h, for 5 days	NR	Within 24 h of stroke onset	No	2289	2289	40- ≥ 85		1457 (63.7)	1443 (63)	NR
Wang 2021 [17]	Cohort	Acute non-lacunar single subcortical infarct associated with mild intracranial atherosclerosis	Argatroban 60 mg/day during the first 2 days, and then twice a day (20 mg/day) for 5 days	Aspirin (100 mg/day) and clopidogrel were given at an initial dose of 300 mg, followed by 7 5 mg/day for 3 weeks	Within 6–12 h of stroke onset	No	136	168	63 (6.4)	62 (7.8)	73 (59.3)	96(57.1)	1)An increase of ≥ 2 points in the total NIHSS score, 2) An increase in the motor score (5a–6b) of ≥ 1 point, or 3) Any new neurological deficit (including deficits that were not measurable by the NIHSS score)
LaMonte 2004 [8]	RCT	Acute ischemic stroke	Argatroban 100 µg/kg bolus over 3 to 5 min, followed by continuous infusion at 3 µg/kg/min for 5 days	Aspirin 81 to 325 mg	Within 12 h of stroke onset	No	59	54	70 (13)	65 (13)	28 (47)	23 (42.6)	NR
Jin 2024 [24]	Retrospective Cohort	Minor ischemic stroke	Argatroban 60 mg/day infusion for 48 h, followed by 10 mg infusion over 3 h twice daily for 5 days	Aspirin 100 mg/day combined with clopidogrel 75 mg/day, clopidogrel loading dose 300 mg	Within 24 h of stroke onset	No	270	363	68 (61–76)	67 (58–75)	93 (34.4)	117 (32.2)	Incremental increase in the NIHSS score of ≥ 1 point in motor power or ≥ 2 points in the total score within 7 days of admission
Zhou 2020 [23]	Retrospective Cohort	Acute minor posterior circulation ischemic stroke	Argatroban infusion with a dose of 60 mg/day for 2 days, followed by 20 or 30 mg/day for 2 to 5 days	Clopidogrel at an initial dose of 300 mg, followed by 75 mg daily plus 100 mg aspirin daily	Within 48 h of stroke onset	No	34	68	61.7 (9.8)	61.7 (10.9)	25 (73.5)	49 (72.1)	2 points or greater increase in NIHSS score at 7 days after hospitalization, compared with baseline
Li 2022 [22]	Cohort	Acute mild to moderate ischemic stroke with large artery atherosclerosis	100 µg/kg argatroban bolus is administrated over 3 to 5 min, followed by an argatroban infusion of 1.0 µg/kg/min	Clopidogrel at an initial dose of 300 mg, followed by 75 mg daily plus 100 mg aspirin daily	Within 48 h of stroke onset	No	120	529	60.9 (8.5)	61.7 (9.8)	88 (73.3)	380 (71.8)	1 or more increase in NIHSS at 7 days compared with the baseline
Xu 2022 [21]	Cohort	Ischemic stroke	NR	NR	Within 48 h of stroke onset	No	25	55	NR	NR	NR	NR	National Institutes of Health Stroke Scale score ≥ 2 increase within 48 h after admission
Adeoye 2024 [10]	RCT	Ischemic stroke	Argatroban bolus (100 µg/kg) followed by a 12-h infusion (3 µg/kg/min)	Eptifibatide bolus (135 µg/kilogram) followed by a 2-h infusion (0.75 µg/kg/min)	Within 3 h of stroke onset	Yesa	59	227	68 (60–79)	68 (58–79)	31 (53)	109 (48)	NR
Sun 2024 [25]	Retrospective Cohort	Acute ischemic stroke	Argatroban 60 mg/day for a duration of 2 days, followed by a maintenance dose ranging between 20 and 30 mg/day for an additional 2 to 5 days	Aspirin: 100 mg daily, Clopidogrel 75 mg daily	Within 72 h of stroke onset	No	57	56	58.71 (10.36)	58.34 (10.52)	26 (45.61)	27 (48.21)	NR
Yan 2025 [20]	Retrospective Cohort	Acute mild-to-moderate ischemic stroke with large artery atherosclerosis	Argatroban intravenous infusion at a dose of 60 mg/day for the first 2 days, followed by 5 days intravenous infusion at a dose of 10 mg/day over 3 h, twice a day	NR	Within 72 h of stroke onset	No	294	289	68.19 (10.83)	71.21 (11.4)	195 (66.3)	172 (59.5)	Increase of 2 or more National Institute of Health Stroke Scale (NIHSS) compared to the best neurological status after stroke within 7 days
Zhang 2024 [13]	RCT	Acute ischemic stroke with early neurological deterioration	Argatroban continuous infusion at a dose of 60 mg/day for 2 days, followed by 20 mg/day for 5 days	NR	Within 48 h of stroke onset	Yesb	302	307	66 (57–74)	66 (56–74)	191 (63.2)	195 (63.5)	Increase of 2 or more points on the National Institutes of Health Stroke Scale within 48 h from symptom onset
Liu 2020 [18]	RCT	Acute paraventricular ischemic stroke	Argatroban 10 mg twice a day for 7 days consecutively	NR	Within 48 h of stroke onset	No	30	30	57.7 (8.8)	57.1 (10.7)	23 (77)	24 (80)	NR
Chen 2018 [1]	Retrospective Cohort	Acute ischemic stroke	Argatroban loading dose of 60 mg daily for 2 days, followed by 20 mg daily	Aspirin 300 mg daily	NR	No	434	1051	65.54 (10.83)	65.14 (11.17)	283 (65.21)	710 (67.55)	NR

RCT randomized controlled trial, NR not reported, SD standard deviation, mg milligrams, kg kilograms

^aAll the patients received intravenous thrombolysis (70% received alteplase, and 30% received tenecteplase), and 225 patients (44%) underwent endovascular thrombectomy

^bIn the argatroban group, 17.5% of patients received intravenous thrombolysis treatment, and 1% underwent endovascular treatment. In the control group, 20.2% received intravenous thrombolysis, and 1.3% underwent endovascular treatment

Quality and Risk of Bias Assessment

According to Rob-2, the four included RCTs were reasoned to have a low risk of bias according to all five domains (Supplementary Fig. 1). According to NOS, the ten included cohort studies were observed to have high quality (Table 2).

Table 2.

Quality assessment of cohort studies using NOS

Study name	The level of representation of the affected cohort (★)	Identification of the unexposed cohort (★)	Determination of exposure (★)	Evidence that the outcome of interest was absent at the commencement of the research (★)	Comparison of cohorts based on design or assessment (max★★)	Was the follow-up duration sufficient for consequences to manifest? (★)	Evaluation of results (★)	Assessment of cohort follow-up sufficiency (★)	Quality level
Oguro 2018 [26]	★	★	★	★	★★	-	★	★	High
Wada 2016 [19]	★	★	★	★	★★	★	★	-	High
Wang 2021 [17]	★	★	★	★	★★	★	★	★	High
Jin 2024 [24]	★	★	★	★	★★	★	-	★	High
Zhou 2020 [23]	★	-	★	★	★★	★	★	-	High
Li 2022 [22]	★	★	★	★	★★	★	★	★	High
Xu 2022 [21]	★	-	★	★	★★	★	-	★	High
Sun 2024 [25]	★	-	★	★	★★	-	★	★	High
Yan 2025 [20]	★	★	★	★	★★	★	★	-	High
Chen 2018 [1]	★	-	★	★	★★	★	-	★	High

Meta-analysis

Primary Outcome: Modified Rankin Scale (mRS) 0–2

The use of argatroban with antiplatelets was also associated with increased incidence of mRS 0–2 compared with the control group with OR = 1.36 (95%CI: 1.05, 1.76, P = 0.02) and I² = 64%, P = 0.005 (Fig. 2a). Sensitivity analysis using leave-one-out showed that Wada 2016 [19] was the cause of heterogeneity, which resolved after its removal due to its large weight (largest sample size) (Supplementary Fig. 2).

Fig. 2.

Comparison between argatroban and control in the incidence of mRS 0–2 (a) and subgroup analysis by different comparisons (b)

By subgroup analysis, in the comparison groups 1 and 3, no significant differences were observed between the argatroban group and control group, while in comparison group 2, the argatroban group showed an increased incidence of mRS 0–2 with OR = 1.57 (95%CI: 1.17, 2.12) and I² = 0%. Also, other antiplatelets (ozagrel and eptifibatide) showed higher mRS 0–2 incidence compared with the argatroban group with OR = 0.65 (95%CI: 0.47, 0.89) and I² = 0% (Fig. 2b).

Secondary Outcomes

mRS 0–1

The use of argatroban with antiplatelets was significantly associated with increased incidence of mRS 0–1 compared with the control group with OR = 1.54 (95%CI: 1.08, 2.2, P = 0.02), I² = 80%, P < 0.0001. By subgrouping according to the comparison groups, no significant difference was observed between argatroban and control in comparison group 1, while argatroban showed an increased incidence of mRS 0–1 compared with the control group in comparison group 2 with OR = 1.19 (95%CI: 1, 3.83, P = 0.05) and I² = 81%, P = 0.001. In the “others” comparison group, the use of other antiplatelets (ozagrel and eptifibatide) was associated with increased incidence of mRS 0–1 compared with the argatroban group with OR = 0.64 (95%CI: 0.46, 0.89, P = 0.008), and I² = 0% (Supplementary Fig. 3).

mRS 3–5

The use of argatroban was associated with decreased risk of mRS 3–5 compared with the control group with OR = 0.76 (95%CI: 0.58, 1.01, P = 0.06); however, this difference was not statistically significant, and the heterogeneity was significant with I² = 61%, P = 0.03 (Supplementary Fig. 4a). However, by the sensitivity analysis using leave-one-out of Wada’s 2016 [19] study (which measured the mRS 3–5 at discharge), the results became statistically significant, and the heterogeneity was resolved and became insignificant (Supplementary Fig. 4c). A complete sensitivity analysis of mRs outcomes is shown in Table 3.

Table 3.

Sensitivity analysis using the leave-one-study-out method for mRS scores 0–1, 0–2, and 3–5 at 90 days

Study removed	Odds ratio (95% Cl)	n, Argatroban/control	df	Test for overall effect	Heterogeneity
mRS 0–1
Total (mRS 0–1)	1.54 [1.08, 2.20]	3105/3581	6	Z = 2.37 (P = 0.02)	Tau2 = 0.15; Chi2 = 30.09, df = 6 (P < 0.0001); I2 = 80%
Jin (2024)	1.63 [1.06, 2.51]	2898/3374	5	Z = 2.24 (P = 0.03)	Tau2 = 0.19; Chi2 = 29.71, df = 5 (P < 0.0001); I2 = 83%
Li (2022)	1.69 [1.09, 2.62]	2985/3052	5	Z = 2.37 (P = 0.02)	Tau2 = 0.20; Chi2 = 30.08, df = 5 (P < 0.0001); I2 = 83%
Liu (2020)	1.49 [1.03, 2.16]	3075/3551	5	Z = 2.10 (P = 0.04)	Tau2 = 0.15; Chi2 = 28.24, df = 5 (P < 0.0001); I2 = 82%
Wada (2016)a	1.75 [1.15, 2.67]	816/1292	5	Z = 2.59 (P = 0.010)	Tau2 = 0.17; Chi2 = 17.33, df = 5 (P = 0.004); I2 = 71%
Wang (2021)	1.25 [0.96, 1.62]	2969/3413	5	Z = 1.63 (P = 0.10)	Tau2 = 0.05; Chi2 = 11.95, df = 5 (P = 0.04); I2 = 58%
Xu (2022)	1.43 [1.03, 2.00]	3080/3526	5	Z = 2.13 (P = 0.03)	Tau2 = 0.12; Chi2 = 24.07, df = 5 (P = 0.0002); I2 = 79%
Zhang (2024)	1.65 [1.04, 2.60]	2807/3278	5	Z = 2.13 (P = 0.03)	Tau2 = 0.22; Chi2 = 28.79, df = 5 (P < 0.0001); I2 = 83%
mRS 0–2
Total (mRS 0–2)	1.36 [1.05, 1.76]	3450/3955	8	Z = 2.31 (P = 0.02)	Tau2 = 0.08; Chi2 = 22.04, df = 8 (P = 0.005); I2 = 64%
Jin (2024)	1.38 [1.03, 1.85]	3243/3748	7	Z = 2.13 (P = 0.03)	Tau2 = 0.09; Chi2 = 21.59, df = 7 (P = 0.003); I2 = 68%
LaMonte (2004)	1.41 [1.07, 1.86]	3403/3908	7	Z = 2.44 (P = 0.01)	Tau2 = 0.08; Chi2 = 21.66, df = 7 (P = 0.003); I2 = 68%
Li (2022)	1.31 [1.00, 1.71]	3330/3426	7	Z = 1.96 (P = 0.05)	Tau2 = 0.07; Chi2 = 18.80, df = 7 (P = 0.009); I2 = 63%
Wada (2016)a	1.48 [1.21, 1.82]	1161/1666	7	Z = 3.81 (P = 0.0001)	Tau2 = 0.01; Chi2 = 7.68, df = 7 (P = 0.36); I2 = 9%
Wang (2021)	1.31 [1.00, 1.71]	3314/3787	7	Z = 1.96 (P = 0.05)	Tau2 = 0.07; Chi2 = 18.90, df = 7 (P = 0.009); I2 = 63%
Xu (2022)	1.33 [1.03, 1.72]	3425/3900	7	Z = 2.19 (P = 0.03)	Tau2 = 0.07; Chi2 = 20.21, df = 7 (P = 0.005); I2 = 65%
Yan (2025)	1.24 [0.98, 1.57]	3156/3666	7	Z = 1.80 (P = 0.07)	Tau2 = 0.04; Chi2 = 13.14, df = 7 (P = 0.07); I2 = 47%
Zhang (2024)	1.40 [1.01, 1.92]	3152/3652	7	Z = 2.04 (P = 0.04)	Tau2 = 0.11; Chi2 = 21.25, df = 7 (P = 0.003); I2 = 67%
Zhou (2020)	1.37 [1.04, 1.81]	3416/3887	7	Z = 2.26 (P = 0.02)	Tau2 = 0.09; Chi2 = 22.02, df = 7 (P = 0.003); I2 = 68%
mRS 3–5
Total (mRS 3–5)	0.76 [0.58, 1.01]	3075/3551	5	Z = 1.90 (P = 0.06)	Tau2 = 0.06; Chi2 = 12.81, df = 5 (P = 0.03); I2 = 61%
Jin (2024)	0.74 [0.53, 1.04]	2868/3344	4	Z = 1.73 (P = 0.08)	Tau2 = 0.08; Chi2 = 12.13, df = 4 (P = 0.02); I2 = 67%
Li (2022)	0.82 [0.62, 1.08]	2955/3022	4	Z = 1.44 (P = 0.15)	Tau2 = 0.05; Chi2 = 9.09, df = 4 (P = 0.06); I2 = 56%
Wada (2016)a	0.70 [0.56, 0.88]	786/1262	4	Z = 3.11 (P = 0.002)	Tau2 = 0.00; Chi2 = 3.54, df = 4 (P = 0.47); I2 = 0%
Wang (2021)	0.82 [0.62, 1.07]	2939/3383	4	Z = 1.45 (P = 0.15)	Tau2 = 0.05; Chi2 = 9.21, df = 4 (P = 0.06); I2 = 57%
Xu (2022)	0.78 [0.60, 1.03]	3050/3496	4	Z = 1.76 (P = 0.08)	Tau2 = 0.05; Chi2 = 10.88, df = 4 (P = 0.03); I2 = 63%
Zhang (2024)	0.72 [0.49, 1.06]	2777/3248	4	Z = 1.68 (P = 0.09)	Tau2 = 0.11; Chi2 = 11.53, df = 4 (P = 0.02); I2 = 65%

^aReported mRS at discharge

By subgroup analysis, the use of argatroban was associated with decreased incidence of mRS 3–5 compared with control in the comparison groups 2 with OR = 0.61 (95%CI: 0.44, 0.83, P = 0.002) and I² = 0%, but no difference was observed in the comparison groups 1 (Supplementary Fig. 4b).

Risk of Early Neurological Deterioration (END)

The use of argatroban was significantly associated with reduced risk of END compared with the control group with OR = 0.42 (95%CI: 0.21, 0.85, P = 0.02) and I² = 76%, P = 0.002 (Fig. 3). This significant heterogeneity was resolved by leave-one-out analysis removing Yan 2025 [20] (the largest weight) (Supplementary Fig. 5).

Fig. 3.

Comparison between argatroban and control in the incidence of early neurological deterioration

National Institutes of Health Stroke Scale (NIHSS)

Argatroban was significantly associated with a more significant decrease in NIHSS compared with the control group with MD =  − 0.52 (95%CI: − 0.89, − 0.15, P = 0.005) and I² = 72%, P = 0.0007 (Fig. 4a). This significant heterogeneity was partially resolved by leave-one-out of Sun 2024 [25] due to the highly significant change they produced as the patients had high baseline NIHSS compared with other studies (Supplementary Fig. 6). By subgroup analysis, argatroban was associated with reduced NIHSS compared with control in comparison groups 1 and 2 with MD =  − 0.48 (95%CI: − 0.76, − 0.19) and − 0.48 (95%CI: − 0.48, − 0.95, 0), respectively with non-significant heterogeneity. However, no difference was observed in comparison group 1 (Fig. 4b).

Fig. 4.

Comparison between argatroban and control in NIHSS change (a) and subgroup analysis by different comparisons (b)

Stroke Recurrence, Intracranial Hemorrhage (ICH), Symptomatic Intracranial Hemorrhage (sICH), and Mortality

No significant differences were observed between argatroban and the control group in risk of stroke recurrence, ICH, sICH, and mortality with OR = 0.83 (95%CI: 0.61, 1.12), OR = 1.52 (95%CI: 0.94, 2.48), OR = 2.22 (95%CI: 0.72, 6.77) and OR = 1.6 (95%CI: 0.69, 3.75), respectively with non-significant heterogeneity in all of them (Figs. 5, 6, and 7).

Fig. 5.

Comparison between argatroban and control in the risk of stroke recurrence

Fig. 6.

Comparison between argatroban and control in the risk of ICH (a) and sICH (b)

Fig. 7.

Comparison between argatroban and control in the risk of mortality

Discussion

The current systematic review and meta-analysis favored the use of argatroban as an additional therapy to antiplatelets used in patients with ischemic stroke. This was evident by more efficacious outcomes presented in the increased incidence of 90-day mRS 0–1 and mRS 0–2, decreased incidence of 90-day mRS 3–5, END, and decreased NIHSS from baseline in patients taking the combination therapy rather than using antiplatelets alone. This shows the effectiveness of argatroban in the improvement of functional parameters in short and long terms in ischemic stroke patients and preventing the major problem of END.

The fear of adding an anticoagulant was also minimized due to the comparable risk of stroke recurrence, ICH, and sICH shown in patients taking argatroban with antiplatelets and those taking antiplatelets alone. According to the subgroup analysis, the use of argatroban with SAPT or DAPT compared with DAPT alone showed the best efficacy in most of the outcomes when compared with other comparisons, including argatroban with SAPT or DAPT vs SAPT or DAPT or argatroban with SAPT vs SAPT alone. This may also be caused by the larger number of studies comparing argatroban with SAPT or DAPT vs. DAPT alone compared with other comparisons.

Antiplatelet therapy is the primary approach for AIS, particularly in cases of major artery atherosclerosis [27, 28]. Despite being utilized for over 50 years, the application of anticoagulant medication remains contentious. An RCT indicated that early administration of low-molecular-weight heparin (LMWH) does not confer a substantial advantage over aspirin in patients with major atherosclerosis [29]. In the International Stroke Trial, individuals assigned to heparin experienced a markedly reduced incidence of recurrent ischemic strokes within 14 days; however, this benefit is accompanied by an elevated risk of hemorrhage [30]. Significant endeavors were undertaken to identify safer alternatives due to the considerable risks associated with these anticoagulants. An RCT (ARGIS-1) [8] conducted in North America involved patients with AIS within 12 h of onset and NIHSS score of 5 to 12. Participants received high-dose (3 µg/kg/min) and low-dose (1 µg/kg/min) argatroban, demonstrating that both dosages were safe and significantly prolonged activated partial thromboplastin time. Numerous randomized trials demonstrated that anticoagulant medication may diminish the recurrence of stroke, pulmonary embolism, and deep vein thrombosis; nevertheless, its effectiveness could be counterbalanced by a significant occurrence of sICH [27]. Consequently, a targeted antithrombotic approach tailored to the distinct etiology of stroke should be the future focus. Numerous studies indicate that patients with large artery atherosclerotic stroke may derive greater benefit from DAPT [31–33]. The present pooled research indicates that argatroban combined with antiplatelets yields superior functional outcomes in both the short and long term compared to antiplatelets alone. No heightened risk of complications, including bleeding, was identified.

Theoretically, assuming effective management of bleeding risk, increased intensity of antithrombotic medication correlates with reduced recurrence and exacerbation of stroke. We contend that the amalgamation of argatroban and DAPT may constitute an effective antithrombotic approach for ischemic stroke, thereby mitigating the elevated incidence of disability and mortality among stroke patients [34].

A 1997 multi-center, randomized, double-blind, placebo-controlled research [35] in Japan was the inaugural investigation to establish the efficacy and safety of argatroban for treating AIS due to large artery atherosclerosis. Subsequent studies have been conducted on the efficacy of argatroban in AIS. A clinical study [18] demonstrated that argatroban enhanced regional drainage of the basal vein of Rosenthal in the ipsilateral hemisphere of patients suffering from acute paraventricular ischemic stroke due to intracranial large artery atherosclerosis; improved the NIHSS score on day 7 post-stroke; and elevated the percentage of patients achieving mRS scores of 0–1 by day 90.

Furthermore, compared to aspirin administered alone, the combination of aspirin and agatroban can markedly decrease serum fibrinogen and neuropeptide Y levels in individuals with AIS accompanied by END while enhancing neurological function [36]. A larger RCT showed that among patients with AIS with END, treatment with argatroban and antiplatelet medication yielded superior functional outcomes at 90 days in both unadjusted and adjusted analyses [13]. A study indicated that, compared to DAPT alone, the combination of DAPT and argatroban decreased the incidence of END and enhanced functional results in branch atheromatous disease without elevating the bleeding risk [21].

Nevertheless, other research indicates that combining argatroban with antiplatelet medication offers little benefit over antiplatelet monotherapy in enhancing neurological function in individuals with AIS [23, 26]. In patients with moderate AIS, therapy protocols involving argatroban and aspirin may be less effective than high-dose aspirin (300 mg daily) in enhancing NIHSS scores [1]. In patients with AIS treated with intravenous alteplase within 4.5 h of symptom onset, the addition of intravenous argatroban (100 µg/kg for 3 to 5 min within 1 h of alteplase administration, followed by a continuous infusion of 1.0 µg/kg per min for 48 h) did not enhance 90-day outcomes [10, 37].

Limitations

This meta-analysis has several limitations. First, many of the included studies were observational or retrospective, and the small number of RCTs increases the risk of bias. Second, multiple studies primarily focused on patients with mild to moderate stroke, making it difficult to generalize our findings regarding hemorrhage risk and treatment efficacy to more severe cases. Differences in baseline stroke severity could also explain the observed improvements in NIHSS scores, as milder strokes tend to recover more quickly. Third, the large timeframe across included studies and the varied definitions and time points used to assess outcomes introduce considerable heterogeneity, further complicating comparisons; for example, mortality has varied from 7 days in Wada (2016) to 90 days in Adeoye (2024). Also, Oguro (2018) and Wada (2016) reported mRS at discharge, while other results showed it at 90 days; hence, it was necessary to perform a sensitivity analysis for this outcome. Similarly, in reporting ICH, Adeoye (2024) reported any ICH within 36 h, centrally read, while LaMonte (2004) reported it within 30 days. Fifth, more importantly, the precision of hemorrhagic complications may be questioned because Wada (2016) used an administrative database, which may have limitations related to the accuracy and granularity of outcome reporting, while Adeoye (2024) conducted a prospective randomized clinical trial with a prospective collection of outcomes. While the former study offered valuable large-scale data on real-world clinical outcomes, the latter provided high-quality, detailed data with greater precision regarding hemorrhagic complications; still, results for both ICH and sICH were comparable and homogenous. Sixth, including different antiplatelet regimens (single vs. dual) with argatroban adds variability and hinders the ability to draw definitive conclusions about the relative efficacy of these combinations.

Conclusion

In patients with AIS, adding argatroban to single or dual antiplatelet therapy was associated with improved functional outcomes, including higher rates of favorable mRS scores (0–1 and 0–2) at 90 days, reduced early neurological deterioration, and greater NIHSS score improvements compared to antiplatelet therapy alone. Despite concerns regarding hemorrhagic events, our findings—mostly from mild to moderate stroke populations—indicate that argatroban did not significantly increase the risk of intracranial hemorrhage, symptomatic intracranial hemorrhage, stroke recurrence, or mortality.

Subgroup analysis suggests that combining argatroban with dual antiplatelet therapy may offer the most pronounced benefits, particularly in minimizing disability and neurological decline, for patients at higher risk of thrombus formation. However, the limited representation of moderate-to-severe strokes in the available data means these results should be interpreted cautiously for more severe cases.

Future RCT should clearly define the timeframe for combination therapy, include moderate-to-severe stroke populations to assess argatroban’s risk–benefit profile in these cases, and establish optimal dosing strategies, patient selection criteria, and timing of initiation. Future work should also evaluate long-term impacts on stroke recurrence and disability and compare argatroban with alternative anticoagulation regimens.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contribution

All authors reviewed the manuscript.

Data Availability

No datasets were generated or analysed during the current study.

Declarations

Ethics Approval and Consent to Participate

Not applicable.

Consent for Publication

Not applicable.

Competing Interests

The authors declare no competing interests.