Acute ischemic stroke (AIS) is a major global health concern associated with significant morbidity, mortality, and long-term disability (1, 2). Rapid and effective reperfusion therapy is crucial for improving clinical outcomes in cases of large vessel occlusion (LVO), a prevalent and severe subtype. Endovascular thrombectomy (EVT) is the standard treatment for eligible patients with large vessel occlusion (LVO), offering better functional recovery than medical therapy alone (3–6).

Intravenous thrombolysis with alteplase continues to be commonly used within 4.5 h of symptom onset in AIS patients (7–11), however, its limited efficacy in achieving recanalization in large vessel occlusions has prompted growing interest in combination strategies. Intravenous thrombolysis followed by thrombectomy has been proposed to enhance early reperfusion, yet its incremental benefit over direct thrombectomy remains uncertain. Several randomized trials have reported inconsistent results regarding whether intravenous alteplase improves outcomes when administered prior to endovascular therapy (12–16). Notably, two studies reported no significant advantage of combining alteplase with EVT compared to EVT alone (13, 16), while three studies suggested equivalence between the two modalities (12, 14, 17). Tenecteplase has recently emerged as a promising alternative (18–24). Preliminary clinical trials have demonstrated encouraging signals, suggesting that tenecteplase may achieve higher rates of early reperfusion and improved functional outcomes in the pre-thrombectomy setting when compared to alteplase (25). However, these randomized trials were conducted in different settings, included patients with non-identical occlusion locations (two trials enrolling both anterior and posterior circulation strokes (25, 26), whereas the others focused on anterior circulation LVO (12–16)), and did not use fully uniform imaging or thrombolytic protocols (e.g., CT/CTA versus perfusion-based selection and low-dose alteplase in SKIP), which may introduce some heterogeneity into the available evidence.

Despite the growing body of literature evaluating these interventions, head-to-head comparisons among thrombectomy alone, plus alteplase with EVT, and tenecteplase with EVT remain limited. A network meta-analysis (NMA) is thus warranted to synthesize both direct and indirect evidence and to provide a comparative framework across multiple treatment arms. This methodology enables a more comprehensive assessment of the relative efficacy and safety of available reperfusion strategies.

This study aims to conduct a network meta-analysis to compare three reperfusion strategies for patients with AIS due to LVO: EVT alone, alteplase with EVT, and tenecteplase with EVT. The findings aim to clarify the comparative effectiveness of these approaches and support clinical decision-making in acute stroke treatment.

This network meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Network Meta-Analyses (PRISMA-NMA) guidelines (26). The study protocol was prospectively registered with PROSPERO (registration number: CRD420251073350).

A total of 3,609 studies were initially identified through the search. A total of 2,773 unique titles and abstracts were screened after eliminating duplicates. Subsequently, 119 articles underwent full-text evaluation, with reasons for exclusion detailed in Figure 1. Ultimately, eight articles comprising seven randomized controlled trials (RCTs) were deemed eligible for data extraction (Figure 1). The DIRECT-SAFE trial was excluded because it did not differentiate between bridging therapy with alteplase or tenecteplase (17). All included studies exhibited a high risk of selection bias owing to inadequate allocation concealment (Figure 1).

We found 5 RCTs comparing EVT and alteplase with EVT, 1 comparing EVT and tenecteplase with EVT, 1 comparing the tenecteplase with EVT and alteplase with EVT. We analyzed data from 2,793 individuals: 1290 received EVT, 379 received tenecteplase with EVT, and 1,124 received alteplase with EVT. The characteristics of the included studies are summarized in the appendix (Table 1). Figure 2 illustrates the network diagram. The appendix (Supplementary Figure 2) provides the results of the heterogeneity test. Heterogeneity was absent for all outcomes except mortality, where significant heterogeneity was observed in the comparisons of tenecteplase with EVT versus EVT (I2 = 73.0%) and EVT plus alteplase versus EVT (I2 = 70.2%). The appendix (Supplementary Figure 3) provides the details of the consistency evaluation by node-splitting. DIC comparisons indicated negligible differences in fit between FE and RE consistency models; thus, RE model results are presented for all analyses (Supplementary Table 3).

In this network meta-analysis, we observed that bridging therapy using intravenous tenecteplase followed by EVT was associated with a significantly higher rate of functional independence (mRS 0–2) at 90 days compared with EVT alone (OR: 1.52; 95% CrI: 1.00–2.36). Tenecteplase bridging therapy also showed a tendency toward greater efficacy compared to alteplase bridging, though this advantage did not reach statistical significance. Importantly, no significant differences were found among the three treatment strategies regarding excellent outcomes (mRS 0–1), sICH, or mortality. These comparative estimates for tenecteplase should, however, be interpreted with caution given the relatively small number of tenecteplase-treated patients and the between-trial differences in design and enrolled populations.

Our findings are broadly consistent with emerging evidence suggesting that tenecteplase may have certain advantages over alteplase in AIS treatment strategies involving bridging therapy, including improved pharmacokinetics, higher fibrin specificity, and ease of administration via single bolus (25, 33). The SUCRA values for tenecteplase combined with EVT (0.97 for mRS 0–2 and 0.96 for mRS 0–1) highlight its potential clinical benefits, but these findings should be interpreted cautiously due to the limited number of direct comparative studies and the reliance on indirect comparisons within our network meta-analysis.

Consistent with our results, recent studies have provided mixed outcomes concerning bridging therapy versus EVT alone. For instance, the DIRECT-MT and DEVT trials demonstrate non-inferiority of EVT alone compared to EVT with alteplase (13, 16). Similarly, the randomized controlled trial conducted by LeCouffe et al. which enrolled 539 European patients with anterior circulation LVO, reported no significant difference in 90-day disability outcomes or rates of sICH between EVT alone and alteplase plus EVT (15). Interestingly, while some studies reported a higher reperfusion rate in the alteplase-bridging group compared to ECT alone, this did not always translate into better clinical outcomes (12). A recent meta-analysis of six studies also failed to establish the non-inferiority of EVT alone compared with combined thrombolysis and EVT (34). Likewise, an individual patient data meta-analysis found no improvement in functional outcomes with bridging therapy (alteplase with EVT) in patients with carotid tandem lesions, reinforcing the ongoing uncertainty regarding the role of intravenous thrombolysis before EVT (35). Although combined intravenous thrombolysis and mechanical thrombectomy could theoretically be disadvantaged by a delay in initiating thrombectomy due to the preparation of thrombolytic therapy, the randomization-to-puncture times did not differ significantly between the two groups (14). Moreover, these mixed results arise from trials that differ in baseline infarct core size, collateral circulation, clot location and workflow times, factors that could not be standardised or fully accounted for in our aggregate-level analysis.

Importantly, our findings support the growing interest in tenecteplase as a thrombolytic agent in AIS. A meta-analysis of 38 observational studies found that combining intravenous thrombolysis with EVT increased the likelihood of achieving functional independence at 3 months compared to EVT alone (17). Although only one RCT in our analysis directly compared tenecteplase with alteplase before EVT, it suggested a trend toward better outcomes with tenecteplase (33). In addition, a recent RCT demonstrated that patients received intravenous tenecteplase was higher rate of functional independence at 90 days compared with EVT alone (25). The total number of tenecteplase-treated patients across these trials remains modest, and the current evidence base remains limited, emphasizing the need for larger, rigorously designed head-to-head trials to clarify tenecteplase’s role in bridging therapy.

Safety remains a critical consideration in thrombolytic strategies. Although no significant differences in sICH were identified, our SUCRA analysis indicated a slightly elevated hemorrhage risk with tenecteplase bridging compared to EVT alone. While this difference was not statistically significant, clinicians should remain vigilant about hemorrhagic complications, especially in high-risk populations. It should be noted that definitions of sICH were not fully uniform across trials. For example, DEVT (13) and Qiu et al. (25)^. defined sICH as symptomatic intracerebral hemorrhage occurring within 48 h, whereas Campbell et al. (33). used narrower time windows for ascertainment (e.g., within 36 h after treatment), which may have influenced the comparability of pooled safety estimates.

This study possesses several strengths. First, we included only randomized controlled trials to ensure methodological rigor and minimize bias. The target population was carefully selected—patients with LVO within 4.5 h of symptom onset and meeting both thrombolysis and thrombectomy eligibility criteria. Despite the limited number of studies directly comparing EVT alone, EVT plus tenecteplase, and EVT plus alteplase, the network meta-analytic approach enabled integration of both direct and indirect comparisons, yielding consistent and robust results.

However, it is important to recognize certain limitations. The number of available RCTs evaluating tenecteplase was small, which may affect the stability of comparative estimates. Moreover, SKIP (14) investigated alteplase at a reduced dose of 0.6 mg/kg, whereas the other included trials used the standard 0.9 mg/kg regimen (12, 13, 15, 16), which may limit the extrapolation of our alteplase estimates beyond the specific dosing schedules evaluated. Furthermore, heterogeneity in enrolled populations—particularly regarding occlusion site (anterior vs. posterior circulation) and ethnicity (e.g., Asian vs. European cohorts)—may limit the generalizability of our findings, and the trials also differed in imaging strategies (e.g., CT/CTA or MRA), thrombectomy techniques and devices, timing of EVT, and non-uniform definitions of sICH. All included studies were open-label, which may introduce performance and detection biases. Additionally, individual patient-level data were unavailable, restricting the possibility of subgroup or adjusted analyses. Several treatment contrasts, particularly those involving tenecteplase, relied largely on indirect evidence, so the assumptions of transitivity and consistency cannot be fully verified and the precision of treatment rankings is limited. Finally, although we performed an extensive literature search, some relevant unpublished or gray literature may have been missed, and evaluation of publication bias in network meta-analysis remains challenging due to the limited number of comparisons per treatment dyad.

Our network meta-analysis indicates that tenecteplase bridging therapy before EVT could lead to better functional outcomes than EVT alone and may be more effective than alteplase bridging therapy for AIS patients with LVO within 4.5 h of symptom onset. Although promising, these findings require confirmation through large-scale, prospective randomized trials before tenecteplase can be definitively recommended as the preferred thrombolytic agent in bridging therapy.

Our network meta-analysis suggests that tenecteplase bridging therapy before EVT may be associated with improved 90-day functional independence in AIS patients with LVO within 4.5 h of symptom onset, compared to EVT alone and potentially surpassing alteplase bridging therapy. However, these signals are based on a limited number of tenecteplase-treated patients, borderline credible intervals and predominantly indirect comparisons, and should therefore be interpreted with caution. Additional large-scale randomized trials are necessary to confirm these findings and develop conclusive treatment guidelines.