Treating Tissue, Not Time: Contemporary Reperfusion Strategies in Ischaemic Stroke

This article reviews contemporary evidence on reperfusion strategies in ischaemic stroke, including intravenous thrombolysis, mechanical thrombectomy, and advanced imaging-based selection, as well as their association with mortality and functional outcomes, drawing on the major randomised controlled trials.

Introduction

Acute ischaemic stroke remains one of the leading causes of death and acquired disability worldwide. Over the past two decades, stroke management has undergone a substantial transformation, shifting from a strategy focused exclusively on intravenous thrombolysis within a limited time window to a more complex model based on advanced imaging selection and the availability of endovascular therapies (Prabhakaran et al. 2026)

The development of mechanical thrombectomy using new-generation devices, together with the systematic use of computed tomography angiography, collateral assessment, and cerebral perfusion imaging, has enabled the identification of patients with viable penumbra beyond traditional temporal limits. This evolution has consolidated the concept of treating tissue, not time, expanding therapeutic opportunities even in patients with unknown onset or extended windows, provided that objective evidence of potentially salvageable tissue is present.

Evolution of Reperfusion in Ischaemic Stroke: From Thrombolysis to Thrombectomy

For many years, intravenous thrombolysis with alteplase represented the only reperfusion therapy with proven efficacy in acute ischaemic stroke, limited by a narrow therapeutic window and a non-negligible risk of haemorrhage. Despite its clinical impact, its ability to achieve effective recanalisation in proximal large-vessel occlusions was consistently suboptimal, contributing to the persistence of severe disability or death in a substantial proportion of patients despite timely treatment. This scenario prompted the development of endovascular strategies aimed at mechanical clot retrieval, with the goal of improving both the rate and speed of reperfusion in proximal occlusions. Earlier large pragmatic trials such as IST-3 (IST-3 Collaborative Group, 2012) contributed to defining the role of thrombolysis across broader patient populations, including older adults and those treated beyond conventional eligibility criteria.

Early endovascular trials failed to demonstrate consistent clinical benefit, partly due to methodological limitations such as inadequate vascular imaging selection, delays to treatment, and the use of first-generation devices with modest recanalisation rates. However, the advent of stent retrievers, combined with routine integration of computed tomography angiography and exclusion criteria for extensive infarction, marked a decisive shift.

The MR CLEAN trial was the first large-scale multicentre study (Berkhemer et al. 2015), to demonstrate that mechanical thrombectomy, added to standard medical management, significantly improves functional outcomes in patients with proximal anterior circulation occlusion treated within six hours. The intervention was associated with a favourable shift in the distribution of the modified Rankin Scale scores at 90 days (shift analysis), as well as an absolute increase in functional independence, without a significant increase in mortality or symptomatic intracranial haemorrhage. This trial consolidated the concept that endovascular reperfusion can translate into measurable clinical benefit when applied to an imaging-selected population within a defined time frame.

Subsequent contemporary trials reinforced and extended these findings, definitively establishing thrombectomy as the standard of care for large-vessel occlusion. ESCAPE introduced a particularly relevant component from a systems management perspective (Goyal et al. 2015): collateral-based selection using multiphase computed tomography angiography and strict optimisation of workflow times. In this trial, thrombectomy not only substantially increased the proportion of patients achieving functional independence, but was also associated with reduced mortality, highlighting that speed and efficiency of care pathways can be decisive determinants of the final benefit of the procedure.

In parallel, EXTEND-IA provided key evidence by demonstrating that perfusion-based selection (Campbell et al. 2015), identifies patients with salvageable tissue in whom thrombectomy achieves higher rates of reperfusion and improved clinical outcomes. Although smaller and terminated early for benefit, this trial was particularly influential in integrating quantitative assessment of ischaemic core and penumbra as eligibility criteria, reinforcing the paradigm that therapeutic indication should be based on tissue viability rather than elapsed time alone.

In SWIFT PRIME, thrombectomy with stent retrievers following intravenous thrombolysis also demonstrated a significant reduction in disability (Saver et al. 2015), with a marked increase in functional independence at 90 days, without a significant increase in symptomatic intracranial haemorrhage. These findings further supported the strategy of combined reperfusion in eligible patients.

Finally, REVASCAT provided additional evidence regarding the efficacy and safety  of thrombectomy (Jovin et al. 2015), up to eight hours after symptom onset in patients with proximal occlusion and no evidence of extensive infarction. A distinctive feature of this trial was its integration within a highly protocolised regional stroke network, which strengthened its external validity and documented consistent clinical benefit in a context closely aligned with routine practice.

Treating Tissue, Not Time

The consolidation of mechanical thrombectomy as standard treatment for large-vessel occlusion in the early hours after stroke onset did not resolve a persistent clinical challenge: a considerable proportion of patients continue to present outside the conventional window for intravenous thrombolysis or with uncertain onset, such as wake-up stroke. Historically, these scenarios were considered beyond therapeutic reach, despite the fact that the pathophysiology of ischaemic stroke suggests that infarct progression is not uniform and depends on individual factors, particularly the status of collateral circulation. This recognition led to the development of a new approach based on the identification of viable brain tissue using advanced imaging, prioritising the presence of salvageable penumbra over chronological time.

One of the trials that marked this conceptual shift was WAKE-UP (Thomalla et al. 2018), which evaluated alteplase thrombolysis in patients with stroke of unknown onset selected using magnetic resonance imaging based on the diffusion-weighted imaging–fluid-attenuated inversion recovery (DWI–FLAIR) mismatch criterion. This pattern is interpreted as an indirect marker of relatively recent ischaemia, allowing inference that the patient may be within a physiological window amenable to reperfusion. The study demonstrated a significant functional benefit at 90 days, albeit accompanied by an increased risk of haemorrhage, confirming that imaging-based selection can extend therapeutic opportunity beyond the traditional time-based model.

Subsequently, EXTEND expanded this approach by using perfusion imaging (computed tomography or magnetic resonance) in patients treated between 4.5 and 9 hours, as well as in wake-up stroke (Ma et al. 2019). The use of quantitative parameters of ischaemic core and penumbra enabled the identification of individuals with potentially salvageable tissue, demonstrating that alteplase can confer functional benefit in an extended window when eligibility is defined by physiological rather than purely chronological criteria. Nevertheless, the study also reinforced that the modest increase in symptomatic intracranial haemorrhage is an unavoidable component of the risk–benefit balance in late thrombolysis, necessitating strict and reproducible selection.

The most decisive advance in expanding ischaemic stroke treatment occurred with trials evaluating mechanical thrombectomy in late windows supported by advanced selection criteria. DEFUSE 3 demonstrated that thrombectomy between 6 and 16 hours substantially improves functional outcomes in patients with a perfusion–core mismatch, consolidating the concept of a tissue window. This study established that, even with prolonged delays from symptom onset, clinical benefit is maintained if objective evidence of viable penumbra and limited core is present (Albers et al. 2018).

The DAWN trial extended the indication for thrombectomy up to 24 hours by selecting patients based on a clinical–imaging mismatch, that is, a neurological deficit disproportionate to the size of the ischaemic core (Nogueira et al. 2018). This design recognised that some patients exhibit slow infarct progression, preserving salvageable tissue for longer periods due to functional collaterals. DAWN confirmed that thrombectomy can translate into marked clinical benefit even in very late windows, provided that patients are selected using strict tissue viability criteria.

Collectively, these trials transformed the organisational model of acute stroke care, as therapeutic decision-making shifted from reliance on rigid time limits to dependence on the system's capacity to perform rapid and standardised vascular and perfusion assessment. In this context, the availability of imaging protocols, reproducible interpretation, and interhospital referral networks became as relevant as the therapy itself.

In parallel, recent years have explored the expansion of intravenous thrombolysis beyond 4.5 hours using perfusion-based criteria, particularly in settings where thrombectomy is unavailable (Zhou et al. 2025). In this regard, HOPE evaluated alteplase in imaging-selected patients between 4.5 and 24 hours, showing that carefully selected populations may derive functional benefit, albeit with increased haemorrhagic events, thereby reinforcing the ongoing need for strict risk–benefit assessment.

More recently, the OPTION trial (Ma et al. 2026), provided relevant evidence in a group historically under-represented in reperfusion studies: patients with non–large-vessel occlusion stroke. Administration of tenecteplase between 4.5 and 24 hours in perfusion-selected individuals increased the likelihood of excellent functional outcome, albeit with a higher rate of symptomatic intracranial haemorrhage. These findings suggest that the extended window could expand even to non-LVO populations when salvageable tissue is documented, opening a new therapeutic field particularly relevant for healthcare systems in which thrombectomy is not feasible for most patients.

Intravenous Thrombolysis

Although alteplase has historically been the standard intravenous thrombolytic in acute ischaemic stroke, its administration requires an initial bolus followed by a continuous infusion over one hour, introducing relevant logistical challenges in high-pressure clinical environments, particularly when interhospital transfer or immediate transition to an endovascular procedure is required. In this context, tenecteplase has emerged as an attractive alternative due to its single-bolus administration, greater fibrin specificity, and prolonged pharmacokinetic profile, characteristics that may facilitate early reperfusion and reduce operational errors associated with infusion.

The integration of tenecteplase into contemporary stroke management has been driven primarily by randomised trials evaluating its performance in patients with large-vessel occlusion and planned thrombectomy. In particular, EXTEND-IA TNK provided evidence that tenecteplase administered prior to thrombectomy can increase rates of early reperfusion and improve clinical outcomes compared with alteplase (Campbell et al. 2018). These findings have been interpreted as supporting a bridging thrombolysis strategy with tenecteplase, in which the thrombolytic is used not only as initial treatment but also as a tool to achieve partial or complete reperfusion before endovascular access, with potential impact on total ischaemic time.

Additional evidence from randomised trials such as ATTEST-2 supports tenecteplase as an alternative to alteplase, demonstrating comparable efficacy and safety within the standard thrombolysis window in patients treated within 4.5 hours of symptom onset (Huang et al. 2022).

However, one of the critical aspects in the clinical implementation of tenecteplase has been uncertainty regarding the optimal dose. This issue was addressed by EXTEND-IA TNK Part 2, which compared 0.40 mg/kg with 0.25 mg/kg in patients with ischaemic stroke due to large-vessel occlusion eligible for intravenous thrombolysis and mechanical thrombectomy within 4.5 hours (Campbell et al. 2020). The higher dose did not demonstrate superiority in pre-thrombectomy cerebral reperfusion nor was it associated with significant functional improvement. Instead, there was a trend towards increased symptomatic intracranial haemorrhage with 0.40 mg/kg. These results suggest that dose escalation does not confer clear clinical advantage and support the use of 0.25 mg/kg as the preferred regimen in planned thrombectomy scenarios.

Beyond potential efficacy, the principal appeal of tenecteplase lies in its operational impact. In healthcare systems aiming to reduce delays to reperfusion and facilitate transfers, single-bolus administration decreases therapeutic complexity, reduces the risk of infusion interruption during transfers, and enables more rapid transition to advanced imaging or endovascular procedures. From a clinical management perspective, these advantages may translate into greater protocol adherence, fewer errors, and potential optimisation of critical time metrics.

The relevance of tenecteplase is not confined to patients with large-vessel occlusion. Indeed, emerging evidence suggests that its utility may extend to scenarios previously considered beyond the scope of intravenous thrombolysis. In the OPTION trial, administration of tenecteplase 0.25 mg/kg in patients with non-LVO stroke, selected by perfusion imaging and treated between 4.5 and 24 hours, increased the likelihood of achieving excellent functional outcome at 90 days. However, this benefit was accompanied by an increase in symptomatic intracranial haemorrhage, underscoring that any expansion of indications must remain strictly linked to imaging-based selection criteria and individual haemorrhagic risk assessment.

Direct Thrombectomy Versus Bridging Therapy

Although the trials that established mechanical thrombectomy as the standard of care for large-vessel occlusion were largely conducted under a model of prior intravenous thrombolysis when no contraindications existed, contemporary practice has faced a relevant operational and clinical question: Should intravenous thrombolysis be systematically administered before thrombectomy, or is it acceptable to proceed directly to mechanical thrombectomy when this can be performed immediately?

The so-called bridging therapy strategy is based on the premise that intravenous thrombolysis may initiate partial or complete recanalisation before endovascular intervention, improve distal microcirculation, and facilitate the procedure. In addition, it provides an immediate therapeutic opportunity while access to the angiography suite is being organised. However, this strategy entails potential risks, including increased haemorrhagic risk, as well as added costs and complexity in patients who will undergo thrombectomy regardless.

By contrast, direct thrombectomy offers important operational advantages: simplification of the therapeutic pathway, reduced exposure to thrombolytics, and potential reduction in haemorrhagic risk. Moreover, in contexts where logistics allow rapid access to thrombectomy, it has been suggested that prior thrombolysis may provide limited incremental benefit, particularly when the thrombus is large or proximal and the probability of recanalisation with thrombolysis alone is low.

Within this framework, the DIRECT-MT trial evaluated direct thrombectomy compared with intravenous thrombolysis followed by thrombectomy in patients with ischaemic stroke due to large-vessel occlusion eligible for endovascular intervention (Yang et al. 2020). The design adopted a non-inferiority approach, seeking to determine whether omitting intravenous thrombolysis did not compromise functional outcome. The results demonstrated that direct thrombectomy was non-inferior in 90-day functional outcome under relatively wide non-inferiority margins. However, interpretation of these findings requires caution, given methodological considerations and the fact that demonstration of non-inferiority does not necessarily imply full clinical equivalence, particularly when accepted margins may substantially influence the final conclusion.

Beyond statistical outcomes, DIRECT-MT and similar trials must be interpreted within the context of healthcare system organisation. In centres with immediate access to thrombectomy and optimised door-to-puncture times, direct thrombectomy may represent a reasonable strategy in selected subgroups. However, in networks where transfer delays, uncertainty regarding suite availability, or operational bottlenecks exist, intravenous thrombolysis continues to offer an initial intervention with potential benefit, particularly when administered early and integrated without delaying thrombectomy (Yang et al. 2020).

In this regard, the decision to adopt a bridging strategy or direct thrombectomy should not be reduced to a dichotomous choice but rather individualised, taking into account factors such as estimated time to endovascular reperfusion, thrombus characteristics, occlusion location, patient haemorrhagic risk profile, availability of advanced imaging, and interhospital transfer capacity. Furthermore, the recent evolution of evidence regarding tenecteplase has reignited this debate, as a single-bolus thrombolytic may facilitate bridging therapy without increasing the logistical complexity associated with alteplase.

Practical and Organisational Implications

Contemporary advances in reperfusion have transformed the prognosis of acute ischaemic stroke, but their real-world impact depends largely on the capacity of healthcare systems to implement efficient care pathways. In practice, the difference between a theoretically available treatment and one that is effectively delivered is defined by operational times, access to advanced imaging, interhospital coordination, and the availability of multidisciplinary teams. In this sense, ischaemic stroke represents a paradigmatic model of a time-dependent disease in which system organisation is a direct determinant of outcome.

Evidence derived from pivotal thrombectomy trials demonstrates that clinical benefit is not confined to the endovascular procedure itself but is closely linked to the speed of large-vessel occlusion identification, efficient transfer to thrombectomy-capable centres, and minimisation of time to reperfusion. Trials such as ESCAPE explicitly highlighted the relevance of care pathways and rapid selection based on vascular imaging and collateral status, reinforcing that the healthcare system can amplify or constrain the effect of the intervention (Goyal et al. 2015).

Imaging Selection: Standardisation as a Priority

The shift from a time to a tissue paradigm implies that the availability and quality of diagnostic imaging have become central components of acute care. Routine implementation of computed tomography angiography to detect proximal occlusion and cerebral perfusion imaging to quantify core and penumbra has enabled expansion of thrombectomy and thrombolysis indications into extended windows. Studies such as EXTEND-IA, DEFUSE 3, and DAWN consolidated that perfusion-based or clinical–imaging mismatch selection allows identification of patients with substantial benefit even many hours after symptom onset (Campbell et al. 2015, Albers et al. 2018; Nogueira et al. 2018).

However, these models also introduce practical challenges. Perfusion interpretation requires standardised protocols, validated software, and reproducible criteria, as technical variability can translate into divergent clinical decisions. Therefore, adoption of clear institutional algorithms with defined thresholds for core volume and mismatch is essential to reduce variability and avoid decisions based solely on subjective interpretation.

Extended Window and Unequal Access: Therapeutic Opportunity and Risk

Expansion of the therapeutic window has important implications for centres with structural limitations. Recent trials have demonstrated that intravenous thrombolysis can confer benefit beyond classical windows in perfusion-selected patients, contingent upon strict selection criteria. OPTION, for example, suggests that tenecteplase in the extended window may benefit non-LVO patients, a population that predominates numerically in everyday practice (Ma et al. 2026).

This finding has relevant operational implications: in systems where access to thrombectomy is limited, the ability to perform cerebral perfusion imaging may become a strategic therapeutic resource. However, this expansion also increases clinical complexity, as benefit is accompanied by a relative increase in symptomatic intracranial haemorrhage, necessitating strict selection protocols, intensive post-treatment monitoring, and clear exclusion criteria.

Tenecteplase as a Tool for Logistical Simplification

In real-world settings, therapeutic decisions depend not only on clinical efficacy but also on operational feasibility. In this respect, tenecteplase has gained interest not only for its potential effectiveness but also for its single-bolus administration. This characteristic reduces complexity during transfers, facilitates emergency department management, and decreases the risk of errors associated with prolonged infusion. Trials such as EXTEND-IA TNK Part 2 support the use of 0.25 mg/kg as the preferred dose in planned thrombectomy scenarios, as no additional benefit was observed with higher doses (Campbell et al. 2018).

These logistical advantages may be particularly relevant in network models, where patients receive thrombolysis at a peripheral hospital and are subsequently transferred to a thrombectomy centre. In such scenarios, simplification of the thrombolytic agent may help maintain therapeutic continuity without interfering with transfer times.

Network Models

Expansion of endovascular indications has strengthened debate regarding the optimal transfer model: direct transport to a thrombectomy-capable centre versus initial thrombolysis at a peripheral centre followed by transfer. The choice depends on geographical variables, estimated transfer times, availability of vascular imaging at first contact, and the capacity to deliver thrombolysis without delay.

Pivotal thrombectomy trials and extended-window studies have demonstrated that late reperfusion can still be beneficial in selected patients, but absolute benefit diminishes as ischaemic time increases. Consequently, networks must prioritise time reduction at every link in the chain: prehospital recognition, stroke code activation, imaging acquisition, therapeutic decision-making, and reperfusion. The summary of trials of reperfusion strategies in ischaemic stroke is shown in Table 1.

Reperfusion is Not the End: Post-treatment Protocols

Even with successful recanalisation, prognosis may deteriorate due to cerebral oedema, haemorrhagic transformation, respiratory or cardiovascular complications. Therefore, a structured post-reperfusion component should be incorporated, including frequent neurological monitoring, haemodynamic control, aspiration prevention, glucose and temperature management, and clear strategies for early detection of symptomatic intracranial haemorrhage. Recent evidence on extended-window thrombolysis reinforces this need, as haemorrhagic risk persists and may be clinically relevant even in patients selected using advanced imaging.

The Role of Early Rehabilitation

The main outcome in most studies evaluating reperfusion strategies in ischaemic stroke is functionality (e.g., mRS); however, this may be influenced by early rehabilitation or its absence.

Evidence from large pragmatic trials such as AVERT suggests that very early and intensive mobilisation may not always confer benefit and may even reduce the likelihood of favourable functional outcomes, these findings emphasise that early rehabilitation should prioritise patient selection and physiological stability rather than a universal early mobilisation approach (Bernhardt et al. 2015).

From the first day, therapeutic positioning and management of silent complications form an integral part of rehabilitation. Strategies such as neutral positioning have demonstrated benefits in comfort and passive mobility in severely immobilised patients (Pickenbrock et al. 2015). Similarly, dysphagia should be addressed as both a safety and early rehabilitation issue, beginning with screening prior to oral intake, a recommendation supported by clinical guidelines and observational evidence suggesting reduced complications (Prabhakaran et al. 2026; Sherman et al. 2021). In selected cases of established dysphagia, specific interventions such as pharyngeal electrical stimulation may modulate swallowing pathophysiology, particularly in clinical subgroups such as tracheostomised patients (Scutt et al. 2015; Dziewas et al. 2018).

In the pharmacological domain, evidence indicates that early use of antidepressants with the intention of enhancing functional recovery does not improve disability or independence and may increase adverse events, underscoring the need to separate treatment of depression from attempts to accelerate neurological recovery through pharmacological means (Prabhakaran et al. 2026). In parallel, rehabilitation should incorporate psychosocial and non-pharmacological strategies focused on self-efficacy, motivation, and self-perception, as these factors influence adherence, therapeutic engagement, and real-world participation (Gangwani et al. 2022; Lo et al. 2022).

From a motor perspective, meaningful functional recovery does not consist solely of reducing spasticity but of increasing the capacity to perform tasks with adaptable variability. Task-oriented training shows consistent effects in improving gait and functional endurance and should be integrated with motor learning principles such as progression, variability, and feedback (Lee et al. 2025). Large pragmatic trials such as LEAPS further demonstrated that different locomotor training strategies may yield comparable functional outcomes, reinforcing the role of individualised rehabilitation approaches (Duncan et al. 2011).

For the upper limb, constraint-induced movement therapy, supported by trials such as EXCITE, represents an intervention with evidence of functional benefit when applied to appropriately selected patients, emphasising the need to match interventions to goals and to measure outcomes systematically (Thrane et al. 2014; Wolf et al. 2006). Other approaches such as robot-assisted therapy and neuromodulation have shown mixed or modest benefits in functional outcomes, as illustrated by trials such as RATULS and VNS-REHAB, highlighting the importance of intervention selection based on patient characteristics and goals (Rodgers et al. 2019; Dawson et al. 2021). Self-directed upper limb programmes such as GRASP have also demonstrated improvements in functional use and task performance, emphasising the importance of repetitive practice and patient engagement (Harris et al. 2009). Mirror therapy has also demonstrated improvements in motor recovery and self-care outcomes in controlled trials (Yavuzer et al. 2008). Bilateral arm training approaches such as BATRAC suggest potential benefits in motor recovery and cortical activation (Whitall et al. 2000).

In the management of foot drop, evidence does not demonstrate clear superiority between ankle–foot orthoses and functional electrical stimulation. Therefore, choice should be based on individual clinical characteristics, tolerance, context of use, and patient functional goals (Nascimento et al. 2020).

Physical exercise occupies a central role in post-stroke rehabilitation. Moderate- to vigorous-intensity aerobic exercise and high-intensity gait training are associated with improvements in functional capacity, cognition, and neural plasticity processes. Evidence suggests that appropriate cardiovascular stimulus dosing and cumulative practice volume are critical determinants of functional response. Furthermore, mechanistic studies indicate that exercise may enhance cortical reorganisation, providing a biological basis for observed improvements (Li et al. 2024; Hornby et al. 2016; Lohse et al. 2014; Ploughman et al. 2015).

Research on bed rest demonstrates that even brief periods of inactivity produce rapid and clinically relevant functional decline in older adults, with loss of muscle mass, strength, and aerobic capacity. This vulnerability to disuse has been confirmed by experimental studies and evidence syntheses, reinforcing the importance of preventing prolonged immobilisation in hospital and rehabilitation contexts (Kortebein et al. 2008; Tanner et al. 2015; Arentson-Lantz et al. 2021).

The summary of early rehabilitation trials in ischaemic stroke is shown in Table 2.

Conclusion

Contemporary management of acute ischaemic stroke has changed significantly owing to randomised controlled trials that consolidated mechanical thrombectomy as the standard of care for large-vessel occlusion and expanded the role of intravenous thrombolysis beyond the classical window through advanced imaging-based selection. Systematic incorporation of angiography and perfusion imaging has reinforced the paradigm of treating tissue, not time, demonstrating benefit even in scenarios of uncertain or late onset. However, the clinical impact of these strategies depends on appropriate patient selection and, above all, on the capacity of healthcare systems to implement rapid, safe, and standardised care pathways.

Conflict of Interest

None.