Introduction
Stroke in patients with active cancer represents an expanding clinical challenge marked by therapeutic uncertainty. Among those presenting with acute ischemic stroke due to large-vessel occlusion (LVO), approximately 6%-10% have a current or recent cancer diagnosis [1-3]. Active cancer—defined as diagnosis within 6 months, ongoing systemic therapy, known metastases, or progressive disease—carries clinical relevance. These patients exhibit distinct mechanisms, including hypercoagulability, nonbacterial thrombotic endocarditis, tumor embolism, and treatment-related vasculopathy [4-6]. Although the term active cancer is widely used in the stroke literature, it is best regarded as an operational category rather than a biologically uniform condition.
In practice, thrombectomy decision-making should distinguish among tumor histology, thrombogenic phenotype, stage, metastatic burden, and recent or ongoing anticancer therapy, as these factors may influence thrombotic risk, bleeding risk, vascular injury, and overall prognosis. Active cancer may be pragmatically defined as cancer diagnosed or treated within the previous 180 days, newly diagnosed cancer identified during the index stroke admission, or recurrent, metastatic, or inoperable cancer present at the time of stroke. Ongoing or recent chemotherapy should not be used as an exclusion criterion. Instead, cancer status should be interpreted within a broader assessment of performance status, treatment intent, expected survival, coagulation profile, and the suspected stroke mechanism, including hypercoagulability, nonbacterial thrombotic endocarditis, tumor embolism, conventional stroke mechanisms, or treatmentrelated vasculopathy. Radiation-associated cerebrovascular injury should be considered separately from hypercoagulable cancerassociated stroke [7,8].
Despite increasing observational data, patients with active cancer were excluded from nearly all pivotal endovascular thrombectomy (EVT) trials and likewise from the recent large-core studies [9-18]. Consequently, clinicians must often make time-critical decisions without randomized evidence, contributing to variability in practice and, in some centers, non-referral for thrombectomy with potentially avoidable loss of independence [1,3]. Guideline statements remain heterogeneous: major stroke guidelines select patients by vessel, time, and imaging but do not list active cancer as a contraindication [19,20].
As global cancer survivorship improves, the number of patients with LVO stroke and active malignancy will continue to rise. To address these uncertainties and the absence of harmonized regional data, the Asia-Pacific Stroke-Cancer Consortium (APEX)— a collaborative network linking tertiary stroke and oncology centers across Australia, Japan, South Korea, Hong Kong, Taiwan, Singapore, Indonesia, Vietnam, and Malaysia—was established. Its purpose is to generate prospective, decision-grade evidence on EVT in patients with active cancer, including those with large ischemic cores, and to develop an evidence-based framework for equitable selection, consent, and aftercare in both well-resourced and under-resourced settings.
This narrative review synthesizes current evidence on the efficacy, safety, and clinical context of thrombectomy in patients with active cancer, integrating multicenter registry findings and meta-analyses, and outlining a pragmatic clinical-ethical-economic framework to guide selection, consent, and future research through the APEX platform.
Why is this question relevant to clinical practice?
Active malignancy represents a meaningful and increasing proportion of patients presenting with LVO stroke. Observational studies suggest that approximately 6%-10% of LVO stroke patients have active cancer [4,21,22]. These patients demonstrate distinct pathophysiological mechanisms, including hypercoagulability, nonbacterial thrombotic endocarditis, tumor embolism, and treatment-related vascular injury [1,5,6].
Registry studies have shown that, when standard imaging and clinical selection criteria are applied, rates of successful reperfusion and symptomatic intracranial hemorrhage (sICH) are similar between cancer and non-cancer populations. However, functional outcomes and survival differ substantially. Patients with active cancer demonstrate lower rates of functional independence and higher mortality, reflecting systemic disease burden rather than procedural complications [1,4,21-25].
In clinical practice, the absence of randomized trial data for this subgroup leads to variability in treatment decisions. Some centers apply standard thrombectomy criteria irrespective of cancer status, whereas others remain cautious, particularly when cancer prognosis is uncertain. This variability highlights the need for a structured synthesis of existing evidence and clearer frameworks for treatment decision-making.
What do the registries show?
Multicenter registry studies consistently demonstrate that technical success rates of EVT are comparable between cancer and non-cancer populations. In the Multicenter Randomized Clinical trial of Endovascular treatment for Acute ischemic stroke in the Netherlands registry substudy, successful reperfusion and sICH rates were similar between groups, although functional independence at 90 days was lower in cancer patients [4].
Similarly, the Selection Criteria in Endovascular Thrombectomy and Thrombolytic Therapy study reported comparable rates of early neurological improvement but higher long-term mortality in patients with active cancer [21]. Data from the Italian national thrombectomy registry also showed similar technical success rates but higher three-month mortality among cancer patients [24].
Across multiple registries, approximately 20%-30% of patients with active cancer achieve functional independence at 90 days, whereas mortality ranges between 35% and 45%. These findings suggest that thrombectomy is technically feasible and procedurally safe in this population, but outcomes are strongly influenced by the underlying malignancy.
What do the meta-analyses show?
Registry, cohort, and meta-analytic findings are summarised in Table 1 to allow direct comparison of reperfusion, sICH, functional independence, mortality, and the principal interpretation across published studies. Across more than 3,600 patients, successful recanalisation and sICH rates are not significantly different between cancer and non-cancer populations, while functional outcomes and survival diverge. In the pooled meta-analysis by Duan et al. [26], odds of good functional outcome were halved (odds ratio [OR]: 0.47; 95% confidence interval [CI] 0.35-0.65) and mortality nearly quadrupled (OR: 3.87; 95% CI 2.64-5.68) among patients with active cancer. Aloizou et al. [3] and Eun et al. [25] confirmed no difference in technical or hemorrhagic outcomes but consistently lower rates of 90-day functional independence and higher mortality across studies. Quantitative synthesis from meta-analyses and systematic reviews is presented in Table 1.
Across nearly all registries, the outcomes that drive real-world decisions—such as early neurological improvement, 90-day performance status, and whether systemic anticancer therapy was initiated or resumed—are seldom prespecified, inconsistently captured, or absent from reporting. This lack of decision-grade endpoints limits translation of procedural results into patientcentered treatment guidance.
Overall, registry data suggest that EVT in active-cancer patients is technically feasible and procedurally safe, but functional and survival outcomes are more heterogeneous—reflecting the influence of systemic cancer burden rather than the thrombectomy itself.
Potential benefits and who stands to gain
EVT can provide meaningful benefit for carefully selected patients with acute ischemic stroke and active cancer, particularly when neurological recovery enables continuation or resumption of oncologic therapy and independent living. Across pooled registries and meta-analyses, successful reperfusion (Thrombolysis in Cerebral Infarction 2b-3) is achieved in approximately 80%-90% of patients with active cancer—rates comparable with those without cancer—while sICH occurs in about 5%- 7% [1-5,21,23,25,26]. Despite this procedural parity, functional recovery remains limited: only 20%-30% of patients with active cancer regain functional independence (modified Rankin Scale [mRS] 0-2) at 90 days, compared with 35%-45% in non-cancer cohorts [1,4,5,21-23,25,26]. Mortality, typically 35%-45% by three months, is driven largely by tumor progression and systemic complications rather than procedural harm [1,3,5,23,26].
These findings suggest that the subset most likely to benefit are those with preserved premorbid independence (mRS 0-2; Eastern Cooperative Oncology Group [ECOG] 0-2), imaging evidence of limited infarct core and viable penumbra, and a realistic expectation of continuing cancer treatment or meaningful life participation. In this group, early reperfusion can restore communication, autonomy, and quality of life sufficient for discharge home and resumption of cancer treatment.
Data from recent large core thrombectomy trials in non-cancer populations indicate that, under defined imaging and workflow criteria, benefit may persist even when early ischemic change is extensive [16-18]. Although these studies did not enrol patients with active cancer, their findings provide cautiously applicable guidance for cases where infarct burden is large, but imaging still shows salvageable tissue. Clinical decisions in such settings should explicitly acknowledge that the supporting evidence originates from non-cancer cohorts [16-18].
Overall, the available data indicate that EVT can restore sufficient neurological function to permit resumption of systemic therapy and prolong survival in a proportion of patients with active cancer, provided selection is individualized and consent reflects both neurological and oncological trajectories.
In patients with active cancer, prognosis after LVO stroke is shaped not only by technical reperfusion success but also by the underlying oncological context. Key determinants of functional outcome and survival include premorbid and oncological performance status, cancer stage and metastatic burden, treatment intent, expected life expectancy, stroke severity, infarct extent, successful reperfusion, and systemic complications. Biomarkers associated with cancer-related coagulopathy, particularly Ddimer, may further help identify patients with biologically aggressive cancer-associated stroke and poorer prognosis. These factors are clinically important not because they justify routine exclusion from thrombectomy, but because they help determine whether reperfusion is likely to translate into meaningful neurological recovery within the patient’s broader cancer trajectory [7,8].
Clinical-ethical-economic pathway for treatment selection
A comprehensive clinical-ethical-economic pathway enables consistent and equitable decision-making for patients with active cancer and acute LVO. The process begins with the simplest viable imaging strategy and concludes with proportionate, values-based consent.
In hospitals with constrained resources or time-critical emergencies— particularly where advanced imaging such as computed tomography (CT) perfusion (CTP) or magnetic resonance imaging is not readily available—non-contrast CT (NCCT) combined with single-phase CT angiography (CTA) should serve as the default diagnostic workflow. In these circumstances, the Alberta Stroke Program Early CT Score on NCCT provides a pragmatic estimate of infarct core and supports timely treatment decisions [19,20]. However, this streamlined approach requires vigilance in cancer-related stroke, where multifocal emboli, hypercoagulable states, or metastatic lesions may not be apparent on NCCT alone.
In our recent mild-stroke study, advanced CTP imaging identified clinically significant tissue-at-risk in nearly one in five patients despite low National Institutes of Health Stroke Scale (NIHSS) scores demonstrating that physiological mismatch can exist even when NCCT appears unremarkable [27]. This reinforces that perfusion-based selection detects salvageable tissue that standard CT may underestimate—a principle directly applicable to cancer-related stroke, where infarct physiology may be heterogeneous or influenced by systemic factors.
Where available, CTP or magnetic resonance perfusion (MRP) (CTP/MRP) remain the preferred imaging modalities because they delineate core, penumbra, mismatch, and collateral status with greater precision, supporting more confident prognostication. Accordingly, CTP/MRP should form the standard imaging tier for both cancer and non-cancer populations whenever feasible.
In settings where advanced imaging is constrained, NCCT combined with CTA provides a pragmatic and ethically sound framework for treatment decisions, provided that the inherent uncertainties are clearly conveyed during consent. This approach should not be regarded as inferior but as context-adapted—reflecting the diversity of healthcare environments represented within the APEX consortium. The emphasis should remain on transparency regarding uncertainty, adherence to consistent selection principles, and equitable access to care rather than presuming a single imaging “gold standard.” This balanced framing broadens the investigative scope of APEX, ensuring inclusivity across hospitals with differing resources while maintaining scientific rigor and ethical parity.
Embedding flexibility within APEX protocols will allow systematic evaluation of imaging pathways—from NCCT±CTA to advanced CTP/MRP—across differing resource environments while preserving safety, ethical integrity, and comparability of outcomes. A streamlined, context-responsive algorithm ensures timely triage without excluding potentially treatable patients, shortens onset-to-puncture time, reduces device use, and enhances efficiency while maintaining clarity in consent about imaging limitations.
Decisions should be made collaboratively, integrating oncologic status, coagulation profile, and overall clinical context while carefully excluding hemorrhagic or metastatic mimics. This balanced approach recognizes that advanced imaging enhances precision but that withholding thrombectomy solely due to absence of CTP/MRP may generate inequity. A context-responsive imaging framework—anchored in safety, feasibility, and fairness— best reflects evolving Asia-Pacific practice and the guiding principles of the APEX registry.
Clinical eligibility should prioritize patients with premorbid independence (mRS 0-2; ECOG 0-2), a definable target occlusion on CTA, and imaging consistent with a limited core and salvageable tissue. Informed consent should emphasize expected benefit, uncertainty related to cancer trajectory, and alternatives such as supportive or palliative care. Discussions should differentiate between procedural feasibility and the probability of meaningful recovery, maintaining realistic expectations for patients and families.
From an economic perspective, resource allocation should be guided by the likelihood of meaningful neurological recovery and overall clinical context rather than by procedural feasibility alone. Cost-effectiveness analyses in non-cancer thrombectomy populations suggest gains in quality-adjusted life-years over medical therapy, with acceptable incremental cost-effectiveness ratios [28,29]. Independent modelling has likewise shown that thrombectomy remains cost-effective under conservative assumptions [28,29]. However, these data derive from broader ischemic stroke cohorts rather than patients with active cancer specifically. The cost-effectiveness of EVT in cancer-associated LVO therefore remains uncertain and should be treated as a priority research question rather than an established assumption. Future studies should integrate clinical, ethical, and health-economic evaluation, particularly in relation to meaningful neurological recovery, discharge disposition, performance status, and the ability to initiate or resume systemic anticancer therapy. Overall, this integrated clinical-ethical-economic framework balances feasibility and fairness—supporting thrombectomy decisions grounded in realistic recovery potential, efficient resource allocation, and respect for patient values.
Practice implications
In clinical practice, decisions regarding EVT in patients with active cancer should align with established LVO criteria while accounting for premorbid function, systemic disease trajectory, and imaging findings. Selection is most appropriate when baseline independence (mRS 0-2; ECOG 0-2) and imaging demonstrate a clear target occlusion with limited core and viable penumbra [1-5,19-21,23,25-27]. Consent discussions should be transparent and patient-centered, outlining potential neurological benefits and realistic limitations related to cancer prognosis [1-8,19-21,23,25,26,30].
Where available, CTP/MRP are preferred because they identify the core-penumbra mismatch and provide stronger prognostic confidence. In under-resourced or emergency settings, however, treatment decisions based on NCCT combined with singlephase CTA remain appropriate when accompanied by transparent discussion of uncertainty and explicit documentation in consent. This pragmatic approach acknowledges ongoing gaps in evidence and ensures that potentially treatable patients are not excluded solely due to imaging limitations [19,20,27].
Bridging intravenous thrombolysis should remain part of the treatment algorithm when indicated and not contraindicated, as available registry data and meta-analyses show no excess sICH or mortality in well-selected cancer populations [31,32]. Attention to clot morphology, thrombus density, and occlusion pattern on baseline imaging can further refine patient selection and procedural planning [33,34].
When early ischemic change is extensive, patient selection may be informed by evidence from large-core thrombectomy trials in non-cancer populations, provided the non-cancer origin of this evidence is clearly stated during consent [16-18].
To enhance decision quality and ethical consistency, integration between stroke and oncology teams is essential—particularly for assessing cancer trajectory, performance status, and feasibility of resuming systemic therapy post-stroke. Data harmonization through regional platforms such as APEX will support this multidisciplinary collaboration and enable region-specific insights into workflow adaptation, imaging strategy, and post-thrombectomy recovery (Figure 1).
Overall, practice decisions should emphasize clinical plausibility and equity rather than categorical exclusion. A contextresponsive strategy—anchored in feasibility, patient values, and fair access—ensures that thrombectomy and adjunct therapies are offered to cancer patients with realistic recovery potential, regardless of local resource limitations. The integration of stroke and oncology perspectives through structured regional frameworks such as the APEX collaboration ensures that decisions remain ethically transparent, evidence-informed, and adaptable across both comprehensive and resource-constrained settings.
Case vignette-early re-occlusion after technically successful reperfusion in metastatic cancer
A 47-year-old man with metastatic malignancy presented with acute left middle cerebral artery (MCA) syndrome (NIHSS: 15). CTP demonstrated a large penumbral mismatch, with a small infarct core (relative cerebral blood flow <30%: 4 mL) and an extensive hypo-perfused region (Tmax >6 seconds: 105 mL; mismatch ratio: 26.3) (Figure 2A).
Digital subtraction angiography confirmed left MCA occlusion (Figure 2B). Mechanical thrombectomy achieved complete reperfusion on the first pass (Figure 2C). Despite initial improvement, the patient presented the following day with recurrent left MCA occlusion (Figure 2D).
This case highlights the challenge of maintaining durable reperfusion in patients with active cancer, where systemic hypercoagulability and endothelial dysfunction can predispose to early re-occlusion.
Within the APEX registry, neuroimaging-based assessment of clot morphology and lesion pattern will be used to explore associations between imaging features, hypercoagulability, and stroke recurrence after EVT in cancer patients [33]. Multidisciplinary integration aims to translate procedural success into sustained clinical recovery by aligning vascular and oncologic management within a unified care framework.
Asia-Pacific Stroke-Cancer Consortium
To address the major evidence gaps in this field, future research must move beyond procedural feasibility alone and focus on decision-grade outcomes, comparative effectiveness, and survivorship-relevant endpoints. Within this context, the APEX is presented as a regional collaborative framework designed to support prospective data harmonization across diverse healthcare settings, rather than as a stand-alone protocol paper. Its purpose is to generate pragmatic evidence on patient selection, imaging pathways, neurological recovery, performance status, and the feasibility of initiating or resuming anticancer therapy after stroke.
A key priority for future registry-based research is to move beyond single-arm EVT reporting and, where feasible, compare EVTtreated and non-EVT patients with acute large-vessel occlusion and active cancer. Any such analysis will require explicit handling of treatment selection bias, including prespecified adjustment for major prognostic variables such as baseline performance status, metastatic burden, cancer stage, and stroke severity. Methods such as propensity-based adjustment or matching may help strengthen comparative inference, although careful interpretation will remain necessary in the absence of randomized data [35,36].
Building upon the regional framework outlined earlier, the APEX will prospectively enrol patients with acute LVO and active malignancy across multiple tertiary centers. Patient selection will use pragmatic imaging criteria based on NCCT, CTA/MRA, and—where available—CTP/MRP [16-20].
The inclusion of patients with large ischemic cores will enable validation of selection signals observed in non-cancer large core thrombectomy trials and clarify how premorbid function, systemic disease burden, and oncologic factors influence outcomes. Primary endpoints will include 24-hour NIHSS change, 90-day functional independence (mRS 0-2), 90-day ECOG performance status, and whether systemic anticancer therapy is initiated or resumed by 90 days [1-6,8,21,23,25,26,30]. Because treatment allocation in registry-based studies is not randomized, comparative analyses between EVT-treated and non-EVT patients will require explicit handling of treatment selection bias. Propensitybased methods, including propensity score matching or inverse probability weighting, together with multivariable regression adjustment for key prognostic variables such as baseline NIHSS, premorbid functional status, cancer stage, metastatic burden, and imaging characteristics, will be used to reduce confounding in observational comparisons.
Imaging and laboratory parameters—core and penumbra volumes, collateral status, platelet count, International Normalized Ratio, fibrinogen, D-dimer, echocardiography, and tumor markers where available—will support mechanistic analyses to distinguish cancer-related cryptogenic and nonbacterial thrombotic endocarditis phenotypes [5,6,8,34].
Three visual elements are proposed for presentation within stroke: (1) a comparative overview of thrombectomy guideline differences, (2) a schematic of the APEX registry framework, and (3) a synthesis of prognostic factors linked with favorable and unfavorable outcomes.
Current data indicate that EVT in patients with active cancer is technically feasible and procedurally safe, achieving reperfusion and hemorrhage rates comparable to those in non-cancer cohorts [1-5,21-26]. However, 90-day functional outcomes and mortality remain more heterogeneous, largely reflecting the underlying malignancy rather than procedural harm [1,4,5,21-23,25,26]. Accordingly, APEX emphasizes individualized selection based on premorbid independence (mRS 0-2; ECOG 0-2), a definable target occlusion, and imaging evidence of salvageable tissue [19,20].
To date, no randomized or observational studies have directly compared EVT with best medical management in patients with active cancer—a key evidence gap that APEX seeks to address.
Conclusions
EVT in patients with active cancer is technically feasible and achieves reperfusion rates similar to those observed in noncancer populations. Functional outcomes and survival remain more heterogeneous, reflecting the trajectory of the underlying malignancy rather than procedural complications. Careful patient selection integrating neurological, oncological, and ethi-cal considerations is therefore essential. Collaborative initiatives such as the APEX aim to generate the evidence needed to guide treatment decisions in this increasingly common clinical scenario.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
