Clinical and Demographic Characteristics, Mechanisms, and Outcomes in Patients With Acute Ischemic Stroke and Newly Diagnosed or Known Active Cancer

Abstract

Background and Objectives

Patients with a new diagnosis of cancer carry an increased risk of acute ischemic stroke (AIS), and this risk varies depending on age, cancer type, stage, and time from diagnosis. Whether patients with AIS with a new diagnosis of neoplasm represent a distinct subset from those with a previously known active malignancy remains unclear. We aimed to estimate the rate of stroke in patients with newly diagnosed cancer (NC) and previously known active cancer (KC) and to compare the demographic and clinical features, stroke mechanisms, and long-term outcomes between groups.

Methods

Using 2003–2021 data from the Acute STroke Registry and Analysis of Lausanne registry, we compared patients with KC with patients with NC (cancer identified during AIS hospitalization or within the following 12 months). Patients with inactive and no history of cancer were excluded. Outcomes were the modified Rankin scale (mRS) score at 3 months and mortality and recurrent stroke at 12 months. We used multivariable regression analyses to compare outcomes between groups while adjusting for important prognostic variables.

Results

Among 6,686 patients with AIS, 362 (5.4%) had active cancer (AC), including 102 (1.5%) with NC. Gastrointestinal and genitourinary cancers were the most frequent cancer types. Among all patients with AC, 152 (42.5%) AISs were classified as cancer related, with nearly half of these cases attributed to hypercoagulability. In multivariable analysis, patients with NC had less prestroke disability (adjusted odds ratio [aOR] 0.62, 95% CI 0.44–0.86) and fewer prior stroke/transient ischemic attack events (aOR 0.43, 95% CI 0.21–0.88) than patients with KC. Three-month mRS scores were similar between cancer groups (aOR 1.27, 95% CI 0.65–2.49) and were predominantly driven by the presence of newly diagnosed brain metastases (aOR 7.22, 95% CI 1.49–43.17) and metastatic cancer (aOR 2.19, 95% CI 1.22–3.97). At 12 months, mortality risk was higher in patients with NC vs patients with KC (hazard ratio [HR] 2.11, 95% CI 1.38–3.21), while recurrent stroke risk was similar between groups (adjusted HR 1.27, 95% CI 0.67–2.43).

Discussion

In a comprehensive institutional registry spanning nearly 2 decades, 5.4% of patients with AIS had AC, a quarter of which were diagnosed during or within 12 months after the index stroke hospitalization. Patients with NC had less disability and prior cerebrovascular disease, but a higher 1-year risk of subsequent death than patients with KC.

According to US administrative claims data, an estimated 10% of patients hospitalized with acute ischemic stroke (AIS) have a history of cancer.¹ A recent meta-analysis and single-center cohort studies suggest that between 1.4% and 2.1% of patients with AIS are diagnosed with cancer in the following year.^2-4 The risk of stroke is increased up to 1 year before a new diagnosis of cancer,⁵ and this risk remains elevated for up to 10 years afterward.⁶ Risk of AIS with cancer varies according to patient age, cancer type, stage, and time from diagnosis.⁷ Comorbid cancer has been shown to increase the risk of early clinical deterioration, disability, recurrence, and mortality after AIS.^8-10

Previous studies suggest that the increased risk of stroke in patients with active cancer (AC) is likely to be multifactorial.^11,12 Besides traditional risk factors shared between patients with cancer and those with noncancer AIS,¹³ people with cancer may develop AIS due to cancer-related mechanisms.¹⁴ These include hypercoagulability, tumor embolism, and oncological treatments, such as prothrombotic effects or cardiotoxicity of cancer therapies or vasculopathy due to radiotherapy. Several studies have highlighted the role of cancer-related AIS mechanisms.^11,15-17 However, most of these analyses focused on small cohorts, lacked information on recent cancer therapeutics such as immunotherapy, and did not analyze the frequency of potential cancer-related and cancer therapy–related AIS mechanisms other than hypercoagulability. Furthermore, previous studies addressed differences between patients with AC and those with inactive or no cancer, and little data exist comparing patients with AIS with previously known active cancer (KC) with those with newly diagnosed cancer (NC), a distinct subset of patients who may have different risks, pathophysiology, and outcomes. In particular, different staging at diagnosis, a varying aggressiveness of cancer phenotypes, or the potentially prothrombotic effects of anticancer drugs^18,19 in treated patients with KC warrant an investigation of differences and similarities between these 2 groups. In this cohort study using data from a large institutional AIS registry spanning nearly 2 decades, we aimed to compare the clinical and laboratory characteristics, mechanisms, clinical outcomes, and recurrent stroke risk between patients with NC and patients with KC.

Methods

Design and Setting

We performed a retrospective single-center cohort study using data from the Acute STroke Registry and Analysis of Lausanne (ASTRAL) registry of the University Hospital of Lausanne. The ASTRAL is a single-center cohort registry of consecutive patients (16 years or older) with AIS admitted to the stroke unit and/or intensive care unit within 24 hours of last-well time.²⁰ Recorded variables and their definitions are prespecified in the ASTRAL registry. The database is approved by our institutional review board as a clinical and research registry. All data stem from routine clinical and radiologic management.

Study Population

This analysis included all patients with AIS in the ASTRAL registry from January 2003 to December 2021. The study population of interest was patients with AC, which we divided into 2 groups: (1) patients with KC and (2) patients with NC. We defined KC as cancer diagnosed within the previous 6 months (excluding nonmelanoma skin cancer, benign meningioma, myelodysplastic syndrome [MDS]); cancer under active treatment or for which treatment had been administered within 6 months; and hematologic cancers that were not in complete remission.²¹ All stages of cancers regardless of being recurrent, regionally advanced, or metastatic were included. NC referred to patients with cancer identified during the admission for the index stroke, within the following 12 months, or during the inpatient workup for suspected cancer complicated by in-hospital stroke. Adapting from the available guidelines and recent literature,^22,23 we considered brain metastases “newly diagnosed” if they were not diagnosed before the current AIS, that is, previous brain imaging was not performed or it was performed but no metastases were found. They were considered “known active” if they had been treated or untreated and they showed signs of imaging activity (progression over time, new lesions) compared with baseline imaging; known brain metastases that received prior treatment and presented no signs of imaging activity were considered “known inactive.”

Patients without cancer and those with a history of cancer not fulfilling these criteria (i.e., remote inactive cancer) were excluded from the analysis. For both cancer groups, we extracted the type and stage of cancer, date of diagnosis, laboratory data, and cancer treatments received. Patients with multiple cancer types were classified as having the most active one (i.e., more advanced stage, active treatment). Cancer was classified based on anatomopathological results; when not available, classification was based on the most likely diagnosis according to clinical and radiologic findings. Aiming to perform a comprehensive analysis, we included all cancer types (including primary brain tumors). We excluded benign meningiomas, which mostly have a limited pathologic potential, MDS, where the threshold of cancer vs noncancer is poorly defined, and nonmelanoma skin cancers, which are generally a localized form of neoplasia.

Patients in the registry consist of two-thirds of local patients (population of 400,000) and one-third of referrals from the secondary catchment area (population of 1.2 million), addressed mostly for IV thrombolysis (IVT) or endovascular thrombectomy (EVT). Our institution has a widely recognized tertiary care oncological department^24,25 with 24 permanent beds, treating inpatients and outpatients from the primary and the secondary catchment area.

Cancer and Stroke Evaluation

The workup for standard stroke mechanisms followed a prespecified institutional protocol. This included acute brain imaging (CT or MRI) with the study of supra-aortic and cerebral arteries, at least 24-hour cardiac rhythm monitoring, and transthoracic echocardiography (unless recently performed or a noncardiac stroke mechanism had already been identified). Further testing included basic blood laboratory tests (comprehensive metabolic panel, blood counts, coagulation factors, lipid profile, d-dimers, and fibrin monomers in some patients) and repeated brain and arterial imaging in case an acute revascularization procedure was performed.

A consensus on when and how to perform a workup for previously unknown cancer in patients with AIS is still lacking.¹⁴ In our institutional protocol, we recommend a cancer diagnostic workup under the following conditions: (1) if clinical clues suggest a possible systemic disease (e.g., unexplained weight loss, fever, and nocturnal sweating); (2) if the standard workup fails to identify AIS mechanisms, especially in patients with few established vascular risk factors; (3) in unexplained multifocal stroke; (4) in simultaneous ischemic and hemorrhagic strokes; (5) unusual laboratory findings (unexplained anemia, polycythemia, or thrombocytosis), and (6) multiple or recent unexplained venous thromboembolic events and at least 1 unexplained stroke. This search is decided by the physician in charge of the patient and consists of a detailed history targeting symptoms of frequent cancers, skin examination, breast and gynecologic examination including transvaginal ultrasound in female patients, prostate examination and prostate-specific antigen in male patients, fibrin monomers and full-body CT scan, and search for genetic abnormalities (such as Janus kinase 2) according to hematologic findings. Transesophageal echocardiography and whole-body positron emission tomography may be added in some situations.

Cancer Treatment

We defined chemotherapy and immunotherapy as being active when AIS appeared during or within 1 cycle after the last application of the specific drug. We considered hormone therapy as active if administered within 1 month before AIS. Active radiotherapy and hematopoietic stem cell transplantation referred to patients who underwent a treatment cycle while developing AIS. Patients were classified as having recent surgery if they developed AIS intraoperatively or within 24 hours of a diagnostic or therapeutic invasive procedure.

Stroke Mechanisms

We categorized stroke subtypes according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) trial classification system²⁶ with the following specific mechanisms added: dissection of supracardiac arteries, patent foramen ovale (PFO) (embolic stroke of undetermined source [ESUS] plus a Risk of Paradoxical Embolism score ≥7²⁷), multiple simultaneous causes, ESUS other than PFO, undetermined non-ESUS, and undetermined mechanism with an incomplete workup.

Within the “rare” subtype category of TOAST, we included cancer-related mechanisms. Based on the available literature, clinical experience, and multidisciplinary discussion with hematologists, oncologists, and pharmacologist consultants at our institution, we defined 3 major cancer-related categories of mechanisms: (1) cancer specific, (2) cancer therapy related, and (3) cancer diagnostic procedure related (additional data are listed in eTable 1, links.lww.com/WNL/C776).

We considered cancer drugs with at least a moderate level of evidence for association with ischemic stroke as a potential stroke mechanism according to a recent publication.¹⁹ For the drugs not included in this review, we performed a search for potential associations with cerebrovascular events using various sources of information, including summaries of product characteristics,²⁸ Micromedex, PubMed, and Meyler's Side Effects of Drugs.²⁹ Cancer drugs were considered prothrombotic applying the same criteria used in this recent review.

Study Outcomes

We investigated 3 study outcomes: (1) 90-day functional outcome defined per the modified Rankin scale (mRS) score; (2) mortality over 12 months; and (3) stroke and transient ischemic attack (TIA) recurrences over 12 months. Recurrences were defined as ischemic or hemorrhagic cerebrovascular events, including TIA and retinal events.

Statistical Analysis

Comparisons between groups were analyzed with the χ² test, independent t test, or Mann-Whitney U test, as appropriate. Continuous and ordinal variables were expressed as medians (interquartile range [IQR]) and categorical variables as absolute counts (percentages).

From all the covariates available in the ASTRAL registry, we manually selected the ones that were clinically/pathophysiologically most plausibly associated with the outcomes examined and had only few missing values, independently from their p values, as described by Harrell.³⁰

We used binomial and ordinal logistic regression to calculate adjusted odds ratios (aORs). We used Cox regression analysis to assess the effect of different variables on the time to a cerebrovascular recurrence or death over the first 12 months and to calculate the adjusted hazard ratios (aHRs). The variables used for the adjustments are listed in the corresponding tables. Missing variables were imputed with multivariate imputations with 5 imputations by chained equations. All analyses were performed using R software 4.1.2 (R Foundation, Vienna, Austria). p Value <0.05 was considered statistically significant.

Standard Protocol Approvals, Registrations, and Patient Consents

Before analysis, study data were anonymized following the principles of the Swiss Human Research Ordinance from 2013 (HRO, Art. 25). Ethical commission approval and patient consent were not required according to the Swiss Federal Act on Research involving Human Beings from 2011 (HRA, Art. 3) because all data were anonymized and the project involved assessing the safety and quality of routine stroke management in the institution.

Anonymized data that support the findings of this study are available from the corresponding author on reasonable request and after signing a data transfer and use agreement. If such data are used for a publication, its methods should be communicated, and internationally recognized authorship rules should be applied.

Data Availability

Anonymized data that support the findings of this study are available from the corresponding author upon reasonable request and after signing a data transfer and use agreement. If such data are used for a publication, its methods should be communicated, and internationally recognized authorship rules should be applied.

Results

Baseline Characteristics

From 2003 to 2021, 362 (5.4%) of 6,686 patients with AIS registered in the ASTRAL registry had KC or NC. The median age of patients with cancer was 74 years (IQR 17), and 138 (38.1%) were women (Table 1). There were 260 (71.8%) patients with KC and 102 (28.2%) patients with NC. The overall incidence of NC in the year after AIS was 1.5% (1.2% diagnosed during hospitalization and 0.3% within the following 12 months). Among the 21/102 (20.6%) patients with AIS diagnosed with NC after hospitalization, 20 (19.6%) received a diagnosis within 6 months and only 1 (1%) afterward.

Table 1.

Baseline Characteristics of Patients With Acute Ischemic Stroke With Newly Diagnosed Cancer vs Those With Known Active Cancer

The median age and sex ratio were similar in both groups (Table 1). Regarding reperfusion therapies, 14.9% of patients received IVT only and 20.2% EVT or bridging therapy (IVT before EVT). We found no differences in the rates of reperfusion times and therapies between the 2 groups.

In multivariable analysis to assess the predictive value of several markers of NC vs KC, we found that patients with NC had less previous stroke/TIA (aOR 0.43, 95% CI 0.21–0.88) and lower prestroke disability (aOR 0.62, 95% CI 0.44–0.86) (Table 2). Patients with NC more frequently had dyslipidemia and PFO at cardiac imaging, and the 2 groups presented several differences regarding laboratory parameters.

Table 2.

Unadjusted and Adjusted Comparison of Demographic, Clinical, and Laboratory Variables Between Patients With Acute Ischemic Stroke With Newly Diagnosed Cancer vs Those With Known Active Cancer

The most frequent tumor types were gastrointestinal (27%) followed by lung (23%) and genitourinary (21%). In the NC group, gastrointestinal cancer was the most frequent (37%), whereas genitourinary cancer (26%) was the most represented in patients with KC (Table 3). In univariable comparison, patients with NC had more gastrointestinal and lung cancers, while patients with KC had more genitourinary cancer. The frequency of metastatic disease (58.0% in NC vs 58.5% in KC) and brain metastasis (7.9% in NC vs 9.8% in KC) was similar between groups. Regarding patients with KC, 50% had recently received cancer treatments, including chemotherapy (29%), hormone therapy (10%), radiotherapy (4%), immunotherapy (4%), and surgery (3%).

Table 3.

Cancer-Related Characteristics

Stroke Mechanisms

Cancer-related stroke mechanisms (rare cancer related or multiple, including cancer related) were more frequent in patients with KC (48% in KC vs 28% in NC) (Table 4). Hypercoagulability (as defined in eTable 1, links.lww.com/WNL/C776) was considered the most frequent cancer-related stroke mechanism in both groups (23% in patients with NC and 23% in patients with KC), followed by chemotherapy in the KC group (22%). In multivariable analysis, female sex (aOR 2.67, 95% CI 1.29–5.62), previous stroke/TIA (aOR 2.21, 95% CI 1.15–4.25), and metastatic cancer (aOR 4.83, 95% CI 2.30–10.7) were associated with a presumed hypercoagulable stroke mechanism, but tumor type and cancer therapies were not (Table 5). In addition, patients with a presumed hypercoagulable stroke mechanism were more likely to be younger (aOR 0.95, 95% CI 0.93–0.98) (Table 5).

Table 4.

Acute Ischemic Stroke Mechanisms

Table 5.

Unadjusted and Adjusted Analyses of Demographic, Clinical, and Oncological Features Associated With a Cancer-Related Hypercoagulability Stroke Mechanism Comparing Patients With Acute Ischemic Stroke With Newly Diagnosed Cancer vs Those With Known Active Cancer

Clinical Outcomes

In multivariable analysis, there was no difference in mRS shift at 3 months between patients with NC vs patients with KC (Figure 1). However, even after adjusting for palliative status (aOR 42.1 95% CI 13.1–191.6), a diagnosis of new brain metastasis was strongly associated with worse functional outcome (aOR 7.22, 95% CI 1.49–43.17) (eTable 2, links.lww.com/WNL/C776). Other significantly associated variables were prestroke disability (aOR 1.98, 95% CI 1.5–2.65), admission NIH Stroke Scale (NIHSS) (aOR 1.12, 95% CI 1.07–1.18), and metastatic cancer (aOR 2.19, 95% CI 1.22–3.97).

Figure 1. mRS Score Shift at 3 Months in Patients With NC vs Patients With Previously KC With AIS.

Comparison by ordinal logistic regression shows the effect of having NC in patients with AIS compared with those with KC and is expressed as an aOR (aOR 1.27, 95% CI 0.65–2.49, p = 0.489). AIS = acute ischemic stroke; aOR = adjusted odds ratio; KC = known active cancer; mRS = modified Rankin scale; NC = newly diagnosed cancer.

Risk of death at 12 months was higher in patients with NC than in patients with KC (adjusted common HR 2.11, 95% CI 1.38–3.21) (Figure 2A and eTable 3, links.lww.com/WNL/C776). Metastatic cancer (aHR 2.71, 95% CI 1.93–3.80) and a palliative status during hospitalization (aHR 4.31, 95% CI 2.82–6.57) were 2 other major predictors of mortality. We did not find a difference in risk of recurrent stroke or TIA between study groups (aHR 1.27, 95% CI 0.67–2.43) (Figure 2B and eTable 4).

Figure 2. Mortality and Cerebrovascular Recurrence at 12 Months in Patients With NC vs Patients With Previously KC.

Patients with NC present a 2-fold increased death risk (adjusted common HR 2.11, 95% CI 1.38–3.21) at 12 months compared with patients with KC (A). Although patients with NC present a 27% higher recurrence time hazard (aHR 1.27, 95% CI 0.67–2.43) at 12 months, this was not statistically significant (B). aHR = aHR = adjusted hazard ratio; KC = known active cancer; NC = newly diagnosed cancer.

Discussion

In a large retrospective cohort study conducted over 18 years, we found that an estimated 5.4% of patients with AIS have KC or NC. Among these patients, approximately 72% had KC and 28% had NC (absolute risk of 1.5%). We found that in the overall cohort and the NC group gastrointestinal cancers were the most common cancer type, whereas in the KC group, genitourinary cancers were the most frequent type. Not surprisingly, patients with NC had less prestroke disability and fewer prior ischemic stroke/TIA events. Regarding AIS mechanisms, more than one-third of patients had some cancer-related stroke mechanism, with stroke presumed to be from hypercoagulability as the most frequent etiology in both groups. Functional outcomes and risk of recurrent cerebrovascular events were similar between study groups; however, patients with NC had an increased long-term risk of death when compared with patients with KC. The presence of newly diagnosed brain metastasis and metastatic cancer were strong predictors of poor functional outcomes in both study groups.

The prevalence of KC or NC in our cohort aligns with previous smaller studies on patients with AIS and cancer, which reported prevalence rates ranging from 4.4% to 6% for AC among multicenter cohorts^4,31,32 and between 1.4% and 2.1% for NC within 1 year of AIS.^2-4 Some studies have reported cancer prevalence rates of up to 12%, but these cohorts included patients with inactive cancer.^1,33

Regarding cancer subtypes, lung^10,34 and genitourinary cancers (including prostate)³¹ were the most frequent tumors in a previous analysis of patients with AC, whereas we found gastrointestinal cancers to be the most frequent overall. Our results and the previous literature^34,35 thus confirm that tumors associated with adenocarcinoma histology are the most represented in patients with AIS. Although we found little difference in tumor subtypes between NC and KC groups, patients with NC may present more frequently with gastrointestinal and lung tumors, indicating that these 2 systems should be addressed first during the diagnostic workup for suspected occult cancer.

Previous studies have shown that patients with KC less frequently receive IVT,^36,37 which was not the case in our cohort neither for IVT nor EVT. Similar to other studies, we did not find significant differences in the time from onset to reperfusion therapies.^36-38

We used a novel categorization of potentially cancer-related stroke mechanisms based on suggestions of several other authors^14,39,40 and pathophysiologic considerations. We identified a high proportion of strokes attributed to cancer itself (27%) or related to a concomitant well-known cause (15%), such as atherosclerosis, cardioembolism, lacunar, dissection, or other rare causes. Hence, the rate of cryptogenic stroke, ranging from 32% to 51% found in other studies,^{11,16,17,31,34} is lower (16%) in our cohort. Using for the first time a comprehensive description of potential cancer-related etiologies (eTable 1, links.lww.com/WNL/C776), we identified hypercoagulability as the most frequent mechanism. Of interest, younger age, female sex, and metastatic cancer were associated with hypercoagulability after adjusting for other potential predictors (Table 5). We also found that nearly 22% of patients with KC have AIS associated with prothrombotic chemotherapy (Table 4). Because this percentage is not negligible, we believe that the choice of whether to discontinue AC therapy during hospitalization for AIS or whether, how, and when to resume it at discharge should be carefully weighed and discussed in a multidisciplinary team considering the type of drug, stage of cancer disease, possible competing mechanisms of AIS, and risk/benefit ratio.

Most other studies assessed functional outcomes at 3 months grouping together patients with KC and NC.^10,36,41 Overall, the functional outcome at 3 months is generally worse in patients with AIS and cancer undergoing reperfusion therapies compared with populations with inactive cancer or no oncological history.³⁶ A recent study reported a lower proportion of patients (46%) achieving independent ambulation (mRS 0–3) at 3 months vs 57% in our population.³⁶ However, admission NIHSS in their cohort was higher (14 [IQR 8] vs 6 [IQR 12]), and it included only patients receiving reperfusion therapies, likely accounting for this difference. When looking for independent predictors of poor functional outcome, we did not find any difference between the NC and KC groups. However, new brain metastasis and metastatic disease were strong independent predictors of worse outcome, even when adjusting for prestroke functional status, admission NIHSS, cancer type, palliative status, and anticoagulation at discharge. This suggests that advanced disease may be a key determinant of prognosis regardless of other clinical and oncological characteristics and secondary preventive strategies. Other smaller studies found no effect of cancer stage on functional outcome, but the sampled population excluded cases with brain metastasis or primary brain tumors.^10,41

Most data on mortality in patients with cancer and AIS are recorded at 3 months, with mortality rates ranging from 27% to 61%.^10,34,36,41 One would expect mortality rates to be even higher at 12 months, but a death rate of 60% may seem relatively low compared with some of the abovementioned studies. However, in these cohorts, the exclusion of some patients with cancer and AIS, such as those with brain metastasis,^10,41 primary brain tumors,^10,41 or hematologic disease,¹⁰ along with variable frequencies of metastatic cancer,³⁶ may partially account for this difference. Compared with patients with KC, patients with NC had a 2-fold increased risk of death at 12 months after adjusting for metastatic disease, cancer type, palliative status, prestroke functional status, and anticoagulation at discharge. This survival advantage could be due to differences in type or intensity of cancer treatment between groups, aggressiveness of cancer phenotypes in the NC group, burden of noncancer causes of death, or other social or economic stressors associated with a new diagnosis of cancer not considered in the analysis. Another interpretation is that patients with KC with the worst prognosis died before presenting a stroke, thus being excluded from our cohort. Consistent with previous analyses,^42,43 metastatic disease was independently associated with a more than 2-fold higher risk of death at 12 months.

The risk of thromboembolic events is higher after AIS in patients with cancer vs those without cancer.^34,42,44 Navi et al.³⁴ reported a 16% recurrence rate of ischemic stroke and TIA at 6 months, with adenocarcinoma histology as the only independent predictor of recurrent thromboembolic events. Another study on cancer-related stroke reported an annual stroke recurrence rate (including ischemic and hemorrhagic stroke and cerebral venous thrombosis) of almost 14%, with AC being an independent predictor.⁴⁴ In our cohort, total stroke (TIA, ischemic stroke, and hemorrhagic stroke) recurrences at 12 months were high (20%), and we did not find differences between the 2 cancer groups. Unfortunately, we did not have information on adenocarcinoma histology in our cohort, but other cancer characteristics such as cancer stage, brain metastasis, and cancer type were not independent predictors.

Our study presents several limitations. First, as a quality assurance project in a single Swiss institution, it may not apply to other settings or patient groups, particularly to non-White elderly populations. Second, despite the large number of patients and detailed information in the registry, the monocentric and retrospective nature of the study may limit the internal and external validity of the results. Third, our AIS cohort spanned a long period (>18 years) during which stroke and oncological care have evolved, potentially introducing confounders regarding the association between cancer and stroke. Fourth, the definition of hypercoagulability also included D-dimer levels or the presence of monomers, which were only measured in a subset of patients, potentially leading to an underestimation of this stroke mechanism. Last, some of the differences we identified in our study may be affected by a survivor bias in that patients with more aggressive cancers may die before an AIS occur. Still, our findings may help inform patients and caregivers about the expected prognosis and influence treatment strategies.

In a large registry-based study of patients with AIS, we found that 5.4% had an active malignancy, with 1.5% having a new diagnosis of cancer during stroke or within the next 12 months. Gastrointestinal tumors were the most frequent cancer type, and hypercoagulability was considered the most common cancer-related stroke mechanism. Long-term mortality was significantly higher in patients with NC compared with patients with KC after adjusting for confounders. Our findings should be corroborated in prospective studies. If confirmed, they may motivate further research on the pathophysiologic mechanisms underlying AIS in patients with cancer, support the early identification of unknown tumors in patients with AIS, and drive the implementation of standardized protocols to screen for cancer in select patients with AIS.

Glossary

AC

active cancer

aHR

adjusted hazard ratio

AIS

acute ischemic stroke

aOR

adjusted odds ratio

ASTRAL

Acute STroke Registry and Analysis of Lausanne

ESUS

embolic stroke of undetermined source

EVT

endovascular thrombectomy

HRA

Human Rights Act

HRO

Human Research Ordinance

IQR

interquartile range

IVT

IV thrombolysis

KC

known active cancer

MDS

myelodysplastic syndrome

mRS

modified Rankin scale

NC

newly diagnosed cancer

NIHSS

NIH Stroke Scale

PFO

patent foramen ovale

TOAST

Trial of Org 10172 in Acute Stroke Treatment

Appendix. Authors

Study Funding

No targeted funding reported.

Disclosure

G. Costamagna is part of the editorial staff of the Resident & Fellow Section of Neurology®. A. Hottinger, H. Milionis, D. Lambrou, A. Salerno, D. Strambo, and F. Livio report no disclosures relevant to the manuscript. B.B. Navi received NIH grants K23NS091395 and R01HL144541, PCORI grant HIS-1511-33024, and NIDILRR grant 90REGE0012-01-11. P. Michel received funding from the National Science Foundation and Swiss Heart Foundation. Go to Neurology.org/N for full disclosures.