Patterns and predictors of cancer-associated ischemic stroke

INTRODUCTION

Cancer and ischemic stroke are global health issues and are the second and fourth leading causes of death.^[1] In India, both contribute significantly to the national health burden, with 1.5 million new cases of each reported annually.^[2] The association between cancer and stroke is bidirectional: Stroke can severely impact a cancer patient’s quality of life, and cancer, especially in its active stages, significantly increases the risk of stroke (hazard ratio of 1.44).^[3] Active cancer is defined as a diagnosis within the past 6 months, ongoing treatment, or cases involving recurrence or metastasis. In some cases, a stroke may reveal a previously undiagnosed cancer or be indicative of relapse of a previously treated cancer.^[4]

Cancer-associated stroke (CAS) differs from traditional stroke due to its unique mechanisms, presentation, and prognosis, often exhibiting higher recurrence and mortality rates.^[5] Cancer-induced hypercoagulability is the primary mechanism. Additional contributing factors include tumor embolism, direct vessel compression, infections, and treatment effects of radiation and chemotherapy. Hypercoagulability is driven by tumor cell activation of platelet aggregation, which leads to prolonged cancer cell survival and metastasis, thereby contributing to thrombotic events such as CAS.^[6] Certain biomarkers, including high D-dimer levels and multiple vascular territory infarction (MVTI), are common in CAS and serve as indicators of this hypercoagulable state. These are also linked with early neurological decline and reflect widespread thrombosis across brain regions.^[7,8]

Our study aimed to investigate CAS patient demographics, associated cancer types, survival, and associations between D-dimer levels, MVTI, and prognosis in CAS patients. Despite the poor prognosis often associated with CAS, there are no specific treatment guidelines for these patients. By identifying critical prognostic factors, this research aims to improve clinical understanding and guide future care strategies for CAS patients.

MATERIALS AND METHODS

This retrospective study analyzed clinical data from consecutive patients diagnosed with all types of acute ischemic stroke (AIS) across inpatient and outpatient settings of a tertiary care center in India between September 2014 and September 2024. Inclusion criteria were: (1) age above 18 years; (2) diagnosis consistent with AIS as per the 2018 American Heart Association/American Stroke Association guidelines; (3) an active cancer diagnosis; and (4) exclusion of other identifiable stroke etiologies, such as dissection or vasculitis.

Clinical data collected at the time of stroke diagnosis included demographic characteristics (age, sex, survival days following diagnosis); stroke-related features (number of vascular territories affected, National Institute of Health Stroke Scale score, modified Rankin Scale [mRS] score at 3 months, assessed through medical records or follow-up phone calls); cancer characteristics (pre-existing or new cancer diagnosis, cancer treatment history, presence of metastasis and tumor type); history of thrombotic disease (e.g., lower extremity deep vein thrombosis and antithrombotic medication); and presence of vascular risk factors (hypertension, dyslipidemia, diabetes, etc). Patients underwent comprehensive clinical evaluation, brain imaging (computed tomography [CT] and/or magnetic resonance imaging [MRI], CT/magnetic resonance angiography), 24-h electrocardiogram, echocardiography, and routine laboratory tests, including D-dimer. A majority of patients in the study received anticoagulation therapy as part of their stroke treatment.

Data were expressed as mean ± standard deviation or median ± interquartile range for continuous variables and as percentages or proportions for categorical variables. Statistical assessments between groups were conducted using t-tests, χ² tests, or Fisher’s exact tests to establish group differences. One-way analysis of variance was applied to compare group means. A binary logistic regression model was used to evaluate the correlation between MVTI and prognosis, calculating the adjusted odds ratio and 95% confidence intervals (CIs). A two-tailed P < 0.05 was considered statistically significant. All statistical analyses were performed using Statistical Package for the Social Sciences version 26.0.

This study was approved by the Institutional Ethics Committee (AIIMSA2863/03.01.2025).

RESULTS

In this study, 92 patients with ischemic stroke and cancer were identified, of whom 54 met the inclusion criteria and were enrolled. Among them, 41 patients completed the 3-month follow-up [Figure 1].

Close

Figure 1:

Screening and inclusion criteria algorithm.

Export to PPT

The median age of the patients was 58 (46–70) years, and 52% were female. Hypertension was the most common traditional vascular risk factor (27%), followed by dyslipidemia (20%), diabetes (19%), and smoking (19%). All patients underwent baseline CT head, while MRI brain was done for 68% (n = 37) of the cohort. The most common stroke etiology was undetermined (62%), followed by large artery atherosclerosis (33.3%) and cardioembolic stroke (3.7%).

The most common cancer types associated with CAS in our cohort were hepatobiliary (18.5%) and gynecological (18.5%), followed by head and neck (16.6%). One-quarter of our patients had metastatic disease and venous thromboembolism (VTE) was observed in 9.2%. In addition, 68% of patients were undergoing cancer-specific treatment at the time of CAS, while 11% were treatment-naïve. On tumor histopathology, 22% of patients had adenocarcinoma and 30% of patients had non-adenocarcinoma lesions. The baseline characteristics of our patients are presented in Table 1.

CAS characteristics [Table 2] revealed that MVTI was present in 42% patients, and the mean D-dimer level was 5.1 ± 4.4 µg/mL (normal < 0.5 µg/mL). Poor functional outcomes (mRS 5–6) were observed in 53% of patients (44% died within 3 months), whereas only 21% were functionally independent (mRS 0–2). Stroke recurrence was noted in 8% of patients [Figure 2].

Close

Figure 2:

Sankey chart showing a distribution comparison between pre-stroke and 3-month post-stroke modified Rankin scale (mRS) scores, with the majority of patients starting with an mRS of 0 or 1 pre-stroke and shifting to an mRS of 6 at 3 months.

Export to PPT

DISCUSSION

Our study aimed at identifying patterns, predictors and outcomes in CAS patients. We found a lower prevalence of traditional vascular risk factors, as cumulative vascular risk factors were below 20% despite accounting for approximately 90% population attributable risk in traditional ischemic stroke patients.^[9-11] This finding aligns with prior CAS research, which indicated that CAS has a distinct pathophysiology compared to non-cancer-related cerebral infarctions.^[12-14] Gynecological and hepatobiliary cancers were the most prevalent malignancies, accounting for 18.5% each, which contrasted with earlier studies where lung cancer was predominant.^[15] Metastatic disease was observed in 25% of our cohort, underscoring the importance of assessing metastasis in patients presenting with CAS.^[16-19]

A significant proportion of our cohort (62%) had cryptogenic strokes, which was much higher than 30% usually seen in the general population and was similar to other CAS studies.^[4,20] Among those with available MRI data, 62% showed MVTI [Figure 3]. While elevated D-dimer levels (>0.5 mg/L) can occur in cancer-free stroke patients, levels exceeding 5.5 mg/L were highly specific for CAS (specificity and positive predictive value > 93%).^[20] This combination of MVTI and elevated D-dimers serves as a crucial marker for CAS and should lead to a thorough cancer evaluation.^[15,20-22] Elevated D-dimers were also significantly higher in patients with metastatic disease and lung adenocarcinoma, reflecting a hypercoagulable state. This finding aligns with some studies but contrasts with others, highlighting the variability in the prognostic utility of D-dimer levels.^[14,23-25]

Close

Figure 3:

Magnetic resonance imaging demonstrates multiple acute infarcts involving both the anterior and posterior circulations across three distinct vascular territories. (a-c) The infarcts appear hyperintense on diffusion-weighted imaging (d-f) hypointense on apparent diffusion coefficient maps.

Export to PPT

In addition to hypercoagulability, radiotherapy-induced vasculopathy is another potential stroke mechanism in CAS patients. Four of our patients with a history of head-and-neck radiotherapy demonstrated significant extracranial internal carotid artery disease despite the absence of traditional vascular risk factors. Prior studies have shown intima-media complex thickening in irradiated patients, suggesting that radiotherapy contributed to stroke risk through vascular changes.^[26]

Stroke recurrence was observed in 8% of the cohort, exclusively in patients with adenocarcinoma, MVTI, and elevated D-dimers, emphasizing the role of hypercoagulability. Recurrence rates were lower than in prior studies, likely due to a high mortality rate (41%) and anticoagulation use for secondary prevention.^[27] Anticoagulation for secondary stroke prevention was done on the lines of previous studies, which revealed hypercoagulability as the driving force for procoagulant state in active cancers.^[27] Further studies with larger cohorts are essential to identify risk factors for ischemic stroke recurrence in CAS. In addition, we identified two cases of occult malignancy during stroke evaluation [Supplementary Table 1], thereby emphasizing the importance of biomarker assessments as tools to detect underlying malignancies that may otherwise remain unidentified.^[28]

Poor outcomes were prevalent in our cohort, with 53% having severe disability or death (mRS 5–6) and a 3-month mortality rate of 40%, aligning with findings from similar studies.^[29,30] These outcomes were strongly associated with metastatic disease, VTE, and MVTI, underscoring the impact of systemic hypercoagulability and advanced cancer stages on prognosis in CAS.

Our study limitations include its single-center and retrospective design with a small sample size. This may limit the statistical power for identifying predictors of outcomes. Further prospective studies with a larger sample size and longer follow-up duration are essential.

CONCLUSION

Our study is the first from India evaluating cancer-associated stroke. CAS is associated with a poor prognosis. In our study, the concurrence of MVTI and elevated D-dimer levels in patients classified under the cryptogenic stroke subtype strongly suggested an underlying or occult malignancy as the etiology.