Cancer and stroke are leading contributors to morbidity and mortality worldwide, and their co-occurrence is increasingly recognized. This nationwide, population-based analysis sought to provide contemporary, population-level estimates of the risk of ischemic stroke (IS) during the first year after diagnosis of a first invasive cancer, overall and across cancer sites, time intervals after diagnosis, cancer stage, and demographic subgroups.
The study used a retrospective cohort design by linking the Israel National Cancer Registry with the Israel Stroke Registry. The eligible population comprised residents aged 40 years and older with a first diagnosis of invasive cancer between 2014 and 2021. Patients were followed for the occurrence of a first-ever ischemic stroke within 1 year of cancer diagnosis.
Stroke risk was assessed using standardized incidence ratios (SIRs) compared with expected rates in the general population; SIRs were standardized by age, sex, and population group. Analyses stratified SIRs by cancer site, time since cancer diagnosis, cancer stage (localized, metastatic, unknown), and demographic subgroups including age categories. The one-year cumulative incidence of IS was estimated with death treated as a competing event.
The cohort included 182,221 patients with a first invasive cancer diagnosis. The overall SIR for first-ever ischemic stroke within 1 year after cancer diagnosis was 1.8 (95% confidence interval [CI], 1.7–1.9), indicating an 80% relative increase in IS incidence compared with the general population.
Temporal patterns showed the highest relative risk in the immediate postdiagnosis period. The SIR peaked during the first 3 months after cancer diagnosis at 2.5 (95% CI, 2.2–2.7). After this early peak the relative risk declined but remained elevated within the first year.
SIRs varied markedly by primary cancer site. The largest relative increases were observed for:
Other cancer sites showed lower but still elevated SIRs in the first year after diagnosis. The one-year cumulative incidence of IS also varied by site, with pancreatic cancer showing the highest cumulative incidence (1.62%) and breast cancer the lowest reported (0.30%).
Relative risk of IS after cancer diagnosis declined with advancing age. Reported SIRs by age category included:
These findings indicate a larger relative increase in stroke risk among younger cancer patients compared with older patients, although absolute stroke incidence remains higher with increasing age.
Stroke risk also increased with more advanced cancer stage. The reported SIRs by stage were:
Patients with metastatic cancer experienced substantially higher relative risk of postdiagnosis IS compared with those with localized disease.
The one-year cumulative incidence of first-ever ischemic stroke for the overall cancer cohort was 0.62% when treating death as a competing event. Site-specific cumulative incidence estimates ranged from 0.30% for breast cancer up to 1.62% for pancreatic cancer. The analysis included competing-risk curves for IS and for death, with an emphasis on the elevated early IS incidence relative to expectations from general population rates.
This population-based study demonstrates an elevated risk of ischemic stroke during the first year after invasive cancer diagnosis, with the greatest relative increase in the first 3 months and with marked variation by cancer site, age, and stage. Pancreatic and lung cancers were associated with the highest relative IS risks, metastatic disease carried higher risk than localized disease, and younger adults showed larger relative increases compared with older adults.
The authors highlight the early postdiagnosis period as a critical window for cerebrovascular risk. They suggest that incorporating targeted cerebrovascular risk assessment into oncologic care may help identify high-risk subgroups and support tailored prevention strategies.
One author reported consulting fees from device companies (Perflow Medical, Pi‐Cardia, and QuantalX Neuroscience); the other authors declared no conflicts of interest.
The source excerpt provides key results, SIRs, cumulative incidence estimates, and broad methodological approach. However, additional granular details that clinicians or analysts might seek—such as the number of strokes by specific cancer site, adjustment covariates beyond standardization, full statistical-model specifications, sensitivity analyses, and explicit limitations—were not reported in the abstract-level excerpt provided here. For those details, the full text should be consulted (the abstract indicates a Free PMC article is available).