A Population Cohort Study to Evaluate the Risk of Ischemic Stroke Among Individuals with a New Diagnosis of Cancer Compared to Matched Cancer-Free Controls: Impact of Prior Stroke History
Bibliographic record
Abstract
Abstract Cancer is a risk factor for thrombosis, but unlike venous thrombosis, its association with ischemic stroke, and its impact on stroke management and post-stroke outcomes are not well characterized. The primary objective of this study was to measure and compare the risk of ischemic stroke in individuals with a new diagnosis of cancer and those without a history of cancer in two separate matched cohorts based on the absence (Matched Cohort 1) or presence (Matched Cohort 2) of a prior diagnosis of ischemic stroke (Figure 1). Methods: We conducted a population-based matched cohort study of adults ≥18 years using linked clinical and administrative health databases in Ontario, Canada (2010 to 2019). Individuals with a new diagnosis of cancer were matched (1:1) to cancer-free controls by age and sex. Cancer diagnoses were determined using ICD-O-3 diagnostic codes (basal and squamous cell carcinoma and primary central nervous system tumors were excluded). For cancer patients, the index date was the diagnosis date and a corresponding dummy index date was assigned to the matched cancer-free control. The primary outcome was the incidence of ischemic stroke (time to ischemic stroke following index) determined using validated ICD-9 or -10 diagnostic codes from hospitalizations or emergency department visits. Analyses were conducted separately, in parallel, for each cohort. Standardized differences were used to compare the distributions of baseline characteristics. Cumulative incidence function (CIF) curves were generated for ischemic stroke and all-cause mortality. Competing risk methods were used to calculate sub-distribution adjusted hazard ratios (aHR) and 95% confidence intervals (CI) for ischemic stroke, where death was treated as a competing event. Interactions with the timeline at 1.5 years after index were incorporated to assess if the hazard ratio for the main exposure (cancer vs cancer-free) varied before and after the 1.5-year mark. Regression analyses were adjusted using baseline characteristics for which there was imbalance between groups. Results: Matched Cohort 1 included 620,647 individuals with cancer and 620,647 controls. Matched Cohort 2 included 13,924 individuals with cancer and 13,924 controls. Baseline characteristics were generally well balanced (Table 1). Use of antithrombotic medications (prior to index) among individuals ≥66 years was similar between the groups (data not shown). CIF curves for ischemic stroke and mortality are shown in Figure 2. In Matched Cohort 1 (no history of stroke), the risk of ischemic stroke was increased in cancer patients compared to cancer-free controls 1.5 years post-index (aHR 1.40, 95%CI 1.34-1.47). From 1.5 years to 5 years, the risk of ischemic stroke was lower in cancer patients compared to controls (aHR 0.72, 95%CI 0.69-0.74). In Matched Cohort 2 (prior history of ischemic stroke), the risk of ischemic stroke was similar at 1.5 years after the index date in individuals with and without cancer (aHR 1.00, 95%CI 0.88-1.14). From 1.5 to 5 years, the risk of ischemic stroke was reduced (aHR 0.53, 95%CI 0.46-0.62) in cancer patients compared to controls. Conclusions: Compared to cancer-free controls, the risk of ischemic stroke among individuals with a new cancer diagnosis depended on the presence or absence of prior history of ischemic stroke, a novel finding not previously reported. In individuals without a prior history of stroke, those with cancer had a 1.5-fold higher risk of ischemic stroke at 1.5 years compared to controls. In individuals with a prior history of stroke, those with cancer had a similar risk of stroke compared to controls. At 5 years post-index, the risk of ischemic stroke was lower in cancer patients in both cohorts which may reflect high early mortality rates and lower stroke risk among long-term survivors of cancer. The risk of death among cancer patients was highest during the first 1.5 years after the index date in both cohorts. However, the magnitude of the risk increase was higher in individuals without a prior history of stroke (10-fold) compared to those with a prior history of cancer (5-fold). Limitations include unknown causes of death and unavailability of some covariates (e.g. smoking and antithrombotic use in individuals aged <65 years). Future analyses will explore risk of ischemic stroke within common cancer subtypes (breast, prostate, colon, pancreas, lymphoma), and assess stroke risk factors, treatments and outcomes. Figure 1 Figure 1. Disclosures Siegal: BMS-Pfizer: Honoraria; Leo Pharma: Honoraria; Novartis: Honoraria; Portola: Honoraria; Servier: Honoraria. Carrier: Pfizer: Honoraria, Research Funding; BMS: Honoraria, Research Funding; Servier: Honoraria; Leo Pharma: Honoraria, Research Funding; Bayer: Honoraria; Sanofi: Honoraria. Gross: Valeo: Honoraria; Leo Pharma: Honoraria; Bayer: Honoraria; BMS-Pfizer: Honoraria.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".