Recurrence patterns and clinical correlates after embolic stroke of undetermined source: a hypothesis-generating retrospective cohort study from Taiwan

Background Embolic stroke of undetermined source (ESUS) carries a substantial risk of recurrence, but the clinical correlates of recurrence in East Asian populations remain poorly characterized. We aimed to describe recurrence patterns after ESUS and to explore, in a hypothesis-generating manner, baseline clinical and metabolic factors associated with recurrent stroke. Methods We conducted a single-center retrospective cohort study of 116 consecutive patients hospitalized with ESUS at Fu Jen Catholic University Hospital, Taiwan, between January 2021 and December 2024, with follow-up until December 31, 2025. Associations with recurrent stroke were examined with univariable Cox proportional hazards regression; Fine–Gray competing-risk regression (death as a competing event) and a 12-month logistic analysis were sensitivity analyses. Bayesian sensitivity analyses used conjugate beta–binomial models with Jeffreys priors. Given the small number of events, all analyses were considered descriptive and hypothesis-generating. Results During a median follow-up of 28.4 months, 17 patients (14.7%) had recurrent ischemic stroke; the cumulative incidence was 10.5% at 12 months and 14.0% at 24 months, and half of the recurrences with a known date occurred within 6 months. No baseline variable was significantly associated with recurrence in the primary analysis; with 17 events, odds ratios of approximately 4 to 7 would have been required for 80% power. Recurrence was numerically more frequent after minor index strokes (National Institutes of Health Stroke Scale [NIHSS] ≤ 3: 12/65, 18.5%) than after more severe strokes (5/51, 9.8%), but the estimate was imprecise (hazard ratio 1.58; 95% confidence interval 0.55–4.50). Among recurrent patients whose index infarct was confined to a single arterial circulation, 6 of 9 with assessable imaging recurred within the same circulation. Conclusions In this small ESUS cohort, recurrent stroke was not uncommon, but no baseline characteristic reliably identified patients at higher risk. The early clustering of recurrence, the numerically higher recurrence after minor strokes, and the recurrence topography are hypothesis-generating observations compatible with several explanations, including non-stenotic atherosclerosis, occult atrial fibrillation, differences in treatment and follow-up, and chance. Larger prospective studies with prolonged rhythm monitoring and vessel-wall imaging are needed.

Authors

Institutions

Publication Details

Journal
BMC Neurology
Published
2026-11-18
DOI
https://doi.org/10.1186/s12883-026-05516-x
Primary Topic
Acute Ischemic Stroke Management
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Recurrence patterns and clinical correlates after embolic stroke of undetermined source: a hypothesis-generating retrospective cohort study from Taiwan

Ben‐Chang Shia, Pao‐Hsien Chu, Yi‐Wei Kao, Chiung-Yun Lo et al.
3 links
BMC Neurology
Acute Ischemic Stroke Management
article

Recurrence patterns and clinical correlates after embolic stroke of undetermined source: a hypothesis-generating retrospective cohort study from Taiwan

Ben‐Chang Shia, Pao‐Hsien Chu, Yi‐Wei Kao, Chiung-Yun Lo, Hai-Pei Hsu, Yen‐Ming Chen, C.-C. Lin
article en

Abstract

Background Embolic stroke of undetermined source (ESUS) carries a substantial risk of recurrence, but the clinical correlates of recurrence in East Asian populations remain poorly characterized. We aimed to describe recurrence patterns after ESUS and to explore, in a hypothesis-generating manner, baseline clinical and metabolic factors associated with recurrent stroke. Methods We conducted a single-center retrospective cohort study of 116 consecutive patients hospitalized with ESUS at Fu Jen Catholic University Hospital, Taiwan, between January 2021 and December 2024, with follow-up until December 31, 2025. Associations with recurrent stroke were examined with univariable Cox proportional hazards regression; Fine–Gray competing-risk regression (death as a competing event) and a 12-month logistic analysis were sensitivity analyses. Bayesian sensitivity analyses used conjugate beta–binomial models with Jeffreys priors. Given the small number of events, all analyses were considered descriptive and hypothesis-generating. Results During a median follow-up of 28.4 months, 17 patients (14.7%) had recurrent ischemic stroke; the cumulative incidence was 10.5% at 12 months and 14.0% at 24 months, and half of the recurrences with a known date occurred within 6 months. No baseline variable was significantly associated with recurrence in the primary analysis; with 17 events, odds ratios of approximately 4 to 7 would have been required for 80% power. Recurrence was numerically more frequent after minor index strokes (National Institutes of Health Stroke Scale [NIHSS] ≤ 3: 12/65, 18.5%) than after more severe strokes (5/51, 9.8%), but the estimate was imprecise (hazard ratio 1.58; 95% confidence interval 0.55–4.50). Among recurrent patients whose index infarct was confined to a single arterial circulation, 6 of 9 with assessable imaging recurred within the same circulation. Conclusions In this small ESUS cohort, recurrent stroke was not uncommon, but no baseline characteristic reliably identified patients at higher risk. The early clustering of recurrence, the numerically higher recurrence after minor strokes, and the recurrence topography are hypothesis-generating observations compatible with several explanations, including non-stenotic atherosclerosis, occult atrial fibrillation, differences in treatment and follow-up, and chance. Larger prospective studies with prolonged rhythm monitoring and vessel-wall imaging are needed.

BMC Neurology
Fu Jen Catholic University (TW), Chang Gung University (TW), Ming Chuan University (TW), United Hospital (US), Chang Gung Memorial Hospital (TW), Tao Yuan General Hospital (TW)
Good health and well-being
Acute Ischemic Stroke Management
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon