Related Experiment Video
Updated: Apr 24, 2026
Author Spotlight: Innovative Techniques and Future Directions in Stroke Research
Published on: May 5, 2023
Ischaemic stroke recurrence in patients with symptomatic intracranial atherosclerotic stenosis in China (PROMISE): a
Chengbei Hou1, Xiao Dong2, Yuanyuan Liu2
1Centre for Evidence-Based Medicine, Xuanwu Hospital Capital Medical University, Beijing, China.
Insights
A new PROMISE model accurately predicts recurrent ischemic stroke risk in patients with symptomatic intracranial atherosclerotic stenosis (ICAS). This tool helps identify high-risk individuals for better stroke management.
Area of Science:
- Neurology
- Cardiovascular Research
- Epidemiology
Background:
- Intracranial atherosclerotic stenosis (ICAS) is a major global cause of stroke with high recurrence rates.
- Predicting individual risk for recurrent ischemic stroke in ICAS patients is crucial for effective management.
Purpose of the Study:
- To develop and validate a simple, effective model for predicting individualized risks of recurrent ischemic stroke in symptomatic ICAS patients.
- To create an accessible tool to aid in the management of patients with symptomatic ICAS.
Main Methods:
- A multivariable prediction model (PROMISE) was developed and validated using data from 2995 participants in the RICA randomized controlled trial.
- The model utilized clinical variables including age, BMI, hypertension, diabetes, smoking status, LDL cholesterol, stenosis location, and degree.
- Model performance was assessed using discrimination (C-index), calibration (Hosmer-Lemeshow test), and clinical utility (decision curve analysis).
Main Results:
- The PROMISE model demonstrated strong discriminative ability with a C-index of 0.81 in the training set and 0.78 in the validation set.
- The model showed satisfactory calibration and provided clinical benefits according to decision curve analysis.
- Kaplan-Meier curves effectively classified participants into low, medium, and high-risk groups for stroke recurrence.
Conclusions:
- The PROMISE model and its online calculator offer a valuable tool for predicting ischemic stroke recurrence in symptomatic ICAS patients.
- The model aids in identifying high-risk individuals, supporting clinical decision-making and patient management.
- Further validation and optimization are recommended to enhance stroke care for ICAS patients.
Background:
Intracranial atherosclerotic stenosis is a prevalent cause of stroke worldwide and carries a high risk of recurrence. This study aimed to develop and validate a simple and effective model for predicting individualised risks of recurrent ischaemic stroke in patients with symptomatic intracranial atherosclerotic stenosis (ICAS).
Methods:
This multivariable prediction model was built and validated using participants with symptomatic ICAS within 30 days of symptom onset from a large randomised controlled trial (RICA). The trial enrolled 3033 participants across 84 hospitals in China between Oct 28, 2015, and Feb 28, 2019. Participants were non-randomly divided by hospitals location into training and validation sets. Eligible participants were aged 40-80 years and had experienced an ischaemic stroke or transient ischaemic attack attributable to 50-99% stenosis of a major intracranial artery. The primary outcome of the model was time to first ischaemic stroke recurrence within 1 year. A Cox proportional hazards model was developed using the Akaike information criterion and validated both internally and externally. Model performance was assessed by discrimination (Harrell's concordance index [C-index]), calibration (modified Hosmer-Lemeshow test and calibration plots), and clinical utility (decision curve analysis and Kaplan-Meier curves).
Findings:
2995 participants from the RICA trial were divided into the training (n=2137) and validation (n=858) sets. The PROMISE model included age, BMI, hypertension, type 1 and type 2 diabetes, current smoking status, LDL cholesterol, location of symptomatic stenosis, and stenosis degree of qualifying artery as predictors. The C-index for the prediction model was 0·81 (95% CI 0·80-0·83) in the training set and 0·78 (0·77-0·84) in the validation set. Model calibration was satisfactory across the full risk profile (modified Hosmer-Lemeshow test χ2=5·32, p=0·38). Furthermore, decision curve analysis curves indicated that this prediction model provided clinical benefits. Additionally, participants were classified into three distinct risk groups (low, medium, and high) by Kaplan-Meier curves based on the prediction model. The corresponding C-indices for these groups were 0·80 (95% CI 0·80-0·86), 0·68 (0·67-0·77), and 0·71 (0·69-0·81) in the training set, and 0·76 (0·75-0·88), 0·67 (0·65-0·82), and 0·68 (0·67-0·86) in the validation set. Based on the final model, an online risk calculator was developed.
Interpretation:
We developed the PROMISE model and an online calculator using accessible clinical variables to predict ischaemic stroke recurrence, identify individuals at high risk, and support symptomatic ICAS patient management. The model had a strong discriminative ability and good calibration. Further validation and model optimisation should be conducted to support stroke care.
Funding:
National Natural Science Foundation of China, Beijing Natural Science Foundation, Noncommunicable Chronic Diseases-National Science and Technology Major Project, and Beijing Physician Scientist Training Project.
