Transcatheter Aortic Valve Implantation (TAVI) is an essential treatment for calcified aortic stenosis. However, accurately assessing patient risk remains a significant clinical challenge. Clinicians seek refined tools for TAVI mortality prediction. Therefore, researchers introduced the novel eCAPRI score. This enhanced tool combines traditional clinical risk factors with automatically measured imaging biomarkers from pre-operative CT scans.
Enhancing TAVI Mortality Prediction with Automated Imaging
The original CAPRI score offered a promising method for predicting one-year all-cause mortality. It did this by integrating manually calculated Thoracic Aortic Calcium (TAC) volume with established clinical data. Conversely, the eCAPRI score leverages deep learning technology to fully automate the TAC volume extraction. Furthermore, this automation significantly reduces the reliance on time-consuming manual measurements. The deep learning model achieved strong segmentation results, demonstrating a mean Dice score of 0.777 ± 0.108. The automation process provides a standardized and highly reproducible component for the risk assessment model.
The Novel Imaging Biomarkers in eCAPRI
A key differentiator of eCAPRI is its incorporation of additional, automatically extracted imaging biomarkers. These novel factors enhance the score’s predictive power over its predecessor. Specifically, the model includes body surface area (BSA)-indexed right ventricle volume and BSA-indexed pulmonary arteries maximum diameter. In addition, the score incorporates the abdominal muscles surface at the L3 vertebral level. Therefore, eCAPRI moves beyond purely structural aortic factors to consider broader cardiometabolic and frailty indicators. These comprehensive metrics improve the overall assessment of patient risk. Existing scores like the EuroSCORE and STS-PROM have demonstrated limited accuracy in predicting mortality for TAVI patients, since they were initially developed for surgical populations.
eCAPRI Outperforms Established Risk Scores
The study rigorously tested eCAPRI against the original CAPRI score and the widely used EuroSCORE Logistic. On the evaluation dataset, eCAPRI demonstrated superior discrimination power, achieving an Area Under the Curve (AUC) of 0.731. Moreover, the CAPRI score had an AUC of 0.669, while the EuroSCORE Logistic lagged considerably at 0.588. This performance difference proved statistically significant (p = 0.034). Consequently, eCAPRI offers better calibration and improved accuracy in determining one-year mortality risk. Therefore, this tool can support standardized risk stratification and more informed clinical decisions for TAVI patients.
Frequently Asked Questions
Q1: What does the eCAPRI score predict?
It predicts one-year all-cause mortality in patients undergoing Transcatheter Aortic Valve Implantation (TAVI) for calcified aortic stenosis. The score’s improved accuracy in this task aids in personalized care planning.
Q2: Which imaging biomarkers are newly incorporated into eCAPRI?
The eCAPRI score includes automatically computed Thoracic Aortic Calcium (TAC) volume, BSA-indexed right ventricle volume, BSA-indexed pulmonary arteries maximum diameter, and abdominal muscles surface at the L3 level.
References
- Lartaud PJ et al. eCAPRI: a novel tool combining clinical and imaging data for post-TAVI mortality prediction. Eur Radiol. 2025 Dec 20. doi: 10.1007/s00330-025-12184-x. PMID: 41420708.
- Mortality risk after transcatheter aortic valve implantation: analysis of the predictive accuracy of the Transcatheter Valve Therapy registry risk assessment model. EuroIntervention. 2017.
- Performance of Surgical Risk Scores to Predict Mortality after Transcatheter Aortic Valve Implantation. Rev Bras Cardiol Invasiva. 2014.
- Significance of the CAPRI risk score to predict heart failure hospitalization post-TAVI: The CAPRI-HF study. Int J Cardiol. 2019.