Related Experiment Video
Updated: Dec 5, 2025
Assessing Cerebral Autoregulation via Oscillatory Lower Body Negative Pressure and Projection Pursuit Regression
Published on: December 10, 2014
Automated Algorithm Using Pre-Intervention Fractional Flow Reserve Pullback Curve to Predict Post-Intervention
Seung Hun Lee1, Doosup Shin2, Joo Myung Lee3
1Division of Cardiology, Department of Internal Medicine, Heart Vascular Stroke Institute, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea.
Objectives:
This study sought to develop an automated algorithm using pre-percutaneous coronary intervention (PCI) fractional flow reserve (FFR) pullback recordings to predict post-PCI physiological results in the pre-PCI phase.
Background:
Both FFR and percent FFR increase measured after PCI showed incremental prognostic implications. However, there is no current method to predict post-PCI physiological results using physiological assessment in the pre-PCI phase.
Methods:
An automated algorithm that analyzes instantaneous FFR gradient per unit time (dFFR(t)/dt) was developed from the derivation cohort (n = 30). Using dFFR(t)/dt, the pattern of atherosclerotic disease in each patient was classified into 3 groups (major, mixed, and minor FFR gradient groups) in both the internal validation cohort with constant pullback method (n = 234) and the external validation cohort with nonstandardized pullback methods (n = 252). All patients in the validation cohorts underwent PCI on the basis of pre-PCI FFR ≤0.80. Suboptimal post-PCI physiological results were defined as both post-PCI FFR <0.84 and percent FFR increase ≤15%. From the derivation cohort, cutoffs of dFFR(t)/dt for major and minor FFR gradient were 0.035/s and 0.015/s, respectively.
Results:
In validation cohorts, dFFR(t)/dt showed significant correlations with percent FFR increase (R = 0.801; p < 0.001) and post-PCI FFR (R = 0.099; p = 0.029). In both the internal and external validation cohorts, the major FFR gradient group showed significantly higher post-PCI FFR and percent FFR increase compared with those in the mixed or minor FFR gradient groups (all p values <0.001). The proportions of suboptimal post-PCI physiological results were significantly different among 3 groups (10.4% vs. 25.8% vs. 45.7% for the major, mixed, and minor FFR gradient groups, respectively; p < 0.001) in validation cohorts. Absence of major FFR gradient lesion (odds ratio: 2.435, 95% [CI]: 1.252 to 4.734; p = 0.009) and presence of minor FFR gradient lesion (odds ratio: 2.756, 95% confidence interval: 1.629 to 4.664; p < 0.001) were independent predictors for suboptimal post-PCI physiological results.
Conclusions:
The automated algorithm analyzing pre-PCI pullback curve was able to predict post-PCI physiological results. The incidence of suboptimal post-PCI physiological results was significantly different according to algorithm-based classifications in the pre-PCI physiological assessment. (Automated Algorithm Detecting Physiologic Major Stenosis and Its Relationship with Post-PCI Clinical Outcomes [Algorithm-PCI]; NCT04304677).
