New findings from the FAVOR III China trial demonstrate that kidney function, followed by body mass index (BMI), is the strongest predictor of serious cardiovascular events two years after quantitative flow ratio-guided (QFR) percutaneous coronary intervention (PCI) treatment.
Researchers also found that QFR can reduce the risk of major cardiovascular events by nearly 15% compared to angiography-guided procedures. The data were presented as late-breaking science at the Society for Cardiovascular Angiography & Interventions (SCAI) 2025 scientific sessions (1–3 May, Washington, DC, USA).
The FAVOR III China trial, a multicentre, randomised, sham-controlled trial conducted at Fuwai Hospital in Beijing, China, explores the long-term outcomes of PCI procedures guided by QFR. Two post-hoc analyses focused on identifying the likelihood of major adverse cardiovascular events (MACE) two years after treatment and highlighting the potential benefits of using physiological assessments to guide PCI treatment decisions.
In the first subset of the FAVOR III trial, researchers explored the impact of using physiological assessments, such as QFR, to guide treatment decisions aimed at restoring blood flow. Before treatment, physicians identified which blood vessels would receive PCI as part of a pre-randomisation treatment plan. Patients were randomly assigned to either the QFR-guided or angiography-guided procedure group. Following treatment, researchers retrospectively evaluated both the original plans and final procedures using QFR to determine whether they aligned with physiological measurements.
Results showed that patients whose treatment strategies were aligned with physiology, particularly those guided by QFR, experienced fewer MACE. Specifically, the QFR-guided group with physiology-concordant plans had a primary MACE rate of 8.9%, compared to 10.4% in the angiography-guided group with physiology-concordant plans.
Among patients whose original treatment plans were not physiologically concordant, only 7.6% in the QFR-guided group experienced MACE, compared to 17% in the angiography-guided group. Furthermore, across all patients, those whose final treatment strategies matched physiological assessments had a lower two-year MACE rate than those whose treatments did not (8.8% vs. 17.2%).
In another subset of the III trial, researchers developed a machine learning prognostic model for identifying the clinical, angiographic, and procedural features of patients likely to have two-year MACE despite online QFR-guided PCI.
Using data from 1,913 patients, researchers tested eight different machine learning algorithms, each trained on 47 clinical features, to predict the likelihood of two-year MACE. Key variables included estimated glomerular filtration rate (eGFR, a measure of kidney function), BMI, LVEF (or how much blood the left ventricle pumps out with each heartbeat), stent length, age, lesion location, anatomic SYNTAX score (a tool used to assess the complexity of coronary artery disease), and history of PCI.
Findings revealed that 161 patients (8.4%) experienced MACE within two years of QFR-guided PCI. Among all variables examined, eGFR emerged as the strongest predictor, followed by BMI. Notably, in a subset of 375 patients whose planned PCI was deferred, researchers identified the lesion location involving LCX, or the location of the coronary blockage, as the most significant risk factor for future events.
“The FAVOR III trial continues to provide valuable information that encourages prioritisation of personalised treatments for patients and the use of artificial intelligence,” said Hao-Yu Wang (Fuwai Hospital, National Center for Cardiovascular Diseases, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China), lead author of the study. “By identifying key risk factors like kidney function, BMI, LVEF, and lesion location involving LCX, we’re moving towards better predicting and preventing heart complications after PCI.”
