Quantitative Flow Ratio on Radial Artery Graft Outcome After Coronary Artery Bypass Grafting

Introduction:

The radial artery (RA) was first used by Carpentier for coronary artery bypass grafting (CABG) in 1971 because of a number of potential advantages, including ease of harvesting, a low propensity for wound infection, a larger diameter than other arterial grafts, and a thick, muscular wall that facilitates the construction of an anastomosis. However, early experience suggested that RA grafts were prone to spasm and functional occlusion, and their use was abandoned for many years. In the past decades, the advent of drug therapy to prevent graft spasm and the adoption of newer harvesting techniques revitalized the interest the use of radial artery. Recently several randomized trials has been conducted to prove the better graft patency over the saphenous vein (SV) and survival benefit was also observed as well. However, due to its unneglectable vulnerability to competitive flow, the recent American and European guidelines for coronary revascularization both limited the use of RA only in significantly stenosed vessels.

Fractional flow reserve (FFR) is the current gold standard to measure the physiological significance of coronary stenosis and the potential for competitive flow. The quantitative flow ratio (QFR) is a novel, intelligent, noninvasive method that enables efficient computation of the FFR from coronary angiography in good concordance with catheter-based FFR. QFR-guided percutaneous coronary intervention (PCI) has been used and showed the improved clinical outcomes in FAVOR III China trial (Comparison of Quantitative Flow Ratio-Guided and Angiography-Guided Percutaneous InterVention in Patients With cORonary Artery Disease).

The Impact of Preoperative Quantitative Flow Ratio on Radial-Artery Graft Outcome after Coronary Artery Bypass Grafting (ASRAB-QUARGO) study aims to investigate whether the preoperative QFR measurement is associated with RA graft patency 6 months after CABG, and to explore the best QFR cut-off value for guiding RA-CABG.

Methods:

Study design- The ASRAB-QUARGO is a single-centre, prospective, double-blind, observational sub-study of the ASRAB-pilot trial (NCT04310995). The study was registered and approved by the ethics committee in out institution. Informed consent was waived under permission from the ethics committee.

Outcome- The primary outcome is the association between preoperative QFR of target vessel and the RA graft outcome at 6 months after CABG. The secondary outcome is the association between preoperative QFR of target vessel and the RA graft outcome at 7 days after CABG.

Study procedures- The complete eligibility criteria for ASRAB-pilot trial is provided in the "eligibility" section. After successfully receiving primary isolated CABG, the patients from the ASRAB-pilot cohort with preoperative CAG images available for QFR analysis were enrolled in this study. QFR analysis was conducted in all vessels grafted based on preoperative CAG images. The images were sent to the core lab (CardHemo, Med-X Research Institute, Shanghai Jiao Tong University, Shanghai, China) for computation of the QFR. The analysis was performed by the experienced analysts using the AngioPlus system (Pulse Medical Imaging Technology, Shanghai, China) as described [11-15]. QFR results were recorded by the core lab and blinded to the patient and surgeon. Apart from the investigational anti-spastic drug from the ASRAB-pilot trial, optimized medical treatment was pursued according to the current American and European guidelines, including smoking cessation counselling and the administration of antiplatelet agents, beta blockers, lipid medications, and angiotensin-converting enzyme inhibitors.

Patients underwent follow-up coronary computer tomography angiography （CCTA）6 months after surgery. Angiographic evaluations were performed by two observers (one radiologist and one cardiac surgeon) blinded to the preoperative QFR values. In the case of ambiguity or disagreement between observers, additional CAG is conducted as possible for further confirmation. For sequential grafts, each segment between two adjacent anastomoses (graft-to-graft anastomosis not included) will be defined as an independent graft and evaluated independently. For composite grafts (eg. T or Y grafts), only RAs directly anastomosed to target vessel were considered as RA grafts. Angiographic patency was graded referring to the FitzGibbon classification: 4 Grade A for widely patent, Grade B for patent with flow limited, Grade S for string sign and Grade O for occluded. Grade B, O and S were considered as diseased. MACE was defined as a composite of all-cause death, nonfatal myocardial infarction, nonfatal stroke and unplanned coronary revascularization.

Sample size- Historical data from our center indicate that graft disease (Grade B, O \& S) occurs in around 20% of grafts 6 months after CABG. We estimated an RA graft disease rate of 10% if QFR ≤ best cut-off value (0.50-0.55 assumed), and 30% if QFR > best cut-off valve (0.5-0.55 assumed). With a power of 0.80 at an alpha level of 0.05, 118 RA grafts would be required to detect a statistically significant difference between groups. Assuming a 10% dropout rate and an average of 1.2 anastomosis per patient, sample size was defined at 110 patients. Therefore, the cohort size of ASRAB-pilot study of 150 patients is deemed adequate.

Analytic design and statistical analysis- Continuous variables were reported as mean, standard deviation, median, interquartile, minimum or maximum. For discrete categorical data, statistical description was presented as count and percentage. Missing data was treated as random missing. Unless otherwise stated, missing data was not filled in in this study.

Considering the possible heterogeneity between the sample population and the actual population for application, 2000 resampling samples were constructed through bootstrapping, and the mean value of QFR cut-off of all samples was taken as the final QFR cut-off. The determination of QFR cut-off value for a single sample adopted the minimum p-value method, taking RA, ITA or SV graft disease at 6 months after operation as endpoint index.

Considering the observational nature of this study, the cohort grouped based on the QFR cut-off may had imbalance of important factors. Therefore, the multivariate regression model and propensity scoring method were used to correct the potential confounding factors to evaluate the difference of outcome indicators between groups after the QFR cut-off grouping. The correlation between QFR and visual estimation was analyzed by the Spearman rank-order correlation coefficient. The predictive value of QFR and visual estimation for graft disease was compared by analyzing their respective ROC curve and corresponding AUC. Taking the QFR and the visual estimation as important prediction variables, a multifactor prediction model was constructed. The prediction accuracy was evaluated and compared by AUC. The prediction model was presented as nomograms, and the calibration curve and decision analysis curve were constructed.