Reproducibility of quantitative flow ratio: the QREP study

Artificial intelligence-based quantitative coronary angiography of major vessels using deep-learning

BACKGROUND

Quantitative coronary angiography (QCA) offers objective and reproducible measures of coronary lesions. However, significant inter- and intra-observer variability and time-consuming processes hinder the practical application of on-site QCA in the current clinical setting. This study proposes a novel method for artificial intelligence-based QCA (AI-QCA) analysis of the major vessels and evaluates its performance.

METHODS

AI-QCA was developed using three deep-learning models trained on 7658 angiographic images from 3129 patients for the precise delineation of lumen boundaries. An automated quantification method, employing refined matching for accurate diameter calculation and iterative updates of diameter trend lines, was embedded in the AI-QCA. A separate dataset of 676 coronary angiography images from 370 patients was retrospectively analyzed to compare AI-QCA with manual QCA performed by expert analysts. A match was considered between manual and AI-QCA lesions when the minimum lumen diameter (MLD) location identified manually coincided with the location identified by AI-QCA. Matched lesions were evaluated in terms of diameter stenosis (DS), MLD, reference lumen diameter (RLD), and lesion length (LL).

RESULTS

AI-QCA exhibited a sensitivity of 89% in lesion detection and strong correlations with manual QCA for DS, MLD, RLD, and LL. Among 995 matched lesions, most cases (892 cases, 80%) exhibited DS differences ≤10%. Multiple lesions of the major vessels were accurately identified and quantitatively analyzed without manual corrections.

CONCLUSION

AI-QCA demonstrates promise as an automated tool for analysis in coronary angiography, offering potential advantages for the quantitative assessment of coronary lesions and clinical decision-making.

Modelled impact of virtual fractional flow reserve in patients undergoing coronary angiography (VIRTU-4)

BACKGROUND

The practical application of 'virtual' (computed) fractional flow reserve (vFFR) based on invasive coronary angiogram (ICA) images is unknown. The objective of this cohort study was to investigate the potential of vFFR to guide the management of unselected patients undergoing ICA. The hypothesis was that it changes management in >10% of cases.

METHODS

vFFR was computed using the Sheffield VIRTUheart system, at five hospitals in the North of England, on 'all-comers' undergoing ICA for non-ST-elevation myocardial infarction acute coronary syndrome (ACS) and chronic coronary syndrome (CCS). The cardiologists' management plan (optimal medical therapy, percutaneous coronary intervention (PCI), coronary artery bypass surgery or 'more information required') and confidence level were recorded after ICA, and again after vFFR disclosure.

RESULTS

517 patients were screened; 320 were recruited: 208 with ACS and 112 with CCS. The median vFFR was 0.82 (0.70-0.91). vFFR disclosure did not change the mean number of significantly stenosed vessels per patient (1.16 (±0.96) visually and 1.18 (±0.92) with vFFR (p=0.79)). A change in intended management following vFFR disclosure occurred in 22% of all patients; in the ACS cohort, there was a 62% increase in the number planned for medical management, and in the CCS cohort, there was a 31% increase in the number planned for PCI. In all patients, vFFR disclosure increased physician confidence from 8 of 10 (7.33-9) to 9 of 10 (8-10) (p<0.001).

CONCLUSION

The addition of vFFR to ICA changed intended management strategy in 22% of patients, provided a detailed and specific 'all-in-one' anatomical and physiological assessment of coronary artery disease, and was accompanied by augmentation of the operator's confidence in the treatment strategy.

Quantitative flow ratio versus fractional flow reserve for Heart Team decision-making in multivessel disease: the randomised, multicentre DECISION QFR trial

BACKGROUND

Vessel-level physiological data derived from pressure wire measurements are one of the important determinant factors in the optimal revascularisation strategy for patients with multivessel disease (MVD). However, these may result in complications and a prolonged procedure time.

AIMS

The feasibility of using the quantitative flow ratio (QFR), an angiography-derived fractional flow reserve (FFR), in Heart Team discussions to determine the optimal revascularisation strategy for patients with MVD was investigated.

METHODS

Two Heart Teams were randomly assigned either QFR- or FFR-based data of the included patients. They then discussed the optimal revascularisation mode (percutaneous coronary intervention [PCI] or coronary artery bypass grafting [CABG]) for each patient and made treatment recommendations. The primary endpoint of the trial was the level of agreement between the treatment recommendations of both teams as assessed using Cohen's kappa.

RESULTS

The trial included 248 patients with MVD from 10 study sites. Cohen's kappa in the recommended revascularisation modes between the QFR and FFR approaches was 0.73 [95% confidence interval {CI} : 0.62-0.83]. As for the revascularisation planning, agreements in the target vessels for PCI and CABG were substantial for both revascularisation modes (Cohen's kappa=0.72 [95% CI: 0.66-0.78] and 0.72 [95% CI: 0.66-0.78], respectively). The team assigned to the QFR approach provided consistent recommended revascularisation modes even after being made aware of the FFR data (Cohen's kappa=0.95 [95% CI:0.90-1.00]).

CONCLUSIONS

QFR provided feasible physiological data in Heart Team discussions to determine the optimal revascularisation strategy for MVD. The QFR and FFR approaches agreed substantially in terms of treatment recommendations.

Consensus document on the clinical application of invasive functional coronary angiography from the Japanese Association of Cardiovascular Intervention and Therapeutics

Invasive functional coronary angiography (FCA), an angiography-derived physiological index of the functional significance of coronary obstruction, is a novel physiological assessment tool for coronary obstruction that does not require the utilization of a pressure wire. This technology enables operators to rapidly evaluate the functional relevance of coronary stenoses during and even after angiography while reducing the burden of cost and complication risks related to the pressure wire. FCA can be used for treatment decision-making for revascularization, strategy planning for percutaneous coronary intervention, and procedure optimization. Currently, various software-computing FCAs are available worldwide, with unique features in their computation algorithms and functions. With the emerging application of this novel technology in various clinical scenarios, the Japanese Association of Cardiovascular Intervention and Therapeutics task force was created to outline expert consensus on the clinical use of FCA. This consensus document advocates optimal clinical applications of FCA according to currently available evidence while summarizing the concept, history, limitations, and future perspectives of FCA along with globally available software.

Accuracy and Reproducibility of Coronary Angiography-Derived Fractional Flow Reserve in the Assessment of Coronary Lesion Severity

Purpose

Coronary angiography-derived fractional flow reserve (caFFR) is a novel computational flow dynamics (CFD)-derived assessment of coronary vessel flow with good diagnostic performance. Herein, we performed a retrospective study to evaluate the reproducibility of caFFR findings between observers and investigate the diagnostic performance of caFFR for coronary stenosis defined as FFR ≤0.80, especially in the grey zone (0.75≤caFFR ≤0.80).

Patients and Methods

A total of 150 patients (167 coronary vessels) underwent caFFR (with FlashAngio used for calculation of flow variables) and subsequent invasive fractional flow reserve (FFR) measurements. Outcomes, including reproducibility, were compared for vessels in and outside the grey zone.

Results

The correlation of caFFR findings was good between the two laboratories (r = 0.723, p<0.001). The AUC of ROC were both high for caFFR-CoreLab1 and caFFR-CoreLab2 (0.975 and 0.883). The diagnostic accuracy, sensitivity, specificity, and negative and positive predictive values were not significantly different between the two laboratories (p>0.05). caFFR had a strong correlation with measures to FFR (r=0.911, p<0.001). There was no systematic difference between caFFR and FFR on Bland-Altman analysis in and outside the grey zone. There was no difference in diagnostic accuracy between the grey and non-grey zones in the prediction of FFR ≤0.80 (p=0.09).

Conclusion

The inter-observer reproducibility for caFFR was high, and the diagnostic accuracy of caFFR was good compared to that of FFR.

Danish study of Non-Invasive Testing in Coronary Artery Disease 3 (Dan-NICAD 3): study design of a controlled study on optimal diagnostic strategy

INTRODUCTION

Current guideline recommend functional imaging for myocardial ischaemia if coronary CT angiography (CTA) has shown coronary artery disease (CAD) of uncertain functional significance. However, diagnostic accuracy of selective myocardial perfusion imaging after coronary CTA is currently unclear. The Danish study of Non-Invasive testing in Coronary Artery Disease 3 trial is designed to evaluate head to head the diagnostic accuracy of myocardial perfusion imaging with positron emission tomography (PET) using the tracers ⁸²Rubidium (⁸²Rb-PET) compared with oxygen-15 labelled water PET (¹⁵O-water-PET) in patients with symptoms of obstructive CAD and a coronary CT scan with suspected obstructive CAD.

METHODS AND ANALYSIS

This prospective, multicentre, cross-sectional study will include approximately 1000 symptomatic patients without previous CAD. Patients are included after referral to coronary CTA. All patients undergo a structured interview and blood is sampled for genetic and proteomic analysis and a coronary CTA. Patients with possible obstructive CAD at coronary CTA are examined with both ⁸²Rb-PET, ¹⁵O-water-PET and invasive coronary angiography with three-vessel fractional flow reserve and thermodilution measurements of coronary flow reserve. After enrolment, patients are followed with Seattle Angina Questionnaires and follow-up PET scans in patients with an initially abnormal PET scan and for cardiovascular events in 10 years.

ETHICS AND DISSEMINATION

Ethical approval was obtained from Danish regional committee on health research ethics. Written informed consent will be provided by all study participants. Results of this study will be disseminated via articles in international peer-reviewed journal.

TRIAL REGISTRATION NUMBER

NCT04707859.

Anonymous Comparison of Various Angiography-Derived Fractional Flow Reserve Software With Pressure-Derived Physiological Assessment

BACKGROUND

Software to compute angiography-derived fractional flow reserve (angio-FFR) have been validated against pressure wire-derived fractional flow reserve (PW-FFR) with an area under the receiver-operating characteristic curve (AUC) of 0.93 to 0.97.

OBJECTIVES

The aim of this study was to investigate diagnostic accuracies of 5 angio-FFR software/methods by an independent core lab in a prospective cohort of 390 vessels with carefully documented sites of PW-FFR and pressure wire-derived instantaneous wave-free ratio.

METHODS

One "matcher investigator" colocalized on angiography the sites of pressure wire measurement with angio-FFR measurements and provided the same 2 optimal angiographic views and frame selection to independent analysts who were blinded to invasive physiological results and results from other software. The results were anonymized and randomly presented. The AUC of each angio-FFR was compared with 2-dimensional quantitative coronary angiography (QCA) percent diameter stenosis (%DS) using a 2-tailed paired comparison of AUC.

RESULTS

All 5 software/methods yielded a high proportion of analyzable vessels (A: 100%, B: 100%, C: 92.1%, D: 99.5%, and E: 92.1%). The AUCs for predicting fractional flow reserve ≤0.8 for software A, B, C, D, E, and 2-dimensional QCA %DS were 0.75, 0.74, 0.74, 0.73, 0.73, and 0.65, respectively. The AUC for each angio-FFR was significantly greater than that for 2-dimensional QCA %DS.

CONCLUSIONS

This head-to-head comparison by an independent core lab demonstrated that the diagnostic accuracy of various angio-FFR software for predicting PW-FFR ≤0.80 was useful, with a higher discrimination compared with 2-dimensional QCA %DS; however, it did not reach the diagnostic accuracy previously reported in validation studies of various vendors. Therefore, the intrinsic clinical value of "angiography-derived fractional flow reserve" requires confirmation in large clinical trials.

Accuracy of the angiography-based quantitative flow ratio in intermediate left main coronary artery lesions and comparison with visual estimation

BACKGROUND

Revascularization of left main coronary artery (LMCA) stenosis is mostly based on angiography. Indices based on angiography might increase accuracy of the decision, although they have been scarcely used in LMCA. The objective of this study is to study the diagnostic agreement of QFR (quantitative flow ratio) with wire-based fractional flow reserve (FFR) in LMCA lesions and to compare with visual severity assessment.

METHODS

In a series of patients with invasive FFR assessment of intermediate LMCA stenoses we retrospectively compared the measured value of QFR with that of FFR and the estimate of significance from angiography.

RESULTS

107 QFR studies were included. The QFR intra-observer and inter-observer agreement was 87% and 82% respectively. The mean QFR-FFR difference was 0.047 ± 0.05 with a concordance of 90.7%, sensitivity 88.1%, specificity 92.3%, positive predictive value 88.1% and negative predictive value 92.3%. All these values were superior to those observed with the visual estimation which showed an intra- and inter-observer agreement of 73% and 72% respectively, besides 78% with the FFR value. The low diagnostic performance of the visual estimation and the acceptable performance of the QFR index measurement were observed in all subgroups analysed.

CONCLUSIONS

QFR allows an acceptable estimate of the FFR obtained with intracoronary pressure guidewire in intermediate LMCA lesions, and clearly superior to the assessment based on angiography alone. The decision to revascularize patients with moderate LMCA lesions should not be based solely on the degree of angiographic stenosis.

Non-invasive Angiographic-based Fractional Flow Reserve: Technical Development, Clinical Implications, and Future Perspectives

New non- and less-invasive techniques have been developed to overcome the procedural and operator related burden of the fractional flow reserve (FFR) for the assessment of potentially significant stenosis in the coronary arteries. Virtual FFR-techniques can obviate the need for the additional flow or pressure wires as used for FFR measurements. This review provides an overview of the developments and validation of the virtual FFR-algorithms, states the challenges, discusses the upcoming clinical trials, and postulates the future role of virtual FFR in the clinical practice.

Diagnostic performance of angiography-derived fractional flow reserve in patients with NSTEMI

Noninvasive methods of estimating invasively measured fractional flow reserve (FFRinvasive ) are actively being explored, aiming to avoid the use of an invasive pressure wire and the administration of hyperemia-inducing drugs. Coronary angiography-derived FFR (FFRangio ) has already demonstrated its diagnostic performance in the context of stable coronary artery disease. However, its applicability in the context of non-ST-segment elevation myocardial infarction (NSTEMI) has yet to be established. We sought to determine the diagnostic performance of FFRangio exclusively in patients presenting with NSTEMI. We performed a prospective, single-center, single-arm, double-blinded study comparing FFR calculated by FFRangio to FFRinvasive in NSTEMI patients. FFRinvasive was measured in all angiographically intermediate lesions (30%-70% stenosis) and was then compared to FFRangio which was calculated at the same position, by a blinded operator. The primary endpoints were the sensitivity and specificity of FFRangio for predicting FFRinvasive using a cut-off value of ≤0.80. Among 100 NSTEMI patients who were screened, 46 patients with 60 vessels in total underwent FFRinvasive and were included in the study. The mean value of FFRinvasive was 0.83 ± 0.3 with 22 (36%) being ≤0.80 while the mean FFRangio was 0.82 ± 0.1 with 22 (36%) being ≤0.80. FFRangio exhibited a sensitivity of 95.5%, a specificity of 97.4%, and a diagnostic accuracy of 96.7%. FFRangio can precisely and noninvasively estimate FFRinvasive in acute coronary syndromes and may have a role in guiding treatment decisions related to angiographically intermediate coronary lesions in this context.

WHAT IS KNOWN

FFRangio has demonstrated its diagnostic performance in validation studies, as a noninvasive and cost-effective method in the context of stable coronary artery disease but its performance has never been exclusively evaluated in NSTEMI patients.

WHAT IS NEW

The present prospective single-center study demonstrates the excellent diagnostic performance of FFRangio in detecting functionally significant coronary artery stenosis in the setting of NSTEMI, providing more confidence in utilizing FFRangio in this population, avoiding the risk of an invasive pressure wire and the administration of hyperemia-inducing drugs.

WHAT IS NEXT

Future randomized trials evaluating FFRangio -guided treatment of coronary artery disease (stable or ACS) are now needed to definitively establish the role of FFRangio in the physiological assessment of coronary lesions.

Diagnostic comparison of automatic and manual TIMI frame-counting-generated quantitative flow ratio (QFR) values

Quantitative flow ratio (QFR) is a computational measurement of FFR (fractional flow reserve), calculated from coronary angiography. Latest QFR software automates TIMI frame counting (TFC), which occurs during the flow step of QFR analyses, making the analysis faster and more reproducible. The objective is to determine the diagnostic performance of QFR values obtained from analyses using automatic TFC compared to those obtained from analyses using manual TFC. This was a single-arm clinical trial that used the prospective analysis of the coronary angiographic image series of 97 patients who underwent evaluation of stable coronary artery disease with FFR/iFR at MedStar Washington Hospital Center in Washington, DC, USA. Automatic and manual TFC QFR values were obtained from the analyses of each of the 97 patients' image series, with manual TFC QFR values as the current gold standard for comparison. The diagnostic performance of automatic TFC QFR values was measured as follows: sensitivity was 0.87 (95% CI 0.66-0.97) and specificity was 1.00 (95% CI 0.9514-1.00), positive predictive value (PPV) was 1.00 (95%CI 1.00-1.00), while the NPV was 0.96 (95% CI 0.96-0.99). Overall accuracy was 96.91% (95% CI 91.23%-99.36%). The agreement as illustrated by the Bland-Altman plot shows a bias of 0.0023 (SD 0.0208) and narrow limits of agreement (LOA): Upper LOA 0.0573 and Lower LOA - 0.0528. The area under curve (AUC) was 0.996. QFR values generated from automatic TFC are comparable to those generated from manual TFC in diagnostic capability. The most recent software update produces values equivalent to those of the previous manual option, and can therefore be used interchangeably.

Angiography-Based Fractional Flow Reserve: State of the Art

Purpose of Review

Three-dimensional quantitative coronary angiography-based methods of fractional flow reserve (FFR) derivation have emerged as an appealing alternative to conventional pressure-wire-based physiological lesion assessment and have the potential to further extend the use of physiology in general. Here, we summarize the current evidence related to angiography-based FFR and perspectives on future developments.

Recent Findings

Growing evidence suggests good diagnostic performance of angiography-based FFR measurements, both in chronic and acute coronary syndromes, as well as in specific lesion subsets, such as long and calcified lesions, left main coronary stenosis, and bifurcations. More recently, promising results on the superiority of angiography-based FFR as compared to angiography-guided PCI have been published.

Summary

Currently available angiography -FFR indices proved to be an excellent alternative to invasive pressure wire-based FFR. Dedicated prospective outcome data comparing these indices to routine guideline recommended PCI including the use of FFR are eagerly awaited.

Heart Team risk assessment with angiography-derived fractional flow reserve determining the optimal revascularization strategy in patients with multivessel disease: Trial design and rationale for the DECISION QFR randomized trial

In patients with multivessel disease (MVD), functional information on lesions improves the prognostic capability of the SYNTAX score. Quantitative flow ratio (QFR®) is an angiography‐derived fractional flow reserve (FFR) that does not require a pressure wire or pharmacological hyperemia. We aimed to investigate the feasibility of QFR‐based patient information in Heart Teams' discussions to determine the optimal revascularization strategy for patients with MVD. We hypothesized that there is an acceptable agreement between treatment recommendations based on the QFR approach and recommendation based on the FFR approach. The DECISION QFR study is a prospective, multicenter, randomized controlled trial that will include patients with MVD who require revascularization. Two Heart Teams comprising cardiologists and cardiac surgeons will be randomized to select a revascularization strategy (percutaneous coronary intervention or coronary artery bypass graft) according to patient information either based on QFR or on FFR. All 260 patients will be assessed by both teams with reference to the anatomical and functional SYNTAX score/SYNTAX score II 2020 derived from the allocated physiological index (QFR or FFR). The primary endpoint of the trial is the level of agreement between the treatment recommendations of both teams, assessed using Cohen's κ. As of March 2022, the patient enrollment has been completed and 230 patients have been discussed in both Heart Teams. The current trial will indicate the usefulness of QFR, which enables a wireless multivessel physiological interrogation, in the discussions of Heart Teams to determine the optimal revascularization strategy for MVD.

Microvascular and Prognostic Effect in Lesions With Different Stent Expansion During Primary PCI for STEMI: Insights From Coronary Physiology and Intravascular Ultrasound

Background

While coronary stent implantation in ST-elevation myocardial infarction (STEMI) can mechanically revascularize culprit epicardial vessels, it might also cause distal embolization. The relationship between geometrical and functional results of stent expansion during the primary percutaneous coronary intervention (pPCI) is unclear.

Objective

We sought to determine the optimal stent expansion strategy in pPCI using novel angiography-based approaches including angiography-derived quantitative flow ratio (QFR)/microcirculatory resistance (MR) and intravascular ultrasound (IVUS).

Methods

Post-hoc analysis was performed in patients with acute STEMI and high thrombus burden from our prior multicenter, prospective cohort study (ChiCTR1800019923). Patients aged 18 years or older with STEMI were eligible. IVUS imaging, QFR, and MR were performed during pPCI, while stent expansion was quantified on IVUS images. The patients were divided into three subgroups depending on the degree of stent expansion as follows: overexpansion (>100%), optimal expansion (80%−100%), and underexpansion (<80%). The patients were followed up for 12 months after PCI. The primary endpoint included sudden cardiac death, myocardial infarction, stroke, unexpected hospitalization or unplanned revascularization, and all-cause death.

Results

A total of 87 patients were enrolled. The average stent expansion degree was 82% (in all patients), 117% (in overexpansion group), 88% (in optimal expansion), and 75% (in under-expansion). QFR, MR, and flow speed increased in all groups after stenting. The overall stent expansion did not affect the final QFR (p = 0.08) or MR (p = 0.09), but it reduced the final flow speed (−0.14 cm/s per 1%, p = 0.02). Under- and overexpansion did not affect final QFR (p = 0.17), MR (p = 0.16), and flow speed (p = 0.10). Multivariable Cox analysis showed that stent expansion was not the risk factor for MACE (hazard ratio, HR = 0.97, p = 0.13); however, stent expansion reduced the risk of MACE (HR = 0.95, p = 0.03) after excluding overexpansion patients. Overexpansion was an independent risk factor for no-reflow (HR = 1.27, p = 0.02) and MACE (HR = 1.45, p = 0.007). Subgroup analysis shows that mild underexpansion of 70%−80% was not a risk factor for MACE (HR = 1.11, p = 0.08) and no-reflow (HR = 1.4, p = 0.08); however, stent expansion <70% increased the risk of MACE (HR = 1.36, p = 0.04).

Conclusions

Stent expansion does not affect final QFR and MR, but it reduces flow speed in STEMI. Appropriate stent underexpansion of 70–80% does not seem to be associated with short-term prognosis, so it may be tolerable as noninferior compared with optimal expansion. Meanwhile, overexpansion and underexpansion of <70% should be avoided due to the independent risk of MACEs and no-reflow events.

Evaluation of the Safety and Efficacy of Coronary Intravascular Lithotripsy for Treatment of Severely Calcified Coronary Stenoses: Evidence From the Serial Disrupt CAD Trials

Background: Previous understanding holds that rotational atherectomy and modified balloons remain the default strategy for severely calcified coronary stenoses. In recent years, coronary intravascular lithotripsy (IVL) provides new ideas. This study was conducted to evaluate the safety and efficacy of IVL for the treatment of severely calcified coronary stenoses. Methods: The serial Disrupt CAD trials (Disrupt CAD I, Disrupt CAD II, Disrupt CAD III, and Disrupt CAD IV) were included in this study. The safety endpoint was freedom from major adverse cardiovascular events (MACE) in hospital, at 30 days, and at 6 months following the index procedure. The efficacy endpoints included procedural success and angiographic success. Optical coherence tomography (OCT) was used to evaluate the mechanism of action of IVL quantifying the coronary artery calcification (CAC) characteristics and calcium plaque fracture. Results: We enrolled a total of 628 patients with a mean age of 71.8 years, 77.1% males. In these patients, the left anterior descending artery and right coronary artery were the most vulnerable vessels. The diameter stenosis was 64.6 ± 11.6% and the lesion length was 24.2 ± 11.4 mm. IVL had a favorable efficacy (93.0% procedural success, 97.5% angiographic success, and 100.0% stent delivery). Among the 628 patients, 568, 568, and 60 reported MACE endpoints in hospital, at 30 days, and at 6 months, respectively. The results showed that 528, 514, and 55 patients were free from MACE in hospital, at 30 days, and at 6 months, respectively. OCT measurements demonstrated that calcium fracture was the underlying mechanism of action for coronary IVL. Conclusions: IVL is safe and efficient for severely calcified coronary stenoses, and, importantly, calcium fracture facilitated increased vessel compliance and favorable stent expansion.