VERMONT: Vessel Fractional Flow Reserve (vFFR) Assessment of Stenosis Severity: A Prospective Study

ABSTRACT

Background

Vessel fractional flow reserve (vFFR) is an angiogram‐based technique validated in early studies; however, data on its real‐time diagnostic performance and integration into routine workflow remain limited.

Aims

The VERMONT study evaluated the diagnostic accuracy and time efficiency of real‐time vFFR versus conventional wire‐based FFR in detecting functionally significant coronary lesions.

Methods

We conducted a prospective, single‐center, blinded study in which vFFR was measured during coronary angiography and compared with simultaneous wire‐based FFR. A wire‐based FFR of ≤ 0.80 defined a physiologically significant lesion.

Results

In 209 patients with 225 intermediate lesions, only 20 (8.9%) of lesions were excluded from vFFR analysis. vFFR demonstrated an AUC of 0.92 (95% CI, 0.89−0.96) for detecting lesions with FFR ≤ 0.80, achieving 90% sensitivity, 79% specificity, a negative predictive value of 93%, and a positive predictive value of 74%. Interobserver agreement was excellent (r = 0.97, p < 0.001). Real‐time vFFR computation was on average 13.9 min faster than wire‐based FFR (p < 0.001).

Conclusion

Real‐time vFFR demonstrated excellent diagnostic performance with high sensitivity and NPV for identifying functionally significant intermediate lesions, supporting its use as a reliable screening tool. Importantly, this was achieved with a low exclusion rate, rapid computation time, and minimal interobserver variability, underscoring its practicality in routine clinical practice.

1. Introduction

Wire‐based fractional flow reserve (FFR) remains the gold standard for assessing the physiological significance of intermediate coronary artery lesions, yet its global adoption remains limited due to procedural complexity, cost, wire‐related complications, and adenosine‐related side‐effects. Angiogram‐based FFR techniques have emerged as promising alternatives, offering functional lesion assessment without the need for pressure wires or pharmacological hyperemia [1, 2, 3].

Vessel fractional flow reserve (vFFR) is a novel angiogram‐based FFR system that performs a three‐dimensional reconstruction of a vessel using two orthogonal cine images, and measures the pressure drop across the vessel through computational fluid dynamics [4]. It relies on linear tapering of the reference vessel diameter with side branches not included in the analysis. As opposed to quantitative flow ratio (QFR), vFFR utilizes an invasively measured aortic root pressure as an input boundary condition and eliminates the need for frame counting. Unlike other angiogram‐based FFR products, assessment of the full cardiac tree is not necessary.

Early industry‐sponsored vFFR validation studies, including FAST I, FAST‐EXTEND, and FAST II, demonstrated strong diagnostic accuracy, sensitivity, and specificity when compared with wire‐based FFR [4, 5, 6]. However, these analyses were performed offline, predominantly by core laboratories under controlled conditions, and involved relatively modest exclusion rates. As a result, there is limited evidence on real‐time diagnostic performance, computational speed, and workflow integration. This knowledge gap is particularly relevant in light of the FAVOR III Europe trial, which challenged the ESC Class IB recommendation for QFR by demonstrating higher rates of revascularization and a greater composite incidence of death, myocardial infarction, and unplanned revascularization at 1 year in the QFR‐guided group compared with the wire‐based FFR group [7].

The VERMONT (VEssel ffR assessMent Of steNosis severiTy) study was designed to address these gaps by evaluating vFFR in a real‐world, prospective setting with a focus on real‐time analysis, low exclusion rates, reproducibility, and workflow integration. By emphasizing clinical practicality, this study provides complementary evidence to prior validation trials and explores the feasibility of integrating vFFR into routine coronary physiology assessment.

2. Methods

2.1. Study Design and Population

The VERMONT study was an investigator‐initiated, single‐center, blinded, prospective observational study. It aimed to evaluate the diagnostic accuracy and time efficacy of real‐time vFFR in identifying physiologically significant coronary artery lesions, using wire‐based FFR ≤ 0.80 as the reference standard. The study was approved by the South‐West‐Sydney Local Health District human research ethics committee (study number: 2021/ETH12209) and was conducted in accordance with the National Statement on Ethical Conduct in Human Research. A full waiver of consent was approved for the study.

All patients with intermediate coronary artery disease (50%–70% stenosis) who underwent wire‐based FFR at Campbelltown Hospital (Sydney, Australia) between February 2022 and August 2023 were included. The decision to perform wire‐based FFR was made by the interventional cardiologist according to guideline‐based standard practice. FFR measurements were taken during diagnostic angiography or before planned percutaneous coronary intervention (PCI) to determine the functional significance of these lesions. Patients were excluded from analysis if their optimized angiogram images continued to display extensive vessel overlap, foreshortening, tortuosity, or technical issues that precluded vFFR computation despite multiple acquisition attempts.

2.2. Study Procedures

All procedures followed guideline‐based, standard clinical practice. Patients with intermediate coronary artery lesions selected for physiological assessment by the interventional cardiologist underwent both vFFR and wire‐based FFR during the same procedure. After recording the aortic root pressure, the coronary artery was engaged with a guide catheter, and intracoronary glyceryl trinitrate (200 mcg) was administered. At least two high‐quality angiographic projections were obtained from different angles at 15 frames per second, with further acquisitions performed as required to secure two optimal frames for vFFR analysis. A pressure wire (PressureWire X; Abbott Laboratories, Abbott Park, IL, USA) was advanced across the lesion, and FFR was measured according to guidelines, with maximal blood flow induced by intravenous adenosine infusion (140 mcg/kg/min).

vFFR and wire‐based FFR were performed independently and simultaneously, with both the interventionalist and vFFR operator blinded to each other's results. All vFFR analyses were performed by the first author (D.A.), with further technical details provided in Appendix 1. Both techniques used the same threshold for a physiologically significant lesion (FFR ≤ 0.80). Interobserver variability was assessed by a second blinded operator who repeated the vFFR analysis postprocedure.

Lesions were categorized as focal (< 20 mm), diffuse (≥ 20 mm), ostial (within 3 mm of the vessel origin), or moderately to severely calcified (visible calcium on angiography before contrast).

2.3. Study Endpoints

The primary outcome was the diagnostic accuracy of vFFR compared with wire‐based FFR for assessing intermediate coronary lesions. A wire‐based FFR of ≤ 0.80 was used to define haemodynamic significance. The secondary outcome was a comparison of the time required to calculate vFFR versus wire‐based FFR.

2.4. Sample Size

Sample size calculations were performed based on results from the FAST II study. In this study, 334 patients were enrolled to assess the diagnostic performance and accuracy of vFFR compared to wire‐based FFR, with 36% of patients having a positive wire‐based FFR of ≤ 0.80 [6]. A vFFR threshold of ≤ 0.80 was associated with a core‐lab sensitivity of 81% and a specificity of 95% in identifying a wire‐based FFR of ≤ 0.80 [6]. We aimed to describe the sensitivity and specificity of vFFR to identify a wire‐based FFR of ≤ 0.80 in the target population with a 95% confidence interval (CI) and a ± 10% margin of error. Based on this data, we aimed to enroll 165 patients. Furthermore, given an area under the curve (AUC) of 0.93 observed in FAST II, this sample size allows estimation of the AUC with a 95% CI and a ± 6.5% margin of error.

2.5. Statistical Analysis

The normality of continuous variables was assessed through visual inspection of histograms and the Shapiro−Wilk test. Variables following a normal distribution are presented as a mean ± standard deviation (SD), while those with non‐normal distributions are reported as medians with interquartile ranges (25th–75th percentile). Categorical variables are summarized as frequencies and percentages. The relationship between vFFR and wire‐based FFR, and between real‐time and postprocedure vFFR analyses, were illustrated using a scatter plot and quantified using Pearson's correlation coefficient (r). The agreement between vFFR and wire‐based FFR, and between real‐time and postprocedure vFFR analyses, were evaluated using Bland−Altman plots with corresponding 95% limits of agreement. The intraclass correlation coefficient (ICC) was used to assess interrater reliability.

The diagnostic performance of vFFR in identifying a wire‐based FFR of ≤ 0.80 was assessed by creating receiver operating characteristic (ROC) curves and calculating the AUC. Mean vFFR and Wire‐based FFR computation times are illustrated on a column graph, and a paired sample t‐test (two‐tailed) was used to evaluate the mean difference. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of vFFR ≤ 0.80 for predicting wire‐based FFR ≤ 0.80 were also determined. All statistical analyses were conducted using SPSS Statistics, Version 27 (IBM Corp., Armonk, NY, USA). Unless stated otherwise, p‐values are two‐sided, with values less than 0.05 considered statistically significant.

3. Results

3.1. Enrollment

Two hundred and nine patients with 225 lesions were enrolled; 20 (8.9%) lesions were excluded due to extensive overlap (n = 10), technical issues (n = 7), foreshortening (n = 2), or tortuosity (n = 1) of the coronary artery (Figure 1). The final analysis was based on 205 lesions (Figure 1).

Figure 1.

Flowchart of patient inclusion and exclusion.

3.2. Baseline Characteristics and Procedural Data

Baseline characteristics and procedural data are listed in Table 1. The median age was 66 years (IQR 59−74), and 150 (73%) were male. The mean body‐mass‐index (BMI) was 29 ± 6 kg/m^2, and 139 (68%) were referred for stable angina. Overall, 157 (77%) patients had hypertension, 173 (84%) had dyslipidaemia, 78 (38%) had diabetes, and 109 (54%) were either current or ex‐smokers. The target vessels were the left‐anterior‐descending/diagonal (81%), the right‐coronary‐artery (12%), and the left‐circumflex/obtuse‐marginal (7%). Coronary lesions were characterized as being diffuse (63%), focal (36%), bifurcation (16%), ostial (11%), and moderate or severely calcified (24%). Mean diameter stenosis was 44 ± 10%, lesion length 23 ± 15 mm, and minimal lumen diameter 1.6 ± 0.4 mm. The mean contour correction percentages were 8 ± 8% for image one and 10 ± 8% for image two.

Table 1.

Baseline characteristics and procedural data.

Demographics	Total (N  = 205 lesions)
Age, years, median (IQR)	66 (59−74)
Male gender, n (%)	150 (73)
Body mass index, kg/m2, mean ± SD	29 ± 6
Cardiovascular risk factors, n (%)
Hypertension	157 (77)
Dyslipidaemia	173 (84)
Type 2 diabetes	78 (38)
Current smoker	32 (16)
Ex‐smoker	77 (38)
Nonsmoker	94 (46)
Family history of ischemic heart disease	47 (23)
Peripheral vascular disease	4 (2)
Cerebrovascular accident	15 (7)
Previous myocardial infarction	24 (12)
Previous percutaneous coronary intervention	43 (21)
Previous percutaneous coronary intervention to same vessel as FFR	16 (8)
Pathology results, mean ± SD
eGFR, mL/min	74.9 (15)
Hemoglobin, g/L	140 (16)
Coronary angiography indication, n (%)
Stable angina	139 (68)
Unstable angina	26 (13)
Non‐ST elevation myocardial infarction	40 (20)
Lesion location, n (%)
Left anterior descending artery/diagonal artery	166 (81)
Right coronary artery	25 (12)
Left circumflex artery/obtuse marginal artery	14 (7)
Lesion characteristics, n (%)
Focal lesion (< 20 mm)	74 (36)
Diffuse lesion (≥ 20 mm)	130 (63)
Bifurcation lesion	32 (16)
Ostial lesion	22 (11)
Tortuous lesion	11 (5)
Moderate or severe calcification	50 (24)
3D quantitative coronary angiography, mean ± SD
Lesion length, mm	23 ± 15
Minimal luminal diameter, mm	1.6 ± 0.4
Minimal luminal area, mm2	2.1 ± 0.9
Diameter stenosis %	44 ± 10
Reference vessel diameter, mm	2.9 ± 0.6
Contour correction
Contour correction image 1, % mean ± SD	8 ± 8
Contour correction image 2, % mean ± SD	10 ± 8
FFR Indices
vFFR, mean ± SD; median (IQR)	0.79 ± 0.09
	0.81 (0.74−0.85)
FFR, mean ± SD; median (IQR)	0.81 ± 0.08
	0.82 (0.76−0.87)
vFFR ≤ 0.80, n (%)	100 (49%)
FFR ≤ 0.80, n (%)	82 (40%)

Abbreviations: eGFR, estimated glomerular filtration rate; FFR, fractional flow reserve; IQR, interquartile range; SD, standard deviation; vFFR, vessel fractional flow reserve.

3.3. Correlation and Diagnostic Performance

The mean vFFR was 0.79 ± 0.09 and wire‐based FFR was 0.81 ± 0.08 (Table 1). Overall, a value of ≤ 80 was obtained in 49% of vFFR and 40% of wire‐based FFR measurements (Table 1). A good correlation was found between vFFR and wire‐based FFR (R = 0.68, p < 0.001) with a mean bias of 0.0184 ± 0.0681 (Figure 2A). ROC curve analysis revealed that vFFR had excellent accuracy in predicting a wire‐based FFR of ≤ 0.80 (AUC 0.92; 95% CI: 0.89−0.96) (Figure 3). A vFFR threshold of ≤ 0.80 produced 90% sensitivity, 79% specificity, 74% PPV, 93% NPV, and 83% diagnostic accuracy (Figure 3). Additional subgroup analysis across specific coronary vessels and patient subsets revealed consistent correlation and AUC findings (Table 2).

Figure 2.

Scatter and Bland−Altman Plots. (A) I, Scatter plot showing the relationship between vFFR and wire‐based FFR with a vFFR threshold of ≤ 0.80. II, Bland−Altman plot of differences against the means. The mean bias is represented by the dashed black line and the 95% CI is represented by the dashed red lines. (B) I, Scatter plot showing the interobserver variability. II, Bland−Altman plot of differences against the means. The mean bias is represented by the dashed black line and the 95% CI is represented by the dashed red lines. FN, false negative; FP, false positive; TN, true negative; TP, true positive. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 3.

vFFR receiver operating characteristic (ROC) curve and diagnostic performance. ROC curve for vFFR compared to wire‐based FFR at a threshold of ≤ 0.80. AUC, area under the curve; NPV, negative predictive value; PPV, positive predictive value. [Color figure can be viewed at wileyonlinelibrary.com]

Table 2.

Subgroup analysis.

	Pearson's R	AUC [95% CI], p value
Left anterior descending artery/diagonal	0.64	0.91 [0.87−0.96], p < 0.001
Right coronary artery	0.60	0.94 [0.84−1], p < 0.05
Left circumflex artery/obtuse marginal	0.62	0.92 [0.74−1], p = 0.09
Type 2 diabetes	0.70	0.89 [0.81−0.96], p < 0.001
Current/ex‐smoker	0.66	0.95 [0.92−0.99], p < 0.001
Current smoker	0.79	0.98 [0.95−1], p < 0.001
Focal lesion (< 20 mm)	0.64	0.91 [0.85−0.97], p < 0.001
Diffuse lesion (≥ 20 mm)	0.70	0.93 [0.90−0.97], p < 0.001
Bifurcation lesion	0.56	0.91 [0.82−1], p < 0.001
Moderate or severe calcification	0.61	0.96 [0.92−1], p < 0.001

A strong correlation was found between real‐time and postprocedural vFFR (R = 0.97, p < 0.001) (Figure 2B) with an ICC of 0.98 (95% CI, 0.98−0.99) revealing excellent agreement between the two blinded observers.

3.4. Time Efficacy

Most lesions required the acquisition of only two additional optimized angiogram images (80%) before vFFR analysis, with 17% requiring three images and 4% requiring four images (Table 3). vFFR image acquisition and computation required a mean time of 3.1 ± 1 min compared to wire‐based FFR analysis which required 16.9 ± 5.4 min (Table 3). A paired sample t‐test revealed that vFFR analysis was significantly faster than wire‐based FFR analysis with a mean difference of 13.9 min (95% CI, 13.1−14.6, p < 0.001) (Figure 4).

Table 3.

Timing and image acquisition data.

Timing (min), mean ± SD
Time to obtain wire‐based FFR	16.9 ± 5.4
Time to obtain optimized vFFR images	0.9 ± 0.6
Time to compute vFFR	2.2 ± 0.7
Time to obtain optimized images and compute vFFR	3.1 ± 1.0
Number of optimized images acquired n (%)
Two	163 (80%)
Three	34 (17%)
≥ Four	8 (4%)

Abbreviations: FFR, fractional flow reserve; SD, standard deviation; vFFR, vessel fractional flow reserve.

Figure 4.

Column graph comparing time efficacy of vFFR to wire‐based FFR. vFFR time: Time taken to obtain optimized angiogram cine images and to compute vFFR. FFR Time: Time from initial FFR decision to final FFR value. The columns represent mean vFFR and FFR computation times, and the error bars represent the 95% confidence interval. The mean difference is represented by the dashed black line.

4. Discussion

Wire‐based FFR remains the gold standard for the physiological assessment of moderate coronary artery lesions [8]. However, its global adoption remains limited, with usage rates estimated at approximately 6% worldwide [9]. The VERMONT study demonstrated that vFFR provides excellent diagnostic accuracy, high sensitivity, and a strong NPV in a prospective, real‐world setting. These results, achieved with a low exclusion rate and rapid computation time, highlight vFFR′s practicality, workflow efficiency, and value as a screening tool to identify physiologically significant coronary artery lesions. Angiography‐derived FFR methods, such as vFFR, offer an attractive alternative to traditional wire‐based assessment by reducing procedural risks and avoiding adenosine‐related side effects. With growing evidence supporting their diagnostic reliability [1, 2, 3], vFFR has the potential to expand access to physiology‐guided PCI, improving patient care and procedural efficiency while enabling more selective use of invasive pressure‐wire measurements.

The VERMONT study demonstrated that vFFR offers excellent rule‐out performance. At a vFFR threshold of ≤ 0.80, sensitivity was 90% and the NPV was 93%. Raising the threshold to ≤ 0.81 improved both measures to 98%, with only two false negatives recorded. These findings exceed the rule‐out performance reported in FAST II and align with pooled sensitivity estimates from a meta‐analysis of angiography‐derived FFR systems [3, 6]. Although VERMONT reported lower specificity and PPV than FAST II, both studies showed similarly high overall diagnostic accuracy (AUC 0.91–0.93) [6]. Clinically, this supports vFFR as an effective first‐line screening tool to safely defer nonsignificant lesions while reserving invasive pressure‐wire assessment for cases with positive vFFR results.

Differences in diagnostic performance between VERMONT and FAST II likely reflect variations in patient and lesion complexity. The VERMONT cohort had a higher proportion of complex lesions, with more diffuse (63% vs. 38%), calcified (24% vs. 14%), and bifurcation (16% vs. 13%) disease, as well as longer lesion lengths, smaller luminal diameters, and more physiologically significant stenoses on vFFR (49% vs. 32%) [6]. These factors, combined with higher rates of diabetes (49% vs. 32%) and smoking, indicate a higher‐risk population with more complex anatomy, which may have contributed to false‐positive vFFR results due to challenges in precise contour delineation. Coronary microvascular dysfunction, expected to be more common in this group, has also been linked to reduced diagnostic performance and lower PPV in angiography‐derived FFR methods such as QFR [10]. Clinically, this supports vFFR as an effective screening tool but reinforces the need for confirmatory pressure‐wire assessment in complex lesions.

Differences in diagnostic performance between VERMONT and FAST II may also reflect variations in vFFR workflow and image quality. FAST II analyses were performed offline under controlled conditions, while VERMONT assessments were performed in real time, mirroring time‐pressured clinical practice. This approach may have led to more manual contour adjustments at the site of stenosis to minimize the risk of physiological underestimation, contributing to the higher rate of positive vFFR findings. VERMONT also included a broader patient spectrum, with a lower exclusion rate (9% vs. 15%) and higher mean BMI (29 ± 6 vs. 27 ± 4 kg/m²), likely resulting in more challenging image acquisition and visualization [6]. These factors highlight that vFFR maintains strong diagnostic accuracy even in complex, real‐world settings, though positive results should be interpreted cautiously and confirmed with wire‐based FFR.

Real‐time vFFR computation integrated seamlessly into the cath lab workflow, saving an average of 13.9 min compared with wire‐based FFR (p < 0.001). This is the first study to report vFFR computation time, which compares favorably with QFR and FFRangio [11, 12]. Most cases (80%) required the acquisition of only two optimized angiographic views, minimizing contrast use, and the exclusion rate (9%) was lower than that reported in FAST II (15%) and pooled QFR validation studies (18%) [6, 13]. vFFR also demonstrated excellent reproducibility, with minimal interobserver variability and strong agreement between real‐time and postprocedural analyses, supporting its reliability. Together, these findings highlight vFFR's potential to streamline cath lab workflows, reduce procedural time, and broaden access to physiology‐guided assessment.

Despite a Class IB recommendation for QFR in the 2024 ESC guidelines, the FAVOR III Europe trial demonstrated that QFR did not meet non‐inferiority criteria compared with wire‐based FFR, driven by higher rates of myocardial infarction and unplanned revascularization in the QFR‐guided group [7, 8]. QFR values averaged lower than FFR (0.81 vs. 0.84), leading to a 21% increase in revascularization [7]. These findings raise questions about potential over‐treatment, lesion misclassification, and interobserver variability, as well as whether these challenges are unique to QFR or inherent in all angiography‐based FFR techniques.

While QFR remains the most extensively studied angiography‐derived FFR modality, vFFR is a newer technology with a limited evidence base. Our prospective, real‐time study demonstrates the potential for vFFR to be utilized as a rapid and reliable screening tool for functional assessment of moderate coronary lesions. Based on our findings, lesions with a negative vFFR (> 0.80) may be safely deferred with optimal medical therapy, while positive results (≤ 0.80) should be confirmed with wire‐based FFR before revascularization. This complementary approach balances efficiency and safety, maximizing the benefits of angiography‐derived FFR while preserving wire‐based FFR as the gold standard.

Finally, these results highlight the need for robust, prospective, real‐time, head‐to‐head outcome trials to define the appropriate role of angiography‐derived FFR in guiding revascularization. Ongoing large‐scale trials, including FAST III and LIPSIA STRATEGY, will be pivotal in defining vFFR′s role in routine practice and informing future guideline recommendations.

5. Limitations

This study has several limitations. First, it was conducted at a single center, which may limit the generalizability of findings to other populations, clinical environments, or procedural settings. Second, vFFR analysis was performed in real time by a single experienced operator (the primary investigator) rather than by an independent core laboratory. Although this reflects routine practice, it may introduce operator bias, particularly in image selection and manual contour adjustments.

Third, although vFFR and wire‐based FFR were performed independently, all clinical decisions were based solely on wire‐based FFR results. This non‐randomized design prevents direct evaluation of patient outcomes or the prognostic value of a vFFR‐guided diagnostic strategy. Fourth, patients were excluded if image quality was inadequate for vFFR computation, which may underestimate challenges encountered in routine practice, especially in those with complex anatomy, severe vessel tortuosity, or suboptimal imaging.

Fifth, most lesions analyzed were located in the left anterior descending artery (81%), which may limit the generalizability of these findings to right coronary or left circumflex/obtuse marginal lesions. Finally, analyses were conducted using a single software platform, which may limit applicability to other angiography‐derived FFR systems.

6. Conclusions

In summary, the VERMONT study demonstrated that real‐time vFFR offers excellent diagnostic accuracy, high sensitivity, and strong NPV for detecting physiologically significant coronary lesions. vFFR may serve as a reliable screening tool that complements wire‐based FFR, which remains the gold standard for guiding revascularization decisions. Its rapid computation, low exclusion rate, and reproducibility support seamless workflow integration and practicality in routine care (Central Illustration 1).

Central Illustration 1.

The VERMONT study demonstrates that real‐time vFFR offers excellent diagnostic accuracy, high sensitivity, and strong negative predictive value for detecting physiologically significant coronary lesions with a low exclusion rate and rapid computation time. FFR, fractional flow reserve; NPV, negative predictive value; PPV, positive predictive value; vFFR, vessel fractional flow reserve. [Color figure can be viewed at wileyonlinelibrary.com]

6.1. Recommendations

• –vFFR may be utilized as a reliable screening tool for rapid assessment of moderate coronary artery lesions.
• –Lesions with a negative vFFR (> 0.80) may be safely deferred and managed with optimal medical therapy.
• –Lesions with a positive vFFR (≤ 0.80) should undergo confirmatory wire‐based FFR before revascularization.

Ethics Statement

Ethical approval was granted by the South Western Sydney Local Health District Human Research Ethics Committee (Study Number: 2021/ETH12209), and the study was conducted in accordance with the National Statement on Ethical Conduct in Human Research. A full waiver of consent was granted.

Conflicts of Interest

The authors declare no conflicts of interest.

Appendix 10.

All procedures were performed as per guideline‐based, standard‐of‐care practice. Patients with intermediate coronary lesions deemed appropriate for FFR by the interventional Cardiologist had both vFFR and wire‐based FFR performed. An invasive aortic root pressure was initially recorded followed by engagement of the coronary ostium with a guide‐catheter and administration of intra‐coronary glyceryl trinitrate (200 mcg). This was followed by the acquisition of two optimized orthogonal angiographic cine images of the vessel in question at least 30° apart. The images were acquired with a frame rate of 15 frames per second and a magnification of 25 cm with no table movement. In cases of significant vessel overlap, foreshortening, or suboptimal contrast opacification, additional acquisitions were obtained as necessary to ensure two optimal frames for vFFR analysis. A pressure wire was then advanced (PressureWire X; Abbott Laboratories, Abbott Park, IL, USA) across the lesion and FFR was performed as per guidelines under maximum hyperemia achieved by continuous intravenous infusion of adenosine at 140 mcg/kg/min through a peripheral intravenous cannula for at least 2 min.

Both vFFR and wire‐based FFR analyses were performed independently and simultaneously in a real‐world setting with both the interventionalist and vFFR operator blinded. All vFFR computations were performed by the first author (D.A.). For patients included in the study, two optimized orthogonal images were transferred to the CAAS Workstation 8.2 (Pie Medical Imaging) for processing. Optimal end‐diastolic frames were initially identified automatically on both images through electrocardiogram triggering, with frame adjustment permitted. Vessel contour tracing was then performed automatically from the ostium to the position of the pressure wire sensor (3 cm from the wire tip). This was followed by manual contour correction as required. An identical anatomical point on both images was then identified, usually at the site of stenosis. Finally, the aortic root pressure was input to obtain the vFFR value and three‐dimensional‐quantitative‐coronary‐angiography (3D‐QCA) anatomical measurements including obstruction length, lesion diameter, mean luminal area, and percentage stenosis. vFFR and wire‐based FFR both had the same threshold value of ≤ 0.80. To assess interobserver variability, a second observer, who was blinded to the preceding results, repeated vFFR computation postprocedure.

The timing of wire‐based FFR commenced from when the decision was made to perform FFR and ended when the final FFR value was achieved. The timing of vFFR commenced when the decision was made to perform FFR and ended when the final vFFR value was obtained, which included the time required to acquire ≥ two optimized angiogram images. A focal lesion was defined as a lesion < 20 mm in length, and a diffuse lesion was defined as a lesion ≥ 20 mm in length without focal lesions. An ostial lesion was defined as a lesion starting up to 3 mm from the coronary origin. Moderate‐severe calcification was defined as radio‐opaque densities visible with or without heart motion before contrast injection.

Akrawi D., Kadappu K., Xu J., et al., “VERMONT: Vessel Fractional Flow Reserve (vFFR) Assessment of Stenosis Severity: A Prospective Study,” Catheterization and Cardiovascular Interventions 106 (2025): 3757‐3765, 10.1002/ccd.70250.