Angiography-Derived Physiology for Coronary Artery Disease Assessment: Expert Opinion From a SCAI Roundtable

Abstract

Angiography-derived physiology (ADP) is an emerging technique that offers streamlined, wire-free assessment of epicardial coronary physiology. This is accomplished by applying computational fluid dynamics or, in some cases, mathematical formulas in conjunction with artificial intelligence to standard coronary angiographic images. This approach represents a potential paradigm shift in the catheterization laboratory workflow. Despite its early promise, obstacles to the adoption and routine use of ADP remain. This document represents expert opinion from a scientific roundtable on ADP for coronary artery assessment sponsored by the Society for Cardiovascular Angiography & Interventions. It provides a comprehensive overview of the underlying concepts of ADP, outlines its optimal clinical application based on current evidence, highlights key advantages and limitations, reviews available software solutions, and discusses future directions.

Introduction

Physiologic lesion assessment, using either fractional flow reserve (FFR) or a nonhyperemic pressure ratio (NHPR), is considered the invasive gold standard for determining the functional significance of intermediate coronary stenoses.^1,2 Despite robust data and guideline recommendations, adoption of pressure wire-based physiology remains low, with utilization rates estimated at only 10% to 20% in the United States and Europe.³ Barriers include technical complexity, the presence of drift and artifact, need for hyperemia and anticoagulation, procedural time, potential complications, cost, limited reimbursement, and insufficient operator training.⁴ Simplifying the evaluation of the functional significance of epicardial coronary lesions is clinically appealing and may increase the adoption of coronary physiologic assessment.

Angiography-derived physiology (ADP) was developed to overcome the above-outlined obstacles by estimating FFR from routine coronary angiography alone. Without requiring a pressure wire, anticoagulation, and hyperemic agents, ADP offers substantial procedural efficiency and accessibility, potentially streamlining workflow in the catheterization laboratory and increasing utilization of physiology (Central Illustration).

Central Illustration.

Angiography-derived physiology: advantages, barriers, and potential clinical applications. AS, aortic stenosis; CABG, coronary artery bypass grafting; LVH, left ventricular hypertrophy; PCI, percutaneous coronary intervention; QCA, quantitative coronary angiography; TAVR, transcatheter aortic valve replacement.

Recognizing the promise and ongoing challenges of ADP, the Society for Cardiovascular Angiography & Interventions (SCAI) convened a roundtable on May 30, 2025, in New York City. This expert panel, comprising 16 interventional cardiologists and 1 noninterventional cardiologist, reviewed the current evidence, evaluated commercial platforms, and discussed clinical applications, best practices, implementation strategies, and areas for future development. This consensus document reflects the roundtable’s key insights and expert opinions, aiming to guide clinicians, researchers, and institutions in understanding the evolving role of ADP, overcoming adoption barriers, and enhancing its clinical utility (Table 1). While SCAI sponsored the roundtable, the content of this report solely reflects the opinions of the roundtable members.

Table 1.

Roundtable expert opinion—key points.

Role of ADP	•Although ADP is not yet a replacement for wire-based physiology, it has the potential to overcome several limitations of invasive assessment, thereby increasing the adoption of physiologic evaluation and streamlining workflow
Clinical applications	•Assessment of functional severity of intermediate stenosis to guide intervention •Selection of lesions for intervention in patients with multivessel disease •Post-PCI assessment for prognostication and for identifying lesions that may warrant additional treatment •Disease pattern analysis and PCI planning using virtual pullback curves integrated with QCA for vessel sizing and lesion length
Potential applications	•Assessment of nonculprit lesions for complete revascularization in acute coronary syndrome •Assessment of coronary lesions in severe aortic stenosis without the use of vasoactive agents •Assessment of lesions in tortuous vessels without wire bias
Efforts to overcome barriers	•Further randomized clinical trials with outcomes are warranted to address evidence gap, especially given conflicting results •Continued technical advancement is required to enhance standardization and reproducibility •Societal efforts are encouraged to improve reimbursement strategies, expand education and hands-on training, and update guidelines as new evidence emerges from ongoing trials

ADP, angiography-derived physiology; PCI, percutaneous coronary intervention; QCA, quantitative coronary angiography.

Technical overview

ADP estimates FFR by integrating coronary anatomy from angiography, assumptions or estimations of coronary blood flow, and mathematical modeling to calculate the pressure gradient down a vessel. Although there are several different ADP systems available with different methodologies, all ADP systems share a common framework: (1) modeling coronary anatomy, (2) estimating flow, (3) calculating the pressure drop across the stenosis, and (4) computing an FFR value.

Modeling coronary anatomy

Most ADP systems first reconstruct a three-dimensional (3D) model of coronary anatomy using 1 or more angiographic projections and require separated views by at least 25° to 30° of angulation. Embedded DICOM metadata is commonly used to calibrate image scale and angulation, enabling the construction of a geometric mesh over which the flow and pressure equations are applied.

Estimating flow

Next, a physiological model of the coronary circulation is applied. Coronary flow is either estimated using contrast flow velocity (eg, Thrombolysis in Myocardial Infarction [TIMI] frame count), inferred from anatomical principles such as Murray’s law (an estimate of relative flow distribution based on vessel diameters), or assumed as a fixed hyperemic value. This flow input is critical for the subsequent pressure calculation, which was originally based on the application of computational fluid dynamics (CFD).

Calculating pressure drop

Although full CFD-based simulations using the Navier-Stokes equations are highly accurate,^5,6 they are computationally intensive and time-consuming. To enable practical use in clinical settings, most contemporary software platforms have adopted mathematical approximation—such as simplified CFD models or resistance-based models—that significantly reduce processing time while maintaining acceptable accuracy. Simplified CFD models use algebraic expressions and empirically derived coefficients to approximate flow and pressure relationships. In contrast, resistance-based models estimate pressure gradients using analogs of Ohm’s law (pressure = flow × resistance), with resistance derived from anatomical and physiological parameters such as vessel diameter, branching patterns, and subtended myocardial mass.

Commercially available ADP systems

Various software-computing systems for ADP have been developed, with unique features in their computation algorithms and functions. Figure 1 provides an overview of several commercially available ADP systems. Currently, 3 ADP systems—FFRangio (Fractional Flow Reserve Derived from Angiography, CAAS vFFR (Cardiovascular Angiographic Analysis System for vessel Fractional Flow Reserve), and QFR (Quantitative Flow Ratio)—are clinically available in the United States. All are reported to demonstrate comparable diagnostic accuracy with wire-based FFR and use the same ischemic threshold of 0.80.

Figure 1.

Comparison of commercially available ADP systems. ADP, angiography-derived physiology; AI, artificial intelligence; CAAS vFFR, Cardiovascular Angiographic Analysis System Vessel Fractional Flow Reserve; FFR, fractional flow reserve; FFRangio, Fractional Flow Reserve Derived from Angiography; QFR, Quantitative Flow Ratio; TIMI, Thrombolysis in Myocardial Infarction.

FFRangio

The CathWorks FFRangio System (CathWorks, Ltd) involves the creation of a 3D representation of the entire coronary arterial tree, using 2 to 3 angiographic views at least 30° apart.7, 8, 9 It applies a resistance-based lumped model, where each stenosis is represented as resistance.⁸ The software simulates hyperemic flow with and without stenoses, and the ratio of these flows yields the FFRangio value.⁸ Each segment of the coronary artery is then evaluated for its contribution to the total flow resistance and converted into a color-mapped coronary tree, with the final result revealing areas of stenosis and estimated FFR values.

FFRangio is the only technology that provides multivessel FFR values across the entire coronary tree, including side branches. It also offers simulated pullback to differentiate functional disease patterns and supports percutaneous coronary interventions (PCI) planning. However, calculation does require multiple projections and invasive aortic pressure input.

In the pivotal FAST-FFR (FFRangio Accuracy versus Standard FFR) study,⁷ diagnostic performance of FFRangio versus pressure wire-based FFR was high (sensitivity 94%, specificity 91%, and diagnostic accuracy 92%). A pooled analysis of 5 prospective cohort studies (n = 700 lesions) confirmed consistent accuracy across subgroups (sensitivity 91%, specificity 94%, and diagnostic accuracy 93%).¹⁰

Clinical outcome data for FFRangio have demonstrated its safety and effectiveness. In a multicenter cohort of 492 patients (552 lesions), FFRangio-guided revascularization showed low 1-year major adverse cardiac event (MACE) rates, consistent with wire-based FFR benchmarks—2.5% in deferred patients and 4.1% in revascularized patients.¹¹ The Prospective Randomized trial of clinical outcomes of angiography-based fractional flow reserve (FFR) guidance Versus wIre-baSed FFR guidance (PROVISION) trial further confirmed noninferior clinical outcomes between FFRangio- and wire-based FFR guided revascularization (1-year MACE: 9.9% vs 12.6%, respectively; hazard ratio [HR] 0.80, 95% CI, 0.42-1.51).¹² The ongoing Advancing Cath Lab Results With FFRangio Coronary Physiology Assessment (ALL-RISE; NCT05893498) trial is a large, multicenter, international, randomized clinical outcomes trial comparing FFRangio-guided with pressure wire-guided PCI and is expected to provide additional validation across diverse patient populations.¹³

CAAS vFFR

CAAS vFFR (Pie Medical Imaging BV) evaluates single-vessel physiology using a 3D quantitative coronary angiography (QCA) model reconstructed from 2 angiographic projections acquired at least 30° apart. A patient-specific coronary flow is estimated based on the reconstructed vessel geometry and the invasively measured aortic pressure.¹⁴ To calculate pressure drop and determine vFFR, the system assumes a constant hyperemic flow and velocity profile along the vessel, applying a simplified fluid dynamics model based on a quadratic relationship between pressure and flow.¹⁴

CAAS vFFR system offers rapid processing time and a straightforward user interface, facilitating seamless integration into catheterization workflows. However, it does not offer full-vessel modeling or advanced PCI simulation tools.

Supported by a strong correlation (r = 0.89; P < .001) and high diagnostic accuracy compared with wire-based FFR (area under the receiver operating characteristic curve [AUC], 0.93; 95% CI, 0.88-0.97) in the FAST (Fast Assessment of STenosis severity) study (n = 100), CAAS vFFR became the first ADP software application to receive clearance from the US Food and Drug Administration.^8,14 In an extended cohort of FAST study (FAST-EXTEND), excellent diagnostic performance and strong correlation between CAAS vFFR and wire-based FFR were confirmed with consistent performance across different vessels and subsets with complex anatomy.¹⁵ These positive findings were subsequently confirmed in the prospective, international, and multicenter FAST II study, which demonstrated a good correlation between vFFR as calculated by a blinded core laboratory and pressure wire-based FFR (r = 0.74; P < .001) and an excellent diagnostic accuracy of vFFR in identifying lesions with an invasive wire-based FFR ≤ 0.80 (AUC, 0.93; 95% CI, 0.90-0.96; P < .001).¹⁶ These results remained consistent in complex lesions, including bifurcations, tortuous, and calcified lesions. The safety and efficacy of a CAAS vFFR compared with an FFR guided revascularization strategy will be assessed in an ongoing multicenter, randomized FAST III trial (NCT04931771).¹⁷

QFR

QFR (Medis Medical Imaging BV) was the first commercially available method to estimate FFR by applying simplified CFD to 3D coronary artery models reconstructed from 2 angiographic projections acquired at least 25° apart. Multiple QFR variants have been developed based on how flow is estimated, with contrast-QFR (cQFR) being the most widely studied in clinical trials. cQFR uses the TIMI frame count method to estimate a patient-specific contrast flow velocity, which serves as a surrogate for hyperemic blood flow.^8,18 The pressure drop across a coronary stenosis is then calculated using a simplified fluid dynamics model that incorporates coefficients derived from the 3D vessel geometry and flow.

QFR does not require invasive pressure measurements, allowing for easier retrospective assessment. It also supports virtual stenting through pre-PCI simulation using residual QFR, whereas dQFR/ds provides detailed insights into the local hemodynamic severity of individual lesions. However, its accuracy depends on high-quality angiographic images and precise TIMI frame count calculations.

Initial validation of QFR was conducted in the Functional Analysis of Vessels Following Revascularization (FAVOR) pilot study,¹⁸ which compared 3 offline QFR algorithms—fixed-QFR, cQFR, and angio-QFR—against 3D-QCA in predicting pressure wire-based FFR, showing higher diagnostic accuracy of QFR (80%, 86%, and 87% vs 65%). Subsequent multicenter studies, FAVOR II China¹⁹ and FAVOR II Europe-Japan,²⁰ assessed online QFR and demonstrated strong diagnostic performance compared with pressure wire-based FFR.

Clinical outcomes associated with QFR have been evaluated in randomized controlled trials (RCTs). In the FAVOR III China randomized trial (n = 3825), QFR-guided PCI significantly reduced 1-year MACE (5.8% vs 8.8%; P < .001) compared with angiography-guided PCI, driven by reductions in periprocedural and spontaneous myocardial infarction (MI) and ischemia-driven revascularization (Table 2).²¹ By contrast, in FAVOR III Europe, a multicenter RCT of 2008 patients comparing QFR-guided PCI with pressure wire-based FFR guidance, QFR guidance led to more revascularizations (54.4% vs 45.8%) with more stents implanted but did not meet the noninferiority margin for the primary end point of 1-year MACE (6.7% vs 4.2%; HR, 1.63; 95% CI, 1.11-2.41), largely driven by higher MI rates (3.7% vs 2.0%; HR, 1.84; 95% CI, 1.07-3.17).²² These contrasting results provoked debate about the generalizability of ADP. In addition, the FAVOR III Europe substudy (REPEAT-QFR) reported a 71.9% diagnostic agreement between in-procedure and core laboratory QFR, highlighting potential variability and limitations of ADP in clinical practice.²³ Although QFR offers clinical utility, particularly where pressure wire use is limited, these results suggest it may not fully substitute for wire-based FFR in all settings. Given that ADP systems differ in reconstruction methods, image requirements, and flow assumptions, study findings from 1 platform do not apply to all systems, and we anxiously await the results of the ALL-RISE and FAST III trials.

Table 2.

Key clinical trials.

Trial name	System	Study design	N	Geography	Comparisons	Primary end point or key outcomes	Results
FAST-FFR7	FFRangio	Prospective, multicenter, observational	301	Global (United States, Europe, Israel)	Binary FFRangio vs pressure wire-based ​FFR	Sensitivity and specificity	Per-vessel sensitivity: 94%; Specificity: 91%; and overall diagnostic accuracy: 92%
PROVISION12,24	FFRangio	Prospective, multicenter, open-label, randomized	401	Single country (Japan)	FFRangio-guided vs pressure wire FFR guided PCI	Noninferiority in revascularization rate Cost Clinical outcomes	Noninferiority met (similar revascularization rate) Lower cost with FFRangio in both deferred and revascularized patients Similar clinical outcomes (MACE: 9.9% for FFRangio and 12.6% for wire-based FFR)
FAST14	CAAS vFFR	In vitro validation with experimental model; clinical validation with Retrospective, observational, study	100	Single-center (the Netherlands)	CAAS vFFR vs pressure wire-based FFR	Correlation between CAAS vFFR and wire-based FFR Binary diagnostic accuracy of CAAS vFFR	Good linear correlation was found between FFR and vFFR (r = 0.89; P < .001). Diagnostic accuracy: AUC, 0.93 (95% CI, 0.88-0.97)
FAST EXTEND15	CAAS vFFR	Retrospective, cohort	294	Single-center (the Netherlands)	CAAS vFFR vs pressure wire-based FFR	Same as above in an extended cohort and complex lesion subset	Strong correlation in the overall cohort (r = 0.89) and in specific lesion subsets (bifurcations [24%], r = 0.90; tortuosity [18%], r = 0.90; calcified lesions [36%], r = 0.86; tandem lesions [14%], r = 0.90; and diffuse disease [41%], r = 0.91).
FAST II16	CAAS vFFR	Prospective, observational, multicenter	334	Global (Europe, USA, Japan)	Core laboratory calculated CAAS vFFR vs pressure wire-based FFR	Binary diagnostic performance of CAAS vFFR	AUC, 0.93 (95% CI, 0.90-0.96) Diagnostic accuracy: 90%; sensitivity: 81%; specificity: 95%; PPV: 90%; NPV:90%.
FAVOR II China19	QFR	Prospective, observational, multicenter	308	Single country (China)	QFR vs pressure wire-based FFR	Binary diagnostic accuracy	Per-vessel diagnostic accuracy: 92.7%.
FAVOR II Europe-Japan20	QFR	Prospective, observational, multicenter	329	Global (Italy, the Netherlands, Germany, Poland, Spain, Japan, Denmark)	QFR vs 2D-QCA to predict positive pressure wire-based FFR	Sensitivity, specificity, and diagnostic accuracy	Sensitivity: 86.5% and specificity: 86.9% AUC, 0.92 (95% CI, 0.89-0.96).
FAVOR III China21	QFR	Multicenter, blinded, randomized, sham-controlled	3825	Single country (China)	QFR-guided vs angiography-guided PCI	1-y MACE (composite of any death, myocardial infarction, or ischemia-driven revascularization)	1-y MACE: 5.8% (QFR) vs 8.8% (angiography); P < .001.
FAVOR III Europe22	QFR	Multicenter, randomized, open-label, noninferiority	2000	European (Denmark, France, Germany, Italy, Latvia, Lithuania, Netherlands, Poland, Spain, Sweden, Switzerland)	QFR-guided- vs pressure wire-based ​FFR guided PCI	Noninferiority in 1-y MACE (composite of death, myocardial infarction, and unplanned revascularization)	QFR did not meet noninferiority to FFR. 1-y MACE: 6.7% (QFR) vs 4.2% (FFR)

AUC, area under the curve; CAAS vFFR, Cardiovascular Angiographic Analysis System Vessel Fractional Flow Reserve; FFR, fractional flow reserve; FFRangio, Fractional Flow Reserve Derived from Angiography; MACE, major adverse cardiac event; PCI, percutaneous coronary intervention; QCA, quantitative coronary angiography; QFR, Quantitative Flow Ratio.

Details of key clinical trials of commercially available ADP systems are listed in Table 2.

Advantages of ADP

The main advantage of ADP over conventional wire-based approaches lies in its minimally invasive nature. By eliminating the need for a guide catheter, pressure wire, anticoagulation, and hyperemic agent (eg, adenosine), ADP simplifies procedural workflow, shortens procedure time, improves safety, and lowers costs associated with traditional invasive physiology. It also reduces radiation exposure and contrast use. In fact, the Prospective Randomized trial of clinical outcomes of angiography-based FFR guidance Versus wIre-baSed fractIOnal flow reserve guidance (PROVISION) trial demonstrated approximately 20% lower radiation exposure with ADP compared with pressure wire-based FFR.^12,24 In addition, ADP is not affected by pressure drift or hydrostatic artifacts and can be performed retrospectively.

Beyond its minimally invasive nature, ADP offers unique advantages by combining physiologic assessment with anatomic detail. Most ADP systems allow for 3D reconstruction of the coronary tree, enabling full-vessel FFR computation and generation of virtual pressure pullback curves. These features enhance real-time clinical decision-making by supporting angiography-physiology coregistration, allowing precise localization of ischemia-producing lesions, and aiding procedural planning. This includes guidance for optimal stent placement and length selection, and in some platforms, prediction of post-PCI FFR values.

ADP also offers particular value in specific scenarios. For example, in the acute phase of ST-segment elevation myocardial infarction (STEMI) with multivessel disease, where transient microvascular dysfunction can impair wire-based FFR lesion assessment, ADP measurements in nonculprit vessels have been shown to outperform FFR itself in predicting reassessed FFR values 1 month later.²⁵ From a practical perspective, ADP allows offline analysis of nonculprit stenoses after primary PCI in STEMI, thereby avoiding recatheterization for invasive evaluation. In addition, ADP is less sensitive to friction losses than NHPR, which contributes to its higher agreement with FFR but may come at the expense of overlooking the contribution of diffuse disease to hemodynamic impairment in some patients—a factor that NHPR can reveal.²⁶ Collectively, these features highlight the potential of ADP to make physiologic lesion assessment safe, accurate, and efficient (Central Illustration). The clinical implications of ADP in specific scenarios are discussed below.

Technical considerations and limitations

ADP has important technical considerations and challenges that should be acknowledged when integrating it into clinical decision-making.

Image quality and operator dependence

ADP accuracy relies heavily on the quality of angiographic images. Poor vessel opacification, overlap, foreshortening, or inadequate separation between views can compromise 3D reconstruction, which is essential for accurate FFR simulation. A substudy of the FAVOR III Europe trial demonstrated that poor-quality angiographic images led to increased variability between in-procedure and core laboratory calculated QFR.²⁴ Variability in the quality of in-procedure QFR analysis may have impacted clinical outcomes, potentially contributing to the inability of FAVOR III Europe to demonstrate noninferiority. Therefore, it is essential for operators to obtain high-quality diagnostic angiograms when applying ADP. Moreover, although ADP is largely automated, reproducibility remains partially dependent on the operator. A study by Westra et al²⁷ showed that the inter- and intra-observer reproducibility of QFR from the same angiographic images varied among observers, influenced by angiographic quality and the severity of coronary artery stenosis as measured by FFR. However, in the FAST-FFR trial, the correlation between FFRangio and pressure wire-based FFR was similar between the site-reported values and blinded core laboratory analyses.⁷

Currently, QFR and FFRangio utilize artificial intelligence (AI) to streamline analysis and reduce user variability. The QFR incorporates AI-based end-diastolic frame selection and employs deep learning to enhance alignment between projections, ensuring accurate 3D reconstructions. The FFRangio simplifies the user experience by automatically categorizing angiograms (left or right), automatically determining the optimal frame, identifying vessels automatically, and generating user-specific prompts to better guide the operator and suggest edits. Further automation and advanced AI integration are expected to further enhance accuracy, reproducibility, and minimize operator dependence.

Discordance with wire-based physiology

Despite strong validation against FFR (AUC, 0.93-0.97), a blinded independent core laboratory retrospective analysis using multiple ADP systems (not including FFRangio) reported lower performance (AUC, 0.73-0.75).²⁸ These findings emphasize the importance of optimal image acquisition and awareness of potential discordance, particularly in gray-zone ADP values. Predictors of false positives include non-left anterior descending artery location, large reference vessel diameter, and increased microvascular resistance, whereas false negatives are more common in left anterior descending artery lesions and small reference vessels.²⁸

Complex lesion subsets and special scenarios

Anatomic lesion characteristics, such as bifurcation or severe calcification, should be considered before applying ADP, as these factors may lead to lower accuracy,²⁹ although some studies have reported reliable diagnostic performance in these settings.³⁰ In aorto-ostial stenosis, guide catheter engagement and contrast backflow can impair 3D reconstruction, potentially affecting the reliability of ADP.²⁹ Therefore, these lesions were excluded from validation studies. Similarly, left main disease was generally excluded; nevertheless, some studies have demonstrated acceptable diagnostic performance.^31,32 Overall, careful consideration is needed when applying ADP in these complex lesion subsets, and additional data are warranted. In addition, evidence remains limited in certain scenarios, such as myocardial bridging, coronary ectasia, anomalous coronary origins, chronic total occlusions with collaterals, or vessels with less than TIMI 3 flow, highlighting the need for further investigation.

Platform-specific considerations

ADP systems are not interchangeable. Differences in computational approaches (eg, flow-based vs resistance-based), input requirements (eg, contrast frame count vs invasive pressure), and software algorithms may result in different diagnostic performance. Clinical data should be interpreted in the context of platform-specific validation.

Systemic barriers

Broader adoption is limited by the lack of universal reimbursement, unclear procedural coding, and variability in software licensing. In addition, accurate interpretation requires familiarity with coronary physiology and system-specific outputs, underscoring the need for structured operator training.

Clinical applications of ADP

The roundtable panel recognized that there are several clinical scenarios in which ADP may be particularly advantageous. These potential implications are extrapolated from invasive physiology, and further studies are warranted to validate and establish their clinical role.

Identifying disease patterns and optimizing PCI planning

In addition to determining lesion significance and appropriateness of revascularization, ADP can help define functional disease patterns, discriminating focal from diffuse disease using the virtual pressure pullback curve (Figure 2).^33,34 This distinction is clinically relevant, as focal disease with a significant step-up in virtual pullback curve may be well suited for PCI with the expectation of significant postintervention physiological improvement (Figure 2A), whereas diffuse disease often yields less physiological gain despite stenting and may be better managed with medical therapy alone or alternative revascularization strategies (Figure 2B).³³

Figure 2.

ADP pullback and disease pattern. Case examples of (A) focal and (B) diffuse disease. (A) Angiogram shows 2 tandem lesions in the left circumflex artery and obtuse marginal branch. FFRangio is 0.62, and the virtual pullback (orange box) demonstrates a focal step-up (orange arrow). (B) Angiogram shows mild-to-moderate diffuse disease in the left anterior descending artery. μQFR is 0.77, and the virtual pullback (orange box) demonstrates diffuse pressure loss without a focal step-up. ADP, angiography-derived physiology; FFR, fractional flow reserve; μQFR, Murray law-based quantitative flow ratio.

ADP-derived virtual pullback enables precise identification of the location and extent of pressure loss along the vessel, as well as quantification of the contribution from individual stenoses in cases of tandem or serial lesions, thereby guiding optimal stent placement. Seki et al³⁵ reported that virtual FFR pullback curves showed moderate agreement with wire-based FFR pullbacks for the evaluation of focal versus diffuse disease (Cohen’s Kappa, 0.31; 95% CI, 0.18-0.45). Certain ADP platforms incorporate virtual stenting tools that simulate the hemodynamic effect of eliminating a pressure gradient by stenting and predict post-PCI ADP values (Figure 3). This capability has been validated against wire-based FFR measurements.^36,37 By guiding PCI planning, it helps achieve favorable post-PCI physiologic outcomes and minimize residual ischemia.

Figure 3.

ADP virtual pullback in serial lesion and PCI planning. Case example demonstrating the virtual stenting function of the FFRangio ADP system. (A) Virtual pullback shows 2 step-ups (orange arrows). The white curve represents the actual baseline FFRangio pullback, with a distal FFRangio value of 0.64. The green curve represents the modified FFRangio pullback after excluding (or treating) proximal lesion 1, yielding a distal value of 0.73. (B) Treating distal lesion 2 results in a modified FFRangio curve (green) with a distal value of 0.80. (C) Treating both lesions increases the distal FFRangio to 0.97, with no residual ischemia or focal step-up in pullback curve (green). ADP, angiography-derived physiology; FFRangio, Fractional Flow Reserve Derived from Angiography.

In addition, ADP software depicts the volumetric distribution of physiologic values on the 3D model. An interactive sizing tool allows measurement of lesion length and reference vessel diameter, with angiographic coregistration ensuring precise anatomical alignment (Figure 4). However, it should be noted that 3D QCA may underestimate vessel size compared with intravascular imaging and provides only an estimation of lumen, not vessel dimensions.

Figure 4.

ADP and anatomical lesion dimensions. Case examples of ADP systems with integrated anatomical lesion measurements. (A) The FFRangio system includes an interactive sizing tool that displays reference vessel diameter and lesion length. (B) Similarly, the CAAS vFFR system provides anatomical lesion dimensions. ADP, angiography-derived physiology; CAAS vFFR, Cardiovascular Angiographic Analysis System Vessel Fractional Flow Reserve; FFRangio, Fractional Flow Reserve Derived from Angiography.

Post-PCI assessment

In addition to predicting post-PCI physiology before intervention, ADP can also assess physiology after PCI (Figure 5). Post-PCI physiologic assessment allows detection of suboptimal results, which have been shown to have prognostic implications,³⁸ thereby providing an opportunity to optimize PCI and improve outcomes. Notably, even after angiographically successful PCI, residual ischemia—evidenced by suboptimal post-PCI instantaneous wave-free ratio values—has been reported in approximately 24% of cases.³⁹ A residual gradient within the stent may indicate the need for further optimization or postdilation, whereas a residual gradient outside the stent may suggest the need for additional stent implantation or the presence of diffuse disease.

Figure 5.

Predicted and actual post-PCI ADP. Pre-PCI ADP not only enables prediction of post-PCI physiology but can also be used to assess actual post-PCI physiology without the need for rewiring with a pressure wire. (A) This case demonstrates focal lesions in the right coronary artery. The virtual stenting function predicts a post-PCI FFRangio value of 0.96 if both lesions are treated. (B) After successful PCI, FFRangio was recalculated from actual post-PCI angiographic images, showing a value of 0.95 with no focal step-up in the pullback curve—closely matching the pre-PCI prediction. ADP, angiography-derived physiology; FFRangio, Fractional Flow Reserve Derived from Angiography; PCI, percutaneous coronary intervention.

However, wire-based post-PCI physiology remains underutilized due to the need for rewiring with or working over a pressure wire, additional procedural time at the end of the case, and operator fatigue. In this regard, ADP may facilitate broader adoption of post-PCI physiologic assessment in daily practice. Post-PCI ADP has been validated against microcatheter-based post-PCI FFR, demonstrating strong correlation and excellent diagnostic accuracy.⁴⁰ In addition, suboptimal post-PCI ADP value was associated with an increased risk of adverse outcomes.⁴¹ These findings support the use of ADP to evaluate post-PCI physiology and guide optimization of stent implantation.

Multivessel disease

Physiological assessment is crucial to guide the management of patients with multivessel disease. ADP may be more efficient and practical than wire-based assessment, especially when evaluating both the right and left coronary circulation, as it eliminates the need for wiring multiple vessels with different guide catheters.

In a study by Asano,⁴² QFR reclassified 26.1% of intermediate- and high risk lesions, emphasizing its potential to change revascularization planning and strategies. In addition, by enabling rapid calculation of the Functional SYNTAX Score—which incorporates only hemodynamically significant lesions—ADP can modify the perceived complexity of disease and, in many cases, may change the optimal revascularization strategy between PCI and coronary artery bypass grafting.⁴² In a Fractional Flow Reserve versus Angiography for Multivessel Evaluation (FAME) 3 substudy, physiologic assessment reclassified 33% of patients initially determined to have 3-vessel disease by anatomy as not having 3-vessel disease, and more than 1-quarter of patients from an SS >22 to an Functional SYNTAX Score ≤22.⁴³ These findings imply that physiologic assessment of multivessel disease, simplified and streamlined by ADP, could significantly impact the selection of optimal revascularization strategy. Further studies are warranted to establish ADP-based functional assessment of multivessel disease and its prognostic implications for clinical decision-making.

Acute coronary syndrome

In hemodynamically stable patients with acute coronary syndrome, revascularization of nonculprit lesions is recommended to improve outcomes,⁴⁴ and approximately 50% of patients presenting with MI have multivessel disease. Evidence from randomized trials has shown that physiology-guided complete revascularization improves clinical outcomes in both STEMI and non-STEMI compared with a culprit-only strategy.^45,46 In addition, the ongoing Physiology-guided vs Angiography-guided Non-Culprit Lesion Complete Revascularization for Acute MI & Multivessel Disease (COMPLETE-2) trial (NCT05701358) aims to determine whether physiology-guided complete revascularization provides additional benefits over an angiography-guided strategy in patients presenting with MI. However, operators are often reluctant to wire intermediate nonculprit lesions during the index procedure because of concerns regarding additional procedure time, contrast load, and procedural risks.

The ADP offers an appealing alternative by enabling wire-free functional assessment of nonculprit lesions directly from angiographic images—even retrospectively for staged procedures—without additional procedure. Importantly, conventional wire-based FFR assessment of nonculprit vessels during the acute MI setting can be affected by the presence of extensive microvascular dysfunction or hemodynamic disturbances like elevated left ventricular end diastolic pressure.⁴⁷ Similarly, wire-based NHPR may be falsely low due to hemodynamic perturbations and increased resting flow in nonculprit vessels during the acute presentation.⁴⁸ In contrast, ADP appears less affected by these acute transient changes. A study by Wang et al²⁵ showed that QFR remained consistent in nonculprit lesions of STEMI patients between the acute and subacute phases (30 days), whereas wire-based FFR values significantly declined. As a result, QFR demonstrated higher diagnostic accuracy in predicting functional significance at follow-up (AUC, 0.977 vs 0.901; P = .047).

In the Functional Assessment in Elderly Myocardial Infarction Patients With Multivessel Disease (FIRE) trial (N = 1445), which demonstrated improved clinical outcomes with physiology-guided complete revascularization compared with culprit-only PCI in elderly patients with MI, ADP was used in 35.2% of nonculprit vessels in the physiology-guided arm.⁴⁵ In a prespecified substudy of the FIRE trial, QFR ≤ 0.80 independently predicted vessel-oriented composite events, and outcomes for vessels assessed by QFR were comparable to those evaluated with wire-based physiology.⁴⁹ These findings suggest the feasibility of an ADP-guided revascularization strategy in patients with acute coronary syndrome with multivessel disease; however, further evidence is warranted to support its routine use given the small number of patients.

In addition, growing evidence supports the prognostic value of ADP in patients with STEMI, particularly the angiography-derived index of microcirculatory resistance (IMR), as well as its predictive value for microvascular obstruction.50, 51, 52, 53

Other clinical scenarios

Severe aortic stenosis (AS) can significantly alter coronary hemodynamics, affecting the physiologic assessment of coronary stenosis. In addition, many operators may be reluctant to use vasoactive agents in this setting, which can further limit adoption of wire-based physiology.⁵⁴ Nevertheless, treating coronary artery disease—particularly lesions with positive FFR or very severe lesions (≥90%)—has been associated with improved outcomes in patients undergoing transcatheter aortic valve replacement for AS, suggesting the clinical utility of physiologic evaluation in these complex hemodynamic conditions.⁵⁵ In this regard, ADP offers a promising modality with good diagnostic yield in assessing the functional significance of coronary lesions in AS patients.^54,56 Nevertheless, diagnostic accuracy may decline in patients with critical AS with an aortic valve area <0.60 cm²,⁵⁴ and further studies, including clinical outcomes, are warranted before routine adoption.

Similarly, coronary tortuosity poses technical limitations for pressure wire-based physiologic assessment due to difficulty navigating tortuous vessels with pressure wires and the potential for pseudostenoses caused by vessel straightening, which can result in falsely low FFR values, overestimating ischemia.⁵⁷ In this regard, ADP avoids wire manipulation and the associated distortion; however, extremely tortuous anatomy can degrade angiographic image quality and reduce ADP accuracy.

Resource-limited settings

ADP enables physiologic assessment across diverse settings, including catheterization laboratories or operators with limited resources who lack access to wire-based physiologic assessment.⁵⁸ It provides a comprehensive functional evaluation during the initial procedure, eliminating the need for interventional consultation in many cases and often avoiding unnecessary staged diagnostic procedures performed solely for invasive physiologic testing. Furthermore, anonymized screenshots and reports can be stored, shared, and accessed on the system, facilitating heart-team discussions and integrating physiologic data into patient care planning.

New technology and innovations

Several technical advancements are on the horizon. New software and systems in development include caFFR (RainMed) and AccuFFRangio (ArteryFlow); μQFR (Murray law-based QFR; Pulse Medical; Figure 2B) and X1-FFR (SpectraWAVE; Figure 6), which estimate FFR from a single angiographic view; and AngioAI+ (AngioInsight), which estimates FFR and assesses for the presence of coronary microvascular dysfunction. The X1-FFR system recently received 510(k) clearance from the US Food and Drug Administration. It analyzes vessel geometry and dynamic flow from a single angiographic view using real-time video streaming, eliminating the need for DICOM file management and enabling a rapid, streamlined workflow in the catheterization laboratory.

Figure 6.

New ADP systems. Two examples of new ADP systems are presented. (A) X1-FFR (SpectraWAVE)—An ADP system that calculates FFR from a single angiographic view using real-time video streaming of angiographic data, eliminating the need for DICOM file transfer and enabling rapid assessment in the cath lab. (B) μQFR (Pulse Medical)—A single-view ADP method that estimates FFR from 1 angiographic projection using advanced vessel reconstruction algorithms and flow estimation using Murray’s law. ADP, angiography-derived physiology; DICOM, Digital Imaging and Communications in Medicine; FFR, fractional flow reserve; μQFR, Murray law-based Quantitative Flow Ratio.

Industry is advancing toward automated and streamlined measurement, leveraging AI to enhance vessel tracing and image analysis. The next generation of ADP systems should aim at minimizing user interaction to improve reproducibility. A novel platform, AutocathFFR (MedHub), is currently being evaluated for fully automated FFR calculation from standard angiographic projections, providing real-time results in the catheterization laboratory without requiring user input.

Advances in ADP technology are also expected to improve accuracy in complex anatomies, with enhanced tools for procedural planning and post-PCI assessment. Broader integration of physiologic indices beyond FFR—including coronary flow reserve, IMR, and pullback pressure gradient—would enable more comprehensive physiologic evaluation. Angiography-derived IMR has been validated against wire-based IMR for assessing coronary microvascular function⁵⁹ and has demonstrated prognostic value in both STEMI⁵⁰ and non-ST segment elevation myocardial infarction.⁶⁰ Pullback pressure gradient can also be derived from ADP,⁶¹ helping us to differentiate focal from diffuse disease and characterize disease patterns.³³ Furthermore, hybrid approaches integrating intravascular imaging with ADP will streamline workflow and allow combined physiologic and anatomical assessment.

Future directions to address barriers and unmet needs

Alongside technical refinements to improve accuracy and reduce variability, the roundtable panel recognizes several key priorities that will be essential to support broader clinical adoption (Figure 7).

Figure 7.

Barriers to adoption of ADP and desired actions. ADP, angiography-derived physiology; AI, artificial intelligence.

Address evidence gaps

Although initial validation studies demonstrate strong correlation between ADP and wire-based physiology, additional data from large randomized controlled trials assessing clinical outcomes are needed. This is underscored by the recent FAVOR III Europe trial, which did not demonstrate noninferiority of a QFR-guided strategy compared with an FFR-guided strategy.²³ In addition, inherent limitations of ADP—such as variability in reproducibility and its strong dependence on imaging quality—necessitate further technological refinement and rigorous validation across diverse lesion types and real-world settings before its universal adoption. Ongoing trials, such as ALL-RISE (NCT05893498) and FAST III (NCT04931771), will provide further evidence for other ADP systems, as findings from 1 platform do not apply universally.

Economic viability and reimbursement

In addition to safety and outcomes, the potential efficiencies and procedural throughput afforded by ADP compared with wire-based physiology are important considerations for hospital decision-makers evaluating adoption. Data from the PROVISION study in Japan are encouraging, with an ADP strategy demonstrating approximately $400 lower resource utilization compared with an FFR strategy while achieving noninferior clinical outcomes.^12,21 Upcoming large trials will further inform the clinical efficacy and cost-effectiveness of ADP.

As the evidence base expands, global reimbursement certifications may follow, driving broader technology adoption. Currently, reimbursement remains variable and is often bundled with angiography, but US-based providers should ensure accurate billing by applying the 0523T code when ADP is used. With increasing use, the frequency threshold for complexity adjustment is likely to be met for most primary Current Procedural Terminology (CPT) codes.

When considering the cost of ADP, one has to consider the upfront cost of the system and the cost of running analyses. The system can be purchased by a hospital wholesale, or in certain circumstances, it can be installed at a lower cost in conjunction with the purchase of a specific number of analyses or duration of subscription. The analyses can be purchased on a “pay-as-you-go” structure, whereby the hospital pays a set price for use in each patient, or they can be purchased on a subscription model, whereby the hospital pays a certain price for a certain duration and can use the system in as many patients as it likes, during that time.

Education and knowledge transfer

The ADP should be systematically integrated into interventional cardiology fellowship training and into the catheterization laboratory rotations of general cardiology fellows. It should also be incorporated into hands-on workshops for practicing interventional cardiologists. Similar to wire-based physiology, completing a minimum number of ADP cases may help trainees become familiar with the software, interface, and workflow, and achieve competency. Beyond this, a thorough understanding of technical limitations—including factors that affect accuracy and variability—is essential to ensure high-quality ADP assessments. In this regard, training should emphasize optimal angiographic technique, including appropriate projections to avoid overlap and foreshortening, use of proper magnification to limit table panning, full-vessel opacification, and appropriate frame rate selection. SCAI, in collaboration with other professional societies, could establish recommendations for fellowship training requirements and incorporate educational and hands-on training sessions within scientific meetings or dedicated fellowship courses.

In addition, training catheterization lab nurses and technologists as super-users and maintaining ongoing education for physicians—including developing physician-champions—will help overcome barriers to adoption and promote consistent utilization.

Guideline recommendations

The roundtable panel recognizes that broader adoption of ADP will eventually require the incorporation of emerging evidence into professional society guidelines. The 2024 European Society of Cardiology (ESC) guidelines for chronic coronary syndromes grant QFR a class I, level B recommendation and other ADP systems a class IIb, level C recommendation for the evaluation of intermediate diameter stenoses.² However, it should be noted that the 2024 ESC Guidelines incorporated the results of the FAVOR III China²² trial but not those of FAVOR III Europe.²³ In addition, several RCTs evaluating different ADP platforms with clinical outcomes are forthcoming. In light of the expanding validation data across multiple platforms, these recommendations should evolve to address ADP as a general concept for coronary physiology assessment, encompassing all validated software. The American College of Cardiology/American Heart Association/SCAI guidelines do not yet explicitly endorse ADP systems in formal recommendations; however, as additional data have emerged since their publication—and with more expected—US guidelines should consider incorporating evidence-based recommendations that support the use of ADP in appropriate patient populations. Under this umbrella, the operator is able to choose their approved system of choice, based on performance, feasibility, and outcome data. A summary of current guideline recommendations for ADP is provided below.

Summary of guidelines for ADP

• 1.2021 American College of Cardiology/American Heart Association Guidelines for Coronary Artery Revascularization¹: ADP has not been incorporated
• 2.
  2024 ESC guidelines for the management of chronic coronary syndrome²:

  • •
    Assessment of functional severity of intermediate diameter stenoses to guide the decision to revascularize (QFR: class 1 level B; angiography-derived vessel FFR: class IIb level C).
  • •
    To guide lesion selection for intervention in patients with multivessel disease (QFR: class 1, level B).
  • •
    Considered at the end of the procedure to identify patients at high risk of persistent angina and subsequent clinical events (QFR: class IIa, level B).
  • •
    Considered at the end of the procedure to identify lesions potentially amenable to treatment with additional PCI (QFR: class IIb, level B).

Conclusion

The roundtable concluded that although ADP is not yet a replacement for wire-based FFR, it represents a significant advance in physiologic lesion evaluation. Given the recent clinical trials showing conflicting outcome results, along with technical limitations and inherent reproducibility issues, caution is warranted to ensure high-quality ADP measurements while awaiting further validation from ongoing and future clinical trials. Nevertheless, current technology demonstrates substantial potential to guide clinical decision-making and reduce the need for pressure wire assessments. Future technological iterations are expected to further simplify its application, making it an automated component of standard angiographic acquisition. Its streamlined workflow and expanding the evidence-based support broader adoption—provided institutions invest in technology integration, protocol standardization, operator training, and data-driven evaluation. With continued refinement and wider implementation, ADP has the potential to improve patient outcomes and optimize procedural decision-making. SCAI remains committed to advancing the science, clinical application, and policy alignment necessary to accelerate the integration of ADP into routine interventional practice.