Beyond the numbers: the challenge of FFR–guided strategies in multivessel disease undergoing CABG—think differently

Abstract

Background

Fractional flow reserve (FFR) is a cornerstone in guiding percutaneous coronary intervention (PCI) for intermediate coronary lesions, supported by robust evidence and guidelines. Its application in coronary artery bypass grafting (CABG), however, remains uncertain, with no validated thresholds or standardized integration into surgical planning.

Case summary

We report the case of a 69-year-old man with three-vessel coronary artery disease who underwent total arterial revascularization using a bilateral internal mammary artery Y-graft. Despite a pre-operative FFR of 0.78 in the left anterior descending artery (LAD), intra-operative transit-time flow measurements revealed severely impaired graft flow, prompting revision of the left internal mammary artery-LAD anastomosis. Post-revision flow remained poor. At 3-month follow-up, angiography confirmed graft failure, yet myocardial perfusion imaging showed no significant ischaemia in the LAD territory.

Discussion

This case underscores the limitations of applying FFR as a primary determinant in surgical revascularization. In the absence of validated cut-offs for CABG, reliance on PCI-based thresholds may lead to inappropriate grafting strategies, competitive flow, and early graft failure. Fractional flow reserve should be seen as a physiological adjunct—not a directive—within a broader framework of clinical, anatomical, and intra-operative assessment. Current evidence does not support replacing angiographic judgment with FFR in surgical decision-making. The integration of FFR into CABG planning requires a nuanced, individualized approach, guided by future trials and multidisciplinary expertise.

Conclusion

In surgical revascularization, the question is not whether a lesion crosses an FFR threshold, but how to best achieve durable myocardial protection.

FFR should guide strategy, not dictate it—think differently.

Learning points.

The excellent outcomes achieved with fractional flow reserve (FFR)–guided percutaneous myocardial revascularization (PCI) in chronic coronary syndrome have led to its application in patients undergoing coronary artery bypass grafting. But was this shift truly justified—even in the absence of randomized clinical trials proving the superiority of FFR guidance over angiographic assessment in surgical candidates?

1. Relying blindly on FFR in surgical revascularization is a dangerous oversimplification—surgeons must trust their eyes, not just the numbers.

2. Coronary physiology matters, but anatomy and surgical strategy matter more—especially when long-term outcomes are on the line.

Think differently: think beyond the number.

Introduction

Fractional flow reserve (FFR) is a well-validated physiological index for guiding percutaneous coronary interventions (PCIs), improving lesion selection and reducing unnecessary stenting.^1,2,3 However, its role in coronary artery bypass grafting (CABG) is less defined and increasingly debated.

Although both PCI and CABG aim to restore blood flow to ischaemic territories, they are based on different pathophysiological principles. Applying a rigid FFR threshold (e.g. 0.80) to surgical planning may fail to account for factors such as conduit behaviour, competitive flow, and future disease progression. Indeed, lesions with a negative FFR may still merit grafting, while those with borderline values may challenge assumptions about long-term graft patency.⁴

This case presents a situation where FFR conflicted with surgical reasoning, highlighting the need for an integrated approach in CABG. Numbers alone are rarely enough. Think differently—FFR should inform, not dictate.

Summary figure

Case presentation

A 69-year-old man with severe obesity (BMI > 40.16 kg/m²), hypertension, dyslipidaemia, and type 2 diabetes presented with effort-induced angina. Coronary angiography revealed triple-vessel disease: a 70%–90% stenosis in the circumflex artery, 90%–99% stenosis of the proximal right coronary artery (RCA) and posterior descending artery, and a 50%–70% stenosis of the proximal left anterior descending artery (LAD). Fractional flow reserve assessment of the LAD lesion yielded a positive result (FFR 0.78; resting value 0.91) (Figure 1).

Figure 1.

Pre-operative coronary angiography with positive fractional flow reserve assessment on IVA (anterior interventricular artery = LAD) (0.78).

Pre-operative echocardiography showed preserved left ventricular ejection fraction (54%) without regional wall motion abnormalities. Following multidisciplinary heart team discussion, complete arterial CABG was planned using a Y-graft configuration:

• LIMA (left internal mammary artery) to LAD

• RIMA (right internal mammary artery) sequentially to OM (obtuse marginal), ramus posterior lateral (RPL), and posterior descending artery (PD)

Haemodynamic conditions during intra-operative TTFM were carefully controlled. Measurements were performed under stable mean arterial pressure (>70 mmHg), without inotropic support, and during both aortic clamping and aortic declamping and after protamine administration Despite well-controlled haemodynamics, TTFM may be affected by metabolic demand, vasoreactivity, and coronary autoregulation. During aortic clamping (in the absence therefore of native coronary flow), the graft to LAD (LIMA-LAD) showed excellent flow (112 mL/min), low pulsatility index (PI: 0.1), and 0% backward flow (Figure 2). Post-declamping and cessation of cardiopulmonary bypass, TTFM revealed adequate flow in the Y-graft common segment (55 mL/min; PI: 1.9; backward flow: 0%) (Figure 3A) and the right Y-branch near the OM anastomosis (45 mL/min; PI: 2.8; backward flow: 0%) (Figure 3B). However, the LIMA-LAD segment showed a low value of flow, 8 mL/min, with elevated resistances (PI: 10.1) and backward flow > 3% (Figure 4A). Therefore, it was decided to re-clamp and revise the LIMA-LAD anastomosis. Despite improvement, the values measured with the TTFM on LIMA-LAD anastomosis (flow: 17 mL/min, PI: 6, backward flow > 3%) (Figure 4B), revealing high flow resistance downstream of the anastomosis and the presence of competitive flow.

Figure 2.

Intra-operative transit-time flow measurement assessment of the left internal mammary artery-left anterior descending artery segment during aortic clamping and vascular clamping on the right branch of the Y-graft.

Figure 3.

Intra-operative transit-time flow measurement assessment of (A) common segment of the Y-graft and (B) right branch right internal mammary artery-obtuse marginal-right coronary artery-posterior descending artery.

Figure 4.

Intra-operative transit-time flow measurement assessment of the left internal mammary artery-left anterior descending artery bypass with vascular clamping on the right branch of the Y-graft. (A) First anastomosis. (B) Revised anastomosis.

To exclude significant flow competition due to the presence of the Y-conduit, measurements were obtained by placing a vascular clamp on the right Y-branch to isolate the LAD graft flow. No further intervention was performed on the anastomosis.

Post-operative recovery was uneventful. Electrocardiogram and echocardiography were normal. Troponin peaked at 6328 pg/mL on Day 0, declining to 415 pg/mL by 72 h.

In view of the intra-operative TTFM values, decide to revise the bypasses with angiography at 3 months which showed

• Proximal RCA occlusion (progression rapid of disease) (Figure 5A)

• Patent right Y-branch

• Lack of opacification in the LIMA-LAD graft (likely from competitive flow) (Figure 5B; Supplementary material online, Video  S1).

Figure 5.

(A) Post-operative coronary angiography at 3 months: occlusion of the proximal segment of the right coronary artery. (B) Post-operative coronary angiography at 3 months: lack of opacification of the left internal mammary artery-left anterior descending artery segment.

To help clarify the situation, a myocardial scintigraphy (Supplementary material online, Figure S1A–D) was requested, which demonstrated a trivial apical ischaemia (<10% of LV myocardium) of low clinical significance, with preserved global ventricular function and no electrical abnormalities. In light of the scintigraphy results, it was decided to proceed with medical therapy alone (ASA 100 mg, clopidogrel 75 mg, bisoprolol 2,5 mg, metformin 1000, Trulicity, gliclazide 60 mg, irbesartan 300 mg, atorvastatin 80 mg).

Discussion

The role of FFR in guiding PCIs is well established; however, its application in CABG remains controversial due to distinct haemodynamic principles and anatomical complexities specific to surgical revascularization. While both techniques share the common goal of restoring blood flow to ischaemic territories, they rely on different pathophysiological foundations. In CABG, key factors must be considered:

• The graft and the native coronary artery form a parallel circuit.

• Graft flow must exceed native coronary flow to remain functional.

In parallel systems, flow preferentially follows the path of least resistance. If the native artery presents lower resistance than the graft, competitive flow may occur, potentially leading to early graft failure. This challenges the direct application of PCI-derived FFR cut-offs to surgical decision-making.⁵

The main question is whether FFR-guided strategies offer benefit in CABG. Without clearly defined criteria, using a fixed threshold (e.g. FFR ≤ 0.80) may result in inappropriate grafting decisions.^5,8

The IMPAG trial,⁴ a prospective study of 65 patients and 114 arterial grafts, demonstrated significantly higher patency at 6 months in grafts to vessels with FFR ≤ 0.78 (95% vs. 47%, P < 0.001). However, its limited sample size, short follow-up, and exclusive focus on arterial conduits restrict generalizability. A subsequent post hoc analysis⁵ refined these findings by differentiating thresholds based on graft configuration and coronary territory. For single arterial grafts, the optimal FFR cut-off was 0.74; for sequential grafts, thresholds were 0.81 for the first and 0.78 for the second anastomosis. Of note, the threshold for the second sequential anastomosis to the right coronary system was even lower (0.71), highlighting territory-specific dynamics and emphasizing that universal cut-offs may be inappropriate in the surgical context.

The GRAFFITI trial,⁶ the first randomized comparison of FFR- vs. angiography-guided CABG, enrolled 172 patients with significant LAD or left main disease and at least one intermediate lesion (30%–90% stenosis). While graft patency at 12 months was similar (FFR: 80% vs. Angio: 81%, P = 0.885), the FFR-guided group had fewer grafts (mean 2.71 vs. 3.14, P = 0.02), more off-pump procedures, and shorter post-operative stays, but no difference in clinical outcomes. Despite these findings, the trial was underpowered, with only 69% of patients undergoing follow-up imaging, and no differences were observed in clinical outcomes.

Similarly, the FARGO trial⁷ (n = 100) found no significant difference in graft patency at 6 months (FFR: 84% vs. Angio: 87%, P = 0.97). Importantly, 21% of lesions with initial FFR > 0.80 exhibited functional progression (FFR ≤ 0.80) at follow-up. This raises a critical concern: reliance on pre-operative FFR alone may lead to omission of grafts for lesions that later become functionally significant.

Collectively, these trials provide an inconclusive yet insightful perspective. The IMPAG data support FFR as a predictor of short-term arterial graft patency, but variability in optimal thresholds underscores the need for individualized assessment. GRAFFITI and FARGO failed to demonstrate clear clinical superiority of FFR-guided surgery, primarily due to sample size limitations and short follow-up durations. Both trials suggest that FFR-guided CABG may reduce the number of grafts, but without tangible benefits in clinical endpoints.

In the present case, the LAD lesion (FFR 0.78) was grafted despite the absence of ischaemia on follow-up. Intra-operative TTFM revealed poor flow and elevated resistance in the LIMA-LAD conduit, consistent with competitive flow dynamics. The anatomical characteristics of the LAD were also a critical factor¹⁰: a moderately stenotic yet large-calibre vessel (>3 mm) likely favoured dominant native perfusion and limited graft recruitment. Moreover, at follow-up coronary angiography, the LAD presented a reference diameter of approximately 3.0–3.5 mm with normal antegrade flow (TIMI 3) and no evidence of homo- or hetero-coronary collateralization. Although no formal quantification of runoff was performed, these findings support the hypothesis that competitive native flow may have contributed to graft dysfunction. Moreover, the RIMA sequentially supplied a larger myocardial territory (OM-RPL-PDA), potentially diverting flow from the LAD branch. This redistribution of flow, despite anastomotic revision, contributed to early graft dysfunction.

Interestingly, the functioning graft to the RCA may have contributed to native vessel thrombosis due to abrupt haemodynamic shifts. Together, these findings underscore how anatomical and physiological factors—not FFR alone—shape graft performance and long-term patency.

Notably, no pre-operative functional imaging, such as myocardial perfusion scintigraphy or stress echocardiography, was performed. This represents a missed opportunity that might have influenced the surgical strategy, potentially favouring the use of a single internal mammary artery instead of a Y-graft, or even the selection of a saphenous vein graft.

In borderline cases (FFR 0.75–0.80), particularly when involving the LAD—a vessel often central to surgical indication—relying solely on FFR without adjunctive anatomical or functional imaging may lead to suboptimal surgical choices. Tools such as myocardial perfusion imaging, Coronary ANGIO-TC, or intravascular imaging can help clarify the true ischaemic burden or stability of the lesion.^9,10 These modalities may help identify silent but clinically relevant ischaemia or, conversely, demonstrate a stable, non-obstructive plaque—both of which are essential to guide an individualized and physiologically sound surgical plan.

This case illustrates how a physiology-only approach can result in unnecessary surgical manoeuvres, increased perioperative risk, and early graft failure, whereas effective revascularization of other territories may ensure sufficient perfusion through collateral networks.

We believe that FFR should not be used to determine whether a vessel should be grafted, but rather how to graft it—informing conduit choice, configuration, and surgical planning based on a full understanding of coronary physiology and anatomy.

Conclusions

This case underscores the need to critically reassess the role of FFR in surgical revascularization, particularly in patients with multivessel coronary disease. Unlike PCI—where FFR quantifies the haemodynamic relevance of a focal lesion—CABG pursues a broader objective: durable myocardial protection based on anatomical complexity, conduit behaviour, and long-term flow dynamics.

The current evidence does not support replacing angiographic and clinical judgment with FFR as the primary determinant of grafting strategy. Rather than serving as a binary filter, FFR should be viewed as a physiological adjunct to inform, not dictate, surgical planning. It may help identify which vessels benefit most from arterial conduits, which may be suitable for venous grafts, and which can be safely managed conservatively—always within an integrated, patient-specific approach.

This case raises important questions that merit further investigation:

• Is the standard FFR threshold of 0.80 appropriate for surgical grafting, or should cut-offs be adapted to the realities of graft physiology and coronary territory?

• Can FFR alone guide revascularization decisions, or does it risk neglecting key anatomical and haemodynamic determinants of graft performance?

• In borderline lesions, should functional imaging or intra-operative assessment take precedence over pre-operative physiology?

• Ultimately, is FFR a decisive parameter, or a contextual tool within a broader, individualized surgical strategy?

Future randomized trials with larger populations, extended follow-up, and integration of imaging and intra-operative flow data will be essential to define the appropriate role of FFR in CABG.

In surgical revascularization, FFR should guide how we graft—not if we graft: think differently.

Take-home message

In CABG, FFR is not a threshold to obey—but a signal to interpret. Its value lies in complementing anatomical and surgical judgment—not in replacing it.

This report does not seek to provide definitive answers, but to provoke meaningful questions. It invites surgeons to rethink the role of physiology in grafting logic—moving from threshold-based decisions to strategy-driven revascularization—while awaiting results from prospective, randomized clinical trials that may better define the role of FFR in surgical planning.

Lead author biography

My name is Francesco Ferraro, a cardiac surgeon with a dedicated interest in coronary pathophysiology and surgical myocardial revascularization. I earned my medical degree from the University of Naples Federico II and completed my residency in cardiac surgery at the Università Cattolica del Sacro Cuore (Rome). My clinical training has focused on the integration of physiological assessment into surgical decision-making, particularly in the context of multivessel coronary artery disease. I am currently serving as a cardiac surgery assistant at CHRU de Nancy, Hôpitaux de Brabois, Université de Lorraine, where I am involved in both clinical practice and academic research.

Supplementary material

Supplementary material is available at European Heart Journal – Case Reports online.

Consent: Written informed consent was obtained from the patient for the publication of this case report and any accompanying images. The authors confirm that this work complies with the Committee on Publication Ethics (COPE) guidelines.

Funding

No funding

Data availability

All data generated or analysed during this study are included in this published article and its supplementary information files.