Author + information
- Published online December 18, 2017.
- Arnold H. Seto, MD, MPA∗ ()
- Department of Medicine, Veterans Administration Long Beach Health Care System, University of California, Irvine, Long Beach, California
- ↵∗Address for correspondence:
Dr. Arnold H. Seto, Veterans Administration Long Beach Health Care System, University of California, Irvine, 5901 East 7th Street, Long Beach, California 90822.
In philosophy, epistemology is the study of knowledge. Are there limits to what we can know, and are some things ultimately unknowable? What are the most reliable sources of knowledge? The rationalist would suggest that our reason is the best way to derive knowledge, as our senses can and often do fool us. The empiricist would argue that only our experience and data bring us closer to reality.
Modern medicine is based on science, and science is based on experiments that provide reproducible empirical evidence that a theory derived from reason corresponds to reality. Science thus uses both our reason and experience to derive knowledge, and has proven to be the most useful and pragmatic system for knowledge creation. Fractional flow reserve (FFR) is an excellent example of science in action. After direct measurements of coronary flow proved unwieldy in detecting ischemia, Nico Pijls and Bernard De Bruyne derived a pressure-based index of the ischemic potential of a stenosis, FFR. At its core is the theory that with (and only with) maximal hyperemia, the resistance of a coronary artery distal to a stenosis is minimized, allowing distal to aortic pressure (Pd/Pa) ratio to approximate the ratio of flow in the stenotic artery compared with a normal one. FFR has been tested in multiple clinical studies and trials, such as the DEFER (Fractional flow reserve to determine the appropriateness of angioplasty in moderate coronary stenosis: a randomized trial), FAME (Fractional Flow Reserve versus Angiography for Multivessel Evaluation), and FAME2 trials, and demonstrably improves clinical outcomes by changing revascularization decisions made by angiography alone.
Recently the instantaneous wave-free ratio (iFR) was introduced to address potential limitations of the FFR technique, namely the need for induction of hyperemia with adenosine. The iFR theory suggests that measuring the distal coronary pressure only during the diastolic wave-free period, when coronary resistance is naturally stable and relatively low (though not minimal), allows measurement of an index of stenosis severity that does not require hyperemia. The iFR and FFR measurements are concordant in approximately 80% of cases. Although some observers suggest that any discordance implies that iFR is inferior to FFR or contrast FFR to estimate ischemia when using FFR as the comparator (1), iFR and FFR have comparable performance when compared with independent tests of ischemia such as myocardial perfusion imaging.
The iFR does not require adenosine administration, and because the response to adenosine varies from patient to patient, iFR may correlate more closely with coronary flow reserve than FFR. In particular, occasional patients with mild angiographic stenoses have exaggerated, high-flow responses to adenosine (e.g., from a typically nonischemic resting Pd/Pa ratio of 0.98 to a hyperemic Pd/Pa ratio or FFR of <0.80), yet are not likely to be at risk of ischemia and worse outcomes. On the other hand, without hyperemia, iFR may miss potentially ischemic lesions, particularly in the left main and left anterior descending arteries. The ischemic potential and outcomes of patients with discordant iFR and FFR values thus are of great interest. Although two large randomized controlled trials, the DEFINE-FLAIR (Functional Lesion Assessment of Intermediate Stenosis to Guide Revascularisation) and IFR-SWEDEHEART (Instantaneous Wave-free Ratio versus Fractional Flow Reserve in Patients with Stable Angina Pectoris or Acute Coronary Syndrome) trials (2,3), have demonstrated noninferior clinical outcomes with iFR-guided percutaneous coronary intervention (PCI) compared with FFR guidance, by design only 1 type of measurement could be obtained, and neither trial could explore the outcomes of patients with iFR-FFR discordance.
In this issue of JACC: Cardiovascular Interventions, Lee et al. (4) examine the outcomes of patients with iFR-FFR discordance from the 3V-FFR FRIENDS (3-vessel fractional flow reserve for the assessment of total stenosis burden and its clinical impact in patients with coronary artery disease) study, in which patients had FFR measurements in all 3 vessels and PCI was recommended for FFR ≤0.80 but ultimately left to the operator discretion. This substudy involved 821 lesions (n = 374) that were deferred from revascularization. The iFR was measured offline. The primary outcome at 2 years was major adverse cardiac events (MACE) including cardiac death, vessel-specific myocardial infarction, and vessel-specific ischemia-driven revascularization. As expected from prior studies, FFR and iFR had a strong correlation, with 91.2% of measurements being concordant (r = 0.746). Both were equally predictive of future clinical events, with concordant FFR ≤0.80 and iFR ≤0.89 being associated with 3-fold higher MACE rates, driven by ischemia-driven revascularization. The major added contribution of this study is that the 8.8% of deferred lesions (72 lesions) with discordant iFR and FFR values were not found to be at higher risk of MACE compared with lesions with concordant normal values.
The major limitation of this study is that this cohort of deferred lesions was small and low risk (average FFR of 0.90), resulting in a very small number of clinical endpoints (17 in total, only 4 in patients with discordant iFR-FFR). The study is thus markedly underpowered to detect potentially subtle differences between iFR and FFR outcomes, if they exist. Discordance between iFR and FFR is more likely to occur in the borderline ischemic zone, where any difference in MACE rates would be small and difficult to demonstrate. Assuming the proportion of events in the DEFINE-FLAIR and IFR-SWEDEHEART trials, an adequately powered study comparing MACE between iFR and FFR could require as many as 290,000 patients.
Nevertheless, this 3V FFR-FRIENDS substudy adds additional empirical evidence that iFR works as well as FFR in predicting clinical outcome from coronary stenoses, supporting the contention from randomized trials that iFR is noninferior to FFR in guiding PCI. Although the numbers are small, there is no clear evidence that iFR-FFR discordance is of clinical concern.
Given that iFR is easier to measure, does this mean that there is no longer any reason to measure FFR? Or should we keep measuring FFR because it is still relatively easy to measure, and the adverse reactions to adenosine are mild and transient? Should we measure both, or would multiple measurements that are occasionally discordant confound the operator who ultimately must make a revascularization decision?
These questions bring us back to what we are capable of knowing about ischemia and revascularization. The truth of ischemia is out there, but we cannot see it directly, only through our imperfect tests and models. There is no agreed-on gold standard for ischemia. Ischemia is a continuous function of demand; even mildly obstructive lesions could potentially cause ischemia given sufficient demand. There is no threshold value in nature for ischemia, FFR, or iFR; rather, there is a continuum of risk, of probabilistic potential of causing adverse events (5). We have an imperfect treatment (PCI) that only treats focal disease rather than the diffuse nature of atherosclerosis. Clinical outcomes are postulated as being the ultimate demonstration of ischemic risk, but are often underpowered and may be confounded by variables unrelated to the hemodynamics such as adherence to medical therapy, quality of PCI performed, and incomplete revascularization.
Speaking in generalities, a rationalist may prefer to measure FFR, as it is derived from an idealized conceptual framework where minimizing resistance with maximal hyperemia is both possible with adenosine and necessary to make pressure measurements approximate flow. From this perspective, resting pressure ratios are poor estimators of ischemia, and clinical outcomes have failed to show the superiority of FFR only because the link between ischemia and outcomes is imperfect and the available studies underpowered. In contrast, in the absence of evidence that one method is clearly superior to the other, an empiricist may favor the simpler and faster method of iFR, as the resistance during the wave-free interval appears to be low enough to obviate the need for adenosine.
In this context, iFR and FFR may represent 2 similarly valid but imperfect methods of approaching the truth of ischemia, akin to different noninvasive stress tests (Figure 1). The initial choice between iFR or FFR may justifiably vary with patient and workflow factors. iFR is a good choice if time is short, multiple vessels or lesions need interrogation, contraindications to adenosine are present, or an attenuated adenosine response is suspected (as in acute coronary syndromes, diabetes, or caffeine use). In other cases, especially high-risk left main lesions, FFR could be preferred to examine the worst possible ischemic scenario. A single test will be sufficiently positive or negative to make a decision in most cases, but in borderline cases both may be helpful to justify a particular action. Either way, after multiple clinical studies, it is increasingly clear that iFR works, and appears to work as well as FFR, to the best of our current knowledge.
↵∗ Editorials published in JACC: Cardiovascular Interventions reflect the views of the authors and do not necessarily represent the views of JACC: Cardiovascular Interventions or the American College of Cardiology.
Dr. Seto has received research grant support and speaker honoraria from Philips Volcano and ACIST Medical Systems.
- 2017 American College of Cardiology Foundation
- Johnson N.P.,
- Jeremias A.,
- Zimmermann F.M.,
- et al.
- Davies J.E.,
- Sen S.,
- Dehbi H.M.,
- et al.
- Götberg M.,
- Christiansen E.H.,
- Gudmundsdottir I.J.,
- et al.,
- iFR-SWEDEHEART Investigators
- Lee J.M.,
- Shin E.-S.,
- Nam C.-W.,
- et al.
- Johnson N.P.,
- Tóth G.G.,
- Lai D.,
- et al.