Li et al. have recently reported the clinical outcome data of all patients who underwent percutaneous coronary intervention (PCI) at the Mayo Clinic between 2002 and 2009, except those undergoing primary PCI for ST-elevation myocardial infarction (STEMI).1 The authors focus on the importance of fractional flow reserve (FFR)-guided decision-making. They show very convincingly that, after excluding patients in whom PCI was deferred in a vessel with an FFR value between 0.75 and 0.80, the incidence of death or MI was significantly lower in the FFR-guided group than in patients in whom PCI was performed, mainly based on the available clinical information and the angiogram. These data confirm many other reports, randomized or not.2–6
The retrospective nature of this single-centre observational registry is considered by the authors themselves to be the first limitation of their work. It is true that the lack of randomization may lead to patient populations with unequal baseline characteristics. One might observe that, indeed, the most sophisticated multivariate and propensity score analyses will not be able to unravel the influence of all biases related to personal decisions of different operators to measure or not measure FFR to guide PCI. One might disdainfully comment that, in these kinds of registries, the guidelines on how to use FFR for individual decision-making are ill defined and evolve over time. Scornfully, one might regret that the endpoints are not pre-defined nor the events adjudicated by an independent event committee and—shockingly!—that the FFR tracings were not checked by a core laboratory. One might indeed regret … what we never do in daily practice.
However, this would be forgetting the extraordinary importance of carefully reporting long-term clinical outcomes of patients treated in ‘daily practice’, i.e. as we all do every day. Two aspects are particularly important. First, the study is based on the clinical database and reporting system used in this laboratory. By itself, this ascertains the consecutive character of the patients included in the analysis. All patients were analysed, not only those who were cherry-picked for some randomized controlled trial. This consecutive character contrasts with the very slow recruitment rate of many randomized controlled trials. In the COURAGE study, the average inclusion rate was <1 patient per month and per centre.7 What became of all the other patients with stable coronary artery disease? These figures strongly suggest the presence of a ‘silent majority’ consisting of those patients not even considered for randomization for all kind of reasons, i.e. biases. This phenomenon drastically limits the generalizability of the conclusions of these trials and—in contrast—emphasizes the importance of well-conducted registries with consecutive patients. Secondly, the completeness of the patients' follow-up in Li's report is exemplary. All patients having undergone a PCI were tracked via telephone calls at 6 and 12 months and annually thereafter. Only 4.2% became lost to follow-up. This is twice as good as in some controlled, randomized trials7 for a similar duration of follow-up. Accurate assessment of long-term outcome of PCI patients is as important a duty as the careful preparation of the procedural details.
In approximately half of the patients with at least one lesion associated with an FFR value between 0.75 and 0.80, often referred to as the ‘grey zone’, the decision was taken by the operator not to proceed with PCI. When these patients were withdrawn from the analysis, the differences between patients in whom FFR-guided PCI was performed and those in whom PCI was performed without FFR measurements became markedly more pronounced. This is an important observation. It indicates that many patients with at least one haemodynamically significant stenosis left untreated did not fare well. This confirms the detrimental influence of incomplete revascularization.8,9 More generally, it also re-establishes the value of 0.80 as the clinically important threshold value for FFR. In the early days of FFR, the threshold value of 0.75 was proposed. The latter corresponded to the optimal diagnostic accuracy of FFR in distinguishing between the presence or absence of ischaemia at non-invasive stress testing.10,11 Although a stringent statistical approach was used in these validation studies, it became apparent that in a sizable number of cases, reversible myocardial ischaemia could be associated with FFR values up to 0.80.11 In addition, Legalery et al.12 indicated that withholding revascularization of stenoses with FFR below 0.80 was associated with poor outcome. Therefore, the clinically important FFR threshold was raised to 0.80.2–5 Li's data tend to confirm the appropriateness of this value.1
Traditionally, the diagnosis/definition of coronary artery disease is based on the presence of at least one stenosis of at least 50% in at least one epicardial artery at coronary angiography. Despite the general awareness that this is a battered gold standard,13 and even though it became commonplace to invoke the need for both anatomy and function for appropriate decision-making,14 the presence of a ≥50% diameter stenosis is still used to justify revascularization, to risk-stratify patients, to validate non-invasive techniques, and to serve as an endpoint in studies on revascularization strategies. Not surprisingly, a recent ‘state-of-the-art review’ by Marzilli et al.15 went on at length about the disappointingly elusive link between the degree of stenosis severity and clinical outcome. These authors called for nothing less than a ‘Copernican revolution’ of our understanding of the mechanistic link between coronary atherosclerosis and clinical outcome. These authors are right to remind us that the microvasculature, the myocardial cell, the platelets and the coagulation, the inflammation, and the endothelial dysfunction all play a role in the pathophysiology of coronary artery disease and its consequences for the patients. It should be understood, however, that the weakness of the definition of a disease is the first cause of the elusiveness of its link with clinical outcome. Thus, using the combination of a high quality angiogram and FFR measurements to quantify accurately and localize precisely the abnormal epicardial resistance should be a first and more realistic step towards a better definition and treatment of coronary artery disease, instead of basing our decisions on a pernicious relic. Li's paper reminds us that these tools have been at hand for several years and that their usage significantly improves patients' outcome.
Why wouldn't we use them? After all these years, finally.
Conflict of interest: none declared.
The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.
. Long-term follow-up after fractional flow reserve-guided treatment strategy in patients with an isolated proximal left anterior descending coronary artery stenosis. JACC Cardiovasc Interv 2011;4:1175-1182.
. Impact of the presence and extent of incomplete angiographic revascularization after percutaneous coronary intervention in acute coronary syndromes: the Acute Catheterization and Urgent Intervention Triage Strategy (ACUITY) trial. Circulation 2012;125:2613-2620.
. The negative impact of incomplete angiographic revascularization on clinical outcomes and its association with total occlusions in the SYNTAX (Synergy Between Percutaneous Coronary Intervention with Taxus and Cardiac Surgery) trial. J Am Coll Cardiol 2012;1097:5372-5377.