Pressure- and flow-derived indices of coronary stenosis severity: old rivals, new allies

KEYWORDS:

• Microcirculation
• coronary flow reserve
• fractional flow reserve
• coronary circulation
• coronary stenosis
• ischaemic heart disease

Although the presence of epicardial stenosis has constituted largely the focus of diagnosis and treatment in patients with coronary artery disease (CAD), myocardial ischemia in ischemic heart disease (IHD) is caused by both obstructive and non-obstructive coronary involvement. Awareness of this fact has generated a growing interest in the diagnosis of microvascular disease (MVD), which has been found to be associated with poor clinical outcomes in both unobstructed [1] and obstructed CAD [2,3] and other cardiac diseases [4–6].

This renewed interest in coronary domains beyond the epicardial vessel occurs at a time in which fractional flow reserve (FFR) has almost completely replaced coronary flow reserve (CFR) as a method of assessing stenosis severity, despite the fact that CFR was the first tool postulated for this purpose [7]. FFR is a pressure-derived estimate of maximum achievable myocardial blood flow in the presence of an epicardial stenosis, as a fraction of the maximal achievable blood flow to the distal myocardium in the absence of a stenosis [8]. FFR is based on the supposition that pressure and flow are linear when coronary resistance is minimal and steady, which justifies the use of hyperemic agents in current clinical practice. While this approach is of great practical value, FFR measurements do not provide insights into the status of non-obstructive IHD or other key contributors to total myocardial flow, such as collateral support or microcirculatory status. On the other hand, CFR infers the hemodynamic relevance of a coronary stenosis from the response of the microcirculation to a hyperemic stimulus [7,9], which will be attenuated if arterioles are already vasodilated as a result of an upstream severe stenosis. This rationale has the obvious caveat that a low CFR also will occur in cases in which the vasodilatory capacity results from microcirculatory dysfunction [10]. CFR can be measured in multiple manners; for instance, noninvasively such as with positron emission tomography and single-photon emission computed tomography. CFR can also be determined with echocardiography using high-frequency fundamental imaging or echo-contrast enhanced harmonic Doppler methods. Obviously, the main advantage of these tools is that they can be measured noninvasively and are quickly available at bedside. This topic, however, reaches beyond the scope of this editorial. For now, we will solely focus on intracoronary measured indices.

1. Interpretation of FFR and CFR discordance in both epicardial and microvascular diseases

When CFR and FFR measurements are classified dichotomously using their cutoffs (e.g. FFR ≤ 0.80 and CFR < 2.0), four different groups can be generated and considered separately. This was proposed by Johnson et al. and results in 4 quadrants [11]. There are two concordant groups or quadrants (those in which both FFR and CFR are either normal or abnormal), which pose no great difficulty in understanding. One quadrant consists of cases with hemodynamically significant stenosis (FFR ≤ 0.80 and CFR < 2.0), and the second one consists of patients with preserved coronary hemodynamics both in epicardial vessels and in the microcirculation (FFR > 0.80 and CFR ≥ 2). On the contrary, the two groups in which CFR and FFR classification disagree deserve a more detailed discussion. Recent studies have shown that the coronary vessels of patients included in these groups have a specific physiological behavior, as we will describe below.

Let’s first consider vessels with a FFR ≤ 0.80 and CFR ≥ 2.0. A key aspect to bear in mind is that FFR is a surrogate of the modification in myocardial flow resulting from the same stenosis in the subtended myocardium: how can the flow surrogate (FFR) contradict the flow estimate (CFR), and what is the physiological explanation? Myocardial function depends on coronary blood flow, and not coronary perfusion pressure alone [9,12,13]. The simplest explanation is that even a relatively mild stenosis will generate a significant translesional pressure gradient if coronary flow increases substantially. This might be the case, for example, in stenoses with a proximal vessel location or a large subtended myocardial mass. Typically, in this scenario, virtually no pressure gradient occurs at rest, but it develops during maximal hyperemia. It has been reported that, in this situation, myocardial ischemia is absent in noninvasive tests [14], supporting the concept that, in this context, FFR abnormality should be interpreted as a false positive result.

Secondly, let’s examine the group of patients with FFR > 0.80 and CFR < 2.0. The most obvious explanation for this pattern of coronary hemodynamics is the existence of MVD causing a decreased CFR with normal epicardial vessels reflecting the normal FFR values. Structural microcirculatory remodeling or microcirculatory plugging may exist downstream in these vessels. An alternative explanation can be found in the context of diffuse epicardial coronary disease, which limits hyperemic flow to the point that the pressure gradient across the stenosis is low or nonexistent. In the presence of diffuse atherosclerotic narrowing, there is a lack of convective accelerative flow and flow separation loss. This causes little pressure drop (normal FFR) in the flow-limiting, interrogated coronary segment, notwithstanding its effect on vessel conductance (which is reflected as a low CFR). Due to low flow in this situation, ∆P through the stenosis is limited and hence FFR remains relatively normal despite the presence of angiographically significant stenosis. A potential solution to differentiate between these two causes of the FFR > 0.80 and CFR < 2.0 pattern is to perform measurements of microcirculatory resistance, as a high microcirculatory resistance seems to be a distinctive feature of those cases with predominantly microcirculatory dysfunction [11,15]. Microvascular resistance can be measured invasively, for instance, by calculation of the index of microcirculatory resistance derived from thermodilution principles (distal coronary pressure divided by hyperemic mean transit time), or by the usage of a Doppler guidewire acquiring the hyperemic microvascular resistance (distal coronary pressure divided by the mean Doppler flow velocity) [16]. In the presence of both epicardial disease and MVD, low flow, and thus decreased pressure drop across the stenosis, is caused by microvascular dysfunction. In this situation, there may be a geometrically significant stenosis, but the FFR stays relatively normal.

2. Prognostic implications of FFR and CFR measurements and future directions

There is limited evidence regarding the clinical implications of the above-described CFR–FFR discordant patterns. Based on the same conceptual plot generated with CFR and FR, Van de Hoef et al. found significantly more major adverse cardiac events in patients with normal FFR and abnormal CFR values at 1, 3, 5, and 10-year follow-up compared to patients with significant stenosis and normal CFR values [17]. Interestingly, in the same study, patients with abnormal FFR but preserved CFR had an excellent prognosis in the long term, supporting the view outlined above on the safety of deferring PCI if a preserved CFR is documented. Based on this information, a prospective clinical trial (DEFINE-FLOW, NCT02328820) [18], which investigates FFR and CFR discordance in patients with CAD, has been launched and will provide insights into the clinical relevance of FFR–CFR classifications.

Perhaps the term ‘discordance’ between CFR and FFR should be abandoned as it has negative connotations; instead, the concept of ‘multimodal physiology’ should be promoted. The combined use of FFR and CFR, envisaged as complementary rather than competing techniques, may help in providing a more comprehensive overview of the coronary status in clinical decision-making, outlining the dominant mechanisms of coronary dysfunction and selecting the most appropriate therapeutic attitude for the individual patient.

Declaration of interest

J Escaned is a consultant for Philips, St Jude Medical, Volcano Corporation and Boston Scientific. N van Royen has received unrestricted research grants from Volcano Corporation. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.