Measuring Coronary Microvascular Function: Is It Finally Ready for Prime Time?

Adrian P. Banning; Giovanni Luigi De Maria


Eur Heart J. 2019;40(28):2360-2362. 

Over the last 50 years, percutaneous coronary intervention has evolved from an unpredictable, potentially hazardous art to a routine and almost predictable science. During that journey, it is easy to see and understand the enduring obsession of the Interventional Cardiologist with the contour of the coronary arteries, and the presence or absence of obstructive atheroma. This focus has been endorsed by the use of a pressure wire to measure either the resting or hyperaemic pressure gradient across atheromatous lesions in order to decide which ones represent potential treatment targets.

Notably during this evolution, relatively little attention has been paid to the function and mechanics of the downstream coronary microcirculation. This is surprising as pathophysiological studies have consistently demonstrated that coronary microcirculation (vessel size < 500 μm) is predominately responsible for the resistance to coronary flow, and that it is coronary microcirculation that ultimately regulates myocardial perfusion via both auto- and metabolic regulation.[1] Notably, the potential for coronary microcirculation dysfunction was initially raised in 1967 by Likoff et al., who described the clinical conundrum of patients with typical angina and unobstructed coronary arteries.[2]

Unfortunately, despite an increasing number of research reports on coronary microcirculation,[3] limited progress has been made in the clinical understanding and effective treatment of patients with coronary microvascular dysfunction. This reality may have multiple explanations, including natural patient variation and the co-existence of multiple pathologies, but perhaps most importantly, the lack of absolute consistency and reproducibility between the available invasive and non-invasive tests that are able to evaluate the coronary microcirculation.

Microvascular resistance (MVR) is expression of the force that coronary microcirculation exerts against coronary flow, and is used as an index of its functional and anatomical integrity. By applying Ohm's law, MVR is expressed as the ratio of delta pressure at both ends of the coronary microvascular bed (arterial and venous) and coronary flow. Assuming that venous pressure is negligible, non-invasive techniques, such as positron emission tomography (PET) and cardiac magnetic resonance, infer the assessment of distal intracoronary pressure and are considered to provide a direct measurement of coronary flow.[4] In contrast, invasive techniques that measure coronary resistance are based on pressure wire technology. These techniques have historically relied on the assessment of surrogates of flow in order to derive MVR and cannot provide direct coronary flow measurement. Hyperaemic microvascular resistance (hMR) is measured with a Doppler wire, and it is expressed as the ratio of distal coronary pressure and hyperaemic average peak flow velocity, with the latter used as a surrogate of coronary flow.[5] In comparison, the index of microcirculatory resistance (IMR) is obtained using thermodilution, and is expressed as the product of the hyperaemic mean transit time of an intracoronary infused bolus of saline and distal coronary pressure.[6] When measuring the IMR, the transit time is used as the surrogate of coronary flow. Both of these indices have been validated against clinical outcomes,[7,8] but these indirect measures present some practical challenges that have limited penetration into routine clinical practice.

In the current issue of the European Heart Journal, Everaars et al. have validated a methodology of 'continuous thermodilution' to assess absolute coronary flow and provided, for the first time, an invasive evaluation of MVR based on actual coronary flow measurement and not on a surrogate.[9] The approach combines a pressure wire with a temperature sensor (Pressure-Wire X, Abbott), a dedicated four-hole microcatheter (Rayflow, Hexacath) for continuous infusion of saline solution at a pre-defined rate, and dedicated analysis software (CoroFlow, Coroventis). The method relies on the observations that a constant infusion of saline produces hyperaemia comparable to intravenous adenosine, and that coronary flow can be calculated by knowing the rate and temperature of infused saline, and the temperature of mixed blood with the infused saline.

The study shows significant agreement between continuous thermodilution-derived absolute flow and PET-derived myocardial blood flow (r = 0.91; P < 0.001; intraclass correlation coefficient 0.90, P < 0.001), and consequently confirms the availability of an invasive measurement of absolute coronary flow.

Inevitably, there are questions about the method and the observations. The observed sample size is small and, in an attempt to work on a 'perfect' model for validating the methodology, MVR assessment was performed only in the left anterior descending and large left circumflex arteries. Continuous thermodilution does not allow the assessment of a commonly used index as coronary flow reserve and the presence of the infusion catheter appears to slightly influence the measurement of fractional flow reserve. Most notably, the authors recognize that continuous thermodilution is not meant to guide contemporary revascularization decisions as absolute flow is dependent on subtending myocardial mass (measured using computed tomography in the study).

The implications of the study are exciting because of the potential of continuous thermodilution to provide a more accurate and reproducible assessment of coronary microvascular status and function. Microvascular dysfunction is considered to be the cause of anginal pain in some patients with ischaemia and no obstructive coronary atheroma. Now that the nature and pathogenesis of this condition is becoming clearer, clinicians are increasingly moving away from using the poorly defined and regrettably titled condition of 'syndrome X'. Continuous thermodilution-derived MVR has the potential to characterize more accurately patients with anginal pain but no obstructive coronary atheromatous disease. The recent CORonary MICrovascular Angina (CorMICA) trial showed that microvascular function assessment combined with test vasoreactivity can help to stratify patients with inducible ischaemia and non-obstructive coronary arteries (INOCA), allowing an ad hoc and personalized therapeutic approach, which resulted in a 11.7 point improvement in symptoms measured by the Seattle Angina Questionnaire.[10]

The potential impact of continuous thermodilution-derived MVR can also extend into the acute coronary syndrome setting by assessing the status and function of coronary microcirculation in patients with myocardial injury with non-obstructive coronary arteries or ST-elevation myocardial infarction (STEMI). Microvascular dysfunction occurs acutely in STEMI patients when, despite relief of the obstruction, coronary flow is impaired, and this is described as 'no reflow'. Unfortunately strategies/therapies designed to prevent and/or improve outcomes for patients with no reflow have not had a meaningful clinical impact. We have used the IMR as a tool to acutely triage patients with STEMI and guide very early (before implantation of the stent) application of a novel therapy in patients at high risk of no reflow.[11] In a non-randomized study, this IMR-based approach has appeared effective in limiting the extension of infarct size.[12] It would be interesting to assess whether continuous thermodilution-derived MVR will offer similar or even superior benefits because of its expected higher accuracy in defining the status of the microvasculature.

Continuous thermodilution promises consistency between non-invasive and invasive techniques for the assessment of coronary flow and microvascular dysfunction. First, we will need to ascertain a 'normal' range of values for continuous thermodilution-derived MVR, and a head-to-head comparison with IMR and/or hMR will be extremely interesting (Figure 1). However, it is possible to speculate that, in the near future, the measurement of MVR will become an integral step of diagnostic procedures in the catheterization laboratory. This may, in turn, facilitate the design of long overdue trials for therapies in patients with INOCA or STEMI who are at risk of 'no reflow'. Ultimately a personalized medicine targeted approach appears to be the best way forward, but on this occasion the target is not an atheromatous lesion but the downstream vascular bed, which is too small to be assessed on the angiogram yet large enough to dramatically affect some of our patients' lives.

Figure 1.

Comparison across main available indices for assessment of coronary microvascular function. APV, average peak velocity; hMR, hyperaemic microvascular resistance; IMR, index of microcirculatory resistance; LAD, left anterior descending; LCx, large left circumflex; MVR, microvascular resistance; Pd, distal pressure; Qb, coronary flow; Qi, saline infusion rate; Ti, temperature of infused saline; tTmean, mean transit time; Tmix, temperature of mixed blood and infused saline.