Statistics from Altmetric.com
The quantification of ischaemic myocardial burden has been the cornerstone guiding clinical decision making in patients with stable coronary artery disease (CAD). This has traditionally been assessed using non-invasive functional testing, which although effective in identifying per-territory or per-patient ischaemia, often fails to identify the lesion responsible for symptoms or an appropriate target for coronary revascularisation.1 Invasive fractional flow reserve (FFR) has filled that void in being a well-validated technique to identify lesion-specific ischaemia. Clinical trials have established the role of FFR in contemporary clinical decision making, with FFR-guided revascularisation associated with improved outcomes and healthcare costs.2 Subsequently, advances in computational fluid dynamics and artificial intelligence have given rise to a non-invasive alternative in the form of FFR derived from a standard CT coronary angiogram (FFRCT).3
There is increasing evidence supporting the accuracy of FFRCT in identifying ischaemic lesions, predicting clinical outcomes, and guiding downstream testing in real-world practice.4–7 Recent data also demonstrate that the numeric value of FFRCT represents a risk continuum, with lower FFRCT values associated with a higher rate of clinical events.6 FFRCT represents one of the many emerging biomarkers that augment the anatomical assessment of CAD on CT coronary angiography (CTCA). The assessment of atherosclerotic plaque morphology and plaque biomechanics also provides incremental clinical value beyond coronary stenosis or traditional population-based cardiovascular risk scores.8 Although CTCA provides accurate assessment of total left ventricular myocardial mass, there is limited validation of techniques to quantify fractional myocardial mass or ischaemic myocardial burden from a standard CTCA. Furthermore, on its own, myocardial mass has limited utility in guiding clinical practice over established non-invasive techniques. Therefore, the role of CT-derived myocardial mass may lie in the integration with other emerging biomarkers such as FFRCT. The potential to incorporate physiological assessment with FFRCT together with CT-derived ischaemic myocardial burden may provide a unique opportunity to improve the diagnosis and prognostication of coronary lesions in a manner not readily available with other invasive or non-invasive tests.
In this issue of Heart, Keulards et al 9 provide the first steps in this path. By comparing relative myocardial mass derived from CTCA to the distribution of myocardial blood flow determined invasively, the authors propose that their study demonstrates the feasibility and validity of determining fractional myocardial mass. These findings pave the way for integrating lesion-specific physiology by way of FFRCT with lesion-specific myocardial territory at risk (figure 1). In this study, 35 patients with normal coronaries or those with minimal atherosclerosis were referred for CTCA and invasive coronary angiography. Assuming that myocardial perfusion is homogenous and that myocardial mass is proportional to myocardial perfusion, the vessel-specific simulation of coronary flow for FFRCT was used to estimate fractional myocardial mass. This measure of fractional mass was validated by comparing against the percentage of hyperaemic myocardial blood flow measured invasively using the method of continuous thermodilution. Overall, there was high agreement (intraclass correlation=0.90) between CT-derived myocardial mass and invasive myocardial perfusion. The authors are to be congratulated on a rigorous methodology and complex invasive validation. The results go further than previous studies by providing an invasive measurement of myocardial perfusion in all three major coronary territories. Overall, the results support the development of prototype techniques for quantifying and visualising CT-derived fractional myocardial mass in a manner that could potentially inform clinical practice in the future.
Notably, the study does have a number of limitations; arguably chief among them is that the assessment was obtained in patients with no or minimal atherosclerotic disease and invasive measurements were only performed in the proximal vessels. The assumption that myocardial perfusion scales in a predictable and homogenous manner in relation to myocardial mass is based predominantly from a validation study conducted in healthy pigs.10 Whether these assumptions hold true in the presence of commonly encountered disease states such as coronary artery stenosis, left ventricular hypertrophy, and/or microvascular dysfunction remains unknown and highlights an area for further investigation.
To appreciate the novelty of these findings first requires an understanding of the differences between the physiological principles underlying invasive FFR and those behind the simulation of FFR on CTCA. It is well established that with invasive physiology, the FFR value is highly dependent on the amount of fractional myocardium supplied by the investigated artery.11 For example, a 50% stenosis in the proximal left anterior descending is more likely to be haemodynamically significant (ie, FFR positive ≤0.80) compared with a 50% stenosis in an obtuse marginal branch, by virtue of differences in the degree of myocardium subtended. In contrast, the modelling of FFRCT is dependent on total left ventricular myocardial mass rather than the fractional amount subtended.3 The total myocardial mass is used to derive total coronary flow and global baseline resistance.3 These global measures of coronary flow and resistance are then distributed along the coronary tree in relation to vessel size to derive FFRCT.3 Therefore, although myocardial mass is an input in the modelling of FFRCT, the FFRCT value on the coronary geometry is not influenced by the degree of fractional myocardial mass subtended. Take for example a scenario in which a small diagonal branch and a large ostial right coronary artery each have a downstream FFRCT of 0.75, respectively. While FFRCT is the same for both lesions and both FFRCT simulations use total myocardial mass as one of many inputs, FFRCT does not quantify the difference in myocardial territories between these two vessels. With this in mind, the findings reported by Keulards et al provide important progress for future FFRCT modelling techniques that could integrate both physiology and myocardium at risk into a single non-invasive test.
There are several potential applications for considering both FFRCT and myocardium at risk on CTCA. (1) Diagnostic accuracy: a recent study demonstrated that integrating ischaemic fractional myocardial mass with FFRCT provided incremental diagnostic performance compared with FFRCT alone.12 The benefit was most pronounced in vessels with grey zone FFRCT values between 0.70 and 0.80. These results were achieved by applying a semiquantitative, ‘one-size fits all’ estimate of myocardium at risk using the APPROACH (Alberta Provincial Project for Outcome Assessment in Coronary Heart Disease) score.12 The findings by Keulards et al provide an opportunity for a far more sophisticated and patient-specific quantification of the myocardium at risk (figure 1). Taking this further, the quantification of fractional myocardial mass could also provide more accurate modelling of vessel-specific coronary flow and therefore has potential to improve the diagnostic performance of future iterations of FFRCT. (2) Clinical utility: the integration of fractional myocardial mass may also assist in the clinical interpretation of FFRCT. As a clinical tool, the location at which FFRCT values should be read is unclear. In many studies thus far, the diagnostic and prognostic value of FFRCT was assessed by taking the most distal value.5 While this strategy can provide important insights on the ‘pattern of disease’ along the course of the vessel (ie, diffuse disease and tandem lesions), it has also been observed that FFR simulation on CTCA is associated with pressure drop-off in the distal vessel in the absence of any measurable coronary stenosis.5 Therefore, taking the most distal measurement can be associated with higher false positive FFRCT values and potentially inappropriate referral for invasive procedures. The integration of CT-derived fractional myocardial mass may provide greater clarity for the interpretation of FFRCT in clinical practice. For example, in a patient with a positive FFRCT of 0.78 in the distal vessel, appreciating that the ischaemic myocardial territory is small could provide greater confidence to the clinician to safely defer invasive angiography while still appreciating the presence of coronary disease that requires aggressive medical therapy. (3) Therapeutic guidance: FFRCT continues to mature into a technology that informs clinicians on decisions regarding both medical therapy and revascularisation. Recently, the application of virtual stenting techniques using FFRCT has been demonstrated to be feasible and highly accurate at predicting the ‘physiological gain’ of percutaneous coronary intervention.13 The integration of fractional myocardial mass with FFRCT allows for the possibility of CT-based virtual stenting to also predict the ‘myocardial gain’ with revascularisation. Therefore, what is the clinical impact of revascularisation in the setting of significant physiological but only minimal myocardial gain? Are these the patients that benefit symptomatically without a prognostic benefit? These questions are yet to be answered, but the results of this study provide the foundation towards developing the tools required to address them.
Finally, it is becoming increasingly evident that the risk of myocardial infarction and cardiac death goes beyond ischaemia alone and is likely an interplay between multiple biomarkers that can now be assessed on CTCA including high-risk plaque, wall shear stress, and inflammation8 (figure 2). The results by Keulards et al provide an important validation of CT-derived fractional myocardial mass and adds another arrow into the increasingly sophisticated diagnostic quiver available from CTCA. The future integration of multiple CT-derived biomarkers will be essential in the quest to provide clinicians with a non-invasive platform that accurately predicts cardiovascular risk and guides decision making in a more focused and impactful manner for our patients.
Contributors SLS and ARI wrote the mansucript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests SLS is supported by fellowships from the Canadian Institutes of Health Research and the Michael Smith Foundation for Health Research. ARI has received consulting honorarium from Boston Scientific and Canon Medical.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Patient consent for publication Not required.
Provenance and peer review Commissioned; internally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.