Author + information
- Published online May 27, 2019.
- aWeill Cornell Medicine, New York, New York
- bIstituto di Ricovero e Cura a Carattere Scientifico (IRCCS) Foundation Policlinico San Matteo, Pavia, Italy
- cMount Sinai School of Medicine, New York, New York
- ↵∗Address for correspondence:
Dr. Leslee J. Shaw, Weill Cornell Medicine, Feil Family Research Building, RR119, 407 East 61st Street, New York, New York 10065.
Heart Failure With and Without Coronary Artery Disease
Differentiation of a flow-limiting, obstructive, epicardial coronary artery has been a central aim of diagnostic coronary angiography and a role that has historically failed to consider the burden of nonobstructive atherosclerosis. Recent evidence reveals a heightened cardiac risk among patients with nonobstructive coronary artery disease (CAD) with substantially worsening prognosis as compared with those without any luminal stenosis (1,2). In the patient with heart failure (HF), notably among those with a reduced left ventricular ejection fraction (HFrEF), the evidence has been lacking as to the prognostic significance of atherosclerotic plaque. Nonobstructive CAD in heart failure is usually equated to nonischemic dilated cardiomyopathy (NIDCM).
In this issue of JACC: Heart Failure, Braga et al. (3) reported an elevated risk of a primary composite endpoint of cardiovascular death, nonfatal myocardial infarction (MI), nonfatal stroke, or HF hospitalization among patients with nonobstructive CAD. From this large cohort of 12,814 patients with HF and an ejection fraction <35%, the relative hazard for nonobstructive CAD was 1.17 for the composite endpoint (p = 0.01) as compared with those with normal coronary arteries. As the authors state, this risk is noteworthy and has been underappreciated within the cardiovascular community (3). Importantly, this risk elevation within an HFrEF cohort extends and validates prior findings of a pattern of worsening outcomes that has been reported in much lower risk patient populations (1,2). However, these findings challenge our classical determination of risk attributed to HF resulting from obstructive CAD and support a progressive or graded increase in risk from nonobstructive to obstructive CAD.
There is emerging evidence as to the role of nonobstructive CAD and its predictive relationship with future diastolic dysfunction (4,5). This research has identified several mechanisms related to HF with preserved ejection fraction risk among patients with nonobstructive CAD including myocardial fibrosis, ventricular stiffness, and coronary microvascular dysfunction (4,6). What is compelling about the current report is that it remains the first study establishing the prognostic risk for nonobstructive CAD among patients with HFrEF. Despite this insight, there are challenges with the current report including the limited detail as to the diffusivity of atherosclerosis, the severity of prior ischemic findings, and the presence of high-risk atherosclerotic plaque features, such as a necrotic core or positive remodeling within a nonobstructive lesion (7). All these latter findings could provide further refinement of risk, as we expect wide heterogeneity in outcomes among varied plaque subgroups with nonobstructive atherosclerosis.
Mechanisms Underlying HF With CAD
Whereas the association of obstructive and extensive CAD with HFrEF is likely mediated through loss of functional myocardial tissue, the precise prognostic link driving an elevated mortality but not MI risk in nonobstructive CAD in this HFrEF cohort is intriguing. Our common thought is that nonobstructive CAD would be associated with progressive disease states and it is known that the culprit lesion for an acute coronary syndrome is often a previously documented nonobstructive stenosis (8). The increase in risk for both all-cause and cardiovascular mortality but not incident MI tacitly supports a role for comorbidity (e.g., chronic kidney or lung disease), adverse remodeling states (e.g., diabetes and hypertension), or microvascular dysfunction, as driving risk. Many of these patients with nonobstructive CAD frequently reported anginal symptoms, which further implicates an ischemic origin. It is expected that 20% to 30% of nonobstructive lesions would demonstrate ischemia (9). Moreover, it would be anticipated that those with serial but mildly stenotic lesions, those with an extensive volume of noncalcified plaque, or those with reduced lumen volume would more often elicit vessel-specific ischemia. An intriguing possibility among this HFrEF subgroup is that nonobstructive CAD would impair flow to an already compromised myocardium and further exacerbate injury leading to an increased risk. However, it would seem only conjecture to support the role of ischemia unless incident MI risk was increased. These inequities in supply or demand could play a role in elevating risk among patients with HFrEF and nonobstructive CAD; however, one would envision an elevated risk of type II MI, but again this was not observed in this cohort.
A deeper review of the prognostic findings also reveals a unique pattern in the relative hazard for events. The hazard for the composite event was 1.17 for nonobstructive CAD, albeit significant (p = 0.01) but only a modest elevation. The stronger link for nonobstructive CAD and cardiovascular mortality, with a hazard ratio of 1.8 (p = 0.001), suggests a potential role for noncoronary atherosclerosis, which may be impacting multiple vascular beds as driving risk. Importantly, for the unique components of the endpoints, including MI and stroke, the lack of risk prediction may be explained due to the separation of fatal from nonfatal events and, for the main analysis, models used the dependent measure of nonfatal MI. Thus, fatal MIs would be censored at the time of their death and the temporal association with risk would be only for the 16 nonfatal MIs. To examine the prognostic significance of coronary and noncoronary atherosclerosis analysis, it would be informative if fatal and nonfatal endpoints were combined. Moreover, in the multivariable analysis, consideration of model overfitting would support inclusion of 2 variables, whereas for this analysis 24 variables were included in the presented model. Thus, we may be overthinking the importance of variability in prognosis for nonobstructive CAD across the varied endpoints. It may be fair to state that the current report on worsening prognosis among patients with nonobstructive CAD remains intriguing and should be viewed as exploratory.
Reconsidering Ischemic and Nonischemic Cardiomyopathy Terminology
Cardiologists have widely used the designations of ischemic and nonischemic dilated cardiomyopathy. The use of ischemic and nonischemic qualification, (particularly in the setting of nonobstructive CAD), contradicts the definition of cardiomyopathy, which is myocardial disease of unknown origin (10–12). The nonischemic adjective is unnecessary for dilated cardiomyopathy. The definition of "ischemic cardiomyopathy" incorporates the diagnostic criteria of obstructive CAD. It is generally believed that if a coronary stenosis does not critically limit the flow and, therefore, myocardial perfusion, it is insufficient to cause systolic dysfunction. Could HF-ObCAD and HF-NObCAD (Heart Failure with Obstructive and Non-Obstructive CAD, respectively) be acceptable terminology (Table 1)? Mechanistic pathways would need to be revealed to define the role of nonobstructive CAD in the evolution of HF.
Prior evidence in non-HF cohorts reveals that the burden of nonobstructive atherosclerosis accelerates disease progression, provokes ischemia, and increases the number of acute ischemic events. Longer-term prognostic follow-up, especially in this high-risk group with HFrEF, may provide further insight into variable risk patterns. Despite the limitations in this report, which exist in all clinical research (in one manner or another), the current findings should prompt a re-orientation of our thoughts with regard to the importance of detecting atherosclerosis, whether it be obstructive or nonobstructive, as precision-based care necessitates a comprehensive patient evaluation.
↵∗ Editorials published in JACC: Heart Failure reflect the views of the author and do not necessarily represent the views of JACC: Heart Failure or the American College of Cardiology.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2019 American College of Cardiology Foundation
- Schulman-Marcus J.,
- Hartaigh B.O.,
- Gransar H.,
- et al.
- Braga J.R.,
- Austin P.C.,
- Ross H.J.,
- Tu J.V.,
- Lee D.S.
- Elgendy I.Y.,
- Pepine C.J.
- Lin F.Y.,
- Zemedkun M.,
- Dunning A.,
- et al.
- Ferencik M.,
- Mayrhofer T.,
- Bittner D.O.,
- et al.
- Chang H.J.,
- Lin F.Y.,
- Lee S.E.,
- et al.
- Ahmadi A.,
- Stone G.W.,
- Leipsic J.,
- et al.
- Arbustini E.,
- Narula N.,
- Dec G.W.,
- et al.
- Maron B.J.,
- Towbin J.A.,
- Thiene G.,
- et al.