Physiological Stratification of Patients With Angina Due to Coronary Microvascular Dysfunction

Background Coronary microvascular dysfunction (CMD) is defined by diminished flow reserve. Functional and structural CMD endotypes have recently been described, with normal and elevated minimal microvascular resistance, respectively. Objectives This study determined the mechanism of altered resting and maximal flow in CMD endotypes. Methods A total of 86 patients with angina but no obstructive coronary disease underwent coronary pressure and flow measurement during rest, exercise, and adenosine-mediated hyperemia and were classified as the reference group or as patients with CMD by a coronary flow reserve threshold of 2.5; functional or structural endotypes were distinguished by a hyperemic microvascular resistance threshold of 2.5 mm Hg/cm/s. Endothelial function was assessed by forearm blood flow (FBF) response to acetylcholine, and nitric oxide synthase (NOS) activity was defined as the inverse of FBF reserve to NG-monomethyl-L-arginine. Results Of the 86 patients, 46 had CMD (28 functional, 18 structural), and 40 patients formed the reference group. Resting coronary blood flow (CBF) (24.6 ± 2.0 cm/s vs. 16.6 ± 3.9 cm/s vs. 15.1 ± 4.7 cm/s; p < 0.001) and NOS activity (2.27 ± 0.96 vs. 1.77 ± 0.59 vs. 1.30 ± 0.16; p < 0.001) were higher in the functional group compared with the structural CMD and reference groups, respectively. The structural group had lower acetylcholine FBF augmentation than the functional or reference group (2.1 ± 1.8 vs. 4.1 ± 1.7 vs. 4.5 ± 2.0; p < 0.001). On exercise, oxygen demand was highest (rate−pressure product: 22,157 ± 5,497 beats/min/mm Hg vs. 19,519 ± 4,653 beats/min/mm Hg vs. 17,530 ± 4,678 beats/min/mm Hg; p = 0.004), but peak CBF was lowest in patients with structural CMD compared with the functional and reference groups. Conclusions Functional CMD is characterized by elevated resting flow that is linked to enhanced NOS activity. Patients with structural CMD have endothelial dysfunction, which leads to diminished peak CBF augmentation and increased demand during exercise. The value of pathophysiologically stratified therapy warrants investigation.

O ne-half of all patients with angina referred for angiography have nonobstructive coronary artery disease (NOCAD); those with coronary microvascular dysfunction (CMD) exhibit a poorer prognosis (1,2). CMD is diagnosed when there is diminished flow augmentation in response to a pharmacological vasodilator or reduced coronary flow reserve (CFR). A CFR <2.5 is associated with abnormal coronary flow with exercise and inducible ischemia on perfusion cardiac magnetic resonance imaging. We recently described 2 endotypes of CMD with distinct processes that contributed to low CFR: patients with functional CMD who have increased baseline flow (due to reduced microvascular resistance at rest); and patients with structural CMD who have reduced hyperemic flow (due to high minimal microvascular resistance) (3). Resting coronary blood flow (CBF) is regulated by nitric oxide synthase (NOS) and its role in the coronary circulation has been extrapolated from effects of vasoactive medication within the forearm circulation (4,5). In healthy patients, physical exercise involves synergistic adaptations between the peripheral and coronary vasculature to match CBF supply with myocardial oxygen demand, a process that is disrupted in structural CMD (3,6). The endothelium is responsible for translating mechanical forces (or shear stress) to smooth muscle dilatation in the coronary circulation and may have a similar role systemically during exercise (6). We sought to unravel the pathobiology of the 2 endotypes and hypothesized that: 1) elevated resting blood flow in functional CMD is a response to increased demand and is mediated by increased activity of the NOS pathway; and 2) structural CMD is reflective of a generalized disorder of vascular dilatation that affects peak flow in response to stress.

METHODS
Patients who underwent elective diagnostic angiography for investigation of exertional chest pain were enrolled in the study. Inclusion criteria were preserved left ventricular systolic function (ejection fraction >50%) and unobstructed coronary arteries  (3,7). Microvascular resistance was calculated as the ratio of the distal mean coronary pressure and the average peak velocity. The supply/demand ratio was estimated as the average peak velocity/ rateÀpressure product (a measure of external stroke work) and denoted as supply/demand EXT . By wave intensity analysis, 4 dominant waves were identified and included in our analysis: 1) the backward compression wave, which causes flow deceleration during isovolumetric contraction in early systole; 2) the forward compression wave, which causes flow acceleration that is associated with peak aortic      Physiological Stratification of Microvascular Dysfunction     (Figure 4).
The response to vasoactive agents was also significantly different across the groups. The increase in FBF in response to acetylcholine was diminished in the structural group compared with the functional

DISCUSSION
This study showed that functional and structural CMD endotypes have inefficient cardiac-coronary coupling during exercise compared with reference group patients with preserved CFR. This was hyperemia ¼ adenosine-induced hyperemia; MR ¼ microvascular resistance; peak ¼ immediately before exercise was discontinued due to exhaustion; Rahman et al.  this may serve as an initially protective mechanism to maintain metabolic vasodilatory reserve (20,21).
Although external myocardial work in the form of rateÀpressure product is not elevated in functional CMD, higher basal metabolic requirements can underlie elevated resting CBF (22). Higher resting wave energy, but no net increment in perfusion efficiency implies an inefficient resting state more pronounced in functional than structural CMD. An inefficient metabolic state may represent the "common soil" of CMD and heart failure with preserved ejection fraction (23). In patients with diabetes, depressed myocardial energetics precedes any overt changes in microvascular function; similarly, diastolic dysfunction occurs late within the disease natural history (24,25). In heart failure with preserved ejection frac-   In one endotype, CBF is impaired at rest due to nitric oxide dysregulation, whereas in the other, endothelial dysfunction limits peak hyperemic blood flow.
TRANSLATIONAL OUTLOOK: Therapeutic studies should stratify patients with microvascular dysfunction based on impairment of resting or hyperemic flow.
Rahman et al.