Abstract
Background: Approximately half of all patients investigated for stable exertional angina are found to have non-obstructive coronary artery disease (NOCAD), the mechanistic basis of symptoms amongst this group remains elusive. In this context, coronary microvascular dysfunction (MVD) is defined when there is a diminished coronary flow reserve (CFR) following adenosine administration, an endothelium-independent pharmacological vasodilator, and indeed MVD is associated with a greater rate of major adverse cardiovascular event (MACE). However, patients with MVD manifest symptoms during physical exercise, a physiologically distinct process to pharmacological vasodilatation. It is unknown if patients with MVD have abnormal coronary blood flow augmentation during physical exercise and whether these changes correlate with global myocardial ischaemia during stress. Additionally, a diminished CFR in MVD can arise from abnormally high resting flow, or diminished maximum flow; these distinct endotypes might be associated with unique mechanisms of disease and hence warrant disparate therapeutic targets. CFR remains the most widely used index to diagnose MVD and there remains ambiguity about what threshold to initiate therapy; this can be clarified by using sensitive tests to identify the onset of exercise physiology abnormalities or myocardial perfusion abnormalities, rather than previous thresholds defined by the probability of MACE. Finally, endothelial function testing in the catheter laboratory does not feature in many guidelines, yet may represent a direct tool for ruling out the presence of ischaemic symptoms when NOCAD is discovered. The following protocol has been designed to answer these clinically important questions.Methods: Patients with angina and NOCAD had simultaneous coronary pressure (Pd) and flow velocity (U) measured using a dual pressure and flow sensor guidewire during rest, supine bicycle exercise and pharmacological vasodilatation with adenosine and acetylcholine (direct smooth muscle and endothelium-dependent vasodilators respectively). CFR was calculated as adenosine-mediated hyperaemic flow velocity / resting flow velocity and microvascular resistance (MR) was calculated as Pd/U. Acetylcholine flow reserve (AcFR) was calculated as acetylcholine-mediated flow divided by resting flow. Wave intensity analysis quantified the proportion of accelerating wave energy (perfusion efficiency) and was a measure of cardiac-coronary coupling. Global myocardial blood flow (MBF) and subendocardial:subepicardial perfusion ratio (endo/epi) were quantified using 3-Tesla cardiac magnetic resonance imaging during hyperaemia and rest, using Fermi-constrained deconvolution. Myocardial perfusion reserve (MPR) was calculated as hyperaemic MBF / resting MBF and inducible ischaemia was defined as hyperaemic endo/epi<1.0. Forearm blood flow (FBF) was measured in response to serial acetylcholine, adenosine and NG-monomethyl-L-arginine (L-NMMA) infusions by venous occlusion plethysmography. Receiver-operating characteristic analysis was used to determine the best adenosine coronary flow reserve (CFR) and acetylcholine flow reserve (AcFR) value for identifying exercise maladaptation or global myocardial ischaemia. Controls were defined as CFR ≥ 2.5, whilst MVD was defined as CFR < 2.5, with functional and structural endotypes as hyperaemic MR < 2.5 and ≥ 2.5mmHg.cm-1s-1 respectively.
Results: 86 study participants were enrolled (78% female, 57±10 years). Using the pre-specified definition of MVD, 45 (53%) were classified as MVD and 40 (47%) as controls. 82% of the MVD group had inducible ischaemia compared to 22% of controls (p<0.001); global MPR was 2.01±0.41 and 2.68±0.49 (p<0.001). In controls, coronary perfusion efficiency improved from rest to exercise (59±11 vs. 65±14%; p=0.02) and was unchanged during hyperaemia (57±18%; p=0.14). In contrast, perfusion efficiency decreased during both forms of stress in MVD (61±12 vs 44±10 vs. 42±11%; both p<0.001). Amongst patients with a CFR<2.5, 62% had “functional MVD”, with normal minimal MR (hyperaemic MR<2.5mmHgcm-1s-1) and 38% had “structural MVD” with elevated hyperaemic MR. Resting MR was lower in those with functional MVD (4.2±1.0mmHgcm-1s-1) compared to structural MVD (6.9±1.7mmHgcm-1s-1) or controls (7.3±2.2mmHgcm-1s-1) (both p<0.001). In response to L-NMMA, the greatest reduction in FBF occurred in the functional group, followed by structural and controls (flow reserve 0.49±0.13 vs. 0.61±0.14 vs. 0.78±0.09; p<0.001 all). Higher resting wave energy was apparent in both functional and structural MVD compared to controls. Systolic blood pressure during peak exercise was higher in structural compared to functional MVD and controls (188±25 vs. 161±27 vs 156±30 mmHg; p<0.005 both vs. structural). Structural MVD had reduced FBF augmentation in response to acetylcholine than functional MVD (flow reserve 2.8±1.8 vs. 4.1±1.7; p=0.04) and controls (flow reserve 4.5±2.0; p=0.02). Functional and structural MVD had similar rates of inducible ischaemia and exercise perfusion efficiency values. The optimal dichotomous threshold for CFR was 2.6 (and thus closer to the upper end of the 2.0-2.5 grey-zone) and for AcFR was 1.5. In patients who received intra-coronary acetylcholine (n = 30), 94% (17/18) of patients with smooth muscle dysfunction also had endothelial dysfunction, 42% (5/12) of patients with normal smooth muscle function had endothelial dysfunction and (58%) 7/12 of patients with normal smooth muscle function had normal endothelial function. The rate of inducible ischaemia in these groups respectively were 73%, 60% and 14% and the change in perfusion efficiency upon exercising were -19%, -4% and +6%. A patient with NOCAD had 47% chance of having inducible ischaemia, compared to 33% if CFR was normal and 14% if both CFR and AcFR were normal.
Conclusion: In patients with angina and NOCAD, diminished CFR characterises a cohort with inducible myocardial ischaemia and a maladaptive response to exercise. We have identified two distinct endotypes of MVD, functional (decreased resting MR) and structural (increased minimal MR). Functional MVD have inefficient cardiac-coronary coupling during rest and exercise but achieve full vascular dilatation during periods of stress; nitric-oxide mediated dilatation seems to underlie this dysregulation in resting flow. Structural MVD additionally have inadequate coronary vascular dilatation in the face of greater myocardial oxygen demand, with impaired endothelial-dependent vasodilatation likely driving both these anomalies. Loss of coronary vasodilator reserve by either mechanism results in pathophysiology during stress but whether the endotypes have a different prognosis or require different treatments merits further investigation; the nitric oxide synthase pathway may form a novel therapeutic target. We therefore demonstrate that assessment of coronary flow reserve and minimal microvascular resistance in patients with NOCAD will inform clinicians about the presence of ischaemia, exercise dysfunction and degree of systemic abnormality. Additionally, a dichotomous CFR threshold of 2.5 more accurately identifies global myocardial ischaemia and abnormal exercise coronary perfusion than the widely-adopted threshold of 2.0. The presence of isolated endothelial dysfunction also identifies a group with a high prevalence of inducible ischaemia and exercise pathology, demonstrating the pivotal role of acetylcholine testing with flow measurements when diagnosing patients with NOCAD. Vasodilator testing with measurement of coronary flow should form an essential toolkit in diagnosing NOCAD in the catheter laboratory and ruling out an ischaemic cause of chest pain.
| Date of Award | 1 Nov 2019 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Divaka Perera (Supervisor), Andrew Webb (Supervisor) & Amedeo Chiribiri (Supervisor) |