Abstract
Background: Up to half of all patients presenting to the catheter laboratory with angina have nonobstructive coronary arteries (ANOCA). ANOCA is an umbrella term comprising several distinct pathophysiological entities, including coronary microvascular disease (CMD) and myocardial bridging (MB). CMD is defined as an inability of the coronary vasculature to adequately augment coronary blood flow (CBF) in response to adenosine, i.e., an impaired coronary flow reserve (CFR). An impaired CFR identifies a substrate for ischaemia and is a sensitive marker that identifies perturbations early in the ischaemic cascade. We have previously demonstrated that patients with an impaired CFR have abnormal exercise physiology and high prevalence of inducible ischaemia on noninvasive imaging. We have also reported that CMD may itself be a heterogenous condition comprising two distinct endotypes, structural and functional CMD, which are phenotypically similar but have distinct underlying pathobiology. However, whether physiology-stratification according to CFR and more granular endotyping leads to improved patient-centric outcomes, such as exercise time on a treadmill and quality of life on angina questionnaires, is not known. Finally, there is now increasing recognition that certain myocardial bridges can cause myocardial ischaemia. However, the mechanisms underpinning this are not well understood. Coronary wave intensity analysis (WIA) has previously been used to identify the mechanisms of ischaemia in different pathophysiological states and represents a powerful tool to study the mechanisms of ischaemia in patients with MB.Methods: Patients with typical and limiting angina underwent coronary angiography with invasive physiology assessment using a Combowire to measure coronary flow and pressure simultaneously in response to adenosine and acetylcholine (clinical protocol) but also supine bicycle exercise and dobutamine (in-lab study protocol). CFR (measure of endothelium-independent microvascular function) and acetylcholine flow reserve (AChFR; measure of endothelium-dependent microvascular function) were calculated as the ratio of CBF in response to the vasoactive agent and the resting CBF. Patients and researchers were blinded to the coronary physiology measurements. All eligible patients were enrolled into the exercise electrocardiogram treadmill test (ETT) study. The accuracy of ischaemic ECG changes during ETT in identifying an underlying ischaemic substrate was compared against the reference standard of coronary endothelium-independent and -dependent microvascular function. The apparent false positive rates with different reference standards were also compared.
Eligible patients were then randomised to a phenotype-blinded crossover therapy trial, which was designed to assess the utility of coronary physiology measurements in predicting response to anti-ischaemic therapy (amlodipine and ranolazine) in patients with ANOCA. The primary outcome was the difference in change in exercise time in response to anti-ischaemic therapy between those with an impaired CFR (coronary microvascular disease group; CMD) and those with a normal CFR (reference). The incremental value of measuring minimal microvascular resistance and acetylcholine flow reserve, in predicting response to anti-ischaemic therapy, was also assessed.
Finally, for our mechanistic in-lab study, coronary perfusion efficiency in response to supine bicycle exercise was calculated as the ratio of accelerating wave energies and the total energy flux. The change in perfusion efficiency during exercise was compared between patients with MB and no MB (the latter being further dichotomised into CMD and reference groups). We also explored the prevalence of endothelial dysfunction in the MB, CMD and reference groups.
Results: One hundred and two patients (65% females; 60±8 years old) were enrolled into the ETT study. Thirty-two patients developed ischaemic ECG changes during their ETT (ischaemic group), whilst 70 patients did not (non-ischaemic group); both groups were phenotypically similar. Ischaemic ECG changes during ETT were 100% specific for underlying endothelium-independent and/or -dependent microvascular dysfunction. AChFR was the strongest predictor of ischaemic ECG changes during exercise. Using endothelium-independent and/or -dependent microvascular dysfunction as the reference standard, the false positive rate of ETTs dropped to 0%.
Eighty-seven patients (62% females, 61±8 years old) underwent randomisation as part of the phenotype-blinded crossover therapy trial (57 impaired CFR (CMD group) and 30 normal CFR (reference group)). Baseline exercise time and Seattle Angina Questionnaire (SAQ) summary scores were similar between the groups. Patients with CMD had a greater increment in exercise time compared to the reference group with both amlodipine (mean difference in change 82 seconds, 95% CI 37 to 126 seconds, p<0.001) and ranolazine (mean difference in change 68 seconds, 95% CI 21 to 115 seconds, p=0.005). The change in SAQ summary score in response to amlodipine was similar between the CMD and reference groups (mean difference in change 2, 95% CI -5 to 8, p=0.549). There was a greater increment in SAQ summary score with ranolazine in the CMD group compared to the reference group (mean difference in change 7, 95% CI 0 to 15, p=0.048). CFR was independently associated with change in exercise time and CFR≤2.5 was the optimal threshold to predict response to therapy. Patients with functional CMD responded equally well to both anti-ischaemic agents, whereas those with structural CMD had a numerically greater response to amlodipine than ranolazine (mean difference in change 46s, 95% CI -2 to 93s, p=0.056). Patients with sole coronary endothelial dysfunction demonstrated a numerical increment in exercise time in response to anti-ischaemic therapy, whereas no such effect was seen in the reference group. These findings support the incremental value of measuring minimal microvascular resistance and AChFR, in addition to CFR, in patients with typical limiting ANOCA.
Ninety-two patients were enrolled into the in-lab mechanistic study (30 MB, 33 CMD and 29 reference). FFR in these 3 groups was 0.86±0.05, 0.92±0.04 and 0.94±0.05; CFR was 2.5±0.5, 2.0±0.3 and 3.2±0.6. Perfusion efficiency improved numerically during exercise in the reference group (65±9% to 69±13%, p=0.063), but decreased in patients with CMD (68±10% to 50±10%, p<0.001) and MB (66±9% to 55±9%, p<0.001). The reduction in perfusion efficiency had distinct causes: in CMD, this was driven predominantly by microcirculation derived energy in early diastole, whereas in MB, this was driven by diminished accelerating energy arising from the upstream epicardial vessel in early systole. 54% of patients with MB, versus 29% reference and 38% CMD, had epicardial endothelial dysfunction. Overall, 93% of patients with a MB had an identifiable ischemic substrate.
Conclusions: Our findings have several important implications for both future research and clinical practice. First, in patients with ANOCA, ischaemic ECG changes on an ETT were always attributable to an underlying ischaemic substrate secondary to abnormalities in the coronary microcirculation. Therefore, a positive ETT (defined as ischaemic ECG changes during exercise) may be an excellent tool to rule-in CMD in patients with typical and limiting angina who have nonobstructive coronary arteries. Second, amongst a phenotypically similar group of patients with ANOCA, only those with an impaired CFR responded to anti-ischaemic therapy. Our data also suggests that measuring minimal microvascular resistance and acetylcholine flow reserve, in addition to CFR, may add incremental value in predicting response to therapy. Third, patients with MB and CMD demonstrated impaired coronary perfusion efficiency during exercise, whereas those with a normal CFR had a numerical increase. The mechanisms driving the attenuated perfusion efficiency during exercise were disparate between patients with MB and CMD, with diminution of accelerating wave energies arising from the epicardial artery during early systole being the predominant mechanism in patients with MB and perturbation of the microcirculation derived wave energies being the predominant mechanism in patients with CMD. Patients with MB also had a high prevalence of epicardial and microvascular endothelial dysfunction. Both mechanisms may lead to ischaemia in patients with MB and represent therapeutic targets.
| Date of Award | 1 Jun 2024 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Divaka Perera (Supervisor) & Andrew Webb (Supervisor) |