Microvascular angina: its diagnosis, pathophysiology and treatment
R.O. Cannon (Bethesda)
Claims for etiologies of chest pain in patients with angiographically normal coronary arteries have been reported by cardiology, gastroenterology, and psychiatry investigators. Thus, it appears likely that this represents a heterogenous population.
The usual approach towards such a patient is reassurance of an excellent prognosis, regardless of the cause for that individual's chest pain syndrome [1, 3]. Unfortunately, many patients are « reassurance resistant», continue to have chest discomfort, and utilize health care resources [4, 5]. Published reports 20 years ago suggested that some of these patients indeed had a cardiac basis for pain [6, 7] although in more recent years emphasis has shifted to noncardiac causes, primarily gastrointestinal [8] (esophageal motility disorders, acid reflux) and psychiatric [9, 10] (psychosomatic disorders, panic attacks, anxiety neurosis). A noncardiac cause for chest pain symptoms in these patients is appealing to cardiologists, because it allows the clinician to direct the patient's attention away from the heart and towards interested specialists.
On occasion, ischemia may be a consequence of undiagnosed valvular, congenital, or cardiomyopathic (especially hypertrophic or hypertensive) heart disease. Mitral valve prolapse is another etiology that has waxed and waned in credibility over the years. Having excluded these considerations by echocardiography and catheterization hemodynamics, we and others have demonstrated in a subset of patients with anginal chest pain responses to pacing stress and to potent coronary arteriolar vasodilators such as dipyridamole, abnormal coronary vascular responses to exercise and nifedipine, and ischemic-appearing ST segment changes during exercise and daily activities [11, 21]. On the other hand, other studies have concluded that there is no evidence for myocardial ischemia in these patients [22, 24]. These studies, however, differed in patient inclusion criteria and study design, as well as in the interpretation of the results. Even the name « Syndrome X » has taken on different meanings to investigators, generally resulting in an unsatisfactory diagnosis for the clinician to make, and for the patient to receive.
We initially separated patients with chest pain despite angiographically normal coronary arteries based on whether or not the stress of rapid atrial pacing provoked the pain described to us upon entry into our study [12, 15, 17]. With pacing to a heart rate of 150, those who experienced their typical chest pain had less of an increase in coronary flow and less of a decrease in coronary vascular resistance compared to those patients experiencing no pain or a non characteristic pain. However, a feature we believe of major importance to our studies, and which distinguishes our work from others, has been the re-assessment of coronary blood
flow and the metabolic response to pacing at the same heart rate following ergonovine administration. After ergonovine 0.15 to 0.3 mg was administered intravenously, repeat atrial pacing was associated with even less of an increase in coronary flow in the patients experiencing chest pain (three-quarters of all patients studied by us) compared to pacing prior to ergonovine.
Further, the absolute differences in flow and coronary resistance, and the metabolic and hemodynamic responses to pacing were more convincing than pacing prior to ergonovine.
Of interest was the observation that coronary resistance during pacing actually increased after ergonovine administration, indicating coronary vasoconstriction in the chest pain group. To our initial surprise, repeat coronary angiography showed only minimal changes in epicardial artery dimensions, too insignificant to explain the resistance changes indicated by our measurements. Thus, we concluded that vasoconstriction must have occurred in vessels too small to be reliably imaged angiographically [25]. We subsequently assessed pharmacologic flow reserve with dipyridamole and found that those patients with ergonovine-provoked chest pain and microvascular constriction also had limited coronary vasodilatation in response to this potent arteriolar vasodilator [17]. We have suggested that « microvascular angina » might be an appropriate name for this syndrome [26].
We subsequently found that patients with microvascular angina, even in the absence of hypertension (present in one-third of our patients), have blunted hyperemic forearm blood flow responses to ischemia compared to normal age and gender-matched controls [27]. Further, the severity of flow limitation in the forearm to ischemic stress correlated directly with the severity of coronary flow limitation to pharmacologic vasodilatation after dipyridamole. Thus, limitation in vasodilator responses appears to co-exist in the coronary and systemic circulations, suggesting abnormal function of vascular smooth muscle. The observation that many patients with microvascular angina also have evidence of esophageal motility dysfunction [28], most commonly «nutcracker esophagus», and methacholine provoked increases in airflow resistance [29], a response similar to asthmatics, has led us to consider a hypothesis of a generalized abnormality of both vascular and non-vascular smooth muscle in some patients. The cause of smooth muscle dysfunction is not known at present, but may be heightened sensitivity to vasoconstrictor stimuli, either neural or humoral, abnormal endothelial-smooth muscle interactions or increased intrinsic tension development in smooth muscle, with inappropriate relaxation in response to vasodilating stimuli.
Approximately one-third of our patient population with microvascular angina have convincingly ischemic ECG changes during treadmill exercise. Thus, although abnormal ECG responses to treadmill exercise might identify patients with micro-vascular angina, the absence of ischemic ST segment changes does not eliminate this diagnosis. We have found instead that a more sensitive and specific noninvasive test for microvascular angina is the left ventricular ejection fraction response to exercise measured by radionuclide angiography. Two-thirds of the patients we studied had either limited (less than 5 %) increase in ejection fraction with exercise, or an actual decrease in ejection fraction during exercise, often associated with wall
motion abnormalities [30]. The absence of convincingly ischemic ECG changes during exercise treadmill testing may relate to the generalized microvascular dysfunction and relative mildness of ischemia in this syndrome. This is in contrast to coronary artery disease, in which flow limitation is more focal and severe. Although the coronary flow reserve studies performed in our catheterization laboratory are impractical for most clinicians, noninvasive techniques are currently being investigated for assessment of coronary flow and may prove useful in the future (e.g., assessment of coronary flow reserve by positron emission tomography before and after dipyridamole).
A recent observation made by Shapiro et al. [31], and confirmed and expanded in our laboratory [32], is of interest, and potentially of clinical importance. Patients with chest pain and normal coronary arteries have a painful sensitivity to catheter manipulation within the heart, a response rarely seen in patients with coronary artery disease or valvular heart disease, and clearly occurring on a nonischemic basis. Further, we have noted that inflation of a balloon within the esophagus commonly provoked chest pain in patients undergoing esophageal manometry, irrespective of whether or not they had any evidence of esophageal motility dysfunction or acid reflux [28].
These observations suggest that a fundamental problem in patients with chest pain and normal coronary arteries may be abnormal afferent sensory receptor activation and nociception, with a painful perception of visceral sensations that would otherwise go unnoted in most individuals. Certainly, some patients within this broad group do have organic dysfunction of the esophagus, or microvascular angina, or both. However, abnormal sensory nerve function and pain perception may explain why high esophageal pressures alone (e.g., nutcracker esophagus), or mild myocardial ischemia (microvascular angina) may cause frequent, severe, prolonged and disabling pain. Either of these conditions in an individual with normal visceral sensory nerve function would probably go unnoted. Further, the anxiety commonly seen in these patients may be a pathogenetic component of this syndrome.
The patient determined in the catheterization laboratory to have angiographically normal coronary arteries, with no evidence of coronary spasm following ergonovine challenge, deserves reassurance, and repeat hospitalizations for chest pain should be discouraged. We do not believe at present that a diagnosis of microvascular angina should be made without demonstration of limited coronary flow reserve response to pacing or pharmacologic stimulation. However, if noninvasive testing suggests myocardial ischemia, associated with the patient's typical chest pain during exercise stress, a cardiac etiology would be supported and an empiric trial of anti-ischemic medications should be considered. We have found most patients to respond to nitrates and calcium channel blockers [33], although a significant minority do not, possibly explained by the recent observation of the inability of nifedipine to dilate coronary arteries of some patients [20].
Also under investigation at our institution is the role of drugs that affect neural pain pathways and processing. Other centers have found beta blockers to be of benefit [34], possibly by reducing myocardial oxygen demands. Recently, Emdin and co-workers have demonstrated improvement in effort duration, pressure-rate
product achieved, and symptom response during bicycle exercise with the abolition of ischemic-appearing ST segment responses, following aminophylline infusion [35]. Whether oral aminophylline preparations will provide the same benefit, while avoiding side effects and toxicity, remains to be determined.
If this treatment trial is ineffective, or if noninvasive testing is normal, the patient should be referred for esophageal testing, including motility testing and 24-hour pH monitoring. An esophageal etiology for pain would be supported by abnormal motility or acid reflux associated with the patient's typical symptoms. If anxiety, depression, or panic attacks appear to be a component of chest pain episodes, consideration should be made for the use of anxiolytic or antidepressant therapy.
References
34. Romeo F et al. (1988) Verapamil versus acebutolol for Syndrome X. Am J Cardiol 62: 312.