Persistent Chest Pain and No Obstructive Coronary Artery Disease

Patients with persistent chest pain and no obstructive coronary artery disease (defined as ≥50% stenosis in ≥1 major coronary artery) are often regarded as having noncardiac chest pain and most often given diagnoses such as gastrointestinal or psychiatric disorders. Approximately 14% to 30% of patients undergoing angiography have no coronary artery disease or other cardiac diagnosis such as coronary spasm, bridging, cardiomyopathy, or hypertrophy, and a majority of these patients are women.^{1 - 3}

Identification of angina caused by microvascular coronary dysfunction is important because of the associated adverse prognosis, including an increased risk of adverse cardiovascular events such as myocardial infarction, congestive heart failure, and sudden cardiac death.^{4 - 5} Abnormal cardiac nociception is another cause for persistent chest pain and contributes, as does microvascular coronary dysfunction, to adverse quality of life, morbidity, and health care costs. An adequate evaluation in patients with persistent chest pain is important for accurate diagnosis and treatment.

In this article, we highlight 2 cases of women with persistent chest pain symptoms and previously inaccurate diagnoses. Both cases illustrate the utility of coronary reactivity testing as a diagnostic tool in differentiating cardiac etiologies for persistent chest pain.

A 45-year-old woman was referred for a second opinion for persistent chest pain. Her medical history was significant for depression, anxiety, and gastroesophageal reflux disease, for which she had been receiving treatment since 2005. She had no cardiac risk factors, including no history of diabetes, hypertension, hyperlipidemia, or smoking, and had no family history of premature coronary artery disease.

She had experienced chest pain her “whole life” and for the past 5 years experienced exertional and stress-related angina. Her chest pain episodes were intermittent, substernal, and left sided, occasionally radiating to the left side of her back, and graded 7 of 10 in severity. Associated symptoms included shortness of breath, lightheadedness, and nausea. Nitroglycerin relieved her chest pains in the past, but she stopped using it because of the adverse effect of headaches. She had frequent emergency department visits to numerous hospitals for chest pain, averaging about once every 3 months. A review of her records at our institution from 2003 to 2007 revealed emergency department diagnoses for her chest pain that included typical and atypical chest pain, unspecified chest pain, gastroesophageal reflux disease, depression, and anxiety disorder.

She had been followed up regularly and treated by a psychiatrist for her anxiety and depression. She was also receiving lansoprazole and ranitidine for her gastroesophageal reflux disease. She self-medicated with low-dose aspirin because of her concern about her symptoms being a myocardial infarction. Despite these treatments, she continued to experience persistent chest pain throughout this period, for which she received 2 coronary angiograms.

Testing conducted in the evaluation of this patient's chest pain included multiple chest radiographs, the results of which were normal, and electrocardiograms that showed normal sinus rhythm with no Q waves or ST- or T-wave abnormalities. Outpatient radionuclide perfusion stress testing, echocardiogram, and ventilation/perfusion scan results were all within normal limits. A cardiac catheterization demonstrated normal coronary arteries, no significant valvular regurgitation, and normal left-sided ventricular function.

Because of persistent symptoms, a second angiogram was performed at an outside hospital, which again found normal coronary arteries. Another echocardiogram was conducted, showing a left ventricular ejection fraction of 67%, normal left ventricular wall motion, no evidence of diastolic dysfunction, and trace mitral, tricuspid, and pulmonic valve regurgitation.

Resting blood pressure was 111/56 mm Hg; pulse rate was 62/min; and cardiac examination revealed no murmur, rubs, gallop, or jugulovenous distension. The remaining examination results were normal.

The patient underwent flow wire testing in the left coronary artery, using previously published methods,⁶ to assess for microvascular coronary dysfunction with intracoronary infusions of adenosine, acetylcholine, and nitroglycerin. She had normal coronary arteries without luminal irregularity; left ventricular end-diastolic filling pressure was normal, at 10 mm Hg. Coronary flow reserve in response to adenosine was normal, at +3.3 (normal, ≥2.5); however, coronary blood flow response to acetylcholine was abnormal, at −11% (normal, ≥50%), endothelial function to acetylcholine was abnormal, at −13% (normal, >0%), and nitroglycerin response was abnormal, at +3% (normal, >20%).

Given the previous normal cardiac evaluation result and the abnormal coronary reactivity testing findings, the patient was diagnosed with microvascular coronary dysfunction as etiologic for her persistent chest pain. She continued receiving her low-dose aspirin and began receiving a statin and carvedilol. She is symptomatically improved by self-report, has not had an emergency department visit in more than 18 months, and is taking less psychiatric medication for her depression and anxiety.

A 43-year-old woman was self-referred for 2 years of persistent chest pain precipitated by exercise, with radiation to her right arm and jaw. These episodes resolved with both rest and nitroglycerin. During the past year, her chest pain episodes had increased in frequency and intensity, and she was experiencing chest pain while walking on level ground; however, she reported no symptoms at rest.

Her family history was positive for premature cardiovascular disease, with her father having a series of strokes starting in his 30s, but negative for the other risk factors. Her medical history included gastroesophageal reflux disease, asthma, and endometriosis.

Previous records reviewed included an exercise stress study with single-photon emission computed tomography myocardial perfusion, which demonstrated an exercise capacity of 7.4 metabolic equivalents on Bruce protocol, with a maximum heart rate of 146/min (82% of maximum predicted heart rate), exercise-induced chest pain, borderline ST-segment depression, and a peak left ventricular ejection fraction of 83%, with no stress-induced reperfusion abnormalities. A computed tomography coronary angiogram was performed, showing normal coronary arteries and an aberrant right subclavian artery.

Resting blood pressure was 120/76 mm Hg, without a difference between arms; pulse rate was 69/min. Her cardiac examination revealed no murmurs, clicks, gallop, or jugulovenous distension, and the remaining examination results were normal.

The patient had normal coronary arteries, without luminal irregularity, but an increased left ventricular end-diastolic pressure of 17 mm Hg. There were normal responses to adenosine, acetylcholine, and nitroglycerin; however, the patient experienced severe chest pain during flow wire manipulation, as well as during the intracoronary infusions. Her chest pain during the procedure was similar to the persistent chest pain episodes she had experienced in the last 2 years. She required 2 doses of 25 μg fentanyl to ameliorate her procedure-related chest pain.

Given the previous negative cardiac evaluation results and chest pain with flow wire manipulation/intracoronary infusions without vasospastic activity consistent with a heightened cardiac pain receptor status, the patient was diagnosed with abnormal cardiac nociception. Quiz Ref IDSubclavian steal syndrome was considered because of the aberrant right subclavian artery; however, this was unlikely because she did not have other associated symptoms such as pain with swallowing, known as dysphagia lusoria, caused by compression of the esophagus from the subclavian artery, or concurrent cerebrovascular symptoms such as vertigo, caused by right-sided common carotid blood flow compromise. The increased left ventricular end-diastolic pressure was of some concern but, in the absence of other evidence of cardiomyopathy, was considered an unlikely etiology for her chest pain. She was treated symptomatically with reassurance and low-dose imipramine. Transcutaneous electrical nerve stimulation was also suggested in the event that her symptoms were refractory. She has not sought additional testing at 18-month follow-up.

Etiologies for persistent chest pain encompass a wide array of possibilities, including both cardiac and noncardiac causes. Oftentimes after initial cardiac evaluation results are negative and coronary angiography reveals normal coronary arteries, patients are dismissed with the diagnosis of noncardiac chest pain, without further cardiac evaluation. Both presented cases are women with normal coronary arteries who had been diagnosed and treated for noncardiac conditions without relief, but were subsequently found to have cardiac causes for their persistent chest pain. These cases emphasize that further cardiac testing can be warranted in light of diagnostic uncertainty of persistent chest pain.

Persistent chest pain with diagnostic uncertainty leads to adverse effects on quality of life, morbidity, and health care costs. Eighty-six percent of patients with persistent chest pain and no coronary artery disease were found to have at least weekly chest pain episodes up to 1 year after angiogram,^{7 - 8} with the intensity of the pain remaining unchanged or worsened.⁷ Persistent chest pain episodes have been found to continue for as long as 4 years in 80% of patients with no obstructive coronary artery disease.⁹ Despite angiograms showing normal or nearly normal coronary arteries, 20% to 50% of patients are rehospitalized for chest pain, with the average lifetime cost for these patients being close to $800 000.¹⁰ Almost half of patients with persistent chest pain and no obstructive coronary artery disease report functional disability,^{7 - 8 ,10} including limitations in activities of daily living¹¹ and inability to work.⁸

Quiz Ref IDCommon noncardiac causes for persistent chest pain include musculoskeletal origin for pain, such as costochondritis and arthritis, and pulmonary causes, including pulmonary embolism and gastrointestinal and psychogenic disorders. Gastrointestinal disorders causing chest pain include gastroesophageal reflux disease, esophageal motility disorder such as nutcracker esophagus, achalasia, and diffuse esophageal spasms. Although esophageal motility and reflux disease are found in patients presenting with chest pain and normal angiogram results,^{12 - 14} they are not necessarily the causative factor for their chest pain.^{13 - 14} Cooke et al¹³ found a poor temporal correlation between gastroesophageal reflux disease and exertional chest pain.

Patients with persistent chest pain and normal coronary angiogram results have significantly higher psychological scores on indices of anxiety and depression compared with patients with chest pain and coronary artery disease,¹⁵ with anxiety disorder being the most common.¹⁶ Again, although up to 40% of patients with persistent chest pain and normal coronary arteries were associated with having panic disorder¹⁷ and major depression, it is unclear whether these are causally related. Patients with persistent chest pain and normal coronary arteries who have psychiatric disorders are no more likely to have exaggerated sensitivity to cardiac pain in response to right-sided ventricular pacing and intracoronary adenosine infusion than those without psychiatric disorders.¹²

In our first case, the patient had concurrent diagnoses of gastroesophageal reflux disease, depression, and anxiety disorder, all of which were appropriately treated, yet she continued to experience chest pain, reinforcing the suggestion that her persistent chest pain required further investigation.

Alternative coronary causes of chest pain include coronary artery vasospasm and coronary bridging. Coronary artery spasms involve abnormal contraction of the large coronary arteries and include Prinzmetal angina, which can present as chest pain symptoms at rest and ischemic changes on electrocardiography and can be induced on angiography with vasoactive stressors. Although the diagnosis can be difficult, contemporary series suggest that Prinzmetal angina is relatively uncommon.¹⁸ Quiz Ref IDMyocardial bridging is a congenital anomaly involving a segment of a large coronary artery tunneled in the myocardium and resulting in compression of the artery during systole.¹⁹ It is detected by routine coronary angiography.²⁰

Almost 50% of women with persistent chest pain and evidence of ischemia in the absence of obstructive coronary artery disease have been found to have microvascular coronary dysfunction.¹¹ These patients tend to be women and younger, with average age in the fifth decade of life.^{1 - 3 ,5} The reason for the sex difference may be related to differences in sex hormones, inflammatory responses, genetic factors, or referral bias related to complaints of chest pain. The coronary microvasculature is not directly evaluated with routine coronary angiography because these vessels are not readily visualized. These vessels are responsible for regulation of coronary blood flow and therefore the delivery of oxygen to myocytes and have both endothelial-dependent and -independent autoregulatory mechanisms. Case 1 illustrates an example of persistent chest pain caused by microvascular coronary dysfunction.

Quiz Ref IDThe Women's Ischemia Syndrome Evaluation study demonstrated that microvascular coronary dysfunction is prevalent in women with the triad of persistent chest pain, evidence of ischemia by stress testing, and no coronary artery disease.²¹ Coronary endothelial dysfunction is an indicator of early coronary atherosclerosis¹⁶ and is independently associated with cardiovascular events.^{4 - 5} Patients with no obstructive coronary artery disease and coronary endothelial dysfunction have an increased risk of adverse cardiovascular events, including sudden cardiac death, myocardial infarction, congestive heart failure, and need for revascularization,^{4 - 6} and have an average 2.5% annual risk per year.²² Although patient 1 had no detectable ischemia on stress testing, it is likely that this reflects the insensitivity of the test, rather than an absence of ischemia.

Quiz Ref IDDuring coronary reactivity testing, a Doppler flow wire inside the coronary artery measures the velocity of blood flow in response to different vasoactive agents such as intracoronary adenosine, acetylcholine, and nitroglycerin.^{5 ,16} Interventional cardiologists capable of performing fractional flow reserve needed for the appropriate use of percutaneous coronary interventions can be trained to perform the related Doppler wire and intracoronary medication injections in coronary reactivity testing.

Four measures of coronary vascular autoregulation are assessed, including coronary flow reserve in response to intracoronary artery adenosine, indicative of nonendothelial microvascular function¹⁷ ; coronary blood flow, indicative of endothelial microvascular function; coronary artery diameter in response to intracoronary acetylcholine, indicative of endothelial macrovascular function; and coronary artery diameter in response to intracoronary nitroglycerin, indicative of nonendothelial macrovascular function. Coronary flow reserve and coronary blood flow are derived by equations obtained from the direct measurements of coronary blood velocity measured by intracoronary Doppler, whereas macrovascular coronary artery diameter in response to agents infused is measured with quantitative coronary angiography. Although coronary flow is primarily controlled by the microvasculature, assessed by coronary flow reserve and coronary blood flow, macrovascular function in the major coronary vessels can also contribute.

The utility of any diagnostic testing involves careful weighing of risks and benefits. Routine coronary angiography carries an approximate 1 in 1000 risk of serious adverse effects such as bleeding, dissection, or embolism. There is a relative paucity of safety data about contemporary coronary reactivity testing. One prospective study evaluated 906 subjects, including obstructive coronary artery disease and cardiac transplant patients undergoing intracoronary Doppler flow measurements with adenosine or papaverine infusions.²³ The authors found a relatively high rate of adverse events in the cardiac transplant patients, with an overall serious adverse event rate in the nontransplant patients (including those with obstructive coronary artery disease) of 2.4%. Although 33% of the patient population in this study had normal coronary angiogram results, the authors did not specifically report the rate of adverse events in this group. Papaverine was not as safe as acetylcholine, and bradycardia with adenosine was found to occur primarily when used in the right coronary artery. A second study reiterated the relative safety of intracoronary acetylcholine infusion in 299 patients undergoing first diagnostic coronary angiogram²⁴ and reported a serious adverse event rate of 0.7%, including that in obstructive coronary artery disease patients.²⁴

As many as 58% of patients with persistent chest pain and no obstructive coronary artery disease have abnormal endothelial coronary function,²⁵ and a synthesis of 15 published reports of endothelial dysfunction demonstrated an increased relative risk ratio of nearly 10-fold (95% confidence interval, 7.8-12.8).²⁶ Further study demonstrated that treatment and restoration of endothelial function are associated with improved outcomes.²⁷ Accordingly, because the significantly increased risk of major adverse events, including death, associated with endothelial dysfunction^{4 - 6 ,26} and microvascular coronary dysfunction is relatively high (2.5% annually)²² compared with the nonfatal adverse event rate (0.7%-2.4%)^{23 - 24} associated with coronary reactivity testing, patients with persistent chest pain, objective evidence of myocardial ischemia, and normal coronary angiogram results should be considered for this testing when diagnostic uncertainty is present.

Given the observed adverse outcomes related to microvascular coronary dysfunction, patients with this diagnosis should be considered for treatment aimed at risk reduction, as well as symptom amelioration. Few studies have specifically addressed microvascular coronary dysfunction treatment. Cardiac syndrome X is defined as patients with angina or angina-like symptoms with exercise, ischemic changes on treadmill stress testing, and normal or nonobstructive coronary arteries on angiography. Data from the Women's Ischemia Syndrome Evaluation and others suggest that at least half of patients fitting the cardiac syndrome X description have microvascular coronary dysfunction.^{17 ,26 ,28}

The current American Heart Association and American College of Cardiology guidelines for management of cardiac syndrome X include medical therapy with nitrates, β-blockers, and calcium antagonists.²⁹ Statins have also been shown to improve endothelial dysfunction in patients with chest pain and normal angiogram results,^{30 - 32} with an increase in coronary flow reserve independent of the magnitude of lipid-level decreases.³³ Angiotensin-converting enzyme inhibitors can also improve coronary flow reserve and exercise duration in cardiac syndrome X patients.³⁴ Kaski et al³⁵ demonstrated that angiotensin-converting enzyme inhibitors decrease exercise-induced ischemia in cardiac syndrome X patients by demonstrating an increase in exercise duration and decrease in the magnitude of ST depression on electrocardiogram. β-Blockers have been effective in decreasing chest pain³⁶ and ischemic episodes³⁷ in patients with cardiac syndrome X.

Calcium channel blockers and imipramine have been shown to improve chest pain symptoms, although calcium channel blockers are less effective than β-blockers in head-to-head comparisons in this population^{36 - 37} Cannon et al³⁸ showed that calcium channel blockers improve chest pain, as well as exercise capacity, in patients with normal angiogram results and limited vasodilatory reserve. Imipramine treatment decreases the frequency of chest pain in patients with chest pain and normal angiogram results by about 50%.¹² Therapies such as aminophylline, doxazosin, and ranolazine may also be efficacious.²⁸ Additional studies with these agents, as well as other novel therapies, are needed.

Heightened cardiac pain perception (nociception) has primarily been studied in patients with cardiac syndrome X.^{12 ,39 - 42} Studies demonstrating abnormal cardiac nociception also included patients with normal coronary angiogram results, without ischemic changes on their stress test,^{12 ,39 ,42} suggesting that this is not a phenomenon particular to patients with chest pain caused by microvascular coronary dysfunction. Case 2 illustrates an example of persistent chest pain most likely caused by abnormal pain perception.

The overall prevalence and the exact mechanism(s) of heightened cardiac pain sensitivity are not fully understood. These patients may have altered intracardiac pain perception³⁹ or increased sensitivity to cardiac pain stimuli.⁴³ The origin of increased pain perception can be due to abnormalities at any point, ranging from cardiac mechano- and chemoreceptors, cardiac afferents, and spinal or cortical connections to central nervous system processing centers for pain.^{39 ,44 - 45}

Characteristic chest pain in patients with normal coronary arteries in the absence of ischemic signs on electrocardiogram has been reproduced with infusions of adenosine^{12 ,43} and dypiramidole.⁴¹ These studies suggest that persistent chest pain in these patients is not due to ischemia but rather to having lower cardiac pain thresholds⁴³ because their pain was reproduced at lower doses of adenosine compared with that of patients with coronary artery disease and that of normal control patients.

Previous studies in populations of patients with chest pain and normal coronary arteries demonstrated reproducible symptoms with stimulus to the right side of the heart, including probing of the right atrium,^{12 ,40 ,42} right atrial saline infusion,⁴² and right ventricular pacing.^{12 ,39} No ischemic changes were observed on electrocardiogram during right-sided heart stimulation,^{39 - 40} again suggesting that ischemia was unlikely the cause of chest pain. Chest pain in these studies may have resulted from mechanical distortion of mechanoreceptors to catheter stimulation.⁴⁰ Case 2 describes a patient with pain to intracoronary Doppler flow wire manipulation, which, to our knowledge, has not previously been reported, although cardiac syndrome X patients have been shown to experience chest pain during coronary angiography⁴⁰ and right-sided catheter manipulation.

Studies have also suggested that chest pain in cardiac syndrome X patients could be due to abnormal pain processing occurring at the brain cortical level. Rosen et al⁴⁴ showed increased regional blood flow by positron emission tomography of the right insular cortex, an area known to receive cardiopulmonary inputs, when these patients were subject to chest pain stimuli. Valeriani et al⁴⁵ investigated abnormalities in electrical cerebral signals to pain stimuli and found decreased habituation to repetitive noxious stimuli in cardiac syndrome X patients.

There are no specific guidelines for treatment of patients with abnormal cardiac pain sensation. Studies have primarily focused on patients who have cardiac syndrome X and abnormalities in cardiac pain perception, and current management guidelines from the American Heart Association and American College of Cardiology include risk factor management and medical therapy with nitrates, β-blockers, and calcium antagonists.²⁹

Imipramine is recommended in patients with continued chest pain despite the abovementioned therapies.²⁹ Cannon et al¹² conducted a randomized, double-blind, placebo-controlled trial of patients who had chest pain and normal coronary angiography results and were treated with placebo, clonidine, or imipramine. They found a significant reduction in chest pain episodes in the imipramine-treated group. The mechanism of action for imipramine is not yet completely understood. Imipramine may interact with pain-modulating neurons through its effect on norepinephrine-uptake–moderating visceral analgesic effects and therefore could benefit patients with heightened cardiac pain sensation and may improve coronary microvascular function via anticholinergic and α-antagonist effects, which have been demonstrated in the coronary and peripheral circulation.⁴⁶

Neurostimulation is an alternative treatment strategy for individuals with persistent chest pain caused by abnormal cardiac nociception and includes transcutaneous electrical nerve stimulation and spinal cord stimulation devices. Transcutaneous electrical nerve stimulation involves the use of electrodes to generate electrical current through the skin for pain control, whereas spinal cord stimulators are implanted devices generating electrical impulses. Patients have reported significantly fewer chest pain episodes after transcutaneous electrical nerve stimulation and spinal cord stimulation treatment.^{47 - 48} Spinal cord stimulation has also been shown to decrease hospitalization in these patients.^{49 - 50} The mechanism of action of neurostimulation is not clear, but it has been found to increase resting blood flow velocity in patients with normal coronary arteries, although not in cardiac transplant patients,⁵¹ suggesting a neural mechanism of action.

Persistent chest pain in patients with no obstructive coronary artery disease continues to be a challenging diagnosis for physicians. Given the increased risk for adverse outcomes in patients with the triad of persistent chest pain, evidence of myocardial ischemia, and no obstructive coronary artery disease related to microvascular coronary dysfunction, appropriate diagnosis and treatment can be vital. Coronary reactivity testing is a diagnostic tool that can be used to distinguish 2 relatively common chest pain etiologies, ie, microvascular coronary dysfunction and abnormal cardiac nociception.

Patients with microvascular coronary dysfunction are at risk for adverse cardiovascular events, and treatment directed at endothelial function can improve their outcomes, whereas abnormal cardiac nociception is benign and a symptom management condition. Therefore, it is important to establish diagnostic certainty in these patients to institute appropriate medical management. Current evidence suggests that the risk associated with coronary reactivity testing is relatively low compared with the adverse prognosis associated with microvascular coronary dysfunction. Further research is needed to understand diagnostic efficacy, elucidate the mechanism of action of current therapies, and investigate the effect of treatment on outcomes in these patients.