One Surprise after Another
http://www.100md.com
《新英格兰医药杂志》
In this Journal feature, information about a real patient is presented in stages (boldface type) to an expert clinician, who responds to the information, sharing his or her reasoning with the reader (regular type). The authors' commentary follows.
A previously healthy 22-year-old man presented to an emergency department reporting that he had had three days of intermittent abdominal pain. Although the pain was initially mild and crampy and was relieved with bismuth subsalicylate, on the day he went to the hospital the patient awoke with severe, midepigastric pain and had two episodes of diarrhea. He reported that he had had nausea and had decreased his oral intake over the past several days. He said there had been no vomiting, dysuria, fever, chills, or ingestion of unusual or undercooked foods and no recent travel. During the preceding week, he had had a productive cough and had noticed a slight decrease in exercise tolerance.
Intermittent epigastric pain with nausea can be due to common gastrointestinal disorders, such as peptic ulcer disease, cholecystitis, pancreatitis, viral gastroenteritis, irritable bowel syndrome, or dyspepsia unrelated to ulcer. My first thought would be that the productive cough is related to bronchitis, pneumonia, or another respiratory infection. I would also consider other conditions characterized by both gastrointestinal and pulmonary disease, such as aspiration from vomiting, gastroesophageal reflux leading to asthma, or congestive heart failure with pulmonary and hepatic congestion. More remote possibilities would be a subdiaphragmatic abscess leading to diaphragmatic irritation or to a pleural effusion or an intestinal parasitic infection (e.g., with strongyloides) that has a pulmonary phase.
Aside from bismuth, the patient's only medication was a daily multivitamin. He had no known drug allergies. He had a two-year history of cigarette use and reported drinking approximately 10 beers each weekend night and several beers each night during the week. He had a history of cocaine use, but not recently. His family history was notable in that his father had been told that he had an "irregular heart rate." The patient also had been told he had an irregular heartbeat at a health fair the preceding summer, but he had never had chest pain or palpitations. He also mentioned a recent upper respiratory tract infection, which he thought he had caught from his fiancée.
Alcohol is a common cause of upper gastrointestinal tract disorders, including gastritis, hepatitis, and pancreatitis. I would initially attribute his abdominal pain and nausea to his heavy alcohol use. His past cocaine use raises the possibility of ongoing polysubstance abuse. Cocaine can cause intestinal ischemia. An irregular heartbeat could represent ectopic beats or an arrhythmia. His father's history raises the possibility of a hereditary cardiac disorder, such as congenital long-QT syndrome or hypertrophic cardiomyopathy, or possibly he and his father have a common risk factor for arrhythmia, such as alcohol abuse.
A case of viral bronchitis that follows an upper respiratory tract infection could certainly account for the patient's productive cough, but an upper respiratory tract infection is often ill defined, can be a sign of acute infection with the human immunodeficiency virus (HIV), or may precede serious viral infections of other organs, such as the thyroid, heart, or meninges. His nonspecific symptoms collectively hint at a multisystem disease; however, each condition can readily be explained by one or more common ailments.
In the emergency department, the patient stated that his symptoms had resolved. On physical examination, he appeared well. His temperature was 37.3°C, blood pressure 147/74 mm Hg, pulse 72 beats per minute, respiratory rate 20 breaths per minute, and oxygen saturation 97 percent while breathing ambient air. He had a pierced tongue, which was free of erythema, and clear lungs. His cardiac examination revealed a sustained point of maximal impulse, a regular rate and rhythm, and the absence of a murmur, rub, or gallop. His abdomen was nontender and nondistended; the bowel sounds were normal, and there was no organomegaly. A specimen of his stool was negative for occult blood. The patient asked to be discharged, since his abdominal pain had resolved.
If the patient's hypertension was confirmed on repeated readings, I might attribute this to his alcohol abuse or possibly cocaine use, although he reported no recent use of the latter. The normal temperature, oxygen saturation, and clear lungs make pneumonia unlikely, but a chest radiograph is still warranted. Given his benign abdominal examination and the spontaneous resolution of his pain, a catastrophic abdominal process is extremely unlikely. The sustained point of maximal impulse is of concern, however, and may suggest a structural abnormality of the heart, particularly hypertrophy or dilatation. His constellation of symptoms may reflect nothing more than a simple viral syndrome; however, his examination and the history of an irregular heart rate and prior cocaine use make further testing warranted, including an assessment of the cardiac enzymes and an electrocardiogram.
Initial laboratory studies revealed a white-cell count of 11,100 cells per cubic millimeter with a normal differential and a hematocrit of 45.6 percent. Measurements of his electrolytes, liver-function tests, coagulation indexes, and the levels of amylase and lipase were normal. Chest radiography showed clear lung fields without effusion or consolidation and a normal cardiac silhouette. An initial electrocardiogram (Figure 1A) showed a normal sinus rhythm, left axis deviation, and Q waves in leads II, III, aVF, and V3. Right-sided leads showed no signs of right ventricular involvement, and posterior leads showed ST-segment elevation in leads V7, V8, and V9. No earlier electrocardiograms were available for comparison. The level of troponin I was elevated at 2.6 ng per milliliter (normal, less than 1.1 ng per milliliter). The patient reported that he still had no abdominal or chest pain.
Figure 1. Electrocardiograms Obtained at the Time of Presentation (Panel A) and Nine Hours Later (Panel B).
Acute myocardial infarction is usually diagnosed when at least two of three criteria — ischemic symptoms, electrocardiographic changes, and elevated cardiac enzyme levels — are met. If this patient's severe epigastric pain turned out to be angina, he would meet all three criteria. Although his young age, the absence of continued pain, and the absence of ST-segment depression in the precordial leads (which might indicate posterior ischemia) make me question this working diagnosis, these features do not eliminate the possibility of an acute myocardial infarction.
What should we do now? Along with standard initial medical treatment for an acute myocardial infarction, I would ask the cardiology service to perform cardiac catheterization.
The patient said he had no personal or family history of hypercoagulable states, coronary artery disease, or hypercholesterolemia. Toxicologic screening of the urine was negative for the presence of cocaine or amphetamines. Aspirin, a beta-blocker, and an angiotensin-converting–enzyme inhibitor were prescribed, and the patient was transferred to the telemetry floor. A repeated electrocardiogram (Figure 1B) revealed new T-wave flattening in the inferior leads and resolution of the peaked T waves in the precordium. The next set of blood tests for cardiac markers showed a rise in the level of troponin I to 5.8 ng per milliliter, a creatine kinase level of 102 units per liter (normal, 55 to 380 U per liter), and a creatine kinase MB fraction of 7 percent (normal 0 to 7 percent). The erythrocyte sedimentation rate was 7 mm per hour. Intravenous heparin was administered.
Given the dynamic changes in the electrocardiogram and the elevated troponin I level, I continue to be concerned about the possibility of an acute coronary syndrome. Whereas the patient's level of troponin I level is rising, his creatine kinase and MB fraction — the traditional but less sensitive markers of myocardial necrosis — are normal. What else might explain an elevated troponin I level? Although troponin T levels can be elevated in association with renal insufficiency (without myocardial damage), this is not the case with troponin I. Occasionally, interfering factors, such as rheumatoid factor or heterophile antibodies, cause false positive troponin I elevations, but the electrocardiographic changes argue against that possibility.
Dilated or hypertropic cardiomyopathy may result in both an elevated troponin I level and an abnormal electrocardiogram. If a cardiomyopathy were the cause of the patient's symptoms, however, in the absence of concomitant ischemia I would expect such changes to be static, not dynamic. Could this patient have had a pulmonary embolus? The ongoing inferior electrocardiographic changes and the elevation of troponin I could reflect increasing strain on his right ventricle. However, he has no known risk factors for venous thromboembolism, the chest pain that is associated with pulmonary embolism is rarely confused with abdominal pain, and he does not have hypoxemia. Pericarditis is usually associated with widespread ST-segment changes — sometimes with subsequent T-wave inversions — but the inflammation can sometimes be focal; irritation of the adjacent muscle can lead to elevation of troponin I levels. Finally, acute myocarditis can cause both electrocardiographic changes and elevation of cardiac enzymes. This man's age, the previous viral infection, and his decreased exercise tolerance are consistent with this diagnosis. However, even young adults with cardiomyopathy or a clinical diagnosis of myocarditis must be proved to be free of atherosclerotic coronary arteries or coronary anomalies. He needs a coronary angiogram.
Cardiac catheterization revealed no evidence of atherosclerotic disease. The coronary anatomy was normal. A left ventriculogram showed a dilated left ventricle with an ejection fraction of 35 percent. An endomyocardial biopsy was performed.
My working diagnosis is now acute myocarditis. The patient's recent upper respiratory tract infection, acute symptoms, dynamic electrocardiographic changes, fluctuating troponin I levels, and decreased left ventricular function suggest an ongoing myocardial disease. The results of catheterization rule out both coronary atherosclerosis and coronary artery anomalies, an important cause of infarction in young patients. The negative urine toxicology report eliminates cocaine as a cause of nonatherosclerotic ischemia. Vasospasm remains possible, but usually ST-segment elevations are associated with ongoing chest pain.
A transthoracic echocardiogram revealed, in addition to the depressed left ventricular function, akinesis of the inferior and posterior walls with apical dyskinesis and elevated left ventricular filling pressures. Telemetric monitoring showed occasional premature ventricular contractions and infrequent episodes of bigeminy. The patient remained symptom-free.
The findings on echocardiography support a diagnosis of myocarditis. Although the list of infections, toxins, drugs, and systemic disorders that can cause myocarditis is long, the most commonly identified trigger is an enteroviral infection, particularly one caused by coxsackievirus. I think that either a viral infection or the patient's heavy alcohol use is the culprit here.
Additional results of laboratory studies included negative tests for HIV, rheumatoid factor, and antinuclear antibodies. Microscopical examination of the cardiac-biopsy specimen taken at the time of catheterization revealed active myocarditis (Figure 2 and Figure 3) with fibrosis, lymphocytic infiltration, and collagen deposition. The overall appearance was consistent with acute viral myocarditis; however, the presence of fibrosis raised the possibility that an acute process was superimposed on chronic cardiomyopathy. Other than the trichrome stain for fibrosis, no special staining was performed. The patient was discharged with directions to continue the treatment with the angiotensin-converting–enzyme inhibitor and a beta-blocker. When he was seen six weeks later for a repeated echocardiographic evaluation, he was found to be symptomatically well. His ejection fraction was marginally lower, at 31 percent, but there was some improvement in the regional wall-motion abnormalities.
Figure 2. Microscopical Views of a Myocardial-Biopsy Specimen.
The low-power view (Panel A) shows myocyte dropout and damage. The high-power view (Panel B) shows lymphocyte infiltration.
Figure 3. Cardiac-Biopsy Specimen with Trichrome Staining Showing Increased Collagen and Fibrosis.
For this patient, in addition to adherence to the prescribed regimen of proper medications, cessation of alcohol consumption is critically important in order to preserve his cardiac function.
Commentary
Clinicians have increasingly come to appreciate the virtues of Bayesian reasoning and to recognize the importance of the prevalence of a given condition (the "pretest probability") in interpreting the results of any diagnostic test.1 This case forced both the clinicians caring for the patient and the discussant to consider relatively low-prevalence diagnoses at various junctures.
The first critical juncture came at the outset. Although neither the patient's clinicians nor the discussant would (or should) have considered a myocardial process a likely cause of abdominal pain in a 22-year-old man, there were some relatively subtle clues that prompted further consideration. The history of the irregular heartbeat and the sustained point of maximal impulse — coupled with the knowledge that inferior wall injury and diaphragmatic irritation can be manifested as epigastric pain — led the patient's medical team to obtain an electrocardiogram and troponin I measurements. It is reasonable to have a low threshold for testing for unlikely but life-threatening and treatable illnesses, especially when the tests are relatively safe and inexpensive.
When both cardiac tests unexpectedly came back positive — and unequivocally so — the physicians had to decide whether the results reflected an acute myocardial infarction with ST-segment elevation or another disorder. In the context of a suggestive electrocardiogram and elevated cardiac enzymes, most physicians proceed as if myocardial infarction is present until it is proven otherwise. Both the clinicians and the discussant, however, remained unsettled by the unusual circumstances. Rather than pursuing an aggressive path (immediate catheterization) or discounting the concerning test results, the clinicians chose a middle ground: medical management and close observation with further invasive testing as warranted. As it happened, further changes in troponin I values and on the electrocardiogram quickly prompted cardiac catheterization.
The fact that the coronary anatomy was normal meant that the most worrisome diagnosis (acute coronary syndrome) could be ruled out. In this patient with normal coronary arteries, striking electrocardiographic changes, fluctuating levels of troponin I, left ventricular dysfunction, and symptoms of an antecedent upper respiratory infection, myocarditis became the most probable diagnosis. As noted by the discussant, other causes, including pericarditis and hypertrophic cardiomyopathy, might have explained the observed electrocardiographic changes and the troponin I elevations, but these potential diagnoses were inconsistent with the other clinical data.
The diagnosis of myocarditis is often challenging to make. In addition to mimicking acute myocardial infarction, this condition can have a spectrum of severity — from a nonspecific viral-like illness to a fulminant cardiomyopathy. Physical findings are often limited and may only reflect decreased contractility, as shown here by the sustained point of maximal impulse. Although several organisms have been implicated, the most common infectious causes in North America are the enteroviruses coxsackievirus A and B and adenoviruses. The initial damage in cases of viral cardiomyopathy is primarily due to the direct cytotoxic effects of the virus on the myocyte.2 However, an autoimmune process soon develops in susceptible patients (cardiac proteins that are normally sequestered intracellularly are released during cell lysis). Thus, even when viral clearance is complete, cytotoxic T-cell–mediated destruction of cardiac myocytes may persist.3
Troponin I levels are elevated in only one third of patients with myocarditis, and thus endomyocardial biopsy remains the gold standard for diagnosis, despite its limited sensitivity and specificity for myocarditis4 and its potential risks (a perforation rate of approximately 0.5 percent). Even when biopsy indicates myocarditis, it infrequently reveals specific causes, such as eosinophilic or giant-cell myocarditis — rare diseases that are usually treated with aggressive immunosuppressive therapy or that prompt referral for transplantation.5 As this case shows, histopathological findings may help establish the chronic nature of the disorder, but they may not do so unambiguously.
The biopsy findings in this case showed a striking amount of fibrosis and lymphocytic infiltration, thus raising the possibility of an underlying cocaine-induced (catecholamine-based) or alcohol-related chronic cardiomyopathy with a superimposed acute viral infection.6 The absence of obvious long-standing symptoms does not fully exclude a chronic cardiomyopathy, since young patients with good physiological reserve may tolerate substantial reductions in the ejection fraction until an acute event supervenes. Fibrosis may in some cases occur early in the course of acute myocarditis; the possibility that an acute viral infection alone accounts for the course of the disease in this case cannot be ruled out.
Whereas most patients with acute viral myocarditis recover spontaneously, a dilated cardiomyopathy develops in approximately one in five patients.7 As for all patients with systolic dysfunction, treatment primarily includes angiotensin-converting–enzyme inhibition and beta-blockade. Given the role of autoimmunity in myocarditis-induced damage, immunosuppressive therapy has also been used. However, randomized studies of azathioprine and cyclosporine have not shown benefits.8 Although this patient's follow-up echocardiogram showed some early signs of improved wall motion, a substantial minority of patients never regain normal cardiac function.
Supported by a Career Development Award from the Health Services Research and Development Program of the Department of Veterans Affairs and a Patient Safety Developmental Center Grant from the Agency for Healthcare Research and Quality (P20-HS11540, to Dr. Saint).
We are indebted to Dr. Kanu Chatterjee for his comments on the manuscript.
Source Information
From the Department of Medicine, University of California, San Francisco, School of Medicine, San Francisco (N.J.L., L.S.W., G.D., R.M.W.); the San Francisco Veterans Affairs Medical Center, San Francisco (G.D.); and the Ann Arbor Veterans Affairs Health Services Research and Development Center of Excellence, and the Patient Safety Enhancement Program, Ann Arbor Veterans Affairs Medical Center and University of Michigan Health System, Ann Arbor, and the Department of Internal Medicine, University of Michigan Medical School — all in Ann Arbor (S.S.).
Address reprint requests to Dr. Wachter at the Department of Medicine, Box 0120, Room M-994, University of California, San Francisco, San Francisco, CA 94143-0120, or at bobw@medicine.ucsf.edu.
References
Sackett DL, Haynes RB, Tugwell P, Guyatt GH. Clinical epidemiology: a basic science for clinical medicine. Boston: Little, Brown, 1991.
Pinney SP, Mancini DP. Myocarditis and specific cardiomyopathies. In: Fuster V, Alexander RW, O'Rourke RA, eds. Hurst's the heart. 11th ed. New York: McGraw-Hill, 2004:1949-74.
Noutsias M, Pauschinger M, Poller WC, Schultheiss HP, Kuhl U. Current insights into the pathogenesis, diagnosis and therapy of inflammatory cardiomyopathy. Heart Fail Monit 2003;3:127-135.
Smith SC, Ladenson JH, Mason JW, Jaffe AS. Elevations of cardiac troponin I associated with myocarditis: experimental and clinical correlates. Circulation 1997;95:163-168.
O'Connell JB, Costanzo-Nordin MR, Subramanian R, Robinson JA. Dilated cardiomyopathy: emerging role of endomyocardial biopsy. Curr Probl Cardiol 1986;11:445-507.
O'Connell JB, Mason JW. Inflammatory myocarditis. In: Hosenpud JD, Greenberg BH, eds. Congestive heart failure: pathophysiology, differential diagnosis, and comprehensive approach to management. New York: Springer-Verlag, 1994:223-33.
D'Ambrosio A, Patti G, Manzoli A, et al. The fate of acute myocarditis between spontaneous improvement and evolution to dilated cardiomyopathy: a review. Heart 2001;85:499-504.
Mason JW, O'Connell JB, Herskowitz A, et al. A clinical trial of immunosuppressive therapy for myocarditis. N Engl J Med 1995;333:269-275.(Nicholas J. Leeper, M.D.,)
A previously healthy 22-year-old man presented to an emergency department reporting that he had had three days of intermittent abdominal pain. Although the pain was initially mild and crampy and was relieved with bismuth subsalicylate, on the day he went to the hospital the patient awoke with severe, midepigastric pain and had two episodes of diarrhea. He reported that he had had nausea and had decreased his oral intake over the past several days. He said there had been no vomiting, dysuria, fever, chills, or ingestion of unusual or undercooked foods and no recent travel. During the preceding week, he had had a productive cough and had noticed a slight decrease in exercise tolerance.
Intermittent epigastric pain with nausea can be due to common gastrointestinal disorders, such as peptic ulcer disease, cholecystitis, pancreatitis, viral gastroenteritis, irritable bowel syndrome, or dyspepsia unrelated to ulcer. My first thought would be that the productive cough is related to bronchitis, pneumonia, or another respiratory infection. I would also consider other conditions characterized by both gastrointestinal and pulmonary disease, such as aspiration from vomiting, gastroesophageal reflux leading to asthma, or congestive heart failure with pulmonary and hepatic congestion. More remote possibilities would be a subdiaphragmatic abscess leading to diaphragmatic irritation or to a pleural effusion or an intestinal parasitic infection (e.g., with strongyloides) that has a pulmonary phase.
Aside from bismuth, the patient's only medication was a daily multivitamin. He had no known drug allergies. He had a two-year history of cigarette use and reported drinking approximately 10 beers each weekend night and several beers each night during the week. He had a history of cocaine use, but not recently. His family history was notable in that his father had been told that he had an "irregular heart rate." The patient also had been told he had an irregular heartbeat at a health fair the preceding summer, but he had never had chest pain or palpitations. He also mentioned a recent upper respiratory tract infection, which he thought he had caught from his fiancée.
Alcohol is a common cause of upper gastrointestinal tract disorders, including gastritis, hepatitis, and pancreatitis. I would initially attribute his abdominal pain and nausea to his heavy alcohol use. His past cocaine use raises the possibility of ongoing polysubstance abuse. Cocaine can cause intestinal ischemia. An irregular heartbeat could represent ectopic beats or an arrhythmia. His father's history raises the possibility of a hereditary cardiac disorder, such as congenital long-QT syndrome or hypertrophic cardiomyopathy, or possibly he and his father have a common risk factor for arrhythmia, such as alcohol abuse.
A case of viral bronchitis that follows an upper respiratory tract infection could certainly account for the patient's productive cough, but an upper respiratory tract infection is often ill defined, can be a sign of acute infection with the human immunodeficiency virus (HIV), or may precede serious viral infections of other organs, such as the thyroid, heart, or meninges. His nonspecific symptoms collectively hint at a multisystem disease; however, each condition can readily be explained by one or more common ailments.
In the emergency department, the patient stated that his symptoms had resolved. On physical examination, he appeared well. His temperature was 37.3°C, blood pressure 147/74 mm Hg, pulse 72 beats per minute, respiratory rate 20 breaths per minute, and oxygen saturation 97 percent while breathing ambient air. He had a pierced tongue, which was free of erythema, and clear lungs. His cardiac examination revealed a sustained point of maximal impulse, a regular rate and rhythm, and the absence of a murmur, rub, or gallop. His abdomen was nontender and nondistended; the bowel sounds were normal, and there was no organomegaly. A specimen of his stool was negative for occult blood. The patient asked to be discharged, since his abdominal pain had resolved.
If the patient's hypertension was confirmed on repeated readings, I might attribute this to his alcohol abuse or possibly cocaine use, although he reported no recent use of the latter. The normal temperature, oxygen saturation, and clear lungs make pneumonia unlikely, but a chest radiograph is still warranted. Given his benign abdominal examination and the spontaneous resolution of his pain, a catastrophic abdominal process is extremely unlikely. The sustained point of maximal impulse is of concern, however, and may suggest a structural abnormality of the heart, particularly hypertrophy or dilatation. His constellation of symptoms may reflect nothing more than a simple viral syndrome; however, his examination and the history of an irregular heart rate and prior cocaine use make further testing warranted, including an assessment of the cardiac enzymes and an electrocardiogram.
Initial laboratory studies revealed a white-cell count of 11,100 cells per cubic millimeter with a normal differential and a hematocrit of 45.6 percent. Measurements of his electrolytes, liver-function tests, coagulation indexes, and the levels of amylase and lipase were normal. Chest radiography showed clear lung fields without effusion or consolidation and a normal cardiac silhouette. An initial electrocardiogram (Figure 1A) showed a normal sinus rhythm, left axis deviation, and Q waves in leads II, III, aVF, and V3. Right-sided leads showed no signs of right ventricular involvement, and posterior leads showed ST-segment elevation in leads V7, V8, and V9. No earlier electrocardiograms were available for comparison. The level of troponin I was elevated at 2.6 ng per milliliter (normal, less than 1.1 ng per milliliter). The patient reported that he still had no abdominal or chest pain.
Figure 1. Electrocardiograms Obtained at the Time of Presentation (Panel A) and Nine Hours Later (Panel B).
Acute myocardial infarction is usually diagnosed when at least two of three criteria — ischemic symptoms, electrocardiographic changes, and elevated cardiac enzyme levels — are met. If this patient's severe epigastric pain turned out to be angina, he would meet all three criteria. Although his young age, the absence of continued pain, and the absence of ST-segment depression in the precordial leads (which might indicate posterior ischemia) make me question this working diagnosis, these features do not eliminate the possibility of an acute myocardial infarction.
What should we do now? Along with standard initial medical treatment for an acute myocardial infarction, I would ask the cardiology service to perform cardiac catheterization.
The patient said he had no personal or family history of hypercoagulable states, coronary artery disease, or hypercholesterolemia. Toxicologic screening of the urine was negative for the presence of cocaine or amphetamines. Aspirin, a beta-blocker, and an angiotensin-converting–enzyme inhibitor were prescribed, and the patient was transferred to the telemetry floor. A repeated electrocardiogram (Figure 1B) revealed new T-wave flattening in the inferior leads and resolution of the peaked T waves in the precordium. The next set of blood tests for cardiac markers showed a rise in the level of troponin I to 5.8 ng per milliliter, a creatine kinase level of 102 units per liter (normal, 55 to 380 U per liter), and a creatine kinase MB fraction of 7 percent (normal 0 to 7 percent). The erythrocyte sedimentation rate was 7 mm per hour. Intravenous heparin was administered.
Given the dynamic changes in the electrocardiogram and the elevated troponin I level, I continue to be concerned about the possibility of an acute coronary syndrome. Whereas the patient's level of troponin I level is rising, his creatine kinase and MB fraction — the traditional but less sensitive markers of myocardial necrosis — are normal. What else might explain an elevated troponin I level? Although troponin T levels can be elevated in association with renal insufficiency (without myocardial damage), this is not the case with troponin I. Occasionally, interfering factors, such as rheumatoid factor or heterophile antibodies, cause false positive troponin I elevations, but the electrocardiographic changes argue against that possibility.
Dilated or hypertropic cardiomyopathy may result in both an elevated troponin I level and an abnormal electrocardiogram. If a cardiomyopathy were the cause of the patient's symptoms, however, in the absence of concomitant ischemia I would expect such changes to be static, not dynamic. Could this patient have had a pulmonary embolus? The ongoing inferior electrocardiographic changes and the elevation of troponin I could reflect increasing strain on his right ventricle. However, he has no known risk factors for venous thromboembolism, the chest pain that is associated with pulmonary embolism is rarely confused with abdominal pain, and he does not have hypoxemia. Pericarditis is usually associated with widespread ST-segment changes — sometimes with subsequent T-wave inversions — but the inflammation can sometimes be focal; irritation of the adjacent muscle can lead to elevation of troponin I levels. Finally, acute myocarditis can cause both electrocardiographic changes and elevation of cardiac enzymes. This man's age, the previous viral infection, and his decreased exercise tolerance are consistent with this diagnosis. However, even young adults with cardiomyopathy or a clinical diagnosis of myocarditis must be proved to be free of atherosclerotic coronary arteries or coronary anomalies. He needs a coronary angiogram.
Cardiac catheterization revealed no evidence of atherosclerotic disease. The coronary anatomy was normal. A left ventriculogram showed a dilated left ventricle with an ejection fraction of 35 percent. An endomyocardial biopsy was performed.
My working diagnosis is now acute myocarditis. The patient's recent upper respiratory tract infection, acute symptoms, dynamic electrocardiographic changes, fluctuating troponin I levels, and decreased left ventricular function suggest an ongoing myocardial disease. The results of catheterization rule out both coronary atherosclerosis and coronary artery anomalies, an important cause of infarction in young patients. The negative urine toxicology report eliminates cocaine as a cause of nonatherosclerotic ischemia. Vasospasm remains possible, but usually ST-segment elevations are associated with ongoing chest pain.
A transthoracic echocardiogram revealed, in addition to the depressed left ventricular function, akinesis of the inferior and posterior walls with apical dyskinesis and elevated left ventricular filling pressures. Telemetric monitoring showed occasional premature ventricular contractions and infrequent episodes of bigeminy. The patient remained symptom-free.
The findings on echocardiography support a diagnosis of myocarditis. Although the list of infections, toxins, drugs, and systemic disorders that can cause myocarditis is long, the most commonly identified trigger is an enteroviral infection, particularly one caused by coxsackievirus. I think that either a viral infection or the patient's heavy alcohol use is the culprit here.
Additional results of laboratory studies included negative tests for HIV, rheumatoid factor, and antinuclear antibodies. Microscopical examination of the cardiac-biopsy specimen taken at the time of catheterization revealed active myocarditis (Figure 2 and Figure 3) with fibrosis, lymphocytic infiltration, and collagen deposition. The overall appearance was consistent with acute viral myocarditis; however, the presence of fibrosis raised the possibility that an acute process was superimposed on chronic cardiomyopathy. Other than the trichrome stain for fibrosis, no special staining was performed. The patient was discharged with directions to continue the treatment with the angiotensin-converting–enzyme inhibitor and a beta-blocker. When he was seen six weeks later for a repeated echocardiographic evaluation, he was found to be symptomatically well. His ejection fraction was marginally lower, at 31 percent, but there was some improvement in the regional wall-motion abnormalities.
Figure 2. Microscopical Views of a Myocardial-Biopsy Specimen.
The low-power view (Panel A) shows myocyte dropout and damage. The high-power view (Panel B) shows lymphocyte infiltration.
Figure 3. Cardiac-Biopsy Specimen with Trichrome Staining Showing Increased Collagen and Fibrosis.
For this patient, in addition to adherence to the prescribed regimen of proper medications, cessation of alcohol consumption is critically important in order to preserve his cardiac function.
Commentary
Clinicians have increasingly come to appreciate the virtues of Bayesian reasoning and to recognize the importance of the prevalence of a given condition (the "pretest probability") in interpreting the results of any diagnostic test.1 This case forced both the clinicians caring for the patient and the discussant to consider relatively low-prevalence diagnoses at various junctures.
The first critical juncture came at the outset. Although neither the patient's clinicians nor the discussant would (or should) have considered a myocardial process a likely cause of abdominal pain in a 22-year-old man, there were some relatively subtle clues that prompted further consideration. The history of the irregular heartbeat and the sustained point of maximal impulse — coupled with the knowledge that inferior wall injury and diaphragmatic irritation can be manifested as epigastric pain — led the patient's medical team to obtain an electrocardiogram and troponin I measurements. It is reasonable to have a low threshold for testing for unlikely but life-threatening and treatable illnesses, especially when the tests are relatively safe and inexpensive.
When both cardiac tests unexpectedly came back positive — and unequivocally so — the physicians had to decide whether the results reflected an acute myocardial infarction with ST-segment elevation or another disorder. In the context of a suggestive electrocardiogram and elevated cardiac enzymes, most physicians proceed as if myocardial infarction is present until it is proven otherwise. Both the clinicians and the discussant, however, remained unsettled by the unusual circumstances. Rather than pursuing an aggressive path (immediate catheterization) or discounting the concerning test results, the clinicians chose a middle ground: medical management and close observation with further invasive testing as warranted. As it happened, further changes in troponin I values and on the electrocardiogram quickly prompted cardiac catheterization.
The fact that the coronary anatomy was normal meant that the most worrisome diagnosis (acute coronary syndrome) could be ruled out. In this patient with normal coronary arteries, striking electrocardiographic changes, fluctuating levels of troponin I, left ventricular dysfunction, and symptoms of an antecedent upper respiratory infection, myocarditis became the most probable diagnosis. As noted by the discussant, other causes, including pericarditis and hypertrophic cardiomyopathy, might have explained the observed electrocardiographic changes and the troponin I elevations, but these potential diagnoses were inconsistent with the other clinical data.
The diagnosis of myocarditis is often challenging to make. In addition to mimicking acute myocardial infarction, this condition can have a spectrum of severity — from a nonspecific viral-like illness to a fulminant cardiomyopathy. Physical findings are often limited and may only reflect decreased contractility, as shown here by the sustained point of maximal impulse. Although several organisms have been implicated, the most common infectious causes in North America are the enteroviruses coxsackievirus A and B and adenoviruses. The initial damage in cases of viral cardiomyopathy is primarily due to the direct cytotoxic effects of the virus on the myocyte.2 However, an autoimmune process soon develops in susceptible patients (cardiac proteins that are normally sequestered intracellularly are released during cell lysis). Thus, even when viral clearance is complete, cytotoxic T-cell–mediated destruction of cardiac myocytes may persist.3
Troponin I levels are elevated in only one third of patients with myocarditis, and thus endomyocardial biopsy remains the gold standard for diagnosis, despite its limited sensitivity and specificity for myocarditis4 and its potential risks (a perforation rate of approximately 0.5 percent). Even when biopsy indicates myocarditis, it infrequently reveals specific causes, such as eosinophilic or giant-cell myocarditis — rare diseases that are usually treated with aggressive immunosuppressive therapy or that prompt referral for transplantation.5 As this case shows, histopathological findings may help establish the chronic nature of the disorder, but they may not do so unambiguously.
The biopsy findings in this case showed a striking amount of fibrosis and lymphocytic infiltration, thus raising the possibility of an underlying cocaine-induced (catecholamine-based) or alcohol-related chronic cardiomyopathy with a superimposed acute viral infection.6 The absence of obvious long-standing symptoms does not fully exclude a chronic cardiomyopathy, since young patients with good physiological reserve may tolerate substantial reductions in the ejection fraction until an acute event supervenes. Fibrosis may in some cases occur early in the course of acute myocarditis; the possibility that an acute viral infection alone accounts for the course of the disease in this case cannot be ruled out.
Whereas most patients with acute viral myocarditis recover spontaneously, a dilated cardiomyopathy develops in approximately one in five patients.7 As for all patients with systolic dysfunction, treatment primarily includes angiotensin-converting–enzyme inhibition and beta-blockade. Given the role of autoimmunity in myocarditis-induced damage, immunosuppressive therapy has also been used. However, randomized studies of azathioprine and cyclosporine have not shown benefits.8 Although this patient's follow-up echocardiogram showed some early signs of improved wall motion, a substantial minority of patients never regain normal cardiac function.
Supported by a Career Development Award from the Health Services Research and Development Program of the Department of Veterans Affairs and a Patient Safety Developmental Center Grant from the Agency for Healthcare Research and Quality (P20-HS11540, to Dr. Saint).
We are indebted to Dr. Kanu Chatterjee for his comments on the manuscript.
Source Information
From the Department of Medicine, University of California, San Francisco, School of Medicine, San Francisco (N.J.L., L.S.W., G.D., R.M.W.); the San Francisco Veterans Affairs Medical Center, San Francisco (G.D.); and the Ann Arbor Veterans Affairs Health Services Research and Development Center of Excellence, and the Patient Safety Enhancement Program, Ann Arbor Veterans Affairs Medical Center and University of Michigan Health System, Ann Arbor, and the Department of Internal Medicine, University of Michigan Medical School — all in Ann Arbor (S.S.).
Address reprint requests to Dr. Wachter at the Department of Medicine, Box 0120, Room M-994, University of California, San Francisco, San Francisco, CA 94143-0120, or at bobw@medicine.ucsf.edu.
References
Sackett DL, Haynes RB, Tugwell P, Guyatt GH. Clinical epidemiology: a basic science for clinical medicine. Boston: Little, Brown, 1991.
Pinney SP, Mancini DP. Myocarditis and specific cardiomyopathies. In: Fuster V, Alexander RW, O'Rourke RA, eds. Hurst's the heart. 11th ed. New York: McGraw-Hill, 2004:1949-74.
Noutsias M, Pauschinger M, Poller WC, Schultheiss HP, Kuhl U. Current insights into the pathogenesis, diagnosis and therapy of inflammatory cardiomyopathy. Heart Fail Monit 2003;3:127-135.
Smith SC, Ladenson JH, Mason JW, Jaffe AS. Elevations of cardiac troponin I associated with myocarditis: experimental and clinical correlates. Circulation 1997;95:163-168.
O'Connell JB, Costanzo-Nordin MR, Subramanian R, Robinson JA. Dilated cardiomyopathy: emerging role of endomyocardial biopsy. Curr Probl Cardiol 1986;11:445-507.
O'Connell JB, Mason JW. Inflammatory myocarditis. In: Hosenpud JD, Greenberg BH, eds. Congestive heart failure: pathophysiology, differential diagnosis, and comprehensive approach to management. New York: Springer-Verlag, 1994:223-33.
D'Ambrosio A, Patti G, Manzoli A, et al. The fate of acute myocarditis between spontaneous improvement and evolution to dilated cardiomyopathy: a review. Heart 2001;85:499-504.
Mason JW, O'Connell JB, Herskowitz A, et al. A clinical trial of immunosuppressive therapy for myocarditis. N Engl J Med 1995;333:269-275.(Nicholas J. Leeper, M.D.,)