Abstract

Coronary embolism is a cause of acute myocardial infarction (AMI)in which obstructive foci enter the coronary circulation, block normal blood flow and precipitate ischemia. Precise studies focusing on patient population affected, pathophysiological mechanisms, and treatment strategies are scanty, in spite of a reported prevalence estimated at 2.9%. As the understanding of myocardial infarction without evidence of coronary artery disease continues to grow, an in-depth review of this previously seldomly reported subtype of coronary ischemia was in order. Patients suffering coronary embolism are 15 to 20 years younger than traditional AMI patients with a slight predominance towards male sex, which resembles the gender data of the populations affected by non-traditional myocardial infarction in published reports. While the expected prevalence rate of cardiovascular disease risk factors such as hypertension and hyperlipidemia are present, this population also has a relatively high prevalence of atrial fibrillation and valve pathology, especially endocarditis. Initial presentation is indistinguishable from other causes of myocardial infarction however fever is commonly present, when endocarditis with valvular involvement is the primary cause of the coronary embolism. Mechanical thrombectomy is the mainstay of treatment, followed by percutaneous coronary intervention. Mortality is the highest in patients who do not receive targeted treatment for the coronary embolism, particularly if only antimicrobial agents or anticoagulation without thrombolytic agents are employed. The unique features of coronary embolism highlighted in this historical study justify further examination in contemporary patient populations.

1. Introduction

Heart disease is the leading cause of morbidity and mortality worldwide ¹. In the United States, heart attack or acute myocardial infarction (AMI) has an annual incidence of about 805,000 cases with a death rate of 96.8 per 100,000 ². Though coronary artery disease (CAD) is the most common cause of AMI ³, infarction can happen without significant coronary artery stenosis. Myocardial infarction with nonobstructive coronary arteries (MINOCA) includes syndromes of the epicardium, microvasculature, and emerging concepts like myocardial necrosis due to oxygen supply-demand mismatch ⁴. A systematic review of 27 clinical trials/AMI registries reports MINOCA prevalence of 6% of AMI cases ⁵.

Coronary embolism (CE) is a heterogenous cause of MINOCA that can be divided into direct, paradoxical, and iatrogenic ⁶. Direct CE is when an embolism enters the coronary circulation, originating from systemic circulation or an intracardiac focus on the left side of the heart. Paradoxical CE enters coronary circulation after passing from venous circulation via a septal defect or patent foramen oval. Iatrogenic encompasses any emboli associated with a procedure, usually cardiothoracic surgery or percutaneous coronary intervention (PCI). There is overlap between the different categories and other organization schema have also been proposed.

In spite of early autopsy series identifying CE ⁷ and advancements in cardiovascular imaging, prevalence of CE has not yet been evaluated across multiple trials/registries. A single center, retrospective study of 1776 de novo AMI patients reports CE prevalence of 2.9% of AMI cases ⁸. As it could represent almost half of MINOCA cases, CE should not be considered a rare entity. We here undertake the first scoping study of CE to understand risk factors, diagnosis, sources of emboli, management, and mortality.

2. Methods

On July 23rd, 2019, a systematic search was conducted using Pubmed, Google Scholar, CINAHL, Cochrane CENTRAL and Web of Science databases. Studies listing the keywords “embolism, myocardial infarction” were used to identify cases of myocardial arrhythmias associated with marijuana use. The reference list of each report was reviewed for potential additional cases. All cases were reviewed in detail. Data reviewed included demographic data, cardiovascular (CV) risk factors, electrocardiography (EKG) findings, troponin levels, transthoracic echocardiography, electrophysiology study, urine drug screen findings, coronary angiogram and management when available.

3. Results

A total of 232 unique cases were identified in 190 publications (Table 1). Males comprised 55.2% of the cases reported and females comprised 44.8%. Overall age range was 4 months to 87 years with mean 50.2 ± 17.9 years and median 51 years; 57.3% of cases were younger than 55 years and 42.7% of cases were 55 years or older. Male age range was 4 months to 87 years with mean 49.6 ± 18.1 years and median 51 years. Female age was 12 to 87 years with mean 51.0 ± 17.9 years and median 51 years (Table 2).

The most prevalent cardiovascular risk factors and comorbidities in the population were as follows: atrial fibrillation 17.2%, hypertension 16.8%, hyperlipidemia 15.9%, prosthetic heart valves 15.9%, chronic valvular disease not treated with valve replacement 15.9%, rheumatic heart disease 12.5%, diabetes 9.91%, obesity 9.05%, history of CAD 9.05%, cerebrovascular disease 9.05%, and smoking 8.62%. Comprehensive risk factors and comorbidities are listed in Table 3.

Patients presented with chest pain 66.8%, fever 19.0%, sudden death 6.03%, hypotension 3.35%, shortness of breath 3.35%, stroke 0.862%, and without symptoms 0.431% (Table 4).

In the 219 cases reporting ECG findings (Table 5), ST elevation was found in 43.8%, ST depression in 11.4%, unspecified “infarct changes” in 26.5%, atrial fibrillation in 14.2%, T wave changes in 10.5%, and Q waves 3.65%.

A total of 173 cases included angiographic findings (Table 6). Occluded arteries were found in 95.4% of cases. Lesions were found in the left anterior descending artery 52.0%, left circumflex artery 17.3%, right main coronary artery 17.3%, left main coronary artery 9.83%, obtuse marginal artery 2.89%, posterolateral artery 2.31%, diagonal artery 1.73%, and intermediate artery 0.578%.

Of the 94 cases reporting TTE, the most common findings were wall motion abnormalities in 47.9% of cases, valvular dysfunction in 29.8%, systolic dysfunction in 25.5%, valve masses or vegetations in 22.3%, intracardiac masses in 18.1%, and chamber enlargement in 11.7%. Normal studies were reported in 9.57% of cases; additional findings are listed in Table 7.

Transesophageal echocardiograms were reported in 63 cases. The most common findings were valves masses or vegetations in 39.7% of cases, intracardiac masses in 33.3%, and valvular dysfunction in 25.4%. Other findings are listed in Table 8.

The reason for CE, as per the case authors, were endocarditis 22.8%, prosthetic heart valve complications 15.1%, thromboembolism 13.4%, atrial fibrillation 12.5%, iatrogenic 12.5%, non-thrombotic embolic sources (including solid tumors, bone marrow, and septic emboli) 9.91%, chronic valvular disease 8.62%, and rheumatic heart disease 6.47%. A comprehensive list of CE causes is listed in Table 9.

The most common interventions performed were thrombectomy 31.0%, PCI 22.0% (stent placed in 31.4% of these cases), anticoagulation 14.7%, thrombolytic therapy 9.48%, and antibiotics/antifungals 8.19%. A comprehensive list of interventions is listed in Table 10.

Death was reported in 35.7% of cases (Table 11), of which 34.9% occurred after no intervention was performed, 15.7% after anticoagulation, 12.0% after antibiotic/antifungal therapy alone were given, 6.02% after surgical intervention, 4.82% after aspiration thrombectomy, 3.61% after valve repair/replacement, and 2.41% after thrombolytic therapy. Recurrence of coronary embolism was reported in 7 cases, 3 of which resulted in death.

4. Discussion

The average age in the study population was approximately 15 years younger for males and 20 years younger for females than expected for all causes of MI, with females also representing a smaller percentage in all age groups relative to men than expected for all causes of MI ⁹. While younger age and skew towards male sex may be due to small sample size and known underdiagnosis of MI in females ¹⁰, these demographic data correlate with findings in larger reviews of MINOCA ⁵. Whether there is a unique patient population for CE should be confirmed in future studies.

Risk factors for CE overlapped with those for MI due to CAD (MI-CAD), specifically hypertension, hyperlipidemia, diabetes, CAD, CVD, and smoking. History of atrial fibrillation (AF), which is more commonly associated with cerebrovascular disease (CVD), was present in 17.2% of patients. Given the risk of thrombus formation in the left atrium in AF and its proximity to the coronary aortic cusps, thromboembolic events can occur in a similar fashion in the coronary circulation as they do in CVD. CE may occur less frequently than CVD in this setting due to the fast rate of flow of blood across the coronary ostia, high resistance to flow of the smaller caliber coronary vessels as dictated by Poiseuille's Law, and the acute angle of the origin of the coronary arteries ¹¹. Interestingly, demand ischemia from AF with rapid ventricular response has also been proposed as a mechanism for MINOCA due to AF ¹².

Additional risk factors in CE patients are chronic valvular disease with or without a history of rheumatic heart disease and prosthetic heart valves. Complications of chronic valvular disease and malfunctioning prosthetic valves usually include heart failure and/or pulmonary hypertension, rather than MI ^{13, 14, 15}. Abnormal blood flow and changes in left atrial volume caused by mitral valve disease, especially in AF, may cause thrombus formation ^{16, 17} thereby contributing to thromboembolic CE. Thrombosed prosthetic valves have the potential to cause thromboembolic events ¹⁸ and may cause CE this way.

Chest pain was the presenting complaint in 66.8% of CE cases, but it is not specific for ischemia and does not distinguish MI due to CE from MI-CAD ¹⁹. Shortness of breath, found in 3.35% of patients, is also a classic symptom for MI-CAD. Fever, found in 19.0% of cases, is an unusual symptom of an underlying cardiac issue except in infective endocarditis (IE) and pericarditis ²⁰. In cases where CE was later determined to have been caused by IE, fever was the initial complaint 69.8% of the time. Stroke was the initial reason for hospitalization in 2 cases, reinforcing that CE can be caused by a similar mechanism as embolic CVD.

Diagnostic investigations included ECG in 94.4% of cases, angiography in 74.6%, TTE in 40.5%, and TEE in 27.2%. The most common ECG finding was ST-elevation, followed by “infarct changes” and AF. These reports did not provide enough data to draw conclusions about how CE presents on ECG versus MI-CAD. Vessel distribution, determined angiographically, followed a similar pattern to MI-CAD ²¹. However, micro-emboli and advancement of CE into small caliber vessels may lead to infarction in territories of angiographically normal vessels, contributing to underdiagnosis ^{22, 23}. TTE identified the probable embolic source in 40.4% of cases when performed and TEE identified the probable source in 76.2%. TEE is 92% sensitive and 98% for detecting thrombi in the left atrial appendage, which is the most common area for thrombus formation ^{24, 25, 26}. TEE is 90% sensitive for native valve endocarditis and 85% sensitive for prosthetic and device-related endocarditis ²⁷.

Causes of CE were varied, but primarily involved pathology of the aortic and mitral valves. Prevalence of endocarditis (22.8%) was high in this population despite less than 4% of patients having a known history of endocarditis, valve vegetations, and intravenous drug use. 44.4% of patients had conditions predisposing to IE, including chronic valvular disease, prosthetic valves, and rheumatic heart disease ²⁸. This association of CE and IE may be a useful step towards creating a predictive tool for CE, especially because the Duke criteria for predicting IE is well-validated ²⁹. Additionally, since the incidence of IE has increased over a 10-year period in the United States ³⁰, it may become easier to assess prevalence of CE and the need for specific management strategies in IE patients.

Iatrogenic CE occurred most commonly in valve repair, valve replacement, and PCI at an overall rate equal to AF. Coronary catheterization is the best studied in this context, having been reported to cause CE with subtherapeutic heparinization or insufficient flushing of coagulated blood in the catheter, as well as incomplete aspiration of air ^{31, 32}. These procedures likely have unique risks for CE, warranting further investigation.

Nearly every case reviewed initially treated the patient for presumptive MI-CAD, with antiplatelet therapy, symptomatic treatment, and angiography playing a central role. Thrombectomy, particularly aspiration thrombectomy was the most frequently chosen intervention, followed by PCI with or without stent placement. Reason for using aspiration thrombectomy versus PCI were not explicitly states in these cases, though previous international guidelines recommend routine use of aspiration thrombectomy with primary PCI and in cases of increased risk factors or high thrombus burden ¹⁷. There was no difference in mortality at 180-days and a slightly higher incidence of stroke for patients who underwent aspiration ³³. ln the present study, all patients who underwent PCI survived, while 4 deaths occurred after aspiration thrombectomy. The ability to pathological examine relatively intact aspiration specimens has enhanced current understanding of the causes of CE, in spite of its questionable survival benefits.

Anticoagulation was the most common medical therapy given, with 14.7% of patients receiving it in the hospital and 43.1% receiving it upon discharge. Thrombolytic therapy was also given, most often as the primary therapy ³⁴, though its use is not well-studied and may even result in distalization of the thromboembolism to a smaller branch ³⁵. In the absence of clear guidelines, therapeutic approach to CE in patients of clinical presentation varies greatly. Reasons for not including other therapies including risk of stent infection in endocarditis ³⁶, lack of evidence for thrombolytic therapy ³⁷, and “free floating” emboli determined unlikely to respond to angioplasty ³⁸.

Death occurred at a higher rate in cases where CE was not explicitly treated. Mortality was 93.5% in patients receiving no intervention or symptomatic treatment, 58.8% in patients treated with antibiotics or antifungals alone, and 48.0% in patients treated with anticoagulation alone. One study ⁸ showed CE MI patients had a significantly higher mortality rate (hazard ratio 3.82) and cardiac death rate (HR, 5.39; 95% CI, 2.38-10.6) than MI-CAD patients, which underscores the need for better understanding of CE.

5. Conclusion

CE is an understudied cause of MI. The at-risk populations appear to be younger and more male than the general MI-CAD population and have unique risk factors in addition to those typically associated MI-CAD, including AF and valvular heart disease. While clinical presentation and ECG was indistinct from MI-CAD, TTE and TEE often demonstrated an embolic focus. When the clinical picture is considered alongside imaging, the most likely causes of CE were determined to be endocarditis with valvular involvement, malfunctioning or thrombosed prosthetic heart valves, intracardiac thrombi, and atrial fibrillation. There are no validated diagnostic algorithms for CE and as such it should be considered a possibility alongside the more traditional causes of MI, especially when the patient has known valve pathology or hemodynamically compromised valves on echocardiography. Definitive management strategies are challenging, in part due to the heterogeneity of causes of CE. The study shows that, in cases where the embolism itself was not directly addressed by thrombectomy, PCI, or thrombolysis, CE mortality was high.

Acknowledgements

This work is supported, in part, by the efforts of Dr. Moro O. Salifu M.D., M.P.H., M.B.A., M.A.C.P., Professor and Chairman of Medicine through NIH Grant number S21MD012474.

References