Lethal Presentations of Coronary Artery Spasm after an Event-Free
Period of Six Years following Initial Diagnosis

Isolated coronary artery spasm without atherosclerotic obstruction is an unusual cause of myocardial infarction (MI). The chest pain and electrocardiographic (ECG) changes can be resolved promptly following administration of nitroglycerin in the emergency department. The recommended longterm therapy includes calcium channel-blockers and nitrates. Although recurrent coronary artery spasm is associated with adverse outcomes,^1,2 data on the management and predictors of long-term outcomes are scant due to the rarity of the condition. Specifically, the optimal duration the patients have to be on medications as well as the risk of noncompliance remain unknown. We present a patient who developed spasminduced MI and cardiac arrest secondary to ventricular fibrillation after an event-free period of 6 years following the initial diagnosis of coronary artery spasm.

Case Report. A middle-aged woman with cardiovascular risk factors of hypertension and hyperlipidemia presented to our institution in 2001 with severe retrosternal chest pain and ECG evidence of acute inferior MI. However, her chest pain resolved with sublingual nitroglycerin, and coronary angiography did not reveal any significant coronary artery disease. Intracoronary administration of ergometrine reproduced the chest pain as well as diffuse coronary artery spasm, which was most prominent at the mid segment of the dominant left circumflex coronary artery. The coronary artery spasm resolved instantly following administration of intracoronary nitroclycerin (Figure 1). Echocardiography showed mild left ventricular systolic function (ejection fraction 45%) with inferior wall hypokinesia. A diagnosis of coronary artery spasm was made and the patient was discharged with calcium channel-blockers and nitroglycerin. She stopped the medication a few months later, but remained clinically stable.

After being well for 6 years, she was admitted again in 2007 with the same type of severe retrosternal chest pain. An ECG showed ST-segment elevation over the inferior leads with reciprocal ST-depression over the precordial leads (Figure 2). Once again, the chest pain resolved with sublingual nitroglycerin and emergency diagnostic coronary angiography showed normal coronary arteries (Figure 3). Subsequent blood tests showed elevated cardiac enzymes, confirming myocardial injury. Echocardiography showed left ventricular systolic dysfunction with an ejection fraction of 30%. A diagnosis of recurrent MI due to coronary artery spasm was made. The patient was discharged with a calcium channel-blocker and nitrates. Despite compliance with the medications, she was readmitted 1 month after discharge due to recurrent chest pain, but the ECG and blood tests were unremarkable.

Two months later, the patient developed another episode of severe retrosternal chest pain at home, followed by cardiac arrest. The family members started cardiopulmonary resuscitation until arrival of the ambulance. Onsite ECG showed ventricular fibrillation, and immediate defibrillation was carried out. She was readmitted to the hospital and recovered over the next few days. There were no identifiable causes such as drug abuse, mental stress or infection that could account for the repeated coronary spasm attacks which suddenly emerged 6 years after the initial diagnosis. In view of the recurrent coronary artery spasm causing MI and ventricular fibrillation, an implantable cardioverter defibrillator (ICD) was placed. The patient was well at 2-month follow up.

Discussion. Coronary artery spasm remains a disease entity that is not well studied. The most common presentation is Prinzmetal angina, which was first described in 1959. Very rarely, coronary artery spasm may present as MI and/or ventricular fibrillation. The diagnosis of coronary vasospasm is based upon the angiographic finding of epicardial coronarynarrowing that is promptly reversed by vasodilators.
Medical therapy with calcium channel-blockers and nitrates are the recommended therapy for patients presenting with Prinzmetal angina. There are few controlled trials assessing the efficacy of these therapies, in isolation or in combination, given the relatively small number of patients with isolated coronary artery vasospasm. Likewise, data are scarce on the optimal therapy as well as the duration of therapy for patients with coronary artery spasm. In a large series, Bory et al reported 277 patients with coronary artery spasm treated with calcium channelblockers.¹ Despite combination therapy at maximal dose, provocative tests remained positive in 12.8% of the patients, representing subtherapeutic medical management. In a median follow up of 89 months, the incidence of recurrent angina was 39%, MI 6.5% and cardiac death 3.6%. Compliance with treatment over this long follow-up period was not mentioned in the study, but it showed that adverse events are common, and coronary artery spasm is not a benign disease. Hypertension and the presence of luminal irregularities in the initial angiogram were found to be predictive factors for adverse events. Finally, coronary stent implantation to the spastic segment was recently reported in patients refractory to medical therapy.^3,4
In the setting of spasm-induced MI, mechanical complications such as free wall or interventricular septal rupture, malignant arrhythmias and sudden death can occur.^1,5 Treatment of ventricular fibrillation due to coronary artery spasmremains empirical. In a small series of 7 patients who were treated with calcium channel-blockers after cardiac arrest due to coronary artery spasm, 1 developed repeated cardiac arrest on long-term follow up.² This study was performed in the era when ICD implantation was not widely available. Although the authors reported “favorable” long-term outcomes, the recurrent cardiac arrest rate of 14.2% (1/7) may seem high in the present-day context. A number of anecdotal reports on the use of ICDs have been published since then.⁶ Given the malignant nature of cardiac arrest due to ventricular fibrillation, randomized trials on ICDs versus medical therapy in this specific patient subgroup are unlikely to be carried out. Implantation of an ICD as a norm in this condition can be expected.
Our case report illustrates that the natural course of coronary artery spasm can be malignant and unpredictable. Recurrent MI or cardiac arrest due to ventricular fibrillation may occur many years following initial presentation with MI. ICD implantation is warranted in patients presenting with cardiac arrest due to coronary spasm.