“Arrhythmogenic” Stroke

Neurologists will tell you that they are unable to identify a cause of stroke in about one-third of their patients who present with an ischemic stroke. They refer to this type of stroke as “cryptogenic”. The word cryptogenic has a Greek origin — kryptos means hidden and genein means to produce. One potential cause of stroke in these patients could be antecedent self-terminating episodes of atrial fibrillation (AF) that lead to left atrial thrombus formation and thromboembolism. These patients could be said to have had an “arrhythmogenic” stroke. 

The current practice is to treat patients with cryptogenic stroke with antiplatelet therapy alone, partly because it would not make sense to expose every patient with a cryptogenic stroke to oral anticoagulation (OAC), and partly because a prior randomized trial did not show a benefit from empiric OAC in these patients. It would make more sense to identify which patients are actually having AF, and target those patients with OAC. Historically, AF is in fact looked for in patients with cryptogenic stroke. These patients are usually placed on a monitor when they are in the hospital and are often given a 24-hour Holter monitor after discharge. A more logical method, however, knowing that AF is often asymptomatic and that AF episodes can be separated by very long intervals of sinus rhythm, would be to monitor patients for an extended amount of time. A safe and effective tool to do this is with an implantable cardiac monitor (ICM, or implantable loop recorder).

A recent study, CRYptogenic STroke and underlying AtriaL Fibrillation (CRYSTAL-AF), randomized 441 patients with cryptogenic stroke to standard AF monitoring versus an ICM. The preliminary results of this study were recently presented by Dr. Richard Bernstein from Northwestern University’s Feinberg School of Medicine at the American Stroke Association’s International Stroke Conference 2014; the long-term CRYSTAL-AF three-year data were presented by Dr. Rod Passman, also from Northwestern, at the American College of Cardiology’s annual scientific sessions in March 2014. The study enrolled patients over age 40 years who had a cryptogenic stroke, with infarct seen on MRI or CT, within the previous 90 days (average of 38 days) and no mechanism determined after a 12-lead ECG, 24-hour ECG monitoring, transesophageal echocardiography, imaging to rule out an arterial source, and screening for a hypercoagulable state in patients <55 years old. A diagnosis of AF was only made if the patient had what appeared to be AF detected for at least 30 seconds. After only one year, AF was detected in over 10% of patients with an ICM — seven times more frequently than in patients who were monitored using conventional means. Furthermore, after three years, AF was found in 30% of patients with an ICM — over nine times more often than in the other patients. Also interesting was that these episodes of AF were not ultra-brief (95% lasted over six minutes), were asymptomatic 75% of the time, and led to initiation of OAC in 89% of patients. 

Based on the results of the CRYSTAL-AF study, it seems worthwhile to perform long-term monitoring with an ICM in every patient with cryptogenic stroke who is a potential candidate for OAC. The recent availability of the injectable ICM makes this idea even more attractive. Certainly the issue of cost-effectiveness will be raised, and some will want to see cost-effectiveness data before adopting the practice. However, with hope, reimbursement for implantation of these devices in a less costly outpatient environment will be established. Regardless, it is hard to believe that, in a cohort of patients with cryptogenic stroke, implantation of a device that can detect AF with a diagnostic yield of 30% in three years would not be cost-effective in any environment.