Is Silent’ Atrial Fibrillation in Diabetics Associated with Cerebral Neurologic Events?
Abstract & Commentary
Edward P. Gerstenfeld, MD
Professor of Medicine, Chief, Cardiac Electrophysiology, University of California, San Francisco
Dr. Gerstenfeld does research for Biosense Webster, Medtronic, and Rhythmia Medical.
Source: Marfella R, et al. Brief episodes of silent atrial fibrillation predict clinical vascular brain disease in type 2 diabetic patients. J Am Coll Cardiol 2013; 62:525-530.
Patients with diabetes have twice the stroke risk of those without diabetes. Yet, only 15% of diabetic patients have symptomatic atrial fibrillation (AF). This study sought to address the question of whether there was a relationship between asymptomatic AF and cerebral events in type 2 diabetics with no evidence of AF or stroke.
The study included two phases: 1) a 4-year recruitment period to enroll patients from a cohort of 1992 type 2 diabetic patients and 2) a 3-year follow-up period. Patients included in the follow-up arm had to meet the following criteria: age < 60 years, successful completion of quarterly 48-hour Holter monitors during the screening phase, and completion of brain magnetic resonance imaging (MRI). Patients with any baseline history of AF, stroke, transient ischemic attack (TIA), or structural heart disease were excluded. Four hundred sixty-four patients were enrolled and compared to 240 healthy subjects without diabetes or any of the exclusion criteria. During the screening phase, patients underwent 48-hour Holter monitors every 3 months with silent AF classified as any episode lasting at least 10 minutes and < 48 hours in duration. Patients were then classified into either a group with silent AF (SAFE group, n = 176) or a group without silent AF (non-SAFE group, n = 288). All patients were treated with aspirin 75-325 mg daily, and if CHADS2 score was > 1, oral anticoagulation with warfarin was initiated. Patients were then followed for 36 months for the occurrence of any clinical neurologic events. The incidence of "silent" AF was much higher among diabetic patients compared to healthy controls (11% vs 1.6%; P < 0.001). The mean AF burden was 21 ± 15 hours in the SAFE group compared to 3 ± 1.4 hours in the healthy control group (P < 0.001). On MRI, a subclinical thromboembolic ischemic event (SCI) was detected more often in the SAFE group compared to the non-SAFE group (61% vs 29%; P < 0.01). The AF burden was correlated to the size and number of SCI events. Using multivariate analysis, silent AF, left atrial size, systolic blood pressure, and duration of diabetes were independently associated with SCI. During the 36-month follow-up, clinical AF episodes occurred in 15% of patients who were excluded from the analysis; 26 patients (15%) in the SAFE group and and 19 patients (7%) in the non-SAFE group were treated with warfarin anticoagulation. Despite this, a clinical ischemic stroke developed in more patients in the SAFE compared to the non-SAFE groups (17.3% vs 5.9%; P < 0.01). No strokes occurred in the healthy control group. Of the 43 stroke events, 42 were ischemic and one was hemorrhagic. The authors concluded that subclinical AF episodes are common in diabetics and predict a higher incidence of subsequent cerebral ischemic events.
Patients with type 2 diabetes have a higher incidence of cerebrovascular events and AF compared to those without diabetes. However, there has not been a definitive link established between the two. Diabetes is also associated with hypercoaguability and vascular endothelial dysfunction that could play a role in SCIs. This ambitious study carried out over 7 years has several important findings. First, the prevalence of silent AF in a population of type 2 diabetics is approximately 10% higher than previously reported. Second, patients with diabetes and silent AF have a very high incidence of SCI on MRI examination (61%). Finally, patients with diabetes and silent AF, even episodes lasting < 48 hours, have a high incidence of strokes during follow-up (17.3%). This should be an eye-opener to cardiologists following patients with diabetes mellitus.
Should we screen our patients with type 2 diabetes for silent AF? It is difficult to know if the patients in the centers enrolling for this study were subject to a referral bias that increased their thromboembolic risk. However, I think the data demonstrating a 10% prevalence of asymptomatic AF and strong association with both silent and overt clinical neurologic events support screening of high-risk diabetic patients for asymptomatic AF. Obviously the cost-effectiveness of this approach has not been established, nor has the efficacy of treating diabetic patients with asymptomatic AF with systemic anticoagulation. While this study was performed before the era of the newer oral anticoagulants, the advent of the oral direct thrombin and factor X inhibitors have certainly lowered the threshold for initiating systemic anticoagulation in AF patients at risk of stroke. One could certainly envision a multicenter, randomized trial screening diabetic patients for brief episodes of asymptomatic AF, and then randomizing them to aspirin or anticoagulation with a novel anticoagulant. However, until more data become available, it certainly is reasonable to have a low threshold to initiate anticoagulation in diabetic patients who have a clinical neurologic event, and to have increased vigilance screening of diabetic patients with multiple AF risk factors for silent AF. We await more data on this important topic.