Smoking, atrial fibrillation, and ischemic stroke

a confluence of epidemics

Abstract

Purpose of review 

Smoking and atrial fibrillation are major contemporary health concerns. They commonly coexist and are frequent causes of ischemic stroke. The purpose of this article is to describe recent scientific investigations about smoking, atrial fibrillation, and ischemic stroke, with a primary focus on prevention.

Recent findings 

Smoking predisposes to atrial fibrillation and is useful for the prediction of future atrial fibrillation. Several recent risk prediction models for adverse events associated with atrial fibrillation include smoking as a component. Smoking status identifies patients at high risk of incident atrial fibrillation, adverse events in an emergency ward after admission with atrial fibrillation, thromboembolic events following a diagnosis of atrial fibrillation, and potentially poor control of vitamin K antagonist treatment.

Summary 

From multiple perspectives of atrial fibrillation, patients who smoke represent a high-risk population. Appropriate preventive measures targeting this endangered population are paramount. These include smoking cessation, appropriate care in the emergency ward, and careful selection of the optimal antithrombotic strategy to reduce the major burden of ischemic stroke attributed to the confluence of the epidemics of smoking and atrial fibrillation.

INTRODUCTION

Atrial fibrillation is the most common cardiac arrhythmia and a frequent and devastating cause of morbidity and mortality from cardioembolic stroke ^[1,2]. The prevalence of atrial fibrillation has taken a daunting rise. It is expected to reach pandemic proportions due to the continued aging of the general population ^[3▪▪]. Atrial fibrillation may be prevalent in up to 1/3 of all patients with incident ischemic stroke ^[4▪].

One potential factor involved in the pathogenesis of atrial fibrillation is smoking ^[5–8]. Smoking causes a broad range of oncogenic ^[9], cardiovascular ^[10], cerebrovascular ^[10], and respiratory disease. Unfortunately, smoking continues to be a dominant health hazard worldwide and responsible for major healthcare costs ^[11].

The purpose of the current article is to describe recent scientific investigations linking the epidemics of smoking, atrial fibrillation, and ischemic stroke, with a primary focus on prevention strategies.

Box 1: no caption available

SMOKING AND PRECIPITATION OF ATRIAL FIBRILLATION

Smoking predisposes to atrial fibrillation ^[6,12–14] in a dose–response manner ^[15▪▪]. Possible mediators include diabetes mellitus ^[16▪], chronic obstructive pulmonary disease/impaired lung function ^[17▪,18], hypertension, and heart failure – all diseases that are likely to be caused by smoking ^[19▪▪], and which in turn can induce atrial fibrillation ^[15▪▪]. In addition, smoking also causes structural remodeling of the myocardium, which has direct toxic effects on the conductive properties and may lead to atrial fibrillation ^{[15▪▪,20]}. Continued tobacco use is also associated with atrial fibrillation after catheter ablation ^[15▪▪].

SMOKING STATUS AND PREDICTION OF INCIDENT ATRIAL FIBRILLATION

Smoking status has also proven valuable for the prediction of future atrial fibrillation. Among five other variables, smoking status is included in the Women's Health Study Atrial Fibrillation risk prediction algorithm identifying women at higher risk for incident atrial fibrillation ^[21] (see Table 1 for risk prediction models covered in this article).

Table 1: Components of various risk prediction models for use in patients with atrial fibrillation that include information on smoking habits

In the Cohorts for Heart and Aging Research in Genomic Epidemiology - Atrial Fibrillation model for the prediction of incident atrial fibrillation, current smoking status is included as a component adding predictive ability independent of other traditional cardiovascular-risk predictors such as age, race, height, weight, systolic and diastolic blood pressure, use of antihypertensive medication, diabetes, and history of myocardial infarction and heart failure ^{[22▪▪,23]}. In summary, smoking not only causes atrial fibrillation, but smoking status also aids in the identification of people at high risk of future incident atrial fibrillation.

REDUCING THE BURDEN

How can we reduce the combined burden caused by the confluence of smoking and atrial fibrillation? Two approaches are immediately evident. First, as smoking predisposes to atrial fibrillation, primary prevention of atrial fibrillation is partly achieved through smoking cessation. Secondly, improving the identification of patients with established atrial fibrillation who stand to gain most from anticoagulant treatment is essential. This is currently provided through the use of risk stratification models used for the prediction of thromboembolic events and anticoagulant-related bleeding in patients with incident atrial fibrillation. The stratification models used for the prediction of thromboembolic events will be briefly introduced in the following section.

ATRIAL FIBRILLATION, STROKE-RISK STRATIFICATION, AND ANTITHROMBOTIC STRATEGIES: WHAT IS KNOWN ALREADY?

Anticoagulant therapy, when used appropriately, has a dramatic effect on the prevention of thromboembolic events – preventing approximately two-thirds of all atrial fibrillation-related strokes ^[24]. Although anticoagulant treatment is highly preventive, not all patients stand to gain. Despite the overall advantage in reducing the risk of stroke, anticoagulant therapy is a double-edged sword, with an inherent risk of potentially fatal bleeding.

RISK STRATIFICATION TOOLS FOR STROKE RISK IN ATRIAL FIBRILLATION

The risk of thromboembolic events is not homogeneous, but varies according to the presence or absence of various lifestyle and clinical-risk factors. These risk factors have formed the basis for stroke-risk stratification schemes and clinical practice guidelines for stroke prevention in patients with atrial fibrillation. The purpose of these schemes is to identify patients who could benefit from oral anticoagulation.

In recent international guidelines ^[25,26], the CHA₂DS₂-VASc score [Cardiac failure or dysfunction, Hypertension, Age ≥75 (doubled), Diabetes, Stroke (doubled)-Vascular disease, Age 65–74 and Sex category (Female)] score ^[27] is recommended as a risk stratification tool for all patients with atrial fibrillation. It is advocated that all patients with a single risk factor from the CHA₂DS₂-VASc score (except women with female sex as their only risk factor) should be considered for preventive anticoagulant treatment. Despite these clear guideline recommendations, many atrial fibrillation patients do not receive oral anticoagulation as often as is clinically indicated ^[4▪,28▪]. It has been suggested that physicians may overestimate bleeding risk from oral anticoagulation and underestimate its benefits in stroke prevention ^[29].

When oral anticoagulation with adjusted-dose vitamin K antagonists is used, the quality of anticoagulation control, as reflected by the time in therapeutic range of the international normalized ratio (INR), is an important determinant of risk of thromboembolism and bleeding ^[30,31]. The Sex [female], Age<60, Medical history, Treatment [interacting drugs], Tobacco use, Race [non-white] (SAMe-TT₂R₂) score is a scheme made to assess the likelihood of poor INR control among atrial fibrillation patients receiving vitamin K antagonists using patient-related clinical parameters ^[32] (Table 1). One of the features found to be associated with poor INR control is tobacco use, and this is given double weight in this risk stratification model. Hence, a patient with incident atrial fibrillation who smokes and has an indication for anticoagulant treatment may benefit most from treatment with one of the non-vitamin K antagonist oral anticoagulants (NOACs).

SMOKING STATUS AND PREDICTION OF THROMBOEMBOLIC EVENTS IN ATRIAL FIBRILLATION

There is an on-going interest in improving the existing thromboembolic-risk prediction models, and recent studies have focused on smoking behavior. Using data from the population-based Diet, Cancer and Health study, a recent investigation explored the potential improvement of prediction of thromboembolic events and death in 3161 patients with incident atrial fibrillation using information on smoking status ^[33▪]. After adjustment for components of the CHA₂DS₂-VASc score and time-varying use of anticoagulant therapy during follow-up, both former and current smokers remained at higher risk of thromboembolism or death in what resembled a dose–response-dependent manner. That smokers represent a high-risk population, also in the setting of atrial fibrillation, should not come as a surprise. The new and interesting evidence provided here is the possible added predictive value to the currently recommended risk stratification tool, the CHA₂DS₂-VASc score. Despite the CHA₂DS₂-VASc including several components that are likely to act as mediators between smoking and thromboembolic events and death (hypertension, heart failure, diabetes mellitus, previous stroke, and vascular disease), smoking status was clearly capable of identifying patients at high risk of future adverse events. This might improve the identification of patients with atrial fibrillation at high risk of thromboembolic events, and thereby increase the use of appropriate anticoagulant treatment in atrial fibrillation. Importantly, the analysis from the Diet, Cancer and Health cohort is a study of prediction, not causality. Therefore, a conclusion of a beneficial effect of smoking cessation on future risk of adverse events in those with atrial fibrillation (although theoretically highly likely) cannot be drawn from the study.

Others have since confirmed the findings from the Danish cohort. In a recent Japanese study, Nakagawa et al.^[34▪] investigated patients already diagnosed with atrial fibrillation and the impact of smoking on the following long-term outcomes: all-cause death, stroke (ischemic and hemorrhagic), cardiac events (myocardial infarction or hospitalization for worsening of heart failure), bleeding (gastrointestinal and intracranial hemorrhages), and cardiovascular event (stroke and cardiac event). They found that a history of smoking, independently of age, antithrombotic treatment and CHADS₂ and CHA₂DS₂-VASc scores, predicted all-cause mortality, death from stroke or death from malignant diseases. Persistent smoking, independently of age or antithrombotic treatment, also predicted intracranial bleeding. Hence, smoking status clearly identifies patients with atrial fibrillation at high risk of subsequent ischemic stroke, but may also identify patients at risk of bleeding during anticoagulant treatment.

The QStroke score is an algorithm developed to estimate the risk of incident stroke or transient ischemic attack ^[35]. It has been validated in populations both with and without atrial fibrillation. It includes major risk factors for cardiovascular disease, including information on smoking divided into five levels. Interestingly, they compare the algorithm with the CHA₂DS₂-VASc score in patients with atrial fibrillation. Investigators have demonstrated some evidence of statistically improved discriminatory performance of QStroke compared with the CHA₂DS₂-VASc score. However, the QStroke algorithm is unlikely to replace the CHA₂DS₂-VASc score for the prediction of thromboembolic events due to its major degree of complexity compared with the CHA₂DS₂-VASc score.

SMOKING STATUS AND PREDICTION OF OTHER EVENTS RELATED TO ATRIAL FIBRILLATION

Another recently introduced risk stratification tool is the Risk Estimator Decision Aid for Atrial Fibrillation (RED-AF) model for estimating risk of experiencing a 30-day adverse event in patients with atrial fibrillation admitted to an emergency department ^[36,37]. RED-AF was derived from a cohort of 832 patients. It assigns points according to smoking status and to age, sex, coexisting disease (e.g., heart failure, hypertension, and chronic obstructive pulmonary disease), medication use (e.g., beta-blockers, diuretics, etc.), physical examination findings (e.g., dyspnea, palpitations, and peripheral edema), and adequacy of emergency department ventricular rate control. Atrial fibrillation combined with just one of these risk factors was independently associated with an adverse event: 30-day emergency department return visit, unscheduled hospitalization, cardiovascular complication, or death. The decision to hospitalize patients with atrial fibrillation is often subjective and multifactorial, according to the patient's acute and chronic conditions. Emergency physicians need to feel confident identifying stable, low-risk patients with atrial fibrillation. The tool is supposed to help guide emergency physicians on which patients with atrial fibrillation can be safely managed as outpatients and which should be hospitalized.

SMOKING AND ISCHEMIC STROKE SUBTYPE

Smoking is a major preventable cause of ischemic stroke. Importantly, though, several subtypes of ischemic stroke exist, for which the optimal preventive pharmacological approach may vary ^[38]. In the general population, smoking causes mainly strokes that are atherothrombotic in nature, whereas strokes caused by atrial fibrillation are of cardioembolic origin. Future studies should address whether the smoking behavior of patients with atrial fibrillation is indeed predictive of cardioembolic stroke, as these strokes are preferably prevented by oral anticoagulant treatment, whereas strokes of atherothrombotic nature are likely to be reduced primarily by the use of antiplatelet therapy ^[39]. If the higher risk of ischemic stroke seen in atrial fibrillation patients who smoke can be attributed to a higher risk of cardioembolic stroke, and not only atherothrombotic stroke, this would clearly establish smoking status as a patient characteristic favoring oral anticoagulant treatment.

SUGGESTIONS FOR FUTURE RESEARCH

Given the current investigations of smoking status as an important risk predictor in patients with atrial fibrillation ^{[33▪,34▪]}, its potential additive predictive value to the CHA₂DS₂-VASc score merits further investigation. Further studies are needed to confirm the relationship between smoking status and risk of cardioembolic stroke in patients with atrial fibrillation, as this may subsequently inform decision-making regarding anticoagulant treatment in everyday clinical practice. Whether smoking status clearly favors the initiation of anticoagulant treatment depends also on the degree of anticoagulant-related bleeding in these patients. Further studies investigating the net clinical benefit of anticoagulant treatment in patients with atrial fibrillation and a history of smoking are needed.

CONCLUSION

Smoking is clearly an important patient characteristic among those with atrial fibrillation. From an etiologic perspective, prevention of incident smoking and promotion of smoking cessation would prevent incident cases of atrial fibrillation. Additionally, smoking status is a critical patient feature in the prediction of the many adverse events related to atrial fibrillation. The identification of smoking status permits the prediction of incident atrial fibrillation, adverse events in an emergency ward after admission with atrial fibrillation, thromboembolic events following a diagnosis of atrial fibrillation, and potentially poor control of vitamin K antagonist treatment. Patients who smoke represent a high-risk population from many angles of adversity related to the growing epidemic of atrial fibrillation. Appropriate measures of prevention targeting this endangered population are called for.

Acknowledgements

None.

Financial support and sponsorship

None.

Conflicts of interest

TBL has served as an investigator for Janssen Scientific Affairs, LLC, and Boehringer Ingelheim and has been on the speaker bureaus for Bayer, BMS/Pfizer, Janssen Pharmaceuticals, Takeda, Roche Diagnostics, and Boehringer Ingelheim. GYHL has served as a consultant for Bayer, Astellas, Merck, Sanofi, BMS/Pfizer, Daiichi-Sankyo, Biotronik, Portola, and Boehringer Ingelheim and has served as a speaker for Bayer, BMS/Pfizer, Boehringer Ingelheim, Daiichi-Sankyo, and Sanofi Aventis. The remaining authors have no conflicts of interest.

REFERENCES AND RECOMMENDED READING

Papers of particular interest, published within the annual period of review, have been highlighted as:

• ▪ of special interest
• ▪▪ of outstanding interest