Paroxysmal atrial fibrillation

INTRODUCTION — Atrial fibrillation (AF) is the most common treated arrhythmia. Its prevalence in the population increases with age, and it is estimated to affect over 4 percent of the population above the age of 60 [1-3]. (See "Epidemiology, risk factors, and prevention of atrial fibrillation", section on 'Epidemiology'.)

This topic will discuss the clinical presentation, etiology, natural history, and management for paroxysmal AF (PAF; also known as intermittent AF) highlighting differences and similarities compared with more sustained forms of AF.

Perioperative AF is discussed separately. (See "Atrial fibrillation in patients undergoing noncardiac surgery" and "Atrial fibrillation and flutter after cardiac surgery".)

DEFINITION AND PREVALENCE — PAF is defined as AF that terminates spontaneously or with intervention within seven days of onset [4]. "Persistent," "longstanding persistent," and "permanent" are terms used for types of AF with episode durations longer than one week. (See "Atrial fibrillation: Overview and management of new-onset atrial fibrillation", section on 'Classification and terminology'.)

PAF has been reported as comprising 25 to 62 percent of AF cases [5]. The prevalence of PAF may be underestimated, as many episodes (including some lasting more than 48 hours) are asymptomatic [6,7]. Also, the duration of recurrent AF episodes vary over time in each individual, and progression to persistent or permanent AF is common. (See 'Recurrence of AF' below and 'Progression to persistent AF' below.)

Risk factors for developing PAF are similar to those associated with sustained AF and include age, hypertension, structural heart disease including valve disease, and obstructive sleep apnea. (See "Epidemiology, risk factors, and prevention of atrial fibrillation".)

PATHOGENESIS — Factors that precipitate PAF, particularly in patients without apparent structural heart disease, are incompletely understood, but are thought to be linked to premature atrial complexes (PACs; also referred to a premature atrial beats, premature supraventricular complexes, or premature supraventricular beats) and alterations in autonomic nervous system activity. (See "Mechanisms of atrial fibrillation", section on 'Mechanisms of atrial fibrillation: triggers and substrates' and "Epidemiology, risk factors, and prevention of atrial fibrillation", section on 'Potentially reversible triggers'.)

Substrate alterations that may contribute to sustained AF are discussed separately. (See "Mechanisms of atrial fibrillation", section on 'Mechanisms of atrial fibrillation: triggers and substrates'.)

Premature atrial complexes — Studies have shown that the majority of episodes of PAF are triggered by PACs [8-10], and greater frequency of PACs is associated with greater risk of AF [11]. PAF episodes are less commonly preceded by atrial flutter, atrial tachycardia, or paroxysmal supraventricular tachycardias [10,12]. Most PACs triggering PAF originate near the ostia of the pulmonary veins (eg, in 89 and 94 percent of cases in two series [8,9]). Less commonly, foci occur in the right/left atria, the vein of Marshall, and the superior vena cava [9,13,14]. The importance of the pulmonary veins in the genesis of PAF is further demonstrated by the beneficial effect of pulmonary vein isolation. (See "Atrial fibrillation: Catheter ablation".)

PACs appear to be most important as triggers of PAF in patients who have structurally normal or near-normal hearts. However, it is unclear if modification of the PAC burden can reduce AF risk. The relative importance of PAC and other triggers versus an abnormal substrate is less clear in patients with significant structural heart disease. (See "Mechanisms of atrial fibrillation", section on 'Mechanisms of atrial fibrillation: triggers and substrates'.)

Autonomic nervous system — The autonomic nervous system may be involved, as both parasympathetic (vagal) and sympathetic (adrenergic) tone [15] promote the development and maintenance of AF [16]. (See "Mechanisms of atrial fibrillation", section on 'Role of the autonomic nervous system'.)

The frequency of autonomic stimuli as a trigger for PAF has not been well studied. In a report from the European Heart Survey on AF, 1517 patients with PAF were categorized according to trigger pattern: adrenergic, vagal, or both (6, 15, and 12 percent of the total group, respectively) [17]. The prevalence of underlying heart disease (heart failure, coronary artery disease, valvular heart disease, or hypertension) was similar in the three groups. In a report of patients referred for radiofrequency ablation due to symptomatic drug-refractory PAF, the prevalence of vagal, adrenergic, or other AF was found 27, 7, and 66 percent of the time, respectively [18].

●Parasympathetic tone – Vagally-mediated AF commonly occurs at night or in the early morning when vagal tone is normally predominant, and it is often seen in athletic young men without apparent heart disease who have slow heart rates during rest or sleep [16,19]. The induction of AF by vagal stimulation may result from shortening of the atrial refractory period in only some areas of the atrial myocardium, thus producing heterogeneity of atrial refractoriness [20]. Acetylcholine and increased vagal tone shorten the atrial myocardial refractory period but vagal innervation of the atria is heterogeneous. Vagal stimulation and associated hypotension may rarely contribute to the development of syncope in association with episodes of AF [21]. (See 'Evaluation' below.)

●Sympathetic tone – Increased adrenergic tone may be associated with AF in patients with underlying heart disease, associated with hyperthyroidism, and during exercise or other activity [16]. Increased sympathetic tone shortens the atrial myocardial refractory period and increases atrial myocardial conduction velocity. However, AF during exercise testing is a rare event; in a retrospective review of 3000 exercise tests, there were only four episodes of AF [22]. Sympathetic stimulation has also been suggested as the cause for AF associated with surgery, particularly cardiac surgery. (See "Cardiovascular effects of hyperthyroidism" and "Atrial fibrillation and flutter after cardiac surgery".)

There is no evidence suggesting that selection of therapy based upon the type of autonomic dysfunction improves outcomes, with the exception of perioperative therapy (such as beta blockers) in patients undergoing cardiac surgery. (See "Atrial fibrillation and flutter after cardiac surgery".)

CLINICAL PRESENTATION

General symptoms and signs — As for more sustained AF, PAF may or may not be accompanied by symptoms, and the spectrum of symptoms is broad. The most common complaints include palpitations, often accompanied by dyspnea (ranging from dyspnea with exertion to dyspnea at rest), a sensation of lightheadedness, fatigue, weakness, or generalized malaise. The severity and extent of symptoms and signs are affected by the patient’s underlying cardiac condition, age, level of physical activity, and rapidity and regularity of the ventricular response. In patients with preexisting heart failure or at risk for heart failure, the loss of atrial contraction and rapid ventricular rate associated with PAF may precipitate heart failure (which may manifest as dyspnea, peripheral edema, and weight gain). (See "The management of atrial fibrillation in patients with heart failure", section on 'mechanisms of cardiac dysfunction'.)

In some patients, a rapid ventricular rate may precipitate angina and/or ischemic electrocardiogram (ECG) changes which may be accompanied by troponin elevation; the presentation may be consistent with an acute coronary syndrome or demand ischemia. (See "Acute coronary syndrome: Terminology and classification" and "Elevated cardiac troponin concentration in the absence of an acute coronary syndrome".)

Syncope is rare — PAF rarely causes syncope [23]. In some case, syncope and PAF are both triggered by another disorder (eg, pulmonary embolism or vagal stimulus) [23,24]. The termination of PAF is occasionally associated with lightheadedness, presyncope or syncope due to a prolonged sinus pause, which may be caused by sinus node dysfunction (waveform 1). Sinus node dysfunction is commonly associated with AF, including PAF [25]. Some patients have progressive atrial structural remodeling that may cause both sinus node dysfunction and AF, while others may develop sinus node dysfunction secondary to electrical remodeling caused by AF. (See "Sinus node dysfunction: Clinical manifestations, diagnosis, and evaluation", section on 'Symptoms' and "Sinus node dysfunction: Epidemiology, etiology, and natural history", section on 'Etiology'.)

Rapid ventricular response in AF alone is rarely the cause of syncope except for patients with Wolff-Parkinson-White syndrome and a rapidly conducting accessory pathway. (See "Syncope in adults: Epidemiology, pathogenesis, and etiologies", section on 'Cardiac arrhythmias' and "Wolff-Parkinson-White syndrome: Anatomy, epidemiology, clinical manifestations, and diagnosis", section on 'Clinical manifestations'.)

EVALUATION — The evaluation of patients with AF includes history, physical examination, ECG, echocardiography, and laboratory tests, as described separately. (See "Atrial fibrillation: Overview and management of new-onset atrial fibrillation", section on 'Evaluation'.)

The ECG in AF displays rapid, low-amplitude, continuously varying fibrillatory (f) waves and no discrete p waves. The ventricular rhythm is generally irregularly irregular (lacking a repetitive pattern); however, AF can have a regular rhythm in the setting of complete heart block. The ECG in patients with AF is described in detail separately (algorithm 1). (See "The electrocardiogram in atrial fibrillation".)

NATURAL HISTORY

Recurrence of AF — PAF is rarely an isolated event. The recurrence rate for PAF is high and has varied among studies. The incidence of recurrence in various reports has ranged from 70 percent at one year (without antiarrhythmic therapy) [26] to 60 to 90 percent at four to six years [27-30]. These reported rates likely underestimate the actual rate of recurrence, since many episodes (including some that last more than 48 hours) are asymptomatic [6,7]; such prolonged asymptomatic episodes occurred in 17 percent of patients in a report using continuous monitoring [6]. That study also showed that 40 percent of patients had episodes of AF-like symptoms in the absence of AF [6].

Risk factors for recurrent AF in patients with PAF are similar to those for recurrence after cardioversion to sinus rhythm in patients with persistent AF. In the Stroke Prevention in Atrial Fibrillation (SPAF) trial, patients with any AF recurrence were more likely to have heart failure (17 versus 8 percent) or a prior myocardial infarction (15 versus 5 percent) than patients without recurrence [31]. Echocardiographic factors associated with AF recurrence included moderate to severe left ventricular dysfunction (12 versus 3 percent in those without recurrence) and larger left atrial diameter. Compared with a normal left atrial diameter of less than 4.0 cm, the relative risk of recurrent AF was 1.6 with a left atrial diameter between 4.1 and 5.0 cm and 4.5 above 5.0 cm.

Progression to persistent AF — By definition, PAF spontaneously reverts to sinus rhythm within seven days of onset. The percent of patients with PAF who progress to persistent or permanent AF increases with time. In different reports, the rate of progression to persistent or permanent AF was 8, 12, 18, and 25 percent at one, two, four, and five years, respectively [30-32]. In a 2016 registry report, approximately 36 percent of patients with PAF had progressed to persistent AF within 10 years [33]. A study of patients with AF and cardiac devices found that in a minority of patients with AF, persistent AF reverted to PAF without any therapeutic intervention, underscoring our limited understanding of the natural history of AF [34].

A number of studies have attempted to identify predictors of progression to persistent AF [30,31]. In the report from SPAF cited above, the following differences were noted between the patients with paroxysmal and persistent AF [31]: patients with PAF were younger, had lower rates of hypertension and heart failure, had smaller left atrial diameters (4.3 versus 4.8 cm), and were less likely to have moderate to severe left ventricular systolic dysfunction (6 versus 18 percent). For every 1 cm increase in left ventricular systolic dimension, there was a 1.8-fold greater risk of developing persistent AF. The risk of progression to persistent AF with older age was quantified in two reports showing a hazard ratio (HR) of 1.41 to 1.82 for each 10-year increase in age [30,32].

The risk of transition from PAF to persistent AF also depends upon the underlying etiology for the arrhythmia. Progression has been reported in approximately 66 percent of patients with rheumatic mitral stenosis, 40 percent with hypertension, and 27 percent with ischemic heart disease [35,36].

Risk of embolization — Studies suggest that the risk of thromboembolic events is higher in patients with permanent or persistent AF, with higher AF burden (ie, daily duration or percentage of time in AF), and in the days immediately after an episode of AF. A potential limitation of prior studies is that they have been conducted in patients with implantable cardiac monitors and/or devices, potentially limiting generalizability to other patient groups. Patients with PAF and long periods of apparent sinus rhythm still may be at significant risk for thromboembolism; PAF episodes and stroke are not always temporally associated [37]. Recurrent episodes of PAF are common and may be asymptomatic [6,7,38]. Also, the surface ECG may not reflect left atrial appendage mechanical function despite the presence of sinus rhythm [39].

The relationship between subclinical AF and cryptogenic stroke is discussed separately. (See "Cryptogenic stroke and embolic stroke of undetermined source (ESUS)", section on 'Occult atrial fibrillation'.)

●PAF versus permanent or persistent AF – A 2015 meta-analysis of 12 studies of 100,000 persons stratified clinical outcomes by AF type (paroxysmal and persistent or permanent) [40]. The incidences of thromboembolism and all-cause mortality were higher with non-paroxysmal AF than with PAF (adjusted HRs 1.38 [95% CI 1.19-1.61] and 1.22 [95% CI 1.09-1.36], respectively). In a community-based (Japan) study (not included in the above meta-analysis) of 1588 patients with PAF and 1716 with sustained AF, PAF was an independent predictor of lower stroke/systemic embolism risk (HR of approximately 0.50 in multiple models) compared with sustained AF [41].

●AF burden – Evidence suggests that AF burden is associated with risk of thromboembolic events, independent of CHADS₂-VASC or ATRIA risk scores. However, the available data do not provide a clear threshold of AF burden or duration for elevated thromboembolic risk. The data suggest that AF episodes lasting longer than 17 to 24 hours impart significant risk, but the risk associated with shorter AF episodes is uncertain.

•Daily duration of AF – Studies have come to different conclusions as to a possible threshold of daily AF duration that confers a long-term risk of embolization. A pooled analysis of data on individuals with AF from five prospective studies (n = 10,016) evaluated the risk of ischemic stroke over a median follow-up of two years with the following cut-off points of AF daily duration: 5 minutes, and 1, 6, 12, and 23 hours [42]. The stroke risks were over twofold for AF duration ≥1 hour compared with shorter durations (HR 2.11, 95% CI 1.22-3.64).

The prospective ASSERT study included 2580 subjects [43]. The initial report indicated that asymptomatic (subclinical) episodes of AF lasting >6 minutes, as compared with no episodes or episodes lasting ≤6 minutes, were associated with an increased risk of ischemic stroke or systemic embolism (adjusted HR 2.50; 95% CI 1.28-4.89). However, a follow-up study reporting on 2455 patients (mean follow-up of 2.5 years) found that only patients with episodes of subclinical AF lasting longer than 24 hours, compared with patients with no subclinical episodes, were at increased risk of subsequent stroke or systemic embolism (adjusted HR 3.24, 95% CI 1.51-6.95) [44].

•Percentage of time in AF

-A retrospective study in 1900 adults with PAF who were not on anticoagulation found that the burden of AF was associated with risk of thromboembolism independent of ATRIA or CHA₂DS₂-VASc risk scores [45]. The median burden of AF was 4.4 percent (interquartile range 1.1 to 17.2 percent). The unadjusted incidence of thromboembolism while not taking anticoagulation was 1.51 per 100 person-years. The highest tertile of AF burden (≥11.4 percent) was associated with a higher risk of thromboembolism (3.16 [95% CI 1.51-6.62]) compared with the lower tertiles.

-In the RATE registry, 5379 patients with pacemakers or implantable cardioverter-defibrillators were followed for a median duration of 22.9 months. Over 37,000 ECGs were adjudicated for the presence of atrial tachycardia (AT) and/or AF [46]. Runs of ≥3 consecutive premature atrial complexes (PACs) were defined as AT/AF, and short episodes were defined as runs <15 to 20 seconds. Clinical events (emergency department visits/hospitalizations for heart failure, atrial or ventricular arrhythmia, stroke or transient ischemic attack [TIA], and syncope) were more likely to occur in patients with long AT/AF episodes as opposed to short AT/AF episodes only. In addition, although not statistically significant, there was a suggestion that patients with only short AT/AF episodes had lower risk of stroke plus TIA.

●Temporal association between AF presentation and stroke onset – Two case-cross-over studies have examined the association between time in AF since onset and stroke risk. One study linked a commercial database of continuous rhythm recording to Veterans Administration clinical records. Stroke risk was highest during the five days immediately after an episode of AF and waned over the 30 days following the episode [47]. Another study linked a large health insurance database with implantable cardiac device data in 890 patients with stroke [48]. This study had similar findings to the above Veterans study in that five hours or more of AF was associated with increased stroke risk (odds ratio [OR], 3.71 95% CI 2.06-6.70). This study also showed that stroke risk was highest during days 1 to 5 following the start of AF (OR 5.00; 95% CI, 2.62-9.55). Furthermore, AF >23 hours per day on a given day was associated with a high risk of stroke (OR 5.00, 95% CI 2.08-12.01).

MANAGEMENT — The management for patients with PAF is similar to that for the general population of patients with AF. Important considerations are the duration of AF and the presence or absence of symptoms during episodes. (See "Atrial fibrillation: Overview and management of new-onset atrial fibrillation".)

Prevention of recurrence — The decision to start antiarrhythmic drug therapy in patients with PAF is similar to that in the general population of patients with AF. (See "Management of atrial fibrillation: Rhythm control versus rate control".)

Patients with frequent or highly symptomatic PAF may require pharmacologic or nonpharmacologic therapy to prevent recurrence. The choice between these therapies is discussed separately (See "Maintenance of sinus rhythm in atrial fibrillation: Catheter ablation versus antiarrhythmic drug therapy".):

●The choice of drug therapy is determined by associated clinical conditions as well as patient preference. (See "Antiarrhythmic drugs to maintain sinus rhythm in patients with atrial fibrillation: Recommendations".)

●Catheter ablation is the primary nonpharmacologic approach to prevent recurrent AF. Surgical-based ablation techniques such as the Maze procedure are generally reserved for PAF patients undergoing other cardiac surgical procedures. . (See "Atrial fibrillation: Catheter ablation" and "Atrial fibrillation: Surgical ablation".)

Rate control — Daily use of a rate-controlling agent such as a beta blocker, calcium channel blocker, or digoxin is generally not needed for PAF. However, for patients with highly symptomatic episodes, such drugs can be used for acute episodes, and some patients are maintained on one of these drugs to control the ventricular rate when PAF occurs. (See "Control of ventricular rate in patients with atrial fibrillation who do not have heart failure: Pharmacologic therapy".)

Anticoagulation — For patients with PAF, the approach for deciding whether to anticoagulate to reduce the risk of thromboembolism is similar to that for patients with persistent or permanent AF. The burden of AF (duration and frequency of episodes) is a factor for decision-making only for selected patients in whom the balance of benefit versus risk of anticoagulation is uncertain, recognizing that it may not be possible to accurately estimate AF burden, as discussed separately. (See "Atrial fibrillation in adults: Selection of candidates for anticoagulation", section on 'Paroxysmal AF'.)

Sinus node dysfunction — Patients with frequent AF episodes complicated by long, symptomatic post-conversion pauses due to sinus dysfunction are commonly treated with a pacemaker (along with rate or rhythm control) (see "Sinus node dysfunction: Treatment", section on 'Permanent pacing'). A possible alternative option for selected patients with PAF and minimally symptomatic sinus node dysfunction with relatively short pauses (eg, 2 to 3 seconds) is catheter ablation to minimize AF episodes and associated post-conversion pauses. However, since ablation is not a cure for AF, a pacemaker may still be required if there are symptomatic pauses. (See "Sinus node dysfunction: Treatment", section on 'Catheter ablation'.)

The management of sinus node dysfunction is discussed further separately. (See "Sinus node dysfunction: Treatment", section on 'Treatment'.)

Management of heart failure — Management of heart failure in patients with AF is discussed separately. (See "The management of atrial fibrillation in patients with heart failure".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Atrial fibrillation" and "Society guideline links: Arrhythmias in adults".)

SUMMARY AND RECOMMENDATIONS

●Definition – Paroxysmal atrial fibrillation (PAF; also known as intermittent AF) is defined as AF that terminates spontaneously or with intervention within seven days of onset. (See 'Definition and prevalence' above.)

●Precipitants – Factors that precipitate PAF are incompletely understood but are thought to be premature atrial complexes (PACs) and alterations in autonomic nervous system activity. (See 'Pathogenesis' above.)

●Symptoms – Like sustained AF, PAF may or may not be accompanied by symptoms, and the spectrum of symptoms is broad. The most common complaints include palpitations, often accompanied by dyspnea, a sensation of lightheadedness, fatigue, weakness, or generalized malaise. PAF may precipitate angina and/or ischemic electrocardiogram changes. PAF rarely causes syncope. (See 'Clinical presentation' above.)

●Recurrence – Most patients with PAF have one or more recurrences in the next year. Over one-third of patients with PAF may progress to persistent AF in 10 years, but the risk of progression increases with older age and presence of cardiovascular conditions such as rheumatic mitral stenosis and hypertension. (See 'Natural history' above.)

●Thromboembolic risk – Studies suggest that the risk of thromboembolic events in patients with PAF is lower than the risk in patients with persistent or permanent AF, is associated with AF burden (ie, percentage of time in AF), and is highest in the days soon after an episode of AF. However, the available data do not provide a clear threshold of AF burden or duration for elevated thromboembolic risk. The data suggest that AF episodes lasting longer than 17 to 24 hours impart significant risk, but the risk associated with shorter AF episodes is uncertain. (See 'Risk of embolization' above.)

●Management – The management of the arrhythmia in patients with PAF is similar to that for the general population of patients with AF. Important considerations are the duration of AF and the presence or absence of symptoms during episodes. (See 'Management' above and "Atrial fibrillation: Overview and management of new-onset atrial fibrillation".)

●Anticoagulation – For patients with PAF, the approach for deciding whether to anticoagulate to reduce the risk of thromboembolism is similar to that for patients with persistent or permanent AF. (See "Atrial fibrillation in adults: Selection of candidates for anticoagulation", section on 'Paroxysmal AF'.)

The burden of AF (duration and frequency of episodes) is a factor for decision-making only for selected patients in whom the balance of benefit versus risk of anticoagulation is uncertain, recognizing that it may not be possible to accurately estimate AF burden, as discussed separately. (See 'Anticoagulation' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff thank Alan Cheng, MD, and Philip J Podrid, MD, FACC, who contributed to earlier versions of this topic review.