Controversies in postoperative atrial fibrillation after noncardiothoracic surgery: clinical and research implications

Abstract

Post operative atrial fibrillation after non‐cardiothoracic surgery is an increasingly common problem. There is little high quality data to guide clinicians in risk prediction as well as short term and long term management. There appears to be a significant risk of recurrent atrial fibrillation and stroke but effective screening strategies have not been tested. In this commentary, we delineate various controversies in the management of post operative atrial fibrillation and briefly review the available evidence. Prospective studies in this clinically important area of cardiology should be encouraged.

1. INTRODUCTION

Atrial fibrillation (AF) affects >2.5 million Americans and is often diagnosed after both cardiac and noncardiac surgeries. Postoperative atrial fibrillation (POAF) is a common complication after cardiothoracic surgery (CTS), with a reported incidence of 30% to 50% and increases mortality, length of hospital stay, and cost of care.1 Prevention and management of POAF after cardiac surgery is guided by the results of several randomized controlled trials.

On the other hand, POAF after noncardiac and nonthoracic surgery (NCTS) is less well studied. Depending on the type of surgery, the incidence varies widely, but it is approximately 3% among unselected patients.2 Given the fact that about 50 million such surgeries are performed every year in the United States alone, this represents an important problem.2 Unlike CTS, knowledge of POAF after NCTS largely relies on retrospective studies and analysis of administrative datasets. Though recent studies have highlighted its clinical implications, there are no high quality prospective studies to direct clinical management of these patients.

2. POAF OCCURRING AFTER CTS AND NCTS

In this article, we compare and contrast the differences between POAF occurring after CTS and NCTS and discuss controversies in management.

2.1. Is the pathophysiology of POAF different after NCTS compared with CTS?

After CTS, it is intuitive that direct handling of the heart (which is likely diseased), elevated cardiac pressures, and/or local inflammation can precipitate POAF.1 As inflammation and local injury subside, POAF often spontaneously resolves. After NCTS, the precipitants of POAF are often more systemic in nature. Some of the known precipitants for POAF after noncardiac surgery are shown in the Table 1.2, 3, 4 Those who develop POAF probably already possess the substrate for AF (atrial fibrosis); enhanced adrenergic tone, along with hemodynamic stressors may precipitate the arrhythmia. Unlike CTS, where risk scores have been validated (eg, the POAF score), there is no established risk score for prediction of POAF after NCTS.5 Similar to CTS, many patients spontaneously achieve sinus rhythm even without any cardiac medications.6

Table 1.

Risk factors for POAF after noncardiac and nonthoracic surgery

Patient‐related factors
Age
Caucasian race2
Male sex3
Preexisting cardiac conditions (HTN, CHF) and cardiac structural abnormalities2
Elevated BNP or NT‐proBNP4
Surgical factors
Emergency surgery
General anesthesia
Site of surgery (abdominal, intracranial)2
Intraoperative hypotension
Inflammatory response (local and systemic)
Postoperative factors
Hypoxia
Positive fluid balance
Pain and nausea
Electrolyte abnormalities
Infection

Abbreviations: BNP, brain natriuretic peptide; CHF, congestive heart failure; HTN, hypertension; NT‐proBNP, N‐terminal probrain natriuretic peptide; POAF, postoperative atrial fibrillation.

2.2. Does POAF impact short‐term outcomes after NCTS?

Evidence suggests that POAF after NCTS is associated with a higher risk of postoperative complications such as bacterial pneumonia and congestive heart failure and leads to increased hospital stay and postoperative mortality.2, 7 In the Perioperative Ischemic Evaluation (POISE) study of perioperative β‐blockade use in NCTS, POAF was associated with an increased risk of stroke at 30 days.8 Based on the factors delineated in table 1, POAF is as much a marker of a sicker patient and a more complicated procedure as it is a direct cause of perioperative morbidity by itself.

2.3. Are there effective preventive strategies for POAF after NCTS?

Studies have been conducted on the role of various medications to prevent POAF after CTS. β‐Blockers, antiarrhythmics (such as oral amiodarone), statins, and electrolyte supplementation (such as magnesium) have been shown to prevent POAF after CTS.9

On the other hand, there is no evidence that these agents can effectively reduce POAF after NCTS. The best evidence comes from the POISE trial, where initiation of a long‐acting formulation of metoprolol before surgery reduced the incidence of POAF compared with placebo (2.2% vs 2.9%; hazard ratio [HR]: 0.75, 95% confidence interval [CI]: 0.58‐0.99).8 However, the dosing strategy used in POISE was associated with an increased incidence of stroke and mortality (likely due to postoperative hypotension), and thus metoprolol or other β‐blockers cannot be recommended for reducing the incidence of POAF after NCTS. In addition, the number needed to treat to see any benefit was large (NNT = 143); thus a better risk‐stratification method is required to guide drug therapy in this setting.

2.4. What is the best strategy for acute management of POAF after NCTS?

In general, the management of POAF after NCTS should be guided by the clinical scenario. Patients with hemodynamically significant POAF should be emergently cardioverted. In the symptomatic but hemodynamically stable patient, either intravenous rate‐ or rhythm‐control strategy can be employed. In patients with mild or no symptoms, oral rate‐control medications can be used.

POAF after CTS is often managed with the use of antiarrhythmic agents, such as amiodarone, despite lack of clear evidence that antiarrhythmic drug therapy is superior to a strategy of rate control alone.10 In the large Cardiothoracic Surgical Trials Network (CTSN) trial in CTS patients with POAF, treatment with amiodarone was not superior to rate control alone in improving mortality or freedom from AF from hospital discharge to 60 days (86.9% vs 84.2%; P = 0.41). However, there was a significant 25% crossover rate, mainly because of drug ineffectiveness (in the rate‐control group) or side effects or adverse drug reactions (in the rhythm‐control group), suggesting that the treatment strategy has to be modified based on response. Another study showed no difference in the rate of persistence of POAF after CTS in patients treated with different duration of anti‐arrhythmic drugs used in patients undergoing coronary artery bypass grafting.11 In a smaller study of patients undergoing lung resection, there was no significant difference between amiodarone and diltiazem therapy with regard to conversion to sinus rhythm.12

In contrast, there are no studies to guide therapy for POAF after NCTS. In practice, POAF after NCTS is often diagnosed and managed differently than POAF after CTS. After NCTS, patients often recuperate in a general ward without continuous cardiac monitoring.13 Thus, the exact timing of onset of POAF is unclear and asymptomatic episodes may never be detected. After the diagnosis is made, nursing protocols often limit the type of intravenous cardiac medications that can be administered. A rhythm‐control strategy would thus necessitate transfer of patients to a higher level of care, and thus increase costs as well as expose patients to the risk of drug‐related side effects. A rate‐control strategy, on the other hand, may also involve intravenous doses of atrioventricular nodal blocking agents such as β‐blockers and calcium channel blockers, which could potentially reduce systemic blood pressure. Thus, a blanket rate‐ or rhythm‐control strategy may not be feasible in NCTS and treatment should be individualized.

The most commonly used antiarrhythmic in the postoperative setting is amiodarone, which is often delivered as intravenous infusion and whose side effects are well known. Despite years of experience, it is well known that antiarrhythmic medications can be proarrhythmic and also cause a variety of noncardiac side effects, some of which can be fatal.14 In instances where a rhythm‐control strategy is necessary postoperatively, the continued use of antiarrhythmic agents should be revisited as the patient recovers. The overall duration of antiarrhythmic therapy should be minimized.

2.5. Is anticoagulation indicated for POAF after NCTS?

Embolic stroke is the most dreaded complication of AF. Data from the POISE study indicate a high risk of stroke within the first 30 days (odds ratio: 3.51, 95% CI: 1.45‐8.52).8 However, it is unknown if the stroke was embolic or ischemic, especially in view of the fact that many patients treated with β‐blockers developed postoperative hypotension. Gialdini et al looked at the risk of stroke from POAF in a large number of patients and found 1.47% cumulative risk of stroke at 1 year after hospitalization in patients undergoing NCTS, compared with 0.36% in patients who did not develop POAF.15 POAF was associated with subsequent stroke both after CTS (HR: 1.3, 95% CI: 1.1‐1.6) and NCTS (HR: 2.0, 95% CI: 1.7‐2.3). In fact, the association was stronger for POAF after NCTS vs CTS (P < 0.001 for interaction).

For POAF after CTS, expert opinion currently suggests that POAF that is short‐lived can be managed without anticoagulation.16, 17 There are no guidelines regarding short‐term anticoagulation for POAF after NCTS. Similarly, there are no studies regarding the use of long‐term anticoagulation to reduce the risk of stroke in patients who develop POAF after NCTS. However, recent studies in patients with implanted cardiac devices suggest that even short‐lived episodes of AF (>5–6 minutes a day) can increase the risk of stroke.18, 19, 20, 21 Based on these observations, it may be prudent to prescribe short‐term anticoagulation (30 days) for POAF, even short‐lasting. However, until more evidence is available, it is difficult to justify long‐term anticoagulation unless recurrent episodes of AF are documented.

In this context, it is very important to understand that the risk of stroke after AF is not entirely related to the presence of atrial thrombus. Patients with POAF often have other risk factors that predispose to ischemic stroke. The CHA₂DS₂‐VASc score essentially captures risk factors for both ischemic and embolic stroke. In fact, AF may be an important marker of risk for stroke (noncausal relationship), in addition to being a more direct cause for stroke.22

We recommend that all patients with short‐lived POAF undergo a thorough risk assessment of vascular risk factors for stroke. Extrapolating from CTS guidelines, routine anticoagulation is not recommended unless the duration of AF is >48 hours and/or AF episodes are frequent. As noted above, because the onset of POAF can sometimes be tough to ascertain in the absence of continuous cardiac monitoring after NCTS, it is best to err on the side of anticoagulation, which should be continued for ≥1 month. Longer‐term anticoagulation may be considered when CHA₂DS₂‐VASc score is elevated (>2) and after an informed discussion with the patient regarding risks and benefits. Modification of stroke risk factors, such as control of hypertension and appropriate use of lipid‐modifying agents, is also important. A randomized controlled trial of anticoagulation in patients with POAF after NCTS (especially those with a high CHA₂DS₂‐VASc score) would be appropriate.

2.6. Should patients with POAF undergo routine long‐term monitoring to detect recurrent asymptomatic AF?

An important question relates to outpatient assessment of a patient who suffered a brief, self‐limited episode of POAF after CTS or NCTS. As noted above, there is no convincing evidence for anticoagulating all patients with short lived POAF during the hospital stay but recurrent atrial fibrillation in follow up would justify anticoagulation. The risk of recurrent AF after POAF was recently investigated by Gialdini et al in a large number of patients. After noncardiac surgery, cumulative 1‐year rates of AF (based on postdischarge encounters with recorded diagnosis) were 37.28% (95% CI: 36.43%‐38.15%) in those with POAF and 1.51% (95% CI: 1.49%‐1.52%) in those without POAF. In contrast, cumulative 1‐year rates of AF after CTS were 22.22% (95% CI: 21.47%‐22.99%) in those with POAF and 4.65% (95% CI: 4.48%‐4.82%) in those without.15

An optimal screening strategy for patients with POAF is unknown. Some would argue that patients with POAF probably had multiple subclinical episodes of AF before the perioperative period and likely afterwards as well so periodic screening may be justified. At a minimum, physical examination and periodic electrocardiography can be used to detect onset of recurrent AF in follow‐up. Periodic Holter monitoring is sometimes utilized in clinical practice. Future studies should assess the role of long‐term monitoring (eg, implantable loop recorders) in detecting recurrent AF in this cohort and its impact on clinical outcomes.

3. CONCLUSION

Despite the increasing NCTS being performed in an increasingly older and frail population who experience frequent episodes of POAF, there is little evidence to guide management of POAF. Current evidence actually suggests that POAF after NCTS is linked to a higher risk of recurrent AF and stroke during the follow‐up period, compared with POAF after CTS. A blanket rhythm‐control strategy is not supported by current evidence and exposes patients to unnecessary medication side effects. POAF after NCTS reflects the presence of an arrhythmic substrate and thus predicts an increased risk of recurrent AF. Prospective studies should be performed to guide long‐term screening and management of patients with POAF after NCTS.

CONFLICTS OF INTEREST

The authors declare no potential conflicts of interest.

Vallurupalli S, Shanbhag A and Mehta JL. Controversies in postoperative atrial fibrillation after noncardiothoracic surgery: clinical and research implications. Clin Cardiol. 2017;40:329–332. 10.1002/clc.22652