Society of Cardiovascular Anesthesiologists

Plasma B-type natriuretic peptide levels predict postoperative atrial fibrillation in patients undergoing cardiac surgery

Wazni O, Martin D, Marrouche N, et al. Circulation 2004; 110:124-7

Reviewer: Mohammed M. Minhaj, MD
Cardiothoracic Anesthesia Fellow
University of Chicago
Chicago, IL

Abstract Excerpt: Postoperative atrial fibrillation (AF) is a known potential complication in patients after cardiac surgery. Recent reports have linked elevated levels of B-type natriuretic peptide (BNP) as being predictive for postoperative AF. This study aimed to validate this potential link. One hundred, eight-seven patients with no previous history of atrial arrhythmia who had preoperative BNP levels obtained and had undergone cardiac surgery were identified and their records reviewed retrospectively to identify the occurrence of atrial fibrillation. Postoperative ECG strips and telemetry strips were analyzed for AF during the entire postoperative period until the time of discharge. Patients who had documented postoperative AF were significantly older than their non-AF counterparts (68±11 vs. 64±14 years, P=0.04), but there was no difference with respect to sex distribution, hypertension, left ventricular (LV) dysfunction, LV hypertrophy (LVH), left atrial size, history of coronary artery disease (CAD), or use of beta-blockers. Preoperative levels of BNP were significantly higher in the postoperative AF patients (615 vs. 444 pg/mL, P=0.005). Furthermore, after a multivariable logistic regression model that adjusted for sex, age, type of surgery (CABG only vs. valvular surgery), hypertension, LV function, LVH, left atrial size, presence of CAD, use of beta-blockers, and plasma BNP, only a BNP level in the upper 50th percentile was associated with a higher risk of postoperative AF. These investigators concluded that an elevated preoperative BNP level is a strong, independent predictor of postoperative AF.

Comments: Postoperative AF is a known and potentially deleterious complication of cardiac surgery. Some studies have estimated that it may occur in as many as 60% of patients after cardiac surgery. While many cases are self-limiting, AF may increase the incidence of postoperative stroke, prolong hospital stay, and result in increased costs. Previously identified risk factors have included older age, low ejection fraction, and type of surgery, though older age is the only factor that has consistently demonstrated an association with postoperative AF.

Another possible predictive factor of AF was also identified in this study: BNP levels. Inherently, this seems to be quite plausible. Stretch of the atria in disease processes such as mitral stenosis has been associated with the onset of AF. This same stretch of the atria is what causes the release of natriuretic peptide. It should follow, therefore, that elevated levels of BNP could be associated with the onset of postoperative AF. Previous studies have demonstrated elevated BNP levels in patients with solitary AF and idiopathic bilateral atrial dilatation. Additionally, BNP has been shown to predict reoccurrence of AF in patients with mild heart failure after cardioversion. Interestingly enough, while this study found a significant difference in BNP levels between the two groups, there was no significant difference in left atrial size. A key finding of this investigation was the increased odds ratio for AF with rising BNP levels (as represented by the multivariable regression analysis after adjustment for several factors, including age). Besides BNP levels and age, valvular surgery was the only other characteristic found to be significantly different between the two groups (75% of postoperative AF patients had valvular surgery whereas only 52% of patients without AF had valvular surgery, P=0.03).

All retrospective reviews such as this have obvious limitations. Although reviewers were blinded as to preoperative BNP values, the presence of such values in the chart could have biased reviewers as to the incidence of postoperative complications. Additionally, other factors that have been previously linked to postoperative AF, such as thyroid disease and postoperative management differences, were not analyzed and could have affected the results.

Overall, given the serious implications of the development of postoperative AF, the identification of potential risk factors is important and this study is a step in that direction. There still exists a need for well-designed, prospective studies, not only to validate these results, but also to implement and gauge the efficacy of prophylactic therapy in patients with elevated BNP levels, with the goal of reducing further complications and health care costs associated with cardiac surgery.

Table of Contents:

PRO/CON: Transesophageal Echocardiography (TEE) should be used routinely in all high risk noncardiac surgery

Pro - Feroze Mahmood, MD
Con - John E. Ellis, MD

Literature Reviews
- Mitral valve surgery for chronic ischemic mitral regurgitation (Mark A. Chaney, MD)
- Plasma B-type natriuretic peptide levels predict postoperative atrial fibrillation in patients undergoing cardiac surgery (Mohammed M. Minhaj, MD)
- Severity of illness and risk of death associated with pulmonary artery catheter use (Michael H. Wall, MD)
- Cardiac replacement with a total artificial heart as a bridge to transplantation (K.W. Tim Park, MD)