Evaluation of brain natriuretic peptide (BNP) role as prognostic factor for cardiovascular outcomes in patients undergoing non-cardiac surgery

Purpose: Cardiovascular events in patients undergoing non-cardiac surgery are higher than other complications; they are associated with high morbidity and mortality, and represent the first cause of death. Clinical evaluation itself is not sufficient to predict them, especially if patients show heart failure experiencing none or few symptoms. The chance of having a simple, fast, and not invasive blood test such as BNP could be an optimal approach to cardiovascular pre-surgery risk stratification. Our aim was to assess the predictive BNP power for adverse cardiac events after non-cardiac surgery. Methods: 205 patients admitted in a general surgery department for major, minor non-cardiac surgery have been enrolled (M/F 91/114, mean age 64±15 yrs). Preoperative data included: patient demographics, vital signs, routine blood samples, plasma BNP before and after surgery (BNP Triage, Biosite), 12-lead ECG, cardiovascular risk evaluation with the Revised Cardiac Risk Index (RCRI) score. All patients were reached by phone 90 days after discharge to investigate the onset of any cardiovascular event. Results: BNP values [median (IQR)] at admission were 28.5 (10.5-90)pg/ml, and at discharge 76.0 (30.2-170.7)pg/ml significantly higher (P<0.0001). Post-discharge cardiac adverse events occurred in 28 patients. Both BNP levels at admission (118 (46-403)pg/ml) and at discharge (211 (82.3-529)pg/ml) were significantly higher (p<0.0001) in patients undergoing post-discharge cardiac events compared to those without events. BNP at admission was significantly correlated to the RCRI index (median 42.0 (16-132)pg/ml for RCRI 2-3, p<0.001). ROC curves for admission and discharge BNP, and cardiac adverse events at 90 days have, respectively, an AUC=0.80, p<0.0001, and an AUC=0.76, p<0.0002. Both admission and discharge BNP levels were more accurate in predicting post-discharge cardiac events than RCRI index (ROC curve for RCRI index AUC=0.57, p=not significant). Multivariate analysis showed that BNP median level >29 pg/ml at admission was the only factor independently correlated with cardiovascular adverse event (p< 0.001, OR=9,4883). Conclusions: Our data demonstrate the strong power of BNP, in non-cardiac surgery, in predicting post-discharge adverse cardiac outcomes, higher than RCRI index, and correlated with anamnestic data of ischemic cardiovascular diseases. Pre- and postoperative BNP have a similar predictive power for cardiovascular adverse outcomes. Using BNP in a non-cardiac surgery setting is useful to identify patients at higher risk for cardiovascular events susceptible for a more accurate therapeutic control.