Abstract 2054: N-Terminal Pro-B-Type Natriuretic Peptide Levels Predict Outcome From Cardiac Surgery: A Prospective Cohort Study.
Background: BNP and NT-proBNP are predictors of cardiovascular outcome in many settings. There are, however, limited data in patients undergoing cardiac surgery. We hypothesized that they would convey useful prognostic data in this setting.
Methods: Eight hundred ninety-two patients undergoing non-emergency cardiac surgery were recruited prospectively. Clinical details and pre-operative blood samples were obtained. The Parsonnet score and EuroSCORE were calculated. NT-proBNP was measured using the Roche Elecsys assay. The primary end-point was all-cause mortality.
Results: During a median of 1.3 years, 45 patients (5%) died. Median NT-proBNP levels were 894pg/mL in patients who died compared to 274pg/mL in survivors (hazard ratio [HR] 1.02 per 200pg/mL, 95% CI 1.01–1.03, p=0.001). NT-proBNP remained an independent predictor of outcome in multivariable models including either the Parsonnet score or EuroSCORE (HR 1.02 per 200pg/mL for both, p=0.007 and p=0.01 respectively). In patients who died within 30-days (n=26) NT-proBNP predicted death with an odds ratio of 1.03 per 200pg/mL (p=0.001) and was again independent of EuroSCORE and Parsonnet score (p=0.005 and 0.009 in models with each of these respectively). NT-proBNP predicted all-cause mortality with an area under the receiver operating characteristic curve of 0.69 (95% CI 0.61– 0.78, p<0.001). The optimal cut-off was 535pg/mL (figure⇓), with a sensitivity of 75% and a specificity of 67%.
Conclusion: NT-proBNP is a powerful predictor of mortality after cardiac surgery, both at 30 days and during medium-term follow-up. It has prognostic utility that is independent of traditional methods of risk stratification.