Background: We tested our hypothesis that serum BNP levels rise in sepsis and septic shock patients as a result of an inflammatory state and not only because of left ventricular dysfunction.
Methods: Twenty-one patients with sepsis or septic shock were enrolled in the study. Echocardiography was performed in every patient on admission and at discharge. Laboratory data were evaluated on admission, during hospitalization, and at discharge. Serum IL-1beta, IL-6, TNFalpha, and BNP concentrations were determined.
Results: BNP values on admission (r=0.47, p=0.03), during hospitalization (r=0.64, p=0.014), and on the day of discharge (r=0.54, p=0.015) were all positively correlated with CRP values. Mean BNP (r=0.07, p=0.006) and BNP level at discharge (r=0.68, p=0.001) were also positively associated with IL-1 at discharge. Mean CRP (17.7 mg/dL+/-1.5 vs. 9.2 mg/dL+/-3.6, p=0.002), IL-6 (46.6 pg/mL+/-2.2 vs. 25.6 pg/mL+/-16.3, p=0.003), and SAPS II levels (41.3+/-4.7 vs. 33.9+/-6.5 p=0.01) were also higher in patients who died versus those who survived. No difference in BNP levels was recorded in subjects who died versus those who survived. There was no clinical or echocardiographic evidence of left ventricular systolic dysfunction (mean EF% on admission 55.1+/-21.7 vs. 61.3+/-8.6 on discharge, p=0.123). Serum BNP levels at discharge were inversely associated with EF values on admission (r=-0.475, p=0.046) and positively associated with E/A ratio on admission (r=0.565, p=0.028). No association was found between BNP values and death.
Conclusion: BNP is positively correlated with CRP levels in septic patients without clinical or echocardiographic evidence of systolic dysfunction. No association was found between death and BNP values. It seems that, in septic patients, BNP is less accurate as a measure of ventricular dysfunction.