Low Serum BNP Predicts Malignant Etiology of Pericardial Effusion

Ali Nassar Cardiovascular Institute, B Padeh Medical Center, Poriya, Israel Azrieli Faculty of Medicine in the Galilee, Bar Ilan University, Zefat, Israel Fabio Kusniec Cardiovascular Institute, B Padeh Medical Center, Poriya, Israel Azrieli Faculty of Medicine in the Galilee, Bar Ilan University, Zefat, Israel Diab Ghanim Cardiovascular Institute, B Padeh Medical Center, Poriya, Israel Azrieli Faculty of Medicine in the Galilee, Bar Ilan University, Zefat, Israel Wadi Kinany Cardiovascular Institute, B Padeh Medical Center, Poriya, Israel Azrieli Faculty of Medicine in the Galilee, Bar Ilan University, Zefat, Israel Doron Sudarsky Cardiovascular Institute, B Padeh Medical Center, Poriya, Israel Azrieli Faculty of Medicine in the Galilee, Bar Ilan University, Zefat, Israel Evgeni Hazanov Cardiovascular Institute, B Padeh Medical Center, Poriya, Israel Azrieli Faculty of Medicine in the Galilee, Bar Ilan University, Zefat, Israel Offer Amir Cardiovascular Institute, B Padeh Medical Center, Poriya, Israel Azrieli Faculty of Medicine in the Galilee, Bar Ilan University, Zefat, Israel Shemy Carasso Cardiovascular Institute, B Padeh Medical Center, Poriya, Israel Azrieli Faculty of Medicine in the Galilee, Bar Ilan University, Zefat, Israel

Background Serum brain natriuretic peptide (BNP) correlates with symptomatic heart failure, with different levels according to etiology and response to treatment. We aimed to evaluate its value in predicting the etiology of pericardial effusions by analyzing BNP levels in serum and pericardial fluid simmultaneously.

Methods 56 successive patients undergoing pericardiocentesis were recruited between 2/2014 and 12/2017. Serum and fluid BNP were tested in addition to standard blood and fluid test. Patients were divided into subgroups according to etiology.

Clinical, echocardiographic, blood and fluid characteristics were compared by ANOVA. Patients were than regrouped into malignant and inflammatory etiologies (excluding iatrogenic related effusions) to find correlates to malignancy related effusions by multivariable regression analysis.

Results Table demonstrates patient and fluid characteristics according to etiology. Pericardial fluid BNP level was indiscriminately high while low serum BNP was found to be the single significant correlate of malignant etiology (p=0.0025 for model). A cut-off point of ≤250ng/dL demonstrated an ROC of 0.67, 90% sensitive and 54% specific for malignancy.

Conclusion Low serum BNP was the single significant factor associated with malignancy in patients undergoing pericardiocentesis for hemodynamically significant pericardial effusions and therefore may guide the diagnostic work-up in these patients. Interestingly, the pericardial fluid BNP was uniformly high and did not contribute to the diagnosis of malignancy. Table