Conference: 2011 PH Professional Network Symposium
Release Date: 09.22.2011
Presentation Type: Abstracts
F. Rahaghi, MD, R. Tofts, MD, H. Zaitoun, MD, R. Schwartz, MD, M. Hernandez, MD, V. Ali, MD, V. Navas, MD, N. Rahaghi, MD, J. Ramirez, MD, L. Smolley, MD, E. Oliveira, MD, M. Areces, MD, C. Scridon, MD, K. Fromkin, MD, G. Novaro, MD, R. Benavides, LPN
Cleveland Clinic Florida - Weston/ Ft. Lauderdale, FL/US
INTRODUCTION: Pulmonary hypertension (PH) secondary to left heart failure (WHO group II) is the most common cause of PH. Right heart catheterization (RHC) is the gold standard in diagnosis. Pulmonary capillary Wedge Pressure ≤15mmHg (PCWP) distinguishes PH due to pulmonary artery hypertension (PAH or non-PVH, WHO Group I) and PH due to Left ventricular disease, or pulmonary venous hypertension (PVH, WHO Group II) with PCWP >15 mmHg. N-terminal pro-brain natriureteic peptide (NT-proBNP) is a reliable indicator of right and left ventricular overload. We hypothesized that the left ventricle will consistently release more NT-proBNP than the right and this difference can predict the etiology of PH. Our objective was to investigate NT-proBNP as a noninvasive guide to identify the etiology in patients with PH.
METHODS: A total of 70 patients underwent RHC and serum measurement of NT-proBNP. Demographic information was collected. All patients with Chronic Kidney Disease (CKD) stage IV (GFR <30) or severe left-sided valvular disease were excluded. The NT-proBNP and PCWP were plotted along receiver operator characteristic (ROC) curves for wedge of ≤15 mmHg and >15 mmHg. Multivariate regression analyses were performed to establish the highest sensitivity and specificity for a given combination of variables.
RESULTS: The study group comprised 57 patients, 22 (39%) men, with an average age of 69 +/- 13 years. Thirty six (63%) had confirmed PH with PCWP <15mmHg (non-PVH), while 21 (37%) had confirmed PH with PCWP >15 mmHg (PVH)). Sixteen (28%) had COPD, and 15 (26%) had collagen vascular disease. The NT-proBNP differed significantly between non-PVH and PVH groups. A NT-proBNP <2500 diagnosed non-PVH with a sensitivity of 94.4%, specificity of 61.09% (OR 27.63, p <0.001.) The ROC curve for this NT-proBNP is 0.787, positive predictive value (PPV) 80.9%, and negative predictive value (NPV) of 86.7%. The echocardiogram improved the predictive model. Patients with NT-ProBNP <2500 without RV dilation had a significantly higher chance of being diagnosed non-PVH versus PVH (OR = 38.50, p <0.001, sensitivity 97.22%, specificity 52.38%, PPV 77.78%, and NPV 91.67%).
LIMITATIONS: Since this study was based on ROC curves based on a smaller academic hospital, the results may neither be exactly extrapolated to larger academic centers or private hospitals. Similarly they may reflect the referral biases of the geographical area and be affected by the pulmonary/ critical care based nature of PAH clinic of the Cleveland Clinic Florida.
CONCLUSIONS: Our study shows NT-proBNP of 2500 reliably distinguishes PH with elevated wedge pressures. The echocardiogram strengthens this predictive model. In certain high-risk patients where right heart catheterization may not be possible, this is a useful tool.