Kraisangka, J., Scott, J., Kanwar, M., Liu, Z., Lin, S., Everett, A., Perer, A., Movahedi, F., Antaki, J., & Benza, R. “LAB parameters in predicting survival in Pulmonary arterial hypertension”. Chest, vol. 160, no. 4, Oct. 2021, p. A2335.
PURPOSE: Lab values are essential sets of parameters for the prognosis of the patients with pulmonary arterial hypertension (PAH).
METHODS: We analyzed the lab variables obtained at baseline in 7 PAH clinical trials: SERAPHIN, BREATHE, CHEST-2, PATENT-2, FREEDOM, TRIUMPH, and the subcutaneous treprostinil study. The outcome of the analysis was mortality at a time to event of each individual trial. We selected only the significant Lab variables (p-value<0.05) to be aggregated across the clinical trials using Stouffer’s method, weighed by their ‘N’ in the clinical trials.
RESULTS: Collective data from 3,387 patients were available at baseline with a total of 45 lab variables. Out of the 45, only 24 variables were still significant (p-value<0.05) after the aggregation of p-values. Elevated Uric acid, leukocyte/neutrophil count, red blood cell count, neutrophil/lymphocyte ratio, bilirubin (direct & total), AST, and BUN are the most predictive of adverse outcomes. Higher Albumin, lymphocyte count, and renal function (Cr, Creatine clearance) are the most predictive of improved outcome.
CONCLUSIONS: To our knowledge, this is the largest aggregated attempt to analyze lab parameters in predicting survival. This analysis will be used to select features for the derivation of machine-learned PAH patient risk stratification models
CLINICAL IMPLICATIONS: Our long-term goal is to generate a comprehensive personalized risk assessment tool from the data-driven variable selections to aid clinical decisions.