ABSTRACT
Purpose There is increasing evidence of adverse cardiovascular morbidity associated with SARS-CoV-2 (COVID-19). Pro-B-type natriuretic peptide (proBNP) is a biomarker of myocardial stress associated with outcomes in various respiratory and cardiac diseases. We hypothesized that proBNP level would be associated with mortality and clinical outcomes in hospitalized COVID-19 patients. Methods We performed a retrospective analysis of hospitalized COVID-19 patients (n=1232) using adjusted logistic and linear regression to assess the association of admission proBNP (analyzed by both categorical cutoff >125 pg/mL and continuous log transformed proBNP) with clinical outcomes. Covariates included age, sex, race, body mass index (BMI), hypertension, coronary artery disease (CAD), diabetes, smoking history, and chronic kidney disease stage (Model 1), with Troponin I added in Model 2. We performed survival analysis by a multivariate Cox proportional hazard model, incorporating log transformed proBNP. We additionally treated BMI, a strong potential confounder of both proBNP levels and COVID-19 outcomes, as an ordinal variable ordered across tertiles. Results Patients were mean age 62.9±17.6, 53.8% male, and 35.9% Black. Preadmission comorbidities were hypertension (57.1%), diabetes (31.6%), CAD (9.0%) and heart failure (HF, 10.6%). In Model 1 and 2, higher proBNP level was significantly associated with death, new HF, length of stay, ICU duration and need for ventilation among hospitalized COVID-19 patients. This significance persisted after ordinal compression of BMI across tertiles. The adjusted hazard ratio of death for log[proBNP] was 1.56 (95% CI: 1.23-1.97;P<0.0001). Conclusion Using a relatively large and racially diverse hospitalized COVID-19 patient cohort, we find that proBNP is associated with adverse clinical outcomes, including mortality and new HF in COVID-19. Further prospective investigation is warranted on the utility of proBNP for clinical prognostication in COVID-19.