Increased Central Venous Pressure Is Associated with Impaired and Worsening Renal Function in Children with Cardiovascular Disease

Background: Low cardiac output leading to poor renal perfusion was previously considered the primary hemodynamic determinant of renal function and injury. However, growing evidence in adults with cardiovascular disease suggests that increased central venous pressure (CVP) is the main driver of worsening renal function (WRF). We sought to describe the relationship between CVP and estimated glomerular filtration rate (eGFR) in children with cardiovascular disease and to determine whether increased CVP is associated with WRF.

Methods: We performed a single center cohort study of patients aged 3 months to 21 years with biventricular circulation who underwent cardiac catheterization between 01/2010 and 08/2016. Patients were included at their first catheterization unless they were transplanted during the study interval, in which case the first post-transplant catheterization was also included. Patients with intracardiac shunting, with pre-existing chronic kidney disease, and those on inotropes or vasopressors were excluded. Data collected from each catheterization included superior vena cava pressure (CVP) and cardiac index (CI) as calculated by the Fick principle. The modified Schwartz equation with serum creatinine drawn within 72 hours prior to catheterization was used to calculate eGFR. WRF was defined as an increase in creatinine of at least 0.3 mg/dL within 1 year following catheterization. We performed univariable and multivariable linear regression and Cox regression to determine the association of CVP with eGFR and WRF.

Results: Data from 335 catheterizations performed on 331 patients were analyzed. The most common indications for catheterization included transplant surveillance (41%), balloon angioplasty/stenting (12%), patent ductus arteriosus occlusion (8%), pre-transplant hemodynamics (7%), and pulmonary hypertension study (6%). Median age at catheterization was 7.1 years (IQR 1.7-14.3 years) and 57% of patients were male. Median CVP was 7 mmHg (IQR, 5-9 mmHg), median AVO2 difference was 26% (IQR, 22-32%), and median CI was 3.5 L/min/m2 (IQR, 2.7-4.5 L/min/m2). The primary dependent variables of interest were eGFR and WRF. The median eGFR at catheterization was 96.8 mL/min/1.73m2 (IQR, 80.5-118.8 mL/min/1.73m2); 72 patients (21.8%) developed WRF. In unadjusted analyses, higher CVP (beta=-2.340; 95% CI -3.168 to -1.513; p<0.001) and AVO2 difference (beta=-0.527; 95% CI -0.910 to -0.143; p=0.007) were associated with lower eGFR. Age, sex, weight, and CI were not associated with eGFR. In adjusted analyses, higher CVP was independently associated with eGFR (beta=-2.189; 95% CI -3.104 to -1.274; p<0.001) when controlling for age, sex, AVO2 difference, and CI. Of these same variables, CVP was the only independent predictor of WRF in multivariate analysis (aHR 1.082 (95% CI 1.027-1.140)).

Conclusion: Among children with cardiovascular disease, higher CVP is associated with lower eGFR and increased risk of developing WRF.