P0387 - Echocardiographic Evaluation of Diastolic Function Can Accurately Predict Simultaneous Cardiac Catheterization Derived Gold Standard Measurements of Diastolic Function After Fluid Bolus

Background: Pediatric echocardiographic studies attempting to predict gold standard cardiac catheterization measures of left ventricular diastolic function (LVDFx) (pulmonary capillary wedge pressure (PCWP) or left ventricular end diastolic pressure (LVEDP)) are often retrospective, evaluated under variable hemodynamic and hydration conditions, and have demonstrated inconsistent results. We are unaware of any pediatric studies comparing echocardiographic measures of LVDFx with PCWP or LVEDP under the same loading conditions. We hypothesize, standard echocardiographic measurements of LVDFx can predict abnormal cardiac catheterization derived LVEDP or PCWP, when obtained under similar loading conditions and following a fluid bolus to optimize fluid status (or potentially unmask abnormal LVDFx).

Methods: We performed a prospective cohort study using a convenience sample of pediatric heart transplant recipients undergoing routinely scheduled cardiac catheterization. Echocardiographic measures of LVDFx were performed with simultaneous catheterization measures of PCWP or LVEDP. Echocardiographic evaluation included volume dependent measures of mitral valve inflow E and A velocities and volume independent tissue doppler indices (TDI) of septal and lateral wall active (E’) and passive (A’) relaxation. Measurements were repeated following a 10 mL/kg (maximum 500mL) normal saline bolus given over 5 minutes. Abnormal LVDFx was defined as a post-bolus PCWP or LVEDP greater than or equal to 15mmHg.

Results: Nineteen patients underwent cardiac catheterization. Three patients were excluded from a bolus due to left heart obstruction, graft anastomosis obstruction, and initial elevation in LVEDP. Median study age was 9.2 years (IQR 3.1-16.4). Median pre-bolus PCWP (n=18) and LVEDP (n=12) were normal with values of 11.5mmHg (IQR 10-12) and 10.0mmHg (IQR 7.75-10.5), respectively. Median post-bolus PCWP (n=16) and LVEDP (n=8) were 15.5mmHg (IQR 12.75-16.25) and 13mmHg (IQR 10-17.5) respectively. Post bolus PCWP and LVEDP were significantly higher than pre-bolus values (PCWP p = 0.0004 per Wilcoxon Rank Sum; LVEDP p = 0.03 per Mann-Whitney U). Only mitral valve inflow velocity (E) and its ratio to septal wall velocity (E’) increased significantly post-bolus. Six echocardiographic parameters held acceptable areas under ROC curves in predicting abnormal LVDFx. Pre- and post-bolus septal and lateral E/E’ were associated with increasing levels of cardiac catheterization defined left ventricular diastolic dysfunction. Pre-bolus septal E/E’ was highly associated with a post-bolus LVEDP or PCWP greater than 15mmHg, (area under the ROC curve, 0.90). Values greater than 8.12 were associated with 90% sensitivity and 83% specificity.

Conclusion: In this cohort, only volume dependent echocardiographic parameters changed significantly following a fluid bolus, supporting the assertion that TDI measures of diastolic function are volume independent. We demonstrated a combination of echocardiographic measures of diastolic function (septal and lateral E/E’) were highly associated with catheterization derived measures of diastolic dysfunction. Additionally, the fluid bolus unmasked previously unrecognized abnormal left ventricular compliance in a majority of patients. Enrollment is ongoing.