Abstract 12217: Accurate Prediction of Systolic Peak-to-Peak Gradient in Children With Pulmonary Valve Stenosis

Objective: To evaluate the correlation between transthoracic echocardiography (TTE) derived gradients and invasive RV-PA peak-to-peak systolic gradient (iPG) and the predictability of these values in children with pulmonary valve stenosis (PVS). Introduction: Historically, correlations have been found between the Doppler-derived peak echo gradient (ePG), using the modified Bernoulli equation and invasive hemodynamic indices. This has led to the ePG being a widely used estimate of the degree of stenosis and an indicator of timing to consider intervention in children with PVS. However, large disparities observed between these invasive and non-invasive measurements have called into question the reliability of ePG to screen patients for consideration of pulmonary valvoplasty. Methods: We performed a retrospective chart review of 119 children who underwent pulmonary valvoplasty between January 2010 and January 2020 and selected 51 patients who had "simple" PVS in whom a TTE was performed within 6 months prior to the procedure. We evaluated the correlation between TTE parameters [Maximum velocity (Vmax), Mean Gradient (MG), ePG], and iPG. We then used the mathematical functions resulting from the correlation models to predict iPG. Results: All the assessed models were statistically significant and presented a strong correlation. When the differences between echocardiographic parameters and iPG were evaluated, ePG demonstrated the least accuracy in predicting iPG, erroneously estimating the real value by >10mm Hg in 66% of cases. The MG presented a more accurate estimation of iPG, showing a narrower range of variance, with an error of >10mm Hg in less than 30% of cases. We then modeled several mathematical functions of Vmax, MG, and ePG to predict iPG, and found that a derived function of Vmax ( f iPG) distributed by BSA (fiPG =12.7[Vmax]-17.3[BSA]) produced the most accurate results. Conclusions: There is a strong correlation between pre-procedural TTE parameters and iPG, but when ePG is used to estimate iPG, up to two-thirds of cases could be inappropriately characterized. We found that MG was a better predictor and that a mathematical function derived from Vmax and BSA ( f iPG) was the most accurate predictor of iPG.