Contrast-enhanced ultrasound evaluation of renal perfusion before angioplasty and its predictive value for hypertension

Abstract

BACKGROUND:

Atherosclerotic renal artery stenosis (ARAS) is a common disease in the elderly population.

OBJECTIVE:

The aim was to develop a contrast-enhanced ultrasound (CEUS)-based model for predicting post-angioplasty improvement in hypertension in patients with severe ARAS.

METHODS:

Thirty-five patients with severe ARAS ( $⩾$ 70%) were included in this study, and 42 renal arteries received percutaneous transluminal renal arterial stenting. An optimal integral formula was developed from pre-interventional color-coded duplex sonography (CCDS) and CEUS parameters using least absolute shrinkage and selection operator (LASSO) regression and receiver operating characteristic (ROC) curve analysis. A model for predicting short-term hypertension improvement was established using the integral formula and clinical risk factors. Bootstrapping was used for internal validation.

RESULTS:

Two integral formulas, LASSO.CCDS and LASSO.CEUS, were established. ROC curves of the two integral formulas showed that LASSO.CEUS was the better formula for predicting hypertension improvement (AUC 0.816, specificity 78.6%). Univariate and multivariate regression analyses showed that duration of hypertension (OR 0.841, $P=$ 0.027), diabetes (OR $=$ 0.019, $P=$ 0.010), and LASSO.CEUS (OR 7.641, $P=$ 0.052) were predictors of short-term hypertension improvement after interventional therapy. Using LASSO.CEUS combined with clinical risk factors, the following prediction model was established: logit (short-term improvement in hypertension) $=$ 1.879–0.173 $\times$ hypertension duration – 3.961 $\times$ diabetes $+$ 2.034 $\times$ LASSO.CEUS (AUC 0.939).

CONCLUSIONS:

The model established using CEUS parameters and clinical risk factors could predict hypertension improvement after interventional therapy, but further research and verification are needed.