The Role of Femoral Artery Ultrasound Measurements in Predicting Restenosis following Endovascular Intervention

Abstract

Background

Although endovascular interventions for peripheral artery disease (PAD) have become increasingly common, they are associated with a significant rate of restenosis. An improved understanding of the risk factors for restenosis is needed to guide the management of patients with PAD.

Methods

We performed a cross-sectional analysis of 101 patients with PAD, to determine the association between clinical predictors, novel imaging measures of vascular function of the femoral artery [intima-media thickness, diameter, and intima-media cross-sectional area, change in diameter (delta diameter), average expansion and average contraction], and risk of reintervention. Univariable and multivariable associations between clinical and imaging measures and the need for reintervention were assessed using logistic regression. The predictive accuracy of the selected regression model and added predictive value of imaging measures were assessed using the area under the Receiver operating characteristic curve (AUC). The added predictive value of ultrasound was assessed by comparing AUC of the selected model with clinical parameters alone with that of the same model with additional ultrasound parameters.

Results

Of the 101 patients with PAD, 20 underwent recent multiple reinterventions. Vessel diameter was statistically significantly smaller in the group who underwent multiple reinterventions, p = 0.0170. Univariate logistic regression revealed that smoking status, hypertension, hypercholesterolemia, and chronic renal insufficiency were clinical parameters that met the statistical cut point of p ≤ 0.20. In the multivariable model, chronic renal insufficiency status (odds ratio (OR) = 8.27, 1.17–58.25), hypertension (OR = 0.145, 0.020–1.062), and femoral artery diameter (OR = 0.375, 0.136–1.031) remained important predictors for reintervention. The AUC for the clinical multivariable logistic regression model was 0.7481 while that for the model additionally including the ultrasound parameters was 0.8325 (p = 0.044).

Conclusions

Including ultrasound parameters in the risk prediction model for restenosis improved the ability to predict restenosis in this group of patients with known PAD. Further study is needed to define its utility in preoperative assessment and risk stratification before revascularization.