Analysis of Threshold Stenosis by Multiplanar Venogram and Intravascular Ultrasound for Predicting Clinical Improvement After Iliofemoral Vein Stenting: Results From the VIDIO Study

Abstract

Selection of patients for iliofemoral vein stenting requires the identification and quantification of stenotic lesions with imaging such as multiplanar venography (MPV) and intravascular ultrasound (IVUS). The Venogram Versus Intravascular Ultrasound for Diagnosing Iliofemoral Vein Obstruction (VIDIO) trial showed IVUS to be more sensitive than MPV for identifying “significant” iliofemoral vein stenosis. To date, the comparative utility of these imaging studies using the customary 50% treatment threshold for stenosis severity has not been prospectively validated against clinical improvement. The multicenter VIDIO trial prospectively enrolled 100 symptomatic patients (clinical, etiologic, anatomic, and pathophysiologic [CEAP] class 4-6) with suspected iliofemoral venous outflow disease. Venous stenting for presumed significant iliofemoral vein stenosis was performed on 68 patients. Each underwent baseline and poststenting MPV and IVUS to compare the diagnostic and clinical utility of the tests. Based on imaging, stenosis was characterized as nonthrombotic (ie, compression) in 48 patients and post-thrombotic (ie, luminal scar, sclerotic vein contraction) in 20 patients. The Revised Venous Clinical Severity Score was the clinical index for which anatomic imaging measurements were consistently related to patient outcome. A 4-point reduction in the Revised Venous Clinical Severity Score between baseline and 6 months was used as an indicator of clinical improvement for calculating sensitivities and specificities for MPV-determined percentage diameter reduction and IVUS-determined percentage diameter and area reduction. Receiver operating characteristic curve analysis was used to determine the optimal diameter and area thresholds for clinical improvement. The validity of each imaging measure was assessed with respective areas under the curve (AUCs), calculating P values estimating the probability that a measure was of greater value than chance alone. Clinical improvement after stenting was best predicted by IVUS-determined baseline stenosis, either by diameter (AUC, .64; P = .046; cutoff >55.7%) or area reduction (AUC, .64; P = .043; cutoff >53.6%). MPV-determined baseline stenosis was not predictive (AUC, .58; P = .290; cutoff >51.9%). Evaluating lumen improvement from baseline to after intervention revealed that IVUS-determined area change was the best measure (AUC, .70; P = .003; cutoff >40.5%), followed by IVUS-determined diameter change (AUC, .66; P = .020; cutoff >45.6%). MPV-determined diameter change did not predict clinical improvement (AUC, .56; P = .370; cutoff >37.5%). In the 48-patient nonthrombotic subset, IVUS diameter measurements of baseline stenosis were significantly predictive (AUC, .67; P = .029; cutoff >61.2%). MPV was not predictive (AUC, .61; P = .17; cutoff >51.9%). The measurement with the most predictive utility for clinical improvement was stenotic change in the nonthrombotic subset, with an AUC of .75 for both IVUS area and diameter measurements (P = .001). MPV-determined diameter change in the nonthrombotic subset had borderline predictive utility for symptom improvement (AUC, .66; P = .05). At the time of intervention, IVUS appears to have better predictive utility than MPV for identifying iliofemoral vein stenosis in CEAP class 4-6 patients that, when treated, results in significant symptom improvement. These findings for the entire cohort corroborate the conventional 50% threshold as defining a clinically significant stenosis that, when stented, has significant predictive value for symptom improvement. In nonthrombotic patients, however, a higher threshold of >61% diameter stenosis may better predict clinical improvement.