OC2. Abstract Title: Comparison of Two Clinical Scales to Assess the Post-Thrombotic Syndrome: Secondary Analysis of a Multicenter Randomized Trial of Pharmacomechanical Catheter-Directed Thrombolysis for Deep Vein Thrombosis

Abstract

<h3>Background</h3> Post-thrombotic syndrome (PTS) occurs in 20-50% of patients following proximal deep vein thrombosis (DVT), despite anticoagulation treatment. The International Society on Thrombosis and Haemostasis recommends using the Villalta scale to standardize the diagnosis of PTS and to quantify its severity. However, many investigators use the Venous Clinical Severity Score (VCSS) to assess PTS. Different to the Villalta score, the VCSS was developed as a measure for chronic venous disease and not PTS specifically. The aim of the study was to determine which of Villalta and VCSS best captures clinically important PTS and PTS severity by analyzing the relationship of each to QoL scores in the Acute Venous Thrombosis: Thrombus Removal with Adjunctive Catheter-Directed Thrombolysis (ATTRACT) trial study population. <h3>Methods</h3> A secondary analysis of the ATTRACT randomized controlled clinical trial was conducted. 621 ATTRACT patients with symptomatic proximal DVT are included in this analysis. Correlations of the Villalta and VCSS scores with QoL scores (Short-Form Health Survey-36 physical component score/mental component score [SF-36 PCS/MCS]; the Venous Insufficiency Epidemiological and Economic Study-QoL/Symptoms [VEINES-QoL/Sym] questionnaire) were examined at study visits. The correlation of the random intercept (mean scores) and random slope (rate of change of the scores) between the Villalta, VCSS and the VEINES-QoL/Sym scores was assessed using a multivariate longitudinal model. <h3>Results</h3> The mean age at enrolment was 51 years of age, 78% were white, 62% were male, the mean BMI was 32 kg/m2, 57% had iliofemoral DVT, 19% had a previous ipsilateral DVT and 93% were taking one or more anticoagulation therapies prior to randomization. The median (range) correlation coefficients across all follow-up visits for the Villalta scale and VEINES-QOL and the Villalta scale and VEINES-Sym were -0.73 (0.51, -0.73) and -0.75 (-0.50, -0.76), respectively. The median (range) correlation coefficients for the Villalta scale and the SF-36 PCS or the SF-36 MCS across all follow-up visits were -0.51 (-0.35, -0.54) and -0.31 (-0.21, -0.38), respectively. In all correlation analyses between the Villalta scale and QoL measures, correlations at baseline were the weakest, and then increased until 6 months, after which the correlation coefficients remained relatively constant. The median correlation coefficient (range) across all follow-up visits for VCSS and VEINES-QOL and for VCSS and VEINES-Sym were -0.39 (-0.37, -0.39) and -0.41 (-0.36, -0.41), respectively. The median correlation (range) between VCSS and SF-36 PCS and for VCSS and MCS were -0.32 (-0.31, -0.32) and -0.12 (-0.12, -0.14), respectively. The magnitude of the correlations remained relatively constant from 6 months to 24 months follow-up visits. The Villalta scale had a strong negative correlation with VEINES-QoL and VEINES-Sym, a moderate negative correlation with SF-36 PCS, and a weak negative correlation with SF-36 MCS. Conversely, the VCSS had a weak negative correlation with VEINES-QoL, VEINES-Sym and SF-36 PCS, and a very weak negative correlation with SF-36 MCS. Adjustment for age, sex, ethnicity, race, extent of DVT and previous ipsilateral DVT did not change our findings. Each pair of disease severity scales and QoL scales were analyzed as a multivariate outcome in a multivariate longitudinal model. A random intercept (mean scores) and random slope (rate of change of the scores) were present in all these multivariate longitudinal models. The impact of covariates (i.e. age, sex and BMI) was also examined. The Villalta score and VCSS score have a positive correlation between their random intercept and random slope. This suggests that patients with a higher average Villalta score also tend to have a higher average VCSS score (correlation is 0.74 and 0.69 if adjusted). Those that have a higher rate of change in Villalta score over time tend to have a higher rate of change in VCSS (correlation is 0.72 and 0.84 if adjusted). The impact from covariate adjustment by age, sex and BMI was minor. <h3>Conclusion</h3> For all measures, the Villalta scale has a substantially higher correlation with QOL than the VCSS. Our findings support the use of the Villalta scale to assess PTS in preference to VCSS, as it better captures the impact of PTS on patient reported QoL, a key consideration in patients with chronic PTS.