Thrombus echogenicity and post-thrombotic syndrome severity after deep vein thrombosis: Insights from residual vein thrombosis assessment

Objective To evaluate the association between thrombus echogenicity in patients with residual venous thrombosis (RVT), assessed by grayscale median (GSM) on duplex ultrasound (DUS), and post-thrombotic syndrome (PTS) severity in patients with prior lower limb deep vein thrombosis (DVT), and to explore its relationship with recurrent venous thromboembolism (RVTE), including the influence of time since the index event. Methods This single-center observational study included 249 adult patients with a history of objectively confirmed lower limb DVT. The study integrated retrospective characterization of the index event, cross-sectional clinical and DUS assessment (2018–2020), and prospective follow-up for RVTE through 2024. RVT was defined using compression ultrasound criteria. GSM was quantified from ultrasound images using both a dedicated software-based method and a more accessible Adobe Photoshop–based approach. PTS was assessed using the Villalta score and categorized by severity. Associations between imaging parameters (GSM and RVT) and outcomes (PTS and RVTE) were analyzed in the overall cohort and across predefined time intervals from DVT to ultrasound (≤24 months, 25–72 months, >72 months). Discriminatory performance was evaluated using receiver operating characteristic (ROC) analysis. Results PTS was present in 67.9% of patients, and RVTE occurred in 25.3% during follow-up. RVT was identified in 57.8% of the cohort and showed a non-significant trend toward higher recurrence (29.9% vs. 19.0%; p=0.053), but was not associated with PTS severity. Lower GSM values measured by the software-based method were significantly associated with greater PTS severity (e.g., median 32.0 vs. 39.0 for moderate-to-severe vs. mild/absent PTS; p=0.005), whereas the Photoshop-based method showed no significant association. This relationship appeared stronger in patients evaluated within 24 months (AUC 0.833, 95% CI: 0.509–1.000), although estimates were imprecise due to small sample size. No significant association was observed between GSM and RVTE, regardless of method or time stratification. Discriminatory performance was modest, with moderate accuracy for moderate-to-severe PTS (AUC 0.644, 95% CI: 0.546–0.736), but consistently poor discrimination for predicting RVTE (AUC 0.552, 95% CI: 0.447–0.663). Conclusions GSM demonstrates a modest, time-dependent association with PTS severity, particularly in the early period after DVT, but does not predict RVTE. These imaging-derived parameters likely reflect aspects of post-thrombotic venous remodeling but have limited utility for individual risk stratification. Their role remains exploratory, and further prospective studies with standardized imaging and advanced analytical approaches are needed to define their potential within integrated risk assessment models.