Classification of Stent Failure and Risk Factors for Further Reintervention in Patients Treated for Deep Venous Obstruction

Deep venous stenting is an established treatment for iliac outflow obstruction, however published data continues to show a high rate of reintervention to maintain stent patency, particularly in chronic disease. The aim of this study was to evaluate causes for reintervention and identify factors that may predict further reintervention.

Methods: Retrospective analysis of a prospective database of deep venous stents placed between 2014-2023 for chronic disease was undertaken. The cause of stent failure was classified using the St. Thomas’ Classification. Kaplan-Meier analysis was carried out for intervention-free survival following reintervention. Univariate Cox’s proportional hazard regression was employed to identify predictor variables. Hazard ratios (HR), 95% confidence intervals and p-values were calculated. Backward selection was used to optimise the final predictive model.

Results: A total of 421 patients underwent deep venous stenting for chronic disease (59.86% female, n=252). The majority were treated for post thrombotic syndrome (348 patients, 82.7%), 66 patients (15.7%) were treated for non-thrombotic iliac vein lesions (NIVL) and 7 (1.7%) for other causes. In 94 patients (22.3%) bilateral procedures were required and a total of 515 limbs were treated. Primary patency at one year was 65.6% and secondary patency 95.1%. Reintervention was performed in 174 patients (203/515 limbs, 39.4%) and not attempted in 6 patients (6/515 limbs 1.17%) totalling 209 limbs. Of these, 90 patients (101/203 limbs, 49.8%) required more than one procedure. The median number of reinterventions was 2 (range 0-14). Causes of stent failure were classified as Flow (101/209, 48.3%), Technical (55/209, 26.3%), Haematological (15/209, 7.2%), and Multifactorial (38/209, 18.2%). Flow represented the largest cause of stent failure in all subsequent reinterventions analysed. Median survival following a technical failure was 927 days vs 215 days for other causes (HR 0.60, 0.41-0.88 p=0.018). Univariate Cox regression identified stenting across the inguinal ligament (HR 1.78, 1.11-3.04 p=0.024), thrombophilia (HR 0.67, 0.44-0.99 p=0.0497), flow related failure (HR 1.72, 1.11-2.77 p=0.019) and profunda vein disease (HR 2.21, 1.50-3.23 p< 0.0001) to be significant predictors. Multiple variable analysis using backward selection was used to build an optimised predictive model. PFV disease and thrombophilia were the best predictors of further reintervention (Wald 21.76, p< 0.0001).

Conclusions: Inflow disease accounts for the largest proportion of stent failures following both the index procedure and subsequent re-intervention. Profunda vein disease and thrombophilia are significant predictors of further reintervention. Addressing venous inflow is a significant unresolved challenge and is an important factor for stent patency. Further work is needed to understand how flow may be optimised to reduce reintervention rates and improve patency.