Ten Years’ Experience in Vena Cava Filter Placement – Sample Characterization and Long-term Follow-up

Abstract

Introduction: Vena cava filter (VCF) has been widely used since it was introduced in 1967. However, there is a general lack of information regarding VCF holders characteristics and their long-term outcomes, and there is not a standardized clinical or imagiological follow-up of these patients. Methods: This is a retrospective analysis of patients who had a VCF placed from 2008 to 2018 at Santa Marta Hospital (Lisbon, PT). Variables collected were patient demographics, indication for filter, filter used. Primary end-point was global survival. Secondary end-point was filter related complications. A substudy was performed to determine late filter related complications in alive patients by abdominal venous duplex ultrasound and abdominal x-ray. Cumulative rates of events were determined by Kaplan-Meier method. Results: From 2008 to 2018, 124 patients underwent VCF placement. Male:female ratio was 1:1. 104 patients (84%) had active venous thromboembolic event (VTE) at time of VCF placement. The most prevalent VTE risk factors were: previous documented VTE (38; 31%), recent hospitalization (63; 51%) and active neoplasm (46; 37%). The most common indications for VCF placement were: anticoagulant contraindication (69; 56%) and persisting pulmonary embolism (PE) despite anticoagulation (29; 23%). Excluding patients whose access site was not specified (16), femoral vein was punctured in 96 patients (89%) and internal jugular vein in 12 patients (11%). VCF was placed on infrarrenal inferior vena cava (IVC) in 100% of patients. Excluding patients whose filter type was not specified, permanent filter was placed in 66 patients (80%). VenaTech LP (B-Braun®) was placed in 34 patients and TrapEase (Cordis®) in 17 patients. A retrievable filter was placed in 17 patients (20%). Celect (Cook®) was placed in 2 patients and OptEase (Cordis®) in 15 patients. There was a VCF retrieval trial in 5 patients, three with success. Survival rate was 86.0±3.2% at 1 month, 63.3±4.5% at 1 year and 55.4±4.8% at 5 years. We found no PE events or VCF related-mortality during follow-up. Patients with active cancer had lower survival rates (28.7±8.5% at 5 years; p<0.001). Survival rates were similar regardless other VTE risk factors and filter indication or type. At present time of our substudy, 32 alive patients were evaluated at median time of 58 months after VCF placement. 2 patients had lower limb DVT in the meantime, one of which have active and the other previous colorrectal cancer. There were no signs of PE. 18 patients (56%) are anticoagulated, 10 with a DOAC (56%). We documented 5 asymptomatic VCF complications: 2 filter occlusions (with patent IVC proximal and distal to VCF), 3 filter tilts (one in a patient that underwent a trial of VCF retrieval). There was no documented filter fracture, penetration or migration. Conclusion: Main absolute indications for VCF placement continue to be VTE in presence of contraindication for anticoagulation and persisting PE despite anticoagulation. Persistent VTE and risk factors for recurrence, lack of follow-up and reported complications of filter retrieval may account for increased use of permanent VCF and retrievable VCF left permanently in place. In spite of known long-term adverse events of VCF (migration, fracture, tilt, penetration, thrombosis), their low incidence and absence of related symptoms challenge a standard of care follow-up for VCF holders. Very low survival rates of VCF holders who have active neoplasm challenge the benefits of VCF placement in patients with dismal prognosis.