Long-term follow-up of Vena Tech-LGM filter: predictors and frequency of caval occlusion.

PURPOSE

To report the frequency of caval occlusion after Vena Tech-LGM filter placement and identify related factors and their potential clinical significance.

MATERIALS AND METHODS

The filter was inserted into 243 patients, 142 of whom met inclusion criteria for this prospective study. Follow-up examinations performed every 2 years included clinical evaluation, plain frontal radiography of the abdomen, duplex scanning of the inferior vena cava (IVC), and/or phlebocavography.

RESULTS

A progressive decrease in IVC patency was observed, reaching 66.8% at 9 years of follow-up. Complete caval occlusion occurred in 28 patients and was significantly (P < 10(-6)) associated with retraction in 24 cases. Caval occlusion was not related to age, sex, pulmonary embolism (PE), deep venous thrombosis level, underlying conditions predisposing to a thromboembolic disease before filter insertion, the level of filter placement, use of anticoagulant therapy, and death during follow-up. PE with anticoagulation failure was a predictive factor (P = .016) of subsequent filter occlusion during follow-up as compared to all other clinical indications for filter placement. Filter patency at 9 years of follow-up was 35.2% in the PE group with anticoagulation failure and 80% for other patients (odds ratio, 2.5; 95% confidence interval 1.16-5.4).

CONCLUSION

PE with anticoagulation failure was the only factor predictive of subsequent caval occlusion observed in patients after Vena Tech-LGM filter placement. Caval occlusion was also related to Vena Tech-LGM filter retraction, which usually occurred at the time of occlusion.