RESUMEN
PURPOSE: To assess the technical success and safety for retrieval of the G2 filter. MATERIALS AND METHODS: The authors performed a prospective, multicenter study of 100 patients with temporary indication for caval interruption. Patients were enrolled consecutively between December 2005 and July 2006. There were 67 men and 33 women with a mean age of 52.1 years (range, 19-82 years). Indications for filter placement were trauma (n = 56), perioperative risk (n = 16), and medical indications (n = 28). Forty-two patients had venous thromboembolism at filter placement. Fifty-eight filters were placed prophylactically. RESULTS: Retrieval was attempted in 61 patients. Fifty-eight of the 61 filters (95%) were successfully retrieved after a mean dwell time of 140 days (range, 5-300 days). In all failed retrievals, the filter tip was against the caval wall. There was no difference in dwell times between successful and unsuccessful retrievals. Although there were no cases of cranial migration, caudal migrations were observed in 12% of cases (10 of 85 patients with a complete data set). Other device-related complications included filter fracture (1/85, 1.2%), filter tilt of more than 15 degrees (15/85, 18%), and leg penetration (16/61, 26%). The recurrent pulmonary embolism (PE) rate was 2%, with no PE in the 30-day period after filter retrieval. CONCLUSIONS: Retrieval of the Recovery G2 filter was safe and successful in most patients. Caudal migration was observed as an unexpected phenomenon.