Abstract
Purpose: To determine the frequency, dimensions, predictors, and sequelae of Günther Tulip filter (GTF) tilt measured at the time of intended retrieval. Materials and Methods: Retrospective review of all medical records and posteroanterior cavograms of 175 patients who underwent both placement and retrieval of the GTF between August 2003 and July 2007 was performed to assess the frequency, dimensions, predictors, and sequelae of tilt. Results: Tilt occurred at the first retrieval attempt in 159 of the 175 patients (91%). The average degree of tilt was 7.1° (range, 0°-30°), with 87 of the 159 filters with tilt (55%) having a rightward tilt. Compared with the femoral approach, filters placed with a jugular approach demonstrated 4.2° (range of the standard deviation, 3.1°-5.3°) greater tilt at the first retrieval attempt (95% confidence interval = 2.6°, 5.7°; P <.001, two-sided Student t test), a greater frequency of tilt of at least 14° (P = .002, two-sided Fisher exact test), and greater rightward tilt predominance (P = .046, one-sided Fisher exact test). Tilt magnitude at the first retrieval attempt correlated positively with the inferior vena cava diameter 40 mm caudal to the renal vein confluence (R = .183, P = .018, Pearson correlation). Within its limitations, this study detected no new cases of pulmonary embolism, caval perforation, or GTF migration. The success rates at the first attempt at retrieval and the cumulative GTF retrieval success rates were 93% (176 of 190 filters) and 97% (181 of 190 filters), respectively. All 29 GTFs with tilt of at least 14° were placed and successfully retrieved by means of a jugular approach with minimal clinical and technical sequelae. Conclusions: Frequent GTF tilt detected at the first retrieval attempt can reach at least 14° and is associated with minimal sequelae. Insertion approach and caval diameter are significant factors in GTF tilt.