2. Van Son JAM, Mierzwa M, Mohr FW. Resection of atherosclerotic aneurysm at origin of aberrant right subclavian artery. Eur J Cardiothorac Surg 1999; 16:576–579. 3. Tosenovsky P, Quigley F, Golledge J. Hybrid repair of an aberrant right subclavian artery with Kommerell’s diverticulum. EJVES 2010; 19:e31–e33.
4. Yang C, Shu C, Li M, Li Q, Kopp R. Aberrant subclavian artery pathologies and Kommerell’s diverticulum: a review and analysis of published endovascular/hybrid treatment options. J Endovasc Ther 2012; 19:373–382.
Failure of Filter Reexpansion during Unsuccessful Retrieval of Option Inferior Vena Cava Filter
and ﬁlter retrieval was aborted. Upon retracting the sheath, the ﬁlter did not return to form as expected but instead remained collapsed within the IVC lumen (Fig 1c). The patient was referred for laser-assisted ﬁlter retrieval, which was successfully performed 1 week later. No adverse events occurred during the interval to laserassisted removal. In the second case, a 54-year-old woman was initially admitted to an outside hospital for a strangulated ventral hernia requiring bowel resection. During a subsequent prolonged hospitalization, the patient was diagnosed with bilateral lower-extremity deep vein thrombosis. According to outside records, even though the patient was already receiving anticoagulation, an Option ﬁlter was placed given her history of anticoagulation failure and concern that a pulmonary embolism would carry substantial mortality risk. The patient presented to our institution for ﬁlter retrieval 11 months after placement. An initial cavogram demonstrated an infrarenal ﬁlter free of thrombus, with a single strut extending 3 mm beyond the caval margin (Fig 2a). With the use of a Cook retrieval set, the ﬁlter hook was successfully snared and the retrieval sheath was advanced over the ﬁlter. The struts were partially reduced into the sheath but remained adherent to the caval wall despite the application of traction. Upon releasing the ﬁlter, the struts remained predominantly collapsed except for their distalmost portions, with suggestion of helical twisting of the struts (Fig 2b). A repeat cavogram demonstrated new focal inward tenting of the caval wall at the level of the strut anchors (Fig 2c). A 14 40-mm angioplasty balloon (Cordis Europa/ Johnson and Johnson, Roden, The Netherlands) was then inﬂated within the remaining expanded portion of the ﬁlter in an attempt to reexpand the ﬁlter, yielding only minimal success. As there was still brisk antegrade ﬂow beyond the point of narrowing, no further intervention was attempted at this time. Oral anticoagulation was resumed, and the patient was referred for laserassisted ﬁlter retrieval, which was successfully performed 2 weeks later. No adverse events occurred during the interval to laser-assisted removal. Here we describe two cases of failure of ﬁlter reexpansion during unsuccessful retrieval of the Option IVC ﬁlter. Each case was attempted by a different subspecialtycertiﬁed interventional radiology attending physician, both of whom have extensive experience in IVC ﬁlter retrieval. In neither case was there caval thrombus or a preexisting caval or pericaval abnormality.
From: Olufoladare G. Olorunsola, MD Maureen P. Kohi, MD Nicholas Fidelman, MD Jeanne M. LaBerge, MD Robert K. Kerlan, MD Department of Radiology and Biomedical Imaging University of California, San Francisco 505 Parnassus Ave. Room 391, Box 0628 San Francisco, CA 94143
Editor: We report two cases of unsuccessful ﬁlter removal involving the Option inferior vena cava (IVC) ﬁlter (Rex Medical, Conshohocken, Pennsylvania). In each case, the ﬁlter was partially sheathed but could not be removed because of ﬁrm attachment to the caval wall. After aborting the retrieval procedure and retracting the sheath, the ﬁlters did not reexpand to their normal conﬁguration as expected, instead remaining collapsed within the IVC lumen. Approval was obtained from our institutional review board, with waiver of informed consent, for the description of the present two cases. In the ﬁrst case, a 57-year-old woman underwent Option IVC ﬁlter placement for pulmonary embolism diagnosed 8 days after a traumatic hepatic laceration. Immediate postplacement ﬂuoroscopy demonstrated an unusual appearance of the ﬁlter, with several of the struts clustered closely together along the right lateral border of the IVC (Fig 1a). A similar appearance was seen on computed tomography (CT) 73 days later (Fig 1b). As a result of discontinuous follow-up, retrieval was not attempted until 277 days (9.2 mo) after placement, with an initial cavogram again demonstrating the unusual conﬁguration of the ﬁlter, which was free of thrombus. With the use of a Cook retrieval set (Cook, Bloomington, Indiana), the ﬁlter hook was snared, and the ﬁlter was nearly completely sheathed but remained adherent to the caval wall despite considerable traction. Subsequent forceful traction elicited abdominal pain, None of the authors have identiﬁed a conﬂict of interest. http://dx.doi.org/10.1016/j.jvir.2013.03.026
Letters to the Editor
Olorunsola et al
Figure 1. (a) Postplacement anteroposterior ﬂuoroscopic image demonstrates an unusual conﬁguration of an Option IVC ﬁlter, with several of the struts clustered along the right lateral border of the IVC (arrowhead). (b) CT obtained 73 days following placement also demonstrates uneven clustering of the struts (arrow). (c) Oblique projection cavogram following unsuccessful attempted ﬁlter retrieval demonstrates retention of the collapsed ﬁlter within a patent IVC.
Figure 2. Anteroposterior-projection cavograms during attempted retrieval in the second case. (a) Infrarenal Option IVC ﬁlter before attempted retrieval shows 3-mm penetration of one of the struts (arrow). (b) After attempted sheathing and subsequent sheath retraction, the ﬁlter remained partially collapsed, with a suggestion of helical twisting of the struts. (c) New focal inward tenting of the caval wall is seen at the level of the strut anchors (arrowhead).
We may speculate as to potential mechanisms governing the observed sequence. The problem of tightly adherent struts to the caval wall is likely related to the prolonged dwell times in both cases (9 and 11 mo, respectively). In one case, there was evidence of strut penetration, possibly playing a role in strut adherence. The subsequent failure of ﬁlter reexpansion is the more unusual observation. The imaging appearance in both cases suggests that this may be attributable to entanglement of ﬁlter struts. In the ﬁrst case, CT and venography before attempted sheathing demonstrate an unusual ﬁlter conﬁguration, with uneven clustering of
several struts along the right lateral caval wall (Fig 1a, b). In the second case, there is suggestion of helical twisting of the struts following attempted sheathing (Fig 2b). Speciﬁc design features could theoretically predispose to limb clustering or entanglement in the Option ﬁlter. Its ultra–low-proﬁle delivery system (6.5 F) necessitates thinner struts, which are perhaps more easily distorted. Additionally, the Option ﬁlter comprises six principal struts, which may be more conducive to interactions between the struts. Crossing of struts at the time of initial deployment of an Option ﬁlter has been reported (1).
Another potential mechanism to explain the ﬁxed ﬁlter collapse is ﬁbrous tissue acting to constrain the struts. Histologic analyses have demonstrated dense ﬁbrous tissue adherent to retrieved ﬁlter specimens (2). It is possible that a ﬁbrous sheath could form in the ﬁlter hook region, which could in turn be pushed over the ﬁlter struts when advancing the retrieval sheath. The potential adverse consequences of this type of ﬁlter dysfunction could be severe. The result may be a ﬁlter left in unstable position with a possibility of ﬁlter embolization, as in the ﬁrst case described here, in which most of the struts appeared to have been detached from the caval wall. A second potential problem is caval stenosis, as seen in the second case, which could result in ﬂow limitation, caval thrombosis, or caval occlusion. Given the potentially serious consequences, it is the authors’ opinion that such occurrences warrant shortterm follow-up for repeat retrieval attempts by using
Durable Plug and Onyx Occlusion of a Refractory Bile Leak From: Kyle A. Wilson, BA Ziv J Haskal, MD Department of Radiology University of Maryland School of Medicine 22 S. Greene St., GK214 Baltimore, MD 21030
Editor: Bile leaks of iatrogenic or traumatic etiology are most commonly treated by biliary decompression, diversion of ﬂow away from the leak, and drainage of adjacent bilomas. Persistent bile leaks have been treated with embolization with ﬁbrin glue, n-butyl cyanoacrylate (NBCA), or microcoils, or with hepatic resection (1–3). Herein, we describe a case of persistent bile leak that was refractory to prolonged diversion and multiple liquid embolizations. A combination of metallic and liquid embolic agents was used to achieve long-term occlusion. This report was exempted from institutional review board approval. In 2009, a 49-year-old man underwent right hepatectomy for curative treatment of synchronous isolated liver metastases from a primary colorectal cancer. Two months later, he presented with a bile leak that was addressed by endoscopic biliary stent placement and percutaneous drainage of the perihepatic biloma. Z.JH. has royalty agreements with Cook (Bloomington, Indiana), is a paid consultant for W.L. Gore and Associates (Flagstaff, Arizona) and Bard Peripheral Vascular (Tempe, Arizona), has research funded by Bard Peripheral Vascular and W.L. Gore and Associates, and owns shares in AngioDynamics (Latham, New York). K.A.W. reports no disclosures. This letter includes Video E1 available online at www.jvir.org http://dx.doi.org/10.1016/j.jvir.2013.03.017
specialized methods such as the laser-assisted retrieval technique. It is unknown with what frequency this complication occurs or whether it is unique to the Option ﬁlter. It is likely that previous such events have occurred with the Option ﬁlter and possibly ﬁlters of other types, but were not reported. Interventional radiologists who retrieve ﬁlters should be aware of this perplexing, alarming, and potentially dangerous problem.
REFERENCES 1. Johnson MS, Nemcek AA Jr, Benenati JF, et al. The safety and effectiveness of the retrievable option inferior vena cava ﬁlter: a United States prospective multicenter clinical study. J Vasc Interv Radiol 2010; 21:1173–1184. 2. Kuo WT, Cupp JS, Louie JD, et al. Complex retrieval of embedded IVC ﬁlters: alternative techniques and histologic tissue analysis. Cardiovasc Intervent Radiol 2012; 35:588–597.
The stent and drain were removed 1 month later. Within 3 months, he returned with a recurrent, large biloma, and was treated with repeat drainage and placement of a percutaneous internal external left biliary catheter for diversion. For the next 12 months, the leak continued, and the biloma drained approximately 100 mL per day. Fistula embolization with ethylene vinyl alcohol copolymer (Onyx; ev3, Plymouth, Minnesota) was performed at 1 month, followed by NBCA embolization at 3, 5, 6.5, and 8 months thereafter. We ﬁrst met the patient 2 weeks later and considered a different approach (Fig 1, Video E1, available online at
Figure 1. Cholangiography through the percutaneous left biliary catheter demonstrates the leak (black arrow) ﬂowing immediately into the right hepatectomy bed. The tip of the lockingloop biloma catheter is seen (white arrow)