Michael L. Martin

Treatment of Type II Endoleaks with Onyx

Endoleaks are defined as persistent perfusion of an abdominal aortic aneurysm (AAA) after endovascular stent-graft deployment. The authors describe their experience treating six endoleaks with the liquid embolic agent Onyx (ethylene-vinyl-alcohol copolymer). Complete endoleak occlusion was achieved in five of six cases. Follow-up imaging has demonstrated decreased aneurysm diameter in all patients 7-29 weeks (mean = 19.2 weeks) after treatment.

Sedation and analgesia in the interventional radiology department.

Administration of sedation and analgesia in the interventional radiology suite is often necessary during painful diagnostic and therapeutic procedures. Although sedative and analgesic agents are generally safe, catastrophic complications related to their use can occur, often as a result of incorrect drug administration or inadequate patient monitoring. The incidence of adverse outcomes related to provision of sedation and analgesia can be reduced with improved understanding of the pharmacology of these medications, by providing adequate monitoring to sedated patients, by recognizing patients who are at increased risk of experiencing an adverse drug reaction, and by early and appropriate management of complications.

Comparison of MRI and Sonography in the Preliminary Evaluation for Fibroid Embolization

OBJECTIVE The purpose of our study was to evaluate whether pelvic MRI provides additional clinically relevant information after sonography in the preprocedure evaluation of uterine artery embolization of fibroids. MATERIALS AND METHODS Forty-nine women who presented for consultation for uterine artery embolization were retrospectively reviewed. The MRI and sonography scans were independently evaluated and compared for uterine size, fibroid size and location (categorized as paraendometrial, intramural, subserosal, or pedunculated) of the four largest fibroids in each patient, and the total number of fibroids present. RESULTS One hundred twenty-two fibroids were measured. The uterine volume was significantly smaller as measured on MRI compared with sonography (p = 0.01). We found good MRI and sonography correlation of the volume of the single largest fibroid in each patient (R = 0.87) but poor correlation of fibroid location (R = 0.17). MRI detected 31 paraendometrial fibroids and three pedunculated fibroids that were thought to be intramural fibroids on sonography. Five fibroids thought to be paraendometrial on sonography were confirmed to be subserosal or intramural on MRI. Discrepancy in the total number of fibroids was noted, with additional fibroids found on MRI in 31 of 49 patients and erroneously suspected on sonography in five of 49 patients. Pelvic MRI affected management in 11 of 49 patients, leading to cancellation of uterine artery embolization in four patients. In another seven patients who were originally thought to be poor candidates on the basis of sonographic findings, uterine artery embolization was performed. MRI did not alter the management plan in 38 patients. CONCLUSION MRI provided considerable additional information compared with sonography and affected clinical decision making in a substantial number of patients. MRI should be considered in all patients being evaluated for uterine artery embolization.

Repeated Günther Tulip Inferior Vena Cava Filter Repositioning to Prolong Implantation Time

A patient presented with iliofemoral deep vein thrombosis, a small pulmonary embolism, and a paradoxic embolus to the axillary artery resulting from a patent foramen ovale (PFO). As prophylaxis against further paradoxic emboli while awaiting percutaneous PFO closure, a Günther Tulip inferior vena cava (IVC) filter was implanted. To prevent incorporation of the IVC filter into the caval wall, it was repositioned twice with use of a filter retrieval set from a transjugular approach. In this way, the implantation time of the filter was extended beyond the recommended period of 10 days. The filter was successfully retrieved 19 days later during percutaneous closure of the PFO.

Selective spermatic venography and varicocele embolization in men with circumaortic left renal veins

The presence of variant venous anatomy may increase the technical difficulty of percutaneous varicocele embolization. The authors review their experience performing venography of the left spermatic vein and varicocele embolization in 18 men with circumaortic left renal veins. Selective catheterization of the left spermatic vein was achieved in all but one patient, and all patients with positive venographic results underwent successful embolization. The site of the confluence of the left spermatic vein and the renal vein was variable and it was necessary to use jugular and femoral venous approaches to achieve selective left spermatic vein catheterization. Familiarity with anatomic variations associated with circumaortic renal veins and with embolization techniques from jugular and femoral venous routes facilitates percutaneous varicocele embolization in patients with this variant.

Pain and anxiety: two problems, two solutions.

tional radiology patients. In this issue of JVIR, Schupp and colleagues (7) describe the administration of a questionnaire to determine preprocedural anxiety levels and demonstrate the impact of anxiety on drug requirements, pain, and anxiety during the procedure and the effectiveness of differing methods of supportive care. The knowledge provided by this study will hopefully improve the ability of interventional radiology practitioners to administer anxiolysis and analgesia (pharmacologic and nonpharmacologic) in a more effective and patientdriven manner.

Percutaneous repair of punctured subclavian artery under fluoroscopy.

T he insertion of central venous catheters is a necessary but occasionally dangerous procedure for critically ill patients. It has been estimated that the complication rates range from 1% to 15% (1). The common approaches include internal and external jugular, subclavian, and femoral. The subclavian route is preferred because of its known lower infection rate and its ease of anchoring the catheter to an immobile surface, which is also more remote from body secretions (2). However, the complication of puncturing the adjacent artery is more serious than with other approaches because it cannot be externally compressed. Thus, the insertion of a catheter via the subclavian route is absolutely contraindicated in patients with coagulopathy. Surgical repair is difficult because of the poor exposure to the deep space behind the subclavian vein and the scalenus anterior muscle (3). We report a case in which the subclavian artery was inadvertently punctured in the intensive care unit (ICU) and was repaired percutaneously under radiologic guidance. A 46 yr-old-man suffered C5–6 quadriplegia after a motor vehicle accident 22 yrs ago. He was obese and had a mild kyphoscoliosis. A chronic diverting colostomy was in place because of past nonhealing sacral decubitus ulcers. The patient was admitted to the hospital because of bilateral pneumonia and was subsequently intubated and ventilated in the ICU. A left subclavian catheter was inserted upon admission without complication. However, he remained septic after 7 days, despite broad-spectrum antibiotic coverage. His temperature was 40°C, his white blood cell count was 14.2, and his hemodynamic status was labile, partly because of his autonomic dysfunction. The decision was made to change the central venous catheter to the right subclavian vein and repeat the septic workup. The patient was sedated, put in Trendelenburg position, prepared, and draped in the usual fashion. Under local anesthesia, the search needle was inserted 2 cm caudal to the clavicle and gradually advanced toward the suprasternal notch. On the second attempt, nonpulsatile blood was obtained and a double lumen catheter was inserted in the usual manner. Soon afterward, it was discovered that normal saline could not be infused through the same catheter via gravity and pulsatility was observed at the interphase between the blood and the crystalloid fluid. Blood gases via the new central catheter showed a pH of 7.47, PO2 of 86, PCO2 of 40, HCO3 of 28, at 40% oxygen. A chest radiograph (Fig. 1) was not helpful in confirming the presence or absence of arterial catheterization. Subsequent angiogram confirmed its erroneous location (Fig. 2). The catheter was kept in its place, and a new femoral venous catheter was inserted without complication. After surgical and radiologic consultations, it was decided that a percutaneous closure device should be tried to close the artery (4), because a CT scan showed that it was cleanly punctured and no adjacent hematoma was found. After obtaining consent from the patient and his wife, the patient was transferred to the angiography suite under sedation and full mechanical ventilation with vascular surgeons standing by. As a precaution, an angiographic catheter was inserted via the right common femoral artery into the right subclavian artery past the point of puncture (Fig. 3). An exchange length guidewire was then placed into the artery. This was done so that an occlusion balloon could be easily and quickly positioned across the subclavian arterial puncture site if it was not completely sealed by the percutaneous sealing device. The misplaced subclavian arterial catheter was withdrawn over a standard angiographic wire, and the artery was sealed with an Angio-Seal device (Sherwood, Davis and Geck, St. Louis, MO) in the following steps. First, the 8-Fr sheath supplied with the kit was advanced into the artery. Through this, a 2 10 1 mm polyglycolic acid suture anchor was placed inside the artery and cinched against the arterial wall by tugging on a 2-0 resorbable suture. While maintaining traction on the suture, a 26-mg collagen plug was tamped against the outer aspect of the arterial wall adjacent to the puncture site, and finally, this was held in place by tension on a tamper placed over the suture (Fig. 4). After 20 mins, the suture was cut and the patient was transferred back to the ICU with no external manifestations of his previous catheter apart from the now-sealed puncture site on the skin. Total procedure time was 32 mins; total fluoroscopy time was 3 mins, 16 secs. During the procedure, the patient was hypertensive, and 100 mg of labetalol was required to maintain his systolic pressure below 170 mm Hg so that hemorrhage from the subclavian artery could be minimized. He recovered uneventfully overnight, and there was no further evidence of bleeding. His hemoglobin remained stable at 140 g/L. An additional 1 L of saline and dopamine at 5 g/kg/min for 4 hrs was given on return to the ICU to counteract the residual effect of the labetalol. Subsequent cultures of blood and catheter tip showed negative results. From the Departments of Medicine (JT) and Radiology (MM, LM), Vancouver General Hospital, UBC Site, Vancouver, BC, Canada.

Common iliac artery occlusion with use of Gianturco coils and ethylene vinyl alcohol liquid embolization agent before aortouniiliac stent-graft deployment.

Editor: Some patients who are otherwise candidates for endovascular abdominal aortic aneurysm (AAA) repair cannot be treated with a bifurcated graft because of severe tortuosity, calcification, or occlusion of a common or external iliac artery. An aortouniiliac device can be used in these patients (1), but it is necessary to occlude the contralateral common iliac artery (CIA) to prevent back perfusion of the aneurysm sac. We describe the use of a liquid embolization agent, Onyx (ethylene vinyl alcohol copolymer; Micro Therapeutics, Irvine, CA), in combination with Gianturco steel coils (Cook, Bloomington, IN) to occlude the contralateral common iliac artery before aortouniiliac endovascular AAA repair. An 82-year-old man presented with an asymptomatic 7.0-cm infrarenal AAA. The presence of a 5-cm-long left external iliac artery occlusion precluded insertion of a bifurcated stent-graft. The patient’s left CIA and internal iliac arteries were patent. It was decided to treat the patient with an aortouniiliac stent-graft (Talent; World Medical, Sunrise, FL) and a right-to-left femorofemoral crossover graft. To achieve left common iliac artery occlusion, we elected to use Gianturco coils (Cook, Bloomington, IN) and Onyx (Micro Therapeutics). Permission to use this investigational liquid embolic agent was received from the Health Protection Board of Canada. The procedure was performed under general anesthesia in the angiography suite. After right common femoral artery cutdown, a 5-F Cobra catheter (Cordis, Miami, FL) was directed from the right iliac system into the left common iliac artery. Four 5-cm 8-mm Gianturco coils were deployed just proximal to the internal iliac artery origin. Although this slowed flow within the CIA, the vessel remained patent. Two milliliters of Onyx were then delivered via a coaxial 3-F microcatheter (Rebar 14; Micro Therapeutics) proximal to the Gianturco coils. Total occlusion of the left common iliac artery was achieved (Fig 1). The stent-graft was delivered through an arteriotomy in the right common femoral artery and was deployed in the usual manner without incident (Fig 2). Completion angiography and computed tomography (CT) of the abdomen performed the next day showed complete exclusion of the AAA with no evidence of endoleaks (Fig 3). There was no flow across the embolized left common iliac artery. Follow-up CT at 3, 6, and 9 months after stent-graft implantation continued to show no evidence of endoleak. The 6and 9-month studies demonstrated reduction in AAA size to 6.7 cm and 6.3 cm, respectively. Several mechanical occlusion devices have been used to occlude the contralateral CIA before endovascular AAA repair with an aortouniiliac graft, but a large delivery device is typically used, which requires ipsilateral delivery through an arteriotomy. Patients with an external iliac artery occlusion cannot be treated in this manner. The use of vascular coils to occlude the CIA has been reported, and this can be performed with use of a contralateral approach. However, recanalization through the interstices of coils can occur, and we were concerned that vascular coils would not provide as robust an occlusion as we required. Concern about the possibility of recanalization prompted us to use a liquid embolic agent as well as coils. We believed that use of a liquid agent would limit the risk of recanalization because a liquid agent Figure 1. Digital subtraction angiogram of the left common iliac artery after embolization showed complete occlusion of the artery. The polymerized Onyx has formed a cast of the artery above the Gianturco steel coils. Letters to the Editor