Arterial Interventions: Embolization

1:15 PM

Abstract No. 40

Refractory Bleeding from Gastroduodenal Ulcers: Arterial Embolization in High-Operative-Risk Patients. R. Loffroy, Bocage Hospital, University Hospital Center, Dijon, Burgundy, France 䡠 B. Guiu 䡠 D. Ben Salem 䡠 F. Ricolfi 䡠 J.P. Cerceuil 䡠 D. Krause PURPOSE: We evaluated the efficacy and medium-term outcomes of transcatheter embolization to control massive bleeding from gastroduodenal ulcers after failed endoscopic treatment in high-operative-risk patients. MATERIALS AND METHODS: Retrospective study of 35 consecutive emergency embolization procedures in hemodynamically unstable patients (24 males, 11 females, mean age 71 ⫾ 11.6 years) referred from 1999 to 2006 for selective angiography after failed endoscopic treatment. Several clinical and procedure-related factors were tested independently to assess their influence on primary clinical failure and 1-month mortality rates. Mean follow-up was 26.9 months. RESULTS: Endovascular treatment was feasible in 33 patients and consistently stopped the bleeding. “Sandwich” coiling of the gastroduodenal artery was done in 11 patients, and superselective occlusion of the terminal feeding artery with glue, coils, or gelatin particles was performed in 22 patients. Early re-bleeding occurred in 6 patients and was managed successfully using endoscopy (n ⫽ 2), re-embolization (n ⫽ 1), or surgery (n ⫽ 3). No major complications related to catheterization occurred. Seven patients died within 30 days of embolization and 3 died later during follow-up, but none of the deaths were due to re-bleeding. None of the tested factors had statistically significant influence on primary clinical failure and 1-month mortality rates. No late bleeding recurrences were reported. CONCLUSION: Selective angiographic embolization is safe and effective for controlling life-threatening bleeding from gastroduodenal ulcers, usually obviating emergency surgery in critically ill patients, whose immediate survival depends on their underlying conditions.

Abstract No. 41

Efficacy of Empiric Gastroduodenal Artery Embolization in Patients with Acute Upper Gastrointestinal Hemorrhage. S.A. Padia, Cleveland Clinic, Cleveland, OH, USA 䡠 M.J. Sands PURPOSE: To assess the effectiveness of empirically embolizing the gastroduodenal artery in patients with acute upper gastrointestinal hemorrhage who are refractory to endoscopic therapy. MATERIALS AND METHODS: Over a 4.5 year period, 52 GDA embolizations were performed on 50 patients. Patient charts were retrospectively reviewed with IRB approval waiving consent. Patients were included if an angiogram demonstrated absence of contrast extravasation into a bowel lumen. Patient demographics, etiology of hemorrhage, endoscopic findings, specifics of the embolization procedure, radiologist, complications from the procedure, requirement for further blood products, need for further endoscopy or surgery, length of stay, and death were evaluated. RESULTS: 50 patients underwent empiric GDA embolization for acute upper GI bleeding (2 patients were embolized a 2nd time). All patients (100%) underwent at least one endoscopy prior to angiogram. 8/50 (16%) had bled secondary to an invasive tumor into the duodenum or stomach. 19/50 (38%) had recently undergone surgery and had a post-operative upper GI bleed. 5/50 (10%) patients had negative endoscopic exams. All patients successfully underwent gastroduodenal artery embolization by one of ten interventional radiologists. There was one direct complication from the procedure (large groin hematoma). After embolization, 20/50 (40%) patients did not have further GI bleeding and blood products were no longer required. Two patients did not have further bleeding but required blood products secondary to coagulopathy. Overall success of empiric GDA embolization was 22/50 (44%). Success rates were not statistically different in patients with bleeding secondary to neoplasm or bleeding which was not localized by endoscopy. 13/50 (26%) underwent further endoscopy after embolization secondary to continued bleeding. 11/50 (22%) eventually required laparatomy secondary to refractory bleeding. 30/50 (60%) patients eventually died. CONCLUSION: In patients with acute upper GI hemorrhage which is refractory to endoscopic therapy and who demonstrate no active bleeding on angiography, empiric GDA embolization can be beneficial and has a 44% success rate in preventing further GI bleeding. 1:39 PM

Abstract No. 42

Outcomes after embolization Procedural success Early clinical success Rebleeding Early re-bleeding (⬍ 30 days) Late re-bleeding (⬎ 30 days) Complications Major complications Minor complications Mortality rate Time within one month after one month Cause recurrent bleeding underlying illness

Number of patients 33/35 26/33 33/33 6/33 0/33

Percentage 94.3% 78.8% 100% 18.2% 0%

2/35 3/35

5.7% 8.6%

7/33 3/26

21.2% 11.5%

0/10 10/10

0% 100%

Experimental Study on Ischemic Bowel Changes Induced by Superselective Embolization at the Vasa Recta Level in Dogs. H.J. Jae, Seoul National University College of Medicine, Seoul, Republic of Korea 䡠 J.W. Chung 䡠 Y.H. So 䡠 W. Lee 䡠 J.H. Park PURPOSE: The purpose of our study was to evaluate the degree of ischemic changes of the bowel after superselective embolization of superior mesenteric artery branches at the vasa recta level with N-butyl cyanoacrylate (NBCA) mixtures in dogs. MATERIALS AND METHODS: In a total of 6 adult mongrel dogs, superselective embolization using NBCA mixtures was performed at the vasa recta level in the five isolated S17

FRIDAY

Friday, March 2, 2007 1:15 PM - 3:15 PM Room: 210

1:27 PM

Scientific Sessions

Scientific Session 5 Arterial Interventions: Embolization

branches of the superior mesenteric artery in each dog. All dogs were sacrificed 24 hours after embolization. According to the extent of the NBCA mixtures in the specimen radiographs, embolized segments were divided into group A (embolization of less than 3 vasa recta) or group B (embolization of more than 4 vasa recta). Histologic evaluation was performed by a pathologist. RESULTS: In Group A (n ⫽ 15), the histologic findings were normal in 7 segments (47%). Mild ischemic changes were noted in the mucosal layer in 8 segments, in the submucosal layer in 4 segments and in the muscle layer in one segment. In group B (n ⫽ 15), ischemic changes were noted in the mucosal layer in all 15 segments, in the submucosal layer in 14 segments and in the muscle layer in 10 segments. The difference of ischemic damage between group A and B was statistically significant (Mann-Whitney U test, p ⬍ .05). CONCLUSION: Superselective embolization involving less than 3 vasa recta of the superior mesenteric artery is relatively tolerable, and embolization involving more than 4 vasa recta carries an increased risk of significant ischemic bowel damage. 1:51 PM

Abstract No. 43

Perioperative Endovascular Internal Iliac Artery Occlusion Balloon Placements in the Management of Placenta Accreta. C.H. Tan, Singapore General Hospital, Singapore, Singapore 䡠 K.H. Tay 䡠 K. Sheah 䡠 K. Kwek 䡠 K. Wong 䡠 B.S. Tan, et al. PURPOSE: To evaluate the efficacy of perioperative placement of occlusion balloons within the internal iliac arteries (IIA) in reducing intraoperative blood loss and transfusion requirements during Caesarean delivery, for women with placenta accreta and its variants. MATERIALS AND METHODS: A total of 11 patients (mean age 32 years, range 27 to 37 years), with placenta accreta and its variants, underwent Caesarean delivery with perioperative IIA occlusion balloon placement, at our institution over a 30-month period from January 2004 to June 2006 (study group). The IIA occlusion balloons were inserted via contralateral femoral artery punctures and positioned in the proximal IIA. The balloons were inflated from the time the umbilical cords were clamped following delivery of the baby to just before skin closure at the end of surgery. The intraoperative blood loss, volume of blood transfused and immediate postoperative change in hemoglobin levels of the patients in the study group were compared with 14 similar patients (mean age 35 years, range 29 to 43 years), who underwent Caesarean delivery and hysterectomy but without IIA occlusion balloon placement, over the preceding 36-month period (control group).

S18

RESULTS: Within the study group, 4 patients underwent hysterectomy, 2 had the placentas delivered while 5 had the placentas retained followed by postoperative gelfoam embolization of the placentas. The mean intraoperative blood loss in the study group (2011 mL, range 400 to 5000 mL) was 39.4% less than in the control group (3316 mL, range 1000 to 4000 mL) (p ⫽ 0.042). The mean volume of blood transfused was 52.1% less in the study group (1058 mL, range 0 to 3600 mL) than in the control group (2211 mL, range 1190 to 3980 mL) (p ⫽ 0.005). There was no significant difference in the immediate postoperative change in hemoglobin levels between the two groups (p ⫽ 0.44).

CONCLUSION: Perioperative IIA occlusion balloon placement is a safe and minimally invasive technique that reduces intraoperative blood loss and transfusion requirements in patients with placenta accreta and its variants undergoing Caesarean delivery. 2:03 PM

Abstract No. 44

Spontaneous Bleeding in Patients Receiving Anticoagulation Therapy: Anatomic Correlation and Management by Embolization. F. Youness, University of Iowa Hospitals and Clinics, Iowa City, IA, USA 䡠 S. Sun 䡠 H. Abada 䡠 S. Laroia 䡠 C. Patel 䡠 J. Golzarian PURPOSE: To determine the sites of bleeding in patients on anticoagulation and to correlate these sites with specific feeding arteries. Also, to report the success of transcatheter embolization in these patients. MATERIALS AND METHODS: Over a three-year period, 24 adults on anticoagulation therapy underwent angiographic evaluation for spontaneous bleeding complicated by hemodynamic instability. Angiography demonstrated bleeding via lumbar arteries in eight cases (33%), inferior epigastric arteries in seven cases (29%), thoracic/intercostal arteries in three patients (12%), superior gluteal artery in two patients (8%), and uterine, circumflex iliac, ileolumbar and renal arteries in one patient each (4%). No contrast extravasation or bleeding site could be demonstrated in two patients (8%). All the bleeding arteries were treated by microcoil and/or Gelfoam and/or PVA particles embolization. Anatomically, all bleeding sites were extraperitoneal except in one case (96%). The rectus sheath muscles were involved in eight patients (33%), the retroperitoneum in 10 patients (42%), the chest wall in two patients (8%), the ileopsoas muscles in three patients (12%), vaginal bleeding in one patient (4%), and peritoneal cavity in 1 patient (4%). RESULTS: Microcoil/Gelfoam/PVA embolization successfully controlled bleeding with hemodynamic stabilization in 19 of the 24 patients (79%). Five patients (19%) died within a week after the procedure from multiorgan system failure. Twenty-three of 24 patients had extraperitoneal bleeding (96%). CONCLUSION: Hemodynamically significant spontaneous bleeding in patients on anticoagulation therapy is mainly extraperitoneal. Transcatheter arterial embolization is an effective method of controlling bleeding in these patients. 2:15 PM

Abstract No. 45

Short- and Long-Term Sequelae of Internal Iliac Artery Embolization in Pelvic Trauma. T. Travis, University of California, Davis, Sacramento, CA, USA 䡠 W.L. Monsky 䡠 M. Danielson 䡠 D. Link 䡠 J. London 䡠 J.A. Wegelin, et al. PURPOSE: To assess the short- and long-term sequelae of internal iliac artery embolization in patients with pelvic trauma. MATERIALS AND METHODS: Patients with pelvic trauma who underwent pelvic angiography from January 1994 through February 2006 were identified in the Interventional Radiology database. A retrospective chart review was performed. RESULTS: 112 patients underwent pelvic angiography for trauma; 105 patient records were available for review. 83/

2:27 PM

Abstract No. 46

Massive Abdominal Wall Hemorrhage from Injury to the Inferior Epigastric Artery: A Retrospective Case Review. P.R. Sobkin, University of California San Francisco, San Francisco, CA, USA 䡠 A.I. Bloom 䡠 M.W. Wilson 䡠 J.M. Laberge 䡠 R.K. Kerlan 䡠 R.L. Gordon PURPOSE: To identify the etiology of inferior epigastric artery injury (IEAI) in patients referred to IR and to determine the efficacy of diagnostic imaging and embolization in these patients.

CONCLUSION: At our institution, IEAI is most often an iatrogenic injury in a coagulopathic patient. Patients undergoing paracentesis appear to be at greatest risk and might benefit from a more focused sonographic search for abdominal wall vessels than is routinely provided. CECT can be diagnostic of active bleeding, but is not sensitive enough to preclude angiography in the setting of significant hemorrhage. Embolization is highly effective at controlling hemorrhage, but affected patients may still die from the complications of massive bleeding, emphasizing the need for prevention, where possible. 2:39 PM

Abstract No. 47

Rectus Sheath Hematoma with Hemodynamic Collapse: Management with Embolization. T. Sanghvi, University of Iowa College of Medicine, Iowa City, IA, USA 䡠 F. Youness 䡠 S. Sun 䡠 H. Mimura 䡠 J. Golzarian PURPOSE: To report our experience regarding the management of unstable rectus sheath hemorrhage with selective transcatheter embolization in the largest patient study population from a single institution to date. MATERIALS AND METHODS: 11 patients underwent angiographic evaluation for rectus sheath hematoma complicated by hemodynamic collapse resulting from anticoagulation (n ⫽ 6), end stage liver disease (n ⫽ 1), trauma and paracentesis in patients with pre-existing coagulopathy (n ⫽ 3) and septic thrombophlebitis (n ⫽ 1). Angiography demonstrated active bleeding from inferior epigastric artery in all patients. RESULTS: Microcoil, PVA or Gelfoam embolization successfully controlled extravasation, with rapid stabilization of hemodynamic parameters in all patients, which is reflected by the decrease in transfusion requirements (19% reduction in FFP and 32% reduction in PRBC). All patients survived the immediate post-procedural interval with three patients subsequently requiring a second embolization. Ultimately, two patients died after care was withdrawn.

MATERIALS AND METHODS: A retrospective review of patients referred to IR at three university-affiliated hospitals from 1995-2006. Patients were identified through case log books and the electronic medical record.

CONCLUSION: Selective transcatheter embolization is an effective method for stabilizing hemodynamic collapse associated with rectus sheath hematoma.

RESULTS: 18 IEAIs causing severe hemorrhage were identified in 17 patients. The etiology of arterial injury was paracentesis in 9 (50%), surgical trauma in 3 (17%), blunt or penetrating trauma in 3 (17%), percutaneous drain placement in 2 (11%) and uncertain in 1 (6%). 12/17 patients (76%) were coagulopathic. Of the 9 injuries caused by paracentesis, 6 had ultrasound marking, but only 1 had the presence of underlying abdominal wall vessels specifically assessed. The diagnosis was confirmed by contrast-enhanced CT (CECT) in 11 (62%), tagged RBC scan in 2 (11%), non-contrast CT in 1 (6%) and ultrasound in 1 (6%). In 3 injuries (17%), there was no confirmatory diagnostic imaging. CECT showed active extravasation in 6/11 (55%); whereas, 10/11 (91%) showed active extravasation on subsequent arteriography. 17/18 injuries (94%) were treated with IEA embolization using microcoils and Gelfoam, with an immediate technical success rate of 100% and no complications. 1 of 18 (6%) was managed surgically. Of the 17 injuries treated with embolization, bleeding was controlled

2:51 PM

Abstract No. 48

Comparison of Splenic Trauma Management Techniques: Surgical, Embolic, and Non-Operative Management. J.A. Requarth, Wake Forest University Baptist Medical Center, Winston-Salem, NC, USA 䡠 P.R. Miller 䡠 M.C. Chang 䡠 J.D. Regan 䡠 M.A. Bettmann PURPOSE: This study was undertaken to compare outcomes of different therapeutic approaches to splenic trauma, specifically comparing the morbidity and mortality of surgical management (SM), angiographic management with splenic artery embolization (AM), and non-operative management (NOM). MATERIALS AND METHODS: Records of all patients who were admitted to this level I trauma center between January 1, 2004, and December 31, 2005, were retrospectively reviewed. All patients were evaluated on an intention to treat S19

FRIDAY

CONCLUSION: Data from retrospective review of 85 patients with pelvic trauma undergoing pelvic angiography demonstrated a significant increase in the incidences of pelvic or perineal infection and skin sloughing or necrosis in patients who had undergone IIA embolization. There was no difference between groups in incidence of sciatic or peroneal nerve damage during the admission or pelvic or perineal ulceration after discharge.

in 13, yielding a clinical success rate of 76%. 4 injuries (24%) resulted in death from persistent hemorrhage and multiorgan failure.

Scientific Sessions

105 patients survived to discharge or transfer; 51% had follow-up. Indications for angiography included pelvic fracture with hemodynamic instability in 97/105 and pelvic hematoma in 8/105. 66/105 underwent embolization: 23/66 of bilateral internal iliac arteries (IIA), 18/66 of a unilateral IIA, and 20/66 of selective IIA branches. Embolic materials included Gelfoam in 46/66, coils in 11/66, and both Gelfoam and coils in 9/66. Short-term sequelae were defined as those occurring during the admission of patients surviving to discharge or transfer. Pelvic or perineal infection were significantly more likely in embolized (12/49) versus nonembolized (3/34) patients (p ⫽ 0.05). Sciatic or peroneal nerve damage were equally likely in embolized (4/49) and non-embolized (3/34) patients (p ⬎ 0.05). Skin necrosis or sloughing in non-degloved areas were seen only in patients who had undergone non-selective IIA embolization (4/32) (p ⬍ 0.05). Long-term sequelae were defined as those occurring after discharge. Preliminary data analysis shows no significant difference in incidence of ulceration in embolized (2/24) versus non-embolized (1/18) patients (p ⬎ 0.05). Final data compilation of long-term sequelae will be completed by March 2007.

basis. Information was obtained from the trauma registry, electronic medical records, and radiographic images. Embolization (distal main splenic artery and/or selective branch) was performed only if intra-parenchymal splenic vascular injury was detected at angiography. RESULTS: A total of 280 patients were admitted with the diagnosis of splenic trauma. NOM was used in 142, SM in 112, and AM in 26. The mortality rate was 6.3%, 18.8%, and 7.7%, respectively. There were statistically significant differences in age, injury severity score, ventilator days, ICU days, hospital days, and American Association for the Surgery of Trauma splenic injury grade between NOM and SM groups (p ⫽ 0.0002, ⬍ 0.0001, 0.0003, ⬍ 0.0001, ⬍ 0.0001, and ⬍ 0.0001, respectively) and between NOM and AM groups (p ⫽ 0.037, 0.0203, 0.0097, 0.0041, 0.0116, and 0.007, respectively). Differences were not significant between AM and SM groups (p ⫽ 0.83, 0.23, 0.54, 0.48, 0.86, and 0.20, respectively). Thus, the SM and AM groups were statistically equivalent, and both the SM and AM groups were statistically more injured than the NOM group. Posttreatment abdominal events/complications occured less frequently for the AM than for SM group (p ⫽ 0.0057), and the frequency of post-treatment splenic complications was not statistically different between the AM and SM groups (p ⫽ 0.1644). CONCLUSION: Although this retrospective study does not differentiate between surgical patients who had multiple intra-abdominal organ injuries and those with only splenic injuries, it does suggest that angiographic management ⫾ embolization can be used in selected critically ill splenic trauma patients with fewer abdominal events/complications and equivalent splenic complications. 3:03 PM

Abstract No. 49

Seven-Year Single-Center Experience with Transcatheter Embolization of Splenic Artery Aneurysms. C.D. Donikyan, Mount Sinai Medical Center, New York, NY, USA 䡠 R.A. Lookstein 䡠 P.A. Stangl 䡠 F.S. Nowakowski 䡠 M.L. Marin 䡠 J.L. Weintraub, et al. PURPOSE: Aneurysms of the splenic artery are rare but can present as life-threatening emergencies. Endovascular repair with transcatheter embolization has been reported as an acceptable treatment option. This study reviews a singlecenter experience in the endovascular repair of splenic artery aneurysms with transcatheter embolization techniques. MATERIALS AND METHODS: Between 1999 and 2006, a total of 22 patients (11 male, 11 female; ages 28-79, mean 58) with splenic artery aneurysms were treated by transcatheter embolization. All patients were scheduled for regular follow-up with CTA or MRA at one month, 6 months, 12 months and annually. Patient demographics, technical success, adverse events, and need for secondary interventions were retrospectively reviewed. RESULTS: 19 true splenic artery aneurysms and 3 splenic artery pseudoaneurysms were treated. The aneurysms ranged in size from 1.5-8.3 cm (mean 3.3 cm). Coil embolization was the initial intervention in 21 cases. One patient underwent embolization of a splenic artery pseudoaneurysm with n-bca tissue adhesive. Initial technical success was documented in 100% of patients. No major adverse events occurred related to the embolization procedures. Follow-up imaging was available for all 21 patients with mean follow-up of 12 months (1-63 months). All patients had stable aneurysm sac diameter at follow-up. Reintervention was S20

required in 3/22 (14%) patients for persistent perfusion of the aneurysm sac (1 successful re-intervention with n-bca tissue adhesive, 1 successful repeat embolization with coils, and 1 unsuccessful repeat coil embolization which ultimately required splenectomy). At follow-up 8/21 (36%) of patients had radiographic evidence of splenic infarcts. No splenic abscesses were seen at follow-up. There were no ruptures observed in the entire cohort at follow-up. CONCLUSION: Transcatheter embolization is a safe and effective method for the treatment of splenic artery aneurysms. The procedure has a high technical success rate with small percentage of patients requiring secondary interventions.

Scientific Session 6 Venous Interventions: Recanalization and Insufficiency Friday, March 2, 2007 1:15 PM - 3:15 PM Room: 205 1:15 PM

Abstract No. 50

Aspiration Thrombectomy for Acute Deep Venous Thrombosis of the Lower-Extremity: Preliminary Results. J.H. Won, Ajou University College of Medicine, Suwon, GyeongGi-Do, Republic of Korea 䡠 G.-S. Jeon 䡠 H.S. Lee 䡠 D.M. Soh PURPOSE: To evaluate the feasibility of aspiration thrombectomy as the sole treatment for acute deep venous thrombosis (DVT) of the lower-extremity. MATERIALS AND METHODS: Twenty-seven patients (18 women, 9 men; mean age, 54 years) with acute DVT of the lower-extremity underwent manual aspiration thrombectomy using a 9 or 10-F guiding catheter. The time from the initiation of symptoms to the procedure ranged from 1-28 days (mean, 8.3). Heparinization was started 0 to 81 hours (mean, 27.4 hours) before the procedure. IVC filter was inserted in 25 patients prior to thrombectomy. The aspirate was filtered and the amount of aspirated blood separated from the clot was measured. Adjuvant thrombolysis with urokinase was performed when the aspirated blood was over 400 mL or when the aspiration thrombectomy was incomplete. RESULTS: In 23 patients (85.2%), aspiration thrombectomy alone successfully restored the patency of the thrombosed veins. In 4 patients, additional thrombolysis was done using urokinase (mean dose, 1.75 million IU). Symptomatic relief was achieved in all patients. Amount of aspirated blood during the procedure ranged from 85 to 350 cc (mean, 210 cc) with decrease of mean hemoglobin level from 11.4 ⫾ 2.1 to 10.2 ⫾ 1.75 g/dL. Venous stenosis was noted in 24 patients and was at left common iliac vein (n ⫽ 20), left common femoral vein (n ⫽ 2), and right common iliac vein (n ⫽ 2). Balloon angioplasty alone (n ⫽ 4) and additional stent placement (n ⫽ 20) were performed after the thrombectomy. Total procedure time excluding the time for urokinase infusion ranged 30-105 min (mean, 63.5 min). Fever developed after the procedure in two patients, but subsided with antipyretics within 3 days. No hemorrhagic complication occurred.