Angiography

Scientific Session 26 Gastrointestinal/Angiography Thursday, April 11, 2002 1:00 PM - 2:30 PM Moderator(s): Jeet Sandhu, MD 1:00PM

Abstract No. 201

Angiographic Evaluation of Patients with Chronic Gastrointestinal Bleeding of Obscure Origin. E.E. Williamson, Mayo Clinic, Rochester, MN, USA oJ.c. Andrews ° M.G. Sarr PURPOSE: I) To determine the efficacy of contemporary digital angiography in detecting the source of chronic or recurrent gastrointestinal bleeding in patients with prior extensive negative evaluation; and 2) to compare our diagnostic efficacy with previous published reports performed with filmscreen angiography. MATERIALS AND METHODS: Angiographic results were reviewed for 102 consecutive patients referred to radiology from January 1996 to June 200 I for the evaluation of chronic or recurrent gastrointestinal bleeding. These results were correlated with surgical and pathological reports where available. Patients with a known source of bleeding were excluded from the study. RESULTS: There were 48 women and 54 men, with a mean age of 59.4 years. Angiographic abnormalities thought to represent a likely source of bleeding were identified in 39/1 02 (38.2%). In the remaining 63 patients, angiography was negative. Surgery was performed in 25 cases following positive angiographic fmdings. Surgical findings included 11 cases of angiodysplasia, 5 cases of arteriovenous malformation, 2 cases ofMeekeIs diverticulum, 2 cases ofdilated submucosal vessels not meeting the pathologic criteria for angiodysplasia, 1 case of cecal divelticulosis and I case of ulcerative colitis. In 3 cases, no lesion was found to correspond with the angiographic abnormality. In the 14 cases that did not proceed to surgery, angiogmphic findings were angiodysplasia (3 cases), arteriovenous malformation (3 cases), nonspecific focal hypervascularity within the colon (3 cases), venous malformations (2 cases), and jejunal artery aneurysm, hepatic artery aneurysm, and mesenteric carcinoid metastasis (I case each). CONCLUSION: I) Angiographic evaluation using contemporary digital imaging techniques established a diagnosis in 35% of patients referred for chronic gastrointestinal bleeding of obscure origin. 2) Our overall diagnostic yield is similar to that seen in prior published studies performed using screen-film techniques. 1:11 PM

Abstract No. 202

Intraarterial TPA Provocation for Localization of Gastrointestinal Hemorrhage. A.J. Beyer III, UniverSity a/North Carolina at Chapel Hill, Chapel Hill, NC, USA ° D. Dascal ° P. Jaques ° 1. Sandhu ° s.M. Weeks oM. Mauro PURPOSE: Localization of intermittent GI bleeding remains problematic. Radionuclide studies and visceral angiography are often unremarkable despite clinical evidence of active hemorrhage. To increase the yield of these studies we have retrospectively investigated the use of catheter directed TPA administration prior to and during nuclear medicine and angiograpruc evaluation of gastrointestinal hemorrhage.

MATERIALS AND METHODS: From 1999 to July 2001,63 patients underwent visceral angiography to localize a source of GI hemorrhage. Of this group, 14 patients underwent catheter directed TPA provocation. TPA was administered directly into the SMA, lMA or both vessels based on the clinically suspected site of bleeding. All patients were systemically heparinized while TPA was administered at doses ranging from I mg/hr - 5-mglhr. Following institution of the TPA infusion most patients were transferred to nuclear medicine for bleeding scans. As soon as the bleeding scan became positive, the patients returned for repeat arteriography. Alternatively, early in our experience, no nuclear medicine examination was obtained and the patients underwent repeat arteriography after one hour ofTPA administration. RESULTS: All 14 patients undergoing TPA provocation had a normal bleeding scans or visceral angiogmm. Following TPA provocation, bleeding scans were positive in 8/14 patients (57%). Of these 8 patients, angiography revealed a specific site of bleeding in 5 (36%). Four of these patients underwent embolization and one was treated surgically. Ofthe remaining 3 patients with positive nuclear medicine scans and negative angiograms, one was treated with empiric transcatheter embolization, one was treated at EGO and one underwent surgical therapy. None of these 8 patients have re-bled. Additionally, none of the six patients with negative studies following TPA provocation have re-bled. CONCLUSION: Catheter directed intraarterial TPA administration for provocation of intermittent GI bleeding significantly increases the yield of positive nuclear medicine and angiographic studies, thereby assisting in determining the appropriate therapy for these patients. 1:22PM

Abstract No. 203

Embolization of Canine Intrahepatic Portosystemic Shunts: A Model for Treatment of Human Portosystemic Malformations. J.A. Solomon, Hospital a/the University a/Pennsylvania, Philadelphia, PA, USA ° C. Weisse oM. Nicholson ° D. Holt ° C. Cope PURPOSE: Five dogs with presumed intrahepatic portosystemic shunts were studied to assess the feasibility of endovascular treatment. MATERIALS AND METHODS: Five encephalopathic dogs were found to have portosystemic shunts by ultrasonography. Shunts were accessed by a jugular approach and venography performed to delineate the anatomy. Pressure measurements were obtained in the right atrium and portal vein with and without temporary occlusion of the shunts. Depending on changes in portal vein pressures, shunts were occluded immediately or in multiple sessions with a combination of stents, stent-grafts, and coils. Dogs were followed clinically as well as with ultrasonography and liver function tests for up to five months. RESULTS: Angiography located the shunts in all dogs. Ofthe lesions seen on ultrasound, one was extrahepatic, located between the splenic vein and IVC. Intrahepatic shunts displayed heterogeneous anatomy and size. Two were large, direct communications between the portal and hepatic veins and were associated with hypoplasia ofthe intrahepatic portal vein. One was complex, with multiple intra and extm hepatic cornrnunications. The final one was a small cornrnunication between the portal vein and intrahepatic vena cava. Endovascular treatment was attempted in four dogs, the extrahepatic shunt was managed surgically. Up to three sessions were required to achieve occlusion. Technical success was 100%. Two dogs improved based on clinical, radiographic

869

and laboratory criteria. One dog failed to respond despite successful occlusion of the shunt and was euthanized. The fmal dog was clinically stable after the first part of a staged procedure.

CONCL USION: Congenital intrahepatic portosystemic shunts are an exceptionally rare vascular anomaly in humans. Such shunts occur more commonly in dogs than in humans but have a similar clinical course and are associated with comparable histopathologic changes in the liver. Surgical ligation is the treatment ofchoice in dogs, but it entails a complex procedure with signifIcant morbidity and mortality. Endovascular techniques may be used to occlude canine intrahepatic shunts. The principles utilized should help guide treatment of similar lesions in humans.

1:33PM

Abstract No. 204

Catheter Directed Intra-Arterial Therapy for Graft Versus Host Disease. A.l. Bloom, Hadassah University Hospital, Jerusalem, Israel· A. Verstandig· r Sasson· Y Kori • 1. Resnick· M.Y Shapira PURPOSE: To evaluate intra-arterial administration of Methylprednisolone and Methotrexate for the treatment of severe graft versus host disease [GVHDj in patients after bone marrow transplantation [BMT). MATERIALS AND METHODS: Fifteen consecutive patients with grade III-IV GVHD involving the liver or intestine, were prospectively enrolled in a pilot study to receive intra-arterial therapy in addition to standard therapy. Local IRB approval was obtained. A minimum pre-existing or corrected platelet count of 40x I09 and a normal or corrected prothrombin time were required. The dominant organ involved by GVHD was targeted for therapy. Therefore the following arteries were selectively catheterized; the PHA/RHNLHA (liver), the SMAlIMA (intestine), and the GDA (stomach). For treatment of the liver, 75mg/m2 Methylprednisolone was injected into the PHA or divided 2/3 to the RHA and 1/3 to the LHA. This was followed by Methotrexate I Omg/m 2 . Drugs were administered over three minutes to facilitate even drug distribution and maximum extraction. For treatment of the intestine or stomach, 60mg ofSolumedrol was injected slowly into each vessel (SMAlIMNGDA). Hemostasis was obtained by direct manual compression. Response was defined as complete, partial or none. A subjective decrease in the quantity and frequency ofdiarrhea was measured for intestinal disease. Decreased serum bilirubin level was the measure of response for liver disease. The procedure was repeated in patients who had a partial response.

S70

RESULTS: Eight male and seven female patients, age range 750 years, median 25 years, underwent 25 procedures (range 1-3). Pre-procedure transfusion of fresh frozen plasma was performed in four cases, platelets in four and cryoprecipitate in three. The intestine was treated in eight patients, the liver in three, and both in four. The stomach was also treated in two patients with severe gastrointestinal symptoms. Variant anatomy, that required selective catheterization of the LHA and RHA for drug administration, was identified in three patients. A replaced RHA from the SMA was seen in one, trifurcation of the CHA in one, and an accessory LHA in one. In two cases prominent SMA to GDA collateral flow via inferior pancreaticoduodenal arteries was identified. The catheter tip was placed distally in the SMA to prevent unintentional administration of drug to the liver. There were two minor technical complications. Vasospasm occurred during catheterization of the LHA (n=l) and of the IMA (n=I). Administration of intra-arterial nitroglycerine resolved the

spasm in the lMA. In the LHA the vasospasm persisted and 1/3 of the planned drug dose to the liver could not be given. Therefore only the right lobe ofthe liver was treated. Response in intestinal disease was complete in eight patients, partial in three, and one patient did not respond. In liver disease, serum bilirubin levels decreased to normal in three patients. A partial response was seen in two and no response in two (bilirubin elevation was noted in the first four weeks of follow up).

CONCLUSION: Intra-arterial therapy for severe GVHD of the intestine and liver appears to effective and safe. It may improve the overall results of treatment in high risk patients. The response of the intestine was superior to that of the liver. Variant arterial blood supply requires a modified technique. Transfusion of blood products in patients with coagulopathy may prevent hemorrhagic complications. Other organs affected by GVHD may benefit from intra-arterial therapy. The technique may also have applications in the treatment of autoimmunity and solid allograft rejection.

1:44PM

Abstract No. 205

Tunneled Peritoneal Catheter Placement for Drainage of Malignant Ascites: A Simplified Percutaneous Approach. rD. Barnell, FF Thompson Hospital, Canandaigua, NY, USA .1. Rubins PURPOSE: To develop a simple, minimally invasive method of draining symptomatic malignant ascites which would allow patients to avoid repeated hospital visits for paracentesis. MATERIALS AND METHODS: Since 1996,28 patients with symptomatic malignant ascites have undergone percutaneous placement of a 2- or 3-cuff multi-sidehole l5-F 42-cm soft silicon rubber catheter (Tenckhoffcatheter) using a simplified tunneling technique. A single skin puncture is performed with an 18-g spinal needle which is advanced in the subcutaneous tissue parallel to the skin. This is followed by passage of a curved 22-g Chiba needle through the·18-g needle and into the peritoneum. There is rio need for a separate incision or a tunneling tool. Ultrasound and fluoroscopic guidance are used. After guidewire placement, the tunnel is dilated to 20-F or, occasionally, 22-F. The peritoneal entry site need only be dilated to l6-F as the cuffs of the Tenckhoff catheter will be passed only into the tunnel and not into the peritoneum. A peel-away sheath is not needed. The Tenckhoff catheter is stiffened with a lubricated 7-F pigtail catheter and advanced over the wire. After catheter placement, gravity drainage is used as often as needed by simply removing the cap. Vacuum bottles are not required. RESULTS: Catheter placement .was successful in all 28 patients. All patients achieved significant symptomatic improvement. Twenty-two of the 28 were outpatients and were able to leave the hospital within several hours of placement. In one patient a new catheter was placed on day 6 because the original one was in a loculation and was draining poorly. Inpatients remained in the hospital for reasons unrelated to catheter placement and most left within several days. Only one patient died in the hospital during the same admission. None devetoped peritonitis. Of the 26 patients who have died, the average survival time was 49 days (9-218 days). All of the deaths were attributable to the patients' underlying malignancy. CONCLUSION: Percutaneous placement of a tunneled catheter using curved and straight coaxial needles is a safe, simple, and effective method for palliative drainage of malignant ascites which allows patients to avoid return hospital visits and to go home quickly.

1:55PM

Abstract No. 206

Fluoroscopic-Guided Balloon Dilation in Corrosive and Postoperative Strictures: The Effectivenss and LongTerm Results. Y. Han, Chonbuk National University Hospital, Chonbuk, Korea· G. Jin • H. Chung PURPOSE: To evaluate the effectiveness and long-term results in the balloon dilation in patients with corrosive and postoperative anastomotic strictures. MATERiALS AND METHODS: Fluoroscopically guided esophageal balloon dilation was performed within 12 years in 34 corrosive strictures and 43 postoperative strictures. In the 34 corrosive stricture, 13 were men and 21 were women and age ranged from 2 to 76 years (mean 41). In the 43 postoperative anastomotic strictures, 36 were men and 7 were women and age ranged from 3 months to 76 years (mean 58). We evaluated frequency ofballoon diJation and balloon session, symptomatic improvement rates, and complication rates between two groups. We analyzed the results statistically by student's paired t-test. RESULTS: The 314 balloon dilations/182 balloon sessions were performed in all patients (174/98 in corrosive strictures, 140/84 in postoperative anastomotic strictures). The symptomatic improvement showed 20/34 (58%) in corrosive and 36/43 (84%) in postoperative strictures. The recurrence of intermittent dysphagia showed 7/34 in corrosive and 5/43 postoperative strictures. Esophageal ruptures occured in 14 (intramural perforation 9, transmural perforation 5: 13 in corrosive, I in postoperative strictures), and were treated with fasting, parenteral alimentation, and antibiotics. Two transmural perforations in corrosive strictures· required operation. The long-term follow-up in 27 corrosive and in 41 postoperative strictures was 13-144 months (mean: 58) and 3-96 months(mean: 27). The balloon session and number were statistically significant between postoperative (1.95 ± l.05/ 3.26 ± 1.87) and corrosive strictures (2.88 ± 1.67/5.12 ± 3.03) (P=0.004,P=0.0014). The balloon session and number was not statistically significant between upper and lower (P=0.17/ 0.08) and. between sewing type and stapler type (P=0.28/ 0.64) in postoperative strictures. CONCLUSION: Fluoroscopic-guided balloon dilation in postoperative anastomotic strictures is safe and effective with good long-term results compared to corrosive strictures. The balloon dilation in cOITosive strictures is effective, but with high recurrence rates. In the corrosive strictures the high rupture rate indicated the need for extra caution.

2:06PM

Abstract 0.207

Results of a Multicenter Clinical Trial with a Covered, Self-Expandable Nitinol Esophageal Stent. 1.E. Lopera, Louisiana State University, New Orleans, LA, USA·o. Alvarez· F Neira· 1. C. Hidalgo' 1.H. Patino' WR. Castaneda PURPOSE: The results ofa multicenter clinical trial with the use ofa self-expandable, covered nitinol stent (Song esophageal stent) for the palliation ofmalignant esophago-gastric tumors are reported. MATERiALS AND METHODS: A prospective, multicenter trial was conducted between June 1999 and March 2001. Sixty-nine patients (mean age 67-years, range 39-86) with malignant esophago-gastric tumors, were treated with fluoroscopic placement ofSong esophageal stents. All patients had inoperable disease and severe dysphagia. Seven patients also presented esophago-respiratory fistulas. Location of the

strictures included: cervical esophagus (n=3), middle esophagus (n=19), distal esophagus and cardias (n=44), esophago-enteric anastomosis (n=3). The stents were placed using a 24 Fr flexible introducer system.

RESULTS: The technical success was 100%. Early «30 days) complications were observed in 16 % of the patients:. sev~re pain (n=6), perforation requiring surgery (n=2), stent ffiJgralion (n=2), aspiration pneumonia (n=I). Late complications were seen in 14%: tumor overgrowth (n=3), torsion/collapse ofthe stent (n=2), food obstruction (n=2), bleeding (n= I), stent migration (n=2). The 30-day mortality was 7%. Death related to stent placement occurred in 1.5%. The esophago-respiratory fistulas were successfully sealed in all seven patients. The average dysphagia score decreased from 3.42 to 1.06 (p<0.000 I). CONCLUSION: Fluoroscopic insertion of Song esophageal stent is a safe and effective method of palliation for patients with inoperable esophago-gastric malignancies, significantly improves dysphagia and is successful in sealing fistulas.

2:17PM

Abstract No. 208

Esophageal Balloon Dilatation under Fluoroscopic Guidance in Children - Experience of 172 Dilatations. 1. Amaral, The Hospitalfor Sick Children, Toronto, ON, Canada' P.G. Chait· B.L. Connolly· M.J. Temple' R. Restrepo· P. Marcon, et al. PURPOSE: Esophageal balloon dilatation under fluoroscopic guidance in children is a well described procedure. It can be used to treat esophageal strictures of different etiologies. The purpose of this paper is to report our technical experience, success rate and complications on esophageal dilatation done by IR in a tertiary care pediatric centre. MATERIALS AND METHODS: Retrospective analysis of 172 records offluoroscopic guided esophageal dilatations done between October 1992 and September 2001 was performed. The total number of patients who had dilatations was 72. Forty-six (64%) were males and 26 (36%) were females. The cause and site of stricture, size of balloons used, incidence of post procedure complications and number of stent placements were reviewed. RESULTS: The two most common causes of stricture were achalasia (32%) and anastomotic stricture post esophageal atresia repair (30%). There were 5 patients with epidermolysis bullosa (7%). Other causes included lye ingestion, reflux and fundoplication. The dilatations were performed using balloons ranging from 3 mm to 40 mm in diameter. The majority of the patients had one (41 %) or two (33%) dilatations. One patient had 12 procedures. 47% (30) ofthe dilatations were performed in the distal esophagus. 98% (169) of the dilatations were successful. Out of 172 dilatations, no major complications such as mediastinitis or pneumomediastinum were noted. Four patients (2.3%) had minimal contrast aspiration during the procedure and 4 patients (2.3%) demonstrated contrast leakage in the area of the stricture after the procedure. Two (2.7%) patients had stents placed in the esophagus. CONCLUSION: Image guided esophageal balloon dilatation and stent placements are safe procedures to treat strictures of different etiologies in children.

S71