Transrectal Endoscopic Ultrasound (TRUS) Elastography in Inflammatory Bowel Disease (IBD

Abstracts

Table 1 OVERALL

Suction

No Suction

p value

Bloody gross appearance High cellularity Diagnostic yield PANCREAS Bloody gross appearance High cellularity Diagnostic yield LYMPH NODES Bloody gross appearance High cellularity Diagnostic yield

72% 50% 78%

50% 59% 81%

0.002 0.592 0.616

61% 57% 81%

40% 51% 85%

0.026 0.551 0.636

85% 56% 70%

67% 69% 74%

0.111 0.305 0.761

W1256 EUS-Guided Selective Embolization of Splenic Vein and Its Branches: A New Therapeutic Option for Variceal Bleeding Sergey V. Kantsevoy, Xavier Dray, Gianfranco Donatelli, Vihar Surti, Richard Ducharme, Devi Mukkai Krishnamurty, Ronald J. Wroblewski, Samuel A. Giday, Andrew Zhigalin, Elena Dubcenco, Jonathan M. Buscaglia, Eun Ji Shin, Paul J. Thuluvath Background: Current endoscopic therapy for variceal bleeding (banding, sclerotherapy) aims at occluding the varix at the site of bleeding. Such therapy does not affect portal blood flow towards the bleeding varix, so even after successful endoscopic hemostasis the risk of rebleeding is not eliminated. Although endoscopic therapy is very successful for control of esophageal variceal bleeding, banding or sclerotherapy is not effective for gastric varices. Aim: To evaluate the feasibility of EUS-guided selective embolization of the splenic vein and its branches delivering portal blood to the gastric varices in a live porcine model. Methods: Linear array echoendoscope (Olympus optical Ltd, Tokyo, Japan) was introduced into the stomach. Under EUS guidance portal vein was punctured with 19g EUS needle (Cook Endoscopy, Winston-Salem, NC) and a guidewire was directed through the needle into the portal vein. Triple lumen sphincterotome was then advanced over the guidewire through portal vein into splenic vein under fluoroscopic guidance. Splenic vein was occluded with 2 occlusion coils (Cook Medical, Bloomington, IN). The first animal was sacrificed after the experiment, the subsequent 6 animals were survived for 2 weeks and then sacrificed. Results: EUSguided portal vein puncture and selective canulation of the splenic vein were easily achieved in all animals. Deployment of the embolization coils was technically easy with total procedure time 53.0  21.7 minutes. There were no problems or complications during these experiments. Immediate post-procedure necropsy (pig #1) and postmortem examination in 2 weeks (pigs #2-7) confirmed the presence of coils with occluding clot in the splenic vein and did not reveal any bleeding or other intraperitoneal complications, Conclusion: EUS-guided selective canulation and embolization of the splenic vein and its branches is feasible, technically easy and may become a new therapeutic option for prevention and control of variceal bleeding.

W1257 Development of a Prototype Trocar for Translumenal Therapeutic Interventions Using EUS Shyam Varadarajulu, Kenneth F. Binmoeller, John Lunsford, Hoang G. Phan, Fiona Sander Introduction: For EUS-guided therapeutic interventions such as pancreatic pseudocyst drainages, current techniques rely on the use of a 19 gauge FNA needle to gain initial access across the GI lumen. However, advancement of catheter-based devices through the incision created by these needles is technically difficult thereby necessitating progressive dilation of the lumenal wall prior to undertaking endotherapy. The EUS 2008 Working Group recommended development of specific devices to facilitate easy endotherapy. An echoendoscope-compatible prototype 19 gauge trocar was designed to create a tissue puncture, significantly larger than its own profile, to facilitate easier passage of larger catheter-based devices for one-step endotherapy. Aim: This study compared a standard 19 gauge FNA needle to three versions of a prototype (‘‘Extending Blade’’) trocar based on: (1) the force required to puncture porcine GI lumenal tissue, and (2) the force required to subsequently pass large caliber catheters through the resulting puncture site. Methods: Ex-vivo porcine stomachs were used to compare puncture forces of a 19 gauge EUS-FNA needle with 1, 2 and 3-blade versions of the prototype trocar passed via the biopsy channel of an echoendoscope. Subsequently, the force required to pass a 2.2 mm dilating catheter and a 6mm balloon catheter through the punctures created by the FNA needle and prototype trocars were compared. Results: The force required for initial puncture of the tissue using all three prototype trocars was equal to that applied using a 19 gauge FNA needle. The force required to advance the 2.2 mm dilating catheter through the tissue punctured using the prototype trocars was 8

AB324 GASTROINTESTINAL ENDOSCOPY Volume 69, No. 5 : 2009

times (range, 6-11) less than the force required to advance them via the 19 gauge needle puncture site. Also, the force required to advance the 6mm balloon catheter through the tissue punctured by the prototype trocars was 9 times (range 7-11) less than the force required to advance them via the 19 gauge needle puncture site. Conclusion: Initial access created by the echoendoscope compatible prototype trocar obviates the need for progressive dilation of the GI lumenal wall and could potentially facilitate easier one-step translumenal endotherapy.

W1258 Transrectal Endoscopic Ultrasound (TRUS) Elastography in Inflammatory Bowel Disease (IBD) Nadan Rustemovic, Silvija Cukovic-Cavka, Davor Radic, Milorad Opacic, Zeljko Krznaric, Irena Hrstic, Roland Pulanic, Boris N. Vucelic Background and Aims: Establishing the diagnosis of Crohn’s disease (MC) or ulcerative colitis (UC) sometimes is very difficult. When IBD is confined to the colon, there is a lack of diagnostic tools for distinction between Crohn’s colitis and ulcerative colitis, which is especially important in the definitive phenotyping before surgical decision. The aim of this study was to assess the potential role of the TRUS elastography in distinction between MC and UC. The idea is based upon the fact that MC is transmural disease, and UC is limited to the mucosa and submucosa. These tissue characteristics are reflected in differences of the elasticity in rectal and perirectal tissue. Changes in the tissue elasticity can be obtained qualitatively by elastography with different coloures (from red-soft tissue to blue-hard tissue) or quantitatively using strain ratio score. Methods and Results: Rectal wall thickness and elastomode of patients were measured by TRUS elastography. Endoscopist was blind for patient diagnosis. SPSS ver. 17 was used for statistical analysis. In pilot study we included 31 patients; 16 patients (52%) with MC and 15 patients (48%) with UC. Average thickness of rectal wall in all study patients was 6.43 mm ( 0.47 SE). In MC group mean rectal thickness was 7.28 mm ( 0.76 SE) compared to 5.52 mm ( 0.44 SE) in UC group. There was no statistical significant difference between MC and UC groups in perirectal thickness (tZ1.97, dfZ29, pZ0.058). Perirectal elastomode showed statistically significant difference between this two groups (c2Z18.6, dfZ2, P!0,001). Twelve (75%) patients in MC group had hard elastomode compared to none of patients in UC group, meaning that hard elastogram had positive predictive value of 100% for patients with MC. We also evaluated strain ratio (SR) of rectal tissue. Strain ratio is ratio of strain between two regions of interest (ROI) in the same image. Mucosal tissue was used as first ROI and perirectal tissue as second. SR was measured 3 times and middle value was used in statistical analysis. Sixteen patients were included in our study; 7 (44%) with MC and 9 (56%) with UC. Mean value of SR was higher in MC group (1.07  0.26) then in UC group (0.26  0.21). This difference was statistically significant (tZ6.85, dfZ14, p!0.001). Conclusion: TRUS elastography provides a valuable information regarding the stiffness of the rectal and perirectal tissue, and can help to differentiate MC from UC. This is a promising new diagnostic tool in the field of IBD. Our study is ongoing and we expect improvement of the method, with increasing number of patients.

W1259 Accuracy of Contrast-Enhanced Harmonic Endoscopic Ultrasound (CEH-EUS) Using DefinityÒ Joseph Romagnuolo, Brenda Hoffman, Stacie A. Vela, Robert Hawes, Shivakumar Vignesh Background: EUS-FNA has limitations in cancer diagnosis and staging. Mucin-filled (mediastinal) cysts can mimic solid lesions. Improvements in contrast agents, transducers and processors have enhanced the potential of CEH-EUS. Aim: To determine optimal settings and accuracy of CEH-EUS using DefinityÒ (Lantheus Medical Imaging, Inc), a prototype linear echoendoscope (XGF-UC180, Olympus) and Aloka ProSound Alpha10 processor, in a pilot. Methods: Patients with esophageal/pancreatic/liver tumors, or adenopathy, without contraindication to EUS or DefinityÒ, were recruited. EKG monitoring was used. DefinityÒ, a 2ndgeneration perflutren lipid microsphere contrast agent, FDA-approved for cardiac imaging, was given (10 uL/kg IV in 1-2 doses) after agitation. Mechanical index (MI) was optimized over 5 cases. Intermittent (1-sec) and continuous imaging was used with extended pure harmonic detection. Perfusion factors were noted (sequence, pattern, washout). Video sequences were recorded (arterial, venous and post-venous phases). Patients without positive tissue were followed for 6 months. Pre-/post-contrast predictions of neoplasia were recorded (5-point Likert scale). Results: An MI of 0.3 appeared optimal: non-linear oscillations without significant disruption, allowing parenchymal (liver/pancreas) perfusion imaging. CEH-EUS took !5 min/site. Thirty sites were imaged in 21 patients (48% male), including 7 nodes and 23 masses (16 pancreatic, 3 hepatic, 2 mediastinal, 1 adrenal and 1 submucosal); 21 sites underwent FNA and 5 were excised in follow-up. CEH-EUS showed a (very) likely benign lesion (diffuse enhancement) in 12, (very) likely neoplastic lesion (perfusion defects) in 18, and 1 indeterminate node. No EUSbenign sites were CEH-EUS-neoplastic; 1 EUS-neoplastic site was CEH-EUS-benign (final diagnosis: likely benign). 4 were EUS-undecided (2 with nonspecific on-site cytology): CEH-EUS appeared correct in 3 and incorrect in 1. Positive and negative predictive values for CEH-EUS were 86.7% and 87.5%. CEH-EUS appeared to

www.giejournal.org