*3467 A NEW DEVICE UTILIZING OPTICAL BIREFRINGENCE FOR THE IMMEDIATE DETECTION OF BILIARY MICROLITHIASIS (BM) DURING ERCP Raman Muthusamy, Univ of CA, San Francisco, San Francisco, CA; Karen Lauer, Chris Darrow, Lawrence Livermore National Lab, Livermore, CA; Andrew Mirhej, Digest Disease Clin, Colorado Springs, CO; David R. Prince, Wayne Adkisson, Uma Mahadevan, Alex Monte, Roberta Caderra, ,fames W. Ostroff, Univ of CA, San Francisco, San Francisco, CA Background: Biliary microlithiasis (BM) is strongly implicated in the etiology of idiopathic recurrent pancreatitis. Microlithiasis is diagnosed by the examination of bile under visible light microscopy (VLM), using a polarized filter. Positive results are based on the identification of BM crystals, which are composed of cholesterol monohydrate (CM) and calcium bilirubinate (CB) in varying proportions. We describe a new device, named a flow lithometer (FL), that detects CM crystals on the basis of optical birefringence and compared these results to those obtained by VLM. Methods: After obtaining Institutional Review Board approval, bile samples were collected from the common bile duct during endoscopic retrograde cholangiopancreatography (ERCP) in 47 patients (18 s/p cholecystectomy). Within one hour of collection, the sample was divided equally and analyzed by the FL and VLM. After the flow lithometer had been calibrated with a standardized cholesterol solution, samples were passed through a glass capillary tube that was placed between crossed polarizing filters. The passage of birefringent particles through a light beam (620 nm) generated electrical impulses at the detector. Pulses above a certain threshold were counted electronically, using LabView software, to quantify the extent of BM. Results: The FL agreed with VLM in 29/47 cases (9 positive BM, 20 negative). Using the FL, birefringent material was detected in 15 samples in which VLM did not detect crystals. In 3 samples, VLM reported crystals not seen by the FL. At ERCP, 7 patients had stones detected. The FL detected 6/7 (86%) of these patients, compared with 4/7 (57%) with VLM. Overall, VLM detected 12 patients with BM (2 CM, 9 CB, 1 CB+CM). The FL identified 9/12 of these patients (3/3 ~vith CM, 7/10 with CB). When compared to VLM, the FL yielded a sensitivity, specificity, and accuracy of 75%, 57%, and 61% respectively. Conclusions: 1) VLM may miss patients with microlithiasis and is dependent upon the individual observer. 2) The flow lithometer offers objectivity and may provide improved sensitivity and efficiency in detecting BM compared to VLM. 3)The incidence of false-positives with the flow lithometer (compared to VLM) cannot be determined at present due to the short follow-up period. 4) Further investigation and followup in this ongoing study will characterize the optical properties of CB and other potential factors affecting the performance characteristics of this device.
*3468 DIAGNOSIS OF HISTOLOGICAL GASTRITIS USING MAGNIFYING ENDOSCOPY Sonichi Nakagawa, Mototsugn Kate, Kaku Hokari, Mitsuru Kawarasaki, Koji Oohira, Jun Ishizuka, Takuji Mizushima, Yoshito Komatsu, Hidetoshi Kagaya, Toshiro Sugiyama, Masahiro Asaka, Hokkaido Univ Sch of Medicine, Sapparo Japan Aims; Collecting venulae, which are a microvascular system in mucosa, can be observed using magnifying endoscopy as a regular arrangement of collecting vascular spots. There is no report on the relationship between H. pylori infection and visibility of collecting venulae. The purpose in this study was to determine the usefulness of magnifying endoscopy for the diagnosis of H. pylori -induced histological gastritis.Methods; Fifty-eight patients who were scheduled to undergo routine endoscopic examination were enrolled. Informed consent was obtained from all of the patients.
VOLUME 53, NO. 5, 2001
After routine endoscopic examination using magnifying endoscopy (GIFQ240Z, Olympus Co. Ltd., Tokyo), magnified observation was performed on three sites of the stomach, greater curvature of the antrum and the lower corpus,and lesser curvature of the lower corpus. The results of visibility of collecting venulae were classified into the following three groups: R group of collecting venulae that showed a regular pattern, I group of collecting venulae that showed an irregular pattern, and D group of collecting venulae that were disappeared. Histological assessment of the observed sites by magnifying endoscopy was performed using an Updated Sydney System. Five morphological parameters were assessed using the Sydney System and graded from 0 to 3. Results; The rates ofH. pylori infection were 0% in the R group, 69.4% in the I group and 75.0% in the D group. H. pylori infection rate in the R group was significantly lower than those in the I and D groups (p<0.01). The scores of all five morphological parameters were significantly lower in the R group than in the I and D groups. (p<0.01) The score of atrophy in the I group was significantly higher than that in the D group. (p<0.01) Conclusion; Visibility of collecting venulae on gastric mucosa is influenced by H. pylori -induced histological gastritis. Magnifying endoscopy is useful for the diagnosis of histological gastritis. H. pylod infeclion rate R group I group D group
Antrum G.C. 0% (0/8) 66.7% (14/21} 65.5% (19/29)
Corpus L.C. ~o (0/16) 73.9% (17123) 78.9% (15/19)
Corpus G.C. 0% (0/16) 64.7%(12/18)
83.3%(20124}
Total 0% (0/40) 69.4%(43162} 75.0%(54/72)
*3469 POLYPECTOMY Mark N. Appleyard, Royal London Hasp, London United Kingdom; Tim Mills, Sandy Masse, Univ Coil, London United Kingdom; Paul Swain, Royal London Hasp, London United Kingdom Background: Analogue computer techniques were applied to measure power delivered and assess electrophysical factors influencing polypectomy and endoscopic submucosal resection. Methods: Total energy, cutting time and tissue effect were measured during snare polypectomy cuts using polyp models formed in porcine gastric tissue as well as 5, 10 and 15 mm polyp stalks fashioned from post mortem porcine muscle. Constant forces were applied to close the snares. The radio frequency generator used was the Erbotom ICC 200 on CUT (1-4), ENDOCUT (a sequential blend of CUT 1-4 and Soft COAG) and COAG (Soft and Forced) settings. Results: Increasing stalk size required a linear energy increase to cut polyps on all settings (p<0.0001). On aCUT, eCUT and fCOAG modes similar amounts of energy were required for 5ram stalks and small gastric polyps. 10 and 15ram stalks required more energy on fCOAG than higher aCUT and eCUT settings (p<0.05). Lower settings for aCUT and eCUT cut poorly on larger polyp models. FCOAG and eCUT3,4 achieved better hemostasis in 4mm vessels than aCUT3,4 (p<0.05). fCOAG was slower than other settings(p<0.001). Thin wire snares cut faster at lower power than thick (p<0.05). Increasing force decreased cutting-time and energy. Measurements during clinical polypectomy validated the models. Conclusions: These experiments measured the effect of waveform, powersetting, polyp stalk -size and snare force on cutting efficiency, hemostasis and tissue effect of endoscopic snare polypectomy. COAG settings may be best for large stalked polyps with a risk of bleeding from a large central vessel. Endoscopic submucosal resection of sessile polyps may be more safely resected using blended waveforms than coagulation alone if there is anxiety about a risk of perforation.
GASTROINTESTINAL ENDOSCOPY
AB127