Injection-incision–assisted snare resection of large sessile colorectal polyps

0016-5107/96/4303-018955.00 + 0 GASTROINTESTINALENDOSCOPY Copyright © 1996 by the American Societyfor Gastrointestinal Endoscopy

Injection-incision-assisted snare resection of large sessile colorectal polyps Toshinari Kanamori, MD, Makoto Itoh, MD, Yoshifumi Yokoyama, MD, Kenji Tsuchida, MD Nagoya, Japan

Background: It can be difficult, even for experienced endoscopists, to completely remove large sessile colorectal polyps. We attempted to remove large sessile colorectal polyps without complication and residual tumors. Method: Our new technique is characterized by submucosal pre-injection with a large volume of saline solution and then circumferential incision outside the lesion before resection using a special needle-tipped snare. The mean size of 33 polyps (including 9 elevated sessile, 20 flat nodular [villous], and 4 flat sessile polyps), was 4.0 cm (range, 3.0-8.5 cm). Results: Twenty-five (76%) were resected piecemeal and the remainder as a single specimen. Mild to moderate bleeding occurred in 3 (9.1%), but there was no clinically significant bleeding or perforation. No residual or recurrent tumors were recognized. Invasive carcinoma was revealed most frequently (44%) in elevated sessile polyps; none occurred in flat nodular polyps. Conclusion: Our removal technique appears to be safe and effective. Flat nodular polyps of any size are a particularly good indication for removal by this technique. (Gastrointest Endosc 1996;43:189-95.) It is generally accepted t h a t intramucosal carcinomas of the colorectum have no likelihood of metastasis.1 This supports the therapeutic effectiveness of endoscopic removal for polyps of any size without invasion. It is difficult, however, to remove large sessile polyps endoscopically without complication, and residual tumor m a y be left behind. To overcome this difficulty and possible incompleteness, we attempted a new modification of the injection polypectomy technique, characterized by submucosal pre-injection with a large volume of normal saline solution followed by circumferential incision a few millimeters outside the lesion edge. A special snare device, which had a 1 mm needle at the tip of a hexagonal snare wire, was used for resection. We applied our technique to 33 cases and evaluated it in terms of rate of complication and occurrence of residual or recurrent tumor.

Received March 7, 1995. For revision June 19, 1995. Accepted October 2, 1995. From the First Department of Internal Medicine, Nagoya City University Medical School, Nagoya, Japan. Reprint requests: Toshinari Kanamori, MD, Internal Medicine, Nagoya City University Medical School, 1 Kawasumi, Mizuho-ku, Nagoya 467, Japan. 37/1/69675

VOLUME 43, NO. 3, 1996

MATERIALS AND METHODS For this study, we have defined a large polyp as a lesion equal to or larger t h a n 3 cm in diameter. According to their shapes, we have classified large polyps into four types (Fig. 1): pedunculated, elevated sessile, flat nodular, or flat sessile. Among these, the l a t t e r t h r e e were collectively grouped as sessile because of t h e i r lack of stalked configuration. During a 3.5-year period from J u l y 1990 to December 1993, 33 colonoscopic removals of large sessile colonic polyps were performed in 32 p a t i e n t s a t the F i r s t Departm e n t of Medicine, Nagoya City University Hospital. The m e a n age of p a t i e n t s was 66.8 y e a r s (range, 44 to 86 years) and there were 23 (69.7%) m e n and 9 (30.3%) women. The symptoms t h a t led p a t i e n t s to undergo colonoscopy were positive test for stool occult blood in 13 of 32 p a t i e n t s (40.6%), constipation in 4 (12.5%), abdominal distention in 4, diarr h e a in 2 (6.3%), history of previous colon cancer in 2, bleeding hemorrhoids in 2, a n e m i a in 2, a n d other indications in the r e m a i n d e r (weight loss, family history of colon cancer, and high value of s e r u m CEA). Of 33 large sessile colonic polyps, 9 were classified as elevated sessile, 20 as flat nodular, a n d 4 as flat sessile polyps. By m e a s u r e m e n t of polyp size on x-ray i~flm, the m e a n size of elevated sessile polyps was 3.5 cm (range, 3.1 to 4.2 cm), flat n o d u l a r polyps 4.4 cm (range, 3.0 to 8.5 cm), and flat sessile polyps 3.2 cm (range, 3.0 to 3.4 cm), the overall m e a n size being 4.0 cm (range, 3.0 to 8.5 cm). As r e g a r d s distribution in the colorectum, the elevated sessile polyps were preGASTROINTESTINAL ENDOSCOPY 189

Table 1. Large sessile colorectal polyps: shape, diameter and distribution Pedunculated

Elevated sessile

Fiat nodular

\

Flat sessile

J

.yr. Sessile

Figure 1. Classification of large sessile colorectal polyps.

~\\\\\\\\\\\\\\\\\N~rnusculalispropria{MP) Submucosal injection volume Snaring site

J small volume

large volume

at the base of the protrusion

at the halfway point of the protrusion below the polyp

below the polyp

~\\NN\\\\\\\NN\\\\'~ ~ Resection

~ Resection ~\\~\\\\\\\\X\\\\\'~

Effect of thermal injury on the MP

near the MP

far from the MP

Figure 2. The effect of submucosal injection volume and snaring site on thermal injury of the muscularis propria. dominantly left-sided (descending colon to rectum), the flat sessile polyps were in the right colon (cecum to transverse colon), and fiat nodular polyps were evenly distributed in both sides (Table 1). All removals were performed in inpatients. Colon preparation was achieved using a combination of 1000 to 1500 mL of polyethylene glycol electrolyte solution, 10 mL of sodium picosulfate, and 90 mL of lactulose. 2 We used a soft-type colonoscope (Olympus colonoscope EVIS 200I, Olympus Corp., Tokyo, Japan), without sedation, but gave scopolamine butylbromide during the procedure. In general, endoscopic resection methods should be evaluated in terms of safety, sureness, and simplicity. We attempted three refinements on the usual removal technique for large sessile colonic polyps, some of which were basically derived from our polypectomy technique 3 for flat colonic lesions (a modified technique from the "strip biopsy" technique of Tada4). First, for safety, we used submucosal pre-injection with a large volume of physiological saline solution and positioned the snare wire halfway up the elevation caused by the submucosal injection. This will effectively prevent the thermal injury from extending to the deep muscle layer, and hence decrease the possibility of perforation (Fig. 2). Second, to be certain of removing the polyp without residual tumor, we used a needle knife (Olympus KD-10Q). We made a circumferential incision a few millimeters outside the lesion edge, which was made more obvious with dye-spraying technique. Removing the entire mucosal tissue inside the circumferential incision ensures complete removal of the lesion apart from any deeply inva-

190 G A S T R O I N T E S T I N A L E N D O S C O P Y

Elevated sessile Flat nodular Flat sessile Total

No. of

Diameter ( m m )

Distribution*

cases

Mean _+SD (Min.-max.)

Right:Left

9

35.3 ± 5.0 (31-42)

3:6

20

44.4 ± 14.7 (30-85)

10:10

4

31.5 ± 1.9 (30-34)

4:0

33

40.3 ± 12.7 (30-85)

17:16

*Right side of colon: cecum to transverse colon; left side of colon: descending colon to rectum.

sive part of the lesion if present. Finally, to make our technique easier we used a special snare device that has a 1 mm needle at the tip of a hexagonal-shaped wire (Fig. 3 A). By sticking the needle into the mucosa, the relative position of the snare wire to the lesion becomes more stable, making resection easier. Moreover, as shown in Figure 3 B and C, when the needle tip is impacted, pushing the snare device inward increases the diameter of the snare loop. This exact control of the loop makes precise regulation of resection size possible. Figure 4 schematically demonstrates the whole process of our removal technique for large sessile colorectal polyps. Submucosal injection (Fig. 4B) is performed at several sites under and around the tumor with a large total volume of physiological saline solution, usually more than 20 ml; the bigger the polyp the larger the volume of injection needed. In order not to obscure the lesion, the injection should be started at the proximal portion of the lesion (i.e., furthest portion from the endoscope) and the injection volume to the proximal portion should be more than the injection volume to the opposite portion. Circumferential incision (Fig. 4C) is made with the needle knife a few millimeters outside the lesion edge on the elevation caused by the submucosal injection. We always used the dye-spraying technique to make the lesion edge more obvious, even if the lesion edge appeared to be distinct with naked eye inspection. Piecemeal resection (Fig. 4D) then began where the polyp seemed most likely to invade deeply or to be most severely dysplastic, because the first cut is usually the best sample for pathologic examination. We tried to resect each piece at a size of 10 to 20 mm because a larger cut may be dangerous but a smaller one less good for pathologic examination. After several cuts are performed (producing the histopathologic specimens), any tiny area of residual mucosal tissue should be removed or destroyed so that only edematous submucosal tissue can be seen inside of the circumferential incision, indicating complete mucosal removal (Fig. 4E). Reconstruction of the resected specimens (Fig. 4F) gives further evidence of complete mucosal removal. We performed all these steps during a single endoscopy procedure because any residual tumor becomes indurated and fixed after electrocoagulation, causing elevation by submucosal injection to be difficult on a subsequent examination. For lesions in the right side of the colon, we usually used

VOLUME 43, NO. 3, 1996

a. 5 cm flat nodular polyp

d. Segmental resections polyp

mucosa

musculari$ propria 5 crn b. S u b m u c o s a l l a r g e - v o l u m e

injection

c. Circumferential incision

~

J

~ ~

~

e. C o m p l e t e r e m o v a l inside the incision

f. R e c o n s t r u c t i o n of s p e c i m e n s

~\\\\\\\\\\\\\\\\\'~"

®

Figure 4. Steps in removal technique for large sessile colorectal polyps.

thrombin solution prophylactically over the base of the polypectomy ulcer. If overt bleeding occurred, we preferred to use endoscopic injection of hypertonic salineepinephrine solution (HSE, consisting of 4.7% NaC1 and 0.005% epinephrine) 5 or pure ethanol. After resection the patient was kept NPO for a day because of the possibility ofa miniperforation. 6 Resected specimens were stretched on a board with pins and fixed in 10% buffered formalin for a day. We examined the specimens under a dissecting microscope to section at the areas with the most dysplastic appearance. We evaluated the usefulness of our technique in terms of safety and completeness. Safety was assessed by the frequency and severity of complications, which were divided into minor and major. Minor complications, meaning trivial problems seen in normal polypectomies, included mild to moderate bleeding or other easily managed events, and major complications included massive (i.e., clinically significant) bleeding, perforation, or other serious events. Completeness was assessed by the frequency of residual or recurrent tumor, diagnosed by endoscopic findings and microscopic examinations of multiple biopsy specimens at the postoperative scar site 3 months later and 1 year later.

RESULTS Figure 3. Special needle-tipped snare device. A, One-millimeter needle at the tip of a hexagonal snare wire. B, In normally opened state the diameter of hexagonal wire is narrow. C, By sticking the needle into the mucosa and pushing the snare device inward, the diameter of hexagonal wire increases, making regulation of resection size possible. a splinting device to make it easier to insert the instrument several times. In addition, we sometimes retrieved two resected specimens at a time using a four-wire grasper, and we sometimes retrieved tiny fragments through the suction channel with a polyp-trap device. We routinely irrigated with 10 to 20 ml of 1000 U/mL

VOLUME 43, NO. 3, 1996

All p r o c e d u r e s w e r e successfully perforwied in a single stage. On t h e whole, it w a s m o r e difficult to rem o v e a polyp in t h e r i g h t side of the colon t h a n in the left side of t h e colon. I t t o o k a b o u t I h o u r to r e m o v e a n a v e r a g e - s i z e d polyp located in t h e r i g h t colon. O f 33 l a r g e sessile colonic polyps, 25 (76%) w e r e r e s e c t e d p i e c e m e a l a n d t h e r e m a i n d e r , in w h i c h all w e r e less t h a n 35 m m in size, as a single specimen. S u b m u c o s a l saline solution injection w a s p e r f o r m e d carefully to avoid leakage, w i t h a m e a n injected v o l u m e of 32 m L (range, 26 to 56 mL). M i n o r complications occurred in 3 of 33 cases, all of which w e r e mild to m o d e r a t e bleeding; no m a j o r complication w a s recognized. All t h e s e m i n o r complications occurred w h e n e l e v a t e d

GASTROINTESTINAL ENDOSCOPY

191

Table 2. Result 1: required resection(s) and complication Required resection(s)

Shape (n)

One 3

Complication*

Multiple 6

Minor 2

Major 0

Elevated sessile (9) Flat nodular 3 17 0 0 (20) Flat sessile 2 2 1 0 (4) Total (33) 8 25 3 0 *Minor: mild to moderate bleeding; major: massivebleeding or perforation.

Case illustration

Figure 5. Case presentation: radiographic findings. Radiograph of flat nodular polyp in the rectum, shewing granular or soap bubble appearance with diameter of 52 mm.

sessile polyps or flat sessile polyps were resected as a single specimen (Table 2). In flat nodular polyps no complication, minor or major, occurred. After resection nothing particular occurred in any patient and all were discharged home 7 to 10 days after the resection. Pathologically (Table 3), all flat nodular polyps consisted of villous or villotubular components. Severe dysplasia (i.e., carcinoma in situ or intramucosal carcinoma) was recognized most commonly in 3 out of 4 (75%) flat sessile polyps, followed by flat nodular polyps (12 out of 20, 60%) and least often in elevated sessile polyps (2 out of 9, 22%). But the frequency of submucosally invasive carcinoma was highest in elevated sessile polyps (4 out of 9, 44%) and lowest in flat nodular polyps (0 out of 20, 0%). In addition, though all submucosally invasive carcinomas in elevated sessile polyps invaded massively, in flat sessile polyps any submucosal invasion was minimal. In follow-up colonoscopies 3 months and i year after the resections (Table 3), no residual or recurrent tumor was recognized endoscopically or microscopically.

192 G A S T R O I N T E S T I N A L E N D O S C O P Y

A 68-year-old woman had a positive test for stool occult blood and then underwent a barium enema, demonstrating a lesion of 52 mm in size in the rectum (Fig. 5). At colonoscopy the lesion was classified as a flat nodular polyp, which was made more obvious with dye-spraying technique (Fig. 6A). Submucosal injection with 38 mL of physiological saline solution and circumferential incision was performed (Fig. 6B). After four snare cuts were performed and some tiny areas of residual tissue were cleaned up, the entire submucosal tissue inside the circumferential incision was clearly seen with the help of sprayed dye (Fig. 6C). Figure 7 shows a dissecting photomicrograph of resected specimens, indicating the size of the four piecemeal cuts to be between 1 and 2 cm in size.

DISCUSSION It is generally accepted t h a t intramucosal carcinomas of the colorectum have no potential for metastasis. 1 This is one of the most important reasons for the development of endoscopic resection of colorectal polyps. However, it is difficult to remove large polyps endoscopically without complications, especially nonstalked lesions. Large polyps are defined as any lesion with a diameter of 2 cm or larger by Christie et al., 7 as 2.5 cm or more by Karita et al., s and as 3 cm or more by Bedogni et al. 9 In this study we defined "large" as a lesion of 3 cm or more because prevalent polypectomy techniques, including simple piecemeal polypectomy, can usually remove lesions of 2 to 3 cm in size without complication or residue. Therefore, lesions of lesser size were thought to be improper for evaluation of the usefulness of a new endoscopic removal technique. On the other hand, pedunculated lesions are also thought to be improper for evaluation because the stalks of pedunculated lesions are usually much smaller than the lesion itself, and resection of a pedunculated lesion rarely causes massive bleeding.

VOLUME 43, NO. 3, 1996

Table 3. Result 2: pathology and follow-up Pathology

Follow-up*

Shape (n) Elevated sessile (9) Flat nodular (20) Flat sessile (4) Total (33)

Adenoma

M-ca

Sm-ca

Residue

Recurrence

3 8 0 11

2 12 3 17

4 0 1 5

0 0 0 0

0 0 0 0

M-ca, Intramucosal carcinoma; Sm-ca, submucosally invasive carcinoma. *Residue was diagnosed 3 months later; recurrence 1 year later.

Therefore, in this study our new technique was evaluated on sessile colonic polyps 3 cm or larger in size. The characteristics of our modified removal technique included, as mentioned above, pre-injection with a large volume of saline solution, circumferential incision to define the lesion edge, and use of our original special snare device with a needle tip. Submucosal pre-injection was originated by Deyhle et al., 1° and in J a p a n the technique was developed into strip biopsy by Tada et al. 4 Recently m a n y variations of strip biopsy technique have been reported, 11-14 with differences mainly in the endoscope, snare device, or injection solution used. In previous reports, 4, 10-14 the volume injected into the submucosa was always up to a maximum of 5 mL of normal saline solution or dilute epinephrine (including HSE) as injection solution. Even in reports evaluating large sessile colonic polyps, the maximum volume injected was also 5 mL. For example, in different publications, 3 to 5 mL of normal saline solution was used for a 50 mm sessile tumor, s 0.5 to 1.0 mL of riSE for a 40 mm sessile tumor, 13 and 2 mL of dilute epinephrine for a sessile tumor over 21 mm in size.14 In our technique a large volume of physiological saline solution (mean volume of 32 mL in this study) was injected in the submucosa and the snare wire was positioned halfway up the elevation caused by submucosal pre-injection, as schematically shown in Figure 2. We think a large volume of pre-injection makes the removal of polyps safer. We always use dye-spray technique even for polyps t h a t appear to be relatively well demarcated because the accurate assessment of the tumor extent is essential to completely remove them. Especially, flatter parts of flat nodular polyps may be easily missed without dye-spray technique. Moreover, we believe our technique of circumferential incision a few millimeters outside the lesion edge plays another import a n t role in removing lesions without leaving residual tumor. To make the incision easier and safer, the polyp should be seen downward (i.e., in the direction of 6 o'clock on the screen). To avoid bleeding or damage of the muscularis propria, the incision must be made superficially on the elevation caused by submucosal in-

VOLUME 43, NO. 3, 1996

jection. Therefore, if an endoscopist is unfamiliar with use of the needle knife, electrocoagulated marker dots with the tip of the snare wire may be a substitute for the incision technique we have employed. Maneuverability of the snare is also very important to resect completely and safely. Our special snare device, which we developed for the endoscopic resection of flat adenomas, 3 is very useful to remove any selected part of a lesion at will and also to regulate resection size. We share the opinion of Karita et al. s that the procedure of piecemeal resection should be repeated as soon as possible because fibrosis under any residual lesion becomes more dense if the interval is longer than 2 weeks. Extending this principle, all our cases were performed in a single endoscopic procedure. It is our opinion that all patients with massively submucosa-invasive carcinoma should undergo surgery without attempting endoscopic removal. In this series, however, we tried to endoscopically remove four cases of massively submucosa-invasive carcinomas. The reasons were advanced age in two (81 and 82 years old, respectively), renal failure in one, and stubborn refusal to undergo surgical operation in one. Of 33 lesions, 3 cases (9.1%) were complicated by mild to moderate bleeding, in all of which bleeding could be stopped easily with submucosal injection of a total of 6 to 10 mL HSE solution (1 to 2 ml HSE per injection). No major complication (massive bleeding or perforation) was recognized at all. According to a previous report, 15the complication rate was calculated as 3.0% by frequency of major complications; in our study there was no major complication. All minor complications in this study occurred when elevated sessile polyps and flat sessile polyps were resected as a single specimen. Therefore, since completing this study we have not resected an elevated sessile or flat sessile polyp with a diameter of 3 cm or more as a single specimen, even though the snare can entrap the whole lesion. in flat nodular polyps, all of which were villous adenomas in this series, no complication occurred, either in cases resected as a single specimen or in cases resected as piecemeal We have the impression that flat

GASTROINTESTINAL ENDOSCOPY

193

Figure 7. Case presentation: dissecting photomicrograph of resected specimens. Four specimens of 1 to 2 cm in size and two tiny tissue residues were examined under a dissecting microscope to section at the most dysplastic-seeming area.

Figure 6. Case presentation: endoscopic findings and resection. A, Dye-spraying technique makes the appearance and edge of the lesion more obvious. B, After submucosal injection with 38 ml of normal saline solution, circumferential incision has been performed. C, After four snare cuts and some tiny areas of mucosal tissue cleaned up, the edematous submucosal tissue inside the circumferential incision is clearly seen with the help of sprayed dilute dye.

nodular polyps are less hemorrhagic in nature t h a n flat sessile or elevated sessile polyps during electrosurgery. In fact, we successfully resected a flat nodular polyp with diameter of 85 mm with about 50 cuts 194

GASTROINTESTINAL ENDOSCOPY

without complication. Because there is probably less bleeding disposition and less frequent invasion (0 out of 20, 0% in this study), flat nodular polyps of any size are a good indication for endoscopic resection unless the patient is unfit for the procedure. In contrast, elevated sessile polyps 3 cm or more in size have the highest likelihood (44%) of invasive carcinoma, so before endoscopic resection, it is desirable to examine them further with endoscopic ultrasonography, and a surgical alternative must also be considered. No tumor persisted or recurred after 1 year in this study, although longer follow-up is necessary to be completely sure of no recurrence. At present all cases have been followed-up for at least 1 year (a maximum for 4 years with a mean of 2.4 years) without any colonoscopic or pathologic sign of recurrence. Although it is difficult to compare recurrence rates in different series because of their different follow-up length, we think our recurrence rate is lower t h a n that of Walsh et a1.15 Therefore, circumferential incision is thought to be useful as a landmark indicating the extent of lesion to be resected. Because circumferential incision is made a few millimeters outside the lesion edge, cleaning up all the mucosal tissue inside the circumferential incision will guarantee eradication of the polyp, assuming no deep invasion is present.

Conclusion Our technique for removal of large sessile colorectal polyps seems useful because only minor complications occurred, with no major complication and no residual or recurrent tumor occurrence. Flat nodular polyps of any size seem to be a particularly good indication for endoscopic resection; in our series there were no complications in their removal and their histology was uniformly benign. VOLUME 43, NO. 3, 1996

ACKNOWLEDGMENT The authors thank Dr. Christopher B. Williams, St. Mark's Hospital, London, U.K., for reviewing the manuscript. REFERENCES 1. Morson BC, Bussey HJR, Samoorian S. Policy of local excision for early cancer of the colorectum. Gut 1977;18:1045-50. 2. Kanamori T, Yokoyama Y, Itoh M, Takeuchi T. A study on colon preparation method for colonoscopy by using 500 ml of polyethylene glycol electrolyte lavage solution. Therapeutic Research 1994;15(suppl 2):11-6 (in Japanese with English abstract). 3. Kanamori T, Itoh M, Yokoyama Y, et al. Studies of endoscopic resection of flat lesions in the colon: an attempt for improvement of removal method. Gastroenterological Endoscopy 1993; 35:1290-8 (in Japanese with English abstract). 4. Tada M, Murakami M, Karita H, Oldta K. Endoscopic resection of early gastric cancer. Endoscopy 1993;25:445-50. 5. Hirao M, Kobayashi T, Masuda K, et al. Endoscopic local injection of hypertonic saline-epinephrinesolution to arrest hemorrhage from the upper gastrointestinal tract. Gastrointest Endosc 1985;31:313-7. 6. Christie JP, Marrazzo JIII. "Mini-perforation" of the colon-not all postpolypectomy perforations require laparotomy. Dis Colon Rectum 1991;34:132-5.

7. Christie JP. Colonoscopic excision of large sessile polyps. Am J Gastroenterol 1977;67:18-20. 8. Karita M, Tada M, Okita K. The successive strip biopsy partial resection technique for large early gastric and colon cancers. Gastrointest Endosc 1992;38:174-8. 9. Bedogni G, Bertoni G, Ricci E, et al. Colonoscopic excision of large and giant colorectal polyps. Technical implications and results over eight years. Dis Colon Rectum 1986;29:831-5. 10. Deyhle P, Largiad~r F, Jenny S, Fumagalli I. A method for endoscopic electroresection of sessile colonic polyps. Endoscopy 1973;5:38-40. 11. McGahan TP, Gilinsky NH. Colonic tumors. Endoscopy 1994; 26:70-87. 12. Fujimori T, Nakamura T, Hirayama D, et al. Endoscopic mucosectomy for early gastric cancer using modified strip biopsy. Endoscopy 1992;24:187-9. 13. Shirai M, Nakamura T, Matsuura A, Ito Y, Kobayashi S. Safer colonoscopic polypectomy with local submucosal injection of hypertonic saline-epinephrine solution. Am J Gastroenterol 1994;89:334-8. 14. Yokota T, Sugihara K, Yoshida S. Endoscopic mucosal resection for colorectal neoplastic lesions. Dis Colon Rectum 1994;37: 1108-11. 15. Walsh RM, Ackroyd FW, Shellito PC. Endoscopic resection of large sessile colorectal polyps. Gastrointest Endosc 1992;38: 303-9.

1-800-55-MOSBY

This number links you to the full text of articles published in over 25,000 journals, including all Mosby journals. M O S B Y Document Express ®, a rapid response information retrieval service, provides quick turnaround, 24-hour availability, and speedy delivery methods. For inquiries and pricing information, call our toll-free, 24-hour order line: 1-800-55-MOSBY; outside the United States:415-259-5046; fax: 415-259-4019; E-mail: mosbyexp @ class.org. MOSBY Document Express® is offered in cooperation with Dynamic Information Corp.

V O L U M E 43, NO. 3, 1996

GASTROINTESTINAL ENDOSCOPY

195