Location in the ascending colon is a predictor of refractory colonic diverticular hemorrhage after endoscopic clipping

ORIGINAL ARTICLE: Clinical Endoscopy

Location in the ascending colon is a predictor of refractory colonic diverticular hemorrhage after endoscopic clipping Naoki Ishii, MD,1,2 Nobuto Hirata, MD,2 Fumio Omata, MD, MPH,1 Toshiyuki Itoh, MD,1 Masayo Uemura, MD,1 Michitaka Matsuda, MD,1 Shoko Suzuki, MD,1 Yusuke Iizuka, MD,1 Katsuyuki Fukuda, MD,1 Yoshiyuki Fujita, MD1 Tokyo, Kamogawa, Japan

Background: Predictors of refractory colonic diverticular hemorrhage after endoscopic clipping (EC) remain unclear. Objective: To elucidate the predictors of uncontrolled bleeding after EC. Design: Retrospective study. Setting: Two tertiary referral centers. Patients: Eighty-nine patients with colonic diverticular hemorrhage who underwent EC as a first-line treatment were included. Interventions: If bleeding remained uncontrolled after 1 or 2 EC sessions, other interventions (transcatheter arterial embolization, endoscopic band ligation, or surgery) were performed. Patients were divided into ECcontrolled and EC-uncontrolled groups; the characteristics of each group were compared. Main Outcome Measurements: Comorbidities, location of bleeding diverticula, and EC technique (direct vs indirect placement). Results: Initial treatment with EC was successful in 87 patients. Early rebleeding (primary failure) occurred in 30 of 87 patients (34%). Secondary failure occurred in 6 of 22 patients treated with reclipping (27%). Cumulatively, 78 patients were successfully managed with EC. Non-EC treatments were required in 11 patients. Location in the right side of the colon, particularly in the ascending colon, was significantly more common in the ECuncontrolled group than in the EC-controlled group (P ⫽ .017 and P ⫽ .0029, respectively). Although the difference was not significant, bleeding was successfully managed in all 13 patients treated with direct placement. Bleeding remained uncontrolled after EC in 11 of 52 ascending cases (21%) treated with indirect placement. Diverticular hemorrhage in other locations was managed regardless of EC technique. Limitations: Retrospective study. Conclusions: Location in the ascending colon is a significant predictor of refractory colonic diverticular hemorrhage after EC. Indirect placement of hemoclips in ascending lesions is ineffective. (Gastrointest Endosc 2012;76:1175-81.)

Lower GI bleeding (LGIB) has several etiologies including diverticula, vascular ectasia, colitis, and neoplasia. Of these, colonic diverticular hemorrhage has been reported

to be the most common cause of LGIB.1-5 Although spontaneous resolution of colonic diverticular hemorrhage occurs in 70% to 80% cases without therapeutic intervention,

Abbreviations: EBL, endoscopic band ligation; EC, endoscopic clipping; LGIB, lower GI bleeding; SRH, stigmata of recent hemorrhage; TAE, transcatheter arterial embolization.

Received March 24, 2012. Accepted July 30, 2012.

DISCLOSURE: The authors disclosed no financial relationships relevant to this publication. Copyright © 2012 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 http://dx.doi.org/10.1016/j.gie.2012.07.040

www.giejournal.org

Current affiliations: Department of Gastroenterology (1), St. Luke’s International Hospital, Tokyo, Japan, Department of Gastroenterology (2), Kameda Medical Center, Kamogawa, Japan. Reprint requests: Naoki Ishii, MD, Department of Gastroenterology, St. Luke’s International Hospital 9-1 Akashi-cho, Chuo-ku, Tokyo 104-8560, Japan.

Volume 76, No. 6 : 2012 GASTROINTESTINAL ENDOSCOPY 1175

EC for colonic diverticular hemorrhage

patients with severe bleeding may require endoscopic, surgical, or angiographic treatment.4,6-10 Recently, several studies recommended urgent colonoscopy for initial evaluation of LGIB. Several endoscopic therapies for colonic diverticular hemorrhage including epinephrine injection, contact thermal therapy, endoscopic clipping (EC), and endoscopic band ligation (EBL) have been used to achieve hemostasis and prevent further bleeding because angiographic treatment may have a risk of ischemic change after embolization (0%-8%).5,9-24 Endoscopic placement of hemoclips on a bleeding vessel offers immediate mechanical hemostasis and the theoretical advantage of causing less injury to the adjacent tissues. In addition, the wall of the right side of the colon is particularly thin; therefore, endoscopic hemostasis with hemoclips has been increasingly used in the treatment of colonic diverticular hemorrhage instead of coagulation therapy.15-18 However, the predictors for uncontrolled diverticular bleeding after EC have not yet been elucidated. The aim of this study was to evaluate the predictors of refractory colonic diverticular bleeding after 1 or 2 EC treatment sessions.

PATIENTS AND METHODS Study setting and patient population Endoscopic records of patients who presented with overt acute hematochezia at 2 tertiary referral centers, St. Luke’s International Hospital in Tokyo (a 520-bed urban tertiary referral hospital) and Kameda Medical Center in Kamogawa (an 800-bed rural tertiary referral hospital) from January 2004 to February 2010, were retrospectively reviewed. More than 4000 colonoscopies are performed every year at each site; urgent colonoscopies (colonoscopy at the endoscopy unit 12-24 hours after hospitalization or diagnosis of hematochezia and within 1 hour after clearance of stool and clots)12 were performed in both hospitals by experienced endoscopists who have performed more than 2000 colonoscopic examinations. In cases of substantial LGIB, urgent colonoscopy was performed. A flow chart of patients with colonic diverticular hemorrhage included in this study is presented in Figure 1. Colonoscopy for evaluation of acute LGIB was performed in 951 patients. Surgery was not initially performed in any patients. First, postpolypectomy bleeding, colorectal cancer, ischemic colitis, and angiodysplasia were ruled out. Stigmata of recent hemorrhage (SRH), which allowed unequivocal identification of specific diverticula as the cause of bleeding, were defined as densely adherent clot after vigorous irrigation, a nonbleeding visible vessel, or active bleeding at colonoscopy12,25; SRH were observed in 100 patients (59 patients at St. Luke’s International Hospital and 41 patients at Kameda Medical Center) who had a diagnosis of definite colonic diverticular hemorrhage.12 After excluding patients who initially underwent EBL or transcatheter arterial embolization (TAE) instead of EC, 89 1176 GASTROINTESTINAL ENDOSCOPY Volume 76, No. 6 : 2012

Ishii et al

Take-home Message ●

●

Location in the ascending colon is a predictor of refractory colonic diverticular hemorrhage after endoscopic clipping. Indirect placement of hemoclips for bleeding diverticula in the ascending colon is ineffective.

consecutive patients who underwent EC as a first-line therapy were included in the retrospective cohort study. This study was approved by the institutional review board of each hospital, and all patients provided informed consent before undergoing the procedure.

Colonoscopy and EC All patients received standard supportive medical care for LGIB, including fluid resuscitation and hemodynamic monitoring. Some patients underwent enhanced helical CT to detect an active bleeding source. After admission, bowel preparation with polyethylene glycol was performed before colonoscopic examinations. A glycerin enema was administered to patients who could not take PEG because of hemodynamic instability. To begin, 50 mg pethidine hydrochloride was administered and an additional dose of midazolam was administered as required based on the endoscopists’ judgment. During colonoscopy, blood pressure, oxygen saturation levels, and electrocardiographic readings were monitored; supplemental oxygen was administered in the event of oxygen desaturation. After identification of bleeding sources, hemoclips (HX-600-090L, HX-600-135, HX-610-090L, or HX-610-135; Olympus Optical Co, Ltd, Tokyo, Japan) were placed directly on the vessel if possible (Fig. 2). When direct placement was not possible because of dome location, massive hemorrhage, or small diverticular orifice, indirect placement was selected, and bleeding diverticula were closed with multiple hemoclips in a zipper fashion (Fig. 3). Monotherapy with clipping was performed, and an injection of epinephrine solution was not administered before EC in all cases. If rebleeding after initial endoscopic treatment occurred, repeat colonoscopy and endoscopic reclipping were first attempted. TAE, colectomy, or EBL was subsequently performed based on the attending gastroenterologist’s judgment when diverticular rebleeding was not controlled because of massive rebleeding from the previously treated diverticula, poor endoscopic view, or hemodynamic instability. Because early rebleeding (clinical evidence of recurrent LGIB within 30 days after initial EC) is more likely to occur from recently treated diverticula,12 patients were closely followed on an inpatient or outpatient basis for at least 30 days after EC, and rebleeding time was confirmed in all patients during this period. www.giejournal.org

Ishii et al

EC for colonic diverticular hemorrhage

Figure 1. Flow chart of patients with colonic diverticular hemorrhage included in this study.

Comparison of EC-controlled and EC-uncontrolled groups Patients were divided into an EC-controlled group in which both initial bleeding and rebleeding were managed with EC, and an EC-uncontrolled group in which TAE, colectomy, or EBL was performed for treatment of either initial bleeding or rebleeding. Comorbid diseases, use of antiplatelet agents or nonsteroidal anti-inflammatory drugs, hematocrit level on admission, location of bleeding diverticula (cecum, ascending colon, transverse colon, descending colon, or sigmoid colon), EC technique (direct vs indirect placement in a zipper fashion), use of transparent caps, procedure time, and complications such as perforation and abscess formation were retrospectively evaluated in each group.

Statistical analysis Statistical analysis was performed by using JMP version 8 (SAS Institute Inc, Cary, NC). Patient age, hematocrit at admission, and procedure time were reported as mean (standard deviation). The Student t test and Fisher exact test were applied for continuous and categorical variables, respectively, and a P ⬍ .05 was considered statistically significant. It is recognized that there was multiple testing www.giejournal.org

of outcome data arising from individual patients. The analyses related to the EC technique and location and SRH as potential clinical predictors of uncontrollable diverticular hemorrhage after EC were considered the main definitive results. It is noted that correction by the method of Benjamini and Hochberg would not have removed statistical significance from any P values for the main results.26 The other statistical tests are to be taken as descriptive only. Thus, all P values are presented uncorrected for multiple testing.

RESULTS Patient characteristics In the 89 consecutive patients who underwent EC as a first-line therapy at both institutions, 63 lesions were located in the right colon (71%). Bleeding points located in the diverticulum were categorized as follows: dome, 11% (10/89); neck, 21% (19/89); and unknown 68% (60/89). The bleeding point in the unknown category was not precisely determined because of a small orifice or active bleeding. Active bleeding was observed in 41 patients (46%). Direct placement of hemoclips on the vessel was performed in 13 patients (cecum, n ⫽ 2; ascending Volume 76, No. 6 : 2012 GASTROINTESTINAL ENDOSCOPY 1177

EC for colonic diverticular hemorrhage

Ishii et al

Figure 2. A, Endoscopic view of a nonbleeding visible vessel located at the neck of the diverticulum (arrow). B, Direct placement of a metal hemoclip on the vessel was performed.

Figure 3. A, Endoscopic view of active bleeding from the diverticulum. B, The bleeding diverticulum was closed with multiple hemoclips in zipper fashion (indirect placement).

colon, n ⫽ 3; transverse colon, n ⫽ 2; descending colon, n ⫽ 3; and sigmoid colon, n ⫽ 3). In 76 patients, indirect placement of hemoclips was performed, and diverticula were closed with multiple hemoclips in a zipper fashion.

patients in total were managed with subsequent non-EC treatment. No rebleeding was observed for at least 30 days after additional non-EC treatment in these patients.

Follow-up after initial EC Details of follow-up after initial EC are presented in Figure 4. Although initial treatment with EC was successful in 87 patients, initial bleeding could not be controlled in 2 ascending colon patients. Therefore, TAE was immediately performed after unsuccessful EC, and hemostasis was achieved in these patients. Early rebleeding (primary failure) after initial EC occurred in 30 of 87 patients (34%). The mean (standard deviation) and median days (range) of rebleeding occurrence after initial EC were 3 (4) and 3 days (range 0.1-18 days), respectively. In 5 patients, early rebleeding was managed conservatively after initial EC. Sudden massive hematochezia developed in 3 patients after initial EC, necessitating TAE (n ⫽ 2) or EBL (n ⫽ 1) (2 hours, 4 days, and 2 days after initial EC, respectively). Reclipping was performed in 22 patients with early rebleeding. In 6 patients with secondary EC failure, right hemicolectomy and/or TAE was performed. Ultimately, 11 1178 GASTROINTESTINAL ENDOSCOPY Volume 76, No. 6 : 2012

Clinical predictors of uncontrollable diverticular hemorrhage after EC The clinical characteristics of 78 patients in the ECcontrolled colonic diverticular hemorrhage (1 or 2 treatment sessions) group and those of the 11 patients in whom further treatment was required are presented in Table 1. Location in the right colon (ascending colon) was significantly more common in the EC-uncontrolled group than in the EC-controlled group (P ⫽ .017 and P ⫽ .0029, respectively). Although no significant differences were observed in terms of the EC technique (P ⫽ .35), bleeding was successfully managed in 13 patients in whom direct placement of hemoclips was performed. In 11 of 52 ascending colon patients (21%) treated with zipper closure, bleeding could not be controlled with EC. In contrast, EC treatment for diverticular hemorrhage in other locations (cecum, transverse through sigmoid colon) was successful regardless of EC technique (direct or indirect placement). www.giejournal.org

Ishii et al

EC for colonic diverticular hemorrhage

Figure 4. Study flow diagram of patients who underwent endoscopic clipping for colonic diverticular hemorrhage. EC, endoscopic clipping; TAE, transcatheter arterial embolization; EBL, endoscopic band ligation.

DISCUSSION This retrospective cohort study of EC for colonic diverticular hemorrhage adds to the existing body of literature regarding endoscopic therapies. These therapies include epinephrine injection, contact thermal therapy, EC, and EBL.5,11-24 In this study, EC as a first-line therapy was attempted in 89 definite colonic diverticular hemorrhage patients with SRH, and patients were closely followed for at least 30 days after EC. In this study, a 98% success rate (87/89) of initial placement of hemoclips with no procedural complications was observed. However, a high rebleeding rate of 34% was noted (30 of 87 cases) in contrast to other more favorable reports,17,18 wherein the overall early rebleeding rate after initially successful hemostasis was excellent (0%, 0/32). In these studies, initial EC was not possible in 3 of 35 patients (9%) and TAE or surgery was required. Kaltenbach et al18 reported a novel approach to improve the visualization of bleeding points in the diverticulum with the use of a transparent cap and clipping technique. However, cap use was not significantly associated with controllability in this study. Technical differences may be responsible for the discrepancy between the results of the current study and those of previous studies. Indirect placement in a zipper fashion was performed in 85% of our patients. Moreover, Olympus clips were used in this study; other newer types of hemoclips may differ in terms of placement depth or duration.27 In addition, anatomic factors such as right-sided localization in 71% of our patients compared with substantially fewer right-sided cowww.giejournal.org

lon bleeds in other reports may be an important prognostic difference. Anatomic features of colonic diverticula include arterial arcades from the diverticulum neck that join and form the artery at the base of the diverticulum.28 When the endoscopic technique of zippering around the diverticulum neck is used, occlusion in the major SRH or the underlying artery is unlikely in patients with rebleeding. Thus, direct placement is considered optimal for adequate arterial occlusion. In addition, initial injection of diluted epinephrine to decrease bleeding is recommended in cases of massive hemorrhage for precise confirmation of the bleeding point and facilitation of direct EC.17,18 In this study, bleeding was managed with EC in all 13 patients in whom direct placement of hemoclips was performed, although no significant difference in rates of uncontrolled bleeding was observed based on hemoclip placement technique (direct vs indirect). However, statistical differences may have been difficult to determine because comparison of these techniques was not included in the study design, and direct placement was only performed in few patients. Although the indirect technique in this study appeared to be suboptimal for hemostasis compared with treatments delivered directly to SRH sites in previous studies,17,18 further studies to determine the differences in EC techniques may be warranted. Several studies have reported a correlation of nonsteroidal anti-inflammatory drug, anticoagulant, and antiplatelet use with colonic diverticular hemorrhage as well as with peptic ulcer disease.29,30 In the current study, nonsteroidal anti-inflammatory drugs or antiplatelet agents were reguVolume 76, No. 6 : 2012 GASTROINTESTINAL ENDOSCOPY 1179

EC for colonic diverticular hemorrhage

Ishii et al

TABLE 1. Characteristics of controlled and uncontrolled diverticular bleeding after endoscopic clipping P value

alternative treatments were required for rebleeding in the right side of the colon, particularly in the ascending colon. Our results corroborate those from previous reports of TAE for colonic diverticular hemorrhage.10,32 Although few cases of diverticular hemorrhage in the ascending cases were treated with direct placement, EC was adequate for management in these cases. On the other hand, treatment by using indirect placement in 21% ascending colon cases was unsuccessful. Therefore, other endoscopic treatments such as EBL may be desirable for treating ascending lesions that could not be treated with direct placement because EBL has been demonstrated to achieve immediate hemostasis even in cases of dome location or massive hemorrhage.19-24 In conclusion, although the study was retrospective and involved a limited number of patients, it revealed that the location of diverticular hemorrhage in the right side of the colon, particularly in the ascending colon, was considered a significant predictor of uncontrolled diverticular bleeding after EC. In addition, indirect placement of hemoclips was unsuccessful in treating bleeding diverticula in the ascending colon.

Controlled (n ⴝ 78)

Uncontrolled (n ⴝ 11)

Sex, M/F

54/24

7/4

.74

Age, mean (SD), y

64 (13)

62 (15)

.63

Cardiopulmonary disease, % (no.)

38 (18/78)

27 (3/11)

.72

Hypertension, % (no.)

46 (36/78)

18 (2/11)

.11

Diabetes mellitus, % (no.)

15 (12/78)

9 (1/11)

1.00

Hyperlipidemia, % (no.)

18 (14/78)

27 (3/11)

.43

Antiplatelet agents or NSAIDs, % (no.)

29 (23/78)

18 (2/11)

.72

Hct on arrival, %, mean (SD)

35.1 (6.4)

33.6 (5.6)

.45

Location in colon (C/A/T/D/S)

5/43/4/9/17

0/11/0/0/0

Location (right/ left)

52/26

11/0

.017

The authors thank Dr Gautam Deshpande for refining our manuscript.

Location (A/ other)

43/35

11/0

.0029

REFERENCES

SRH (AB/ NBVV⫹AC)

34/44

7/4

.17

Placement of hemoclips (direct/indirect)

13/65

0/11

.35

Use of transparent cap, % (no.)

64 (50/78)

64 (7/11)

.65

Procedure time, min, mean (SD)

45 (24)

49 (29)

.64

ACKNOWLEDGMENTS

M, Male; F, female; SD, standard deviation; NSAIDs, nonsteroidal antiinflammatory drugs; Hct, hematocrit; C, cecum; A, ascending colon; T, transverse colon; D, descending colon; S, sigmoid colon; SRH, stigmata of recent hemorrhage; AB, active bleeding; NBVV, nonbleeding visible vessel; AC, adherent clot.

larly and consistently administered in 25 of 89 patients (28%). However, use of these agents was not associated with adverse outcomes of EC for diverticular hemorrhage, which simulates the results of a previous report on bleeding peptic ulcer.31 Right-sided diverticulosis has been demonstrated to be associated with more severe hemorrhage and rebleeding after TAE compared with left-sided diverticulosis.10,32 In this study, rebleeding in the left side of the colon was well managed by EC regardless of the technique. However, 1180 GASTROINTESTINAL ENDOSCOPY Volume 76, No. 6 : 2012

1. Burkitt DP, Walker AR, Painter NS. Effect of dietary fibre on stools and the transient-times, and its role in the causation of disease. Lancet 1972;2: 1408-12. 2. Ohi G, Minowa K, Oyama T, et al. Changes in dietary fiber intake among Japanese in the 20th century: a relationship to the prevalence of diverticular disease. Am J Clin Nutr 1983;38:115-21. 3. Lee YS. Diverticular disease of the largest bowel in Singapore: an autopsy survey. Dis Colon Rectum 1986;29:330-5. 4. Longstreth GF. Epidemiology and outcome of patients hospitalized with acute lower gastrointestinal hemorrhage: a population-based study. Am J Gastroenterol 1997;92:419-24. 5. Chaudhry V, Hyser MJ, Gracias VH, et al. Colonoscopy: the initial test for acute lower gastrointestinal bleeding. Am Surg 1998;64:723-8. 6. McGuire HH Jr. Bleeding colonic diverticula. A reappraisal of natural history and management. Ann Surg 1994;220:653-6. 7. Suzman MS, Talmor M, Jennis R, et al. Accurate localization and surgical management of active lower gastrointestinal hemorrhage with technetium-labeled erythrocyte scintigraphy. Ann Surg 1996;224: 29-36. 8. Farrell JJ, Friedman LS. Gastrointestinal bleeding in the elderly. Gastroenterol Clin North Am 2001;30:377-407. 9. Funaki B, Kostelic JK, Lorenz J, et al. Superselective microcoli embolization of colonic hemorrhage. AJR Am J Roentgenol 2001;177:829-36. 10. Tan KK, Nallathamby V, Wong D, et al. Can superselective embolization be definitive for colonic diverticular hemorrhage? An institution’s experience over 9 years. J Gastrointest Surg 2010;14:112-8. 11. Jensen DM, Machicado GA. Colonoscopy for diagnosis and treatment of severe lower gastrointestinal bleeding. Routine outcomes and cost analysis. Gastrointest Endosc Clin N Am 1997;7:477-98. 12. Jensen DM, Machicado GA, Jutabha R, et al. Urgent colonoscopy for the diagnosis and treatment of severe diverticular hemorrhage. N Engl J Med 2000;342:78-82.

www.giejournal.org

Ishii et al

EC for colonic diverticular hemorrhage

13. Bloomfeld RS, Rockey DC, Shetzline MA. Endoscopic therapy of acute diverticular hemorrhage. Am J Gastroenterol 2001;96:2367-72. 14. Mauldin JL. Therapeutic use of colonoscopy in active diverticular bleeding. Gastrointest Endosc 1985;31:290-1. 15. Hokama A, Uehara T, Nakayoshi T, et al. Utility of endoscopic hemoclipping for colonic diverticular bleeding. Am J Gastroenterol 1997;92:543-6. 16. Simpson PW, Nguyen MH, Lim JK, et al. Use of endoclips in the treatment of massive colonic diverticular bleeding. Gastrointest Endosc 2004;59:433-7. 17. Yen EF, Ladabaum U, Muthusamy VR, et al. Colonoscopic treatment of acute diverticular hemorrhage using endoclips. Dig Dis Sci 2008;53:2480-5. 18. Kaltenbach T, Watson R, Shah J, et al. Colonoscopy with clipping is useful in the diagnosis and treatment of diverticular bleeding. Clin Gastroenterol Hepatol 2012;10:131-7. 19. Witte JT. Band ligation for colonic bleeding: modification of multiband ligating devices for use with a colonoscope. Gastrointest Endosc 2000;52:762-5. 20. Farrell JJ, Graeme-Cook F, Kelsey PB. Treatment of bleeding colonic diverticula by endoscopic band ligation: an in-vivo and ex-vivo pilot study. Endoscopy 2003;35:823-9. 21. Ishii N, Itoh T, Uemura M, et al. Endoscopic band ligation with a water-jet scope for the treatment of colonic diverticular hemorrhage. Dig Endosc 2010;22:232-5. 22. Ishii N, Uemura M, Itoh T, et al. Endoscopic band ligation for the treatment of bleeding colonic and ileal diverticula. Endoscopy 2010;42:82-3. 23. Setoyama T, Ishii N, Fujita Y. Endoscopic band ligation (EBL) is superior to endoscopic clipping for the treatment of colonic diverticular hemorrhage. Surg Endosc 2011;25:3574-8.

24. Ishii N, Setoyama T, Deshpande GA, et al. Endoscopic band ligation (EBL) for colonic diverticular hemorrhage. Gastrointest Endosc 2012;75:382-7. 25. Foutch PG. Diverticular bleeding: are nonsteroidal anti-inflammatory drugs risk factors for hemorrhage and can colonoscopy predict outcome for patients? Am J Gastorenterol 1995;90:1779-84. 26. Benjamini Y, Hochberg Y. On the adaptive control of the false discovery rate in multiple testing with independent statistics. J Educ Behav Stat 2000;25:60-83. 27. Jensen DM, Machicado GA. Hemoclipping of chronic canine ulcers: a randomized, prospective study of initial deployment success, clip retention rates, and ulcer healing. Gastrointest Endosc 2009;70:969-75. 28. Meyers MA, Alonso DR, Gray GF, et al. Pathogenesis of bleeding colonic diverticulosis. Gastroenterology 1976;71:577-83. 29. Aldoori WH, Giovannucci EL, Rimm EB, et al. Use of acetaminophen and nonsteroidal anti-inflammatory drugs: a prospective study and the risk of symptomatic diverticular disease in men. Arch Fam Med 1998;7: 255-60. 30. Yamada A, Sugimoto T, Kondo S, et al. Assessment of the risk factors for colonic diverticular hemorrhage. Dis Colon Rectum 2008;51:116-20. 31. Choudari CP, Rajgopal C, Elton RA, et al. Failures of endoscopic therapy for bleeding peptic ulcers: an analysis of risk factors. Am J Gastroenterol 1994;89:1968-72. 32. Wong SK, Ho YH, Leong AP, et al. et al. Clinical behavior of complicated right-sided and left-sided diverticulosis. Dis Colon Rectum 1997;40: 344-8.

Availability of Journal back issues As a service to our subscribers, copies of back issues of Gastrointestinal Endoscopy for the preceding 5 years are maintained and are available for purchase from Elsevier until inventory is depleted. Please write to Elsevier Inc., Subscription Customer Service, 3251 Riverport Lane, Maryland Heights, MO 63043 or call 800-654-2452 or 314-447-8871 for information on availability of particular issues and prices.

www.giejournal.org

Volume 76, No. 6 : 2012 GASTROINTESTINAL ENDOSCOPY 1181