Approximately 5% of patients presenting with gastrointestinal

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2009;7:664 669 Long-Term Outcomes After Double-Balloon Enteroscopy for Obscure Gastrointestinal Bleeding LAUREN B. GERSON,* MELISSA A. BATENIC, SHARESE L. NEWSOM, ANDREW ROSS, and CAROL E. SEMRAD *Division of Gastroenterology and Hepatology, Stanford University, Stanford, California; and Division of Gastroenterology, University of Chicago, Chicago, Illinois Background & Aims: Long-term outcomes after doubleballoon enteroscopy (DBE), performed for the evaluation of obscure gastrointestinal bleeding, have not been determined. Methods: We invited 274 patients undergoing DBE at Stanford University or the University of Chicago between 2004 and 2006 to participate in the study; 135 (49%) agreed (mean age, 64 14.8; range, 23 90 years). Telephone interviews were conducted at a mean of 11 and 30 months after DBE. Results: Arteriovenous malformations (AVMs) were detected in 43% of the cohort. One hundred one patients (37%; 56 with overt, 45 with occult bleeding) were interviewed 12 5 (range, 6 26) months after DBE. At 12 months, 43% reported no further overt bleeding or iron/ transfusion needs, 23% reported overt bleeding, and 35% reported ongoing iron and/or transfusions. Eighty-five patients (31%) participated in the second survey, conducted 30 5.7 (range, 19 51) months after DBE. Fifty (59%) reported no overt bleeding or iron/transfusion needs, 20 (24%) reported overt bleeding, and 15 (18%) reported ongoing iron and/or transfusions. In the 40 patients evaluated after endoscopic treatment for AVMs, 17 (43%) reported no bleeding or iron therapy at 12 months and 16 of 29 (55%) at 30 months. Of the 40 patients with normal DBE examinations to the depth of insertion, 19 of 40 (48%) reported no bleeding or iron/transfusion needs at 12 months and 25 of 43 (58%) at 30 months. Conclusions: At 30 months after DBE, up to 60% of patients report no further bleeding. Patients with AVMs or normal examinations to the depth of insertion are most likely to report recurrent hemorrhage. Approximately 5% of patients presenting with gastrointestinal bleeding have no source found by upper endoscopy and colonoscopy. 1 In approximately 75% of these patients, responsible lesions can be detected in the small bowel. 2,3 Current options for diagnosis and management of small bowel lesions include push enteroscopy, capsule endoscopy (CE), doubleballoon enteroscopy (DBE), and intraoperative enteroscopy. CE is the examination recommended for patients with obscure gastrointestinal bleeding after negative upper and lower endoscopy but does not allow for therapeutic intervention. Multiple studies to date have demonstrated a 25% 55% higher diagnostic yield of CE compared with push enteroscopy, ileocolonoscopy, and small bowel series for patients with obscure gastrointestinal hemorrhage. 4 DBE, first described by Yamamoto et al 5 in 2001, allows for complete endoscopic visualization of the small intestine and therapeutics. Previously published studies reported that obscure bleeding was the indication for DBE in 36% 100% of examinations. The overall diagnostic yield from DBE ranged from 43% 80%. 6 8 Diagnostic or therapeutic success was reported in 55% 75% of examinations. Long-term outcomes after DBE for suspected small bowel bleeding have not been well-studied. Potential bleeding sources in the small bowel include vascular lesions, ulcerations, tumors, and diverticular disease. The purpose of this retrospective study was to determine the outcome of DBE on subsequent rates of recurrent bleeding, and iron and/or transfusion requirements in patients with obscure gastrointestinal bleeding. Methods Institutional Review Board approval was obtained at Stanford University Hospital and the University of Chicago, tertiary referral centers for DBE. We invited patients who underwent DBE for obscure overt or occult bleeding at Stanford University and the University of Chicago between August 2004 and November 2006 to participate in the study. Patients were sent a letter of invitation or called to explain the purpose of the study. To participate in the study, we required at least 6 months after a patient s DBE examination. Double-Balloon Enteroscopy Examinations For all patients, initial DBE approach was directed by findings from prior imaging studies including CE, radiographic examinations, or angiogram. When a prior study did not reveal a potential bleeding source, the oral approach was initiated empirically or determined by the type of overt bleeding. India ink tattooing was performed to mark the extent of the initial approach. The patient was offered repeat exam via the opposite approach if the initial exam was normal and a prior examination showed a potential bleeding site, or patients had ongoing bleeding or high suspicion for a small bowel bleeding source. The methodology proposed by May et al 9 was used to estimate depth of insertion into the small bowel. Technical failure was defined as inability to traverse past the ligament of Treitz during an upper examination or inability to intubate the ileum more than 20 cm during a retrograde examination. Total enteroscopy was defined by identification of cecum from the oral approach or identification of a tattoo placed at the distal extent of an initial approach on DBE examination from the opposite approach. In most humans, the length of Abbreviations used in this paper: AVM, arteriovenous malformation; CE, capsule endoscopy; CT, computed tomography; DBE, double-balloon enteroscopy; EGD, esophagogastroduodenoscopy; SD, standard deviation. 2009 by the AGA Institute 1542-3565/09/$36.00 doi:10.1016/j.cgh.2009.01.021

June 2009 DBE FOR OBSCURE GASTROINTESTINAL BLEEDING 665 the small intestine ranges between 300 and 1000 cm and is correlated with the height of the individual. 10 Potential or active bleeding lesions found at DBE were treated endoscopically when possible. Follow-Up Assessments The initial follow-up assessment occurred after a mean standard deviation (SD) follow-up period of 12 5 months (range, 6 26 months). The second assessment occurred after a mean follow-up of 30 5.7 months (range, 19 51 months). At each follow-up period, patients who agreed were interviewed by telephone after DBE examination. Patients were asked whether they experienced recurrent or new-onset overt bleeding or required post-dbe testing (including repeat upper or lower endoscopy, CE, DBE, and/or radiographic studies), surgery, and/or need for transfusions or iron therapy. We did not contact the patients primary gastroenterologists or primary care physicians. We did not obtain follow-up laboratory data on patients referred from outside medical facilities and, therefore, were unable to verify whether patients were taking iron for documented anemia and/or iron deficiency. In those who required continued iron or transfusion therapy, we did not determine whether needs were lessened after DBE. Statistical Analysis The clinical characteristics of patients in this study were compared, with mean and SD calculations compiled by using Microsoft Excel 2000 (Microsoft Corporation, Redmond, WA). Categorical data analysis was conducted by using the 2 test performed with Statistical Analyzing System software (SAS 9.1 for Windows; SAS Institute Inc, Cary, NC). The level of significance was set at P.05. Results We invited 274 patients (342 DBE examinations) to participate in the study; 135 (49%) patients agreed to participate, with a mean SD age of 64 14.8 years (range, 23 90 years). Sixty-eight (50%) were female, and 93 (69%) were white. The mean duration of bleeding before DBE examination was 12.6 18 months (range, 0.25 120 months) in the 89 patients with available information. Twenty-six patients initially did not sign informed consent forms or were unable to be contacted for the follow-up telephone interview after agreeing to participate. Eight patients with less than 6 months of follow-up after DBE were excluded. One hundred one (37%) patients were enrolled and completed the 12-month survey, including 56 patients with obscure overt bleeding and 45 patients with occult bleeding and/or iron deficiency anemia. At the second interview conducted at a mean of 30 months after DBE, 85 (31%) patients were able to be contacted including 45 patients with overt bleeding and 40 with occult bleeding and/or iron deficiency anemia (Figure 1). A total of 177 DBEs were performed including 117 (66%) anterograde and 60 (34%) retrograde examinations. The mean number of DBEs per patient was 1.5 0.8 (range, 1 5). Total enteroscopy was performed in only 8 (6%) patients. The cecum was not reached in any patients undergoing DBE via the anterograde approach. The majority (117, 66%) of the patients underwent initial oral DBE on the basis of CE examination reports that suggested pathology in the proximal two thirds of the small intestine by using small bowel transit time. Primary CE videos were reviewed when possible by the investigators before the DBE examinations. We did not collect information regarding the timing of the DBE examination and the capsule study or last episode of overt bleeding. The mean distance traversed during each DBE examination and the associated findings are shown in Table 1. Vascular lesions were more likely to be detected during the anterograde approach (43% on the oral approach versus 9% for retrograde, P.001). Technical failure was more common for rectal DBE procedures. We analyzed findings for the 139 patients who did not participate. Results from the DBE examinations included arteriovenous malformations (AVMs) in 24 (17%), examinations that were normal to depth of insertion in 90 (65%), and other findings in 25 (18%). There were significantly more patients who participated in the study with AVMs (P.001), whereas normal studies were more common in patients who did not participate (P.001). There were no differences between the groups for patients with other findings (P.06). Outcomes Data at Follow-Up Assessments Twelve-month follow-up period. One hundred one patients participated in the initial follow-up period that occurred at a mean SD of 12 5 months (range, 6 26 months) after DBE including 56 patients who presented with obscure Figure 1. Study flow diagram.

666 GERSON ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 7, No. 6 Table 1. Results From DBE Examinations (N 177 DBE Examinations on 135 Patients) Oral DBE, N 117 DBE results a (66% initial) Rectal DBE, N 60 P value b Mean SD distance (cm) (range) 340 164 (50 795) 238 193 (0 730) AVM 50 (43%) 4 (9%).001 Normal mucosa to depth of insertion 46 (39%) 28 (47%) Polyp or inflammatory tissue 8 (7%) 1 (2%) Ulcerative disorders c 5 (4%) 6 (10%) Dieulafoy s lesions 2 (2%) 1 (2%) Blood, unknown source 2 (2%) 2 (3%) Diverticulosis 2 (2%) 3 (5%) Blue rubber nevus 1 (0.8%) 0 (0%) Failed examination d 1 (0.8%) 15 (25%).001 a Total enteroscopy performed in only 8 patients. b Fisher exact test. All P values not significant unless otherwise noted. c Anastomotic ulcerations (N 2), NSAID-induced (N 1), Crohn s disease (N 1), radiation enteritis (N 1), chemotherapy-induced (N 1), and idiopathic (N 5). d Defined as inability to pass the ligament of Treitz on anterograde approach or failure to advance more than 20 cm in the distal ileum on retrograde examination. overt and 45 patients with obscure occult bleeding. Results for both patient cohorts are shown in Table 2. Overall, 43% of the cohort (19 [34%] patients with overt bleeding and 24 [53%] with occult hemorrhage) reported no overt bleeding or need for iron or transfusion therapy. Recurrent or new overt bleeding was reported in 23% of the cohort, and 35% reported transfusions or ongoing iron therapy. Six (13%) patients who initially presented with occult bleeding before DBE examination developed overt bleeding during the 12-month follow-up period. In the 40 patients with small bowel AVMs treated with argon plasma coagulation therapy during the initial DBE examinations, 17 (43%) reported no recurrence of bleeding or need for iron/transfusion therapy, 11 (28%) reported recurrent overt bleeding, and 12 (30%) reported ongoing need for iron or Table 2. Outcomes Data at 12-Month Follow-Up Period (N 101) DBE finding Treatment No recurrence a Overt bleeding b requiring transfusions c Iron therapy or Overt bleeding (N 56) Angiodysplasia (N 24) APC, 24 11 (46%) 9 (38%) 4 (16%) Ulceration (N 3) Chemotherapy, 1; none, 1; 3 surgery, 1 Dieulafoy s lesion (N 2) APC, 2 2 Gastropathy (N 2) None, 2 2 Colonic diverticulosis (N 2) None, 1 1 Blue rubber nevus (N 1) None, 1 1 1 Blood, unknown source (N 1) None, 1 1 Failed retrograde (N 3) None, 3 2 1 Normal to depth of insertion (N 18) None, 18 5 (28%) 6 (33%) 7 (39%) Total, overt 19/56 (34%) 17/56 (30%) 20/56 (36%) Occult bleeding (N 45) d Angiodysplasia (N 16) APC, 16 6 (38%) 2 (12%) 8 (50%) Ulceration (N 2) Crohn s therapy, 1; none, 1 2 Dieulafoy s lesion (N 1) APC, 1 1 Paraesophageal hernia (N 1) None, 1 1 Vascular lesion (N 1) Surgery, 1 1 Carcinoid (N 1) Surgery, 1 1 Failed retrograde (N 1) None, 1 1 Normal to depth of insertion (N 22) None, 22 14 (64%) 3 (14%) 5 (23%) Total, occult 24/45 (53%) 6/45 (13%) 15/45 (33%) Summary 12-month outcomes 43/101 (43%) 23/101 (23%) 35/101 (35%) APC, argon plasma coagulation. a Defined as absence of overt bleeding or iron therapy. b Defined as active bleeding with or without iron therapy and transfusions. c Without overt bleeding. d Initial presentation before DBE with occult bleeding.

June 2009 DBE FOR OBSCURE GASTROINTESTINAL BLEEDING 667 Table 3. Subsequent Procedures After DBE and Findings AVM Normal Other a Recurrent bleeding/anemia at 12 months (N 34/101) Angiography 4 1 2 DBE 1 7 CE 1 3 EGD and/or colonoscopy 1 9 Surgery/intraoperative DBE 1 2 Recurrent bleeding/anemia at 30 months (N 19/73) DBE 3 CE EGD/colonoscopy 1 Surgery 7 Total 9 23 11 a Angiography findings included gastrointestinal stromal tumor (1) and vasculitis (1). Surgery findings included gastrointestinal stromal tumor (1), paraesophageal hernia (1), small bowel resection with Crohn s disease (1), small bowel resection with metastatic carcinoid (2), ruptured appendix (1), mesorenal shunt for portal hypertension (1), surgical resection for radiation enteritis (1), replacement of St Jude s valve with porcine valve (1). transfusion therapy during the mean follow-up period of 12 5 months. Thirty-two procedures were performed in 34 patients with persistent anemia or bleeding after DBE, with the findings shown in Table 3. Twenty (62%) of the subsequent procedures were normal. Eight of 23 patients (35%) with recurrent anemia or bleeding after DBE treatment of AVMs by using argon plasma coagulation therapy had subsequent endoscopic or radiographic procedures. Findings that were treated included AVMs on angiography (4), gastric AVMs (1), small bowel AVMs on DBE (2), and other findings requiring surgery in 4 patients. Seven patients had died at the time of the initial interview but had follow-up information available. Causes of death included renal failure (1), myocardial infarction (2), lung cancer (2), unknown (1), and 1 patient died of gastrointestinal hemorrhage complications after refusing further endoscopic or surgical therapy for angiodysplastic lesions. Thirty-month follow-up period. Eighty-five patients (31% of the patients invited) were contacted at the second follow-up period, which occurred at a mean of 30 5.7 months (range, 19 51 months). Patients who participated in the 30- month survey included 65 patients who had been interviewed at 12 months and an additional 20 patients who had agreed to participate but were unable to be contacted at 12 months (N 15) or patients who had less than 6 months of follow-up and were excluded from the 12-month analysis (N 5). Table 4 lists the outcomes data for the 30-month follow-up period. Overall, 50 (59%) of the cohort reported no overt bleeding or need for iron and/or transfusion therapy. Overt bleeding was noted in 20 (34%), and 15 (18%) reported ongoing iron or transfusion needs. Cessation of bleeding was reported in 16 of 29 (55%) patients with AVMs on initial DBE and in 25 of 43 (58%) patients with examinations that were normal to the extent of insertion of the endoscope. AVMs that were found during subsequent testing were treated. Thirteen of 29 (45%) patients with AVMs reported overt bleeding or need for iron and/or transfusions despite treatment at 30 months. Follow-up procedures at the 30-month follow-up period are listed in Table 4. Four patients with initial normal DBE examinations underwent subsequent partial small bowel resections, with findings including pancreatic mass, Crohn s disease, metastatic carcinoid, and ruptured appendicitis. Seven of the patients with AVMs initially diagnosed on DBE had subsequent overt bleeding and underwent other procedures including placement of a mesorenal shunt for portal hypertension (N 1), esophagogastroduodenoscopy (EGD) or DBE with cautery of AVMs (N 4), replacement of a St Jude s valve with a porcine valve replacement (N 1), and resection of carcinoid tumor (1). Persistent anemia was explained by the development of chronic renal insufficiency in 2 patients and myelodysplastic syndrome in another patient. At the time of the second assessment, 7 more patients had died. Causes of death included cardiac (3), jejunal cancer complications (1), complications after resection for Crohn s disease (1), myelodysplastic syndrome and complications in a patient who received chemotherapy for non-hodgkin s lymphoma (1), and septic shock in a patient with Rendu-Osler-Weber syndrome (1). Discussion We performed this retrospective study to determine patient outcomes after DBE performed for obscure gastrointestinal bleeding. The most common findings on DBE were AVMs detected within an anterograde approach or normal mucosa to the depth of insertion. However, total enteroscopy was only possible in a small number of patients (8%) when attempted, and not all patients returned for total enteroscopy from the opposite approach when recommended. In our study, 43% of patients reported cessation of bleeding at 12 months and 59% at 30 months. The higher rate of bleeding cessation at 30 months was likely due to definitive DBE endoscopic therapy or surgical procedures after the initial 12-month assessment. The rates of recurrent overt bleeding and/or need for ongoing iron therapy or transfusions were highest in patients with AVMs or examinations that showed normal mucosa to the extent of insertion of the endoscope. We could not reliably determine whether iron and transfusion requirements lessened after DBE therapy in those with ongoing needs. A prior study published by Sun et al 11 in 2006 examined outcomes in patients undergoing DBE for obscure gastrointestinal bleeding. They studied 152 patients who had undergone 191 DBE procedures and found small bowel tumors in 39% of the cohort and AVMs in 30%. At a mean follow-up of 16 months in 119 patients, 85 (89%) had no further bleeding reported. Of their cohort, 38% underwent surgery, and 16% had endoscopic management. Our study showed higher rebleeding rates, likely as a result of the low prevalence of tumors that would be amenable to surgical resection and higher rates of AVMs. The high rate of cessation of bleeding in patients with normal DBE examinations to the extent of insertion of the endoscope might have been due to false-positive or equivocal capsule or radiographic findings, missed AVMs that spontaneously stopped bleeding, healed lesions, or cessation of anticoagulation. Limitations of our study included the low patient response rate, the small number of patients undergoing total enteroscopy, lack of knowledge regarding ongoing usage or discontinuation of warfarin, aspirin, and/or other antiplatelet agents, and lack of specific data for transfusion requirements, hematocrit, and iron levels after

668 GERSON ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 7, No. 6 Table 4. Outcomes Data at 30-Month Follow-Up After DBE (N 85) DBE findings Additional treatment since Iron therapy or 11-month follow-up No recurrence a Overt bleeding b requiring transfusions c Overt bleeding (N 45) Angiodysplasia (N 14) EGD with APC, 1; valve replacement, 1; mesorenal shunt portal hypertension, 1 8 (57%) 6 (43%) Ulcerations (N 2) None, 2 2 Dieulafoy s lesion (N 1) None, 1 1 Meckel s diverticulum (N 1) Surgery, 1 1 Clots, no source (N 1) Surgery, gastrointestinal stromal tumor, 1 1 Blue rubber nevus (N 1) Octreotide, 1 1 Gastropathy (N 1) None, 1 1 Normal (N 24) Small bowel resections, 4 e 13 (54%) 5 (22%) 6 (25%) Total, overt 27/45 (60%) 12/45 (27%) 6/45 (13%) Occult bleeding (N 40) d Angiodysplasia (N 14) APC, 3; carcinoid 8 (57%) 2 (14%) 4 (29%) resection, 1 Ulcerations (N 3) Small bowel resection, 1 1 1 1 Dieulafoy s lesion (N 1) None, 1 1 Vascular (N 1) None, 1 1 Carcinoid (N 1) None, 1 1 Paraesophageal hernia (N 1) None, 1 1 Normal (N 19) None, 15 12 (63%) 5 (26%) 2 (10%) Total, occult 23/40 (58%) 8/40 (20%) 9/40 (35%) Summary 30-month outcomes 50/85 (59%) 20/85 (24%) 15/85 (18%) APC, argon plasma coagulation. a Defined as absence of overt bleeding or iron therapy. b Defined as active bleeding with or without iron therapy and transfusions. c Without overt bleeding. d Initial presentation with occult bleeding. e Small bowel resections for pancreatic mass (1), Crohn s disease (1), carcinoid (1), and ruptured appendicitis with small bowel involvement (1). DBE. Most of the patients in this study were referred by other gastroenterologists for DBE, which limited follow-up and complete chart reviews for medication and laboratory data. DBE examinations included those performed at the early part of the learning curve, which likely impacted on the number of incomplete retrograde examinations and completed enteroscopies. Previous abdominal surgery with fixed bowel and large body habitus found in approximately one third of our patients 6 also impacted on the low number of completed enteroscopies. The factors that trigger bleeding from AVMs and the natural history of vascular lesions in the small bowel have not been well-characterized. Estimates of spontaneous cessation rate can be derived primarily from clinical trials comparing hormonal therapy with placebo for the treatment of AVMs in the gastrointestinal tract. The spontaneous cessation rate in these patients has been reported to range from 40% to 50% per year. 12,13 Despite the fact that AVMs were often found during subsequent testing and treated in our patients, 13 of 29 (45%) reported overt bleeding or need for iron and/or transfusions at 30 months. In a recent study examining short-term outcomes after DBE, 30 patients with positive findings on DBE were followed for a mean of 2 months. Therapeutic interventions were performed in 18 of 30 (60%). The rate of rebleeding was similar for patients, regardless of endoscopic therapy (20% in the treated group compared with 18% in the nontreated cohort). 14 In some patients, ongoing bleeding might be due to the need for continued anticoagulation therapy. Risk factors for recurrent bleeding in patients with small bowel AVMs deserve further study. Our findings are similar to an outcomes study after CE performed for obscure bleeding published by Lai et al, 15 demonstrating that the highest 1-year rebleeding rates were in patients with small bowel AVMs (58%) or when bleeding from an unknown source (54%) was observed on CE examination. As demonstrated in our results (Table 1), the majority of the AVMs were detected on anterograde DBE examination, consistent with prior data that most AVMs are located within the proximal two thirds of the small bowel. 16 In patients at greatest risk for small bowel AVMs presenting with obscure overt bleeding, empiric anterograde DBE approach without initial CE has been shown to be cost-effective but also results in increased endoscopic workload and potential for complications. 17 The reasons for the higher rebleeding rates in the patients with small bowel AVMs might include inability to reach the lesions by DBE examination, recurrence of AVMs after initial endoscopic therapy, or submucosal location of AVMs rendering them inaccessible to endoscopic therapy. In the current study when patients had AVMs that were not detected or accessible by DBE, angiography was successful in 4 of the patients who exhibited recurrent bleeding after DBE. Computed tomography (CT) angiography is an emerging option for visualization of AVMs when CE or DBE examinations are negative. 18 We treated all small bowel AVMs, regardless of whether there was active bleeding. A randomized controlled trial compar-

June 2009 DBE FOR OBSCURE GASTROINTESTINAL BLEEDING 669 ing endoscopic therapy with DBE versus sham therapy for AVMs is needed to determine the impact of endoscopic therapy on such lesions. In 21 (25%) of the 84 capsule examinations, there were findings on the CE examination that were not detected on subsequent DBE examination including bleeding without a source in 4 patients and potential AVMs in 10 patients. Although polyps and masses seen on CE might be false positives, depending on the experience of the CE reader, the findings of blood on CE without a source warrant further investigation with either radiographic studies or repeat endoscopic procedures, given the high risk of rebleeding in this cohort. In addition, our results showed that patients with incomplete or normal DBE examinations should undergo further evaluation in the setting of recurrent anemia and/or overt bleeding, because a large percentage of these patients will have AVMs or extraluminal sources that can be detected via radiographic imaging studies. A recent meta-analysis determined that the yields of CE and DBE were equivalent for vascular lesions as well as for inflammatory lesions and polyps or tumors. 19 Although we did not interpret the majority of the CE studies, we found that in 36 of the CE studies in which findings were present, there was agreement with DBE in 94%. In 9 studies, findings on CE were discordant with DBE findings. In 10% of examinations, there were findings on DBE after normal CE examinations. Prior comparative studies have demonstrated miss rates for CE and DBE ranging between 20% and 30%. 6 Particularly in the setting of active bleeding, a DBE examination might be a reasonable approach, even after a negative CE. In summary, this retrospective study of patients undergoing DBE for obscure gastrointestinal bleeding showed that approximately 60% of patients stopped bleeding at an average of 30- month follow-up. Patients with small bowel AVMs and patients with DBE examinations that were normal to the extent of insertion of the endoscope were most likely to experience recurrent obscure bleeding. Rebleeding occurred in approximately one third of the cohort at the 12-month and 30-month assessment periods. In the challenging subset of patients with recurrent bleeding after DBE, options to consider include repeat CE to examine for pathologic findings distal to the tattoo site, DBE via the opposite approach, and/or abdominal imaging with CT enterography or angiography. However, in the setting of small bowel AVMs, it is unclear whether ongoing diagnostic testing and therapy will change the rebleeding rates. Supplementary Data Note: To access the supplementary material accompanying this article, visit the online version of Clinical Gastroenterology and Hepatology at www.cghjournal.org, and at doi:10.1016/j.cgh.2009.01.021. References 1. Szold A, Katz LB, Lewis BS. Surgical approach to occult gastrointestinal bleeding. Am J Surg 1992;163:90 93. 2. Pennazio M, Arrigoni A, Risio M, et al. Clinical evaluation of push-type enteroscopy. Endoscopy 1995;27:164 170. 3. Chong J, Tagle M, Barkin JS, et al. Small bowel push-type fiberoptic enteroscopy for patients with occult gastrointestinal bleeding or suspected small bowel pathology. Am J Gastroenterol 1994;89:2143 2146. 4. Lewis BS, Eisen GM, Friedman S. A pooled analysis to evaluate results of capsule endoscopy trials. Endoscopy 2005;37:960 965. 5. Yamamoto H, Sekine Y, Sato Y, et al. Total enteroscopy with a nonsurgical steerable double-balloon method. Gastrointest Endosc 2001;53:216 220. 6. Mehdizadeh S, Ross A, Gerson L, et al. What is the learning curve associated with double-balloon enteroscopy? Technical details and early experience in 6 U.S. tertiary care centers. Gastrointest Endosc 2006;64:740 750. 7. Monkemuller K, Weigt J, Treiber G, et al. Diagnostic and therapeutic impact of double-balloon enteroscopy. Endoscopy 2006; 38:67 72. 8. Di Caro S, May A, Heine DG, et al. The European experience with double-balloon enteroscopy: indications, methodology, safety, and clinical impact. Gastrointest Endosc 2005;62:545 550. 9. May A, Nachbar L, Schneider M, et al. Push-and-pull enteroscopy using the double-balloon technique: method of assessing depth of insertion and training of the enteroscopy technique using the Erlangen Endo-Trainer. Endoscopy 2005;37:66 70. 10. Underhill BM. Intestinal length in man. Br Med J 1955;2:1243 1246. 11. Sun B, Rajan E, Cheng S, et al. Diagnostic yield and therapeutic impact of double-balloon enteroscopy in a large cohort of patients with obscure gastrointestinal bleeding. Am J Gastroenterol 2006; 101:2011 2015. 12. Junquera F, Feu F, Papo M, et al. A multicenter, randomized, clinical trial of hormonal therapy in the prevention of rebleeding from gastrointestinal angiodysplasia. Gastroenterology 2001; 121:1073 1079. 13. Lewis BS, Salomon P, Rivera-MacMurray S, et al. Does hormonal therapy have any benefit for bleeding angiodysplasia? J Clin Gastroenterol 1992;15:99 103. 14. Madisch A, Schmolders J, Bruckner S, et al. Less favorable clinical outcome after diagnostic and interventional double balloon enteroscopy in patients with suspected small-bowel bleeding? Endoscopy 2008;40:731 734. 15. Lai LH, Wong GL, Chow DK, et al. Long-term follow-up of patients with obscure gastrointestinal bleeding after negative capsule endoscopy. Am J Gastroenterol 2006;101:1224 1228. 16. Gay G, Delvaux M, Fassler I. Outcome of capsule endoscopy in determining indication and route for push-and-pull enteroscopy. Endoscopy 2006;38:49 58. 17. Gerson L, Kamal A. Cost-effectiveness analysis of management strategies for obscure GI bleeding. Gastrointest Endosc 2008; 68:920 936. 18. Huprich JE, Fletcher JG, Alexander JA, et al. Obscure gastrointestinal bleeding: evaluation with 64-section multiphase CT enterography initial experience. Radiology 2008;246:562 571. 19. Pasha SF, Leighton JA, Das A, et al. Double-balloon enteroscopy and capsule endoscopy have comparable diagnostic yield in small-bowel disease: a meta-analysis. Clin Gastroenterol Hepatol 2008;6:671 676. Reprint requests Address requests for reprints to: Lauren B. Gerson, MD, MSc, Division of Gastroenterology and Hepatology, Stanford University, A149, 300 Pasteur Drive, Stanford, California 94305-5202. e-mail: lgerson@ stanford.edu; fax: (650) 723-8305. Conflicts of Interest These authors disclose the following: Drs Gerson, Semrad, and Ross have received equipment and speaker s honorarium from Fujinon Inc. Dr Gerson has received speaker s honorarium from Given Imaging Inc. The remaining authors disclose no conflicts.