- Email: [email protected]
Is there a role for second-look capsule endoscopy in patients with obscure GI bleeding after a nondiagnostic first test?
ORIGINAL ARTICLE: Clinical Endoscopy
Is there a role for second-look capsule endoscopy in patients with obscure GI bleeding after a nondiagnostic first test? Nikos Viazis, MD, Kostis Papaxoinis, MD, John Vlachogiannakos, MD, Alkiviadis Efthymiou, MD, Ioannis Theodoropoulos, MD, Dimitrios G. Karamanolis, MD Athens, Greece
Background: Long-term follow-up data on patients with obscure GI bleeding subjected to capsule endoscopy (CE) are missing. Objective: Our purpose was to follow up patients with a nondiagnostic test and determine whether a secondlook CE would be beneficial. Patients: We enrolled 293 subjects. CE studies were classified as diagnostic (positive findings) or nondiagnostic (findings of uncertain significance/no findings). Patients were followed up for a mean (SD) 24.8 (5.2) months. Outcome was defined as continued or complete resolution of bleeding. Interventions: Patients with a nondiagnostic test were subjected to a repeat CE if they manifested a new bleeding episode or a drop in hemoglobin R2 g/dL. Results: Positive findings, findings of uncertain significance, and no findings were identified in 41.6%, 16.0%, and 42.3% of our patients, respectively. Therapeutic intervention was possible in 72.1% of those with positive findings and in 30% of those with findings of uncertain significance. Complete resolution of bleeding occurred more often in patients with a diagnostic test (65.2%) compared with those with a nondiagnostic test (35.4%, P! .001). Second-look CE was performed in a subgroup of our patients (n Z 76) and was diagnostic in those whose presentation changed from occult to overt or those whose hemoglobin dropped R4 g/dL. Conclusions: In patients with obscure GI bleeding, a diagnostic CE leads to therapeutic interventions and a favorable outcome. Patients with a nondiagnostic test would definitely benefit from a second-look CE if the bleeding presentation changes from occult to overt or if the hemoglobin value drops R4 g/dL. (Gastrointest Endosc 2009;69:850-6.)
Copyright ª 2009 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 doi:10.1016/j.gie.2008.05.053
ranges from 45% to 75%, clearly reflecting differences in image interpretation and the diagnostic criteria used for the identification of the bleeding condition.4-12 We have previously reported that positive findings, defined by strict diagnostic criteria, were identified in 41.7% of patients subjected to CE for the investigation of OGIB.13 After a median follow-up period of 14 months, those patients were more likely to receive some kind of therapeutic intervention for their lesions and eventually had a favorable outcome. On the contrary, the outcome of those patients with findings of uncertain significance or no findings on CE was not changed in a positive way. In clinical practice, the most appropriate management of patients with OGIB and a nondiagnostic CE test remains elusive, and we do not know whether these subjects would benefit from a second-look CE. Meanwhile, data on follow-up longer than 1 year are sparse in the literature, further obscuring our knowledge on the best therapeutic algorithm that needs to be followed in these cases.
850 GASTROINTESTINAL ENDOSCOPY Volume 69, No. 4 : 2009
www.giejournal.org
Obscure GI bleeding (OGIB) remains a major clinical challenge. The majority of OGIB cases originate from the small intestine; therefore, wireless capsule endoscopy (CE) is a valuable tool in the investigation of these patients.1-3 Although sound literature data have proven the superiority of CE compared with all other diagnostic modalities in identifying bleeding lesions,4-12 the true diagnostic yield of this test in patients with OGIB is still not known. Numerous previous reports have reported conflicting results, and according to the available data the diagnostic yield of CE Abbreviations: CE, capsule endoscopy; DBE, double-balloon enteroscopy; NSAIDs, nonsteroidal anti-inflammatory drugs; OGIB, obscure GI bleeding. DISCLOSURE: All authors disclosed no financial relationships relevant to this publication.
Viazis et al
Second-look capsule endoscopy after nondiagnostic first test
TABLE 1. Characteristics of the study population
Capsule Summary
Age (y) (mean [SD])
What is already known on this topic
Male/female Hemoglobin (g/dL) (mean [SD], minimum-maximum) Overt GI bleeding of obscure origin
59.5 (14.4) 169:124 9.9 (1.68), 7.2-12.4 146
Hematochezia
61
Melena
85
Occult GI bleeding of obscure origin Iron deficiency anemia Positive fecal occult blood test
d
147 108
Despite capsule endoscopy (CE), a substantial number of patients with occult GI bleeding (OGIB) do not receive a diagnosis, and a second CE may be used.
What this study adds to our knowledge d
d
In 293 subjects at a single institution, 65% of individuals with a nondiagnostic initial CE continued to manifest OGIB after a mean follow-up period of 24 months. Development of overt bleeding and a hemoglobin drop of 4 g/dL or more were significant predictive factors of a diagnostic second CE.
39
From September 2002 to January 2007, all patients referred to our unit for CE because of overt or occult OGIB were considered for entry into the study. We excluded patients with known or suspected GI obstruction, strictures or fistulae; patients with a cardiac pacemaker, defibrillator, or other implanted electromedical device; pregnant women; patients under 18 years of age; and patients who refused consent. Patients with suspected GI obstruction were excluded because the capsule can become stuck at points of luminal narrowing, with resultant acute bowel obstruction. Patients with a pacemaker were excluded because of fear of interference of the pacemaker by the capsule, although, so far, such an incidence has not been reported. We defined bleeding of obscure origin as bleeding of unknown origin that persists or recurs (ie, recurrent or persistent iron deficiency anemia, fecal occult blood test positivity, or visible bleeding) after a negative initial workout including gastroscopy, colonoscopy, small bowel barium follow-through or enteroclysis, and push enteroscopy. Bleeding of obscure origin was further subdivided in 2 clinical entities: (1) obscure: occult, as manifested by recurrent iron-deficiency anemia or recurrent positive fecal occult blood test results and (2) obscure: overt, with recurrent passage of visible blood (hematochezia/ melena). CE was performed with the Given M2A video capsule system (Given Imaging, Yogneam, Israel), as reported previously.14 All patients ingested the capsule at 9 AM after an
overnight fast. They were also given a bowel preparation with 2 L of polyethylene glycol electrolyte lavage solution 16 hours before the investigation because this has been shown to increase the diagnostic yield of CE.15 In addition, they were advised to consume clear liquids from the time they started receiving the bowel purge until the time they went to bed that same night. After ingestion, nonhospitalized patients were free to go home and continue their usual activities, but they were asked to contact us if they noted abdominal pain, nausea, or vomiting. All patients were advised not to drink and eat for a 3- and a 4-hour period, respectively. At the end of the recording period, the equipment was removed. Two investigators (N. V., K. P.) independently evaluated the video capsule endoscopy images at a movie speed of 10 frames per second. The investigators were blinded as to whether the patients were referred for the evaluation of overt or occult bleeding of obscure origin. Findings were considered positive if the investigators could explain the patient’s symptoms or help further management or if the findings were later confirmed by other diagnostic modalities. Findings were considered of uncertain significance if they failed to completely explain the patient’s symptoms, thus necessitating further investigation. Finally, when no abnormality was detected, despite a definite indication of an existing lesion, the test was considered as having no findings. All patients were monitored closely at 3- and 6- month intervals and every 6 months thereafter for a survey of their final outcome. Outcome was described as either continued bleeding or complete resolution of bleeding (regardless of whether treatment was given). For all patients the following follow-up information was obtained: use of medical, endoscopic, or surgical treatment; further bleeding episodes; persistence of iron deficiency anemia; or positivity of fecal occult blood testing. This information was sought through telephone interviews, contact of the referring physicians, or follow-up visits of the patients themselves.
www.giejournal.org
Volume 69, No. 4 : 2009 GASTROINTESTINAL ENDOSCOPY 851
We therefore prospectively evaluated all patients with OGIB who underwent CE at our institution to determine their outcomes and to investigate whether those with a nondiagnostic test would benefit from a second-look CE.
PATIENTS AND METHODS
Second-look capsule endoscopy after nondiagnostic first test
TABLE 2. CE findings
TABLE 3. Outcome data for patients who received treatment because of the results of CE Overt GI Occult GI bleeding of bleeding of obscure origin obscure origin (No. of patients) (No. of patients)
Positive findings
58
64
Multiple or bleeding angiodysplasias
50
36
Multiple apthoid ulcerations
0
10
Mucosal ulcers
4
8
Bleeding polyps
4
4
Small bowel tumors
0
6
29
18
Isolated, nonbleeding angiodysplasias
15
8
Isolated apthoid ulceration
4
0
Mucosal erosion
4
4
Small nonbleeding polyp
0
6
Blood within the small bowel lumen
6
0
59
65
Findings of uncertain significance
No findings
Viazis et al
Treatment (No. of patients)
Resolution of bleeding (No. of patients)
Multiple or bleeding angiodysplasias
52
36
Multiple apthoid ulcerations
10
10
Mucosal ulcers
12
10
Bleeding polyps
8
8
Small bowel tumors
6
6
10
2
4
2
Capsule findings Positive findings
Findings of uncertain significance Isolated nonbleeding angiodysplasias Mucosal erosions
dependent predictors of a definite CE diagnosis. The various factors examined for correlation with the diagnostic yield at a second CE test were defined before the study. A P! .05 was considered statistically significant. Statistical analysis was performed with SPSS 11.0 for Windows (SPSS, Chicago, Ill).
RESULTS
Values are expressed as mean SD. Categorical variables were compared with the c2 test with Yates’ correction or the Fisher exact test, as appropriate. Logistic regression was used to identify the variables that were in-
During the study period, 305 patients were referred to our unit for CE because of OGIB. Six with subclinical small intestinal stenosis detected during small bowel followthrough were excluded; 2 patients were younger than 18 years old and were excluded; 1 had a pacemaker and was excluded; and 3 refused consent. Therefore, the study sample consisted of 293 patients. The characteristics of the study population are presented in Table 1. All patients ingested the capsule without difficulty, and the test was well tolerated without adverse events. The capsule reached the cecum (ie, the complete small bowel was visualized) in 220 patients (75.1%). Positive findings were identified in 122 of our patients, leading to an overall diagnostic yield of 41.6%. More precisely, positive findings were identified in 58 patients with overt GI bleeding (diagnostic yield 39.7%) compared with 64 patients with OGIB of obscure origin (diagnostic yield 43.5%) (not significant). As previously noted, the most common positive finding was angiodysplasia of the small intestine. Findings of uncertain significance were identified in 47 patients (16%, overt 29/occult 18), whereas in 124 patients the
852 GASTROINTESTINAL ENDOSCOPY Volume 69, No. 4 : 2009
www.giejournal.org
During the follow-up period, a second-look CE was performed in those patients with a nondiagnostic first test (ie, findings of uncertain significance, no findings) when a new bleeding episode (hematochezia/melena) or a drop in hemoglobin of 2 g/dL or more were reported. The video capsule images were evaluated by the same two investigators who had evaluated the initial capsule study, and they were blinded as to whether the patients had a new bleeding episode or were subjected to the test for a drop in hemoglobin. The study protocol adhered to the principles of the Declaration of Helsinki and was approved by the ethics committee of our hospital. All patients gave written informed consent at the time of enrollment.
Statistical analysis
Viazis et al
Second-look capsule endoscopy after nondiagnostic first test
TABLE 4. Second-look CE findings Overt GI bleeding of obscure origin (No. of patients)
Occult GI bleeding of obscure origin (No. of patients)
20
17
15
12
Multiple apthoid ulcerations
0
0
Mucosal ulcers
2
3
Bleeding polyps
2
1
Small-bowel tumors
1
1
4
18
Isolated, nonbleeding angiodysplasias
1
7
Isolated apthoid ulceration
1
5
Mucosal erosion
1
4
Small nonbleeding polyp
0
2
Blood within the small-bowel lumen
1
0
4
13
Positive findings Multiple or bleeding angiodysplasias
Findings of uncertain significance
No findings
test did not reveal any findings (42.3%, overt 59/occult 65). The types of lesions seen are presented in Table 2. The degree of agreement between the 2 endoscopists regarding the identification of the bleeding site condition was 95% in studies with positive findings, 90% in studies with findings of uncertain significance, and 94% in studies with no findings. Of the 122 patients with positive findings at CE, 88 (72.1%) received some kind of treatment for their lesions or their management was changed accordingly (Table 3). Twenty-six patients with angiodysplasia of the duodenum and proximal jejunum underwent endoscopic treatment with argon plasma coagulation; 10 patients with bleeding angiodysplasia of the distal jejunum and ileum underwent surgery; 16 patients with multiple angiodysplasias received estrogen therapy (although this strategy is of unproven value); 10 patients with findings suggestive of Crohn’s disease were begun on corticosteroids; 12 patients with small bowel ulcers were advised to discontinue nonsteroidal anti-inflammatory drug (NSAID) ingestion; and finally, 14 patients with bleeding polyps or tumors underwent surgical resection. Of the 47 patients with findings of uncertain significance at CE, 14 (30%) either received some kind of treatment for their lesions or their management was changed accordingly. Ten patients with angiodysplasia of the proximal jejunum underwent endoscopic treatment with argon plasma coagulation and 4 patients with mucosal erosions were advised to discontinue NSAIDs (Table 3). Outcome data are available for 279 patients (positive findings 118/findings of uncertain significance 47/no find-
ings 114) because 14 patients were lost at follow-up. The mean (SD) follow-up time was 24.8 (5.2) months (range 7-29 months). From the 118 patients with a diagnostic test (positive findings), complete resolution of bleeding occurred in 77 (65.2%) compared with 57 of the 161 patients (35.4%) with a nondiagnostic test (findings of uncertain significance or no findings) (P ! .001). During the follow-up period, of the 104 patients with a nondiagnostic first test whose bleeding did not resolve, 76 were subjected to a second-look CE, whereas the remaining 28 patients were not because the degree of their anemia was considered to be stable (ie, drop in hemoglobin less than 2 g/dL). Of the 76 patients subjected to the secondlook capsule study, 28 had a new bleeding episode (ie, hematochezia/melena) regardless of whether the hemoglobin value dropped, and 48 had a drop in hemoglobin of 2 g/dL or more. From the 28 patients with an overt new bleeding episode, 17 had overt OGIB at initial presentation and 11 had occult OGIB at initial presentation. All 48 patients who were subjected to a second CE test because of a drop in hemoglobin of 2 g/dL or more had occult OGIB at initial presentation. The results of the second-look CE test are presented in Table 4. Once again, the most common positive finding was angiodysplasia of the small intestine, whereas mucosal ulcers, a bleeding polyp, and a small bowel tumor of the terminal ileum (in a patient whose first CE did not reach the cecum) were also seen. It is of note that among the 37 patients with a diagnostic second-look CE, 7 had an incomplete first test (ie, the capsule did not reach the cecum), whereas the relevant number among the 39 with a nondiagnostic second-look CE was 13 (P Z .19).
www.giejournal.org
Volume 69, No. 4 : 2009 GASTROINTESTINAL ENDOSCOPY 853
Second-look capsule endoscopy after nondiagnostic first test
Viazis et al
TABLE 5. Logistic regression analysis of the predictive values of the evaluated parameters regarding second-look CE diagnosis certainty 95% Confidence limits Estimated coefficient
SE
Wald statistic
Significance
Odds ratio
Lower
Upper
Age (y)
0.015
0.025
0.381
.537
0.985
0.938
1.034
Sex (male)
0.756
0.618
1.497
.221
2.130
0.634
7.152
OGIB initial presentation (overt)
0.970
0.603
2.585
.108
2.638
0.809
8.610
Incomplete first CE
1.117
0.836
1.785
.182
3.055
0.594
15.727
Agreement between examiners
0.851
0.637
1.781
.182
2.341
0.671
8.166
Treatment after first CE
0.856
0.644
1.767
.184
2.353
0.666
8.308
OGIB relapse presentation (overt)
0.775
0.654
1.407
.236
2.171
0.603
7.818
OGIB change (occult to overt)
2.526
0.985
6.572
.010
0.080
0.012
0.552
Hemoglobin drop (more than 4 mg/dL)
2.095
0.862
5.902
.015
0.123
0.023
0.667
A substantial number of patients subjected to CE for the investigation of OGIB will end up having a nondiagnostic test. Literature data regarding the long-term outcome of these patients and options for their best management are missing. According to our results, 65% of patients with a nondiagnostic first CE test continued to manifest OGIB after a mean follow-up period of 24.8 months, and a change in their bleeding presentation from occult to overt or a drop of their hemoglobin level of 4g/dL or more were both significant predictive factors of a diagnostic second CE. Management of patients with OGIB remains a challenging problem in clinical practice. In our study obscure bleeding was defined according to the American Gastroen-
terological Association technical review16; however, we have extended our initial workup to include not only colonoscopy and upper endoscopy but also small-bowel follow-through and push enteroscopy. Therefore, our study population consisted of patients with bleeding who had failed to reach a diagnosis although they had been subjected to all conventional radiographic and endoscopic techniques. In clinical practice, when a bleeding source in not identified in upper or lower GI endoscopy, the small bowel is interrogated and wireless CE has now become the method of choice for this purpose. Indeed, the diagnostic yield of CE has been demonstrated in a number of comparative studies that show that CE is superior to push enteroscopy (50% vs 24%),17 barium follow-through (31% vs 5%),4 CT enteroclysis (59% vs 36%),18 intraoperative enteroscopy (74.4% vs 68%),19 double-balloon enteroscopy (DBE) (59% vs 42%),20 and angiography (72% vs 56%).21 With any new technology the question of whether patient outcomes are improved arises; however, few outcome data on the use of CE in patients with OGIB have so far been published. In the report by Pennazio et al,7 CE significantly affected outcome in patients with unexplained continuing overt bleeding; however, results were less favorable for patients with previous overt and occult bleeding of obscure origin. In a recent study by Neu et al,22 56 patients were followed up for a mean period of 13 months. Outcome change (no further bleeding) was noted in 55% of patients with tumors, erosions, or ulcers; in 56% of those with angiodysplasia; and in 78% of those with a negative test. Finally in a trial by Lai et al,23 49 patients who underwent CE for overt GI bleeding were followed up for a mean period of 19 months. According to this study, 58.3% of patients with angiodysplasia and 25% of those with ulcers rebled, whereas the rebleeding
854 GASTROINTESTINAL ENDOSCOPY Volume 69, No. 4 : 2009
www.giejournal.org
All 104 patients were stratified according to the degree of hemoglobin drop, as follows: less than 2 g/dL (28 patients not subjected to CE), 2 to 2.9 g/dL (19 patients), 3 to 3.9 g/dL (20 patients), 4 g/dL or more (37 patients). Positive findings were identified in 6, 7, and 24 patients in each category, respectively (P Z .022). It is of interest that all 11 patients whose OGIB presentation changed from occult to overt had a positive finding identified in the second-look CE. On the contrary, only 17 of 48 and 9 of 17 patients whose OGIB remained occult or overt, respectively, had a definite diagnosis (P ! .001). We performed a logistic regression analysis to estimate the parameters that might be associated with a diagnostic second-look CE (Table 5). According to our results, the change of the bleeding presentation from occult to overt, and the drop in hemoglobin of 4 g/dL or more were both significant predictive factors of a diagnostic test.
DISCUSSION
Viazis et al
Second-look capsule endoscopy after nondiagnostic first test
rate was 53.8% and 5.6% for those patients with active bleeding and no identifiable cause in CE and those with a negative test, respectively. Our study is among the largest published series to date, reporting on the outcome of a substantial number of patients after a 2-year followup period. According to our results, CE appears to affect long-term outcome in a positive way (ie, complete resolution of further bleeding) only in those patients with definitive lesions, lesions that can clearly be responsible for the bleeding episode and therefore characterize a diagnostic test. As expected, definitive lesions are more likely to lead to therapeutic interventions and are finally associated with resolution of bleeding. Despite the undisputable progress in visualizing lesions of the small bowel brought in by CE, a substantial number of patients with OGIB still remain without a diagnosis. The best management of this subset of patients is not known, and their management poses a real challenge in clinical practice. Because CE has proven its superiority versus all other tests available for the investigation of the small bowel, one reasonably wonders whether a second CE test would benefit these patients. According to our knowledge, our study is the first to report on this matter, and our results clearly show that there is a subgroup of this population (namely, those whose presentation changed from occult to overt and those with a drop in hemoglobin of 4 g/dL or more) that will mostly benefit from a secondlook CE. What is the best management for the remaining patients is not known as yet. Could DBE be of any value in this setting? The consensus report from the First International Workshop on Double Balloon Endoscopy recently presented an algorithm for evaluating patients with OGIB.24 It suggests proceeding directly to oral DBE in patients with continuing overt obscure bleeding. For other patients, it suggests initial CE, with positive results prompting a DBE (direction determined by location of the lesion) and negative results prompting observation only. We believe that further studies comparing DBE versus CE in the subgroup of patients with a nondiagnostic first capsule study would be of great value. Although it is logical to wonder why bleeding lesions seen on the second-look CE were missed by the first test, no certain explanation can be given. Could it be that as CE propels itself through the small intestine, certain areas of the jejunal or ileal mucosa are not seen? This is a logical explanation, albeit difficult to prove. Could it be that lesions in the terminal ileum are missed in cases that CE does not reach the cecum during the life span of its battery? This was indeed the case in certain instances; however, incomplete visualization of the small intestine was not a predictive factor of a diagnostic secondlook CE test. Could it be that certain lesions, such as angiodysplasias, are not bleeding at the time of the CE and therefore are not seen? This is also a plausible explanation because angiodysplasias of the small intestine were the most common finding during the second-look CE, and
it can be hypothesized that they were not bleeding ‘‘enough’’ during the first test to be visualized. The fact that only patients with a significant drop in hemoglobin values or those whose bleeding presentation changed from occult to overt had a diagnostic second-look test theoretically could give some support to the above hypothesis. In conclusion, our study verifies the usefulness of CE in the evaluation of patients with OGIB. Patients with a nondiagnostic first test should definitely be subjected to a second-look CE if the presentation changes from occult to overt or if they manifest a drop in hemoglobin of 4 g/dL or more. For the remaining patients, although a second test might be diagnostic too, best management remains elusive as yet.
www.giejournal.org
Volume 69, No. 4 : 2009 GASTROINTESTINAL ENDOSCOPY 855
REFERENCES 1. Pennazio M. Small bowel endoscopy. Endoscopy 2004;36:32-41. 2. Rey J, Ladas S, Alhassani A, et al. Video capsule endoscopy: update to guidelines (May 2006). Endoscopy 2006;38:1047-53. 3. Concha R, Amaro R, Barkin J. Obscure gastrointestinal bleeding: diagnostic and therapeutic approach. J Clin Gastroenterol 2007;41:242-51. 4. Costamagna G, Shah S, Riccioni M, et al. A prospective trial comparing small bowel radiographs and video capsule endoscopy for suspected small bowel disease. Gastroenterology 2002;123:999-1005. 5. Ell C, Remke S, May A, et al. The first prospective controlled trial comparing wireless capsule endoscopy with push enteroscopy in chronic gastrointestinal bleeding. Endoscopy 2002;34:685-9. 6. Mylonaki M, Fritscher-Ravens A, Swain P. Wireless capsule endoscopy: a comparison with push enteroscopy in patients with gastroscopy and colonoscopy negative gastrointestinal bleeding. Gut 2003;52:1122-6. 7. Pennazio M, Santucci R, Rondonotti E, et al. Outcome of patients with obscure gastrointestinal bleeding after capsule endoscopy: report of 100 consecutive cases. Gastroenterology 2004;126:643-53. 8. Triester S, Leighton J, Leontiadis G, et al. A meta-analysis of the yield of capsule endoscopy compared to other diagnostic modalities in patients with obscure gastrointestinal bleeding. Am J Gastroenterol 2005;100:2407-18. 9. Estevez E, Gonzalez-Conde B, Vazquez-Iglesias J, et al. Diagnostic yield and clinical outcomes after capsule endoscopy in 100 consecutive patients with obscure gastrointestinal bleeding. Eur J Gastroenterol Hepatol 2006;18:881-8. 10. Leighton J, Sharma V, Hentz J, et al. Capsule endoscopy versus push enteroscopy for evaluation of obscure gastrointestinal bleeding with 1-year outcomes. Dig Dis Sci 2006;51:891-9. 11. Redondo-Cerezo E, Perez-Vigara G, Perez-Sola A, et al. Diagnostic yield and impact of capsule endoscopy on management of patients with gastrointestinal bleeding of obscure origin. Dig Dis Sci 2007;52: 1376-81. 12. Carey E, Leighton J, Heigh R, et al. A single-center experience of 260 consecutive patients undergoing capsule endoscopy for obscure gastrointestinal bleeding. Am J Gastroenterol 2007;102:89-95. 13. Viazis N, Papaxoinis K, Theodoropoulos I, et al. Impact of capsule endoscopy in obscure small-bowel bleeding: defining strict diagnostic criteria for a favorable outcome. Gastrointest Endosc 2005;62:717-22. 14. Swain P. Wireless capsule endoscopy. Gut 2003;52(4 Suppl):48-50. 15. Viazis N, Sgouros S, Papaxoinis K, et al. Bowel preparation increases the diagnostic yield of wireless capsule endoscopy: a prospective, randomized, controlled study. Gastrointest Endosc 2004;60:534-8. 16. Zuckerman G, Prakash C, Askin M, et al. AGA technical review on the evaluation and management of occult and obscure gastrointestinal bleeding. Gastroenterology 2000;118:201-21.
Second-look capsule endoscopy after nondiagnostic first test
Viazis et al
17. de Leusse A, Vahedi K, Edery J, et al. Capsule endoscopy or push enteroscopy for first-line exploration of obscure gastrointestinal bleeding? Gastroenterology 2007;132:855-62. 18. Voderholzer W, Beinhoelzl J, Rogalla P, et al. Diagnostic yield of wireless capsule enteroscopy in comparison with computed tomography enteroclysis. Endoscopy 2003;35:1009-14. 19. Hartmann D, Schmidt H, Schilling D, et al. Follow-up of patients with obscure gastrointestinal bleeding after capsule endoscopy and intraoperative enteroscopy. Hepatogastroenterology 2007;54: 780-3. 20. Nakamura M, Niwa Y, Ohmiya M, et al. Preliminary comparison of capsule endoscopy and double-balloon enteroscopy in patients with suspected small-bowel bleeding. Endoscopy 2006;38:59-66. 21. Saperas E, Dot J, Videla J, et al. Capsule endoscopy versus computed tomographic or standard angiography for the diagnosis of obscure gastrointestinal bleeding. Am J Gastroenterol 2007;102:731-7. 22. Neu B, Ell C, May A, et al. Capsule endoscopy versus standard tests in influencing management of obscure digestive bleeding: results
from a German multicenter trial. Am J Gastroenterol 2005;100: 1736-42. 23. Lai L, Wong G, Chow D, et al. Long term follow up of patients with obscure gastrointestinal bleeding after negative capsule endoscopy. Am J Gastroenterol 2006;101:1224-8. 24. Suzuki T, Matsushima M, Okita I, et al. Clinical utility of double balloon enteroscopy for small intestinal bleeding. Dig Dis Sci 2007;52:1914-8.
856 GASTROINTESTINAL ENDOSCOPY Volume 69, No. 4 : 2009
www.giejournal.org
Received September 3, 2007. Accepted May 18, 2008. Current affiliations: Second Department of Gastroenterology, Evangelismos Hospital, Athens, Greece. Reprint requests: Nikos Viazis, MD, 59 Niriidon St, 17561 P Faliro, Athens, Greece. If you want to chat with an author of this article, you may contact him at [email protected].
Is there a role for second-look capsule endoscopy in patients with obscure GI bleeding after a nondiagnostic first test? Nikos Viazis, MD, Kostis Papaxoinis, MD, John Vlachogiannakos, MD, Alkiviadis Efthymiou, MD, Ioannis Theodoropoulos, MD, Dimitrios G. Karamanolis, MD Athens, Greece
Background: Long-term follow-up data on patients with obscure GI bleeding subjected to capsule endoscopy (CE) are missing. Objective: Our purpose was to follow up patients with a nondiagnostic test and determine whether a secondlook CE would be beneficial. Patients: We enrolled 293 subjects. CE studies were classified as diagnostic (positive findings) or nondiagnostic (findings of uncertain significance/no findings). Patients were followed up for a mean (SD) 24.8 (5.2) months. Outcome was defined as continued or complete resolution of bleeding. Interventions: Patients with a nondiagnostic test were subjected to a repeat CE if they manifested a new bleeding episode or a drop in hemoglobin R2 g/dL. Results: Positive findings, findings of uncertain significance, and no findings were identified in 41.6%, 16.0%, and 42.3% of our patients, respectively. Therapeutic intervention was possible in 72.1% of those with positive findings and in 30% of those with findings of uncertain significance. Complete resolution of bleeding occurred more often in patients with a diagnostic test (65.2%) compared with those with a nondiagnostic test (35.4%, P! .001). Second-look CE was performed in a subgroup of our patients (n Z 76) and was diagnostic in those whose presentation changed from occult to overt or those whose hemoglobin dropped R4 g/dL. Conclusions: In patients with obscure GI bleeding, a diagnostic CE leads to therapeutic interventions and a favorable outcome. Patients with a nondiagnostic test would definitely benefit from a second-look CE if the bleeding presentation changes from occult to overt or if the hemoglobin value drops R4 g/dL. (Gastrointest Endosc 2009;69:850-6.)
Copyright ª 2009 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 doi:10.1016/j.gie.2008.05.053
ranges from 45% to 75%, clearly reflecting differences in image interpretation and the diagnostic criteria used for the identification of the bleeding condition.4-12 We have previously reported that positive findings, defined by strict diagnostic criteria, were identified in 41.7% of patients subjected to CE for the investigation of OGIB.13 After a median follow-up period of 14 months, those patients were more likely to receive some kind of therapeutic intervention for their lesions and eventually had a favorable outcome. On the contrary, the outcome of those patients with findings of uncertain significance or no findings on CE was not changed in a positive way. In clinical practice, the most appropriate management of patients with OGIB and a nondiagnostic CE test remains elusive, and we do not know whether these subjects would benefit from a second-look CE. Meanwhile, data on follow-up longer than 1 year are sparse in the literature, further obscuring our knowledge on the best therapeutic algorithm that needs to be followed in these cases.
850 GASTROINTESTINAL ENDOSCOPY Volume 69, No. 4 : 2009
www.giejournal.org
Obscure GI bleeding (OGIB) remains a major clinical challenge. The majority of OGIB cases originate from the small intestine; therefore, wireless capsule endoscopy (CE) is a valuable tool in the investigation of these patients.1-3 Although sound literature data have proven the superiority of CE compared with all other diagnostic modalities in identifying bleeding lesions,4-12 the true diagnostic yield of this test in patients with OGIB is still not known. Numerous previous reports have reported conflicting results, and according to the available data the diagnostic yield of CE Abbreviations: CE, capsule endoscopy; DBE, double-balloon enteroscopy; NSAIDs, nonsteroidal anti-inflammatory drugs; OGIB, obscure GI bleeding. DISCLOSURE: All authors disclosed no financial relationships relevant to this publication.
Viazis et al
Second-look capsule endoscopy after nondiagnostic first test
TABLE 1. Characteristics of the study population
Capsule Summary
Age (y) (mean [SD])
What is already known on this topic
Male/female Hemoglobin (g/dL) (mean [SD], minimum-maximum) Overt GI bleeding of obscure origin
59.5 (14.4) 169:124 9.9 (1.68), 7.2-12.4 146
Hematochezia
61
Melena
85
Occult GI bleeding of obscure origin Iron deficiency anemia Positive fecal occult blood test
d
147 108
Despite capsule endoscopy (CE), a substantial number of patients with occult GI bleeding (OGIB) do not receive a diagnosis, and a second CE may be used.
What this study adds to our knowledge d
d
In 293 subjects at a single institution, 65% of individuals with a nondiagnostic initial CE continued to manifest OGIB after a mean follow-up period of 24 months. Development of overt bleeding and a hemoglobin drop of 4 g/dL or more were significant predictive factors of a diagnostic second CE.
39
From September 2002 to January 2007, all patients referred to our unit for CE because of overt or occult OGIB were considered for entry into the study. We excluded patients with known or suspected GI obstruction, strictures or fistulae; patients with a cardiac pacemaker, defibrillator, or other implanted electromedical device; pregnant women; patients under 18 years of age; and patients who refused consent. Patients with suspected GI obstruction were excluded because the capsule can become stuck at points of luminal narrowing, with resultant acute bowel obstruction. Patients with a pacemaker were excluded because of fear of interference of the pacemaker by the capsule, although, so far, such an incidence has not been reported. We defined bleeding of obscure origin as bleeding of unknown origin that persists or recurs (ie, recurrent or persistent iron deficiency anemia, fecal occult blood test positivity, or visible bleeding) after a negative initial workout including gastroscopy, colonoscopy, small bowel barium follow-through or enteroclysis, and push enteroscopy. Bleeding of obscure origin was further subdivided in 2 clinical entities: (1) obscure: occult, as manifested by recurrent iron-deficiency anemia or recurrent positive fecal occult blood test results and (2) obscure: overt, with recurrent passage of visible blood (hematochezia/ melena). CE was performed with the Given M2A video capsule system (Given Imaging, Yogneam, Israel), as reported previously.14 All patients ingested the capsule at 9 AM after an
overnight fast. They were also given a bowel preparation with 2 L of polyethylene glycol electrolyte lavage solution 16 hours before the investigation because this has been shown to increase the diagnostic yield of CE.15 In addition, they were advised to consume clear liquids from the time they started receiving the bowel purge until the time they went to bed that same night. After ingestion, nonhospitalized patients were free to go home and continue their usual activities, but they were asked to contact us if they noted abdominal pain, nausea, or vomiting. All patients were advised not to drink and eat for a 3- and a 4-hour period, respectively. At the end of the recording period, the equipment was removed. Two investigators (N. V., K. P.) independently evaluated the video capsule endoscopy images at a movie speed of 10 frames per second. The investigators were blinded as to whether the patients were referred for the evaluation of overt or occult bleeding of obscure origin. Findings were considered positive if the investigators could explain the patient’s symptoms or help further management or if the findings were later confirmed by other diagnostic modalities. Findings were considered of uncertain significance if they failed to completely explain the patient’s symptoms, thus necessitating further investigation. Finally, when no abnormality was detected, despite a definite indication of an existing lesion, the test was considered as having no findings. All patients were monitored closely at 3- and 6- month intervals and every 6 months thereafter for a survey of their final outcome. Outcome was described as either continued bleeding or complete resolution of bleeding (regardless of whether treatment was given). For all patients the following follow-up information was obtained: use of medical, endoscopic, or surgical treatment; further bleeding episodes; persistence of iron deficiency anemia; or positivity of fecal occult blood testing. This information was sought through telephone interviews, contact of the referring physicians, or follow-up visits of the patients themselves.
www.giejournal.org
Volume 69, No. 4 : 2009 GASTROINTESTINAL ENDOSCOPY 851
We therefore prospectively evaluated all patients with OGIB who underwent CE at our institution to determine their outcomes and to investigate whether those with a nondiagnostic test would benefit from a second-look CE.
PATIENTS AND METHODS
Second-look capsule endoscopy after nondiagnostic first test
TABLE 2. CE findings
TABLE 3. Outcome data for patients who received treatment because of the results of CE Overt GI Occult GI bleeding of bleeding of obscure origin obscure origin (No. of patients) (No. of patients)
Positive findings
58
64
Multiple or bleeding angiodysplasias
50
36
Multiple apthoid ulcerations
0
10
Mucosal ulcers
4
8
Bleeding polyps
4
4
Small bowel tumors
0
6
29
18
Isolated, nonbleeding angiodysplasias
15
8
Isolated apthoid ulceration
4
0
Mucosal erosion
4
4
Small nonbleeding polyp
0
6
Blood within the small bowel lumen
6
0
59
65
Findings of uncertain significance
No findings
Viazis et al
Treatment (No. of patients)
Resolution of bleeding (No. of patients)
Multiple or bleeding angiodysplasias
52
36
Multiple apthoid ulcerations
10
10
Mucosal ulcers
12
10
Bleeding polyps
8
8
Small bowel tumors
6
6
10
2
4
2
Capsule findings Positive findings
Findings of uncertain significance Isolated nonbleeding angiodysplasias Mucosal erosions
dependent predictors of a definite CE diagnosis. The various factors examined for correlation with the diagnostic yield at a second CE test were defined before the study. A P! .05 was considered statistically significant. Statistical analysis was performed with SPSS 11.0 for Windows (SPSS, Chicago, Ill).
RESULTS
Values are expressed as mean SD. Categorical variables were compared with the c2 test with Yates’ correction or the Fisher exact test, as appropriate. Logistic regression was used to identify the variables that were in-
During the study period, 305 patients were referred to our unit for CE because of OGIB. Six with subclinical small intestinal stenosis detected during small bowel followthrough were excluded; 2 patients were younger than 18 years old and were excluded; 1 had a pacemaker and was excluded; and 3 refused consent. Therefore, the study sample consisted of 293 patients. The characteristics of the study population are presented in Table 1. All patients ingested the capsule without difficulty, and the test was well tolerated without adverse events. The capsule reached the cecum (ie, the complete small bowel was visualized) in 220 patients (75.1%). Positive findings were identified in 122 of our patients, leading to an overall diagnostic yield of 41.6%. More precisely, positive findings were identified in 58 patients with overt GI bleeding (diagnostic yield 39.7%) compared with 64 patients with OGIB of obscure origin (diagnostic yield 43.5%) (not significant). As previously noted, the most common positive finding was angiodysplasia of the small intestine. Findings of uncertain significance were identified in 47 patients (16%, overt 29/occult 18), whereas in 124 patients the
852 GASTROINTESTINAL ENDOSCOPY Volume 69, No. 4 : 2009
www.giejournal.org
During the follow-up period, a second-look CE was performed in those patients with a nondiagnostic first test (ie, findings of uncertain significance, no findings) when a new bleeding episode (hematochezia/melena) or a drop in hemoglobin of 2 g/dL or more were reported. The video capsule images were evaluated by the same two investigators who had evaluated the initial capsule study, and they were blinded as to whether the patients had a new bleeding episode or were subjected to the test for a drop in hemoglobin. The study protocol adhered to the principles of the Declaration of Helsinki and was approved by the ethics committee of our hospital. All patients gave written informed consent at the time of enrollment.
Statistical analysis
Viazis et al
Second-look capsule endoscopy after nondiagnostic first test
TABLE 4. Second-look CE findings Overt GI bleeding of obscure origin (No. of patients)
Occult GI bleeding of obscure origin (No. of patients)
20
17
15
12
Multiple apthoid ulcerations
0
0
Mucosal ulcers
2
3
Bleeding polyps
2
1
Small-bowel tumors
1
1
4
18
Isolated, nonbleeding angiodysplasias
1
7
Isolated apthoid ulceration
1
5
Mucosal erosion
1
4
Small nonbleeding polyp
0
2
Blood within the small-bowel lumen
1
0
4
13
Positive findings Multiple or bleeding angiodysplasias
Findings of uncertain significance
No findings
test did not reveal any findings (42.3%, overt 59/occult 65). The types of lesions seen are presented in Table 2. The degree of agreement between the 2 endoscopists regarding the identification of the bleeding site condition was 95% in studies with positive findings, 90% in studies with findings of uncertain significance, and 94% in studies with no findings. Of the 122 patients with positive findings at CE, 88 (72.1%) received some kind of treatment for their lesions or their management was changed accordingly (Table 3). Twenty-six patients with angiodysplasia of the duodenum and proximal jejunum underwent endoscopic treatment with argon plasma coagulation; 10 patients with bleeding angiodysplasia of the distal jejunum and ileum underwent surgery; 16 patients with multiple angiodysplasias received estrogen therapy (although this strategy is of unproven value); 10 patients with findings suggestive of Crohn’s disease were begun on corticosteroids; 12 patients with small bowel ulcers were advised to discontinue nonsteroidal anti-inflammatory drug (NSAID) ingestion; and finally, 14 patients with bleeding polyps or tumors underwent surgical resection. Of the 47 patients with findings of uncertain significance at CE, 14 (30%) either received some kind of treatment for their lesions or their management was changed accordingly. Ten patients with angiodysplasia of the proximal jejunum underwent endoscopic treatment with argon plasma coagulation and 4 patients with mucosal erosions were advised to discontinue NSAIDs (Table 3). Outcome data are available for 279 patients (positive findings 118/findings of uncertain significance 47/no find-
ings 114) because 14 patients were lost at follow-up. The mean (SD) follow-up time was 24.8 (5.2) months (range 7-29 months). From the 118 patients with a diagnostic test (positive findings), complete resolution of bleeding occurred in 77 (65.2%) compared with 57 of the 161 patients (35.4%) with a nondiagnostic test (findings of uncertain significance or no findings) (P ! .001). During the follow-up period, of the 104 patients with a nondiagnostic first test whose bleeding did not resolve, 76 were subjected to a second-look CE, whereas the remaining 28 patients were not because the degree of their anemia was considered to be stable (ie, drop in hemoglobin less than 2 g/dL). Of the 76 patients subjected to the secondlook capsule study, 28 had a new bleeding episode (ie, hematochezia/melena) regardless of whether the hemoglobin value dropped, and 48 had a drop in hemoglobin of 2 g/dL or more. From the 28 patients with an overt new bleeding episode, 17 had overt OGIB at initial presentation and 11 had occult OGIB at initial presentation. All 48 patients who were subjected to a second CE test because of a drop in hemoglobin of 2 g/dL or more had occult OGIB at initial presentation. The results of the second-look CE test are presented in Table 4. Once again, the most common positive finding was angiodysplasia of the small intestine, whereas mucosal ulcers, a bleeding polyp, and a small bowel tumor of the terminal ileum (in a patient whose first CE did not reach the cecum) were also seen. It is of note that among the 37 patients with a diagnostic second-look CE, 7 had an incomplete first test (ie, the capsule did not reach the cecum), whereas the relevant number among the 39 with a nondiagnostic second-look CE was 13 (P Z .19).
www.giejournal.org
Volume 69, No. 4 : 2009 GASTROINTESTINAL ENDOSCOPY 853
Second-look capsule endoscopy after nondiagnostic first test
Viazis et al
TABLE 5. Logistic regression analysis of the predictive values of the evaluated parameters regarding second-look CE diagnosis certainty 95% Confidence limits Estimated coefficient
SE
Wald statistic
Significance
Odds ratio
Lower
Upper
Age (y)
0.015
0.025
0.381
.537
0.985
0.938
1.034
Sex (male)
0.756
0.618
1.497
.221
2.130
0.634
7.152
OGIB initial presentation (overt)
0.970
0.603
2.585
.108
2.638
0.809
8.610
Incomplete first CE
1.117
0.836
1.785
.182
3.055
0.594
15.727
Agreement between examiners
0.851
0.637
1.781
.182
2.341
0.671
8.166
Treatment after first CE
0.856
0.644
1.767
.184
2.353
0.666
8.308
OGIB relapse presentation (overt)
0.775
0.654
1.407
.236
2.171
0.603
7.818
OGIB change (occult to overt)
2.526
0.985
6.572
.010
0.080
0.012
0.552
Hemoglobin drop (more than 4 mg/dL)
2.095
0.862
5.902
.015
0.123
0.023
0.667
A substantial number of patients subjected to CE for the investigation of OGIB will end up having a nondiagnostic test. Literature data regarding the long-term outcome of these patients and options for their best management are missing. According to our results, 65% of patients with a nondiagnostic first CE test continued to manifest OGIB after a mean follow-up period of 24.8 months, and a change in their bleeding presentation from occult to overt or a drop of their hemoglobin level of 4g/dL or more were both significant predictive factors of a diagnostic second CE. Management of patients with OGIB remains a challenging problem in clinical practice. In our study obscure bleeding was defined according to the American Gastroen-
terological Association technical review16; however, we have extended our initial workup to include not only colonoscopy and upper endoscopy but also small-bowel follow-through and push enteroscopy. Therefore, our study population consisted of patients with bleeding who had failed to reach a diagnosis although they had been subjected to all conventional radiographic and endoscopic techniques. In clinical practice, when a bleeding source in not identified in upper or lower GI endoscopy, the small bowel is interrogated and wireless CE has now become the method of choice for this purpose. Indeed, the diagnostic yield of CE has been demonstrated in a number of comparative studies that show that CE is superior to push enteroscopy (50% vs 24%),17 barium follow-through (31% vs 5%),4 CT enteroclysis (59% vs 36%),18 intraoperative enteroscopy (74.4% vs 68%),19 double-balloon enteroscopy (DBE) (59% vs 42%),20 and angiography (72% vs 56%).21 With any new technology the question of whether patient outcomes are improved arises; however, few outcome data on the use of CE in patients with OGIB have so far been published. In the report by Pennazio et al,7 CE significantly affected outcome in patients with unexplained continuing overt bleeding; however, results were less favorable for patients with previous overt and occult bleeding of obscure origin. In a recent study by Neu et al,22 56 patients were followed up for a mean period of 13 months. Outcome change (no further bleeding) was noted in 55% of patients with tumors, erosions, or ulcers; in 56% of those with angiodysplasia; and in 78% of those with a negative test. Finally in a trial by Lai et al,23 49 patients who underwent CE for overt GI bleeding were followed up for a mean period of 19 months. According to this study, 58.3% of patients with angiodysplasia and 25% of those with ulcers rebled, whereas the rebleeding
854 GASTROINTESTINAL ENDOSCOPY Volume 69, No. 4 : 2009
www.giejournal.org
All 104 patients were stratified according to the degree of hemoglobin drop, as follows: less than 2 g/dL (28 patients not subjected to CE), 2 to 2.9 g/dL (19 patients), 3 to 3.9 g/dL (20 patients), 4 g/dL or more (37 patients). Positive findings were identified in 6, 7, and 24 patients in each category, respectively (P Z .022). It is of interest that all 11 patients whose OGIB presentation changed from occult to overt had a positive finding identified in the second-look CE. On the contrary, only 17 of 48 and 9 of 17 patients whose OGIB remained occult or overt, respectively, had a definite diagnosis (P ! .001). We performed a logistic regression analysis to estimate the parameters that might be associated with a diagnostic second-look CE (Table 5). According to our results, the change of the bleeding presentation from occult to overt, and the drop in hemoglobin of 4 g/dL or more were both significant predictive factors of a diagnostic test.
DISCUSSION
Viazis et al
Second-look capsule endoscopy after nondiagnostic first test
rate was 53.8% and 5.6% for those patients with active bleeding and no identifiable cause in CE and those with a negative test, respectively. Our study is among the largest published series to date, reporting on the outcome of a substantial number of patients after a 2-year followup period. According to our results, CE appears to affect long-term outcome in a positive way (ie, complete resolution of further bleeding) only in those patients with definitive lesions, lesions that can clearly be responsible for the bleeding episode and therefore characterize a diagnostic test. As expected, definitive lesions are more likely to lead to therapeutic interventions and are finally associated with resolution of bleeding. Despite the undisputable progress in visualizing lesions of the small bowel brought in by CE, a substantial number of patients with OGIB still remain without a diagnosis. The best management of this subset of patients is not known, and their management poses a real challenge in clinical practice. Because CE has proven its superiority versus all other tests available for the investigation of the small bowel, one reasonably wonders whether a second CE test would benefit these patients. According to our knowledge, our study is the first to report on this matter, and our results clearly show that there is a subgroup of this population (namely, those whose presentation changed from occult to overt and those with a drop in hemoglobin of 4 g/dL or more) that will mostly benefit from a secondlook CE. What is the best management for the remaining patients is not known as yet. Could DBE be of any value in this setting? The consensus report from the First International Workshop on Double Balloon Endoscopy recently presented an algorithm for evaluating patients with OGIB.24 It suggests proceeding directly to oral DBE in patients with continuing overt obscure bleeding. For other patients, it suggests initial CE, with positive results prompting a DBE (direction determined by location of the lesion) and negative results prompting observation only. We believe that further studies comparing DBE versus CE in the subgroup of patients with a nondiagnostic first capsule study would be of great value. Although it is logical to wonder why bleeding lesions seen on the second-look CE were missed by the first test, no certain explanation can be given. Could it be that as CE propels itself through the small intestine, certain areas of the jejunal or ileal mucosa are not seen? This is a logical explanation, albeit difficult to prove. Could it be that lesions in the terminal ileum are missed in cases that CE does not reach the cecum during the life span of its battery? This was indeed the case in certain instances; however, incomplete visualization of the small intestine was not a predictive factor of a diagnostic secondlook CE test. Could it be that certain lesions, such as angiodysplasias, are not bleeding at the time of the CE and therefore are not seen? This is also a plausible explanation because angiodysplasias of the small intestine were the most common finding during the second-look CE, and
it can be hypothesized that they were not bleeding ‘‘enough’’ during the first test to be visualized. The fact that only patients with a significant drop in hemoglobin values or those whose bleeding presentation changed from occult to overt had a diagnostic second-look test theoretically could give some support to the above hypothesis. In conclusion, our study verifies the usefulness of CE in the evaluation of patients with OGIB. Patients with a nondiagnostic first test should definitely be subjected to a second-look CE if the presentation changes from occult to overt or if they manifest a drop in hemoglobin of 4 g/dL or more. For the remaining patients, although a second test might be diagnostic too, best management remains elusive as yet.
www.giejournal.org
Volume 69, No. 4 : 2009 GASTROINTESTINAL ENDOSCOPY 855
REFERENCES 1. Pennazio M. Small bowel endoscopy. Endoscopy 2004;36:32-41. 2. Rey J, Ladas S, Alhassani A, et al. Video capsule endoscopy: update to guidelines (May 2006). Endoscopy 2006;38:1047-53. 3. Concha R, Amaro R, Barkin J. Obscure gastrointestinal bleeding: diagnostic and therapeutic approach. J Clin Gastroenterol 2007;41:242-51. 4. Costamagna G, Shah S, Riccioni M, et al. A prospective trial comparing small bowel radiographs and video capsule endoscopy for suspected small bowel disease. Gastroenterology 2002;123:999-1005. 5. Ell C, Remke S, May A, et al. The first prospective controlled trial comparing wireless capsule endoscopy with push enteroscopy in chronic gastrointestinal bleeding. Endoscopy 2002;34:685-9. 6. Mylonaki M, Fritscher-Ravens A, Swain P. Wireless capsule endoscopy: a comparison with push enteroscopy in patients with gastroscopy and colonoscopy negative gastrointestinal bleeding. Gut 2003;52:1122-6. 7. Pennazio M, Santucci R, Rondonotti E, et al. Outcome of patients with obscure gastrointestinal bleeding after capsule endoscopy: report of 100 consecutive cases. Gastroenterology 2004;126:643-53. 8. Triester S, Leighton J, Leontiadis G, et al. A meta-analysis of the yield of capsule endoscopy compared to other diagnostic modalities in patients with obscure gastrointestinal bleeding. Am J Gastroenterol 2005;100:2407-18. 9. Estevez E, Gonzalez-Conde B, Vazquez-Iglesias J, et al. Diagnostic yield and clinical outcomes after capsule endoscopy in 100 consecutive patients with obscure gastrointestinal bleeding. Eur J Gastroenterol Hepatol 2006;18:881-8. 10. Leighton J, Sharma V, Hentz J, et al. Capsule endoscopy versus push enteroscopy for evaluation of obscure gastrointestinal bleeding with 1-year outcomes. Dig Dis Sci 2006;51:891-9. 11. Redondo-Cerezo E, Perez-Vigara G, Perez-Sola A, et al. Diagnostic yield and impact of capsule endoscopy on management of patients with gastrointestinal bleeding of obscure origin. Dig Dis Sci 2007;52: 1376-81. 12. Carey E, Leighton J, Heigh R, et al. A single-center experience of 260 consecutive patients undergoing capsule endoscopy for obscure gastrointestinal bleeding. Am J Gastroenterol 2007;102:89-95. 13. Viazis N, Papaxoinis K, Theodoropoulos I, et al. Impact of capsule endoscopy in obscure small-bowel bleeding: defining strict diagnostic criteria for a favorable outcome. Gastrointest Endosc 2005;62:717-22. 14. Swain P. Wireless capsule endoscopy. Gut 2003;52(4 Suppl):48-50. 15. Viazis N, Sgouros S, Papaxoinis K, et al. Bowel preparation increases the diagnostic yield of wireless capsule endoscopy: a prospective, randomized, controlled study. Gastrointest Endosc 2004;60:534-8. 16. Zuckerman G, Prakash C, Askin M, et al. AGA technical review on the evaluation and management of occult and obscure gastrointestinal bleeding. Gastroenterology 2000;118:201-21.
Second-look capsule endoscopy after nondiagnostic first test
Viazis et al
17. de Leusse A, Vahedi K, Edery J, et al. Capsule endoscopy or push enteroscopy for first-line exploration of obscure gastrointestinal bleeding? Gastroenterology 2007;132:855-62. 18. Voderholzer W, Beinhoelzl J, Rogalla P, et al. Diagnostic yield of wireless capsule enteroscopy in comparison with computed tomography enteroclysis. Endoscopy 2003;35:1009-14. 19. Hartmann D, Schmidt H, Schilling D, et al. Follow-up of patients with obscure gastrointestinal bleeding after capsule endoscopy and intraoperative enteroscopy. Hepatogastroenterology 2007;54: 780-3. 20. Nakamura M, Niwa Y, Ohmiya M, et al. Preliminary comparison of capsule endoscopy and double-balloon enteroscopy in patients with suspected small-bowel bleeding. Endoscopy 2006;38:59-66. 21. Saperas E, Dot J, Videla J, et al. Capsule endoscopy versus computed tomographic or standard angiography for the diagnosis of obscure gastrointestinal bleeding. Am J Gastroenterol 2007;102:731-7. 22. Neu B, Ell C, May A, et al. Capsule endoscopy versus standard tests in influencing management of obscure digestive bleeding: results
from a German multicenter trial. Am J Gastroenterol 2005;100: 1736-42. 23. Lai L, Wong G, Chow D, et al. Long term follow up of patients with obscure gastrointestinal bleeding after negative capsule endoscopy. Am J Gastroenterol 2006;101:1224-8. 24. Suzuki T, Matsushima M, Okita I, et al. Clinical utility of double balloon enteroscopy for small intestinal bleeding. Dig Dis Sci 2007;52:1914-8.
856 GASTROINTESTINAL ENDOSCOPY Volume 69, No. 4 : 2009
www.giejournal.org
Received September 3, 2007. Accepted May 18, 2008. Current affiliations: Second Department of Gastroenterology, Evangelismos Hospital, Athens, Greece. Reprint requests: Nikos Viazis, MD, 59 Niriidon St, 17561 P Faliro, Athens, Greece. If you want to chat with an author of this article, you may contact him at [email protected].