REVIEWS endoscopy for upper gastrointestinal bleeding: is routine second-look necessary? Kelvin K. F. Tsoi, Philip W. Y. Chiu and Joseph J. Y. Sung abstract | The benefit of routine repeat endoscopy after endoscopic hemostasis in the management of peptic ulcer bleeding is controversial. The aim of this Review is to evaluate the efficacy of second-look endoscopy by examining the evidence from published, randomized, clinical trials. Outcome measurements included recurrent bleeding, surgery, mortality, blood transfusion, and length of hospital stay. Studies were categorized into those in which endoscopy was performed with endoscopic injection or thermal coagulation. On the basis of existing evidence, second-look endoscopy with heater probe reduces the risk of recurrent bleeding, but has no effect on overall mortality or the need for surgery. Therefore, routine second-look endoscopy cannot be recommended. Selected high-risk patients may benefit from second-look endoscopy, but the use of high-dose intravenous PPIs may obviate the need for this procedure. Tsoi, K. K. F. et al. Nat. Rev. Gastroenterol. Hepatol. 6, 717–722 (2009); doi:10.1038/nrgastro.2009.186
Continuing Medical Education online This activity has been planned and implemented in accordance with the Essential Areas and policies of the Accreditation Council for Continuing Medical Education through the joint sponsorship of MedscapeCME and Nature Publishing Group. MedscapeCME is accredited by the Accreditation Council for Continuing Medical Education (ACCME) to provide continuing medical education for physicians. MedscapeCME designates this educational activity for a maximum of 0.5 aMa pra Category 1 CreditstM. Physicians should only claim credit commensurate with the extent of their participation in the activity. All other clinicians completing this activity will be issued a certificate of participation. To participate in this journal CME activity: (1) review the learning objectives and author disclosures; (2) study the education content; (3) take the post-test and/or complete the evaluation at http://www.medscapecme.com/journal/ nrgastro; and (4) view/print certificate.
learning objectives Upon completion of this activity, participants should be able to: 1 Describe the use of second-look endoscopy among patients with upper gastrointestinal bleeding. 2 Identify outcomes improved with second-look endoscopy. 3 Compare second-look endoscopy with high-dose protonpump inhibitor therapy among patients with upper gastrointestinal bleeding. 4 Discriminate when to use second-look endoscopy.
Introduction
Peptic ulcer bleeding is a serious clinical problem leading to significant morbidity and mortality. large, populationbased studies from europe estimate that the annual incidence of acute ulcer bleeding is 100 to 190 per 100,000 Competing interests J. J. Y. Sung declares associations with the following companies: AstraZeneca, Bristol-Myers Squibb, GlaxoSmithKline, Nycomed, Roche. See the article online for full details of the relationships. The other authors, the Journal Editor N. Wood and the CME questions author C. P. Vega declare no competing interests.
individuals,1 with a mortality rate of 1.7–10.8%,2 either directly caused by the bleeding episode or indirectly related to concurrent medical illnesses. endoscopic therapy is generally recommended as the first-line treatment for upper gastrointestinal bleeding as it is shown to reduce recurrent bleeding, the requirement for surgical intervention, and mortality.3 However, 9–18% of patients still develop recurrent bleeding after single endoscopic hemostasis.2 a repeat endoscopic evaluation and retreatment of a bleeding lesion on a subsequent follow-up in the absence of clinical evidence of rebleeding is commonly called second-look endoscopy. the aim is to detect those ulcers with persistent endoscopic stigmata of bleeding and prevent severe rebleeding before it actually occurs. some descriptive studies in the 1980s mentioned that repeating injection therapy was useful to arrest hemorrhage from actively bleeding ulcers.4–6 However, the evidence was empirical, as these studies did not include groups of control patients. since 1994, a few randomized, controlled trials have formally evaluated the clinical benefit of second-look endoscopy, but the results are conflicting. Factors such as small sample size, selection of different groups of patients, and variations in endoscopic therapy used may contribute to the discrepant results. a scheduled second endoscopy and reapplication of endoscopic therapy has been proposed by saeed to reduce recurrent bleeding and improve clinical outcome.7 However, an international multidisciplinary consensus group from 11 national societies does not recommend use of second-look endoscopy as routine practice for the management of peptic ulcer bleeding.8 the american society of Gastrointestinal endoscopy has not given any specific recommendations on this practice,9 and the British society of Gastroenterology guidelines state that repeat endoscopy should only be considered when there is
nature reviews | gastroenterology & hepatology © 2009 Macmillan Publishers Limited. All rights reserved
Department of Medicine and Therapeutics, (K. K. F. tsoi, J. J. y. sung), Department of Surgery (p. W. y. Chiu), Chinese University of Hong Kong, Hong Kong. Correspondence: J. J. Y. Sung, Department of Medicine and Therapeutics, Prince of Wales Hospital, Shatin, NT, Hong Kong
[email protected]
volume 6 | DeCemBer 2009 | 717
REVIEWS Key points ■ Second-look endoscopy with repeated thermal coagulation for bleeding peptic ulcers reduces the risk of recurrent bleeding, but does not have an impact on surgical requirement or overall mortality ■ Second-look endoscopy cannot be recommended as routine practice ■ Selected groups of high-risk patients, with advanced age and/or presence of a concurrent acute life-threatening illness, may benefit from second-look endoscopy ■ High-dose intravenous PPIs may obviate the need for second-look endoscopy
clinical evidence of active rebleeding or initial endoscopic therapy is unsatisfactory.10 we searched the literature and abstracts published at major international conferences (including Digestive Disease week, united european Gastroenterology week and asian Pacific Digestive week) over the past 5 years for studies on second-look endoscopy. of the 552 studies identified, 11 were deemed appropriate for further scrutiny. we excluded two narrative reviews that focused on the discussion of benefits and drawbacks of second-look endoscopy,11,12 one retrospective survey evaluating the need for second-look endoscopy,13 one prospective trial comparing two forms of second-look endoscopy using different injection therapies,14 and two abstracts without full-text publications.15,16 this leaves five randomized, controlled trials published between 1994 and 2003 to be included in this review.17–21 substantial heterogeneity was found in endoscopic treatment, adjuvant therapy, and definitions of recurrent bleeding among these trials (table 1). in general, second-look endoscopy repeats the same hemostatic modality as the initial endoscopy, namely endoscopic injection or thermal coagulation.
Injection therapy
endoscopic injection is considered an effective and safe treatment for upper gastrointestinal bleeding. injection of epinephrine has the effects of vasoconstriction, vessel tamponade and platelet aggregation.19,22 of these, immediate compression of the bleeding vessel is thought to be most important for the initial control of bleeding.23,24 in a prospective, randomized, controlled trial, villanueva et al. recruited a total of 104 adult patients with active arterial bleeding or a nonbleeding visible vessel.17 a total of 52 patients were randomly assigned to receive second-look endoscopy with retreatment between 18 and 24 h after the emergency procedure, while 52 patients did not receive second-look endoscopy. Both groups were comparable in age and sex, but fewer patients in the second-look endoscopy group were taking nsaiDs (44%) than the control group (59%). all patients were treated by adrenaline injection and were given ranitidine after therapeutic endoscopy. a second injection was performed only in those with visible vessels identified on second-look endoscopy. there was no difference in the incidence of recurrent bleeding, need for surgery, mortality, blood transfusion, or length of hospital stay between the two groups (table 2). the rebleeding rate
among those who received a second adrenaline injection was still high (17%). the authors concluded that secondlook endoscopy showed no distinct benefit and hence should not be recommended routinely for the treatment of upper gastrointestinal bleeding. a clinical study has demonstrated that injection therapy with fibrin glue is superior to the injection of diluted epinephrine alone.25 rutgeerts et al. conducted a prospective, multicenter, randomized, controlled trial in seven countries to investigate second-look endoscopy using fibrin glue injection.19 Patients with peptic ulcers that were spurting, oozing, or showing nonbleeding visible vessels were randomly assigned to repeated fibrin glue injections, single injection of fibrin glue or single injection of polidocanol. all patients were observed for at least 5 days after initial treatment and underwent daily repeat endoscopies. mean age, gender, ulcer history, location of ulcer, and intake of nsaiDs were comparable between treatment groups. all patients received concomitant therapy with ranitidine. Patients who received repeated fibrin glue injection had a marginally significant benefit on the reduction of recurrent bleeding (15% in repeated injection versus 21% in single injection, odds ratio [or] 0.68, 95% Ci 0.46–1.00). there were no differences in mortality, need for surgery or blood transfusion (table 2). unlike all other studies, second-look endoscopy was offered to every patient in this trial, but no repeat treatment was offered to patients in the control groups. the authors concluded that repeated injections of fibrin are marginally more effective than single injections, especially to prevent rebleeding from gastroduodenal ulcers. However, this is a single study showing the efficacy of fibrin glue, and as it is not clear whether the effects came from routine second-look endoscopy or fibrin glue, the latter has not been widely adopted in clinical practice. messmann et al. studied the clinical outcomes of 107 patients with peptic ulcers that were either actively bleeding or with signs of recent hemorrhage.20 Patients with a hospital admission for upper gastrointestinal bleeding and who had undergone emergency endoscopy within 4 h of admission were recruited to the study. a total of 52 patients were randomly assigned to receive second-look endoscopy with injection of epinephrine and fibrin or thrombin between 16 and 24 h, whereas 53 patients did not receive second-look endoscopy. all patients received omeprazole as adjuvant therapy. Patient demographics were well matched between the two groups with respect to age, sex, intake of nsaiDs, initial hemoglobin levels, severity of bleeding reflected by hemodynamic instability, location and size of ulcer, and Forrest classification. the overall incidence of adverse outcomes, including rebleeding, need for surgery, and mortality, was similar (table 2). transfusion requirement and median length of hospital stay were comparable. the authors concluded that scheduled second-look endoscopy cannot be recommended after an initial successful endoscopic treatment of peptic ulcer bleeding.
718 | DECEMBER 2009 | voluME 6
www.nature.com/nrgastro © 2009 Macmillan Publishers Limited. All rights reserved
REVIEWS Table 1 | Five studies evaluating the benefit of second-look endoscopy* study
study period
patients
time of second-look endoscopy
endoscopic therapy
adjuvant therapy (dosage)
Definition of rebleeding
Follow-up period
Villanueva et al.17
1990– 1991
With active bleeding or visible vessel
18–24 h after first endoscopic treatment
Adrenaline injection alone
Ranitidine (IV 50 mg every 6 h)
Endoscopic or clinical rebleeding
Until discharge or death
Saeed et al.18
1991– 1993
With active bleeding, visible vessel or adherent clot
24 h after initial endoscopic hemostasis
Heater probe alone or preceded by epinephrine injection
Ranitidine (nonspecified)
Clinical rebleeding
Until discharge or death
Rutgeerts et al.19
NA
With active bleeding or visible vessel
Daily until ulcer base was clean
Epinephrine injection followed by fibrin glue or polidocanol
Ranitidine (nonspecified)
Endoscopic or clinical rebleeding
30 days
Messmann et al.20
1994– 1996
With active bleeding or signs of recent bleeding
18–24 h after first endoscopic treatment
Epinephrine injection preceded by fibrin glue
Omeprazole (IV 40 mg every 12 h)
Endoscopic or clinical rebleeding
4 weeks
Chiu et al.21
1999– 2001
With active bleeding, visible vessel of adherent clot
16–24 h after initial hemostasis
Adrenaline injection followed by heater probe
Omeprazole (IV 40 mg every 12 h)
Clinical rebleeding, confirmed by endoscopy
30 days
*All trials were randomized, controlled trials. Abbreviations: IV, intravenous; NA, not available.
Table 2 | Clinical outcomes from five studies comparing single endoscopy with second-look endoscopy studies
endoscopic therapy
n (second-look vs single endoscopy)
Villanueva et al.17
Injection therapy
Rutgeerts et al.19
second-look endoscopy vs single endoscopy rebleeding (%), odds ratio (95% CI)
surgery (%)
Death (%)
Mean unit of blood transfusion
Mean or median days of hospital stay
104 (52 vs 52)
21% vs 29% 0.66 (0.27–1.62)
8% vs 15%
2% vs 4%
1.7 vs 2.5
9.3 vs 11.8
Injection therapy
790 (270 vs 520*)
15% vs 21% 0.68 (0.46–1.00)
5% vs 3%
4% vs 5%
3.7 vs 3.2
NA
Messman et al.20
Injection therapy
105 (52 vs 53)
27% vs 21% 1.41 (0.57–3.47)
6% vs 4%
6% vs 4%
3.5 vs 3.1
14 vs 12
Saeed et al.18
Thermal coagulation
40 (19 vs 21)
0% vs 24%‡
0% vs 0%
5% vs 10%
0 vs 0.9‡
NA
Chiu et al.21
Thermal coagulation
194 (100 vs 94)
5% vs 14% 0.33 (0.11–0.96)
1% vs 6%
2% vs 2%
1.9 vs 2.1
4 vs 4
*Single therapy included polidocanol and fibrin glue. ‡P 5 or Baylor score (postendoscopy) >10. ‡Low and medium risks are defined as Rockall score ≤6 or Baylor score (pre-endoscopy) ≤5 or Baylor score (postendoscopy) ≤10.
defined as a rockall score of more than six. the rockall score is more accurate than the Baylor score in predicting rebleeding in low- and intermediate-risk patients.33 Future studies need to be performed on a larger scale, and need to focus on identifying high-risk patients. simplifying the scoring system, if possible, would also be desirable to make it more user-friendly, rather than including many clinical and endoscopic parameters.
PPIs versus second-look endoscopy
endoscopic treatment with intravenous administration of high-dose PPis has been demonstrated to significantly reduce the incidence of recurrent bleeding.34 the adjuvant use of high-dose PPis in endoscopic therapy has been endorsed in two international consensus recommendations.6,8 Could the use of PPis obviate the need for repeat endoscopy and retreatment? this notion is addressed in a randomized, controlled trial that compares high-dose intravenous infusion of PPis versus scheduled secondlook endoscopy.35 the study is not yet published in a full manuscript version, hence limited details are provided. in total, 305 patients were randomly allocated to receive highdose intravenous omeprazole with a single endoscopy, or routine second-look endoscopy without PPis. unlike previous studies by messmann20 and Chiu,21 an optimal dose of PPis (80 mg bolus followed by 8 mg per hour infusion) was used in this study. recurrent bleeding occurred in 6.5% of patients who received PPis and 7.9% who received second-look endoscopy (or 1.23, 95% Ci 0.51–2.93). this study indicates that there is no advantage of repeat endoscopy and retreatment if PPis are used as an adjuvant therapy with endoscopic hemostasis. Furthermore, routine scheduled second-look endoscopy incurs a higher hospital administrative cost than infusion of PPis.
Conclusions
after initial endoscopy and endoscopic hemostasis, and in the absence of clinical signs of rebleeding, a repeat endoscopic procedure using thermal coagulation may benefit a select group of high-risk patients—for example, those with advanced age and/or presence of a concurrent acute life-threatening illness. on the basis of existing evidence from the published literature, however, routine repeat endoscopy for peptic ulcer bleeding after initial endoscopic hemostasis cannot be recommended. the use of PPis may further reduce the need for second-look endoscopy. Future studies should focus on finding an improved method of selecting high-risk patients for second-look endoscopy. on the basis of the available scoring systems, we propose an algorithm to identify patients with a high risk of rebleeding who might benefit from second-look endoscopy (Figure 1).
Review criteria We searched the literature using the keywords “secondlook”, “repeated endoscopy”, “re-treatment” and ”ulcer bleeding” to identify English-language, full articles or abstracts from five computerized databases: MEDLINE (1950–December 2008), EMBASE (1980–December 2008), the Cochrane Central Register of Controlled Trials (October 2008), the Database of Abstracts of Reviews of Effects (October 2008), and the International Pharmaceutical Abstracts (1970–December 2008). Manual searches for abstracts published at major international conferences, including DDW, UEGW and APDW, over the past 5 years, and the scanning of articles from the reference lists of retrieved studies and other clinical reviews was also conducted.
nature reviews | gastroenterology & hepatology © 2009 Macmillan Publishers Limited. All rights reserved
volume 6 | DeCemBer 2009 | 721
REVIEWS 1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
Sung, J. J., Kuipers, E. J. & El-Serag, H. B. Systematic review: the global incidence and prevalence of peptic ulcer disease. Aliment. Pharmacol. Ther. 29, 938–946 (2009). Lau, J. Y., Sung, J. J., Metz, D. C. & Howden, C. W. Systematic review of the epidemiology of complicated peptic ulcer: incidence, recurrence, risk factors and mortality. Gastroenterology 134, A32 (2008). Cook, D. J., Gayatt, G. H., Salena, B. J. & Laine, L. A. Endoscopic therapy for acute nonvariceal hemorrhage: a meta-analysis. Gastroenterology 102, 139–148 (1992). Hirao, M. et al. Endoscopic local injection of hypertonic saline–epinephrine solution to arrest hemorrhage from the upper gastrointestinal tract. Gastrointest. Endosc. 31, 313–317 (1985). Chen, P. C., Wu, C. S. & Liaw, Y. F. Hemostatic effect of endoscopic local injection with hypertonic saline-epinephrine solution and pure ethanol for digestive tract bleeding. Gastrointest. Endosc. 32, 319–323 (1986). Steele, R. J., Park, K. J. & Crofts, T. J. Adrenaline injection for endoscopic haemostasis in non-variceal upper gastrointestinal haemorrhage. Br. J. Surg. 78, 477–479 (1991). Saeed, Z. A. Second thoughts about secondlook endoscopy for ulcer bleeding? Endoscopy 30, 650–652 (1998). Barkun, A. et al. Consensus recommendations for managing patients with nonvariceal upper gastrointestinal bleeding. Ann. Intern. Med. 139, 843–857 (2003). Adler, D. G. et al. ASGE guideline: The role of endoscopy in acute non-variceal upper-GI hemorrhage. Gastrointest. Endosc. 60, 497–504 (2004). British Society of Gastroenterology Endoscopy Committee. Non-variceal upper gastrointestinal haemorrhage: guidelines. Gut 51 (Suppl. 4), iv1–6 (2002). Chung, S. C. Preventing ulcer rebleeding: the role of second-look endoscopy. Can. J. Gastroenterol. 13, 409–411 (1999). Seewald, S. et al. Interventional endoscopic treatment of upper gastrointestinal bleeding— when, how and how often. Langenbeck Arch. Surg. 386, 88–97 (2001). Trap, R., Skarbye, M. & Rosenberg, J. Planned second look endoscopy in patients with bleeding duodenal or gastric ulcers. Dan. Med. Bull. 47, 220–223 (2000).
14. Babicki, A., Dobosz, M., Marczewski, R. & Wajda, Z. Evaluation of using fibrin tissue adhesive (Beriplast) and preparations of thrombin and adrenalin in injection hemostasis methods for gastric and duodenal ulcer hemorrhage. Randomized, prospective clinical trial [Polish]. Wiad. Lek. 50 (Suppl. 1, Pt 2), 383–387 (1997). 15. Ell, C. Scheduled endoscopic retreatment vs. single injection therapy in bleeding gastroduodenal ulcers: results of a multicenter study. Gastrointest. Endosc. 47, AB83 (1998). 16. Lin, C. K. et al. The value of second-look endoscopy after endoscopic injection for bleeding peptic ulcer. Gastroenterology 110, A177 (1996). 17. Villanueva, C. et al. Value of second-look endoscopy after injection therapy for bleeding peptic ulcer: A prospective and randomized trial. Gastrointest. Endosc. 40, 34–39 (1994). 18. Saeed, Z. A. et al. Endoscopic retreatment after successful initial hemostasis prevents ulcer rebleeding: a prospective randomized trial. Endoscopy 28, 288–294 (1996). 19. Rutgeerts, P. et al. Randomised trial of single and repeated fibrin glue compared with injection of polidocanol in treatment of bleeding peptic ulcer. Lancet 350, 692–696 (1997). 20. Messmann, H. et al. Effect of programmed endoscopic follow-up examinations on the rebleeding rate of gastric or duodenal peptic ulcers treated by injection therapy: a prospective, randomized controlled trial. Endoscopy 30, 583–589 (1998). 21. Chiu, P. W. et al. Effect of scheduled second therapeutic endoscopy on peptic ulcer rebleeding: a prospective randomised trial. Gut 52, 1403–1407 (2003). 22. Lai, K. H. et al. Endoscopic injection for the treatment of bleeding ulcers: local tamponade or drug effect. Endoscopy 26, 338–341 (1994). 23. Chung, S. C., Leung, F. W. & Leung, J. W. Is vasoconstriction the mechanism of hemostasis in bleeding ulcers with adrenalin? A study using reflectance spectrophotometry. Gastrointest. Endosc. 34, A174 (1988). 24. O’Brien, J. R. Some effects of adrenaline and anti-adrenaline compounds on platelets in vitro and in vivo. Nature 200, 763–764 (1963). 25. Kubba, A. K., Murphy, W. & Palmer, K. R. Endoscopic injection for bleeding peptic ulcer: a comparison of adrenaline alone with adrenaline plus human thrombin. Gastroenterology 111, 623–628 (1996). 26. Chung, S. C. et al. Randomised comparison between adrenaline injection alone and
722 | DECEMBER 2009 | voluME 6
27.
28.
29.
30.
31.
32.
33.
34.
35.
adrenaline injection plus heat probe treatment for actively bleeding ulcers. Br. Med. J. 314, 1307–1311 (1997). Chung, S. C., Lau, J. Y., Rutgeerts, P. & Fennerty, M. B. Thermal coagulation for nonvariceal bleeding. Endoscopy 34, 89–92 (2002). Saeed, Z. A., Winchesterm, C. B., Michaletz, P. A., Woods, K. L. & Graham, D. Y. A scoring system to predict rebleeding after endoscopic therapy of nonvariceal upper gastrointestinal haemorrhage, with a comparison of heat probe and ethanol injection. Am. J. Gastroenterol. 88, 1842–1849 (1993). Saeed, Z. A., Ramirez, F. C., Hepps, K. S., Cole, R. A. & Graham, D. Y. Prospective validation of the Baylor bleeding score for predicting the likelihood of rebleeding after endoscopic hemostasis of peptic ulcers. Gastrointest. Endosc. 41, 561–565 (1995). Marmo, R. et al. Dual therapy versus monotherapy in the endoscopic treatment of high-risk bleeding ulcers: a meta-analysis of controlled trials. Am. J. Gastroenterol. 102, 279–289 (2007). Spiegel, B. M., Ofman, J. J., Woods, K. & Vakil, N. B. Minimizing recurrent peptic ulcer hemorrhage after endoscopic hemostasis: the cost-effectiveness of competing strategies. Am. J. Gastroenterol. 98, 86–97 (2003). Rockall, T. A., Logan, R. F., Devlin, H. B. & Northfield, T. C. Risk assessment following acute gastrointestinal hemorrhage. Gut 38, 316–321 (1996). Camellini, L. et al. Comparison of three different risk scoring systems in non-variceal upper gastrointestinal bleeding. Dig. Liver Dis. 36, 271–277 (2004). Sung, J. J. et al. Intravenous esomeprazole for prevention of recurrent peptic ulcer bleeding: a randomized trial. Ann. Intern. Med. 150, 455–464 (2009). Chiu, P. W. et al. Scheduled 2nd endoscopy or high dose omeprazole infusion in preventing peptic ulcer rebleeding—a prospective randomized study. Gastroenterology 136, A43 (2009).
acknowledgments Charles P. Vega, University of California, Irvine, CA, is the author of and is solely responsible for the content of the learning objectives, questions and answers of the MedscapeCME-accredited continuing medical education activity associated with this article.
www.nature.com/nrgastro © 2009 Macmillan Publishers Limited. All rights reserved
