Drugs 2008; 68 (15): 2105-2111

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2105 1. Initial Management of the Patient with Upper Gastrointestinal Bleeding . . . . . . . . . . . . . . . . . . . . . . 2106 2. Histamine H2-Receptor Antagonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2106 3. Proton Pump Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2107 4. Somatostatin and Octreotide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2109 5. Tranexamic Acid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2110 6. Conclusions and Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2110 Upper gastrointestinal (UGI) bleeding occurs frequently and results in substanAbstract tial patient morbidity, mortality and medical expense. After initial resuscitation to stabilize the patient, carefully performed endoscopy provides an accurate diagnosis and can identify high-risk subgroups in ulcer patients who are likely to rebleed with medical therapy alone and would benefit most from endoscopic haemostasis. Several different pharmacological therapies have been used for patients with bleeding ulcers, including intravenous histamine H2-receptor antagonists, proton pump inhibitors, somatostatin and octreotide, and tranexamic acid. The results of several studies and meta-analyses favour high-dose, intravenous proton pump inhibitors, such as omeprazole or pantoprazole, after successful endoscopic haemostasis. For patients with ulcer bleeding and low-risk endoscopic stigmata, high-dose oral proton pump inhibitor therapy is suggested. Medical management with proton pump inhibitors is not a substitute for appropriate endoscopic therapy for patients with UGI bleeding and high-risk ulcer stigmata. Upper gastrointestinal (UGI) bleeding occurs frecases, and UGI bleeding is the most common comquently and is a common cause of hospitalization or plication of peptic ulcer disease.[1] Although other inpatient bleeding. Such bleeding results in substannon-variceal conditions, such as Mallory-Weiss tial patient morbidity, mortality and healthcare extear, angiodysplasia, watermelon stomach or Dieupense. Ulcer disease is the most common cause of lafoy’s lesion, may also cause UGI haemorrhage, severe UGI haemorrhage, causing about 40–50% of these occur much less frequently.[2] This article re2106 Kovacs & Jensen views the important aspects of the short-term ration of hospitalization and need for interventions medical management of UGI bleeding, secondary to (endoscopy, angiography or surgery). Histamine ulcers and other peptic or non-variceal lesions. H2-receptor antagonists (H2RAs), proton pump inhibitors (PPIs), somatostatin and octreotide, and tranexamic acid have been the most extensively stud1. Initial Management of the Patient with ied in the medical management of non-variceal UGI Upper Gastrointestinal Bleeding bleeding. The initial management of the patient with UGI The reason for the use of acid-reducing medicableeding should include evaluation of severity of the tions is based on studies showing that acid and haemorrhage, patient resuscitation, a brief medical pepsin interfere with the haemostatic process of history and physical examination, and consideration ulcers and non-variceal UGI lesions. In vitro studies of possible interventions.[1] The initial clinical asdemonstrated that (i) both the intrinsic and extrinsic sessment should focus on the patient’s haemodynapathways were adversely influenced by an acid enmic state, with a view to early resuscitation. Initial vironment; (ii) platelet aggregation was inhibited by medical therapy should be aimed at restoring blood acid; and (iii) pepsin activity was also highly acid volume by fluid replacement to ensure that tissue sensitive, with maximal clot lysis at a pH of 2, but perfusion and oxygen delivery are not comprolimited effect at pH above 5.[6] These results sugmised. Airway protection with endotracheal intubagested that increasing intragastric pH to greater than tion should be strongly considered in patients with 6 could improve the coagulation process. Furtherongoing haematemesis, altered mental or respiratory more, since clinical trials have shown that ulcer status, or severe neuromuscular disorders that prerebleeding occurs mainly during the first 72 hours, vent aspiration.[1,2] acid suppression should be maintained for at least After patient resuscitation and stabilization, en72 hours after haemostasis of non-variceal UGI doscopy is the preferred procedure for diagnosis and haemorrhage because any formed blood clots could treatment because of its high accuracy (diagnostic in be lysed with a return of an acidic intragastric enviabout 95% of severe UGI haemorrhage patients) and ronment. low complication rate. Endoscopy may also show After the initial bleed has been treated endoscopistigmata of haemorrhage on ulcers that have imporcally and haemostasis has been achieved, medical tant prognostic value for the risk of rebleeding.[1,3,4] management is recommended with PPIs for Patients with high-risk endoscopic findings, such as 6–8 weeks, unless the patient is also Helicobacter active bleeding, non-bleeding visible vessel or adpylori-positive, requires low-dose aspirin maintenherent clot, are likely to rebleed with medical therance or uses a non-selective NSAID. H. pyloriapy alone and would benefit from endoscopic positive patients should receive eradication therapy haemostasis in combination with medical theraand should be retested to document H. pylori eradipies.[1,3,4] Improved outcomes have included control cation 6–10 weeks after completion of antibacterial of active bleeding, and decreased rebleeding, transtherapy. Patients needing long-term aspirin or fusion requirements, duration of hospitalization and NSAIDs should receive PPI maintenance treatment mortality.[1,3-5] Rebleeding after endoscopic therapy to indefinitely reduce ulcer recurrence.[1,3] of UGI ulcers occurs in 10–25% of patients and represents a challenging problem.[5] Patients with a 2. Histamine H2-Receptor Antagonists clean-based ulcer or a flat dark spot rarely rebleed and do not require endoscopic haemostasis. These There have been many randomized trials using patients are best managed medically. H2RAs as medical treatment for acute UGI bleedAs with endoscopic haemostasis, the main goals ing because they were the first medications available of medical management include reduced morbidity, to inhibit acid secretion. In 1985, Collins and mortality, risk of rebleeding, transfusion needs, duLangman[7] reported the results of a meta-analysis of © 2008 Adis Data Information BV. All rights reserved. Drugs 2008; 68 (15) Medical Management of Non-Variceal UGI Bleeding 2107 27 randomized controlled trials on the use of H2RAs Several randomized controlled trials have demversus placebo in patients with ulcer bleeding. They onstrated the efficacy of high-dose PPI infusion showed that there was no effect on rebleeding and (omeprazole 80 mg bolus followed by 8 mg/h) for only a minor decrease in the need for surgery or 3 days after successful endoscopic treatment of pamortality.[7] Subgroup analysis demonstrated that tients with bleeding ulcers and high-risk stigmata of there was a significant reduction in rebleeding, morhaemorrhage.[15-17] Lau and co-workers[17] showed tality and surgery for gastric ulcer patients, but not that after primary haemostasis had been achieved by for duodenal ulcer patients.[7] A more recent metaendoscopic coagulation, high-dose omeprazole infuanalysis of H2RAs compared with placebo for peptic sion reduced the rate of rebleeding, transfusion reulcer haemorrhage showed no benefit on rebleeding, quirements and duration of hospitalization. Sung mortality or the need for surgery.[8] It should be and colleagues[18] reported similar prevention of renoted that these meta-analyses are limited because current bleeding in ulcer patients with non-bleeding most patients in the older studies did not receive visible vessels and adherent clots with combination endoscopic haemostasis for high-risk ulcer stigmata. endoscopic therapy and omeprazole infusion, comThe reason for these poor results is that intravepared with omeprazole infusion alone. These studies nous H2RAs are ineffective in maintaining a susillustrated the benefits of combination intravenous tained high gastric pH. Bolus injections, as frequentPPI infusion after endoscopic haemostasis, but not ly as every 4 hours (ranitidine 50 mg), produced an as a stand-alone therapy. intragastric pH >6 only 35% of the time.[9] Although More recently, several reviews and meta-analycontinuous intravenous infusion of H2RAs could ses of PPI use in peptic ulcer bleeding confirm that increase intragastric pH above 6, tolerance develPPIs reduce rebleeding, surgery, transfusion reoped within the first 12 hours of the infusion.[10] quirements and duration of hospitalization without This pharmacological tolerance could not be overdecreasing mortality.[19-24] come, even with large doses (ranitidine >500 mg/ Further review of the available studies suggested 24 h) and occurred with each of the other H2RAs important differences in outcomes of Asian versus (cimetidine and famotidine).[11] non-Asian patients in randomized controlled In summary, the results of both pharmacokinetic trials.[23] There was a significant reduction in 30-day studies and clinical trials do not support the use of mortality with PPI therapy for bleeding ulcers in the intravenous H2RAs for the medical management of Asian trials, but not in the non-Asian studies. The non-variceal UGI bleeding. effects of PPI therapy on rebleeding and the reduced need for surgery were also markedly greater in the 3. Proton Pump Inhibitors Asian than non-Asian trials.[23] Possible reasons to account for these different outcomes include: (i) younger age of Asian patients (57 years vs PPIs reduce both basal and stimulated acid secre66 years in non-Asian trials) with fewer co-morbidition by inhibiting the H+K+-adenosine triphosphaties; (ii) a lower parietal cell mass in Asian patients tase, the proton pump of the parietal cell.[12] Several leading to a more profound decrease in acid secrestudies have shown that the infusion of PPIs protion; (iii) a higher rate of H. pylori infection in Asian vides sustained, high intragastric pH,[11,13] and that patients, which is associated with a greater PPI an omeprazole infusion (80 mg bolus followed by effect on acid suppression;[25] and (iv) Asian patients 8 mg/h) can maintain intragastric pH steadily above are more likely to be slow metabolizers of PPIs.[26] 6 during a 72-hour period[11] without the developEach of these factors would produce an increased ment of tolerance. In the US, the only PPIs available antisecretory effect of PPI treatment in Asian comas intravenous formulation are pantoprazole and pared with non-Asian or heterogeneous study esomeprazole. These PPIs have been shown to progroups. vide potent and sustained acid suppression.[14] © 2008 Adis Data Information BV. All rights reserved. Drugs 2008; 68 (15) 2108 Kovacs & Jensen Three recent studies compared intravenous PPIs valent dose of intravenous PPI. Howden et al.[31] (pantoprazole 80 mg bolus follow by 8 mg/h continshowed that bolus intravenous lansoprazole 90 mg uous infusion for up to 72 hours) with intravenous followed by an intravenous infusion of 9 mg/h in H2RAs (ranitidine) in the management of ulcer paH. pylori-negative subjects maintained intragastric tients with high-risk stigmata who had been successpH >6 for only 36% of patients during the first fully treated with endoscopic haemostasis. In one 24-hour period and only 61% during the second study,[27] there was no benefit in rebleeding or mor24-hour period. Another PPI, pantoprazole (intravetality with PPIs over H2RAs. In another US study, nous 80 mg bolus followed by an intravenous infuthere was a trend to lower rebleeding with pantoprasion of 8 mg/h for 24 hours) produced intragastric zole than with ranitidine.[28] However, the small pH >6 for only 28% of the 24-hour observation number of patients may have limited the ability of period.[32] Since this was the pantoprazole dose that the trial to detect a true treatment difference. A was used in the two negative trials (pantoprazole vs Chinese study compared pantoprazole (40 mg bolus ranitidine),[27,28] the lack of effect in rebleeding may followed by 40 mg every 12 hours for 3 days) with have been secondary to ineffective acid suppression ranitidine (50 mg bolus followed by 50 mg every in the heterogeneous, non-Asian populations. This 18 hours for 3 days) and reported significantly lower may explain why these two trials did not provide the rebleeding rates after endoscopic haemostasis with beneficial clinical outcomes in UGI bleeding noted pantoprazole compared with ranitidine, but similar in other Asian studies. rates of transfusion requirements, hospital stay, need Two other aspects of PPI use in the medical for surgery and mortality.[29] management of non-variceal UGI bleeding have Two of these studies[27,28] included only H. pylorecently been considered. The first includes the pori-negative patients, and PPIs are reported to be less tential role of PPI use before endoscopy and the effective in reducing acid secretion in H. pylorisecond is that of oral administration. negative than in H. pylori-positive individuals.[25] In a retrospective report, Canadian authors found Also, most subjects were rapid metabolizers of PPIs that an intravenous PPI infusion prior to endoscopy according to cytochrome P450 2C19 status. These in patients with UGI bleeding and high-risk ulcer two studies may be more generalizable to the US stigmata lowered the proportion of actively bleeding and other heterogeneous populations, where most ulcers noted at endoscopy, without altering rebleedpatients with ulcer bleeding are likely to be H. pyloing, surgery or mortality rates compared with those ri-negative[30] and the majority of people extensively patients receiving PPIs after endoscopy.[33] Another metabolize PPIs. retrospective report suggested that pre-endoscopic Based on published randomized clinical trials, use of PPIs (both intravenous and oral) in patients the recommended dose of PPIs for patients with with ulcer haemorrhage significantly reduced adhigh-risk endoscopic findings is the equivalent of verse outcomes such as rebleeding, surgery, durabolus intravenous omeprazole 80 mg followed by an tion of hospitalization and mortality.[34] A recent 8 mg/h intravenous infusion for 72 hours.[21] Howprospective, randomized, placebo-controlled study ever, PPIs are not approved by the US FDA for such in Hong Kong showed that an intravenous omepramedical therapy of either UGI or peptic ulcer bleedzole bolus and infusion before endoscopy in patients ing. After the patient’s condition stabilizes, intravewith UGI haemorrhage decreased the need for endonous PPI therapy may be switched to oral PPI therscopic therapy, decreased the number of actively apy. Patients with low-risk endoscopic findings bleeding peptic ulcers and decreased the duration of (clean ulcer base or flat spot) should be treated with hospitalization.[35] These results raise the possibility high-dose (double the standard dose) oral PPIs. that pre-endoscopic high-dose PPI infusion may improve outcomes. A meta-analysis including a total However, recent studies suggest that North of 1512 patients confirmed that PPI therapy prior to American subjects may require an even higher equi© 2008 Adis Data Information BV. All rights reserved. Drugs 2008; 68 (15) Medical Management of Non-Variceal UGI Bleeding 2109 endoscopy in patients with UGI bleeding signifiendoscopic therapy in the treatment of oesophageal cantly decreased the proportion of patients with and gastric variceal bleeding. They act by decreasstigmata of haemorrhage, but did not demonstrate ing splanchnic blood flow. In addition to reducing any significant benefit in important clinical outgastroduodenal mucosal blood flow, somatostatin comes such as mortality, rebleeding or surgery.[36] In and octreotide also decrease gastric acid secretion patients with UGI haemorrhage, intravenous PPI by inhibiting gastrin release from antral G cells and therapy before endoscopy appears reasonable in histamine from enterochromaffine-like cells, deview of the previously documented benefits and crease pepsin secretion, and stimulate mucus pronegligible risks associated with PPI infusion. duction.[44] The effect on inhibition of pepsin secreUse of oral PPIs in non-variceal UGI bleeding is tion is a potential advantage of somatostatin and a controversial aspect. In an Asian population, a analogues over H2RAs and PPIs, which do not afhigh dose of oral omeprazole (40 mg twice daily fect pepsin secretion. compared with placebo) significantly reduced Somatostatin and its analogues have been used in rebleeding in ulcer patients with a visible vessel or the treatment of patients with ulcer bleeding to take adherent clots.[37] More recent trials have suggested advantage of these effects on blood flow and gastric that (i) high-dose oral PPI (pantoprazole 40 mg secretion. A small study found that in patients with twice daily[38] or omeprazole 40 mg/day)[39] was just bleeding peptic ulcers, somatostatin 250 μg bolus as effective as an intravenous infusion[38,39] after injection followed by 500 μg/h infusion was more endoscopy therapy; (ii) oral PPI (omeprazole 40 mg effective than pantoprazole 80 mg bolus followed by twice daily) was as effective as intravenous 8 mg/h infusion in maintaining high intragastric pH, omeprazole in ulcer patients with low-risk stigmata especially in the first 12 hours.[45] In another trial of haemorrhage;[40] and (iii) oral PPI (rabeprazole comparing intravenous somatostatin with intrave20 mg twice daily) was as effective as endoscopic nous pantoprazole after endoscopic treatment, sotreatment with haemoclips.[41] High-dose intravematostatin was not as effective in preventing renous PPI treatment is expensive, while oral PPIs are bleeding.[46] Mortality and surgery rates were simimuch less costly. Cost-effectiveness analyses in palar in the two treatment groups. However, this tients with high-risk endoscopic stigmata who had study[46] used a lower dose of somatostatin (250 μg successful endoscopic therapy have shown that both bolus followed by 250 μg/h infusion) than the prior intravenous and oral PPI treatment are very effective one that showed superior control of acid secreand less expensive than intravenous H2RAs or tion.[45] placebo.[43] When intravenous PPI was compared Octreotide has usually been administered as a with oral PPI, divergent results were obtained, with 50–100 μg bolus followed by a continuous infusion one analysis favouring intravenous administraof 25 μg/h for up to 3 days. A meta-analysis of tion[43] and the other supporting oral administraavailable trials comparing somatostatin and octreotion.[42] tide with placebo or H2RAs in ulcer bleeding reported that somatostatin and octreotide reduced the risk 4. Somatostatin and Octreotide of continued bleeding or rebleeding from ulcers.[47] Surgery was not significantly reduced by somatostaSomatostatin is a 14 amino-acid peptide hortin or octreotide, and mortality was not assessed.[47] mone, which is widely distributed in the stomach, None of the trials in this meta-analysis had endogastrointestinal tract and pancreas, as well as the scopic therapy as part of the management, which CNS. Octreotide, a synthetic analogue of somatostalimited the interpretation and relevance to current tin has similar biological effects and a longer halfmanagement. life. Octreotide is available in the US, but somatosAt present, although there may be theoretical tatin is not. Somatostatin and octreotide have been advantages to their use, there is no firm evidence to successfully used either alone or combined with © 2008 Adis Data Information BV. All rights reserved. Drugs 2008; 68 (15) 2110 Kovacs & Jensen recommend somatostatin or octreotide infusion over References 1. Kovacs TOG, Jensen DM. Recent advances in the endoscopicPPI therapy for non-variceal UGI haemorrhage. Fur-diagnosis and therapy of upper gastrointestinal, small intesti-thermore, somatostatin and octreotide infusions arenal, and colonic bleeding. Med Clin N Am 2002; 86: 1319-56 2. Kovacs TOG. Mallory-Weiss tears, angiodysplasia, watermelonmore expensive than PPI infusions.[44]stomach and Dieuafoy’s: a potpourri. Tech Gastrointest En-dosc 2005; 7: 139-475. Tranexamic Acid3. Kovacs TOG, Jensen DM. Endoscopic treatment of peptic ulcerbleeding. Curr Treat Opt Gastroenterol 2007; 10: 143-8Tranexamic acid is an antifibrinolytic agent that 4. Jensen DM, Kovacs TOG, Jutabha R, et al. Randomized trial ofmedical or endoscopic therapy to prevent recurrent ulcer hem-inhibits plasminogen activators. It is not approvedorrhage in patients with adherent clots. Gastroenterology 2002;by the FDA for the therapy of non-variceal UGI or123: 407-13peptic ulcer haemorrhage. A meta-analysis of six 5. Jensen DM. Treatment of patients at high risk for recurrentbleeding from a peptic ulcer. Ann Intern Med 2003; 139: 294-5randomized trials of patients with UGI bleeding6. Green Jr FW, Kaplan MM, Curtis LE, et al. Effects of acid andreported that tranexamic acid significantly reducedpepsin on blood coagulation and platelet aggregation: a poss-mortality compared with placebo, without reducingible contributor to prolonged gastroduodenal mucosal hemor-rhage. Gastroenterology 1978; 74: 38-43rebleeding or the need for surgery.[48] However,7. Collins R, Langman M. Treatment with histamine H2-antago-when the results were reviewed and re-analysed, thenists in acute upper gastrointestinal hemorrhage: implicationsof randomized trials. N Engl J Med 1985; 313: 660-6mortality effect was no longer significant. Also,8. Levine JE, Leonitiadis GI, Sharma VK, et al. Meta-analysis: thethese trials did not include endoscopic haemostasisefficacy of intravenous H2-receptor antagonists in bleedingfor high-risk ulcer lesions. Tranexamic acid shouldpeptic ulcer. Aliment Pharmacol Ther 2002; 16: 1137-42 9. Lanas A, Artal A, Blas JM, et al. Effect of parenteral, omepra-not be used in the management of ulcer haemor-zole, and ranitidine on gastric pH and the outcome of bleedingrhage.peptic ulcer. J Clin Gastroenterol 1995; 21: 103-6 10. Labenz J, Peitz U, Leusing C, et al. Efficacy of primed infusions6. Conclusions and Recommendationswith high dose ranitidine and omeprazole to maintain highintragastric pH in patients with peptic ulcer bleeding: a pro-spective, randomized, controlled study. Gut 1997; 40: 36-41UGI bleeding, secondary to peptic ulcer bleed-11. Netzer P, Gaia C, Sando Z, et al. Effect of repeated injection anding, is a common cause of hospitalization and inpa-continuous infusion of omeprazole and ranitidine on intragas-tient bleeding, and results in substantial patient mor-tric pH over 72 hours. Am J Gastroenterol 1999; 94: 351-7 12. Sachs G, Shin JM, Howden CW. The clinical pharmacology ofbidity and mortality. The results of randomized con-proton pump inhibitors. Aliment Pharmacol Ther 2006; 23trolled trials and meta-analyses have shown thatSuppl. 2: 2-8PPIs improve clinical outcomes in patients with 13. Merki HS, Wilder-Smith CH. Do continuous infusions ofomeprazole and ranitidine retain their effect after prolongedulcer haemorrhage. Patients with ulcers with high-dosing? Gastroenterology 1994; 106: 60-4risk endoscopic stigmata should receive high-dose 14. Van Rensburg CJ, Hartmann M, Thorpe A, et al. Intragastric pHduring continous infusion with pantoprazole in patients withintravenous PPI therapy after successful endoscopicbleeding peptic ulcer. Am J Gastroenterol 2003; 98: 2635-41treatment. Patients with low-risk endoscopic stig-15. Hasselgren G, Lind T, Lundell L, et al. Continuous intravenousmata should receive an oral PPI at twice the usualinfusion of omeprazole in elderly patients with peptic ulcerbleeding: results of a placebo-controlled multicenter study.clinical dose. High-dose intravenous PPI therapyScand J Gastroenterol 1997; 32: 328-33prior to endoscopy appears reasonable but is expen16. Lin HJ, Lo WC, Cheng YC, et al. Role of intravenousomeprazole in patients with highrisk peptic ulcer bleedingsive. The use of oral PPI administration soon afterafter successful endoscopic epinephrine injection: a prospec-successful endoscopic treatment is still controver-tive, randomized, comparative trial. Am J Gastroenterol 2006;sial.101: 500-5 17. Lau JY, Sung JJ, Lee KK, et al. Effects of intravenousomeprazole on recurrent bleeding after endoscopic treatmentAcknowledgementsof bleeding peptic ulcers. N Engl J Med 2000; 343: 310-6 18. Sung JJY, Chan FKL, Lau JYW. The effect of endoscopicNo sources of funding were used to assist in the prepara-therapy in patients receiving omeprazole for bleeding ulcerstion of this article. Dr Jensen has acted as a consultant forwith non-bleeding visible vessels or adherent clots. Ann InternAstraZeneca and has received both honoraria and researchMed 2003; 139: 237-43grant support from AstraZeneca. Dr Kovacs has no conflicts 19. Leontiadis GI, Sharma VK, Howden CW. Systematic reviewof interest that are directly relevant to the content of thisand meta-analysis of proton pump inhibitor therapy in pepticreview.ulcer bleeding. BMJ 2005; 330: 568-70 © 2008 Adis Data Information BV. All rights reserved.Drugs 2008; 68 (15) Medical Management of Non-Variceal UGI Bleeding2111 20. Andriulli A, Annese V, Caruso N, et al. Proton-pump inhibitors 36. Dorward S, Sreedharan A, Leontiadis GI, et al. Proton pumpand outcome of endoscopic hemostasis in bleeding pepticinhibitor treatment initiated prior to endoscopic diagnosis inulcers: a series of meta-analyses. Am J Gastroenterol 2005;upper gastrointestinal bleeding. Cochrane Database of Syst100: 207-19Rev 2006; (4): CD005415 21. Barkun A, Bardou M, Marshall JK, et al. Nonvariceal Upper GI 37. Khuroo MS, Yattoo GN, Javid G, et al. A comparison ofBleeding Consensus Conference Group. Consensus recom-omeprazole and placebo for bleeding peptic ulcer. N Engl Jmendations for managing patients with nonvariceal upper gas-Med 1997; 336: 1054-8trointestinal bleeding. Ann Intern Med 2003; 139: 843-5738. Bajaj JS, Dua KS, Hanson K, et al. Prospective, randomized trial22. Bardou M, Toubouti Y, Benhaberou-Brun D, et al. Meta-ana-comparing effect of oral versus intravenous pantoprazole onlysis: proton-pump inhibition in high-risk patients with pepticrebleeding after nonvariceal upper gastrointestinal bleeding: aulcer bleeding. Aliment Pharmacol Ther 2005; 21: 677-86pilot study. Dig Dis Sci 2007; 52: 2190-423. Leontiadis GI, Sharma VK, Howden CW. Systematic reviewand meta-analysis. Enhanced efficacy of proton pump inhibitor 39. Murthy S, Keyvani L, Leeson S, et al. Intravenous versus high-therapy for peptic ulcer bleeding in Asia: a post hoc analysisdose oral proton pump inhibitor therapy after endoscopic he-from the Cochrane Collaboration. Aliment Pharmacol Thermostasis of high-risk lesions in patients with acute non-2005; 21: 1055-61variceal upper gastrointestinal bleeding. Dig Dis Sci 2007; 52:24. Leontiadis GI, Sharma VK, Howden CW. Proton pump inhibi-1685-90tor treatment for acute peptic ulcer bleeding. Cochrane Data40. Yilmaz S, Bayan K, Juzunm Y, et al. A head to head comparisonbase Syst Rev 2006; (1): CD002094of oral vs intravenous omeprazole for patients with bleeding25. van Herwaarden MA, Samson M, van Nispen CHM, et al. Thepeptic ulcers with a clean base, flat spots, and adherent clots.effect of Helicobacter pylori eradication on intragastric pHWorld J Gastroenterol 2006; 12: 7837-43during dosing with lansoprazole or ranitidine. Aliment41. Kim JI, Cheung DY, Cho SL, et al. Oral proton pump inhibitiorsPharmacol Ther 1999; 13: 731-40are as effective as endoscopic treatment for bleeding peptic26. Caraco Y, Lagerstrom PO, Wood AJJ. Ethnic and genetic deter-ulcer: a prospective, randomized, controlled trial. Dig Dis Sciminants of omeprazole disposition and effect. Clin Pharmacol2007; 52: 3371-76Ther 1996; 60: 157-67 27. Barkun A, Racz I, van Rensburg C, et al. Prevention of peptic 42. Spiegel BMR, Dulai GS, Lim BS, et al. The cost-effectivenessulcer rebleeding using continuous infusion of pantoprazole vsand budget impact of intravenous versus oral proton pumpranitidine: a multicenter, multinational, randomized, doubleinhibitors in peptic ulcer hemorrhage. Clin Gastroenterolblind, parallel-group comparison [abstract]. GastroenterologyHepatol 2006; 4: 988-972004; 126: A-7843. Barkun AN, Herba K, Adam U, et al. The cost-effectiveness of28. Jensen DM, Pace SC, Soffer E, et al. Continuous infusion ofhigh-dose oral proton pump inhibition after endoscopy in thepantoprazole versus ranitidine for prevention of ulcer rebleed-acute treatment of peptic ulcer bleeding. Aliment Pharmacoling: a US multicenter, randomized, double-blind study. Am JTher 2004; 20: 195-202Gastroenterol 2006; 101: 1991-944. Sgouros SN, Bergele C, Viozis N, et al. Somatostain and its29. Hsu PI, Lo GH, Lo CC, et al. Intravenous pantoprazole versusanalogues in peptic ulcer bleeding: facts and pathophysiologi-ranitidine for prevention of rebleeding after endoscopic he-cal aspects. Dig Liver Dis 2006; 38: 143-8mostasis of bleeding peptic ulcers. World J Gastroenterol2004; 10: 3666-945. Averinos A, Sgouros S, Viazis N, et al. Somatostatin inhibits30. Jensen DM, Savides T, Sitzer M, et al. Demographics, riskgastric acid secretion more effectively than pantoprazole infactors, and outcomes of peptic ulcer hemorrhage in the Unitedpatients with peptic ulcer bleeding: a prospective, randomized,States: results of a large multicenter study [abstract]. Gastroen-placebo-controlled trial. Scand J Gastroenterol 2005; 40: 515-terology 2003; 124: A-172231. Howden CW, Metz DC, Hunt B, et al. Dose-response evaluation 46. Tsibouris P, Zintzaras E, Loppos C, et al. High-dose pantopra-of the antisecretory effect of continuous infusion intravenouszole continuous infusion is superior to somatostain after endo-lansoprazole regimen over 48 h. Aliment Pharmacol Therscopic hemostasis in patients with peptic ulcer bleeding. Am J2006; 23: 975-84Gastroenterol 2007; 102: 1192-932. Metz DC, Amer F, Hunt B, et al. Lansoprazole regimen that47. Imperiale TF, Birgisson S. Somatostatin or octreotide comparedsustain intragastric ph >6.0: an evaluation of intermittent oralwith H2 antagonists and placebo in the management of acuteand continuous intravenous infusion dosages. Alimentnonvariceal upper gastrointestinal hemorrhage: a meta-ana-Pharmacol Ther 2006; 23: 985-95lysis. Ann Intern Med 1997; 127: 1062-7133. Andrews CN, Levy A, Fishman M, et al. Intravenous protonpump inhibitors before endoscopy in bleeding peptic ulcers 48. Henry DA, O’Connell D. Effects of fibrinolytics inhibitors onwith high-risk stigmata: a multi-center comparative study. Canmortality from upper gastrointestinal hemorrhage. BMJ 1989;J Gastroenterol 2005; 19: 667-71298: 1142-634. Keyvani L, Murthy S, Leeson S, et al. Pre-endoscopic protonpump inhibitor therapy reduces recurrent adverse gastrointesti-nal outcomes in patients with acute non-variceal upper gastroCorrespondence: Dr Thomas O.G. Kovacs, CURE Digestiveintestinal bleeding. Aliment Pharmacol Ther 2006; 24: 1247Disease Research Center, VA Greater Los Angeles Health-55care System, Building 115, Room 212, 11301 Wilshire35. Lau JY, Leung WK, Wu JCY, et al. Omeprazole before endos-Boulevard, Los Angeles, CA 90073-1003, USA.copy in patients with gastrointestinal bleeding. N Engl J MedE-mail: [email protected]; 356: 1631-40 © 2008 Adis Data Information BV. All rights reserved.Drugs 2008; 68 (15)