You are on page 1of 8

American Journal of Gastroenterology


C 2008 by Am. Coll. of Gastroenterology
Published by Blackwell Publishing

ISSN 0002-9270
doi: 10.1111/j.1572-0241.2008.02149.x

ORIGINAL CONTRIBUTIONS
Endoscopy

High- Versus Low-Dose Proton Pump Inhibitors After
Endoscopic Hemostasis in Patients With Peptic Ulcer
Bleeding: A Multicentre, Randomized Study
Angelo Andriulli, M.D.,1 Silvano Loperfido, M.D.,2 Rosaria Focareta, M.D.,3 Pietro Leo, M.D.,4
Fabio Fornari, M.D.,5 Antonietta Garripoli, M.D.,6 Paolo Tonti, M.D.,7 Sergio Peyre, M.D.,8
Antonio Spadaccini, M.D.,9 Riccardo Marmo, M.D.,10 Antonio Merla, M.D.,1 Alessandro Caroli, M.D.,2
Gian Battista Forte, M.D.,3 Angelo Belmonte, M.D.,4 Giovanni Aragona, M.D.,5 Gianni Imperiali, M.D.,11
Fabrizio Forte, M.D.,3 Fabio Monica, M.D.,2 Nazario Caruso, M.D.,1 and Francesco Perri, M.D.1
Divisions of Gastroenterology, 1 “Casa Sollievo della Sofferenza” Hospital, IRCCS, San Giovanni, Rotondo,
Italy, 2 Regional Hospital, Treviso, Italy, 3 “ S. Sebastiano” Hospital, Caserta, Italy, 4 P.O. Annunziata Hospital,
Cosenza, Italy, 5 “G. da Saliceto” Hospital, Piacenza, Italy, 6 “ Maria Vittoria” Hospital, Torino, Italy,
7
Ospedali Riuniti, Foggia, Italy, 8 Ivrea Hospital, Ivrea, Italy, 9 “S. Pio da Pietrelcina” Hospital, Vasto, Italy,
10
“L. Curto” Hospital, Polla, Italy, and 11 Valduce Hospital, Como, Italy
BACKGROUND:

The most effective schedule of proton pump inhibitor (PPI) administration following endoscopic
hemostasis of bleeding ulcers remains uncertain.
METHODS:
Patients with actively bleeding ulcers and those with nonbleeding visible vessel or adherent clot
were treated with epinephrine injection and/or thermal coagulation, and randomized to receive
intravenous PPIs according to an intensive regimen (80 mg bolus followed by 8 mg/h as
continuous infusion for 72 h) or a standard regimen (40 mg bolus daily followed by saline infusion
for 72 h). After the infusion, all patients were given 20 mg PPI twice daily orally. The primary end
point was the in-hospital rebleeding rate, as ascertained at the repeat endoscopy.
RESULTS:
Bleeding recurred in 28 of 238 patients (11.8%) receiving the intensive regimen, and in 19 of 236
(8.1%) patients receiving the standard regimen (P = 0.18). Most rebleeding episodes occurred
during the initial 72-h infusion: 18 (7.6%) and 19 events (8.1%) in the intensive and standard
groups, respectively (P = 0.32). Mean units of blood transfused were 1.7 ± 2.1 in the intensive and
1.5 ± 2.1 in the standard regimen group (P = 0.34). The duration of hospital stay was <5 days for
88 (37.0%) and 111 patients (47.0%) in the intensive and standard groups (P = 0.03). There were
fewer surgical interventions in the standard versus intensive regimen (1 vs 3). Five patients in each
treatment group died.
CONCLUSIONS: Following endoscopic hemostasis of bleeding ulcers, standard-dose PPIs infusion was as effective
as a high-dose regimen in reducing the risk of recurrent bleeding. (ClinicalTrials.gov number,
NCT00374101).
(Am J Gastroenterol 2008;103:3011–3018)

INTRODUCTION
For patients with bleeding peptic ulcers that display major
endoscopic stigmata of recent hemorrhage, a combination of
endoscopic and pharmacologic therapy is the current standard
management; however, the optimal regimen for administration of proton pump inhibitors (PPIs) remains controversial.
Two consensus documents have endorsed a high-dose PPI
regimen (80 mg stat followed by an infusion of 8 mg/h for
72 h) (1, 2). The biologically plausible mechanism of benefit
of such a high-dose regimen is to promote clot stability by
sustaining the intragastric pH above 6 (3, 4). Once primary
hemostasis is achieved by endoscopic therapy, clinical trials
show that a high-dose PPI infusion is superior to placebo

(5, 6); however, when the comparator is a standard-dose PPI
regimen, four prospective trials (7–10) and a meta-analysis
(11) report no difference in the magnitude of risk reduction
between the intensive- and the low-dose regimens.
We conducted a head-to-head study, comparing two strategies for intravenous PPI administration in the prevention
of rebleeding, surgery, and death in patients with highrisk bleeding peptic ulcers in whom successful endoscopic
hemostasis was achieved.

METHODS
A randomized, double-blind study of PPI therapy for the
prevention of ulcer rebleeding in high-risk patients after

3011

05. Schedule of Intravenous PPIs As this study was not supported financially by a pharmaceutical company. designing. Patients clinically suspected of rebleeding were those who after an initial stabilization presented with at least one of the following signs: decrease in blood pressure (≤100 mmHg). two bolus injections of saline were administered on the second and third day) or the standard regimen (40 mg bolus of PPI once daily. power = 0. transfusion of ≥2 units of packed blood cells or hemoglobin ≤10 g/dL. Patients with a clinical suspicion of rebleeding underwent a second attempt at endoscopic hemostasis: rebleeding was diagnosed if the ulcer was actively bleeding or if there was fresh blood in the stomach.5). 8) that randomized patients with bleeding ulcers after successful endoscopic intervention to highdose or standard PPI regimen have reported rebleeding rates of 8.000. calculated on patients’ demographic and clinical characteristics or volume and rapidity of blood loss: age ≥70 yr. patients were switched to oral PPI therapy (20 mg twice daily) until discharge. and conducting of the trial or in the analysis of the experimental data was pursued. need for continuous anticoagulation. Investigators were provided with a computer-generated list of random numbers in blocks of 30.2%. Patients in the lowrisk group (Rockall score <6) were offered early resumption of feeding and monitored clinically. and those who received PPI therapy before the index endoscopy. Exclusion criteria were malignant-appearing ulcers. individual investigators were allowed to use either omeprazole or pantoprazole. endoscopic hemostasis was conducted in 11 Italian centres on a nonprofit. a 13. The choice was independent of investigators’ preference. a sample size of 235 patients per each treatment arm was estimated (α = 0. According to Rockall’s criteria (12). Hemostasis was considered to be established if bleeding had stopped. planning. or when bleeding started if the patient was already in hospital for other reasons. was at the discretion of the treating endoscopist. Secondary outcomes included the need for surgery. ulcers with a flat spot or a clean base. Moreover. diagnosed at a repeat endoscopy. Clots covering ulcer lesions were washed by a water pump. Patients with unsuccessful endoscopic hemostasis were not randomized. in order to assure blindness. as determined by a Rockall ≥6 score at admission. the attending nurse in the endoscopic suite prepared the locally available PPI in identically appearing solutions and started an infusion of either the high-dose intensive regimen (loading dose of 80 mg on the first day followed by continuous infusion of 8 mg/h for 72 h. or reappearance of overt bleeding (new hematemesis or melena). the only two intravenous formulations commercially available in Italy at the time this trial was planned. The present trial was designed as a superiority study. and hemodynamic instability. decline in hemoglobin (>20 g/L). Hemodynamically unstable patients were initially resuscitated and then considered for enrolment if their condition stabilized. and was based on the agents’ acquisition costs at the pharmacy department of the local hospital. defined either by hypotension (systolic blood pressure ≤90 mmHg) or tachycardia (heart rate ≥100 beats per min).2% and 12. Patient Population Patients who presented with overt gastrointestinal bleeding or a recent history (<24 h before presentation) of hematemesis and/or melena to their hospital emergency departments were eligible. voluntary basis after the protocol received approval by local ethics committees. severe comorbid conditions. Sample Size Estimate Two previous trials (7. All study subjects gave written informed consent and were enrolled between October 2004 and March 2007. Patients were then transferred to a gastroenterology ward for monitoring and continuation of therapy. whether the epinephrine injection (1:10. and underlying nonbleeding visible vessels or adherent clots received endoscopic treatment. Eligible patients were required to have an ulcer with either active bleeding (spurting arterial or persistent oozing) or a nonbleeding lesion (nonbleeding visible vessel or adherent clot) at endoscopy. increase in pulse rate (≥100 beats per min). no change in hemoglobin levels with red blood cell transfusions. 1–1. followed by continuous infusion of saline for 72 h) according to the list of randomization. severe coagulopathy (platelet count <100. After randomization. respectively. The reference time was the onset of symptoms and signs. and mortality. defined as a medical history of chronic illness or presence of acute medical condition. . length of hospital stay.3012 Andriulli et al.5 cc/injection) was administered as monotherapy (unimodal) or in association with either thermal or mechanical therapy (multimodal). transfusion requirement. After the initial 72 h.2% probability of rebleeding was calculated in a meta-analysis of eight trials on PPI infusion with or without endoscopic therapy (13). Endoscopists who carried out the initial examination and the nurses and physicians attending the individual patients in the ward were unaware of the group assignment. Endoscopic Hemostasis The modality of endoscopic hemostasis. To show a statistically significant difference in rebleeding rate between an expected 15% in the standard arm and 7% in the high-dose group. as were patients whose ulcer hemorrhage started after hospitalization for an unrelated medical or surgical condition.80). concomitant illness. without stratification. patients determined to be at high clinical risk for rebleeding clinically were those with a Rockall score ≥6. but was mandatory in all patients with a suspected rebleeding.000 dilution in saline. Outcomes The primary end point was the in-hospital occurrence of rebleeding. No industry support in the ideation. A repeat endoscopy was selectively used in high-risk patients. international normalized ratio >1.

Analyses were performed using SPSS. clinical characteristics. anti-inflammatory drugs. statistical software (Chicago. withdrawal of consent (N = 1).425 patients were screened for inclusion. 1). and continuous data as means ± standard deviation (SD).and low-dose regimen) were excluded before the first injection of drugs for uncertainty about the success of endotherapy (N = 3). The treatment groups were similar with respect to patients demographics. and use of aspirin or nonsteroidal. RESULTS A total of 1. endoscopic modality.Proton-Pump Inhibitors After Endoscopic Hemostasis 3013 1425 Evaluated 943 Excluded: 125 declined participation 383 F2c / F3 ulcers 74 Mallory–Weiss or gastric erosions 33 cancer ulcers 12 unstable condition or moribond 84 Had begun to receive PPI 27 unable to provide written informed consent 43 on anticoagulation therapy 22 failed endoscopic therapy 140 Excluded for other reasons 482 Randomly assigned to a study group 243 Assigned to 239 Assigned to an intensive regimen standard regimen 5 Excluded from 3 Excluded from analysis analysis 1 Withdrew voluntarily 3 Withdrew 1 Had a final diagnosis of voluntarily pancreatic gastrinoma 2 Had warfarin 1 Had warfarin anticoagulation anticoagulation 238 included in 236 included in analyses analyses Figure 1. and protocol violation (N = 4) (Fig. type of the index bleeding (whether active or inactive). the twosample t-test was used for continuous variables and the Fisher exact test used for discrete variables. PPI administration strategy (standard or intensive regimen). For other P values. and 482 patients underwent randomization. Hazard ratios and their 95% confidence intervals (CI) were calculated using es- timated regression coefficients (B) and their standard errors (SE) in the Cox regression analysis. Inc. A prediction model for rebleeding was developed including the following covariates: Rockall score (≥ or <6). P values for the primary end point were obtained from the two-sided χ 2 Pearson’s test. 8 randomized patients (3 and 5 patients in the high. Flow chart of patients throughout the study. modality of endoscopic therapy (unimodal or multimodal). Statistical Analysis All patients who took at least one dose of drugs were included into the analysis. type of PPI administered (omeprazole or pantoprazole). Categorical data were expressed as proportions (%). IL). and PPI .

(%) Severe comorbidity—no. Between the two treatment groups.6) 73 (30.9) 35 (14.3014 Andriulli et al.8%.7) 151 (63.4) 146 (61.4) 75 (31.0) 21 (8. (Table 1).32). and more patients had a Rockall score ≥6.0) in the intensive regimen group and in 19 (8.7) 77 (32.4) 32 (13.8) 163 (69. rebleeding rates did not differ for gastric and duodenal ulcers . presented with a severe comorbidity (P = 0.2) 49 (20.0) 136 (57.5) 72 (30.8 ± 16.3 64 (26.7 ± 1.3 ± 15.9) 85 (36.3) 30 (12.1) 9. (%) Previous bleeding—no. Characteristics of the 474 Patients Included in the Analysis Characteristic Age—yr Age > 70 yr—no. Recurrent bleeding was most common during the period of intravenous administration: within the first 3 days. about half were older than 70 yr. CI 4.7) 10 (4.5) 81 (34. 9 lesions were spurting ulcers.9) 36 (15.4) 167 (70.6) 93 (39. 13 had a nonbleeding visible vessel.9) in the intensive regimen group and 22 (9. or adherent clot was not significantly different. (%) Bleeding during hospitalization no. 70% of patients received omeprazole and the remaining patients received pantoprazole. At the repeat endoscopy.4 66. (%) 1 a—spurting 1 b – oozing 2 a—nonbleeding visible vessel 2 b—adherent clot Endoscopic hemostasis modality—no. Bleeding recurred in 28 patients (11. respectively. CI 5.2) 15 (6.1 ± 2.5–13. (%) Duodenal ulcer Gastric ulcer Ulcer size ≥ 20 mm—no.6 116 (48.4) 15 (6. and a majority had taken aspirin or nonsteroidal anti-inflammatory drugs before hospitalization.5–11. About half of ulcers were injected with epinephrine solution.1) 78 (32. (%) Hemoglobin—g/dL Patients with hemoglobin ≤ 10 g/dL—no.0) 88 (37.2) 71 (29.4) 74 (31.8) 143 (60.1%.1) 32 (13.2 155 (65. and prevalence of active arterial bleeding. Nineteen and 30 patients in the intensive and standard regimen groups.1) 73 (30.4) 82 (34.9) NSAID = nonsteroidal anti-inflammatory drug.6) 100 (42. a difference of 3.7 121 (51. (%) Omeprazole Pantoprazole Intensive Regimen (N = 238) Standard Regimen (N = 236) 66. SD = standard deviation.09).5 81 (34. An equal number of patients had duodenal and gastric ulcers. all but 2 of previous lesions were successfully controlled by a second attempt at endoscoscopic hemostasis.8) 33 (13.5) 82 (34.8% (CI –1.18) (Table 2). and the remaining received epinephrine in association with thermal therapy (bipolar electrocoagulation or argon plasma coagulation) or apposition of clips.4) 25 (10. 20 oozing.0) 5.3) 74 (31.1) 20 (8.3) 156 (66. mean Rockall score higher. (%) Men—no.0) 53 (22. the remaining 3 lesions were classified as Forrest 2c or 3 and were not treated endoscopically.8) 4. 16 patients (6. (%) Unimodal Multimodal Type of PPI administered—no. CI 3.0) 119 (50.5 ± 2.4) 25 (10.1%) (P = 0.3) 14 (5.0) 25 (10.6–16.5–9.3%. CI 7. (%) Peptic ulcer as source of bleeding—no.5) 119 (50. nonbleeding visible vessel.4) 9. Signs of hemodynamic instability were more frequent in the standard regimen: shock was more common. oozing.0 ± 1. (%) Forrest classification of endoscopic stigmata of hemorrhage—no. All patients completed the assigned schedule of PPIs administration.7%.5) in the standard regimen group.9) 150 (63. (%) Rockall score Mean ± SD ≥6 points—no. No drug-related side effects were reported in either group. (%) Shock at presentation—no. therapy (Table 1): two-thirds of patients were men. In the two treatment groups. Equal proportions of patients bled during their hospital stay for unrelated illnesses. (%) Use of NSAID Use of aspirin Time from bleeding to endoscopy ≤6 h Between 6 and 12 h Between 12 and 24 h >24 h Previous ulcer disease—no.9) 19 (8.7 to 9. Table 1. (%) Risk factors for bleeding peptic ulcer—no.0) 50 (21. PPI = proton pump inhibitor. 3 an adherent clot.1) in the standard regimen experienced rebleeding (P = 0.1) 55 (23.5) 76 (32.

61 10 (4.76.9 and 11. CI 3.7 111 (47. HR: 2. There were no differences in rebleeding rates between the two treatment groups for each of the endoscopic stigmata of recent hemorrhage.7 ± 2.6) No.5) 6 (2.7) 8 (3.4 <5 days—no.5) 0. respectively (P = 0.8) Rebleeding according to Forrest classification of endoscopic stigmata of hemorrhage—no.3) Other causes 2 (0. Clinical Outcomes for the 47 Patients Included in the Analysis Characteristic Intensive Regimen (N = 238) Rebleeding—no. the independent predictive factors for rebleeding were a Rockall score ≥ 6 at presentation (B = 0.5 ± 2. (%) 1 a—spurting 1 b—oozing 11 (4.88.4) 0. (%) ≤24 h 13 (5. The proportion of patients who required two or more units of blood transfused were not statistically different between treatment groups (P = 0. repeat endoscopic hemostasis stopped bleeding (P = 0. In Cox regression analysis. bleeding recurred in 29 patients (14.14.6) Rebleeding according to ulcer location: no.1) Rebleeding according to time from index endoscopy—no. we found no evidence that an intensive highdose PPI regimen reduced rebleeding compared with a lessintensive regimen of bolus injections: bleeding recurred in . HR: 1.8) Standard Regimen (N = 236) P Value 19 (8.1) Gastric 11 (4. of units of blood transfused Mean ± SD 1. SE = 0. SD = standard deviation.8) 0.80). and for patients with baseline Rockall score >6 versus <6.1) 0.31. and for ulcer size smaller or larger than 2 cm (P = 0.7%.6 days in the two treatment groups. P = 0.18 5 (2.9 ± 4.7%] with the standard regimen). The mean (±SD) number of units of blood transfused was 1. (%) Omeprazole 17 (7. 95% CI 1. (%) Unimodal 13 (5. Three patients (1. (%) <6 points 17 (7.31.31).13). respectively.007). In patients with rebleeding (25 of 28 patients [89.3%] with the intensive regimen and in 18 of 19 patients [94.7) 8 (3. (%) 28 (11.1) 0. Among patients with active bleeding.3) Death—no.32).1) 1 (0.6 ± 3.Proton-Pump Inhibitors After Endoscopic Hemostasis 3015 Table 2. P = 0.49.1%) died during the hospital stay.1 in the intensive regimen group and 1.00 PPI = proton pump inhibitor. Five patients in each group (2.3 ± 4.03).3) 2 (0. (%) 116 (48.3) Rebleeding according to PPI administered: no.0) Surgery—no. SE = 0. (%) 88 (37.2 ± 5. as compared with 18 of 269 patients (6.3) Rebleeding according to Rockall score: no.5) Multimodal 15 (6. 95% CI 1.21 11 (4.3%) and one patient (0. In 205 patients with actively bleeding ulcers.62 3 (1. CI 9.0).84 13 (5.5 days for 98 patients in the standard group (P = 0.63.03 0.4%) in the intensive and standard groups.75 97 (41.2–19.8) 1.64).017) and active index bleeding (B = 0.1) ≥6 points 11 (4.14 11 (4. In patients with in-hospital bleeding.5 ± 2.6) 2 a—nonbleeding visible vessel 7 (2. (P = 0.84 1.34). A higher percentage of patients in the intensive regimen group had a prolonged (>5 days) hospitalization (P = 0.5) >24 h to 7 days 15 (6.8) with nonbleeding lesions (P = 0.4) 0. for patients who received endoscopic monotherapy versus those who underwent multimodal intervention.1 ± 4.84). underwent surgery because of failure to achieve hemostasis at repeat endoscopy (P = 0.4) 0.1 0.4) 11 (4.043).4) 5 (2.1 ± 5.2) 9 (3.1%.1) 8 (3.0) 1 (0.09 6. (%) 3 (1.70). the mean hospital stay was 10.98. (%) Duodenal 17 (7.15 – 3.1 Patients with ≥2 units of blood transfused no.7) 0. the cause of death was linked to the bleeding event in three patients in each group.7 ± 2. as compared with 7.9) 2 b—adherent clot 5 (2.1 in the standard regimen group (P = 0.7) Hospital stay—days Mean ± SD 7.1) Pantoprazole 11 (4.68 8 (3.36 0.8 days.7–9. the mean hospital stay for 107 patients in the intensive regimen group was 6.4) 5 (2.6) Rebleeding according to endoscopic heemostasis: no.02 – 3.8) 0. (%) Because of rebleeding 3 (1. DISCUSSION In patients with ulcer bleeding successfully treated at endoscopy.

or dualendoscopic treatment. with a corresponding 3. However. or death rate. to be costeffective the extra cost of the medication must be offset by a reduction in the occurrence of important adverse clinical outcomes. or death. However. transfusion requirements. however. 14).7 to 9. therefore precluding the evaluation of its impact on the therapeutic outcomes of the study. such an event is unlikely when considering that bleeding recurred predominantly within the first 72 h either in this study as well as in previous reports (6. 17). Although it is possible that some patients may have experienced rebleeding after discharge. who are typically Helicoacter pylori negative and rapid PPI metabolizers. A final shortcoming might be reporting only in-hospital events and excluding data on 30-day outcomes. . are in keeping with the 13. the currently accepted high-dose PPI regimen maintained intragastric pH above 6 for only 30% of the time (24. need for surgery. Findings from this investigation indicate that we may not need intragastric pH maintained at <6 at all. need for surgery.1%) difference in proportions. pylori-infected patients may respond differently to PPIs therapy than noninfected patients (30).8% (CI –1. The rationale for the need for profound gastric acid suppression in peptic ulcer bleeding is extrapolated largely from in vitro evidences that an acidic environment impairs platelet function and clot stabilization (3. but are at odds with the Lin report (15). Pharmacologic data indicate a class effect of PPI therapy as inhibitors of gastric secretion. is more effective in patients who have received concomitant endoscopic treatment. patients who received the high-dose PPI regimen had no advantage with respect to transfusion requirement. length of hospital stay.3016 Andriulli et al. 11. As intravenous PPI therapy is expensive. a similar rate of rebleeding has been found in a headto-head comparative trial of omeprazole versus pantoprazole (27). in our vision. Standard PPI dosing has the obvious advantage of reduced cost and was not shown to increase the risk of nonulcer deaths (15. a combination of endoscopic and pharmacologic therapy is the current standard management. but this information could be retrieved only for a third of enrolled patients in our study. Furthermore. it is not always clear whether subtle variations in the pharmacokinetics and pharmacodynamics of individual PPIs are necessarily of clinical importance (26). Indeed. reported in this study. as it better approximates the translation of results into real-life clinical practice than the outcomes emerging from a structured trial. two trials reported opposite results with the intensive PPI dosing versus placebo (5. in patients with bleeding peptic ulcers with successful endoscopic hemostasis the high-dose intensive PPI regimen had no advantage with respect to in-hospital rates of rebleeding rates. we feel confident the number of missed negative outcomes was very low. 25). Our negative results are in keeping with negative data from four previous trials that compared the two PPI strategies (7–10). As for the claim of a better outcome after dual endoscopic therapies as opposed to monotherapy with epinephrine injection (28. a single report showed a benefit of highdose PPI therapy in comparison to H2R antagonists (19–21). the divergent results between our study and previous trials are unlikely to reflect major differences in patients’ populations or endoscopic interventions. and four studies did not show a significant difference between an intensive regimen and standard bolus PPI injections (7–10). we cannot endorse the reccommendation issued by two consensus statements on the routine use of the intensive PPI regimen for peptic ulcer bleeding (2. 31). in healthy non-Asian volunteers. Therefore.2% probability of rebleeding reported in a pooled analysis of data from eight trials on high-dose PPI infusion (13). respectively. oral (22) or intravenous (23). Indeed. In conclusion. STUDY HIGHLIGHTS What Is Current Knowledge r For patients with bleeding peptic ulcers and major endoscopic stigmata of recent hemorrhage.1% rebleeding rates with intensive or standard regimens. 4). 17) or H2-receptor antagonists (18). Two consensus documents have endorsed a high-dose PPI regimen (80 mg stat followed by an infusion of 8 mg/h for 72 h). To the ongoing debate. it has been shown that H. cost-effectiveness analyses have yielded contrasting results regarding which of the two PPIs administration strategies. We acknowledge that in this study the type of PPI and the modality of endoscopic intervention used has varied across participating centres. 6). The statistically equal rebleeding rate between the two intravenous PPI strategies is of relevance considering that patients treated with bolus PPIs presented more frequently with shock and signs of hemodynamic instability. Reasons for the discrepancy might be revealed by carefully inspecting the comparator used in previous trials: whenever endoscopic hemostasis was not applied. The 11. Moreover. we add our finding of similar rebleeding rates after hemostasis being achieved with either mono. have not affected the overall validity of the findings.0% of patients. In our study. Consequently.8% and 8. intravenous PPI therapy reduced the rate of rebleeding when compared with placebo (16. in patients who underwent simultaneous endoscopic hemostasis. length of hospital stay. Considering the mean duration of hospital stay of our patients. published data are not consistent. the difference in practice patterns at the different participating hospitals might constitute a particular strength of this work. however. 29). two predictors of worse outcome (12. These in vitro data have been taken as supportive evidences for the in vivo clinical benefit.8% and 8. For an investigator-driven study these methodological shortcomings may be considered as inevitable and. the optimal regimen for administration of proton pump inhibitors (PPIs) remains controversial. thus raising the possibility that we may not be achieving the theoretically desirable pH even with intensive therapy. 3).

25. Muttillo IA. Amer F. lansoprazole. et al. Peitz U. Schaffalitzky de Muckadell UB. 27. et al. Rotondano G. Received November 9. Intravenous omeprazole and pantoprazole after endoscopic treatment of bleeding peptic ulcers. et al. Lundell L. Consensus recommendations for managing patients with nonvariceal upper gastrointestinal bleeding. Lo WC. et al.158:54–8. Pannewick U.20:195–202. accepted May 5. N Engl J Med 2000. et al. Lin HJ. Effects of omeprazole on the outcome of endoscopically treated peptic ulcers. Treatment with proton pump inhibitors in acute non-variceal upper gastrointestinal bleeding: A meta-analysis.57: AB150. Efficacy of primed infusions with high dose ranitidine and omeprazole to maintain high intragastric pH in patients with peptic ulcer bleeding: A prospective randomised controlled study. 2008.36:1332–8. et al. Bajaj JS. “Casa Sollievo della Sofferenza” Hospital. supporting the use of a low-dose regimen for its cost-saving potential. et al. Effect of acid and pepsin on blood coagulation and platelet aggregation. Logan RF. Sandoz M. et al. A possible contributor prolonged gastroduodenal mucosal hemorrhage. Comparison of intravenous pantoprazole with intravenous ranitidine in peptic ulcer bleeding. Gut 2002. 14. Lind T. 20. 6.40:36– 41. Am J Gastroenterol 1999. Bolus intravenous omeprazole b.Proton-Pump Inhibitors After Endoscopic Hemostasis What Is New Here r We report the results of a multicentre randomized controlled trial comparing an intensive high-dose acid-inhibiting regimen versus a low-dose standard regimen in the management of high-risk peptic ulcerrelated bleeding after performance of appropriate endoscopic hemostasis: no significant difference in outcomes was registered between the two regimens. Netzer P. 16.D. Non-variceal upper gastrointestinal hemorrhage: Guidelines. Zancanella L. Gut 1999. Curtis LE. Lansoprazole regimens that sustain intragastric pH >6: AN evaluation of intermittent oral and continuous intravenous lansoprazole or ranitidine. Preliminary results. Hunt B. Clin Gastroenterol Hepatol 2006. continuous intravenous omeprazole infusion combined with endoscopic hemostasis in the treatment of peptic ulcer bleeding. Rotondo. Prospective validation of the Baylor bleeding score for predicting the likelihood of rebleeding after endoscopic hemostasis of peptic ulcers. 21. Lazaraki GI. 8. 10. Udd M. Dua KS. Dokas SM. et al. Arch Intern Med 1998. Labenz J.1:CD0020947. Comparison of two doses of intravenous pantoprazole in peptic ulcer bleeding. J Gastroenterol Hepatol 2005. 18. Adam V. Marmo R.32:328–33. Proton-pump inhibitors and outcome of endoscopic hemostasis in bleeding 17. pantoprazole. Effect of intravenous omeprazole on recurrent bleeding after endoscopic treatment of bleeding peptic ulcers. Hepps KS. Ramirez FC. Khuroo MS. 19. et al.139:843–57. Prospective. Aliment Pharmacol Ther 1999.40: 52–62. Palmu A.i. et al. 12. Hasselgren G. Continuous intravenous infusion of omeprazole in elderly patients with peptic ulcer bleeding: Results of a placebo controlled multicenter study. et al. Andriulli A. 15. Piscopo R. Farahat KL. Gastroenterology 1999. Leusing C. Hanson K. Am J Gastroenterol 2006. Barkun AN. Lim BS. 2007. 5. Spiegel BM.. Marshall JK. Ahrens H. Am J Gastroenterol 2005. et al. Dual therapy versus monotherapy in the endoscopic treatment of .20:11–25.52:2190–4. Lee KKC.51(Suppl 4):iv1–6. Cardi M. Piazzi L. Regular-dose versus high-dose omeprazole in peptic ulcer bleeding: A prospective randomized double-blind study. The cost-effectiveness and budget impact of intravenous versus oral proton pump inhibitors in peptic ulcer hemorrhage.343:310– 6. Lin HJ.74:38–43. Caruso N. Gastrointest Endosc 2003.d vs. Kagevi IE. Gut 1989. Ann Chir 1997. Beglinger C. Chen YC. A prospective randomised comparative trial showing tat omeprazole prevents rebleeding in patients with bleeding peptic ulcer after endoscopic therapy. Saeed ZA. J Am Pharm Assoc 2000. 22. REFERENCES 1. Green FW Jr. Gut 1997. et al. Welage LS. Leontiadis GI. Role of intravenous omeprazole in patients with high-risk of peptic ulcer bleeding after successful endoscopic epinephrine injection: A prospective randomized comparative trial. Meier R.38:316–21.45(Suppl V):A100. 13. Barkun A.30:1074–7. randomized trial comparing effect of oral versus intravenous pantoprazole on rebleeding after nonvariceal upper gastrointestinal bleeding: A pilot study. 9. 4.. Sung JJY. 28. Fried R. Gut 1996. for the Nonvariceal Upper GI Bleeding Consensus Conference Group. 7. Gut 2004. Miettinen P.101:500–5. 2. Om´eprazole versus ranitidine intraveineux dans le traitment de l’ulc`ere duod´enal h´emorragique: Une e´ tude randomis´ee prospective.23:985–95. 23. Dulai GS. The cost-effectiveness of high-dose oral proton pump inhibition after endoscopy in the acute treatment of peptic ulcer bleeding. These data demonstrate that 3-day PPI infusion can be replaced with a once-daily intravenous bolus.4:988–97. 3017 peptic ulcer: A series of meta-analysis. Reprint requests and correspondence: Angelo Andriulli. Scand J Gastroenterol 1997. Schonekas H. et al. Italy. et al. Proton pump inhibitor treatment for acute peptic ulcer bleeding. Enright H. 3. Division of Gastroenterology.41:561–5. Amadori L. Chilovi F. et al. Scand J Gastroenterol 1997. Kaplan MM. Afdhal N. Patchett SE. 71013. Lo WC.116:A305.32:320–7.53(Suppl VI):A290. Sharma VK. Gastrointest Endosc 1995. 26. Gastroenterology 1978. Lau JYW. et al. Harling H. Kontoninas Z. et al. Howden CW. Berardi RR.94:351– 7. Evaluation of omeprazole. Effect of repeated injection and continuous infusion of omeprazole and ranitidine on intragastric pH over 72 h. San Giovanni. M. Rockall TA. Bardou M. IRCCS.51:136–9. Devlin HB. Clot lysis by gastric juice: An in vitro study. Risk assessment after acute upper gastrointestinal haemorrhage. Am Intern Med 2003. Scand J Gastroenterol 2001. Aliment Pharmacol Ther 2004. et al. et al. and rabeprazole in the treatment of acid-related diseases. Gaia C. Herba K. Annese V. 24. British Society of Gastroenterology Endoscopy Committee.100:207–19. Dig Dis Sci 2007. Metz DC. Havelund T. Lee FY. Cochrane Database Syst Rev 2006. et al. Perng CL. 11.

Nazario Caruso agreed on the final version of the protocol of the study. enrolled patients to the study. et al.D. Antonietta Garritoli. Am J Gastroenterol 2007. Alessandro Caroli. Gianni Imperiali. discussed it with the Ethics committee of the coordinating centre. Specific author contributions: Angelo Andriulli designed the study. Silvano Loperfido discussed and amended the original draft of the protocol. and agreed on the final version of the manuscript. Gian Battista Forte. et al. Rosaria Focareta. Pietro Leo. Longoa MG. Continuous infusion of pantoprazole versus ranitidine for prevention of ulcer rebleeding: A U. 30.36:666–70. Giovanni Aragona. Pace SC. . Paolo Tonti. and wrote the manuscript. Fabio Monica. Helicobacter pylori infection and the prevention of peptic ulcer with proton pump inhibitors in elderly subjects taking low-dose aspirin. Pilotto A. 31. Jensen DM. M. Dig Liver Dis 2004. wrote the original draft of the protocol. Soffer E. Am J Gastroenterol 2006. high-risk bleeding ulcers: A meta-analysis of controlled trial. collected data from contributing investigators. discussed the protocol with their local Ethics committees.102:279–89. and reviewed the final version of the manuscript. Financial support: None. multicenter randomized. Angelo Belmonte. 60:497–504. ASGE guideline: The role of endoscopy in acute nonvariceal upper-GI hemorrhage. Sergio Peyre. Francesco Perri ran the statistical analysis of experimental data.101:1991–9. enrolled patients to the study. CONFLICT OF INTEREST Guarantor of the article: Angelo Andriulli. Potential competing interests: None. double-blind study. acted as the coordinator of the entire study. Franceschi M. Fabio Fornari. Riccardo Marmo. Antonio Merla enrolled patients to the study.3018 Andriulli et al. supervised the statistical analysis of trial data. and managed the electronic database. presented and discussed the study with the local Ethics committee. Fabrizio Forte. and members of the 315 Study Group.S.. Gastrointest Endosc 2004. Antonio Spadaccini. 29.