Slow recruitment

OBJECTIVES: Up to 15% of patients with peptic ulcer bleeding will develop rebleeding, mainly in those with ulcers of higher-risk stigmata (i.e. Forrest class Ia to IIb). Randomized trials show that second-look endoscopy is effective in reducing rebleeding rate. However, whether to withhold aspirin or other anti-platelet agents (for the treatment of established cardiovascular or cerebrovascular diseases) remains controversial. Studies have shown that although continuation of anti-platelet agents reduces mortality rate due to reduced cardiovascular and cerebrovascular events, there is a marginal increase in rebleeding risk.

HYPOTHESIS: We hypothesize that continuation of aspirin or other anti-platelet agents coupled with second-look endoscopy could reduce the rebleeding rate without increasing the risk of thromboembolic events in high-risk patients.

For patients who are found to have high-risk peptic ulcers on endoscopy, they will be treated by endoscopic combination therapy including adrenaline injection and followed by either thermal coagulation or haemoclips to achieve haemostasis. Peptic ulcers of high-risk stigmata are ulcers with active bleeding, non-bleeding visible blood vessels, or adherent blood clots according to Forrest classification (Ia to IIb).

After initial haemostasis for high-risk ulcers, patients who continue to require anti-platelet agents (namely, aspirin, clopidogrel, prasugrel, or ticagrelor) for treatment or prophylaxis of cardiovascular or cerebrovascular diseases will be eligible for inclusion into study.

Eligible patients will be randomly allocated to have a scheduled second-look endoscopy 16-24 hours after initial endoscopy (i.e. the intervention group), or to receive conventional standard care (i.e. the control group). Randomization will be performed in a 1:1 ratio by computer generated random sequences.

For patients who are allocated to second-look endoscopy group, endoscopic retreatment will be performed if the ulcer shows persistence of high-risk stigmata. Retreatment again will be performed by combination of adrenaline injection follows by thermal coagulation or haemoclips. Standard conventional care will be offered to the other group of patients, in which endoscopy is indicated only in case of suspected clinical bleeding as defined in the subsequent section.

Both the intervention and control groups will receive bolus injection of proton-pump inhibitor (esomeprazole 80mg), follow by high dose infusion of 8mg/h continuously for 72 hours after initial endoscopy. Thereafter, both groups will receive oral form of proton-pump inhibitor (esomeprazole 40mg/d)

Rebleeding is suspected clinically by the presence of fresh haematemesis, or haematochezia/melaena after a normal stool, or unstable haemodynamics with systolic blood pressure ≤ 90 mmHg or pulse ≥ 100 beats/min (after ruling out other causes of shock, e.g. cardiogenic or septic shock), or a drop in haemoglobin level by 2 g/dL or more within 24 hours despite transfusion of 2 or more units of blood during the same period. If any one of the features is present, we will perform emergency endoscopy to confirm the diagnosis of recurrent peptic ulcer bleeding by either the persistence of ulcers of high-risk stigma (Forrest I to IIb) or fresh blood in the stomach). Rebleeding is only defined if it is confirmed by the presence of both clinical and endoscopic features. The findings of endoscopic high-risk stigmata on second-look endoscopy alone without the above clinical features would not be labelled as rebleeding.

During hospitalization, all patients will have close monitoring of vital signs including blood pressure, pulse and oxygen saturation. Blood tests will be taken daily for haemoglobin and renal function including urea and creatinine for the first 3 days. A designated team of gastroenterologists will assess the patients twice daily for clinical features of rebleeding, abdominal complications or development of thromboembolic complications. Cardiac enzymes (creatine kinase and troponin T), electrocardiogram and computed tomography scan of the brain will be ordered if there is any clinical suspicion of cardiovascular or cerebrovascular complications.

After observation for more than 72 hours in our medical unit, patients who have successful endoscopic haemostasis without clinical evidence of rebleeding will be either discharged or transferred to convalescent hospitals for further rehabilitation. All patients will be given esomeprazole 40mg/d on discharge and their prior anti-platelet therapies will be resumed. Patient will be contacted by our research staff at day 7 by telephone for any symptoms or discomfort. Patients can also contact us via telephone hotlines for suspected rebleeding or other urgent inquires. They will have follow-up at our specialized outpatient clinic 30 days after the primary endoscopy. Blood tests will be repeated at day 30 and drug compliance of anti-platelet therapies and esomeprazole will be checked by pill counts. If patients default their follow-up, we will contact them via phone call. In addition, we will identify any post-discharge hospitalization within the first 30 days via electronic medical records of the Hospital Authority.

Inclusion Criteria:

• Antiplatelet for both primary and secondary prophylaxis
• Peptic ulcers of Forrest class Ia (spurting of blood), Ib (oozing of blood), IIa (visible vessel) & IIb (adherent clot)

Exclusion Criteria:

• Peptic ulcers of class IIc (pigmented ulcer base) & III (clean ulcer base)
• Unsuccessful endoscopic hemostasis
• Ulcer perforation
• Gastric outlet obstruction precluding passage of scope to D2
• Malignant ulcers
• Proton pump inhibitor (PPI) allergy
• Concomitant anticoagulants
• Pregnancy