Re-EValuating the Inhibition of Stress Erosions (REVISE) Trial (REVISE)

Re-EValuating the Inhibition of Stress Erosions: Prophylaxis Against Gastrointestinal Bleeding in the Critically Ill (The REVISE) Trial

Canadian Institutes of Health Research (CIHR), Canadian Critical Care Trials Group, Australian and New Zealand Intensive Care Society Clinical Trials Group, National Health and Medical Research Council, Australia

Patients who are critically ill in the in the Intensive Care Unit (ICU), especially those who need a breathing machine, can develop ulcers in the stomach that bleed. To prevent bleeding, many such patients around the world receive a drug called pantoprazole that decreases acid production. However, today, compared to decades ago, critically ill patients rarely develop gastrointestinal bleeding. This decrease is likely due to modern medicine, better resuscitation and earlier feeding. There may also be harms associated with pantoprazole and other drugs that reduce acid levels in the stomach including lung infections (pneumonia) and bowel infections (clostridium difficile). Studies in this area are old and of modest quality. Therefore, it is difficult to know whether pantoprazole does decrease stomach bleeding these days, or whether the possible harms of lung and bowel infections are actually more common and more serious problems. The goal of this international study is to determine if, in critically ill patients using breathing machines, the use of pantoprazole is effective in preventing bleeding from stomach ulcers or whether it causes more problems such as lung infection (pneumonia) and bowel infection (Clostridium difficile), or whether pantoprazole has no effect at all. Whether the harms are worth the benefits, and whether the benefits are worth the costs, will be determined by an economic analysis to inform patients, families, clinicians, and healthcare systems globally.

Background: For 40 years, pharmacologic prevention of stress ulcer-related gastrointestinal (GI) bleeding with acid suppression has been the standard of care for mechanically ventilated ICU patients. Worldwide, proton pump inhibitors (PPIs) are more commonly used than histamine-2-receptor antagonists. Observational studies and the latest network meta-analysis suggest that PPIs increase the risk of ventilator-associated pneumonia (VAP) and Clostridioides difficile infection (CDI). However, a recent large randomized trial showed that pantoprazole had no impact on the primary outcome of 90-day mortality. The secondary outcome (a composite of pneumonia, gastrointestinal bleeding, CDI and acute myocardial ischemia) was not different between the groups. Although pantoprazole was associated with a significantly lower rate of GI bleeding, some bleeding events required no blood transfusion, endoscopy or other diagnostic or therapeutic interventions, calling into question whether these bleeds were truly patient-important. Further, patients with high illness severity on pantoprazole had a significant, unexplainable higher risk of death than those receiving placebo. REVISE Pilot Trial: We completed the 91-patient REVISE Pilot Trial in Canada, Australia and Saudi Arabia, demonstrating a high consent rate (77.8%); recruitment rate (2.6 patients/month/center); and protocol adherence (96.8%), thereby successfully establishing the feasibility of a larger REVISE Trial. Objectives of the REVISE Trial: To determine, among mechanically ventilated patients, the effect of pantoprazole versus placebo on the primary efficacy outcome of clinically important upper GI bleeding, and the primary safety outcome of 90-day mortality. Secondary outcomes are VAP, CDI, acute kidney injury, hospital mortality and patient-important GI bleeding. Tertiary outcomes are transfused packed red blood cells, serum creatinine, duration of mechanical ventilation, duration of ICU stay and duration of hospital stay. Methods: We will include 4,800 ICU patients >18 years old who have an anticipated duration of mechanical ventilation of ≥48 hours. Exclusion criteria are acute or recent GI bleeding, dual antiplatelet therapy, combined antiplatelet and anticoagulant therapy, hopeless prognosis or intent to withdraw advanced life support, and previous enrolment in this or a confounding trial. Patients will be randomized in a fixed 1:1 allocation, stratified by center and pre-ICU acid suppression ('start or no start' strata, and 'continue or discontinue' strata). Research Coordinators will obtain informed consent using a deferred or a priori consent model. Study Pharmacists will obtain concealed allocation from the REVISE website; all research team and clinical team members, patients and families will be blinded. Patients will receive pantoprazole 40 mg or identical placebo intravenously daily while in ICU up to 90 days or until: 1) successful discontinuation of mechanical ventilation for >48 hours; 2) development of clinically important GI bleeding, or 3) death in ICU. Analyses will be by intention-to-treat and per protocol. Collaborations: REVISE will be conducted in collaboration with the Canadian Critical Care Trials Group, the Australian and New Zealand Critical Care Trials Group, other consortia and interested investigators, under the auspices of the International Forum for Acute Care Trialists. Ethical Imperative: Many factors converge to underscore the ethical imperative for REVISE. Critical care has evolved, research standards have improved, epidemiology may have changed, the risk:benefit and cost:benefit ratios of prophylaxis may have shifted. Thus, stress ulcer prophylaxis may need to be REVISED. Relevance: Most mechanically ventilated patients around the world receive daily stress ulcer prophylaxis, although variation exists such that some centers do not use any. Many RCTs of stress ulcer prophylaxis were conducted 10-30 years ago, several are at moderate or high risk of bias, and cointerventions may not reflect current critical care. A recent large trial raised concerns about death associated with pantoprazole in the most severely ill patient subgroup. Although GI bleeding events were less frequent, they may not have been patient-important, and there were no other benefits observed. Consensus in the scientific and clinical community is that equipoise remains regarding whether routine use of pantoprazole should continue for stress ulcer prophylaxis during critical illness. The question is especially important for patients who are receiving acid suppression pre-ICU, those who are receiving enteral nutrition, and those at high risk of death. Today, infectious complications of PPIs have emerged as potentially more common and serious than upper GI bleeding. The number needed to prophylax to prevent 1 GI bleed and the cost per GI bleed averted may be very high; furthermore, the number needed to harm to cause 1 episode of VAP or CDI may be low. Recent practice guidelines are conflicting. Remaining doubts about the effectiveness and safety of PPIs urge re-examination of universal prophylaxis for possible de-adoption. Aligned with the 'Choosing Wisely' Campaign, REVISE and the companion economic evaluation (E-REVISE) will be incorporated into guidelines to inform global practice.

A prospective, international, multicentre, parallel group, concealed, blinded, randomized trial in critically ill mechanically ventilated adults

Clinically important GI bleeding requires the presence of overt GI bleeding which is defined as one of the following; Hematemesis Overt nasogastric bleeding Melena Hematochezia PLUS (in the absence of another cause), at least one of the following in the 24 hours following overt GI bleeding: Haemodynamic change defined as a spontaneous decrease in invasively monitored mean arterial pressure or non-invasive systolic or diastolic blood pressure of 20 mmHg or more or an orthostatic increase in pulse rate of 20 beats/minute and a decrease in systolic blood pressure of 10 mmHg, with or without vasopressor initiation, or increase Vasopressor initiation A decrease in haemoglobin of ≥ 20 g/l in a 24-hour period or less, Transfusion of ≥2 units of packed red blood cells within 24 hours of bleeding to maintain stable haemoglobin or haemodynamics, or Need for therapeutic intervention (e.g. angiography, surgery or endoscopic treatment of bleeding).

Diagnostic criteria for VAP include: previous mechanical ventilation for at least 48 hours, a new, progressive or persistent radiographic infiltrate on chest X-ray (without other obvious cause) plus at least 2 of the following 4 features: fever or hypothermia (temperature >38 °C or <36 °C) relative leukopenia or leukocytosis (WBC<4.0 or >12 x 10^9/L) purulent sputum gas exchange deterioration

We will use clinical features (diarrhea, ileus, toxic megacolon) and either microbiological evidence of toxin producing Clostridium difficile or Pseudomembranous colitis on colonoscopy.

Rate of patient important GI bleeding in ICU or resulting in ICU readmission, censored at 90 days after randomization

Patient important GI bleeding in ICU or resulting in ICU readmission, censored at 90 days after randomization

Inclusion Criteria: Age 18 years or more. Receiving invasive mechanical ventilation in an ICU and in the opinion of the treating ICU physician mechanical ventilation will not be discontinued before the end of the day after tomorrow. Exclusion Criteria: The treating clinician considers either Pantoprazole or placebo are indicated or contraindicated for this patient. Pantoprazole contraindicated for patient due to local product information; Australia/New Zealand; being treated with HIV protease inhibitors atazanavir or nelfinavir being treated with high dose methotrexate (i.e., greater than 300 mg as part of a chemotherapy regimen). documented cirrhosis or severe liver disease (for example as indicated by an INR greater than 5.0 due to underlying liver disease). Canada; being treated with rilpivirine or atazanavir patients who are hypersensitive to pantoprazole, substituted benzimidazoles, or to any ingredient in the formulation Patients in whom a PPI or histamine 2 receptor antagonist (H2RA) is indicated due to active bleeding or increased bleeding risk, defined as patients with acute GI bleeding, severe oesophagitis or peptic ulcer disease within the previous 8 weeks, Zollinger Ellison syndrome, Barrett's oesophagus or any previous admission to hospital because of upper GI bleeding (patients receiving PPIs for mild dyspepsia or mild gastroesophageal reflux disease or an uncertain indication are not excluded). Received invasive mechanical ventilation during this ICU admission for 72 hours or more. Patients who have received more than 24 hours treatment (i.e., more than one daily dose equivalent) with a PPI or H2RA during this ICU admission. Being treated with or need for dual anti-platelet therapy. Admitted for palliative care or the ICU physician is not committed to continuing life-sustaining therapies at the time of enrolment. Known or suspected pregnancy. Physician, patient, or substitute decision maker (SDM) declines. Previously enrolled in the REVISE trial Enrolled in another trial for which co-enrolment is not approved.

Following the publication of REVISE, the dataset will be used for secondary observational studies addressing additional hypothesis-driven questions (e.g., predictors of gastrointestinal bleeding). Access by REVISE investigators will follow a submitted rationale, analysis plan and approval by the Management Committee. Requests for access to the dataset by external investigators will be considered following a submitted rationale, analysis plan and approval by the Management Committee and research ethics boards as relevant. Requirements will be stipulated in a pre-specified data sharing agreement. Only de-identified data will be provided and will be transferred via a secure web portal.

Following the publication of REVISE, the dataset will be used for secondary observational studies addressing additional hypothesis-driven questions (e.g., predictors of gastrointestinal bleeding). Access by REVISE investigators will follow a submitted rationale, analysis plan and approval by the Management Committee. Requests for access to the dataset by external investigators will be considered following a submitted rationale, analysis plan and approval by the Management Committee and research ethics boards as relevant. Requirements will be stipulated in a pre-specified data sharing agreement. Only de-identified data will be provided and will be transferred via a secure web portal.