Intensive Care Med (2019) 45:609–618 https://doi.org/10.1007/s00134-019-05589-y ORIGINAL Pantoprazole prophylaxis in ICU patients with high severity of disease: a post hoc analysis of the placebo-controlled SUP-ICU trial Søren Marker1,2* , Anders Perner1,2, Jørn Wetterslev2,3, Mette Krag1,2, Theis Lange4,5, Matt P. Wise6, Mark Borthwick7, Stepani Bendel8, Frederik Keus9, Anne Berit Guttormsen10,11, Joerg C. Schefold12, Morten Hylander Møller1,2 and The SUP-ICU investigators © 2019 Springer-Verlag GmbH Germany, part of Springer Nature Abstract Purpose: In the subgroup of patients with Simplifed Acute Physiology Score (SAPS) II > 53 in the Stress Ulcer Prophy- laxis in Intensive Care Unit (SUP-ICU) trial, there was interaction (P 0.049) suggesting increased mortality in patients allocated to pantoprazole as compared with placebo. We aimed to= explore this further. Methods: The SUP-ICU trial allocated acutely admitted adults at risk of gastrointestinal bleeding to pantoprazole vs placebo. In this post hoc study, we repeated all the preplanned analyses of SUP-ICU in patients with baseline SAPS II > 53. Results: A total of 1140 patients had a complete SAPS II > 53 and were included. At 90 days, 272/579 patients (47%) assigned to pantoprazole had died, as compared with 229/558 patients (41%) assigned to placebo [relative risk 1.13; 95% confdence interval (CI) 1.00–1.29]. This was supported by sensitivity analyses adjusted for risk factors and those in the per-protocol population. When accounting for patients with incomplete SAPS II in two additional analyses, the relative risk was 1.08; 95% CI 0.96–1.22 and 1.10; 95% CI 0.97–1.25. This was also observed for the secondary outcome days alive without life support. There were no diferences between the intervention groups in the other secondary outcomes. Conclusions: In this post hoc analysis of patients with high disease severity included in the SUP-ICU trial, we observed higher 90-day mortality and fewer days alive without life support with pantoprazole vs placebo. Some of this may have been explained by missing SAPS II data, but further research is needed to draw frm conclusions. ClinicalTrials.gov: ClinicalTrials.gov No. NCT02467621. Keywords: Stress ulcer prophylaxis, Stress ulceration, Intensive care unit, Gastrointestinal bleeding, All-cause mortality *Correspondence: [email protected] 1 Department of Intensive Care 4131, Copenhagen University Hospital, Rigshospitalet, Blegdamsvej 9, 2100 Copenhagen, Denmark Full author information is available at the end of the article 610 Introduction Take‑home message Critically ill patients in the intensive care unit (ICU) are In this post-hoc analysis of a predefned subgroup population in at risk of developing clinically important gastrointestinal the Stress Ulcer Prophylaxis in the Intensive Care Unit (SUP-ICU) trial, (GI) bleeding due to stress ulcers [1–3]. GI bleeding in we observed higher 90-day mortality and fewer days alive without the use of life support in patients with a complete Simplifed Acute critically ill patients has been suggested to be associated Physiology Score (SAPS) II > 53 points allocated to pantoprazole. with adverse outcomes, including a 2–4-fold increased We found no explanations for these observations in baseline data, risk of death and increased length of ICU stay [3]. To process variables or the other trial outcome measures–however, missing SAPS II data for some patients seem to explain a part of the prevent GI bleeding, stress ulcer prophylaxis is currently observed diference. recommended in international guidelines [4] and acid suppressants are frequently used in ICU patients [2, 5]. In the international Stress Ulcer Prophylaxis in the The SUP‑ICU trial Intensive Care Unit (SUP-ICU) trial, 3350 patients at risk Te SUP-ICU trial was a randomised, multicenter, strati- of GI bleeding were randomised to receive either panto- fed, parallel-group, placebo-controlled, blinded clini- prazole or placebo during the ICU stay [1]. We observed cal trial recruiting 3350 patients (3291 were available no diference between the pantoprazole and placebo for data analyses) at risk for GI bleeding in 33 ICUs in groups in the primary outcome measure—90-day mortal- Denmark, Finland, Norway, Switzerland, the Netherlands ity—but a predefned subgroup analysis suggested higher and UK from 4 January 2016 to 22 October 2017 [7]. Eli- 90-day mortality in the pantoprazole group among gible subjects were patients aged 18 years or older, who patients who had higher baseline disease severity, i.e., were acutely admitted to the ICU and had at least one Simplifed Acute Physiology Score (SAPS) II > 53 points risk factor for clinically important GI bleeding, includ- (test of interaction, p value 0.049) [6, 7]. Tis warrants ing any history of liver disease, any history of or ongoing further investigation, because it may be a chance fnding coagulopathy, shock, or use of anticoagulant agents, renal or driven by baseline imbalance or diferences in with- replacement therapy, or mechanical ventilation expected drawals or follow-up in the subgroup, or missing data for to last > 24 h [1]. SAPS II elements in some patients. We excluded patients with GI bleeding during cur- In this post hoc study of the SUP-ICU trial, we per- rent hospital admission, ongoing daily treatment with formed the full set of preplanned outcome analyses in all acid suppressants, patients who were withdrawn from patients with baseline SAPS II > 53 allocated to pantopra- active treatment or brain dead, had organ transplanta- zole or placebo on all trial outcomes to search for expla- tion during current hospital admission, had peptic ulcer nations to substantiate or refute the suggested harm from confrmed by endoscopy or other method during cur- pantoprazole in the most severely ill. We hypothesised rent hospital admission, had a contraindication to panto- that the increased 90-day mortality observed in patients prazole, were pregnant or where consent for enrolment with SAPS II > 53 allocated to pantoprazole would be could not be obtained. Te full defnitions of the inclu- substantiated in sensitivity analyses and in those of the sion and exclusion criteria are presented in the ESM. remaining outcomes. Patients were randomly assigned to daily intravenous pantoprazole 40 mg or a matching placebo during the Methods ICU stay. Randomisation was stratifed according to the Study oversight presence or absence of hematological malignancy and Tis study is a post hoc analysis of the SUP-ICU trial trial site. Te intervention period lasted until discharge [7]. Te trial protocol, statistical analysis plan and pri- from the ICU to a maximum of 90 days. Te protocol was mary results have been published elsewhere [1, 7, 8]. reinitiated in case of readmissions to trial ICUs within Te SUP-ICU trial was approved by all relevant institu- this period. Apart from the use of stress ulcer prophy- tions prior to randomisation of the frst patient. No addi- laxis, all other parts of patient care were at the discretion tional approvals were required for this study. We have of the clinicians. prepared this manuscript according to the Strengthen- At baseline we collected data on demographics and ing the Reporting of Observational Studies in Epidemi- clinical characteristics. Daily recordings during the ology (STROBE) statement [9], and the flled-in checklist entire admission to the ICU included GI bleeding events, is available in the Electronic Supplementary Material infectious adverse events (pneumonia and Clostridium (ESM). difcile infection) and data on the usage of life support 611 modalities. Time of death was registered for all patients computed from odds ratios [8]. In a sensitivity analysis, at a maximum follow-up of 90 days. Observations for we compared the primary outcome in the per-protocol GI bleeding and infectious adverse events stopped at the population (excluding patients having one or more major time of discharge from the ICU. protocol violations—defnition in ESM), and in the origi- Informed consent was obtained prior to randomisation nally pre-specifed subgroups, defned by the presence from all participants or their legal substitutes according or absence of any history of liver disease, the presence to national legislation. or absence of any history of or ongoing coagulopathy, the type of ICU admission (medical versus surgical), Outcomes the presence or absence of shock and the use or not of Te primary outcome was 90-day mortality. Secondary mechanical ventilation [8]. outcomes were the proportion of patients with clinically In the secondary analyses, we compared all dichoto- important events (a composite outcome of clinically sig- mous outcomes using binary logistic regression analyses nifcant GI bleeding, pneumonia, Clostridium difcile adjusted for stratifcation variables and predefned risk infection and myocardial ischemia), clinically important factors at baseline (age, type of admission (medical, elec- GI bleeding, infectious adverse events (pneumonia or tive surgery or emergency surgery) and the Sepsis-related Clostridium difcile infection) and serious adverse reac- Organ Failure Assessment (SOFA) score (Table S8 in the tions (SARs) and the percentage of days alive without the ESM) assessed in the 24 h before randomisation). We use of life support (mechanical ventilation, renal replace- analysed percentages of days alive without life support in ment therapy or inotropes/vasopressor) [1]. the 90-day period by use of the van Elteren test (adjusted for site only), as the assumptions