June 4, 2019

EJRC - One size does not suit all?

Benefit-risk balance: in patients with high severity disease scores, pantoprazole prophylaxis might be related to poorer outcomes


This subgroup study comes from an international multicentre trial, the SUP-ICU study (Stress Ulcer Prophylaxis in the Intensive Care Unit), in which 3350 patients at risk of gastrointestinal bleeding were randomised with a daily prophylactic dose of 40 mg of pantoprazole versus placebo.

The aim was to evaluate the impact on 90-days mortality, on a composite outcome of clinically important events and serious adverse reactions, and the percentage of days alive without the use of life support.

Results in a subgroup of patients with higher severity disease suggested higher 90-mortality rates when exposed to pantoprazole prophylaxis. These findings were sufficiently alarming to repeat the analyses on this specific population and seek further assessment.

Out of the 1140 patients with a complete SAPS II > 53 at baseline, 580 had been randomised in the pantoprazole group and 560 in the placebo group; 90-day vital status was obtained for 579 patients in the first group and 558 in the control group.

No significant differences were observed between them, besides a higher coagulopathy in the pantoprazole group, but it did not affect the risk estimate of mortality in the sensitivity analysis.

The findings of these post hoc analyses were consistent with the initial trend, confirming higher mortality rate and fewer days alive without the use of life support when exposed to pantoprazole prophylaxis in this population (272 patients [47%] died out of 579 patients of the pantoprazole group, and  229  out of 558[ 41%] in the placebo group [relative risk, 1.13; 95% CI 1.00-1.29])



The main limitation is the potential impact of missing SAPS II data in some patients of the full SUP-ICU trial.  However, the observed relative risk increase can only be partly explained by this imbalance.

The strength of this study, on the other hand, is to be issued from a large, consistent study; all analyses were conducted in accordance with the pre-defined analysis plan of the initial trial population, minimising selection bias and confirming initial findings.



Although in the initial SUP-ICU overall study no difference was observed between the pantoprazole and the placebo group, the findings in a specific population of highly severe patients showed higher mortality rate and lower percentage of days alive without the use of life support, without a particular explanation.


This article review was prepared and submitted by Silvia Calvino Gunther from the N&AHP section, on behalf of Journal Review Club.



1) Marker, S., Perner, A., Wetterslev, J. et al., Pantoprazole prophylaxis in ICU patients with high severity of disease: a post hoc analysis of the placebo-controlled SUP-ICU trial, Intensive Care Med (2019) 45: 609. https://doi.org/10.1007/s00134-019-05589-y

Comment on this news