Gastrointestinal bleeding in critically ill immunocompromised patients

Jennifer Catano, Sophie‐Caroline Sacleux, Jean‐Marc Gornet, Marine Camus, Naïke Bigé, Faouzi Saliba, Élie Azoulay, Guillaume Dumas, Lara ZafraniAnnals 2021 · DOI

Abstract

Mortality is high in immunocompromised patients with gastrointestinal bleeding (GIB) in the ICU, especially in those receiving long-term corticosteroids. However, the mortality of GIB in this population is not significantly different from that of non-immunocompromised patients in the ICU. The prophylactic administration of proton pump inhibitors (PPIs) should be strongly considered in this high-risk population.

Introduction

Acute gastrointestinal bleeding (GIB) is a life-threatening emergency in the Intensive Care Unit (ICU). While management guidelines exist for the general population, immunocompromised patients—including those with hematological malignancies, organ transplants, or HIV—present unique clinical challenges. These patients are often coagulopathic, susceptible to opportunistic infections, and may suffer from direct tumor infiltration of the gastrointestinal tract. Prior to this study, data regarding GIB in this specific demographic were limited to small case reports. This research aims to define the clinical spectrum of GIB in immunocompromised patients, identify predictors of "severe GIB," and determine if immunocompromised status independently worsens prognosis compared to non-immunocompromised ICU patients.

Methods

This retrospective multicenter study analyzed data from three hospitals in France over a ten-year period (2010–2019). The primary cohort consisted of 141 immunocompromised adults admitted to the ICU for GIB at Saint-Louis University Hospital.

Fig. 1 Flow chart of patient selection and outcomes.

To evaluate the impact of immune status on outcomes, researchers performed a 1:1 matched comparison with a control group of 151 non-immunocompromised patients from Paul Brousse and Saint-Antoine University Hospitals. Matching criteria included age, hemoglobin levels, bleeding location, SOFA score (organ failure severity), and prior use of proton pump inhibitors (PPIs). Multivariate logistic regression was employed to isolate risk factors for "severe GIB" (defined by hemorrhagic shock, high lactate, or requirement of >5 units of red blood cells) and 90-day mortality.

Results

Clinical Characteristics and Risk Factors

Upper GIB was significantly more prevalent (73%) than lower GIB (27%). The study identified several independent risk factors for severe bleeding and mortality:

  • Severe GIB Predictors: Male gender (OR 4.48), upper GIB (OR 2.88), and digestive malignant infiltration (OR 5.85). Notably, 26% of patients bled due to direct cancer infiltration (primarily lymphomas) in the gut. This finding is distinct from the general ICU population, where peptic ulcers are the primary cause of GIB.
  • Mortality Predictors: 90-day mortality was independently associated with hemorrhagic shock (OR 2.91), upper GIB (OR 4.33), and long-term corticosteroid therapy (OR 2.98). Conversely, higher albumin levels—a marker of nutritional status and liver function—were protective (OR 0.54 per 5 g/l increase).

The Role of Prophylaxis: The PPI Shield

One of the most clinically actionable findings was the protective effect of Proton Pump Inhibitors (PPIs). Patients receiving PPIs prior to hospitalization had a significantly lower risk of developing severe GIB (OR 0.25). Despite this clear benefit, only approximately half of the high-risk patients in the study were on PPI prophylaxis. This suggests a major gap in preventative care for immunocompromised patients, where early administration could mitigate the severity of potential bleeding events.

Comparative Outcomes: Severity vs. Mortality

When compared to the matched non-immunocompromised control group, immunocompromised patients experienced more severe bleeding episodes (76% vs. 60%, p=0.01). However, ICU mortality rates were comparable between the two groups (23% vs. 25%). This suggests that while the acute bleeding event may be more severe in immunocompromised patients—likely due to factors like coagulopathy or tumor infiltration—overall mortality is driven more by the severity of the acute illness and the underlying disease rather than immune status alone.

Discussion and Conclusion

This study underscores that while immunocompromised patients with GIB face high mortality (38% at 90 days), their prognosis in the ICU is not inherently worse than that of non-immunocompromised patients with similar clinical presentations. The identification of long-term corticosteroid use and malignant infiltration as major risk factors provides critical targets for clinical vigilance.

The significant protective effect of pre-hospital PPI use suggests that broader prophylactic application could improve outcomes. Limitations of this study include its retrospective nature and the inability to compare specific endoscopic intervention techniques. Future research should focus on optimizing PPI dosing and establishing standardized endoscopic protocols for managing tumor-related gastrointestinal bleeding.

Summary updated

Abstract

Mortality is high in immunocompromised patients with GIB in ICU, especially in patients receiving long term corticosteroids. Mortality of GIB is not different from mortality of non-immunocompromised patients in ICU. The prophylactic administration of proton pump inhibitors should be considered in this population.
BibTeX
@article{catano2021gastrointestinal,
  author = {Jennifer Catano and Sophie‐Caroline Sacleux and Jean‐Marc Gornet and Marine Camus and Naïke Bigé and Faouzi Saliba and Élie Azoulay and Guillaume Dumas and Lara Zafrani},
  title = {{Gastrointestinal bleeding in critically ill immunocompromised patients}},
  journal = {Annals},
  year = {2021},
  doi = {10.1186/s13613-021-00913-6},
}
Figure