Multidrug-resistant organism colonization in critically ill cirrhotic patients: Marker or mediator of mortality?

Abstract

Bacterial infections are a key precipitant of acute decompensation and acute-on-chronic liver failure in cirrhotic patients. The rising prevalence of multidrug-resistant organisms complicates intensive care unit management, making colonization screening increasingly important. In this issue, Kosuta et al report that one-third of cirrhotic intensive care unit patients were colonized with multidrug-resistant organisms, with an 82% concordance between colonizing and infecting strains. Yet colonization did not independently predict infection or short-term mortality, which were instead driven by the severity of organ dysfunction. These findings highlight host vulnerability as the main determinant of mortality, while reinforcing colonization’s role in guiding empiric therapy and regional stewardship strategies.

Key Words: Multidrug-resistant organisms; Cirrhosis; Acute-on-chronic liver failure; Intensive care unit; Mortality

Core Tip: Multidrug-resistant organism colonization is frequent in critically ill cirrhotic patients and strongly predicts subsequent infection. However, colonization alone does not determine mortality, which is primarily driven by organ dysfunction severity. Colonization status remains an important tool for guiding empiric antimicrobial therapy and tailoring stewardship strategies, particularly in regions with distinct epidemiological patterns.

TO THE EDITOR

Bacterial infections are a major precipitant of acute decompensation and acute-on-chronic liver failure (ACLF) in cirrhotic patients[1], particularly in the intensive care unit (ICU). While timely recognition and prompt antimicrobial therapy remain central to improving survival, the growing prevalence of multidrug-resistant organisms (MDROs) complicates empiric management and raises concerns for transplant eligibility and long-term outcomes[2]. Screening for gastrointestinal colonization at admission has become a common practice, based on evidence that colonization often precedes infection[3]. However, whether colonization itself drives poor prognosis or simply reflects an already fragile clinical state remains debated.

Kosuta et al[4] report important findings from a Southeastern European ICU setting characterized by a high burden of carbapenem-resistant organisms, particularly Acinetobacter baumannii and OXA-48-producing Klebsiella pneumoniae. Among 107 cirrhotic patients admitted to the ICU, nearly one-third were colonized with MDROs at baseline. Strikingly, when infection occurred, concordance between colonizing and infecting strains reached 82%, highlighting colonization as a reservoir for subsequent invasive disease. Yet colonization did not independently predict either infection risk or short-term mortality; instead, mortality was primarily determined by the degree of organ dysfunction, as reflected in sequential organ failure assessment and chronic liver failure consortium-ACLF scores[5].

This disconnect between colonization and outcomes merits closer consideration. Colonization is a dynamic interplay between microbial pressure, mucosal barrier integrity, and host immunity[6]. In cirrhosis, small intestinal bacterial overgrowth, increased intestinal permeability, and impaired reticuloendothelial clearance[7] create a permissive environment for microbial translocation. MDRO colonization amplifies this risk by introducing pathogens with enhanced virulence and resistance traits. However, the transition from colonization to infection depends critically on host factors[8]: Immune paralysis, neutrophil dysfunction, complement deficiency, and the extent of organ failure. The findings of Kosuta et al[4], showing that colonization alone did not dictate mortality, reinforce the notion that in cirrhosis, the host milieu—rather than microbial carriage—shapes the clinical trajectory.

Still, the high concordance between colonizing and infecting strains has important clinical implications. Even if colonization does not independently predict mortality, it remains a valuable guide for empiric antimicrobial therapy when infection is suspected, particularly in settings with unique epidemiological patterns. The predominance of Acinetobacter baumannii and OXA-48 Klebsiella pneumoniae in this cohort contrasts with Western Europe, where extended-spectrum β-lactamase-producing Escherichia coli and vancomycin-resistant enterococci are more prevalent[9,10]. This variability argues strongly for region-specific colonization surveillance to inform empiric antibiotic policies and stewardship strategies.

The study by Kosuta et al[4] also raises practical questions. Should screening be repeated beyond ICU admission to capture dynamic changes in microbial carriage? Could targeting high-risk patients, such as those with ACLF, improve predictive value? Might molecular typing distinguish transient carriage from clinically significant colonization[11], refining both infection-control strategies and empiric treatment algorithms?

From a therapeutic perspective, colonization data can help tailor empiric antibiotics, reducing treatment failure and unnecessary broad-spectrum use, thereby supporting antimicrobial stewardship[12]. Screening could also guide infection-control measures, such as isolation or cohorting, especially in ICUs with high MDRO prevalence. In transplant candidates, colonization status may influence perioperative prophylaxis and immunosuppressive strategies. The potential for “decolonization” strategies—whether selective digestive decontamination, probiotics, or faecal microbiota transplantation[13]—warrants exploration, though careful assessment of risks, including further resistance, is essential in this vulnerable population.

Several limitations of the study deserve emphasis. Colonization was assessed only at ICU admission, whereas serial screening could identify new acquisitions under broad-spectrum antibiotic pressure. Key confounders such as prior antibiotic exposure, invasive device use, and infection-control practices were not fully accounted for. Molecular typing was not performed, so while phenotypic concordance was high, genetic relatedness could not be confirmed. Finally, the study focused on short-term mortality; longer-term outcomes such as transplant-free survival, recurrent infections, or quality of life remain unexplored.

Future research should adopt a multipronged approach. Prospective longitudinal studies are needed to assess MDRO colonization dynamics. Incorporating molecular epidemiology, including whole-genome sequencing, would clarify whether infections arise from colonizing strains or exogenous acquisition. Mechanistic studies should examine how colonization interacts with cirrhosis-associated immune dysfunction[14], including neutrophil oxidative burst, monocyte cytokine production, and T-cell exhaustion. Colonization status should be evaluated alongside clinical severity indices to determine additive or synergistic prognostic value[15].

In summary, Kosuta et al[4] provide valuable evidence from a region with high carbapenem resistance, showing that MDRO colonization is common and highly concordant with infecting strains but does not independently drive short-term mortality in critically ill cirrhotic patients. Colonization remains clinically informative for guiding empiric therapy but must be interpreted within the context of organ failure and immune dysfunction.

The challenge is to move beyond descriptive associations. Multicenter collaborations[16] should prioritize longitudinal surveillance, mechanistic exploration, and interventional trials to determine whether colonization can be modified, monitored, or integrated into personalized management strategies. Only then will we know whether colonization is a passive marker or an actionable mediator in the complex interplay between cirrhosis, infection, and mortality.