The intestine has long been hypothesized as "the engine" of critical illness, but its clinical importance needs to be better defined. The gut microbiome is severely altered in multiple disease states, including critical illness, where the health-inducing microbiome becomes a disease-promoting pathobiome that makes the patient more vulnerable to nosocomial infections, sepsis, and multiple organ failure [20]. So far, only a few studies have analyzed the gut microbiome in critically ill patients, and they have confirmed a state of dysbiosis [21,22,23]. Moreover, recent studies in intensive care unit (ICU) patients observed a gradual worsening of the dysbiosis during their stay in the ICU [22,24,25,26]. The most relevant changes in the microbiome can be seen in the largest study to date that examined the sequencing of the 16S rRNA gene from multiple body sites (skin, oral, and feces) from 115 ICU patients and compared it with 1242 healthy volunteers [22]. At the intestinal level there was a low prevalence of the Firmicutes and Bacteroidetes phyla, and a greater richness of Proteobacteria in comparison to healthy individuals. At the genus level, there was a lower prevalence of key commensal genera (such as Faecalibacterium—an anti-inflammatory organism, Blautia, and Ruminococcus), and in some cases, an overgrowth (over 50% relative abundance) of genera with pathogenic properties, such as Escherichia/Shigella, Salmonella, Enterococcus, Clostridium difficile, or Staphylococcus [21,22,23,24,27]. It has been proposed that changes in the Firmicutes/Bacteroidetes ratio can predict patient outcome [24], although further work is required to validate these findings.

Overall, critically ill patients admitted to the ICU present a gut microbiome characterized by lower bacterial diversity and large inter-individual variation. A study of 14 ICU patients also reported the emergence of ultra-low-diversity communities in 35% of patients who only presented one to four bacterial taxa [27]. In previous studies, low microbial diversity has been associated with an increased risk of mortality [28,29], and the domination of certain pathogens have been identified as an independent risk factor for adverse outcomes [30,31,32].

Considering all the studies on critically ill patients, Proteobacteria is the dominant phyla, and Firmicutes is reduced, whereas Enterococcus, Staphylococcus, and Enterobacter are increased in septic patients. In these patients with sepsis, the focus often lies on the identification of a single pathogen as the causative agent. However, there is an increasing belief that most infections have “polymicrobial” phenotypes that depend on the microbiome status of the patient. Thus, the initial state of the microbiome can determine both the susceptibility to infection [33] and its severity [34]. The composition and functions of the intestinal microbiome of critically ill patients and healthy humans are summarized in Figure 1.

Furthermore, migration of microorganisms between the intestinal and pulmonary microbiome has been reported in critically ill patients. A recently published study highlights the impact of the gut microbiome on the pulmonary microbiome. It was observed that the pulmonary microbiome in both murine sepsis and human acute respiratory distress syndrome (ARDS) was enriched with bacteria associated with the intestine. An operative taxonomic unit of Bacteroides was detected in the bronchoalveolar fluid (BAL) samples from 41% of patients with ARDS compared to 3% in healthy patients. Moreover, the systemic and alveolar levels of tumor necrosis factor (TNF)-α in patients with ARDS were markedly increased by the presence of organisms derived from the intestine in the BAL. However, the precise route by which the intestinal microorganisms reached the lungs of the mice with sepsis has not been identified [35].

The alteration of the microbiome in critically ill patients is multicausal. Critical disease leads to profound modifications in the gut microbiome, caused by general alterations in the host environment [36,37] including enhanced virulence of the bacteria due to the expression of ancestral or newly acquired genes [38].

In addition, many treatments administered to patients in the ICU, like antibiotics, proton pump inhibitors, vasopressors, and opioids, produce harmful effects outside their target organ, which directly affect the microbiome. The most significant alterations are probably related to antibiotic treatments since they indiscriminately ablate the commensal microbiome, favoring the intrusion of secondary pathogenic microorganisms and the enrichment of antibiotic resistance genes [39]. Thus, antibiotic therapies may aggravate the alteration of the microbiome caused by the different pathologies. In fact, the use of antibiotics in ICUs is very frequent, with 71% of the patients receiving antibiotic treatment, according to data from the Centers for Disease Control and Prevention in the United Sates [40]; although it is estimated that 35% of the antibiotic regimens are unnecessary, according to the latest recommendations. As a result, the increase in mortality and morbidity associated with these alterations leads to an additional increase in the cost and care related to the ICU.

Besides, the inappropriate use of antibiotics is considered to be responsible for the increasing emergence of multidrug-resistant bacteria (MDR), and it is important to consider that nosocomial infections represent an additional complication in critically ill patients. The incidence of MDR infections is escalating rapidly all over the world [41]. In fact, a recent publication calculated that infections with antibiotic-resistant Clostridium difficile occur in more than 450,000 patients per year, in the United States [41]. In addition, MDR infections are increasingly lethal for hospitalized patients. It is calculated that C. difficile contributes to more than 30,000 deaths per year in the United States [41,42]. Consequently, the implementation of new antibiotic drugs has not significantly improved the survival to infectious diseases in developed countries, but has instead contributed to the emergence of increasingly aggressive MDR organisms.

Moreover, nutrition is another key factor for the gut microbiome homeostasis since it primarily depends on the availability of enteral nutrients for survival. Thus, the nutritional components (carbohydrates, lipids, and proteins) and the route of administration (enteral/parenteral) might also alter the health of the microbiome [43,44,45]. In addition, pharmacological interventions can modify the specific conditions of the body site (for example, skin decontamination with chlorhexidine) and invasive procedures may alter the natural barrier mechanisms (e.g., endotracheal intubation, intravascular catheters) that could facilitate the access and proliferation of microbes [46]. The factors that may alter the microbiome are shown in Figure 2. Therefore, the impact of ICU care on the microbiome should be further explored.

Critical illness induces hyper-permeability of the gut barrier that begins as early as one hour after the onset of sepsis or trauma and lasts for at least 48 h [47,48]. Mucus also plays a crucial role in the defence of the host by preventing bacteria and digestive enzymes from contacting the intestinal epithelium as a result of mucus hydrophobia that reduces the absorption of toxic molecules. In critical disease, the mucus layer is affected, which induces the dysfunction of epithelial cells. Actually, in these patients, it is very common to find a reduced intestinal reperfusion that can lessen the hydrophobicity of the mucus layer and alter the intestinal permeability [49].

One of the mechanisms responsible for the epithelial defects in critically ill patients could be the impairment of short chain fatty acids (SCFA) production. One important metabolic function of the gut microbiome is the fermentation of dietary fiber and production of SCFA, including butyrate, which serves as the primary energy source for the colonic epithelium and preserves its integrity [50]. During sepsis, a rapid and persistent fall in the concentration of SCFA takes place [51] and as a consequence, the mucosal epithelial barrier is impaired due to epithelial apoptosis resulting in poor absorption of nutrients, diarrhea, loss of fecal energy, and pathogen translocation [22,23]. However, it has been described in a graft versus host disease mouse model that when bacterial strains, which are capable of producing large amounts of SCFA, are ingested, the severity of the disease decreases. This is explained by the high intestinal concentrations of butyrate that improve the epithelial barrier by enhancing intercellular junctions and decreasing cell apoptosis [52].

Considering the microbiome as an internalized organ with important physiological functions, it is evident that its alteration might be as harmful as other “organ failures” in ICU patients. The possible damage could be caused by both the loss of “organ” function and also the aberrant physiology replacing its function. In this context, the “lost organ” is the commensal microbial community that helps to metabolize drugs, nutrients, and hormones, modulate immune responses, and maintain the mucosal barrier homeostasis. By losing commensal microbes, the host also loses protection against invading pathogens by different mechanisms. The gut microbiome is the main activator of the host immunity against infections, which involves innate (stimulation of granulopoiesis, production of antimicrobial peptides (bacteriocins), and nutrient source competition) [39,50] and adaptive (regulation and differentiation of Th17 cells) mechanisms [53]. The “aberrant physiology” is represented by emerging pathogens that dominate microbial communities and cause dysregulated inflammatory responses, excessive inflammation leading to multiple organ dysfunction, and eventually, immune depletion due to the loss of specific microbial signals necessary for the maintenance of normal T cell function in the gut, that could facilitate the emergence of super-infections [54] and ultimately sepsis [55].

Hence, recent preclinical data derived from animal models suggest that the intestinal microbiome plays a protective role in the host defence against sepsis [34,58,59]. In murine models of Gram-positive and Gram-negative pneumosepsis, it has been shown that antibiotics can induce the disruption of the gut microbiome, which increases inflammation and bacterial spread [34,59]. Similarly, data from ICU patients indicate that the loss of microbiome diversity implies an increased length of stay in the ICU, which further highlights the potential clinical relevance of the intestinal microbiome for critically ill patients [22].