2. Gut Microbiota Dysbiosis in Neonates Receiving PN
Over the last several decades, progress in perinatal and postnatal care have increased the survival of neonates, although morbidity later in life has increased
[19]. Most investigative efforts have focused specifically on extreme preterm infants, given that around 50% of these neonates present neurologic and pulmonary complications, a three-fold increased risk of developing chronic kidney disease
[20] and a low birth weight, as well as growth failure or postnatal growth restriction
[21]. This raises the need for carefully designed early nutritional support, in terms of the optimal nutrient intake and the route of administration, to ensure a normal growth and a healthy development in preterm neonates, thus preventing early malnutrition-related adverse psychomotor and mental disorders later in life
[22]. Optimal nutritional support is also mandatory in critically ill neonates who are admitted to the NICU due to these patients having limited macronutrient stores and relatively higher energy requirements
[23]. In this sense, enteral nutrition (EN) is generally preferred for its additional physiological contribution in infant development and lower related complications
[24,25][24][25]. Nevertheless, this type of nutritional support is not sufficient to cover the preterm infant´s needs due to their gastrointestinal tract immaturity and critically ill conditions. Consequently, TPN is often initiated to supplement the insufficient EN. It is well established that early- and long-term TPN in preterm and critically ill neonates show health benefits on the survival rates, an optimal weight gain in the NICU and the improvement of long-term neurodevelopment and motor development
[26,27,28][26][27][28]; however, the specific health conditions of these patients make them more susceptible to neonatal morbidities, including bronchopulmonary dysplasia, late-onset sepsis, NEC-associated intestinal failure or the rapid onset of PNALD
[1,28][1][28]. Thus, a balanced PN with an early and “aggressive” approach, either as a transition to or in combination with EN, must be used to limit the growth retardation and its related long-term consequences
[29].
Recently, there is growing interest to evaluate the potential adverse effects of TPN on the gut microbiota’s composition and function, which is highly compromised in both critically ill and preterm infants who are usually exposed to aggressive treatments, the NICU environment or antibiotics, and how these changes may affect the development and progression of TPN-related comorbidities. Thus, for example, evidence suggest that preterm neonates with an increased risk of PNALD usually show a structural and functional gut microbiota dysbiosis and a subsequent potential “gut–brain axis” malfunction, immune system alterations and the development of non-communicable diseases during childhood and adult life
[1]. In general, the studies mostly carried out in animal models support that PN dramatically changes the gut microbiota’s structure, with low abundances of
Firmicutes and a high prevalence of
Bacteroidetes and
Proteobacteria phylum, as well as
Actinobacteria phylum and
Akkermansia muciniphila, but to a lesser extent
[30,31][30][31]. Unfortunately, these changes adversely affect gastrointestinal health. On the one hand, it is well known that
Bacteroidetes phylum promotes an intestinal inflammation and increases the intestinal permeability, which can drive a bacterial flux across the mucosa and result in a cytokines-mediated hepatocellular injury
[32,33][32][33]. On the other hand, unlike
Firmicutes phylum,
Proteobacteria can metabolize the host-derived substrates in the absence of enteral feeding and incorporated them into gut microbial organisms, including
Enterobacteriaceae of the
Proteobacteria phylum, thus increasing its starvation resistance
[34]. It is also important to highlight that PN increases the growth of opportunistic pathogens including
E. coli,
Salmonella,
Yersinia,
Helicobacter and
Vibrio [35[35][36],
36], and decreases the abundance of commensal microbials, such as
Bacteroides fragilis [37]. All of these mentioned microbial compositional changes occur along with the PN-associated adverse effects on the gastrointestinal immunity as well as the cellular and chemical barriers, which in turn exacerbate intestinal failure and comorbidities during long-term PN
[15].
In the light of these findings, there is no doubt that the understanding of the mechanisms and process derived from PN in critically ill neonates still raises many challenges and unique considerations. In fact, although the current guidelines support the safe use of this feeding route, improvements in the PN formulations, the timing of initiation, the advancement of nutritional support and clear individualized goals are still needed. For instance, a high risk of infection during TPN is considered one of the central current and future challenges in pediatric clinical research. In this regard, a higher risk of nosocomial infection noted among long-term TPN patients involves the need of prolonged antibiotic use, which has also been identified as a key factor in the gut microbiota’s modulation. Thus, recent data support that long-term antibiotic therapy profoundly decreased the relative abundance of potential probiotic candidates such as
Lactobacillus and
Enterococcus in those preterm patients receiving PN support
[18]. These findings thereby support the need to take into account the duration of antibiotic therapy in the development of the optimal strategies for improving the gut microbiota’s composition. Consequently, the published guidelines by the Centers for Disease Control and Prevention
[38] recommend extreme caution when performing a peripheral insertion to prevent intravascular catheter-related infections. Unfortunately, both the efficacy of these recommendations and the need for sterile barrier precautions during the subsequent changing of PN bags have yet to be thoroughly researched
[39]. In this regard, specific strategies, such as the changing of PN bags every 48 h with the maximal sterile barrier precautions, seem to reduce the risk of bacteremia and mortality in preterm infants
[40], but further randomized and controlled trials involving unmeasured or unknown confounders are still needed to verify its effectiveness. On the other hand, although its safety has been previously tested, another key challenge is to identify the most effective and useful probiotic strain in the prevention of severe NEC, late-onset sepsis and all-cause mortality in preterm neonates. This knowledge undoubtedly is of vital importance to better understand the exact mechanisms of action involved in the health beneficial effects of probiotics
[41]. In fact, in a recent network meta-analysis of fifty-one randomized controlled trials (RCTs) involving 11,231 preterm infants, the overlap of strains with an effective result on multiple domains was not found, highlighting once again the need for more large and adequately powered RCTs aimed to evaluate the optimal probiotic-based treatment strategies
[42].
3. Influence of Gastrointestinal Diseases in Infants and Children Receiving PN on Gut Microbiome: Potential Use of Pre-, Pro- or Postbiotics Therapies
3.1. Inflammatory Bowel Disease
The term inflammatory bowel disease (IBD) implies various chronic and relapsing inflammatory intestinal disorders with a low mortality, such as ulcerative colitis (UC), Crohn’s disease (CD) and IBD-unclassified (IBD-U), that primarily affect the small intestine and colon, although these disorders clearly differ in the location and severity of the lesion
[43]. Both its incidence and prevalence are growing globally, and they are expected to continue increasing over the next few years, particularly in industrialized countries
[44]. Previous epidemiological studies have also reported that about 25% of patients with IBD have their first symptoms in childhood and, subsequently, IBD incidence is greater in the pediatric population than the adult population
[45,46][45][46]. As a result, IBD poses a major challenge for health care systems that are unable to deal with a staggering increase in the burden of this disease
[47]. From an etiological point of view, IBD is defined as a multifactorial inflammatory or inflammation-associated disease that involves a complex interaction between genetic predisposition and immunological abnormalities, the gut microbiota and the environmental influences, although neither factor in itself is sufficient for IBD development
[48,49][48][49]. Among these environmental factors, several perinatal (prenatal diseases, smoking during pregnancy and maternal age) and postnatal exposures (domestic hygiene, an urban environment, a diet high in proteins and total fats, infections and the abuse of antibiotics) have been clearly associated with IBD development
[50]. However, the potential role of other perinatal factors such as prematurity and their potential relationship with various confounders in IBD development later in life remains unclear
[51,52,53][51][52][53]. Recently, special interest has focused on the potential role of medical nutrition as a risk factor in patients with active IBD. The current ESPEN practical guidelines about clinical nutrition in IBD recommends EN based on formulas or liquids as a supportive therapy when oral feeding is not sufficient, while PN is only indicated in patients with advanced-stage and complicated disease
[54]. This is of particular importance for preterm infants in which, due to their intolerance to enteral food, PN could be implanted shortly after birth to preserve the metabolic and hemodynamic stability. Nevertheless, studies with animals receiving PN showed a reduced gut growth, villous height, mucosal mass, protein mass, cell proliferation and mucosal immunity
[55]. Furthermore, these deficiencies are implicated in the development of intestinal permeability, a bacterial translocation and a high risk of sepsis
[56]. Despite the fact that these events may compromise the integrity of the gut in the neonate
[57], the PN effects on IBD development remain unclear, and further well-designed clinical studies in humans are still needed
[58].
It is well established that environmental, genetic and immune factors can directly or indirectly lead to gut microbiota dysbiosis, which has been proposed as the key risk factor for IBD development in pediatric patients and adults
[59]. In this regard, the results obtained from large human cohort studies indicate that commensal bacteria from the
Firmicutes and
Bacteriodetes phyla as well a bacterial species from the genera
Bacteroides,
Lactobacillus, Eubacterium,
Faecalibacterium and
Roseburia are generally decreased in IBD patients. Conversely, these patients show a relative increase in the bacteria belonging to the phylum
Proteobacteria (mainly
Escherichia coli,
Enterobacteriaceae,
Klebsiella and
Proteus spp.) and
Fusobacterium [48,59,60][48][59][60]. In preterm infants, their poor somatic growth and subsequent need for PN support may further exacerbate the mentioned changes in the gut microbiota’s composition
[61], but scarce information is available about this topic. Theoretically, the gut microbiota of PN-receiving patients should be characterized by the lower abundance of commensal bacteria (mainly
Bacteriodes and
Bifidobacterium) as well as the increased prevalence of potentially pathogenic Gram-negative bacteria
[30,31,62][30][31][62]. These changes not only impair a healthy gut colonization, but also could interact with the host epigenome in order to predispose a gut infection and the high risk of diseases, including IBD
[63,64][63][64]. Likewise, taking into account that the gut microbiota interacts with the host through metabolites, there is a growing interest to better understand the potential role of these signals in IBD development as well as their influence on immune maturation and homeostasis, the host energy metabolism and the maintenance of mucosal integrity
[60]. In fact, patients with IBD presented alterations in their metabolite profiles as well as a perturbated interaction between the diet and gut microbiota. Moreover, specific classes of metabolites, particularly bile acids (BAs), lower levels of SCFAs (mainly acetate, butyrate and propionate) as well as the disruption of the tryptophan metabolism have been implicated in the pathogenesis of IBD
[60,65,66][60][65][66]. Finally, it is important to note that these changes in the microbiota’s composition and gut metabolome involve a significant impairment in the host immune response, mucosal homeostasis and the energy metabolism, which further increases both the incidence and severity of disease in both adults and pediatric patients
[59,60][59][60].
3.2. Necrotizing Enterocolitis (NEC)
Necrotizing enterocolitis (NEC) is one of the most frequent and fatal intestinal disorders in preterm infants which is characterized by variable intestinal injury, from epithelial injury to transmural involvement and perforation, accompanied by intestinal inflammation and often bacterial invasion
[89][67]. While its incidence is extremely rare in term infants, and usually related to congenital anomalies, sepsis or hypotension, NEC affects about 5–12% of neonates born at a very-low birth weight (VLBW; <1500 g)
[90][68], with an associated mortality rate of nearly 30–50% for preterm infants of an extremely low weight (<1000 g) and 10–30% for VLBW neonates
[91][69]. Its pathogenesis has been charged to a multifactorial origin marked by a low gestational age and weight at birth, formula feeding and intestinal dysbiosis, although maternal factors such as chorioamnionitis, a high BMI, preeclampsia, or smoking during pregnancy also seem to be involved in NEC development
[92,93,94][70][71][72].
Recently, there is growing evidence supporting the significant relationship between NEC onset and progression and gut microbiota dysbiosis. In this regard, gut microbiota from preterm neonates suffering from NEC is characterized by a low bacteria diversity and commensal bacteria abundance, as well as an overgrowth of pathogenic bacteria causing concomitant infections
[95][73]. However, due to heterogenicity in molecular methods used for identification and detection, no common microbial pattern has been consistently identified. While Mai et al.
[96][74] reported a high abundance of
γ-Proteobacteria, Normann et al.
[97][75] found an increased abundance of
Bacillales and
Enterobacteriaceae in the stool samples obtained from preterm infants who suffer from NEC. Moreover, patterns of gut microbiota dysbiosis seem to change according to the time of the NEC onset, with a high
Firmicutes and
Clostridia abundance reported in its early onset in contrast to the predominance of
Entorobacteriaceae,
Escherichia/Shigella and
Cronobacter in those cases of later onset NEC
[98,99][76][77]. It is also important to note that NEC-related gut microbiota dysbiosis may occur several weeks prior to the onset of the disease, suggesting a time frame in which gut microbiota-targeted therapy could positively influence the clinical outcomes
[100][78].
In addition to these changes in the gut microbiota’s composition, the studies carried out to date suggest a potential involvement of specific genetic variants regulators in NEC pathogenesis, including the nuclear factor κB1 (NF- κB1), the co-receptor molecule lymphocyte antigen 96 (MD-2 co-receptor), the small glycolipid transport protein ganglioside GM2 activator and the interleukin (IL)-1 related receptor (IL-1R). These genetic variations are related to the upregulation of the Toll-like receptor-4 (TLR-4)-dependent signaling pathway, thus increasing the intestinal inflammatory response
[101,102][79][80]. Interestingly, TLR4 activation also leads to the impairment of the epithelial barrier and a subsequent luminal bacterial translocation, which results in the recognition of gut the microbiota by TLR-4 expressed in mesenteric blood vessels, favoring vasoconstriction, intestinal ischemia and NEC
[103][81]. All of these changes, along with the presence of unusual intestinal microbial species and the overall reduction in the gut microbiota’s community diversity, may explain why preterm neonates who develop NEC also have a high susceptibility to infectious diseases
[104][82].
As mentioned above, both the mode of feeding (enteral versus parenteral feeding) received by preterm infants and how the transition between both of the feeding routes occurs, also seems to be involved in pathogenesis and the clinical outcomes of NEC. Thus, studies performed in animal models suggest that long-term TPN may predispose to TLR-4-dependent NEC lesions
[105][83]. Likewise, clinical strategies based on acute enteral refeeding seem not to be effective in preventing small intestinal mucosa homeostasis, including intestinal epithelial cell apoptosis, the loss of the epithelial barrier function and the failure of the leucine rich repeat-containing G protein-coupled receptor 5-positive stem cell expression
[106][84]. Overall, these findings also provide evidence showing that that a switch from parenteral to enteral nutrition may rapidly induce diet-dependent histopathological, functional and proinflammatory insults to the immature intestine. Consequently, special attention should be given to the speed of feeding progression, although the results achieved to date are contradictory. In fact, Roze et al.
[107][85] observed that a higher speed of feeding progression is not a risk factor for NEC but relates rather to a shorter time with PN. Conversely, Ou et al.
[108][86] found no effects of faster advancing feeds on late-onset incidence. Moreover, although PN is initiated in
nil per os patients (or “nothing to mouth”) following the NEC diagnosis
[108][86], it is also important to highlight that this feeding route at the NECs onset seems not to improve the clinical outcomes (the rates of surgical intervention or in-hospital mortality) in those who are premature with a low birth weight
[109][87].
3.3. Parenteral Nutrition-Associated Liver Disease (PNALD)
It is well established that long-term PN (>27 days) causes a complex wide spectrum of liver function alterations, commonly named PNALD, which is also referred to as PN-associated cholestasis (PNAC), PN-associated liver injury (PNALI) or, more recently, intestinal failure-associated liver disease (IFALD). Among these concepts, both PNALD and PNAC usually refer to liver disease related to the potential toxic compounds present in PN, while IFALD is normally used to specify the hepatobiliary dysfunction caused by intestinal failure
[122][88].
PNALD/IFALD are clinically manifest with intrahepatic cholestasis (conjugated bilirubin levels > 2 mg/dL) in the absence of any other liver etiology, hepato-steatosis and altered biochemical markers of liver damage. These hepatic complications can lead to fibrosis and cirrhosis in those cases with prolonged PN, which can variably progress to end-stage liver disease and thus requiring a liver transplantation, or death
[123,124][89][90]. Its reported incidence varies considerably depending on the diagnostic criteria used and the age groups, with a higher incidence in infants and child patients (25–60%) compared to adults (15–40%). Moreover, its incidence is particularly common in LBW premature infants with long-term PN (>85%)
[125,126][91][92]. Consequently, research efforts have primarily focused on elucidating the complex pathophysiological mechanisms involved in PNALD/IFALD and its most effective treatment. In this sense, PNALD/IFALD has a multifactorial origin involving both nutrition-, patient- and nutrition-related risk factors
[1,125][1][91]. Among the latter, the immaturity of the liver function in preterm and LBW neonates, there is a high risk of recurrent bacterial infections and NEC observed on these patients, as well as SBS and the associated intestinal comorbidities, have been identified as potent risk factors for PNALD/IFALD development
[124,127,128,129][90][93][94][95].
Regarding the nutrition-related risk factors, the inability to successfully implement enteral nutrition in preterm and critically ill infants, and consequent long-term PN, have been established as key factors for PNALD/IFALD development
[123,127,129][89][93][95]. In fact, the lack of EN impairs a gastrointestinal hormones secretion (gastrin, motilin, secretin and glucagon, among others), thus leading to important abnormalities in intestinal motility, gallbladder contractility, enterohepatic circulation and bile acids secretion/absorption, all of which potentially increase the risk of cholestasis and subsequent PNALD/IFALD
[124,127,130][90][93][96]. On the other hand, prolonged PN also adversely impacts the hepatobiliary system as a direct consequence of the immaturity of organ systems and their resulting inability to detoxify certain toxic minerals (mainly aluminum, copper and manganese) present in parenteral products, causing or aggravating cholestasis
[123,129][89][95]. However, other components and nutritional features of PN have been also implicated in mediating PNALD/IFALD pathogenesis, including an excessive calorie intake, a high protein content (>2.5 g/kg/day) as well as certain amino acid deficiencies (taurine, choline and glutamine) in PN solutions
[123,129,131][89][95][97]. Nevertheless, both sources and amounts of intravenous lipid emulsions (ILEs) have acquired the greatest interest as key factors in PNALD/IFALD pathogenesis
[124,127][90][93]. Compared to fish oil-based ILEs (FO-ILEs), traditional soybean oil-based ILEs (SO-ILEs) are strongly discouraged due to the hepatotoxic effects caused by its high abundance in phytosterols, plant-based cholesterol-like compounds and pro-inflammatory omega-6 polyunsaturated fatty acids (ω-6 PUFAs). Moreover, SO-ILEs also contain relatively low amounts of the antioxidant α-tocopherol as well as anti-inflammatory ω-3 PUFAs (mainly docosahexaenoic (DHA) and eicosapentaenoic (EPA) acids)
[125,126,129][91][92][95]. Due to these nutritional characteristics, long-term PN programs using SO-ILEs have been associated with altered bile acid homeostasis, reduced cholesterol synthesis and bile flow. All of these conditions ultimately promote hepatic and liver inflammation via macrophage-derived IL-1β/NF-kB signaling, as well as cholestasis, in both pediatric patients
[132,133][98][99] and murine models of human IFALD
[134,135][100][101].
3.4. Gut Mucosal Atrophy
It is well known that gut mucosal has the ability to respond to a wide range of internal and external environmental stimuli through diverse physiological, cellular and molecular mechanisms controlling its morphology and function. Therefore, mucosal adaption is critical to gut homeostasis and the subsequent host health
[156][102]. However, specific pathological or nutritional conditions (the absence of enteral nutrition as well as long-term periods of starvation or parenteral nutrition) induce gut mucosal atrophy. This condition is mainly characterized by a marked decrease in the intestinal function and profound morphological changes in terms of a decreased villous height, crypt depth, surface area and epithelial cell numbers
[157][103]. In this regard, the results obtained from animal models support an association between TPN and mucosal atrophy, even if this route of feeding is properly provided as a life-support system for neonates, infants and children with gastrointestinal disorders
[158][104]. Overall, TPN-associated mucosal atrophy is mainly caused by nutrients deprivation in the luminal content and subsequent mucosal hypoplasia via the TNF-α/EGF signaling pathway; nevertheless, intestinal barrier dysfunction is also involved through different mechanisms such as an altered peristaltic compression and villus motility, a decreased enterocytic proliferation/differentiation and an increased enterocyte apoptosis. Taken together, these changes result in a loss of the overall barrier function and subsequent bacterial translocation
[158][104]. Moreover, there is also growing evidence showing that gut mucosal atrophy is driven by TPN-related gut microbiota dysbiosis, characterized by a decreased α-diversity, a lower abundance of
Firmicutes as well as an increased prevalence of potentially pathogenic Gram-negative bacteria, mainly belonging to
Proteobacteria phylum
[1,16,159,160,161][1][16][105][106][107]. Interestingly, the studies carried out to date also suggest that these changes in bacterial diversity and richness are positively related to the longer duration of parenteral nutrition and its related consequences, such as the lack of fermentable fiber and the depletion of beneficial SCFAs, further increasing the abundance of potentially harmful bacteria
[16,159,161,162][16][105][107][108]. This TPN-related shift in the gut microbiota’s composition is strongly suspected to trigger a TLR-dependent proinflammatory response in the gut with the consequence being a loss in the epithelial integrity, thus causing morphological alterations, and a loss in the barrier function
[1].
Significant efforts have been made to improve or prevent TPN-associated intestinal mucosal atrophy. In this sense, the studies performed in animal models of gut mucosal atrophy reported that the use of different growth and stimulation factors, including the epithelial growth factor
[162][108], glucagon-like-protein-2
[163][109], hepatocyte growth factor
[164][110], ghrelin
[165][111], glutamate
[166][112], arginine
[167][113] or PUFA emulsions
[168,169][114][115], can improve intestinal development and its function, thus preventing TPN-associated intestinal mucosal atrophy. Interestingly, prebiotic-, probiotic- and postbiotic-based treatments have been recently used in order to achieve this purpose. Thus, the experimental studies conducted in both rodents and piglets that received PN supplemented with SCFAs, mainly butyrate, reported lower rates of infection associated with an improved mucosal immunity
[170][116]. Likewise, the therapeutic use of PN enriched with butyric acid is also supported by its moderate but positive effects on the recovery of intestinal mucosa
[171,172][117][118] and mucosal protein synthesis
[173][119]. Moreover, using a piglet model of intestinal failure, treatment based on partial EN supplemented with short-chain fructo-oligosaccharides (scFOS), was more effective than a probiotic treatment with
Lactobacillus rhamnosus GG for an intestinal adaptation
[174][120]. Although these results are promising, studies on TPN-associated gut mucosal atrophy and the potential use of pre- and probiotic therapy in humans are very limited, particularly in children and even more in premature ones. This may be due to the fact that those children who require TPN also have a high incidence of gastrointestinal disorders. Therefore, it is certainly difficult to discriminate whether gut mucosal atrophy is the cause or consequence of the type of nutrition received or the existing gastrointestinal disease. Consequently, further studies are needed to clarify the potential relationship between TPN, gut mucosal atrophy and gut microbiota dysbiosis in humans, which will open up new clinical and therapeutic avenues based on the use of pre- and probiotics.