Intestinal microbiota and blue baby syndrome

Abstract

Necrotizing enterocolitis (NEC) is the most common intestinal emergency among premature infants. Risk factors in premature infants include immature intestinal immunity and an intestinal microbiota dominated by hospital-acquired bacteria. Some probiotics have been shown to decrease the incidence of NEC in premature infants. Among term infants, NEC is rare. However, among term infants with cyanotic congenital heart disease (CCHD), the incidence of NEC is similar to that of premature infants but with even greater mortality rates. Mechanisms by which NEC occurs in term infants with CCHD are unknown. Of central interest is the potential role of changes in the intestinal microbiota and whether these can be modified with probiotic bacteria; accordingly, we review the literature, propose hypotheses and present the rationale for future studies involving preliminary probiotic clinical trials.

Introduction

Necrotizing enterocolitis (NEC) is the most common and destructive intestinal disease in premature infants, but rare and sporadic in term infants. Term infants with cyanotic congenital heart disease (CCHD) have a disturbingly high incidence of NEC. While the pathophysiology of NEC in premature infants has been much studied, processes by which NEC develops in term infants remain unclear. The most promising development in the prevention of NEC in premature infants over the last decade has been the use of probiotics. In this manuscript, we briefly review what is known about intestinal microbiota, NEC and probiotics in premature infants; with this background we then propose hypotheses and present the rationale for future clinical trials of probiotics in term infants with CCHD.

Cyanotic Congenital Heart Disease (CCHD): The ‘Blue Baby Syndrome’ Historical Perspective

In 1944 at the Johns Hopkins Hospital, Blalock and Thomas successfully performed an experimental operation to improve pulmonary blood flow in a cyanotic (blue) baby with tetralogy of Fallot, a common form of CCHD.¹ Their surgical shunt technique consisted of anastomosis of the subclavian artery to the pulmonary artery. This pioneering work paved the way for the development of surgical interventions for other forms of CCHD. Fortunately, over 65 years after that first life-saving operation, pediatric cardiothoracic surgery and cardiopulmonary bypass procedures for the early definitive correction of tetralogy of Fallot have mortality rates of 3% or less and surgeries for all but the most severe forms of CCHD are generally successful.² The inspiring history and evolution of this field have been well summarized.^3,4

Common Forms of CCHD Include:

Left-sided obstructive lesions

• Critical aortic stenosis

• Coarctation of the aorta

• Interrupted aortic arch

Right-sided obstructive lesions

• Tetralogy of Fallot

• Critical pulmonic stenosis

• Ebstein's anomaly

Transposition of the great arteries

Single ventricle physiology

• Hypoplastic left heart

• Double outlet right ventricle

• Tricuspid atresia

• Pulmonary atresia with intact ventricular septum

• Truncus arteriosus

Total anomalous pulmonary venous return

Premature Infants and Necrotizing Enterocolitis

With the development of neonatology and the increased survival of small premature infants, NEC has become the most common gastrointestinal emergency in the premature infant. The incidence of NEC is 3–10% in infants with birth weight less than 1,500 grams.⁵ This amounts to 10,000 infants diagnosed with NEC annually in the United States with peak incidence occurring 2–6 weeks after birth in premature infants.⁶ Death rates for NEC in premature infants vary depending on the level of prematurity and concurrent medical conditions, but several reports range from 9–30%.^7,8 Mortality rates are highest in the most premature infants and in those forms of NEC characterized by full-thickness necrosis and cardiovascular complications.⁹

Previous reviews have detailed the complex pathophysiologic mechanisms of NEC in premature infants.^10,11 In brief, the primary event of NEC in the premature infant appears to involve opportunistic pathogenic bacteria and translocation of bacterial antigens from the intestinal lumen triggering a systemic inflammatory cascade, intestinal necrosis and often cardiovascular instability. The premature infant appears to be at increased risk due to two major factors: (1) gut immaturity—the barrier function and innate immune system of the premature intestines are immature; and (2) gut dysbiosis—the premature infant becomes colonized with an intestinal microbiota that differs markedly from that of the healthy term infant. These differences are due in part to immature intestinal immunity¹² as well as the frequent use of antibiotics, the presence of indwelling feeding tubes, lack of contact with the maternal microbiota and prolonged hospital stays.

Fecal microbiota analyses by conventional culturing techniques demonstrate that the intestinal microbiota of premature infants, compared to healthy term infants, show delayed colonization of beneficial Gram-positive bacteria and increased colonization of pathogenic and potentially pathogenic Gram-negative and Gram-positive bacteria.¹³ A culture-independent 16S ribosomal RNA gene-based study revealed that preterm infants that developed NEC, compared to preterm infants that did not, had less community diversity, greater abundance of Gammaproteobacteria, and had received more days of antibiotics prior to the development of NEC.¹⁴

Clinical staging of NEC is generally based on modified Bell criteria (Fig. 1).¹⁵ NEC is usually managed conservatively with cessation of enteral feedings, continuous gastric suctioning and broad-spectrum intravenous antibiotics, however, 20–40% of infants undergo surgery, most commonly due to evidence of intestinal perforation. These infants often require resection of large segments of necrotic bowel leading to long-term difficulties with feeding intolerance and poor growth and abnormal neurodevelopment.¹⁶ In spite of decades of research, the incidence and mortality rates of NEC in the US have not decreased, due to the rapid onset and progression of this disease and the lack of successful interventions and prevention strategies.^17,18

Figure 1.

Modified Bell criteria for clinical staging of necrotizing enterocolitis (NEC).¹⁵

Several promising prevention strategies have been explored to date in premature infants. Feeding of human milk either from the infant's mother¹⁹ or from screened donors²⁰ decreases the risk of NEC. Enteral antibiotics²¹ and enteral infusions of HCl²² have been shown to prevent NEC, but concerns of safety and feasibility have limited further exploration of these modalities. Epidermal growth factor and heparin-binding epidermal growth factor consistently prevent NEC in animal models²³ but have not been studied in premature infants. Probiotics are the most promising intervention to date in this population.^24–26 Routine use of probiotics for premature infants in the US is not currently recommended due to uncertainty regarding dosage, organism and safety.^27,28

Term Infants, NEC and CCHD

Epidemiology.

NEC among term infants has mostly been reported in case studies and case series from single institutions, therefore determining the incidence in this group is challenging. The most comprehensive estimate comes from a review of 29 cases of NEC in term infants in Australia, representing one case of NEC for every 20,000 live term births.²⁹ Several authors have estimated that 10% of all NEC cases occur in term infants.^30,31

Risk factors. Predisposing factors for NEC in these term infants include CCHD,³² polycythemia treated with partial exchange transfusion,³³ Hirschsprung's disease,³⁴ endocrine disease,²⁹ early onset bacterial sepsis, hypotension, perinatal asphyxia, hypoglycemia, respiratory distress syndrome, protracted diarrhea, pre-eclampsia, cocaine abuse during pregnancy, cows' milk allergy and anti-C Rhesus incompatibility.^35,36 In the two largest case series of term infants with NEC, the number of infants with CCHD and NEC were similar: 10 of 29 infants in Australia²⁹ and 8 of 30 infants in Utah.³⁵

Among term infants with CCHD, the incidence of NEC has been reported to be 3–7%, which is similar to the incidence among premature infants.^32,37 Mortality rates among infants with CCHD who develop NEC, however, are even greater than in premature infants^31,38 with one study reporting mortality as distressingly high as 71%.³⁹ The highest risk seems to be in infants with (1) hypoplastic left heart,⁴⁰ where the heart's left side (aorta, aortic valve, left ventricle and mitral valve) is underdeveloped and (2) truncus arteriosus,⁴¹ which is characterized by a large ventricular septal defect over which a large, single great blood vessel (truncus) gives rise to both the pulmonary arteries and the aorta.³²

Proposed gut pathophysiology.

The exact mechanisms by which NEC develops in infants with CCHD are unknown. In term infants, intestinal function and immunity is not completely mature. In addition, in term infants with CCHD it is likely that both the acute mesenteric blood-flow reduction (hypoperfusion) and chronic cyanosis (hypoxia) may alter the development of the gut microbiota and the integrity of the gut barrier, increasing risk for NEC. Additional factors may include alterations in coagulation, apoptosis, peristalsis and motility, nutrient and water absorptive balances, tissue oxidative metabolism, cellular reduction-oxidation potential and regulation of reactive oxygen species generation and inflammatory responses.^30,42–44 It is likely that, as in premature infants, disruption of the mucin layer and/or the epithelial monolayer separating the intestinal lumen from the lamina propria results in translocation of intestinal microbial products such as lipopolysaccharide (LPS), flagella, peptidoglycans, nucleic acids and other antigens, triggering local and systemic inflammation, intestinal necrosis and cardiovascular instability.^30,44 The effect of persistent T-cell hyperactivation may also contribute to the pathologic outcome as seen in other microbial translocation-based diseases such as HIV/AIDS.⁴⁵

The care of term infants with CCHD may also play a signifi- cant role in the development of NEC in this fragile population including: prolonged hospital stays and exposure to nosocomial microbes, delays in feeding, decreased rates of human milk feeding, poor nutritional status, courses of broad-spectrum antibiotics,¹⁴ and the inflammation and stress associated with cardiopulmonary bypass and cardiac surgery.³⁰

Of central interest is the role of changes in the microbial community structure and function within the gut among these term infants with CCHD and whether any observed changes can be modified with probiotic bacteria.

Animal study insights. Animal models of NEC have predominantly been developed to study the immature gut (simulating the premature infant). In these models, the role of the innate and adaptive immune response in NEC has been well characterized including recognition of bacterial antigens by Toll-like receptors (TLRs), particularly TLR-2 and -4, leading to activation of transcription factors (e.g., NFκB), resulting in increased secretion of pro-inflammatory chemokines and cytokines (e.g., TNFα and numerous interleukins) as well as platelet activating factor.^43,46 It is presumed that similar increases in pro-inflammatory signaling cascades occur in NEC in term infants with CCHD, though this has not been well studied.

The intestines are highly susceptible to hypoperfusion-induced pathologies due to a high critical oxygen requirement and the mucosal micro-circulatory network. An intriguing animal model that sheds some light on NEC in the term infant with CCHD is the adult rat hemorrhagic shock and resuscitation (HS/R) model. This model creates acute micro-circulatory mesenteric hypoperfusion resulting in intestinal mucosal injury. Interestingly, intravenous heparin-binding epidermal growth factor treatment attenuated injury of the intestinal mucosa and increased intestinal restitution. This protection occurs, at least in part, by reversing post-HS/R mesenteric vasoconstriction leading to significantly increased villous micro-circulatory blood flow. Multiple vasodilatory mechanisms may mediate these processes including the release of nitric oxide by endothelial cells and a reduction of the polymorphonuclear leukocyte/endothelial cell interactions associated with the inflammatory response.⁴⁷ Changes in the intestinal microbiota in this model have not been reported. Acute hypoxia appeared to have no effect on microbiota colonization and translocation in the newborn rabbit; however, this study used only aerobic culture techniques.⁴⁸ The impact of chronic hypoxia on the integrity of the mucosal barrier, on gut permeability and microbial translocation, and on the molecular genetic architecture of gut microbial communities, has not been determined.

Intestinal Microbiota and Probiotics: Potential Role in Term Infants with CCHD

Advances in the study of human microbial ecology.

The development of culture-independent methods to identify, examine and explore microbes based on molecular sequence analysis^49–51 has dramatically increased momentum in this area of investigation and indeed made this “an exciting time to study gastrointestinal microbiology.”⁵² The recently published draft of the International MetaHIT (Metagenomics of the Human Intestinal Tract) project created a gene catalogue of the human gut microbiota (the microbiome).⁵³ This is one of many large scale studies of the human microbial ecosystem, including the NIH Human Microbiome Project,⁵⁴ which are defining the role of both individual pathogens⁵⁵ and entire microbial communities¹⁴ in the development of diseases such as NEC.

The gut microbiota of term infants.

The fetal gut is generally believed to be sterile until rupture of the fetal membranes. Microbial colonization of the healthy term newborn gut progresses quickly with the initial wave of colonists arriving predominantly from the maternal genito-urinary tract followed by a more stable population. Mode of delivery and exposure to antibiotics influence the intestinal microbiota briefly, but then a stable fairly simple community is established which is mostly influenced by feeding (breast vs. formula) and other environmental factors. There is significant inter-individual variation at lower-taxonomic levels (genus and species/strain) in the healthy term infant, but at the higher-taxonomic phylum-level, the rank abundance of this community is dominated mostly by members of the phyla Proteobacteria, Firmicutes and Actinobacteria (mixed results) with little from the Bacteroidetes or other phyla. This community does not shift significantly until the introduction of solid foods at 4 to 8 months of age, at which time diversity increases and leads to a gut microbiota composition similar to the adult, dominated by the phyla Firmicutes and Bacteroidetes with a small fraction allotted to the Proteobacteria, Actinobacteria and other phyla, by 2 years of age.

The gut microbiota of term infants with CCHD.

We found no studies that characterize the intestinal microbiota of the term infant with CCHD in either the NIH-National Library of Medicine (Pubmed) database or the ClinicalTrials.gov registry. The human intestinal microbiota is critical to homeostasis and particularly relevant to infants with CCHD due to many co-host functioning processes including: regulation of angiogenesis in the gut microvascular network, maintenance of proper reductive- to oxidative-stress and anti- to pro-inflammatory balances, epithelial turnover and genetic/epigenetic regulation, intestinal pH and motility, synthesis of micronutrients like vitamin K and several of the B vitamins, regulation of bile salt and lipid metabolism and fat deposition, nitrogen metabolism, fermentation of polysaccharides and absorption of digestible microbial products, absorption of electrolytes and trace minerals and bioenergetic/thermodynamic regulation.^56,57 In addition, gut microbial communities can protect the host and the ecosystem's vital diversity through competition with opportunist colonization as well as through interfering with receptor binding and virulence factor production.^58,59

Growth. The specific role of intestinal microbiota in energy regulation has particular relevance in term infants with CCHD, who are at high risk for poor growth. Gut microbes participate in normal digestion and metabolism of nutrients even during periods of nutrient deprivation.⁶⁰ Conventionally reared mice have a 40% higher body fat content than germ-free mice even though the latter consume more food. When the distal gut microbiota from normal mice is transplanted into germ free mice, there is a 60% increase in body fat within 2 weeks without any increase in food consumption or obvious difference in energy expenditure.^61,62 Microbes generate short-chain fatty acids and ferment indigestible polysaccharides to digestible monosaccharides increasing the amount of energy extracted from the diet. Metagenomic studies of genetically obese mice and humans demonstrate correlations between obesity and an ‘energy-harvesting’ gut microbiome composition.⁶³ Conversely dietary manipulation can rapidly alter the gut microbiome composition despite host colonization history⁶⁴ and independent of obesity.⁶⁵ Since term infants with CCHD, like premature infants, are at high risk for poor growth, they require maximal caloric intake to ensure adequate weight and length gains. Furthermore, growth is a significant challenge for infants with CCHD even in the context of very high caloric intake. In a recent study in Egypt, 18 infants with CCHD (ten with tetralogy of Fallot, four with tricuspid atresia and four with pulmonary atresia), in comparison to ageand sex-matched healthy controls, had a mean weight and height at 46 and 81%, respectively, of their counterparts.⁶⁶

Chronic hypoxia also influences growth. In a lamb model of chronic hypoxia, the pulmonary artery was partially occluded and an atrial septostomy performed to create an aortic oxygen saturation of 60–74%. After two weeks of stable chronic hypoxia, somatic growth in this model was decreased to 60% of control lambs and intestinal length and weight were unchanged. Specific activity of the digestive enzyme lactase, the principal disaccharidase of the infant lamb intestine, and total small intestinal contents of lactase, were significantly curtailed in the hypoxemic animals.⁶⁷ The intestinal microbiota was not analyzed in this study. In a mouse model of chronic hypoxia (11% oxygen for 4 weeks), somatic weight gain was decreased to 70% of controls, heart weight increased, but weights of liver, kidney and brain decreased compared to controls. Intestinal weights and lengths and the composition of the intestinal microbiota were not reported.⁶⁸ Similar studies have been performed in rats with similar results.^69,70

Attempts to manipulate the intestinal microbiota with probiotics to improve growth in premature infants have been successful in some populations but not in others.^71,72 Similar studies in term infants with CCHD have not yet been attempted.

Probiotics for term infants with CCHD.

The joint Food and Agriculture Organization of the United Nations/World Health Organization defines probiotics as “live microorganisms, which when administered in adequate amounts, confer a health benefit on the host”.⁷³ These products can be sub-categorized to include dietary supplements, medical foods or designer probiotics that are genetically altered.⁷⁴ Probiotics have been shown to provide diverse human health benefits,^75–78 particularly in the pediatric population.^79–81 Clinical recommendations regarding the use of probiotics in term and pre-term infants remain unclear due to the extensive array of probiotic products and doses administered. Clinical trials in the neonatal and pediatric population have included single organisms and combinations, most commonly lactobacilli, bifidobacteria, streptococci, enterococci, clostridia, species of the genera Bacillus, E. coli and the yeast Saccharomyces boulardii.^82,83

Many probiotic species administered are already inhabitants of the healthy gut and their function when ingested is similar to the homeostatic functions of the ‘normal’ microbiota. However, some probiotic species have originated very specific and unique metabolic capabilities, possibly in co-evolutionary response to persistent selective pressures exerted by host biology and ingested matter. A compelling example in the neonate is the development of specific enzymes (sialidase and fucosidase) in the genome of Bifidobacterium longum ssp. infantis which allow deconstruction and consumption of human milk oligosaccharides.⁸⁴

Potential mechanisms. In vitro and animal studies suggest several mechanisms of action of probiotics: competitive inhibition of pathogen-adhesion receptor-binding sites, growth promotion of partnering commensals and mutualists, induction of transforming growth factor-β and modulation of cellular apoptosis/proliferation, decreased expression of pro-inflammatory cytokines, increased expression of anti-inflammatory cytokines, secretion of enzymes and beneficial metabolites, modulation of cellular redox potential, functional enhancement of epithelial tight junctions, induction of intestinal mucin production, signaling interactions with cell surface molecules invoking cytoprotective and immune responses (with local and systemic effects), increased expression of innate immune antimicrobial peptides, and possibly even evoking neuro-endocrine and hormonal responses involved in growth and development.^{57,78,85–88}

Probiotics and cardiovascular health.

In adults, a potential role of intestinal microbiota and probiotics in prevention/treatment of atherosclerosis and coronary artery disease has been proposed.^86,89–91 Animal trials of probiotics in coronary heart disease have focused mostly on lowering serum lipoprotein cholesterol.⁹² A rat model of heart attacks or myocardial infarctions (MI) demonstrated that a combination of two probiotics prior to and following the induced MI had less post-MI apoptosis in certain cerebral regions (e.g., limbic system) than the controls. The proposed mechanism is a decrease in circulating pro-inflammatory cytokines as a result of probiotic ingestion.⁹³

Unanswered Questions and Hypothesis

Many questions remain unanswered regarding the use of probiotics in high-risk term infants with CCHD. Figure 2 summarizes some of the proposed mechanisms of action of probiotics in preventing NEC in premature infants. The following questions remain unanswered in term infants with CCHD (Fig. 3):

1. How does the intestinal microbiota of the term infant with CCHD differ from that of the healthy breast-fed term infant?

2. Does the administration of probiotics change the intestinal microbiota of the term infant with CCHD?

3. Does administration of probiotics to term infants with CCHD change fecal or plasma levels of pro-inflammatory Gram-negative bacterial antigens (e.g., LPS) or cytokines?

4. Does feeding human milk to infants with CCHD decrease the risk of NEC?

5. Does administration of probiotics and/or prebiotics to infants with CCHD decrease the risk of NEC?

Figure 2.

Proposed mechanisms for prevention of NEC by probiotics in premature infants.

Figure 3.

Unanswered questions regarding NEC onset in term infants with cyanotic congenital heart disease (CCHD) and the potential role of probiotics.

Future studies could include experimental investigation of the following overarching hypotheses: CCHD alters the intestinal microbiota composition of term infants increasing the risk of NEC and prophylactic administration of probiotics alters the intestinal microbiota in this high-risk population decreasing the NEC risk. Pilot probiotic clinical trials in infants with CCHD could include analysis of fecal microbiota and fecal LPS levels with comparisons made between the probiotic group and the placebo group and between the placebo group and age- and sex-matched healthy control infants. Such studies would be essential to determine feasibility and sample size for a larger multi-center trial of NEC prevention. Parallel animal studies to elucidate the impact of chronic hypoxia on the development of the intestinal microbiota as well as intestinal permeability and immune function would also be of interest.

Potential Barriers to Research Progress

Guidelines to designing, conducting, publishing and communicating results of clinical trials involving probiotic applications in humans have recently been published.⁹⁴ The following concerns are of particular relevance in the context of infants with CCHD.

Safety.

The safety of probiotics has been extensively reviewed.^95,96 Probiotics have been added to infant formulas in Europe and Asia for many years without evidence of harm.⁹⁷ Probiotic containing formulas are now available in North America. Potential safety concerns include the following:

1. Potential for systemic infection including septicemia and endocarditis from probiotic species. Rare cases of bacteremia and fungemia have been reported, particularly for species of lactobacilli⁹⁸ and Saccharomyces.⁹⁹ However population studies in countries with high probiotic use have not shown any increased risk of bacteremia or endocarditis due to probiotics.^100,101 Probiotics appear to actually decrease rather than increase bacterial translocation in both animal models^102,103 and humans.¹⁰⁴

2. Antibiotic resistance. Testing of 41 probiotic supplements and foods demonstrated high sensitivity to chloramphenicol, tetracycline, ampicillin, amoxicillin/clavulanic acid, cephalothin and imipenem, moderate sensitivity to vancomycin, rifampicin, streptomycin, bacitracin and erythromycin and high levels of resistance to ciprofloxacin, amikacin, trimethoprim/sulphamethoxazole and gentamicin.¹⁰⁵ Bifidobacteria appear to have minimal antibiotic resistance. Twenty-six strains of B. breve showed susceptibility to all antibiotics tested (17 total) except streptomycin.¹⁰⁶

3. Purity and viability of commercial probiotic products. Several studies have demonstrated that labeling of commercial probiotic products is often inaccurate. Some commercial products contain organisms not listed and others do not contain the specific strain listed.¹⁰⁷ In addition, stability of probiotics changes over time so that it is challenging to be certain of the dose provided. Ideally, strains with a known genomic sequence and sensitivity pattern should be used⁸⁴ and regular testing of viability performed.

Onset of disease.

The onset of NEC in term infants is very early (average 4 days of life). This may not apply to infants with CCHD, as most of these infants develop NEC post-operatively. Early onset of NEC suggests that attempts to change the intestinal microbiota and thereby prevent this disease may not be effective.

Potential confounders.

Infants with CCHD often receive prolonged courses of antibiotics, which may negate any positive effects of the probiotic administered. In addition, infant feeding patterns have a significant effect on the intestinal microbiota. Individual institution protocols for care may also impact outcomes.

Specimen collection.

Stool has often been assumed to represent the microbiota of more distal intestinal sites (e.g., ileum to anus) and, therefore, may not provide an accurate assessment of the microbiota inhabiting more proximal sites. In addition, stool specimens can also dominantly represent allochthonous and transient species (also called liquid-phase microbes), as opposed to the autochthonous and adherent biofilms (also called particle-phase microbes) interacting directly with the host gut wall.¹⁰⁸ The abundance and diversity of these species may be underestimated in fecal surveys as shown by mucosal biopsy-specific detection of microbes¹⁰⁹ implicated in other inflammatory bowel diseases. Qualitatively, spatial and temporal studies of human intestinal ecology as well as comparative studies of intestinal sample collection/preparation methods produce differing results. Standardization of sampling and laboratory approaches will add clarity to this complex field.

Conclusion

Necrotizing enterocolitis is a devastating complication in term neonates with cyanotic congenital heart disease, often resulting in great morbidity and mortality among these high risk ‘blue babies’. Whether intestinal hypoxia and hypoperfusion alters the intestinal microbiota increasing the risk of NEC is unknown. Whether administration of probiotics to these infants alters their intestinal microbiota, decreases the risk of NEC, and/or improves weight gain also is unknown. If a simple preventive approach, like probiotic therapy, were effective, this would be a significant breakthrough and shift in the treatment paradigm for these high risk infants.

Abbreviations

NEC

necrotizing enterocolitis

CCHD

cyanotic congenital heart disease

LPS

lipopolysaccharide

TLR

toll-like receptor