Necrotizing enterocolitis (NEC) is a devastating inflammatory disease of the neonatal bowel characterized by serious morbidity and a mortality rate of 20% to 30% (1). It is the most common gastrointestinal (GI) tract emergency of the neonatal period and a major cause of infantile short gut syndrome (2). The etiology is multifactorial and under intensive investigation. The primary risk factor is prematurity, with bowel ischemia, enteral feedings, and bacterial colonization considered as additional risk factors (3).
The fetal gut is sterile, yet the adult GI tract houses up to 100 trillion microbes (4). Birth marks the beginning of the newborn's microbial colonization process that evolves continuously into adult life. The predominant bacterial phyla found in the GI tract are Firmicutes, Bacteroidetes, and Proteobacteria(4). Recent studies report decreased microbial diversity and increased relative abundance of Proteobacteria in infants with NEC (5,6).
External factors, for example, mode of delivery or type of enteral feed, may affect colonization by influencing microbial exposure and the bacterial environment. Increasing evidence suggests that gastric pH may also influence the GI bacterial milieu (7). Breast milk has a poorer buffering capacity than formula. The resulting lower colonic pH in breast-fed infants promotes the growth of commensal organisms such as Bifidobacterium spp and lactobacilli, and may help explain the protective benefit of breast milk in the prevention of NEC (8–10). In a prospective, double-blind study, acid supplementation of milk resulted in lower gastric pH levels, less gastric colonization with enteric Gram-negative bacteria, and a reduced incidence of NEC (11).
Histamine-2 receptor (H2-) blockers have been used extensively in neonates to reduce gastric acidity for a variety of clinical indications. They are given to minimize esophagitis in severe gastroesophageal reflux, to prevent and treat steroid-associated gastritis and as an additive to total parenteral nutrition to reduce stress-associated gastritis; however, recent observations suggest caution may be indicated in their administration. Using a large database from the National Institute of Child Health and Human Development Neonatal Research Network, Guillet et al (12) found H2-blocker therapy to be associated with higher rates of NEC. They postulated that a less acidic gastric milieu may lead to bacterial overgrowth and promote the development of NEC. In a large, prospective, multicenter study, Terrin et al (13) found a >6-fold increased frequency of NEC, as well as a higher mortality and more prolonged hospitalization in very-low-birth-weight infants treated with ranitidine. They also hypothesized that ranitidine-induced gastric hypochlorhydria may have altered the intestinal microflora.
The effect of H2-blockers upon colonic bacterial colonization is unclear. The purpose of the present study was to test the hypothesis that H2-blockers alter colonic bacterial colonization by analyzing and comparing the fecal microbiota in premature infants with and without H2-blocker therapy using sensitive molecular biological techniques.
This case control study is part of an ongoing multicenter investigation of factors affecting the postnatal development of fecal microbiota in premature infants. It was approved by the institutional review boards of Louisiana State University Health Sciences Center, Touro Infirmary, East Jefferson General Hospital and Children's Hospital of New Orleans.
Following parental consent, 76 premature infants ≤1500 g or ≤34 weeks of gestation were enrolled. Twenty-five infants received H2-blockers (ranitidine 1–4 mg · kg−1 · day−1 intravenously or 6 mg · kg−1 · day−1 enterally) for an average of 19 days (range 3–58 days) before stool collection (H2-Pos); 51 babies had never received H2-blockers (H2-Neg). One stool sample from each subject and control was used for the cross-sectional comparison. There were no significant differences in most clinical parameters between the 2 groups (Table 1). H2-Pos infants were less likely to have been exposed to antenatal steroids and more likely to be delivered vaginally than controls. All infants received enteral feedings of formula, breast milk, or a mixed diet of both breast milk and formula. The majority of neonates in each group had been treated with antibiotics early in their neonatal intensive care unit course. Three of 25 infants in the H2-Pos group (12%) and 8 of 51 infants in the H2-Neg group (16%) developed either culture-proven sepsis or urinary tract infection at some time during their hospitalization (P = 1.000, not significant). No infant in either group had culture-proven sepsis or urinary tract infection during the study. No child had received antibiotics for at least 8 days before study. No infant developed NEC.
Stool samples were collected from each infant, placed in buffered ethanol (50% ethanol, 10 mmol/L Tris-HCl, 1 mMol/L ethylenediaminetetraacetic acid), and frozen before analysis. DNA was extracted from specimens using QIAamp DNA micro kit according to the manufacturer's instructions (Qiagen Inc, Valencia, CA). An initial bacterial cell lysis step using lysozyme (20 mg/mL at 37oC for 1 hour) was included. Bacterial tag–encoded FLX amplicon pyrosequencing was performed by the Research and Testing Laboratory (Lubbock, TX) using broad-range PCR amplification of the V1–V3 region of the 16S rRNA gene with primers 28F: GAGTTTGATCNTGGCTCAG and 519R: GWNTTACNGCGGCKGCTG. Sequences that were <200 bases in length, contained ≥1 ambiguous bases, had mean quality scores <Q25, or had a homopolymer region >6 nucleotides were excluded. Taxonomic assignment of the pyrosequencing reads was performed using RDP classifier. The resulting sequences were analyzed using the QIIME pipeline (14). The total number of reads used in the study was 420,203, averaging 5529 reads per specimen with an average length of 355 bases.
Community composition and Principal Coordinates Analysis (PCoA) were performed using the QIIME pipeline (14). Sequences were clustered at 97% similarity using UCLUST (15), aligned using PyNast (16) and the Greengenes core reference alignment (17) and classified using the RDP Classifier 2.2 (18). A phylogenetic tree was created using FastTree 2.1.3 (19) and β-diversity was calculated using UniFrac (20). The QIIME normalized data were further analyzed to determine relative abundance at various taxonomic levels and Shannon Diversity Index (DI). Depending on data distribution, differences between the groups were tested by t test or the Mann-Whitney nonparametric test, (GraphPad Instat 3.0, GraphPad Software, La Jolla, CA). In further analysis, PROC REG and PROC GLM in the Statistical Analysis System version 9.3 (SAS Institute, Cary, NC) were used to fit an analysis of covariance model to the percent relative abundances of Proteobacteria. The model included dichotomous effects for H2-blocker therapy (coded 0 = no and 1 = yes), delivery method (coded 0 = caesarean section and 1 = vaginal), and continuous predictors for the total number of days since delivery and number of days since last antibiotic use. The overall model was significant (F = 5.67, P = 0.0005) with an R2 value of 0.24.
Proteobacteria and Firmicutes were the major phyla contributing to fecal microbial community composition in the premature infants studied. Wide inter-individual variations were found. No apparent community differences were detected by PCoA; however, microbial diversity was lower (P = 0.018, t test) in the H2-Pos (DI = 0.323 ± 0.263 [standard deviation, SD]) than in the H2-Neg group (DI = 0.480 ± 0.266 [SD]). Comparison of individual taxa revealed that the mean relative abundance of Proteobacteria was increased, whereas that of Firmicutes was decreased in the H2-Pos compared with the H2-Neg infants (Fig. 1).
The analysis of covariance revealed that H2-blocker therapy was positively associated with increased mean relative abundance of Proteobacteria (P = 0.0014) after adjusting for all other variables in the model. In addition, vaginal delivery was associated with a decreased mean relative abundance of Proteobacteria compared with caesarean section delivery (P = 0.0216). The days since delivery were positively associated with mean relative abundance of Proteobacteria (P = 0.0356), but days since last antibiotic use were not a significant factor in the model.
Further taxonomic differentiation detected a greater relative abundance of gamma-Proteobacteria of the family Enterobacteriaceae in the neonates receiving H2-blockers compared with H2-Neg babies (Fig. 2).
One infant who had received 2 separate courses of H2-blockers was followed longitudinally with serial fecal analyses (Fig. 3). The clinical picture was complicated by initial nil per os status and use of antibiotics at the time of the first 2 sample collections. Nonetheless, the pattern of change in fecal flora supports the concept that use of H2-blockers shifts the relative microbial abundance toward the phylum Proteobacteria and away from Firmicutes. Cessation of H2-blocker therapy was associated with a reversal of this shift. Reinstitution of H2-blocker therapy resulted in a renewed shift toward Proteobacteria.
Gastric acidity acts as a natural defense against bacterial growth, specifically inhibiting growth of many Proteobacteria (21). Although gastric acid production is low immediately after birth, premature infants develop and maintain a pH <4 within the first few days of life (22,23). H2-blockers significantly increase gastric pH within hours of administration (24). Alteration of the naturally occurring acidic gastric defense barrier changes the gastric floral pattern (21). Our study suggests that it also lowers microbial diversity and changes the fecal microfloral pattern, shifting it toward greater predominance of Proteobacteria.
Previous studies have demonstrated an association of diminished diversity (5) and overabundance of Proteobacteria in fecal samples from premature infants with NEC (5,6). Our study was too limited and not designed to look specifically at the effect of H2-blockers on the incidence of NEC. None of our infants developed NEC; however, the previous detection of an increased incidence of NEC in premature infants receiving H2-blockers (12) and clinical studies (11) clearly demonstrating that higher gastric pH favors colonization with Gram-negative organisms, raises concern. Many Proteobacteria, especially of the family Enterobacteriaceae, are known pathogens, such as Klebsiella, Shigella, Escherichia coli, Citrobacter and the powdered formula contaminant Cronobacter sakazakii(25). They are Gram-negative, facultative anaerobes, often motile, and capable of producing toxins (26), adhesins (27), and capsular antigens (28). They have the ability to undergo antigenic phase variation (29), type III secretion (30), and exchange antimicrobial resistance genes (31). Some have been associated with epidemics or anecdotal cases of NEC (32). An overabundance of such organisms in an immature GI tract is cause for concern.
Our findings suggest that H2-blockers lower microbial diversity and shift the fecal microbial pattern toward Proteobacteria. These alterations in fecal microbiota may predispose the vulnerable premature gut to NEC. Until further data are available, it would seem prudent to limit the widespread use of these drugs in the premature infant.
The authors thank Dr Donald Mercante, Director of Biostatistics, Louisiana State University Health Sciences Center School of Public Health, for expert help in statistical analysis.
1. Fitzgibbons SC, Ching Y, Yu D, et al. Mortality of necrotizing enterocolitis expressed by birth weight categories. J Pediatr Surg
2. Hunter CJ, Upperman JS, Ford HR, et al. Understanding the susceptibility of the premature infant to necrotizing enterocolitis (NEC). Pediatr Res
3. Neu J, Walker WA. Necrotizing enterocolitis. N Engl J Med
4. Palmer C, Bik EM, DiGiulio DB, et al. Development of the human infant intestinal microbiota. PLoS Biol
5. Wang Y, Hoenig JD, Malin KJ, et al. 16S rRNA gene-based analysis of fecal microbiota from preterm infants with and without necrotizing enterocolitis. Isme J
6. Mai V, Young CM, Ukhanova M, et al. Fecal microbiota in premature infants prior to necrotizing enterocolitis. PLoS One
7. Bergholz TM, Whittam TS. Variation in acid resistance among enterohaemorrhagic Escherichia coli in a simulated gastric environment. J Appl Microbiol
8. Wilson M. The gastrointestinal tract and its indigenous microbiota. In: Microbial Inhabitants of Humans: Their Ecology and Role in Health and Disease
. London: Cambridge University Press; 2005:278–282.
9. Iseki K. [Development of intestinal flora in neonates]. Hokkaido Igaku Zasshi
10. Bullen JJ. Iron-binding proteins and other factors in milk responsible for resistance to Escherichia coli
. Ciba Found Symp
11. Carrion V, Egan EA. Prevention of neonatal necrotizing enterocolitis. J Pediatr Gastroenterol Nutr
12. Guillet R, Stoll BJ, Cotten CM, et al. Association of H2-blocker therapy and higher incidence of necrotizing enterocolitis in very low birth weight infants. Pediatrics
13. Terrin G, Passariello A, De Curtis M, et al. Ranitidine is associated with infections, necrotizing enterocolitis, and fatal outcome in newborns. Pediatrics
14. Caporaso JG, Kuczynski J, Stombaugh J, et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods
15. Edgar RC. Search and clustering orders of magnitude faster than BLAST. Bioinformatics
16. Caporaso JG, Bittinger K, Bushman FD, et al. PyNAST: a flexible tool for aligning sequences to a template alignment. Bioinformatics
17. DeSantis TZ, Hugenholtz P, Larsen N, et al. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl Environ Microbiol
18. Wang Q, Garrity GM, Tiedje JM, et al. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol
19. Price MN, Dehal PS, Arkin AP. FastTree 2--approximately maximum-likelihood trees for large alignments. PLoS One
20. Lozupone C, Knight R. UniFrac: a new phylogenetic method for comparing microbial communities. Appl Environ Microbiol
21. O’May GA, Reynolds N, Macfarlane GT. Effect of pH on an in vitro model of gastric microbiota in enteral nutrition patients. Appl Environ Microbiol
22. Kelly EJ, Newell SJ, Brownlee KG, et al. Gastric acid secretion in preterm infants. Early Hum Dev
23. Hyman PE, Clarke DD, Everett SL, et al. Gastric acid secretory function in preterm infants. J Pediatr
24. Kuusela AL. Long-term gastric pH monitoring for determining optimal dose of ranitidine for critically ill preterm and term neonates. Arch Dis Child Fetal Neonatal Ed
25. Joseph S, Forsythe SJ. Predominance of Cronobacter sakazakii
sequence type 4 in neonatal infections. Emerg Infect Dis
26. Pagotto FJ, Nazarowec-White M, Bidawid S, et al. Enterobacter sakazakii: infectivity and enterotoxin production in vitro and in vivo. J Food Prot
27. Oelschlaeger TA, Dobrindt U, Hacker J. Virulence factors of uropathogens. Curr Opin Urol
28. Korhonen TK, Valtonen MV, Parkkinen J, et al. Serotypes, hemolysin production, and receptor recognition of Escherichia coli strains associated with neonatal sepsis and meningitis. Infect Immun
29. Hallet B. Playing Dr Jekyll and Mr Hyde: combined mechanisms of phase variation in bacteria. Curr Opin Microbiol
30. Young GM, Schmiel DH, Miller VL. A new pathway for the secretion of virulence factors by bacteria: the flagellar export apparatus functions as a protein-secretion system. Proc Natl Acad Sci U S A
31. Hansen LH, Bentzon-Tilia M, Bentzon-Tilia S, et al. Design and synthesis of a quintessential self-transmissible IncX1 plasmid, pX1.0. PLoS One
32. Boccia D, Stolfi I, Lana S, et al. Nosocomial necrotising enterocolitis outbreaks: epidemiology and control measures. Eur J Pediatr