Secondary Logo

Journal Logo

Society Papers

Probiotics and Preterm Infants

A Position Paper by the European Society for Paediatric Gastroenterology Hepatology and Nutrition Committee on Nutrition and the European Society for Paediatric Gastroenterology Hepatology and Nutrition Working Group for Probiotics and Prebiotics

van den Akker, Chris H.P.*; van Goudoever, Johannes B.*; Shamir, Raanan; Domellöf, Magnus; Embleton, Nicholas D.§; Hojsak, Iva||; Lapillonne, Alexandre¶,#; Mihatsch, Walter A.**; Berni Canani, Roberto††; Bronsky, Jiri‡‡; Campoy, Cristina§§; Fewtrell, Mary S.||||; Fidler Mis, Nataša¶¶; Guarino, Alfredo††; Hulst, Jessie M.##; Indrio, Flavia***; Kolaček, Sanja||; Orel, Rok||||; Vandenplas, Yvan†††; Weizman, Zvi‡‡‡; Szajewska, Hania§§§

Author Information
Journal of Pediatric Gastroenterology and Nutrition: May 2020 - Volume 70 - Issue 5 - p 664-680
doi: 10.1097/MPG.0000000000002655

Abstract

What Is Known

  • Probiotics might be a potential therapy for preterm infants to reduce morbidity and mortality.
  • Only a limited number of different strains have shown preliminary potential effectiveness.

What Is New

  • We provide advice, which specific strains might potentially be used and which strains should not be used for preterm neonates.
  • Several safety issues are addressed to which probiotic products and their supplementation for preterm infants should fulfil.

Infants born prematurely have high rates of mortality, septicaemia, and gastrointestinal morbidities, such as necrotising enterocolitis (NEC). The exact aetiology of these morbidities is unknown, but include intestinal immaturity with increased permeability and an immature immune system (1–3). Enteral tolerance is frequently reduced in preterm infants, and most require parenteral nutrition. Feeding preterm infants nonpasteurized own mother's milk is the best feeding strategy to reduce neonatal mortality and many morbidities (4).

Over the last 2 decades, certain probiotic strains, either single or in combination, have been administered in clinical trials in an attempt to reduce NEC and late-onset sepsis, and to improve feed-related outcomes, such as time to full feeds. Whilst multiple potential mechanisms of how probiotics may exert their beneficial effect have been postulated (5–8), very few, if any, mechanistic studies exist in this patient group. Results of individual trials have varied, but almost all systematic reviews and meta-analyses have shown positive effects on reducing the incidence of a range of adverse outcomes when studies with different strains are combined and analysed as a single group (9–23). Importantly, however, long-term neurodevelopmental follow-up has neither shown beneficial nor detrimental effects of probiotics in preterm neonates in a recent meta-analysis based on 5 studies in 1637 infants (24).

Whilst many have strongly argued for their routine use (25–28), other groups including the European Society for Paediatric Gastroenterology Hepatology and Nutrition (ESPGHAN) and the American Academy of Pediatrics (AAP), have been more cautious, noting some of the major limitations in many of the studies, methodological differences in study design, and pointing out that probiotic efficacy may vary widely (29–36). This was emphasised by a recent high-quality study in the UK that showed no effect for a specific strain of Bifidobacterium breve (BBG-001) on mortality or NEC in a large group of preterm infants (37). The importance of strain specificity is further exemplified by the fact that within the species Escherichia coli, certain strains may cause haemolytic uremic syndrome (strain O157:H7), whereas others are considered probiotic supplements (strain Nissle 1917). On the other hand, several probiotic genera or species share underlying mechanistic characteristics that are beneficial (38), which would favour the argument of pooling the data of several strains together. So far, heterogeneity of organisms and dosing regimens studied have prevented strain-specific treatment recommendations from being made.

Recently, the ESPGHAN Working Group for Probiotics and Prebiotics published a document using a network meta-analysis (NMA) approach to identify strains with greatest potential efficacy for preventing major neonatal morbidities in preterm infants (39). Following this publication, the ESPGHAN Committee on Nutrition and the ESPGHAN Working Group for Probiotics and Prebiotics aimed to develop a document that might serve as a guide for the possible use of probiotics in preterm infants.

METHODS

An ESPGHAN Position Paper addresses a topic for which guidance is necessary but there is only limited scientific evidence, and therefore, the recommendations are mostly based on expert opinion. A writing consensus group was convened to support the development of this document. This group included experts in the fields of neonatology, paediatric gastroenterology, and nutrition. All members of the group disclosed any potential conflicts of interest. No funding for the development of this document was received.

Defining the Clinical Questions

The first stage of the development of this position paper involved specifying the clinical questions:

  1. Are probiotics safe enough for administration to preterm infants?
  2. Should probiotics be used in preterm infants? If yes, which probiotics (single or combinations) should be used in what dose?
  3. Are combinations of species more effective than the use of a single strain to reduce the risk of NEC (stage 2 or 3)?
  4. Which dose of a probiotic strain or combination of strains should be administered?
  5. What should be the duration of administering probiotics?
  6. Is it appropriate to administer other strains than those studied in large well-conducted randomised controlled trials (RCTs)?

Methodology for Synthesis and Grading of Recommendations

The panel decided as a primary starting point that any recommendations for the use of probiotics should be specified at strain level because of the strain-specific effects. This relates particularly to clinical question 2 and means that for studies in which the probiotic was only specified at the species level (without strain designation), no recommendation could be derived. Furthermore, it was decided that recommendations should be based primarily on the results from RCTs, and that evidence from cohort studies is only used for interpretation and discussion of the recommendation. Thus, our recently published probiotic strain-specific systematic review and NMA (39) could form the direct basis. Although in our previous paper, we defined prematurity as a gestational age (GA) of less than 37 weeks’ gestation, the recommendations posed here are only applicable to infants being born at less than 32 weeks’ gestation. Furthermore, the panel decided that we could only make proper recommendations on the use of a specific intervention that was tested in RCTs with adequate cumulative power for at least 1 of the 3 outcomes of particular interest, namely mortality, NEC, or late-onset sepsis. This reduces the chance of making recommendations based on type 1 errors (false positive). Sample size calculations are depicted in Table 1 for each of the 3 outcome domains, with an α of 0.05 and 1-β of 0.80. The panel acknowledges that proposed reductions in mortality, NEC stage ≥2, or late-onset sepsis are arbitrary. We, however, deliberately chose high baseline rates and optimistic reductions in order to achieve realistic sample sizes. This translated to a minimum of 247 infants (per group) corresponding to a sepsis reduction from 25% to 15% with 80% power, being the least number of infants that needed to be studied before recommendations were made. For each outcome domain, we assessed power separately to take this into account when formulating the recommendations (by downgrading certainty of evidence on imprecision, see below).

TABLE 1
TABLE 1:
Sample size calculations for each outcome domain (α = 0.05; 1-β = 0.80; 2-sided)

To grade the recommendations, the GRADEpro software was used, developed by the Grading of Recommendations, Assessment Development and Evaluations Working Group (40). GRADE assesses evidence quality by grading risk of bias, inconsistency, indirectness, and imprecision each as not serious, serious, or very serious. On the basis of these assessments any observed risk reduction is categorized as high, moderate, low, or very low certainty of evidence. The GRADEpro system offers 2 categories for the strength of the final recommendation (strong or conditional). The strength of a recommendation was graded as strong when the evidence showed a clear benefit or absence of benefit of the intervention based on moderate or high certainty of evidence. The strength of a recommendation was graded as conditional when the trade-offs were less certain, either because of the low certainty of evidence or as the evidence suggested that desirable and undesirable effects were closely balanced.

Unfortunately, only clinical questions 1 and 3 could be answered from systematic PICO (population, intervention, comparison, outcome) questions where RCTs assessed our patient group of interest. Regarding clinical question 2, the final proposed recommendations for strains (single or combinations) are based on the combined evidence on mortality, NEC stage 2 or 3, and late-onset sepsis rates, together with its quality grading (certainty assessment) and are depicted in GRADE tables. Effect sizes are reported as a relative risk (RR) versus placebo with its 95% credible interval (CrI). For each recommendation, we provide the dose (or the range in which it was used) of the probiotic strains (single or combination) that exerted the effect in the available studies.

To answer clinical question 3, additional models were constructed in our NMA database using the same methodology as previously used (39). First, we compared placebo versus administration of a single probiotic strain/species and versus multiple strains/species on the incidence of NEC stage 2 or 3. Second, we compared placebo versus administration of any single/multiple Lactobacillus probiotic(s), versus any single/multiple Bifidobacterium probiotic(s), versus the combination of any Lactobacillus and Bifidobacterium probiotics. Effect sizes are reported as a RR versus placebo with its 95% CrI. As these analyses are not strain-specific, these data are only hypothesis-generating. Therefore, these recommendations were rated as conditional and based on very low certainty of evidence.

The other clinical questions (1, 4, 5, and 6) are each discussed based on the known literature (mainly case series and the expertise of the authors). Because this is regarded as indirect evidence, these recommendations were also rated as conditional and based on very low certainty of evidence.

Probiotic Nomenclature

For the remainder of this manuscript, probiotic species are truncated at their genus: Bifidobacterium, Escherichia, Lactobacillus, Saccharomyces, and Streptococcus are denoted by B, E, L, S, and Str, respectively. In addition, subspecies (subsp) names are truncated as well: Bifidobacterium animalis subsp lactis is denoted as B lactis; Bifidobacterium longum subsp infantis as B infantis; B longum subsp longum as B longum, and Streptococcus salivarius subsp thermophilus as Str thermophilus. Over the past decades. multiple reclassifications in taxonomy have been proposed and designations in the historical publications may no longer be accurate. We therefore adhered to the latest nomenclature we were aware of, so that for example B bifidum Bb-12 is designated as B lactis Bb-12 (41). Although in our recent NMA, we analysed some strains together because of their relative resemblance, we here chose to be truly strain-specific. Results from the Lactobacillus reuteri DSM 17938 strain are thus separated from L reuteri ATCC 55730, and B lactis B94 is now separated from B lactis Bb-12.

Document Review

The manuscript and the recommendations were drafted first by the writing committee of the group (Chris H.P. van den Akker, Johannes B. van Goudoever, Hania Szajewska, and Raanan Shamir). Then, several other members of the author group (Magnus Domellöf, Nicholas D. Embleton, Iva Hojsak, Alexandre Lapillonne, and Walter A. Mihatsch) reviewed and discussed the evidence, reviewed the drafted recommendations, and reached a consensus on the strength of each recommendation. As a next step of the consensus development process, the manuscript with its draft recommendations was then submitted for review to the other members of the ESPGHAN Committee on Nutrition and the ESPGHAN Working Group for Probiotics and Prebiotics (Roberto Berni Canani, Jiri Bronsky, Cristina Campoy, Mary S. Fewtrell, Nataša Fidler Mis, Alfredo Guarino, Jessie M. Hulst, Flavia Indrio, Sanja Kolaček, Rok Orel, Yvan Vandenplas, and Zvi Weizman). Then, the finalized manuscript was sent to all aforementioned people together with an invitation to vote the recommendations. The ideal was to reach 100% consensus, but 85% agreement was considered acceptable as is proposed by the general ESPGHAN Guideline Development Group. All of the comments were considered, and revisions were made in response to peer-reviewer's comments until the desired 85% threshold was reached. If consensus was not reached within a maximum of 3 voting rounds, the recommendation was not accepted. A finalised document was submitted to the ESPGHAN Council for peer review before publication.

Updating

The group will monitor new publications and evidence made available and decide whether and when it is necessary to update the recommendations. In any case, the results will be reviewed within 5 years from publication.

SUMMARY OF EVIDENCE, INTERPRETATION, AND RECOMMENDATIONS

Are Probiotics Safe Enough for Administration to Preterm Infants?

Probiotics may theoretically be responsible for at least 5 types of side effects: systemic infections, deleterious metabolic activities, excessive immune stimulation, antibiotic resistance gene transfer, and gastrointestinal side effects, such as intestinal gas formation (42,43). Most of the RCTs conducted in preterm infants or other patient groups, however, did not adequately monitor or report these side effects (44). Other safety issues might more be related to quality control of the probiotic supplementation. Several issues will be elaborated below.

Probiotic sepsis in premature infants could be particularly important, as they represent an immunocompromised patient group. Furthermore, probiotic bacteraemia may be hard to detect with classic culture methods especially in single paediatric culture bottles, as strictly anaerobic strains are difficult to grow. Yet, multiple case reports have described single or multiple cases of bacteraemia (sometimes in conjunction with NEC) in premature infants (45). In particular, B infantis(46–49) and L rhamnosus GG (50–55) bacteraemia have been described in premature neonates, but other cultured probiotic strains include L reuteri(56), Saccharomyces boulardii(57,58), B breve BBG-001 (59), and E. coli Nissle 1917 (60). Probiotic bacteraemia may occur not only because of intestinal translocation, but also because of contamination from probiotic preparation and subsequent line handling. Especially if probiotics are prepared on the ward from powder sachets or capsules that are opened, probiotic spills and contamination may occur to other surface areas, medications, or intravenous catheter sites, or cross colonisation to other infants on the neonatal ward (61,62). Although the cross colonisation may not necessarily be seen as an adverse effect (in case of a safe product), it illustrates how easily living organisms may spread and warrants extreme caution when preparing and supplementing a probiotic supplement. This is exemplified in a recent paper on 3 cases of L rhamnosus GG bacteraemia in preterm neonates (55), in which only 1 infant actually received the particular probiotic strain, whereas the other 2 infants (who also had a central line) were only hospitalised in the same room as where other infants were supplemented with probiotics. Simultaneously, a study was published describing 6 cases of L rhamnosus GG bacteraemia out of a cohort of 522 patients on a paediatric intensive care unit who also received a probiotic supplement containing L rhamnosus GG (63). None of the infected infants were immunocompromised or had known bowel disintegrity, but contamination of their central line was suspected. Similarly, because of the risk of contamination, the European Medicine Agency even amended a contra-indication to the use of S boulardii in patients (not specifically neonates) who are critically ill or are immunocompromised, or those who have a central venous catheter (64).

Examples of deleterious metabolic activities include increased D-lactate and biogenic amines production or bile salt hydrolysis activity affecting cholesterol metabolism and lipid uptake. Because of a complete lack of data in infants and children on the latter examples, only the issue of D-lactate is elaborated on here as this has been studied in older infants. Whereas some Lactobacilli strains produce mainly L-lactate, many produce a mixture, and some predominantly produce D-lactate. From the Lactobacilli that are described in the next clinical question, L rhamnosus GG ATCC 53103 produces almost only lactate in its L-isoform, but fermentation by L reuteri DSM 17938 or L acidophilus NCDO 1748 (ATCC 4356, LA37, or NCIMB 30316) yields larger proportions of D-lactate. Although the amount of D-lactate that is produced may quantitatively be relatively small, D-lactate is difficult to dispose of after enteral uptake, which could be even more problematic in premature infants (65). Not only do most premature infants already have the tendency to be acidotic, D-lactate cannot routinely be measured in blood gases, making it very difficult to suspect or detect. In healthy 6-month old infants (66), children (67), or adults (68), D-lactate formation by probiotics is probably not much of a clinical issue. However, in term-born infants elevated urinary D-lactate concentrations were found in the first 2 weeks of life after being fed a L reuteri DSM 17938-containing formula (69), although there were no signs of blood acidosis. Yet, several case reports have appeared describing D-lactate acidosis in short bowel syndrome infants (70,71). To avoid any risks, it is stated in the Codex Alimentarius that if probiotics are added to infant formulas, they may only contain L-lactate-producing cultures (72). On the other hand, L reuteri DSM 17938 has been reviewed as GRAS (generally recognized as safe) for the use in term infant formula by the Food and Drug Administration (FDA) (73). In premature infants, the issue of D-lactate has not been systematically researched, but it seems prudent to select only those Lactobacilli that are predominantly L-lactate producers in preterm infants, until further specific safety data is available in this specific patient group. This may be especially important in infants during kidney failure or with short bowel syndrome (e.g. after extensive NEC surgery).

Although there are some indications that meconium is not sterile (74), these findings are challenged by others (75). Yet, the vast majority of gastrointestinal colonisation of the microbiome occurs in the weeks after birth (76). Albeit premature infants on a NICU by definition have an abnormal colonization because of an immature immune system, less parental skin-to-skin contact, and frequent antibiotic exposure, supplementing 1 or few probiotic strains soon after (preterm) birth influences colonization as well. It is currently unknown if this effect only lasts during supplementation or has longer influences and if any effect is positive in later life with potential excessive immune stimulation or allergy in susceptible individuals, although this has not been systematically researched. Augmentation of natural killer activity, T-cell functions, and cytokine production are some of the plausible mechanisms underlying the immune regulatory activities of probiotics (43,77).

The gut harbours simultaneously with its microbiome, a natural reservoir of antibiotic resistance genes, which appear to increase upon increased antibiotic exposure (78,79). This can be beneficial during administration of antibiotics, as it preserves some protection to the bacterial microbiome. Also, many commercially available probiotics carry some antibiotic resistance genes (80,81). As long as these genes in probiotic products are not transferable through plasmids to other more pathogenic bacteria, these risks are probably limited. Yet, there are several examples of probiotic strains with potentially transferable genes (82–85). Especially under antibiotic pressure such as on the NICU, risks of horizontal gene transfer might be higher and contribute to increased antibiotic resistance (86–88). A vancomycin-resistant enterococcus outbreak on a Turkish NICU was linked to the provision of certain probiotics in a recent study (89), although another report showed no greater antibiotic resistome in infants that had received probiotics (90). As most preterm infants will receive concomitant antibiotics during some period on their NICU stay, it is prudent to select only those probiotic strains with known safety profile on gene transfer (91).

Other potential gastrointestinal side-effects of probiotics, such as intestinal gas formation are even less studied, especially in premature infants. Therefore, and because potential adverse effects are probably less severe than potential benefits, these will not be addressed here further.

Probiotics are usually marketed as nutritional supplements rather than as drugs and, thus, form an unregulated market where manufacturers may change product contents and/or the production process without properly addressing these issues (92). Previously, the ESPGHAN has also called for more stringent controls of the production of probiotics, especially in premature neonates (93,94). Ascertaining product safety and quality is of specific concern here, as preterm infants frequently have the need for indwelling catheters and nasogastric tubes, and they do not have an adequate immune response. For example, a fatal case of gastrointestinal mucormycosis in a preterm infant has been described following contamination of a combination of 3 probiotic strains (95). The caveats in quality control of probiotics should thus be more stringent to ensure that the probiotic content as mentioned on the label meets the actual content throughout the shelf life of the product, while no contamination is present. Several reports, however, show that product labels on commercial or medical probiotic products frequently do not match actual contents in terms of species identity and bacterial count, or contained contamination with nonprobiotic bacteria (96–98). Even in a more recent report, only 1 out of 16 tested commercial probiotic products (including those marketed specifically for infants) contained the correct probiotics at subspecies level as claimed on the product label (99). Ensuring correct product identity at strain level is essential, not only during research but also during actual clinical implementation, in order to match achieved trial results to clinical practice (100). Probiotic products for premature infants should, therefore, be manufactured according to current Good Manufacturing Practice (cGMP) guidelines. In addition, manufacturers should provide certificates of compliance and analysis to be able to address at least strain identity, purity, viability at end of shelf life, and antibiotic susceptibility and resistance profiles.

Because of all of these potential safety and quality issues, we suggest that if a NICU is implementing probiotics as part of standard care, parents must be actively informed. Communication on the potential benefits and risks of probiotic administration is best undertaken face to face and supplemented with the use of written materials appropriate to the local context.

Recommendations
  1. The panel conditionally recommends that in case of implementing a probiotic product, the local microbiologists should be informed and they should confirm the ability to routinely detect probiotic bacteraemia/fungaemia with standard culture methods (very low certainty of evidence).
  2. The panel conditionally recommends not to provide probiotic strains, which produce D-lactate, as its potential risk or safety has not been adequately studied in preterm infants and remains uncertain (very low certainty of evidence).
  3. The panel conditionally recommends only the use of strains devoid of any plasmids containing transferable antibiotic resistance genes (very low certainty of evidence). This information should be confirmed and provided by the manufacturer.
  4. The panel conditionally recommends only the use of probiotic products manufactured according to cGMP to ensure correct strain identity with lack of contamination (very low certainty of evidence). Certificates of analysis should address at least strain identity, purity, viability, and antibiotic susceptibility and resistance profiles.
  5. The panel conditionally recommends to provide parents of preterm infants with sufficient information so they can understand the potential benefits and risks of probiotic administration (very low certainty of evidence). Communication is best undertaken face to face and supplemented with the use of written materials appropriate to the local context.

Should Probiotics Be Used in Preterm Infants? If Yes, Which Probiotics (Single or Combinations) Should be Used in What Dose?

The following probiotic strains (or combination of strains) fulfilled the criteria of being defined at strain level and were tested in at least 247 infants (per group) in RCTs: B breve BBG-001 (YIT4010), L reuteri DSM 17938, L rhamnosus GG ATCC 53103, S boulardii CNCM I-745, the combination of B bifidum NCDO 1453 with L acidophilus NCDO 1748 (ATCC 4356, LA37, or NCIMB 30316), and the combination of B infantis Bb-02, B lactis Bb-12, and Str thermophilus TH-4. In our previous NMA (39), we analysed results from the strains B lactis Bb-12 and B94 together, yielding reduced NEC rates. These results, however, were largely based on the single trial (101) that investigated the B94 strain in 200 infants, which is lower than the required power to assess sepsis. The B lactis Bb-12 strain was assessed in 219 infants (102–105) and did not result in reduced mortality or morbidity incidence, although the power was thus also lower than our threshold. As our aim was to give strain-specific recommendations, these 2 B lactis strains are, therefore, not further assessed. Other probiotic strains that have been previously studied in RCTs and were summarised in our prior NMA, but were not specified at strain level or did not reach the threshold of 247 infants in each group, are: Bacillus clausii (4 strains: O/C, N/R84, T84, and Sin8); Bacillus coagulans (previously L. sporogenes); combination of Ba subtilis R0179 and E faecium R0026; B bifidum OLB6378; combination of B bifidum, B infantis, B longum, and L acidophilus; combination of B bifidum, B lactis, B longum, and L acidophilus; B breve M-16 V, combination of B breve and L casei; combination of B infantis ATCC 15697 and L acidophilus ATCC 4356; combination of B infantis, L acidophilus, L casei, L plantarum, L rhamnosus, and Str thermophilus; combination of B infantis PTA-5843, E faecium PTA-5844, and L gasseri PTA-5845; combination of B lactis Bb-12 and B longum BB536; B longum BB536; combination of B longum BB536 and L rhamnosus GG; combination of B longum 35624 and L rhamnosus GG; combination of B longum R00175, L helveticus R0052, L rhamnosus R0011, and S boulardii CNCM I-1079; L acidophilus Lb; L acidophilus LA-5 (DSM 13241); and S boulardii CNCM I-3799. These probiotic strains are thus not discussed further.

The Following Strains (or Combinations of Strains) Have a Conditional Positive Recommendation

The GRADE evidence Table as to whether L rhamnosus GG (LGG) ATCC 53103 versus usual care should be used for preterm infants is depicted in Table 2. Mortality and sepsis did not show any clear direction in effect size, especially if the CrI is taken in consideration (very low and low certainty of evidence, respectively). Trials on sepsis (106–113) contained sufficient numbers of infants to rule out a significant beneficial effect of administration of L rhamnosus GG ATCC 53103, whereas the outcome on mortality was highly underpowered (106–108). The RR for NEC is, however, clearly reduced: 1507 infants studied in total (107–111,114); RR 0.240 (CrI 0.064--0.670); low certainty of evidence. Remarkably, both the control and intervention groups contained very few events (2.3% and 0.8%, respectively), whereas the NEC rate in the control groups of all 51 RCTs combined was 6.1% on average (39). In addition, mortality and sepsis rates in the studies evaluating L rhamnosus GG ATCC 53103 were very low. The reasons for the very low event rates in the control group could include the fact that relatively older infants were included in the 6 RCTs (mean GA ranged from 29–34 weeks; mean BW ranged from 1150–1950 g). Furthermore, in the study by Manzoni et al in 2009/2014, both the control and intervention groups received bovine lactoferrin in addition to either the placebo or LGG. This may explain the low event rates in the studies, even in the control groups, although a recent large RCT demonstrated no effect of enteral bovine lactoferrin supplemented solely (115). The number needed to treat is thus very high, despite a considerably low RR. Although our predefined sample size calculations predicted enough power with 431 infants in each arm, these calculations were performed with an expected NEC reduction from 10% to 5%. Thus, although enough infants were included in the 6 RCTs as defined in the method section, the observed reduction from 2.3% to 0.8% has only 63% power to predict a true effect in the reduction of NEC.

TABLE 2
TABLE 2:
GRADE table summarizing the evidence on the use of Lactobacillus rhamnosus GG ATCC 53013 compared with usual care in preterm infants

On the basis of the RCTs described above, the use of L rhamnosus GG ATCC 53103 at a dose ranging from 1 × 109 to 6 × 109 colony-forming units (CFU) may conditionally be recommended in preterm infants, as there is low quality evidence it might reduce NEC stage 2 or 3.

Considering evidence from non-RCTs, a pre-post cohort study in 221 infants on LGG ATCC 53103, weighing 900 g on average and who survived until discharge, could not clearly confirm a reduction in the NEC rate, as significance turned to P = 0.07 after adjusting for confounders (116). Another study compared morbidity and mortality rates after implementing simultaneous administration of both L rhamnosus GG together with bovine lactoferrin (50). In a timeframe of 11 years 835 infants, weighing approximately 1300 g on average at birth, NEC rates decreased from 3% to 1% after implementation of the combined strategy, whereas sepsis and mortality rates were unaltered. Remarkably, 2 other retrospective cohort studies reported higher NEC rates after implementing routine administration of L rhamnosus GG to very low birth weight (VLBW) infants. In 1 study, NEC rate (stage ≥2) amounted 3.2% out of 1900 infants without probiotics, and 4.6% out of 418 infants with LGG supplementation (117). In the other more recent cohort (465 infants without; 175 with LGG), NEC stage ≥2 incidence increased from 10% to 19% (118).

Recommendation

If all safety conditions are met, the panel conditionally recommends the use of L rhamnosus GG ATCC 53103 at a dose ranging from 1 × 109 CFU to 6 × 109 CFU as it might reduce NEC stage 2 or 3 (low certainty of evidence).

The GRADE evidence Table as to whether the combination of B infantis Bb-02, B lactis Bb-12, and Str thermophilus TH-4 versus usual care should be used for preterm infants is depicted in Table 3. Mortality and sepsis did not show any clear direction in effect size, especially if the CrI is taken in consideration (very low and low certainty of evidence, respectively). The administration of these 3 strains, however, did seem to significantly reduce rates of NEC stage 2 and 3 (RR 0.29 [0.073–0.78]). The evidence base was made up of 1 larger (119) and 1 smaller (120) RCT, with the inclusion of a total of 1244 infants with an average birth weight of approximately 1050 g.

TABLE 3
TABLE 3:
GRADE table summarizing the evidence on the use of the combination of Bifidobacterium infantis Bb-02, Bifidobacterium lactis Bb-12, and Streptococcus thermophilus TH-4 compared with usual care in preterm infants

Based on the RCTs described above, a conditional recommendation can be made for the use of a combination of B infantis Bb-02, B lactis Bb-12, and Str thermophilus TH-4 at a dose of 3.0 to 3.5 × 108 CFU (of each strain) in preterm infants as there is low-quality evidence it might reduce NEC stage 2 or 3.

However, a beneficial effect of these 3 strains (at a dose of 1.75 to 3.5 × 108 CFU of each strain) on reducing NEC could not be demonstrated in a retrospective cohort of 580 infants weighing approximately 1100 g on average at birth (121).

Recommendation

If all safety conditions are met, the panel conditionally recommends using the combination of B infantis Bb-02, B lactis Bb-12, and Str thermophilus TH-4 at a dose of 3.0 to 3.5 × 108 CFU (of each strain) as it might reduce NEC stage 2 or 3 (low certainty of evidence).

The Following Strains (or Combinations of Strains) Have a Conditional Neutral or Negative Recommendation

The GRADE evidence Table as to whether L reuteri DSM 17938 in a dose ranging from 4 × 107 to 2 × 108 CFU versus usual care should be used for preterm infants is depicted in Table 4. Previously, in our NMA we showed a significant reduction in NEC rates after combining the results from the L reuteri ATCC 55730 and DSM 17938 strains (1459 infants; 4 studies; RR 0.43 [0.16–0.98]). For mortality and sepsis rates, we could not demonstrate a reduction in our NMA. On the basis of panel discussions, however, we decided to omit the results from the single small study that used L reuteri ATCC 55730 (108) to be able to give truly strain-specific recommendations on the DSM 17938 strain, despite strains being very similar (82). In addition, in hindsight, one of the studies from our NMA also included stage 1 NEC, so we furthermore decided to exclude that small study as well for the NEC analysis only (122). On the other hand, three very recently published studies using the DSM 17938 strain could be added to our table (123–125). The GRADE evidence table, thus, does not represent previously published RRs from our NMA, but uses traditional RevMan forest plot-derived RRs (see also Supplemental Figure S1a--c, Supplemental Digital Content, http://links.lww.com/MPG/B785).

TABLE 4
TABLE 4:
GRADE table summarizing evidence on the use of L reuteri DSM 17938 compared to usual care in preterm infants

For none of the outcome domains, a irrefutably reduced event rate was noted, although the RR for reducing NEC stage ≥2 approached significance (RR 0.65 [95% CI 0.40--1.07]). If we would had added the results from the trial with the similar ATCC 55730 strain (82,108), results would not have been any different. It must be noted, however, that 4 studies investigating the use of L reuteri DSM 17938 included relatively larger preterm infants with average birth weights ranging from 1400 to 1700 g approximately (122,124–126). Remarkably, these RCTs showed most efficacious results from supplementing L reuteri DSM 17938, whereas in the 2 studies in which average birth weights amounted approximately 750 g (123) and 1050 g (127), NEC rates were not reduced (Supplemental Figure S1b, Supplemental Digital Content, http://links.lww.com/MPG/B785).

On the basis of the RCTs described above, no recommendation can be made in either direction for using L reuteri DSM 17938 at a dose ranging from 4 × 107 to 2 × 108 CFU in preterm infants (very low to low certainty of evidence).

The panel also noted 2 epoch cohort studies. The first analysed 311 infants (232 before and 79 after introduction) weighing on average 750 g and showed highly significant results, as the NEC rate decreased from 15.1% to 2.5% after L reuteri DSM 17938 administration (6 × 107) was routinely initiated (128). Sepsis rates were not different between both epochs. Another recent study including those born <33 weeks gestation compared 330 infants who did not receive probiotics to 1027 infants who received L reuteri DSM 17938 after a policy change (129). NEC rates were significantly reduced amongst all subgroups (also those <26 weeks), but nosocomial sepsis and mortality rates were unaltered.

Recommendation

The panel concludes that no recommendation can be made in either direction regarding the use of L reuteri DSM 17938 in preterm infants to reduce the risk of mortality, NEC stage 2 or 3, or sepsis (very low certainty of evidence). Additionally, L reuteri DSM 17938 is a partially D-lactate-producing strain for which there is insufficient safety data available in preterm infants.

The GRADE evidence Table as to whether the combination of B bifidum NCDO 1453 (currently reclassified as B longum) with L acidophilus NCDO 1748 (ATCC 4356, LA37, or NCIMB 30316) versus usual care should be used for preterm infants is based on 2 studies (130,131) and depicted in Table 5. Very low certainty evidence showed that mortality rates were lower in the probiotics group. Yet, NEC rates only showed a trend towards reduced risk, whereas the point estimate for sepsis rates showed an increased risk.

TABLE 5
TABLE 5:
GRADE table summarizing the evidence on the use of the combination of Bifidobacterium bifidum NCDO 1453 and L acidophilus NCDO 1748 (ATCC 4356, LA37, or NCIMB 30316) compared with usual care in preterm infants

Based on the RCTs described above, no recommendation can be made in either direction for using the combination of B bifidum NCDO 1453 with L acidophilus NCDO 1748 (ATCC 4356, LA37, or NCIMB 30316) at a dose of 1 × 109 CFU (of each strain) in preterm infants (based upon very low-to-moderate certainty of evidence).

Evidence from 2 recent nonrandomised trials show conflicting results. A large pre-post implementation cohort study (n = 1288 before and n = 673 after) that used these 2 strains found no reduction in rates of mortality, NEC, or sepsis after correction for confounders (132). A study, however, with similar design and strains (n = 170 before and 3 = 346 after) found a doubling of NEC rates after implementation, but a 16% reduction in late-onset sepsis rates (133).

Recommendation

The panel concludes that no recommendation can be made in either direction regarding the use of the combination of B bifidum NCDO 1453 (currently reclassified as B longum) with L acidophilus NCDO 1748 (ATCC 4356, LA37, or NCIMB 30316) in preterm infants to reduce the risk of mortality, NEC stage 2 or 3, or sepsis (very low-to-moderate certainty of evidence). Additionally, L acidophilus NCDO 1748 (ATCC 4356, LA37, or NCIMB 30316) is a partially D-lactate-producing strain for which there is insufficient safety data available in preterm infants.

The GRADE evidence table as to whether B breve BBG-001 (YIT4010) in a dose of 7 × 108 CFU versus usual care should be used for preterm infants is depicted in Table 6. There appears no clear direction in effect on any of the described outcomes (mortality, NEC stage ≥2, and sepsis). The evidence is derived from a single, large, well performed RCT in 1310 infants with a median GA of 28 weeks and higher than average event rates of NEC and sepsis (37).

Recommendation

The panel conditionally recommends against using B breve BBG-001 to reduce the risk of mortality, NEC stage 2 or 3, or sepsis (low-to-moderate certainty of evidence).

TABLE 6
TABLE 6:
GRADE table summarizing the evidence on the use of Bifidobacterium breve BBG-001 compared with usual care in preterm infants

The GRADE evidence Table as to whether S boulardii CNCM I-745 versus usual care should be used for preterm infants is depicted in Table 7(134–138). None of the 3 outcomes show a clear direction of effect when credible intervals are considered, although both the outcomes on mortality and NEC were underpowered.

TABLE 7
TABLE 7:
GRADE table summarizing the evidence on the use of Saccharomyces boulardii CNCM I-745 compared with usual care in preterm infants

On the basis of the RCTs described above, no recommendation can be made in either direction for using S boulardii CNCM I-745 at a dose ranging from 1 × 109 to 5 × 109 CFU in preterm infants (based upon very low to low certainty of evidence).

We found only 1 small cohort study that included only preterm infants with birth weight between 1 and 2 kg and in which this strain was investigated (139). Mortality and NEC stage 2 rate amounted 10.3% and 7.7%, respectively, in the 39 infants without probiotics and 0% in the 46 infants who had received the S boulardii.

Regarding safety, the European Medicine Agency recently amended a contra-indication to the use of S boulardii in patients with a central venous catheter, in critically ill patients, or in immunocompromised patients because of a risk of fungaemia (64).

Recommendation

The panel does not recommend the routine use of S boulardii for safety reasons (in line with the position of the European Medicine Agency which contraindicates the use of S boulardii in patients with a central venous catheter, in critically ill patients, or in immunocompromised patients because of a risk of fungaemia) as well as lack of evidence of efficacy (very low to low certainty of evidence).

Are Combinations of Species More Effective Than the Use of a Single Strain to Reduce the Risk of NEC (Stage 2 or 3)?

Several classic meta-analyses have shown decreased morbidity rates after supplementing with multiple strains versus a single strain (18,21–23). These meta-analyses, however, were not genus-specific, species-specific, or strain-specific. Therefore, it is not appropriate to extrapolate or determine whether the beneficial effect in the ‘multiple strain’ group was because of the chance that more effective strains were used in that group versus the strains used in the ‘single strain’ group. It mainly comes down to which strain is used. Use of a single strain with proven effectiveness is likely to be more efficacious than use of a combination of strains without proven effectiveness. On the other hand, a combination of 2 or more independently proven efficacious strains may be more efficacious than a single efficacious strain, provided no antagonistic mechanisms exists.

Figure 1 shows additional models run from the database in our NMA (39) to gather more formal evidence. It is shown, that from all neonatal trials combined, there is no a priori advantage of administering multiple strains versus a single strain. Also, there appears to be no benefit of selecting a specific genus (Bifidobacterium or Lactobacillus) or a combination of these 2.

FIGURE 1
FIGURE 1:
Relative effect plots depicting risk ratios on reducing necrotising enterocolitis stage 2 or 3 after supplementing (A) a single probiotic strain, 2 strains, or 3 or more strains, versus placebo care; and (B) one or more Bifidobacterium strains, one or more Lactobacillus strains, or a combination of the 2 versus placebo care.

Thus, these data, although not-strain specific, do not support the notion that administration of multiple strains or combinations of species (from a different genus) is more effective than the administration of a single probiotic strain.

Recommendation

The panel conditionally recommends that when considering the use of probiotics, a strain (or combination of strains) with proven effectiveness and established safety profile should be selected, rather than focussing on administering multiple strains from different genera (very low certainty of evidence).

Which Dose of a Probiotic Strain or Combination of Strains Should Be Administered?

The administered dose of probiotic strains used in premature neonates differed widely, as can be seen in our database of 51 RCTs (39). Usually, doses were in the range of 108 to 109 CFU, but doses ranging from as low as 105 CFU (140) to as high as 1010 CFU (103) have also been used. Even between trials investigating the same strain, administered doses varied widely in different trials. In the 6 trials that studied the effects of the strain B lactis Bb-12 (102–105,119,120), the supplemented dose differed 600-fold, ranging from 2.0 × 107 to 1.2 × 1010 CFU. For other well-studied strains, the administered doses differed less; from 2 × 108 to 6 × 109 CFU (107–114,141,142) for L rhamnosus GG, from 4 × 107 to 2 × 108 CFU (122–127,143) for L reuteri DSM 17938, and from 1 × 109 to 5 × 109 CFU for S boulardii I-745 (134–138).

A small trial (n = 149) in total showed no clear differences in colonization between dosing either 1 × 109 or 1 × 1010 CFU of the same strain daily (144). Another trial, however, showed that administration of a daily dose of 1 × 109 CFU of 2 probiotic strains was more effective in terms of colonization than a weekly or bi-weekly dose of the same strains (145).

Recently, a systematic review was published on dose-responses of probiotics in different clinical settings and patient groups (146). Only for antibiotic-associated diarrhoea was a dose-response observed, although this was not strain-specific and analysed with all probiotic strains simultaneously. For NEC, no such relation could be demonstrated in preterm infants. The author, however, also notes that this issue is highly understudied throughout all clinical settings.

Apart from designated doses on the product label, it is well-known that actual viable bacterial counts are frequently much lower, sometimes only a few percent of what is claimed on the product packaging (80,96). Suppliers of probiotics should thus always provide reports on the number of viable bacterial counts in their product including a stability analysis. In addition, probiotic viability is highly affected whether it is dissolved in water, breast milk or formula (147). In conclusion, data do not support the notion that a higher dose of probiotics is more effective than a lower dose and the optimal dose for most species and strains remains undetermined.

Recommendation

The panel conditionally recommends that, if probiotics are administered, to use similar doses as applied in relevant RCTs (very low certainty of evidence). Probiotic products should be accompanied with formal quality reports that ascertain product viability until the end of shelf life.

What Should be the Duration of Administering Probiotics?

This issue has not been systematically researched. The times after birth at which probiotics are started vary widely, as well as the total duration of probiotic administration (39). Several studies started probiotics immediately after birth, whereas others waited for up to a week after birth. In some studies, probiotic administration was stopped after 2 weeks. In most studies, however, probiotic administration lasted 4 to 6 weeks or up until discharge.

With strain-proven efficacy, it would be common sense to administer probiotics before and during the period when NEC risk is the highest, so relatively fast following birth. Yet, it is unknown if very early administration of a high dose of 1 or more probiotic strains might be harmful, when ‘natural’ (breast milk driven) colonisation is just beginning, and when the immune system is underdeveloped, and gastrointestinal barrier function is impaired by inadequate tight junctions and reduced mucus layer.

Data do not provide clear evidence as to when probiotic supplementation should be started or ceased. The rationale though does exist, that prolonged use may prevent more ‘natural’ colonisation and/or that the risk:benefit ratio might be the lowest when used in the period when NEC risk is the highest.

Recommendation

The available data do not clearly indicate an optimal start or length of treatment. The panel conditionally recommends individual units determine treatment duration based on the population who will receive them and their ongoing risk of diseases, such as NEC (very low certainty of evidence).

Is it Appropriate to Administer Other Strains Than Those Studied in Large Well-conducted Randomised Controlled Trials?

It has been suggested that based on the consistently decreased risk of NEC in RCTs using variable probiotic regimens, it is time we accept that commonly used probiotic strains share pathways of benefits providing ‘non-specific’ protection (27,38). Our recent NMA clearly, however, shows that the results of RCTs on different probiotic strains largely differ with regard to the 3 analysed outcomes (39). Whether this is truly a reflection of strain-specific benefits (8,32,33,148), internal and external study validity, or a power issue remains to be elucidated. Considering the vulnerable patient population, and aforementioned potential safety and product quality issues, however, only high-quality, safe, and evidence-based strains can be recommended for clinical use.

Recommendation

The panel conditionally recommends that in the clinical setting, the use of a single strain or combination of strains should be practise-based on positive results from well-conducted RCTs (very low certainty of evidence). In research settings, however, it is appropriate to test new strains or new combinations of strains.

DISCUSSION

The gastrointestinal-related intervention that is both the most safe and efficacious in reducing morbidity and mortality would absolutely be to stimulate the use of unpasteurised own mother's milk. However, especially in NICUs with a high NEC incidence, the use of prophylactic probiotic therapy might be considered as well. This position paper aimed to provide some guidance on which probiotic strains have proven efficacy while addressing safety issues as well. Others have also come up with the 10 golden rules of safe introduction of probiotics (100). Unfortunately, current available evidence appears only marginally enough to conditionally recommend 1 or 2 therapeutic options that are evidence-based on RCTs. We only advise the routine use of certain strains of probiotics that have been shown to be safe and efficacious and that have been studied in a large number of VLBW infants. Thus, there is still a need for well-designed and carefully conducted RCTs, with relevant inclusion/exclusion criteria and adequate sample sizes. We specifically encourage undertaking trials that aim to include extremely premature infants (particularly those <26 weeks gestation), as these infants are relatively understudied so far. Whilst these infants have the highest risk of NEC, the risk of harm from probiotics might be the greatest as well. Such trials should define the optimal doses and intake durations, as well as providing more information about the long-term safety of probiotics. Probiotic products that are used should be submitted to systematic quality control procedures by the respective authorities to confirm the viability and identify the strain-level(s) of the active ingredient(s). As most of the trials published so far have been company-funded, independent trials, preferentially financed jointly by national/governmental/European Union bodies and other international organisations, would be desirable. Finally, long-term follow-up is warranted, not only from a neurodevelopmental perspective (24) but also regarding safety and immunity (35).

Another major problem in many of the RCTs is the definition of NEC. Probably, only surgically proven NEC is a reliable outcome and this should always be separately reported in future trials. In quite a few of the trials, blinding is an important issue and stage 2 NEC is not an exact diagnosis.

Other open questions not addressed here and in many studies are the optimal matrix of the probiotic supplement (powder, capsules, or liquids) as well as the concomitant feeding strategy (own or donor human milk or formula), despite the fact that they may affect outcomes. The times at which probiotics are added to either human milk or formula could affect strain viability at the time of ingestion, for example (147,149). Other reviews have suggested that probiotics might be more effective in infants fed human milk, rather than preterm formula (9,16,22), despite that human milk itself already lowers the incidence of sepsis and NEC. Whether this finding is a coincidence of having clustered nonefficacious strains in formula-fed infants versus more efficacious strains in the human milk-fed group, or whether there is a biological rationale remains unknown. One of the explanations could be that the human milk-fed infants respond better to probiotics because of the fact that only human milk contain human milk oligosaccharides (HMOs) from which Bifidobacteriae benefit, especially B infantis(150). Whether there is a further potential difference in effectiveness between own mother's milk or donor milk remains unknown. Although donor milk still contains HMOs, all beneficial bacteria that fresh human milk normally harbours (151) are destroyed in the process of pasteurization. On the other hand, mothers with antibiotics have less Bifidobacteriae in their milk as well (152). Yet, in a recent Cochrane, it was not recommended for mothers of preterm infants to use probiotics (153).

Exciting new areas of research are the study of killed (ghost) probiotics or closely related postbiotics, which might still harbour beneficial immunological effects but eliminates the risk of, for example, sepsis or contamination (154–157).

RESEARCH GAPS

The following additional clinical and research questions were also posed and voted upon with high agreement (>85%):

  1. Placebo-controlled studies on promising specific strains for different outcomes are still needed, as no single strain has been studied in individual adequately powered studies. These studies could be conducted by head-to-head comparisons in trials that include a placebo arm.
  2. Appropriately designed and powered studies that determine the optimal dosing, optimal time of initiation, and duration treatment of effective probiotics are needed.
  3. The number of extremely preterm infants (<28 weeks GA) and infants with a birth weight below 1000 g included in the current studies is limited, whereas NEC and mortality rates are the highest in that population. Studies specifically focussed on these groups are needed. Within the future studies, stratification should be based on the quality of the enteral feeding (own mother's milk, donor milk, or formula).
  4. The efficacy and safety of different modes of administration (powder, liquid, added to formula by manufacturer) should be a topic of investigation.
  5. Long-term safety including the effects of probiotic administration on metabolic, endocrine, immunological, and behavioural parameters should be a topic of investigation.
  6. In-hospital safety of used probiotics should be assessed by determination of “probiotic sepsis rates” by a microbiology department that is equipped to evaluate these infections.
  7. Attention should be paid to characteristics of the population studied. Gender, ethnicity, region of birth, composition of diet, and antibiotic use are just a few factors that might have an impact on the safety and efficacy of specific strains.

Recommendations

  1. The panel conditionally recommends that in case of implementing a probiotic product, the local microbiologists should be informed and they should confirm the ability to routinely detect probiotic bacteraemia/fungaemia with standard culture methods (very low certainty of evidence).
  2. The panel conditionally recommends not to provide probiotic strains, which produce D-lactate, as its potential risk or safety has not been adequately studied in preterm infants and remains uncertain (very low certainty of evidence).
  3. The panel conditionally recommends only the use of strains devoid of any plasmids containing transferable antibiotic resistance genes (very low certainty of evidence). This information should be confirmed and provided by the manufacturer.
  4. The panel conditionally recommends only the use of probiotic products manufactured according to cGMP to ensure correct strain identity with lack of contamination (very low certainty of evidence). Certificates of analysis should address at least strain identity, purity, viability, and antibiotic susceptibility, and resistance profiles.
  5. The panel conditionally recommends to provide parents of preterm infants with sufficient information so they can understand the potential benefits and risks of probiotic administration (very low certainty of evidence). Communication is best undertaken face to face and supplemented with the use of written materials appropriate to the local context.
  6. If all safety conditions are met, the panel conditionally recommends the use of L rhamnosus GG ATCC 53103 at a dose ranging from 1 × 109 CFU to 6 × 109 CFU as it might reduce NEC stage 2 or 3 (low certainty of evidence).
  7. If all safety conditions are met, the panel conditionally recommends using the combination of B infantis Bb-02, B lactis Bb-12, and Str thermophilus TH-4 at a dose of 3.0 to 3.5 × 108 CFU (of each strain) as it might reduce NEC stage 2 or 3 (low certainty of evidence).
  8. The panel concludes that no recommendation can be made in either direction regarding the use of L reuteri DSM 17938 in preterm infants to reduce the risk of mortality, NEC stage 2 or 3, or sepsis (very low certainty of evidence). Additionally, L reuteri DSM 17938 is a partially D-lactate producing strain for which there is insufficient safety data available in preterm infants.
  9. The panel concludes that no recommendation can be made in either direction regarding the use of the combination of B bifidum NCDO 1453 (currently reclassified as B longum) with L acidophilus NCDO 1748 (ATCC 4356, LA37, or NCIMB 30316) in preterm infants to reduce the risk of mortality, NEC stage 2 or 3, or sepsis (very low to moderate certainty of evidence). Additionally, L acidophilus NCDO 1748 (ATCC 4356, LA37, or NCIMB 30316) is a partially D-lactate-producing strain for which there is insufficient safety data available in preterm infants.
  10. The panel conditionally recommends against using B breve BBG-001 to reduce the risk of mortality, NEC stage 2 or 3, or sepsis (low-to-moderate certainty of evidence).
  11. The panel does not recommend the routine use of S boulardii for safety reasons (in line with the position of the European Medicine Agency, which contraindicates the use of S boulardii in patients with a central venous catheter, in critically ill patients, or in immunocompromised patients because of a risk of fungaemia) as well as lack of evidence of efficacy (very low to low certainty of evidence).
  12. The panel conditionally recommends that whenever considering the use of probiotics, a strain (or combination of strains) with proven effectiveness and established safety profile should be selected, rather than focussing on administering multiple strains from different genera (very low certainty of evidence).
  13. The panel conditionally recommends that, if probiotics are administered, to use similar doses as applied in relevant RCTs (very low certainty of evidence). Probiotic products should be accompanied with formal quality reports that ascertain product viability until the end of shelf life.
  14. The available data do not clearly indicate an optimal start or length of treatment. The panel conditionally recommends individual units determine treatment duration based on the population who will receive them and their ongoing risk of diseases, such as NEC (very low certainty of evidence).
  15. The panel conditionally recommends that in the clinical setting, the use of a single strain or combination of strains should be practise-based on positive results from well-conducted RCTs (very low certainty of evidence). In research settings, however, it is appropriate to test new strains or new combinations of strains.

REFERENCES

1. Isani MA, Delaplain PT, Grishin A, et al. Evolving understanding of neonatal necrotizing enterocolitis. Curr Opin Pediatr 2018; 30:417–423.
2. Schreurs R, Baumdick ME, Sagebiel AF, et al. Human fetal TNF-alpha-cytokine-producing CD4(+) effector memory T cells promote intestinal development and mediate inflammation early in life. Immunity 2019; 50:462.e8–476.e8.
3. Battersby C, Modi N. Challenges in advancing necrotizing enterocolitis research. Clin Perinatol 2019; 46:19–27.
4. Corpeleijn WE, Kouwenhoven SM, Paap MC, et al. Intake of own mother's milk during the first days of life is associated with decreased morbidity and mortality in very low birth weight infants during the first 60 days of life. Neonatology 2012; 102:276–281.
5. Bermudez-Brito M, Plaza-Diaz J, Munoz-Quezada S, et al. Probiotic mechanisms of action. Ann Nutr Metab 2012; 61:160–174.
6. Hidalgo-Cantabrana C, Delgado S, Ruiz L, et al. Bifidobacteria and their health-promoting effects. Microbiol Spectr 2017; 5:1–19.
7. Hill C, Guarner F, Reid G, et al. Expert consensus document. The International Scientific Association for Probiotics and Prebiotics consensus statement on the scope and appropriate use of the term probiotic. Nat Rev Gastroenterol Hepatol 2014; 11:506–514.
8. Plaza-Diaz J, Ruiz-Ojeda FJ, Gil-Campos M, et al. Mechanisms of action of probiotics. Adv Nutr 2019; 10: (Suppl_1): S49–S66.
9. Aceti A, Gori D, Barone G, et al. Probiotics and time to achieve full enteral feeding in human milk-fed and formula-fed preterm infants: systematic review and meta-analysis. Nutrients 2016; 8:pii: E471.
10. Baucells BJ, Mercadal Hally M, Alvarez Sanchez AT, et al. Probiotic associations in the prevention of necrotising enterocolitis and the reduction of late-onset sepsis and neonatal mortality in preterm infants under 1,500 g: a systematic review. An Pediatr (Barc) 2016; 85:247–255.
11. Billimoria ZC, Pandya S, Bhatt P, et al. Probiotics-to use, or not to use? An updated meta-analysis. Clin Pediatr (Phila) 2016; 55:1242–1244.
12. Hu HJ, Zhang GQ, Zhang Q, et al. Probiotics prevent candida colonization and invasive fungal sepsis in preterm neonates: a systematic review and meta-analysis of randomized controlled trials. Pediatr Neonatol 2017; 58:103–110.
13. Rao SC, Athalye-Jape GK, Deshpande GC, et al. Probiotic supplementation and late-onset sepsis in preterm infants: a meta-analysis. Pediatrics 2016; 137:e20153684.
14. Sawh SC, Deshpande S, Jansen S, et al. Prevention of necrotizing enterocolitis with probiotics: a systematic review and meta-analysis. PeerJ 2016; 4 (e2429):1–29.
15. Zhang GQ, Hu HJ, Liu CY, et al. Probiotics for preventing late-onset sepsis in preterm neonates: a PRISMA-compliant systematic review and meta-analysis of randomized controlled trials. Medicine (Baltimore) 2016; 95:e2581.
16. Aceti A, Maggio L, Beghetti I, et al. Italian Society of Neonatology. Probiotics prevent late-onset sepsis in human milk-fed, very low birth weight preterm infants: systematic review and meta-analysis. Nutrients 2017; 9: pii: E904.
17. Cavallaro G, Villamor-Martinez E, Filippi L, et al. Probiotic supplementation in preterm infants does not affect the risk of retinopathy of prematurity: a meta-analysis of randomized controlled trials. Sci Rep 2017; 7:13014.
18. Chang HY, Chen JH, Chang JH, et al. Multiple strains probiotics appear to be the most effective probiotics in the prevention of necrotizing enterocolitis and mortality: an updated meta-analysis. PLoS One 2017; 12:e0171579.
19. Dermyshi E, Wang Y, Yan C, et al. The “golden age” of probiotics: a systematic review and meta-analysis of randomized and observational studies in preterm infants. Neonatology 2017; 112:9–23.
20. Rees CM, Hall NJ, Fleming P, et al. Probiotics for the prevention of surgical necrotising enterocolitis: systematic review and meta-analysis. BMJ Paediatr Open 2017; 1:e000066.
21. Sun J, Marwah G, Westgarth M, et al. Effects of probiotics on necrotizing enterocolitis, sepsis, intraventricular hemorrhage, mortality, length of hospital stay, and weight gain in very preterm infants: a meta-analysis. Adv Nutr 2017; 8:749–763.
22. Thomas JP, Raine T, Reddy S, et al. Probiotics for the prevention of necrotising enterocolitis in very low-birth-weight infants: a meta-analysis and systematic review. Acta Paediatr 2017; 106:1729–1741.
23. Villamorse-Martinez E, Pierro M, Cavallaro G, et al. Probiotic Supplementation in Preterm Infants Does Not Affect the Risk of Bronchopulmonary Dysplasia: A Meta-Analysis of Randomized Controlled Trials. Nutrients 2017; 9:
24. Upadhyay RP, Taneja S, Chowdhury R, et al. Effect of prebiotic and probiotic supplementation on neurodevelopment in preterm very low birth weight infants: findings from a meta-analysis. Pediatr Res 2018; [Epub ahead of print].
25. Ofek Shlomai N, Deshpande G, Rao S, et al. Probiotics for preterm neonates: what will it take to change clinical practice? Neonatology 2014; 105:64–70.
26. Tarnow-Mordi WO, Wilkinson D, Trivedi A, et al. Probiotics reduce all-cause mortality and necrotizing enterocolitis: it is time to change practice. Pediatrics 2010; 125:1068–1070.
27. Athalye-Jape G, Patole S. Probiotics for preterm infants - time to end all controversies. Microb Biotechnol 2019; 12:249–253.
28. Underwood MA. Arguments for routine administration of probiotics for NEC prevention. Curr Opin Pediatr 2019; 31:188–194.
29. Agostoni C, Buonocore G, Carnielli VP, et al. Enteral nutrient supply for preterm infants: commentary from the European Society of Paediatric Gastroenterology, Hepatology and Nutrition Committee on Nutrition. J Pediatr Gastroenterol Nutr 2010; 50:85–91.
30. Mihatsch WA, Braegger CP, Decsi T, et al. Critical systematic review of the level of evidence for routine use of probiotics for reduction of mortality and prevention of necrotizing enterocolitis and sepsis in preterm infants. Clin Nutr 2012; 31:6–15.
31. Thomas DW, Greer FR, et al. American Academy of Pediatrics Committee on Nutrition. Probiotics and prebiotics in pediatrics. Pediatrics 2010; 126:1217–1231.
32. Pell LG, Loutet MG, Roth DE, et al. Arguments against routine administration of probiotics for NEC prevention. Curr Opin Pediatr 2019; 31:195–201.
33. McFarland LV, Evans CT, Goldstein EJC. Strain-specificity and disease-specificity of probiotic efficacy: a systematic review and meta-analysis. Front Med (Lausanne) 2018; 5:124.
34. Berrington JE, Zalewski S. The future of probiotics in the preterm infant. Early Hum Dev 2019; 135:75–81.
35. Suez J, Zmora N, Segal E, et al. The pros, cons, and many unknowns of probiotics. Nat Med 2019; 25:716–729.
36. Lerner A, Shoenfeld Y, Matthias T. Probiotics: if it does not help it does not do any harm. really? Microorganisms 2019; 7:104.
37. Costeloe K, Hardy P, Juszczak E, et al. Bifidobacterium breve BBG-001 in very preterm infants: a randomised controlled phase 3 trial. Lancet 2016; 387:649–660.
38. Sanders ME, Benson A, Lebeer S, et al. Shared mechanisms among probiotic taxa: implications for general probiotic claims. Curr Opin Biotechnol 2018; 49:207–216.
39. van den Akker CHP, van Goudoever JB, Szajewska H, et al. ESPGHAN Working Group for Probiotics, Prebiotics; Committee on Nutrition. Probiotics for preterm infants: a strain-specific systematic review and network meta-analysis. J Pediatr Gastroenterol Nutr 2018; 67:103–122.
40. GRADEpro GDT: GRADEpro Guideline Development Tool [Software]. 2015 Developed by Evidence Prime, Inc. Available from gradepro.org. [Computer Program]. McMaster University.
41. European Food Safety Authority (EFSA). Scientific Opinion on the substantiation of health claims related to Bifidobacterium animalis ssp. lactis Bb-12. EFSA J 2011; 9:2047–2066.
42. Doron S, Snydman DR. Risk and safety of probiotics. Clin Infect Dis 2015; 60: (Suppl 2): S129–S134.
43. Kothari D, Patel S, Kim SK. Probiotic supplements might not be universally-effective and safe: a review. Biomed Pharmacother 2019; 111:537–547.
44. Bafeta A, Koh M, Riveros C, et al. Harms reporting in randomized controlled trials of interventions aimed at modifying microbiota: a systematic review. Ann Intern Med 2018; 169:240–247.
45. Costa RL, Moreira J, Lorenzo A, et al. Infectious complications following probiotic ingestion: a potentially underestimated problem? A systematic review of reports and case series. BMC Complement Altern Med 2018; 18:329.
46. Zbinden A, Zbinden R, Berger C, et al. Case series of Bifidobacterium longum bacteremia in three preterm infants on probiotic therapy. Neonatology 2015; 107:56–59.
47. Bertelli C, Pillonel T, Torregrossa A, et al. Bifidobacterium longum bacteremia in preterm infants receiving probiotics. Clin Infect Dis 2015; 60:924–927.
48. Esaiassen E, Cavanagh P, Hjerde E, et al. Bifidobacterium longum Subspecies infantis bacteremia in 3 extremely preterm infants receiving probiotics. Emerg Infect Dis 2016; 22:1664–1666.
49. Jenke A, Ruf EM, Hoppe T, et al. Bifidobacterium septicaemia in an extremely low-birthweight infant under probiotic therapy. Arch Dis Child Fetal Neonatal Ed 2012; 97:F217–F218.
50. Meyer MP, Alexander T. Reduction in necrotizing enterocolitis and improved outcomes in preterm infants following routine supplementation with Lactobacillus GG in combination with bovine lactoferrin. J Neonatal Perinatal Med 2017; 10:249–255.
51. Brecht M, Garg A, Longstaff K, et al. Lactobacillus sepsis following a laparotomy in a preterm infant: a note of caution. Neonatology 2016; 109:186–189.
52. Kunz AN, Noel JM, Fairchok MP. Two cases of Lactobacillus bacteremia during probiotic treatment of short gut syndrome. J Pediatr Gastroenterol Nutr 2004; 38:457–458.
53. Dani C, Coviello CC, Corsini II, et al. Lactobacillus sepsis and probiotic therapy in newborns: two new cases and literature review. AJP Rep 2016; 6:e25–e29.
54. Sadowska-Krawczenko I, Paprzycka M, Korbal P, et al. Lactobacillus rhamnosus GG suspected infection in a newborn with intrauterine growth restriction. Benef Microbes 2014; 5:397–402.
55. Cavicchiolo ME, Magnani M, Calgaro S, et al. Neonatal sepsis associated with Lactobacillus supplementation. J Perinat Med 2019; 48:87–88.
56. Celis Castaneda LA, Morales Camacho WJ, Duran Ochoa NM. Sepsis due to Lactobacillus reuteri in an extreme preterm newborn: case report. Arch Argent Pediatr 2019; 117:e509–e513.
57. Roy U, Jessani LG, Rudramurthy SM, et al. Seven cases of Saccharomyces fungaemia related to use of probiotics. Mycoses 2017; 60:375–380.
58. Lungarotti MS, Mezzetti D, Radicioni M. Methaemoglobinaemia with concurrent blood isolation of Saccharomyces and Candida. Arch Dis Child Fetal Neonatal Ed 2003; 88:F446.
59. Ohishi A, Takahashi S, Ito Y, et al. Bifidobacterium septicemia associated with postoperative probiotic therapy in a neonate with omphalocele. J Pediatr 2010; 156:679–681.
60. Guenther K, Straube E, Pfister W, et al. Sever sepsis after probiotic treatment with Escherichia coli NISSLE 1917. Pediatr Infect Dis J 2010; 29:188–189.
61. Hickey L, Garland SM, Jacobs SE, et al. Cross-colonization of infants with probiotic organisms in a neonatal unit. J Hosp Infect 2014; 88:226–229.
62. Gengaimuthu K. The cross contamination (cross colonization) phenomenon of probiotic use in neonatal intensive care units: putative mechanisms and clinical and research implications. Cureus 2018; 10:e2691.
63. Yelin I, Flett KB, Merakou C, et al. Genomic and epidemiological evidence of bacterial transmission from probiotic capsule to blood in ICU patients. Nat Med 2019; 25:1728–1732.
64. EMA (European Medicines Agency). PSUSA/00009284/201702: Saccharomyces boulardii: CMDh (Coordination Group for Mutual Recognition and Decentralised Procedures - Human) scientific conclusions and grounds for the variation, amendments to the product information and timetable for the implementation. 2017.
65. Mack DR. D(-)-lactic acid-producing probiotics, D(-)-lactic acidosis and infants. Can J Gastroenterol 2004; 18:671–675.
66. Connolly E, Abrahamsson T, Bjorksten B. Safety of D(-)-lactic acid producing bacteria in the human infant. J Pediatr Gastroenterol Nutr 2005; 41:489–492.
67. Lukasik J, Salminen S, Szajewska H. Rapid review shows that probiotics and fermented infant formulas do not cause d-lactic acidosis in healthy children. Acta Paediatr 2018; 107:1322–1326.
68. Vitetta L, Coulson S, Thomsen M, et al. Probiotics, D-Lactic acidosis, oxidative stress and strain specificity. Gut Microbes 2017; 8:311–322.
69. Papagaroufalis K, Fotiou A, Egli D, et al. A randomized double blind controlled safety trial evaluating d-lactic acid production in healthy infants fed a Lactobacillus reuteri-containing formula. Nutr Metab Insights 2014; 7:19–27.
70. Reddy VS, Patole SK, Rao S. Role of probiotics in short bowel syndrome in infants and children--a systematic review. Nutrients 2013; 5:679–699.
71. Munakata S, Arakawa C, Kohira R, et al. A case of D-lactic acid encephalopathy associated with use of probiotics. Brain Dev 2010; 32:691–694.
72. FAO/WHO Codex Alimentarius Commission; International Food Standards. Codex Stan 72-1981: Standard for Infant Formula and Formulas for Special Medical Purposes Intended for Infants. 2016:1-17.
73. FDA. GRAS Notification: Lactobacillus reuteri strain DSM 17938. 2012:GRN 000410.
74. Borghi E, Massa V, Severgnini M, et al. Antenatal microbial colonization of mammalian gut. Reprod Sci 2018; 26:1045–1053.
75. Perez-Munoz ME, Arrieta MC, Ramer-Tait AE, et al. A critical assessment of the “sterile womb” and “in utero colonization” hypotheses: implications for research on the pioneer infant microbiome. Microbiome 2017; 5:48.
76. Korpela K, Blakstad EW, Moltu SJ, et al. Intestinal microbiota development and gestational age in preterm neonates. Sci Rep 2018; 8:2453.
77. Underwood MA. Probiotics and innate and adaptive immune responses in premature infants. For Immunopathol Dis Therap 2016; 7:1–15.
78. Hu Y, Yang X, Qin J, et al. Metagenome-wide analysis of antibiotic resistance genes in a large cohort of human gut microbiota. Nat Commun 2013; 4:2151.
79. Lerner A, Matthias T, Aminov R. Potential effects of horizontal gene exchange in the human gut. Front Immunol 2017; 8:1630.
80. Wong A, Ngu DYS, Dan LA, et al. Detection of antibiotic resistance in probiotics of dietary supplements. Nutr J 2015; 14:95.
81. Sharma P, Tomar SK, Goswami P, et al. Antibiotic resistance among commercially available probiotics. Food Res Int 2014; 57:176–195.
82. Rosander A, Connolly E, Roos S. Removal of antibiotic resistance gene-carrying plasmids from Lactobacillus reuteri ATCC 55730 and characterization of the resulting daughter strain, L reuteri DSM 17938. Appl Environ Microbiol 2008; 74:6032–6040.
83. Wei YX, Zhang ZY, Liu C, et al. Safety assessment of Bifidobacterium longum JDM301 based on complete genome sequences. World J Gastroenterol 2012; 18:479–488.
84. Gevers D, Huys G, Swings J. In vitro conjugal transfer of tetracycline resistance from Lactobacillus isolates to other Gram-positive bacteria. FEMS Microbiol Lett 2003; 225:125–130.
85. Gueimonde M, Sanchez B, C GdLR-G, et al. Antibiotic resistance in probiotic bacteria. Front Microbiol 2013; 4:202.
86. Zheng M, Zhang R, Tian X, et al. Assessing the risk of probiotic dietary supplements in the context of antibiotic resistance. Front Microbiol 2017; 8:908.
87. Imperial IC, Ibana JA. Addressing the antibiotic resistance problem with probiotics: reducing the risk of its double-edged sword effect. Front Microbiol 1983; 7:1–10.
88. Hagbo M, Ravi A, Angell IL, et al. Experimental support for multidrug resistance transfer potential in the preterm infant gut microbiota. Pediatr Res 2019; [Epub ahead of print].
89. Topcuoglu S, Gursoy T, Ovali F, et al. A new risk factor for neonatal vancomycin-resistant Enterococcus colonisation: bacterial probiotics. J Matern Fetal Neonatal Med 2015; 28:1491–1494.
90. Esaiassen E, Hjerde E, Cavanagh JP, et al. Effects of probiotic supplementation on the gut microbiota and antibiotic resistome development in preterm infants. Front Pediatr 2018; 6:347.
91. Varankovich NV, Nickerson MT, Korber DR. Probiotic-based strategies for therapeutic and prophylactic use against multiple gastrointestinal diseases. Front Microbiol 2015; 6:685.
92. de Simone C. The unregulated probiotic market. Clin Gastroenterol Hepatol 2019; 17:809–817.
93. Hojsak I, Fabiano V, Pop TL, et al. Guidance on the use of probiotics in clinical practice in children with selected clinical conditions and in specific vulnerable groups. Acta Paediatr 2018; 107:927–937.
94. Kolacek S, Hojsak I, Berni Canani R, et al. Commercial probiotic products: a call for improved quality control. a position paper by the ESPGHAN Working Group for Probiotics and Prebiotics. J Pediatr Gastroenterol Nutr 2017; 65:117–124.
95. Vallabhaneni S, Walker TA, Lockhart SR, et al. Centers for Disease Control and Prevention (CDC). Notes from the field: Fatal gastrointestinal mucormycosis in a premature infant associated with a contaminated dietary supplement--Connecticut, 2014. MMWR Morb Mortal Wkly Rep 2015; 64:155–156.
96. Toscano M, de Vecchi E, Rodighiero V, et al. Microbiological and genetic identification of some probiotics proposed for medical use in 2011. J Chemother 2013; 25:156–161.
97. Marcobal A, Underwood MA, Mills DA. Rapid determination of the bacterial composition of commercial probiotic products by terminal restriction fragment length polymorphism analysis. J Pediatr Gastroenterol Nutr 2008; 46:608–611.
98. Vermeulen MJ, Luijendijk A, van Toledo L, et al. Quality of probiotic products for preterm infants: contamination and missing strains. Acta Paediatr 2019; 109:276–279.
99. Lewis ZT, Shani G, Masarweh CF, et al. Validating bifidobacterial species and subspecies identity in commercial probiotic products. Pediatr Res 2015; 79:445.
100. Toscano M, De Grandi R, Pastorelli L, et al. A consumer's guide for probiotics: 10 golden rules for a correct use. Dig Liver Dis 2017; 49:1177–1184.
101. Dilli D, Aydin B, Fettah ND, et al. The propre-save study: effects of probiotics and prebiotics alone or combined on necrotizing enterocolitis in very low birth weight infants. J Pediatr 2015; 166:545–551. e1.
102. Hays S, Jacquot A, Gauthier H, et al. Probiotics and growth in preterm infants: a randomized controlled trial, PREMAPRO study. Clin Nutr 2016; 35:802–811.
103. Mihatsch WA, Vossbeck S, Eikmanns B, et al. Effect of Bifidobacterium lactis on the incidence of nosocomial infections in very-low-birth-weight infants: a randomized controlled trial. Neonatology 2010; 98:156–163.
104. Mohan R, Koebnick C, Schildt J, et al. Effects of Bifidobacterium lactis Bb12 supplementation on intestinal microbiota of preterm infants: a double-blind, placebo-controlled, randomized study. J Clin Microbiol 2006; 44:4025–4031.
105. Stratiki Z, Costalos C, Sevastiadou S, et al. The effect of a bifidobacter supplemented bovine milk on intestinal permeability of preterm infants. Early Hum Dev 2007; 83:575–579.
106. Manzoni P, Rinaldi M, Cattani S, et al. Italian Task Force for the Study and Prevention of Neonatal Fungal Infections, Italian Society of Neonatology. Bovine lactoferrin supplementation for prevention of late-onset sepsis in very low-birth-weight neonates: a randomized trial. JAMA 2009; 302:1421–1428.
107. Manzoni P, Mostert M, Leonessa ML, et al. Oral supplementation with Lactobacillus casei subspecies rhamnosus prevents enteric colonization by Candida species in preterm neonates: a randomized study. Clin Infect Dis 2006; 42:1735–1742.
108. Romeo MG, Romeo DM, Trovato L, et al. Role of probiotics in the prevention of the enteric colonization by Candida in preterm newborns: incidence of late-onset sepsis and neurological outcome. J Perinatol 2011; 31:63–69.
109. Dani C, Biadaioli R, Bertini G, et al. Probiotics feeding in prevention of urinary tract infection, bacterial sepsis and necrotizing enterocolitis in preterm infants. A prospective double-blind study. Biol Neonate 2002; 82:103–108.
110. Underwood MA, Salzman NH, Bennett SH, et al. A randomized placebo-controlled comparison of 2 prebiotic/probiotic combinations in preterm infants: impact on weight gain, intestinal microbiota, and fecal short-chain fatty acids. J Pediatr Gastroenterol Nutr 2009; 48:216–225.
111. Chrzanowska-Liszewska D, Seliga-Siwecka J, Kornacka MK. The effect of Lactobacillus rhamnosus GG supplemented enteral feeding on the microbiotic flora of preterm infants-double blinded randomized control trial. Early Hum Dev 2012; 88:57–60.
112. Pärtty A, Luoto R, Kalliomaki M, et al. Effects of early prebiotic and probiotic supplementation on development of gut microbiota and fussing and crying in preterm infants: a randomized, double-blind, placebo-controlled trial. J Pediatr 2013; 163:1272.e1–1277.e2.
113. Millar MR, Bacon C, Smith SL, et al. Enteral feeding of premature infants with Lactobacillus GG. Arch Dis Child 1993; 69 ((5 Spec No)):483–487.
114. Manzoni P, Meyer M, Stolfi I, et al. Bovine lactoferrin supplementation for prevention of necrotizing enterocolitis in very-low-birth-weight neonates: a randomized clinical trial. Early Hum Dev 2014; 90: (Suppl 1): S60–S65.
115. Griffiths J, Jenkins P, Vargova M, et al. Enteral lactoferrin to prevent infection for very preterm infants: the ELFIN RCT. Health Technol Assess 2018; 22:1–60.
116. Dang S, Shook L, Garlitz K, et al. Nutritional outcomes with implementation of probiotics in preterm infants. J Perinatol 2015; 35:447–450.
117. Luoto R, Matomaki J, Isolauri E, et al. Incidence of necrotizing enterocolitis in very-low-birth-weight infants related to the use of Lactobacillus GG. Acta Paediatr 2010; 99:1135–1138.
118. Kane AF, Bhatia AD, Denning PW, et al. Routine supplementation of Lactobacillus rhamnosus GG and risk of necrotizing enterocolitis in very low birth weight infants. J Pediatr 2018; 195:73.e2–79.e2.
119. Jacobs SE, Tobin JM, Opie GF, et al. ProPrems Study Group. Probiotic effects on late-onset sepsis in very preterm infants: a randomized controlled trial. Pediatrics 2013; 132:1055–1062.
120. Bin-Nun A, Bromiker R, Wilschanski M, et al. Oral probiotics prevent necrotizing enterocolitis in very low birth weight neonates. J Pediatr 2005; 147:192–196.
121. Li D, Rosito G, Slagle T. Probiotics for the prevention of necrotizing enterocolitis in neonates: an 8-year retrospective cohort study. J Clin Pharm Ther 2013; 38:445–449.
122. Shadkam MN, Jalalizadeh F, Nasiriani K. Effects of probiotic Lactobacillus Reuteri (DSM 17938) on the incidence of necrotizing enterocolitis in very low birth weight premature infants. Iranian J Neonatol 2015; 6:15–20.
123. Wejryd E, Marchini G, Frimmel V, et al. Probiotics promoted head growth in extremely low birthweight infants in a double-blind placebo-controlled trial. Acta Paediatr 2019; 108:62–69.
124. Cui X, Shi Y, Gao S, et al. Effects of Lactobacillus reuteri DSM 17938 in preterm infants: a double-blinded randomized controlled study. Ital J Pediatr 2019; 45:140.
125. Kaban RK, Wardhana, Hegar B, et al. Lactobacillus reuteri DSM 17938 improves feeding intolerance in preterm infants. Pediatr Gastroenterol Hepatol Nutr 2019; 22:545–553.
126. Rojas MA, Lozano JM, Rojas MX, et al. Prophylactic probiotics to prevent death and nosocomial infection in preterm infants. Pediatrics 2012; 130:e1113–e1120.
127. Oncel MY, Sari FN, Arayici S, et al. Lactobacillus Reuteri for the prevention of necrotising enterocolitis in very low birthweight infants: a randomised controlled trial. Arch Dis Child Fetal Neonatal Ed 2014; 99:F110–F115.
128. Hunter C, Dimaguila MA, Gal P, et al. Effect of routine probiotic, Lactobacillus reuteri DSM 17938, use on rates of necrotizing enterocolitis in neonates with birthweight < 1000 grams: a sequential analysis. BMC Pediatr 2012; 12:142.
129. Rolnitsky A, Ng E, Asztalos E, et al. A quality improvement intervention to reduce necrotizing enterocolitis in premature infants with probiotic supplementation. Pediatr Qual Saf 2019; 4:e201.
130. Lin HC, Hsu CH, Chen HL, et al. Oral probiotics prevent necrotizing enterocolitis in very low birth weight preterm infants: a multicenter, randomized, controlled trial. Pediatrics 2008; 122:693–700.
131. Saengtawesin V, Tangpolkaiwalsak R, Kanjanapattankul W. Effect of oral probiotics supplementation in the prevention of necrotizing enterocolitis among very low birth weight preterm infants. J Med Assoc Thai 2014; 97: (Suppl 6): S20–S25.
132. Samuels N, van de Graaf R, Been JV, et al. Necrotising enterocolitis and mortality in preterm infants after introduction of probiotics: a quasi-experimental study. Sci Rep 2016; 6:31643.
133. Escribano E, Zozaya C, Madero R, et al. Increased incidence of necrotizing enterocolitis associated with routine administration of Infloran in extremely preterm infants. Benef Microbes 2018; 9:683–690.
134. Demirel G, Erdeve O, Celik IH, et al. Saccharomyces boulardii for prevention of necrotizing enterocolitis in preterm infants: a randomized, controlled study. Acta Paediatr 2013; 102:e560–e565.
135. Serce O, Benzer D, Gursoy T, et al. Efficacy of Saccharomyces boulardii on necrotizing enterocolitis or sepsis in very low birth weight infants: a randomised controlled trial. Early Hum Dev 2013; 89:1033–1036.
136. Costalos C, Skouteri V, Gounaris A, et al. Enteral feeding of premature infants with Saccharomyces boulardii. Early Hum Dev 2003; 74:89–96.
137. Xu L, Wang Y, Wang Y, et al. A double-blinded randomized trial on growth and feeding tolerance with Saccharomyces boulardii CNCM I-745 in formula-fed preterm infants. J Pediatr (Rio J) 2016; 92:296–301.
138. Zhang AM, Sun ZQ, Zhang LM. Mosapride combined with probiotics on gastrointestinal function and growth in premature infants. Exp Ther Med 2017; 13:2675–2680.
139. Karthikeyan G, Govindarajan M, Veerasekar G. Routine probiotic supplementation (Saccharomyces boulardii) of neonates with birth weight 1000–1999 g: a cohort study. Int J Sci Study 2015; 3:127–131.
140. Fernández-Carrocera LA, Solis-Herrera A, Cabanillas-Ayon M, et al. Double-blind, randomised clinical assay to evaluate the efficacy of probiotics in preterm newborns weighing less than 1500 g in the prevention of necrotising enterocolitis. Arch Dis Child Fetal Neonatal Ed 2013; 98:F5–F9.
141. Al-Hosni M, Duenas M, Hawk M, et al. Probiotics-supplemented feeding in extremely low-birth-weight infants. J Perinatol 2012; 32:253–259.
142. Van Niekerk E, Nel DG, Blaauw R, et al. Probiotics reduce necrotizing enterocolitis severity in HIV-exposed premature infants. J Trop Pediatr 2015; 61:155–164.
143. Indrio F, Riezzo G, Tafuri S, et al. Probiotic supplementation in preterm: feeding intolerance and hospital cost. Nutrients 2017; 9: pii: E965.
144. Dutta S, Ray P, Narang A. Comparison of stool colonization in premature infants by three dose regimes of a probiotic combination: a randomized controlled trial. Am J Perinatol 2015; 32:733–740.
145. Watkins C, Murphy K, Dempsey EM, et al. Dose-interval study of a dual probiotic in preterm infants. Arch Dis Child Fetal Neonatal Ed 2019; 104:F159–F164.
146. Ouwehand AC. A review of dose-responses of probiotics in human studies. Benef Microbes 2017; 8:143–151.
147. Watkins C, Murphy K, Dempsey EM, et al. The viability of probiotics in water, breast milk, and infant formula. Eur J Pediatr 2018; 177:867–870.
148. Wu RY, Pasyk M, Wang B, et al. Spatiotemporal maps reveal regional differences in the effects on gut motility for Lactobacillus reuteri and rhamnosus strains. Neurogastroenterol Motil 2013; 25:e205–e214.
149. Mantziari A, Aakko J, Kumar H, et al. The impact of storage conditions on the stability of Lactobacillus rhamnosus GG and Bifidobacterium animalis subsp. lactis Bb12 in human milk. Breastfeed Med 2017; 12:566–569.
150. Underwood MA, Davis JCC, Kalanetra KM, et al. Digestion of human milk oligosaccharides by Bifidobacterium breve in the premature infant. J Pediatr Gastroenterol Nutr 2017; 65:449–455.
151. Le Doare K, Holder B, Bassett A, et al. Mother's milk: a purposeful contribution to the development of the infant microbiota and immunity. Front Immunol 2018; 9:361.
152. Padilha M, Iaucci JM, Cabral VP, et al. Maternal antibiotic prophylaxis affects Bifidobacterium spp. counts in the human milk, during the first week after delivery. Benef Microbes 2019; 10:155–163.
153. Grev J, Berg M, Soll R. Maternal probiotic supplementation for prevention of morbidity and mortality in preterm infants. Cochrane Database Syst Rev 2018; 12:CD012519.
154. Deshpande G, Athalye-Jape G, Patole S. Para-probiotics for preterm neonates-the next frontier. Nutrients 2018; 10: pii: E871.
155. Aceti A, Beghetti I, Maggio L, et al. Filling the gaps: current research directions for a rational use of probiotics in preterm infants. Nutrients 2018; 10: pii: E1472.
156. Aguilar-Toaláa JE, García-Varela R, García HS, et al. Postbiotics: An evolving term within the functional foods field. Trends Food Sci Tech 2018; 75:105–114.
157. Pique N, Berlanga M, Minana-Galbis D. Health benefits of heat-killed (tyndallized) probiotics: an overview. Int J Mol Sci 2019; 20: pii: E2534.
Keywords:

Bifidobacterium; Lactobacillus; microbiome; necrotizing enterocolitis; premature neonate; preterm infant; probiotics; sepsis

Supplemental Digital Content

© 2020 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology,