Probiotics for Preventing Necrotizing Enterocolitis in Preterm Infants: A Network Meta-Analysis

Abstract

Background: Recent evidence supports a role of probiotics in preventing necrotizing enterocolitis (NEC) in preterm infants.

Methods: A systematic review and network meta-analysis of randomized controlled trials (RCTs) on the role of probiotics in preventing NEC in preterm infants, focusing on the differential effect of type of feeding, was performed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. A random-effects model was used; a subgroup analysis on exclusively human milk (HM)-fed infants vs. infants receiving formula (alone or with HM) was performed.

Results: Fifty-one trials were included (10,664 infants, 29 probiotic interventions); 31 studies (19 different probiotic regimens) were suitable for subgroup analysis according to feeding. In the overall analysis, Lactobacillus acidophilus LB revealed the most promising effect for reducing NEC risk (odds ratio (OR), 0.03; 95% credible intervals (CrIs), 0.00-0.21). The subgroup analysis showed that Bifidobacterium lactis Bb-12/B94 was associated with a reduced risk of NEC stage ≥2 in both feeding type populations, with a discrepancy in the relative effect size in favour of exclusively HM-fed infants (OR 0.04; 95% CrIs <0.01-0.49 vs. OR 0.32; 95% CrIs 0.10-0.36).

Conclusions: B. lactis Bb-12/B94 could reduce NEC risk with a different size effect according to feeding type. Of note, most probiotic strains are evaluated in few trials and relatively small populations, and outcome data according to feeding type are not available for all RCTs. Further trials are needed to confirm the present findings.

Keywords: necrotizing enterocolitis; network meta-analysis; preterm infants; probiotics; systematic review.

Figure 1

Flow diagram of search strategy and study selection.

Figure 2

Network diagram. The area of the nodes is based on the total number of patients for each treatment among all trials. The thickness of the lines represents the total number of studies comparing the two treatments/nodes. B.b: B. breve BBG YIT4010, B. breve BBG-001, B. breve M-16V; B.la: Bifidobacterium lactis Bb-12 OR B94; B.lBB536: B. longum BB536; L.r: Lactobacillus reuteri DSM 17938, L. reuteri ATCC 55730; LGG53103: L. rhamnosus GG ATCC 53103; L.a: L. acidophilus LB; LCR35: L. casei var. rhamnosus (LCR 35); Ba.co: Bacillus coagulans (L. sporogenes); Ba.c: Ba. clausii (four strains); Sa.b: Saccharomyces boulardii CNCM I-745, Sa. boulardii CNCMI-3799; B.MS: B. lactis Bb-12 + B. longum BB536; B.l + LGG: B. longum 35,624 + L. rhamnosus GG, B. longum BB536 + L. rhamnosus GG; MG: multi-genus probiotic group.

Figure 3

Bar chart of surface under the cumulative ranking curve (SUCRA) scores (A) and forest plot of relative effect sizes compared to placebo (B) for each treatment under study.

Figure 4

Bar chart of SUCRA scores (A) and forest plot of relative effect sizes compared to placebo (B) for each treatment under study. Only trials that made information on type of feeding available were investigated.

Figure 5

Bar chart of SUCRA scores (A) and forest plot of relative effect sizes compared to placebo (B) for each treatment under study by type of feeding (purple: exclusively human milk (HM); orange: exclusively formula milk and non-exclusive HM). NA, not available.