Author(s): Simon Eaton 1 [*] , Nigel J Hall 2
intestinal microbiota; meta-analysis; necrotizing enterocolitis; premature infants; prevention; probiotics; randomized controlled trials
There has recently been a great interest in the potential of probiotics to prevent necrotizing enterocolitis (NEC). Combined data from heterogeneous randomized controlled trials (RCTs) suggest that probiotics may decrease the incidence of NEC. However, as these trials used a variety of probiotic products, we do not know which probiotic is the optimum. In addition, the group at greatest risk of NEC, those with a birthweight of <1000 g, is relatively underrepresented in these trials, so we do not have adequate evidence of either efficacy or safety to recommend universal prophylactic administration of probiotics to premature infants. Nevertheless, some experimental evidence of the mechanism by which probiotics may be effective is starting to accumulate. This may allow development of surrogate end points of effectiveness, refinement of probiotic regimes or even development of pharmacological agents that may act through the same mechanism. Hence, although routine probiotic administration is controversial, studies of probiotic effects may ultimately lead us to effective means of preventing this devastating disease.
Necrotizing enterocolitis is one of the most common life-threatening diseases affecting neonates, predominantly premature infants, with a prevalence of 5-10% in infants of birth weight <1500 g and a mortality rate of 20-30%  . The multifactorial etiology is unclear, but involves: intestinal immaturity, gut microbial colonization and localized ischemia. Gut inflammation leads to necrosis that may require intestinal resection. At present, there are no specific therapies; supportive treatment comprises intestinal rest, antibiotics and medical treatment of systemic consequences. Survivors of acute NEC may have long-term morbidities related both to gut necrosis (e.g., short bowel syndrome  ) and systemic sequelae (e.g., adverse neurodevelopmental outcome  ). Given the morbidity and mortality and the lack of specific treatment, there is an urgent need for development of preventative strategies.
Recently, probiotic administration to premature infants with the aim of preventing NEC has attracted much interest. Investigators have hypothesized that by altering the balance of the intestinal microbiome away from 'unfriendly' pathogenic bacteria in favor of 'friendly' bacteria, the incidence of NEC and NEC-related mortality may be reduced. Many RCTs have been presented, and meta-analyses of these trials suggest that probiotics may reduce the incidence of NEC and NEC-related mortality [4,5,6] . However, caution has been advised that routine probiotics use cannot yet be recommended. Closer examination reveals just reason for this caution. Despite these studies and meta-analyses, we still do not know the answer to these fundamental questions:
* which probiotic product?
* which dose and when?
* to which infants?
* is routine probiotic administration safe?
Of the trials meta-analyzed, many have methodological flaws and used different probiotic strains (single agents or combinations). A critical systematic review concluded that first, there were too few studies with similar probiotic strains to show that any single intervention is effective, and second, the overall level of evidence is no better than 2a (a meta-analysis of homogeneous well-conducted trials would achieve a level of evidence of...