To evaluate outcome data in an observational cohort of very low birth weight infants of the German Neonatal Network stratified to prophylactic use of Lactobacillus acidophilus/Bifidobacterium infantis probiotics.
Within the observational period (September 1, 2010, until December 31, 2012, n = 5351 infants) study centers were categorized into 3 groups based on their choice of Lactobacillus acidophilus/Bifidobacterium infantis use: (1) no prophylactic use (12 centers); (2 a/b) change of strategy nonuser to user during observational period (13 centers); and (3) use before start of observation (21 centers). Primary outcome data of all eligible infants were determined according to center-specific strategy.
The use of probiotics was associated with a reduced risk for necrotizing enterocolitis surgery (group 1 vs group 3: 4.2 vs 2.6%, P = .028; change of strategy: 6.2 vs 4.0%, P < .001), any abdominal surgery, and hospital mortality. Infants treated with probiotics had improved weight gain/day, and probiotics had no effect on the risk of blood-culture confirmed sepsis. In a multivariable logistic regression analysis, probiotics were protective for necrotizing enterocolitis surgery (OR 0.58, 95% CI 0.37-0.91; P = .017), any abdominal surgery (OR 0.7, 95% CI 0.51-0.95; P = .02), and the combined outcome abdominal surgery and/or death (OR 0.43; 95% CI 0.33-0.56; P < .001).
Our observational data support the use of Lactobacillus acidophilus/Bifidobacterium infantis probiotics to reduce the risk for gastrointestinal morbidity but not sepsis in very low birth weight infants.
Nosocomial infection and necrotizing enterocolitis (NEC) are leading causes of neonatal morbidity and mortality in very low birth weight (VLBW) infants with an estimated rate of associated death up to 30% and significant impact on long-term neurodevelopment.
For both entities, a multifactorial pathophysiology has been hypothesized, including endogenous host factors, eg, gestational age and immaturity of the immune response. Moreover, environmental factors such as enteral feeding and exposure to the endemic hospital milieu are important as these issues influence abnormal gastrointestinal colonization and translocation of enteropathogenic bacteria through vulnerable intestinal mucosa.
The clinical courses of sepsis and NEC often are fulminant, and the effectiveness of therapeutic interventions is limited.
Consequently, there is an urgent need to improve prevention strategies of NEC and sepsis. In recent years, several meta-analyses have been published that demonstrated that probiotics are beneficial to preterm infants (ie, by reducing the risk for NEC and/or death but not nosocomial sepsis).
These reports prompted the majority of neonatal intensive care units (NICUs) collaborating in the German Neonatal Network (GNN) to implement probiotic prophylaxis Lactobacillus acidophilus/Bifidobacterium infantis into their clinical care. However, the discussion—whether probiotics are to be recommended for routine use—is still controversial. This controversy is related to concerns regarding efficacy and safety in populations of greatest vulnerability; comparability of study designs as the result of differences in patient cohorts, dosage, and composition of probiotics; and, last but not least, lack of knowledge on the evolution of the gut microbiota in individual infants.
Our aim was to evaluate outcome data in a large cohort of VLBW infants born in GNN centers stratified to prophylactic use of Lactobacillus acidophilus/Bifidobacterium infantis probiotics, including infants born in 13 participating NICUs that changed their strategy within the observational period.
We performed an observational study on the effects of the prophylactic use of Lactobacillus acidophilus/Bifidobacterium infantis (Infloran; Berna, Berne, Switzerland) probiotics in VLBW infants cared for in 46 NICUs in Germany (GNN). Within the study period, the data were collected prospectively from infants born between September 1, 2010, and December 31, 2012. We also evaluated a primary data set of each group (according to strategy of probiotic use) before the study period which was defined as “baseline” (n = 2828). The study parts were approved by the local committee on research in human subjects of the University of Lübeck (08-022; 03.12.2010) and the local ethical committees at the other study centers.
The inclusion criteria were as follows: birth weight <1500 g and gestational age >22 + 6 and <32 + 0 weeks. Exclusion criteria were lethal malformations (eg, trisomy 13 and trisomy 18). In all infants born in GNN centers during the study period, a primary data set documented the important outcome measures (n = 5351). After written informed consent was given by the parents, a predefined GNN data set with additional variables including antenatal/postnatal treatment (eg individual data on probiotic use, or weight gain) and outcome data were recorded (n = 3527). After discharge, data sheets were sent to the GNN center in Lübeck.
We categorized the participating study centers into 3 groups based on their choice for one of the following strategies of probiotic use: group 1, no Lactobacillus acidophilus/Bifidobacterium infantis probiotics prophylactically (n = 12); group 2, changed strategy during the observational period (n = 13; infants were divided based on date of birth; 2a, before prophylactic use based on date of birth; 2b, after prophylactic use); group 3, adopted prophylactic use before the observational study period (n = 21). In group 2b and 3 centers, there was some variability regarding dosage and time of Lactobacillus acidophilus/Bifidobacterium infantis administration. Most study centers administered Lactobacillus acidophilus/Bifidobacterium infantis to VLBW infants (1 × 1 capsule/day or 2 × 1/2 capsule/day) from day 2 or 3 of life for 14 days or until full enteral feeds (150 mL/kg/d) were tolerated. Some centers restricted prophylactic use of probiotics to infants with birth weight <1000g.
Gestational age was calculated from the best obstetric estimate based on early prenatal ultrasound and obstetric examination. NEC surgery was defined according to modified Bell criteria (≥stage 2) requiring surgery.
Any abdominal surgery was defined as required abdominal surgery for NEC, focal intestinal perforation, peritoneal adhesions, volvolus, and meconium obstruction. We excluded herniotomies, pyloromyotomies, or surgery for other congenital malformations (eg, intestinal atresia, gastroschisis, large omphalocele, or diaphragmatic hernia) from this definition. Blood-culture confirmed sepsis was defined as clinical sepsis with at least two signs (temperature >38°C or <36.5°C, tachycardia >200/min, new onset or increased frequency of bradycardias or apneas, hyperglycemia >140 mg/dL, base excess < −10 mval/L, changed skin color, increased oxygen requirements) and proof of causative agent in blood culture and one laboratory sign (C-reactive protein >2 mg/dL, immature/neutrophil ratio >0.2, white blood cell count <5/nL, platelet count < 100/nL).
All-cause mortality was defined as death occurring after admission to NICU before discharge home.
Data analysis was performed using the SPSS 20.0 data analysis package (Munich, Germany). Hypotheses were evaluated with χ2 test, Fisher exact test, and Mann-Whitney U test. P < .05 was considered as statistically significant for single tests.
We used a predefined set of well known confounding risk factors for any abdominal surgery, NEC surgery, and the combined outcome of death and/or abdominal surgery ie, gestational age, small for gestational age (SGA), inborn delivery and included Lactobacillus acidophilus/Bifidobacterium infantis prophylaxis given to the individual infant (all infants with full GNN data set, n = 3229).
From September 1, 2010, until December 31, 2012, 5351 VLBW infants were born in 46 tertiary level NICUs. The Figure (available at www.jpeds.com) demonstrates that probiotics were given across all gestational age groups and birth weight classes, and lower-risk preterm infants >30 weeks were less frequently treated. The primary baseline data sets before the study period for each group of study centers are presented (n = 2828 VLBW infants born in GNN centers, n = 1565 born January 1, 2009 to December 31, 2009, n = 1263 VLBW infants born January 1, 2010 to August 31, 2010; Table I). Interestingly, nonusers of Lactobacillus acidophilus/Bifidobacterium infantis probiotics had a relatively low rate of NEC requiring surgery compared with those centers who chose to adopt prophylactic use of probiotics before our observational study (group 1 vs 3, 3.7% vs 5.0%). The same trend was observed for the combined outcome “any abdominal surgery or death” (group 1 vs 3, 12.9% vs 14.7%).
Table IClinical characteristics of VLBW cohort before observational period (baseline)
Prophylactic use of Lactobacillus acidophilus/Bifidobacterium infantis probiotics was associated with a reduced risk for surgery for NEC, any abdominal surgery, and all-cause mortality (Table II). Notably, group 1 had a greater percentage of SGA infants than group 3. SGA is associated with adverse outcome. In those centers with a change of strategy during the study period, we observed a remarkable decrease in NEC surgery (6.2 vs 4.0%, P < .001), any abdominal surgery (8.2 vs 6.3%, P = .03), and death (10.0 vs 8.3%, P = .017). There was only a trend towards less blood-culture confirmed sepsis (15.4 vs 10.6%, P = .05) after the introduction of probiotics (group 2a vs group 2b, respectively).
Table IIClinical characteristics of VLBW cohort categorized according to the centers' choice for prophylactic use of Lactobacillus acidophilus/Bifidobacterium infantis probiotics
In a multivariable logistic regression analysis, probiotics were protective for NEC surgery (OR 0.58, 95% CI 0.37-0.91; P = .017), any abdominal surgery (OR 0.7, 95% CI 0.51-0.95; P = .02), and the combined outcome abdominal surgery and/or death (OR 0.43; 95% CI 0.33-0.56; P < .001; Table III).
Table IIIRisk factors for NEC surgery, abdominal surgery, and/or death from multiple logistic regression analysis
Clinical Characteristics of GNN Enrolled Infants Stratified to Center-specific Strategy
GNN enrolled infants prophylactically treated with Lactobacillus acidophilus/Bifidobacterium infantis probiotics had a shorter stay in hospital and less exposure to glycopeptide antibiotics (Table IV). The use of Lactobacillus acidophilus/Bifidobacterium infantis probiotics was associated with increased weight gain per day (group 1, 20.8 g/d vs group 3, 22.2 g/d; P < .001; group 2 before change, 21.5 g/d vs after change, 22.7 g/d; P = .01). This was not related to improved feeding tolerance, as infants born in group 1 centers needed less time to establish full enteral feeds (150 mL/kg/d) compared with group 3 infants. In a subgroup of infants that excluded those who had any abdominal surgery and/or died (n = 302), we confirmed the association of probiotics intake and improved weight gain (group 1, 21.3 [±6.1] g/d vs group 3, 22.5 [±4.8] g/d; P < .001; group 2a before change, 21.8 [±5.1] g/d vs group 2b after change, 23.3 [±4.3] g/d; P < .001).
Table IVClinical characteristics of GNN enrolled VLBW infants categorized according to the centers' choice for prophylactic use of Lactobacillus acidophilus/Bifidobacterium infantis probiotics
Probiotics have entered routine clinical use in the majority of NICUs in the GNN (34/46 centers), with 62% of VLBW infants 23 + 0 < 32 + 0 weeks of gestation prophylactically treated with Lactobacillus acidophilus/Bifidobacterium infantum probiotics. This development is in association with the publication of 2 meta-analyses on the beneficial effects of probiotics to reduce the risk of NEC and/or death.
We evaluated the outcome of infants born in participating study centers that were categorized according to their choice for probiotic use over time. Our baseline data before the study period suggested that probiotic use was related to greater baseline rate of NEC requiring surgery in group 3 (5.0%) compared with group 1 (3.7%). Within the study period, we observed that infants born in centers that used probiotics had a decreased risk for surgery for NEC compared with infants born in centers without probiotic use. The risk for sepsis was not influenced by center-specific strategy.
Because there were greater proportions of SGA infants in group 1 centers, we performed a multiple logistic regression analysis that included known risk factors of greater-stage NEC (gestational age, SGA, outborn delivery), as recently published by investigators of the Canadian Neonatal Network.
In this multivariate analysis, administration of Lactobacillus acidophilus/Bifidusbacterium infantum was protective against NEC requiring surgery, any abdominal surgery, and the composite outcome any abdominal surgery and/or death.
What is the consequence of our observational data? We focused on surgical NEC, because the diagnostic variability of medical NEC is well known but the decision for surgery eliminates many of the milder, questionable cases of NEC. In general, our data reflect the wide variation among centers, as seen in other networks (Eunice Kennedy Shriver National Institute of Child Health and Human Development, Canadian Neonatal Network).
Within a multicenter collaboration, centers with greater incidence of adverse outcomes—as seen for sepsis and NEC in group 2 centers before change to probiotic prophylaxis—are more likely to benefit from implementation of new strategies than those centers with low incidences. On the other hand, group 2 centers may also have improved by the feedback given through annual reports and by benchmarking approaches including optimized infection control protocols.
The major limitation of our study is its observational design. To diminish the risk of bias, we implemented the comparison between centers using probiotics, those not using probiotics, and those who changed their practice during the data collection period. Thus, our data represent current clinical practice and support results from previous randomized controlled trials. We also noted improved weight gain/day in infants treated with probiotics, which may affect secondary outcomes, eg, bronchopulmonary dysplasia, which are partially explained by impaired growth/deficient utilization of nutrients as well as inflammation (“new bronchopulmonary dysplasia”).
Probiotics may be the influence of the cross-talk between the developing immune system and the microbiota that interferes with growth and susceptibility to long-term morbidity. Animal models have demonstrated a direct link between body composition and gut microbiota.
Microbial patterns of initial colonization of the intestine are known to affect host metabolic and endocrine functions (fat deposition, leptin, and insulin levels), although the variables have not been studied in preterm infants. The finding of shorter stay of probiotics-treated VLBWs in hospital requires further investigation. Based on the hypothesis that “protective microbiota” result in an improved weight gain and shorter stay in hospital, the impact of probiotics on families' quality of life, risk for nosocomial infections, and health care costs should be investigated as important study end points.
Neonatal networks such as GNN provide a platform to study aspects of probiotic prophylaxis that have not yet been considered, eg, probiotic administration to the subgroup of patients with stoma who have different colonization patterns and a significant exposure to long-term antibiotics.
Future studies are needed that include data on human milk feeding, evaluation of stool cultures, and culture-independent methods to determine the gut microbiota in extremely vulnerable infants. It would be an important objective to include this additional information into future guidelines on feeding and probiotic administration in the individual VLBW infant.
We are grateful to the infants, parents, and health care providers who supported our study.
Additional investigators of GNN include: Stefan Avenarius, MD, Department of Pediatrics, University of Magdeburg; Kai Bockenholt, MD, Children's Hospital Cologne-Amsterdamer Strasse; Bettina Bohnhorst, MD, Department of Neonatology, Hanover Medical School; Michael Dördelmann, MD, Diakonissenhospital Flensburg; Silke Ehlers, MD, Bürgerhospital Frankfurt; Ursula Felderhoff-Müser, MD, PhD, Department of Pediatrics, University of Essen; Corinna Gebauer, MD, Department of Pediatrics, University of Leipzig; Hubert Gerleve, MD, Children's Hospital Coesfeld; Ludwig Gortner, MD, Department of Pediatrics, University of Homburg; Peter Groneck, MD, Children's Hospital Leverkusen; Friedhelm Heitmann, MD, Children's Hospital Dortmund; Georg Hillebrand, MD, Children's Hospital Itzehoe; Thomas Höhn, MD, PhD, Department of Pediatrics, University of Düsseldorf; Mechthild Hubert, MD, Children's Hospital Siegen; Helmut Hummler, MD, PhD, Department of Neonatology, University of Ulm; Andreas Jenke, MD, Children's Hospital Wuppertal; Reinhard Jensen, MD, Children's Hospital Heide; Olaf Kannt, MD, Children's Hospital Schwerin; Angela Kribs, MD, Department of Pediatrics, University of Cologne; Helmut Küster, MD, Department of Pediatrics, University of Göttingen; Reinhard Laux, MD, Children's Hospital Hamburg-Barmbek; Ursula Lieser, MD, Department of Pediatrics, University of Halle; Michael Mögel, MD, Department of Pediatrics, University of Dresden; Jens Möller, MD, Children's Hospital Saarbrücken; Dirk Müller, MD, Children's Hospital Kassel; Werner Nikischin, MD, Department of Pediatrics, University of Kiel; Dirk Olbertz, MD, Children's Hospital Rostock-Südstadt; Thorsten Orlikowsky, MD, Department of Pediatrics, University of Aachen; Jochen Reese, MD, Children's Hospital Eutin; Claudia Roll, MD, PhD, Department of Pediatrics, University Witten-Herdecke, Children's Hospital Datteln; Thomas Schaible, MD, Department of Pediatrics, University of Mannheim; Stefan Seeliger, MD, Children's Hospital Neuburg/Ingolstadt; Hugo Segerer, MD, Children's Hospital St. Hedwig, Regensburg; Norbert Teig, MD, Department of Pediatrics, University of Bochum; Ursula Weller, MD, Children's Hospital Bielefeld; Matthias Vochem, MD, Olgahospital Stuttgart; and Jürgen Wintgens, MD, Children's Hospital Mönchengladbach.
Deaths in preterm infants: changing pathology over two decades.