Probiotics and the EARly Life effects on intestinal bacteria and inflammation in children with cystic fibrosis (PEARL-CF) study

Abstract

Background: Probiotics in cystic fibrosis (CF) have potential benefits on gut inflammation and clinical outcomes [1], however their application in early childhood remains unexplored.We evaluated probiotics on gut microbiota and inflammation in children with cystic fibrosis (CwCF). Methods: PEARL-CF is a multicenter, double-blind, randomized, placebocontrolled probiotic trial in CwCF aged 0–6y comparing 3 cohorts: (i) probiotic (15 Lactobacillus & Bifidobacterium strains; 10.2 billion colony forming units (CFU)) daily during a 12 month (12 m) intervention period (INT); (ii) placebo for 12 m, and (iii) healthy non-CF controls. Stool samples and clinical datawere collected during the INTand for 12 mpost INT (POSTINT). Analyses of stool samples and pulmonary infections are shown. Linear mixed models estimate and standard error (SE) were used for longitudinal analyses. A post hoc subgroup analyses of ≥ 4 years (y) and < 4y at INT start was performed due to the plateau in stool bacterial richness at 4y. Results: 38 CwCF received the probiotic (mean age 3.2 ± 1.6y) and 39 the placebo (mean age 3.1 ± 1.7y). No baseline demographic, anthropometric or stool diversity differences were found between groups. Fecal bacterial alpha diversity did not differ between the probiotic and placebo groups during the INT (Fig. 1A) or POST-INT. Fecal bacterial community structure (Bray-Curtis dissimilarity) did not differ between the probiotic and placebo groups at the end of the INT (Fig. 1B) or end of the POST-INT. Fewer Haemophilus influenzae pulmonary infections occurred during the INT for the probiotic vs placebo groups (27% vs 56%, p = 0.03). CwCF starting < 4y in the probiotic group had fewer H. influenzae infections vs placebo (18% vs 62%, p = 0.006). Rates of hospitalisations, pulmonary exacerbations or antibiotic courses did not differ between groups. The change in fecal M2 pyruvate kinase (M2PK) from baseline to 12 mwas lower in the probiotic vs placebo group, −1.3 U/mL (−4.3–1.1) vs 3.9 U/mL (−1.0–11.9), p = 0.002 (Fig. 1D). Change in fecal calprotectin did not differ between groups. Post hoc analysis of CwCF ≥4y during the INT showed the probiotic group had a higher relative abundance of two probiotic strains compared to the placebo group: (i) Lactobacillus paracasei (estimate (SE) 0.25% (0.1), p = 0.03) and (ii) Lactobacillus rhamnosus (estimate (SE) 0.44% (0.2), p = 0.049). CwCF ≥4y on the probiotic had higher fecal bacterial richness vs placebo during: (i) INT (estimate (SE) 26 (10) species, p = 0.02; Fig. 1C); and (ii) POST-INT (estimate (SE) 49 species (18), p = 0.02). In CwCF < 4y we observed that during the INT, fecal: (i) relative abundance of Agathobaculum butyriciproducens (butyrate producer) was lower in the probiotic vs placebo (estimate 0.13% (0.04), p = 0.001); and (ii) butyric acidwas lower in the probiotic vs placebo (estimate 2125 μg/mg (818) dry weight, p = 0.01). Conclusions: CwCF starting a multi-strain probiotic < 4y of age had fewer pulmonary infections with H. influenzae. Probiotics improved fecal M2PK (i. e. inflammation) and in CwCF ≥4y improved fecal bacterial diversity. Further understanding of the long-term potential clinical effects of changes is required. Acknowledgements: This work was supported by: Cystic Fibrosis Foundation; NHMRC, Australia (2020/GNT1194358); SCHF; Thrasher Fund; Evolution Health Pty Ltd; Dr Charles Foster & Prof. William Rawlinson (VRL).