Treatment-Specific Composition of the Gut Microbiota Is Associated With Disease Remission in a Pediatric Crohn’s Disease Cohort

Abstract

Background The beneficial effects of antibiotics in Crohn’s disease (CD) depend in part on the gut microbiota but are inadequately understood. We investigated the impact of metronidazole (MET) and metronidazole plus azithromycin (MET+AZ) on the microbiota in pediatric CD and the use of microbiota features as classifiers or predictors of disease remission. Methods 16S rRNA-based microbiota profiling was performed on stool samples from 67 patients in a multinational, randomized, controlled, longitudinal, 12-week trial of MET vs MET+AZ in children with mild to moderate CD. Profiles were analyzed together with disease activity, and then used to construct random forest models to classify remission or predict treatment response. Results Both MET and MET+AZ significantly decreased diversity of the microbiota and caused large treatment-specific shifts in microbiota structure at week 4. Disease remission was associated with a treatment-specific microbiota configuration. Random forest models constructed from microbiota profiles before and during antibiotic treatment with metronidazole accurately classified disease remission in this treatment group (area under the curve [AUC], 0.879; 95% confidence interval, 0.683–0.9877; sensitivity, 0.7778; specificity, 1.000; P < 0.001). A random forest model trained on pre-antibiotic microbiota profiles predicted disease remission at week 4 with modest accuracy (AUC, 0.8; P = 0.24). Conclusions MET and MET+AZ antibiotic regimens in pediatric CD lead to distinct gut microbiota structures at remission. It may be possible to classify and predict remission based in part on microbiota profiles, but larger cohorts will be needed to realize this goal.

Predictive Metagenomic Analysis of Autoimmune Disease Identifies Robust Autoimmunity and Disease Specific Microbial Signatures

“…Using an expansive literature search we identified a total of 132 autoimmunity studies fulfilling our criteria ( Supplementary Figure 1 ). Following filtering based on unique data, age (2 years or older), metadata and raw file availability and sequencing depth we were able to successfully download raw (FASTQ)16S rRNA and/or shotgun metagenomics data from 42 studies, 30 with 16S rRNA sequencing data ( Hevia et al, 2014 ; Mejía-León et al, 2014 ; Stoll et al, 2014 ; Consolandi et al, 2015 ; Miyake et al, 2015 ; Chen et al, 2016a , b ; Di Paola et al, 2016 ; Dunn et al, 2016 ; Jangi et al, 2016 ; Mar et al, 2016 ; Shaw et al, 2016 ; Tejesvi et al, 2016 ; Bajer et al, 2017 ; Halfvarson et al, 2017 ; Jacob et al, 2017 ; Pascal et al, 2017 ; Goyal et al, 2018 ; Luo et al, 2018 ; Manasson et al, 2018 ; Moris et al, 2018 ; Braun et al, 2019 ; Kozhieva et al, 2019 ; Lee et al, 2019 ; Li et al, 2019 ; Ruff et al, 2019 ; Sprockett et al, 2019 ; Sun et al, 2019 ; Zegarra-Ruiz et al, 2019 ; Choileáin et al, 2020 ) and 9 studies with shotgun metagenomics data ( Lewis et al, 2015 ; Heintz-Buschart et al, 2016 ; Ananthakrishnan et al, 2017 ; Hall et al, 2017 ; Wen et al, 2017 ; Ye et al, 2018 ; Proctor et al, 2019 ; Ventura et al, 2019 ; Zhou et al, 2020 ), and 3 studies with both ( Scher et al, 2013 ; Cekanaviciute et al, 2017 ; Connors et al, 2020 ; Supplementary Table 1 and Supplementary Figure 1 ). These included studies on Inflammatory Bowel Disease (IBD, N = 14 ), Multiple Sclerosis (MS, N = 7 ), Rheumatoid Arthritis (RA, N = 5 ), Juvenile Idiopathic Arthritis (JIA, N = 3 ), Systemic Lupus Erythematosus (SLE, ...…”

“…Following quality control (QC) and filtering, 33 studies containing 16S rRNA ( Scher et al, 2013 ; Hevia et al, 2014 ; Mejía-León et al, 2014 ; Stoll et al, 2014 ; Consolandi et al, 2015 ; Miyake et al, 2015 ; Chen et al, 2016a , b ; Di Paola et al, 2016 ; Dunn et al, 2016 ; Jangi et al, 2016 ; Mar et al, 2016 ; Shaw et al, 2016 ; Tejesvi et al, 2016 ; Bajer et al, 2017 ; Cekanaviciute et al, 2017 ; Halfvarson et al, 2017 ; Jacob et al, 2017 ; Pascal et al, 2017 ; Goyal et al, 2018 ; Luo et al, 2018 ; Manasson et al, 2018 ; Moris et al, 2018 ; Braun et al, 2019 ; Kozhieva et al, 2019 ; Lee et al, 2019 ; Li et al, 2019 ; Ruff et al, 2019 ; Sprockett et al, 2019 ; Sun et al, 2019 ; Zegarra-Ruiz et al, 2019 ; Choileáin et al, 2020 ; Connors et al, 2020 ) and 12 studies containing metagenomics ( Scher et al, 2013 ; Lewis et al, 2015 ; Heintz-Buschart et al, 2016 ; Ananthakrishnan et al, 2017 ; Cekanaviciute et al, 2017 ; Hall et al, 2017 ; Wen et al, 2017 ; Ye et al, 2018 ; Proctor et al, 2019 ; Ventura et al, 2019 ; Connors et al, 2020 ; Zhou et al, 2020 ) data remained for downstream analysis ( Figure 1 and Supplementary Figure 2 ). Notably, 10 out of the 33 16S rRNA and 5 of the 12 metagenomics data sets investigated the role of the human gut microbiome in IBD, due in part to its relatively high prevalence in 1.3% of US adults ( CDC, 2019 ).…”

“…However, we were also able to acquire data from studies of more rare autoimmune diseases including Behçet’s Syndrome, which results from inflammation of the blood vessels ( Consolandi et al, 2015 ), Myasthenia Gravis, a neuromuscular disorder, and Reactive Arthritis. A portion of these studies contained significantly more disease samples than the healthy samples, with Halfvarson et al (2017) having 10 times more samples from individuals with autoimmune disease than from healthy controls, and with 6 other studies ( Lewis et al, 2015 ; Ananthakrishnan et al, 2017 ; Jacob et al, 2017 ; Lee et al, 2019 ; Sprockett et al, 2019 ; Connors et al, 2020 ) containing only disease samples ( Figure 2 ). For this reason, we included healthy samples from three additional studies which investigated non-autoimmune diseases ( Huttenhower et al, 2012 ; Giloteaux et al, 2016 ; Whisner et al, 2018 ), which after QC and preprocessing resulted in additional 232 16S and 322 shotgun metagenomics samples.…”

Nutritional Therapy Strategies in Pediatric Crohn’s Disease

“…Extensive studies have been performed to relate microbiome changes with active disease and/or response to treatment [ 23 , 24 , 25 , 26 , 27 , 28 , 29 ]. Indeed, microbiome manipulation by means of antibiotics has shown promise as a therapeutic strategy for treating pediatric CD in a randomized controlled trial (RCT) of azithromycin + metronidazole for luminal CD [ 9 , 19 , 30 , 31 ]. Antibiotics are also indicated to help maintain remission to anti-TNF in perianal CD [ 9 , 32 , 33 ].…”

Shrinkage improves estimation of microbial associations under different normalization methods

“…In many studies, microbial counting strategy has transitioned away from the use of OTUs and toward ASVs ( 3 ). We suspect that shrinkage may also improve association learning in these contexts where variance-stabilizing transformations are often used in analysis ( 49 ).…”