Gut Microbiota Offers Universal Biomarkers across Ethnicity in Inflammatory Bowel Disease Diagnosis and Infliximab Response Prediction

Abstract

Gut microbiota dysbiosis contributes to the onset and perpetuation of inflammatory bowel disease (IBD). Given that gut microbiotas vary across geography and ethnicity, it remains obscure whether any universal microbial signatures for IBD diagnosis and prognosis evaluation exist irrespective of populations. Here we profiled the fecal microbiota of a series of Chinese IBD patients and combined them with two Western IBD cohorts, PRISM and RISK, for meta-analyses. We found that the gut microbial alteration patterns in IBD are similar among Chinese and Westerners. Our prediction model based on gut microbiome for IBD diagnosis is robust across the cohorts, which showed 87.5% and 79.1% prediction accuracy in Crohn's disease (CD) and ulcerative colitis (UC) patients, respectively. A relative increase in the levels of Actinobacteria and Proteobacteria (Enterobacteriaceae) and a relative decrease in the levels of Firmicutes (Clostridiales) were strongly correlated with IBD severity (P < 0.05). Additionally, restoration of gut microbiota diversity and a significant increase in Clostridiales relative abundance were found in patients responding to infliximab (IFX [Remicade]) treatment compared to those in relapse. Moreover, certain microbes, mainly Clostridiales, predicted the treatment effectiveness with 86.5% accuracy alone and 93.8% accuracy in combination with calprotectin levels and Crohn's disease activity index (CDAI). Taking the results together, we conclude that gut microbiota can offer a set of universal biomarkers for diagnosis, disease activity evaluation, and infliximab treatment response prediction in IBD. IMPORTANCE In the present report, we show that the human fecal microbiota contains promising and universal biomarkers for the noninvasive evaluation of inflammatory bowel disease severity and IFX treatment efficacy, emphasizing the potential ability to mine the gut microbiota as a modality to stratify IBD patients and apply personalized therapy for optimal outcomes.

Keywords: disease activity; gut microbiota; inflammatory bowel disease; infliximab treatment.

FIG 1

Dysbiosis of gut microbiota patterns in Chinese patients with IBD. (A) Comparison of Shannon index values between HC, CD, and UC groups. (B) Taxa listed according to their linear discriminant analysis (LDA) values determined from comparisons between the HC and IBD groups as computed by the use of the LEfSe algorithm. (C) Taxa illustrated according to their taxonomic relationship using a cladogram, showing the discriminative patterns in taxonomic lineages. The LDA cutoff values for panels A and B were set at 3.5. (D) Heat map showing abundance distributions for the 18 operational taxonomic units (OTUs) identified as key variables using the Kruskal-Wallis test (after Bonferroni correction) among the HC, CD, and UC groups. Lachnospiraceae OTUs are colored in green, and those of Ruminococcaceae are colored in red. (E) Weighted UniFrac PCoA data showing grouping patterns of HC, CD, and UC. Each dot represents a sample, with shapes indicating groups and colors the abundance of Clostridiales.

FIG 2

The similar shifts of gut microbiota in IBD across cohorts. Comparisons of OTUs (A and B) and predicted KOs (C and D) differentiated between healthy controls and subjects with IBD in the current study and the RISK and PRISM cohorts with biopsy samples. Each dot represents an OTU or KO that differed significantly between healthy and disease samples in Chinese cohorts. Axes indicates the log2-fold changes of the levels of these OTU/KO abundances between subjects with disease and healthy individuals in the Chinese cohort (x axis) and the Western cohorts (y axis), with the RISK cohort indicated in green and the PRISM cohort in blue. The correlation coefficients and the P values determined from comparisons between the cohorts are labeled on the plot.

FIG 3

Gut microbiotas distinguish diseases from health similarly across cohorts. The black ROC curve indicates the accuracy of the classification model built on the Chinese cohort for classifying HC versus CD (A) and HC versus UC (B). Colored curves are the classification accuracies when these models are applied to the other cohorts.

FIG 4

Bacterial biomarkers associated with disease severity. (A) Cladogram representing taxa with different abundances for CD activity. The size of each circle is proportionate to the abundance of the taxon. (B and C) Relative abundances of Clostridiales (B) and Enterobacteriaceae (C) in CD activity groups are shown. (D) Cladogram representing taxa with different abundances for UC activity. (E and F) Relative abundances of Clostridiales (E) and Enterobacteriaceae (F) in UC activity groups are shown. The statistical significance was determined with the Wilcoxon test and was adjusted using the false-discovery rate (FDR). **, P value < 0.01; ***, P value < 0.001.

FIG 5

The microbiotas differ in the response and the relapse groups of CD patients treated with infliximab (IFX). The alpha diversity, measured by Shannon index (A) and PD whole tree (B), Clostridiales abundance (C), and calprotectin abundance (D), was restored in the response group after treatment compared to the relapse group. Stars (*) indicate P values of <0.05. (E and F) Supervised prognostic prediction of Crohn’s disease progression (IFX response or relapse). (E) The accuracy of prediction was best using the combination of microbiota, calprotectin, and CDAI (93.8%). The use of the microbiota alone (86.5%) still outperformed the results seen with the traditional clinic markers of CDAI (58.7%) or calprotectin (62.5%). (F) The top informative OTUs that contribute to the classification model. Clostridiales OTUs, colored in orange, are the most informative OTUs.