Different gut microbial communities correlate with efficacy of albendazole-ivermectin against soil-transmitted helminthiases

Abstract

Soil-transmitted helminth infections represent a large burden with over a quarter of the world's population at risk. Low cure rates are observed with standard of care (albendazole); therefore, a more effective combination therapy (albendazole and ivermectin) is being investigated but showed variable treatment efficacies without evidence of intrinsic parasite resistance. Here, we analyzed the microbiome of Trichuris trichiura and hookworm-infected patients and found an association of different enterotypes with treatment efficacy. 80 T. trichiura-infected patients with hookworm co-infections from Pak-Khan, Laos, received either albendazole (n = 41) or albendazole and ivermectin combination therapy (n = 39). Pre-/post-treatment stool samples were collected to monitor treatment efficacy and microbial communities were profiled using 16S rRNA gene sequencing, qPCR, and shotgun sequencing. We identified three bacterial enterotypes and show that pre-treatment enterotype is associated with efficacy of the combination treatment for both T. trichiura (CRET1 = 5.8%; CRET2 = 16.6%; CRET3 = 68.8%) and hookworm (CRET1 = 31.3%; CRET2 = 16.6%; CRET3 = 78.6%). This study shows that pre-treatment enterotype enables predicting treatment outcome of combination therapy for T. trichiura and hookworm infections.Trial registration: ClinicalTrials.gov, NCT03527732. Registered 17 May 2018, https://ichgcp.net/clinical-trials-registry/NCT03527732 .

Conflict of interest statement

The authors declare no competing interests.

© 2022. The Author(s).

Figures

Fig. 1. Trial profile.

Trial profile depicting the number of patients included for microbiome analyses.

Fig. 2. Underlying compositional structures and their taxonomic features.

A Gut microbial community composition of patients infected with Trichuris trichiura before treatment. The cladogram was generated using Bray–Curtis dissimilarity. B Performance of classification in this dataset using a random forest model. Taxonomic features are ranked according to their individual contribution to sample classification. C Bacterial genera found to be enriched in one of the enterotype using a Kruskal–Wallis test for group comparison combined with a linear discriminant analysis for effect size (Lefse). D Non-metric multidimensional scaling ordination plot of baseline samples using Bray–Curtis dissimilarity. The labeled genera were found to be enriched in either enterotype. Treatment arm A = albendazole and ivermectin, treatment arm B = albendazole, ET enterotype.

Fig. 3. Taxon-specific and total bacterial qPCR to classify pre-treatment sample in a treatment-relevant category.

A Total bacteria and taxon-specific density measured by quantitative PCR (qPCR), by enterotype. Comparison of absolute abundances between enterotype 1 (n = 17), enterotype 2 (n = 6), and enterotype 3 (n = 16) was conducted using two-sided Mann-Whitney tests. The lower and upper bound of each box represent the 25th and 75th percentiles, respectively, and the line within indicates the median. The whiskers represent the minimum and maximum values. B Classification sensitivity and specificity into the enterotype using qPCR values. PREV Prevotella genus qPCR target, FAEC Faecalibacterium genus qPCR target, ESCH Escherichia coli qPCR target; 16S total bacteria qPCR target.

Fig. 4. Association between soil-transmitted helminths cure and pre-treatment enterotype (ET) by treatment arm and species-level characteristics of each enterotype.

A Association between treatment outcome of Trichuris trichiura and ET at baseline. B Association measured between treatment outcome of hookworm and ET before treatment. The cure rate is defined as the presence or absence of eggs in stool between day 14 and day 28 after treatment (= average eggs per gram of stool in samples collected between days 14 and 28 after treatment). C Species-level differences between compositional clusters (= enterotypes). Fisher’s exact tests were two-sided and the 95% confidence interval of the odds ratios is shown in the bracket. Labels on the pie charts represent the number of patients in each group. n number of patients, OR odds ratios.

Fig. 5. Relationship between pre-treatment enterotype and daily post-treatment egg counts of Trichuris trichiura and hookworm.

A Kaplan–Meier curve showing the remaining proportion of infected patients over time, stratified by baseline enterotype. T. trichiura cure is shown on the left, while hookworm cure is shown on the right. B Averaged daily egg counts of T. trichiura (upper two panels) and hookworm (lower two panels) stratified by pre-treatment enterotype. Daily egg counts are shown for the combination therapy (albendazole and ivermectin; left half panels) and monotherapy (albendazole; right half panels). The black dashed lines indicate the lower and upper thresholds of moderate infections for each parasite. ND not detected, post-tx post-treatment, ALB albendazole, IVE ivermectin.

Fig. 6. Comparison of metabolic potential between enterotypes.

A Venn diagram showing qualitative overlap of KEGG orthology terms (KO) between treatment-relevant enterotypes. B Volcano plots highlight quantitative differences (differences in normalized counts) at the KO level between enterotypes. The comparison was performed using Deseq2 and P-values were adjusted for multiple testing bias using the Benjamini–Hochberg procedure. C Principal component analysis to display community-level differences of KO. A PERMANOVA analysis was conducted to quantify these differences. D Bar chart showing the average relative abundances of high-level KEGG pathways, by enterotype. For clarity, only proportions of sequences above 2% are shown. E Comparison between enterotypes of relative abundances of metabolic pathways associated with positive soil-transmitted helminth infection status in previous studies–. A Kruskal–Wallis test was used in combination with the Benjamini–Hochberg multiple testing bias corrections to identify differences and fold changes were calculated to show the corresponding effect sizes. ET1–3 enterotypes 1–3, FC fold change, Ref. Reference.