Gut microbiota diversity after autologous fecal microbiota transfer in acute myeloid leukemia patients

Abstract

Acute myeloid leukemia (AML) intensive chemotherapy combined with broad-spectrum antibiotics, leads to gut microbiota dysbiosis promoting pathological conditions and an increased incidence of complications. Here we report findings from a phase II single-arm, multicenter study evaluating autologous fecal microbiota transfer (AFMT) in 25 AML patients treated with intensive chemotherapy and antibiotics (ClinicalTrials.gov number: NCT02928523). The co-primary outcomes of the study are to evaluate the efficacy of AFMT in dysbiosis correction and multidrug-resistant bacteria eradication. The main secondary outcomes are to define a dysbiosis biosignature, to evaluate the effect of dysbiosis correction on patient clinical status, to assess the short and mid-term safety of AFMT in this immunocompromised population, and to evaluate the feasibility of the AFMT procedure and acceptability by the patient. Intensive induction chemotherapy induces a dramatic decrease of α-diversity indices, and a microbial dysbiosis with a significant shift of the microbial communities and domination of pro-inflammatory families. After AFMT treatment, α-diversity indices return to their initial mean levels and the similarity index shows the restoration of microbial communities. The trial meets pre-specified endpoints. AFMT appears to be safe and may be effective for gut microbiota restoration in AML patients receiving intensive chemotherapy and antibiotics, with an excellent gut microbiota reconstruction based on both richness and diversity indices at the species level.

Fig. 1. Study flow chart.

AML is for acute myeloid leukemia. AFMT, autologuous fecal microbiota transfer; D, day.

Fig. 2. Characterization of the fecal microbiota from visit 1 (V1) to visit 4 (V4) in per-protocol patients (n = 20).

a Alpha-diversity measured at species level with, from left to right: Shannon and Inverse Simpson indexes (P-value were determined by two-sided signed-rank Wilcoxon paired test, no adjustments were made for multiple comparisons, error bars indicate median and interquartile range). b Beta-diversity measured at species level with the Bray–Curtis index comparison between visits and Principal Coordinates Analysis based on Bray–Curtis similarity at each visit (P-value were determined by two-sided signed-rank Wilcoxon paired test, no adjustments were made for multiple comparisons, error bars indicate median and interquartile range). Source data are provided as a Source Data file.

Fig. 3. Family profile heatmap.

The relative abundances of the families found in the different samples are represented in shades of blue. Each column corresponds to a patient and the order of the columns is respected between the different panels. Only patients for which a complete kinetics from visit 1 (V1) to visit 4 (V4) is available are shown (n = 12). For each sample, the associated values of the Shannon index at the species level and the Bray-Curtis similarity with respect to V1 at the species level are represented. For Shannon index, ranges were based on the Shannon value of the study population: low < first quartile, medium ≥ first quartile and < third quartile, high ≥ third quartile. For Bray–Curtis similarity index: low < 0.4, medium ≥ 0.4 and < 0.6 and high ≥ 0.6. The choice of this mathematically determination was made before the analysis of the results.

Fig. 4. Number of reads mapped against antibiotic resistance genes identified through metagenomic sequencing from V1 to V4 for the per-protocol population (n = 20).

P-values were determined by two-sided signed-rank Wilcoxon paired test, no adjustments were made for multiple comparisons, error bars indicate median and interquartile range. ABR is for antibiotic resistance genes. Source data are provided as a Source Data file.

Fig. 5. Evolution of biochemical and immunological parameters in per-protocol patients (n = 20).

a Systemic level in blood, from left to right: neopterin, C-Reactive Protein (P-value were determined by two-sided signed-rank Wilcoxon paired test, no adjustments were made for multiple comparisons, error bars indicate median and interquartile range, for neopterin n = 10). b Local levels in feces: neopterin, secretory immunoglobulin A (IgA, P-value were determined by two-sided signed-rank Wilcoxon paired test, no adjustments were made for multiple comparisons, error bars indicate median and interquartile range). Source data are provided as a Source Data file.

Fig. 6. Correlation between host and microbiota parameters (n = 20).

Spearman pairwise correlation. Only significant correlations are depicted. Blue = positive correlation; Red = negative correlation. Correlations are calculated from the values measured in all patients at all visits (except for TGFβ2 for which only the values measured at visit V3 are considered). CRP is for C-Reactive Protein; IgA, immunoglobulin A; TGFβ2_V3, Transforming growth factor-beta 2 at visit 3; CD14_pos, cluster of differentiation 14 positive. Source data are provided as a Source Data file.

Fig. 7. CONSORT diagram.

AFMT is for autologous fecal microbiota transfer; D, day; N, number of patients, V1 to V6, visit 1 to visit 6.