Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 51 of the 51 references it cites were found in a registry or an index; 18 of the 89 sources its claims cite were checked for support.

A non-antibiotic-disrupted gut microbiome is associated with clinical responses to CD19-CAR-T cell cancer immunotherapy

Open access

Abstract

Increasing evidence suggests that the gut microbiome may modulate the efficacy of cancer immunotherapy. In a B cell lymphoma patient cohort from five centers in Germany and the United States (Germany, n = 66; United States, n = 106; total, n = 172), we demonstrate that wide-spectrum antibiotics treatment ('high-risk antibiotics') prior to CD19-targeted chimeric antigen receptor (CAR)-T cell therapy is associated with adverse outcomes, but this effect is likely to be confounded by an increased pretreatment tumor burden and systemic inflammation in patients pretreated with high-risk antibiotics. To resolve this confounding effect and gain insights into antibiotics-masked microbiome signals impacting CAR-T efficacy, we focused on the high-risk antibiotics non-exposed patient population. Indeed, in these patients, significant correlations were noted between pre-CAR-T infusion Bifidobacterium longum and microbiome-encoded peptidoglycan biosynthesis, and CAR-T treatment-associated 6-month survival or lymphoma progression. Furthermore, predictive pre-CAR-T treatment microbiome-based machine learning algorithms trained on the high-risk antibiotics non-exposed German cohort and validated by the respective US cohort robustly segregated long-term responders from non-responders. Bacteroides, Ruminococcus, Eubacterium and Akkermansia were most important in determining CAR-T responsiveness, with Akkermansia also being associated with pre-infusion peripheral T cell levels in these patients. Collectively, we identify conserved microbiome features across clinical and geographical variations, which may enable cross-cohort microbiome-based predictions of outcomes in CAR-T cell immunotherapy.

Overview

CoChat Summary

TL;DR

In a five-center German/US lymphoma cohort, broad-spectrum antibiotics before CD19 CAR-T were linked to worse outcomes, but this appeared largely explained by greater pre-treatment tumor burden and inflammation. Among patients not exposed to those antibiotics, baseline gut microbiome features—especially Bifidobacterium longum, peptidoglycan biosynthesis, and taxa including Akkermansia—correlated with response/survival and supported cross-cohort machine-learning prediction.

Key findings

  • The study enrolled 172 patients with relapsed/refractory B-cell lymphoma (Germany n=66; US n=106) across five centers; median age was 62.0 years, 55 were female and 117 male, and patients had received a median of three prior systemic therapy lines. Day-90 best overall response was 58.1% (100/172) and complete response was 47.6% (82/172).
  • Any antibiotic exposure within three weeks before infusion was associated with shorter progression-free survival (HR=2.04), and the defined “high-risk antibiotics” group (meropenem, cefepime, ceftazidime, piperacillin-tazobactam) had higher progression (OR=2.60), shorter PFS (HR=3.05), and shorter overall survival (HR=2.99). High-risk antibiotic recipients also had more any-grade ICANS (23/36, 63.8%) than others (42/110, 38.2%), but they had higher LDH, metabolic tumor volume, CRP, and IL-6, plus lower peripheral CD4 T cells; in multivariate analysis, high-risk antibiotics remained associated with PFS (HR=2.19) but lost association with OS (HR=1.96).
  • After excluding high-risk antibiotic-exposed samples, Bifidobacterium longum and microbiome-encoded peptidoglycan biosynthesis were associated with 6-month survival or early lymphoma progression. Machine-learning models trained on the German cohort and validated in the US cohort predicted day-180 complete response better after excluding high-risk antibiotic samples (probabilistic AUROC=0.67±0.04; deterministic AUROC=0.69) than models including all samples (probabilistic AUROC=0.52±0.05; deterministic AUROC=0.64).
  • Bacteroides, Ruminococcus, Eubacterium, and Akkermansia were among the taxa most important for predicting CAR-T response; Akkermansia muciniphila abundance was also associated with pre-infusion peripheral T cell levels.

Methods

  • Prospective international cohort of patients receiving CD19-targeted CAR-T therapy for B-cell lymphoma between 2018 and 2021 at three German and two US tertiary centers; products included axi-cel (n=122), tisa-cel (n=49), and liso-cel (n=1). Clinical data included antibiotic exposure in the three weeks before infusion, tumor burden markers, inflammatory markers, T cell counts, and outcomes including response at 90/180 days, PFS, OS, CRS, and ICANS.
  • Stool samples were collected serially (351 samples from 116 of 172 patients) and analyzed by shotgun metagenomic sequencing, with taxonomic and functional profiling using MetaPhlAn3 and HUMAnN3. Analyses included alpha diversity, Bray-Curtis beta diversity, differential abundance, PERMANOVA, and Cox proportional hazards models.
  • Prediction used supervised machine learning with the German cohort for training and the US cohort for external validation. Models included deterministic logistic regression with elastic-net regularization and a probabilistic Bayesian logistic regression, using baseline species abundances plus age and gender, and were run both with all baseline samples and after excluding high-risk antibiotic-exposed samples.

Limitations

  • The paper states that sample-size differences and constraints between the German and US cohorts limited Cox proportional hazards regressions for PFS; that the observational design cannot demonstrate causality; and that center-specific antibiotic use, especially cefepime being given only in US patients, could confound microbiome signatures.
  • Evident from the text: high-risk antibiotic exposure is strongly confounded by worse pre-treatment disease status, higher tumor burden, greater systemic inflammation, and lower T cell counts, so antibiotic-outcome associations should not be read as causal. Species-level survival analyses also had small per-group sample sizes and some borderline results, and no direct mechanistic validation is reported.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1038/s41591-023-02234-6

Cites

51 of 51 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

10 Valid · 0 Not valid · 41 Not checked · 166 Own results and interpretations

95 more claims

Create a free account to see all 217 claims and to verify the 71 references still open.

See all claims — free account
  • Probably not a substantial claim. 66
  • It may state two separate claims as one. 65
  • The quote only partly states this claim. 3

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Citations

Cites

Show all 49 references

Cited by

See all 271

Metadata

Authors
Stein-Thoeringer, C. K., Saini, N. Y., Zamir, E., Blumenberg, V., Schubert, M.-L., Mor, U., Fante, M. A., Schmidt, S., Hayase, E., Hayase, T., Rohrbach, R., Chang, C.-C., McDaniel, L., Flores, I., Gaiser, R., Edinger, M., Wolff, D., Heidenreich, M., Strati, P., Nair, R., Chihara, D., Fayad, L. E., Ahmed, S., Iyer, S. P., Steiner, R. E., Jain, P., Nastoupil, L. J., Westin, J., Arora, R., Wang, M. L., Turner, J., Menges, M., Hidalgo-Vargas, M., Reid, K., Dreger, P., Schmitt, A., Müller-Tidow, C., Locke, F. L., Davila, M. L., Champlin, R. E., Flowers, C. R., Shpall, E. J., Poeck, H., Neelapu, S. S., Schmitt, M., Subklewe, M., Jain, M. D., Jenq, R. R., & Elinav, E.
Affiliations
  • The University of Texas MD Anderson Cancer Center (US)
  • German Cancer Research Center (DE)
  • LMU Klinikum (DE)
  • University Hospital Heidelberg (DE)
  • Weizmann Institute of Science (IL)
  • University Hospital Regensburg (DE)
  • Leibniz Institute for Immunotherapy (DE)
  • Moffitt Cancer Center (US)
  • University of South Florida (US)
Corresponding author
  • Saini, Neeraj Y.
  • Jain, Michael D.
  • Jenq, Robert R.
  • Elinav, Eran
Venue
Nature Medicine
Published
2023-03-13
Volume
29
Issue
4
Pages
906-916
ISSN
  • 1078-8956 (ISSN-L)
  • 1546-170X
OpenAlex
W4324044824
Cited by
255
References
49
Citations by year
2026: 67 · 2025: 85 · 2024: 63 · 2023: 40
FWCI
31.10
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Medicine (publishedVersion)
  • PubMed (publishedVersion)
  • PubMed Central (acceptedVersion)
  • Digital Commons-TMC (Texas Medical Center) (submittedVersion)
  • University of Regensburg Publication Server (University of Regensburg) (submittedVersion)
Funders
Mark Foundation Endeavor Award
Topics
  • Health Sciences › Medicine › Oncology › CAR-T cell therapy research
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gut microbiota and health
  • Health Sciences › Medicine › Oncology › Cancer Immunotherapy and Biomarkers
Keywords
gut microbiota, CD19 CAR T-cell therapy, B-cell lymphoma, antibiotic exposure, Bifidobacterium longum, peptidoglycan biosynthesis, Akkermansia, Bacteroides, Ruminococcus, Eubacterium, treatment response prediction, peripheral T cells, systemic inflammation, tumor burden
MeSH terms
Gastrointestinal Microbiome, Receptors, Chimeric Antigen, Humans, Immunotherapy, T-Lymphocytes, Immunotherapy, Adoptive, Lymphoma, B-Cell, Antigens, CD19

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-07, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-07 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol and qwen3.8-omni-flash).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: A non-antibiotic-disrupted gut microbiome is associated with clinical responses to CD19-CAR-T cell cancer immunotherapy. https://app.cochat.ai/papers/10.1038/s41591-023-02234-6
The paper
Stein-Thoeringer, C. K., Saini, N. Y., Zamir, E., Blumenberg, V., Schubert, M.-L., Mor, U., Fante, M. A., Schmidt, S., Hayase, E., Hayase, T., Rohrbach, R., Chang, C.-C., McDaniel, L., Flores, I., Gaiser, R., Edinger, M., Wolff, D., Heidenreich, M., Strati, P., . . . Elinav, E. (2023). A non-antibiotic-disrupted gut microbiome is associated with clinical responses to CD19-CAR-T cell cancer immunotherapy. Nature Medicine. https://doi.org/10.1038/s41591-023-02234-6

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.