CoChat

The DOI matches its Crossref record; 153 of the 157 references it cites were found in a registry or an index; 29 of the 42 sources its claims cite were checked for support.

Microbiome meets immunotherapy: unlocking the hidden predictors of immune checkpoint inhibitors

Open access
Huang, Lihaoyun, Li, Yu, Zhang, Chunyan, Jiang, Aimin, Zhu, Lingxuan, Mou, Weiming, Li, Kailai, Zhang, Jian, Cui, Cui, Cui, Xinfang, Lin, Anqi, Luo, Peng, Wei, Ting · 2025 · npj Biofilms and Microbiomes

Abstract

Immune checkpoint inhibitors (ICIs) have revolutionized cancer immunotherapy, with the microbiome significantly influencing treatment outcomes. Analysis of 4663 studies (2014.01-2024.10) identified 71 eligible randomized controlled trials (RCTs) and cohort studies (41 viral, 30 bacterial). Analyses included subgroup assessments by cancer type, microbial taxa, and ICI regimens. Among 4663 identified studies, 71 met inclusion criteria (41 viral, 30 bacterial). Viral status, particularly hepatitis B virus (HBV) and human papillomavirus (HPV), significantly associated with ORR and DCR. Bacterial enrichment correlated with improved survival in hepatobiliary (OS: HR = 4.33, 95%CI: 2.20-8.50) and lung cancers (PFS: HR = 1.70, 95%CI: 1.04-2.78). Multi-microbiome models demonstrated superior outcome prediction, with microbial diversity correlating with improved PFS (HR = 0.64, 95%CI: 0.42-0.98). Viral status showed cancer-specific associations with SAEs. The microbiome serves as a valuable predictor of ICI outcomes. Future studies should emphasize large-scale RCTs, standardized assessment methods, and host-microbiome interactions.

Overview

CoChat Summary

TL;DR

This PRISMA meta-analysis of 71 studies (41 viral, 30 bacterial) found baseline microbiome status predicts ICI efficacy, especially viral status for ORR, with cancer- and regimen-specific effects. Bacterial enrichment and multi-microbiome models showed tumor-specific survival associations, while safety associations were limited and heterogeneous.

Key findings

  • Among 71 included studies (41 viral, 30 bacterial), baseline microbiota was associated with ORR: pooled RR 1.29 (95% CI 1.07–1.55, P = 0.049, N = 34, I² = 55.6%); virus-related cohorts had RR 1.28 (95% CI 1.11–1.47, P < 0.001, N = 27, I² = 47.9%).
  • Viral status predicted ORR (RR 1.36, 95% CI 1.09–1.69, P = 0.008, N = 29), with HBV RR 1.36 (95% CI 1.10–1.67, P = 0.005, N = 9) and HPV RR 1.61 (95% CI 1.19–2.18, P = 0.002, N = 7); hepatobiliary cancers RR 1.71 (95% CI 1.22–2.38, P = 0.002) and monotherapy RR 1.46 (95% CI 1.22–1.74, P < 0.001, N = 16). DCR pooled RR was 0.99 (95% CI 0.71–1.39, I² = 82.7%), but HBV was positive (RR 1.29, 95% CI 1.10–1.50, P = 0.001, N = 4) and EBV negative (RR 0.63, 95% CI 0.52–0.77, P < 0.001, N = 3).
  • Bacterial enrichment was associated with OS in hepatobiliary cancers (HR 4.33, 95% CI 2.20–8.50, P < 0.001, N = 3) and PFS in hepatobiliary cancers (HR 2.33, 95% CI 1.54–3.53, P < 0.001, N = 3) and lung cancer (HR 1.70, 95% CI 1.04–2.78, P = 0.035, N = 2); gastrointestinal cancer PFS was negatively associated (HR 0.82, 95% CI 0.70–0.96, P = 0.016, N = 4). Multi-microbiome models were associated with PFS (HR 0.64, 95% CI 0.42–0.98, P = 0.039, N = 2).
  • Safety: viral status was not significantly associated with overall irAEs (RR 1.04, 95% CI 0.92–1.17, P > 0.05) or SAEs (RR 1.10, 95% CI 0.57–2.25, P = 0.570, I² = 72.8%); hepatobiliary SAEs showed RR 0.62 (95% CI 0.39–1.00, P = 0.048, N = 3) and monotherapy SAEs RR 2.31 (95% CI 1.35–3.94, P = 0.002, N = 2).

Methods

  • PRISMA-compliant meta-analysis registered in PROSPERO (CRD42025630951); searched PubMed, Web of Science, Medline, Embase, Cochrane Library, Wiley Online Library, ClinicalTrials.gov, and ESMO/ASCO/AACR conference abstracts from January 1, 2014, to October 9, 2024; human studies only, with no restrictions on publication type or language.
  • Included RCTs or cohort studies of pathologically confirmed malignancies with in vivo microbiome profiling and associations with ICI outcomes; required ORR, DCR, PR, CR, SD, PFS, OS, or unadjusted HRs, and/or irAEs/SAEs; excluded animal studies, studies lacking pre-treatment microbiome profiles or specific microbiome classifications, and studies reporting only adjusted HRs.
  • Three researchers independently screened and extracted data; quality assessed with Newcastle–Ottawa Scale for cohorts and Cochrane ROB for RCTs; publication bias assessed with funnel plots and Egger’s test; pooled RRs for ORR/DCR and HRs for PFS/OS using RevMan 5.4 and R 4.4.1, with fixed-effect models for I² < 50% and random-effect models for I² > 50%, significance at P < 0.05.

Limitations

  • Stated: the analysis primarily included moderate- to high-quality cohort studies with a lack of RCTs, limiting confounding control and causal inference; endpoints were incomplete and heterogeneous across studies (viral studies lacked long-term follow-up; bacterial studies lacked AE data); bacterial presence/enrichment was often analyzed using binary classifications without standardized sequencing or categorization; sample sizes and cohorts were limited, most bacterial samples were fecal, and few were from extracellular vesicles or tumor tissues; subgroup analyses were insufficiently deep.
  • Evident: many subgroup estimates relied on single cohorts (e.g., hepatobiliary DCR, melanoma/mesothelioma OS, phylum-level bacterial results), and some findings had confidence intervals whose upper bound equaled 1.0 (hepatobiliary SAEs RR 0.62, 95% CI 0.39–1.00; multiple bacterial taxa OS HR 0.54, 95% CI 0.29–1.0), requiring validation. High heterogeneity was present in some pooled analyses (DCR I² = 82.7%; SAEs I² = 72.8%; multiple bacterial communities PFS I² = 81.3%).
  • Evident: publication bias tests were not statistically significant for ORR/OS/PFS, but funnel plots showed asymmetry and substantial between-study heterogeneity (τ² values), so results should be interpreted cautiously.

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

Integrity signals

Registry

Crossref · 10.1038/s41522-025-00819-2

Cites

153 of 153 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig. 1 Workflow of identification of studies through databases following PRISMA guidelines.
Fig. 1 Workflow of identification of studies through databases following PRISMA guidelines.CC BY-NC-ND · PubMed Central
Fig. 2 Microbial predictors of immunotherapy efficacy based on objective response rate (ORR) and risk ratio (RR). A Association of microbiota with immunotherapy efficacy. B Association of viruses with immunotherapy efficacy across different virus types, tumor types, and immunotherapy strategies.
Fig. 2 Microbial predictors of immunotherapy efficacy based on objective response rate (ORR) and risk ratio (RR). A Association of microbiota with immunotherapy efficacy. B Association of viruses with immunotherapy efficacy across different virus types, tumor types, and immunotherapy strategies.CC BY-NC-ND · PubMed Central
Fig. 3 Virus-associated prediction of immunotherapy efficacy measured by disease control rate (DCR) and risk ratio (RR), across different virus types, tumor types, and immunotherapy strategies.
Fig. 3 Virus-associated prediction of immunotherapy efficacy measured by disease control rate (DCR) and risk ratio (RR), across different virus types, tumor types, and immunotherapy strategies.CC BY-NC-ND · PubMed Central
Fig. 4 Bacterial predictors of immunotherapy outcomes across various tumor types. A Overall survival(OS); B Progression free survival(PFS).
Fig. 4 Bacterial predictors of immunotherapy outcomes across various tumor types. A Overall survival(OS); B Progression free survival(PFS).CC BY-NC-ND · PubMed Central
Fig. 5 Bacteria as predictive factors for immunotherapy outcomes in various bacterial phyla and mono-/poly-bacterial models. Association of different bacterial phyla with overall survival (OS) (A) and progression free survival (PFS) (B). Comparison of mono-bacterial vs. poly-bacterial models for OS (C) and PFS (D).
Fig. 5 Bacteria as predictive factors for immunotherapy outcomes in various bacterial phyla and mono-/poly-bacterial models. Association of different bacterial phyla with overall survival (OS) (A) and progression free survival (PFS) (B). Comparison of mono-bacterial vs. poly-bacterial models for OS (C) and PFS (D).CC BY-NC-ND · PubMed Central

Verification

Claims the paper makes

21 Valid · 4 Not valid · 4 Not checked · 81 Own results and interpretations

54 more claims · 7 with their cited sources checked

Create a free account to see all 110 claims and to verify the 2 references still open.

See all claims — free account
  • It may state two separate claims as one. 42
  • Probably not a substantial claim. 14
  • Borderline: one of the tests for a substantial claim is unclear. 1

Create a free account to see these claims.

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

2 of 2

Citations

Cites

Show all 157 references

Research works printed but not linked: 1

  1. Routy, B. et al. Gut microbiome influences efficacy of PD-1–based immunotherapy against epithelial tumors. Science 359, 91–97 (2018). · DOI · Paper page

Cited by

See all 28

Metadata

Authors
Huang, L., Li, Y., Zhang, C., Jiang, A., Zhu, L., Mou, W., Li, K., Zhang, J., Cui, C., Cui, X., Lin, A., Luo, P., & Wei, T.
Affiliations
  • Zhujiang Hospital (CN)
  • Second Military Medical University (CN)
  • Shanghai Jiao Tong University (CN)
  • Sun Yat-sen University Cancer Center (CN)
Corresponding author
  • Lin, Anqi
  • Luo, Peng
  • Wei, Ting
Venue
npj Biofilms and Microbiomes
Published
2025-09-02
Volume
11
Issue
1
Pages
180-180
ISSN
2055-5008 (ISSN-L)
OpenAlex
W4413903278
Cited by
28
References
157
Citations by year
2026: 21 · 2025: 7
FWCI
11.50
Type
journal article
Language
en
Open access
Licence
Locations
  • npj Biofilms and Microbiomes (publishedVersion, cc-by-nc-nd)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by-nc-nd)
  • PubMed (publishedVersion)
  • PubMed Central (publishedVersion, cc-by-nc-nd)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • Guangdong Provincial Medical Science and Technology Research (B2023341)
  • National Natural Science Foundation of China (82172750)
Topics
  • Health Sciences › Medicine › Oncology › Cancer Immunotherapy and Biomarkers
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gut microbiota and health
  • Life Sciences › Immunology and Microbiology › Immunology › Immunotherapy and Immune Responses
Keywords
gut microbiota, immune checkpoint inhibitors, cancer immunotherapy, treatment response prediction, hepatitis B virus, human papillomavirus, bacterial flora, hepatobiliary cancer, lung cancer, microbiome diversity, progression-free survival, overall survival, adverse events, host-microbiome interactions
MeSH terms
Immune Checkpoint Inhibitors, Bacteria, Humans, Immunotherapy, Neoplasms, Randomized Controlled Trials as Topic, Treatment Outcome, Microbiota

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 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

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

Cited sources are checked with a language model’s help (gpt-6-sol).

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: Microbiome meets immunotherapy: unlocking the hidden predictors of immune checkpoint inhibitors. https://app.cochat.ai/papers/10.1038/s41522-025-00819-2
The paper
Huang, L., Li, Y., Zhang, C., Jiang, A., Zhu, L., Mou, W., Li, K., Zhang, J., Cui, C., Cui, X., Lin, A., Luo, P., & Wei, T. (2025). Microbiome meets immunotherapy: unlocking the hidden predictors of immune checkpoint inhibitors. npj Biofilms and Microbiomes. https://doi.org/10.1038/s41522-025-00819-2

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.