Causal relationship between gut microbiota and malignant lymphoma: a two-way two-sample mendelian randomization study

Introduction

Malignant lymphoma, a highly heterogeneous malignancy arising from lymphoid tissues, is classified into Hodgkin lymphoma (HL) and non-Hodgkin lymphoma (NHL) based on histopathological features and clinical manifestations (1,2). Patients with malignant lymphoma exhibit varying degrees of immune dysfunction. Characterized by immune dysregulation, its pathogenesis involves complex interactions between genetic susceptibility, environmental exposures, and host-microbiota crosstalk (3). While an observational study has reported associations between gut microbiota alterations and lymphoma risk, the causal relationship remains controversial due to inherent limitations: (I) confounding factors (e.g., chemotherapy-induced microbiota shifts in lymphoma patients); (II) reverse causation (lymphoma-driven immune dysfunction altering microbiota composition); and (III) heterogeneity in microbiota measurement methods (stool vs. mucosal sampling) (4). This ambiguity underscores the need for causal inference approaches. For instance, Xu et al. (5) found Fusobacteria enrichment in diffuse large B-cell lymphoma (DLBCL) patients, while relapse refractory DLBCL patients showed increased Enterococcus and cytochrome P450-related genes. However, population-based cohorts identified reduced microbial diversity in lymphoma patients, while others suggested that chemotherapy or lymphoma-induced immune alterations may secondarily disrupt gut microbiota (6).

The human body harbors hundreds of millions of microorganisms both internally and externally, which constitutes a small ecosystem. The gastrointestinal tract is home to numerous microbial populations, primarily bacteria, whose encoded genomes are over a hundred times larger than the that of the human genome (7,8). The interaction between the intestinal microbiota and the host has been extensively studied in various human cohorts and animal experiments, with some groups shown to participate in a variety of physiological regulatory activities of the host (9,10). Intestinal dysbiosis refers to alterations in the diversity and abundance of intestinal microbiota, which hold the potential to actively or passively affect the emergence and progression of an assortment of ailments, including malignant tumors. In recent years, an increasing number of studies have pointed to a clear association between the gut microbiota and malignant lymphoma, although the causal relationship and potential directionality have not been fully established (11,12). Notably, recent clinical studies suggest bidirectional interactions: lymphoma treatments (e.g., autologous hematopoietic stem cell transplantation) profoundly alter gut microbiota (13), while microbiota-derived signals may modulate lymphoma progression via immune checkpoint pathways (14). However, establishing causality in human studies remains challenging due to methodological constraints.

Mendelian randomization (MR) is an emerging epidemiological research method that employs genetic variation, such as single-nucleotide polymorphism (SNP) and instrumental variable (IV) to characterize the causal relationships between genetic susceptibility to relevant risk factors and the outcomes of interest (15-18). This method is particularly suited for disentangling the gut microbiota-lymphoma axis, as germline genetic variations precede disease onset and are less susceptible to reverse causality. Here, we hypothesize a bidirectional causal relationship: (I) gut microbiota composition causally influences lymphoma risk; (II) genetic predisposition to lymphoma alters microbiota profiles via disease-associated.

In this study, a two-sample MR approach was used to examine the causative link between gut microbiota and malignant lymphoma. Furthermore, we aimed to determine whether a genetic predisposition to lymphoma risk is causally related to the gut microbiota. The ultimate purpose of this study was to lay a reliable foundation for understanding the pathogenesis of intestinal microbiota-associated lymphoma and elucidating the bidirectional signaling mechanism of the intestinal microbiota-gut-lymphoma axis. We present this article in accordance with the STROBE-MR reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-303/rc).

Methods

The design and assumptions of the MR analysis

This MR analysis study was designed to meet the three key hypotheses of MR analyses. These hypotheses include a strong association between the IV and exposure, the absence of association between the IV and confounding factors, and the IV’s effect on the outcome solely through exposure (19). In this study, we employed forward MR analysis, treating each bacterial taxon within the gut microbiota as a separate exposure. Two-sample MR analysis was then conducted to identify the bacterial taxa causally related to malignant lymphoma. Subsequently, reverse MR analysis was performed, in which malignant lymphoma was treated as the exposure, while the bacterial taxa found in the forward MR analysis as being associated with malignant lymphoma were treated as the outcome. An overview of the study design is shown in Figure 1.

Figure 1 The design of MR study. GWAS, genome-wide association study; IEU, Integrative Epidemiology Unit; MR, Mendelian randomization.

Data sources

The genome-wide association study (GWAS) summary statistics for the 211 intestinal flora taxa were derived from the MiBioGen consortium (https://mibiogen.gcc.rug.nl/), a multiethnic population-scale project analyzing 16S rRNA gene sequencing profiles and genotyping data from 18,340 participants across Europe, North America, East Asia, and other regions. To ensure analytical consistency and maximize statistical power, we focused on the European subpopulation within MiBioGen for microbiome-host variation analyses, as this cohort provided the largest homogenous sample size across available datasets. After excluding 15 bacterial taxa lacking species-level classification, we retained 196 taxa (spanning 9 phyla, 16 orders, 32 families, and 119 genera) for downstream integration. The GWAS summary statistics for lymphoma were obtained from a cross-cohort meta-analysis combining data from the UK Biobank (UKB; application No. 47821) and FinnGen Release 3 (20). This integrated dataset is publicly available through the Integrative Epidemiology Unit (IEU) OpenGWAS database (https://gwas.mrcieu.ac.uk/datasets/, GWAS ID: ebi-a-GCST90018878), comprising 490,803 European-ancestry participants. Although detailed lymphoma subtyping was not explicitly provided in the original study, standardized case definitions were implemented in both cohorts. Notably, the core lymphoma subtypes in both datasets were HL and NHL. In the UKB cohort, malignant lymphoma phenotypes were classified using the ICD-10 coding system (C81–C96), with case data integrated from hospital inpatient records, primary care registries, and national death certification systems, undergoing rigorous quality control through triple validation. The FinnGen cohort employed a hybrid classification system combining FinnGen endpoint codes (LYMPHOMA_MALIGNANT) with ICD-10 equivalents (C81–C96), thus maintaining consistent subtyping architecture across cohorts. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

IV screening criteria

To ensure the robustness and effectiveness of the two-sample MR analysis methods, the IV must adhere to the assumptions of association, independence, and exclusivity, thereby guaranteeing the reliability of causal inference (15). In this study, we employed a bidirectional MR analysis, in which the IV was SNP, and the following selection criteria were employed:

• When the gut microbiota was the exposure, the SNP was required to meet the genome-wide significance threshold (P<1e−5). When malignant lymphoma was the exposure, the SNP was required to meet the genome-wide significance threshold (P<5e−8).
• The minor allele frequency (MAF) of the SNP was required to be greater than 0.05.
• The SNP could not be in a state of linkage disequilibrium (LD). SNPs in LD, determined by a LD parameter (r²) of 0.001 and a genetic distance of 10,000 kb, were excluded.
• Palindromic SNPs were not used as IVs and were directly removed.
• The F-statistic of the SNP was required to be greater than 10 to avoid the causal bias caused by weak IV.

  $$F=\left(\frac{R^2}{R^2-1}\right)\times \left(\frac{N-K-1}{K}\right)$$[1]

• The IV could not be correlated with confounding factors, and the selected IV could not be correlated with other factors that could have a causal relationship with the outcome.

The IV could only be associated with the outcome caused by the exposure, and it could not be directly associated with lymphoma. The IV could only influence the outcome of lymphoma through the exposure pathway.

MR analysis

This study employed the inverse variance-weighted (IVW) method as the primary approach for estimating causal associations. The IVW method yields unbiased and dependable estimates of causal associations when horizontal pleiotropy is absent. The IVW method through the incorporation of Wald estimates of each SNP of gut microbiota into the meta-analysis, which implements a weighted linear regression of IV and malignant lymphoma associations to obtain an MR estimate of the causal effect between gut microbiota and malignant lymphoma (21). Additionally, the MR Egger, weighted median, simple mode, and weighted mode were used as supplementary approaches for estimating causal associations, offering a broader confidence interval (CI) (22). In cases in which the causal association results obtained from these five methods were contradictory, priority was given to the IVW method in determining whether a positive result was present.

Pleiotropic analysis

When the pleiotropic effect of IV is present, it impacts the outcome through factors that are not related to the exposure factors. This results in the violation of the assumptions of independence and exclusivity. In order to address this issue, we leveraged the independent of direct effects (InSIDE) assumption, which assumes that IV is InSIDEs. The P value of the MR-Egger intercept test and the MR pleiotropy residual sum and outlier (MR-PRESSO) global test were used to initially identify the presence of horizontal pleiotropy. A P value greater than 0.05 indicated the absence of horizontal pleiotropy. Additionally, the MR-PRESSO analysis was conducted based on the InSIDE hypothesis to detect and mitigate horizontal pleiotropy via the removal of significant outliers. The NbDistribution parameter in the MR-PRESSO analysis was adjusted to 1,000, and outliers were detected and removed to account for level pleiotropy.

Heterogeneity analysis

The potential heterogeneity arising from variations in IV across analysis platforms, experiments, populations, and other factors can exert an impact on the outcomes of MR analyses. To assess this heterogeneity, we employed the IVW and MR-Egger tests, with a significance level of P<0.05 indicating the presence of heterogeneity in the study. Additionally, the Cochran IVW Q statistic test was used to evaluate the heterogeneity of the IVs. To ensure the robustness of the test outliers and validation results, a leave-one-out analysis was conducted by systematically omitting each SNP in turn.

Reverse MR analysis

In order to clarify the causal relationship between the gut microbiota and malignant lymphoma, we additionally conducted a reverse MR analysis of the bacterial composition that was found to have a causal association with malignant lymphoma in the forward MR analysis. The approach and parameters employed were aligned with those used in the initial MR analysis.

Statistical analysis

The statistical analysis in this study was conducted using R software version 4.3.1 (The R Foundation for Statistical Computing, Vienna, Austria). Additionally, MR analysis was performed using the R package “TwosampleMR” (version 0.5.7).

Results

Based on the experimental conditions set in this study, the exposure and outcome data were filtered, which included a genome-wide significance threshold, LD testing, MR-PRESSO testing, and validation of F statistics. To ensure the robustness and reliability of the study results, all SNPs with an F statistic less than 10 and SNP detected as outliers by MR-PRESSO were removed. Additionally, it should be noted that there was a hierarchical relationship between gut microbiota classifications. For example, the family Rikenellaceae belongs to the bacterial group within the phylum Bacteroidetes. Therefore, there may be overlaps in the SNP and their associated orders between the two.

This study conducted five MR analyses and found seven causal relationships between the gut microbiota and malignant lymphoma (Figure 2). The results revealed that certain phyla, classes, families, and genera of bacteria were associated with increased odds of malignant lymphoma emergence. The seven gut microbiota taxa linked to and malignant lymphoma were as follows: the phylum Bacteroidetes [odds ratio (OR) =1.31; 95% CI: 1.02–1.68; P=0.03], the class Bacilli (OR =1.22; 95% CI: 1.00–1.49; P=0.048), the family Rikenellaceae (OR =1.27; 95% CI: 1.04–1.55; P=0.02), the gennus Eubacterium nodatum group (OR =1.13; 95% CI: 1.00–1.27; P=0.046), the genus Oxalobacter (OR =1.23; 95% CI: 1.06–1.43; P=0.006), the genus Parabacteroides (OR =1.41; 95% CI: 1.00–1.99; P=0.049), and the genus Sellimonas (OR =1.18; 95% CI: 1.03–1.35; P=0.02) (Figure 2 and Table 1). The causal estimation direction indicated by the MR-Egger method, the weighted median method, the simple mode method, and the weighted mode method analyses aligned with the above-described results, but not significantly so (or only nominally significant) (Figure 3).

Figure 2 Causal analysis of intestinal microbiota and malignant lymphoma (P<0.05). Blue dots: OR >1 (positive association); red dots: OR ≤1 (negative/neutral association). IVW, inverse variance weighting; MR, Mendelian randomization; OR, odds ratio.

Figure 3 MR analysis scatter plot for the associations between seven gut bacterial taxa and malignant lymphoma. Relationship between malignant lymphoma and (A) phylum Bacteroidetes, (B) class Bacilli, (C) family Rikenellaceae, (D) genus Eubacterium nodatum, (E) genus Oxalobacter, (F) genus Parabacteroides, and (G) genus Sellimonas. MR, Mendelian randomization; SNP, single-nucleotide polymorphism.

The sensitivity analysis was conducted on the seven bacterial traits found in the study (Table 2). The results from the intercept of the MR-Egger regression and the MR-PRESSO method indicated that these SNPs did not exhibit horizontal pleiotropy or outliers (P>0.05). Furthermore, the Cochrane IVW Q test results (P>0.05) suggested no significant heterogeneity among these SNP. The robustness of the findings was further validated by the leave-one-out analysis (Figure 4). Overall, the results of the MR analysis in this study are robust and credible.

Figure 4 Forest plot of the causal effect of seven gut bacterial taxa associated SNPs on malignant lymphoma. The leave-one-out sensitive analysis was conducted to assess the association between gut microbiota and malignant lymphoma via the systematic removal of individual SNP (P<1×10⁻⁵). Relationship between malignant lymphoma and (A) phylum Bacteroidetes, (B) class Bacilli, (C) family Rikenellaceae, (D) genus Eubacterium nodatum, (E) genus Oxalobacter, (F) genus Parabacteroides, and (G) genus Sellimonas. MR, Mendelian randomization; SNP, single-nucleotide polymorphism.

The reverse MR analysis carried out in this study, focusing on malignant lymphoma and the seven bacterial taxa, revealed no noteworthy evidence of reverse causality between these variables.

Discussion

We completed an MR analysis to investigate the potential causal relationship between gut microbiota and malignant lymphoma at the level of genetic prediction. Previous investigations on the link between the gut microbiota and malignant lymphoma have primarily been conducted on a small scale, employing omics analysis in animal models and limited clinical cohorts. These studies, however, lack the representativeness required for general population analysis. In contrast, our study leveraged extensive data on intestinal microbiota and malignant lymphoma in multiethnic, European populations and employed MR analysis to ascertain the relationship between changes in specific intestinal microbiota abundance and the onset and progression of malignant lymphoma. By using genetic variables as a tool, our study revealed a potential association between the gut microbiota and malignant lymphoma, while mitigating the impact of confounding factors on the causal relationship. Moreover, our findings offer robust genetic evidence supporting the role of the gut microbiota in malignant lymphoma. Specifically, our two-sample MR analysis found seven bacterial taxa, including one phylum (phylum Bacteroidetes), one class (class Bacilli), one family (family Rikenellaceae), four genera (gennus Eubacterium nodatum group, genus Oxalobacter, genus Parabacteroides, and genus Sellimonas), which exhibited a positive correlation with the risk of malignant lymphoma. This novel genetic evidence adds to our understanding of the causative link between the intestinal microbiome and malignant lymphoma.

The findings of this bidirectional two-sample MR analysis provide further support to the existing body of research on the association of gut microbiota with lymphoma. Recently, there has been a growing interest in the connection between the microbiome and human health. Numerous studies have indicated a significant correlation between the balance of the gut microbiota and various malignant tumors, hematologic malignancies, metabolic disorders, and immune-related diseases (23-25). Research has also indicated a close relationship between specific components of the gut microbiota and the pathogenesis of malignant lymphoma, influencing disease progression and patient prognosis (26,27). Bacteroidetes is widespread in the distal intestine of humans (28), and studies have reported an increased abundance of Parabacteroides phylum bacteria among lymphoma patients, which has been associated with sepsis, abdominal infections, and a poor prognosis (29-31). Moreover, in vivo evidence suggests that the gut microbiome can impact the efficacy of chimeric antigen receptor T-cell (CAR-T) therapy for hematologic malignancies by altering the microbial niche or producing metabolites (32). In a predictive model, Bacteroides, a taxonomic group, was identified as a key predictor of long-term response to CAR-T cell therapy (33). Family Rikenellaceae, a member of the Bacteroidetes phylum, is negatively correlated with T-cell activation and blood lymphocyte concentration (34). Safety bacillus, a group within the Bacilli class, has been linked to malignant T-cell activation and cytokine secretion, triggering malignant lymphoma in patients with cutaneous T-cell lymphoma (35). Certain bacteria belonging to the Pseudobacterium genus have been found to inhibit or mitigate the development of lymphoma through butyrate production (12). However, in a study investigating the intestinal flora associated with NLH of the small intestine, the disease group exhibited a notably higher relative abundance of the Eubacterium nodatum group in comparison to the control group (36). NLH is a rare benign lesion with the potential to transform into lymphoma. Its pathogenesis is caused by the accumulation of plasma cell precursors due to a maturation defect in B-lymphocyte development. Alterations in the relative abundance of the Eubacterium nodatum group can potentially increase the transformation of NLH into lymphoma (36,37). The increased relative abundance of Parabacteroides species among patients with lymphoma contributes to dysbiosis, leading to severe abdominal infections and impacting prognosis (29). Although these previous studies have examined the association between intestinal flora and lymphoma, the direct causal relationship between the two cannot be definitively characterized due to the limitations in their respective research methods.

This study represents an innovative analysis that employed a two-way, two-sample MR approach to investigate the causa link between the gut microbiome and malignant lymphoma. The study’s credibility and robustness are reinforced by three key hypotheses upon which the MR analysis is built. Using SNP as the IV, our study effectively eliminated the potential confounding influence of environmental factors and other variables on the observed results. The findings of this study offer a novel theoretical framework for comprehending the microbe-gut-lymphoma axis and highlight the potential of microbiome-oriented lymphoma treatment strategies. Furthermore, these results contribute to the identification of new biomarkers for malignant lymphoma. Certainly, to conclusively establish the potential causal link between these seven gut microbiota and malignant lymphoma, it is imperative to conduct more exhaustive and rigorous research endeavors. This includes rigorously adhering to Koch’s postulates and other pertinent scientific criteria to ascertain the specificity, pathogenicity, and reproducibility of the causative agent. Furthermore, it is necessary to validate and refine this biological association within the framework of real-world epidemiological studies, utilizing large-scale cohort studies, case-control studies, and incorporating advanced bioinformatics analysis and statistical methodologies. Additionally, animal model experiments and in vitro cell culture experiments should be employed to further investigate the mechanisms underlying the relationship between these microbiota and malignant lymphoma.

While this two-sample MR study provides novel insights into bidirectional causal relationships between gut microbiota and malignant lymphoma, several limitations warrant acknowledgment to guide future research. Population stratification is a critical concern. Both GWAS datasets (MiBioGen and OpenGWAS) were predominantly derived from European-descent cohorts, which may limit the generalizability of findings to globally diverse populations. Ethnicity-specific genetic architectures of the gut microbiome and lymphoma susceptibility could introduce biases, as allelic frequencies and microbiota compositions vary across ancestral backgrounds. While multi-ethnic cohorts are increasingly prioritized in genetic studies, the current reliance on European data underscores the need for replication in non-European populations to assess cross-population validity. Conservative instrument selection criteria (e.g., F-statistic >10, P<5×10⁻⁸) may have excluded weaker yet biologically relevant associations. Such thresholds balance statistical power against instrument validity but risk discarding SNP with modest effects. Future work could employ more permissive criteria or leverage polygenic scores to capture broader microbial-host interactions, provided rigorous sensitivity analyses confirm robustness. Biological validation remains incomplete. While MR analyses support causal inference, mechanistic links between implicated taxa and lymphoma pathogenesis remain unresolved. Functional studies—including metagenomic characterization, metabolomic profiling, and animal models—are necessary to elucidate whether these associations reflect direct microbial roles (e.g., toxin production, immune modulation) or indirect effects (e.g., microbiome-mediated drug metabolism). The taxonomic hierarchy and instrument overlap pose methodological challenges. For example, SNP associated with the Bacteroidetes phylum may also influence lower-level taxa (e.g., Bacteroides genus), complicating the interpretation of distinct causal effects. The potential redundancy introduced by hierarchical bacterial classifications (e.g., phylum vs. genus). To mitigate this, we applied stringent quality control during instrument selection, prioritizing taxa with distinct genomic features and specifically: we used independent GWAS SNP (r²<0.001) to avoid correlation between instruments for different taxa. The con minimal overlap in SNP instruments. Consistency of results across multiple MR methods (IVW, MR-Egger, weighted median) suggests robustness despite taxonomic overlap. Future studies should harmonize taxonomic classifications and employ methods like hierarchical MR or taxon-specific GWAS to disentangle these relationships. What is more, participant overlap between discovery cohorts could not be excluded due to incomplete metadata. While MiBioGen and OpenGWAS are distinct consortia, shared contributors might inflate effect estimates. Cross-database collaboration to clarify participant demographics and improve data transparency would mitigate this risk. Collectively, these limitations highlight priorities for translational follow-up: (I) expansion to multiethnic cohorts, (II) integration of multi-omics data to validate microbial mechanisms, and (III) prospective validation of causal taxa in clinical or preclinical models. Such efforts will be critical to advancing microbiome-targeted strategies for lymphoma prevention or management.

In summary, our study represents a comprehensive exploration of the possible causal link between the gut microbiota and malignant lymphoma. Our findings may provide a novel theoretical framework for the investigation of lymphoma-related intestinal flora in future research.

Conclusions

Our study examined the potential causal connection between the gut microbiota and malignant lymphoma. Our findings contribute to the theoretical foundation of future research on the gut microbiota’s relationship with lymphoma, facilitating potential advancements in lymphoma diagnosis, treatment, and prevention strategies in clinic.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE-MR checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-303/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-303/prf

Funding: The study was supported by the Guangxi Zhuang Autonomous Region Medical Health Appropriate Technology Development and Application Promotion Project Commission (No. S2020038).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-303/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Elenitoba-Johnson KSJ, Lim MS. New Insights into Lymphoma Pathogenesis. Annu Rev Pathol 2018;13:193-217. [Crossref] [PubMed]
2. Swerdlow SH, Campo E, Pileri SA, et al. The 2016 revision of the World Health Organization classification of lymphoid neoplasms. Blood 2016;127:2375-90. [Crossref] [PubMed]
3. Sheikh IN, Elgehiny A, Ragoonanan D, et al. Management of Aggressive Non-Hodgkin Lymphomas in the Pediatric, Adolescent, and Young Adult Population: An Adult vs. Pediatric Perspective. Cancers (Basel) 2022;14:2912. [Crossref] [PubMed]
4. Hrubesz G, Leigh J, Ng TL. Understanding the relationship between breast cancer, immune checkpoint inhibitors, and gut microbiota: a narrative review. Transl Breast Cancer Res 2024;5:31. [Crossref] [PubMed]
5. Xu Y, Shi C, Qian J, et al. The gut microbiota is altered significantly in primary diffuse large b-cell lymphoma patients and relapse refractory diffuse large b-cell lymphoma patients. Clin Transl Oncol 2024; Epub ahead of print. [Crossref]
6. Guevara-Ramírez P, Cadena-Ullauri S, Paz-Cruz E, et al. Gut Microbiota Disruption in Hematologic Cancer Therapy: Molecular Insights and Implications for Treatment Efficacy. Int J Mol Sci 2024;25:10255. [Crossref] [PubMed]
7. Structure, function and diversity of the healthy human microbiome. Nature 2012;486:207-14. [Crossref] [PubMed]
8. Qin J, Li R, Raes J, et al. A human gut microbial gene catalogue established by metagenomic sequencing. Nature 2010;464:59-65. [Crossref] [PubMed]
9. Geva-Zatorsky N, Sefik E, Kua L, et al. Mining the Human Gut Microbiota for Immunomodulatory Organisms. Cell 2017;168:928-943.e11. [Crossref] [PubMed]
10. Smith K, McCoy KD, Macpherson AJ. Use of axenic animals in studying the adaptation of mammals to their commensal intestinal microbiota. Semin Immunol 2007;19:59-69. [Crossref] [PubMed]
11. Shi Z, Hu G, Li MW, et al. Gut microbiota as non-invasive diagnostic and prognostic biomarkers for natural killer/T-cell lymphoma. Gut 2023;72:1999-2002. [Crossref] [PubMed]
12. Lu H, Xu X, Fu D, et al. Butyrate-producing Eubacterium rectale suppresses lymphomagenesis by alleviating the TNF-induced TLR4/MyD88/NF-κB axis. Cell Host Microbe 2022;30:1139-1150.e7. [Crossref] [PubMed]
13. Yu H, Fu L, Liao X, et al. Efficacy and safety of autologous hematopoietic stem cell transplantation in the treatment of malignant lymphoma after chemotherapy: a systematic review and meta-analysis. Transl Cancer Res 2022;11:888-96. [Crossref] [PubMed]
14. Nisar MF, Yan T, Cai Y, et al. Immuno-oncological Challenges and Chemoresistance in Veterinary Medicine: Probiotics as a New Strategic Tool. Probiotics Antimicrob Proteins 2025; Epub ahead of print. [Crossref]
15. Sekula P, Del Greco M F, Pattaro C, et al. Mendelian Randomization as an Approach to Assess Causality Using Observational Data. J Am Soc Nephrol 2016;27:3253-65. [Crossref] [PubMed]
16. Jin Q, Ren F, Dai D, et al. The causality between intestinal flora and allergic diseases: Insights from a bi-directional two-sample Mendelian randomization analysis. Front Immunol 2023;14:1121273. [Crossref] [PubMed]
17. Li Z, Zhu G, Lei X, et al. Genetic support of the causal association between gut microbiome and COVID-19: a bidirectional Mendelian randomization study. Front Immunol 2023;14:1217615. [Crossref] [PubMed]
18. Long Y, Tang L, Zhou Y, et al. Causal relationship between gut microbiota and cancers: a two-sample Mendelian randomisation study. BMC Med 2023;21:66. [Crossref] [PubMed]
19. Emdin CA, Khera AV, Kathiresan S. Mendelian Randomization. JAMA 2017;318:1925-6. [Crossref] [PubMed]
20. Kurilshikov A, Medina-Gomez C, Bacigalupe R, et al. Large-scale association analyses identify host factors influencing human gut microbiome composition. Nat Genet 2021;53:156-65. [Crossref] [PubMed]
21. Burgess S, Labrecque JA. Mendelian randomization with a binary exposure variable: interpretation and presentation of causal estimates. Eur J Epidemiol 2018;33:947-52. [Crossref] [PubMed]
22. Slob EAW, Burgess S. A comparison of robust Mendelian randomization methods using summary data. Genet Epidemiol 2020;44:313-29. [Crossref] [PubMed]
23. Savage DC. Microbial ecology of the gastrointestinal tract. Annu Rev Microbiol 1977;31:107-33. [Crossref] [PubMed]
24. Frick JS, Autenrieth IB. The gut microflora and its variety of roles in health and disease. Curr Top Microbiol Immunol 2013;358:273-89. [Crossref] [PubMed]
25. D’Angelo CR, Sudakaran S, Callander NS. Clinical effects and applications of the gut microbiome in hematologic malignancies. Cancer 2021;127:679-87. [Crossref] [PubMed]
26. Yuan L, Wang W, Zhang W, et al. Gut Microbiota in Untreated Diffuse Large B Cell Lymphoma Patients. Front Microbiol 2021;12:646361. [Crossref] [PubMed]
27. Wotherspoon AC, Ortiz-Hidalgo C, Falzon MR, et al. Helicobacter pylori-associated gastritis and primary B-cell gastric lymphoma. Lancet 1991;338:1175-6. [Crossref] [PubMed]
28. Bäckhed F, Ley RE, Sonnenburg JL, et al. Host-bacterial mutualism in the human intestine. Science 2005;307:1915-20. [Crossref] [PubMed]
29. Cobo F, Pérez-Carrasco V, Gómez-Vicente E, et al. Parabacteroides goldsteinii abdominal infection in a patient with lymphoma. Infect Dis Now 2022;52:117-9. [Crossref] [PubMed]
30. Appelbaum PC, Ballard JO, Eyster ME. Septicemia due to Capnocytophaga (Bacteroides ocharaceus) in Hodgkin’s disease. Ann Intern Med 1979;90:716-7. [Crossref] [PubMed]
31. Bilgrami S, Bergstrom SK, Peterson DE, et al. Capnocytophaga bacteremia in a patient with Hodgkin’s disease following bone marrow transplantation: case report and review. Clin Infect Dis 1992;14:1045-9. [Crossref] [PubMed]
32. Hu Y, Li J, Ni F, et al. CAR-T cell therapy-related cytokine release syndrome and therapeutic response is modulated by the gut microbiome in hematologic malignancies. Nat Commun 2022;13:5313. [Crossref] [PubMed]
33. Stein-Thoeringer CK, Saini NY, Zamir E, et al. A non-antibiotic-disrupted gut microbiome is associated with clinical responses to CD19-CAR-T cell cancer immunotherapy. Nat Med 2023;29:906-16. [Crossref] [PubMed]
34. Riazati N, Kable ME, Stephensen CB. Association of intestinal bacteria with immune activation in a cohort of healthy adults. Microbiol Spectr 2023;11:e0102723. [Crossref] [PubMed]
35. Dehner CA, Ruff WE, Greiling T, et al. Malignant T Cell Activation by a Bacillus Species Isolated from Cutaneous T-Cell Lymphoma Lesions. JID Innov 2022;2:100084. [Crossref] [PubMed]
36. Jiang QL, Lu Y, Zhang MJ, et al. Mucosal bacterial dysbiosis in patients with nodular lymphoid hyperplasia in the terminal ileum. World J Gastroenterol 2022;28:811-24. [Crossref] [PubMed]
37. Volle RL. Blockade by 9-aminoacridine of potassium fluxes in frog sartorius muscle. Biochem Pharmacol 1971;20:315-24. [Crossref] [PubMed]

(English Language Editor: J. Gray)