Intratumor microbiota as a novel potential prognostic indicator in mesothelioma

Our plain-English summary of Intratumor microbiota as a novel potential prognostic indicator in mesothelioma, by F. Pentimalli, M. Krstic-Demonacos, C. Costa, et al (Frontiers in Immunology, 2023).

The study

• Authors: F. Pentimalli, M. Krstic-Demonacos, C. Costa, L. Mutti, E.Y. Bakker
• Journal: Frontiers in Immunology
• Published: 2023;14:1129513
• Cited by: 8 articles, as counted by NIH iCite
• Read it: PubMed · Publisher · Free full text

Abstract

Introduction: Despite increased attention on immunotherapy, primarily immune checkpoint blockade, as a therapeutic approach for mesothelioma (MMe), its efficacy and tolerability remain questioned. One potential explanation for different responses to immunotherapy is the gut and intratumor microbiota; however, these remain an underexplored facet of MMe. This article highlights the cancer intratumor microbiota as a novel potential prognostic indicator in MMe.

Methods: TCGA data on 86 MMe patients from cBioPortal underwent bespoke analysis. Median overall survival was used to divide patients into "Low Survivors" and "High Survivors". Comparison of these groups generated Kaplan-Meier survival analysis, differentially expressed genes (DEGs), and identification of differentially abundant microbiome signatures. Decontamination analysis refined the list of signatures, which were validated as an independent prognostic indicator through multiple linear regression modelling and Cox proportional hazards modelling. Finally, functional annotation analysis on the list of DEGs was performed to link the data together.

Results: 107 genera signatures were significantly associated with patient survival (positively or negatively), whilst clinical characteristic comparison between the two groups demonstrated that epithelioid histology was more common in "High Survivors" versus biphasic in "Low Survivors". Of the 107 genera, 27 had published articles related to cancer, whilst only one (Klebsiella) had MMe-related published articles. Functional annotation analysis of the DEGs between the two groups highlighted fatty acid metabolism as the most enriched term in "High Survivors", whilst for "Low Survivors" the enriched terms primarily related to cell cycle/division. Linking these ideas and findings together is that the microbiome influences, and is influenced by, lipid metabolism. Finally, to validate the independent prognostic value of the microbiome, multiple linear regression modelling as well as Cox proportional hazards modelling were employed, with both approaches demonstrating that the microbiome was a better prognostic indicator than patient age or stage of the cancer.

Discussion: The findings presented herein, alongside the very limited literature from scoping searches to validate the genera, highlight the microbiome and microbiota as a potentially rich source of fundamental analysis and prognostic value. Further in vitro studies are needed to elucidate the molecular mechanisms and functional links that may lead to altered survival.

Copyright © 2023 Pentimalli, Krstic-Demonacos, Costa, Mutti and Bakker.

In plain English

This was a computer analysis of public data on 86 mesothelioma patients in the 'Mesothelioma (TCGA, PanCancer Atlas)' dataset on cBioPortal. The authors state that only pleural mesothelioma was covered. Asbestos exposure was not examined. Patients were split at the median overall survival into 'Low Survivors' and 'High Survivors'. The groups were compared for clinical features, gene activity, and microbe signals in tumor samples. After likely contaminants were removed, 107 microbe groups (genera) differed between the groups. Of these, 103 were more abundant in High Survivors and four, including Klebsiella, in Low Survivors. Biphasic tumors were more common among Low Survivors. In a survival model, each one-unit rise in a summed abundance score (log RNA Seq CPM) for the 103 genera was linked to lower risk of death (hazard ratio 0.986, 95% confidence interval 0.978 to 0.995). Non-epithelioid histology was linked to a higher risk (hazard ratio 1.781, 1.013 to 3.131). Age and stage were not statistically significant. The authors called the microbiome a potential prognostic indicator needing laboratory study. They noted the microbe data's source study was later retracted, so specific genera may be uncertain.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Retrospective bioinformatic analysis of existing public patient data (exploratory study), with Kaplan-Meier, differential gene expression, multiple linear regression and Cox proportional hazards analyses
• Who was studied: 86 mesothelioma patients in the 'Mesothelioma (TCGA, PanCancer Atlas)' study on cBioPortal (accessed 05 April 2022), after one patient without survival data was excluded. The patients were divided at the median overall survival into 'Low Survivors' and 'High Survivors'. The authors state only pleural mesothelioma was explored. The places and dates of patient diagnosis are not given.
• Exposure: Not an asbestos exposure study. The factor studied was intratumor microbiome signatures (abundance of microbial genera detected in tumor RNA sequencing data, in log RNA Seq CPM), compared with age, stage and tumor histology.
• Main results:
  • 175 genera initially differed in abundance between the groups (q less than 0.05). After decontamination analysis, 107 remained: 103 more abundant in High Survivors and 4 more abundant in Low Survivors.
  • Klebsiella was more abundant in Low Survivors. It was the only one of the 107 genera with any published articles in connection with mesothelioma (three papers). The authors found none of those papers linked Klebsiella to the intratumor microbiome or to the cancer.
  • Of the 107 genera, 27 returned published papers on cancer in general.
  • Histological subtype was the only clinical characteristic that differed significantly between the groups. Biphasic mesothelioma was more common among Low Survivors, and epithelioid was less common among Low Survivors. Age, stage and gender did not differ significantly.
  • 60 genes were more active in High Survivors and 274 in Low Survivors. Fatty acid metabolic process was the most enriched term in High Survivors (Metascape). Terms related to cell cycle, cell division and DNA repair were enriched in Low Survivors.
  • In the final multiple linear regression model (overall survival in months), the Positive Microbiome Value (p 0.000318305) and histology (p 0.0064759) remained. The model's adjusted R2 was 0.188463827, meaning the variables explained at most 18.8% of the variation in survival.
  • In the Cox model, the Positive Microbiome Value had a hazard ratio of 0.986 (95% CI 0.978 to 0.995), p 0.001. This hazard ratio applies to each one-unit increase in a per-patient score made by summing the log RNA Seq CPM abundance values of the 103 genera more abundant in High Survivors. It is not a comparison of High Survivors with Low Survivors.
  • In the Cox model, non-epithelioid histology (versus epithelioid) had a hazard ratio of 1.781 (95% CI 1.013 to 3.131), p 0.045. Age (hazard ratio 1.011, p 0.467), stage (hazard ratio 0.916, p 0.534) and the Negative Microbiome Value (hazard ratio 1.097, p 0.097) were not statistically significant.
• The authors' conclusion: The authors concluded that they had identified 107 tumor microbiome genera relevant to mesothelioma patient survival, opening a new research area. They wrote that the microbiome was more statistically significant than age, stage and histology, though histology had the larger effect size. They called for laboratory studies to establish mechanisms.
• Limitations the authors state:
  • Only pleural mesothelioma was explored. No peritoneal mesothelioma data were available in cBioPortal.
  • The microbiome signature data came from another study that has since been retracted. Specific findings such as the exact genera may therefore be uncertain.
  • The associations found may be correlational rather than causal.
  • The patients were relatively uniform in age and unevenly distributed by stage. The authors suggest this may explain why age and stage did not predict survival.

Topics

The National Library of Medicine indexes this study under subjects our site covers: Mesothelioma.

Medical terms in this study, explained in our Lexicon: Mesothelioma, Immunotherapy, Median Overall Survival, Epithelioid, Histology, Biphasic Mesothelioma.

Related studies

• Pleural fluid microbiota as a biomarker for malignancy and prognosis — B. Kwok, B.G. Wu, et al. (2023)
• Microbiomes, Epigenomics, Immune Response, and Splicing Signatures Interplay: Potential Use of Combination of Regulatory Pathways as Targets for Malignant Mesothelioma — B.P. Setlai, Z.L. Mkhize-Kwitshana, et al. (2022)
• Pan-cancer analyses of bromodomain containing 9 as a novel therapeutic target reveals its diagnostic, prognostic potential and biological mechanism in human tumours — Y. Chen, Z. Gao, et al. (2024)
• Midkine is a potential novel marker for malignant mesothelioma with different prognostic and diagnostic values from mesothelin — G. Ak, Y. Tada, et al. (2017)
• Causal effect of gut microbiota on the risk of cancer and potential mediation by inflammatory proteins — Y. Wang, W. Liu, et al. (2025)
• Airways microbiota: Hidden Trojan horses in asbestos exposed individuals? — D.E. Magouliotis, V.S. Tasiopoulou, et al. (2014)

Browse more research in Frontiers in Immunology, or return to medical research.