Ning Wang
Arcus Biosciences

Immuno-Oncology Insights 2022; 3(4), –97

Understanding the prognostic & predictive values of TME components

In recent years, extensive works have attempted to understand prognostic values of cell populations within the TME. By definition, a prognostic biomarker provides information on patient health outcome regardless of therapy, whereas a predictive biomarker gives information about the benefit to the patient from the therapy intervention. Compared with previous reports, here I emphasize identifying prognostic and predictive roles of the TME by considering multiple components together instead of considering individual components each time (so called ‘context dependent manner’). More and more evidence is emerging and revealing that multiple TME components together contribute to patient prognosis [5]Roumenina LT, Daugan MV, Petitprez F, Sautès-Fridman C, Fridman WH. Context-dependent roles of complement in cancer. Nat. Rev. Cancer 2019; 19, 698–715.. In addition, given the ever-growing data on identifying prognostic values of the TME [6]Bruni D, Angell HK, Galon J. The immune contexture and Immunoscore in cancer prognosis and therapeutic efficacy. Nat. Rev. Cancer 2020 Nov;20(11):662–80., I also advocate the urgent need to identify prognostic and predictive values of the TME across multiple cancer types, and identify cancer type specific features (so called ‘cancer type specific manner’).

Context dependent manner

In the past few years, enormous effort has been made to identify the prognostic value of TME components. The classical paradigm of anti-tumor response to eliminate tumor cells is mainly considered via cytotoxic T lymphocytes (CTLs), with other immune cell populations either stimulating or suppressing CTL mediated tumor killing. Several immunosuppressive cells have been identified in this process, including myeloid-derived suppressor cells (MDSCs), regulatory T (T_reg) cells, M2 tumor-associated macrophages (TAMs), and cancer associated fibroblasts (CAFs) [4]Ho WJ, Jaffee EM, Zheng L. The tumour microenvironment in pancreatic cancer—clinical challenges and opportunities. Nat. Rev. Clin. Oncol. 2020; 17, 527–40, [6]Bruni D, Angell HK, Galon J. The immune contexture and Immunoscore in cancer prognosis and therapeutic efficacy. Nat. Rev. Cancer 2020 Nov;20(11):662–80..

However, recent work has suggested that multiple TME components interact with each other and together contribute to patient prognosis [5]Roumenina LT, Daugan MV, Petitprez F, Sautès-Fridman C, Fridman WH. Context-dependent roles of complement in cancer. Nat. Rev. Cancer 2019; 19, 698–715.. Individual TME components could be associated with either positive or negative prognosis depending on the other TME components [6]Bruni D, Angell HK, Galon J. The immune contexture and Immunoscore in cancer prognosis and therapeutic efficacy. Nat. Rev. Cancer 2020 Nov;20(11):662–80.. Although several studies are attempting to delineate regulation mechanisms to achieve such context dependent regulation [7–9], it remains largely unclear.

Cytokines and metabolites have been considered as the examples of the mechanisms to modulate the TME and determine tumor growth and progression. For example, the adenosinergic pathway is one of the identified immune suppressive pathways [10]Allard B, Allard D, Buisseret L, Stagg J. The adenosine pathway in immuno-oncology. Nat. Rev. Clin. Oncol. 2020; 17, 611–29. , which could be induced by tumor hypoxia, inflammation, or specific oncogenic pathways. Extracellular adenosine has been shown to modulate broad immune cell populations within the TME, including effector T cells, NK cells, dendritic cells, macrophages, regulatory T (T_reg) cells, B cells, and neutrophils through adenosine receptors A2A and A2b. Activation of the adenosine pathway could also lead to the release of inflammatory cytokines, like tumor necrosis factor (TNF), IL-6, IL-10, CXCL9, and CXCL10. The mechanisms by which cytokines and metabolites modulate the TME and impact antitumor immunity remain to be further explored.

So far, this review has been focused on the prognostic value of the TME components regardless of immunotherapy treatment. It would also be important to identify the predictive value of the TME components, particularly to checkpoint inhibitor immunotherapy, that is, which TME components can specifically predict patient survival with CPI treatment, and which TME components are associated with better CPI response? How does CPI modulate the TME? How does CPI impact cell-cell interactions among the TME components? Substantial work is ongoing to address these questions [11–13]. However, they still remain unknown due to the innate complexity of the TME and limited data (including single cell RNA-seq) from CPI treated patients.

Cancer type specific manner

Identifying the cancer type and patient cohort that could potentially respond to an oncology drug is one of the most important questions in oncology drug development. Thus, profiling and quantifying the TME in a cancer type specific manner is critical. Colorectal, gastric, and lung cancers are among the cancer types with the most well quantified TME [6]Bruni D, Angell HK, Galon J. The immune contexture and Immunoscore in cancer prognosis and therapeutic efficacy. Nat. Rev. Cancer 2020 Nov;20(11):662–80., [14]Fridman WH, Zitvogel L, Sautès–Fridman C, Kroemer G. The immune contexture in cancer prognosis and treatment. Nat. Rev. Clin. Oncol. 2017; 14, 717–34.. However, it is important to take a pan-cancer view and compare TME across multiple cancer types and identify cancer type specific features to guide I-O agent clinical development. Previously, several works summarized the TME across different cancer types, and specified the prognostic values of immune components in the TME [6]Bruni D, Angell HK, Galon J. The immune contexture and Immunoscore in cancer prognosis and therapeutic efficacy. Nat. Rev. Cancer 2020 Nov;20(11):662–80., [14]Fridman WH, Zitvogel L, Sautès–Fridman C, Kroemer G. The immune contexture in cancer prognosis and treatment. Nat. Rev. Clin. Oncol. 2017; 14, 717–34.. However, those summaries are based on examining previously published studies instead of quantifying the TME from unified datasets using a data-driven approach.

Notably, a recent work on profiling T cell states across multiple cancer types reveals that the states of the potentially tumor-reactive T cell (pTRT) populations varied dramatically in the TME of different cancer types [15]Zheng L, Qin S, Si W et al. Pan-cancer single-cell landscape of tumor-infiltrating T cells. Science 2021; 374, abe6474. . Another work from the same research group also performed a pan-cancer analysis of single myeloid cells, and revealed that mast cells with a high ratio of TNF+/VEGFA+ cells are associated with better prognosis in nasopharyngeal cancer [16]Cheng S, Li Z, Gao R et al. A pan-cancer single-cell transcriptional atlas of tumor infiltrating myeloid cells. Cell 2021;184, 792–809.. Particularly, although more and more studies are coming out to integrate multiple single cell RNA-seq on immune cells [17]Nieto P, Elosua-Bayes M, Trincado JL et al. A single-cell tumor immune atlas for precision oncology. Genome research 2021; 31, 1913–26., not much work has been reported on integrating and examining CAFs, which is another important part of the TME across cancer types.

Those works focused on integrating single cell RNA-seq datasets from patients to 1) profile of the TME components across different cancer types, 2) identify the TME components that are significantly associated with patient survival or treatment response, 3) discover new sub cell populations that are associated with survival or response to treatment.

In addition to profiling the TME using single cell RNA-seq, researchers are also actively quantifying the TME from bulk RNA-seq using deconvolution methods [18–20]. Among existing deconvolution methods, EcoTyper is a recently published tool and is specialized in quantifying cell types and states in tumor biopsies. Notably, it provides not only fraction of individual defined cell types and states but also their gene expressions. Based on the deconvoluted cell type gene expression, they have identified 10 shared multi-cell communities (so called ‘ecosystems’) across 16 cancer types. Interestingly, such identified ecosystems shared two similar types with previous reported work [18, 20]: TGFβ dominant and IFNγ dominant in the TME.

Challenges

To develop effective cancer immunotherapies, quantifying the TME is expected to answer two questions: 1) how to identify effective prognostic and predictive biomarkers? 2) how to identify potential I-O agent combinations or new I-O drug targets?

With regard to the first question, although the full landscape of prognostic value of individual TME components still remains to be captured, accumulating evidence illustrates multi-faceted prognostic and predictive values of the TME components in several cancer types and even in sub patient cohorts level. Future work should therefore focus on investigating and integrating datasets across multiple cancer types [21]Wang N, Next steps in unlocking the power of computational biology for I-O target identification and validation. Immuno-Oncology Insights 2021; 2, 241–245.

However, the second question still remains challenging to the field due to multiple factors: 1) Although accumulating clinical trial results have been released on PD-1/PD-L1 inhibitors, not enough clinical data (particularly Phase 3 data) has been reported for other immune checkpoint inhibitors and other immune modulators for combination. This fact limits us in fully understanding the landscape of I-O drugs and combination efficacy at the patient level. 2) We are still on the way to fully understanding the mechanism of response and resistance to PD-1/PD-L1. Recently, several studies identified different T cell populations that may potentially be responsible for PD-1/PD-L1 anti-tumor response [Two subsets of stem-like CD8+ memory T cell progenitors with distinct fate commitments in humans; clonal replacement of tumor-specific T cells following PD-1 blockade]. 3) Current approaches to quantify the TME are not sufficient to gain mechanistic understanding of interactions within the TME components from bulk and single cell RNA-seq datasets.

With more and more patient data with PD-1/PD-L1 inhibitor treatment expected to be released in the near future, together with our quickly accumulated knowledge of resistance mechanisms of immune checkpoints, it is expected that more effective prognostic and predictive biomarkers are going to be developed by illustrating clinical genomic datasets with advanced computational platforms.

Biography

Ning Wang Ning is a bioinformatics scientist at Arcus Biosciences, where he is responsible for preclinical and clinical studies in multiple immune-oncology drug programs. Notably, Ning supported bioinformatics analysis for the first A2AR/A2BR inhibitor (Etrumadenant, now in phase 2 clinical trial). He also pioneered the 3D organoids based drug response prediction technology, which largely facilitated cancer model selection. He is now leading clinical genomic real world data efforts to support immune-oncology agent clinical development. Ning holds a PhD in Bioinformatics from UCLA.

References

1.Sharma P, Hu-Lieskovan S, Wargo JA, Ribas A. Primary, adaptive, and acquired resistance to cancer immunotherapy. Cell 2017;168, 707–23.  Crossref

2.Kalbasi A, Ribas A. Tumour-intrinsic resistance to immune checkpoint blockade. Nat. Rev. Immunol. 2020; 25–39.  Crossref

3.Gohil SH, Iorgulescu JB, Braun DA, Keskin DB, Livak KJ. Applying high-dimensional single-cell technologies to the analysis of cancer immunotherapy. Nat. Rev. Clin. Oncol. 2021; 18, 244–56.  Crossref

4.Ho WJ, Jaffee EM, Zheng L. The tumour microenvironment in pancreatic cancer—clinical challenges and opportunities. Nat. Rev. Clin. Oncol. 2020; 17, 527–40  Crossref

5.Roumenina LT, Daugan MV, Petitprez F, Sautès-Fridman C, Fridman WH. Context-dependent roles of complement in cancer. Nat. Rev. Cancer 2019; 19, 698–715.  Crossref

6.Bruni D, Angell HK, Galon J. The immune contexture and Immunoscore in cancer prognosis and therapeutic efficacy. Nat. Rev. Cancer 2020 Nov;20(11):662–80.  Crossref

7.Dominguez CX, Müller S, Keerthivasan S et al. Single-cell RNA sequencing reveals stromal evolution into LRRC15+ myofibroblasts as a determinant of patient response to cancer immunotherapy. Cancer discovery 2020 Feb 1;10(2):232–53.  Crossref

8.Wang N, Lefaudeux D, Mazumder A, Li JJ, Hoffmann A. Identifying the combinatorial control of signal-dependent transcription factors. PLoS computational biology 2021; 17: e1009095.  Crossref

9.Ge X, Chen YE, Song D et al. Clipper: p-value-free FDR control on high-throughput data from two conditions. Genome biology 2021; 22: 1–29.  Crossref

10.Allard B, Allard D, Buisseret L, Stagg J. The adenosine pathway in immuno-oncology. Nat. Rev. Clin. Oncol. 2020; 17, 611–29.  Crossref

11.Riaz N, Havel JJ, Makarov V et al. Tumor and microenvironment evolution during immunotherapy with nivolumab. Cell 2017;171, 934–49.  Crossref

12.Caushi JX, Zhang J, Ji Z et al. Transcriptional programs of neoantigen-specific TIL in anti-PD-1-treated lung cancers. Nature 2021; 596, 126–32.  Crossref

13.Galletti G, De Simone G, Mazza EM et al. Two subsets of stem-like CD8+ memory T cell progenitors with distinct fate commitments in humans. Nature Immunol. 2020; 21,1552–62.  Crossref

14.Fridman WH, Zitvogel L, Sautès–Fridman C, Kroemer G. The immune contexture in cancer prognosis and treatment. Nat. Rev. Clin. Oncol. 2017; 14, 717–34.  Crossref

15.Zheng L, Qin S, Si W et al. Pan-cancer single-cell landscape of tumor-infiltrating T cells. Science 2021; 374, abe6474.  Crossref

16.Cheng S, Li Z, Gao R et al. A pan-cancer single-cell transcriptional atlas of tumor infiltrating myeloid cells. Cell 2021;184, 792–809.  Crossref

17.Nieto P, Elosua-Bayes M, Trincado JL et al. A single-cell tumor immune atlas for precision oncology. Genome research 2021; 31, 1913–26.  Crossref

18.Luca BA, Steen CB, Matusiak M et al. Atlas of clinically distinct cell states and ecosystems across human solid tumors. Cell 2021; 184, 5482–96.  Crossref

19.Newman AM, Steen CB, Liu CL et al. Determining cell type abundance and expression from bulk tissues with digital cytometry. Nature biotechnol. 2019; 37, 773–82.  Crossref

20.Thorsson V, Gibbs DL, Brown SD et al.The immune landscape of cancer. Immunity 2018; 48, 812–30.  Crossref

21.Wang N, Next steps in unlocking the power of computational biology for I-O target identification and validation. Immuno-Oncology Insights 2021; 2, 241–245  Crossref

Affiliation

Ning Wang
Bioinformatics Scientist
Arcus Biosciences

Authorship & Conflict of Interest

Contributions: All named authors take responsibility for the integrity of the work as a whole, and have given their approval for this version to be published.

Acknowledgements: None.

Disclosure and potential conflicts of interest: The author is a bioinformatics scientist at Arcus Biosciences. But this work is not related with his work at Arcus, nor supported by Arcus Biosciences. These are his personal views and not made on behalf of Arcus Biosciences. The author owns stock or stock options in Arcus Biosciences. The author declares that they have no conflicts of interest.

Funding declaration: The author received no financial support for the research, authorship and/or publication of this article. 

Article & copyright information

Copyright: Published by Immuno-Oncology Insights under Creative Commons License Deed CC BY NC ND 4.0 which allows anyone to copy, distribute, and transmit the article provided it is properly attributed in the manner specified below. No commercial use without permission.

Attribution: Copyright © 2022 Wang N. Published by Immuno-Oncology Insights under Creative Commons License Deed CC BY NC ND 4.0.

Article source: Invited.

Submitted for peer review: Jan 19 2022;Revised manuscript received:Feb 27 2022;Publication date:Mar 15 2022.