Cancer-immune coevolution dictated by antigenic mutation accumulation
Curation statements for this article:-
Curated by eLife
eLife Assessment
This important work presents a stochastic branching process model of tumour-immune coevolution, incorporating stochastic antigenic mutation accumulation and escape within the cancer cell population. They then used this model to investigate how tumour-immune interactions influence tumour outcome and the summary statistics of sequencing data of bulk and single-cell sequencing of a tumour. The evidence is currently incomplete: statistical comparisons between the observed mutational burden distribution and theoretical predictions in the absence of immune selection should be carried out. Conclusions should be tested extensively for robustness/sensitivity to parameters.
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
Abstract
The immune system is one of the first lines of defence against the emergence of cancer. When effector cells attempt to suppress the tumour, the cancer cells can respond in kind by evolving methods of escape or inhibition. Knowledge of this coevolutionary system and the selection taking place within it can help us understand tumour-immune dynamics both during tumorigenesis but also when treatments such as immunotherapies are applied. Here, we present an individual-based branching process model of mutation accumulation, where random mutations arising in cancer cells trigger corresponding specialised immune responses. Different from previous research, we explicitly model interactions between cancer and effector cells, while incorporating stochastic effects, which are especially important for the expansion and extinction of small populations. We find that the parameters governing interactions between the cancer and effector cells induce different outcomes of tumour progress, such as suppression and evasion. While it is hard to measure the cancer-immune dynamics directly in patients, genetic information of the cancer may indicate the presence of such interactions. Our model demonstrates signatures of selection in sequencing-derived summary statistics, such as the single-cell mutational burden. Thus, bulk and single-cell sequencing of a tumour may give information about the coevolutionary dynamics.
Article activity feed
-
eLife Assessment
This important work presents a stochastic branching process model of tumour-immune coevolution, incorporating stochastic antigenic mutation accumulation and escape within the cancer cell population. They then used this model to investigate how tumour-immune interactions influence tumour outcome and the summary statistics of sequencing data of bulk and single-cell sequencing of a tumour. The evidence is currently incomplete: statistical comparisons between the observed mutational burden distribution and theoretical predictions in the absence of immune selection should be carried out. Conclusions should be tested extensively for robustness/sensitivity to parameters.
-
Reviewer #1 (Public review):
Summary:
The topic of tumor-immune co-evolution is an important, understudied topic with, as the authors noted, a general dearth of good models in this space. The authors have made important progress on the topic by introducing a stochastic branching process model of antigenicity/immunogenicity and measuring the proportion of simulated tumors that go extinct. The model is extensively explored, and the authors provide some nice theoretical results in addition to simulated results.
Major comments
The text in lines 183-191 is intuitively and nicely explained. However, I am not sure all of it follows from the figure panels in Figure 2. For example, the authors refer to a mutation that has a large immunogenicity, but it's not shown how many mutations, or the relative size of the mutations in Figure 2. The same …
Reviewer #1 (Public review):
Summary:
The topic of tumor-immune co-evolution is an important, understudied topic with, as the authors noted, a general dearth of good models in this space. The authors have made important progress on the topic by introducing a stochastic branching process model of antigenicity/immunogenicity and measuring the proportion of simulated tumors that go extinct. The model is extensively explored, and the authors provide some nice theoretical results in addition to simulated results.
Major comments
The text in lines 183-191 is intuitively and nicely explained. However, I am not sure all of it follows from the figure panels in Figure 2. For example, the authors refer to a mutation that has a large immunogenicity, but it's not shown how many mutations, or the relative size of the mutations in Figure 2. The same comment holds true for the claim that spikes also arise for mutations with low antigenicity.
-
Reviewer #2 (Public review):
Summary:
In this work, the authors developed a model of tumour-immune dynamics, incorporating stochastic antigenic mutation accumulation and escape within the cancer cell population. They then used this model to investigate how tumour-immune interactions influence tumour outcome and summary statistics of sequencing data.
Strengths:
This novel modeling framework addresses an important and timely topic. The authors consider the useful question of how bulk and single-cell sequencing may provide insights into the tumour-immune interactions and selection processes.
Weaknesses:
One set of conclusions presented in the paper is the presence of cyclic dynamics between effector/cancer cells, antigenicity, and immunogenicity. However, these conclusions are supported in the manuscript by two sample trajectories of …
Reviewer #2 (Public review):
Summary:
In this work, the authors developed a model of tumour-immune dynamics, incorporating stochastic antigenic mutation accumulation and escape within the cancer cell population. They then used this model to investigate how tumour-immune interactions influence tumour outcome and summary statistics of sequencing data.
Strengths:
This novel modeling framework addresses an important and timely topic. The authors consider the useful question of how bulk and single-cell sequencing may provide insights into the tumour-immune interactions and selection processes.
Weaknesses:
One set of conclusions presented in the paper is the presence of cyclic dynamics between effector/cancer cells, antigenicity, and immunogenicity. However, these conclusions are supported in the manuscript by two sample trajectories of stochastic simulations, and these provide mixed support for the conclusions (i.e. the phasing asynchrony described in the text does not seem to apply to Figure 2C). Similarly, the authors also find immune selection effects on the shape of the mutational burden in Figure 5 D/H using a qualitative comparison between the distributions and theoretical predictions in the absence of immune response. However the discrepancy appears quite small in panel D, and there are no quantitative comparisons provided to evaluate the significance. An analysis of the robustness of all the conclusions to parameter variation is missing. Lastly, the role of the Appendix results in the main messages of the paper is unclear.
-
-
-
-