Creemers2021 - A tipping point in cancer-immune dynamics leads to divergent immunotherapy responses and hampers biomarker discovery
August 2021, Model of the Month by Emilia Chen
Original model - BIOMD0000001022
Introduction
Until recent years, there were only three main methods of treating cancer: surgery, radiotherapy and chemotherapy. Currently, immunotherapy is emerging as a promising means for treating cancer. In order to harness its potential, the understanding of tumour-immune dynamics and the discovery of new predictive biomarkers is necessary. Creemers et al. present a mechanistic ordinary differential equation (ODE)-based model representing critical tumour-immune interactions and dynamics. With this model, the authors analysed patient responses to immune checkpoint inhibition (ICI) therapy and the predictive power of biomarkers by considering the critical threshold for survival – a tipping point revealed by the model.
Model
Creemers et al. constructed a two-compartment ODE model to investigate the core mechanisms of tumour-immune dynamics. The mathematical model captured essential interactions between cancer and immune cells during tumorigenesis by considering processes within the tumour microenvironment and lymph nodes (Figure 1). In the tumour microenvironment, the growth and death of tumour cells (Equation 1) and intratumoural T cells (Equation 2) were modelled. In the lymph node, primed antigen-specific and naïve T cells (Equations 3 and 4, respectively) were described.
Figure 1. A schematic diagram of the ODE model describing fundamental processes in the tumour microenvironment and lymph nodes.
Results and Discussion
The model simulated three end scenarios as shown in Figure 2. These outcomes lead to two clinical observations: (1) the tumour is eradicated (Figure 2a) or controlled (Figure 2b) by the immune system (immune control), or (2) the tumour becomes clinically detectable (Figure 2c, immune evasion). By varying the tumour growth rate (ρ) and T cell killing rate (ξ), the authors identified a critical threshold (tipping point) where a sudden state transition between immune control and evasion occurred (Figure 3). In these specific simulations, tipping points were observed at ρ ≈ 3 (Figure 3a) and ξ ≈ 0.032 (Figure 3b). Clinically, the presence of these tipping points indicate that small modifications to tumour growth rate or T cell killing rate near a threshold can result in substantial survival differences in patients.
Figure 2. Disease courses of cancer patients simulated by the model yields three possible scenarios (a) Eradication of tumour cells by an effective anti-tumour immune response can occur before clinical presentation. (b) The immune system can suppress and control tumour growth in a subclinical state. (c) Tumour growth can outpace control by the immune system, giving rise to a tumour that can be clinically diagnosed and ultimately lead to a cancer-related death.
Figure 3. A tipping point in tumour-immune dynamics is revealed by simulations where tumour growth rate or T cell killing rate were varied. (a) Increasing tumour growth rate revealed a tipping point, where immune control of tumour progression is replaced with immune evasion. (b) Similarly, increasing T cell killing rate shows a tipping point from immune evasion to immune control.
The implications of this on patient response to ICI and survival dynamics revealed the extent to which a survival benefit will be induced depends on the distance to a tipping point. The presence of a tipping point can mechanistically explain the dichotomous clinical outcomes following ICI treatment. Specifically, ICI treatment can induce a survival benefit if it is potent enough to shift a patient over a tipping point. If the treatment effect is sustained long enough for a patient to benefit, ICI can result in long-term survival and immune control. Thus, tipping points in tumour-immune dynamics are crucial in shaping survival kinetics, therefore can influence treatment response and the prognosis of patients.
The authors advocate for biomarker discovery studies to integrate both tumour and immune markers into a biomarker panel in order to better take into account tipping points. However, the dynamic behaviour of cancer and immune interactions results in varying patient disease trajectories. To improve the predictive power of biomarkers in highly dynamic disease courses, continuous monitoring and modelling is beneficial. This could aid the implementation of adaptive treatment strategies, which may lead to better outcomes than purely basing patient stratification at baseline.
Conclusion
Creemers et al. used a mathematical model of tumour-immune dynamics to reveal a tipping point between immune control and evasion and investigated its consequence on the outcome of cancer immunotherapies. In silico experiments revealed that tipping points can mechanistically explain the observation of ICI-induced dichotomous clinical outcomes. The authors also used simulations combined with analysis of clinical data, to assess the implications of tipping points in guiding the future of biomarker research.
References
1. National Cancer Institute. 2021., 'Immunotherapy for Cancer.', Available at: https://www.cancer.gov/about-cancer/treatment/types/immunotherapy.
2. Hu-Lieskovan S, Bhaumik S, Dhodapkar K, et al., 'SITC cancer immunotherapy resource document: a compass in the land of biomarker discovery.', Journal for ImmunoTherapy of Cancer. 2020 Dec.;8:e000705. doi: 10.1136/jitc-2020-000705.
3. Darvin, P., Toor, S.M., Sasidharan Nair, V. et al., 'Immune checkpoint inhibitors: recent progress and potential biomarkers.' Exp Mol Med. 2018 Dec.;50:1–11. doi: 10.1038/s12276-018-0191-1
4. JHA. Creemers et al., ‘A tipping point in cancer-immune dynamics leads to divergent immunotherapy responses and hampers biomarker discovery.’, Journal for ImmunoTherapy of Cancer. 2021 May;9:e002032. doi: 10.1136/jitc-2020-002032.