Spatial modelling and quantification of T cell exhaustion in the tumour microenvironment of oesophageal cancer

背景。最近，免疫检查点抑制剂对食管腺癌（OAC）的免疫治疗已被证明可以改善少数接受治疗的患者的治疗反应的深度和持久性。免疫系统成功控制和消除癌症需要将活化的肿瘤抗原特异性细胞毒性T淋巴细胞运输和浸润到肿瘤中，然后识别和杀死癌细胞，正如癌症免疫周期所描述的那样。这个过程可以在多个点上中断，导致淋巴细胞驱动的癌症有效破坏的破坏和预防。其中突出的是抑制性微环境的影响，它可以抑制肿瘤特异性的、活化的、细胞毒性的CD8+ T淋巴细胞的扩张、迁移到肿瘤并杀死癌细胞；以及淋巴细胞衰竭的影响。这些免疫“检查点”阻止了抗肿瘤反应的发展，并挫败了抗肿瘤反应的免疫治疗激活。因此，确定对靶向免疫治疗的原发性耐药机制对进展至关重要。多种免疫抑制网络已被描述，包括与调节性T细胞（T-regs）和癌症辅助成纤维细胞（CAFs）的相互作用。在OAC中，T细胞浸润的总体计数与临床结果相关性较差，而CAFs与细胞毒性T细胞抑制密切相关。总之，这些观察结果表明，不同类型T细胞和CAFs之间的空间关系对于理解OAC中的免疫抑制非常重要。该项目的目的是分析和量化涉及OAC患者免疫细胞的空间域，以便根据空间统计制定新的指标，从而改进诊断、预后和免疫治疗策略。研究方法新颖性。我们将开发一个空间分辨的、多尺度的计算模型，描述OAC的体内生长及其与CAFs和不同T细胞亚型的相互作用。亚细胞变量将代表每个T细胞的耗竭水平作为一个连续值，这将决定其杀死OAC细胞的功效。反过来，T细胞衰竭将通过与caf、T regs和免疫疗法的相互作用而改变。模拟输出将使用一套空间统计分析工具进行定量描述。这些工具也将应用于由Elliott实验室生成的OAC的mIHC图像，现有的面板针对CD8+亚群、T-regs和CAFs进行了优化。因此，生物医学图像和ABM模拟之间的直接定量比较将成为可能。我们将识别空间统计的组合，以区分通过不同参数生成的模拟，并将这些应用于免疫组化样本，以预测患者的结果和对治疗的反应。提出的结果。该项目将提供一个多功能的多尺度计算模型，模拟免疫细胞亚群和OAC之间的相互作用，并生成用于与mIHC图像比较的合成空间数据。该模型将为OAC内驱动T细胞衰竭和免疫抑制的过程提供新的机制理解，并作为识别潜在的新免疫疗法的工具。通过将模型结果与mIHC数据进行比较，我们将确定细胞共定位的统计描述，它作为成像生物标志物，可以区分哪些患者将受益于免疫治疗，哪些患者不会受益于免疫治疗。纳菲尔德医学院Tim Elliott教授的实验室将合作提供OAC的多重免疫组织学（Vectra）图像。该项目属于EPSRC数学生物学研究领域。

Background. Recently, immunotherapy of oesophageal adenocarcenoma (OAC) with immune checkpoint inhibitors has been shown to improve depth and durability of therapeutic responses for a significant minority of treated patients. Successful control and elimination of a cancer by the immune system requires the trafficking and infiltration of activated tumour antigen specific cytotoxic T lymphocytes into tumours, followed by recognition and killing of cancer cells, as described in the cancer-immunity cycle. This process can be disrupted at multiple points, leading to the breakdown and prevention of effective lymphocyte-driven destruction of cancer. Prominent among these is the influence of a suppressive microenvironment that can inhibit tumour-specific, activated, cytotoxic CD8+ T lymphocytes from expanding, migrating to the tumour, and killing cancer cells; and the impact of lymphocyte exhaustion. These immunological "checkpoints" prevent an anti-tumour response from developing and frustrate immunotherapeutic activation of the anti-tumour response. Thus, identifying the mechanisms underlying primary resistance to targeted immunotherapies is vital for progress. Multiple immunosuppressive networks have been described, including interactions with regulatory T cells (T-regs) and cancer assisted fibroblasts (CAFs). In OAC, bulk counts of T cell infiltration correlate poorly with clinical outcomes, and CAFs are strongly implicated in cytotoxic T cell suppression. Together, these observations suggest that spatial relations between different types of T cells and CAFs are important for understanding immunosuppression in OAC. The aim of this project is to analyse and quantify spatial domains involving immune cells from patients with OAC in order to develop new metrics, based on spatial statistics, that will improve diagnostics, prognostics, and immunotherapeutic strategies. Novelty of research methodology. We will develop a spatially-resolved, multiscale computational model describing the in vivo growth of OAC and its interactions with CAFs and different T cell subtypes, adapting an existing model. Subcellular variables will represent each T cell's level of exhaustion as a continuous value, which will determine its efficacy for killing OAC cells. In turn, T cell exhaustion will be altered through interactions with CAFs and T regs and immunotherapy. Simulation outputs will be described quantitatively using a suite of spatial statistical analysis tools. These tools will also be applied to mIHC images of OAC, generated by the Elliott lab, with existing panel optimised for CD8+ subsets, T-regs and CAFs. direct quantitative comparison will therefore be possible between biomedical images and ABM simulations. We will identify combinations of spatial statistics which distinguish between simulations generated via different parameters, and apply these to IHC samples to predict patient outcomes and responses to therapy. Proposed outcomes. The project will deliver a versatile multiscale computational model that simulates interactions between immune cell subsets and OAC, and that generates synthetic spatial data for comparison with mIHC images. The model will provide new mechanistic understanding of processes driving T cell exhaustion and immunosuppression within OAC and also serve as tool for identifying potential new immunotherapies. By comparing model outcomes with mIHC data, we will identify statistical descriptions of cell colocalization which act as imaging biomarkers and which can distinguish patients who would benefit from immunotherapeutic treatments from those who would not. Professor Tim Elliott's lab at the Nuffield Department of Medicine will be collaborating to provide Multiplex Immunohistological (Vectra) images of OAC. This project falls within the EPSRC Mathematical Biology research area.