PROJECT 1: TIME-Based Spatiotemporal Cancer Immunograms Predictive for Immunotherapy-Targeted Therapy Sequential Combinations

项目1摘要/摘要 将免疫疗法与其他治疗方案相结合，特别是靶向治疗，是一个非常活跃的领域 探索以提高抗肿瘤疗效为目标，将治疗惠及更多患者或 肿瘤类型。作为首个突变免疫共靶向治疗，抗PD-L1与 BRAFV600MUT和MEK抑制剂(所谓的三联疗法)已被批准用于BRAFV600MUT患者 黑色素瘤。然而，关于这个三联体的数据似乎喜忧参半，其他试验没有达到关键终点， 这表明同时结合并不是最优的。我们最近对同基因小鼠黑色素瘤的研究 模型一致地显示，在不同驱动因素突变和癌症组织学的肿瘤模型中， 1周的抗PD-1/L1(±抗CTLA-4)预处理方案通过以下方式增加三胞胎治疗的疗效 增强MAPKi的耐用性，显著抑制黑色素瘤的脑转移。改进的治疗方法 促进肿瘤相关巨噬细胞的促炎极化和 肿瘤免疫中T细胞克隆性扩增和克隆性收敛的激发 抗PD-1/L1诱导的微环境(时间)。这与临床上观察到的情况一致。 试验数据表明，在MAPKi之前进行免疫治疗与改善无进展生存率有关。这些 结果强调了每个治疗组件的顺序/时机在合理设计 联合疗法，并指出需要对早期阶段的影响进行机械性的理解 每一次联合治疗的时间。 然而，这种序贯联合治疗试验的设计是具有挑战性的，因为数量太多。 要测试的变量(顺序、剂量和时间)。复杂性的程度需要一种预测性的 该框架显著减少了参数空间和通知识别的有效序贯 免疫疗法-靶向抑制剂组合。在这里，我们假设一个时空的，多组学的 分析早期(几天)单一疗法引起的时间变化可以提供深刻的见解 为了大大简化免疫治疗靶向抑制剂序贯联合试验的设计。这个 项目1的目标是提供一个可以挖掘的数据集，以便为有效的顺序组合设计提供信息 养生法。我们将利用最先进的空间多组学组织分析工具来构建时空 已建立的同基因黑色素瘤模型及其相关脑中进化时间的“电影” 转移，在用每种组合治疗成分治疗后。由此产生的时空 多组数据将被分析以提取一些基于代理的高度信息量的时间特征 可以构建预测有效的序贯联合方案的模型(项目2)。回顾 临床肿瘤活检的研究被用来验证模型的发现。

Project 1 Summary/Abstract Combining immunotherapy with other therapy regimens, particularly targeted therapy, is a highly active area of exploration with the goal of improving anti-tumor efficacy and extending therapeutic benefits to more patients or tumor types. As the first mutation-immune co-targeted therapy, the simultaneous combination of anti-PD-L1 with BRAFV600MUT and MEK inhibitors (so-called “triplet” therapy) has been approved for patients with BRAFV600MUT melanoma. However, the data on this triplet appear mixed, with other trials not meeting key endpoints, suggesting that simultaneous combination is not optimal. Our recent work in syngeneic murine melanoma models showed uniformly, across tumor models of distinct driver mutations and cancer histologies, that a regimen of 1-week anti-PD-1/L1 (± anti-CTLA-4) pretreatment augments the efficacy of triplet therapy by enhancing MAPKi durability and dramatically suppressing melanoma brain metastasis. The improved therapy efficacy resulted from the promotion of pro-inflammatory polarization of tumor-associated macrophages and the elicitation of robust T cell clonal expansion and clonotypic convergence within the tumor-immune microenvironment (TIME) induced by the anti-PD-1/L1 lead-in. This is consistent with observations in the clinical trial data that prior immunotherapy before MAPKi is associated with improved progression-free survival. These results highlight the vital role of the sequence/timing of each therapy component in the rational design of combination therapies and also point to the need for a mechanistic understanding of the early-stage impact of each combinatorial therapy component on the TIME. However, the design of such sequential combination therapy trials is challenging because of the sheer number of variables (sequence order, dosing, and timing) to be tested. The level of complexity calls for a predictive framework to significantly reduce the parameter space and inform the identification of effective sequential immunotherapy-targeted inhibitor combinations. Herein, we hypothesize that a spatiotemporal, multi-omics analysis of early-stage (few days) monotherapy-induced changes in the TIME can provide deep insights for greatly simplifying the design of immunotherapy-targeted inhibitor sequential combination trials. The goal of Project 1 is to provide a data set that can be mined to inform the design of effective sequential combination regimens. We will leverage state-of-the-art, spatial multi-omics tissue profiling tools to build a spatiotemporal “movie” of the evolving TIME in established syngeneic melanoma tumor models, and their associated brain metastases, after treatment with each of the combinatorial therapy components. The resultant spatiotemporal multi-omic data will be analyzed to extract a number of highly informative TIME features from which agent-based models (Project 2) for predicting effective sequential combination regimens can be constructed. Retrospective studies of clinical tumor biopsies are proposed to validate the model findings.