GLIOMATCH: The malignant Glioma immuno-oncology matchmaker: towards data-driven precision medicine using spatially resolved radio-multiomics

成人和儿童恶性胶质瘤（GBM和pHGG）仍然是最难治疗的癌症之一，尽管进行了强化标准治疗，5年生存率仍<5%。患者之间的差异以及每个个体肿瘤的异质性和可塑性导致过去20年中所有治疗临床试验均失败。最近，免疫疗法已经显示出巨大的希望，但仅限于患者的子集。识别这些患者不能先验地完成，因为生物标志物仍然很大程度上缺失，当患者接受治疗时，我们也无法跟踪治疗效果。GLIOMATCH项目旨在改善GBM/pHGG患者的临床结局，方法是通过基于免疫学的患者分层来实现适当免疫疗法的个性化匹配，同时改善对现有/新型疗法的临床应答的随访。这将通过整合空间分辨的多层组织图（使用整合的单细胞多组学）与非侵入性MRI图像来实现。这种整合将推动形成一个新的MRI放射多组学中心，该中心将提供给临床专业人员，通过该中心，他们可以执行基于肿瘤宿主的患者分层和个性化治疗匹配，同时解释纵向随访和治疗效果。提出的数据驱动模型将通过分析最大的免疫肿瘤学（I/O）治疗的GBM/pHGG患者（n>300，包括治疗前后样本）与匹配的对照组（n>300）和异常长期存活的GBM患者（n~140）来开发，其中将研究各种肿瘤宿主小生境对I/O扰动的反应并改善临床结局。这将通过部署一个与UCAN兼容的数据湖来实现，其中增量的重复收集将用于进一步完善机器学习模型，同时提出新的治疗方案。这一行动是癌症使命“理解（肿瘤-宿主相互作用）"项目组的一部分。

Adult and paediatric malignant glioma (GBM and pHGG) remain among the most difficult-to-treat cancers with 5-year survival ratesof <5% despite intensive standard-of-care therapy. The differences among patients and the heterogeneous and plastic nature of eachindividual tumour have resulted in all therapeutic clinical trials failing during the past 20 years. Recently, immunotherapy has beenshowing great promise, but only in subsets of patients. Identifying those patients cannot be done a priori as biomarkers are still largelymissing, nor are we able to follow-up on therapeutic efficacy when patients get treated. The GLIOMATCH project aims at improvingthe clinical outcome of GBM/pHGG patients by enabling immunology-based patient stratification to empower personalised matchingof appropriate immunotherapy, while improving follow-up of clinical responses to existing/novel therapeutics. This will be achievedby integrating spatially resolved, multi-layered tissue maps (using integrated single-cell multiomics), with non-invasive MRI images.This integration will fuel into a novel MRI Radio-multiomics hub, that will be made available to clinical professionals through whichthey can perform tumour-host based patient stratification and personalised therapy matching while interpreting longitudinal follow-upand treatment efficacy. The proposed data-driven models will be developed by analysing the largest cohort of immuno-oncology (I/O)treated GBM/pHGG patients (n>300, including pre-post treatment samples) with matched controls (n>300) and exceptionally long-termsurviving GBM patients (n~140), in which various tumour-host niches will be studied in how they respond to I/O perturbations andlead to improved clinical outcome. This will be empowered by deploying an UNCAN-compatible data lake, to which incremental datacollection will be used to further refine the machine learning models, while proposing novel treatment options. This action is part of theCancer Mission cluster of projects on “Understanding (tumour-host interactions)".