Integrative molecular dissection of acquired resistance to PD1/PD-L1 blockade in localized and metastatic urothelial carcinoma

转移性尿路上皮癌（mUC）通常无法治愈，生存率不高 一线顺铂化疗和铂类后PD 1/L1抑制剂治疗提供的获益。最优 对于PD 1/L1抑制剂治疗后进行性mUC，有必要使用新药物进行治疗 （恩福妥单抗vedotin，erdafitinib）提供适度的增量益处，同时伴有毒性。有限数据 存在用于分析PD 1/L1抑制剂全身治疗后获得的肿瘤，这对于确定 反应和抵抗的原因。事实上，PD 1/L1抑制剂后UC肿瘤的多平台分析 还没有报道。这项研究的首要假设是，肿瘤，免疫和基质， 微环境因素决定了对PD-1/L1治疗mUC的耐药性， 并且治疗后肿瘤的单细胞分子解剖将揭示这些特征。我们提出 创新和潜在的变革性多平台纵向获得的散装和/或单细胞（sc） 来自肌肉浸润性膀胱癌（MIBC）的PD 1/L1抑制剂治疗前和治疗后肿瘤组织和血浆的分析 和mUC，以深入了解肿瘤生物学和耐药性以及可用于 开发新的疗法。我们将对2个单独的队列进行详细的多平台分析， 170例患者：1）PD 1/L1抑制剂暴露前后获得的肿瘤组织或未治疗对照 （n=130）和2）在PD 1/L1抑制剂暴露前后或未经处理的血浆无细胞（cf）-DNA 对照组（n=40）。全外显子组测序（WES）、RNA-Seq和正交多参数免疫组化 进行基线经尿道膀胱肿瘤切除术（TURBT）福尔马林固定石蜡 包埋（FFPE）标本和治疗后肿瘤。纳入的患者来自评估 PD 1/L1抑制剂（n=70）作为新辅助治疗，在RC治疗MIBC之前，或接受PD 1/L1抑制剂治疗 mUC（n=50）和对照组未治疗和暂时分离的连续FFPE肿瘤，无干预治疗 （n=10）。对PD 1/L1后mUC肿瘤的亚组（n=20）进行scRNAseq，以分析肿瘤、间质和细胞因子。 免疫细胞。使用获得的质量控制的血浆cfDNA进行WES用于免疫基因组学分析。 PD 1/L1抑制剂治疗前后或对照未治疗患者（n=40）。的功能相关性 通过检测功能性免疫应答来研究特异性基因组改变和由此产生的预测的新抗原。 和基质读出，这将有助于了解基因组改变的免疫原性， 然后为治疗的进步提供基础。我们将开发一种新的集成模型， 学习架构以解释分子肿瘤、间质和免疫系统的静止快照， 配置文件和临床数据的基于时间的轨迹，以计算方式聚合异构数据类型。 总之，这项创新性研究有望发现UC的决定因素 PD 1/L1抑制剂治疗后出现进展，这是一种无法治愈的疾病，存在大量未满足的需求。

Project Summary/Abstract Metastatic urothelial carcinoma (mUC) is generally incurable with modest survival benefit provided by first-line cisplatin-based chemotherapy and post-platinum PD1/L1 inhibitor therapy. Optimal therapy for progressive mUC following PD1/L1 inhibitors is necessary with new agents for this setting (enfortumab vedotin, erdafitinib) providing modest incremental benefits coupled with toxicities. Limited data exist for analysis of tumor acquired after PD1/L1 inhibitor systemic therapy, which is necessary to determine causes of both response and resistance. Indeed, multiplatform analyses of post-PD1/L1 inhibitor UC tumor have not been reported. The overarching hypothesis of this study is that tumor, immune and stromal microenvironmental factors determine resistance to PD-1/L1 therapy mUC, and that integrated bulk and single cell molecular dissection of post-treatment tumors will reveal these features. We propose innovative and potentially transformative multiplatform longitudinally obtained bulk and/or single cell (sc) analysis of pre and post PD1/L1 inhibitor tumor tissue and plasma from muscle invasive bladder cancer (MIBC) and mUC to shed deep insights regarding tumor biology and resistance and information that can be used to develop new therapies. We will perform detailed multiplatform analyses on 2 separate cohorts totaling 170 patients: 1) tumor tissue obtained before and after PD1/L1 inhibitor exposure or untreated controls (n=130) and 2) plasma cell-free (cf)-DNA obtained before and after PD1/L1 inhibitor exposure or untreated controls (n=40). Whole exome sequencing (WES), RNA-Seq and orthogonal multiparametric IHC for immune markers is performed of baseline transurethral resection of bladder tumor (TURBT) formalin fixed paraffin embedded (FFPE) specimens and post-therapy tumors. Patients included are from trials that evaluated a PD1/L1 inhibitor (n=70) as neoadjuvant therapy preceding RC for MIBC, or who received a PD1/L1 inhibitor for mUC (n=50) and control untreated and temporally separated serial FFPE tumors without intervening therapy (n=10). scRNAseq is performed on a subset of post-PD1/L1 mUC tumors (n=20) to analyze tumor, stromal and immune cells. WES is performed for immunogenomic analyses using quality-controlled plasma cfDNA obtained before and after PD1/L1 inhibitor therapy or control untreated patients (n=40). The functional relevance of specific genomic alterations and resultant predicted neoantigens is studied by examining functional immune and stromal readouts, which will shed insights regarding the immunogenicity of genomic alterations that can then provide the foundation for therapeutic advances. We will develop a novel integrated model using deep learning architecture to account for the stationary snapshot of the molecular tumor, stromal and immune profiles, and the time-based trajectory of clinical data to computationally aggregate heterogeneous data types. To summarize, this innovative study is expected to provide discoveries regarding determinants of UC progression following PD1/L1 inhibitors, an incurable disease setting with substantial unmet needs.