PROJECT SUMMARY This application aims to address the urgent need for precision in intercepting hereditary diffuse gastric cancer (HDGC) associated with pathogenic germline CDH1 mutations, elevating lifetime cancer risk up to 80% for developing advanced diffuse gastric cancer (AGC), particularly at a young age. HDGC patients pose unique challenges in cancer interception and risk reduction due to the aggressive nature of AGC, which is associated with a poor prognosis. Consequently, those with pathogenic CDH1 mutation are advised to undergo prophylactic total gastrectomies (PTG), However, the incidence of AGC in CDH1 carries varies significantly, with only 25-45% of mutation carries develop AGC while many patients remain negative during surveillance (normal-appearing mucosa, N) or stable at the T1a stage (intramucosal stage which is considered as the “precancerous” lesion of AGC in the context of HDGC) without progression to AGC in years or even lifetime. These facts indicate that PTG, a major operation associated with chronic physical and psychosocial disabilities, may not be necessary for all CDH1 mutation carries, however, previous attempts to stratify risk based on specific variants or family cancer history have not been successful and tools to identify patients at high risk of AGC development, potentially requiring PTG, are currently lacking. This is largely due to our limited understanding of the earliest events that drive T1a lesion development and progression. To fill this void, we assembled the Center for Gastric Pre-Cancer Atlas of Multidimensional Evolution in 3D (GAME3D) that is composed of scientists who work closely together bridging multiple disciplines, have longstanding expertise in pathobiology of HDGC, and who developed innovative workflows for multimodal characterization and modeling of the N-T1a-AGC pathologic continuum in CDH1 germline mutation carriers. The overall objective of our GAME3D Research Center is to develop a multimodal, single-cell, dynamic, and 3D spatial atlas of normal-appearing mucosa, T1a lesion, and AGC through longitudinal samples from CDH1 mutation carriers that will inform of transition of at-risk mucosa to T1a lesions and then to AGC. Aim 1 will build a multimodal, dynamic, 2D spatial atlas that maps the transition of T1a lesion from at-risk mucosa and its subsequent evolution to AGC. Aim 2 will employ revolutionary technical and computational approaches and build a spatiotemporal three-dimensional (3D) atlas at single-cell resolution that captures the N-T1a-AGC pathologic continuum. Aim 3 will leverage complete use cases and 2D and 3D atlases to build computational models of tumor initiation in normal-appearing gastric mucosa, pathogenesis and progression of T1a lesions, and, ultimately, the N-T1a-AGC pathologic continuum. Our atlas is of high-value to the current HTAN repertoire as it will provide an exceptional view into the single-cell spatial landscape of the N- T1a-AGC pathologic continuum, thereby generating an invaluable resource for HTAN and the field to understand the earliest phases in human AGC development in high-risk CDH1 mutation carriers and thus identify targets that are ideal to intercept this trajectory.

NARRATIVE Pathogenic germline mutations in the CDH1 gene significantly elevate the risk of hereditary diffuse gastric cancer, prompting many carriers to undergo prophylactic total gastrectomy at a young age due to the disease’s poor survival rates. However, the lack of fundamental understanding of the molecular and cellular mechanisms underlying premalignant disease initiation and evolution poses a barrier to precision approaches for early interception. To address this gap, we propose to construct a multimodal, spatial, and 3-dimensional (3D) atlas utilizing advanced techniques in 3D reconstruction, visualization, and artificial intelligence-driven predictive modeling.