Abstract PR08: Spatial mapping of transcriptomic and lineage diversity in metastatic pancreatic cancer

Jimin Min,Guangsheng Pei,Vittorio Branchi,Kimal I. Rajapakshe,Paul M. Grandgenett,Michael A. Hollingsworth,Paola A. Guerrero,Anirban Maitra,Linghua Wang
DOI: https://doi.org/10.1158/1538-7445.panca2023-pr08
IF: 11.2
2024-01-18
Cancer Research
Abstract:Patients with treatment refractory pancreatic ductal adenocarcinoma (PDAC) often succumb to widespread systemic metastases; however, the current understanding of transcriptomic heterogeneity that underlies recalcitrance to therapy remains understudied. Here we constructed high-resolution maps of clonal architecture and lineage states using spatially resolved transcriptomic profiles from 13 primary PDAC and corresponding liver, lung, and peritoneal metastases (total of 49 sites), using samples obtained from the UNMC Rapid Autopsy Program (RAP). We demonstrate discernible transcriptomic shifts, with the most pronounced distinctions between liver and lung loci, and fluctuations in the expression of basal and classical markers during the transition from primary tumor to organ-specific metastases. Through inferring copy number variations, we revealed clonal heterogeneity within both primary tumors and metastases and the diversity of spatial clonal distributions. Construction of phylogenetic trees facilitated the mapping of patient-specific clonal hierarchies and a variety of evolutionary trajectories. These included five discrete evolutionary modes within primary tumors and selective clonal expansions in divergent spatial regions of metastases, even within the same organ site, particularly notable in the liver. We identified the coexistence of multiple tumor lineages within each tissue and observed site-specific lineage plasticity. Lung metastases presented a marked lineage transition, while liver and peritoneal metastases frequently reflected the lineage composition of the primary tumor. Two metastatic loci within the liver demonstrated spatial divergence in tumor lineages within the same organ site. Comparative transcriptomic analysis with the imputation of super-resolution gene expression revealed that tumor regions in an intermediate state, a bi-phenotypic status bridging basal and classical subtypes, harbored a potent immunogenic gene expression signature and antigen-presenting cancer-associated fibroblasts (CAFs). We noted a unique distribution of myofibroblastic CAFs (myCAFs) and angiogenesis-related tumor-associated macrophages (TAMs) in the basal lineage tumor regions. While both cell populations were abundant within the basal lineage tumors, an increased presence of myCAFs was also observed in the tumor microenvironment (TME) juxta-tumoral to neoplastic cells. In the majority of tissues, irrespective of site and lineage, we observed an exclusion of plasma cells from the tumor bed and adjacent TME, areas where myCAFs with abundant TGFB1 expression are present. Strong CXCL12 expression co-localized with plasma cells, suggesting the potential for interstitial CAFs-induced plasma cell exclusion via CXCL12 in PDAC and its metastases. Collectively, our findings underscore the potential of spatial molecular state mapping to predict regional changes and corresponding TME dynamics. This insight holds promise for enabling personalized stratification of PDAC patients with multi-organ metastasis and for pioneering more effective therapeutic avenues. Citation Format: Jimin Min, Guangsheng Pei, Vittorio Branchi, Kimal I. Rajapakshe, Paul M. Grandgenett, Michael A. Hollingsworth, Paola A. Guerrero, Anirban Maitra, Linghua Wang. Spatial mapping of transcriptomic and lineage diversity in metastatic pancreatic cancer [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Pancreatic Cancer; 2023 Sep 27-30; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(2 Suppl) nr PR08.
oncology
What problem does this paper attempt to address?