Abstract 5073: Whole transcriptome spatial molecular imaging in matched pre and post treatment specimens to identify treatment-associated effects in pancreatic adenocarcinoma

Abstract Pancreatic ductal adenocarcinoma (PDAC) is among the deadliest cancer types with a poor therapeutic response. The molecular heterogeneity of PDAC has been well-studied using dissociated single-cell approaches, but clinically relevant spatial architecture remains poorly understood. Innovations in spatial profiling approaches over the years have facilitated in situ annotation and mapping of cell types in PDAC tumors. Nevertheless, spatial imaging at subcellular resolution has been limited to only a few thousand targets at a time, due to the challenges in attaining acceptable resolution as target panel sizes increase. Moreover, increased data sparsity with higher panel sizes greatly increases the challenges of cellular annotation. Consequently, the analysis of cellular programs has been limited to preselected targets. A whole-transcriptome (WTx; 18, 985 targets) panel for spatial molecular imaging has recently been developed that enables unbiased discovery and spatial annotation of novel gene programs that may be missed with smaller panel sizes. These signatures could parse out nuanced clinically relevant transcriptional features, including subtype and intercellular interactions. We previously applied a 1000-plex (1K) spatial transcriptomics panel to evaluate post-treatment tissue remodeling in 13 unmatched primary PDAC specimens. Here, we analyzed five matched pre-treatment core biopsies and post-treatment resection specimens at subcellular resolution at both 1K panel and whole transcriptome levels on consecutive sections. This analysis provides the first direct window into treatment-induced effects in PDAC. With the WTx panel, we identified a mean of >1000 unique genes per cell, and up to 1, 700 transcripts, with up to 125, 000 cells analyzed per matched pair. To overcome the challenge of cellular annotation in the context of increased sparsity in the WTx data, we first annotated cells using the 1K panel, then used these labels to guide the WTx annotations. Merging our transcriptional annotations with hematoxylin and eosin-staining of FFPE sections annotated by a board-certified pathologist gave insight into the cellular composition of pre-and post-treatment samples and potential intercellular interactions around specific features and cell populations of interest. This study highlights new strategies for analyzing whole-transcriptome spatial transcriptomics at subcellular resolution and the potential thereof to identify novel treatment-associated effects. Citation Format: Beatrice W. Awasthi, Dennis Gong, Nicole Lester, Jung Woo Bae, Xunqin Yin, Ashley Heck, Elizabeth Stirling, Maximilian Walter, Megan Vandenberg, Margaret Hoang, Prajan Divakar, Joe Beechem, David T. Ting, William L. Hwang. Whole transcriptome spatial molecular imaging in matched pre and post treatment specimens to identify treatment-associated effects in pancreatic adenocarcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 5073.