Project Summary
Pancreatic ductal adenocarcinoma (PDAC), a traditionally non-immunogenic tumor type, has shown limited
beneficial response to immunotherapy. We were the first to demonstrate that it is possible to convert PDAC to
an immunogenic state following GM-CSF-secreting allogeneic vaccine (GVAX) treatment which promotes
Dendritic Cell (DC) precursor expansion and formation of tertiary lymphoid aggregates (TLAs). However, immune
regulatory mechanisms are preventing any significant clinical benefit. DC paucity gives rise to dysfunctional
immune surveillance and can establish an immunosuppressive TME in PDAC that prevents lymphoid cell
activation and immune invasion. With recently emerged single cell and spatial omics, we now have the ability to
study cancer immunology at unprecedented scale and resolution. We propose to generate single cell and spatial
transcriptomic and proteomic data to study systemic responses and local immunological activities in vaccine
primed PDAC. Specifically, we hypothesize that DC state transitions and interactions with other TME cell types
can delineate immunologic responses to immunotherapies in vaccine primed PDAC. To address this hypothesis,
we propose two specific aims. We will first determine the distinct immunologic effects of vaccine and immune
checkpoint inhibition combination regimens on peripheral DC state transitions in PDAC patients (Aim 1). To
accomplish this we will develop a novel single cell proteomic trajectory analysis pipeline that computes cell
phenotypes using continuous variables. This will allow us to study phenotypic transitions using unsupervised
approaches that are less discernible in discrete cell type analyses (technological sub-aim). We will then apply
our pipeline on DCs by implementing mass cytometry to capture DC state transitions in peripheral blood
mononuclear cells (PBMCs) from vaccine clinical trials by assessing baseline and on-treatment samples
(biological sub-aim). To delineate spatial factors influencing the immune dynamics of TLA formation after vaccine
priming, we will evaluate the role of DCs in the formation and regulation of lymphoid aggregates in PDAC (Aim
2). To achieve this, we will spatially resolve TLAs in vaccine (GVAX) primed human PDAC tumors at the RNA
and protein levels using Visium spatial transcriptomics and imaging mass cytometry (IMC). We will employ matrix
factorization methods to learn gene and protein expression patterns in both of the spatial data modalities to
discern gene expression patterns unique to TLAs and evaluate their expression of DC markers. By
understanding DC state transitions directly within the TME, the findings from this Aim will synergize with Aim 1.
Completion of these aims will deliver potential new immunotherapy strategies in PDAC patients, as well as
develop novel open-source software for mass cytometry analysis. The skills I obtain from this work will prepare
me to pursue a career as a cross-trained cancer immunologist and computational biologist, delineating immune
responses to empower precision immunotherapy.