Introduction: A single arm trial (NCT007773097) and a double-blind, placebo controlled randomized trial (NCT02134925) were conducted in patients with newly diagnosed advanced colonic adenomas to test the safety and immunogenicity of the MUC1 antigen vaccine and its potential to prevent new adenoma formation. These are the first trials of a non-viral cancer vaccine administered in the absence of cancer. In both trials, the vaccine was safe and strongly immunogenic in 43% and 25% of participants (Responders), respectively. The lack of robust response in a significant number of participants suggested, for the first time, that even in a premalignant setting, the immune system may have already been exposed to regulatory influences that, in the case of the vaccine, determine who does and who does not respond. We hypothesized that there could be molecular and cellular differences in the immune competence between vaccine responders and non-responders, and that they could be identified by studying their pre-vaccination peripheral blood mononuclear cells (PBMCs). Methods: The two MUC1 vaccine trials are described in https://doi.org/10.1101/2022.10.05.22280474 and https://doi.org/10.1158%2F1940-6207.CAPR-12-0275. We performed single cell RNA-sequencing (scRNAseq) on banked pre-vaccination PBMCs from 16 Responders and 16 Non-Responders, determined by anti-MUC1 IgG response. Using differential gene expression (DGE), pathway enrichment, and network estimation analyses, we identified specific cell types, genes, and pathways that differ between responders and non-responders. Results: Pre-vaccination PBMCs from Responders contained a significantly higher percentage of CD4+ naive T cells, while Non-Responders showed significantly higher percentage of CD8+ T effector memory (TEM) cells and a higher percentage of CD16+ monocytes. DGE and gene interaction network analysis showed a higher level of expression of T cell activation genes, such as Fos and Jun, in the CD4+ naive T cells in Responders. Further network analysis showed that these genes were directly connected to response. We also found pre-vaccination specific gene ontology (GO) pathways for translational and transcriptional activity enriched in all cell types in Responders compared to Non-Responders. Conclusion: Our analyses identified candidate biomarkers that are predictive of a preventative cancer vaccine response. Thus, our results can be used for patient selection for vaccine administration. Furthermore, we identified cell type differences and transcriptional pathways that provide information of possible mechanisms of vaccine response. Citation Format: Daniel Y. Yuan, Michelle L. McKeague, Matthew T. Dracz, Olivera J. Finn, Panayiotis V. Benos. Single cell transcriptomics uncovers cellular and molecular differences in PBMCs of responders and non-responders to the MUC1 cancer vaccine given in the preventative setting [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 6537.

As the cost of high-throughput genomic sequencing technology declines, its application in clinical research becomes increasingly popular. The collected datasets often contain tens or hundreds of thousands of biological features that need to be mined to extract meaningful information. One area of particular interest is discovering underlying causal mechanisms of disease outcomes. Over the past few decades, causal discovery algorithms have been developed and expanded to infer such relationships. However, these algorithms suffer from the curse of dimensionality and multicollinearity. A recently introduced, non-orthogonal, general empirical Bayes approach to matrix factorization has been demonstrated to successfully infer latent factors with interpretable structures from observed variables. We hypothesize that applying this strategy to causal discovery algorithms can solve both the high dimensionality and collinearity problems, inherent to most biomedical datasets. We evaluate this strategy on simulated data and apply it to two real-world datasets. In a breast cancer dataset, we identified important survival-associated latent factors and biologically meaningful enriched pathways within factors related to important clinical features. In a SARS-CoV-2 dataset, we were able to predict whether a patient (1) had COVID-19 and (2) would enter the ICU. Furthermore, we were able to associate factors with known COVID-19 related biological pathways.