Search

Recent advances in immunotherapy demonstrate the need to further understand the characteristics of an individual cancer patient’s immune system and how it influences responses to cancer treatment. Here, we developed an immunoprofiling platform to evaluate the features in the blood of cancer patients to test the hypothesis that peripheral immune cell heterogeneity could be used to stratify these patients into different categories or immunotypes to monitor disease progression and treatment response. To that end, we established a unique diagnostic immunoprofiling assay and analytical framework based on the analysis of leukocytes in the peripheral blood using multiparameter flow cytometry. Supervised manual gating of flow cytometry data from a cohort of 50 healthy donors identified 415 cell types and immune activation states that were used to train and later independently validate machine learning models to automatically identify immune cell subsets from raw cytometry data. By applying this tool to peripheral blood samples from a mixed cohort of 299 healthy donors and 323 cancer patients, we developed a machine-learning classification model that can differentiate between these two groups with 93% accuracy. This model was further refined using spectral clustering with bootstrapping, revealing 5 clusters, or immunotypes, characterized by specific physiological immune profiles: (1) Myeloid-derived suppressor/NK cell, (2) Terminally-differentiated CD8+ T cells, (3) Mixed CD4+ T helper cells, (4) CD4+ Th1 & CD8+ T cell memory, and (5) Naive T and B lymphocytes. Interestingly, very few healthy donors could be found in clusters 1 and 2 but were assigned most frequently to cluster 5. Matched RNA-seq was used to further validate these profiles using the cellular deconvolution algorithm, Kassandra, and differential gene expression analysis revealed immunotype-specific signatures that are consistent with immune response potential. Patients in the terminally-differentiated CD8+ T cell cluster had a narrower range of HLA-types than the other clusters, and TCR repertoire analysis indicated significantly increased clonality and reduced clonotype diversity. Within this cluster there was a high degree of overlap between TCR sequences in the peripheral blood and the tumor, indicating a relationship between peripheral blood immunotype and tumor infiltration. Altogether, the establishment of these immunotypes using peripheral blood immunoprofiling represents a promising signature that can be used to identify and stratify cancer patients that will benefit from immune-based therapies. Citation Format: Daniiar Dyikanov, Iris Wang, Tatiana Vasileva, Polina Shpudeiko, Polina Turova, Arseniy A. Sokolov, Olga Golubeva, Evgenii Tikhonov, Anna Kamysheva, Ilya Krauz, Mary Abdou, Madison Chasse, Tori Conroy, Nicholas R. Merriam, Boris Shpak, Anastasia Radko, Anastasiia Kilina, Lira Nigmatullina, Linda Balabanian, Christopher J. Davitt, Alexander A. Ryabykh, Olga Kudryashova, Cagdas Tazearslan, Ravshan Ataullakhanov, Alexander Bagaev, Aleksandr Zaitsev, Nathan Fowler, Michael F. Goldberg. Comprehensive immunoprofiling of peripheral blood reveals five conserved immunotypes with implications for immunotherapy in cancer patients [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 6664.

Here, we show that Porphyromonas gingivalis ( Pg ), an endogenous oral pathogen, dampens all aspects of interferon (IFN) signaling in a manner that is strikingly similar to IFN suppression employed by multiple viral pathogens. Pg suppressed IFN production by down-regulating several IFN regulatory factors (IRFs 1, 3, 7, and 9), proteolytically degrading STAT1 and suppressing the nuclear translocation of the ISGF3 complex, resulting in profound and systemic repression of multiple interferon-stimulated genes. Pg -induced IFN paralysis was not limited to murine models but was also observed in the oral tissues of human periodontal disease patients, where overabundance of Pg correlated with suppressed IFN generation. Mechanistically, multiple virulence factors and secreted proteases produced by Pg transcriptionally suppressed IFN promoters and also cleaved IFN receptors, making cells refractory to exogenous IFN and inducing a state of broad IFN paralysis. Thus, our data show a bacterial pathogen with equivalence to viruses in the down-regulation of host IFN signaling.

Cellular deconvolution algorithms virtually reconstruct tissue composition by analyzing the gene expression of complex tissues. We present the decision tree machine learning algorithm, Kassandra, trained on a broad collection of9,400 tissue and blood sorted cell RNA profiles incorporated into millions of artificial transcriptomes to accurately reconstruct the tumor microenvironment (TME). Bioinformatics correction for technical and biological variability, aberrant cancer cell expression inclusion, and accurate quantification and normalization of transcript expression increased Kassandra stability and robustness. Performance was validated on 4,000 HE slides and 1,000 tissues by comparison with cytometric, immunohistochemical, or single-cell RNA-seq measurements. Kassandra accurately deconvolved TME elements, showing the role of these populations in tumor pathogenesis and other biological processes. Digital TME reconstruction revealed that the presence of PD-1-positive CD8

e15050 Background: Complex tumor tissue is composed of malignant cells and diverse tumor microenvironment (TME) cellular populations. Depending on the cancer type, the percentage of malignant cells present in tumor tissue varies, with percentages sometimes below 10%. TME cellular transcripts may comprise the majority of the total transcripts in a tumor, potentially resulting in biases during biomarker development and for clinical decision-making. However, TME gene expression can be subtracted from total tumor gene expression, resulting in only malignant cell expression. Methods: A computational tool was developed for the “subtraction” of TME-specific gene expression from total gene expression in an array of solid tumors, producing in silico “purified'' malignant cell gene expression. To extract the malignant cell RNA expression values, a machine learning model was trained on an artificial transcriptome dataset created from different solid tumor cancer cell lines with the addition of various TME cellular proportions. The artificial transcriptomes, both training and test, were composed of 7,114 samples of purified TME cell types and 3,143 cancer cell lines. The final model relied on the following three major parameters: 1) RNA percentages of deconvolved TME cell types; 2) the weighted sum of the average expression of a malignant cell target gene produced by different TME cell populations where the RNA percentages (weights in the weighted sum) of the TME cell populations were predicted by deconvolution; and 3) a set of genes expressed predominantly in the TME. To experimentally validate the computational model, different proportions of COLO829 (cutaneous melanoma), MCF-7 (invasive ductal carcinoma), and K562 (chronic myeloid leukemia) cell lines were mixed in vitro with PBMCs, creating controlled representations of tumor tissue, and RNA was extracted and sequenced. Gene expression was quantified and analyzed with the computational model, with comparisons to pure cancer cell line expression. Results: The expression of at least 120 clinically relevant biomarkers was reconstructed by applying the model to artificial transcriptomes in which the percentage of malignant cells varied from 10% to 90%. The concordance correlation coefficient between pure cancer cell lines and the extracted malignant cell expression increased on average from 0.75 to 0.9 compared to unprocessed data (e.g., PTEN from 0.2 to 0.88, RB1 from 0.57 to 0.89, ERBB2 from 0.83 to 0.99). In vitro validation showed that the tool improved the concordance correlation coefficient and mean absolute error (MAE) for many tumor biomarkers. For example, the PTEN correlation coefficient increased by 0.33 and its MAE was reduced 3-fold. Conclusions: This novel computational tool can aid in treatment decision-making based on malignant cell expression, promoting the use of gene expression for personalized therapeutics.