There is currently substantial interest in liquid biopsy approaches to profile tumors. Profiling of the exosome surfaceome and cargo has yielded immuno-oncology biomarkers that inform about tumor status and potential response to immunotherapy.

Exosomes, a subset of small extracellular vesicles (EVs), are emerging as key components of liquid tumor biopsy in cancer and as biomarkers of immune competence. Plasma-derived exosomes can be separated by immune capture into tumor cell-derived and T cell-derived fractions. Each fraction carries information about the tumor or immune competence of the cancer patient, respectively. Simultaneous real-time molecular profiling of the tumor and of the host immune status by plasma-derived exosomes suggest that sEVs are highly promising biomarkers in immuno-oncology.

The emergence of immunotherapy in oncology requires the discovery, validation and subsequent adoption of robust, sensitive and specific predictive and prognostic biomarkers for daily practice. The use of a liquid biopsy could provide an important complementary or alternative added value to PD-L1 detection in tissue biopsy. In my talk, I will discuss how liquid biopsy could be used in the field of immuno-oncology to predict response or relapse for patients undergoing immune-checkpoint inhibitor therapy.

Using data generated from immune profiling of peripheral blood samples collected from ICB-treated patients, we applied a novel supervised statistical learning approach to classify phenotypes and determine their association with survival and treatment response. We will describe how baseline characteristics of T cells in the peripheral blood associate with clinical outcomes.

B- and T- cell acute lymphoblastic leukemia (B/T-ALL) may be refractory or recur after therapy by suppressing host anti-cancer immune surveillance mediated specifically by natural killer (NK) cells. In order to elucidate the role of NK cells in sustaining ALL regression, we delineated the phenotypic and functional defects in NK cells of high-risk B/T-ALL patients using mass, flow, and in silico cytometry. We found that, compared to normal counterparts, NK cells in B/T-ALL patients exhibit an activated signature characterized by high CD56 and high CD69 expression, production of activated NK-origin cytokines, and calcium signaling. However, we demonstrated that defective maturation of NK cells into cytotoxic effectors prevents NK cells of ALL patients from lysing NK-sensitive targets as efficiently as normal NK cells, despite their activated phenotype. Additionally, we showed that NK cells in ALL are exhausted, which is likely caused by their chronic activation. We found that increased frequencies of activated cytokine-producing NK cells are associated with increased disease severity and independently predict poor clinical outcome in ALL patients. The comprehensive analysis of the NK cell phenotype and functions in ALL patients highlighted the relevance of developing NK cells as diagnostic and prognostic tool to predict clinical outcome and underscore the clinical potential of allogeneic NK infusions to prevent ALL recurrence.

Utility of multiplex immunohistochemistry/immunofluorescence (mIHC/IF) highlights the spatial distribution of infiltrating immune cells within the tumor immune microenvironment that allow a detailed characterization of specific cell phenotypes defined by co- or lack of expression of multiple markers that may help in predicting clinical responses and mechanisms of resistance to therapy. In this presentation we will show a unique workflow supporting whole slide imaging of a high-plex mIF and traditional same slide H&E for comprehensive tissue immunophenotyping analysis.

TCR/BCR repertoire profiling holds great potential for understanding disease mechanisms. We introduce a novel technology for profiling of all human TCR and BCR variable regions and phenotypic characterization of immune cells in bulk and at the single-cell level in PBMCs and immune cell fractions. Preliminary data shows that TCR/BCR clonotype analysis combined with targeted expression profiling of immune cells can be applied for large-scale discovery in several immune-responsive model systems. 

Pan-cancer molecular biomarkers of immunotherapy response can identify patients likely to derive benefit from PD-1/PD-L1-directed monotherapy. Clinical genomic data has also illuminated mechanisms of resistance to immunotherapy which can be leveraged to drive rational combinations and novel drug discovery. Data will be presented showing key determinants of response to pembrolizumab across multiple tumor types, as well as emerging resistance signatures, which provide a framework for novel approaches to cancer therapy.

Several biomarkers associated with response and resistance to immune checkpoint blockade in urothelial cancer have been described. However, how these different cancer cell, stromal, and immune cell features are interrelated, and which are independently associated with outcomes, are not well defined. Approaches to identify independent features associated with immune checkpoint blockade outcomes for prioritization as biomarkers and therapeutic targets will be described.

Tumor biopsies and advanced multiplex immunofluorescence enable quantification of immune cell subsets including their precise placement within the tumor microenvironment—yet getting samples can put patients at risk. Liquid biopsies are less invasive and enable monitoring of circulating biomarkers. In this presentation, we will show an innovative strategy utilizing liquid biopsies to assess biomarkers in both circulating tumor cells (CTCs) and peripheral immune cells from the same patient sample.

AI and Machine Learning (AI/ML) have enabled the development of diagnostic tests using unbiased -omics level data. In this presentation, we focus on the use of unbiased liquid biopsy-based proteomics for the development of diagnostic tests that predict response of lung cancer patients to immune checkpoint inhibitors (ICIs). Finally, we will introduce methods in translational science to examine the biologic underpinnings that drive test outcomes.

JTX-8064 is a LILRB2/ILT4 antagonist mAb that may relieve myeloid cell immuno-suppression by shifting macrophages to an M1-like state and activating T cells. Using histoculture, pharmacodynamic (PD) responses to JTX-8064 can be measured preclinically in a human ex vivo tumor system. Ex vivo evaluation of human tumors identified hypotheses for both predictive and pharmacodynamic markers that can be evaluated in the clinical development of JTX-8064.

In patients with rigorously defined anti-PD-1 refractory advanced melanoma, we show that increasing intratumoral IL-12 through plasmid electroporation induced deep, durable, systemic responses including regression of visceral lesions with nominal systemic toxicity. This work represents the final stage of translation of our earlier biomarker findings using an approach that adds negligible toxicity to standard-of-care checkpoint inhibitor therapy.

Signal transduction pathways control the functional state of the different immune cell types. OncoSignal RT-qPCR tests quantify the activity of signaling pathways in immune cells from blood or tissue samples. Tests can be used to assess activation of the innate and adaptive immune response, enabling use to predict and monitor response to immunotherapy. OncoSignal-based companion diagnostics will be commercialized in a partnership with QIAGEN.

Prostate cancer (PCa) is primarily driven by androgen receptor (AR) signaling, with treatment dominated by androgen depravation therapies (ADT). Although the prognosis for low-grade localized PCa is favorable, about half of patients with high-grade localized disease develop metastatic castration resistant PCa (mCRPC). The 5-year survival of mCRPC is only 30%, highlighting an unmet therapeutic need for this subset of patients. PCa is characterized by a low level of tumor infiltrating lymphocytes and is considered a “cold” tumor. Even when CD8+ cells are present in the tumor and stroma, they have been shown to poorly correlate with prognosis, perhaps due to non-active or senescent cell states. Tregs, TAMS, reactive fibroblasts and myofibroblasts, growth factors and cytokines also contribute to the immuno-suppressive environment, tumor invasion, angiogenesis and tumor-stromal cross-talk of PCa. Through an integrative tumor and immunological molecular profiling of primary low- and high-grade treatment-naïve PCa samples, we have identified a subset of high-grade PCa patients who relapse under SOC that may benefit from the addition of adjuvant immune checkpoint blockade (ICB) to SOC to overcome immune suppression. Our analysis suggests that high-grade PCa is characterized by low antigenicity as assessed by loss of MHC-I protein expression and an immunosuppressive microenvironment rich in CAFs, macrophages, T-regs and T-cell exclusion phenotypes. Unlike low-grade, high-grade PCA can have a poor prognosis (within 5 year relapse). However, a subset of high-grade PCa patients that metastasized while on SOC exhibited a biomarker profile that favors the combination of SOC with ICB (lower tumor AR expression, retained tumor MHC-I expression, moderate CD8+ T-cell infiltration and a high IFN-γ RNA signature).

Despite progress in melanoma treatment with immunotherapy, many patients develop resistance and succumb to their disease. Analysis of patient-derived tumors and blood samples have allowed us to understand mechanisms of therapeutic response and resistance. We are using multi-omic strategies to identify novel therapeutic targets and biomarkers to assist in risk stratification and disease management.

Several biomarkers of response to immune checkpoint inhibitors show potential but are not yet scalable to the clinic. We developed a pipeline that integrates deep learning on histology specimens with clinical data to predict ICI response in melanoma. We built a multivariable classifier that predicted response with AUC 0.80. With prospective validation, we believe our approach has potential for integration into clinical practice.