Merkel cell polyomavirus-specific CD8 T cells in Merkel cell carcinoma: T cell receptor diversity & novel immune therapies

Thesis (Ph.D.)--University of Washington, 2016-06

Merkel cell carcinoma (MCC) is an aggressive skin cancer with a disease-specific mortality of 46%. Unfortunately, no FDA-approved treatments for advanced disease exist. Over 80% of MCCs are caused by persistent expression of T-antigen oncoproteins from the Merkel cell polyomavirus (MCPyV). Systemic immunity is strongly linked to both MCC incidence and disease-specific survival, with an especially important role played by the adaptive immune response. Additional background about the biology and immune response to MCC is provided in the Introductory Chapter 1. The current standard of care for advanced MCC, cytotoxic chemotherapy, elicits tumor responses in approximately half of patients, but responses are not durable with a median progression-free survival among responders of only 3 months. In Chapter 2, we review the literature surrounding such conventional therapeutics and outline the scope of emerging immune-based therapies. Immune-stimulating agents are thought to hold great therapeutic promise for MCC. Over half of patients with MCCs caused by MCPyV (‘virus-positive MCC’) develop robust oncoprotein-specific antibodies and/or T cell responses to persistently expressed T antigen (T-Ag) of MCPyV. However, though many patients have cytotoxic CD8+ T cells specific for these oncoproteins, MCCs still develop and persist. Previous studies found that robust infiltration of CD8+ lymphocytes into the tumor is associated with 100% MCC-specific survival, yet unfortunately infiltration is observed in only 4-18% of MCCs. We explored two distinct mechanisms that we hypothesized may contribute this observed immune dysfunction. Importantly, we hoped to uncover ways in which this immune evasion may be therapeutically targeted. First, we explored the hypothesis that there is significant genetic and functional diversity among T cell receptors (TCRs) recognizing one common HLA type/MCPyV epitope, and that TCR repertoire differences may be linked to patient outcomes. Indeed, detailed in Chapter 3, we found almost 400 different TCRβ clonotypes among epitope-specific T cells from 12 patients. Only one of these unique TCRβ clonotypes was detected in more than one patient, highlighting the diversity of the T cell response to this epitope. Additionally, there is significantly improved MCC-specific survival among patients with increased infiltration of epitope-specific T cells within their primary tumors. By studying the functional properties associated with individual TCRs among select patients, we found that patients with improved MCC-specific survival have more functionally avid TCRs. This work has identified avid TCRs with potential for use in transgenic T-cell therapy for MCC. Secondly, we hypothesized that MCPyV-specific T cells are exhausted. In Chapter 4 we found that MCPyV-specific peripheral blood mononuclear cells (PBMC) and tumor infiltrating lymphocytes (TIL) express high levels of inhibitory receptors that inhibit TCR signaling (PD-1 and Tim-3) directly ex vivo. In addition, we show that while PBMC do not produce the effector cytokine IFN-γ in response to viral peptide stimulation directly ex vivo, they are able to regain effector function after culture, which is enhanced by the addition of antibodies that block these inhibitory receptors such as anti-PD-1. Remarkably, since the review of emerging immune-stimulating agents (Chapter 2) was published in Spring of 2013, many of these novel immune-stimulating agents have now been evaluated in formal studies and have showed impressive efficacy. A retrospective analysis of the use of single-fraction radiation therapy in metastatic MCC patients is presented in Chapter 5, demonstrating the 94% efficacy at palliating target tumor lesions with few side effects and impressive durability, especially among patients with intact systemic immunity. In addition, PD-1 blocking antibodies are now being utilized therapeutically, and in Chapter 6 we present the results of the first clinical trial of a PD-1 agent, pembrolizumab, in MCC. Excitingly, 54% of MCC patients in the trial responded to this drug, and 67% of responding patients remained progression-free at 6 months. In this chapter we also detail our finding that both virus-positive and virus-negative MCC respond to pembrolizumab, and that many of the biomarkers used for other cancers to predict response to PD-1 agents, such as tumoral PD-L1 expression or CD8 infiltration, may not be relevant in MCC. We further perform correlative studies on patient tumor and PBMC samples from these clinical trial patients throughout their therapeutic course, and our preliminary findings are presented in Chapter 7. In summary, we have used our extensive and unique repository of clinically annotated MCC blood and tumor specimens to further elucidate aspects of the CD8+ T cell response to MCC such as TCR diversity, inhibitory mechanisms, and augmentation with immunotherapy. These studies have helped us define the mechanisms by which some patients mediate superior disease outcomes, and have contributed to our goal of improving therapeutic options for patients with advanced MCC.