Epitope and HLA-type independent monitoring of antigen-specific T-cells after treatment with dendritic cells presenting full-length tumor antigens

Abstract

The efficacy of cancer immunotherapy can be improved by treatment with full-length tumor antigen and by combining several antigens. This approach allows the induction of a broad immune response irrespective of the patient's HLA type which at the same time challenges immune monitoring. Also, the number of available lymphocytes is most often limited and minimal in vitro restimulations of the lymphocytes should maintain information about the actual in vivo situation.

To overcome these hurdles, we developed a method to measure the CD8⁺ and CD4⁺ T-cell responses directly ex vivo. Skin biopsies taken from dendritic cell (DC)-induced DTH reactions from melanoma patients participating in a DC-clinical trial served as lymphocyte source. Antigen-specificity of skin infiltrating lymphocytes was investigated by coculture with antigen-presenting autologous B cells and assessed for CD137 upregulation and enhanced cytokine secretion.

Using this approach we could detect treatment-specific CD8⁺ T-cells without restimulation in vitro. Upregulation of the activation marker CD137 correlated with the upregulation of the lytic marker CD107a. CD137 upregulation by treatment-specific CD4⁺ lymphocytes however was more pronounced after antigen-specific in vitro restimulation. Both CD8⁺ and CD4⁺ lymphocytes could be further expanded using the same B cells as for screening allowing characterization of the recognized antigenic region. In addition, this technique can be extended to detect a broader array of T-cell functions and to monitor a large cohort of patients.

We believe that this approach of direct ex vivo monitoring, irrespective of the patient's HLA-type or the recognized peptide, and using a limited number of lymphocytes is a valuable tool in the immune monitoring of current cellular immunotherapies.

Introduction

Many studies, both present and past, evaluate the contribution of antigen-specific immunotherapy to control diseases such as autoimmunity and cancer. In the case of cancer, antigen specific immunization, in particular through dendritic cell (DC)-based-immunotherapy has been shown to be safe and feasible while making step by step progress in clinical outcome (Kantoff et al., 2010, Schwartzentruber et al., 2011). Besides clinical follow-up, monitoring of the cellular immune response is a prerequisite to understanding the underlying mechanisms responsible for the presence or absence of clinical efficacy and to allow further therapy improvements.

Ideally, tumor infiltrating lymphocytes (TIL) should be monitored (Haanen et al., 2006, Ladanyi et al., 2007, Tjin et al., 2011). However, tumor material is often not available or accessible. Recently, a positive correlation has been shown between the progression free survival of stage IV melanoma patients (clinical outcome) and the presence of functional tumor-antigen specific lymphocytes in the skin after induction of delayed-type hypersensitivity responses (DTH) (de Vries et al., 2005, Aarntzen et al., 2008). As such, DTH or skin infiltrating lymphocytes (SKILs) can be monitored as a good surrogate for TIL. Unfortunately, only a low number of SKILs can be obtained from a DTH biopsy.

DC-based immunotherapy increasingly uses DCs loaded with multiple and full-length tumor antigens (Van Tendeloo et al., 2010, Wilgenhof et al., 2011). This allows stimulation of a broad anti-tumor T-cell repertoire and provides the opportunity for more cancer patients to enter immunotherapy trials independent of their HLA-type (Wilgenhof et al., 2011, Van Nuffel et al., in press).

The monitoring of cellular immune responses against a broad array of possible epitopes becomes a complex matter, especially when limited testing material is available and when patients are not selected for particular HLA-types. Both the CD8⁺ and CD4⁺ T-cell response should be monitored maintaining as much information as possible about the actual in vivo situation, thus with as little in vitro restimulations as possible. Once a cellular response is detected, one should be able to characterize the fine specificity and the functionality of the induced antigen-specific T-cells. Also, monitoring should be performed with minimal use of patient's material necessary to generate the therapeutic product, such as peripheral blood mononuclear cells (PBMC). Finally, monitoring should not be time-consuming to allow screening of a larger cohort of patients in a clinical trial.

Monitoring antigen-specific cellular immune responses commonly includes detection of T-cell receptor specificity, measurement of the antigen-specific T-cell-proliferation, measurement of cytotoxicity, and of secretion of various factors such as cytokines and chemokines. To this end, techniques such as HLA-peptide multimer staining, ³H-thymidine incorporation, CFSE dilution assay, ⁵¹Cr-release assay, CD107a assay, FATT-CTL, ELISA, ELISpot assays, intra-cellular cytokine and chemokine staining, antigen-specific real time RT-PCR, etc. are being used (Housseau et al., 2002, Hernandez-Fuentes et al., 2003, Keilholz et al., 2006). Variations of these techniques have been described (Schultz-Thater et al., 2008, Hadrup et al., 2009, Newell et al., 2009, Van Camp et al., 2010) but still none of them fulfill all of the above mentioned requirements.

To address these issues, we developed a novel approach to monitor CD8⁺ and CD4⁺ T-cell responses. This assay combines the use of autologous EBV-immortalized B cells (aEBV-B-cells) as antigen presenting cells (Issekutz et al., 1982, Topalian et al., 1994) and CD137 upregulation (Wolfl et al., 2007, Wehler et al., 2008) together with cytokine secretion (IFN-γ and TNF-α) to detect antigen-specific, activated and functional T-cells. This approach allows combination with other flow cytometric markers as well as extensive quantification of secreted soluble factors to further characterize the T-cells in a single sample. The same tumor antigen expressing aEBV-B-cells can be used to further expand the antigen-specific T-cells in vitro in order to determine the recognized antigenic region. We further validate the feasibility of our technique by monitoring the presence of treatment antigen-specific skin infiltrating lymphocytes (SKILs) in melanoma patients enrolled in a pilot trial of TriMix-DC loaded with gp100, tyrosinase, MAGE-C2 and MAGE-A3 (Bonehill et al., 2009, Wilgenhof et al., 2011).

Together, we believe this technique is a useful tool to monitor immunotherapy trials with complex treatment strategies.