Tumor associated antigen specific T-cell populations identified in ex vivo expanded TIL cultures

Abstract

Ex vivo expanded tumor infiltrating lymphocytes (TILs) from malignant melanoma (MM) and head & neck squamous cell carcinoma (HNSCC) share a similar oligoclonal composition of T effector memory cells, with HLA class I restricted lysis of tumor cell lines. In this study we show that ex vivo expanded TILs from MM and HNSCC demonstrate a heterogeneous composition in frequency and magnitude of tumor associated antigen specific populations by Elispot IFNγ quantitation. TILs from MM and HNSCC shared reactivity towards NY ESO-1, cyclin B1 and Bcl-x derived peptides. Additionally we show that dominating T-cell clones and functionality persists through out expansion among an oligoclonal composition of T-cells. Our findings mirror prior results on the oligoclonal composition of TIL cultures, further indicating a potential for a broader repertoire of specific effector cells recognizing the heterogeneous tumors upon adoptive transfer; increasing the probability of tumor control by minimizing immune evasion by tumor cell escape variants.

Introduction

Head & neck squamous cell carcinoma (HNSCC) is the most frequent cancer derived from the aero-digestive tract. Patients with end stage (stage III–IV) and recurrent disease have a very poor prognosis when radiotherapy and/or surgery are no longer an option. Malignant melanoma (MM) is also a poorly controllable disease once it has reached an advanced stage. Unlike HNSCC, MM is often asymptomatic which allows for a disease spread prior to diagnosis. In spite of massive experimental clinical trials in both patient groups, survival has not improved significantly over the past two decades. Meanwhile, a recent phase III clinical trial has shown promising results in MM patients treated with Ipilimumab (CTLA-4 antibody) by significantly increased survival in the Ipilimumab treated patient group [1]. Although the molecular and cellular background for efficacy is largely unknown these data suggest that manipulation with cellular suppressive mechanisms may tip the balance from immune tolerance of cancer towards immunological attack.

Tumors are frequently infiltrated by cells of the immune system, which controls tumor growth at an early stage. Due to genetic instability the tumor cells undergo mutations leading to escape variants and progression [2]. In addition the tumor acquires immune suppressive mechanisms at a cellular (Treg [3], MDSC [4]), soluble (TH2 cytokines) and metabolic (ex. IDO [5]) level. Thereby the tumor environment is skewed towards a tolerogenic state.

In an attempt to tip the balance towards a tumor immunogenic state, different approaches of cancer immunotherapy have been tested including administration of cytokines and therapeutic vaccination. To the latter, although measurable immune responses were induced, limited clinical responses were obtained. However, recent studies in adoptive T-cell therapy have shown very promising results where 50% of MM patients with advanced disease obtained long lasting clinical objective responses after preconditioning lymphodepletion, infusion of ex vivo expanded tumor infiltrating lymphocytes (TILs) and high dose IL-2 [6]. TILs comprised T-cell populations with specificity towards autologous tumor, serving as a source of unknown tumor associated antigens (TAAs) [7]. Regarding identification of TAA peptide antigen specific T-cells among TILs, most attention has been directed at single epitopes derived from differentiation antigens such as MART-1 in MM patients. Thus, the antigen specific repertoire of TILs remains unknown and further attention is needed to identify tumor targets of potential clinical importance in adoptive T-cell therapy.

On the basis of the published methods from Dudley et al. we have established a method for successful expansion of TILs from both MM and HNSCC patients [8]. This has let us to initiate clinical pilot trials in MM patients (NCT00937625) and HNSCC patients (NCT00937300).

To identify the TAA specific populations in the TILs, we screened for reactivity against a panel of peptides covering three main categories of over expressed (OE), cancer testis (CT) and differentiation (D) TAAs. Belonging to the first category, the BCL-2 family proteins (Mcl-1, Bcl-2 and Bclx L) and survivin inhibits apoptosis and are in many cases crucial for tumor cell survival [9], [10], [11], [12]. Telomerase is another over expressed protein that ensures the capacity of cancer cells to continued proliferation [13], [14], while cyclin B1 ensures sustained mitosis [15], [16] and RhoC facilitates increased metastatic potential [17], [18]. Cancer testis antigens (CT) (MAGE 1, MAGE 3 and NY ESO-1) are silenced in normal tissue and upregulated in MM [19] and HNSCC [20], [21], rendering them suited targets for immune recognition. Finally, the differentiation antigen MART-1 is recognized as a highly immunogenic target in MM [22]. Specific T-cell populations for all the included peptides have previously been identified in the circulation of cancer patients. Here we show an occurrence of a broad repertoire of TAA specific T-cell populations from multiple individually established TIL cultures from MM and HNSCC patients. Although there is a tendency of an unequal distribution of TAA specific subgroups between TIL cultures and between patients, there is an equal broad representation of responses towards over expressed antigens (e.g. cyclin B1 and Bcl-x) and cancer testis antigens (e.g. NY ESO-1) in both patient categories.