Towards Phase 2/3 Trials for Epstein - Barr Virus (EBV)-Associated Malignancies,

About 30% of lymphomas arising in immunocompetent individuals as well as >90% of undifferentiated nasopharyngeal carcinomas (NPCs) carry the EBV genome and express four of about 90 potential viral antigens (type 2 latency). We have targeted two of these antigens (LMP1 and LMP2) in clinical trials for the treatment of multiply-relapsed and refractory lymphoma and NPC. LMP-specific T cells had minimal short or long-term toxicities and produced tumor responses in over 70% patients with relapsed lymphoma and complete response in over 60%. In patients with NPC it was more difficult to manufacture LMP-specific T cells. 15/34 (45%) of infused patients had tumor responses including 5 CRs, which were predominantly observed in patients with locoregional disease (4/5).

LMP-specific T cells were activated using an initial stimulation of patient PBMCs with autologous dendritic cells (DCs) transduced with an adenoviral vector expressing LMP1 and LMP2 (Ad5f35-LMP1-IRES-LMP2). Nine days later and weekly thereafter responder T cells were restimulated with autologous EBV-transformed B lymphoblastoid cell lines (LCLs) overexpressing LMP1 and LMP2 from the same Ad vector until sufficient cells for the dose, QC and research were expanded. This manufacturing strategy presents several critical barriers to its transition to pivotal phase 2 and 3 clinical trials. First, the live virus (EBV) and viral vectors (adenovirus) used in our current manufacturing process will significantly increase the regulatory hurdles. Second, autologous LCL lines required as an antigen-presenting cell (APC) from lymphoma patients are difficult to establish now that Rituximab is used as standard therapy for lymphoma. Third, T cells specific for adenovirus antigens, derived from vector (virion proteins) and EBV antigens expressed in the LCL, but not in type 2 latency tumors can dominate the cultures; a particular problem in patients with NPC. Fourth, the CTL manufacturing is complex and takes a minimum of 12 weeks. We have developed a new manufacturing strategy that addresses these problems.

To manufacture type 2 latency antigen-specific T cells without using adenovirus or EBV, PBMCs are stimulated with overlapping peptide libraries (20 amino acids overlapping by 15) spanning the entire protein sequences of LMP1, LMP2 and EBNA1 in the presence of cytokines. By day 9, the T cells require restimulation, but B cells and monocytes that can present antigens effectively are depleted from the responding population. Therefore we used as APCs, autologous activated T cells (T-APCs) that express HLA class I and class II antigens pulsed with the same pepmixes together with an HLA-negative cell line (K562) expressing costimulatory molecules to prevent anergy induction.

This combination of antigen and antigen-presenting cells produces antigen-specific T cell expansion without reducing the T cell repertoire, which remains broad with subsequent stimulations. Competition between peptides within the libraries ensures that the avidity of the responding T cells remains high. By eliminating the requirement for EBV-LCLs we have reduced the minimum generation time from 10 weeks to ∼30 days. There were no detectable differences in the phenotype or function of these T cells, but the specificity was more focused on relevant antigens. In patients with relapsed NPC the frequency of LMP/EBNA1-specific T cells was increased in the presence of blocking antibodies to PD-L1. This phenomenon was not found in healthy donors, suggesting that type 2 latency antigen-specific T cells might be anergic in NPC patients.

We have now developed a strategy that eliminates live virus and viral vectors from the manufacturing process, eliminates the requirement for robust patient B cells, eliminates the reactivation of adenovirus-specific T cells and irrelevant EBV-specific T cells, increases the frequency of LMP-specific T cells in the cultures and adds EBNA1 as an additional target antigen. The clinical equivalency of these T cells in patients with NPC and lymphoma remains to be tested in clinical trials.

No relevant conflicts of interest to declare.