Establishment of a Highly Characterized Third-Party Virus-Specific T Lymphocyte Bank for Treatment of EBV+ Lymphoma

Although autologous or donor-derived EBV-specific T cells (EBVSTs) have remarkable clinical efficacy for Hodgkin and non-Hodgkin lymphoma, the lengthy manufacturing process and anergy of patient tumor-specific T cells reduce the feasibility of widespread use.   An alternative approach is establishment of a pre-made third party bank that can provide access to VSTs for almost any patient.  We previously evaluated multivirus-specific third-party T cells for refractory viral infections after HSCT.  Of 50 recipients, 74% responded, even when T cells were HLA-matched at only a single allele. Although established banks consist primarily of donors from national marrow registries or family members according to HLA genotype, there are many limitations to this approach, including the generation of lines with limited specificity for the targeted antigens.  To circumvent these drawbacks, we have developed a strategic approach to generating third-party lines for treatment of EBV+ lymphoma.  Our bank has been generated from blood-bank eligible donors, initially chosen based on racial diversity thus predicted to have HLA haplotypes representative of our diverse patient population.  Donors were subsequently screened for specificity to the EBV Type 2 latency antigens (LMP1/2, EBNA1, and BARF1) with IFN-g Elispot assays using overlapping peptide libraries (pepmixes) spanning each antigen.  The EBVSTs exhibited significant specificity (with >50% of donors recognizing 3 of the 4 antigens) and cytotoxicity to both pepmix-pulsed autologous activated T cells (ATCs) and HLA-matched EBV-lymphoblastoid cell lines (LCLs) that naturally express viral antigens.  To ensure that the best EBVST line is chosen for each recipient, it is important to ensure antigen-specific activity restricted by the alleles shared between donor and recipient. To characterize HLA restriction we use peptides or pepmix-pulsed PHA blasts or LCLs that are HLA matched at a single class I or II allele as target cells in cytotoxicity assays. We then constructed a database in which the HLA restriction of the antigen specificity of each T cell line is listed. Based on data from our third party trivirus-specific T cell study, we expect that our projected bank of 30 highly characterized EBVSTs will cover >95% of referred patients (with matching at ≥ 2 class I and/or class II alleles) since our screening strategy ensures that a wide range of HLA haplotypes is covered.  Additionally, given the significantly increased virus specificity of our highly characterized donors, we expect to see significant anti-tumor activity.  This study will therefore allow us to determine whether third-party T cells can be effective outside the HSCT setting in patients who are not immunocompromised.