Safety and Clinical Efficacy of Rapidly-Generated Trivirus-Directed T Cells After Allogeneic Hematopoietic Stem Cell Transplant

Abstract 223

We have previously demonstrated that small numbers of ex vivo-expanded, trivirus-specific T cells targeting Epstein Barr virus (EBV), cytomegalovirus (CMV), and Adenovirus (Adv) are safe, proliferate in vivo and protect human subjects against all 3 viruses following HSCT. However, broader implementation is limited by the need for infectious virus (EBV) to establish an EBV-transformed B lymphoblastoid cell line (EBV-LCL), for clinical grade adenoviral vector, and by prolonged (6wks for EBV-LCL and 6wks for T cells) and complex manufacture. Moreover, competition between viral antigens limits extension to additional viruses. We now evaluate whether it is possible to make clinically effective T cell lines using methods that exclude all viral components and utilize simplified manufacturing technology.

With NHLBI-Production Assistance for Cellular Therapies (PACT) support, 29 clinical-grade rCTL lines have been generated. From an initial 15×10⁶ PBMCs, we prepared a median of 214±88 × 10⁶ T cells (range 100–420×10⁶) over 9–11 days by using dendritic cells (DCs) nucleofected with DNA plasmids encoding immunogenic EBV (LMP2, EBNA1 and BZLF1), Adv (Hexon and Penton), and CMV (pp65 and IE1) antigens, and expanding them with IL4+7 in gas permeable (G-Rex) devices. The rCTL lines were polyclonal, comprising both CD4+ (33±3%) and CD8+ (60.5±3%) cells, that expressed the activation and memory markers CD45RO+ CD62L+ (64.3±26.6%) and CD45RO+ CD62L- (17.4±14%). Twenty lines generated from donors that were seropositive for all three viruses demonstrated activity against all 3 targets - CMV (IE1: 359±100; pp65: 637±177 SFC/2×10⁵), EBV (LMP2: 217±60, EBNA1: 67±19 and BZLF1: 111±31) and Adv (Hexon: 265±74, Penton: 191±53) - while 9 lines generated from donors who were CMV seronegative demonstrated activity exclusively against EBV (LMP2: 197±70, EBNA1: 145±51 and BZLF1: 239±84) and Adv (Hexon: 271±96, Penton: 254±90). None of the lines reacted against recipient PHA blasts (median spontaneous Cr⁵¹ release of 0% at a 20:1 effector to target cell ratio).

To date we have administered these lines to 10 recipients of allogeneic HSCT. Five patients received dose level (DL) 1 (5×10⁶/m²), 2 received DL2 (1×10⁷/m²) and 3 had DL3 (2×10⁷/m²) of this phase I/II study. Three patients were infused as treatment for CMV, 2 for Adv, 2 for EBV, 1 for EBV+Adv, and 2 for CMV+Adv. Our major anticipated concern was that these once stimulated, unselected rCTLs might cause GvHD in vivo, but that was not the case. One patient developed a skin rash 2 weeks after rCTLs but no other toxicity related to the infused cells was observed. Eight of the 10 treated patients including one patient with a biopsy-proven EBV lymphoma and the 3 patients with double reactivations had complete responses to rCTL therapy with a return of viral load to normal and resolution of all other symptoms. Response was associated with an increase in the frequency of virus-specific T cells detected in the peripheral blood against the infecting virus. For CMV there was an increase from a median of 0.5 to 96 and 1 to 277 SFC/4×10⁵ IE1 and pp65-specific T cells respectively 3–6wks post-infusion; for Adv an increase from a mean of 0.5 to 137 and 0.5 to 99 SFC/4×10⁵ Hexon and Penton-specific cells 2wks post-infusion, respectively, and for EBV an increase from 2.75 to 227, 1.5 to 39, and 1 to 188.5 SFC/4×10⁵ EBNA1, LMP2, and BZLF1-specific T cells 2–4wks post-infusion, respectively. Two patients failed to respond. The first had a 3 year history of persistent CMV colitis despite high circulating CMV-specific precursors (297 IE1-specific and 193 pp65-specific T cells/4×10⁵ PBMCs). Post rCTL we saw no increase in T cell precursor levels and no clinical improvement. The second was treated for an elevated EBV viral load but also had high pre-existing EBV-specific T cell precursors (60, 23, and 240 SFC/4×10⁵ EBNA1, LMP2, and BZLF1-specific T cells). Again, post-rCTL we did not detect an increase in EBV-specific precursors and no response. Thus, infusion of rCTLs has been safe and in 8/10 patients was associated with the appearance of virus-reactive T cells directed against the infecting virus in peripheral blood and subsequent virus clearance. rCTLs have the potential to increase the availability of cell products for HSCT recipients and we are currently extending this platform to additional viruses, thereby broadening the spectrum of pathogens that can be targeted by adoptive transfer of a single T cell line.