Multicenter Study of “off-the-Shelf” Third Party Virus-Specific T Cells (VSTs) to Treat Adenovirus (Adv), Cytomegalovirus (CMV) or Epstein Barr Virus (EBV) Infection After Hemopoietic Stem Cell Transplantation (HSCT)

Abstract 457

Adoptive transfer of virus-specific cytotoxic T lymphocytes from stem-cell-donors can avert or treat serious viral infections in recipients of allogeneic HSCT. Unfortunately, implementation is limited by the time and cost of T cell preparation for individual donor-recipient pairs, and by the inability to generate virus-specific T cells from cord-blood and other seronegative donors. A potential alternative is to prepare a bank of “off the shelf” virus-specific T cells (VSTs) from third party donors that will share one or more HLA polymorphisms. We evaluated the feasibility, safety and effectiveness of this strategy in a multicenter study through the NHLBI Specialized Centers for Cell-Based Therapy (SCCT) program. Using the NHLBI Production Assistance for Cellular Therapies (PACT) program we used our established techniques to prepare a bank of 32 VSTs from normal donors, which were directed to EBV, CMV and Adv. All the lines were polyclonal comprising CD4+ (median 10%, range 1.5–99%) and CD8+ (median 83%, range 0–96%) T cells, with populations expressing activation and memory markers [CD45RO+ CD62L− (median 68%) and CD45RO+ CD62L+ (median 29%)]. Specificity for the target viruses was shown by IFNg ELIspot; (CMV - mean 827 SFC, EBV – mean 463 SFC and Adv – mean 287 SFC/1×10⁵ cells). These lines were administered to subjects with persistent viral infection, unresponsive to alternative therapies, and we selected the line for administration based on specificity for the target virus through a shared allele and on the overall level of HLA match. Patients were eligible for a subsequent VST infusion in the event of a partial response. Of 81 screened patients after marrow (n=23), blood (n=33) or cord (12 single, 13 double) transplantation, a line was identified for 73. Of patients with an identified line, 24 did not receive a VST because they improved (n=14), became ineligible (n=6), declined treatment (n=2) or died (n=2). Forty nine patients received a total of 18 VSTs at a dose of up to 2×10⁷ cells/m². The VSTs matched the recipient at 1 (n=11), 2 (n=18), 3 (n=13) or 4 (n=2) HLA antigens and were given as one (n= 27), two (n=11), 3 (n=6), or 5 (n=1) separate infusions. Five patients were subsequently excluded from analysis as they withdrew from the study or died of their underlying disease or an intercurrent infection with another virus within 7 days of infusion. Of the 44 subjects included in the long-term analysis, 9 received VSTs for refractory EBV-PTLD, 19 for persistent CMV, and 16 for persistent Adv. There were no immediate adverse effects related to infusion. The overall cumulative incidence of first CR/PR in these 44 patients based on an independent committee adjudicated review by day 42 post-infusion was 81.8% (89.5% for CMV, 66.7% for EBV and 81.3% for Adv). We also saw tissue responses with resolution of CMV colitis and retinitis and complete responses on imaging for four patients with widespread EBV-PTLD (one CD20-negative and 3 refractory to rituximab). Of note complete antiviral responses occurred even in recipients of lines matching at only a single HLA antigen. In approx. half of the subjects who responded to VST therapy we saw an increase in virus-specific T cells (CMV – mean 106±58 SFC pre and 193±123 SFC/4×10⁵ post-infusion; Adv - 6±4 SFC pre and 125±94 SFC/4×10⁵ post-infusion; EBV 43±22 SFC pre and 104±43 SFC/4×10⁵ post-infusion) and increased virus-specific activity was associated with the appearance of donor VST-derived TCR sequences in peripheral blood, as detected by sequential TCRvβ sequence analysis. Despite the HLA disparity of VSTs and recipients, de novo GvHD occurred in only 2 subjects following infusion of these banked cells. Thus of the eight patients who developed acute GVHD within 45 days of their first infusion (6 Grade I, 1 Grade 2 and 1 Grade 3), 6 had a previous history of GVHD prior to VST infusion. One additional patient had a flare of chronic skin GVHD. Two patients developed transplant-associated microangiopathy but had other risk factors including sirolimus use. These results show that it is feasible and safe to implement adoptive immunotherapy for CMV, EBV, and Adv using banked allogeneic VSTs, and that the approach may be both practicable and effective for a high proportion of recipients who otherwise lack options for the treatment of intractable virus infections after stem cell transplantation.