Safety and clinical efficacy of rapidly-generated trivirus-directed T cells as treatment for adenovirus, EBV, and CMV infections after allogeneic hematopoietic stem cell transplant

Abstract

Adoptive transfer of virus-specific T cells can prevent and treat serious infections with Epstein-Barr virus (EBV), cytomegalovirus (CMV), and adenovirus (Adv) after allogeneic hematopoietic stem cell transplant. It has, however, proved difficult to make this approach widely available since infectious virus and viral vectors are required for T cell activation, followed by an intensive and prolonged culture period extending over several months. We now show that T cells targeting a range of viral antigens derived from EBV, CMV, and Adv can be reproducibly generated in a single culture over a 2-3-week period, using methods that exclude all viral components and employ a much-simplified culture technology. When administered to recipients of haploidentical (n = 5), matched unrelated (n = 3), mismatched unrelated (n = 1) or matched related (n = 1) transplants with active CMV (n = 3), Adv (n = 1), EBV (n = 2), EBV+Adv (n = 2) or CMV+Adv (n = 2) infections, the cells produced complete virological responses in 80%, including all patients with dual infections. In each case, a decrease in viral load correlated with an increase in the frequency of T cells directed against the infecting virus(es); both immediate and delayed toxicities were absent. This approach should increase both the feasibility and applicability of T cell therapy. The trial was registered at www.clinicaltrials.gov as NCT01070797.

Figures

Figure 1

Cell expansion and immunophenotype of rCTL generated for clinical use. Plasmid-activated rCTL were expanded in the G-Rex in the presence of IL4+7 for 9–11 days. Panel a shows overall T cell expansion, based on cell counting using trypan blue exclusion. Each symbol represents an individual line, and data for 36 rCTL lines is presented. Panel b shows the phenotype of the rCTL on the day of cryopreservation. Reactivity of CTL lines (n = 36) with antibodies against the T cell surface antigens CD3, CD4, CD8, and CD56, and the activation/memory markers CD45RO and CD62L is shown. The mean for each condition is represented as a black line.

Figure 2

Virus-specific activity and lack of alloreactivity in rCTL. Panel a shows the frequency of Adv (Hexon and Penton), EBV (LMP2, EBNA1, BZLF1), and CMV (IE1 and pp65) reactive T cells in rCTLs using IFNγ ELIspot as readout. The top panel shows the specificity of 22 lines generated from CMV seropositive donors, while the bottom panel shows the activity against Adv (Hexon and Penton) and EBV (LMP2, EBNA1, BZLF1) in 14 lines generated from CMV seronegative donors. Results represent the mean ± SEM SFC/2 × 105 input cells. Control was IFNγ release in response to stimulation with irrelevant/no pepmix. Panel b51Cr release at 4 hours after coincubation of rCTL lines with recipient/haploidentical PHA blasts (allogeneic targets). These data are mean percent specific lysis (± SD) of targets by 35 rCTL lines at E:T ratios of 40:1, 20:1, 10:1, and 5:1. This analysis could not be performed for one individual who was post-transplant and lacked a haploidentical donor from which PHA blasts could be generated.

Figure 3

In vivo expansion and clinical benefits of rCTL in subjects infected with one virus. Panels a and b show the CMV viral load in blood (copies/ml), indicated by the dashed line (---), and frequency of T cells specific for CMV pp65 and IE1 pre and postinfusion in two subjects (UPN 1927 and UPN 2472) who were treated for a CMV reactivation. T cell frequencies are shown as a solid line. Panel c shows the Adv load in blood (copies/ml – dashed line) and frequency of T cells specific for Adv Hexon and Penton pre and postinfusion (solid lines) in subject UPN 2530 who was treated for an Adv infection. This patient received two rCTL infusions, one on day 0 and one 4 weeks after the initial infusion, which is indicated by the arrow. Panel d shows the EBV viral load (copies/µg DNA isolated from PBMCs – dashed line) and frequency of T cells specific for EBV LMP2, EBNA1, and BZLF1 (solid lines) in UPN 2639 treated for an EBV reactivation. In all cases the frequency of specific T cells was measured by IFNγ ELIspot and results are presented as average SFC/4 × 105 PBMCs.

Figure 4

In vivo expansion and clinical benefits of rCTL in subjects with dual infections. Panel a shows the EBV and Adv load (dashed lines) detected in the peripheral blood of UPN 2632. This patient was initially infused with rCTL as treatment for an EBV reactivation and subsequently developed an Adv infection. The frequency of T cells directed to EBV (LMP2, EBNA1, BZLF1) and Adv (Hexon and Penton) (shown as solid lines) was measured over time by IFNγ ELIspot and results are presented as average SFC/4 × 105 PBMCs. Panel b shows the Adv and EBV load (dashed lines) detected in the peripheral blood of UPN 2784. This patient was initially infused with rCTL as treatment for an Adv infection and subsequently reactivated EBV. The frequency of T cells directed to Adv (Hexon and Penton) and EBV (LMP2, EBNA1, BZLF1) (solid lines) was measured over time by IFN ELIspot and results are presented as average SFC/4 × 105 PBMCs. Panels c and d show the CMV and Adv load detected in the peripheral blood of UPN 2537 and UPN 2763 (dashed lines). Both these patients were infused with rCTL as treatment for a CMV reactivation and subsequently developed an Adv infection. The frequency of T cells directed to CMV (IE1 and pp65) and Adv (Hexon and Penton) (solid lines) was measured over time by IFNγ ELIspot and results are presented as average SFC/4 × 105 PBMCs.