Third-Party Donor Virus-Specific T Cells Are Efficacious in the Treatment of Refractory Viral Infection Following Allogeneic HSCT, but May Not Persist Post-Infusion

The efficacy of adoptive transfer of stem cell donor-derived virus-specific T cells (VST) to prevent or treat viral infection in allogeneic HSCT is well established. However, this approach has some limitations. It is prerequisite on the donor cells being accessible, and virus seropositive. The generation of a product applicable to a single intended recipient is cost- and labour-intensive, and may not be immediately available in urgent clinical scenarios. Creation of a bank of cryopreserved third party donor-derived VST addresses these limitations.

We established such a bank of 177 VST, specific for cytomegalovirus (CMV n=77), Epstein Barr virus (EBV n=55), and adenovirus (AdV n = 47). To date, 19 allogeneic HSCT patients have been treated with partially HLA-matched VST for viral reactivation or disease (CMV=17 (2 with CMV colitis), EBV=1, AdV=1); persistent despite at least 14 days of antiviral treatment. Patients were transplanted from unrelated (n=12), sibling (n=4), or haploidentical (n=3) donors. The graft source was PBSC (n= 15), cord (n=1), or bone marrow (n=3), and 12/19 patients underwent in vivo T cell depletion. The 17 patients treated for CMV received a median of 31 (15-113) days of antiviral therapy prior to infusion, and 14/17 were transplanted from a CMV negative donor. Patients were eligible for up to 4 infusions of 2x10⁷/m²VST if viral persistence was documented 2 weeks after initial infusion. VST were matched at a minimum of one HLA allele between VST and recipient, and chosen on the basis of greatest number of HLA matches with preference for products with viral activity restricted through a shared HLA. A total of 31 VST have been infused; 1 patient received 4 infusions, 1 patient received 3 infusions, 7 patients received 2 infusions, and 10 patients received 1 infusion. VST were matched at 1/6 (n=4) to 4/6 HLA alleles, with 24/31 infusions matched at both a class I and class II HLA allele. The antiviral activity of VST products (determined by MHC tetramer staining or cytokine response to epitope stimulation) generated from 12 third-party donors was predominantly restricted through class I HLA alleles: A1 (n=4), A2 (n=2), A2 + B7 (n=1), A2 + DRB04:01 (n=1), A2 + B35 + DRB01:01 (n=1), A24 (EBV-specific, n=1), B7 (n=1), and DRB01:01 (Adv-specific, n=1). Best viral response post-initial VST was assessed in 18 evaluable patients. A complete response (CR) was documented if virus PCR negativity was achieved at any time, and partial response (PR) for a greater than 50% reduction from the pre-administration titre. At a median follow-up of 6.3 (1-14.8) m, a CR was attained in 14/18 (78%) patients and PR in 4/18 (22%) patients. Both patients with CMV colitis achieved CR and symptoms resolved. Reinitiation of antiviral therapy after cessation was not required in 15/18 patients post-final VST infusion. The median number of months free from antiviral therapy was 5.1 (0-14.6). No immediate infusion-related toxicities have occurred. One patient with skin GVHD prior to enrolment developed grade II acute skin GVHD post-infusion. Chronic GVHD has been documented in 4 patients (2 had preceding acute GVHD pre-infusion, 1 occurred following donor lymphocyte infusion with no preinfusion GVHD, 1 developed mild skin cGVHD with no preceding GVHD).  One patient treated with EBV-specific VST was not evaluated for response as he died from relapsed disease 14 days after VST infusion. A further 4 deaths have occurred; 1 due to presumed CMV disease following a PR to VST; 1 bacterial pneumonia; 1 relapsed disease; 1 due to CNS EBV PLTD in a patient with a CR to CMV-specific VST. T cell receptor deep sequencing was performed to investigate the persistence of infused cells in treated patients, but no clones detected in individual VST products were shown to persist beyond 24hrs post-infusion in the 3 patients examined.

Overall, third party VST appear to be safe and efficacious with a 100% overall response rate and durable responses demonstrated in the 18 evaluable patients reported. The use of an ‘off the shelf’ product has allowed prompt treatment of viral infection in a number of of allogeneic HSCT patients in whom the HSCT donor would not have been suitable or available for generation of T cells.  It appears that expansion and persistence of infused cells may not be the predominant mechanism for achieving the demonstrated clinical responses, and investigations are underway to examine alternative potential mechanisms of VST mediated immune modulation.