Virus-specific T cells for adenovirus infection after stem cell transplantation are highly effective and class II HLA restricted

Abstract

Infection with adenoviruses is a common and significant complication in pediatric patients after allogeneic hematopoietic stem cell transplantation. Treatment options with traditional antivirals are limited by poor efficacy and significant toxicities. T-cell reconstitution is critical for the management of adenoviral infections, but it generally takes place months after transplantation. Ex vivo-generated virus-specific T cells (VSTs) are an alternative approach for viral control and can be rapidly generated from either a stem cell donor or a healthy third-party donor. In the context of a single-center phase 1/2 clinical trial, we treated 30 patients with a total of 43 infusions of VSTs for adenoviremia and/or adenoviral disease. Seven patients received donor-derived VSTs, 21 patients received third-party VSTs, and 2 received VSTs from both donor sources. Clinical responses were observed in 81% of patients, with a complete response in 58%. Epitope prediction and potential epitope identification for common HLA molecules helped elucidate HLA restriction in a subset of patients receiving third-party products. Intracellular interferon-γ expression in T cells in response to single peptides and response to cell lines stably transfected with a single HLA molecule demonstrated HLA-restricted CD4+ T-cell response, and these results correlated with clinical outcomes. Taken together, these data suggest that VSTs are a highly safe and effective therapy for the management of adenoviral infection in immunocompromised hosts. The trials were registered at www.clinicaltrials.gov as #NCT02048332 and #NCT02532452.

Graphical abstract

Figure 1.

Expansion of adenovirus-specific T cells in peripheral blood of recipients of VSTs by ELISpot. (A) Quantitation of the number of adenoviral T cells before VSTs compared with the peak number in the first 4 weeks after infusion presented as SFCs. Graph shows data from patients with clinical response and an increase in T-cell number after infusion. For patients with multiple infusions, data reflect T-cell increase after the first infusion that resulted in a response. (B-C) Inverse relationship between frequency of circulating VSTs and adenoviral burden. (B) Representative examples of 3 VST recipients who had CRs after infusion with a corresponding increase in T-cell number by ELISpot. (C) ELISpot in 1 patient with no clinical response; note small scale of the y-axis demonstrating minimal T-cell numbers. Open symbols denote number of SFCs per 400 000 PBMCs. Solid symbols denote adenoviral burden in peripheral blood. UPN, unique patient number.

Figure 2.

In silico predicted adenoviral epitopes elicit a VST response by intracellular flow cytometry. (A) Representative example of VST product was stimulated with both the entire adenoviral Pepmix and single peptides synthesized on the basis of epitope prediction. Each single peptide is predicted to be presented by the noted HLA molecule present within the cellular product. Differential degree of IFN-γ positivity seen by intracellular flow cytometry, indicating that multiple epitopes are involved in provoking a T-cell response in this product. CD4-predominant response is seen, but little CD8 response is observed.

Figure 2.

In silico predicted adenoviral epitopes elicit a VST response by intracellular flow cytometry. (B) Representative example of a product that had been given to a patient with a complete clinical response. This product has a response to peptides predicted to be presented by DRB1*07:01 but not to the peptide predicted to be presented by DRB1*15:01. (C) Quantitation of the number of adenoviral T cells by ELISpot in 2 recipients of third-party VSTs. PBMCs were stimulated with either adenoviral Pepmix or single peptides with the HLA predicted to present the epitope indicated in the table. At all timepoints, the background media spot counts have been subtracted.

Figure 3.

SALs for the direct determination of HLA restriction. (A) Schematic demonstrating the model system. Mouse fibroblasts are stably transfected with a single HLA molecule, and Pepmix is added to these cells in culture. After incubation, Pepmix is removed so that the only antigen present is what is bound to and presented by the SAL. SALs are then cocultured with VSTs and intracellular flow cytometry is performed. (B-C) Two retrospective examples of using SALs to determine HLA restriction and correlate with clinical outcomes. In panel B, the coculture of a VST product given to patient UPN-69 shows IFN-γ positivity only with DRB1*03:01 SALs, indicating restriction through that allele. The patient and product did not match at DRB1*03:01 and had no clinical response. By contrast, in Figure 2C, a coculture of the VST product given to patient UPN-32 shows restriction by SALs through both DRB1*03:01 and DRB1*07:01. The patient and product matched at DRB1*07:01 and had a robust clinical response.