Third-party CMV- and EBV-specific T-cells for first viral reactivation after allogeneic stem cell transplant

Abstract

Virus-specific T-cells (VSTs) from third-party donors mediate short- and long-term antiviral effects in allogeneic hematopoietic stem cell transplant (HSCT) recipients with relapsed or refractory viral infections. We investigated early administration of third-party VSTs, together with antiviral therapy in patients requiring treatment for first cytomegalovirus (CMV) or Epstein-Barr virus (EBV) infection. Thirty HSCT patients were treated with 1 to 4 VST infusions (2 × 107 cells/m2; CMV n=27, EBV n=3) at a median of 4 days after initiation of antiviral treatment. The overall viral response rate was 100%, with a complete response (CR) rate of 94%. Of the 28 patients who achieved a CR, 23 remained virus PCR negative (n=9) or below quantitation limit (n=14) for the duration of follow-up. Four patients had brief episodes of quantifiable reactivation not requiring additional therapy, and one required a second infusion after initial CR, remaining PCR negative thereafter. All 3 patients treated for EBV post-transplant lymphoproliferative disorder achieved sustained CR. Rates of aGVHD and cGVHD after infusion were 13% and 23%, respectively. There were no serious infusion-related adverse events. VST infusion was associated with rapid recovery of CD8+CD45RA-CD62L- and a slower recovery of CD4+CD45RA-CD62L- effector memory T-cells; CMV-specific T-cells comprised up to 13% of CD8+ cells. At 1 year post-transplant, non-relapse mortality was 10%, cumulative incidence of relapse was 7%, overall survival was 88% and 25 of 27 patients had ECOG status of 0 or 1. Early administration of third-party VSTs in conjunction with antiviral treatment appears safe and leads to excellent viral control and clinical outcomes. Registered on Australian New Zealand Clinical Trials Registry as #ACTRN12618000343202.

Graphical abstract

Figure 1.

Clinical outcomes. (A) Cumulative incidence of complete response (CR) adjusted for competing risks, over the 6-month follow-up period after VST infusion. (B) Cumulative incidence of disease relapse and non-relapse mortality in the first 12 months after transplant. (C) Overall survival in the first 12 months after transplant. N = 30 for all.

Figure 2.

Composite swimmer plot showing the clinical course of patients over the 12 months after VST infusion. Patients appear in order of recruitment on the y-axis. Patients treated for EBV infection are noted (n = 3); all others were treated for CMV infection (n = 27). All patients received the first VST infusion at day 0 (green line), and their subsequent clinical courses are illustrated. Antiviral treatment is shown before and after infusion: blue boxes (CMV therapy) and purple vertical lines (rituximab doses for EBV).

Figure 3.

Immune reconstitution of 25 patients by flow cytometry, showing major cell populations and their subsets over time. Median T-cell subsets are shown, with the predominant CD4⁺ and CD8⁺ subset being CD45RA^-CD62L^- effector memory cells. Additional subsets are visualized in supplemental Figure 1. DPI, day post 1st VST infusion; br, bright; cm, central memory; Conv, conventional; em, effector memory; emra, terminal effector memory.

Figure 4.

Virus-specific T-cell responses and cell persistence in post-infusion samples. (A) ELISpot assay showing preinfusion and maximum CMV-induced IFN-γ production within 100 days after infusion. Statistical significance was evaluated with a 2-tailed Wilcoxon matched-pairs signed-rank test. (B) Tetramer responses for 12 patients with >1% expression, showing elevation and persistence over time. (C) Tetramer responses for patient 8 shown in composite at 4 time points. For each time point, biaxial flow plots, tSNE of major cell populations, and blood viral load (below each tSNE plot) are shown. (D) Microchimerism assessed by ddPCR in 15 patients. Informative insertion-deletion polymorphisms (indels) that distinguished the third-party VST donor from the transplant donor and recipient were identified. Adoptively transferred VST product was detected in the peripheral blood of all patients up to day 120 after infusion, all quantified below 3 copies per 100 ng of DNA. Day 0 results were from samples taken 2 hours after VST administration. Patient 6 withdrew from the study and had no further time points after day 92. tSNE, t-distributed stochastic neighbor embedding.

Figure 5.

Correlations between AUC of the viral load and various clinical and immunological outcomes, by Spearman rank correlation coefficient. (A) Nonsignificant correlation with CD8⁺ T-cell count at the time of best viral response (n = 25). (B) Significant negative correlation with mean CD8⁺ T-cells across the follow-up period (n = 25). (C) Significant positive correlation with total days in hospital from transplant to 1 year later (n = 30). (D) Significant positive correlation with total days in hospital from first VST infusion to 6 months after infusion (n = 30). (E) Nonsignificant correlation with total days of antiviral treatment (n = 30).