Antiviral cellular therapy for enhancing T-cell reconstitution before or after hematopoietic stem cell transplantation (ACES): a two-arm, open label phase II interventional trial of pediatric patients with risk factor assessment

Abstract

Viral infections remain a major risk in immunocompromised pediatric patients, and virus-specific T cell (VST) therapy has been successful for treatment of refractory viral infections in prior studies. We performed a phase II multicenter study (NCT03475212) for the treatment of pediatric patients with inborn errors of immunity and/or post allogeneic hematopoietic stem cell transplant with refractory viral infections using partially-HLA matched VSTs targeting cytomegalovirus, Epstein-Barr virus, or adenovirus. Primary endpoints were feasibility, safety, and clinical responses (>1 log reduction in viremia at 28 days). Secondary endpoints were reconstitution of antiviral immunity and persistence of the infused VSTs. Suitable VST products were identified for 75 of 77 clinical queries. Clinical responses were achieved in 29 of 47 (62%) of patients post-HSCT including 73% of patients evaluable at 1-month post-infusion, meeting the primary efficacy endpoint (>52%). Secondary graft rejection occurred in one child following VST infusion as described in a companion article. Corticosteroids, graft-versus-host disease, transplant-associated thrombotic microangiopathy, and eculizumab treatment correlated with poor response, while uptrending absolute lymphocyte and CD8 T cell counts correlated with good response. This study highlights key clinical factors that impact response to VSTs and demonstrates the feasibility and efficacy of this therapy in pediatric HSCT.

Fig. 1. Phenotype and antiviral activity of Virus-specific T cell products (n = 26).

A Cellular phenotype of VST products by flow cytometry. T cell memory sub-phenotyping (n = 6): Stem cell memory: CD45RO−/CCR7 + /CD95+; Central memory: CD45RO+/CCR7+; Effector memory: CD45RO+/CCR7−. B Antiviral responses of VSTs against CMV (pp65, IE1), EBV (EBNA1, LMP2), and adenovirus (Hexon, Penton) antigens by IFN-g ELISpot. SFC Spot forming colonies. Line: median values; whiskers: standard deviation.

Fig. 2. Longitudinal viral loads and antiviral medications post VST therapy.

A EBV viral loads (red) and Rituxan use in patient 41, who received VST therapy on day +147. B PET CT (positron emission tomography/computed tomography) imaging before and after VST therapy in Patient 41. C Adenoviral load (red) in Patient 15, who required extracorporeal membrane oxygenation (ECMO) for severe adenoviral disease after unconditioned matched sibling marrow infusion, and subsequent methylprednisolone (MP) course prior to de-cannulation. D–G Viral loads pre and post- VST therapy and antiviral medications over time in study patients 28, 38, 42, and 47 (FOS foscarnet, GAN ganciclovir, VALGAN valganciclovir, CID cidofovir).

Fig. 3. Overall survival and treatment schema in study patients.

A Overall survival in responders (CR complete responders, PR partial responders) versus non-responders (NR) following VST therapy (n = 51, survival curves were compared by log-rank test, p = 1.06 × 10⁻⁷). B Overall survival by patient acuity following VST infusion for Arm A (n = 47, p = 0.0017). High acuity was defined as infusion in the intensive care unit and/or respiratory failure, renal failure, veno-occlusive disease, or transplant-associated microangiopathy. Survival curves were compared by log-rank test. C Infusion schema and responses by number of VST doses and product details. Patients without data at day +28 post-infusion are listed as not evaluable (NE).

Fig. 4. Impact of clinical factors on antiviral responses post VST therapy.

Odds ratios (OR) in favor of antiviral responses (complete or partial response at day+28) based on a univariate logistic regression model are shown (n = 43). CI confidence interval, RIC reduced intensity conditioning, FK tacrolimus, CsA cyclosporin A, MMF mycophenolate mofetil, ECP extracorporeal photopheresis, VL viral load, VOD veno-occlusive disease, TMA transplant-associated thrombotic microangiopathy. Low grade GVHD: grades I-II; high grade GVHD: grade III-IV.

Fig. 5. Immune reconstitution post VST therapy.

A Trends in best response pre/post VST therapy against targeted viral antigens by IFN-γ Elispot were compared by 2-way ANOVA (CMV: n = 11; adenovirus: n = 15). SFC Spot forming colonies; *p = 0.045; **p = 0.006. B–F Virus-specific T cell reconstitution in recipients was evaluated by intracellular cytokine staining following a 7-day ex vivo expansion against targeted viral antigens. Longitudinal CD4⁺ and CD8+ responses against CMV (B, C, n = 10), adenovirus (D, E, n = 6), and EBV (F) were evaluated. Bar: mean; whiskers: standard deviation. G Peripheral frequencies of CMV-specific T cell receptor beta (TCR) clonotypes associated with the infused VST products were tracked relative to the first infusion in three recipients.