Phase I/II Clinical Trials of Donor-Derived Purified Regulatory T Cells for the Treatment of Steroid-Refractory Chronic Graft Versus Host Disease

Introduction

Chronic Graft versus Host Disease (cGvHD) remains a major complication of hematopoietic stem cells transplantation (HSCT) causing significant morbidity and mortality. We performed Phase I/II clinical trials to access the feasibility, safety and efficacy of infusing donor regulatory T cells (Treg) for the treatment of steroid-refractory/dependent cGvHD under the auspices of the EC-funded consortium TREGeneration.This is the first presentation of the Lisbon (iMM) and Seville (SAS) cohorts.

Methods

Treg were purified using CliniMACS (Miltenyi Biotec^®) from fresh leukapharesis from the original HSCT donor by CD8 and CD20 depletion, followed by CD25bright positive selection. The dose escalation protocol comprised sequential groups of 3 (SAS) or 5 (iMM) patients per center receiving 0.5x10⁶ (Dose A), 1x10⁶ (Dose B) or 2-3x10⁶ (Dose C) donor Treg/kg. After Phase I, 5 patients per center were included at the MTD, as a preliminary Phase II trial. We report feasibility and toxicity results on 33 patients (n=19, iMM; n=14, SAS) and efficacy on 31 patients (n=19, iMM; n=12, SAS) evaluable at month 12 post infusion (n=8, moderate; n=23, severe).

Patients were classified according to the NIH 2014 cGVHD criteria (Jagasia, MH et al, BBMT, 2015). Scores for the skin, eye, modified oral mucosa, GI tract, liver, lung, joint/fascia and cGvHD severity were used to calculate clinical responses at 3, 6 and 12 months, whereupon the final response was recorded. cGVHD treatment at iMM comprised primarily steroids and MMF, while all patients at SAS were on a ruxolitinib-containing regimen and had not obtained at least a partial response. Blood was collected for lymphocyte subsets monitoring by flow cytometry, evaluation of tissue damage biomarkers, homeostatic, pro-inflammatory and suppressive cytokines (sIL-2Ra, CXCL9/MIG, MMP3, Osteopontin, CXCL10/IP-10, CSTB, sCD13, Elafin, sTM, ST2, BAFF, IL-1R antagonist, IL-2, IL-7, IL-10, IFN-γ, TNF-α, CXCL11/I-TAC, IL-6, IL-17A/CTLA8 and TGF-β) by Multiplex, and tracking of infused Treg clonotypes by deep-level TCR RNA-based Next Generation Sequencing (NGS).

Univariate tests were carried out using Fisher’s exact test for categorical variables. t-test or Mann-Whitney tests were used for continuous variables as per normality analysis. Since no multiple testing correction was carried out, there were no corrections for multiple comparison or adjustments for other covariates.

Results and Conclusions

Donor Treg products preparation was feasible. Nevertheless, due to the paucity of circulating Treg, dose C required 2 separate leucapheresis in most patients. At SAS, achieving more than 1.5x10⁶ donor Treg/kg was possible for only 4 patients. At iMM, the QC criteria were met in 20/20 products (mean Treg purity 67%; CD8 depletion 4.5 and 4 times for CD20). At SAS, 1 product did not meet the purity criteria and was not infused, the remaining 14 products were infused (mean purity 73%; CD8 log depletion of 4.9 and of 6.5 times for CD20). Infused cell doses were calculated based on Treg purity (CD4+CD25++CD127lowFoxP3+).

There were no dose limiting toxicities and dose C was used whenever reached in the Phase II trial. Efficacy analysis revealed that 22 out of 31 evaluable patients (70.9%) showed either complete response (CR) or partial response (PR) at month 12, whereas 29% progressed (P) or did not respond (NR). The median time from infusion to response was 6 months. The same efficacy trend was observed irrespective of a previous failure to ruxolitinib.

Time from cGvHD diagnosis to Treg infusion was significantly associated with response in the iMM cohort (p=0.003).

Interestingly, the Treg dose influenced outcomes. Hence, when patients receiving A+B doses (<1 x10⁶Treg/Kg) were compared to patients receiving >1.3x10⁶ Treg/Kg, 10/18 (56%) of patients in lower doses responded versus 12/13 (92%) in higher Treg doses (p=0.045). Furthermore, 39% of patients had the daily dose of prednisolone tapered by at least 50%.

Exploratory TCR tracking analysis revealed the persistence of infused Treg clones in 18 of the 20 patients analyzed thus far up to one year after infusion.

In summary, we report the feasibility and safety of donor Treg infusion in patients with moderate/severe cGVHD. Meaningful responses were observed in over 2/3 of the patients, further suggesting that early treatment with Treg is associated to improved clinical outcomes. Our results set the stage for larger Phase II trial.