Reproducible Transient, Significant, Early Donor-Derived T-Cell Expansion, Induced By Pooled Granulocyte Transfusion, during Mismatch T-Cell Replete Cord Blood Transplant - an Optimal Platform for Allogeneic Graft Versus Leukaemia: Results of a Prospective Phase I/II Clinical Trial

INTRODUCTION

T-replete cord blood transplant (TCBT) reduces relapse and improve survival in difficult-to-cure AML. This is likely mediated by CB T-cell Graft versus Leukaemia (GvL). We have shown that third party pooled granulocyte transfusion during TCBT induce a transient, donor-derived T-cell expansion, associated with a cytokine release syndrome (CRS) and propose that such an expansion might augment GVL. We report our experience of 24 patients treated prospectively with granulocyte administration during TCBT.

METHODS

20 patients were recruited to the GRANS trial, a prospective, single-centre, phase I/II clinical trial (registration: NCT05425043) and 4 were treated off trial using the same protocol. Patients were referred through a national leukaemia MDT. Patients had either refractory disease, early post-transplant relapse, or both. Patients received T-replete, HLA-mismatched (5-7/8) CBT and seven doses of granulocytes (10ml/kg up to 200ml per dose), starting from day -1. Immune suppression was with MMF and ciclosporin, but these were withdrawn early in the absence of GVHD, stopping from 14 and 28 days respectively. Comparison was made with a cohort of AML patients who received TCBT without granulocyte infusion at our centre between 2015 to 2023. Statistical analysis was performed in R version 4.4.1.

RESULTS

24 patients received granulocytes on the trial protocol. All 24 patients had relapsed and/or refractory disease with 20/24 (83.3%) relapsed after a previous transplant, compared to 3/28 in the control group. In the study group 14/24 were MRD positive at transplant compared to 15/28 in the control group. 17/24 (70.8%) received MAC conditioning, usually with fludarabine, treosulfan and thiotepa. All received a single CB unit which was HLA-mismatched using allele-level typing in all patients (7/8 HLA-matching in 5/24 patients, 6/8 in 13/24 and 5/8 in 6/24). Granulocyte infusions were commenced between day -1 and day +1 (median: day -1) of transplant.

1. Engraftment, T cell expansion and CRS

Granulocytes were well tolerated and 23/24 patients in the study group received all 7 doses.

Within the study group, graft failure was noted in 3/24 (12.5%) vs. 0/28 (0%) in the control group (p = 0.09). For those who engrafted, time to neutrophil engraftment (the first of three consecutive days with ANC >0.5 x10⁹/L) was comparable between groups.

All patients that engrafted experienced CRS with fever, rash and elevated CRP. No patients needed inotropes, and no patients were admitted to PICU. The median peak CRP within the first 28 days post-HSCT was 203 and was 139 in the control group (p <0.01).

All patients that did not reject experienced significant, transient early T-cell expansion. The median peak absolute lymphocyte count (ALC) within the first 28 days post-HSCT was 1.19 [range: 0.13 to 4.00] vs. 0.59 [range: 0.10 to 2.36] in the control group (p<0.01). The median day post-HSCT of peak ALC was day +9 in the study group and day +26 in the control (p<0.001). There was a phenotype switch from naïve to effector memory/TEMRA cells, and there was both CD4 and CD8 expansion.

2. Acute and Chronic GvHD

Grade 3/4 acute GvHD was not increased, present in 5/24 (20.8%) vs. 5/28 (17.9%) in the control group (N.S.). Chronic GvHD amongst those alive at day +90 was present in 2/21 (9.5%) vs. 4/24 (16.7%) in the control group (N.). Chronic GvHD within the study group was noted to be mild and limited with both requiring only topical treatments.

3. Disease response

22 of 24 had T-cell expansion induced by granulocyte infusion during transplant (including one of the 3 patients that rejected their first transplant, during a second TCBT procedure).

21 of these 22 became MRD negative after transplant. One patient was always MRD positive, and 2 others that rejected also died of disease. 11 patients remain in remission with a median follow up of 200 days, and 9 patients have relapsed with a median time to relapse of 183 days (range 55 to 832 days). One patient experienced TRM in CR.

CONCLUSIONS

We optimise GVL conditions with mismatched T-cell replete transplant with T-cell expansion and early cessation of IS. The protocol is well tolerated with TRM less than 5%, no admissions to PICU with CRS and GVHD rates not increased compared to a control group. In a very high-risk cohort of AML, most patients enter remission, and this is sustained in many. Continuing work to understand the GVL response will allow further optimisation of this protocol.