Phase I Study of CTL Anti-DP Infusion Post-hematopoietic Stem Cell Transplantation

Rationale Despite graft-versus-tumor effect, relapse remains one of the main causes of morbidity and mortality in allo-HSCT recipients. Forty to 50% of deaths following allo-HSCT are due to disease relapse. In case of relapse, the prognosis is very poor and disease burden remains a challenge for the use of adoptive cellular therapy alone. The 3 year overall survival (OS) in case of post-transplant relapse is dismal. The post-transplant period is characterized by a prolonged phase of immunodeficiency leading to increased vulnerability to infections and risk of relapse. For this reason, maintenance or pre-emptive therapies for patients in CR (complete remission) are now considered to prevent future relapse. Following allo-HSCT, the recognition by donor T lymphocytes of recipient HLA antigens may result in different consequences. On one side, T cells may recognize antigens present on malignant cells and eradicate residual disease or prevent tumor relapse. On the other side, injection of unselected T cells may induce a graft-versus-host-disease (GVHD).

HLA-DPB1 is one of MHC class II molecule lying centromeric to other class II loci on chromosome 6p21.3. Increase recombination events are found in the region between the HLA-DP loci and other class II loci, explaining the relative lack of linkage disequilibrium (LD) between HLA-DP (* HLA- DP: Human Leucocytes Antigen (DP allele))and the rest of MHC haplotype. For this reason, it is difficult to find a donor matched for DPB1 in addition to other classic HLA molecules. In sibling donors, the rate of incompatibility has been estimated to be as high as 10.9% and in unrelated donors a mismatch rate can be up to 89%. HLA-DPB1 is often not taken into consideration in donor selection. However HLA-DPB1 matching status has an impact on GVL and GVHD. In recipients of HSCT, a match for DPB1 is associated with a significantly increased risk of disease relapse, irrespective of the matching status of other HLA molecules. HLA class II molecules expression is mainly restricted to hematopoietic cells. Therefore, one could anticipate that a mismatched of HLA class II could induce a selective GVL reactivity without GVHD. However, HLA class II expression can be upregulated on various tissues following exposure to pro-inflammatory cytokines with a risk of GVHD as it is the case following some conditioning regimens or infections.

The frequency of the different HLA-DPB1 alleles in the general population is well known: HLA-DPB1*04:01 is the most frequent (70.5%) HLA-DPB1*02:01 and and HLA-DPB1*03:01 represent 32% and 20% respectively. 96% of leukemic cells could potentially be targeted with only three CTL clones directed against HLA-DPB1*04:01, 03:01 and 02:01.

HLA-DP-expressing B cell and myeloid malignancies can be recognized and lysed by HLA-DP-specific CD4+ cells ( CD4+ : cluster of differentiation 4+). The majority of leukemic cells (AML, ALL, CLL) express HLA-DP. CD4+ cytotoxic T cell (CTL) clones recognizing specifically HLA-DPB1*04:01 can be identified and have been demonstrated to be able to kill HLA-DPB1*04:01 positive leukemic cells.

In addition, it has already been shown that HLA-DP-specific CD4+ T cells can induce graft-versus-leukemia reactivity in the presence or absence of graft-versus-host disease. In this study the presence of HLA-DP-specific CD4+ T cells correlated with the clinical response to DLI.

The team of the Inserm unit 1232 (H. Vié, B. Clemenceau, both co-investigators of this project) has developed a suicide gene-transduced CD4+ T cell clone that recognizes the HLA-DPB1*401, which is the most frequent HLA-DPB1 allele expressed by leukemic blasts (70%). This clone has been described in detailed in :"The doubling potential of T lymphocytes allows clinical-grade production of a bank of genetically modified monoclonal T-cell populations" by Vivien R et al. Cytotherapy. 2018 Mar;20(3):436-452.

Several clinical trials have evaluated the possibility to inject, following allo-HSCT, donor lymphocytes transduced with a suicide gene either to treat tumor relapse or to accelerate the immune reconstitution. No acute infusion-related toxicity has been reported. Ganciclovir was used in some patients to control GVHD, leading to a rapid elimination of TK+ cells (tyrosine kinase +).

The clone was obtained by performing a mixed lymphocytes culture (MLR) between two populations differing only by HLA-DP, and subsequently by transducing reactive T cells with high efficiency, and finally cloning them directly, before selecting each clone for the desired characteristics. Thanks to a clinical grade Herpes-simplex-virus-TK vector, the clone harbors a suicide gene and can be killed in presence of ganciclovir (GCV).

This clone presents several important characteristics in terms of efficiency and safety. The clone is stable following thawing. It can be grown and amplified in vitro following thawing (at least more than one million times), while maintaining its cytotoxic capacity. It produces TH1-type cytokines in large amounts.

Regarding the safety of the CTL antiDP under study, we emphasized on two major points: Specificity and sensitivity to ganciclovir:

Specificity. The clone is specific for HLA-DPB1*04:01. The clone was selected against a donor homozygous for HLA-DPB1*04:01 (and identical for HLA-A, B, C, DQ, DR). Specificity testing confirmed the recognition of HLA-DPB1*:04:01. Yet, HLA-DPB1 alleles described are numerous (447 proteins to date IMGT (ImMunoGeneTics database)/HLA release, www.ebi.ac.uk/ipd/index.html), and it is just impossible to anticipate the cross reactions exhaustively (beyond some against HLA-DPB1*04:02 and HLA-DPB1*05:01 that have been observed against particular cell lines). The risk, since we are in a context of allo-HSCT, would be the recognition of donor cells and thus a possibility of graft rejection. For this reason, we will perform a pre-inclusion testing where donor cells will be used as targets for the clone.

Sensitivity to ganciclovir (GCV) : The proliferation tests in the presence of GCV confirms the efficacy of GCV with sufficient margin according to GCV blood levels reached during a conventional treatment.

Data from our colleagues show a low number of T cells detectable during the first months post-transplant, with a relative increased of Treg cells meaning that the clone may not be eliminated rapidly by donor T-cells.

The reasons to administer a suicide gene-transduced CD4+ T cell clone recognizing HLA-DPB1*04:01 following allogeneic transplantation can be summarized in the following points :

1. Relapse of the hematological malignancy remains a serious concern in these types of transplantation.
2. The immune reconstitution is very delayed allowing for injection of a third party T cell clone.
3. In case of GVHD following the CTL infusion, the CTL clone can be rapidly eliminated using ganciclovir.

Hypothesis We hypothesize that the infusion of a third party suicide gene-transduced T cell clone directed against HLA-DPB1*04:01 following allogeneic transplantation can be safe and might protect against possible relapse of hematological malignancies

Detailed description of the methodology (number of necessary subjects) Patients candidate for allogeneic transplantation who are both HLA-DPB1*04:01 and with aHLA-DPB1*04:01-expressing hematological malignancy (almost 100% of cases) with a donor HLA-DPB1*04:01 negative, will be proposed to receive one single infusion of the T cell clone at 4-5 months post-transplantation, once the immunosuppression by cyclosporine and/or mycophenolate mofetil has been discontinued. The expression of HLA-DPB1* by tumor cells will be checked by cytometry or immunohistochemistry. Any possible cross-reactivity of the clone against donor cells will also be excluded

A standard phase 1 dose-escalation study will be used:

Level 1: 1 x 104 cells/kg of recipient, Level 2: 5 x 104 cells/kg, Level 3: 25 x 104 cells/kg, Level 4: 50 x 104 cells/kg, Level 5: 100 x 104 cells/kg. After study of toxicity of the 4 injected patients, choice has been made to stop the CRM method to choose the dose and to test only 2 doses for each last patients : 100 x 10^4 cells/kg (actual level 5) and 500 x 10^4 cells/kg (choice based on compassionnal injections data at the level 5). All the more, if a DLT was observed at a dose of 100.10^4 cells/kg (see definition in paragraph 2.1.2), and after advice from the DSMB, a 2nd patient would be included and injected at the same dose, to confirm that the MTD had indeed been reached.

• Should MTD be confirmed in this 2nd patient, the study would be terminated.
• In the absence of MTD in this 2nd patient, and on the advice of the DSMB, the study would continue at the last dose of 500.10^4 cells/kg, and then be terminated (cell line exhausted).

Conditioning regimens : no restriction

The use of DLI in case of mixed chimerism or relapse is permitted after the clone infusion if necessary. In case of acute GVHD post CTL clone infusion, ganciclovir will be administered (at the dose of 5 mg/kg twice daily) for 14 days.