The curative principle behind allogeneic hematopoietic stem cell transplantation (HSCT) is eradication of the malignant cells of the patient (recipient) by donor graft cells, a process termed graft-versus-leukemia (GVL) effect. GVL is traditionally mediated by donor αβ T cells in an immunological process driven by genetical differences between individuals, i.e. an allogeneic response. For this reason, αβ T cells also cause an unwanted and dangerous complication of HSCT called graft-versus-host disease (GVHD) in which healthy recipient cells are targeted by donor cells with great risk of morbidity and mortality to the patient. In addition to αβ T cells, other cells from the donor stem cell graft, termed innate effector lymphocytes, can contribute to the GVL effect. These are termed natural killer (NK) cells and T-cell receptor (TCR) γδ cells, the latter being a subset of T cells. NK and TCR γδ cells can recognize and eliminate leukemic cells in a direct tumor response independent of conventional allogeneicity. Therefore, opposite αβ T cells, innate effector lymphocytes cells can mediate GVL but are not likely to cause GVHD. The main indications for HSCT in adults are acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS). Approximately 50% of AML/MDS transplant patients experience significant acute GVHD and 30% experience relapse of the malignant disease. Prospective clinical studies from the research group of the investigators have shown that patients with high doses of innate lymphocytes in stem cell grafts and during early immune reconstitution after HSCT have a reduced risk of both GVHD and relapse. The aim of this clinical trial is therefore to administer innate donor lymphocyte infusion (iDLI) enriched in NK and TCR γδ cells and depleted of αβ T cells in patients early after HSCT. By improving the HSCT procedure with iDLI cell therapy the scope is less GVHD and less relapse of the malignant disease and thereby improved survival and life quality in AML/MDS patients.

PROJECT DESCRIPTION PURPOSE: To address whether transplant outcomes in patients treated with HSCT for AML/MDS can be improved by addition of innate donor lymphocyte infusion (iDLI) early after transplantation. The primary outcomes are acute GVHD, relapse and relapse-free survival after HSCT.

BACKGROUND Allogeneic hematopoietic stem cell transplantation (HSCT) is a potential curative treatment for malignant hematologic diseases. The main indication for HSCT in adults is acute leukemia, followed by myelodysplastic syndrome. In HSCT, the bone marrow and immune system of the patient is replaced with that from a donor, during an extensive clinical procedure carried out in a highly specialized transplant center. The curable principle of HSCT is an immune-based cell-to-cell killing of residual leukemic cells mediated by donor cells termed the graft-versus-leukemia (GVL) effect.

The basic principles of HSCT are given below:

Conditioning regimen Prior to donor graft infusion, a conditioning regimen is given in order to reduce the tumor burden and weaken the recipient immune system to allow engraftment of donor cells. The conditioning regimen can contain different degrees of myeloablation depending on the desired grade of immune suppression suitable for the disease, patient age and comorbidity.

Stem cell donors and grafts The stem cell graft is most often obtained from an HLA-matched sibling (25% of patients) or an HLA-matched unrelated register donor.

The graft is obtained from the donor either by aspiration of bone marrow or by leukapheresis (cell separation and removal of leucocytes from the remaining blood) after administration of granulocyte-colony stimulating factor (G-CSF), which mobilizes stem cells to the blood. The latter is known as a peripheral blood stem cell (PBSC) graft and is by far the most common graft type used in current era.

Graft contents and patient immune reconstitution The stem cell graft contains stem cells that establish donor derived hematopoiesis including a functional immune system in the patient, but also mature immune cells responsible for early elimination of residual tumor cells through the graft-versus-leukemia (GVL) effect. Immune reconstitution is dependent on the dose and contents of the transplanted donor cells and both the graft contents and patient immune reconstitution is crucial for the overall outcome of the transplantation.

GVL and GVHD The GVL effect is carried out by lymphocytotoxic donor cells that kill residual malignant cells of the patient. GVL is traditionally mediated by donor αβ T cells in an immunological process driven by HLA-differences between individuals, i.e. an allogeneic response.

In acute graft versus-host-disease (GVHD), also involving HLA allo-recognition, donor αβ T cells attack the normal tissue in skin, liver and gastro-intestinal system of the recipient, leading to a potential deadly immune disease. Acute GVHD occurs during the first 3 months after HSCT, and the clinical course is highly variable, from a mild curable presentation to a full-blown immune attack, which is fatal in 80-90% of the patients (fig 2). Acute GVHD is seen in 40-60% of the patients.

Relapse after HSCT Relapse of the hematologic cancer after HSCT is a major cause of mortality. In patients with AML/MDS approximately 30% experience disease recurrence after HSCT. The prognosis in patients relapsed after HSCT is dismal.

Donor lymphocyte infusion In order to prevent relapse or treat manifest relapse after HSCT a dose of additional cells from the original stem cell donors, donor lymphocyte infusion (DLI), can be infused into the patient with the purpose of initiating a renewed/boosted GVL effect and eradicate remaining malignant cells. However, the effect on the leukemic cells of conventional DLI, containing primarily TCR αβ cells from the donor, is uncertain and limited by the risk of lethal GVHD why new methods of DLI are constantly under development.

Natural killer cells and T-cell receptor γδ cells During recent years natural killer (NK) cells and T-cell receptor (TCR) γδ cells, also termed innate effector cells, have proven capable of eliminating cancer cells and thereby contributing to the GVL-effect in HSCT. NK cells are innate lymphocytes, comprising approximately 10% of all lymphocytes, and purposed to eradicate "stressed", malignant and virus-infected cells. They are able to lyse and eliminate tumor cells of leukemia through activating receptors of the natural killer (NK) and natural cytotoxicity receptor (NCR) families independent of HLA disparities.

T cells of the TCR αβ type (CD4 and CD8 T cells) comprise the majority of all circulating T cells and lymphocytes in general. TCR γδ cells are a subgroup of T cells, different from αβ T cells, which constitute approximately 5% of all T cells. The repertoire of TCR γδ cell functions combine features of the adoptive immune system with innate-like responses. They are involved in control of inflammation, fighting of pathogens and maintaining tolerance towards self-antigens. TCR γδ are also capable of recognizing tumor cells through their TCR and activating receptors shared with NK cells (e.g. NKG2D) and their anti-tumor effects are being increasingly documented against leukemic cells. NK and TCR γδ cell recognition is independent of HLA-disparities, making their contribution to GVHD less likely, and their immunoregulatory features may actually protect against GVHD. These innate effector cells therefore have the potential to mediate GVL without GVHD in the HSCT setting.

Results from the research group of the investigators:

Under this hypothesis, our group conducted a prospective clinical study throughout 2015-2018 investigating concentrations of NK and TCR γδ cells in stem cell grafts and during immune reconstitution early after transplantation and associated these with patient outcomes in terms of GVHD, relapse and relapse-free survival. Results showed that patients transplanted with high doses of NK cells in the graft and high concentrations of NK cells in peripheral blood early after transplantation (day 28 and 56) had significantly improved relapse-free survival compared to patients with low transplant doses and early post-transplant concentrations of NK cells. Regarding TCR γδ cells, patients with high concentration of TCR γδ cells in peripheral blood early (day 28 and 56) after transplantation has significantly less GVHD and less relapse resulting in improved overall survival and relapse-free survival compared to patients with low post-transplant TCR γδ cell concentrations.

Based on these results, and existing literature, the aim of this study is infusion of an innate DLI (iDLI) containing purified NK and TCR γδ cell from the original stem cell donor early after HSCT. The expected effect of additional cell therapy with these innate effector cells during early immune reconstitution are reduced relapse as well as GVHD rates and thereby improved outcomes after HSCT in AML/MDS patients.

Procedure This is a collaborative project between the Bone Marrow Transplant Unit at the Department of Hematology and the Department of Clinical Immunology at Rigshospitalet, Copenhagen University Hospital.

Approximately 60 patients per year receive PBSC-HSCT for the diagnoses of AML or MDS at our center. One out of four patients has an HLA-identical sibling while the remaining patients receive stem cell grafts from matched unrelated donors from international donor registries. For this study, genetic randomization will be applied so that patients with HLA-identical siblings will be included for intervention while patients with unrelated donors receiving a standard-HSCT will be included in the control group. The background for the design is donor availability; only sibling donors can be recruited to donate iDLI. With an estimated recruiting rate of approximately 85%, 13-15 patient/donor pairs in the intervention group are expected to be enrolled per year over a total of three years, giving at total of 39-45 pairs. In a previous prospective study, the investigators obtained a recruiting rate of 97%. Historically, a higher incidence of GVHD was observed in transplants using an unrelated donor, but due to more precise HLA-typing and better protocols for immune suppression, GVHD and relapse incidence are nowadays comparable between related and unrelated transplants. This is also the case at the transplant center at Rigshospitalet, as previously published6.

For intervention:

Donors will undergo health screening at the Cell Therapy Facility, Department of Clinical Immunology, within 30 days before donation (standard procedure). Day 7 before HSCT: Start of the conditioning regimen preparing the patient for HSCT (standard procedure) Day 5 before HSCT: Start of G-CSF stimulation of the donor in order to mobilize stem cells for leukapheresis (standard procedure).

Donor leukapheresis and graft infusion to the patients, termed day 0 of HSCT (standard procedure) Day 14 after HSCT: donor leukapheresis (without preceding G-CSF stimulation), product manipulation and transfusion of iDLI to the patient (intervention).

Control group: Standard HSCT-procedure as described above in patients receiving grafts from unrelated donors.

Procedure of modified DLI = iDLI:

White blood cells are obtained from the donor by leukapheresis using the Spectra Optia Apheresis. System (Terumo BCT) at the Blood Bank Unit, Department of Clinical Immunology, Rigshospitalet. Purification of innate effector cells is carried out by TCRαβ/CD19-depletion of the fresh leukapheresis product with the CliniMACS Prodigy System (Miltenyi Biotec). This procedure removes TCRαβ T cells together with B-cells (CD19positive cells) rendering a cell product enriched in NK cells and TCR γδ cells and is routinely used at Rigshospitalet for procedures involved in pediatric transplantations.

The intended dose of harvested cells is 108 lymphocytes/kg resulting in expected cell doses of 107 NK cells/kg and 1-5 x106 TCR γδ cells/kg after the depletion. The iDLI product is freshly infused into the patient at the Department of Hematology shortly after the depletion procedure.

Immune phenotyping:

Immune phenotyping of the iDLI product as well as the original stem cell graft will be performed using a highly standardized customer designed freeze dried flowcytometry panel (DuraClone product from Beckman Coulter) developed and routinely used for immune function monitoring at the Department of Clinical Immunology. Two tubes will be included for detailed lymphocyte phenotyping in this study.

Tube 1 includes markers of overall T cells (CD3, CD4, CD8), NK cell subtypes (CD16, CD56), B cells (CD19) and monocytes (CD14). Tube 2 includes markers of overall TCR α/β cells, overall TCR γ/δ cells, the two main TCR γ/δ subtypes (TCR Vδ1, TCR Vδ2) and TCR γ/δ cell activation markers (HLA-DR, CD8, CD8beta).

The same panel will be used for characterization of immune reconstitution in peripheral blood samples in patients at the day of iDLI infusion and 14 and 21 days hereafter (i.e. day 14, 28 and 56 after transplantation. Viable cells will be cryopreserved for later functional studies.

Data management plan Clinical data on patient and transplant outcomes will be acquired from the clinical transplant database (IBMTR 2019) situated and administrated at the Transplant Unit, Department of Hematology. Flowcytometry data from characterization of iDLI products, stem cell grafts and patient immune reconstitution will be acquired from the Department of Clinical Immunology and processed by appropriate analysis software (Kaluza, Flow Cytometry Analysis Software, Beckman Coulter). Clinical and laboratory data will be assembled, structured and processed in SPSS (SPSS® Statistics, IBM) witch together with R (The R Project for Statistical Computing, www.r-project.org) will be used for statistical analyses. All data will be stored at Rigshospitalet, Copenhagen University Hospital in accordance with data protection rules of The Danish Data Protection Agency.

Statistical Analyses The cumulative incidence of acute GVHD and relapse in patients treated with iDLI compared to patients receiving a standard HSCT will be compared used Gray's competing risk analysis. Relapse-free survival between the two groups will be compared using the Kaplan Meier survival analysis and Cox proportional hazard models. Cell doses and immune phenotyping of iDLI and original stem cells grafts and cell concentrations and phenotyping during patient immune reconstitution will be entered using descriptive statistics. SPSS (SPSS® Statistics, IBM) and R (The R Project for Statistical Computing, www.r-project.org) will be applied for statistical computing.

Power Based on results from previous study from the investigators research group describing GVHD, relapse and relapse-free survival in patients with high versus low doses of NK/TCR γδ cells in transplant grafts and during early immune reconstitution. The investigators anticipate a reduction in relapse incidence from 30% to 5% and a reduction in acute GVHD incidence from 50% to 20%. With genetic randomization, patients with HLA-identical siblings will be included for intervention while patients with unrelated donors receiving a standard-HSCT will be included as controls. Tested at a significance level of 0.05 with a power of 80% (beta 0.2) a sample size of 35-38 patients in each group is appropriate for the study.

Inclusion Criteria:

• Diagnoses: AML, MDS
• Age: ≥18 years
• Graft type: PBSC
• Donor: ≥18 years
• Informed consent from both donor and recipient

Exclusion Criteria:

• Donors with need for central venous access for the leukapheresis procedure

Open label.