Graft versus Leukaemia (GvL): Identification & characterisation of GVL antigens and cognate T cell responses in Acute Myeloid Leukaemia

The most aggressive blood cancers often cannot be cured by chemotherapy alone. For these cancers a curative treatment called allogenic stem cell transplantation or allo-SCT is used. Allo-SCT involves transfer of blood stem and immune cells from a healthy person (donor) to a patient (recipient). Allo-SCT is well-established and in routine clinical practice since the 1960s. ~20000 allo-SCTs are performed worldwide annually. The most common disease treated by allo-SCT is Acute Myeloid Leukaemia (AML), an aggressive adult blood cancer.

How does allo-SCT cure patients? Though donor cells are matched closely to patient cells to avoid the donor cells being rejected by the patient, donor cells are not identical to patient cells. Therefore, the donor cells can "see" the patient's cells as foreign and attack them. This has both beneficial and adverse consequences. The benefit arises when donor immune cells attack and eradicate the patient's blood cancer. This is called graft versus leukaemia (GvL). However, donor immune cells can also attack and damage the patient's normal healthy tissue. This is known as graft versus host disease (GvHD). The principal immune cells mediating GvL and GvHD are called T cells.

Though allo-SCT is curative it has serious side effects. Given this, allo-SCT is usually restricted to patients under the age of 70 years of age; but even then 10-40% of patients die of procedure-related mortality, often related to GvHD or infection. Furthermore, ~20-40% of patients still die of disease relapse. Thus, only 30-70% of patients are cured after allo-SCT. The reason why allo-SCT is performed is because without it most patients will die within 6 months.

So how can we improve allo-SCT outcomes? One of the most important gaps in our knowledge is that we do not know the identity of the proteins on a patient's cells that trigger donor immune cells to attack them resulting in GvL and/or GvHD. Proteins that recognise and activate immune cells are called antigens. The central aim of this project is to identify the antigens that trigger GvL and GvHD in an AML patient who has received an allo-SCT.

The laboratories I will be working in have developed a novel, innovative approach to identify antigens that trigger GvL and GvHD, in an unbiased manner, using samples from AML patients who have had a successful allo-SCT, and their donors. In two patient-donor pairs they have shown these antigens on the patient's cells are recognised by donor T cells. They have also identified the proteins on the T cells (T cell receptors) that recognise these antigens. Finally, when they introduced these antigen-reactive T cell receptors into T cells from a different donor, they could make the new donor T cells respond to antigen. In this way the laboratories have developed a complete workflow to identify antigens that may trigger GvL and GvHD and the T cells populations that mediate GvL and GvHD.

What will I do in this project? I will now test this workflow on a larger group of 13 AML patient-donor pairs. I will identify all the antigens that could be recognised by donor T cells, that are present on patient cells. I will then test if these antigens are clinically important - could prevent disease relapse (i.e. may be triggering GvL) and/or could be causing GvHD by testing a very large cohort of patient-donor pairs. I will then isolate the T cell receptors recognising 3 of the most clinically important antigens that may be triggering GvL. Finally, I will test if they recognise the antigen and if that leads to killing of AML cells.

This project could lead to more precise matching of donors for patients - donors that could provide a curative GvL response whilst avoiding/minimising GvHD. It could also allow detailed tracking of T cells post allo-SCT that cause GvL and GvHD. Finally, it could result in new T cell therapies that just promote GvL. The benefits of this work are set out more completely in the Academic Beneficiaries section.

Allo-SCT the most established, routine, curative cellular immunotherapy is commonly used to treat AML. Alloreactive donor T cells recognise and eradicate the patient's cancer cells (graft versus leukaemia, GvL), but can also harm normal tissue (graft versus host disease, GvHD). Remarkably, the antigenic basis of GvL and GvHD is mainly unknown, mitigating a fundamental understanding of the immunology of GvL and GvHD.

The laboratories I am joining developed a novel, innovative integrated workflow to study a unique set of 15 AML patient-donor bone marrow and blood samples collected over 15 years. They combined whole exome and RNA sequencing to identify expressed coding variants mismatched between each patient and their donor. They used HLA-binding prediction algorithms to identify putative HLA-binding peptides with mismatched variants. In 2 patient-donor pairs they showed specific peptides, with mismatched variants, elicit T cell responses from patients' blood, post allo-SCT, using a sensitive cultured IFNg ELISPOT method. They determined the HLA restriction of these responses. They then isolated T cell receptors and clonal T cell populations that recognise these alloantigens.

I now will implement this workflow on the remaining 13 patient-donor pairs. I will also define the tissue-specific pattern of alloantigen expression. I will test which antigens are clinically relevant in two large cohorts of patient and donors. Finally, for 3 clinically relevant antigens I will test if the cognate TCRs recognise and kill primary AML cells that express the allo-antigen and have the appropriate HLA restriction, when introduced into primary human T cells.

This work has the potential to provide fundamental insight into GvL and GvHD, allow selection of donors that promote GvL but not GvHD, and light a path to T cell therapies that specifically recognise and kill AML cells.