An international research program

What biomarkers for diagnosis and monitoring of treatment of GvHD?

The use of allogeneic hematopoietic cell transplantation (HCT) is the most potent immunotherapeutic treatment for hematologic diseases, but its practical use is impeded by acute and chronic graft-versus-host disease (GvHD). The diagnosis of GvHD relies almost entirely on clinical signs, and can only be confirmed by biopsy of target organs. Neither pre-transplant clinical characteristics nor transplant characteristics are reliably predictive of GvHD outcomes. Discovery and availability of biomarkers suitable for screening high risk patients using early, non-invasive, blood tests specific for GvHD-therapy resistance would significantly aid in the management of patients with this disease and would allow for rational drug development and drug approval by regulatory authorities. Using proteomics, it has been shown that plasma concentration of Suppression of Tumorigenicity 2 (ST2), accurately stratified patients according to risk of non-response to acute GvHD therapy in large retrospective sets. In recent experiments, we have also found that a four protein panel including ST2, CXCL9, MMP3 (matrix metaaloproteinase) and osteopontin is associated with the diagnosis of chronic GvHD. Based on our preliminary data, we propose to validate these biomarkers to predict response to aGvHD and cGvHD therapy and monitor response. The rationale for the proposed research is that when we are able to identify patients who are at particularly high risk for unresponsiveness to standard treatment and subsequent mortality, we can porpose customized treatment plans. We plan to use a well-established ELISA workflow biomarkers on aGvHD and cGvHD samples. ST2, Reg3α, TNFR1, IL-6 will be validated on 300 matched aGvHD patient samples pre-treatment and 1month post-treatment. The pre-treatment samples will allow assessing of their predictive values. We will then create an informative and clinically useful biomarker panel to the prediction of responsiveness to therapy using a statistical approach to integrate the information gathered for the biomarkers in this aim. Regarding cGvHD, the candidate biomarkers are ST2, CXCL9, MMP3 and OPN. We will validate this predictive panel on 300 matched cGvHD patients samples pre-treatment and 3 months post-treatment. The same statistical approach as for aGvHD will be used. We anticipate that these aims will yield the following expected outcomes : First, validating multilayer biomarker panel that predicts response to aGvHD and cGvHD therapy in a European multicenter cohort. Second, we will determine the threshold of the different biomarkers that will provide the best sensitivity for future use in clinical trials. These outcomes are expected to have a major impact, because the panels of biomarkers are likely to be applied in future GvHD therapeutic trials as surrogate marker for clinical response, and the biomarkers have potential as GvHD-specific therapeutic targets.

Project leaders
Dr Etienne Daguindau (Hematology Department, CHRU Besançon), and Pr Pierre‐Simon Rohrlich (Pediatric Hematology-Oncology Unit, Archet Hospital, CHU Nice)

Key figures
Project duration: 2 years
Number of patients involved: 227
Total number of samples: 500 (Plasma)
Budget: € 27,388

Are other atypical white blood cells an asset to address complications?

The mucosa-associated invariant T (MAIT) cells are innate-like T cells with restricted T cell receptor (TCR) usage, which are preferentially localized in mucosal tissues (liver, lung and gut) and respond to microbial infection by rapidly producing cytokines and cytotoxic effectors. They recognize the non-classical MHC-related molecule MR1, which binds an presents a novel class of antigens, namely precursor derivatives of vitamin B2 (riboflavin), which are found in most bacteria and yeasts. Upon bacterial stimulation, MAIT cells display immediate effector activity, namely production of pro-inflammatory cytokines (TNFα, IFNγ, IL-17) and cytotoxicity. In certain circumstances, MAIT cells are also able to produce IL-22, a cytokine involved in gut epithelial protection during inflammatory intestinal damage. Given the potential importance of MAIT cells in control of microbial infections and protection of epithelial surfaces, we investigated MAIT cell reconstitution in 43 children who underwent hematopoietic stem cell transplantation (HSCT) after myeloablative conditioning for the treatment of malignant hematological disease. MAIT cell numbers remained very low during the first 12 months after HSCT in recipients of matched-related or unrelated donor, and were almost undetectable in cord blood recipients. However, large interindividual variability was observed, suggesting that peritransplant factors might influence MAIT cell recovery. Furthermore, no export of naïve (thymic) MAIT cells was observed during the 12 months follow-up. We want to rapidly confirm and extend these preliminary results. The CRYOSTEM collection offers us a unique opportunity to obtain rapid results, in order to: 1/ confirm the kinetics of MAIT cell recovery and maturation over the first year after HSCT in matched-related donor recipients with myeloablative conditioning; 2/ determine if early MAIT cell recovery, and export of naïve MAIT cells from the thymus, are improved in matched-related donor recipients with reduced intensity conditioning compared to those with myeloablative conditioning; 3/ to determine if an early and efficient MAIT cell recovery is associated with clinical outcomes, in particular acute GVHD or bacterial infections. Given our strong expertise in MAIT cell analysis, the proposed research should be completed within 6 months, and should lead to a comprehensive assessment of the role of MAIT cells in HSCT recipients.

Project leaders
Pr Jean-Hugues DALLE (Pediatric Hematology-Oncology Unit, Robert Debré Hospital AP-HP) and Pr Sophie CAILLAT-ZUCMAN (Center of Research on Inflammation, Inserm, CNRS, Paris Diderot University)

Key figures
Project duration: 1 year
Number of patients involved: 55
Total number of samples: 468 (Cells in DMSO)
Budget: € 19,521

 A new class of immune cells to predict GvHD?

Recently, a non-T cell lymphoid population was identified as another source of cytokines, the Innate Lymphoid Cells. These cells represent new effectors in the establishment of the mucosal immune response. In close contact with the epithelial cells of respiratory and intestinal mucosa, they are first in line to respond quickly to any disruption of the environment. Our team contributed to the discovery of these new lymphocyte populations in humans. ILCs can be categorized as cytotoxic ILCs, represented by NK cells, and helper-like ILCs, represented by the ILC1, ILC2 nd ILC3 subsets. The ILC3 produce IL-17 and/or IL-22. They are important in the epithelial regeneration and mucosal immunity by their secretion of antimicrobial peptides. By restoring hematopoietis, hematopoiteic stem cells (HSC) transplantation is the reference treatment of severe, non-malignant and malignant blood diseases. For the latter, it allows the patient to benefit from the maximum anti-leukemic conditioning « dose effect » (HD chemotherapy and total body irradiation) and the anti-leukemic immunological effect of the graft on minor antigens (mHAgs) expressing- leukemic cells («Graft versus Leukemia » effect). But preparing regimens (conditioning) can cause serious harm to mucosal tissues by inducing apoptosis of rapidly dividing epithelial cells. Moreover, the immunological effect exposes the patient to graft against the mHAgs of host cells, leading to graft versus host disease (GvHD). GvHD and immune deficiency inherent to the kinetics of reconstitution are two major immunological complications of HSC transplantation, and represent the main cause of mortality and post-transplant morbidity. A recent report from Hanash et al. Showed that the type 3 cytokine IL-22 mediates a protective effect on immune-mediated tissue damage occuring after bone marrow trannsplant in the mice. The source of IL-22 was shown to be ILC. This prompted researchers to investigate ILC in patients receiving HSCT. It was recently shown the apparition of activated NCR+ ILC3, not present in healthy persons, after conditioning regimen and allogeneic HSCT in a very small cohort of 6 patients. Interestingly these cells also expressed homing receptors to skin and gut which are the main sites of GvHD. Their data suggest that ILC recovery affects the development of GvHD. We propose to determine whether the appearance of distinct ILC subsets in peripheral blood, particularly the activated NCR+ ILC3, and the cytokine secretion signature of these subsets can be associated with clinical events post-graft such as GvHD. At our knwoledge, it is the first study studying the contribution to GvHD of peripheral blood ILC subsets in a large cohort of adult and pediatric patients. Currently there is no way to efficiently predict the occurrence of GvHD after HSC transplantation. If it is confirmed that the appearance of an ILC subset post-conditioning can be associated with a reduced risk of GvHD, this will lead to modulate preventive treatment in patients with a high risk of GvHD. Indeed, these ILC subsets could be good candidates for autologous cell therapy approach after in vitro expansion.

Project leader
Dr Frédéric Vély (Immunology Department – Hôpital de la Conception, Marseille)

Key figures
Project duration: 1 year
Number of patients involved: 124
Total number of samples: 1 056 (Cells in DMSO)
Budget: € 33,556

The occurrence of EBV-associated PTLDs is generally preceded by an increase in viral reactivation- related viral EBV load and an increase in the number of infected B cells. Regular monitoring of the EBV viral load by PCR in the whole blood or plasma of allograft patients is therefore currently recommended to allow for pre-emptive therapy based on immunosupressive modulation and / Use of a monoclonal anti-CD20 antibody targeting B lymphocytes (Rituximab). This measure of viral load in the blood is not standardized and lacks both sensitivity and specificity, which leads to too late diagnoses of PTLD or to Unnecessary treatment. Other virological or immunological biomarkers are necessary to complete the measurement of the EBV blood viral load and to improve its predictive value. Recent results from our laboratory showed that this protein could be “excreted” and detected simply by an enzyme immunoassay in serum of transplant patients suffering from PTLD. For the first time we succeeded to detect the soluble ZEBRA (s-ZEBRA) protein in serum from transplant patients (measured by an antibody-based ELISA). The s-ZEBRA (>100 ng/mL) was predictive in 80% PTLD-patients within ten weeks, prior the PTLD diagnosis (p<0,0001) (Retrospective clinical study with 66 transplant patients. We hypothesize that early EBV replication may occur in the tumor or its environment, and eventually release of the ZEBRA in the bloodstream. At the end, this phenomenon could lead to the secretion of cytokines and factors promoting angiogenesis, B-cell proliferation, and thereby further aggravating the immunosuppressive environment.

The objective of this project is to confirm the clinical utility of this new biomarker (s-ZEBRA) as a surrogate of the EBV load measured by qPCR in a larger retrospective follow up study (HSC transplant patients). The ultimate goal will be to (i) predict not only malignant lymphoproliferation but also pejorative events in such patients (PTLD, GvHD, severe viral syndromes), (ii) define the first warning signal to reduce the dose of immunosuppressive drugs (the rituximab depletes B cells harbouring the latent EBV). Methods: 11 HSC transplant patients with PTLD and 40 GHVD patients were selected from the Cryostem database, in addition to 15 other HSC transplant patients from University Medical Center (Utrecht, NL). 45 solid organ transplant patients have been selected from other database in France (CIC Thorax, collection of CHU Strasbourg, DIVAT collection CHU Nantes) and will be investigated in parallel. From these collections, the serum samples will be tested for the soluble ZEBRA protein. Briefly an antigen-capture ELISA method was developed specifically for measuring sZEBRA in serum samples, using two specific monoclonal antibodies. A standard curve was obtained based on serial dilutions of r-ZEBRA, ranging from 1ng/mL to 250ng/mL. The results were expressed as concentrations of ZEBRA (ng/mL) extrapolated from the standard curve. In view of determining the detection specificity, three positive serum samples were submitted to a neutralization test.

Expected results: The test will be a new approach in monitoring EBV-related pathologies and adapting immunosuppressive therapy to transplantation. In addition, this new test can be used in the therapeutic follow-up of PTLD management, allowing individual evaluation of treatment efficacy, as well as in the development of new anticipated therapeutic approaches. Early detection will make it possible to adapt the doses of immunosuppressive drugs and / or to initiate an anti-CD20 treatment targeting the tumor cells.

Project leader
Pr Emmanuel Drouet (Joseph Fourier University  – Grenoble 1)

Key figures
Project duration: 18 months
Number of patients involved: 61
Total number of samples: 132 (Plasma)
Budget: € 6,845.67

Death following acute GVHD is the major cause of non-relapse mortality after allogeneic stem cell transplantation (alloSCT). Our previous reports provided evidence that this complication is associated with endothelial cell dysfunction that can either be demonstrated at onset of acute GVHD or even be predicted prior to conditioning therapy for alloSCT. Predictive endothelial associated markers included angiopoietin-2, serum nitrates, suppressor of tumorigenicity (ST)-2 and SNPs in the thrombomodulin gene. The close relation of transplant-associated microangiopathy (TAM) and death of acute GVHD led us to construct the Endothelial Activation and Stress Index (EASIX): ‘LDH(U/l) x creatinine (mg(dl) / thrombocytes (/nl)’. This simple formula was validated in 3 independent cohorts and predicts death after acute GVHD at disease onset and prior to alloSCT (manuscript submitted). However, TAM associates with only half of the deaths from acute high grade GVHD, and patients with low EASIX scores still have a considerable risk of dying from this condition. Our search for this alternative pathomechanism now led to the proposal of a dysfunctional CXCR3 axis as the potential cause of death. CXCR3 is a chemokine receptor expressed on activated T lymphocytes, in particular on Th1 cells, NK cells, dendritic cells, and subsets of epithelial and endothelial cells. Here we have studied if a dysfunctional CXCR3 axis might be involved in GVHD pathogenesis and could link endothelial and T cell pathology in acute GVHD.

We assessed concentrations of the CXCR ligands CXCL9, CXCL10 and CXCL11 as well as anti-CXCR3 autoantibodies in 119 patients with high grade (3-4) acute intestinal GVHD for whom serum was available at GVHD onset. Furthermore, anti-CXCR3 autoantibodies and CXCL9 levels were measured in sera stored before conditioning therapy. All variables were tested for influence on post-GVHD survival using cause-specific Cox regression analysis.

At GVHD onset, high serum levels of CXCL9/MIG and a high EASIX score were strongest predictors of NRM. In patients with low EASIX, high CXCL9/MIG and low anti-CXCR3 serum levels independently predicted outcome in multivariable analyses. A score based on CXCL9, anti-CXCR3, and EASIX allowed an effective prediction of acute GVHD outcome ranging from mortality >90% (high CXCL9 + high EASIX) to mortality <20% (low CXCL9, low EASIX, high anti-CXCR3.

Our data suggest a strong role for the CXCR3 axis in the pathology of acute high grade GVHD that complements the established pathomechanism of endothelial cell dysfunction. The opposing effects of CXCL9 and anti-CXCR3 indicate a functional, attenuating role for these auto-antibodies. The overall prognostic impact of the immune-modulating CXCR3 axis appears to depend on the underlying integrity of the patients’ endothelial homeostasis.

Project leader
Dr Thomas Luft  (Department of Internal Medicine V: Hematology, Oncology and Rheumatology – University Hospital Heidelberg)

Key figures
Project duration: 6 months
Number of patients involved: 136
Total number of samples: 136 (Plasma)
Budget: € NA

Unmanipulated haplo-identical transplantations combining Cyclophosphamide in high doses with the immunosuppressive agent, mycophenolate have now proved their efficacy. They appear to have similar results to geno-identical transplantation in terms of overall survival, NRM and acute GVH. However, although NRM is low, early viral infections have been reported by different teams, either with CMV reactivation or with polyomavirus infections. Recent data have suggested that the HLA-G and HLA-E molecules have both humoral and cellular anti-inflammatory and immunosuppressive properties. HLA-G and HLA-E polymorphisms have been shown associated with various chronic viral infections including human immunodeficiency virus, papillomavirus, cytomegalovirus (CMV) and hepatitis viruses. Several recent studies have suggested that HLA-E and HLA-G polymorphisms were correlated with the occurrence of HSCT patients. Our Preliminary data on cohort of 108 patients with haplo-identical transplantation showed that the polymorphism HLA-E*01:03 homozygous in donor or recipient were associated with occurrence to chronic GVH.

Thus, the objective of this study is to determine the clinical relevance of the HLA-G / HLA-E genetic status of the donor / recipient pair in haplo-identical transplantations without T-depletion with cyclophosphamide at J3-J4.

This is a 7-month retrospective study of adult patients receiving haplo-identical transplants whose clinical data and at least one dry pellet of the donor / recipient couples are available at the CRYOSTEM Biological Resource Center . In practice, this study will take place in the EFS HLA laboratory and the “Blood Group Biology” department led by Dr. Christophe Picard, according to the following points: i) To extract of the DNAs from the dry pellets and aliquot. ii/ To Perform HLA classical, HLA-G and HLA-E as well as the HLA-G regulatory regions typing in the donors / recipients by two NGS techniques detecting the polymorphisms on all genes, one developed within the team of “Blood Group Biology” and one commercial (NGS-Go, EFS) iii / To determine in silico HLA-G/-E alleles and haplotypes-G iv / To determine impact of HLA-G and/or HLA-E polymorphisms and haplotype-G in the donor and recipient and their compatibility on NRM, viral infection, overall survival, relapse-free survival, acute disease and chronic graft against the host.

This study may help to introduce additional criteria into the donor selection algorithm and also to propose a different therapeutic strategy according to the non-classical HLA status for patients who have received a haplo-identical transplant.

Project leader
Dr Christophe Picard  (UMR 7268 ADÈS Aix Marseille University / EFS/ CNRS Marseille)

Key figures
Project duration: 7 months
Number of patients involved: 330
Total number of samples: 660  (Dried white blood cell pellets)
Budget: € 116,164

Despite new drugs, allogeneic hematopoietic stem cells transplantation (HSCT) is frequently indicated in different hematologic disorders. The mean goal is to obtain allogeneic effect by donor immunologic system. If this effect is clearly demonstrated, some hematological malignancies appear less susceptible, in particular in lymphoid malignancies as non-Hodgkin’s large B-cell lymphomas (DLBCL) and Hodgkin lymphomas. The differences between these diseases are not clearly explained finally. Immune tolerance as described in the carcinogenesis mechanisms is potentially considered.

PD-1 (Programmed Death 1)(CD279) is an inhibitor receptor who attenuate the signal of the TCR(T-cell Receptor). With his ligand, PD-L1, they have a role to inhibit immune response and induce immune tolerance. Aberrant expression in hematologic disorders can inhibit T-cells in the tumor microenvironment and participate for the development of the tumor. This model is described in different pathologies as Hodgkin disease.

In otherwise, PD-1/PD-L1 could have a role for immunomodulation after allogeneic HSCT. It could be participated to inhibit Graft Versus Host Disease (GvHD) and anti-tumor effect, principal goal of HSCT. Currently, no data are described in vivo.

The aim of our study is to investigate the expression of PD-1 expression on the T-cells and plasma level of PD-L1 during allogeneic HSC transplants for the treatment of hematological malignancies and to determine whether there is an association between expression and prognosis after transplant HSC. Of the final goal is to define how use or not in some patients anti-PD1 therapy.

Project leader
Dr Jérome Cornillon  (Clinical hematology department, Institut de Cancérologie Lucien Neuwirth, University Hospital Saint Etienne)

Key figures
Project duration: 1 year
Number of patients involved: 60
Total number of samples: 540  (Viable cells in DMSO, blood plasma)
Budget: € 49,375