Immune Pathophysiology of Aplastic Anemia and Immunosuppressive Treatments

National Heart, Lung, And Blood Institute

Investigators

Linked publications & trials

Abstract

In aplastic anemia, the bone marrow is replaced by fat, and peripheral blood fall to extremely low levels, leading to death from anemia, bleeding or infection. Aplastic anemia is a disease of young persons and when severe is almost invariably fatal untreated. Aplastic anemia has been linked to chemical exposures, in particular benzene; it is an idiosyncratic complication of some medical drug use; it occurs as a rare event in pregnancy and following seronegative hepatitis; and the diseases associated with certain immunologic conditions. The chance observation that some patients post-bone marrow transplant recovered their own marrow function led to the inference that the immunosuppressive conditioning regimen might have treated an underlying immune-mediated pathophysiology. Purposeful administration of anti-thymocyte globulin (ATG) has led to hematologic recovery in the majority of treated patients. Laboratory data have also revealed abnormalities of the immune system: lymphocyte populations that induce apoptosis in hematopoietic target cells by the Fas-mediated pathway, and oligoclones of effector T cells which express type 1 cytokines, especially gamma-interferon. More recently, eltrombopag, a thrombopoietin mimetic, was shown to be effective in improving clinical outcomes; clinical data from our Branch was the basis for approval by the FDA for use of this drug in both refractory and treatment-nave severe aplastic anemia. Our section in the Hematology Branch has been a leader in both scientific research and medical studies of aplastic anemia pathophysiology and treatment. In clinical work, our protocol for severe, treatment-nave aplastic anemia was extended to collect more data and subsequently for purposes of ancillary research laboratory studies, both aimed in particular at long-term outcomes. Our large cohort is of unique value in providing accurate estimates of response rates, relapse, and evolution to myeloid malignancies with now standard therapy. Overall and complete response rates are consistent with the initial report, and survival remains very high. Retrospective analysis of pediatric patients has not shown benefit for addition of eltrombopag to standard immunosuppression, of importance due to the favorable outcomof transplantation from alternative donors in children, Evolution to myelodysplastic syndrome and acute myeloid leukemia, the most serious late consequence of acquired aplastic anemia, appears to be equivalent or lower than historical results: high risk cytogenetic transformation (monosomy 7 and complex abnormalities) at a cumulative incidence of <5% at >2 years follow-up from treatment. In a recently approved second protocol for treatment-nave disease, now actively accruing patients, we will initiate low risk oral therapy, low dose cyclosporine and eltrombobag, in order to introduce immunosuoppression and stem cell stimulation early and to prevent loss of hematopoietic cells to T cell killing. Telemedicine will be employed to enroll and follow patients until they are admitted to the Clinical Center for definitive anti-thymocyte globulin infusion (if needed). The endpoints of this novel approach are unexpected adverse events, especially drug toxicity, errors in diagnosis, and patient compliance to outpatient visits and physician directives. Relapse remains a serious complication after immunosuppression, In subjects treated with horse ATG, cyclosporine and eltrombopag on our phase II clinical trial (12-H-0150), long term complications and outcomes were analyzed at a median follow up time of 2 years. Among the 6 month responders, the cumulative incidence of relapse was 42%. Relapse occurred predominantly at 6 months and 2 years after treatment; two-thirds of relapsed patients had blood count recovery upon reinitiation of therapies. High risk clonal evolution, isolated acquisition of chromosome 7 abnormalities or complex cytogenetic changes, or a morphological diagnosis of MDS/AML, has occurred in 6% of total subjects. Older age was the major risk factor for both relapse and clonal evolution. Our ongoing protocol, now more than half accrued, is assessing rapamycin (sirolimus) as an agent to induce tolerance and to prevent relapse, based on preclinical data from animal models. Patients are randomized to 3 months of sirolimus versus no therapy and followed for two years, with relapse the primary end point. Study drug has been well tolerated. There are early signals from the numbers of relapses, deaths, transfusions, and platelet counts that of a difference between the study arms. Future protocols, preferably multicenter, would introduce as replacement for cyclosporine after discontinuation of eltrombopag and prior to dose reduction of cyclosporine. Our mouse model of immune marrow failure has continued to yield interesting data. Based on our finding of a role of macrophages and tumor necrosis factor in the pathologic immune response, experiments have been performed to test other components of the innate immune system. Using combinations of mice genetically deficient in several toll-like receptors (2 and 4) and IL-6 as donors and recipients in the model, we were unable to detect significant or consistent effects on the production of marrow failure. However, depletion of neutrophils and myeloid-derived suppressive cells using an anti-Ly-6G antibody worsened immune-mediated damage to hematopoietic cells, with accompanying augmented expansion and activation of cytotoxic T lymphocytes. In collaboration with Drs. Mock and Trinchieri we have explored the microbiota in relationship to murine hematopoiesis and in the immune marrow failure model. Attenuation of immune bone marrow damage was observed in conventionally housed mice, and conventionally co-housed mice had increased gut microbiota diversity, increased expression of genes associated with hematopoietic cell survive, increased myeloid cells and effector memory T cells, and enhanced reconstitution of Lin-Sca-1+Kit+ hematopoietic stem and progenitor cells. We have extended our investigations of sirolimus in non-immune mediated hematopoietic injury models: sirolimus augments hematopoietic stem and progenitor cell regeneration following hematopoietic insults. We are planning to test therapeutic efficacy of three commercial Jak1/Jak2 inhibitors in murine models of immune-mediated bone marrow failure. In translational experiments utilizing patient samples, we have emphasized RNA sequencing in single cells (scRNAseq). Previously, we published our analyses of scRNAseq in normal and aneuploid human CD34 bone marrow cells. Our analysis of lymphocytes in a large cohort of patients with large granular lymphocytosis, before and after treatment, is ongoing. scRNAseq has confirmed by sequencing alone the characteristic mono- and oligoclonality of cytotoxic lymphocytes; successful treatment with alemtuzumab, a monoclonal antibody which we have shown to be effective in refractory disease, did not greatly alter T cell clonality or cell number but markedly affected lymphocyte activation. We have examined loss of HLA expression in hematopoietic cells of patients with aplastic anemia; loss of heterozygosity of the 6p chromosome and somatic mutations of HLC class I genes are frequent in this disease, likely as an escape mechanism from immune attack on marrow target cells; we have found mutations to be equivalently prevalent for multiple different HLA-A and -B alleles. Leukocytes lacking surface HLA class I allele expression due to somatic mutation and/or loss of heterozygosity were observed in 92 of 414 patients with immune aplastic anemia; the presence of HLA class I allele loss associated with significantly more high-risk clonal evolution after immunosuppressive therapy. Immunophenotyping of circulating and bone marrow B cells revealed markedly reduced regulatory B cells in immune aplastic anemia,