1. Characterizing T cells, NK cells, and the cytokine milieu in AML to explore immune editing. Our studies support the hypothesis that AML creates an immune sanctuary favorable for leukemia growth and persistence through extensive immune-editing of T cells and NK cells. T cell profile: In collaboration with Dr Strickland (Vanderbilt University) samples were prospectively collected from 32 patients (median age 63 years; range 24-80) with newly diagnosed AML or high risk MDS under the IRB approved protocol VICCHEM 1073. Mononuclear cells were isolated from both blood and marrow at the onset of leukemia, 14-21 days (early phase) and 28-35 days (late phase) after the initiation of induction chemotherapy. Healthy donor peripheral blood from age and gender matched population (n=31, median age 61 years) were simultaneously analyzed as controls. An extensive multi-color flow cytometry analysis was performed to characterize the subset of T cell, Tregs, NK cells, NKT cell and gamma-delta (γδ) T cell with various functional markers including T cell exhaustion probes (PD-1, LAG3, TIM-3, PDL-1) and NK cell activating or inhibitory molecules (KIR, Lir-1, NKG2A, NKG2D). T cells showed persisting abnormalities from diagnosis to late phase after induction chemotherapy suggesting immune editing by AML. Major finings were 1) an exhausted phenotype with significantly high expression of PD-1 in both CD4 and CD8 T accompanied with high LAG3 and TIM-3 expressions; 2) higher frequency of FoxP3+ CD4 cells in both peripheral blood and bone marrow due mainly to augmentation in the HeliosloFoxP3+Treg (p<0.01); 3) predominance of immune-regulatory cytokines of IL-27 and IL-35. Since IL-27 induced PD-L1 expression in CD4+ nave and central memory cells, we hypothesize that IL-27 production induced by AML cells causes exhaustion of CD4+ T-cells through increased PD-L1 expression. NK cell profiles: NK cells also showed important changes suggesting immune editing: 1) a maturation arrest with significantly lower frequency of the mature CD56dim NKG2A+KIR-CD57- NK cells at diagnosis increasing to more normal frequencies after induction chemotherapy; 2) reductions in activating molecules NKG2D; and higher expression of Lir-1, and other NK cell inhibitory receptors. In a separate study, in collaboration with Stringaris and Rezvani (London), NK cell phenotype and function were measured in 32 consecutive AML patients at presentation, including 12 achieving complete remission. The most striking finding was that AML NK cells had reduced cytotoxicity against autologous blasts and K562 targets and co-incubation of AML blasts with healthy donor NK cells also impaired NK cell TNF-alpha; and IFN-gamma; production against K562 targets. The inhibitory effect was primarily mediated by IL-10. These results suggest that AML blasts reduce the competence of the innate immune system, favoring leukemia survival. 2. Mechanisms of immune suppression: During studies exploring the usefulness of the human leukemia cell line K562 as an artificial antigen presenting cell (aAPC), we found that HLA-A*02:01 and CD89 transduced K562 pulsed with CMV pp65 and IE-1 peptides failed to elicit antigen-specific CD8 T cells. All K562 lines strongly inhibited T cell proliferation to superantigen stimuli and MLR and abrogation of the effect was only achieved by brief fixation of K562 with 0.1% formaldehyde. Subsequently, we demonstrated that myeloid leukemia lines and primary AML cells suppressed CD4 T cell proliferation. Recent murine and human studies suggest that the STAT3 pathway not only promotes leukemia survival but also plays a key role in the immunosuppressive function of myeloid derived suppressor cells (MDSC). Therefore, we hypothesized that myeloid leukemia might suppress T lymphocytes through the STAT3 pathway. Phosphorylated STAT3 was markedly activated in suppressive leukemia lines and prior exposure to STAT3 inhibitors almost completely abrogated their suppressive effect on CD4 T cell proliferation. These findings confirmed that myeloid leukemia, like MDSC, directly suppress T cells through a STAT3 pathway and may contribute to an immune-editing effect of myeloid leukemia. Kortylewski (City of Hope, CA) has developed a siRNA STAT3 blocker conjugated to CpG, which binds preferentially to TLR9 on AML cells, sparing the STAT3 pathway in T cells or NK cells. He showed that STAT3 blockade could induce remission in leukemic mice by relieving immune suppression as well as by blocking leukemic proliferation. In collaboration with Dr. Kortylewski, we demonstrated that his selective STAT3 inhibitor effectively abrogates the immune-suppressive function of myeloid leukemia cells . Our results suggest that a STAT3 inhibitor could be used to augment leukemia-targeted immunotherapy. 3. Post-transplant relapse and immunotherapeutic strategies. Relapse after allo-HSCT remains a major cause of treatment failure occurring 20-40% in standard risk and up to 40-80% in high-risk patients. Our approach has been to offer second HSCT to relapsing patients but in a recent analysis we found the second allo-HSCT did not prolong survival compared with other treatments in either early or later relapses and no patient survived beyond 1y if relapse occurred within 6 months post-transplant.Since no progress has been made in the treatment of relapsed leukemia after HSCT we focused on strategies to predict and prevent relapse. Using an extended multigene array for AML markers developed by Dr. Hourigan, we retrospectively tested pre-transplant blood of 74 AML patients who underwent allogeneic allo-HSCT between 1994-2012. This approach predicted all clinical relapses occurring in the first months after allo-HSCT.This study provides proof of principle that a multiple gene approach may be superior to the use of WT1 expression alone for AML residual disease detection and emphasized the need for disease control prior to the conditioning regimen. Because of the ease of tracking residual disease using BCR/ABL transcripts, CML represents a useful model to study disease control and the stem cell nature of residual leukemia after HSCT. We retrospectively analyzed clinical and biological parameters determining post-relapse outcome of patients with CML relapsing after allo-HSCT and found that patients with BCR/ABL expression in their early hematopoietic stem cell compartment (HSC: Lineage-CD34+CD38-CD90+) had worse survival irrespective of the disease status emphasizing the need to target early CML progenitors.Recent observations that cytomegalovirus (CMV) reactivation was protective against relapse in myeloid leukemias prompted us to analyze this unique favorable risk factor of relapse in our HSCT cohort. Early CMV reactivation was an independent factor for decreased relapse in CML, but not other malignancies. These findings stimulated us to study interactions between CMV-exposed lymphocytes and myeloid leukemia stem cells (LSC). Such studies are hampered by failure of primary leukemias to survive in culture. We developed an in vitro LSC culture system supported by MSC and demonstrated long-term maintenance of LSC.59 This approach may facilitate experiments to understand interactions between LSC and immune cells and the marrow stroma. 4. Kinetics of T cell and NK cells after high dose lenalidomide. A phase I trial (12-H-0146) in patients with high risk MDS, CML, and AML evaluates clofarabine (5mg/m2/d x5 iv) followed by oral lenalidomide with dose escalation to a maximum of 2 cycles 50 mg daily for 28 days. We hypothesize that the cytoreductive effect of clofarabine will create a favorable platform for sequential therapy with lenalidomide to favorably modulate T and NK cell function. Study endpoints are treatment toxicity and lymphocyte subset and serum cytokine profiles to characterize the effect of lenalidomide.
Showing the most recent 10 out of 43 publications
