Cellular Immunotherapies for Leukemia Patients

Abstract

leukemia: leukemiaHigh dose therapy and allogeneic hematopoietic stem cell transplantation (HCT) began at Seattle's Fred Hutchinson Cancer Research Center in the 1960s, led by Nobel Prize–winning E. Donnall Thomas, MD, and represented one of the first curative treatments for relapsed leukemia. Using reduced intensity conditioning, Rainer Storb, MD, and colleagues were among the first to show that donor immune T cells play a major role in successful allogeneic HCT through graft-versus-leukemia effects (GVL). Further proof of the immune system's leukemia-eliminating potential comes from the activity of donor lymphocyte infusions (DLI) in some patients who relapse with leukemia after HCT. Unfortunately, allogeneic HCT and DLI can also cause graft-versus-host disease (GVHD), in which the donor immune system attacks the patient's normal tissues. Depletion of all donor T cells from the graft reduces GVHD, but also limits GVL and increases the incidence of serious infections and disease relapse. Engineering Improved HCT Outcomes One approach to potentially improve HCT outcomes is to selectively remove the cells that cause GVHD, sparing GVL. Preclinical studies by Marie Bleakley, MD, PhD, MMsc, and Stanley Riddell, MD, showed that CD45RA+ “naïve” T cells (TN) are enriched for the capacity to cause GVHD while donor grafts depleted of CD45RA+TN can protect from infection and potentially deliver a GVL effect with less risk of severe GVHD. This graft engineering strategy has been translated from the laboratory to the clinic in phase I/II trials in HLA-matched, peripheral blood HCT for adults (NCT02220985) and children (NCT01858740) with high-risk acute leukemia or chronic myelogenous leukemia in accelerated phase/blast crisis. In the first 35 patients on these protocols, engraftment was rapid, acute GVHD has been universally corticosteroid-responsive and the risk of chronic GVHD markedly decreased. Patients appear to have low rates of infection and relapse (J Clin Invest 2015;125(7):2677-89). Brenda Sandmaier, MD, and Ann Woolfrey, MD, are testing an alternative HCT approach to reduce GVHD while protecting against infection and disease relapse. T-cell depleted grafts from partially HLA-mismatched, related (haploidentical) donors are delivered to adult (NCT01744223) and pediatric (NCT01744223) leukemia patients, followed by infusions of donor T cells genetically modified to contain a “safety switch” that can be specifically drug-activated to rapidly ablate the cells should severe GVHD arise. This strategy could make mismatched HCT safer and available to more patients. Adoptive T-Cell Therapy In the 1990s, Fred Hutch investigators began isolating T cells that target specific antigens, including native, tumor-infiltrating lymphocytes (TILs) that can be collected, enriched, and expanded from tumor biopsies and reinfused to amplify virus- and cancer-fighting capacities. However, endogenous antitumor T cells are difficult to expand ex vivo and often insufficient for the eradication of established tumors after infusion, as also evidenced by the failure of cancer vaccination and checkpoint blockade strategies against select tumors, including leukemias. This recognition led to approaches to genetically-modify CD8+ cytotoxic T cells (CTL) using innovative T-cell receptor (TCR)-based and chimeric antigen receptor (CAR) strategies to increase T-cell persistence and anti-tumor activity. Philip Greenberg, MD, Head of the Fred Hutch Program in Immunology, and Riddell are world-recognized leaders in these efforts. TCR-Based WT1-Specific Immunotherapy TCRs are the natural molecular structures by which T cells bind target peptides “presented” by specific major histocompatibility complex molecules (HLA) on the cell surface. Because TCR's can recognize processed peptides from extracellular or intracellular proteins, TCR-based therapies can target a broader range of proteins than surface antigen-targeting CARs, across many tumor types. Greenberg's team is leading the way in developing TCR-based therapies. They are identifying immunogenic cancer-associated proteins that promote tumor survival and growth, obtaining high-affinity antigen-specific TCRs in HLA-matched screens and purifying the relevant TCR-encoding genes. The TCR genes are then transferred into patient or donor T cells, which are rapidly expanded to generate large numbers of tumor-reactive T cells for adoptive transfer. For acute myeloid leukemia (AML), one target has been Wilms tumor antigen 1 (WT1), a highly overexpressed “self-antigen” that contributes to the malignant phenotype. WT1-specific CD8+ T cells can distinguish overexpressing targets and lyse leukemic, but not normal CD34+ cells. Aude Chapuis, MD, and colleagues undertook a trial for AML patients who had received an allogeneic HCT. WT1-specifc CD8+ CTL clones with the highest-avidity were selected and expanded for each HLA-A∗0201 (A2)-matched patient-donor pair. Among 11 relapsed or high-risk leukemia patients, CTL clones proved safe and evidence of antileukemic activity was noted in two patients. However, variable functional avidities and short persistence limited anti-leukemic activity (Sci Transl Med 2013;5(174):174ra27). To overcome this drawback, a high affinity WT1-specific TCR (TCRC4) was isolated from a screen of healthy HLA-A2+ donors. TCRC4 was then genetically inserted into virus-specific (EBV or CMV) donor T cells, since the native TCR, being virus specific, should not cause GVHD. These virus-specific T cells, especially the EBV-specific populations, also contain high frequencies of central memory T (TCM) cells associated with prolonged persistence of transferred T-cells. In an initial clinical trial, 11 AML patients who were early post-HCT, with no measurable disease but at high risk for disease relapse, received WT1-specific TCRC4-transduced, virus-specific CTL as a prophylaxis to prevent relapse (Blood 2016;128(22):1001). All continue in complete remission at a median follow-up of 27 months post-HCT, whereas 27 percent of a concurrent high-risk cohort given HCT alone had relapsed by 16 months. Preliminary results suggest that continued WT1-antigen encounter drove CTL proliferation/persistence. There was evidence of a direct antitumor effect for another 11 patients who had relapsed post-HCT and received WT-1 specific TCRC4-transduced CTL, but there was no overall long-term survival advantage. Four of these patients had received CMV-specific substrate cells, which persisted less well than EBV-specific cells and may in part explain the lack of long-term survival benefit in these patients. A follow-on trial is planned for HCT-ineligible AML patients who complete induction/consolidation and are at high risk of relapse. Polyclonal, autologous CD8+ TN, TCM and EBV-specific T cells will be genetically modified to express the WT1-specific TCRC4 and infused into patients. Aims of this study include identifying the optimal T cell subset for WT1-specific TCR-based immunotherapy. This approach may have broader utility as WT1 is also overexpressed in other malignancies, including solid cancers. mH Antigen Targeting Based on the observation of hematopoietic-restricted minor histocompatibility (mH) antigen mismatching between donors and recipients and a lower risk of leukemic relapse after HCT, E.H. Warren, MD, PhD, Riddell, and Fred Hutch colleagues selected donor CD8+ CTL specific for hematopoietic-restricted recipient mH antigens, expanded the clones ex vivo and infused seven patients who had relapsed post-HCT (Blood 2010;115(19):3869-78). Five patients had transient remissions, but poor CTL persistence. Seeking effective and safe CTL targets, Bleakley's team is focused on the HA-1H mH antigen, expressed predominantly/exclusively on hematopoietic cells. They have identified and selected a HA-1H-specific TCR that recognizes the HA-1/HLA-A∗0201 complex, but not the corresponding non-immunogenic HA-1R complex, and produces efficient leukemia-specific killing by transduced T cells. Clinical trials are anticipated in 2017. CAR-Mediated Anti-CD19 Immunotherapy As in other centers, Fred Hutch researchers are designing antigen-specific CAR constructs to redirect T cell immune responses, leveraging decades of therapeutic antibody development. Current CARs typically encode an extracellular, antigen-recognition module of single-chain antibody domains (variable domains of the light and heavy chains; i.e. scFv) linked to a stalk-like region, a transmembrane region and intracellular T cell signaling domain(s). With scFv antigen-binding, CARs can only recognize proteins expressed on a cell's surface. However, CAR-directed responses are not HLA-restricted, such that one antigen-specific CAR is expected to have broader utility than any one TCR recognizing the same target protein. CARs directed against CD19 are the most clinically developed. CD19 is a B-cell lineage-specific antigen expressed on normal and malignant B cells in patients with B-cell acute lymphoblastic leukemia (B-ALL), chronic lymphocytic leukemia (CLL) and non-Hodgkin lymphoma (NHL). Lymphodepletion chemotherapy followed by infusion of CD19-specific CAR-T cells has produced remarkably high rates of complete remission in ALL and encouraging responses in NHL and CLL, including durable responses. As Cameron Turtle, MBBS, PhD, FRACP, FRCPA, Riddell, and I described in a recent review of CD19-targeting CAR therapies, there are many differences in CAR-T cell approaches across centers (Clin Pharmacol Ther 2016;100(3):252-8). Differences include the use of retroviral- versus lentiviral-based CAR constructs, inclusion of CD28 or 4-1BB costimulatory sequences, which T cell subset(s) are used for production and which lymphodepletion chemotherapy regimen is used. Despite this variation, these “second generation” CD19 CARs have produced dramatic remission rates in early trials. In addition to the expected adverse effects of lymphodepletion chemotherapy, CAR-T immunotherapy can be complicated by cytokine release syndrome (CRS), manifesting as fever, hypotension, capillary leak and coagulopathy and due to T-cell activation, tumor cell killing and the release of inflammatory mediators. Reversible neurologic toxicity has also been observed, mostly presenting as delirium and somnolence, although seizures and/or stroke-like phenomena can occur. In most cases, CAR-T toxicities have been transient and manageable, using the anti-IL-6 receptor antibody, tocilizumab, and/or corticosteroids. Our studies have suggested that tumor burden and associated levels of tumor antigen can drive CD19 CAR-T expansion and toxicity. Most centers manufacture CAR-T cell products from bulk T cells, despite substantial variation in the frequencies of specific T-cell subsets. Based on preclinical results, we initiated a clinical trial using cyclophosphamide (Cy)-based lymphodepletion with or without fludarabine (Flu), followed by infusion of CD19 CAR-T cells in a 1:1 ratio of CD4+:CD8+ CAR-T cells (NCT01865617). The lentiviral CAR construct included a CD19 specific murine scFv, a 4-1BB costimulatory domain and a CD3z signaling sequence. Cells were marked with a truncated human epidermal growth factor receptor, allowing precise enumeration, and formulation of CAR-T cells. We initially reported results for 29 adults with relapsed/refractory CD19-expressing B-ALL and 28 adults with relapsed/refractory B-NHL or CLL (Blood 2015;126(23):3773, Blood 2015;126(23):184). Twenty-four of 26 restaged ALL patients (93%) achieved CR, with no evidence of leukemia in the bone marrow. Clinical responses were observed in 50 percent of 12 NHL patients who received Cy-only lymphodepletion (1 CR, 5 partial responses/PR). To minimize rejection of the murine-based CAR, Flu was added to the lymphodepletion regimen for the subsequent 16 NHL patients, increasing the CR rate from 8 percent to 42 percent. Responses among the first six CLL patients included three CR and one PR. In an updated report, among ibrutinib-refractory (n=10) or intolerant (n=3) CLL patients, the overall response rate was 77 percent (7 PR, 3 CR). We have now reported results for adults with ALL, NHL or CLL treated with a defined 1:1 ratio of CD8+:CD4+ anti-CD19 CAR-T cells, at one of three dose levels following lymphodepletion (Blood 2016;128(22):1852). Among patients (45 ALL, 47 NHL, 17 CLL) who have completed toxicity and response assessments, we have established relationships between the infused CAR-T cell dose, in vivo CAR-T cell expansion and subsequent risk of toxicity. The maximum tolerated dose was 2x106 CAR-T for NHL and CLL. For ALL patients, toxicities were correlated with the percentage of marrow blasts and the CAR-T cell dose. For patients with ≥5% marrow blasts, CRS and neurotoxicity was mitigated by administering 2x105 CAR-T cells/kg, without compromising efficacy. We conclude that adoptive CD19 CAR-T cell therapy with defined subset composition is feasible, with potent anti-tumor activity and toxicity related to cell dose. Cy/Flu lymphodepletion results in greater CAR-T cell expansion and persistence, and can improve the rate of CR. We measured higher levels of IL-15, IL-6, IL-8, IL-10, soluble TNF receptor type 1, and IFN-γ on day 1 post-infusion in patients who subsequently developed severe CRS, including patients with and without neurotoxicity. This finding suggests that targeted intervention strategies might be applied early to reduce severe toxicity in patients at high risk of relapse. We are applying these findings to develop other new CAR-based therapies for patients with B-cell malignancies, including leukemias that may downregulate CD19 levels during CD19 CAR-T cell therapy. For example, ROR1 is a surface protein expressed on CLL and mantle cell lymphoma (MCL), and on a subset of B-ALLs. As ROR1 is not expressed on mature B cells, R0R1 targeting may avoid the depletion of normal CD19+ B cells that is observed after CD19 CAR-T cells. The Riddell group has developed a ROR1 specific CAR construct that is now in a phase I clinical trial (NCT02706392) for patients with ROR1+ CLL, MCL or ALL that is refractory to conventional therapy. New Targets & Improved Immunotherapy Novel cellular immunotherapy requires translation from the research laboratory to the clinic. Monitoring of these trials requires sample collection from patients and analysis back in the research laboratories. With close collaboration, using preclinical advances in vitro and in animal studies along with clinical trials, we are beginning to learn how to make cell-based therapies safer and more effective and available to more patients. Ongoing approaches include using longer lived T-cell subsets, engineering constructs with specific function-enhancing elements and suicide genes that can be activated should serious toxicities arise, manipulating the “tumor microenvironment” to overcome otherwise immune-inhibiting processes, combining multiple immunotherapy approaches, and combining T-cell therapies with targeted drug therapies. It is important that we continue to learn to identify, treat and eventually prevent toxicity. Fred Hutch investigators are also continuing to identifying novel antigen targets for both TCR- and CAR-based therapies. Early in 2016, the Fred Hutch and the Seattle Cancer Care Alliance (SCCA) launched an Integrated Immunotherapy Research Center, led by Riddell. In November, we opened a first-of-its-kind clinic, “The Bezos Family Immunotherapy Clinic” at the SCCA, committed to develop and offer innovative cellular immunotherapies. Lessons from allogeneic HCT first demonstrated the power of the immune system to eradicate leukemia and the recent unprecedented response rates to CD19 CAR-T cell therapy in patients with refractory ALL is now leading the way to targeted cancer therapies with curative potential. DAVID G. MALONEY, MD, PHD, is a member of the Clinical Research Division, Leonard and Norma Klorfine Endowed Chair for Clinical Research; Medical Director, Cellular Immunotherapy, Fred Hutchinson Cancer Research Center; Professor of Medicine, Division of Oncology, University of Washington; and Medical Director, Bezos Family Immunotherapy Clinic, Seattle Cancer Care Alliance. DEBORAH E. BANKER, PHD, is Senior Science Writer, Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle.David G. Maloney, MD, PhD: David G. Maloney, MD, PhD