The Hematologist

September-October 2017, Volume 14, Issue 5

The Potential Utility of Immunotherapy for AML

Omar Abdel-Wahab, MD Assistant Member, Attending Physician
Memorial Sloan Kettering Cancer Center, New York, NY
Justin Taylor, MD Hematology/Oncology Fellow
Memorial Sloan Kettering Cancer Center, New York, NY

Published on: August 08, 2017

Study Title: A Phase 1B Study Evaluating the Safety and Pharmacology of Atezolizumab (Anti-PD-L1 Antibody) Administered in Combination With Immunomodulatory Agents in Participants With Acute Myeloid Leukemia (AML) Identifier: NCT02892318

Sponsor: Hoffmann-La Roche

Participating Centers: Multicenter national trial with 10 open sites across the U.S.

Study Design: This study is a phase Ib, nonrandomized, open-label trial of the safety of combining the anti-PD-L1 antibody atezolizumab with the novel hypomethylating agent guadecitabine in older adults with acute myeloid leukemia (AML), who are not eligible for standard induction therapy. There are two safety cohorts in this trial consisting of treatment-naïve patients who are deemed unfit for induction chemotherapy by the following standards: age either > 70 or 65 to 69 with ECOG score of 2, intermediate- or adverse-risk cytogenetics, or otherwise deemed unfit for chemotherapy, and a second cohort for relapsed or refractory AML of any age or fitness. Both of these safety cohorts have planned expansion cohorts, and the trial is actually intended to have an umbrella design where other immunomodulatory agents can be combined with atezolizumab in the future.

Rationale: Harnessing the immune system to target cancer is an idea that has been around for decades but that recently came to the forefront of cancer research with the success of so-called “checkpoint blockers” — antibodies that block inhibitory signals on T-cells allowing for release of the “brakes” on anticancer cytotoxic T-cells. The most widely successful target for this new immunotherapy is the PD-1/PD-L1 axis. Programmed cell death-1 (PD-1) is a cell surface receptor on T-cells that negatively autoregulates T-cells after binding its ligands (primarily PD-L1, but also PD-L2), which are upregulated by inflammatory cytokines to prevent autoimmunity. Many cancers have been shown to co-opt this system by expressing PD-L1 to shield themselves from immune attack. Following this logic, cancers with high PD-L1 expression generally respond to checkpoint blockers, though some responder’s tumors do not express PD-L1. While the available data on PD-L1 expression in AML is equivocal, studies have shown that hypomethylating agents (HMAs) can cause upregulation of inhibitory immune checkpoint proteins such as PD-L1, thus potentially sensitizing to blockers of the PD-1/PD-L1 axis. Additionally, recent translational research has found that HMAs also re-induce expression of silenced endogenous retroviruses that could provoke an immune response, providing further rationale for this combination. Guadecitabine is a novel HMA currently in phase III clinical trials as monotherapy that is resistant to cytidine deaminase degradation, thus providing for higher sustained levels of its metabolite decitabine.

Comments: There are at least 15 ongoing trials using checkpoint inhibitors with or without hypomethylating agents in AML. AML in complete remission (CR) with minimal residual disease or relapse after allogeneic stem cell transplant are clinical categories being explored further for use in monotherapy with some success. For example, single-agent CTLA-4 blockade with ipilimumab for relapsed AML post-SCT showed responses in 42 percent of subjects, that were durable for more than one year.1 However, monotherapy with the current immune checkpoint inhibitors is not thought to be a viable option for patients with more than minimal disease because of the usually rapid growth of AML and the typical amount of time needed for response to immunotherapy. Therefore, the focus has been on combination therapy, usually with agents other than chemotherapy, perhaps because chemotherapy for AML is usually associated with infection/inflammation and could likely be quite toxic to combine with immunotherapy. The first of these combination studies to report results was presented at the 2016 ASH Annual Meeting2 and was recently updated at the 2017 American Society of Clinical Oncology Meeting.3 This trial evaluated 53 patients treated with the combination of the HMA 5-azacitidine and the anti-PD-1 antibody nivolumab. Thus far, the overall response rate is 35 percent, with 21 percent achieving CR/CRi that mostly lasted longer than one year; however, the median survival of all patients was still 5.7 months. Preclinical investigations conducted in solid tumors has suggested that blocking the ligand PD-L1 can be more effective than blocking the PD-1 receptor. Besides the trial highlighted here with atezolizumab, there is also ongoing investigation of another PD-L1 inhibitor, durvalumab, in combination with HMA.

Other exciting developments in immunotherapy for AML include a trial of the PD-1/PD-L1 blockers with the next generation of checkpoint inhibitors such as TIM-3 antibodies (NCT03066648). TIM-3 is another T-cell surface receptor that contributes to T-cell exhaustion upon binding its ligand, phosphatidylserine. Other combinations could be considered, such as an inhibitory signal antagonist with a stimulatory signal agonist such as OX40. These stimulatory T-cell signals are the same as those utilized for current CAR-T cell construction and cause growth and survival promoting autoregulatory effects on the T-cells. Thinking beyond T-cells, other killer cells of the immune system, such as natural killer (NK) cells or macrophages could be exploited. Lirilumab is an antibody against KIR inhibitory receptors on NK cells that failed to show significant activity over placebo in a Phase II trial in AML this year,4 but appears to boost the efficacy of PD-1 blockade in head and neck squamous-cell carcinoma. Another approach is to help macrophages and dendritic cells overcome suppression by tumors that express CD-47, a ligand for the SIRP-α receptor that facilitates phagocytosis, also called the “don’t eat me” signal. The safety and efficacy of this is currently being tested in a phase I study (NCT02641002). Although the results of these clinical trials are still quite nascent, the experience with immunotherapeutics for many forms of cancer to date, combined with the heterogeneity of AML, suggest that success with any of these agents may be found only in particular subgroups of AML patients. Therefore, a critical step in future development of these drugs will be identifying the patients in whom immunotherapy will be most useful and when to use it.


  1. Davids MS, Kim HT, Bachireddy P, et al. Ipilimumab for patients with relapse after allogeneic transplantation. N Engl J Med. 2016;375:143-153.
  2. Daver N, Basu S, Garcia-Manero G, et al. Defining the immune checkpoint landscape in patients (pts) with acute myeloid leukemia (AML). Blood. 2016;128:2900.
  3. Daver NG, Basu S, Garcia-Manero G, et al. Phase IB/II study of nivolumab with azacytidine (AZA) in patients (pts) with relapsed AML. J Clin Oncol. 2017;35:7026.
  4. Press Release, Innate Pharmaceuticals. Last accessed July 18, 2017.

Conflict of Interests

Dr. Taylor and Dr. Abdel-Wahab indicated no relevant conflicts of interest. back to top