At the 2018 American Society of Hematology Annual Meeting, the AML Global Portal (AGP) attended a poster discussion session. At this session, Jacalyn Rosenblatt, Beth Israel Deaconess Medical Center, US, discussed immune checkpoints inhibition in acute myeloid leukemia (AML).
Dr. Rosenblatt began her talk by highlighting the promise of immunotherapy in AML. Immune therapy has the potential to “selectively target the malignant clones” and the capability to target tumor heterogeneity broadly, including malignant stem cell populations. Moreover, the immune response has the potential to induce memory response, which can provide long-term immune protection from relapse. However, in order to develop an effective immune therapy for leukemia, it is “really critical to think about factors that must be overcome with a potent immune therapy”. Some of these factors include ineffective antigen presentation by leukemia cells, lack of an effective T-cell repertoire and regulation of negative checkpoint pathways. The speaker then discussed three abstracts (#7016, #7015 and #7014) presented at the ASCO hematology poster session.
Abstract 7016: Treg infiltration and the expression of immune checkpoints associated with T cell exhaustion in AML
Patrick Williams and colleagues from the MD Anderson Cancer Center (MDACC) aimed to better define the immune checkpoint landscape in AML including further defining T cell subsets, checkpoints on T cell subsets, and ligands on AML blast. To do this, the researchers performed flow cytometry (17 color) analysis on bone marrow (BM) aspirates from 107 AML patients treated at the MDACC. The expression of inhibitory (PD1, CTLA4, LAG3, TIM3) and stimulatory receptors (GITR, OX40, 41BB, ICOS) on T cell subsets and the expression of their ligands (41BBL, B7-1, B7-2, ICOSL, PDL1, PDL2, and OX40L) on AML blasts were evaluated.
Compared to healthy controls (n = 8), patients with new (n = 39) and relapsed (n = 68) AML have an increased frequency in PD1+ and OX40 positive CD8 T cells. Additionally, it was observed that there was an increased frequency of Tregs in AML BM compared to healthy controls. Furthermore, patients with AML have a higher frequency of exhausted PD1-TIM3 and PD1-LAG3 double positive T cells in the BM. In TP53 mutated AML and patients with adverse cytogenetics, PD-L1 expression in the BM is increased.
In summary, there is an increased Treg infiltration in the BM of AML. The authors concluded that the “dual contributions from Tregs and checkpoint ligand expression by blasts to immune suppression indicate that these pathways may play an important role in AML survival and therefore patients may benefit from checkpoint antibody therapy”.
Jacalyn Rosenblatt added that the increased PD-LI expression suggests that “this pathway may play a role in muting T cell responses and may be a way of overcoming some of these adverse prognostic features”. She further added that this study provides a “rationale for incorporating checkpoint blockade into the treatment of leukemia”.
The next abstract discussed at this poster session by Paola Dama and colleagues from the University of Chicago, Chicago, IL. The researchers aimed to characterize immune checkpoint pathways in AML patients treated on the prospective study (NCT02573363) of selinexor (a first in class SINE XPO1 antagonist) in combination with high-dose cytarabine (HiDAC) with mitoxantrone (Mito). Bone marrow and peripheral blood samples at diagnosis and following remission induction therapy were analyzed in 26 AML patients. Of these patients treated in this prospective study, 16 patients achieved complete remission (CR) and 10 patients experienced treatment failure. Immune markers were compared between responders and non-responders.
In patients experiencing treatment failure after chemotherapy, the percentage of CD34- Gal9+ cells were significantly higher compared to patients in CR. At the time of remission, a significant increase of expression of PD-L1 on BM CD34-cells was observed. Markers of exhaustion were detected, both in the peripheral blood and the marrow, in patients who achieved remission with high levels of TIM-3 and PD-1 expression in the marrow as well as in the peripheral blood.
This small study demonstrated “high-level expression of Gal9 in CD34- cells in the BM at diagnosis in patients who failed induction chemotherapy. Increased expression of Tim 3 on CD4 and CD8 T cells in the BM and high PD-1 in peripheral CD4+ T cell at the disease remission suggested an exhausted immune status”.
Jacalyn Rosenblatt noted that the data from this study support an exhausted immune status in AML and suggests that the TIM-3/PD-L1 pathway “may play a role in patients in remission and muting ongoing immune surveillance”. Thus, this study provides a rationale for “incorporating checkpoint blockade once patients achieve remission, to provide ongoing immune surveillance”.
Abstract 7014: Nivolumab (Nivo) maintenance (maint) in high-risk (HR) acute myeloid leukemia (AML) patients
The final abstract discussed in this session was by Tapan Kadia and colleagues, MDACC, on a phase II study (NCT02532231) evaluating nivolumab maintenance in high-risk AML patients in CR ineligible for standard chemotherapy. The primary endpoint of the study was relapse-free survival. Secondary endpoints of the study include toxicity, immune response, minimal residual disease (MRD) and overall survival (OS).
In total, 14 high-risk AML patients (median age = 57 years, range 32–71) were enrolled in this study. Patients received nivolumab at two-week intervals for the first three doses, then monthly for a year, and then every three months until relapse.
Treatment was well tolerated. Grade 3–4 immune-related adverse events occurred in five patients, one patient had autoimmune hemolytic anemia and came off study. MRD (data in 14 patients) and cytogenetic abnormality clearance were seen in one patient. With a median follow-up of 19.3 months, the median OS was not reached and the median CR duration was 8.3 months.
In conclusion, maintenance therapy with nivolumab is safe in AML patients treated in remission. Interestingly, clinical outcomes observed in this study surpassed expected historical controls.
Jacalyn Rosenblatt concluded this poster discussion session by highlighting that “each of these studies really speaks to the potential important role that these pathways, including the PD-1/PD-L1 pathway, plays in muting activated T cell responses in leukemia”. Moreover, “there is a great potential for incorporating checkpoint blockade in the treatment of leukemia patients”. However, she noted that in order to “carefully inform the best timing of therapy and the best combinations of therapy, it would be critical to carefully characterize factors in the microenvironment, as well as on the leukemia cells at diagnosis, serially, throughout therapy, and at relapse”.
Questions on the use of immune checkpoint blockade in AML still remain. The speaker proposed two questions, “Will leukemia be like Hodgkin's disease, where there are really dramatic results with single-agent activity?” or “will leukemia be more like multiple myeloma, where there is not as dense a native T cell infiltrate at the tumor bed?”.