Page 64 - Read Online

Gurska et al. Cancer Drug Resist 2023;6:674-87 https://dx.doi.org/10.20517/cdr.2023.39 Page 680

CLINICAL INVESTIGATION OF IMMUNE CHECKPOINT INHIBITORS IN AML
Several strategies that incorporate checkpoint inhibitors have been tested in AML in clinical trials, and
several more clinical trials are underway [Table 1]. In a phase 1/1b clinical trial of ipilimumab (anti-CTLA4)
in patients with hematologic malignancies that relapsed after allogeneic stem cell transplantation, analysis of
the AML subset (12/28 patients) showed that 5/12 patients achieved complete remission following
treatment, and this was accompanied by a reduction in the frequency of circulating Tregs compared to non-
[54]
responders .
Furthermore, in a Phase II study investigating the combination of high-dose cytarabine and pembrolizumab
(anti-PD1) in relapsed/refractory AML patients, 14 out of 37 patients achieved complete remission (CR).
Interestingly, of the patients that achieved a CR, TCR signaling identified a trend towards increased TCR
diversity in these patients, as well as decreased regulatory T cell and increased CD8+ T cell frequencies . Of
[55]
note, RNA-seq analysis of AML blasts from these patients revealed that increased MHC expression was
significantly upregulated at baseline in patients who achieved CR compared to non-responders .
[55]

Interestingly, recent data suggests that PD1 signaling may be implicated in the poor response to
hypomethylating agents (HMAs), including azacitidine and decitabine, as patients who are resistant to
HMAs show higher expression of PD-L1, PD-L2, and CTLA4 [56-58] . On the other hand, preclinical findings
from single-agent immune checkpoint inhibitor trials in AML have demonstrated limited efficacy. This
has prompted the investigation of checkpoint inhibitors in combination with HMAs . In a phase 1b
[57]
trial investigating the combination of ipilimumab with decitabine in relapsed/refractory AML, patients who
were transplant naïve (N = 23) observed a higher response rate than those who relapsed following
stem cell transplantation (N = 20) (CR + CRi + mCR 52% vs. 20%, P = 0.034; median overall survival 16.2
[59]
months vs. 8.6 months) . Not surprisingly, when performing integrative transcriptome-based analysis of
bone marrow infiltrating cells from participating patients, a high baseline ratio of T cells to AML cells was
[60]
associated with higher response rates . The authors speculated that the inadequate clearance of the
immature LSC population triggered relapse in patients following stem cell transplantation, but also noted
that ipilimumab exposure resulted in increased memory T cell bone marrow infiltration and high
expression of CTLA4 and FOXP3, suggesting that the efficacy of ipilimumab and decitabine may be
impacted by these immune evasion mechanisms employed by LSCs . The results of the ipilimumab
[60]
and decitabine combination studies also highlight the limitations of ICIs in AML. A comparison of the
memory and exhaustion gene scores associated with CD8+ T cells from AML bone marrow with those
from CD8+ TILs isolated from solid tumors, in which ipilimumab demonstrates high clinical activity,
revealed higher exhaustion profiles and checkpoint expression in solid tumor-derived T cells .
[60]

In two ongoing trials testing the combination of pembrolizumab and decitabine in relapsed/refractory
AML, interim results showed a tolerable safety profile with promising efficacy data [56,61] . Furthermore,
through the generation of RNA expression datasets from patients who were treated with conventional
cytotoxic chemotherapy or with pembrolizumab and azacitidine in relapsed/refractory AML, Rutella et al.
revealed a newly defined CD8+ T cell senescent gene population with a distinct gene expression
signature . These cells were impaired in their ability to kill AML blasts isolated from the same patient
[62]
[62]
sample, and their frequency negatively correlated with overall survival . However, there is still promise for
the combination of PD1 blockade and HMA, as results from the Phase II trial investigating nivolumab and
azacitidine in relapsed/refractory AML yielded a 33% overall response rate, with a higher response rate in
HMA naïve vs. HMA pre-treated patients (58% vs. 22%) . Based on these clinical trials, the possible
[63]
predictors of response to immune checkpoint inhibitors are summarized in Table 2. Overall, given these
data, the field is anxiously awaiting the results of additional clinical trials currently that are investigating