Page 55 - Read Online

Page 130 Saliba et al. Cancer Drug Resist 2021;4:125-42 I http://dx.doi.org/10.20517/cdr.2020.95

the U.S. Food and Drug Administration (FDA) in September 2020 approved once-daily oral azacitidine as
maintenance therapy for the continued treatment of AML in adults who achieve their first CR/CRi with
induction chemotherapy and are not candidates for further intensive consolidation therapy with HSCT [71,72] .

A common theme across the clinical trials of HMAs in AML is a consistent trend towards improved
survival despite the low CR rates of < 20%. This led to a number of studies examining the impact of adding
various agents to HMAs, ultimately resulting in the HMA/venetoclax regimen described in a subsequent
section of this review.

Mechanisms of resistance to hypomethylating agents
When HMAs are used as single agents in myeloid malignancies, rates of remission are low, and durations
of response are often short. Resistance, whether primary or secondary, is a central challenge in the
treatment of MDS and AML. Therefore, potential mechanisms of resistance to HMAs have been extensively
investigated. In brief, for HMAs to be active, they must accumulate in cells, undergo phosphorylation to
the nucleoside triphosphates, avoid metabolic inactivation, and get incorporated into DNA. Resistance has
been observed at each of these steps.

Resistance to HMAs can be mediated by membrane proteins that are involved in drug uptake. Human
equilibrative nucleoside transporter-1 (hENT1) facilitates transport of nucleosides, including decitabine
[88]
and azacitidine, in a bidirectional manner across the cell membrane . Consistent with a role for hENT1 in
HMA cellular uptake, the level of hENT1 mRNA expression is significantly higher in patients who respond
[89]
to decitabine . Interestingly, the decreased activity of azacitidine in cells with low hENT activity can be
overcome by the use of an elaic ester of azacitidine, CP-4200, with reduced dependence on membrane
transporters .
[90]

The balance between HMA phosphorylation and enzymatic inactivation also plays a role in response.
Cell lines that are resistant to decitabine have not only low expression of hENT1 and hENT2 but also low
expression of DCK and high expression of CDA [36,89] . Both of these changes are thought to play a role in
HMA resistance. Low expression of DCK diminishes the activating phosphorylation of decitabine, and
high CDA results in rapid catabolism of decitabine and azacitidine to uridine counterparts that do not
contribute to DNMT inhibition [91,92] . Consistent with these observations, malignant cell lines engineered
to overexpress CDA are decitabine resistant. Azacitidine and decitabine have poor oral availability, likely
reflecting high CDA expression in gastrointestinal and liver cells [91,93,94] .

In the clinical setting, it has been reported that lower expression of UCK1 is associated with blunted
response to azacitidine and that DCK is significantly reduced in the bone marrow and peripheral blood
[95]
[89]
specimens from MDS patients who relapsed after decitabine , highlighting the importance of metabolic
activation in the action of HMAs. These findings, however, have been far from universal. It has also been
reported that expression of DCK or CDA does not distinguish MDS cases that respond to decitabine from
[96]
those that do not, but the ratio of CDA to DCK was elevated in nonresponding cases .

Emerging evidence suggests that metabolism of pyrimidines, including HMAs, reflects an adaptive network
[35]
of enzymes that detect nucleotide levels and respond in a compensatory manner [Figure 1] . For example,
DCK protein is diminished in MDS samples with decitabine resistance and upregulated in azacitidine-
resistant cells. Conversely, UCK2 protein is upregulated in decitabine-resistant cells and diminished in
azacitidine-resistant cells . These observations raise the possibility of tailoring therapy based on DCK and
[35]
UCK levels or diminishing HMA resistance by alternating HMAs [35,97] .

After HMAs are metabolically activated, they are incorporated into DNA in place of dCTP. The enzyme
carbamoyl-phosphate synthetase (CAD) facilitates the synthesis of dCTP, which competes with aza-dCTP,