Factors affecting response to 5-azacytidine and prognosis of myelodysplastic syndrome. Is long-term survival a realistic goal?
Panagiotis T. Diamantopoulos *, Nora-Athina Viniou
First Department of Internal Medicine, Laikon General Hospital, National and Kapodistrian University of Athens, Athens, Greece

A R T I C L E I N F O

Keywords:
Myelodysplastic syndrome 5-azacytidine
Prognosis Hypomethylating agents Response

A B S T R A C T

The introduction of hypomethylating agents (HMAs) 5-azacytidine and decitabine has altered the prognosis of patients with myelodysplastic syndrome (MDS). Over the past few years, the International Prognostic Scoring System (IPSS) and the revised IPSS (IPSS-R) have been used both to define the prognosis of patients with MDS and to select patients to be treated with HMAs. Nevertheless, the prognosis of individual patients with MDS can differ considerably from the one calculated with the use of the above-mentioned prognostic systems. Thus, some patients may achieve long-term survival irrespective of their initial prognostic score. Several factors besides those used to define the IPSS/IPSS-R are analyzed in this review article; these include age and gender, the baseline hematologic characteristics, the comorbidities, the cytogenetic and molecular profile of the patients, as well as their response to treatment with 5-azacytidine. Thus, insight into a more personalized way of managing patients with MDS is given and long-term survival is set as a more realistic goal of treatment with 5-azacytidine.

1. Introduction
Defining the prognosis of patients with myelodysplastic syndrome (MDS) has dramatically changed following a systematic effort to quan- tify the risk for acute myelogenous leukemia (AML) transformation and death. This effort resulted in the introduction of the International Prognostic Scoring System (IPSS) [1] in 1997 and was followed a few years later by the revised IPSS (IPSS-R) [2] and the World Health Or- ganization (WHO) Classification-Based PSS (WPSS) [3]. The use of these prognostic scores allowed for a better selection of patients to be treated with hypomethylating agents (HMAs). Thus, HMAs have been used since 2004 in the treatment of intermediate 2 and high-risk patients per the IPSS. These agents have altered the prognosis of MDS, offering signifi- cant increases in the progression free survival (PFS) and overall survival (OS) of the patients when compared to conventional treatment choices, such as low-dose cytarabine and best supportive care (BSC), while being less toXic compared to intensive chemotherapy [4,5]. 5-Azacytidine is a chemical pyrimidine nucleoside analog approved by the US Food and Drug Administration for the treatment of MDS and chronic myelomo- nocytic leukemia and widely used in AML [6]. 5-azacytidine is first anabolized to its nucleoside monophosphate by uridine-cytidine kinase (UCK) and is eventually incorporated into the RNA in its triphosphate

form [7]. About 80 %–90 % of 5-azacytidine embodies into the RNA, thus disrupting the nucleic acid and protein metabolism and leading to apoptosis. A smaller fraction of 10 %–20 % of the drug is converted by ribonucleotide reductase (RR) to 5-aza-2′-deoXycyridine triphosphate
(DAC-TP) and incorporates to the DNA leading to interception of DNA synthesis [7]. The phosphorylated form of 5-azacytidine binds to DNA methyltransferases, hereby inhibiting the activity of these proteins, resulting in the synthesis of hypomethylated DNA and changes in gene transcription and expression. Moreover, in vitro, 5-azacytidine leads to decondensation of the chromatin structure, chromosomal derangement, and extension of the replication time of heterochromatin [8].
Large clinical trials have showed that patients with MDS treated with 5-azacytidine achieve a median OS of around 20–24 months [4,9]. AZA-001, the pivotal trial which lead to the approval of 5-azacytidine for the treatment of higher risk MDS, was a phase III, multicenter,
controlled trial comparing 5-azacytidine to conventional care (low-dose cytarabine, intensive chemotherapy or best supportive care) in 358 patients with higher risk MDS. The median OS for the 5-azacytidine arm was 24.5 months versus 15.0 months for the conventional care arm (p 0.0001). Nevertheless, such favorable outcomes were not confirmed by all subsequent studies or in real-life settings; thus in a retrospective analysis of 1187 older patients with MDS treated with HMAs, OS

* Corresponding author at: 1st Department of Internal Medicine, Hematology Unit, National and Kapodistrian University of Athens, “Laikon” General Hospital, Athens, 11527, Greece.
E-mail address: [email protected] (P.T. Diamantopoulos).

https://doi.org/10.1016/j.leukres.2021.106543

Received 9 December 2020; Received in revised form 15 February 2021; Accepted 18 February 2021
Available online 20 February 2021
0145-2126/© 2021 Elsevier Ltd. All rights reserved.

reached only 14 months [10], while similar rates of OS were reported in smaller patient series as well [11,12]. Moreover, it seems that the range of survival is wide, with some patients achieving long remissions and high survival rates irrespective of their initial prognostic characteristics. Defining the profile of a patient with the potential to respond to treatment with 5-azacytidine and thus survive longer is rather difficult since there are very few factors correlated with response to HMAs. Nevertheless, several clinical and laboratory/molecular factors have been proposed as predictive of response to treatment. The complex ac- tions of 5-azacytidine and its interaction with several proteins such as UCK and RR have led to an effort of identifying possible biomarkers of response to the drug, mainly associated with its mechanism of action. Moreover, the OS of patients with MDS is mainly affected by the inherent features of the disease as these have been defined in the prognostic scores (cytopenias, cytogenetic risk, bone marrow blast percentage). However, due to the fact that MDS usually affects the elderly, factors such as comorbidities and poor performance status may adversely affect prognosis and should be taken into consideration when setting the goals of treatment with HMAs. With the exception of age that has been used as a cofactor in the age-adjusted IPSS-R [2], these factors have not been incorporated in any of the prognostic scores that are
widely used today, probably due to the fact that their impact cannot be
easily quantified. In the following paragraphs, we will provide published data on the clinical, hematologic, cytogenetic, and molecular charac- teristics that have been found to affect response to 5-azacytiditine and overall survival. These factors have also been listed in Tables 1 and 2.
1.1. Age and gender
Although the gender has never been included in any of the prognostic scores, a recent population study evaluating over 34,000 patients with MDS showed that male patients have a significant survival disadvantage compared to female patients [13]. The prognostic impact of age and gender was evaluated in a study on 897 untreated patients with MDS. The authors extended the prognostic model of IPSS by including gender and age (as a categorized variable) and they re-estimated the survival risks in several newly formed categories. Thus, in low-risk MDS, male
patients had a less favorable prognosis, while a particularly unfavorable outcome was found in younger (<66 years) high-risk female patients [14]. The male gender has been also shown to be an adverse prognostic
factor in a study of 203 untreated patients with MDS, but this study was conducted before the era of hypomethylating agents [15]. Although none of these studies assessed the impact of gender on the response to hypomethylating agents, a recent study proposes a differential response of male and female patients to 5-azacytidine and decitabine [16]. In this

Table 1

study of 642 patients with higher-risk MDS, it was found that female patients treated with decitabine had a higher OS than those treated with 5-azacytidine (21.1 months versus 13.2 months, p 0.0014), while no differenced was noted for male patients. Age as a prognostic factor has been evaluated and incorporated in the IPSS-R (as age-adjusted IPSS-R or IPSS-RA) [17]. The authors propose that age is an important albeit optional feature to assess survival prediction and point out that it has some prognostic influence in all risk groups, with a considerable impact in lower rather than higher risk patients. Nevertheless, in a subgroup analysis of AZA-001 comprising 87 patients 75 years, it was shown that 5-azacytidine was as effective and safe in patients over 75 years as it
was in younger patients [18]. In fact, the 2-year survival rate was 55 % versus 15 % for conventional care regimens (p < 0.001). The same conclusion was reached in a study of 102 patients, aged over 80 years, treated with 5-azacytidine [19]. Nevertheless, a recent study focusing on the value of comprehensive geriatric assessment (CGA) in 98 patients
with MDS treated with 5-azacytidine found that CGA detected geriatric-related health issues, predicted poor survival, and identified patients less likely to continue treatment, proposing that CGA should be included in the treatment decision algorithm of older patients with MDS [20].
1.2. Baseline hematologic characteristics
In a recent study evaluating factors affecting long-term survival in patients with MDS treated with 5-azacytidine, it became clear that the disease categories of the 2008 WHO classification were not correlated with the chance of long-term survival. This finding denotes that the individual subgroup of an MDS is not a significant prognostic factor for survival. Even the blast percentage as a single variable was not strong enough in predicting long-term survival [21].
Several authors have proposed the hemoglobin level, as well as the neutrophil and platelet count as single independent prognostic factors of survival in different groups of patients with MDS.
Anemia, the main clinical manifestation of MDS, seems to be multifactorial, due to chromosomal abnormalities, mitochondrial dysfunction, acquired abnormalities in the hemoglobin synthesis, and abnormal expression of cytokines and growth factors [22]. There is one study proposing that anemia per se has additive prognostic value to IPSS with regard to overall survival [23]. The authors evaluated the effect of anemia after adjustment for gender and IPSS and reported that low hemoglobin levels remained a significant predictor of OS but only in lower risk patients. The authors also suggest that the effects of chronic anemia on the heart may enhance cardiac morbidity and mortality. Moreover, transfusion dependence has already been used as a variable in

Epidemiologic and clinical/hematological factors found to affect prognosis of patients with MDS, eligibility for treatment with hypomethylating agents and treatment response.

Factor Effect on prognosis Effect on eligibility for treatment with HMAs and treatment response
Gender Male gender – survival disadvantage [13,14,15] Female patients – better response to decitabine than 5-azacytidine [16]

Age IPSS-RA – prognostic impact mainly in lower risk patients [17] Additive prognostic value to IPSS [23]
Transfusion dependence has a detrimental effect on prognosis [3,24,25]

5-azacytidine effective and safe in patients >75 years [18] and >80 years [19] Comprehensive geriatric assessment detects health issues predicting poor survival
in patients treated with 5-azacytidine [20]

Thrombocytopenia Predictive of poor OS [26] Not predictive or response to 5-azacytidine [26] Neutropenia Correlated with lower OS [27]
Monocytopenia Correlated with lower OS in univariate analysis [28]
Circulating blasts Independent prognostic factor for OS in patients treated with 5-azacyti-
MDS-CI used to detect patients with worse prognosis [32] MDS-CI useful in identifying patients’ chances to respond to 5-azacytidine [23]

Comorbidities

An eGFR<45 ml/min/1.73m2 increases the predictive value of IPSS-R in patient treated with 5-azacytidine [36]

5-azacytidine administration feasible in patients with CKD [33,34,35]

HMAs, hypomethylating agents; IPSS-RA, age adjusted international prognostic scoring system; OS, overall survival; MDS-CI, MDS comorbidity index; eGFR, estimated glomerular filtration rate; CKD, chronic kidney disease.

Table 2
Cytogenetic and molecular factors found to affect prognosis and response to hypomethylating agents in patients with MDS.
Factor Effect on prognosis Effect on treatment response

Cytogenetic abnormalities

Monosomal karyotype (MK)
Chromosome 7 abnormalities
Chromosome 17 abnormalities
Chromosome 3 abnormalities

correlated with low OS [37], especially in patients without complex karyotype [38]
Addition of MK improves IPSS-R stratification [39]

Found mainly in the context of CK and carry its poor prognosis [42]

5-azacytidine better than BSC, especially with complex karyotype [40]
Worse response to 5-azacytidine [41]

Lower response rate to 5-azacytidine [43]

Translocations Rare but correlated with lower OS [44] Impressive impact of 5-azacytidine on survival of translocation
Molecular factors
Low expression levels correlated with response to 5-azacytidine [48]
Better response to treatment [49,50,51] Not predictive of response [52,55,57]
DNMT3A mutations Independent predictor of response [51]
STAT3/5 signaling profiles Correlated with response to 5-azacytidine [52]

mRNA levels correlated with higher survival rate in patients treated with 5- azacytidine [53]

mRNA levels correlated with better response to 5-azacytidine [54]

ASXL mutations Adversely correlated with response to 5-azacytidine [55]
Correlated with high response rates to 5-azacytidine [56] Not correlated with response to 5-azacytidine [57]
Methylation level High number of methylated genes correlated with shorter OS [58] High methylation status of RRM1 correlated with response to 5-

IDO-1 IDO-1 positivity correlated with shorter OS in patients treated with 5-azacytidine
High GATA2 expression correlated to adverse prognosis in patients treated with 5-azacytidine [61]

IDO-1 positivity correlated with 5-azacytidine failure [60]
GATA1 and FLI1 mRNA expression predict response to 5-azacyti- dine [61]

sncRNAs Some expression patterns predict response to 5-azacytidin [62]
MK, monosomal karyotype; IPSS-R, revised international prognostic scoring system; BSC, best supportive care; CK, complex karyotype; UCK1, uridine-cytidine kinase- 1; TET2, ten-eleven translocation 2; STAT, signal transducer and activator of transcription; PARP1, poly (ADP-ribose) polymerase 1; RRM1, rivonucleotide reductase subunit 1.

the WPSS [3], while subsequent studies have confirmed the detrimental effect of transfusion dependence on AML transformation and overall survival, mainly in patients with lower risk MDS [24,25]. No effect of
transfusion dependence on response to 5-azacytidine has been reported so far.
Severe thrombocytopenia (i.e. < 30 109/L) has been evaluated as a
prognostic factor in a study of 225 higher-risk MDS patients treated with 5-azacytidine and although it was not predictive of response to treat- ment, it predicted poor OS [26]. EXcept for the absolute platelet count, it
has been shown that a relative platelet count drop >25 % within 6
months from diagnosis of lower-risk MDS is correlated with shorter OS (5-year OS 21.9 % versus 48.6 % for patients without such a drop, p < 0.0001) [24].
On the other hand, neutropenia has been independently correlated with lower OS in 358 patients with hypocellular MDS, along with older age, low serum albumin, and prior transfusions [27]. Furthermore, in a recent study on 976 patients with MDS, monocytopenia was correlated with lower overall survival in univariate analysis. The authors of the study suggest that an absolute monocyte count within the third quartile of values was independently predictive of longer OS compared to the remaining quartiles [28].
Finally, the prognostic significance of circulating myeloblasts was studied in 56 patients with MDS treated with 5-azacytidine and was found to be an independent prognostic factor for OS, especially in pa- tients with lower IPSS-R scores [29].

1.3. Comorbidities
In general, comorbidities have been evaluated only in a few studies and are believed not to influence success of treatment with 5-azacyti- dine, but to negatively affect OS. The need for specific comorbidity

indexes has been emphasized since the first years of the introduction of 5-azacytidine as first line treatment for MDS [30,31]. The so-called MDS comorbidity index (MDS-CI) has been used to identify patients with
worse outcome, while it has been recently shown to be useful in iden- tifying patients’ chances to respond to 5-azacytidine [32].
As mentioned earlier, anemia has been proposed as an independent
prognostic factor for patients with MDS. It should also be kept in mind that anemia of chronic kidney disease is rather common in the elderly and may affect survival. Although renal impairment was not addressed as a prognostic or limiting factor for the use of 5-azacytidine in the approval trials of the drug, there are a few studies supporting the feasibility of treatment with 5-azacytidine in patients with chronic kidney disease [33,34], even in patients under renal replacement ther- apy [35]. The incidence of toXicities may be higher in patients with severe renal impairment, thus dose reductions and the use of growth factors is recommended by the authors. Moreover, a recent study assessing the estimated glomerular filtration rate (eGFR) as a prognostic
factor in 536 patients with MDS treated with 5-azacytidine found that an eGFR <45 ml/min/1.73 m2 independently predicts worse response and lower OS. The authors proposed that the addition of eGFR <45 ml/min/1.73 m2 in the IPSS-R increased the predictive value of the
model [36]. With the exception of renal impairment, there are no studies focusing on the impact of individual comorbidities on response to treatment and OS.

1.4. Cytogenetics
The cytogenetic risk is considered as the most important feature of the IPSS-R and it has been found to be the strongest single-variable predictor of lower OS in large scale clinical trials. There are several studies focusing on the impact of individual cytogenetic abnormalities

on the prognosis of patients with MDS.
The prognostic impact of monosomal karyotype (MK) has been investigated in three studies. In the first study, MK was identified in 16.3
% of 405 higher-risk MDS patients and was correlated with reduced OS. Furthermore, MK was found to distinguish patients with worse prognosis among those with high and very high IPSS-R [37]. In the second study of 243 patients treated with 5-azacytidine, MK was identified in 90 patients and was an important prognostic parameter of worse outcome only in patients without a complex karyotype [38]. Moreover, reclassification of 154 patients treated with 5-azacytidine based on the IPSS-R and the presence of MK showed that MK improves IPSS-R stratification [39].
The prognostic impact of chromosome 7 abnormalities has been investigated in 235 high-risk MDS patients treated with 5-azacytidine or BSC, showing that 5-azacytidine had favorable impact on OS compared to BSC, especially in patients with complex karyotype [40]. In another study of 113 patients treated with 5-azacytidine, the presence of chro- mosome 7 abnormalities and 17p deletion were associated with worse response to the drug [41].
The prognostic impact of chromosome 17 abnormalities was studied in 548 MDS patients treated with 5-azacytidine. Chromosome 17 ab- normalities were identified in 32 patients and correlated with poor prognostic features and low OS, but the majority of them were found in the context of a complex karyotype. The authors concluded that chro- mosome 17 abnormalities carry the poor prognosis of complex karyo- type [42].
Finally, chromosome 3 abnormalities were found to be correlated with low response rates to 5-azacytidine and poor outcomes in a cohort of 413 patients with MDS, MDS/MPN, and AML treated with 5-azacyti- dine [43].
On the other hand, chromosomal translocations are rare in MDS, but in one study on 751 patients, they were identified in 5.3 % of them and were independently correlated with lower OS, while treatment with 5- azacytidine had an impressive positive impact on survival of trans- location carriers, whereas this survival difference was not observed for patients without translocations. Moreover, no correlation of specific chromosomal translocations with response to treatment were found [44].
1.5. Molecular factors
Clonal hematopoiesis (CH) is a common age-related phenomenon referring to the emergence of clones with distinct acquired mutations of leukemia-associated genes, such as DNMT3A, TET2, ASXL1, and others in otherwise healthy persons. The emergence of cytopenias in patients with CHIP defines clonal cytopenia of uncertain significance (CCUS). Both these conditions must be differentiated from MDS, since they are not characterized by dysplasia or excess blast counts, although they present a small but considerable potential for progression to MDS, varying with the specific mutations, the number of mutations, and the variant allele frequency [45]. In patients with MDS, the underline mo- lecular architecture is correlated with several of the aforementioned factors such as the age and the gender of the patient. For example, DNTT3A mutations are more common in older patients [46], while SRSF2 and U2AF1 mutations are strongly associated with male gender [47]. Moreover, except for isolated del(5q), there is a male predomi- nance for most MDS categories. There is extensive interest on the mo- lecular characteristics of patients with MDS and their impact on prognosis. Several genes have been found to be mutated in patients with MDS and some of these mutations have been correlated with response to treatment and overall survival. Nevertheless, the accumulation of data on the identification of a solid biomarker of response to 5-azacytidine
has not yet been fruitful enough, thus no widely accepted biomarkers
exist to date.
Low levels of expression of UCK1 have been found to be correlated with response to 5-azacytidine [48], while ten-eleven-translocation 2 (TET2) mutations have been correlated to better treatment response in 2

studies [49,50], and TET2 and DNMT3A mutations were found to be independent predictors of response in another study [51]. Moreover, the pretreatment signaling profiles of Signal Transducer and Activator of Transcription 3 and 5 (STAT3/5) in CD34 cells were correlated with response to 5-azacytidine and independently predicted, event-free sur- vival, while in the same study TET2 mutations were not predictive of response [52]. Finally, high (ADP-ribose) polymerase 1 (PARP1) mRNA levels have been correlated with better response [53] and higher sur- vival rates after 5-azacytidine initiation [54].
In a study of 121 patients with MDS treated with an HMA, ASXL mutations were adversely correlated with response to HMAs, while contrary to other reports, no correlation was found between TET2 mu- tations and response to HMAs [55]. In 128 patients with MDS or AML treated with 5-azacytidine, none out of fourteen commonly encountered gene mutations was correlated to response, while in high-risk groups TP53 mutations were correlated with a high response rate [56]. In another study on 39 patients with high-risk MDS and secondary AML treated with 5-azacytidine, 95 % of the patients harbored at least one mutation. No correlations of TET2 and TP53 mutations were found with response to 5-azacytidine, but patients with TP53 mutations had shorter OS in univariate analysis [57].
The number of methylated genes as a prognostic factor has been evaluated in a study of 63 patients with MDS and AML treated with 5- azacytidine. The presence of a high number ( 2) of methylated genes was independently correlated with shorter OS, but the aberrant methylation status did not correlate with response to treatment [58]. Moreover, in a study of 98 patients with MDS treated with 5-azacytidine, high methylation status of RR subunit 1 (RRM1) was observed in re- sponders, thus making RNR a possible biomarker for predicting success of epigenetic treatment [59].
Finally, there are several recent studies assessing the prognostic role of other molecules, not directly correlated with the pathogenesis of MDS. Thus, the immunomodulator indoleamine-2,3-dioXygenase (IDO- 1) was studied in a cohort of 95 patients with MDS, secondary AML, and MDS/MPN treated with 5-azacytidine. The authors reported that IDO-1 positivity, found in more than one third of the patients, was correlated with treatment failure and shorter OS. The rationale behind this finding seems to be that IDO-1 expression induces an immunosuppressive microenvironment leading to 5-azacytidine failure [60]. The prognostic role of the transcription factors GATA1, GATA2, and FLI1 mRNA expression was studied in 56 patients with MDS and low blast count AML treated with 5-azacytidine. High GATA2 expression was found to be an adverse prognostic factor for OS while GATA1 and FLI1 mRNA expres- sion may predict response to treatment [61]. Finally, in a recent study, it was shown that circulating, small, noncoding RNAs (sncRNAs) have specific expression patterns in the plasma of patients with MDS related to the presence of somatic mutations in the SF3B1 and DNMT3A genes and that the combined expression of some of them can predict response to 5-azacytidine as well as define survival [62].
1.6. Response to treatment
Response to 5-azacytidine has been shown to be a decisive factor of prognosis. The impact of response to 5-azacytidine has been assessed in the AZA-001 study. Achievement of a response (complete remission, partial remission, hematologic improvement) was associated with an 84
% reduced risk of death (HR 0.16, 95 % Confidence Interval, CI,
0.07 0.37, p < 0.0001). Moreover, the median OS was not reached for patients responding at 3-month, 6-month, and 9-month landmarks, and the median 2-year survival rate for responders was 83 % [63].
Recent reports, though, imply that stable disease (SD) is not as bad as previously considered. Even in the AZA-001 study that led to the approval of 5-azacytidine in higher risk MDS, patients with stable dis- ease as an initial response achieved a better response with subsequent treatment cycles and this was correlated with a positive impact on sur- vival [9]. Moreover, in a study on 291 patients treated with

hypomethylating agents, 20 % of patients achieving an SD at 4–6 months achieved a better response at a later time point with continuing treatment [64]. Even so, the outcome of patients achieving an SD is
largely unclear. A recent study from our center on 353 patients with higher risk MDS showed that 51.6 % of patients achieving an SD and continuing treatment with 5-azacytidine showed a lower risk of AML transformation and increased OS, compared to patients with SD who discontinued treatment [65]. No studies focusing on the prognostic significance of hematologic improvement as a response to treatment with hypomethylating agents could be found in the literature.
Finally, it has been recently shown in a study of 93 patients with MDS that the relative dose intensity of treatment with 5-azacyditine affects survival, especially if dose reductions are undertaken before the achievement of an objective response [66]. The effect of treatment de- lays and dose reductions on the prognosis of patients with MDS treated with 5-azacytidine was assessed in a large study of 897 patients. The authors reported that delays during the first two cycles of treatment adversely affected survival independently of the IPSS while delays before achieving an objective response adversely affected response to treatment [67].
2. Conclusions
The aforementioned factors affect prognosis irrespective of the already established prognostic scoring systems and may help the clini- cian to define the prognosis of individual patients based on the accu- mulation of adverse or favorable prognostic factors. Based on the IPSS, IPSS-R and WPSS and taking into consideration the age, the comorbid- ities, the cytogenetic and molecular profile, the clinician can set the goals of treatment, give an estimation of the prognosis, and define realistic expectations for the patient. The incorporation of fluorescent in situ hybridization and next generation sequencing in the investigation of patients with MDS may help the treating physician more accurately estimate the prognosis of the patient, define the goals of treatment with HMAs, and decide further actions such as proceeding with allogeneic stem cell transplantation or including the patient in a clinical trial, either upfront or after 5-azacytidine failure. Moreover, since aberrant methylation is a central mechanism in the pathogenesis and progression to AML, methylation assays may play in the future an important role in treatment choice for patients with MDS. However, data on DNA methylation analysis to discover unknown epigenetic changes or to assess methylation of particular regulatory genes of interest is still young and largely inconclusive.
Finally, the available data on the effect of clinical and/or genetic
factors on response to 5-azacytidine is often contradictory, so that no clear-cut guidelines can be given to clinicians about the decision of administering 5-azacytidine based on the baseline characteristics of the patient and/or disease. The clinician should be aware though that pa- tients with adverse baseline prognostic factors such as a complex or monosomal karyotype, chromosome 7 or 3 abnormalities, TP53 and ASXL mutations, or even low levels of PARP1 expression should be managed more aggressively, preferably in the context of clinical trials. The investigation of additional clinical and laboratory factors affecting response to 5-azacytidine is a fertile field with ongoing research on the molecular factors that define the chances of response to 5-azacytidine. Moreover, since HMAs are now administered in combination with other agents such as lenalidomide, the bcl2 inhibitor venetoclax, the anti-CD47 monoclonal antibody magrolimab, the programmed cell death protein 1 (PD1) inhibitors nivolumab and pembrolizumab and its ligand (PDL1) inhibitors atezolizumab and durvalumab, the selective NEDD8 inhibitor pevonedistat and others, mainly in the context of clinical trials, new prognostic markers may emerge through the gath- ering of information on the synergistic action of HMAs with other agents, especially since molecular profiling of patients with MDS is a common practice both in clinical trials and in real-world settings.

Funding
No funding sources were used.

Availability of data and material
Not applicable

Authors’ contributions
PD conceived the idea and drafted the manuscript NAV critically reviewed the manuscript

Declaration of Competing Interest
PD and NAV report personal fees for honoraria from Genesis Pharma, Roche, Sandoz, and Novartis.

Acknowledgment
We would like to thank Mrs. Evita Alexopoulos for copy-editing the final manuscript.

References
[1] P. Greenberg, C. CoX, M.M. LeBeau, et al., International scoring system for
evaluating prognosis in myelodysplastic syndromes, Blood 89 (6) (1997)
2079–2088.
[2] P.L. Greenberg, H. Tuechler, J. Schanz, et al., Revised international prognostic scoring system for myelodysplastic syndromes, Blood 120 (12) (2012) 2454–2465.
[3] L. Malcovati, U. Germing, A. Kuendgen, et al., Time-dependent prognostic scoring
system for predicting survival and leukemic evolution in myelodysplastic syndromes, J. Clin. Oncol. 25 (23) (2007) 3503–3510.
[4] L.R. Silverman, E.P. Demakos, B.L. Peterson, et al., Randomized controlled trial of azacitidine in patients with the myelodysplastic syndrome: a study of the cancer
and leukemia group B, J. Clin. Oncol. 20 (10) (2002) 2429–2440.
[5] H. Kantarjian, J.P. Issa, C.S. Rosenfeld, et al., Decitabine improves patient outcomes in myelodysplastic syndromes: results of a phase III randomized study,
Cancer 106 (8) (2006) 1794–1803.
[6] J.R. Kuykendall, 5-azacytidine and decitabine monotherapies of myelodysplastic
disorders, Ann. Pharmacother. 39 (10) (2005) 1700–1709.
[7] J. Aimiuwu, H. Wang, P. Chen, et al., RNA-dependent inhibition of ribonucleotide reductase is a major pathway for 5-azacytidine activity in acute myeloid leukemia,
Blood 119 (22) (2012) 5229–5238.
[8] K. Raj, G.J. Mufti, Azacytidine (Vidaza®) in the treatment of myelodysplastic
syndromes, Ther. Clin. Risk Manag. 2 (4) (2006) 377–388.
[9] P. FenauX, G.J. Mufti, E.A. Hellstrom-Lindberg, et al., Efficacy of azacitidine compared with that of conventional care regimens in the treatment of higher-risk
myelodysplastic syndromes: a randomised, open-label, phase III study, Lancet Oncol. 10 (2009) 223–232.
[10] A.M. Zeidan, M. Stahl, X. Hu, et al., Long-term survival of older patients with MDS treated with HMA therapy without subsequent stem cell transplantation, Blood 131
(7) (2018) 818–821.
[11] R. Itzykson, S. The´pot, B. Quesnel, et al., Prognostic factors for response and overall survival in 282 patients with higher-risk myelodysplastic syndromes
treated with azacitidine, Blood 117 (2) (2011) 403–411.
[12] R. Itzykson, S. The´pot, B. Quesnel, et al., Long-term outcome of higher-risk MDS
patients treated with azacitidine: an update of the GFM compassionate program cohort, Blood 119 (25) (2012) 6172–6173.
[13] F. Wang, J. Ni, L. Wu, Y. Wang, B. He, D. Yu, Gender disparity in the survival of
patients with primary myelodysplastic syndrome, J. Cancer 10 (5) (2019) 1325–1332.
[14] T. No¨sslinger, H. Tüchler, U. Germing, et al., Prognostic impact of age and gender
in 897 untreated patients with primary myelodysplastic syndromes, Ann. Oncol. 21 (1) (2010) 120–125.
[15] P. Morel, C. Declercq, M. Hebbar, F. Bauters, P. FenauX, Prognostic factors in myelodysplastic syndromes: critical analysis of the impact of age and gender and
failure to identify a very-low-risk group using standard mortality ratio techniques, Br. J. Haematol. 94 (1) (1996) 116–119.
[16] A.E. DeZern, A.M. Zeidan, J. Barnard, et al., Differential response to hypomethylating agents based on sex: a report on behalf of the MDS Clinical
Research Consortium (MDS CRC), Leuk. Lymphoma 58 (6) (2017) 1325–1331.
[17] P.L. Greenberg, H. Tuechler, J. Schanz, et al., Revised international prognostic scoring system for myelodysplastic syndromes, Blood 120 (12) (2012) 2454–2465.
[18] J.F. Seymour, P. FenauX, L.R. Silverman, et al., Effects of azacitidine compared with conventional care regimens in elderly ( 75 years) patients with higher-risk myelodysplastic syndromes, Crit. Rev. Oncol. Hematol. 76 (3) (2010) 218–227.

[19] S.G. Papageorgiou, C.K. Kontos, I. Kotsianidis, et al., Effectiveness of 5-Azacytidine in older patients with high-risk myelodysplastic syndromes and oligoblastic acute myeloid leukemia: a retrospective analysis of the Hellenic (Greek) MDS Study
Group, J. Geriatr. Oncol. 11 (1) (2020) 121–124.
[20] A. Molga, M. Wall, R. Chhetri, et al., Comprehensive geriatric assessment predicts azacitidine treatment duration and survival in older patients with myelodysplastic
syndromes, J. Geriatr. Oncol. 11 (1) (2020) 114–120.
[21] P.T. Diamantopoulos, V. Pappa, A. Symeonidis, et al. Characteristics of long-term survivors with Myelodysplastic syndrome treated with 5-azacyditine. Results from the Hellenic 5-azacytidine Registry. https://doi.org/10.1016/j.clml.2019.09.614.
[22] L. Malcovati, M.G. Della Porta, C. Strupp, et al., Impact of the degree of anemia on the outcome of patients with myelodysplastic syndrome and its integration into the WHO classification-based Prognostic Scoring System (WPSS), Haematologica. 96
(10) (2011) 1433–1440.
[23] J.M. Kao, A. McMillan, P.L. Greenberg, International MDS risk analysis workshop
(IMRAW)/IPSS reanalyzed: impact of cytopenias on clinical outcomes in myelodysplastic syndromes, Am. J. Hematol. 83 (10) (2008) 765–770.
[24] R. Itzykson, S. Crouch, E. Travaglino, et al., Early platelet count kinetics has
prognostic value in lower-risk myelodysplastic syndromes, Blood Adv. 2 (16) (2018) 2079–2089.
[25] E. Kaphan, D. Laurin, B. Lafeuillade, P. Drillat, S. Park, Impact of transfusion on survival in patients with myelodysplastic syndromes: current knowledge, new insights and transfusion clinical practice, Blood Rev. 41 (2020), 100649.
[26] A. Galanopoulos, E. Kamouza, C. Kontos, et al., Prognostic significance of severe thrombocytopenia in overall survival of patients with myelodysplastic syndromes
treated with azacytidine. A multicenter study by the Hellenic MDS study group [abstract], in: In: 60th Annual Meeting of the American Society of Hematology,
Dec, 2018. Abstract nr 1822.
[27] J. Jungmin, B. Sung Hwa, P. Jinny, et al., Prognostic factors and response to the hypomethylating agents in hypocellular myelodysplastic syndromes: a retrospective study from the Korean Society of Hematology AML/MDS working party [abstract], in: 59th Annual Meeting of the American Society of Hematology, Dec, 2017.
[28] T. Silzle, E. Schuler, J. Kaivers, et al., Absolute monocyte count affects prognosis in myelodysplastic syndromes independently from the IPSS-R. [abstract], in: 24Th
Congress of the European Hematology Association, Jun 15, 2019, 266957; PS1340.
[29] T. Kamiya, T. Nakazato, The prognostic value of circulating myeloblasts in patients with myelodysplastic syndromes treated with azacitidine, Med. Oncol. 36 (3) (2019) 25.
[30] A. Sanna, A. Gozzini, D. Gioia, et al., Comorbidities influence prognosis in MDS high-risk patients treated with 5-azacitidine [abstract], 16th Congress of the European Hematology Association (2011).
[31] A. Sanna, L. Cannella, A. Gozzini, et al., Comorbidities indexes in patients treated
with 5-azacitidine are a useful and easily applicable tool to refine prognostic
evaluation [abstract], in: 52Nd Annual Meeting of the American Society of Hematology, Dec, 2010.
[32] M. Breccia, A. Salaroli, G. Loglisci, P. Finsinger, A. Serrao, G. Alimena, MDS- specific comorbidity index is useful to identify myelodysplastic patients who can have better outcome with 5-azacitidine, Haematologica 97 (2) (2012) e2.
[33] G.N. Batty, H. Kantarjian, J.P. Issa, et al., Feasibility of therapy with hypomethylating agents in patients with renal insufficiency, Clin. Lymphoma
Myeloma Leuk. 10 (3) (2010) 205–210.
[34] E. Douvali, M. Papoutselis, T.P. Vassilakopoulos, et al., Safety and efficacy of 5- azacytidine treatment in myelodysplastic syndrome patients with moderate and
mild renal impairment, Leuk. Res. 37 (8) (2013) 889–893.
[35] T. Yoshihiro, T. Muta, K. Aoki, S. Shimamoto, Y. Tamura, R. Ogawa, Efficacy and
adverse events of azacitidine in the treatment of hemodialysis patients with high- risk myelodysplastic syndrome, Rinsho Ketsueki 57 (8) (2016) 1004–1010.
[36] V. Papadopoulos, P.T. Diamantopoulos, S.G. Papageorgiou, et al., Estimated glomerular filtration rate independently predicts outcome of azacitidine therapy in higher-risk Myelodysplastic syndromes. Results from 536 patients of the Hellenic
National Registry of Myelodysplastic and Hypoplastic syndromes, Hematol. Oncol. 38 (4) (2020) 541–553.
[37] S.G. Papageorgiou, D. Vasilatou, C.K. Kontos, et al., The prognostic value of monosomal karyotype (MK) in higher-risk patients with myelodysplastic syndromes treated with 5-Azacitidine: a retrospective analysis of the Hellenic
(Greek) Myelodysplastic syndromes Study Group, Am. J. Hematol. 93 (7) (2018) 895–901.
[38] K.L. Hwang, M.K. Song, H.J. Shin, et al., Monosomal and complex karyotypes as prognostic parameters in patients with International Prognostic Scoring System higher risk myelodysplastic syndrome treated with azacitidine, Blood Res. 49 (4)
(2014) 234–240.
[39] T. Cluzeau, N. Mounier, J.M. Karsenti, et al., Monosomal karyotype improves IPSS-
R stratification in MDS and AML patients treated with Azacitidine, Am. J. Hematol. 88 (9) (2013) 780–783.
[40] M. Díez-Campelo, J.I. Lorenzo, R. Itzykson, et al., Azacitidine improves outcome in higher-risk MDS patients with chromosome 7 abnormalities: a retrospective comparison of GESMD and GFM registries, Br. J. Haematol. 181 (3) (2018)
350–359.
[41] I. Martín, B. Navarro, A. Serrano, et al., Impact of clinical features, cytogenetics,
genetic mutations, and methylation dynamics of CDKN2B and DLC-1 promoters on treatment response to azacitidine, Ann. Hematol. 99 (3) (2020) 527–537.
[42] P. Diamantopoulos, D. Koumbi, I. Kotsianidis, et al., The prognostic significance of chromosome 17 abnormalities in patients with myelodysplastic syndrome treated

with 5-azacytidine: results from the Hellenic 5-azacytidine registry, Cancer Med. 8 (5) (2019) 2056–2063.
[43] D.A. Sallman, J. Barnard, N.H. Al Ali, et al., Hypomethylating agent therapy in
myelodysplastic syndromes with chromosome 3 abnormalities, Clin. Lymphoma Myeloma Leuk. 20 (9) (2020) e597–e605.
[44] K.J. Suh, J.W. Cheong, I. Kim, et al., Prognostic impact of IPSS-R and chromosomal translocations in 751 korean patients with primary myelodysplastic syndrome, PLoS One 11 (11) (2016), e0166245.
[45] R. Bejar, CHIP, ICUS, CCUS and other four-letter words, Leukemia 31 (9) (2017) 1869–1871.
[46] M.J. Walter, L. Ding, D. Shen, et al., Recurrent DNMT3A mutations in patients with
myelodysplastic syndromes, Leukemia 25 (7) (2011) 1153–1158.
[47] X. Wang, X. Song, X. Yan, Effect of RNA splicing machinery gene mutations on prognosis of patients with MDS: a meta-analysis, Medicine (Baltimore) 98 (21) (2019), e15743.
[48] A. Valencia, E. Masala, A. Rossi, et al., EXpression of nucleoside-metabolizing enzymes in myelodysplastic syndromes and modulation of response to azacitidine,
Leukemia 28 (3) (2014) 621–628.
[49] R. Itzykson, O. Kosmider, T. Cluzeau, et al., Impact of TET2 mutations on response rate to azacitidine in myelodysplastic syndromes and low blast count acute
myeloid leukemias, Leukemia 25 (7) (2011) 1147–1152.
[50] R. Bejar, A. Lord, K. Stevenson, et al., TET2 mutations predict response to
hypomethylating agents in myelodysplastic syndrome patients, Blood 124 (17) (2014) 2705–2712.
[51] F. Traina, V. Visconte, P. Elson, et al., Impact of molecular mutations on treatment
response to DNMT inhibitors in myelodysplasia and related neoplasms, Leukemia 28 (1) (2014) 78–87.
[52] P. Miltiades, E. Lamprianidou, T.P. Vassilakopoulos, et al., The Stat3/5 signaling biosignature in hematopoietic stem/progenitor cells predicts response and outcome in myelodysplastic syndrome patients treated with azacitidine, Clin.
Cancer Res. 22 (8) (2016) 1958–1968.
[53] P.T. Diamantopoulos, C.N. Kontandreopoulou, A. Symeonidis, et al., Bone marrow
PARP1 mRNA levels predict response to treatment with 5-azacytidine in patients with myelodysplastic syndrome, Ann. Hematol. 98 (6) (2019) 1383–1392.
[54] P. Diamantopoulos, K. Zervakis, P. Zervakis, et al., Poly (ADP-ribose) polymerase 1 mRNA levels strongly correlate with the prognosis of myelodysplastic syndromes, Blood Cancer J. 7 (2) (2017) e533.
[55] D. Idossa, T.L. Lasho, C.M. Finke, et al., Mutations and karyotype predict treatment response in myelodysplastic syndromes, Am. J. Hematol. 93 (11) (2018)
1420–1426.
[56] A. Kuendgen, C. Müller-Thomas, M. Lauseker, et al., Efficacy of azacitidine is independent of molecular and clinical characteristics - an analysis of 128 patients
with myelodysplastic syndromes or acute myeloid leukemia and a review of the literature, Oncotarget 9 (45) (2018) 27882–27894.
[57] M. Cabezo´n, J. Bargay, B. Xicoy, et al., Impact of mutational studies on the
diagnosis and the outcome of high-risk myelodysplastic syndromes and secondary acute myeloid leukemia patients treated with 5-azacytidine, Oncotarget 9 (27) (2018) 19342–19355.
[58] M. Abaigar, F. Ramos, R. Benito, et al., Prognostic impact of the number of
methylated genes in myelodysplastic syndromes and acute myeloid leukemias treated with azacytidine, Ann. Hematol. 92 (11) (2013) 1543–1552.
[59] C.N. Kontandreopoulou, P.T. Diamantopoulos, S. Symeonidis, et al., Bone marrow ribonucleotide reductase mRNA levels and methylation status as a prognostic factor in patients with myelodysplastic syndrome treated with 5-Azacytidine, Blood 134 (Supplement_1) (2019) 1721.
[60] C. Müller-Thomas, M. Heider, G. Piontek, et al., Prognostic value of indoleamine
2,3 dioXygenase in patients with higher-risk myelodysplastic syndromes treated with azacytidine, Br. J. Haematol. 190 (3) (2020) 361–370.
[61] G. Falconi, E. Fabiani, M. Criscuolo, et al., Transcription factors implicated in late megakaryopoiesis as markers of outcome after azacitidine and allogeneic stem cell transplantation in myelodysplastic syndrome, Leuk. Res. 84 (2019), 106191.
[62] A. Hrustincova, Z. Krejcik, D. Kundrat, et al., Circulating small noncoding RNAs have specific expression patterns in plasma and extracellular vesicles in myelodysplastic syndromes and are predictive of patient outcome, Cells 9 (4) (2020) 794.
[63] S.D. Gore, P. FenauX, V. Santini, et al., A multivariate analysis of the relationship between response and survival among patients with higher-risk myelodysplastic syndromes treated within azacitidine or conventional care regimens in the
randomized AZA-001 trial, Haematologica 98 (7) (2013) 1067–1072.
[64] A. Nazha, M.A. Sekeres, G. Garcia-Manero, et al., Outcomes of patients with myelodysplastic syndromes who achieve stable disease after treatment with
hypomethylating agents, Leuk. Res. 41 (2016) 43–47.
[65] S.G. Papageorgiou, C.K. Kontos, I. Kotsianidis, et al., The outcome of patients with high-risk MDS achieving stable disease after treatment with 5-Azacytidine. A retrospective analysis of the Hellenic (Greek) MDS Study Group, Hematol. Oncol.
36 (4) (2018) 693–700.
[66] K. Laribi, D. Bolle, M. Alani, et al., Impact of the relative dose intensity on survival of patients with high-risk myelodysplastic syndromes treated with Azacitidine,
Cancer Med. 8 (5) (2019) 2188–2195.
[67] P.T. Diamantopoulos, A. Symeonidis, V. Pappa, et al., The effect of 5-azacytidine treatment delays and dose reductions on the prognosis of patients with myelodysplastic syndrome: how to optimize treatment results and outcomes [published online ahead of print, 2020 Aug 30], Br. J. Haematol. (2020), 10.1111/ bjh.17062.