Efficacy and safety of an HDACi- and HMA-based protocol in adults with acute myeloid leukemia of intermediate- and adverse-risk categories: a retrospective study

ABSTRACT

Objective

Anthracyclines and cytarabine have comprised standard induction therapy for acute myeloid leukemia (AML) for decades. Low overall survival of AML is due to non-remission or relapse after remission. Hypomethylating agent (HMA) decitabine combined with low-dose chemotherapy or other targeted agents has shown promising effect for AML in clinical trials, especially in t(8;21) acute myeloid leukemia. We previously investigated histone deacetylase inhibitor (HDACi) chidamide could regulate Wnt/β-catenin signaling pathway in leukemia cell lines.

Methods

Adult patients with de novo or relapsed/refractory AML who were treated with chidamide and decitabine in combination with chemotherapy (chidamide group, n = 23) or only decitabine combination with chemotherapy (decitabine group, n = 17) were analyzed.

Results

Chidamide group represented higher complete response rate (82.6% and 52.9%, p = 0.0430, vs. decitabine group), progression-free survival and overall survival rates (p = 0.0088 and p = 0.0139, respectively), especially for patients with de novo AML. Hematological toxicity and infections were the most common adverse events (AEs) in both groups, and they were manageable by supportive treatments.

Conclusions

This HDACi- and HMA-based protocol is an effective and tolerable therapy for patients with AML. The comprehensive mechanism and effects of chidamide in combination with decitabine are worth to be further explored in AML.

KEYWORDS:

• Acute myeloid leukemia
• HDACi
• HMA
• chidamide
• decitabine
• chemotherapy
• combination therapy

Introduction

Acute myeloid leukemia (AML, not acute promyelocytic leukemia [non-APL]), which is a most common adult hematological malignancy, is considered to be probably caused by abnormal genetic or epigenetic regulation of hematopoietic stem cells [1,2]. Anthracyclines and cytarabine have comprised the standard AML induction therapy for decades. More than 80% of young patients with AML could achieve complete responses (CRs) [3], whereas only 24–29.5% of adult patients could achieve 5-year overall survival due to non-remission or relapse after remission [4–7]. Further exploration is strongly encouraged to increase overall response rate (ORR), as well as to seek opportunities for consolidation treatments or hematopoietic stem cell transplantation (HSCT).

Genomic and epigenetic alterations participate in the occurrence and progression of hematological malignancies [8–11]. Hypomethylating agent (HMA) decitabine works by binding to DNA methyltransferase [12], thus specifically inducing cell cycle arrest, activating tumor suppressor genes, and inhibiting the proliferation of tumor cells [13,14]. Mutations such as DNMT3A, TET2 are related to functions of DNA methylation, and HMA has shown efficacy for patients with mutation mentioned above [15]. Decitabine has been approved as a favorable agent by the Food and Drug Administration for clinical treatments of AML (combined with venetoclax for patients intolerant for intense chemotherapy), myelodysplastic syndromes. Clinical trials related to decitabine combined with chemotherapy (such as FLAG, DA and CAG) for newly diagnosed or relapsed/refractory AML (R/R AML) achieved better outcomes [16–18].

Histone deacetylases (HDACs) are essential for maintaining the balance of histone acetylation, which induces gene transcriptional repression through chromatin condensation, and the overexpression of HDACs is implicated in oncogenesis [19,20]. HDAC inhibitors (HDACis) have emerged as new therapeutic options that modulate anti-tumor mechanisms and exert anti-tumor effects synergistically with other therapeutic protocols [21–23]. Chidamide (CS055/HBI-8000) is a subtype-selective HDACi with the ability to inhibit HDAC1, HDAC2, HDAC3, and HDAC10. Chidamide could target leukemia cells by inhibiting Wnt/β-catenin, ERK1/2 and JAK2/STAT3 signaling pathway [24–26], which could also synergize with decitabine to inhibit the proliferation of AML cells and significantly enhance the cytotoxicity in AML cells [27–30]. We conducted a retrospective study on chidamide and decitabine combined with chemotherapy in adults with AML and evaluated the efficacy and safety of this therapeutic regimen.

Materials and methods

Patients and treatments

We retrospectively analyzed treatment efforts on de novo or relapsed adult AML (non-APL) patients treated between December 2016 and October 2019 in the Department of Hematology, First Affiliated Hospital of Harbin Medical University, who have undergone a complete inductive course with evaluable efficacy. The diagnosis was made according to 2016 World Health Organization AML criteria [31]. Due to the fact that none of these patients were given FLT3 inhibitors, FLT-3 mutation was still seemed as adverse prognosis, and risk stratification was conducted according to the 2017 European Leukemia Net (ELN) recommendations for diagnosis and management of adults with AML [32]. This study was approved by ethics committees of the First Affiliated Hospital of Harbin Medical University (Approval No.201609).

The regimens were chidamide and decitabine combined with chemotherapy (based on anthracyclines and cytarabine, shown in Sup Table 1,2). Therapeutic protocols: (1) Chidamide group: oral chidamide given at 30 mg twice a week for 2 weeks starting on day 1, decitabine administered at 20 mg/(m²·day) via intravenous infusion on days 1–5, and combination chemotherapy; or (2) Decitabine group: decitabine administered at 20 mg/(m²·day) via intravenous infusion on days 1–5 and combination chemotherapy. Both protocols were followed by consolidation therapies and symptomatic therapy and supportive care were adequately provided.

Assessment of efficacy and safety

Treatment responses and outcomes were evaluated at the completion of each cycle on the basis of NCCN clinical practice guidelines [33]. The primary efficacy evaluation included CR, CR with incomplete blood count recovery (CRi), partial response (PR), and ORR. ORR was calculated as CR+CRi+PR. The secondary efficacy evaluation included progression-free survival (PFS), duration of response (DOR), and overall survival (OS). MRD and gene mutations were measured by flow cytometry (FCM) and next-generation sequencing, respectively.

The incidence and severity of adverse events (AEs) were monitored and recorded by the investigators throughout the process. Toxicities were evaluated and graded using the National Cancer Institute Common Terminology Criteria for Adverse Events version 5.0 and compared between two groups. Post-therapy follow-up and detailed examinations of patients with AML were performed every 3 months until November 20, 2019.

Statistical analysis

All statistical analyses were conducted using SAS software version 9.3. When data followed a normal distribution, Student’s t-test was performed to compare quantitative variables between both groups. Otherwise, quantitative variables were compared using the rank sum test. PFS and OS were estimated using Kaplan–Meier method, and differences were compared by applying the log-rank test. The hazard ratio (HR) was estimated using Cox proportional-hazards model, and survival outcomes were presented as HRs with 95% confidence intervals (CIs), with statistical significance evaluated at 0.05 alpha level.

Results

Patient characteristics

We enrolled 40 AML patients between December 2016 and October 2019 in this study, including 26 men and 14 women with a median age of 46.2 ± 15.7 years. Twenty-three patients (23/40, 57.5%) were treated with chidamide and decitabine combined with chemotherapy (chidamide group), and 17 patients (17/40, 42.5%) received decitabine and chemotherapy (decitabine group). Baseline characteristics of patients in both groups are listed in Sup Table 3. In total, 22 patients were newly diagnosed with AML (22/40, 55.0%), and 18 patients with R/R AML (18/40, 45.0%). Five patients (5/40, 12.5%) were intermediate-risk and 35 patients (35/40, 87.5%) were adverse risk according to the 2017 ELN recommendations. Of 18 patients with R/R AML, 6 (6/18, 33.3%) had received other salvage therapeutic regimens (e.g. FLAG, CAG, HMA) for R/R AML before the study. Excluding sex, age, and disease status, there were no significant differences in baseline characteristics between two groups (p > 0.05). Among 32 patients with accessible cytogenetic analysis, 16 patients were evaluated as intermediate risk, and other 16 patients were adverse risk (detailed in Sup Table 3). Gene mutations, including FLT3-ITD (12/33, 36.4%), DNMT3A (8/33, 24.2%), NPM1 (7/33, 21.2%), CEBPA (7/33, 21.2%), IDH2 (6/33, 18.2%), NRAS (6/33, 18.2%), TET2 (5/33, 15.2%), KRAS (5/33,15.2%), RUNX1 (5/33, 15.2%), GATA2 (3/33, 9.1%), and ASXL1 mutations (2/33, 6.1%), were detected in 33 patients (Sup Fig. 1a). FLT3-ITD, NPM1, NRAS, DNMT3A, CEBPA, IDH2, and RUNX1 mutations were mainly identified in patients with newly diagnosed AML, whereas FLT3-ITD, DNMT3A, TET2, CEBPA, NPM1, IDH2, and GATA2 mutations were more common in patients with R/R AML (Sup Fig. 1b). FLT3-ITD, DNMT3A, NPM1, CEBPA, IDH2, NRAS, TET2, and RUNX1 mutations were present in both groups before treatment (Sup Fig. 1c).

Treatment responses

Chidamide group (n = 23) included 9 patients with newly diagnosed AML and 14 patients with R/R AML. CR or CRi was reported in 19 patients in this group (19/23, 82.6%), including CRs in 16 patients (16/23, 69.6%, 6 newly diagnosed patients [6/9, 66.7%] and 10 relapsed/refractory patients [10/14, 71.4%]) and CRis in 3 patients (3/23, 13.0%, 2 newly diagnosed patients [2/9, 22.2%] and 1 relapsed/refractory patient [1/14, 7.1%]). No patients displayed a PR, and 4 patients (4/23, 17.4%) experienced treatment failure. ORR in this group was 82.6% (19/23). Of 19 patients with CRs/CRis, 3 patients achieved CR after two cycles of treatment, whereas the other 16 patients achieved CR/CRi after one cycle. Fifteen patients (15/23, 65.2%) were MRD negative (MRD⁻, <1 × 10⁻⁴) according to FCM after one or two cycles of treatment. Thirteen patients with newly diagnosed AML and four patients with R/R AML were investigated in decitabine group (n = 17). CRs or CRis were observed in nine patients in decitabine group (9/17, 52.9%), including CRs in three patients with newly diagnosed AML (3/13, 23.1%), three patients with R/R AML (3/4, 75.0%) and CRis in three patients with newly diagnosed AML (3/13, 23.1%). Two patients (2/17, 11.8%) had PRs, and treatment failure was observed in six patients (6/17, 35.3%). Of all nine patients with CRs/CRis, one patient achieved a CR after two cycles of treatment, whereas the other eight patients achieved a CR/CRi after one cycle. ORR was 64.7% (11/17), and seven patients (7/17, 41.2%) achieved MRD− (Sup Table 4). Twenty patients (13 in chidamide group and 7 in decitabine group) who achieved CRs/CR were continued to receive consolidation chemotherapy.

Patients in chidamide group exhibited higher CR/CRi rates (82.6% vs. 52.9%, p = 0.0430) and improved DOR (4.9 months vs. 1.1 months, p = 0.0274) than decitabine group. Median durations of complete responses before relapse were 7.0 months (range: 2.9–11.9 months) and 2.9 months (range: 0.9–8.8 months) in chidamide and decitabine groups, respectively (p = 0.1853, Sup Table 4). ORR was not significantly different between groups (p = 0.2743). Patients with de novo AML showed higher CR/CRi rates (88.9% vs. 46.2%, p = 0.0405) and prolonged median DOR (7.5 months vs. 5.7 months) in chidamide group. However, there were no significant differences in CR/CRi rates in patients with R/R AML (78.6% vs. 75.0%, p = 0.8796). Extramedullary infiltration lesions disappeared in two patients and significantly declined in one patient in chidamide group, whereas no obvious improvement of extramedullary infiltration lesions was noted in decitabine group. Thirteen patients in both groups experienced disease progression over the study period.

Genetic abnormality and responses

Genetic abnormality might predict the response to treatment in patients with AML. Chidamide combined with decitabine and chemotherapy seemed to overcome adverse effects of karyotypes leading to CR/CRi in 8/8 patients all with complex karyotype, while 2/8 patients in decitabine group (complex karyotype in 2 CR/CRi patients, one with −7 and 5 with complex karyotype in 6 PR or NR patients, Sup Table 3). While gene mutations affected prognosis at the same time. Patients with DMNT3A, IDH2, and TET2 gene mutations generally experienced disease remission in both groups, as presented in Sup Figure 2a–b. In chidamide group, 66.7% of patients (4/6) with FLT3-ITD mutations alone or combined with other gene mutations achieved CR/CRi. Two patients exhibited treatment failure, including one patient with FLT3-ITD, NRAS, PTPN11, CEBPA, and WT1 gene mutations and one with an FLT3-ITD mutation alone, but it is worth noted that extramedullary disease was reduced after one cycle of treatment. Meanwhile, some satisfactory responses were observed after chidamide treatment in patients with FLT3-ITD, AXSL1, and DNA methylation-related gene mutations (e.g. DNMT3A, IDH2, TET2). In decitabine group, CRs were observed in 83.3% of patients (5/6) with FLT3-ITD gene mutations, and patients with DNMT3A, IDH2, or TET2 gene mutations achieved disease remission. However, one patient with a TP53 gene mutation displayed an unfavorable response to decitabine regimen. A majority of patients with NPM1 gene mutations achieved CR/CRi (66.7% [2/3] in chidamide group vs. 75.0% [3/4] in decitabine group). Patients with RUNX1 gene mutations achieved CR/CRi after one or two cycles of treatment, but a short DOR and poor prognosis were observed in both groups. Chidamide in combination with decitabine and chemotherapy produced better effects on NRAS and KRAS gene mutations than regimen without chidamide, and 66.7% of patients (4/6) with at least one of two aforementioned gene mutations achieved remission. Variations of gene mutations before and after treatment were also delineated in 25 patients (Sup Fig. 3).

Survival analysis

In total, eight patients in chidamide group and three patients in decitabine group were still alive at the last follow-up, and the cumulative mortality rates in these groups were 65.2% (15/23) and 82.4% (14/17), respectively. Median OS was 9.7 months in chidamide group, versus 8.5 months in decitabine group (log-rank test, p = 0.0684, Figure 1(a)). Patients with newly diagnosed AML who received chidamide exhibited an OS improvement (log-rank test, p = 0.0450, Figure 1(b)), whereas no significant OS improvement was noted in patients with R/RAML in either group (Figure 1(c)). Median PFS was 6.4 months in chidamide group compared with 2.6 months in decitabine group (log-rank test, p = 0.0242, Figure 1(d)). Median PFS among patients with newly diagnosed AML was 12.9 months in chidamide group, versus 2.0 months in decitabine group (log-rank test, p = 0.0135, Figure 1(e)), whereas no significant difference in PFS was noted among patients with R/R AML (7.2 months vs. 3.3 months; log-rank test, p = 0.5124, Figure 1(f)). OS and PFS benefits were analyzed and compared using Cox proportional-hazards model (Sup Tables 4 and 5). OS and PFS among patients with newly diagnosed AML were significantly improved in chidamide group (Figure 2(a,c)). However, few survival benefits were observed among patients with R/R AML because of the limited sample size (Figure 2(b,d)). It was also estimated that sex, age, and adverse risk stratification were crucial factors.

Figure 1. Kaplan–Meier analysis of survival in two treatment groups and MRD status. (a) OS analysis among all patients in two therapeutic groups. (b, c) OS analysis separately in patients with newly diagnosed acute myeloid leukemia (AML) and R/R AML. (d) PFS analysis among all patients in therapeutic groups. (e, f) PFS analysis separately in patients with newly diagnosed acute myeloid leukemia (AML) and R/RAML. (g) OS analysis according to MRD status in two therapeutic groups. MRD⁻ was associated with significantly improved OS (p < 0.0001). (h, i) OS analysis in chidamide and decitabine groups, respectively. MRD⁻ had a positive effect on OS among patients in decitabine group (p = 0.0014), but there was no significant difference in OS in chidamide group (p = 0.1128). (j) PFS analysis according to MRD status in two therapeutic groups. MRD⁻ was associated with significantly improved PFS (p < 0.0001). (k, l) PFS analysis in chidamide and decitabine groups, respectively. Median PFS was significantly longer among patients with AML and MRD⁻than among those without MRD⁻ in two groups (p = 0.0220 and p = 0.0023, respectively). Differences were compared by applying log-rank test. Statistical significance was set at p < 0.05.

Figure 2. Overall survival (OS) and progression-free survival (PFS) analysis by acute myeloid leukemia (AML) subgroup (newly diagnosed and R/R AML) using Cox proportional-hazards model. (a, b) OS analysis in newly diagnosed and R/R AML subgroups. OS and PFS were more significantly improved by chidamide combination regimen than by decitabine combination regimen among patients with newly diagnosed AML. (c, d) PFS analysis in newly diagnosed and R/RAML subgroups. Few survival benefits were observed among patients with R/RAML in chidamide group.

In this study, the incidence of MRD⁻ was higher in chidamide group, but no significant difference was observed (65.2% vs. 41.2%, p = 0.1308, Table 1). We found that MRD⁻ could significantly improve OS and PFS (both p < 0.0001, Figure 1(g,j)). MRD− had a positive effect on OS among patients in decitabine group (log-rank test, p = 0.0014), but no significant difference was recorded in chidamide group (log-rank test, p = 0.1128, Figure 1(h,i)). Median PFS among patients with MRD⁻ (9.9 months in chidamide group vs. 10.3 months in decitabine group) was significantly longer than that among patients who did not achieve MRD⁻ (p = 0.0220 and p = 0.0023, respectively, Figure 1(k,l)).

Table 1. Treatment responses in two therapeutic groups.

	Total (n = 40)	Chidamide group (n = 23)	Decitabine group (n = 17)	P value
Treatment response				0.0462*
CR/CRi	28 (70.0)	19 (82.6)	9 (52.9)	0.0430^a
PR	10 (25.0)	0 (0.0)	2 (11.8)
Treatment failure	2 (5.0)	4 (17.4)	6 (35.3)
ORR	30 (75.0)	19 (82.6)	11 (64.7)	0.2743
DOR, months	3.1 (0.7, 8.5)	4.9 (2.3, 8.8)	1.1 (0, 5.8)	0.0274*
Duration of remission before relapse, months	3.7 (2.3, 9.0)	7.0 (2.9, 11.9)	2.9 (0.9, 8.8)	0.1853
PFS, months	4.6 (2.6, 9.7)	6.4 (3.7, 10.0)	2.6 (1.7, 6.8)	0.0138*
OS, months	9.1 (4.3, 12.2)	9.7 (5.4, 12.7)	8.5 (2.5, 10.2)	0.1255
MRD^-	22 (55.0)	15 (65.2)	7 (41.2)	0.1308
Disease progression	31 (77.5)	17 (73.9)	14 (82.4)	0.7071
Death	29 (72.5)	15 (65.2)	14 (82.4)	0.2973
Myelosuppression, days	15 (12, 18)	14 (12, 17)	16 (13, 18)	0.4096

Note: CR: complete remission; CRi: CR with incomplete blood count recovery; PR: partial response; ORR: overall response rate; DOR: duration of response; PFS: progression-free survival; OS: overall survival; MRD: minimal residual disease. * p<0.05

Safety and toxicities

AEs of any grade are presented in Table 2. Hematological toxicities were most common all-cause AEs reported in both groups. Of note, grade 3 or 4 hematological AEs occurred in nearly all patients in both groups. The incidence of grade 3 or higher febrile neutropenia was slightly higher in decitabine group (64.7% vs. 47.8%). The most common grade 1 or 2 non-hematological AE was infection (including pneumonitis, sepsis, and anal mucositis). A higher incidence of sepsis was observed in chidamide group (30.4% [7/23] vs. 5.9% [1/17]); however, serious sepsis caused 1 death in decitabine group (1/17, 5.9%). However, there were no significantly differences in myelosuppression time (15.57 ± 5.36 vs.15.18 ± 5.29, p = 0.8209). The high rates of sepsis in chidamide group might be related with more frequencies of severer neutropenia (grade 3 or higher), but more were mild (grade 2 or less). Other severe (grade 3 or higher) non-hematological AEs occurred more frequently in decitabine group. Most patients tolerated treatment, and the majority of AEs were managed and cured using symptomatic and supportive treatments, such as transfusions, subcutaneous injection of granulocyte colony-stimulating factor (G-CSF), anti-inflammatory treatment, and correction of electrolyte disturbance. During one or two cycles of treatment, no patients died in chidamide group, whereas three patients (3/17, 17.6%) died in decitabine group (one case each of severe infection, tumor lysis syndrome, and possible cerebral hemorrhage).

Table 2. Adverse events in two therapeutic groups.

	Chidamide group (n = 23)				Decitabine group (n = 17)
	Grade 1–2	Grade 3	Grade 4	Grade 5	Grade 1-2	Grade 3	Grade 4	Grade 5
Leucopenia	0	6 (26.1)	17 (73.9)	0	0	5 (29.4)	11 (64.7)	0
Neutropenia	0	0	23 (100.0)	0	0	0	16 (94.1)	0
Thrombocytopenia	0	2 (8.7)	21 (91.3)	0	0	3 (17.6)	14 (82.4)	0
Anemia	9 (39.1)	14 (60.9)	0	0	4 (23.5)	13 (76.5)	0	0
Febrile neutropenia	0	11 (47.8)	0		0	10 (58.8)	1 (5.9)	0
Fatigue	22 (95.7)	0	0	0	14 (82.4)	3 (17.6)	0	0
Fever	20 (87.0)	0	0	0	15 (88.2)	2 (11.8)	0	0
Anorexia	15 (65.2)	0	0	0	11 (64.7)	0	0	0
Hypokalemia	12 (52.2)	0	0	0	9 (52.9)	1 (5.9)	0	0
Hypoalbuminemia	11 (47.8)	0	0	0	12 (70.6)	1 (5.9)	0	0
Nausea	10 (43.5)	0	0	0	3 (17.6)	0	0	0
Pneumonitis	7 (30.4)	3 (13.0)	0	0	8 (47.1)	2 (11.8)	1(5.9)	0
Alanine aminotransferase increased	8 (34.8)	0	0	0	3 (17.6)	0	0	0
Aspartate aminotransferase increased	5 (21.7)	0	0	0	2 (11.8)	0	0	0
Sepsis	0	7 (30.4)	0	0	0	0	0	1 (5.9)^a
Pleural effusion	3 (13.0)	0	0	0	6 (35.3)	0	0	0
Cough	6 (26.1)	0	0	0	3 (17.6)	0	0	0
Productive cough	3 (13.0)	0	0	0	2 (11.8)	0	0	0
Diarrhea	3 (13.0)	0	0	0	3 (17.6)	0	0	0
Menorrhagia	3(13.0)	0	0	0	0	1 (5.9)	0	0
Anal mucositis	4 (17.4)	0	0	0	0	0	0	0
Pharyngolaryngeal pain	2 (8.7)	0	0	0	1 (5.9)	0	0	0
Abdominal pain	2 (8.7)	0	0	0	0	0	0	0
γ-Glutamyl transferase increased	1 (4.3)	0	0	0	1 (5.9)	0	0	0
Hyponatremia	1 (4.3)	0	0	0	4 (23.5)	0	0	0
Abdominal distension	1 (4.3)	0	0	0	1 (5.9)	0	0	0
Hypocalcemia	0	0	0	0	2 (11.8)	0	0	0
Tumor lysis syndrome	0	0	0	0	0	0	0	1 (5.9)^a
Respiratory failure	0	0	0	0	0	0	0	1 (5.9)^a
Atrial fibrillation	0	1 (4.3)	0	0	0	0	0	0
Soft tissue infection	0	1 (4.3)	0	0	0	0	0	0
Hyperuricemia	0	0	0	0	1 (5.9)	0	0	0

Note: Each patient was counted only once according to the most severe adverse event throughout the process; ^aDeaths attributable to adverse events.

Discussion

Achieving remission is crucial for patients with newly diagnosed or R/R AML to increase ORR and enhance survival benefits in patients, as well as for consolidation treatments or HSCT. Epigenetic therapy has widely attracted attention for the treatment of cancers in recent years. Hypomethylating agents (HMAs) such as decitabine and azacitidine were used for elder AML or adverse risk MDS patients. Interestingly, regimens of HMA combined with DA, HAAG, CAG and FLAG have shown better effect in R/R or de novo AML in clinical trials [16–18,34,35]. Hypomethylating agents or low dose- cytarabine combined with BCL-2 inhibitor has become promising regimen for AML patients intolerant for intensive chemotherapy, while there were also limitations in treating resistance or AML relapse [36]. Chidamide has shown significant inhibition of AML cell lines and exerted synergistic effects with HMAs [37,38]. A phase II study on chidamide combined with decitabine and CAG in patients with relapsed/refractory acute myeloid leukemia has shown tolerated and effective outcomes [39]. Chidamide also has been reported to overcome venetoclax plus azacitidine regimen-resistant AML [40], and increase the sensitivity of refractory or relapsed AML cells to anthracyclines [41]. Therefore, chidamide combined with HMA and chemotherapy seems to have great potential for AML patients especially resistant to conventional chemotherapy or targeted therapy.

In this study, we analyzed 18 patients with R/R AML to compare efficacy between two salvage regimens. The CR/CRi rate was 78.6% in chidamide group, which was higher than that in decitabine group, and median OS (9.7 months) and PFS (6.4 months) were also prolonged. Compared with results in decitabine group, CR/CRi rates and DOR of patients with newly diagnosed AML were significantly improved in chidamide group, and extramedullary disease disappeared in two patients. Moreover, OS and PFS among patients with newly diagnosed AML were also significantly prolonged with chidamide combination regimen. Although R/R AML patients in chidamide group had no significant improvement in OS and PFS, results in those patients displayed an increased CR/CRi rates, which provided an opportunity for those preparing for bridging HSCT. However, in this study, it is a pity that none of these patients chose to conduct HSCT due to financial reasons. During treatments, hematological toxicities were the most common AEs reported in both groups. Grade 3 or higher hematological AEs were observed in almost all patients in both groups. Non-hematological AEs, especially several infections, gastrointestinal disorders, general disorders, and metabolic disorders, occurred more frequently in decitabine group. Most AEs could be managed using symptomatic and supportive treatments. Hence, chidamide-based regimens might be regarded as effective and tolerable therapeutic protocols.

Genetic abnormalities are closely correlated with risk categories, treatment options, and prognosis in AML with evolution of whole-genome sequencing technology. Chidamide regimen could overcome karyotype adverse conditions with higher CR/CRi rates in our study, while less effective for certain genetic mutations. TP53 mutations are associated with complex karyotypes, monosomal karyotypes, and some aneuploid karyotypes, which are suggestive of unfavorable prognoses in AML [42]. Patients with RUNX1-mutant AML generally have poor outcomes [9]. ASXL1 mutations are more common in elderly patients with AML, and they are negatively correlated with long-term survival [43]. In our study, frequently mutated genes included FLT3-ITD, DNMT3A, NPM1, CEBPA, IDH2, NRAS, TET2, KRAS, RUNX1, GATA2, and ASXL1. Patients with DNA methylation-related gene mutations (DMNT3A, IDH2, and TET2) generally achieved remission under HMA- and HDACi-based treatment. FLT3-ITD, ASXL1, RUNX1, and GATA2 gene mutations are considered poor risk factors for AML. Although some satisfactory responses were observed in patients with FLT3-ITD, AXSL1, and DNA methylation-related gene mutations in chidamide group, the combination of an HDACi and HMA exerted significant effects on patients with NRAS and KRAS gene mutations compared with the effects of HMA monotherapy. In this study, MRD⁻ rate was higher in chidamide group than in decitabine group, which was correlated with improved PFS and OS survival. However, the efficacy of combination of a HDACi and HMA should be examined using greater numbers of clinical samples for clarification and these findings in this cohort need to be replicated in a larger one.

There were several limitations in our study: (1) no significant differences of OS and FPS for R/R AML were detected in both groups, which should be re-evaluated if samples and follow-up time are sufficient; (2) did not evaluation HDACi and HMA combination with other targeted drugs, such as BCL2 inhibitor and FLT-3 inhibitors, which could be encouraging. Further prospective or retrospective studies of targeted drug combinations will help improve outcomes and reduce toxicity for AML patients.

Conclusion

The combination of an HDACi, HMA, and chemotherapy is an effective therapy with tolerable toxicities in adults with intermediate- and adverse- risk AML in this cohort, especially for those with newly diagnosed disease. There remain challenges and opportunities for AML treatments, and further explorations to clarify mechanisms of HDACis and HMAs in AML will be critical.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.