Real-world outcomes of azacitidine plus venetoclax in acute myeloid leukemia: a multicenter retrospective cohort study from Thailand

Abstract

Background:

Azacitidine (AZA) plus venetoclax (VEN) has emerged as a widely accepted treatment option for acute myeloid leukemia (AML), particularly in older patients or those unfit for intensive chemotherapy. However, real-world data on AZA + VEN efficacy and safety in Southeast Asia remain limited.

Objectives:

To evaluate the real-world effectiveness and safety of AZA + VEN in newly diagnosed (ND) and relapsed/refractory (R/R) AML patients in Thailand.

Design:

A retrospective observational multicenter study.

Methods:

This is a multicenter retrospective study included ND and R/R AML patients treated between 2021 and 2024 at three tertiary hospitals in Thailand. All patients received AZA at 75 mg/m² for 7 days per cycle. VEN dosing and duration were individualized based on physician judgment, drug availability, and patient affordability. Data collection included clinical characteristics, cytogenetics, treatment details, response rates, survival outcomes, and toxicities.

Results:

A total of 81 patients were analyzed, included 54 ND and 27 R/R AML cases, with a median age of 65 years. Based on European LeukemiaNet 2022 classification, 51.9% had intermediate risk, and 33.3% had adverse risk. The composite complete remission was 56.8% (ND: 64.8%, R/R: 40.7%). VEN was administered at a median dose of 100 mg for 28 days, combined with potent CYP3A4 inhibitor of antifungal prophylaxis (posaconazole 51.0%, voriconazole 30.4%, itraconazole 17.7%). The median overall survival was 9.2 months and relapse-free survival was 8.1 months. Grades 3–4 neutropenia and febrile neutropenia occurred in 93.8% and 60.5% of patients, respectively.

Conclusion:

This real-world practice highlights the feasibility and effectiveness of AZA-VEN in combination with antifungal prophylaxis for elderly or unfit AML patients in resource-limited countries. However, infectious complications remain a concern with this low-intensity regimen.

Plain language summary

Real-world outcomes of azacitidine and venetoclax in Thai patients with acute myeloid leukemia

Acute myeloid leukemia (AML) is a type of blood cancer that mainly affects older adults and people who are too weak to receive strong chemotherapy. A new treatment using two medicine azacitidine and venetoclax has been shown to help in clinical trials. However, there is limited information on how well this combination works in real-world settings, especially in Southeast Asia. We studied 81 adults in Thailand with either newly diagnosed or previously treated (relapsed/refractory) AML who received azacitidine plus venetoclax between 2021 and 2024. The goal was to understand how effective and safe the treatment was in routine hospital settings. We also looked at the side effects and how long patients lived after treatment. About two-thirds of the patients had newly diagnosed AML, and the rest had relapsed or difficult-to-treat disease. Most patients were around 65 years old. After treatment, more than half went into remission (the cancer was no longer detectable), and those with newly diagnosed AML had better outcomes than those with relapsed disease. On average, patients lived around 9 months after starting treatment. Those with better initial responses tended to live longer. Side effects included low blood counts, infections, and fevers, which were common but manageable with hospital care. This study shows that azacitidine and venetoclax can be used safely and effectively in Thai hospitals, even when lower medicine doses are used to reduce costs and drug interactions. The results are similar to what has been seen in Western clinical trials. The findings may help guide doctors treating leukemia patients in low- and middle-income countries.

Keywords

acute myeloid leukemia azacitidine real-world data venetoclax

Introduction

Acute myeloid leukemia (AML) is a highly complex malignancy with a persistently poor prognosis. A study of AML patients in Thailand revealed that the overall survival (OS) rates at 1 year and 2 years were only 31.9% and 29.6%, respectively.¹ Among elderly AML patients in Thailand (age >60 years), only 15% received intensive induction chemotherapy that reflecting both clinical fragility and limited treatment access.² Prognosis is particularly dismal in patients who are unfit for intensive chemotherapy, such as the elderly, those with significant comorbidities that increase the risk of treatment-related complications, or patients with relapsed or refractory AML (R/R AML).^2,3

Recently, venetoclax (VEN), a selective inhibitor of B-cell lymphoma-2 (BCL-2), has emerged as a promising therapeutic option in AML. BCL-2 is an antiapoptotic protein that is frequently overexpressed in leukemic stem cells, playing a critical role in evading apoptosis and sustain leukemic proliferation. Hypomethylating agents (HMAs) such as azacitidine (AZA) or decitabine (DEC) sensitize leukemic cells to apoptosis by reactivating silenced pro-apoptotic genes and altering mitochondrial priming. The combination of VEN with HMAs synergistically promote mitochondrial outer membrane permeabilization, leading to enhanced apoptosis of leukemic stem cells. This synergistic mechanism of action has significantly transformed the treatment landscape for AML.^4,5

In a phase Ib clinical trial, the combination of VEN and HMAs in elderly patients with newly diagnosed AML (ND AML) demonstrated remarkable efficacy, with composite complete remission (cCR) rates exceeding 70%. This includes both complete remission (CR) and complete remission with incomplete blood count recovery (CRi). Median OS was reported as 16.9 months for AZA and 16.2 months for DEC in combination with VEN.⁶ These findings were subsequently confirmed by the VIALE-A phase III trial, which reported cCR (CR/CRi) rate of 66.4% and a median OS of 14.7 months with AZA + VEN in ND AML patients deemed unfit for intensive chemotherapy. These results have established AZA + VEN as a standard of care for elderly (⩾75 years) or unfit patients with ND AML.^7–10

Encouraged by these outcomes, VEN combined with HMAs has been increasingly adopted in clinical practice as a first-line treatment for patients unfit for intensive chemotherapy, those ineligible for hematopoietic stem cell transplantation including for patients with R/R AML. Real-world studies have reported cCR rates ranging from 40% to 86.8% in ND AML patients, and median OS spanning from 8.6 and 33.8 months with an average of six treatment cycles (range 1–35).^11–19 In R/R AML, where the treatment options are limited and outcomes are generally poor, VEN + HMA therapy has been used off-label with reported cCR between 33% and 71.4%, and median OS ranging from 4 to 16 months with an average of four treatment cycles (range 1–24).^20–25 However, the limited sample sizes and heterogeneity of study populations highlight the need for additional real-world data, particularly from Southeast Asia, where access to novel therapies and supportive care may differ significantly from Western settings.

This study aims to evaluate the real-world outcomes of AZA + VEN on the response rate, survival outcomes, adverse events, and factors associated with treatment response and survival in Thai patients with ND and R/R AML. The findings will intend to inform current practice and guide treatment strategies to improve the quality of care for AML patients.

Methods

Study design and participants

This study was a multicenter retrospective observational analysis conducted at three tertiary hospitals in Thailand: Chiang Mai University Hospital, King Chulalongkorn Memorial Hospital, and Phramongkutklao Hospital. The study period spanned from January 1, 2021, to December 31, 2024. Inclusion criteria included adult patients aged 18 years or older with ND AML or R/R AML regardless of age or Eastern Cooperative Oncology Group (ECOG) performance status who received at least one dose of AZA plus VEN. The decision to use AZA + VEN was made on a case-by-case basis, considering factors such as comorbidities, prior therapies, transplant eligibility, and drug accessibility. In several cases of ND AML, younger or fit patients received AZA + VEN following the shared decision-making between physicians, patients, and their families when intensive chemotherapy was declined or considered inappropriate for personal or contextual reasons. Most patients in the R/R AML group were younger or fit individuals who had previously received intensive chemotherapy, and some patients experienced relapse following allogeneic stem cell transplantation (AlloSCT) were also included in this study. In Thailand, AZA is reimbursed only under the Civil Servant Medical Benefit Scheme (CSMBS). Patients covered by other national health program including the Universal Coverage Scheme (UCS) and the Social Security Scheme (SSS) must pay for AZA out of pocket. VEN is not reimbursed under any of the three national health insurance schemes. All the patients must purchase VEN at their own expense. Comprehensive data on clinical characteristics, cytogenetics, mutational profiles, risk stratification per the European LeukemiaNet (ELN) 2022 AML risk classification,²⁶ and ELN risk classification for AML patients receiving less-intensive therapies (ELN 2024 less-intensive).²⁷ Treatment regimens (dosage, duration, and total cycles of AZA plus VEN), antifungal prophylaxis based on the national guidelines for the treatment and prevention of febrile neutropenia in hematologic malignancies of Thailand²⁸ (posaconazole, voriconazole, or itraconazole—based on institutional availability and physician preference), response rates, percentage of AlloSCT, and survival outcomes. Patients who developed breakthrough fungal infection despite prophylaxis were managed in accordance with the national guidance for treatment of invasive fungal diseases.²⁹ Survival outcomes included OS and relapse-free survival (RFS). OS was defined as the time from AML diagnosis to death from any cause. Patients who were alive at their last follow-up were censored on the date they were last known to be alive. RFS was defined as the time from the first documentation of cCR (achieving CR or CRi) to relapse or death from any cause, whichever occurred first. Patients without relapse or death were censored at the date of last follow-up.²⁶ Treatment responses were categorized according to ELN2022 criteria.²⁶ CR was defined as an absolute neutrophil count (ANC) ⩾1.0 × 10⁹/L, a platelet count ⩾100 × 10⁹/L, red blood cell transfusion independence, and bone marrow blasts <5%. CRi was defined as meeting all the criteria for CR, except for persistent neutropenia (ANC < 1.0 × 10⁹/L) and/or thrombocytopenia (platelet count, <100 × 10⁹/L). Partial remission (PR) was defined as a decrease of ⩾50% in bone marrow blasts to a level between 5% and 25%, in addition to meeting the hematologic criteria for CR. The overall response rate (ORR) was calculated as the proportion of patients who achieved CR, CRi, and PR. The duration of remission (DOR) was defined as the length of time from first documentation of CR/CRi until relapse, death from any cause, or last of follow-up, whichever occurred first. Exclusion criteria included patients with acute promyelocytic leukemia, mixed phenotypic acute leukemia (MPAL), insufficient clinical information, or those receiving palliative cares. This was retrospective observational study may have the potential for selection and information bias, particularly due to incomplete or missing data in some cases. To minimize these effects, this study included all consecutive eligible patients during study period and performed cross-verification of clinical records across multicenter whenever possible to ensure data accuracy and consistency. Due to the retrospective nature of data collection from anonymized electronic medical records, the requirement for informed consent was waived by the institutional review board approval (Research Ethics Committee Panel 5 of the Faculty of Medicine, Chiang Mai University, Thailand (study code: MED-2567-0690)).

Study objectives

The primary objective was to evaluate the ORR of the AZA plus VEN regimen in adult AML patients in Thailand. Secondary objectives included assessments of complete remission (CR/CRi) rate, OS, DOR, toxicities, and factors associated with response and survival outcomes in this real-world setting.

Statistical analysis

Continuous variables were summarized as medians with ranges or interquartile ranges (IQRs), while categorical variables were expressed as percentages. Comparisons between categorical variables were performed using the Chi-square test or Fisher’s exact test, and continuous variables were compared using the Mann–Whitney U test. Survival analyses were conducted using the Kaplan-Meier method, with groups differences assessed via the log-rank test. Logistic regression was used to estimate odds ratios (ORs) with 95% confidence intervals (95% CIs) for ORR prediction. Cox proportional hazards modeling was applied to estimate hazard ratios (HRs) with 95% CIs in time-to-event analyses for OS and RFS prediction. Variables with a p value <0.10 in univariable analyses were included in the multivariable model. Missing data were addressed using a complete case analysis approach. Sensitivity analysis was not performed due to the exploratory nature of the study. Statistical analysis was performed using Stata/MP version 17.0 for Mac (StataCorp, College Station, TX, USA), Microsoft Excel for Mac version 16.94 (Microsoft), and RStudio version 2024.12.1+563 (Posit, PBC). Sample size calculation was based on the primary outcome of an estimated ORR of approximately 65% for AZA plus VEN in adult AML patients.⁸ To achieve a 10% margin of error with a 95% CI and calculated power of study of 81.6%, a minimum of 88 patients was required.

Results

Patients’ characteristics

A total of 88 patients were initially recruited during study period. Two patients had MPAL, and five patients had incomplete clinical data. The final study cohort comprised 81 patients with AML who received at least one cycle of AZA + VEN, including 54 (66.7%) with ND AML patients and 27 (33.3%) with R/R AML. The median age of the entire cohort was 65 years (IQR: 54–70), with the ND group being older (68.5 years, IQR: 64–71) than the R/R group (52 years, IQR: 38–61; p < 0.001). According to the ELN 2022 risk classification, 14.8% of patients were categorized as favorable risk, 51.9% as intermediate risk, and 33.3% as adverse risk. Secondary AML from prior myelodysplastic syndrome (MDS) or myeloproliferative neoplasm (MPN) was observed in 27.2% of cases. Therapy-related AML was reported in 3.7% of patients. The median bone marrow blast percentage was significantly higher in the ND group compared to the R/R group (73.5% vs 43%, p = 0.046). Cytogenetic analysis revealed a normal karyotype in 45.7% of patients, while 11.1% exhibited complex karyotypes. For the ELN 2024 classification specific to AML patients receiving less-intensive therapies, risk classification was possible in 44 patients who had undergone mutation testing. Among them, 57.8% were classified as favorable risk, 37.8% as intermediate risk, and 4.4% as adverse risk. The median number of mutations per patient was 2 (range 0–8), with a median of 2 (range 0–8) in ND AML, and 1 (range 1–6) in R/R AML. The most frequent mutations identified were FLT3-ITD (14.8%), IDH1/IDH2 (9.9%), and NPM1 (7.4%). TP53 mutation was found in 2.5%. Among patients with adverse-risk mutations (excluding TP53 mutations), the most frequently detected were RUNX1 (16.1%), SRSF2 (12.5%), ASXL1 (10.7%), SF3B1 (8.9%), and STAG2 (5.4%) (Figure 1). The most common co-mutation involved NPM1 with DNMT3A was observed in 4 cases. Other notable co-mutations were RUNX1 with FLT3-ITD, RUNX1 with SRSF2, and SRSF2 with TET2 in three patients each. Other co-mutations were observed in 1–2 cases each and the details information provided in Supplemental Material.

Figure 1.

Number of patients with detected mutations in ND AML and R/R AML patients treated with azacitidine plus venetoclax.

In the R/R AML group, 51.9% had primary refractory disease prior to AZA + VEN initiation. The majority (96.3%) had received at least one prior line of therapy, with 51.9% receiving one line, 48.1% receiving two or more lines of treatment. Moreover, six patients (22.2% of R/R AML) had relapsed after AlloSCT and then receiving AZA + VEN (Table 1). Prior AlloSCT patients had a median age of 49 years (IQR: 38–61) with and a median bone marrow blast of 26% (IQR: 17–50). The ELN 2022 risk classification showed that 16.7% of patients were at favorable risk, 66.6% were at intermediate risk, and 16.7% were at adverse risk. Secondary AML from prior MDS or MPN was observed in 33.3% of cases. Cytogenetic studies showed a complex karyotype in 33.3%. Gene mutation testing was not performed in any of the prior AlloSCT patients.

Table 1.

Baseline characteristics of ND AML and R/R AML treated with AZA + VEN.

Characteristics	Total (n = 81)	ND AML (n = 54)	R/R AML (n = 27)	p Value
Age, years (median, IQR)	65 (54–70)	68.5 (64–71)	52 (38–61)	<0.001
⩾65 years, n (%)	42 (51.9)	39 (72.2)	3 (11.1)	<0.001
⩾75 years, n (%)	7 (8.6)	7 (12.9)	0	0.050
Male sex, n (%)	42 (51.9)	31 (57.4)	11 (40.7)	0.157
ECOG performance status, n (%)
0–1	63 (77.8)	38 (70.4)	25 (92.6)	0.177
2	18 (22.2)	16 (29.7)	2 (7.4)	0.052
Hemoglobin, g/dL, median (IQR)	7.5 (6.6–9.1)	7.1 (6.1–8.7)	8.4 (7.3–9.8)	0.010
WBC count, ×10⁹/L, median (IQR)	10.6 (2.7–31.7)	14.3 (3.8–39.5)	3.2 (2.2–20.6)	0.018
Platelets, ×10⁹/L, median (IQR)	63 (30–107)	71 (35–132)	40 (10–90)	0.171
BM blast, %, median (IQR)	60 (30–84)	73.5 (35–86)	43 (22–70)	0.046
Cytogenetics, n (%)
Normal	37 (45.7)	28 (51.9)	9 (33.3)	0.115
Other intermediate	25 (30.9)	15 (27.8)	10 (37.0)	0.395
t(8;21)	4 (4.9)	3 (5.6)	1 (3.7)	0.717
t(6;9)	2 (2.5)	1 (1.9)	1 (3.7)	0.613
−7/7q-	4 (4.9)	2 (3.7)	2 (7.4)	0.468
Complex	9 (11.1)	4 (7.4)	5 (18.5)	0.134
ELN 2022 risk, n (%)				0.268
Favorable	12 (14.8)	10 (18.5)	2 (7.4)
Intermediate	42 (51.9)	25 (46.3)	17 (62.9)
Adverse	27 (33.3)	19 (35.2)	8 (29.6)
Previous MDS/MPN, n (%)	22 (27.2)	15 (27.8)	7 (25.9)	0.860
Therapy-related AML, n (%)	3 (3.7)	3 (5.6)	0	0.212
AML status of R/R AML, n (%)
Primary refractory	—	—	14 (51.9)	—
Number of prior lines*				—
1	—	—	14 (51.9)
⩾2	—	—	13 (48.1)
Prior azacitidine exposure, n (%)	—	—	10 (37.0)	—
Prior AlloSCT,** n (%)	—	—	6 (22.2)	—

Details of prior lines of treatment; 1 prior line (n = 14): four patients received a single course of 7+3 chemotherapy, four patients received 7+3 and intermediate dose cytarabine consolidation, six patients received azacitidine alone; ⩾2 prior lines (n = 13): five patients received 7+3 induction and intermediate dose cytarabine consolidation then received MEC (mitoxantrone, etoposide and cytarabine) at relapse, four patients received 7+3 induction and intermediate dose cytarabine consolidation then received FLAG-Ida at relapse, one patient received 7+3 then received azacitidine alone after refractory, one patient received HyperCVAD induction then received 7+3 at relapse, one patient received high dose of cytarabine induction then received 7+3 at relapse, and one patient received azacitidine alone, 7+3 after relapse and FLAG-Ida after refractory.

Prior AlloSCT; all six patients received AlloSCT with matched-sibling donor.

AlloSCT, allogeneic stem cell transplantation; AML, acute myeloid leukemia; AZA + VEN, azacitidine plus venetoclax; BM, bone marrow; ECOG, Eastern Cooperative Oncology Group; ELN, European Leukemia Net; IQR, interquartile range; MDS, myelodysplastic syndrome; MPN, myeloproliferative neoplasm; ND, newly diagnosed; R/R, relapsed/refractory; WBC, white blood cell.

Treatment, response rate, and survival outcomes

All patients received AZA at standard dose of 75 mg/m² for 7 days per cycle, while the median VEN dose was 100 mg (range: 50–300) for a median duration of 28 days (IQR: 14–28). Antifungal prophylaxis was administered to 79 patients (97.5%), with posaconazole (51.0%), voriconazole 30.4%, and itraconazole 17.7%. Two additional patients received fluconazole and clarithromycin concomitantly with VEN, respectively. The median number of AZA + VEN cycles was 3 (range: 1–24). Treatment delays before the initiation of subsequent cycles were reported in 61.7% of patients, with prolonged neutropenia as the major reason (54.3%). Details on treatment and cycle information were presented in Table 2.

Table 2.

Treatment duration of azacitidine plus venetoclax in ND and R/R AML.

Treatment details	Total (n = 81)	ND AML (n = 54)	R/R AML (n = 27)	p Value
Venetoclax duration (days), median (IQR)	28 (14–28)	28 (14–28)	28 (14–28)	0.734
Total number of AZA + VEN cycles, (median, range)	3 (1–24)	4 (1–24)	3 (1–15)	0.221
Delay before next cycle, n (%)	50 (61.7)	34 (62.9)	16 (59.3)	0.491
Cycle duration (days), median (IQR)	35 (31.7–42.0)	35.0 (32.7–42.0)	35.5 (30.4–39.7)	0.579
G-CSF prophylaxis after CR/CRi, n (%)	36 (47.4)	28/50 (56.0)	8/26 (30.8)	0.037

AML, acute myeloid leukemia; AZA + VEN, azacitidine plus venetoclax; CR, complete remission; CRi, complete remission with incomplete blood count recovery; IQR, interquartile range; ND, newly diagnosed; R/R, relapsed/refractory.

The ORR after the first cycle was 49.4% (ND: 55.6%, R/R: 37.0%), with a best ORR of 64.2% (ND: 72.2%, R/R: 48.1%; Figure 2). The composite complete remission (CR/CRi) at best response in ND AML and R/R AML were 64.8% and 40.7%. The median time to first response was 33.5 days (IQR: 27–39.5), and the median time to best response was 40.5 days (IQR: 28.5–62.5). The median DOR was 3.8 months (IQR: 2.1–10.9; Table 3). Among patients who achieved a response, relapse occurred in 51.6%, with a higher rate observed in the R/R group (73.7%) compared to the ND group (41.9%). AlloSCT was subsequently performed in 8.6% of patients (ND: 9.3%, R/R: 7.4%) and AZA + VEN was stopped after AlloSCT in all patients.

Figure 2.

Best response achieved in ND and R/R AML patients treated with AZA + VEN.

Table 3.

Response rates and duration of remission in ND and R/R AML treated with AZA + VEN.

Response and survival outcomes	Total (n = 81)	ND AML (n = 54)	R/R AML (n = 27)
Overall response (ORR), n (%)
After first cycle	40 (49.4)	30 (55.6)	10 (37.0)
Best response	52 (64.2)	39 (72.2)	13 (48.1)
Type of response after first cycle, n (%)
CR	8 (9.9)	6 (11.1)	2 (7.4)
CRi	24 (29.6)	19 (35.2)	5 (18.5)
PR	8 (9.9)	5 (9.3)	3 (11.1)
Best response, n (%)
CR	24 (29.6)	20 (37.0)	4 (14.8)
CRi	22 (27.2)	15 (27.8)	7 (25.9)
PR	6 (7.4)	4 (7.4)	2 (7.4)
Number of cycles at best response, n (%)
1	50 (62.5)	31 (57.4)	19 (70.4)
2	16 (20.0)	13 (24.1)	3 (11.1)
3	8 (10.0)	6 (11.1)	2 (7.4)
4	6 (7.5)	4 (7.4)	2 (7.4)
Time to first response (days), median (IQR)	33.5 (27–39.5)	30.5 (27–40)	34.5 (28–39)
Time to best response (days), median (IQR)	40.5 (28.5–62.5)	46.5 (30–66.5)	34 (27.5–58.5)
Duration of remission (months), median (IQR)	3.8 (2.1–10.9)	4.4 (2.1–11.7)	3.7 (2.1–9.9)

With a median follow-up duration of 7.1 months (IQR: 3.7–13.2), the median OS for the entire cohort was 9.2 months (95% CI: 8.0–14.7). OS was significantly longer in the ND group (14.4 months, 95% CI: 8.7–28.0) compared to the R/R group (7.8 months, 95% CI: 5.3–9.4; log-rank test, p = 0.037; Figure 3(a)). The 1-year OS rate was 42.2% in the entire cohort, 53.1% in the ND group, and 23.1% in the R/R group. There were four induction-related deaths (4.9%), and the overall mortality rate was 58.0% (ND: 50.0%, R/R: 74.1%). Patients in the R/R group experienced a higher incidence of mortality from disease progression compared to those in the ND group (73.7% vs 41.9%). Median RFS was 8.1 months (95% CI: 3.8–25.7), with a median RFS of 8.1 months (95% CI: 4.4–26.2) in the ND group and 3.8 months (95% CI: 2.0–19.8) in the R/R group (log-rank test, p = 0.249). The 1-year RFS was 39.1% for the overall cohort, 45.1% for ND AML patients, and 23.4% for R/R AML patients (Figure 3(b)). A detail of the visualization of individual patients’ outcomes, including treatment duration time to response, and survival status, was presented in the swimmer plot (Figure S1).

Figure 3.

Kaplan–Meier curve for OS and RFS in newly ND AML and R/R AML treated with AZA + VEN. (a) Median OS of ND AML and R/R AML were 14.4 and 7.8 months, respectively (log-rank test, p = 0.037). (b) Median RFS of ND AML and R/R AML were 8.1 and 3.8 months, respectively (log-rank test, p = 0.249).

Predictors of response, survival, and relapse

In the multivariable logistic regression analysis, R/R AML status was an independent negative predictor of ORR, significantly reducing the likelihood of achieving remission (adjusted odds ratio: 0.219, 95% CI: 0.058–0.815, p = 0.024; Table S1). Poor ECOG performance status (⩾2) was associated with inferior OS and RFS in univariable analysis; however, this lost statistical significance in the multivariable model.

Regarding molecular predictors, the ASXL1 mutation was significantly associated with worse OS (HR: 4.131, 95% CI: 1.458–11.703, p = 0.008) and higher relapse risk (HR: 13.581, 95% CI: 2.241–82.322, p = 0.005) in univariable analysis, but this association was not retained in multivariable analysis. In contrast, adverse-risk per ELN 2022 and FLT3-ITD mutation remained independent predictors of relapse in the multivariable Cox regression for RFS. Specifically, adverse-risk ELN 2022 patients had a significantly higher relapse risk (adjusted HR (aHR): 4.756, 95% CI: 1.750–6.926, p = 0.031), and patients with FLT3-ITD mutation conferred a similarly elevated relapse risk (aHR: 2.833, 95% CI: 1.550–5.176, p = 0.024).

Achieving an ORR was the strongest independent predictor of improved survival outcomes. It significantly reduced both the risk of death (aHR: 0.039, 95% CI: 0.005–0.296, p = 0.002) and relapse (aHR: 0.081, 95% CI: 0.016–0.345, p = 0.020). In univariable analysis, relapse strongly increased mortality risk (HR: 6.791, 95% CI: 2.376–19.415, p < 0.001), although this effect diminished after adjusting for other factors. Furthermore, patients with favorable-risk AML per ELN 2022 showed a significantly lower risk of relapse (aHR: 0.119, 95% CI: 0.014–0.690, p = 0.032; Tables S2 and S3).

Adverse events

Hematologic toxicities were the most frequently observed adverse events, with anemia (96.3%), neutropenia (96.3%), and thrombocytopenia (92.6%) being the most frequently observed. Grades 3–4 neutropenia occurred in 93.8% of patients and grades 3–4 thrombocytopenia was observed in 75.3%. Febrile neutropenia developed in 60.5% of patients, and infection-related complications occurred in 62.9% of the cohort (Table 4). Among these, invasive fungal infections were identified in nine patients (11.1%). Invasive pulmonary aspergillosis accounted for eight patients, despite antifungal prophylaxis. All patients with breakthrough fungal infections recovered with appropriate antifungal therapy. Notably, no cases of tumor lysis syndrome were reported. Additionally, there was no any grade of gastrointestinal toxicities observed during the treatment period, including nausea, vomiting, or elevated transaminase levels.

Table 4.

Adverse events in ND and R/R AML patients treated with AZA + VEN.

Adverse events	Total (n = 81)	ND AML (n = 54)	R/R AML (n = 27)
Hematologic AEs, n (%)
Anemia
Overall	78 (96.3)	53 (98.2)	25 (92.6)
Grades 3–4	59 (72.8)	41 (75.9)	19 (66.7)
Neutropenia
Overall	78 (96.3)	44 (96.3)	26 (96.3)
Grades 3–4	76 (93.8)	50 (92.6)	26 (96.3)
Thrombocytopenia
Overall	75 (92.6)	51 (94.4)	24 (88.9)
Grades 3–4	61 (75.3)	41 (75.9)	20 (74.1)
Febrile neutropenia, n (%)	49 (60.5)	36 (66.7)	13 (48.2)
Infection, n (%)	51 (62.9)	37 (68.5)	14 (51.9)
Fungal infection	9 (11.1)	5 (9.3)	4 (14.8)
Bacteremia	7 (8.6)	4 (7.1)	3 (11.1)
Perianal abscess	7 (8.6)	6 (11.1)	1 (3.7)

AEs, adverse events; AML, acute myeloid leukemia; AZA + VEN, azacitidine plus venetoclax; ND, newly diagnosed; R/R, relapse/refractory.

Discussion

This study provided real-world evidence on the effectiveness and safety of AZA + VEN in Thai patients with AML, highlighting both treatment response and survival outcomes. The ORR was 64.2%, with a composite CR (CR/CRi) of 56.8%. Among ND AML patients, the median age was 68.5 years (IQR: 64–71), with an ORR of 72.2% and a composite CR of 64.8%, resulting in a median OS of 14.4 months. In R/R AML patients, the median age was 52 years (IQR: 38–61), with an ORR of 48.1%, composite CR of 40.7%, and a median OS of 7.8 months. The DOR following CR/CRi was 4.4 months for ND AML and 3.7 months for R/R AML. Hematologic toxicities were prominent, with grades 3–4 neutropenia occurring in 93.8% and febrile neutropenia in 60.5%. Fungal infections were documented in 11.1% of cases despite antifungal prophylaxis, while bacteremia was observed in 8.6% of patients.

A key distinguishing feature of this study was the age distribution of the study population. While most landmark clinical trials and real-world studies have focused on older populations (aged ⩾75 years), this study enrolled a relatively younger ND-AML cohort. Specifically, 72.2% of the patients were aged ⩾65 years, but only 12.9% were aged ⩾75 years. This pattern reflects with current clinical practice in Thailand, where ND-AML patients up to 60–65 years of age are typically treated with intensive induction chemotherapy (e.g., the 7+3 regimen), whereas those older than 65–70 years are more often managed with palliative approaches.² In contrast, the VIALE-A trial enrolled ND AML patients who were predominantly aged ⩾75 years and unfit for intensive chemotherapy, reporting a median OS of 14.7 months and a composite CR rate of 66.4% with AZA + VEN therapy.⁸ Despite the younger median age in our ND AML group (68.5 years), the response rates and survival outcomes were remarkably comparable, with a median OS of 14.4 months and composite CR of 64.8%. Similarly, a long-term follow-up of VIALE-A confirmed sustained survival benefits with AZA + VEN.¹⁰ The real-world studies on ND AML treated with AZA + VEN reported ORR 40%–86.8% with median OS between 8.6 and 33.8 months, and an average of six treatment cycles (range 1–35).^11–19 This suggested that AZA + VEN maintains its effectiveness even in relatively younger, but still unfit patients in real-world settings.

Our findings also emphasize the potential role of AZA + VEN in an age group that might traditionally be considered for intensive chemotherapy. In Thailand, limited access to transplantation and socioeconomic barriers often preclude more aggressive treatment options, making AZA + VEN a feasible alternative. However, treatment decisions are also shaped by insurance coverage and out-of-pocket costs. AZA is reimbursed only patients under the CSMBS, while those covered by the UCS and the SSS must pay for AZA by themselves. VEN is not reimbursed under any national health insurance plan, meaning all patients must pay VEN at their own expense. These financial limitations may significantly impact treatment accessibility and continuity, especially in low- and middle-income countries. Notably, most patients in our study achieved a response most commonly CR or CRi within the first one to two cycles of AZA + VEN. Specifically, 62.5% of patients reached their best response after the first cycle, and an additional 20% responded by the second cycle (82.5% in total after 1–2 cycles of AZA + VEN). Comparatively, historical data on Thai ND AML patients treated with intensive chemotherapy showed slightly lower ORR (60.2%) compared to this study (72.2%), with a shorter 1-year OS (31.9% vs 53.1%).¹ Moreover, A single-center study comparing AZA + VEN with historical IC-treated controls in ND AML found a higher rate of measurable residual disease (MRD)-negative CR in the AZA + VEN group, along with lower transfusion dependency, infection rates, gastrointestinal side effects, and overall hospitalization costs.³⁰ These findings suggested that AZA + VEN may provide tolerable alternative to IC while maintaining favorable response rates, particular in resource-limited settings.

Moreover, despite utilizing a reduced VEN dose (100 vs 400 mg/day in VIALE-A), response rates in this study were comparable to those observed in clinical trials. The lower VEN dose was necessitated by concomitant use of CYP3A4 inhibitors for antifungal prophylaxis, which increases VEN plasma levels. Previous studies have demonstrated that VEN dose adjustments when combined with strong CYP3A4 inhibitors (e.g., azoles) maintain therapeutic efficacy while potentially reducing treatment-related costs.³¹ Our data supported this approach, highlighting the feasibility of optimized VEN dosing in real-world practice, especially where invasive fungal infections remain a concern.

In R/R AML, our observed median OS of 7.8 months and ORR of 48.1% were consistent with real-world studies, which describe a median OS of 4–16 months and ORR of 30%–50% in similar patient populations.^21,24 Despite the limitations of this regimen in heavily pretreated patients, AZA + VEN appeared to provide meaningful clinical benefit, even in those with poor-risk features. Furthermore, meta-analyses have demonstrated a pooled ORR of 60%–70% for ND AML and 30%–50% for R/R AML patients receiving AZA + VEN, consistent with our findings.^15,17,19

Notably, multivariable analysis in our study identified ELN 2022 adverse-risk classification and FLT3-ITD mutation as independent predictors of relapse, aligning with previous reports particularly in the absence of NPM1 co-mutation or without AlloSCT.²⁶ FLT3-ITD is well established as a driver mutation linked to high leukemic burden, increase relapse risk, poor long-term outcomes even in the era of targeted therapies. While the addition of FLT3 inhibitor may improve outcomes in intensive regimen,^32,33 their role in lower-intensity approach such as AZA + VEN remains unclear. The ELN 2024 risk classification in AML patients that specifically developed for patients receiving less-intensive therapy did not demonstrate a significant association with survival outcomes in our cohort. This finding suggested that while ELN 2024 represented an effort to refine risk prediction in older or unfit patients, its value in mixed or real-world cohorts that combining of ND AML and R/R AML across age ranges may require further validation. Moreover, TP53 mutation is widely recognized as a marker of induction failure, high relapse rate, and dismal OS across therapeutic setting.^26,27 In our study, the TP53 mutated patients were infrequent and precluded robust statistical significance in prognostication due to limited sample size. However, their biological relevance is still important and warrants larger with molecularly annotated cohorts.

A key strength of this study was the real-world representation of AML patients in Southeast Asia, addressing a gap in literature on AZA + VEN outcomes in this population. Additionally, the study highlighted the feasibility of administering a lower fixed-dose VEN strategy with potent CYP3A4 inhibitors to optimize treatment efficacy, decrease fungal infection while alleviating economic burden. However, limitations included a relatively short follow-up period compared to other studies, restricting long-term survival assessment, and lack of MRD assessment in this study. The low rate of AlloSCT (8.6%) also impacts survival outcomes, reflecting challenges in transplantation for unfit elderly patients including the donor availability and reimbursement issues in Thailand. Furthermore, the limited number of cases which did not reach the initially calculated sample size may reduce the precision of findings. However, a post hoc power calculation based on the actual sample of 81 patients yielded an estimated power of 78.3% which remains reasonably high and acceptable for observational research. The modest sample size also limited the statistical power to detect the association between specific mutations or co-mutations and treatment outcomes; however, it was not the primary objective of this study. Given the high rates of infection and febrile neutropenia observed, routine use of granulocyte-colony stimulating factor (G-CSF) prophylaxis should be considered in future treatment protocols. Notably, the fungal infection rate in this study (11.1%) appeared slightly lower than reported rate with AML patients received intensive chemotherapy in Thailand (15.8%)³⁴ and AZA + VEN treated AML patients (17%–30.8%).^31,35 In Thailand, all AML patients received intensive chemotherapy and low-intensity regimen (AZA + VEN) in our study received the same antifungal prophylaxis strategy in accordance with national guidelines. Therefore, the observed difference in fungal infection rates may reflect the differences in the intensity of bone marrow suppression. Furthermore, longer follow-up studies and comprehensive cost-effectiveness analyses are needed to refine treatment strategies and improve patient outcomes in resource-limited settings.

Conclusion

This study provides real-world evidence on the effectiveness and safety of AZA + VEN in Thai patients with AML, demonstrating comparable response rates and survival outcomes to international clinical trials despite dose modifications due to economic constraints. The findings highlight the feasibility of a lower VEN dose with CYP3A4 inhibitors while maintaining treatment efficacy. High rates of hematologic toxicity and infection emphasize the need for optimized supportive care, including G-CSF and antifungal prophylaxis. Limited access to AlloSCT and the proper assessment tool for selecting unfit patients for transplantation remains a key challenge affecting survival outcomes. Future research should focus on extended follow-up, cost-effectiveness analyses, and strategies to enhance treatment accessibility and prolong survival in resource-limited countries.

Supplemental Material

sj-docx-1-tah-10.1177_20406207251372770 – Supplemental material for Real-world outcomes of azacitidine plus venetoclax in acute myeloid leukemia: a multicenter retrospective cohort study from Thailand

Supplemental material, sj-docx-1-tah-10.1177_20406207251372770 for Real-world outcomes of azacitidine plus venetoclax in acute myeloid leukemia: a multicenter retrospective cohort study from Thailand by Thanawat Rattanathammethee, Chantiya Chanswangphuwana, Panachai Silpsamrit, Kannadit Prayongratana, Sirichai Srichairatanakool, Teerachat Punnachet, Nonthakorn Hantrakun, Pokpong Piriyakhuntorn, Sasinee Hantrakool, Chatree Chai-Adisaksopha, Ekarat Rattarittamrong, Lalita Norasetthada and Adisak Tantiworawit in Therapeutic Advances in Hematology

Supplemental Material

sj-docx-2-tah-10.1177_20406207251372770 – Supplemental material for Real-world outcomes of azacitidine plus venetoclax in acute myeloid leukemia: a multicenter retrospective cohort study from Thailand

Supplemental material, sj-docx-2-tah-10.1177_20406207251372770 for Real-world outcomes of azacitidine plus venetoclax in acute myeloid leukemia: a multicenter retrospective cohort study from Thailand by Thanawat Rattanathammethee, Chantiya Chanswangphuwana, Panachai Silpsamrit, Kannadit Prayongratana, Sirichai Srichairatanakool, Teerachat Punnachet, Nonthakorn Hantrakun, Pokpong Piriyakhuntorn, Sasinee Hantrakool, Chatree Chai-Adisaksopha, Ekarat Rattarittamrong, Lalita Norasetthada and Adisak Tantiworawit in Therapeutic Advances in Hematology

Footnotes

Acknowledgements

The authors thank all the medical staffs in three tertiary hospitals in Thailand: Chiang Mai University Hospital, King Chulalongkorn Memorial Hospital, and Phramongkutklao Hospital, who contributed to patient care in this study.

Declarations

Supplemental material

Supplemental material for this article is available online.

References

Wanitpongpun

Utchariyaprasit

Owattanapanich

, et al. Types, clinical features, and survival outcomes of patients with acute myeloid leukemia in Thailand: a 3-year prospective multicenter study from the Thai Acute Leukemia Study Group (TALSG). Clin Lymphoma Myeloma Leuk 2021; 21(7): e635–e643.

Owattanapanich

Utchariyaprasit

Tantiworawit

, et al. Improved Survival of Elderly-fit Patients With acute myeloid leukemia requiring intensive therapy: 3-year multicenter analysis from TALWG. Clin Lymphoma Myeloma Leuk 2018; 18(12): e509–e514.

Estey

Karp

Emadi

, et al. Recent drug approvals for newly diagnosed acute myeloid leukemia: gifts or a Trojan horse? Leukemia 2020; 34(3): 671–681.

Pollyea

Stevens

Jones

, et al. Venetoclax with azacitidine disrupts energy metabolism and targets leukemia stem cells in patients with acute myeloid leukemia. Nat Med 2018; 24(12): 1859–1866.

Pollyea

Amaya

Strati

, et al. Venetoclax for AML: changing the treatment paradigm. Blood Adv 2019; 3(24): 4326–4335.

DiNardo

Pratz

Letai

, et al. Safety and preliminary efficacy of venetoclax with decitabine or azacitidine in elderly patients with previously untreated acute myeloid leukaemia: a non-randomised, open-label, phase 1b study. Lancet Oncol 2018; 19(2): 216–228.

DiNardo

Pratz

Pullarkat

, et al. Venetoclax combined with decitabine or azacitidine in treatment-naive, elderly patients with acute myeloid leukemia. Blood 2019; 133(1): 7–17.

DiNardo

Jonas

Pullarkat

, et al. Azacitidine and venetoclax in previously untreated acute myeloid leukemia. N Engl J Med 2020; 383(7): 617–629.

Manda

Anz

3rd Benton

, et al. A phase 3b study of venetoclax and azacitidine or decitabine in an outpatient setting in patients with acute myeloid leukemia. Hematol Oncol 2024; 42(3): e3274.

10.

Pratz

Jonas

Pullarkat

, et al. Long-term follow-up of VIALE-A: venetoclax and azacitidine in chemotherapy-ineligible untreated acute myeloid leukemia. Am J Hematol 2024; 99(4): 615–624.

11.

Vachhani

Flahavan

, et al. Venetoclax and hypomethylating agents as first-line treatment in newly diagnosed patients with AML in a predominately community setting in the US. Oncologist 2022; 27(11): 907–918.

12.

Gershon

, et al. Early real-world first-line treatment with venetoclax plus HMAs versus HMA monotherapy among patients with AML in a predominately US community setting. Clin Lymphoma Myeloma Leuk 2023; 23(5): e222–e231.

13.

Hoff

Patel

Belli

, et al. Real-world outcomes of frontline venetoclax-based therapy in older adults with acute myeloid leukemia: an analysis utilizing EHR data. Leuk Lymphoma 2023; 64(6): 1123–1128.

14.

Qin

Kuang

Liu

Venetoclax combined with hypomethylating agents or low-dose cytarabine as induction chemotherapy for patients with untreated acute myeloid leukemia ineligible for intensive chemotherapy: a systematic review and meta-analysis. Clin Exp Med 2023; 23(2): 219–227.

15.

Ucciero

Pagnoni

Scotti

, et al. Venetoclax with hypomethylating agents in newly diagnosed acute myeloid leukemia: a systematic review and meta-analysis of survival data from real-world studies. Cancers (Basel) 2023; 15(18): 4618.

16.

Chatzilygeroudi

Darmani

El Gkotmi

, et al. Real-life multicenter experience of venetoclax in combination with hypomethylating agents in previously untreated adult patients with acute myeloid leukemia in Greece. J Clin Med 2024; 13(2): 584.

17.

Wen

Zheng

Efficacy and safety of venetoclax-based combination therapy for previously untreated acute myeloid leukemia: a meta-analysis. Hematology 2024; 29(1): 2343604.

18.

Othman

Lam

HPJ

Leong

, et al. Real-world outcomes of newly diagnosed AML treated with venetoclax and azacitidine or low-dose cytarabine in the UK NHS. Blood Neoplasia 2024; 1(3): 100017.

19.

Qureshi

Altaf

Jamil

, et al. Safety, Efficacy, and predictive factors of venetoclax-based regimens in elderly acute myeloid leukemia patients: a meta-analysis. Clin Lymphoma Myeloma Leuk 2024; 24(11): e835–e851.

20.

Bewersdorf

Giri

Wang

, et al. Venetoclax as monotherapy and in combination with hypomethylating agents or low dose cytarabine in relapsed and treatment refractory acute myeloid leukemia: a systematic review and meta-analysis. Haematologica 2020; 105(11): 2659–2663.

21.

Morsia

McCullough

Joshi

, et al. Venetoclax and hypomethylating agents in acute myeloid leukemia: Mayo Clinic series on 86 patients. Am J Hematol 2020; 95(12): 1511–1521.

22.

Graveno

Carulli

Freyer

, et al. Venetoclax in combination with hypomethylating agents or low dose cytarabine for relapsed and refractory acute myeloid leukemia. Leuk Lymphoma 2022; 63(7): 1645–1650.

23.

Laloi

Billotey

Dumas

, et al. Retrospective, real-life study of venetoclax plus azacitidine or low-dose cytarabine in French patients with acute myeloid leukemia ineligible for intensive chemotherapy. Cancer Med 2023; 12(6): 7175–7181.

24.

Todisco

Papayannidis

Fracchiolla

, et al. AVALON: the Italian cohort study on real-life efficacy of hypomethylating agents plus venetoclax in newly diagnosed or relapsed/refractory patients with acute myeloid leukemia. Cancer 2023; 129(7): 992–1004.

25.

Baba

Hida

Sambe

, et al. Efficacy of Venetoclax and azacitidine in acute myeloid leukemia compared to azacitidine monotherapy: real-world experience. Anticancer Res 2024; 44(5): 2003–2007.

26.

Dohner

Wei

Appelbaum

, et al. Diagnosis and management of AML in adults: 2022 recommendations from an international expert panel on behalf of the ELN. Blood 2022; 140(12): 1345–1377.

27.

Dohner

DiNardo

Appelbaum

, et al. Genetic risk classification for adults with AML receiving less-intensive therapies: the 2024 ELN recommendations. Blood 2024; 144(21): 2169–2173.

28.

Thai Society of Hematology, Infectious Disease Association of Thailand. Guideline for treatment and prevention of febrile neutropenia in hematologic malignancies, https://www.tsh.or.th/file_upload/files/FN%20guidance%202024.pdf (2024, accessed 26 June 2025).

29.

Infectious Disease Association of Thailand. Guidance for treatment of invasive fungal diseases 2025, https://idthai.org/Contents/Views/?d=!17!9!!1059! (2025, accessed 26 June 2025).

30.

Liu

Chen

, et al. Single-center experience of venetoclax combined with azacitidine in young patients with newly diagnosed acute myeloid leukemia. Ther Adv Hematol 2025; 16: 20406207241311776.

31.

Zhong

Wang

Dai

, et al. Combining azole antifungals with venetoclax plus azacitidine in patients with newly diagnosed acute myeloid leukemia. Hematology 2024; 29(1): 2433172.

32.

Stone

Mandrekar

Sanford

, et al. Midostaurin plus chemotherapy for acute myeloid leukemia with a FLT3 mutation. N Engl J Med 2017; 377(5): 454–464.

33.

Erba

Montesinos

Kim

, et al. Quizartinib plus chemotherapy in newly diagnosed patients with FLT3-internal-tandem-duplication-positive acute myeloid leukaemia (QuANTUM-First): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet 2023; 401(10388): 1571–1583.

34.

Rattanathammethee

Munsamai

Punnachet

, et al. D-index and invasive fungal infections (IFIs) in adult acute myeloid leukemia (AML) patients with the first episode of febrile neutropenia. PLoS One 2023; 18(5): e0286089.

35.

Zhang

Johnson

Abbott

, et al. Incidence of invasive fungal infections in patients with previously untreated acute myeloid leukemia receiving venetoclax and azacitidine. Open Forum Infect Dis 2022; 9(10): ofac486.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB

0.25 MB