Reduced duration and dosage of venetoclax is efficient in newly diagnosed patients with acute myeloid leukemia

ABSTRACT

Objectives

The combination of Venetoclax (VEN) and Azacitidine (AZA) increases survival outcomes and yields excellent responses in patients with acute myeloid leukemia (AML). However, dose reduction (or discontinuation) is commonly encountered due to therapy-related toxicity. Thus, this study aimed to investigate the efficiency and safety of a lower dosage of venetoclax for the treatment of AML.

Methods

This observational study analyzed the characteristics and outcomes of newly diagnosed AML patients who received 100 mg VEN combined with AZA for 14 days at our institution.

Results

A total of 36 patients were enrolled, and the median age at diagnosis was 64 years. After a median follow-up of 15 (range 4–29) months, the median overall survival (OS) and progression-free survival (PFS) for the whole cohort were 17 (4–29) months and 12 (1–28) months, respectively. Meanwhile, the overall response rate (ORR) was 69.4%, and the CRc rate was 66.7% in the whole cohort. Subgroup analysis revealed that NPM1 mutations and FAB-M5 subtype were associated with higher response rates, whereas the adverse ELN risk group was predictive of an inferior response. Moreover, ASXL1, NPM1, and IDH1/2 mutations negatively impacted PFS.

Discussion

Our study optimized the administration of venetoclax plus azacytidine for the treatment of AML patients. Response rates were favorable, with median survival in agreement with the findings of earlier reports, offering valuable insights for optimizing VEN-based regimens.

Conclusion

In summary, the VEN combination regimen is effective for the treatment of newly diagnosed AML patients in the real world despite VEN dose reductions .

KEYWORDS:

• Venetoclax
• azacytidine
• dose and duration optimization
• acute myeloid leukemia
• myelosuppression management
• Newly diagnosed
• retrospective study
• efficiency and safety

Introduction

Acute myeloid leukemia (AML) is a highly aggressive and heterogeneous hematopoietic malignancy characterized by a poor prognosis [1]. Venetoclax (VEN), an oral and highly selective BCL2 inhibitor, in combination therapy with hypomethylating agents (HMAs), has recently revolutionized the treatment paradigm for AML [2]. Phase III clinical trials have highlighted the encouraging efficacy and safety of VEN in combination with HMAs in newly diagnosed AML patients who are older or ineligible for intensive induction chemotherapy [3]. However, compared with clinical trials, the real-world utilization of VEN + HMAs is more diverse and intricate, especially in highly heterogeneous populations such as AML patients. Recently, several retrospective studies have investigated the real-world safety and efficacy of VEN + HMAs for the treatment of newly diagnosed AML. Surprisingly, the majority of patients experienced VEN dose reductions or cycle delays, primarily due to therapy-related toxicities. Zhu, L.X. et al. reported that thirty-two out of 59 patients experienced VEN schedule delay, whilst 71.2% of patients experienced VEN dose reduction attributable to hematological and infectious adverse events (AEs) [4]. Similarly, the Mayo Clinic documented that VEN treatment cycle delays were frequently encountered in roughly half of patients, and VEN dose reduction was considered in 70% of the patients [5]. At the same time, Gangat N et al. showed that the median VEN dose was 200 mg, and 86% of patients adjusted their VEN dose following the first cycle due to AEs. Besides, the duration of VEN treatment was reduced in 80% of patients, ranging from 7 to 21 days [6]. The Italian cohort study revealed that 102/190 patients underwent VEN dose modifications due to AEs, including dose reduction and permanent discontinuation in 78.4% and 21.6% of cases, respectively [7]. A retrospective study conducted in the US enrolling 169 AML patients highlighted that VEN treatment modifications did not affect treatment duration. Importantly, patients with treatment schedule modifications and alterations in VEN doses had a longer median OS compared with those without [8]. Consequently, optimal VEN dose scheduling may enhance clinical outcomes while mitigating the severity and frequency of therapy-related hematological and infectious AEs [9–13]. In this retrospective study, the real-world efficiency and safety of the 100 mg VEN combined with AZA regimen for a 14-day schedule was investigated, aiming to provide valuable insights into optimizing VEN-based treatment regimens in a real-world clinical setting.

Materials and methods

Study design and patients

Clinical, biological, and molecular data were retrospectively collected from 36 patients treated with VEN + AZA between January 2020 and October 2022. The inclusion criteria were as follows: adult patients with newly diagnosed AML who received at least one cycle treatment with VEN + AZA and a follow-up duration exceeding 2 months. The exclusion criteria were as follows: patients with missing or unavailable results or those with a diagnosis of acute promyelocytic leukemia. 100 mg of VEN was orally administered daily for 14 days, while 75 mg/m² 5-azacitidine was subcutaneously administered for 7 days. All patients in the study did not use azoles as prophylaxis. The diagnostic criteria for AML were based on the World Health Organization classification [14]. Genetic risk was classified in accordance with the 2022 ELN risk stratification [1]. The study was approved by the Ethics Committee (2022-053, approval date: 13 July 2022).

Outcomes and assessments

Response was assessed according to the 2022 ELN response criteria [1]. All patients (n = 36) included in our study underwent bone marrow cell morphology after each treatment cycle, as per the physician’s discretion. CRc comprised both complete response (CR) and CR with incomplete hematologic recovery (CRi). The overall response rate (ORR) encompassed CRc and partial remission (PR), while NR denoted response to the therapy. OS was defined as the period from the initiation of VEN treatment to the last follow-up or death from any cause, and PFS was defined as the period from the date of achieving the first CRc until the date of hematologic relapse or death from any cause.

Safety analyses

Adverse events (AEs) occurring from the first cycle until 30 days after treatment discontinuation were recorded. Hematologic and non-hematologic toxicity were evaluated based on the Common Terminology Criteria and Adverse Events classification (CTCAE v5.0).

Statistical analysis

Baseline variables of patients were described as numbers and percentages for categorical variables or median and range (or interquartile range) for continuous variables. Categorical variables were compared using the X2 test or Fisher’s exact test, whereas continuous variables were compared using the nonparametric Mann–Whitney U test. Overall survival was assessed using the Kaplan-Meier method, and group differences were compared using the log-rank test. Median follow-up was estimated by the reverse Kaplan–Meier method. A two-sided p < 0.05 was considered statistically significant. All statistical analyses were conducted using GraphPad Prism 8.0.

Result

Baseline characteristics

A total of 36 patients were enrolled in this retrospective study, including 18 males and 18 females, with a median age of 64 (44–86) years. The majority of patients were de novo AML (n = 34, 94%); 86% of patients had a documented Eastern Cooperative Oncology Group (ECOG) score of 0 or 1, whilst 14% of patients had an ECOG ≧2. Risk stratification according to the current ELN criteria 2022 categorized patients into favorable, intermediate, and adverse risk groups, comprising 5 (14%), 10 (28%), and 19 (53%) patients, respectively (Table 1). As anticipated, the FAB-M5 subtype was the most common subtype of leukemia, accounting for 69% of cases. The whole molecular alteration is illustrated in Figure 1. Notably, FLT3-ITD/TKD (36%), NPM1 (33%), and IDH1/2 (33%) were the top three mutations in our study.

Figure 1. Bar graph exhibiting the gene mutational characteristics of all patients.

Table 1. Baseline features of all patients.

Characteristic	Total (n = 36)	%
Median age	64 (44–86)	–
Age < 65	17	47%
Age> = 65	19	53%
Sex(male/female)	18/18	50%/50%
AML type
De novo	34	94%
Secondary to MDS	2	6%
ELN risk stratification
Favorable	5	14%
Intermediate	10	28%
Adverse	19	53%
unknown	2	5%
FAB type
M5	25	69%
non-M5	11	31%
ECOG score
0	13	36%
1	18	50%
≧2	5	14%
Baseline counts, median
WBC ×10^9/L	9.1 (0.31–258)	–
Bone marrow blast (%)	70 (30–95)	–
Cycles of VEN (median)	7 (2–15)	–
1–3	9	25%
4–9	15	42%
≧10	12	33%

Treatment

The thirty-six patients received a total of 260 cycles, with a median therapy cycle of 7 (range, 2–15). Noteworthily, 6% of patients received less than 14 days of VEN during the first cycle due to severe myelosuppression. The median time between the first and second cycles was 28 days (range 16–46). The median duration of hospitalization for cycle 1 was 14 days (ranges 7–33). Of note, one patient was transitioned to alloHSCT following VEN treatment. At the time of analysis, 56% of patients were still receiving VEN therapy. Most patients (89%) continued VEN into the third cycle, and 33% of patients underwent more than 10 cycles of VEN + AZA therapy. Importantly, four patients who achieved sustained remission prolonged their therapy intermission to 2 months after receiving more than 10 cycles of therapy. Detailed information is displayed in Supplementary Table 1.

Response and survival

Responses to treatment are presented in Table 2. The final ORR was 69.4%, including 66.7% (n = 25) CR + CRi and 2.7% (n = 1) PR; the median time to first response was 1 month (range, 0.5–2), and the median duration of response was 10 months (range, 1–28). Among the 24 patients with CRc, 25% (n = 6) of patients relapsed. The response rates at cycles 1 and 2 are displayed in Figure 2. ORR was achieved in 2 NR patients (5.5% of the total population), whilst CRc was achieved in six PR patients (16.7%) following a second course. Subgroup analysis exposed that the presence of NPM1 mutation (CRc 91.7% vs. 52%, P = 0.018) and the M5 subtype (CRc 82.6% vs. 38.5%, P = 0.011) was associated with a favorable response, while adverse ELN risk (50% vs. 87.5%, P = 0.032) predicted inferior response. Lastly, age, gender, WBC count at diagnosis, ECOG score, as well as IDH1/2, FLT3-ITD, DNMT3A, and RUNX1 mutations had no significant impact on response to VEN-based therapy (Figure 3).

Figure 2. Bar plot reflecting the proportion of patients achieving response at cycles 1 and 2.

Figure 3. Subgroup analysis of composite complete remission (CRc) based on clinical characteristics and gene mutations.

Table 2. Efficacy of reduced VEN-AZA combinations in 36 patients with acute myeloid leukemia

Outcomes	All patients (n = 36)
Responses, N (%)
ORR (CR + CRi + PR)	25 (69.4%)
CRc (CR + CRi)	24 (66.7%)
PR	1 (2.7%)
Median time to response, months	1 (0.5–2)
Relapse, months
Relapse rate of patients with CRc	6 (25%)
Median duration of response	10 (1–28)
Survival, months
Median overall survival (OS)	17 (3-NR)
Median progression free survival (PFS)	12 (1–28)
Median OS for patients achieving CR/CRi	NR
Median OS for patients not achieving CR/CRi	8.5 (2–21)

With a median follow-up of 15 months (range: 4–29 months), 16 (44.4%) patients died, including 6 owing to relapses, 3 ascribed to disease progression, 3 due to adverse infection, 2 attributed to cerebral hemorrhage, 1 due to respiratory failure and 1 due to gastrointestinal tract hemorrhage. In the total population, 1- and 2-year OS were 64.6% and 39.9% (Figure 4A), respectively. The median OS for the entire cohort was 17 months. OS was not reached in patients achieving CRc, while it was 8.5 months in patients who did not achieve CRc (Figure 4C). Patients above 65 years had a significantly lower median months compared to younger patients (12 months vs 23 months, P = 0.047, Figure 5A). Likewise, patients in the adverse ELN 2017 risk group and carrying the DNMT3A mutation were associated with shorter OS, yet the difference was not statistically significant (Figure 5B,C). Conversely, patients harboring NPM1 and IDH1/2 mutations had a numerically longer median OS versus those without mutations, but the difference was not statistically significant (Figure 5D,E). PFS for the entire cohort was 12 months and was not achieved in patients with CRc (Figure 4B, D). Age <65y vs >=65: 15 months vs 10 months (P = 0.044), ELN 2022 risk category (adverse vs no adverse events: 10 months vs not reached) (P = 0.034), ASXL1 mut vs wt: 4.5 months vs 14 months (P = 0.009), NPM1 mut vs wt: not reached vs 12 months (P = 0.08), IDH1/2 mut vs wt: not reached vs 9 months (P = 0.08) was the statistically significant impact the PFS duration (Figure 6). Given that patients with the FAB-M5 subtypes achieved higher CRc rates, the prognosis of patients with the M5 and non-M5 subtypes was analyzed. As depicted in Supplementary Figure 1, the median OS and PFS of patients with the M5 subtype were 17 and 12 months, which were comparable to those of the whole cohort. While patients with the M5 subtype had a numerically longer median OS than those without the M5 subtype, the difference was not statistically significant. In addition, the relapse rate for patients with the M5 subtype was 20.8%.

Figure 4. Survival. (A) Overall survival (OS) of the whole cohort (N = 36). (B) progression-free survival (PFS) of the whole cohort (C) OS based on composite complete remission (CRc) (D) PFS based on CRc.

Figure 5. Overall survival of patients with AML according to clinical characteristics and gene mutation status (A) OS stratified by age (age < 65y vs age > = 65y) (B) OS according to the occurrence of adverse events (non-adverse events vs adverse events) (C) OS stratified by DNMT3A mut status (DNMT3A mut vs wt) (D) OS stratified by IDH1/2 mut status (IDH1/2 mut vs wt) (E) OS stratified by NPM1 mut status (NPM1 mut vs wt) (F) OS stratified by ASXL1 mut status (ASXL1 mut vs wt).

Figure 6. Progression-free survival of patients with AML according to clinical characteristics and gene mutation status (A) PFS stratified by age (age < 65y vs age > = 65y) (B) PFS according to the occurrence of adverse events (non-adverse events vs adverse events) (C) PFS stratified by ASXL1 mut status (ASXL1 mut vs wt) (D) PFS stratified by IDH1/2 mut status (IDH1/2 mut vs wt) (E) OS stratified by NPM1 mut status (NPM1 mut vs wt) (F) PFS stratified by DNMT3A mut status (DNMT3A mut vs wt).

Adverse events

Common adverse events are listed in Table 3. Myelosuppression was the most commonly occurring AE. In the first cycle, grade 3 and 4 hematologic side effects, including neutropenia, thrombocytopenia, and anemia, accounted for 36%, 25%, and 44% of all AEs, respectively. 19% (n = 7) of patients developed febrile neutropenia (≥ Grade 3). Six patients experienced grade ≥ 3 infection. The incidence of invasive fungal infections (n = 1,3%) and tumor lysis syndrome (n = 1,3%) were relatively low. Nevertheless, it is worthwhile pointing out that several patients (n = 4,11%) required a delay in the treatment cycle or intermittent therapy. No early deaths were noted within the first 30 days following treatment initiation in the current study. In the second cycle, the grade and severity of most AEs were lower.

Table 3. Adverse events.

Treatment-related AEs	Cycle 1	Cycle 2
neutropenia ≥ grade 3	10 (28%)	6 (17%)
thrombocytopenia ≥ grade 3	9 (25%)	4 (11%)
anemia ≥ grade 3	6 (17%)	1 (3%)
Febrile neutropenia ≥ grade 3	7 (19%)	3 (8%)
Bleeding ≥ grade 3	0	0
Infection ≥ grade 3	6 (17%)	2 (6%)
Invasive fungal infection	1 (3%)	1 (3%)
Neutropenia ≥30 d	1 (3%)	0
Tumor lysis syndrome	1 (3%)	0
Treatment adjustment due to AE	4 (11%)	0
30-day mortality	0	0

Discussion

To the best of our knowledge, this is the first retrospective study to assess the safety and efficacy of shortening VEN administration to 100 mg for 14 days from cycle 1 in AML patients treated with VEN + AZA. Similar to previous studies, the majority of our patients had de novo-AML (94%), with intermediate-high risk cytogenetics (81%). However, the median age of our cohort was 64 years, which was lower than earlier studies [15,16]. The majority of patients (69%) were of the FAB-M5 subtype. The median duration of VEN administration was 7 cycles herein, which was consistent with the phase III VIALE-A trial. Contrastingly, the median OS (17 months) was longer than that reported in the phase III VIALE-A trial (14.7 months) [3], and other retrospective studies have documented comparable median OS ranging from 10 to 12 months [5,7,11,15,17], likely attributable to differences in patient population and limitations. In terms of prognostic determinants of survival, patients carrying NPM1 and IDH1/2 mutations achieved longer OS and PFS, which was in line with the findings of previous research [6,7,18,19]. Nevertheless, it is worthwhile acknowledging that the OS was comparable in our cohort, possibly due to the limited number of enrolled patients. Importantly, ASXL1 mutation was associated with a shorter PFS, in agreement with the observation of Garciaz S. et al., who postulated that the presence of ASXL1 mutation was predictive of poor outcomes [16]. On the other hand, prior studies described that ASXL1 mutations were associated with a higher response rate [6].

Regarding efficacy outcomes, the ORR and CRc rates in our study were 69.4% and 66.7%, respectively, similar to those reported in the VIALE-A trial (66.4%) and several retrospective cohorts [10,15]. Remarkable, the median time to CRc was merely 1 month. A favorable CRc rate of 91.7% was achieved in cases with NPM1 mutations, which is in line with the results of prior published trials [7,19–21]. A novel finding of this study was that patients with the FAB-M5 subtype had a superior response than other subtypes of AML, which was inconsistent with the observations of previous research, which observed that the FAB-M5 subtype was associated with disease relapse and that the OS of patients with the M5 subtype was significantly shorter than that of non-M5 patients [22]. The non-M5 subtyped just accounted for 28% (n = 10) in our cohort; additional clinical data and larger sample sizes are warranted to validate our conclusion. As expected, patients in the ELN adverse group exhibited lower response rates.

Prolonged myelosuppression and infections are complications related to VEN treatment. Despite lowering the dosage and duration of VEN treatment, a number of patients experienced cytopenia and infection. Grades 3 and 4 hematologic adverse events in our study and VIALE-A encompassed thrombocytopenia (25% vs 45%), neutropenia (28% vs 42%), febrile neutropenia (19% vs 42%), and anemia (17% vs 26%). It is worthwhile emphasizing that the incidence of adverse infections (grade > 3) in VIALE-A was significantly higher than that in our study (64% vs 17%). Moreover, the overall mortality rate was also marginally lower in our study compared with the VIALE-A trial (47% vs. 56.2%). Several expert recommendations have been proposed to manage those common side effects, including shortening VEN duration and adjusting the VEN or HMA dose. A recently published study demonstrated that shortening the duration of VEN administration to 14 days may lower the risk of complications and achieve similar CRc and survival rates compared with the 28-day regimen [23]. Mirgh S. et al. also described that a shorter 21-day duration of VEN administration in induction therapy leads to a similar response rate, early blood cell count recovery, and a shorter duration of intravenous antibiotic use [24]. In a French multiple-center study, limiting VEN treatment to 7 days acquired a similar response rate to the recommended 28-day VEN administration. Additionally, dose reduction of VEN appears to be associated with favorable outcomes in responders [12].

The limitations of the current single-center study are as follows: firstly, this was a retrospective study with a limited number of patients. Secondly, a control group was not established herein, and response assessment was neither centralized nor blinded. In addition, important clinical data, including minimal residual disease measurements, overall hospital resource utilization, and real-world quality of life outcomes, were not collected in the database, rendering them unavailable for analysis.

In summary, limiting the duration and dosage of VEN yielded similar response and OS rates to those reported in previous studies, thereby paving the way for future clinic studies investigating different dosages and durations of VEN administration. Taken together, this study offers a rationale for employing a reduced AZA-VEN regimen in a clinical trial evaluating triplet combination.

Author contributions

[Guolin Yuan] contributed to the study conception and design. Material preparation, data collection and analysis were performed by [Jingying Cui], [Chunfang Li] and [Qiong Yan] and the first draft of the manuscript was written by [Xuexing Chen]. These four authors contributed equally to this work and should be considered co-first authors. All authors read and approved the final manuscript.

Acknowledgments

Authors acknowledge all the staff and team members of Department of Hematology, Xiangyang Central Hospital.

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.