Finite‐Duration Venetoclax/Azacitidine Followed by Umbilical Cord Blood Infusion Improves Outcomes in Frail Elderly Patients With Untreated Acute Myeloid Leukemia—A Pilot Study

ABSTRACT

Background

Continuous venetoclax‐azacitidine (VA) therapy is currently the major intervention for elderly or unfit acute myeloid leukemia (AML) patients. However, moderate chemotherapy with finite‐duration VA could achieve comparable or superior efficacy to infinite‐duration VA. Additionally, umbilical cord blood (UCB) transfusion improves clinical safety for elderly patients. Therefore, this single‐center study presents an improved approach: A finite‐duration VA and decitabine priming with intermediate‐dose cytarabine followed by UCB infusion.

Methods

We included elderly AML patients who underwent the treatment between March 2021 and May 2024. Induction and maintenance regimens were VA. Our consolidation therapy involved intravenous decitabine (20 mg/m² d1–3), intermediate‐dose cytarabine (1.0 g/m² q12h d4–5), with subsequent UCB infusion (1.5% ± 25% × 10⁷/kg) d7. Our endpoints included duration of response (DOR), overall survival (OS), and safety.

Results

We analyzed 16 patients with a median follow‐up of 53.6 months (95% CI: 36.2–71.0 months), and 7 out of 16 participants were at adverse risk. The median VA cycle number prior to disease relapse was 5 (IQR: 4–6). The median DOR was 17.4 months (95% CI: 7.3–27.5 months), with a 1‐year DOR rate of 68.8% and a 2‐year DOR rate of 35.2%. The median OS was 24.9 months (95% CI: 9.8–40.1 months), with a 1‐ and 2‐year OS rate of 100% and 55.6%. Specifically, among the 8 patients with IDH1/2 mutations, 4 experienced sustained remission, with 1‐year OS of 100%. Hematological and non‐hematological toxicities remained manageable during the consolidation phase.

Conclusions

We demonstrated prolonged survival, particularly in patients with IDH mutations, using the modified intervention. This approach holds significant value for elderly patients with untreated AML who are unfit for standard treatments.

1. Introduction

Venetoclax‐azacitidine (VA) is an approved drug for elderly acute myeloid leukemia (AML) patients who are unfit for intensive chemotherapy. Emerging evidence suggests that VA induction produces an overall 50%–70% response rate, and patients in remission undergo continuous VA therapy until disease relapse [1, 2, 3, 4, 5, 6, 7, 8]. Nonetheless, continuous intervention can be economically cumbersome and have augmented complications, including neutropenia, infection, and nausea [4]. Moreover, some reports suggested that relapse is also possible among continuous VA therapy patients. Thus, it is crucial to explore other forms of therapy, such as finite‐duration VA. Post‐remission moderate intensity chemotherapy is a robust consolidation therapy for managing minimal residual disease, and it can produce comparable or superior efficacy to sustained VA therapy. Herein, we established a new intervention strategy consisting of finite‐duration VA with moderate intensity chemotherapy consolidation for adoption among frail elderly AML patients.

Integrating hypomethylating agents, namely, decitabine, with chemotherapy directly and simultaneously affected leukemia cell apoptosis and tumor‐specific cytotoxic T lymphocyte responses, as demonstrated in both animal and human cellular models [9]. Based on several clinical reports, appropriate deployment of epigenetic priming, along with intensive chemotherapy, can increase CR rates [10, 11]. Hence, herein we employed a consolidation therapy involving decitabine and cytarabine. Multiple investigations reported the successful use of umbilical cord blood (UCB) in treating hematologic malignancies across multiple patient populations, namely children, young adults, and the elderly [12, 13, 14, 15, 16, 17]. Moreover, a phase II single‐arm investigation suggested that UCB infusion along with chemotherapy is an ideal and safe consolidation therapy, with tolerated myelosuppressive duration among elderly AML patients [18].

Considering strong evidence from previous publications, we established a new intervention, consisting of the finite‐duration VA plus moderate intensity chemotherapy consolidation to manage frail elderly AML patients.

2. Materials and Methods

2.1. Clinical Information

This work received ethical approval from the First Affiliated Hospital, College of Medicine, Zhejiang University (No. 20240575A). The study included untreated and frail AML patients between 60 to 80 years of age, who received a finite‐duration VA regimen (≥ 2 cycles). Only those patients who achieved complete remission were included after signing informed consent. Before consolidation therapy, cases with one or more of the following conditions were excluded from data analysis: (1) Eastern Cooperative Oncology Group (ECOG) performance status ≥ 3, or severe comorbid cardiac, pulmonary, neurologic, or metabolic diseases; (2) with other malignant tumors; (3) having received other regimens besides the VA regimen. The patients were then followed by decitabine priming combined with intermediate‐dose cytarabine, and then sequential UBC infusion as consolidation therapy (Figure 1).

FIGURE 1.

Flow diagram of selecting patients with AML in this study.

2.2. Treatment Regimens

The newly proposed intervention included induction, consolidation, and maintenance therapy. In the induction phase, the patients received venetoclax and azacitidine; venetoclax was provided at 100 mg on day 1, 200 mg on day 2, and 400 mg from day 3–28 in cycle 1, then 400 mg daily in subsequent cycles; azacitidine was subcutaneously injected at 75 mg/m² for the first 7 days of all 28‐day cycles. Those who achieved complete remission (CR) were examined in terms of chemotherapeutic tolerance, using the Ferrara criteria [19]. Fit patients underwent consolidation therapy, which included decitabine (20 mg/m² intravenous administration for 3 consecutive days, from day 1–3) and intermediate‐dose cytarabine (1.0 g/m² every 12 h for 2 consecutive days, from day 4–5), and subsequent UCB infusion (1.5% ± 25% × 10⁷/kg) on day 7. Cycle 2 of consolidation therapy is repeated every 2 months or less. Lastly, the finite‐duration VA therapy was sustained as maintenance therapy (Figure 2). We adopted a therapy cycle repetition of every 4–6 weeks, in the absence of myelosuppression. G‐CSF was provided (150 ug twice a day) when the neutrophil count dipped to < 1 × 10⁹/L, and the treatment was terminated once the neutrophil count was 2 × 10⁹/L. Supportive care, such as standard antiemetic therapy, blood transfusions, and antimicrobial therapy, was administered as deemed appropriate by the attending physician. Following the disease relapse, the physician managed the subsequent patient care.

FIGURE 2.

The improved therapeutic regimen.

2.3. Selection of UCB and Chimeric Analysis

We conducted high‐resolution HLA typing targeting the HLA‐A, B, and DR loci among the participants. UCB was acquired from the China Cord Blood Bank Network and was selected based on the following criteria: Serological matching for 3–5 out of 6 HLAs, possessed 1.5% ± 25% × 10⁷ nucleated cells/kg prior to freezing. Following intervention, we extracted patient peripheral blood cells for chimeric analysis using a real‐time quantitative PCR (qPCR) detection system and a sensitivity of 0.01%.

2.4. Monitoring of Minimal Residual Disease (MRD)

To conduct leukemia‐related immunophenotyping at diagnosis, we employed monoclonal antibodies against 10 antigens, namely, CD34, CD38, CD117, HLA‐DR, CD13, CD33, CD7, CD19, CD10, and CD45. MRD (MFC) negativity was defined as a 0.01% threshold value. In the case of 2 AML patients with recurring cytogenetic aberrations (CBFβ‐MYH11), the MRD content was described via two distinct protocols: MFC and qPCR. MRD (qPCR) negativity was also defined as a 0.01% threshold value.

2.5. Patient Follow‐Up

Subjects were followed up over 53.6 months (95% CI: 36.2–71.0 months) from treatment date to the last follow‐up in March 2026. The last follow‐up was conducted by investigators who were blinded to prior data. DOR was described as the duration between CR and disease relapse. OS was described as the duration between the treatment date and expiry date, regardless of cause (Table S1).

2.6. Assessment of Therapeutic Response and Toxicity

We employed the 2022 European LeukemiaNet (ELN) guidelines to assess therapeutic response, and simultaneously examined CR [20]. The severity of bone marrow suppression was determined using the National Cancer Institute Common Terminology Criteria for Adverse Event version 5.0 (CTCAE v5.0). The duration of hematopoietic recovery was defined as the chemotherapeutic termination date to neutrophil and platelet counts > 0.5 × 10⁹/L and > 20 × 10⁹/L, respectively, without transfusion.

2.7. Statistical Analysis

Descriptive statistics are presented as median and range in case of variables exhibiting a non‐Gaussian distribution. Follow‐up time was estimated using the reverse Kaplan–Meier method. Survival curves were generated using Kaplan–Meier analysis. All analyses employed IBM SPSS 22.0, and all tests were two‐tailed.

3. Results

3.1. Patient Demographics

In all, we analyzed 16 patients, with 68 years median age (ranging between 60 to 77 years). Table 1 and (Figure 3A,B) summarize the primary patient clinical details. These patients had an ECOG performance status score of 3 points or other comorbidities. According to the ELN risk classification, 7 out of 16 participants experienced high risk. The most frequent genetic mutations were IDH 1/2 (8 of 16) and DNMT3A (4 of 16), and 2 patients exhibited CBFβ‐MYH11 fusion (Patients 5 and 11). The median VA cycle number prior to disease relapse was 5 (IQR: 4–6). Overall, 13 patients received 2 consolidation therapy cycles, whereas 3 patients (Patients 7, 10, and 16) received only 1, according to the patients' willingness. Patient 7 refused due to prolonged neutropenia with fever during the previous cycle, Patient 10 because of ES during the previous cycle, and Patient 16 opted for allogeneic transplantation instead. Among the 13 patients, the median interval between the two infusions was 1.4 months (IQR: 1.2–1.9 months, not exceeding 2 months) (Table S1). All patients voluntarily terminated therapy following finite‐duration VA maintenance intervention.

TABLE 1.

clinical characteristic of 16 patients.

No	Age/Sex	diagnosis	WBC (×109/L)	BM blasts(%)	Chromosome karyotype	2022 ELN risk
1	68/F	M2	0.8	42	46,XX[20]	Adverse
2	74/M	M2	1.5	23	47,XY,del(7)(q21),+8[6]/48,idem,del(20)(q11q13)[1]	Adverse
3	70/M	M2	12.3	71	46,XY,del(20)(q11q13)[1]/46,XY[2]	Favorable
4	70/M	M0	1.4	35	46,XY[20]	Intermediate
5	62/F	M4	2.0	63	46,XY[20]	Favorable
6	72/M	M5	3.0	30	46,XY,t(12;22)(p13;q12)[10]/46,XY[2]	Adverse
7	62/F	M2	1.0	53	46,XX,+8[1]/47,idem,+X,+mar[3]/46,XX[16]	Adverse
8	73/F	M5	1.3	36	48,XX,+X,+8[2]/46,XY[1]	Adverse
9	65/M	M5	14.0	70	46,XY[20]	Favorable
10	60/M	M2	5.3	21	47,XY,+8[2]/47,idem,del(10)(p12)[1]/46,XY[7]	Intermediate
11	62/M	M5	6.5	45	46,XY[20]	Favorable
12	77/F	M2	44.5	59	46,XX[20]	Favorable
13	68/M	M0	1.6	84	46,XY[20]	Adverse
14	69/F	M5	0.9	57.5	46,XX,del(5)(q21q34)[1]/46,idem,del(12)(p13)[9]	Adverse
15	67/F	M5	1.3	83.5	46,XY[20]	Favorable
16	63/M	M5	1.1	72.5	46,XY[20]	Intermediate

FIGURE 3.

(A) Patient gene mutations identified via next‐generation sequencing technology. (B) Clinical outcomes of 16 elderly patients. (C) Kaplan–Meier curve of DOR. (D) Kaplan–Meier curve of OS.

3.2. UCB Infusion and Chimeric Results

In the first UCB infusion, 16 patients were included. Among these, 3 patients had 5 antigen matches, 8 had 4 matches, and 5 had 3 matches. Post‐freezing analysis revealed a median nucleated cell count of 1.87 × 10⁷/kg (IQR: 1.42–2.34) and a CD34+ cell count of 0.57 × 10⁵/kg (IQR: 0.36–0.68). In the second infusion, 13 patients received UCB infusions, while 3 patients opted out voluntarily. Within this group, there was 1 patient with 5 antigen matches, 6 with 4 matches, and 6 with 3 matches. The post‐freezing analysis showed a median nucleated cell count of 1.92 × 10⁷/kg (IQR: 1.62–2.03) and a CD34+ cell count of 0.44 × 10⁵/kg (IQR: 0.35–0.81). Chimerism was routinely monitored on day 7 post‐UCB infusion. Among the 16 patients, only one (6.3%) established a mixed chimerism level of 7.59%. Two weeks later, the chimerism level was re‐evaluated at 0.56%. Notably, the patient had 5 antigen matches and the highest CD34+ cell count post‐freezing at 0.99 × 10⁵/kg. The remaining 15 patients exhibited very low levels of microchimerism, ranging from 0% to 0.04%.

3.3. Overal Outcomes

All 16 participants achieved sustained CR with MRD (MFC) negativity prior to treatment discontinuation. Patients 5 and 11, respectively, exhibited 0.24% and 0.04% MRD (qPCR) with a downward trend of CBFβ‐MYH11 fusion prior to treatment discontinuation, although they did not reach the level of 0.01%. In the median follow‐up time of 53.6 months, there were 10 patients who relapsed. The median DOR was 17.4 months (95% CI: 7.3–27.5 months), with a 1‐year DOR rate of 68.8% and a 2‐year DOR rate of 35.2% (Figure 3C). The 2 patients with MRD (qPCR) positivity had DORs of 12.1 and 12.6 months, respectively. Subsequent post‐relapse intervention included VA (Patients 1, 6, 7, 13), venetoclax with chemotherapy (Patients 3, 5, 9, and 15), chemotherapy (Patient 8), and targeted therapy (Patient 14). Of the 10 patients who received treatment for relapse, 7 achieved CR again. Overall, the median OS was 24.9 months (95% CI: 9.8–40.1 months), with a 1‐ and 2‐year OS rate of 100% and 55.6%, respectively (Figure 3D). The 2 patients with MRD (qPCR) positivity had OS of 53.6 and 14.6 months, respectively. In particular, among the 8 patients with IDH1/2 mutations, 4 experienced continuing CR. The median OS was 31.2 months (95% CI: Not reached), with a 1‐ and 2‐year OS rate of 100% and 62.5%.

3.4. Adverse Clinical Outcomes During Consolidation Therapy

Table 2 summarizes the adverse clinical outcomes of patients during the consolidation therapy. In all, 16 subjects experienced grade 3–4 neutropenia or thrombocytopenia. The duration between chemotherapy initiation and leukocyte recovery was 18 days and 19 days during the first and second cycles of consolidation therapies, respectively. The median respective neutropenia durations were 6 (IQR: 4–7) and 7 days (IQR: 4–9), while the median thrombocytopenia durations were 6 (IQR: 4–8) and 6 days (IQR: 4–7). Among the prevalent hematological and non‐hematological toxicities were febrile neutropenia and infection.

TABLE 2.

A summary of toxicity to the consolidation therapy.

	The first cycle (n = 16)	The second cycle (n = 13)
The time from chemotherapy to leukocyte recovery (> 1 × 109/L) (IQR; days)	18 (14‐20)	19 (17‐22)
Median duration of neutropenia (IQR; days) a	6 (4‐7)	7 (4‐9)
Median duration of thrombocytopenia (IQR; days) b	6 (4‐8)	6 (4‐7)
Platelet transfusion (IQR; units)	11 (10‐16)	17 (0‐28)
Febrile neutropenia (Grade 3)	10 (62.5%)	5 (38.5%)
Upper respiratory infection (Grade 3)	1 (6.3%)	0
Lung infection (Grade 3)	0	2 (15.4%)
Nasal infection (Grade 3)	0	1 (7.7%)
Atrial fibrillation (Grade 3)	0	1 (7.7%)
Engraftment syndrome (Grade 3)	1 (6.3%)	0

^a Neutropenia indicated < 0.5 × 10⁹/L.

^b Thrombocytopenia indicated < 20 × 10⁹/L.

During the first consolidation therapeutic cycle (n = 16), 10 subjects experienced Grade 3 febrile neutropenia, 1 (Patient 2) suffered from upper respiratory inflammation, and 1 (Patient 10) exhibited engraftment syndrome (ES). The patient developed ES 10 days post‐infusion, coinciding with a rise in white blood cell count. The syndrome presented as a rash, fever, and abnormal liver function, but improved after one week of glucocorticoid treatment. Upon re‐examination two weeks later, the chimerism rate decreased from 7.59% to 0.56%. Patients 7, 10, and 16 did not provide consent for a second round of consolidation therapy. During the second consolidation therapy (n = 13), 5 subjects experienced Grade 3 febrile neutropenia, 2 (Patients 6 and 11) contracted lung infection, and 1 (Patient 8) exhibited Grade 3 nasal infection. 1 (Patient 12) experienced atrial fibrillation.

4. Discussion

Prior investigations reported strong efficiency of continuous VA‐based treatments in managing elderly AML patients. DiNardo et al. reported a 73% CR/CRi rate associated with a median DOR of 12.5 months and 2‐year OS of 51% among 60 patients who underwent a median of 5 VA maintenance cycles [21]. Candoni et al. revealed an overall response rate (ORR) of 72% and a median OS of 11 months among 132 recipients of continuous VA therapy [2]. Matthews et al. also examined retrospective prognoses of 488 elderly patients who were administered venetoclax, along with hypomethylating agents, and reported a median OS of 10 months and a 2‐year OS of 25% [5]. Apel et al. achieved a 61% CR rate, 25% relapse rate, and an 11.7‐month median event‐free survival among responders [1]. In the large‐scale VIALE‐A study, involving 286 participants receiving a median of 7.0 VA cycles of VA, patients achieved a 66.4% CR rate and a median DOR of 17.5 months [4]. During the median 43.2 months of follow‐up, the aforementioned study demonstrated a median OS of 14.7 months and a 2‐year OS rate of 37.5% [7]. Moreover, emerging evidence suggests that certain mutations, namely, ASXL1, NPM1, IDH2, and DDX41, are robust biomarkers of a favorable response, whereas FLT3‐ITD, RUNX1, and TP53 mutations often result in poor responses [22, 23, 24, 25, 26].

Herein, we utilized the modified approach that consisted of a VA‐based induction therapy, 2 moderate intensity consolidation chemotherapy cycles, followed by VA therapy (median: 5 cycles). All 16 patients achieved a similar DOR of 17.4 months, compared to a DOR of 17.5 months reported in the VIALE‐A study. Significantly, there was an increased risk among patients in our study, with 7 out of 16 (43.8%) experiencing complications, compared to 36% in the VIALE‐A study. Furthermore, the median OS was 24.9 months with a 2‐year OS rate of 55.6%. This median OS was longer than that observed in studies examining continuous VA therapy. Meanwhile, patients with IDH1/2 mutations had a 1‐year OS of 100%, representing an improvement compared to the 1‐year OS of 66.8% reported in the VIALE‐A trial. Owing to the introduction of moderate consolidation chemotherapy, along with a finite VA therapy, the adverse side effects and accumulated myelotoxicity risks were lowered dramatically, which, in turn, enhanced patient prognosis and quality of life, relative to continuous VA therapy. As indicated by the Todisco et al. study, patients who experience relapse can still achieve remission with venetoclax‐based treatments, with about 40% response rates [27]. In this study, relapsed patients received various treatments: VA (n = 4), venetoclax combined with chemotherapy (n = 4), and chemotherapy or target therapy (n = 2). Notably, 7 out of the 10 patients achieved CR again.

Earlier investigations reported great success of consolidation therapy, which was corroborated in the current study [9, 18]. Qin et al. demonstrated that decitabine and cytarabine co‐administration generated a synergistic effect on human leukemia cell apoptosis in vitro [28]. Notably, cord blood T cells have augmented anti‐tumor influences, relative to adult peripheral blood T cells [29]. Prior investigations revealed robust clinical outcomes for older AML patients administered with decitabine plus intermediate‐dose cytarabine, along with UCB infusion [18]. Moreover, the consolidation therapy was generally safe and well‐tolerated by patients, and no fatal complications were reported. Overall, neutrophil and platelet count recovery was rapid, with appropriate UCB infusion likely because of the presence of hematopoietic stem cells in UCB, along with influences from decitabine, which augments platelet release and megakaryocyte maturation [30]. The prompt blood cell restoration, in turn, prevented further complications, namely, infection and bleeding. Herein, only mild or moderate infections (those impacting the upper respiratory tract, lung, and nose) were observed. In a phase 2 study, wang et al. [31] found that consolidation chemotherapy combined with UCB infusion achieved a median event‐free survival of 72.2 months and a 2‐year OS rate of 76.9% in elderly AML patients. Compared with our study, the 2‐year survival data were relatively better, possibly due to the enrolled patients having received induction and maintenance chemotherapy. Furthermore, this study revealed the mechanism of action of umbilical cord blood through single‐cell RNA sequencing, showing enhanced anti‐tumor and anti‐aging properties that are manifested through activation of immune responses and telomere synthesis/maintenance. However, it should be emphasized that the mechanistic role of umbilical cord blood infusion in this combination regimen remains uncertain, and the observed clinical outcomes cannot be definitively attributed to this component. That said, the consolidation treatment with UCB infusion could still be viewed as one potentially effective and safe post‐remission adjuvant approach for elderly AML patients.

In the study, one patient experienced symptoms of ES 10 days after transplantation. ES is a common occurrence post‐transplant and is linked to a higher likelihood of acute graft‐versus‐host disease (GVHD), chronic GVHD, non‐relapse mortality, and reduced overall survival (OS) [32]. The patient's symptoms were likely influenced by a five‐loci match between their HLA typing and the cord blood, alongside the highest recorded number of CD34 cells. Following one week of glucocorticoid treatment, the patient showed significant improvement. As a result, our center decided to prioritize matching at 3 or 4 loci, in conjunction with the median number of CD34 cells for future patients. To date, there have been no additional cases of ES.

Our analyses and conclusions have several important limitations that should be acknowledged. First, the sample population was relatively small, which may have introduced bias and limited the statistical power of our findings. Second, the absence of a control group precludes definitive conclusions regarding comparative efficacy and represents a major limitation. Third, the single‐arm, non‐randomized design potentially introduced selection bias, as the inclusion of frail elderly patients from a single center may limit the generalizability of our results. Consequently, the present findings should be interpreted as hypothesis‐generating, and larger, controlled studies are needed to confirm the observed outcomes. Of note, despite these limitations, the current work has inspired an ongoing phase II clinical trial examining the safety and efficacy of another regimen involving decitabine priming with cytarabine, followed by UCB infusion after VA‐induced remission.

Considering the strong evidence presented in this study, we speculate that the new technique involving a finite duration VA and decitabine priming with intermediate‐dose cytarabine, with subsequent UCB infusion, can improve survival, particularly in patients with IDH mutations. Thus, this is an excellent candidate intervention for frail elderly AML patients. This work would require additional validation with large‐scale and multi‐center patient populations.

Author Contributions

Liya Ma: resources. Xinping Zhou: data curation. Li Ye: writing – original draft, data curation, formal analysis, investigation. Yanling Ren: resources. Gaixiang Xu: data curation, investigation, formal analysis, writing – original draft. Chunmei Yang: resources. Jie Jin: project administration. Chao Hu: data curation. Chen Mei: data curation. Mixue Xie: data curation, investigation, writing – original draft, formal analysis. Jiejing Qian: project administration. Hongyan Tong: funding acquisition, resources, supervision, writing – review and editing.

Funding

This work was supported by the National Natural Science Foundation of China (82270146) and the Key R&D Program of Zhejiang, China (2024C03164).

Ethics Statement

This study was approved by the ethics committee of the First Affiliated Hospital, College of Medicine, Zhejiang University (IIT20240575A), and was conducted in accordance with the principles of the Declaration of Helsinki and Good Clinical Practice guidelines issued by the International Conference of Harmonization.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

TABLE S1: The date of patient enrollment, consolidation therapy, relapse, or death.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.