Venetoclax Combined with Azacitidine and Homoharringtonine in Relapsed/Refractory AML: A Multicenter, Phase 2 Trial

In this multicenter, phase 2 trial, patients were enrolled at ten hospitals in China. Patients with R/R AML (aged 18–65 years) who had an Eastern Cooperative Oncology Group performance status of 0–2 were eligible for this trial. Refractory AML was defined as no CRc and a reduction in bone marrow (BM) blasts of less than 50% after one cycle or no CRc after two cycles. Relapsed AML was defined as recurrence of blasts in the peripheral blood (PB) or BM blasts ≥ 5% or development of extramedullary disease after achieving CRc [21]. Patients were excluded if they previously received venetoclax-based treatment, had acute promyelocytic leukemia, pregnancy, active acute or chronic graft-versus-host disease (GVHD), clinically significant coagulation abnormalities, clinically significant cardiovascular disease, uncontrolled active infection, substantial history of renal, neurological, pulmonary, psychiatric, endocrine, metabolic, immunological, hepatic, or any other medical conditions not suitable for the trial. Active acute GVHD or chronic GVHD was defined as GVHD requiring either at least 1 mg/kg per day of prednisone (or equivalent) or treatment beyond systemic corticosteroids [22]. The study protocol (Additional file 2: Appendix) was approved by the ethics committee review board at each of the ten hospitals, and written informed consent was obtained from patients or guardians in accordance with the Declaration of Helsinki before the initiation of the study.

VAH regimen consisted of 14-day courses of venetoclax and 7-day courses of azacitidine and homoharringtonine. Venetoclax began at 100 mg on day 1 and increased stepwise over 3 days to reach the target dose of 400 mg (100, 200, and 400 mg); dosing was continued at 400 mg per day from day 4 through day 14; azacitidine (75 mg/m²) and homoharringtonine (1 mg/m²) were administered subcutaneously on days 1 to 7. All patients were hospitalized during the treatments. Venetoclax dose was reduced by at least 50% in the patients receiving concomitant moderate or strong CYP3A4 inhibitors (e.g., azole antifungals) [23]. Fms-related receptor tyrosine kinase 3 (FLT3) inhibitors were recommended in the patients with FLT3 mutations before allo-HSCT. For the patients undergoing allo-HSCT, FLT3 inhibitors maintenance post-transplantation was recommended regardless of FLT3 being mutated or not. Once CRc was achieved, patients received allogeneic hematopoietic stem cell transplantation (allo-HSCT) if donors were available. If donors were unavailable, patients received one course of original therapy again and two or three cycles of cytarabine-based consolidation therapy. After consolidation, if patients have actionable target, they will receive targeted inhibitor maintenance such as FLT3 mutated patients receive sorafenib maintenance. If not, patients will receive pre-emptive treatment according to MRD detection. If patients did not reach CRc after two courses, they proceed to allo-HSCT if donors were available. If donors were unavailable, the patients might receive other salvage therapy. For patients who relapsed after allo-HSCT, donor lymphocyte infusion (DLI) was administered on day 15 after the initiation of the original therapy if donors were available, and the second DLI depended on GVHD status. The CD3⁺T cell count for each DLI was 3.0 × 10⁷/kg of the recipient weight. DLI was given monthly until GVHD occurred or MRD became negative or for a total of four times [24].

Criteria for removing patients from trial treatment included the development of intolerable adverse events related to study treatment, patient withdrew informed consent, and completion of the protocol therapy and evaluation period. The cytogenetic evaluation was used with standard metaphase karyotype and fluorescence in situ hybridization analysis. Molecular analysis via polymerase chain reaction and 167-gene institutional next-generation sequencing platform was performed at study enrollment. Measurable residual disease (MRD) was assessed by 8-color multiparameter flow cytometry (FC) using leukemia-associated immunophenotype or different from normal assessment with a minimum sensitivity of 10^–4. The MRD levels of 0.01% were used as a threshold to distinguish MRD positive (MRD⁺) from MRD negative (MRD⁻) [25, 26].

For response assessments, BM was evaluated at cycle 1 day 28 and again 1–2 weeks after hematologic recovery if on day 28 BM was aplastic. Subsequent BM evaluations were done before and after cycle 2 and then, as clinically needed. Morphologic, cytogenetic, and MRD assessments were done during each BM assessment. Response criteria were defined by the European LeukemiaNet (ELN) 2017 guidelines. CR was defined as an absolute neutrophil count of more than 1000 cells per cubic millimeter, a platelet count of more than 100,000 per cubic millimeter, red-cell transfusion independence, and BM with less than 5% blasts. CRi was defined as all the criteria for CR, except for neutropenia (absolute neutrophil count, ≤ 1000 per cubic millimeter) or thrombocytopenia (platelet count, ≤ 100,000 per cubic millimeter) [21]. Partial remission (PR) was a minimal residual disease of 5% to 25% with a greater than 50% decrease in leukemic blast percentage. Non-remission (NR) was defined as a failure to obtain CRc or PR [21]. CRc comprised CR and CRi, and overall response rate (ORR) comprised CRc and PR.

Adverse events were defined as those that occurred from the first dose until 28 days after the discontinuation of treatment and were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0 [27]. An independent study adjudication committee (consisting of experts in hematology, infection, pathology, pharmacy, and statistics) judged whether adverse events were treatment-related or non-treatment-related.

The primary endpoint was CRc after 2 cycles of trial treatment. The secondary endpoints were safety, overall survival (OS), event-free survival (EFS), disease-free survival (DFS), and relapse. OS was defined as the time from treatment until death or censored at the last follow-up. EFS was defined as the time from treatment until documented failure to achieve CRc, relapse after CRc, or death from any cause, whichever occurred first. DFS was defined as the time from the date of CRc until relapse, death from any cause, or censored at the last follow-up. Data for each patient were censored at the date of the last visit or the date on which the patient was last known to be alive.

The sample size calculation for the trial was based on the assumption that the VAH regimen would achieve a higher CRc rate compared with historical CRc rate of 45% (on the basis of venetoclax in combination with HMAs study by our previous and others in R/R AML) [11, 20]. To identify a 15% absolute improvement in CRc with VAH regimen, a total of 87 patients were required to provide the study with a significance level of 5% and a power of 80%. After adjusting for a 10% dropout, the total planned sample size was 96 patients. The sample size calculation was done using PASS software, version 11·0.

The clinical data cutoff date was June 30, 2022. The descriptive analysis of patient characteristics included median and interquartile range (IQR) for continuous variables, and absolute and relative frequencies for categorical variables. The time-to-event endpoints including OS, EFS and DFS were analyzed by the Kaplan–Meier method and compared using the log-rank tests. The corresponding hazard ratio (HR) and 95% CI were estimated using the Cox proportional hazards model. The cumulative incidences of relapse were calculated by accounting for competing risks, and non-relapse mortality was a competing risk for relapse. The comparison of the cumulative incidence in the presence of a competing risk was done using the Fine and Gray model [28]. All variables in the Cox models were tested for proportional hazards assumption. Variables included in the univariable analysis were age, gender, AML status, ELN classification, MRD status, and allo-HSCT. Factors that were significant at the 0.1 level from the univariable model were included in the multivariable model. All statistics were analyzed in software R version 4.1.0 (R Development Core Team, Vienna, Austria) or Stata version 15.1 (StataCorp 4905 Lakeway Dr College Station, TX77845, USA) or SPSS version 22.0 (SPSS, Chicago, IL). All statistical tests were two-tailed with a significance level of 0.05. This trial is registered with ClinicalTrials.gov (NCT04424147).

Between May 27, 2020, and June 16, 2021, 108 patients with R/R AML were assessed for eligibility, 96 of whom were enrolled, including 37 (38.5%) patients with primary refractory AML and 59 (61.5%) with relapsed AML (16 (16.7%) relapsed after chemotherapy and 43 (44.8%) after allo-HSCT) (Fig. 1). There were 51 male (53.1%) and 45 female (46.9%), with a median age of 45 (IQR, 33–55) years at enrollment. Patient characteristics are shown in Table 1. Twenty-one (56.8%) patients with refractory AML had received one cycle, and 16 (43.2%) patients received two or more cycles of inducing therapy. Among patients with relapse after chemotherapy, 14 (87.5%) patients were first relapse, and 2 (12.5%) were second relapse. Among patients with relapse after allo-HSCT, 2 patients (4.65%) relapsed after a second allo-HSCT. Of the 96 patients enrolled, 45 patients (46.9%) received one cycle and 51 (53.1%) received more than one cycle, including 42 (82.4%) two and 9 (17.6%) three cycles as trail treatment. Among 19 patients with FLT3 mutations, 17 patients (89.5%) received sorafenib, and 2 (10.5%) gilteritinib treatment. More patients relapsed after allo-HSCT received more cycles of VEN-based treatments other than second allo-HSCT compared with primary refractory patients and patients relapsed after chemotherapy (P = 0.01 and P < 0.01).

The responses of treatments are summarized in Table 2 and Fig. 2. CRc rate at the end of cycle 2 was 70.8% (68 of 96 patients; 95% CI 60.8–79.2), with CRc at the end of cycle 1 was 58.3% (56 of 96 patients; 95% CI 48.1–67.9). In the patients with CRc, MRD⁻ was attained in 58.8% of CRc patients. Accordingly, ORR at the end of cycle 2 was 78.1% (75 of 96 patients; 95% CI 68.6–85.4), with ORR at the end of cycle 1 was 71.9% (69 of 96 patients; 95%CI 61.9–80.1). There were no difference in CRc and ORR among patients with or without prior HMAs exposure (CRc, 63.9% (95%CI 46.8–78.1%) vs 75.0% (62.3–84.5%), p = 0.246; ORR, 75.0% (95%CI 58.0–86.7%) vs 80.0% (67.7–88.4%), p = 0.566). Of the 68 patients reached CRc, 36 patients (52.9%) proceeded to allo-HSCT, 31 (45.6%) received a median of 3 courses (range 1–4) of consolidation chemotherapy, and 1 patient (1.5%) gave up further consolidation treatment due to personal reason. Of the 28 patients who did not obtain CRc, 20 patients (71.4%) received other salvage treatments (16 chemotherapy and 4 allo-HSCT), and 8 (28.6%) gave up treatments or died. Of the 20 patients receiving other salvage treatments, 6 patients (30.0%) obtained CRc (4 allo-HSCT and 2 chemotherapy). Among the 40 patients bridged to allo-HSCT, 36 (90.0%) received sorafenib maintenance post-transplantation, except 4 patients (10.0%) (2 patients GVHD and 2 patients haematotoxicity). Of the 43 patients relapsed after allo-HSCT, 34 patients (79.1%) received a median of two DLI (range 1–2). Twenty-six patients (76.5%) developed the acute GVHD, and six patients (17.6%) developed the extensive chronic GVHD among 34 patients who received DLI treatment.

The results of cytogenetic and molecular subgroup analyses with respect to CRc are shown in Additional file 1: Table S1 and Fig. 3. CRc was 94.1% (95%CI 66.0–99.2), 63.6% (41.6–81.2) and 66.7% (53.3–77.8) in the ELN favorable-, intermediate- and adverse-risk groups, respectively (P = 0.064). The CRc in favorable-risk group was significantly better than intermediate- or adverse-risk groups (P = 0.025; P = 0.025, respectively). In the cytogenetic subgroups, CRc was 100%, 77.4% (63.9–86.8) and 55.6% (36.3–73.2) in the favorable-, intermediate- and poor-risk groups, respectively (P = 0.066). The CRc in poor-risk group was interior than favorable- or intermediate-risk groups (P = 0.044; P = 0.041, respectively). We further analyzed the molecular mutation subgroups which mutation rate was more than 5% according to related literature report and our samples size [29]. Nine patients (9.4%) harbored K/NRAS^mut and 7 patients (7.3%) harbored MLL^mut. The CRc was 44.4% (16.3–76.7) for K/NRAS^mut and 42.9% (12.7–79.4) for MLL^mut. Except for K/NRAS^mut and MLL^mut, the CRc were above 50% in all the other molecular mutation. The CRc was not significantly different between patients with and without these mutations, except for K/NRAS^mut and MLL^mut (Additional file 1: Fig. S2).

Within a median follow-up of 14.7 months (IQR, 6.6–22.8), 43 patients (44.8%) died. The most common causes of death were relapse/disease progression (n = 36, 83.7%), infections (sepsis [n = 2], pneumonia [n = 1], febrile neutropenia [n = 1]), GVHD (n = 2), and heart failure (n = 1). The median OS was 22.1 months (95% CI 12.7–Not estimated (NE)), including the median OS was not reached (95% CI 22.1–NE) for patients achieving CRc and 3.8 months (95% CI 3.0–6.3) for patients who did not achieve CRc. The median EFS was 14.3 months (95% CI 7.0–NE). The 1-year OS was 61.5% (95% CI 51.0–70.4), and 1-year EFS was 51.0% (95% CI 40.7–60.5) (Table 2 and Fig. 4A, B). With the median time of 5.4 months (IQR 2.4–6.9), 14 patients (20.6%) relapsed among the 68 patients who achieved CRc. The 1-year cumulative incidence of relapse was 19.4% (95% CI 11.2–29.4). Accordingly, 1-year DFS was 69.4% (95% CI 57.4–78.7) (Fig. 4C, D). The OS, EFS, relapse and DFS of disease subgroup are shown in Additional file 1: Fig. S3. Of the 96 patients, 40 patients (41.7%) bridged to allo-HSCT, including 36 patients (90.0%) achieved CR and 4 patients (10.0%) did not achieve CR. The 1-year OS was 85.0% (95% CI 69.6–93.0) versus 44.6% (95% CI 31.4–57.0) for patients who received or did not receive allo-HSCT (P < 0.001, Fig. 5A) The 1-year EFS was 75.0% (95% CI 58.5–85.7) versus 33.9% (95% CI 22.0–46.3) (P < 0.001). The 1-year cumulative incidence of relapse was 15.0% (95% CI 6.0–27.8) and 25.0% (95% CI 11.6–41.0) for patients who received or did not receive allo-HSCT (P = 0.375). Accordingly, 1-year DFS was 77.5% (95% CI 61.2–87.6) versus 59.4% (95% CI 40.5–74.0) (P = 0.074) (Fig. 5B–D). Among patients who reached CRc, the 1-year OS was 91.7% (95% CI 76.4–97.2) and 71.9% (95% CI 52.9–84.3; P = 0.023) for patients who received or did not receive allo-HSCT. Accordingly, the 1-year DFS was 83.3% (95% CI 66.6–92.1) and 59.4% (95% CI 40.5–74.0), respectively (P = 0.016) (Additional file 1: Fig. S4A, B). Exploratory subgroup analyses showed that 1-year OS was 58.4% (95% CI 46.6–68.5) for patients without FLT3^mut and 73.7% (95% CI 47.9–88.1) for patients with FLT3^mut (P = 0.266). The 1-year EFS was 51.9% (95% CI 40.3–62.4) and 47.4% (95% CI 24.4–67.3), respectively (P = 0.823). The 1-year cumulative incidence of relapse was 16.4% (95% CI 8.0–27.4) for patients without FLT3^mut and 29.4% (95% CI 10.1–52.0) for patients with FLT3^mut (P = 0.248). Accordingly, 1-year DFS was 72.7% (95% CI 58.9–82.6) and 58.8% (95% CI 32.5–77.8), respectively (P = 0.258). (Additional file 1: Fig. 5A–D). A post hoc multivariable analysis showed that proceeding to allo-HSCT and MRD⁻ were protective factors for OS (HR 0.36 (95%CI 0.13–0.98); P = 0.046, and HR 0.35 (95%CI 0.13–0.93); P = 0.035. Additional file 1: Table S2).

Common adverse events are summarized in Table 3. The most frequently reported hematologic adverse events of grade 3 or higher included neutropenia (82.3%), thrombocytopenia (75.0%), anemia (66.7%), and febrile neutropenia (37.4%). Gastrointestinal adverse events of any grade were common and predominantly included nausea (26.0%), constipation (12.5%), diarrhea (11.5%), and vomiting (12.5%). Notable serious adverse events (grade ≥ 3) were febrile neutropenia (37.4%), sepsis (11.4%), pneumonia (21.9%), and heart failure (4.1%). Tumor lysis syndrome was reported during the ramp-up period (on days 1 through 3 when the dose of venetoclax was increased) in 2 patients (2.0%).

The percentage of patients who discontinued VAH owing to adverse events was 9.4%, including sepsis (n = 3), pneumonia (n = 3), heart failure (n = 2), and bleeding (n = 1). The delay between cycles and dose reduction owing to adverse events occurred in 29.2% of the patients. The delay between cycles and dose reduction was primarily because of myelosuppression, including neutropenia (5.2%), febrile neutropenia (19.8%), and thrombocytopenia (4.2%). Treatment-related death was 4.2%, including sepsis (n = 1), pneumonia (n = 1), febrile neutropenia (n = 1), and heart failure (n = 1).