AML patients who had failed at least one cycle of intensive induction chemotherapy or developed relapse prior to receiving venetoclax combination therapy were eligible for this study. All enrolled patients were treated at our hospital between March 2021 and March 2023, voluntarily enrolled in this clinical trial, and provided informed consent on specified forms themselves or through their legal guardians. The patients were screened according to the following criteria: AML diagnosis was in accordance with the 5th edition of the World Health Organization (WHO) diagnostic criteria for 2016; patients had received at least 1 cycle of treatment with the VEN-DCAG regimen; an Eastern Cooperative Oncology Group (ECOG) score ≤3; ages between 10 and 70 years at the time of screening, regardless of gender; no severe allergies or psychiatric-related diseases; liver function: ALT or AST ≤ 2.5 times the upper limit of normal, bilirubin ≤ 2.0 times the upper limit of normal; renal function: creatinine within the upper limit of normal; no significant cardiac dysfunction (e.g., severe cardiac failure or cardiac arrhythmia); and no uncontrollable active infection. Exclusion criteria for patients included acute promyelocytic leukemia, pregnancy, breastfeeding, a history of autoimmune disease, chronic smoking or alcohol use that could interfere with trial evaluation, and investigator-determined unsuitability for the clinical trial (e.g., poor medical adherence, surgery within the last 6 weeks, etc.). All patients were diagnosed through bone marrow cytomorphology, leukemia immunophenotyping, cytogenetics, and molecular biology.

This study aimed to explore the efficacy and safety of the combination of venetoclax, hypomethylating drugs, and CAG regimen in the treatment of R/R AML patients. It received approval from the Ethics Committee of the General Hospital of the People’s Liberation Army (S2022-740-01) and was registered as a clinical trial (ChiCTR2300075985).

Twenty patients with R/R AML were enrolled: 17 receiving the VEN + decitabine (DAC) + CAG regimen, and 3 receiving the VEN + azacitidine (AZA) + CAG regimen. The VEN administration consisted of 400 mg/d or 100 mg/d (combined with azole antifungals), d1-14, taken orally. DAC 20 mg/m², d1-5, IV; AZA 50-75 mg/m², d1-7, subcutaneously; CAG regimen: cytarabine 50-100 mg/m²/d, d1-5, IV; aclarubicin 20 mg/d (6 cases d1/3/5, 14 cases d1-5), IV; G-CSF 300 μg/d or polyethylene glycol long-acting colony-stimulating factor (PEG-CSF) 6 mg once a week; and G-CSF or PEG-CSF was discontinued when the white blood cell count (WBC) was ≥10×10⁹/L. Dose reduction in cytarabine or azacitidine were considered for patients with particular conditions: bone marrow hypoproliferation; advanced age (>60), or those unable to tolerate standard-dose chemotherapy, as well as individuals with heart, lung and kidney complications. Treatment was continued as long as the patient demonstrated clinical benefit, and symptomatic supportive treatment, such as blood products transfusion or anti-infection measures, was offered according to the actual situation.

Patients were primarily assessed for 1-cycle efficacy and total efficacy, which included evaluating the complete remission/complete remission with incomplete hematologic recovery (CR/CRi) rate, partial remission (PR), no remission (NR), measurable residual disease (MRD) negative rate, overall survival (OS), event-free survival (EFS), and success rate of subsequent bridging HSCT. The safety assessment primarily included examining the occurrence of adverse events (AEs) taking place after VEN-DCAG treatment, with determinations made through clinical symptoms, laboratory tests, imaging examinations, and others. The adverse event severity grading standard was adopted from the National Cancer Institute (NCI) adverse event grading (CTC-AE) version 4.03 (12).

The main definitions are as follows: CR: Disappearance of signs and symptoms of leukemia, absence of leukemic cells in the leukocyte classification, <5% primitive cells in the bone marrow, absence of extramedullary leukemia, and neutrophil count ≥1.0×10⁹/L, PLT ≥100×10⁹/L. CRi: Disappearance of signs and symptoms of leukemia, absence of leukemic cells in the leukocyte classification, <5% primitive cells in the bone marrow, absence of extramedullary leukemia, and neutrophil count <1.0×10⁹/L or PLT <100×10⁹/L. PR: Bone marrow primitive cells decreased to 6%-20% or decreased over 50% from pretreatment. NR: Non-fulfillment of the above criteria. MRD negative status was defined as negative leukemia immunotyping by flow cytometry (FCM). Relapse was defined as the presence of >5% bone marrow primitive cells or the presence of extramedullary infiltrates after achieved CR (13). Refractory was defined as failure to achieve CR with at least one courses of previous-line induction chemotherapy. OS: The time from entry into VEN combination therapy until death from any cause or last follow-up. EFS: The time from entry into VEN combination therapy until disease progression, relapse, death from any cause, or last follow-up. The overall response rate (ORR) was calculated as CR + CRi + PR.

Patient statuses were acquired through telephone follow-ups and a review of outpatient or inpatient records. The follow-up cutoff date was April 1, 2023. The endpoint of the follow-up was either the follow-up cutoff or death from any cause. Data were analyzed using SPSS 26.0 software, and outcome data have been expressed as the median (range) or number (percentage). Survival curves were plotted using the Kaplan−Meier method, and survival was compared between groups using the log-rank test. A P-value < 0.05 was considered to indicate a statistically significant difference.

The median age of the 20 patients was 40 years (10-70), including 11 (55%) male and 9 (45%) female ( Table 1 ). There were 17 cases of primary AML, 1 case was secondary to myelodysplastic syndrome (MDS); 1 case emerged 5 years after aplastic anemia unrelated donor transplantation secondary to AML, and 1 case pertained to treatment-related AML (anterior mediastinal germ cell tumor). The patients were grouped according to AML disease stage: 13 had refractory disease, 7 showed relapse, with 2 cases of relapse occurring after allo-HSCT. According to the ELN 2022 version of AML genetic risk stratification, 5 cases in the favorable prognosis group, 2 cases in the intermediate prognosis group, and 13 cases in the adverse prognosis group. Among these patients, 11 had received 1 course of previous treatment, 9 had received >1 course of treatment, 12 had received HMA treatment, 3 had received venetoclax, and 2 cases had received allo-HSCT. The median percentage of bone marrow (BM) primitive cells before treatment was 27.65 (4.0-76.8)%, the median hemoglobin (HGB) level was 78 g/L (45-135), the median white blood cell (WBC) count was 2.46 (0.59-64.51) × 10⁹/L, and the median platelet (PLT) count was 48 (6-873) × 10⁹/L. Chromosomal karyotyping showed 2 cases with a favorable prognosis karyotype, 13 cases with an intermediate prognosis karyotype, and 4 cases with an adverse prognosis karyotype. Reverse transcription RT−PCR molecular biology detected the AML1-ETO fusion gene in 3 cases each, the MLL-AF6 fusion gene and MLL-MLL fusion gene in 2 cases each, and the MLL-SEPTIN9 fusion gene and FUS-ERG fusion gene in 1 case each. Second-generation gene sequencing (NGS) detected 7 cases of IDH1/2 mutations, 6 cases each of DNMT3A and WT1 mutations, 3 cases of U2AF1 mutations, 2 cases each of NARS, ASXL1, TP53, BCOR, TET2, RUNX1, and 1 case each of BCORL1, NMP1, CEBPA, C-KIT and EP300 (nonsignificant mutations have been omitted) ( Figure 1 ).

The ORR of 20 patients after 1 cycle of treatment was 90% (18/20), including 17 cases of CR/CRi (85%), 1 case of PR (5%), 2 cases of NR (10%), and 10 cases of MRD-CR (58.8%). Twelve patients (7 MRD-CR, 4 MRD+CR, 1 NR) received the follow-up VEN-DCAG regimen, with 10 patients undergoing 2 cycles, 1 patient completing 4 cycles, and 1 patient undergoing 5 cycles. Three of the 4 patients with 1-course MRD+CR were converted to MRD-CR after the 2nd course of VEN-DCAG, and 1 patient with NR had disease progression during the follow-up treatment. Finally, of the 20 patients, 17 achieved CR/CRi and 13 reached MRD-CR (76.5%) ( Table 2 ).

There were 7 patients in the relapse group, and all 7 patients achieved CR after 1 cycle, including 5 MRD-CR and 2 MRD+CR (1 MRD+CR was converted to MRD-CR with subsequent VEN-DCAG treatment). Among the two patients who had relapsed after allo-HSCT (both 5 years after transplantation), MRD-CR was reached after 1 cycle of VEN-DCAG. There were 13 patients in the refractory group, 10 patients (5 MRD-CR, 5 MRD+CR) achieved CR, 1 reached PR, and 2 witnessed NR after 1 cycle of treatment. Eight patients (4 MRD-CR, 3 MRD+CR, 1 NR) received follow-up treatment, of whom 2 MRD+CR patients converted to MRD-CR after the second cycle. 1 NR patient deteriorated during follow-up treatment. Eventually, the MRD negative remission rates in the relapse group and refractory group were 85.7% and 70.0%, respectively, with P > 0.05.

According to the ELN 2022 hazard stratified subgroup analysis, the MRD negative remission rates in the favorable prognosis, intermediate prognosis, and adverse prognosis groups were found to be 40.0%, 100.0%, and 90.0%, P=0.070, respectively, after treatment with VEN-DCAG. Further analysis of the differences in gene mutations’ effect on efficacy demonstrated no statistically significant variations between the different mutations due to the sample size limitations. Analyzing the effect of concomitant MLL fusion gene mutations on the efficacy, the results showed that the ORR with or without this fusion gene was 60.0% and 100%, P=0.053, and the CR rate was 60.0% and 93.3%, P=0.140, respectively. Following the chromosomal karyotype subgroup analysis results, the MRD negative remission rates after 1 course of treatment with this combination regimen in the favorable prognosis karyotype, intermediate prognosis karyotype, and complex karyotype groups were 100.0%, 25.0%, and 66.7%, respectively (P=0.011), and the final MRD negative remission rates were 100.0%, 50.0%, and 100.0%, respectively (P=0.086). Analysis of whether prior medication with VEN or HMAs affected efficacy highlighted that the receipt of HMAs was not significantly correlated with efficacy, whereas the ORR and CR rate were significantly higher in patients with no prior VEN treatment than in those who had received VEN treatment (ORR: 100.0% and 33.3%, P=0.016, CR rate: 94.1% and 33.3%, P=0.046). Further, analysis of whether age (≥50 years), disease type, and number of prior courses (≥2) affected efficacy showed no statistically significant difference in CR rate or MRD-CR rate.

Among the 3 patients without CR, there was 1 NR patient, a male, 70 years old, positive for MLL-AF6 fusion gene, 76% quantification of fusion gene, NR after 2 cycles of VEN-AZA treatment, NR after the 3rd course of VEN-DAC-HHT (homoharringtonine), and still NR after the 4th course adjusted to VEN-DCAG regimen, who finally died of disease progression. Another NR patient, a male, 37 years old, M4, with TET2, ASXL1 and U2AF1 mutations, VAF values of 50.95%, 35.78% and 42.47%, respectively, and positive MLL/MLL fusion gene, PR after the 1st course of VEN-DHAG, NR after the 2nd course of VEN-AZA-CAG, and still NR after the 3rd course of adjustment to VEN-DCAG, finally exhibited disease progression, multiple organ failure and death. A PR patient, a male, 20 years old, with an anterior mediastinal germ cell tumor with multiple metastases in the lung, was diagnosed with AML in the 2nd year of treatment by bone marrow (BM) examination due to thrombocytopenia. The chromosome examination revealed a complex karyotype (46-50, xy, +x [28], -4[28], del(5)(q31)[27], del(6)(q?23)[28], +9[27],?del(10)(p13)[5], +11[27], add(12)(q?24)[27],?add(15)(q22)[27], +21[20],+22[27][cp30]) and del(5q) positivity by FISH. The mutation sites detected by NGS were TP53 Exon8 c.817C>T p. R273C, TP53 Exon5 c.396G>T p. K132N, EP300 Exon6 c.1519A>G p. S507G, VAF values were 40.2%, 37.8%, 64.3%, respectively. There was NR after the 1st course of AZA-CAG and PR after the 2nd course of VEN-DCAG, but the anterior mediastinal tumor enlarged rapidly during treatment; both malignant tumors progressed, and the patient died during subsequent treatment.

Adverse events (AEs) were observed in all patients ( Table 3 ). Grade ≥3 AEs were predominantly hematologic AEs, including leukopenia in 20 cases (100%), neutropenia in 20 cases (100%), thrombocytopenia in 19 cases (95%), and hemoglobin reduction in 19 cases (95%). The primary nonhematological AEs were 17 (85%) fevers in the granulocyte deficiency phase, 7 (35%) infections, including 4 cases of pneumonia, 2 skin and soft tissue infections, and 1 positive blood culture; 12 (60%) gastrointestinal discomfort reactions, including nausea, vomiting, diarrhea, abdominal pain, and constipation; 7 (35%) electrolyte disturbances, 4 (20%) skin rashes, 3 (15%) cases of skin/mucosal or nasal bleeding, and 2 (10%) cases each of abnormal liver function and weakness. The lowest median HGB value after treatment was 59.5 g/L (45.0-84.0), the lowest median WBC value was 0.22 × 10⁹/L (0.02-1.00), and the lowest median PLT value was 5 × 10⁹/L (1-97). Two NR patients had no signs of improvements after being given blood products and anti-infective treatment and died within 1 month after VEN-DCAG treatment while the remaining patients showed improvements after aggressive symptomatic treatment.

The median cycle of VEN-DCAG treatment for all patients was 2 (1-5) cycles, with a median follow-up time of 10.4 (0.7-21.8) months and a median EFS of 9.2 (0.7-21.8) months. Eleven patients (7 MRD-CR, 4 MRD+CR) were subsequently bridged to allo-HSCT.

By the end of follow-up, 14 of 20 patients survived (12 disease-free and 2 relapsed), with a median follow-up of 12.7 (2.6-21.8) months in 14 surviving patients; 6 died (1 after transplantation and 5 without transplantation), with a median OS of 4.35 (0.7-18.1) months in 6 deceased patients. 7 patients in the relapse group had no death and 6 of 13 patients in the refractory group died, P=0.051. Further, 10 of 11 patients who received allo-HSCT survived, and 4 of 9 patients who did not receive allo-HSCT survived (90.9% vs. 44.4%, P=0.05) (details in Table S1 ). Survival analysis showed no statistically significant difference (P > 0.05) in the comparison of OS and EFS between patients in the relapse and refractory groups ( Figure 2 ) but a statistically significant difference (P = 0.048) in the comparison of OS between those who received allo-HSCT and those who did not receive allo-HSCT ( Figure 3 ).