Venetoclax combined with escalating doses of homoharringtonine, low-dose cytarabine, and granulocyte colony-stimulating factor demonstrates feasibility and tolerability for remission induction in pediatric acute myeloid leukemia

Advances in supportive care and intensified treatment regimens, including hematopoietic stem cell transplantation (HSCT), have markedly improved outcomes for pediatric acute myeloid leukemia (AML). However, event-free and overall survival rates have plateaued at approximately 65% and 80%, respectively.^1,2 Relapsed disease, as well as acute and long-term toxicities, remain significant challenges.

Our AML collaborative group has prioritized developing and evaluating low-dose chemotherapy (LDC) regimens for remission induction in pediatric AML.³ A recent randomized study demonstrated that an LDC regimen for remission induction was non-inferior to standard-dose chemotherapy.⁴ Nevertheless, exposure to anthracyclines and the number of patients undergoing HSCT remain high. We are currently exploring alternatives to reduce toxicity by incorporating agents with more favorable efficacy and toxicity profiles. Homoharringtonine (HHT), a plant-derived alkaloid (from Cephalotaxus) that inhibits protein synthesis by targeting ribosomes, has been widely used in treating adult and pediatric AML in China.^5-7 Preclinical studies have shown that HHT peak plasma concentrations of 3 mg/m²/day and 5 mg/m²/day exceed levels required to inhibit 50% of HL-60 leukemia cell growth.⁸ Further preclinical evidence suggests that HHT and venetoclax can synergistically promote apoptosis by inhibiting the MAPK/ERK and the PI3K/ AKT pathways while activating the p53 pathway.⁹

Clinically, HHT has been successfully integrated into standard-dose regimens of cytarabine and daunorubicin, demonstrating feasibility and efficacy.⁷ Given the potential synergy between HHT, venetoclax, and cytarabine, we explored replacing mitoxantrone with venetoclax and HHT in the LDC regimen for remission induction. Based on relapse risk, patients received two to three additional courses of standard chemotherapy as consolidation therapy. Details of this new regimen (V-HAG) are provided in Online Supplementary Table S1. Patients at high risk of relapse (criteria outlined in Online Supplementary Table S1) were considered for HSCT. Homoharringtonine was administered in a 3×3 dose escalation design at 1 mg/m² (dose level 1), 2 mg/m² (dose level 2), and 3 mg/m² (dose level 3) daily for ten days to determine the maximum tolerated dose within the context of this regimen. The study is registered under clinicaltrials. gov identifier ChiCTR2200064901. This study was approved by the institutional review board of the Children’s Hospital of Soochow University, and conducted in accordance with the Declaration of Helsinki. Patient data were maintained with strict confidentiality.

Between October 2022 and June 2023, 12 consecutive patients were enrolled in this phase I feasibility study, with 3 patients assigned to dose level (DL) 1, 3 patients to DL 2, and 6 patients to DL 3. The cohort included 7 males and 5 females, with a median age of 8.3 years (range, 3.3-12.7 years). The most common fusion gene identified was RUNX1::RUNXT1, present in 5 patients, followed by KMT2A rearrangements in 2 patients, with one case each of KMT2A::MLLT4 and KMT2A::MLLT10. Other detected fusion genes included CBFB::MYH11 (N=1) and NUP98::NSD1 (N=1). The most frequently identified gene variants were NRAS (N=3), CEBPA double mutant (N=2), and KIT (N=3), located in exon 17 in 2 cases and exon 8 in one case. Additionally, 2 patients had CEBPA single mutations (non-bZip) and 2 had KRAS mutations (Table 1).

All patients received at least one cycle of V-HAG and were evaluated for toxicity and response to Induction I. One patient in DL1 was classified as a non-responder after Induction I, with more than 20% blasts observed in the bone marrow on day 22. This patient subsequently withdrew from the trial. The remaining 11 patients completed both induction courses and achieved remission. These patients proceeded to a median of three consolidation cycles (range, 2-4 cycles), with 5 undergoing allogeneic HSCT. As of August 1, 2024, at a median follow-up of 18 months (range, 11-19 months), all patients were alive and disease-free, including the one patient who withdrew from the protocol. Relevant demographic data are summarized in Table 1.

No dose-limiting toxicities (DLT) or deaths were observed within the first 30 days following the initiation of Induction I. The most common non-hematologic toxicities were febrile neutropenia, nausea or vomiting, lung infections, electrocardiogram (ECG) T-wave changes (inverted T-waves, flattened T-waves, and bidirectional changes), and sinus tachycardia (Table 2). Prolongation of the QT interval was not observed in any of the patients. The median duration of neutropenia (<0.5x10⁹/L) and thrombocytopenia (<20x10⁹/L) during Induction I was 22 days (range, 14-38 days) and 16 days (range, 10-28 days), respectively (Online Supplementary Table S2).

Table 1.Patients’ characteristics and outcome.

Similarly, no DLT were observed among the 11 patients who received Induction II. The most frequent non-hematologic toxicities during Induction II were febrile neutropenia, nausea or vomiting, ECG T-wave changes, sinus tachycardia, and lung infections (Table 2). Importantly, no impairment in cardiac ejection function was observed during Induction II. The median duration of neutropenia and platelet recovery during Induction II was 18 days (range, 7-34 days) and 15 days (range, 0-29 days), respectively (Online Supplementary Table S2). The overall response rate after Induction II was 100%. Furthermore, no severe adverse events (grade 4-5) occurred during either induction phase.

The venetoclax concentration values measured using liquid chromatography-tandem mass spectrometry and peak-to-trough concentration ratios are shown in Figure 1. Venetoclax concentrations were assessed 5-7 days after treatment initiation. The trough concentration was measured 30 minutes before the next dose, while the peak concentration was measured 6 hours (hr) post administration. During Induction I, the median venetoclax peak concentration was 1,375 ng/ mL, and the median trough concentration was 415 ng/mL, yielding a peak-to-trough ratio of 4.1. During Induction II, the median peak concentration increased to 1,740 ng/mL, while the trough concentration was 385 ng/mL, resulting in a peak-to-trough ratio of 4.86. These ratios were used to categorize patients into high- and low-ratio groups. No significant differences between the high- and low-ratio groups were observed in hematologic or non-hematologic toxicities. Regarding treatment response, although a higher proportion of patients in the high-ratio group achieved minimal residual disease (MRD) <1% in Induction I, this difference was not statistically significant (Online Supplementary Table S3).

Our study demonstrates the feasibility of integrating venetoclax and HHT into a regimen of low-dose cytarabine and G-CSF for remission induction in children with de novo AML. This combination was well-tolerated, and the dose-limiting toxicity of HHT was not reached. Therefore, we recommend a 3 mg/m² dose over ten days for future studies. No severe complications, such as septicemia and acute cardiac toxicity, were observed. These findings are consistent with several multicenter clinical trials in China, which also reported the safety and efficacy of HHT in children with AML.^5,6

Table 2.Hematologic and non-hematologic toxicities during Induction I and II.

Cardiovascular side effects of HHT, including heart rhythm abnormalities, transient hypotension, and chronic cardiotoxicity, are rare.¹⁰ The incidence of these effects appears to be influenced by infusion duration and cumulative dosage. Patients receiving continuous HHT infusions experienced fewer cardiovascular complications than those receiving bolus injections.¹¹ Furthermore, patients who were administered a high cumulative dosage of HHT exhibited a higher incidence of cardiac complications than the lowdose group.¹¹ Our study mandated a minimum intravenous infusion time of 4 hr at a constant infusion rate (Online Supplementary Table S1). No abnormal left ventricular ejection fraction changes were detected on the echocardiograph throughout the treatment course. Additionally, no elevations in cardiac enzyme levels, including troponin T, were observed. These findings suggest that cumulative doses of HHT up to 30 mg/m² during remission induction are safe for pediatric patients. However, ongoing monitoring and long-term follow-up are essential to assess the potential delayed cardiac effects.

Figure 1.Venetoclax concentration in plasma during Induction therapies. (A) Venetoclax concentration in plasma during Induction I. (B) Venetoclax concentration in plasma during Induction II.

Venetoclax, a BCL-2 inhibitor, is widely used in adults with de novo or secondary AML, typically in combination with azacitidine or low-dose cytarabine, and has shown favorable tolerability and efficacy.^12,13 Similarly, notable responses have been observed in pediatric relapsed AML when venetoclax was combined with intensive chemotherapy.¹⁴ A retrospective multicenter study evaluating salvage therapy with venetoclax combined with conventional chemotherapy in 31 previously treated children with AML or myelodysplastic syndromes reported an overall response rate of approximately 70% and a complete remission rate of 51%.¹⁵ However, its use as a front-line treatment for pediatric AML remains limited.

In our protocol, venetoclax was well-tolerated at the administered doses. Although there was marked inter-patient variability in plasma concentrations, no correlation was observed between venetoclax plasma levels and toxicity. Furthermore, neutrophil and platelet recovery times showed no association with venetoclax plasma concentrations. After two induction courses, all patients, regardless of high or low venetoclax ratios, achieved negative MRD.

In conclusion, the combination of venetoclax, homoharringtonine, low-dose cytarabine, and granulocyte colony-stimulating factor represents a safe and promising minimally myelosuppressive regimen. This anthracycline-free approach for remission induction is currently under investigation in a multicenter study of children with de novo AML.

Footnotes

• Received October 31, 2024
• Accepted February 28, 2025

Correspondence

*SC, LG, JLu, YH, YW and HH contributed equally as first authors

S. HU - hushaoyan@suda.edu.cn

P. XIAO - xpfdr@163.com

R.C. RIBEIRO - Raul.Ribeiro@stjude.org

Disclosures

No conflicts of interest to disclose.

Contributions

Funding

This study was supported by the following grants: The National Key Research and Development Program of China (n. 2022YFC2502700), the National Natural Science Foundation of China (NSFC, 82170218, 82470221 to SH, NSFC 82200177 to LG, NSFC 82470127 to YH, NSFC 82300244 to BL, NSFC 82400264 to YZ, NSFC 82300182 to SW), Suzhou Projects (DZXYJ202305, GSWS2023048, 2020ZKPB02 to SH), and the Suzhou Municipal Key Laboratory (SZS201615, SKY2022012, SZS2023014 to SH). Soochow University of Medical School, ML13101223 to SH. Children’s Hospital of Soochow University, 2023QN07 to SC. RCR was partially funded by grants from the National Institutes of Health, the National Cancer Institute (CA21765), the National Institute of General Medical Sciences (P50GM115279), American Lebanese Syrian Associated Charities (ALSAC), and the St. Jude Departments of Oncology and Global Pediatric Medicine.