A Phase 3 Randomized Trial for Patients with De Novo AML Comparing Standard Therapy Including Gemtuzumab Ozogamicin (GO) to CPX-351 with GO – a Report from the Children’s Oncology Group

Introduction:

Survival in childhood AML has plateaued and new therapeutic strategies are essential. CPX-351 (Vyxeos) is a liposomal preparation of cytarabine and daunorubicin with superior pharmacokinetic properties when compared to free drug. The Children’s Oncology Group (COG) Phase III clinical trial AAML1831 tested the hypothesis that CPX-351 given during Inductions 1 and 2 would improve treatment outcomes and reduce long-term toxicities compared to a standard chemotherapy.

Methods:

COG AAML1831 randomized patients (pts) <22 years of age with de novo AML to 2 different induction regimens. Pts with FLT3 mutations were offered enrollment onto gilteritinib-containing regimens and are not included in this analysis. Arm A patients received cytarabine, daunorubicin and gemtuzumab ozogamicin (DA+GO) during Induction 1 and DA during Induction 2. Patients on Arm B received, during Induction 1, CPX-351 on days 1, 3 and 5 and GO on day 6; the CPX-351 dosing was repeated in Induction 2. Centralized risk stratification occurred after Induction 1 and was based on cytomolecular findings and end of induction 1 (EOI 1) minimal residual disease (MRD) >0.05% by central multidimensional flow cytometry. Pts were designated as low risk 1 (LR1; favorable genetics and MRD negative), low risk 2(LR2; pts not in LR1 or HR), or high risk (HR; MRD+ without favorable genetics or any high risk genetics) and received 4 (LR1) versus 5 (LR2) cycles of chemotherapy or 2-3 cycles followed by hematopoietic stem cell transplant (HSCT; HR).

Results:

A total of 721 eligible pts without FLT3 mutations were randomized to Arm A (n=358) or Arm B (n=363). Age, race, ethnicity and sex were similar for the 2 arms; distribution of risk-defining cytomolecular features were also comparable, with exception of higher rates of inv(16)(p13.1q22.1) (Arm A: 10% vs Arm B: 6%; p=0.03) but not t(16;16)(p13.1;q22) CBFB-MYH11 AML. Pts on Arm A were also more likely to be CNS negative at time of initial assessment (Arm A: 86%; Arm B: 79%; p=0.026). EOI 1 MRD was comparable (Arm A: 23%; Arm B: 23%) and elective withdrawal rates at EOI1 did not differ significantly between the 2 arms (Arm A: 9%; Arm B: 11%). However, Induction 2 treatment toxicities [grade 3 infection (Arm A: 7.4%; Arm B: 27.2%), grade 3 sepsis (Arm A: 3.5%; Arm B: 5.1%)] and median cycle length (Arm A: 35 days; Arm B: 44 days) were higher in Arm B though this did not translate into higher Induction 2 treatment related mortality.

Protocol-specified interim analyses using data available on 12/31/23 were performed to monitor efficacy and futility of event-free survival (EFS) for Arm A and Arm B patients without FLT3 mutations. Closure of this randomization was recommended given the futility monitoring rule was crossed. Using a 6/30/24 data cutoff, 2-year EFS from study entry for Arm A was 60.9% (95% CI: 54.3-66.8%) vs. 51.2% (95% CI: 44.7-57.4%) for Arm B (p=0.019) though overall survival (OS) was not significantly different [Arm A: 75.5% (95% CI-69.3-80.5%); Arm B: 74.5% (95% CI – 68.4-79.6%]; p=0.782]. Outcomes for HR patients were comparable for both arms whereas EFS was significantly lower and relapse rates (RR) higher for LR patients on Arm B [2-yr EFS from EOI1: Arm A: 71.9% vs. Arm B 57.8% (p=0.006); 2-year RR Arm A: 16.1% vs. Arm B: 25.4%, p=0.018). EFS from EOI1 in the LR1 patients was particularly impacted (Arm A: 78.3%; Arm B: 62.6%, p=0.029) while LR2 EFS from EOI is Arm A: 67.2 % vs. Arm B: 54.6% (p=0.079); most events reported for either group were relapse.

Conclusion

The primary randomization of the Phase III study COG AAML1831, a comparison between standard induction therapy and CPX-351, was terminated early because the futility boundary was crossed. The difference in EFS observed was driven mainly by events in LR pts, particularly those that received 4 cycles of chemotherapy (LR1). Further analysis is underway to determine additional factors that may have contributed to the inferior outcomes observed for pts enrolled on Arm B.