Fludarabine and cytarabine is not superior to high dose cytarabine as consolidation therapy in standard risk AML in children: Results from the international Phase III MyeChild01 trial

Abstract Introduction MRC AML 15 reported that high dose cytarabine (HD AraC) was not inferior to an anthracycline containing regimen as consolidation therapy in patients with non-high risk AML (Burnett, JCO 2013). Fludarabine and cytarabine (FLA) is a commonly used regimen particularly in relapsed AML. MyeChild01 aimed to test if FLA was superior to HD AraC as consolidation therapy in standard risk (SR) patients; both regimens anthracycline sparing and reducing the risk of cardiotoxicity. This paediatric trial for AML, high risk myelodysplastic syndrome (MDS > 10% blasts) or isolated myeloid sarcoma (IMS) delivered risk stratified treatment based on genetic subtype, remission status post course 1 and measurable residual disease (MRD). 360 patients were randomised to FLA or HD AraC as course 3 and 4 in consolidation from Jan-2017 to Oct-2022; 180 to each arm. Methods At diagnosis 97% patients were allocated mitoxantrone and cytarabine (MA) with 79% receiving 1 or 3 doses of GO in course 1 (Gibson, Blood; 144, Suppl 1, 2024). Subsequent treatment was stratified by genetic subtype, remission status post course 1 and MRD. SR patients had either good risk (GR) genetics and MRD negativity post course 2, or intermediate risk (IR) genetics and MRD negativity post course 1 and 2. SR patients received a second course of MA followed by two courses of either HD Ara C (3g/m2bd on days 1, 3, 5) or FLA (Fludarabine 30 mg/m2 and cytarabine 2g/m2on days 1-5) based on randomisation. Randomisation was stratified by age; diagnosis (AML, MDS, and IMS); disease type (de novo, secondary); number of gemtuzumab ozogamicin (GO) doses. Statistical analyses were Bayesian with primary outcome of relapse-free survival (RFS) defined as time from randomisation to relapse or death from any cause. Results Patient characteristics were: males 59%; median age 8yr; AML 96%, MDS 2%, IMS 2%; de novo 99.7%; genetic GR 54%, IR 44%. Induction therapy was MA 97%; with: no GO 14%; 1 dose GO 38%; 2 doses GO 7%; 3 doses GO 41%. Of the 360 patients 178 received HD AraC and 180 FLA; baseline characteristics were well balanced. Efficacy analysis included all patients;7 patients were excluded from safety summaries due to ineligibility/not receiving treatment. Median follow-up is 3.6 years. The 2yr RFS for HD AraC was 78% (72%, 85%) and for FLA 70% (64%, 77%) with a 2yr overall survival (OS) for HD AraC of 97% (94%, 99%) and for FLA 90% (86%, 95%). The Bayesian probability that FLA was superior to HD AraC was 7% for RFS and 8% for OS. No substantial differences in RFS were seen within subgroups of age, diagnosis, disease type, GO doses and genetic risk group. The 2yr cumulative incidence of relapse (CIR) for HD AraC was 22% (16%, 28%) and for FLA 29% (22%, 35%) indicating no difference (p=0.17). Only 2 deaths in remission were reported, both patients received FLA. Analysed by genetic risk group 97 GR and 78 IR patients received HD AraC while 96 GR and 80 IR received FLA; 9 patients without a risk group were excluded. In GR patients, 2yr RFS was 84% (78%, 92%) and 79% (71%, 88%), 2yr OS 98% (95%, 100%) and 98% (95%, 100%) and CIR 16% (9%, 23%) and 21% (14%, 30%) for HD AraC and FLA respectively. IR patients had a 2yr RFS of 69% (59%, 80%) and 62% (52%, 74%), 2yr OS of 95% (90%, 100%) and 81% (72%, 90%) and CIR of 31% (21%, 42%) and 35% (25%, 46%) for HD AraC and FLA respectively. IR patients consistently did worse than GR patients irrespective of treatment but had substantially worse OS rates with FLA. 67 patients with a KMT2A rearrangement were considered IR. At 2yr those who received HD AraC had RFS of 73% (60%, 89%), OS of 92% (84%, 100%) and CIR of 27% (14%, 42%) while those receiving FLA had RFS of 50% (35%, 72%), OS of 73% (58%, 91%) and CIR of 47% (28%, 63%). Incidence of grade ≥3 adverse reactions or any grade serious adverse events (SAE) for HD AraC was 56% and for FLA 53%. 54%of patients receiving HD AraC and 44% receiving FLA had at least one grade ≥3 SAE. There were two grade 5 SAEs both in the FLA arm; heart failure unrelated to FLA and CNS infection possibly related to FLA. The median time from the start of course 3 to course 4 was 38 days for HD AraC and 41 for FLA. Conclusion FLA is not superior to HD AraC as consolidation therapy in children with SR AML. Patients with IR genetics, particularly those with a non-high risk KMT2A rearrangement had a worse outcome with FLA. High Dose AraC should remain standard of care in consolidation for patients with SR AML.