Introduction: The 2022 European Leukemia Network (ELN) recommendations defined several efficacy endpoints for acute myeloid leukemia (AML) treatment, including complete remission (CR), CR with partial hematologic recovery (CRh), CR with incomplete hematologic recovery (CRi), and morphological leukemia-free state (MLFS). In pediatric AML, subsequent chemotherapy courses are often initiated before full hematologic recovery as part of treatment intensification. However, regulatory agencies frequently regard CRh, CRi, and MLFS as treatment failure due to survival differences observed in adult AML, which poses challenges for pediatric drug development. This study evaluated whether survival outcomes differ according to the degree of hematologic recovery in pediatric AML.

Methods: We analyzed pediatric patients (< 18 years) with newly diagnosed AML who were enrolled in two nationwide multicenter clinical trials conducted in Japan: JPLSG AML-05 (UMIN000000511) and AML-12 (UMIN000013288/jRCTs041180128). Both trials employed two courses of induction therapy. The initial induction therapy was standard-dose cytarabine (ECM) in AML-05, and either ECM or high-dose cytarabine (HD-ECM) in AML-12 based on randomization. In both trials, these were followed by a second induction therapy (HCEI). In AML-12, flow cytometry-based measurable residual disease (MRD) was centrally monitored at the end of inductions 1 and 2, although MRD results did not guide subsequent therapies. MRD ≥ 0.1% was considered positive. Intensification therapies were initiated following hematologic recovery, defined as an absolute neutrophil count ≥ 0.5 × 109/L and a platelet count ≥ 75–80 × 109/L. A total of 757 patients (422 from AML-05 and 335 from AML-12) who underwent response assessment after induction 2 were included in this analysis. Treatment response after induction 2 was re-evaluated according to the ELN2022 criteria. Survival comparisons between response groups were performed using the log-rank test, Gray's test, and multivariable Cox regression analysis.

Results: After induction 2, 313 patients achieved CR, 352 achieved CRh, 14 achieved CRi, 4 were in MLFS, and 74 did not achieve remission (NR). Compared with the 5-year event-free survival (EFS) probability from registration in patients with CR (63.4%; 95% confidence interval [CI], 57.4%–68.7%), the EFS in those with CRh (65.8%; 95% CI, 60.3%–70.7%; P = 0.434) and CRi (57.1%; 95% CI, 28.4%–78.0%; P = 0.484) did not differ significantly. In contrast, patients with MLFS (5 years not attained; P = 0.029) and NR (0.0%; P < 0.001) had significantly worse EFS. Similarly, the 5-year overall survival (OS) probability from registration in patients with CR was 79.1% (95% CI, 73.7–83.5%). There was no significant difference in OS for patients with CRh (79.7%; 95% CI, 74.4%–84.1%; P = 0.619) or CRi (67.7%; 95% CI, 34.4%–86.7%; P = 0.275), whereas OS was significantly lower in the MLFS group (0.0%; P = 0.002) and NR group (29.2%; 95% CI, 17.9%–41.5%; P < 0.001). The cumulative incidence of relapse (CIR) did not differ significantly between those with CR and CRh, CRi, or MLFS. The cumulative incidence of non-relapse mortality was also not significantly different between the CR group and the CRh or CRi groups, but was significantly higher in the MLFS group compared with the CR group (P < 0.001). In the AML-12 cohort, among both the CR and incomplete CR (CRh, CRi, and MLFS) groups, patients who were MRD-positive at the end of induction 2 had significantly inferior 5-year disease-free survival (DFS) and OS compared to MRD-negative patients (P < 0.001, multivariable Cox regression). Furthermore, among MRD-negative patients, no statistically significant differences were observed between the CR and incomplete CR groups in terms of 5-year DFS (65.6% vs. 76.9%, P = 0.080) or OS (84.2% vs. 85.7%, P = 0.778).

Conclusions: In pediatric AML, EFS and OS in patients with CRh and CRi were comparable to those in CR, supporting the validity as effective treatment responses. In contrast, MLFS was associated with inferior EFS and OS, a comparable CIR, and a higher incidence of non-relapse mortality compared to CR. This suggests that inadequate hematologic recovery in MLFS may contribute to complications such as infections; however, caution is warranted due to the limited number of cases analyzed. Moreover, MRD should be further explored as a more precise surrogate endpoint for assessing treatment response in pediatric AML.

This content is only available as a PDF.
2025
Sign in via your Institution