This study aims to analyze the clinical and genetic characteristics of pediatric AML patients, evaluate treatment responses and survival under AML-BFM 2004 and 2012 protocols, identify factors affecting survival, and compare findings with international data to provide context for results.

In this study, 49 pediatric patients under 18 years of age diagnosed with AML at a tertiary university hospital in Turkey between January 2010 and December 2023 were retrospectively reviewed. Demographic data, clinical findings at initial diagnosis, use of leukapheresis, administration of cytoreductive therapy prior to induction, induction response, relapse status, outcomes of stem cell transplantation, and current status (alive/deceased) of these patients were evaluated. Age at diagnosis was stratified into three groups: <2 years, 2–13 years, and ≥14 years. Hemoglobin level (g/dL), leukocyte count (x109/L), platelet count (x109/L), and blast percentages in bone marrow aspirate and peripheral smear (%) at diagnosis were recorded. Leukocyte count was categorized as <50 × 109/L or ≥50 × 109/L.Patients were treated according to the AML-Berlin–Frankfurt–Munster 2004 and AML-BFM 2012 protocols. In the AML-BFM 2004 protocol, patients are stratified into standard-risk (SR) and high-risk (HR) groups based on their treatment response after induction and genetic features at the time of initial diagnosis. The AML-BFM 2012 protocol also stratifies risk as SR, intermediate-risk (IR), and HR.In addition, patients were compared in two groups: AML M3 (acute promyelocytic leukemia) and non-AML M3 (AML M1, M2, M4-7, biphenotypic, undifferentiated type).The AML-BFM 2004 protocol was used between 2010 and 2015, and the AML-BFM 2012 protocol was used between 2016 and 2023. Treatment efficacy was assessed at the morphological level using bone marrow aspiration at the end of chemotherapy blocks. Complete remission was defined as normocellular marrow with ≤5% blasts, peripheral neutrophil count ≥1 × 109/L, platelet count ≥80 × 109/L, absence of extramedullary disease, and no blasts in the central nervous system.

Data were expressed as mean ± SD or median (IQR) based on normality. Intergroup comparisons were performed using Chi-square, Fisher’s exact, and t-tests. Non-normally distributed data were analyzed using the Mann–Whitney U test. Survival rates were calculated using the Kaplan–Meier method. Receiver operating characteristic (ROC) curve analysis evaluated prognostic prediction performance. Statistical analyses were conducted using IBM SPSS Statistics Data Editor. A p-value <0.05 was considered statistically significant.

The median age was 12 years, with the most frequent age group being 2–13 years. Twenty-five patients (51%) were female. The most frequent morphological subtype was AML M3 in 16 cases (32.7%). Risk stratification classified 28 patients (57.2%) as SR, 6 (12.2%) as IR, and 15 (30.6%) as HR. The most frequent genetic mutation observed was t(15;17) in 16 patients (32.6%), followed by t(8;21) in 9 patients (18.3%). Seventeen patients (34.7%) were treated under the AML-BFM 2004 protocol, and 32 under the AML-BFM 2012 protocol. Leukapheresis was performed in 8 (16.3%), with all leukapheresis patients achieving a response. Complete remission (bone marrow blasts <5%) after induction was achieved in 40 patients (81.6%). Six patients underwent bone marrow transplantation (BMT), of whom 3 (50%) died due to post-transplant relapse. Thirteen patients (26.5%) experienced relapse: One in the AML M3 group and 12 (24.5%) in the non-AML M3 group. Twelve relapsed patients (92.3%) died. Among BMT recipients, one patient underwent transplantation for HR disease without prior relapse; however, later died due to post-transplant relapse. Overall, 19 patients (38.8%) died. Survival rates were significantly lower in patients with leukocyte counts ≥50 × 109/L (p=0.002). ROC curve analysis revealed an area under the curve (AUC) of 0.744 for leukocyte count (p < 0.005), with a cut-off value >42,850 × 109 /L. Lower hemoglobin levels at diagnosis were associated with reduced survival (p=0.030). The mean overall survival (OS) for AML-M3 patients was 113 months ±%6.8 months (95% CI: 99.3-126.3), whereas the median OS for non-AML M3 patients was 28±%17.2 months (95% CI: 0.00-61.8) (p=0.002). Survival duration was shorter in IR and HR groups than the SR group.Cumulative survival rates at 1 and 5 years were significantly longer in AML-M3 patients than non-AML M3 patients. For AML-M3, cumulative survival was 94%±%6 at both time points. In non-AML M3 patients, 1-year and 5-year cumulative survival rates were 73% and 43%, respectively (p = 0.002). In relapsed patients, median OS after relapse was 3±0.8 months (95% CI:1.23-4.76) (p=0.000). No significant difference in survival rates was observed between the AML-BFM 2004 and AML-BFM 2012 protocols.

In the present study, key factors influencing survival in pediatric AML included risk group, age at diagnosis, induction response at diagnosis, and time-to-relapse. Among relapsed patients, the initial risk group also affected survival. Leukapheresis had no impact on survival. Mortality remains high in non-AML M3 cases. Further research is required to develop genetically defined treatment subgroups in pediatric AML. We recommend more stringent risk stratification for IR patients under the AML-BFM 2012 protocol, and advocate for larger studies aimed at creating standardized, personalized treatment protocols for all patients.