841P - Transforming the landscape of pediatric AML treatment: A cutting-edge SCT prognostic model

Background

The study aims to improve the efficacy of stem cell transplantation (SCT) for pediatric acute myeloid leukemia (pAML) by developing specific model.

Methods

A total of 1941 pAML patients were enrolled in this study, of whom 308 underwent SCT. A 7:3 ratio was used to generate training and internal validation subsets. External validation included 408 patients from four clinical centers. The primary endpoints were overall survival (OS) and event-free survival (EFS). Prognostic models were developed using univariate and multivariate Cox analyses, and the performance of the models was assessed by the consistency index (C-index) and the area under the curve (AUC).

Results

The pAML SCT model based on the three clinical factors effectively categorized patients into low-risk, intermediate-risk, and high-risk groups (p = 0.024), and also had applicability in terms of EFS (p < 0.001). The pAML SCT model was confirmed to be valid in external validation at 4 clinical centers in China. Notably, the model's outperformed the mainstream clinical prognostic tool and was comparable to another cytogenetic risk-based SCT pAML prognostic model in terms of C-index and AUC. The DCA curves suggested that the pAML SCT model used for clinical decision-making with optimal accuracy and efficiency. Finally, after further improving the model by comparing it with non-SCT patients, the Final pAML SCT model distinguished patients into four groups: the No benefit, EFS prolongation, OS and EFS prolongation, and Unsuitable groups. SCT does not improve the prognosis of the No benefit group. The OS and EFS prolongation group among SCT patients had the best prognosis, while the Unsuitable group had the worst prognosis. Notably, the survival rate of patients in the no-benefit group was comparable to that of the non-transplanted population, suggesting that this group of patients may be at risk of overtreatment.

Conclusions

We have developed an effective prognostic model for pAML patients undergoing SCT. The model provides comprehensive prognostic information and higher predictive validity than existing mainstream pAML prognostic models, and has the potential to identify patients who are overtreated or unsuitable for SCT.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

National Natural Science Foundation of China (82203844).

Disclosure

All authors have declared no conflicts of interest.