A Unified Prognostic Data Architecture for Risk Stratification in Pediatric Acute Lymphoblastic Leukemia

KÖSE U., CEYLAN O., SÜRÜCÜ E. B.

4th Cognitive Models and Artificial Intelligence Conference, AICCONF 2026, Prague, Çek Cumhuriyeti, 24 - 25 Nisan 2026, (Tam Metin Bildiri)

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Doi Numarası: 10.1109/aicconf69182.2026.11600718
  • Basıldığı Şehir: Prague
  • Basıldığı Ülke: Çek Cumhuriyeti
  • Anahtar Kelimeler: acute lymphoblastic leukemia (ALL), berlin-frankfurt-münster (BFM), childhood leukemia, children's oncology group (COG), data architecture, national cancer institute (NCI), prognostic modeling, risk classification
  • Süleyman Demirel Üniversitesi Adresli: Evet

Özet

Risk stratification for childhood acute lymphoblastic leukemia (childhood-ALL) forms the basis of modern risk-based treatment regimens. Large international consortia, such as the Children's Oncology Group (COG), the Berlin-Frankfurt-Münster (BFM) group, and the National Cancer Institute (NCI) working groups, have defined current standards for risk-appropriate treatment. These groups determine treatment intensity using various variables such as age, baseline white blood cell count (WBC), genetic characteristics, and minimal residual disease (MRD) levels. However, there are differences in threshold values and categorical definitions among the criteria established by these groups. This heterogeneity creates difficulties in data processing, multicenter analyses, and data-driven modeling studies. This study proposes a unified prognostic data architecture integrating parameters at the time of diagnosis and early treatment response indicators. The NCI, COG, and BFM risk systems are matched under a common data representation and harmonized with different risk categories. In addition, the approximate 5-year event-independent survival (EFS) and overall survival (OS) ranges reported in the literature have also been correlated with risk levels. The proposed structure provides a scalable basis for data standardization and decision support systems while maintaining clinical validity.