Abstract B023: From Mutational Processes to Precision Therapy: AI-Guided Drug Repositioning Using COSMIC Signatures in Pediatric Blast-Crisis Chronic Myeloid Leukemia

Pediatric blast-crisis chronic myeloid leukemia (PBC-CML) is a rare, aggressive malignancy with limited therapeutic options. We evaluated an AI-guided framework integrating whole-exome sequencing, COSMIC mutational signatures, and pharmacogenomic drug prioritization to identify patient-specific therapeutic vulnerabilities. Distinct mutational profiles revealed heterogeneous drivers of disease progression. AI-guided analyses identified therapeutic opportunities linked to DNA repair deficiency, replication stress, epigenetic dysregulation, and inflammatory signaling. These findings support precision oncology and drug repurposing in pediatric BC-CML. Introduction: Blast-crisis transformation remains the major cause of treatment failure in pediatric CML [1,2]. Current therapeutic approaches are largely derived from adult studies and may not adequately address biological heterogeneity. COSMIC mutational signatures capture mutagenic processes driving cancer evolution and may reveal actionable vulnerabilities beyond individual mutations [3]. We investigated whether integrating mutational signatures with AI-guided pharmacogenomics could improve therapeutic prioritization in pediatric BC-CML [4]. Patients and Methods: Whole-exome sequencing data from pediatric BC-CML cases were analyzed to identify somatic variants and COSMIC mutational signatures [3,5]. AI-guided drug prioritization integrated genomic alterations, mutational signatures, and pharmacogenomic evidence to generate individualized therapeutic rankings [4,6]. Results: Each patient demonstrated a unique COSMIC signature profile, reflecting substantial biological heterogeneity. Signature-defined pathways included DNA repair deficiency, replication stress, epigenetic dysregulation, and inflammatory signaling [3,5]. AI-guided integration prioritized patient-specific therapeutic candidates, including PARP inhibitors, ATR inhibitors, IDH-targeted therapies, CDK inhibitors, and JAK/STAT-directed agents. No two patients shared identical therapeutic rankings, supporting a true N-of-1 precision oncology framework. Actionable therapeutic opportunities were identified even in cases lacking directly targetable driver mutations [4,6]. Discussion: This study presents a precision oncology framework integrating COSMIC mutational signatures, AI, and pharmacogenomic drug prioritization in pediatric BC-CML. By focusing on mutational processes rather than individual mutations, the approach enables identification of patient-specific therapeutic vulnerabilities. The findings support individualized treatment strategies and suggest applicability to other rare and treatment-refractory malignancies [6,7]. Conflict of Interest: Authors declare no conflict of interest. References:[1] Chen J et al. Blood. 2026;147(4):379-390. [2] Apperley JF et al. Leukemia. 2025. [3] Moiz A et al. Clin Case Rep. 2025. [4] Carrasco R et al. NPJ Precis Oncol. 2026. [5] Yoo W et al. Precision Oncology in the Age of AI. 2026. [6] Wang X et al. J Cancer Res Clin Oncol. 2026. [7] Saha S et al. Pediatr Blood Cancer. 2026.

Citation Format:

Zafar Iqbal, Abdulkareem AlGarni, Nasser AlQahtani, Nawaf Alanazi, Tahani Al-Qurashi, Rizwan Naeem, Masood Shammas, Sohail Rao, Muhammad F. Sabar. From Mutational Processes to Precision Therapy: AI-Guided Drug Repositioning Using COSMIC Signatures in Pediatric Blast-Crisis Chronic Myeloid Leukemia [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Bridging Discovery and Clinical Impact in Pediatric Cancer; 2026 Sep 22-25; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2026;86(18_Suppl_1):Abstract nr B023.