Somatic mutations and outcomes in chronic myeloid leukemia adolescent and young adults compared to children, adults, and BCR::ABL1-positive acute lymphoblastic leukemia

Abstract

Adolescent and young adults (AYAs) with chronic myeloid leukemia in chronic phase (CML-CP) reportedly respond worse to tyrosine kinase inhibitors (TKIs) than adults, potentially due to additional genetic abnormalities, including mutations in cancer-related genes (CRGs). This real-life study compared mutation profiles and their impact on outcomes in 80 AYA, 97 adult, and 16 pediatric CML-CP patients, alongside 81 BCR::ABL1-positive acute lymphoblastic leukemia (Ph+ ALL) patients. CRG mutations were more frequent in AYAs (25.0%) than in adults (19.6%) or children (12.5%). AYAs with Ph+ ALL exhibited higher mutational frequencies (53.3%) compared to children (26.7%) and adults (38.9%). At diagnosis, mutations in ASXL1, DNMT3A, and TET2 dominated in CML-CP and RUNX1, IKZF1, and BCR::ABL1 in Ph+ ALL. ASXL1 mutations correlated with reduced progression-free survival (PFS) in AYAs and adults. Unlike adults, AYAs showed no increase in BCR::ABL1 kinase domain mutations during TKI therapy. Nilotinib improved PFS in AYAs with ASXL1 mutations, highlighting the efficacy of higher-generation TKIs. ASXL1 mutations also impaired erythropoiesis, warranting further validation. Despite a higher mutational burden, AYAs did not exhibit worse prognoses than adults. Lower mutation rates at follow-up suggest potential impact of nilotinib. Mutation profiling and optimized TKI use are crucial to mitigate progression risks in CRG-mutated patients.

Fig. 1. Age- and disease-related spectrum of somatic mutations at diagnosis.

The frequency of patients with mutations according to age and disease (A) CML and (B) Ph+ ALL. Patients were divided into subgroups based on age at diagnosis: children (0–17 years), AYA (18–39 years), and adults (>40 years). Spectrum of somatic mutations in (C) CML and (D) Ph+ ALL patients according to age subgroups. The total number of mutations identified is shown in tables.

Fig. 2. Spectrum of somatic mutations observed in AYA and adult TKI non-responders.

The bars represent the frequency of patients with somatic mutations at diagnosis (A), TKI follow-up (B), mutations persisting during TKI treatment (C), and those acquired de novo during TKI treatment (D). The tables below indicate the total number of mutations detected.

Fig. 3. Impact of somatic mutations identified at diagnosis on PFS according to type of mutations and age groups.

PFS of (A) AYA and (B) adult CML patients with any mutation identified at diagnosis compared to patients with no mutation. Effect of ASXL1 mutations observed at diagnosis on PFS in (C) AYA and (D) adult patients. Hazard R (95% CI) derived from Cox proportional hazard regression models and the p-value calculated by the Log-rank test are shown. The number of patients at risk is shown in tables below. Allmut—patients with any mutation at diagnosis; wt - patients with no mutation during the whole TKI follow-up.

Fig. 4. Cumulative incidence of BCR::ABL1 KD mutations in adult CML patients according to the presence of ASXL1 mutation at the time of diagnosis.

The number of patients at risk is shown on the table below. ASXL1mut ASXL1 mutation at diagnosis, wt no mutation at diagnosis.