Living without treatment has become a realistic target for patients with chronic myeloid leukemia (CML) who achieved durable deep molecular remission under treatment with tyrosine kinase inhibitors (TKI). Recently, we have identified novel BCR-ABL-independent gene mutations in newly diagnosed CML patients that may function as important cofactors in the evolution of CML and persistence of the disease (Schmidt M, Rinke J, et al. Leukemia 2014; 28(12): 2292-9). Here, we sought to investigate the dynamics of such mutations during a 2-year treatment with the second generation TKI nilotinib. To check if BCR-ABL independent mutations are an age associated phenomenon we also screened a cohort of children and adolescents with CML. A total of 51 chronic phase CML patients were analyzed including 31 adults (male, n=23; median age 57 years, range 42 to 73 years) and 20 children/adolescents (male, n= 14; median age 14 years, range 10 months to 27 years). For each adult patient, samples were analyzed at diagnosis, in subsequent complete cytogenetic remission (CCyR) after 3 months and in deep molecular remission (≤ MMR) after 24 months of nilotinib therapy. Children/adolescents samples were investigated at diagnosis only. Targeted deep next-generation sequencing (NGS) was used to analyze a panel of 30 commonly mutated genes in myeloid disorders. Constitutional DNA obtained from buccal swabs at diagnosis was investigated to check whether mutations are of somatic origin. BCR-ABL independent gene mutations were found in 8/31 (26%) adult CML patients at diagnosis affecting the genes ASXL1 (n=3), DNMT3A (n=3), RUNX1 (n=1) and RUNX1 plus TET2 (n=1). No mutation was recognized in corresponding constitutional DNA specimens indicating that all mutations were somatically acquired. Analysis of individual hematopoietic colonies from three adult patients revealed that most mutations were part of the BCR-ABL-positive clone. NGS of subsequent samples obtained after three months of TKI therapy identified two DNMT3A mutations in BCR-ABL-negative cells that were also present in BCR-ABL-positive cells at diagnosis, implying that this mutation preceded the BCR-ABL rearrangement. After 24 months of TKI treatment, NGS revealed these two DNMT3A mutations again as well as a novel low-level EZH2 mutation in a patient with deep molecular remission (MR4.5) that was neither detectable at diagnosis nor after 3 months of treatment indicating the emergence of a BCR-ABL-independent subclone in BCR-ABL-negative cells during long-term nilotinib treatment. Within the children/adolescents cohort BCR-ABL independent gene mutations were identified in 5/20 (25%) subjects all affecting the ASXL1 gene. All mutations were stop codon or frameshift mutations implying loss of ASXL1 gene function. We conclude that BCR-ABL independent gene mutations can be found frequently in CML patients and affect predominately epigenetic modifier genes. The finding of frequent mutations also in children/adolescents implicates that BCR-ABL independent gene mutations are not just age-related events. Such molecular aberrations may play an important role in the evolution and persistence of the disease and may affect therapy in both children and adults with CML. These findings may provide important novel genetic information regarding CML biology and for the design and performance of discontinuation trials.


Hochhaus:Novartis: Honoraria, Research Funding; Pfizer: Honoraria, Research Funding; BMS: Honoraria, Research Funding; ARIAD: Honoraria, Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.