Frequency and clinical impact of compound mutation in chronic myeloid leukemia patients resistant to tyrosine kinase inhibitor

Background: Chronic myeloid leukemia is characterized by onco-protein BCR-ABL1 which is constitutively activated. This particularity allowed the development of the first specific tyrosine kinase inhibitor (TKI) which changed the course of this disease. However, resistance to TKI is still a growing clinical problem. The canonical mechanism of resistance is point mutation in ABL1 kinase domain (KD). After the introduction of second and third TKI, the presence of multiple mutations was grown and the clinical importance is still controversial. Methods: We included 21 resistant patients with two BCR-ABL1 KD mutations detected by direct sequencing. 43% failed to two therapies, 8 to three and 4 failed only to imatinib. For two patients we had sequential samples. We cloned ABL1 region and sequenced minimum of 10 clones. In order to evaluate the frequency of false compound mutation we cloned sample with only T315I mutation, other with F317L and 1:1 dilution of both samples. Results: We observed that 66.7% of double mutations detected by direct sequencing were compound mutations. Our ratio of false positive compound mutation was 9%. T315I plus other mutation was the most frequently compound mutation in our cohort. Patients who failed to more than one TKI had higher frequency of compound mutation. Analyzing sequential samples we observed mutated clones in samples up to 5 months earlier than direct sequencing. The frequency of clones harboring compound mutations with more than two missense mutations was low (6%), comparing with silent mutations, suggesting a limited tolerance for BCR-ABL1 KD missense mutations. The number of transition was higher than transversion. Among patients with compound mutations: 5/7 progressed to advanced phases presented compound mutation, 1/2 died, 5/8 achieved molecular response (MR) during second or third line therapy and 3/4 did not achieve MR (median follow up 26 months). Conclusions: We did not find correlation between presence of compound mutation and probability to respond therapy or progression. Analyze higher number of sample is needed. Indeed, the importance to detect clones harboring high resistance mutation such T315I earlier could help the clinician to choose the therapy to avoid clone progression. Legal entity responsible for the study: Instituto Nacional do Câncer. Funding: Has not received any funding. Disclosure: All authors have declared no conflicts of interest.