BCR/ABL1-independent markers of resistance in patients affected by chronic myeloid leukemia (CML) receiving tyrosine kinase inhibitors (TKIs).

Chronic Myeloid Leukemia (CML) is a chronic myeloproliferative neoplasm derived from the neoplastic transformation of the pluripotent stem cell characterized by the reciprocal translocation between chromosome 9 (9q) and 22 (22q) t (9; 22) (q34; q11), which leads to the formation of a new BCR-ABL1 gene that causes expansion of the pathological clone. The introduction of tyrosine kinase inhibitor drugs (TKIs) has revolutionized the treatment of this neoplasm. However, about a third of patients have to suspend treatment for resistance or intolerance to TKIs. The aim of our study was to identify the factors that affect the efficacy and toxicity of TKIs, evaluating biological mechanisms that cause BCR/ABL1-independent resistance in the leukemic cell, in order to find new markers of prognosis and for a target therapy. The study was divided in several part: 1) to evaluate the influence of polymorphisms (SNPs) of membrane transporters hOCT1, ABCB1, ABCG2, SLCO1B3 on the pharmacokinetics of Imatinib and Nilotinib; 2) to evaluate if aspects of epigenetics, such as the expression of the Polycomb group genes (PcGs), can correlate with the response to Imatinib; 3) evaluation of BMI1 as marker of LSCs; 4) the study of the gene expression profile of pathways of PcGs, JAK/STAT and Wnt/b-catenin, involved in the resistance and progression of the disease; 5) identification of CML-related proteins in a different compartment from peripheral blood or bone marrow (salivary proteomic). From the data collected on patients enrolled in the TIKlet protocol, we have shown that the polymorphism of hOCT1c.480C>G significantly influences the clearance of Imatinib. About Nilotinib, thanks to a pharmacokinetic model that we have developed, it was possible to highlight how the elimination rate of nilotinib was influenced by the polymorphism ABCB1 c.2677G>T/A. Furthermore, SLCO1B3 c.334T>G is significantly associated with the bioavailability of Nilotinib. At the same time, patients with the SCLO1B3 c.521C allele will take longer to achieve the complete cytogenetic response or MR3. In addition, we demonstrated in our patients the impact of the PcGs, epigenetic repressors involved in the progression of CML and resistance to therapy. We measured the expression of 8 PcGs before and after three months of treatment with Imatinib from bone marrow samples of 30 patients. The expression of BMI1, PHC3, CBX6 and CBX7 was significantly increased during treatment with TKI; in particular, we shown that the increase in BMI1 had a negative impact on the prognosis, even in the subgroup of patients with unfavorable hOCT1. BMI1 is responsible for the resistance to the TKIs in a BCR-ABL1-independent way, even if is unknown where BMI1 in CML is expressed (in progenitors or more mature cells). We decided, therefore, to evaluate if and where the BMI1 protein is located, focusing on the CD34+/CD38-/CD26+ CML progenitors. We measured, by flow cytometry, the proportion of CD34+/CD26+ cells in bone marrow of 20 CML patients, at diagnosis and during treatment with imatinib. After that the bone marrow blood smears were stained with antibodies anti-CD26, BCR-ABL1, and BMI1. These smears were observed by a confocal laser microscope and we showed that the BMI1 protein was co-expressed in the cytoplasm of the CD26+ precursors. In parallel, we performed the expression analysis of 255 genes, related to the WNT, JAK/STAT and PcGs pathways. We measured genes de- regulation after 6 months of therapy in 11 CML patients. In order to find possible correlations with quality of response to treatment and event-free-survival (EFS). These results were then re-analyzed by the principal component method (PCA) for tempting to better cluster resistant cases. WNT was the most responsible pathway for the TKIs resistance. Indeed, 100% of patients with a “low” up-regulation of this pathway achieved an optimal response versus 33% of those who showed a “high” gene over-expression (p=0.016). Analogously, the 24-months EFS resulted significantly influenced by the degree of up-regulation of the WNT signaling: all patients with a “low” up-regulation were event- free versus 33% of those who presented a “high” gene expression (p=0.05). In particular, the PCA analysis confirmed the role of WNT pathway and showed that the most significantly up-regulated genes with negative prognostic value were DKK, WNT6, WISP1 and FZD8. Moreover, we focused on the study of salivary proteomic, and although our study was preliminary, the analysis of proteomic expression identified proteins related to the bone marrow niche, differentiation and stemness. In particular, there was a significantly different proteomic profile of resistant in comparison to the sensitive patients. This data could allow the assessment of the salivary compartment as an alternative to those already used for the monitoring of patients with CML. In conclusion, our results sustain the need of a wide and multitasking approach in order to understand the resistance mechanisms in CML. This study allowed us to demonstrate how pharmacokinetic and pharmacogenomic aspects can be important in the management of the patient with CML, for the choice of the type of drug to be used in the first line to reduce toxicity, that negative impacts on the quality of life, to promote adherence to treatment and prevent the loss of response. TKIs act on intracellular pathways other than BCR-ABL1 and, even if in a small cohort of CML patients, we found a significant de- regulation of WNT pathway and we demonstrated that its activation could affect the sensitivity to TKIs and the clinical outcome. Once translated into the clinical context, our results would suggest the combination of the WNT/β-catenin with TKIs. Then, if BMI1 is a marker of stemness, also combination with TKIs and BMI1 inhibitors could be a promising way to overcome resistance in CML patients.