ABCG2 but Not ABCB1 RNA Levels at Diagnosis and during Treatment Correlates with Response to Imatinib and 2G-Tkis Front-Line Therapy in Chronic Phase CML Patients

Introduction. Overexpression of multidrug resistance proteins ABCB1 and ABCG2 confers resistance to anticancer drugs, including tyrosine kinase inhibitors (TKIs). However, little is known about the significance of ABCB1 and ABCG2 expression during TKI treatment for chronic myeloid leukemia (CML) and possible correlation with response to therapy.

Aims. To investigate the role of ABCB1 and ABCG2 expression during the first 12 months of treatment with imatinib and 2G-TKIs for chronic phase CML patients and correlation with molecular response.

Methods. We analysed expression of ABCB1 and ABCG2 RNA at diagnosis and after 3, 6 and 12 months of TKI therapy in 20 patients with CML, divided in two groups according to response to treatment. Group A (n=10) comprised patients treated with imatinib that displayed optimal response at 3 months (defined as BCR/ABL <10%), 6 months (BCR/ABL < 1%) and at 12 months (BCR/ABL < 0.1%); group B (n=10) included patients treated with imatinib (n=5) or 2G-TKIs (nilotinib in 3 and dasatinib in 2 cases, respectively) who did not meet the criteria for optimal molecular response at all three timepoint or that switched to alternative treatment after at least the 3^rd month. RNA levels were expressed using the 2^-ΔΔCt method proposed by Livak and Schmittgen, using ABL as normalizing gene and the expression level of the K562 cell line as calibrator, in order to standardize the results.

Results. ABCB1 expression at diagnosis was similar in group A (mean±SD = 0.21±0.22) and in group B (0.15±0.20; p=0.48); similarly, expression during treatment was similar in responding and non-responding patients at 3 (0.82±0.32 vs 0.83±0.41; p=0.66) 6 (0.85±0.22 vs 0.90±0.50; p=0.65) and 12 (1.05±0.24 vs 0.77±0.42; p=0.21) months. Conversely, ABCG2 expression at diagnosis was lower in group A (0.85±0.71) compared to group B (2.83±2.84; p=0.06). RNA levels increased in both groups during treatment, but were always lower in patients with optimal response compared to cases without response or treatment failure: at 3 months (5.12±3.31 vs 49.44±82.70; p=0.13), at 6 months (8.23±8.55 vs 12.71±10.04; p=0.32) and at 12 months (6.05±4.87 vs 13.00±9.93; p=0.05).

Considering response to therapy, all patients in group A are still in imatinib with sustained optimal response; all patients in group B receiving imatinib switched to alternative therapy (nilotinib = 2, dasatinib = 1, allogeneic SCT for evolution to blast phase = 1) or died for non-CML causes while not in complete cytogenetic response (n=1), while 3/5 patients receiving 2G-TKI changed therapy (nilotinib to dasatinib = 1, dasatinib to ponatinib = 1, nilotinib to allogeneic SCT for aplasia = 1) and 2/5 patients continued the same drug achieving MMR beyond the 12^th month.

Discussion. With the limits of a small number of patients, our data suggest that high ABCG2 levels at diagnosis seems to be associated with non-optimal response to imatinib and 2G-TKIs, and an increase in ABCG2 during the first months of treatment is a hallmark of poor response.