ORIGINAL ARTICLE OCT1 genetic variants are associated with long term outcomes in imatinib treated chronic myeloid leukemia patients Maya Koren-Michowitz 1,2 , Zehavit Buzaglo 1 , Elena Ribakovsky 1 , Michaela Schwarz 3 , Ilias Pessach 1 , Avichai Shimoni 1,2 , Katia Beider 1 , Ninette Amariglio 1 , Philipp le Coutre 3 , Arnon Nagler 1,2 1 Division of Hematology, Chaim Sheba Medical Center, Tel Hashomer; 2 Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel; 3 Campus Virchow Klinikum Charit e, Berlin, Germany Abstract Objectives: One third of CML patients treated with first line imatinib have suboptimal responses or treatment failures with increased risk for disease progression. Imatinib is actively transported into cells by the SLC22A1 transporter (hOCT1) and its genetic variants may affect intracellular drug import. We studied the effect of SLC22A1 genetic variants on long-term outcomes of imatinib treated patients. Methods: A total of 167 patients, 94% in chronic phase, were analyzed for rs41267797, rs683369, rs12208357, and rs628031 variants using the Sequenom MassARRAY platform. Results: Rates of CHR, MCyR, CCyR, and MMolR were not significantly different according to allelic variants. However, patients with AA or GA rs628031 genotypes had a higher incidence of poor response to imatinib compared to the GG genotype (47% compared to 29%, P = 0.06), and a higher rate of KD mutation discovery (8/16 vs. 5/27, P = 0.04), suggesting that secondary resistance was more common in these genotypes. Median EFS was shorter for rs628031 genotype AA/AG compared with the GG genotype (61 months and not reached, respectively, P = 0.05), and 5 yr OS rates were lower for patients with the rs628031 genotypes AA/AG compared with the GG genotype (88% and 97%, respectively, P = 0.03). Patients with AA/GA rs628031 and additional rare genotypes had worse EFS and OS compared to patients with only AA/GA rs628031 (P = 0.02 for EFS and 0.01 for OS). There was no difference in pretreatment SLC22A1 mRNA expression levels in patients with rs628031 genotypes GG/AA or GA. Conclusions: Studying SLC22A1 genetic variants prior to TKI initiation could influence treatment decisions. Key words CML; tyrosine kinase inhibitor; hOCT1; resistance; kinase domain mutations *Correspondence Maya Koren-Michowitz, Division of Hematology, The Chaim Sheba Medical Center, Tel-Hashomer, Ramat Gan, Israel 52621. Tel: 972 3 5302342; Fax: 972 3 5305343; e-mail: m.koren.michowitz@gmail.com There are no financial disclosures for any author.This study was supported (in part) by a generous grant from the Naor family (A.N.) and a Sarousy Foundation grant (A.N). Accepted for publication 6 November 2013 doi:10.1111/ejh.12235 Treatment with the tyrosine kinase inhibitor (TKI) imatinib has become the cornerstone of CML therapy and has mark- edly changed the outcome of this disease (1). Although sec- ond generation TKI including nilotinib (2, 3) and dasatinib (4) are being introduced into the first line therapy arsenal, imatinib is still the most commonly used TKI for the upfront treatment of CML worldwide. While most patients in early chronic phase analyzed in the IRIS trial with first line imatinib achieved major responses (82% achieving CCyR at the last, 6-yr follow-up report), only 63% of patients remain- ing on study at 6 yr were still in a CCyR (5). Cytogenetic and, recently, molecular responses at early time points during therapy with imatinib are predictive of event free survival and survival without progression to an accelerated phase (AP) or blast crisis (BC) (6, 7). Imatinib plasma levels during therapy can also predict response to therapy (8), and are a © 2013 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 283 European Journal of Haematology 92 (283–288)