Tyrosine kinase inhibitors (TKIs) that preferentially inhibit the oncogenic Bcr-Abl fusion protein have radically changed the treatment of chronic myeloid leukaemia (CML). Imatinib mesylate ((IM) Gleevec®, Glivec®, STI571; Novartis) has been shown to induce almost complete hematological responses in patients after 12 months and major cytogenetic responses in 69%. However, the proportion of patients achieving molecular responses or having undetectable Bcr-Abl transcripts remains low. In addition to this failure to induce molecular response, resistance occurs in 10–15% of patients. Nilotinib ((NIL) AMN107; Novartis) has approximately 20-fold greater potency in in vitro assays but despite this improved potency in vitro and on bulk primary CML cells we have found that NIL is no better than IM in inducing apoptosis of the primitive CD34+ CML cell population. We previously showed that as with IM, CD34+ CML cells persist after NIL treatment and that NIL also has an anti-proliferative rather than pro-apoptotic effect, resulting in the accumulation of quiescent CD34+ cells. It has been proposed that this population of TKI-insensitive primitive cells may form a pool of disease in patients under treatment and contribute to the Bcr-Abl molecular signal detected in the majority of patients. This population must therefore, be specifically targeted in order to eradicate the disease in patients and to result in cure rather than control of the disease. One possible reason for the failure of NIL to kill CD34+ CML cells is that the cells do not accumulate sufficient intracellular levels of the drug due to either inadequate active uptake via the SLC transporter family, or to efflux via multidrug resistance proteins of the ATP-binding cassette (ABC) family. In order to determine the interaction of NIL with major clinically implicated drug transporters we have investigated interactions with ABCB1, MRP1 (ABCC1), ABCG2 and hOCT1 in CML cell lines and primitive (CD34+) primary CML cells.

Determination of the distribution coefficient (logD) showed that NIL is more lipophilic than IM (logD of 2.4 vs 0.8 respectively) and assays using 14C-NIL in OCT1 over expressing KCL22 CML cells showed that accumulation of NIL is neither temperature dependent nor reduced by the inhibitors prazosin or amatadine. These data demonstrate that NIL uptake is not dependent on active import by hOCT1, however NIL did inhibit the uptake of a known OCT1 substrate tetraethylammonium bromide (TEA).

Efflux assay in cell lines over-expressing MDR1, MRP1 or ABCG2 showed that NIL is not effluxed by any of these proteins. Over expression &/or the addition of specific inhibitors (PSC833, MK571, fumitremorgin C respectively) did not alter the cellular accumulation of NIL as assayed by radiolabelled drug accumulation assays, transepithelial transport, HPLC analysis or when p-CrkL was measured as a surrogate assay for Bcr-Abl inhibition. Using efflux assays with known substrates we found that NIL inhibited rhodamine efflux via MDR1, or BODIPY-prazosin efflux via ABCG2, but had no effect on Fluo-3 efflux via MRP1. Thus, rather than being a substrate NIL acts as an inhibitor of MDR1 and ABCG2 in these assays.

To conclusively assess the effect of transporter activity in primary cells we first measured the expression of the transporters in primitive CD34+ CML cells. Unlike OCT1, the efflux transporters MDR1, MRP1 and ABCG2 are all expressed on CML CD34+ cells. When compared to normal CD34+ cells ABCG2 was over expressed (291%), MRP1 was expressed at very similar levels (108%) and surprisingly MDR1 was expressed at a much lower level in CML (13.5%). We treated CD34+ CML cells with NIL for 72hrs in the presence or absence of the transporter inhibitors. NIL reduced the total cell number and p-CrkL activity as expected however, these effects were not potentiated by any of the transporter inhibitors. Furthermore, NIL increased the number of quiescent CD34+ cells remaining after treatment but again transporter inhibitors did not modulate this effect.

Therefore, we have found no evidence for either active uptake of NIL via hOCT1 or efflux via MDR1, MRP1 or ABCG2. It is unlikely that these transporters will have any effect on the clinical response to this drug in either MNCs or the more resistant CD34+ stem cell population.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author