Combining Dose-Response Slope with in Vitro Sensitivity to ABL Tyrosine Kinase Inhibitors As an Improved Correlative Metric for Predicting Clinical Response in Chronic Myeloid Leukemia

Abstract 1670

BCR-ABL mutations result in clinical resistance to ABL tyrosine kinase inhibitors (TKIs) in CML. Although in vitro IC₅₀ values for specific mutations have been suggested to guide TKI choice in the clinic, quantitative relationship between IC₅₀ and clinical response has never been demonstrated. Notably, IC₅₀ value constitutes only one point on the dose-response curve for a given drug. Most dose-response curves can be described by Hill's equation (equation 1), which incorporates both IC₅₀ and slope (m) parameters:

Here, f_a is cell fraction affected by treatment and D is drug dose. We report estimation of the slope of in vitro dose-response curves for wild-type and kinase domain-mutant BCR-ABL against clinical ABL TKIs for CML and examine the value of this incorporated parameter for predicting clinical response. Dose response curves for imatinib, nilotinib and dasatinib were determined by methanethiosulfonate-based cell viability assay in Ba/F3 cells expressing wild-type BCR-ABL or each of 15 specific kinase domain point mutations (O'Hare T. et al Cancer Res 65:4500-5). The parameters m and IC₅₀ were determined for each mutation and drug by fitting the log-transformed equation 1 to the respective dose-response curve. Excellent goodness of fit (r² range 0.94–0.99) was observed for all drug-mutation pairings, confirming the adequacy of Hill's equation to describe the effect of ABL TKIs on cellular viability in vitro. Inhibitory potential at peak concentration (IPP) was subsequently calculated as:

Here, f_u is cell fraction unaffected by treatment, and D is mean peak concentration (C_max) reported in plasma (Laneuville P. et al J Clin Oncol 28:e169-71). IPP and IC₅₀ values for each Ba/F3 BCR-ABL mutant were compared with previously reported CCyR rates for nilotinib (Kantarjian H. et al Blood 110:3540-6) and dasatinib (Muller M.C. et al Blood 114:4944-53). Consistent with the particularly negative effect of P-loop mutations on drug binding and clinical outcome with imatinib, we found that 4 of 7 these mutations tested (G250E, Y253H, E255K, E255V) showed lower dose-response slope relative to wild-type BCR-ABL in addition to high IC₅₀, while all other mutations showed variably increased slopes. Furthermore, the range of IPPs of imatinib for these mutations was lower than (and not overlapping with) all other mutations (0.084–1.66 vs 2.93–5.59; p=6*10⁻⁶). Slope variability provided particular additional interpretive value in cases where in vitro IC₅₀ and clinical response are disparate. For example, G250E and V379I mutants feature comparable cellular IC₅₀ values for imatinib (1184 and 1140 nM, respectively), but only G250E harbors worse clinical prognosis, arguably due to a lower dose-response slope of G250E, as also reflected in a lower IPP value compared to V379I (1.66 vs 3.04). Similarly, the M244V mutant does not confer substantial clinical resistance to imatinib despite increased IC₅₀, possibly due to an exceptionally high slope value (m=5 vs. m=1.87 for wild-type BCR-ABL) reflecting a very steep dose-response curve which may render patients with this mutation particularly vulnerable to consequences of unfavorable imatinib pharmacokinetic profile or reduced compliance. Lastly, we examined whether higher IPP values were predictive of better clinical response to nilotinib and dasatinib in second-line clinical setting. Dasatinib-treated patients with mutations resulting in high IPP (IPP>7) had a significantly higher mean CCyR rate than those patients with lower IPPs (53% vs 31%; p=0.038). In contrast, this relationship was not evident when IC₅₀ alone was used. In nilotinib-treated patients the difference in mean CCyR rate between patients with values above or below the median IPP or IC₅₀ value approached but did not reach statistical significance (p=0.055 for both cases), potentially due to the lower number of patients for which response data by mutation has been reported (7 vs 11 mutations, 65 vs 295 patients with mutations, for nilotinib and dasatinib respectively). Taken together, our data suggest an integrated metric such as IPP may have both further relevance in conjunction with individual pharmacokinetic measurements and application to improved interpretation of mutationally-guided TKI treatment selection in CML and other malignancies.