Crenolanib Inhibits the Drug-Resistant PDGFRA D842V Mutation Associated with Imatinib-Resistant Gastrointestinal Stromal Tumors

Published OnlineFirst June 27, 2012; DOI: 10.1158/1078-0432.CCR-12-0625

Clinical Cancer Cancer Therapy: Preclinical Research

Michael C. Heinrich1, Diana Grifﬁth1, Arin McKinley1, Janice Patterson1, Ajia Presnell1, Abhijit Ramachandran2, and Maria Debiec-Rychter3

Abstract Purpose: To determine the potential of crenolanib, a potent inhibitor of PDGFRA, to treat malignancies driven by mutant PDGFRA. Experimental Design: The biochemical activity of crenolanib was compared with imatinib using a panel of PDGFRA-mutant kinases expressed in several different cell line models, including primary gastrointestinal stromal tumors (GIST) cells. The antiproliferative activity of crenolanib was also studied in several cell lines with PDGFRA-dependent growth. Results: Crenolanib was signiﬁcantly more potent than imatinib in inhibiting the kinase activity of imatinib-resistant PDGFRA kinases (D842I, D842V, D842Y, DI842-843IM, and deletion I843). For example, crenolanib was 135-fold more potent than imatinib against D842V in our isogenic model system,

with an IC50 of approximately 10 nmol/L. The relative potency of crenolanib was further conﬁrmed in BaF3 and primary GIST cells expressing PDGFRA D842V. In contrast, imatinib was at least 10-fold more potent than crenolanib in inhibiting the V561D mutation. For all other tested PDGFRA mutations, crenolanib and imatinib had comparable potency. Conclusions: Crenolanib is a potent inhibitor of imatinib-resistant PDGFRA kinases associated with GIST, including the PDGFRA D842V mutation found in approximately 5% of GISTs. The spectrum of activity of crenolanib suggests that this drug is a type I inhibitor (inhibitor of activated conformation of kinase). Based in part on these results, a phase II clinical study of this agent to treat GIST with the PDGFRA D842V mutation has been initiated. Clin Cancer Res; 18(16); 4375–84. 2012 AACR.

Introduction produces high rates of clinical beneﬁt (PR þ SD) and Activating mutations of receptor tyrosine kinases KIT or progression-free survival in the range of 18 to 24 months. PDGFRA are fundamental to the pathogenesis of most In addition, median overall survival of patients with met- gastrointestinal stromal tumors (GIST). More than 80% of astatic GIST is now approximately 50 to 60 months (2–4). GISTs express mutated, constitutively active KIT receptor, In the second-line setting, treatment with sunitinib another 5% to 10% express mutated PDGFRA, whereas 10% increases progression-free and overall survival of patients to 15% of cases have no mutations in either of these kinases with imatinib-resistant disease compared with placebo (5). (wild-type GIST, WT; refs. 1, 2). Patients with primary KIT exon 11 mutations have the The use of KIT/PDGFRA tyrosine kinase inhibitors (TKI) longest progression-free and overall survival during imati- has transformed the treatment of localized and advanced nib treatment (3, 6). Congruent with this observation, KIT GIST. Front-line treatment of metastatic GIST with imatinib exon 11–mutant kinases are potently inhibited by imatinib and sunitinib in vitro (7). However, treatment of PDGFRA-mutant GIST with currently available PDGFRA TKIs (imatinib, sunitinib, and 1 Authors' Afﬁliations: Portland VA Medical Center and OHSU Knight others) has yielded mixed results, with outcomes being Cancer Institute, Portland, Oregon; 2AROG Pharmaceuticals LLC, Dallas, Texas; and 3Department of Human Genetics, K.U.Leuven and University closely correlated with the intrinsic sensitivity of different Hospitals, Leuven, Belgium mutant PDGFRA oncoproteins to various inhibitors. In Corresponding Author: Michael C. Heinrich, Division of Hematology/ vitro, some PDGFRA-mutant kinases (e.g., V561D) are Oncology, Departments of Medicine and Cell and Developmental Biology, extremely sensitive to imatinib, and patients with these Portland VA Medical Center and OHSU Knight Cancer Institute, Oregon PDGFRA Health & Science University, R&D-19 3710 U.S. Veterans Hospital Road, underlying imatinib-sensitive mutations seem to Portland, OR 97239. Phone: 503-220-3405; Fax: 503-273-5158; E-mail: have similar clinical outcomes as patients with KIT exon 11 [email protected] mutations (7–10). doi: 10.1158/1078-0432.CCR-12-0625 In contrast, the most common PDGFRA mutation asso- 2012 American Association for Cancer Research. ciated with GIST, D842V, is strongly resistant to inhibition

www.aacrjournals.org 4375

Heinrich et al.

tested the activity of crenolanib against GIST-associated Translational Relevance PDGFRA-mutant kinases in a variety of models. We iden- Approximately 5% to 10% of gastrointestinal stromal tified crenolanib as the most potent PDGFRA D842V kinase tumors (GIST) have pathogenetic activating mutations inhibitor described to date with an IC50 in the range of of the PDGFRA receptor tyrosine kinase. The most 10 nmol/L. Notably, crenolanib was at least 100-fold more common of these mutant kinases, PDGFRA D842V, is potent than imatinib for biochemical inhibition of resistant to inhibition by imatinib or sunitinib, which PDGFRA D842V kinase. We also report herein the biochem- are currently approved as first-line and second-line ical activity of crenolanib against a larger panel of imatinib- treatment of advanced GIST, respectively. In addition, sensitive and imatinib-resistant PDGFRA-mutant kinases. there are other PDGFRA mutations that are also resis- On the basis of our results, we hypothesize that crenolanib tant to imatinib. Patients whose GIST have a PDGFRA could be an effective treatment for patients with PDGFRA- D842V-mutant kinase have a markedly inferior pro- mutant GIST, particularly those patients whose GIST gression-free and overall survival compared with express the PDGFRA D842V–mutant oncogenic kinase. patients with KIT-mutant GIST. In this study, we identified crenolanib as a potent inhibitor of PDGFRA Materials and Methods D842V-mutant kinase with an IC50 of approximately DiscoveRx–KINOMEscan: Kd ELECT 10 nmol/L. Thus, crenolanib is 100- to 150-fold more Crenolanib was profiled using the commercially avail- potent than imatinib against PDGFRA D842V. In addi- able KINOMEscan: Kd ELECT screening service (Disco- tion, crenolanib has similar potency against less com- veRx–KINOMEscan) as previously described (18, 19). PDGFRA mon imatinib-resistant mutations. On the KINOMEscan: Kd ELECT is based on a competition-binding basis of these results, a phase II study of crenolanib assay that quantitatively measures the ability of a com- for advanced PDGFRA D842V-mutant GIST has been pound to compete with an immobilized, active site–direct- initiated (NCT01243346). ed ligand. The assay was carried out by combining 3 components: DNA-tagged kinase, immobilized ligand, and a test compound. The ability of the test compound to compete with the immobilized ligand was measured via by imatinib or sunitinib (7, 8, 11, 12). In addition to the quantitative PCR of the DNA tag. Inhibitor binding con- D842V mutation, there are other less common PDGFRA- stants (Kd values) were calculated from duplicate 11-point mutant kinases that are also resistant to imatinib in vitro dose–response curves. Measurements were made under (e.g., PDGFRA D842Y or PDGFR DI842-843IM; ref. 13). optimized conditions that generate true thermodynamic Patients with D842V-mutant GIST have a very low rate of Kd values that facilitate direct comparison of inhibitor clinical benefit from imatinib treatment. For example, Biron affinity across kinases. and colleagues conducted a survey of international GIST treatment centers and collected data on 20 patients with Reagents and antibodies locally advanced or metastatic GIST treated with front-line Imatinib was obtained commercially from LC Laborato- imatinib. In this series, no objective responses were seen. In ries and reconstituted in PBS to yield a 10 mmol/L stock. addition, the median progression-free survival was a mere Crenolanib was obtained from AROG Pharmaceuticals, 2.8 months, and the median overall survival was only 12.7 LLC and reconstituted in dimethyl sulfoxide (DMSO) to months (14). Similar results were seen in smaller numbers yield a 10 mmol/L stock. Working solutions of imatinib or of PDGFRA D842V–mutant GIST patients treated as part of crenolanib were diluted in growth media or PBS for use in phase II–III trials of imatinib for advanced GIST (6, 8–10). experiments. For PDGFRA immunoprecipitation experi- The PDGFRA D842V–mutant kinase is also strongly resis- ments, we used rabbit polyclonal anti-PDGFRA antibody tant to sunitinib in vitro. Limited clinical data suggest that (Santa Cruz Biotechnology, SC-20) and Protein A/G beads sunitinib has no activity against GIST with primary or (Santa Cruz). For detection of phospho-PDGFRA and total secondary PDGFRA D842V mutations (7). PDGFRA, we used anti-phosphotyrosine monoclonal anti- The above data indicate that patients with TKI-resistant body (PY-20, 1:500; BD Transduction Labs) and anti- PDGFRA-mutant kinases do not benefit from standard TKI PDGFRA rabbit polyclonal antibody (SC-20, 1:500; Santa therapy. On the basis of underlying molecular similarity of Cruz), respectively. KIT immunoprecipitation experiments PDGFRA- and KIT-mutant GIST, it is predicted that GIST were carried out as previously described (7). For detection of with drug-resistant PDGFRA D842V mutations would phospho- and total-KIT, we used rabbit polyclonal antibody respond to effective kinase inhibition (15). However, cur- to P-KIT Tyr 719 (Cell Signaling Technology; catalog no. rently there is no proven effective medical treatment for 3391, 1:500) or total KIT (Santa Cruz; C-19, 1:500), patients with PDGFRA D842V–mutant GIST. respectively. Crenolanib (formerly CP-868,596), an orally bioavail- able benzimidazole, is a selective and potent inhibitor of Cell lines PDGFRA and PDGFRB. This agent has been tested in both The BaF3 V561D and BaF3 D842V cell lines have been phase I single agent and phase Ib combination therapy previously described (13). These cell lines were generated by clinical studies and found to be well tolerated (16, 17). We transfection of mutant PDGFRA isoforms into BaF3 cells.

4376 Clin Cancer Res; 18(16) August 15, 2012 Clinical Cancer Research

Crenolanib Inhibits PDGFRA D842V

Clones were selected for G418 resistance and interleukin-3 Proliferation assays independence. DNA from stable lines was checked to Cells were added to 96-well plates at densities of 20,000 ensure that the expected mutant PDGFRA sequence was cells per well and incubated with imatinib or crenolanib for present, and expression was verified by immunoblotting for 72 hours before measuring cellular proliferation using a 2,3- total and tyrosyl phosphorylated forms of PDGFRA. The bis[2-methoxyl-4-nitro-5-sulfophenyl]-2H-tetrazolium-5- GIST T1 cell line was a generous gift of Dr. Taguchi (Kochi carboxanilide (XTT)–based assay (Roche Molecular Bio- University, Kochi, Japan). This cell line was established chemicals; ref. 13). from an untreated GIST and expresses an exon 11 deletion constitutively active KIT kinase (20). The EOL-1 cell line was Results derived from a patient with chronic eosinophilic leukemia and expresses the constitutively activated FIP1L1-PDGFRA DiscoveRx–KINOMEscan: Kd ELECT of crenolanib fusion kinase (DSMZ Tissue Bank; ref. 21). The H1703 Crenolanib was developed as an orally available inhibitor non–small cell lung cancer (NSCLC) cell line was obtained of PDGFR kinases with more than 100-fold selectivity from the American Tissue Type Collection. The H1703 cell versus a limited number of profiled kinases (VEGFR2, line has been previously reported to have 24-fold amplifi- TIE-2, and FGFR-2). To better characterize the activity of cation of the 4q12 region that contains the PDGFRA crenolanib, we carried out competitive binding assays to locus (22). quantify the affinity of crenolanib for a panel of recombinant kinases (DiscoveRx–KINOMEscan: Kd ELECT). Biochemical assessment of PDGFRA/KIT kinase activity The results showed that crenolanib is a specific and Chinese hamster ovary (CHO) cells were transiently potent inhibitor of class III receptor tyrosine kinases transfected with mutated KIT or PDGFRA cDNA con- (RTK). The inhibitor-binding constant (Kd) of crenolanib structs and treated with various concentrations of for the wild-type receptors PDGFRA, PDGFRB, and FLT3 imatinib or crenolanib as previously described (7, 13). was 3.2, 2.1, and 0.74 nmol/L, respectively (Table 1). In Experiments involving recombinant DNA were carried contrast to other known inhibitors of PDGFR (e.g., ima- out using biosafety level 2 conditions in accordance with tinib, sunitinib, and dasatinib), crenolanib has 25-fold published guidelines. Protein lysates from cell lines were more affinity for PDGFRA/B compared with KIT. Besides prepared and subjected to immunoprecipitation using class III RTKs, crenolanib does not inhibit any other anti-KIT or anti-PDGFRA antibodies followed by sequen- known tyrosine or serine/threonine kinases at clinically tial immunoblotting for phospho-KIT and total KIT, or achievable concentrations. Notably, crenolanib was iden- phosphotyrosine or total PDGFRA, respectively, as pre- tified in this study as an extremely potent inhibitor of viously reported (7, 13). Densitometry was carried out to FLT3; the activity of crenolanib against FLT3 has not been quantify drug effect using Photoshop 5.1 software, with previously reported. the level of phospho-KIT or phospho-PDGFRA normal- ized to total protein. Densitometry and proliferation experimental results were analyzed using Calcusyn 2.1 software (Biosoft) to mathematically determine the IC50 Table 1. The inhibitor binding constant (Kd)of values. The Wilcoxon rank sum test was used to compare crenolanib for RTKs the IC50 values of imatinib and crenolanib for a given mutation. Crenolanib RTK K (nmol/L) Ex vivo assay using primary GIST cells d Surgical specimens of primary GIST from imatinib-naive CFS1R 30 patients were used for the primary cell cultures, as previ- FLT3 0.74 ously described (12). The DNA was isolated and PDGFRA KIT 78 D842V mutation was identified by direct sequencing, PDGFRA 2.1 according to standard procedures. For Western immuno- PDGFRB 3.2 blotting, primary GIST cells obtained from collagenase- NOTE: Crenolanib was profiled using the commercially avail- disaggregated tumor specimens were seeded in duplicate able KINOMEscan: Kd ELECT screening service as previ- at 80% confluence in 25-mm diameter cell culture dishes ously described (18, 19). Kd measurements were based on a (Corning Inc.) and grown in Dulbecco’s modified Eagle’s competition-binding assay that quantitatively measures the medium supplemented with 10% FBS, 1.0 mmol/L non- ability of a compound to compete with an immobilized, essential amino acids, and 1.0 mmol/L sodium pyruvate for active-site directed ligand. Inhibitor binding constants (K d 24 hours at þ37 C. Next, the cells were exposed to a values) were calculated from duplicate 11-point dose– crenolanib in different concentrations, or to vehicle alone response curves. Measurements were made under opti- þ (DMSO), and incubated for 2 hours at 37 C. After a mized conditions that generate true thermodynamic Kd wash in ice-cold PBS, cells were lysed and immunoblotted values that facilitate direct comparison of inhibitor affinity using anti-phospho-PDGFRA(Tyr754) and anti-PDGFRA across kinases. antibodies.

www.aacrjournals.org Clin Cancer Res; 18(16) August 15, 2012 4377

Heinrich et al.

In vitro measure of IC50 activity of crenolanib or 6 to 10 times more potent than crenolanib for inhibiting imatinib against specific mutants in an isogenic model PDGFRA mutations involving the juxtamembrane domain system (V561D, RV560-561RERV). There is no difference in poten- The above results indicate that crenolanib is a fairly cy between imatinib or crenolanib for inhibiting an ATP- specific kinase inhibitor with only minimal predicted binding pocket activating PDGFRA mutation (N659K, "off-target" effects. However, these binding assays are not exon 14). always predictive of biochemical activity (23). Therefore, we We and others have previously noted that mutations sought to confirm these results in biochemical assays using involving PDGFRA exon 18, which encode the PDGFRA full-length kinase. To determine the potential clinical effi- activation loop, can be divided into groups of mutations cacy of crenolanib in the treatment of GIST, we measured that are sensitive to imatinib (e.g., deletion DIMH842-845) the activity of crenolanib against gain-of-function PDGFRA and those that are resistant (e.g., D842V; refs. 8, 11, 12). For mutations associated with GIST. In particular, we analyzed the imatinib-sensitive mutations that we tested (deletion the activity of crenolanib against imatinib-resistant DIMH842-845, deletion I843, deletion HDSN845-848R, PDGFRA oncoproteins. D846Y, and N848K), there was no significant difference Mutant isoforms corresponding to selected human between imatinib and crenolanib for inhibition of kinase PDGFRA mutations were created by site-specific mutagen- activity (Table 2, Fig. 1). In contrast, crenolanib is signifi- esis and expressed in CHO cells by transient transfection as cantly more potent than imatinib for inhibiting mutant previously described (8, 13). All of the expressed mutants kinases that were previously reported to be imatinib resis- showed phosphorylation of tyrosine residues in the absence tant (PDGFRA D842I, D842V, D842Y, and DI842-843IM). of PDGF-AA ligand, implying constitutive activation of their Notably, crenolanib is approximately 135-fold more potent kinase domains. We directly compared the biochemical than imatinib for inhibiting the PDGFRA D842V mutation potency of imatinib and crenolanib to inhibit PDGFRA in this model system. activity (as assessed by inhibition of autophosphorylation) Previous studies of crenolanib have indicated that it is in the same experiment. significantly less potent against KIT than against PDGFRA Consistent with previous data, both imatinib and creno- (16). To confirm these results, we compared the activity of lanib are potent inhibitors of PDGF-AA ligand stimulated imatinib and crenolanib against KIT-mutant isoforms asso- WT PDGFRA (7, 8, 16). Both compounds have IC50 values ciated with GIST or mastocytosis (Table 2). The IC50 for in the very low nanomolar range with no difference in crenolanib for a KIT exon 11 deletion mutant kinase is potency between these drugs (Table 2, Fig. 1). Imatinib is greater than 1,000 versus 8 nmol/L for imatinib. Neither

Table 2. Biochemical IC50 values for inhibition of kinase activity in cells expressing single mutation kinases

Kinase Exon Model Imatinib Crenolanib P FIP1L1-PDGFRA 12 EOL-1 1 0.2 21 13 0.13 PDGFRA WT None CHO 9 411 10 0.62 PDGFRA V561D 12 CHO 32 18 319 75 0.006a PDGFRA N659K 14 CHO 56 37 76 67 0.77 PDGFRA RV560-561RERV 14 CHO 5 344 19 0.11 PDGFRA D842V 18 CHO 1,353 311 9 3 <0.001a PDGFRA D842I 18 CHO >1,000 124 36 0.02a PDGFRA D842Y 18 CHO 388 137 88 23 0.06 PDGFRA deletion DIMH842-845 18 CHO 21 323 9 0.75 PDGFRA DI842-843IM 18 CHO 781 342 10 2 0.003a PDGFRA deletion I843 18 CHO 67 43 197 71 0.12 PDGFRA deletion HDSN8455-848P 18 CHO 266 105 87 52 0.30 PDGFRA D846Y 18 CHO 29 13 18 6 0.67 PDGFRA N848K 18 CHO 22 739 16 0.16

KIT exon 11 deletion 11 GIST-T1 8 4 >1,000 0.05a KIT D816V 17 CHO >1,000 >1,000 1.00 PDGFRA V561D 12 BaF3 13 12 134 71 0.03a PDGFRA D842V 18 BaF3 272 163 2 2 0.002a

NOTE: The values for crenolanib and imatinib represent the biochemical IC50 expressed in nmol/L units the SEM. Values represent the data from at least 3 replicate experiments per mutation. aP < 0.05 by Wilcoxon rank sum test.

4378 Clin Cancer Res; 18(16) August 15, 2012 Clinical Cancer Research

Crenolanib Inhibits PDGFRA D842V

Figure 1. Sensitivity of single mutant PDGFRA isoforms to imatinib or crenolanib. CHO cells were transfected with PDGFRA constructs and 24 hours later were incubated for 90 minutes in the absence or presence of imatinib or crenolanib at the indicated concentrations. PDGFRA was immunoprecipitated from cell extracts and analyzed by immunoblotting using anti-PDGFRA (for total PDGFRA) and anti-phosphotyrosine (for phospho-PDGFRA) antibodies. Representative results from a minimum of 3 replicate experiments per mutant kinase are shown.

www.aacrjournals.org Clin Cancer Res; 18(16) August 15, 2012 4379

Heinrich et al.

Table 3. Biochemical IC50 values for inhibition of PDGFRA kinase activity in CHO cells expressing compound-mutant kinases

Kinase Exons Imatinib Crenolanib P V561D þT674I 12 þ 14 >1,000 >1,000 0.99 V561D þ D842V 12 þ 18 >1,000 26 6 0.0002 T674I þ D842V 14 þ 18 >1,000 24 12 0.006

NOTE: The values for crenolanib and imatinib represent the biochemical IC50 expressed in nmol/L units the SEM. Values represent the data from at least 3 replicate experiments per mutation.

crenolanib nor imatinib had any signiﬁcant biochemical unexpected, as the isolated V561D mutation is relatively activity against the D816V mutation, which is homologous resistant to crenolanib. We also found that addition of the to the PDGFRA D842V mutation associated with mastocy- gatekeeper T674I mutation to the D842V mutation had no tosis (IC50 > 1,000 nmol/L for both; ref. 15). signiﬁcant effect on the potency of crenolanib, indicating that The molecular mechanisms underlying the secondary crenolanib binding is not dependent upon the T674 residue. clinical progression of imatinib-sensitive PDGFRA-mutant GIST are poorly understood. However, one reported sec- Comparative activity of crenolanib and imatinib in ondary mutation has been described—acquisition of a cellular models secondary PDGFRA D842V mutation (7). This same muta- The above studies were conducted in a CHO transient tion has also been described as secondary drug resistance transfection system that allows us to directly compare the mutation in patients with hypereosinophillic syndrome activity of different TKIs on the same mutant kinase in an with a FIP1L1-PDGFRA fusion who are treated with imati- isogenic system. To extend these results, we also compared nib (24). In addition, the secondary gatekeeper mutation the potency of crenolanib and imatinib in a panel of (PDGFRA T674I) has been reported in a patient with PDGFRA-dependent cell lines. These cell lines allowed us FIP1L1-PDGFRA–positive hematologic neoplasm treated to not only assess biochemical potency of the 2 inhibitors with PDGFRA inhibitors (25). but also to measure the effect of PDGFRA kinase inhibition We expressed the compound PDGFRA mutants V561D þ on cellular proliferation. T674I (exon 12 þ gatekeeper), V561D þ D842V (exon 12 þ The EOL-1 cell line is derived from a patient with chronic 18), and T674I þ D842V (gatekeeper þ exon 18) in CHO eosinophilic leukemia and expresses the constitutively acti- cells and compared the activity of imatinib or crenolanib vated FIP1L1-PDGFRA fusion kinase (21). Both imatinib against these compound-mutant kinases (Table 3, Fig. 2). and crenolanib potently inhibit the kinase activity of the Neither imatinib nor crenolanib has any activity against the fusion oncogene with IC50 values of 1 and 21 nmol/L, V561D þ T674I-mutant kinase (IC50 > 1,000 nmol/L for respectively (P ¼ 0.13; Fig. 3A). In addition, both com- each). However, crenolanib has low nanomolar potency pounds potently inhibit the proliferation of EOL-1 cells against the V561D þ D842V-mutant kinase that is similar with a 50% growth inhibitory concentration of 0.2 pmol/L to its potency against the isolated D842V mutation. This was (Fig. 3B).

Figure 2. Sensitivity of compound-mutant PDGFRA isoforms to imatinib or crenolanib. CHO cells were transfected with PDGFRA constructs and analyzed exactly as described in Fig. 1. Representative results from a minimum of 3 replicate experiments per mutant kinase are shown.

4380 Clin Cancer Res; 18(16) August 15, 2012 Clinical Cancer Research

Crenolanib Inhibits PDGFRA D842V

Figure 3. A, biochemical activity of imatinib or crenolanib to inhibit mutant kinases expressed in PDGFRA-dependent cell lines. Cell lines were incubated for 90 minutes in the absence or presence of imatinib or crenolanib at the indicated concentrations. PDGFRA was immunoprecipitated from cell extracts and analyzed by immunoblotting using anti-PDGFRA (for total PDGFRA) and anti-phosphotyrosine (for phospho-PDGFRA) antibodies as in Fig. 1. FIP1L1- PDGFRA panel indicates experiments using EOL-1 cells. Representative results from a minimum of 3 replicate experiments per mutant kinase are shown. B, antiproliferative activity of imatinib or crenolanib against mutant kinases expressed in PDGFRA-dependent cell lines. Cell lines were incubated for 72 hours in the absence or presence of imatinib or crenolanib at the indicated concentrations. Proliferation/cell number was assessed using an XTT-based assay system as previously described. Error bars indicate the SD for the depicted experimental conditions. Each TKI drug concentration was assayed in triplicate. Representative results from a minimum of 3 independent experiments per cell line are depicted.

We also tested the comparative potency of imatinib and the PDGFRA locus. This cell line is dependent upon crenolanib to inhibit the activation of V561D- or D842V- PDGFRA kinase activity for proliferation, as evidenced by mutant kinases expressed in BaF3 cells. Consistent with our inhibition of growth by short hairpin RNA against PDGFRA results in the transient transfection model (CHO cells), or PDGFRA TKIs (imatinib, sunitinib; ref. 22). Both ima- imatinib is more potent than crenolanib in inhibiting the tinib and crenolanib inhibited PDGFRA activation in this activity of V561D (IC50 11 vs. 85 nmol/L, P ¼ 0.05). In cell line with IC50 values of 93 and 26 nmol/L, respectively addition, we confirmed that crenolanib is 100-fold more (Fig. 3B, P ¼ 0.4). Both drugs also inhibited the prolifer- potent than imatinib for inhibiting activity of the D842V ation of these cells in a similar dose range (data not shown). mutation (IC50 2 vs. 272 nmol/L, P ¼ 0.002; Fig. 3A). Finally, we tested the biochemical activity of crenolanib in Similar to the biochemical data, crenolanib is significantly 2 primary GIST cells with PDGFRA D842V mutation by ex more potent than imatinib for inhibiting proliferation of vivo assay. Crenolanib, at a concentration of 10 nmol/L, the BaF3 D842V cell line, with IC50 values of 22 and 1,510 completely inhibits PDGFRA D842V activation, as assessed nmol/L, respectively (Fig. 3B). by autophosphorylation (Fig. 4). Due to the limited number PDGFRA genomic amplification leading to constitutive of primary cells available, we did not carry out comparative PDGFRA activation has been reported in a number of studies against imatinib nor conducted experiments to assess malignancies including NSCLC and gliomas (22, 26–28). the effect of crenolanib on cellular proliferation. Notably, in We tested the activity of imatinib and crenolanib against the previously published studies using these cells, concentra- H1703 NSCLC cell line that has been previously reported to tions of imatinib up to 5,000 nmol/L had only a minimal have 24-fold amplification of the 4q12 region that contains effect on PDGFRA D842V autophosphorylation (29).

www.aacrjournals.org Clin Cancer Res; 18(16) August 15, 2012 4381

Heinrich et al.

The D842V mutation is the most common primary PDGFRA mutation found in GIST but can also develop as a secondary resistance mutation following imatinib treatment of PDGFRA-mutant tumors (7). Crenolanib potently inhibited the activity of the compound V561D þ D842V- mutant kinase and retained activity when the gatekeeper T674I mutation was added to the D842V mutation. Figure 4. Crenolanib inhibits PDGFRA D842V in primary GIST cells by ex Zhang and colleagues have classified TKIs into type I and vivo assay. Two independently established primary GIST cell lines were type II inhibitors. Type I inhibitors can only bind to the exposed to a crenolanib in different concentrations, or to vehicle alone active kinase conformation (30). In contrast, type II inhi- (DMSO), and incubated for 2 hours at 37 C. Cell extracts were analyzed bitors, such as imatinib and sunitinib, can only bind to the by Western blotting using anti-PDGFRA (for total PDGFRA) and anti- phospho-PDGFRA(Tyr754) antibodies. inactive conformation of the kinase and function by pre- venting the enzyme from assuming the active conformation (30, 31). Of note, the D842V mutation is felt to stabilize the Discussion PDGFRA activation loop in the active conformation; this TKI inhibitor therapy of GIST has revolutionized the conformation cannot bind imatinib or sunitinib (15, 32). treatment of advanced metastatic or unresectable GIST. However, in this study, we found that this mutation is very Before the introduction of TKI therapy, there was no effec- potently inhibited by crenolanib. tive medical therapy for GIST as these tumors are uniformly The spectrum of activity of crenolanib suggests that it is a resistant to chemotherapy. However, the benefits of TKI type I inhibitor. The ability of both imatinib and crenolanib therapy strongly correlate with the underlying biology of the to inhibit certain mutations with equivalent potency (e.g., tumor. Notably, the best results are seen for treatment of KIT del DIMH842-845), suggests that these particular mutant exon 11–mutant GISTs that express mutant kinases that are kinases are in an equilibrium with roughly equal popula- extremely sensitive to in vitro inhibition by imatinib or tions of inactive and active kinase isoforms, thus allowing sunitinib (2). either type I and II inhibitors to bind and thereby inhibit Although KIT is the most commonly mutated kinase in kinase activity. On the basis of this model, the relative GIST, approximately 5% to 10% of GISTs express exclusive- potency of imatinib and crenolanib against PDGFRA- ly a mutated PDGFRA kinase (2). The response of patients mutant isoforms may provide insight into how different with PDGFRA-mutant GIST to TKIs correlates with intrinsic mutations affect the distribution of protein isoforms sensitivity of the mutant kinase expressed by tumor cells. between the inactive and active state. In addition, further Patients whose GIST have an imatinib-sensitive PDGFRA structure-based studies of crenolanib and related com- mutation (e.g., PDGFRA exon 12 V561D mutation) seem to pounds may aid in the development of even more potent have similar clinical outcomes as patients whose tumor has type I PDGFRA kinase inhibitors. a KIT exon 11 mutation. In contrast, the most common We hypothesize that crenolanib may be useful for the PDGFRA mutation associated with GIST (PDGFRA exon 18 treatment of GIST with an associated PDGFRA mutation. In D842V) has been shown to be resistant to imatinib and phase I–II studies, orally administered creonolanib was well sunitinib in vitro. Consequently, patients with PDGFRA tolerated and exhibited favorable pharmacokinetic proper- D842V–mutant GIST treated with conventional GIST TKIs ties. The dose-limiting toxicities included hematuria, have a very low rate of clinical benefit, and the median increased liver function tests, and nausea/vomiting (16, overall survival for these patients is approximately 1 year, 17). Based in part on these results and the prior clinical compared with the median of 4 to 5 years seen for patients experience with crenolanib, a multicenter phase II study of with KIT exon 11–mutant GIST. In addition to the PDGFRA crenolanib for treatment of PDGFRA D842V–mutant GIST D842V mutation, there are also a number of less common has been initiated (NCT01243346). PDGFRA mutations that also are resistant to imatinib in vitro In addition to a subset of GIST, genomic alterations (6, 8–10, 13, 14). resulting in PDGFRA or PDGFRB activation have been We tested the activity of crenolanib, a potent PDGFRA/B reported in other human malignancies, including myelo- TKI to inhibit a panel of PDGFRA-mutant kinases. Creno- proliferative disorders (e.g., F1P1L1-PDGFRA, BCR- lanib was significantly more potent than imatinib in inhi- PDGFRA, and ETV6-PDGFRB), gliomas (amplification of biting the kinase activity of imatinib-resistant PDGFRA PDGFRA), dermatofibrosarcoma protuberans (activation kinases (D842I, D842V, D842Y, DI842-843IM, and dele- of PDGFRB due to paracrine secretion of COL1A1-PDGFB tion I843). For example, crenolanib was 135-fold more fusion protein), and NSCLC (genomic amplification of potent than imatinib at D842V kinase activity in our iso- PDGFRA). Also, PDGFRB activation may contribute to genic model system. The relative potency of crenolanib was tumor angiogenesis either directly (endothelial cells) or further confirmed in BaF3 and primary GIST cells expressing indirectly (support of pericytes). Therefore, crenolanib may PDGFRA D842V. In contrast, imatinib was 6- to 10-fold be useful in the treatment of cancers that have abnormal more potent than crenolanib in inhibiting the V561D activation of PDGFRA or PDGFRB. On the basis of prior mutation. For all other tested PDGFRA mutations, creno- studies of the role of PDGFRA in the pathogenesis/biology lanib and imatinib had comparable kinase activity. of gliomas, several phase II studies of crenolanib for

4382 Clin Cancer Res; 18(16) August 15, 2012 Clinical Cancer Research

Crenolanib Inhibits PDGFRA D842V

treatment of pediatric or adult glioma have been initiated Writing, review, and/or revision of the manuscript: M.C. Heinrich, D. Griffith, A. Ramachandran, M. Debiec-Rychter (NCT01229644, NCT01393912; refs. 22, 26–28, 33, 34). Administrative, technical, or material support (i.e., reporting or orga- nizing data, constructing databases): M.C. Heinrich, D. Griffith, A. Disclosure of Potential Conflicts of Interest McKinley, A. Presnell, A. Ramachandran M.C. Heinrich has received commercial research grants from AROG, Study supervision: M.C. Heinrich, M. Debiec-Rychter Novartis, Imclone, and Ariad, and honoraria from speakers bureau of Novartis. He also has ownership interest (including patents) from Molecu- larMD and is a consultant and an advisory board member of Novartis, Pfizer, and MolecularMD. No potential conflicts of interest were disclosed by the Grant Support other authors. The work was supported in part from funding from a Veterans Affairs Merit Review Grant (M.C. Heinrich), AROG Pharmaceuticals LLC, the Life Authors' Contributions Raft Group (M.C. Heinrich, M. Debiec-Rychter), and the GIST Cancer Conception and design: M.C. Heinrich, J. Patterson Research Fund (M.C. Heinrich). Development of methodology: M.C. Heinrich, M. Debiec-Rychter The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked Acquisition of data (provided animals, acquired and managed patients, advertisement provided facilities, etc.): M.C. Heinrich, D. Griffith, A. McKinley, J. Pat- in accordance with 18 U.S.C. Section 1734 solely to indicate terson, A. Ramachandran, M. Debiec-Rychter this fact. Analysis and interpretation of data (e.g., statistical analysis, biosta- tistics, computational analysis): M.C. Heinrich, D. Griffith, A. McKinley, J. Received February 22, 2012; revised May 24, 2012; accepted June 11, 2012; Patterson, A. Ramachandran, M. Debiec-Rychter published OnlineFirst June 27, 2012.

References 1. Corless CL, Heinrich MC. Molecular pathobiology of gastrointestinal stromal tumors and activity of the PKC412 inhibitor against imati- stromal sarcomas. Annu Rev Pathol 2008;3:557–86. nib-resistant mutants. Gastroenterology 2005;128:270–9. 2. Corless CL, Barnett CM, Heinrich MC. Gastrointestinal stromal tumours: 13. Corless CL, Schroeder A, Griffith D, Town A, McGreevey L, Harrell P, origin and molecular oncology. Nat Rev Cancer 2011;11:865–78. et al. PDGFRA mutations in gastrointestinal stromal tumors: frequen- 3. Gastrointestinal Stromal Tumor Meta-Analysis Group (MetaGIST). cy, spectrum and in vitro sensitivity to imatinib. J Clin Oncol 2005;23: Comparison of two doses of imatinib for the treatment of unresectable 5357–64. or metastatic gastrointestinal stromal tumors: a meta-analysis of 1,640 14. Biron P, Cassier PA, Fumagalli E, Blesius M, Debiec-Rychter M, Adenis patients. J Clin Oncol 2010;28:1247–53. A, et al. Outcome of patients (pts) with PDGFRA D842V mutant 4. Blanke CD, Demetri GD, von MM, Heinrich MC, Eisenberg B, Fletcher gastrointestinal stromal tumor (GIST) treated with imatinib (IM) for JA, et al. Long-term results from a randomized phase II trial of advanced disease. J Clin Oncol 2010;28:15s. standard- versus higher-dose imatinib mesylate for patients with 15. Heinrich MC, Corless CL, Duensing A, McGreevey L, Chen CJ, Joseph unresectable or metastatic gastrointestinal stromal tumors expressing N, et al. PDGFRA activating mutations in gastrointestinal stromal KIT. J Clin Oncol 2008;26:620–5. tumors. Science 2003;299:708–10. 5. Demetri GD, Van Oosterom AT, Garrett CR, Blackstein ME, Shah MH, 16. Lewis NL, Lewis LD, Eder JP, Reddy NJ, Guo F, Pierce KJ, et al. Phase I Verweij J, et al. Efficacy and safety of sunitinib in patients with study of the safety, tolerability, and pharmacokinetics of oral CP- advanced gastrointestinal stromal tumour after failure of imatinib: a 868,596, a highly specific platelet-derived growth factor receptor randomised controlled trial. Lancet 2006;368:1329–38. tyrosine kinase inhibitor in patients with advanced cancers. J Clin 6. Heinrich MC, Owzar K, Corless CL, Hollis D, Borden EC, Fletcher CD, Oncol 2009;27:5262–9. et al. Correlation of kinase genotype and clinical outcome in the North 17. Michael M, Vlahovic G, Khamly K, Pierce KJ, Guo F, Olszanski AJ. American Intergroup Phase III Trial of imatinib mesylate for treatment of Phase Ib study of CP-868,596, a PDGFR inhibitor, combined with advanced gastrointestinal stromal tumor: CALGB 150105 Study by docetaxel with or without axitinib, a VEGFR inhibitor. Br J Cancer Cancer and Leukemia Group B and Southwest Oncology Group. J Clin 2010;103:1554–61. Oncol 2008;26:5360–7. 18. Fabian MA, Biggs WH III, Treiber DK, Atteridge CE, Azimioara MD, 7. Heinrich MC, Maki RG, Corless CL, Antonescu CR, Harlow A, Griffith D, Benedetti MG, et al. A small molecule-kinase interaction map for et al. Primary and secondary kinase genotypes correlate with the clinical kinase inhibitors. Nat Biotechnol 2005;23:329–36. biological and clinical activity of sunitinib in imatinib-resistant gastro- 19. Karaman MW, Herrgard S, Treiber DK, Gallant P, Atteridge CE, Camp- intestinal stromal tumor. J Clin Oncol 2008;26:5352–9. bell BT, et al. A quantitative analysis of kinase inhibitor selectivity. Nat 8. Heinrich MC, Corless CL, Demetri GD, Blanke CD, von Mehren M, Biotechnol 2008;26:127–32. Joensuu H, et al. Kinase mutations and imatinib response in patients 20. Taguchi T, Sonobe H, Toyonaga S, Yamasaki I, Shuin T, Takano A, with metastatic gastrointestinal stromal tumor. J Clin Oncol 2003;21: et al. Conventional and molecular cytogenetic characterization of a 4342–9. new human cell line, GIST-T1, established from gastrointestinal stro- 9. Debiec-Rychter M, Dumez H, Judson I, Wasag B, Verweij J, Brown M, mal tumor. Lab Invest 2002;82:663–5. et al. Use of c-KIT/PDGFRA mutational analysis to predict the clinical 21. Griffin JH, Leung J, Bruner RJ, Caligiuri MA, Briesewitz R. Discovery of response to imatinib in patients with advanced gastrointestinal stromal a fusion kinase in EOL-1 cells and idiopathic hypereosinophilic syn- tumours entered on phase I and II studies of the EORTC Soft Tissue drome. Proc Natl Acad Sci U S A 2003;100:7830–5. and Bone Sarcoma Group. Eur J Cancer 2004;40:689–95. 22. Ramos AH, Dutt A, Mermel C, Perner S, Cho J, Lafargue CJ, et al. 10. Debiec-Rychter M, Sciot R, Le CA, Schlemmer M, Hohenberger P, Van Amplification of chromosomal segment 4q12 in non-small cell lung Oosterom AT, et al. KIT mutations and dose selection for imatinib in cancer. Cancer Biol Ther 2009;8:2042–50. patients with advanced gastrointestinal stromal tumours. Eur J Cancer 23. Patricelli MP, Nomanbhoy TK, Wu J, Brown H, Zhou D, Zhang J, et al. In 2006;42:1093–103. situ kinase profiling reveals functionally relevant properties of native 11. Hirota S, Ohashi A, Nishida T, Isozaki K, Kinoshita K, Shinomura Y, kinases. Chem Biol 2011;18:699–710. et al. Gain-of-function mutations of platelet-derived growth factor 24. Lierman E, Michaux L, Beullens E, Pierre P, Marynen P, Cools J, receptor alpha gene in gastrointestinal stromal tumors. Gastroenter- et al. FIP1L1-PDGFRalpha D842V, a novel panresistant mutant, ology 2003;125:660–7. emerging after treatment of FIP1L1-PDGFRalpha T674I eosino- 12. Debiec-Rychter M, Cools J, Dumez H, Sciot R, Stul M, Mentens N, et al. philic leukemia with single agent sorafenib. Leukemia 2009;23: Mechanisms of resistance to imatinib mesylate in gastrointestinal 845–51.

www.aacrjournals.org Clin Cancer Res; 18(16) August 15, 2012 4383

Heinrich et al.

25. Lierman E, Folens C, Stover EH, Mentens N, Van MH, Scheers W, et al. tumor-associated PDGFRAD842V mutation. Clin Cancer Res 2008;14: Sorafenib is a potent inhibitor of FIP1L1-PDGFRalpha and the imati- 5749–58. nib-resistant FIP1L1-PDGFRalpha T674I mutant. Blood 2006;108: 30. Zhang J, Yang PL, Gray NS. Targeting cancer with small molecule 1374–6. kinase inhibitors. Nat Rev Cancer 2009;9:28–39. 26. LoRusso PM, Eder JP. Therapeutic potential of novel selective-spec- 31. Gajiwala KS, Wu JC, Christensen J, Deshmukh GD, Diehl W, DiNitto trum kinase inhibitors in oncology. Expert Opin Investig Drugs JP, et al. KIT kinase mutants show unique mechanisms of drug 2008;17:1013–28. resistance to imatinib and sunitinib in gastrointestinal stromal tumor 27. Zarghooni M, Bartels U, Lee E, Buczkowicz P, Morrison A, Huang A, patients. Proc Natl Acad Sci U S A 2009;106:1542–7. et al. Whole-genome proﬁling of pediatric diffuse intrinsic pontine 32. Foster R, Grifﬁth R, Ferrao P, Ashman L. Molecular basis of the gliomas highlights platelet-derived growth factor receptor alpha and constitutive activity and STI571 resistance of Asp816Val mutant KIT poly (ADP-ribose) polymerase as potential therapeutic targets. J Clin receptor tyrosine kinase. J Mol Graph Model 2004;23:139–52. Oncol 2010;28:1337–44. 33. Appelmann I, Liersch R, Kessler T, Mesters RM, Berdel WE. Angio- 28. Martinho O, Longatto-Filho A, Lambros MB, Martins A, Pinheiro C, genesis inhibition in cancer therapy: platelet-derived growth factor Silva A, et al. Expression, mutation and copy number analysis of (PDGF) and vascular endothelial growth factor (VEGF) and their recep- platelet-derived growth factor receptor A (PDGFRA) and its ligand tors: biological functions and role in malignancy. Recent Results PDGFA in gliomas. Br J Cancer 2009;101:973–82. Cancer Res 2010;180:51–81. 29. Dewaele B, Wasag B, Cools J, Sciot R, Prenen H, Vandenberghe P, 34. Tefferi A. Molecular drug targets in myeloproliferative neoplasms: et al. Activity of dasatinib, a dual SRC/ABL kinase inhibitor, and IPI- mutant ABL1, JAK2, MPL, KIT, PDGFRA, PDGFRB and FGFR1. J Cell 504, a heat shock protein 90 inhibitor, against gastrointestinal stromal Mol Med 2009;13:215–37.

4384 Clin Cancer Res; 18(16) August 15, 2012 Clinical Cancer Research

Crenolanib Inhibits the Drug-Resistant PDGFRA D842V Mutation Associated with Imatinib-Resistant Gastrointestinal Stromal Tumors

Michael C. Heinrich, Diana Griffith, Arin McKinley, et al.

Clin Cancer Res 2012;18:4375-4384. Published OnlineFirst June 27, 2012.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-12-0625

Cited articles This article cites 34 articles, 13 of which you can access for free at: http://clincancerres.aacrjournals.org/content/18/16/4375.full#ref-list-1

Citing articles This article has been cited by 27 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/18/16/4375.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/18/16/4375. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.