Alectinib Shows Potent Antitumor Activity Against RET-Rearranged Non–Small Cell Lung Cancer

Published OnlineFirst October 27, 2014; DOI: 10.1158/1535-7163.MCT-14-0274

Molecular Cancer Small Molecule Therapeutics Therapeutics

Alectinib Shows Potent Antitumor Activity against RET-Rearranged Non–Small Cell Lung Cancer

Tatsushi Kodama, Toshiyuki Tsukaguchi, Yasuko Satoh, Miyuki Yoshida, Yoshiaki Watanabe, Osamu Kondoh, and Hiroshi Sakamoto

Abstract Alectinib/CH5424802 is a known inhibitor of anaplastic lymphoma kinase (ALK) and is being evaluated in clinical trials for the treatment of ALK fusion–positive non–small cell lung cancer (NSCLC). Recently, some RET and ROS1 fusion genes have been implicated as driver oncogenes in NSCLC and have become molecular targets for antitumor agents. This study aims to explore additional target indications of alectinib by testing its ability to inhibit the activity of kinases other than ALK. We newly veriﬁed that alectinib inhibited RET kinase activity and the growth of RET fusion–positive cells by suppressing RET phosphorylation. In contrast, alectinib hardly inhibited ROS1 kinase activity unlike other ALK/ROS1 inhibitors such as crizotinib and LDK378. It also showed antitumor activity in mouse models of tumors driven by the RET fusion. In addition, alectinib showed kinase inhibitory activity against RET gatekeeper mutations (RET V804L and V804M) and blocked cell growth driven by the KIF5B-RET V804L and V804M. Our results suggest that alectinib is effective against RET fusion– positive tumors. Thus, alectinib might be a therapeutic option for patients with RET fusion–positive NSCLC. Mol Cancer Ther; 13(12); 2910–8. 2014 AACR.

Introduction kinases, and in fact, some inhibitors are approved or are Rearranged during transfection (RET) is an oncogene currently in clinical trials for several indications. For that when genetically altered, is involved in the develop- example, imatinib, which has inhibitory activity against ABL and KIT, was approved for use in the treatment of ment of several human cancers. Activation of the RET BCR-ABL point mutation is associated with medullary thyroid car- Philadelphia chromosome ( fusion)–positive cinoma (MTC; ref. 1), and RET rearrangement has been chronic myeloid leukemia and KIT-positive metastatic detected in papillary thyroid carcinoma (2). In addition, gastrointestinal stromal tumor (8). In addition, crizotinib, some RET fusion genes such as KIF5B-RET and CCDC6- which was granted accelerated approval by the FDA in RET have also recently been identified as driver onco- 2011 as the first anaplastic lymphoma kinase (ALK) inhib- genes in non–small cell lung cancer (NSCLC), with the itor for advanced ALK-positive NSCLC patients (9), inhi- associated fusion genes present in approximately 1% to bits not only ALK but several other kinases including 2% of cases (3–5). Vandetanib and cabozantinib, which are MET, ROS1, and RON (10). Clinical trials with crizotinib have recently demonstrated the drug’s efficacy in the multikinase inhibitors targeting RET, were approved by ROS1 the FDA in 2011 and 2012, respectively, for the treatment treatment of NSCLC with translocations (11–13). of patients with MTC (6). In a prospective phase II trial In contrast, off-target toxicities are generally observed that included RET fusion–positive NSCLC patients, 2 when therapeutic agents inhibit kinases that were not patients with RET fusion–positive NSCLC responded to intended as targets (14). For example, anticancer therapies cabozantinib (7). Thus, the genetic alterations of RET are involving the use of multikinase inhibitors that target considered to be promising targets for anticancer therapy. KDR are associated with the development of hypertension Most of the clinically approved small-molecule inhibi- and proteinuria. These side effects commonly disappear tors that are currently available are able to inhibit multiple upon drug withdrawal (15, 16). Alectinib is a potent ALK inhibitor that exhibits antitumor activity against cancers with ALK gene alterations Research Division, Chugai Pharmaceutical Co., Ltd., Kamakura, Japan. (17). A recent report on a phase I/II clinical study shows that alectinib is well tolerated and highly active in patients Note: Supplementary data for this article are available at Molecular Cancer ALK Therapeutics Online (http://mct.aacrjournals.org/). with advanced -rearranged NSCLC (18). One ongo- ing clinical study is testing the activity of alectinib in Corresponding Author: Hiroshi Sakamoto, Research Division, Chugai Pharmaceutical Co., Ltd., 200 Kajiwara, Kamakura, Kanagawa 247- patients in whom crizotinib treatment failed (Trial regis- 8530, Japan. Phone: 81-467-47-6257; Fax: 81-467-47-6561; E-mail: tration ID: NCT01588028). In this study, we explored [email protected] indications for alectinib treatment in non–ALK-positive doi: 10.1158/1535-7163.MCT-14-0274 cancers, and provide evidence that alectinib is a potent 2014 American Association for Cancer Research. RET inhibitor for RET fusion–driven NSCLC. Alectinib

2910 Mol Cancer Ther; 13(12) December 2014

Alectinib Is a Potent Inhibitor of RET

inhibited RET kinase activity and RET fusion–driven cell pound for the indicated time. The viable cells were mea- growth by suppressing phospho-RET. Alectinib also sured by the CellTiter-Glo luminescent cell viability assay RET showed antitumor activity in mouse models of a (Promega). The IC50 values were calculated using XLfit fusion–positive tumor. software. The caspase-3/7 assay was evaluated using the Caspase-Glo 3/7 Assay Kit (Promega). Fluorescence was Materials and Methods quantified using Envision (PerkinElmer). Compounds and cell lines Alectinib was synthesized at Chugai Pharmaceutical Transcriptome profiling and qRT-PCR Co. Ltd. according to the procedure described in patent Total RNA was extracted using the RNeasy mini-kit publication WO2010143664. Crizotinib, cabozantinib, and (Qiagen), and cDNA libraries were generated using the vandetanib were purchased from Selleck Chemicals; gefi- TruSeq RNA sample preparation kit (Illumina). The tinib was purchased from Kemprotec Limited; and cDNA libraries were sequenced by an Illumina HiSeq LDK378 was purchased from Active Biochemicals. 2000 instrument according to the manufacturer’s instruc- LC-2/ad cells were obtained from RIKEN in September tions. RSEM software (19) was used to align reads against 2011, NCI-H2228 cells were obtained from American Type the RefSeq transcript and to calculate the expression value Culture Collection (ATCC) in February 2008, NCI-H522 for each gene. For qRT-PCR, RNA was amplified by cells were obtained from ATCC in July 2001, and Ba/F3 QuantiFast Multiplex RT-PCR (Qiagen) using a Universal cells were obtained from RIKEN in July 2008. Each cell line probe library (Roche Applied Science) and the LightCy- was cultured using the medium recommended by the cler System (Roche). GAPDH served as an internal suppliers. The cell lines have not been authenticated by control. the authors. Immunoblotting Cells were lysed in Cell Lysis Buffer (Cell Signaling Kinase inhibitory assays Technology) containing 1 mmol/L PMSF, 1% (v/v) phos- Recombinant human RET, mutated RET (G691S, Y791F, phate inhibitor cocktail 2 (Sigma), 1% (v/v) phosphate V804L, V804M, S891A, and M918T), ROS1, RON, EGFR, inhibitor cocktail 3 (Sigma), and Complete Mini, EDTA- KDR, PDGFRb, FGFR2, KIT, SRC, HER2, MET, RAF1, and Free (Roche). Cell lysates were subjected to SDS-PAGE, MEK1 were purchased from Carna Biosciences or Merck. and the separated proteins were electrophoretically trans- Inhibitory activity against each kinase was evaluated using ferred to Immobilon-P membranes (Millipore). After the time-resolved fluorescence resonance energy transfer blocking in Blocking One (Nacalai Tesque, Inc.), the assay to examine each compound’s ability to phosphory- membranes were incubated independently in primary late substrate peptide Biotin-EGPWLEEEEEAYGWMDF in antibody diluted with anti-RET (Santa Cruz Biotechnol- the presence of the drug and 30 mmol/L ATP (17). The IC 50 ogy, sc-167 or Cell Signaling Technology, #3341), anti- values were calculated using XLfit software (ID Business Phospho-RET (Tyr 905; Cell Signaling Technology, #3221), Solutions). anti-STAT3 (Cell Signaling Technology, #9132), anti-Phos- In silico pho-STAT3 (Tyr 705; Cell Signaling Technology, #9131), modeling anti-AKT (Cell Signaling Technology, #9277), anti-Phos- The structure models of RET with alectinib, crizotinib, pho-AKT (Ser 473; Cell Signaling Technology, #9271), and LDK378 were modeled by Discovery Studio 3.5 anti-p44/42 MAP Kinase (ERK1/2; Cell Signaling Tech- (Accelrys). All figures were drawn using PyMol software € nology, #9102), anti-Phospho-ERK1/2 (Thr 202/Tyr 204; (Schrodinger K.K.). Cell Signaling Technology, #9101), anti-Cleaved PARP (Asp214; Cell Signaling Technology, #9546), anti-BIM Generation of Ba/F3 cells expressing KIF5B-RET and (Cell Signaling Technology, #2819), anti-MCL-1 (Cell Sig- mutated KIF5B-RET naling Technology, #4572), anti-BAX (Cell Signaling Tech- The plasmids that express KIF5B-RET, KIF5B-RET nology, #2774), or anti-b-actin (Sigma, A5441). To detect V804L, and KIF5B-RET V804M, respectively, named phosphorylated RET, cell lysates were immunoprecipi- CS-GS104J-M67, CS-GS104J-M67-m1, and CS-GS104J- tated with anti-phosphotyrosine (PY-20) antibody (BD M67-m2, were purchased from Genecopoeia. To generate Biosciences). The membranes were incubated with an Ba/F3 cells that stably expressed KIF5B-RET and mutated anti-rabbit or anti-mouse IgG, horseradish peroxidase- KIF5B-RET, Ba/F3 cells were transfected with the appro- linked antibody (Cell Signaling Technology). The bands priate expression plasmids using a Nucleofector device were detected using Chemi-Lumi One Super (Nacalai (Amaxa). Stable transfectants were then isolated from the Tesque, Inc.) technology with LAS-4000 (Fujifilm). cultured medium without IL3. In vivo studies Cell growth inhibition and caspase-3/7 assay Male SCID mice (C.B-17/Icr-scid/scidJc, 5-week-old) Cells were cultured in 96-well spheroid plates (Sumilon were obtained from CLEA Japan, Inc. Cells (LC-2/ad cells Celltight Spheroid 96U; Sumitomo Bakelite Inc.) over- at 8.4 106 or Ba/F3 cells expressing KIF5B-RET at night and incubated with various concentrations of com- 5.0 106) were grown as subcutaneous tumors in SCID

www.aacrjournals.org Mol Cancer Ther; 13(12) December 2014 2911

Kodama et al.

mice. Mice were randomized to treatment groups to three kinases: ALK, LTK, and GAK (17). In this study, we receive vehicle or alectinib (oral, qd) for the indicated found that in addition to ALK and LTK, alectinib at 100 duration of treatment. The final concentration of vehicle nmol/L bound to CHEK2, FLT3 (D835Y), PHKG2, RET, was 0.02N HCl, 10% DMSO, 10% Cremophor EL, 15% and RET (M918T) with greater than 95% inhibition (data PEG400, and 15% HPCD (2-hydroxypropyl-b-cyclodex- not shown). Because several RET fusion genes have trin). The length (L)andwidth(W)ofthetumormass recently been identified as driver oncogenes in NSCLC were measured, and tumor volume (TV) was calculated (3–5), we focused on the potential of alectinib as a RET 2 as: TV ¼ (L W )/2.Therateofchangeinbodyweight inhibitor. The results of a kinase inhibitory assay revealed (BW) was calculated using the following formula: BW ¼ that alectinib strongly inhibited RET kinase activity W W W W ¼ / 0 100, where and 0 are the body weight on a (IC50 4.8 nmol/L; Table 1) and RET kinases with point ¼ specific experimental day and on the first day of treat- mutations that were identified in MTC (IC50 5.7 53 ment, respectively. All animal experiments in this study nmol/L, Table 1; refs. 20–22). In contrast, crizotinib hardly ¼ were performed in accordance with protocols approved inhibited RET (IC50 3,200 nmol/L) and LDK378, a sec- by the Institutional Animal Care and Use Committee of ond-generation ALK inhibitor (23, 24), did not inhibit RET > Chugai Pharmaceutical Co., Ltd. kinase activity (IC50, 5,000 nmol/L). In addition, ATP- competitive binding assay showed that alectinib bound to Results RET at a dissociation constant (KD) value of 7.6 nmol/L. Inhibition of RET kinase activity mediated by The kinase domain of RET shares 36.6% amino acid alectinib sequence homology with ALK in the kinase domains, and The ALK inhibitor crizotinib inhibits several kinases astructure-basedin silico analysis revealed that common including ALK, MET, ROS1, and RON (10). Several ROS1 hydrophobic interactions of alectinib bound to ALK (25) fusion genes have recently been identified in NSCLC and RET (Fig. 1B). In addition, the differences among ALK (4, 11, 12), and crizotinib has recently shown efficacy in inhibitors show that alectinib has a unique chemical scaf- the treatment of NSCLC with ROS1 translocations (13). fold (it is a benzo[b]carbazole derivative) that is unlike the Having previously reported that alectinib potently inhib- scaffold used in ALK inhibitors of other chemical classes, ¼ in ited ALK (IC50 1.9 nmol/L) but not MET kinase (IC50, such as crizotinib and LDK378 (17, 23, 26, 27). Moreover, >5,000 nmol/L; ref. 17), we further checked the inhibitory silico analysis suggested that RET would cause steric hin- activity of alectinib for ROS1 and RON. However, alecti- drance, thereby interfering with the action of crizotinib and ¼ nib hardly inhibited ROS1 (IC50 3,700 nmol/L) and did LDK378 but not with that of alectinib (Fig. 1C). > not inhibit RON (IC50, 5,000 nmol/L; Fig. 1A; Table 1). To examine the additional kinase inhibitory activity of Alectinib mediated inhibition of RET fusion–positive alectinib, we evaluated 451 biochemical kinases, includ- cell growth ing several mutated kinases, using KINOMEscan technol- Next, to investigate the sensitivity of RET fusion–positive ogy (DiscoveRx). Using this technology, we had previ- cells to alectinib, we evaluated alectinib’s capacity to inhibit ously found that alectinib at 10 nmol/L bound to only cell growth in LC-2/ad NSCLC cells harboring the CCDC6-RET fusion gene (28). Alectinib, similarly to well-known RET inhibitors cabozantinib and vandetanib, Table 1. Kinase inhibitory activity of alectinib inhibited the growth of LC-2/ad cells (Fig. 2A). However, against RET, ROS1, and RON other ALK inhibitors (crizotinib and LDK378) and the EGFR inhibitor gefitinib did not inhibit cell growth. To investigate the sensitivity of another line of RET fusion– Kinase IC50 (nmol/L) positive cells to alectinib, we generated Ba/F3 cells expres- ALK 1.9 sing KIF5B-RET. An established Ba/F3 transfectant expres- RET 4.8 sing KIF5B-RET shows IL3 independent growth and that ROS1 3,700 the cells depend on KIF5B-RET. Alectinib, cabozantinib, > RON 5,000 and vandetanib were also effective against Ba/F3 cells RET G691S 9.5 expressing KIF5B-RET (Fig. 2B). From public information, RET Y791F 14 ponatinib also inhibited Ba/F3 cells expressing KIF5B-RET RET V804L 32 ¼ (IC50 11 nmol/L; ref. 29), consistent with the potency of RET V804M 53 enzyme inhibition on RET (Supplementary Fig. S1). In RET S891A 8.3 contrast, gefitinib and crizotinib did not inhibit these cells RET M918T 5.7 and LDK378 hardly inhibited these cells (Fig. 2B and NOTE: The in vitro kinase inhibitory assays of purified native Supplementary Fig. S2). Moreover, alectinib induced cas- RET and mutated RET (amino acids 658–1114), ROS1 (amino pase-3/7 activation in LC-2/ad cells as well as in NCI- acids 1883–2347), and RON (amino acids 979–1400) fused H2228 cells harboring EML4-ALK (Fig. 2C), indicating that to GST in the presence of alectinib were carried out as apoptosis is involved in the antitumor activity of alectinib. described in Materials and Methods. To understand the effects of phospho-RET suppression and the contribution of factors downstream to RET, we

2912 Mol Cancer Ther; 13(12) December 2014 Molecular Cancer Therapeutics

Alectinib Is a Potent Inhibitor of RET

O N A N N N

B Alectinib - ALK Alectinib - RET

Figure 1. Structural model of K1150 K758 alectinib with RET. A, chemical E1167 E775 structure of alectinib. B, X-ray M1199 A807 structure of alectinib with ALK at D892 the ATP-binding site and 3D model D1270 of alectinib with RET. Amino acid G1269 S891 residues, which form hydrogen bonds to the ligand directly or via R1253 R878 water molecules, are depicted. Alectinib and amino acid residues are shown in stick form (C in yellow, C Alectinib - ALK Crizotinib - ALK LDK378 - ALK white, and green; O in red; and N in blue). C, X-ray structure of alectinib with ALK (PDB ID: 3AOX), crizotinib with ALK (PDB ID: 2XP2), and LDK378 with ALK (PDB ID: 4MKC) and structural models of RET (PDB ID: 2K2K) with them. Alectinib (C in yellow), crizotinib (C in pink), and LDK378 (C in aqua) are shown in stick form. Alectinib - RET Crizotinib - RET LDK378 - RET

conducted a cellular phosphorylation assay using LC-2/ad that certain MAPK-related genes, such as DUSP6 and cells that had been treated with alectinib. Alectinib sup- EREG, are suppressed by a knockdown of RET using pressed the autophosphorylation of RET in a concentra- siRNA in LC-2/ad cells (34). To validate our data, we tion-dependent manner (Fig. 2D). Alectinib also sup- conducted qRT-PCR and conﬁrmed signiﬁcant decreases pressed the phosphorylation of ERK1/2, but not of STAT3 in the expression of these genes (e.g., DUSP2, EREG, and AKT (Fig. 2D), suggesting that the proliferation of LC- ETV5, and FOS; Fig. 3A). We also examined the effects 2/ad cells requires the MAPK signaling pathways. Taken of alectinib on the expression of apoptosis-related pro- together, these results indicate that alectinib exhibits potent teins in LC-2/ad cells. Alectinib induced PARP cleavage inhibitory activity against RET fusion–positive cells by in LC-2/ad cells and increased the abundance of BIM, a suppressing RET phosphorylation. key proapoptotic member of the BCL-2 family of proteins, whereas the levels of BCL-2 family members MCL-1 and Downstream signaling pathway in LC-2/ad cells BAX remained unaffected (Fig. 3B), suggesting that alec- harboring CCDC6-RET tinib induces apoptosis through the upregulation of BIM To understand the downstream signal pathway of via inhibition of the MAPK signaling pathway. In Ba/F3 CCDC6-RET in NSCLC, we performed comprehensive cells expressing KIF5B-RET, alectinib also suppressed gene expression analysis of alectinib-treated LC-2/ad phosphorylation of ERK and increased the abundance of cells based on Illumina HiSeq 2000 next-generation BIM (Supplementary Fig. S3). These data were consistent sequencing. The majority of genes downregulated by with a previous report that inhibition of the MAPK sig- alectinib were negative feedback regulators of ERK such naling pathway contributes to EGFR inhibitor-induced as DUSP6 (30) and DUSP2 (31) and MAPK pathway BIM upregulation in EGFR-mutated NSCLC cells (35). downstream genes such as ETV1, ETV4, ETV5, FOS However, we recognize that the full downstream signal (32), and EREG (33). A previous report has demonstrated pathway of RET fusion protein in NSCLC, including that

www.aacrjournals.org Mol Cancer Ther; 13(12) December 2014 2913

Kodama et al.

A B LC-2/ad (CCDC6-RET) Ba/F3 KIF5B-RET 120 120

100 100

80 80

Alectinib 60 60 Alectinib Cabozantinib Cabozantinib Vandetanib 40 Cell viability (%) 40 Cell viability (%) Crizotinib Vandetanib LDK378 20 20 Gefitinib Gefitinib

0 0 1 10 100 1,000 1 10 100 1,000 Drug conc. (nmol/L) Drug conc. (nmol/L) C D Alectinib 5 (nmol/L) 0 100 1,000 pRET

4 Alectinib RET 00 nM nmol/L pSTAT3 3 300300 nM nmol/L STAT3

2 pAKT

AKT 1 Caspase-3/7 activity (fold) pERK1/2

0 ERK1/2 LC-2/ad NCI-H2228 NCI-H522

Figure 2. Inhibition of RET-driven tumors by alectinib. A, growth inhibition of LC-2/ad cells by alectinib. Cells were seeded in a spheroid culture plate and incubated overnight, and then treated with the indicated concentrations of alectinib, cabozantinib, vandetanib, crizotinib, LDK378, and geﬁtinib. The viable cells were measured using a CellTiter-Glo luminescent cell viability assay after 5 days of treatment. Data are shown as mean SD (n ¼ 3per group). B, effect of alectinib against Ba/F3 cells expressing KIF5B-RET. Cells were seeded and then treated with various concentrations of alectinib, cabozantinib, vandetanib, and geﬁtinib for 2 days. The viable cells were measured using the CellTiter-Glo luminescent cell viability assay. Data are shown as mean SD (n ¼ 3). C, alectinib-mediated increase in caspase-3/7 activity. Cells were seeded in spheroid culture plate and incubated overnight, then treated with alectinib at the indicated concentrations. The viable cells were measured using a Caspase-Glo 3/7 assay kit after 2 days of treatment. Data are shown as mean SD (n ¼ 3 per group). D, effects of alectinib on the phosphorylation of RET, STAT3, AKT, and ERK1/2 in LC-2/ad cells. LC-2/ad cells were treated with alectinib for 2 hours at the indicated concentrations. Cell lysates were subjected to immunoprecipitation with anti- phosphorylated tyrosine antibodies, and the level of RET phosphorylation was detected by immunoblot analysis using anti-RET antibody. The expression level of RET, phospho-STAT3 (Tyr 705), STAT3, phospho-AKT (Ser 473), AKT, phospho-ERK1/2 (Thr 202/Tyr 204), and ERK1/2 were detected by immunoblot analysis with the appropriate antibodies.

in LC-2/ad cells, remains unknown. Further detailed xenograft model of LC-2/ad cells expressing CCDC6- studies are needed to elucidate the downstream signal RET. Although the engraftment of LC-2/ad cells using pathway of RET fusion protein in NSCLC to explore SCID mice was slow, RET break-apart FISH analysis options for combination therapy. using the RET Split Dual Color FISH Probe (ver. 1 SP018, GSP Lab., Inc.; ref. 3) showed that LC-2/ad xenograft Antitumor activity of alectinib against RET fusion– tumors continued to harbor the RET rearrangement up positive tumors to 141 days after the inoculation (data not shown). Using To evaluate the in vivo antitumor activity of alectinib this xenograft model, we conﬁrmed that once-daily oral against RET fusion–positive tumors, we used a mouse administration of alectinib resulted in remarkable

2914 Mol Cancer Ther; 13(12) December 2014 Molecular Cancer Therapeutics

Alectinib Is a Potent Inhibitor of RET

A DUSP2 EREG B 1.2 1.2 Figure 3. Downstream signaling 1 1 pathway in LC-2/ad cells harboring Alectinib CCDC6-RET 0.8 0.8 . A, the levels of each (nmol/L) 0 100 1,000 transcript were measured by 0.6 0.6 Cleaved qRT-PCR. Data are shown as mean 0.4 0.4 SD (n ¼ 2). Student t test: * PARP

(EREG/GAPDH) 0.2 0.2 * (DUSP2/GAPDH) , P < 0.05, versus DMSO Relative expression Relative expression 0 treatment. B, effects of alectinib on 0 BIM DMSO Alectinib the expression levels of cleaved DMSO Alectinib PARP, BIM, MCL-1, and BAX in MCL-1 LC-2/ad cells. LC-2/ad cells were ETV5 FOS treated with alectinib for 24 hours 1.2 BAX at the indicated concentrations. 1.2 The expression level of cleaved 1 1 PARP, BIM, MCL-1, BAX, and 0.8 0.8 β-Actin b -actin were detected by 0.6 0.6 immunoblot analysis using the 0.4 * 0.4 * appropriate antibodies.

0.2 (FOS/GAPDH) 0.2 (ETV5/GAPDH) Relative expression

Relative expression 0 0 DMSO Alectinib DMSO Alectinib

tumor regression at all doses without significant weight (6.5- to 8.1-fold) was lower than that of cabozantinib (15- to loss (Fig. 4A). Similar efficacy was also observed in the 25-fold) and vandetanib (50- to 57-fold; Fig. 5A). The case of treatment with vandetanib at 50 mg/kg, but degree of cell sensitivity to these drugs was consistent significant body weight loss was seen at this dose level with the suppression of phospho-RET (Fig. 5B). (Fig. 4A). To understand the difference in the inhibitory effects We next tested a subcutaneous mouse model of Ba/F3 on RET V804L and V804M as exhibited by alectinib cells expressing KIF5B-RET. We found that treatment versus vandetanib, we examined an in silico model with alectinib at 60 mg/kg significantly inhibited the based on the crystal structures of alectinib with ALK growth of KIF5B-RET–driven tumors (Fig. 4B). In a phos- (PDB ID: 3AOX). This model was then superimposed on phorylation assay of the Ba/F3 cells expressing KIF5B- the crystal structure of RET bound to vandetanib. Both RET tumors, alectinib suppressed autophosphorylation of V804L and V804M mutation would cause steric hin- RET in vivo (Fig. 4C). These in vitro and in vivo findings drance that interfered with the action of vandetanib but confirmed that alectinib is effective against RET fusion– not with that of alectinib (Supplementary Fig. S4). These positive tumors. results indicated potential antitumor activity of alectinib against tumors harboring gatekeeper-mutated RET Alectinib is active in cancers driven by RET fusion genes. gatekeeper mutations Point mutations in the kinase domain are known as a mechanism of acquired resistance to small-molecule Discussion kinase inhibitors. In particular, gatekeeper mutations, Discovery of the RET fusion gene in NSCLC has led to such as T790M in EGFR, T315I in ABL, and L1196M in the development of RET inhibitors for patients with ALK, are one of the most frequent causes of resistance (36– NSCLC with RET fusion–positive tumors. Clinical trials 38). To evaluate the inhibitory effect of alectinib on the have been designed to investigate the therapeutic effects most predictable resistant RET mutations (V804L and of multikinase inhibitors targeting RET, such as vandeta- V804M), which are known as gatekeeper mutations nib (Trial registration ID: NCT01823068), cabozantinib (20), we measured the kinase inhibitory activity using (NCT01639508), sunitinib (NCT01829217), lenvatinib recombinant glutathione S-transferase (GST)-fused RET (NCT01877083), dovitinib (NCT01831726), and ponatinib V804L and V804M. Alectinib had substantial inhibitory (NCT01813734). In a recent study of cabozantinib in 3 potency against both RET gatekeeper mutants and the patients with NSCLC with the RET fusion gene, cabozan- IC50 values on RET V804L and V804M were 32 and tinib achieved partial responses in 2 patients and stable 53 nmol/L, respectively (Table 1). Next, to investigate the disease in 1 patient (7). Of those patients, grade 3 toxicities sensitivity of these mutant-driven cells to alectinib, we of fatigue, proteinuria, and hypertension were observed, generated Ba/F3 cells expressing KIF5B-RET V804L and and 2 patients treated with cabozantinib required dose V804M. Alectinib was effective against each mutant-driv- reduction of cabozantinib. To understand the differences en cell, and the IC50 ratio of alectinib against the mutated between alectinib and other RET inhibitors, we confirmed KIF5B-RET as compared with the nonmutated KIF5B-RET the associated kinase selectivity profiles. Cabozantinib,

www.aacrjournals.org Mol Cancer Ther; 13(12) December 2014 2915

Kodama et al.

A 400 B 1,200 Vehicle )

) 1,000 300 3 3 Vehicle 800

200 600 *** Alectinib 400 Tumor volume (mm Tumor

Tumor volume (mm Tumor 20 mg/kg *** 100 Alectinib Vandetanib *** 200 60 mg/kg Alectinib 50 mg/kg 60 mg/kg *** 0 0 10 12 14 16 18 20 116 118 120 122 124 126 128 130 Days after tumor implantation Days after tumor implantation C 120 Vehicle Alectinib 110 N.S. 100 pRET N.S. 90

Body weight * RET

change rate (%) 80 116 118 120 122 124 126 128 130 Days after tumor implantation

Figure 4. Antitumor activity of alectinib against RET fusion–driven tumors. A, mice bearing LC-2/ad cells were administered alectinib or vandetanib orally at the indicated doses, once daily for 14 days. Tumor volume and changes in body weight were measured for each dose group. Data are shown as mean SD (n ¼ 5 per group). Parametric Dunnett test: , P < 0.001; , P < 0.05; N.S., not significant, versus treatment with vehicle of alectinib on the final day of the experiment. B, mice bearing Ba/F3 cells expressing KIF5B-RET were treated with vehicle, 60 mg/kg of alectinib orally once daily for 10 days. Tumor volume was measured for each dose group. Data are shown as mean SD (n ¼ 5 per group). Parametric Dunnett' test: , P < 0.001, versus vehicle treatment on the final day of the experiment. C, mice bearing Ba/F3 cells expressing KIF5B-RET were orally administered a dose of 0 (vehicle) or 60 mg/kg alectinib, and the tumors were collected and lysed at 4 hours postdosing. The expression levels of phospho-RET and RET were detected by immunoblot analysis using the appropriate antibodies.

vandetanib, sunitinib, sorafenib, ponatinib, lenvatinib, tumor growth in a mouse xenograft model of TT cells (42). and dovitinib inhibited not only RET kinase, but also Thus, the growth of TT xenograft tumors might depend other kinases such as KDR, which is a receptor for VEGF not only on RET, but also on other kinases such as KDR. (Supplementary Fig. S1; refs. 39, 40). In contrast, alectinib Further investigation is needed to conﬁrm the mechan- only slightly inhibited KDR kinase activity (Supplemen- isms underlying the effects of several RET inhibitors in tary Fig. S1). The major toxicities of KDR tyrosine kinase RET-mutated MTC xenograft models. inhibitors are associated with the development of hyper- In addition to the RET fusion gene, the ROS1 fusion tension and proteinuria and they are commonly regres- gene has been identiﬁed as an oncogenic driver that is sive on drug withdrawal (15, 16). In a clinical trial of found in approximately 1% to 2% of NSCLC (11, 12). alectinib in patients with ALK-rearranged NSCLC, alecti- Alectinib hardly inhibited ROS1 kinase activity (Table nib was generally tolerable with no dose-limiting toxici- 1) and did not inhibit the growth of HCC78 cells harboring SLC34A2-ROS1 > ties observed and KDR-related toxicities such as hyper- the fusion gene (IC50, 1,000 nmol/L; tension observed rarely (18). Therefore, potent RET inhi- refs. 11, 12, 43), although crizotinib strongly inhibited bitors with weak KDR inhibition may offer a clinical growth, and LDK378 weakly inhibited the growth of these advantage in the treatment of RET fusion–positive cells under three-dimensional (3D) spheroid culture con- NSCLC. ditions. In addition, in silico analysis suggested that ROS1 Although alectinib inhibited the growth of RET C634W would cause steric hindrance, thus interfering with the mutation-positive MTC TT cells (data not shown; ref. 41), action of alectinib but not with that of crizotinib or this effect was weak in a mouse xenograft model. Never- LDK378 (Supplementary Fig. S5). In contrast, alectinib theless, alectinib inhibited RET phosphorylation in inhibited the growth of RET fusion–driven cells but cri- these tumor samples (data not shown). In contrast, a zotinib and LDK378 did not (Fig. 2A and B). Alectinib also multikinase inhibitor, sorafenib, showed tumor growth showed antitumor activity in mouse models of RET inhibition against this mouse model and inhibited RET fusion–driven tumors by suppressing phospho-RET (Fig. phosphorylation in these tumor samples (data not 4A and B). On the basis of our studies, in contrast with shown). In a previous report, cabozantinib also inhibited other ALK inhibitors, alectinib would not be useful for the

2916 Mol Cancer Ther; 13(12) December 2014 Molecular Cancer Therapeutics

Alectinib Is a Potent Inhibitor of RET

A >1,500 nmol/L 1,500

1,250 Figure 5. Inhibition of RET gatekeeper mutants by alectinib, vandetanib, or cabozantinib. 1,000 KIF5B-RET A, effect of alectinib, vandetanib, or KIF5B-RET V804L cabozantinib against Ba/F3 cells expressing KIF5B-RET and (nmol/L) 750 KIF5B-RET V804M gatekeeper mutations (V804L and 50 V804M). Cells were seeded and IC then treated with various 500 concentrations of alectinib, vandetanib, or cabozantinib for 2 250 days. The viable cells were measured using a CellTiter-Glo luminescent cell viability assay. 0 IC values were determined by 50 Alectinib Cabozantinib Vandetanib plotting the drug concentration versus the percentage of cell B KIF5B-RET V804L KIF5B-RET V804M growth inhibition. Data are shown KIF5B-RET as mean SD (n ¼ 3). B, effects of Ale Van Cab AleVan Cab Ale CabVan alectinib on the phosphorylation (nmol/L) 0 100 1,000 1,000 1,000 0 100 1,000 1,000 01,000 100 1,000 1,000 1,000 levels of RET in Ba/F3 cells expressing KIF5B-RET and pRET gatekeeper mutations (V804L and V804M). Ba/F3 cells expressing KIF5B-RET were treated with β-Actin alectinib, vandetanib, or cabozantinib for 2 hours at the indicated concentrations. The expression levels of phospho-RET, pAKT phospho-AKT, AKT, phospho- ERK1/2, ERK1/2, and b-actin were detected by immunoblot analysis AKT using the corresponding antibodies. pERK

ERK

treatment of ROS1 fusion-positive NSCLC but would be Writing, review, and/or revision of the manuscript: T. Kodama, RET H. Sakamoto useful for the treatment of fusion–positive NSCLC. Study supervision: O. Kondoh, H. Sakamoto

Disclosure of Potential Conﬂicts of Interest No potential conﬂicts of interest were disclosed. Acknowledgments The authors thank Y. Tachibana and M. Izawa for the biologic assay and in vivo Authors' Contributions A. Nakagawa for the study. Conception and design: T. Kodama, H. Sakamoto The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked Development of methodology: Y. Watanabe advertisement Acquisition of data (provided animals, acquired and managed patients, in accordance with 18 U.S.C. Section 1734 solely to indicate provided facilities, etc.): T. Kodama, T. Tsukaguchi, Y. Satoh, M. Yoshida, this fact. Y. Watanabe Analysis and interpretation of data (e.g., statistical analysis, biostatis- Received March 31, 2014; revised September 19, 2014; accepted Sep- tics, computational analysis): T. Kodama, Y. Watanabe tember 25, 2014; published OnlineFirst October 27, 2014.

References 1. Alberti L, Carniti C, Miranda C, Roccato E, Pierotti MA. RET and NTRK1 2. Bongarzone I, Vigneri P, Mariani L, Collini P, Pilotti S, Pierotti MA. RET/ proto-oncogenes in human diseases. J Cell Physiol 2003;195:168–86. NTRK1 rearrangements in thyroid gland tumors of the papillary

www.aacrjournals.org Mol Cancer Ther; 13(12) December 2014 2917

Kodama et al.

carcinoma family: correlation with clinicopathological features. Clin 25. Wells SA Jr, Gosnell JE, Gagel RF, Moley J, Pfister D, Sosa JA, et al. Cancer Res 1998;4:223–8. Vandetanib for the treatment of patients with locally advanced or 3. Kohno T, Ichikawa H, Totoki Y, Yasuda K, Hiramoto M, Nammo T, et al. metastatic hereditary medullary thyroid cancer. J Clin Oncol 2010; KIF5B-RET fusions in lung adenocarcinoma. Nat Med 2012;18:375–7. 28:767–72. 4. Takeuchi K, Soda M, Togashi Y, Suzuki R, Sakata S, Hatano S, et al. 26. Lovly CM, Heuckmann JM, de Stanchina E, Chen H, Thomas RK, Liang RET, ROS1 and ALK fusions in lung cancer. Nat Med 2012;18:378–81. C, et al. Insights into ALK-driven cancers revealed through develop- 5. Lipson D, Capelletti M, Yelensky R, Otto G, Parker A, Jarosz M, et al. ment of novel ALK tyrosine kinase inhibitors. Cancer Res 2011;71: Identification of new ALK and RET gene fusions from colorectal and 4920–31. lung cancer biopsies. Nat Med 2012;18:382–4. 27. Friboulet L, Li N, Katayama R, Lee CC, Gainor JF, Crystal AS, et al. The 6. Hart CD, De Boer RH. Profile of cabozantinib and its potential in the ALK inhibitor ceritinib overcomes crizotinib resistance in non-small cell treatment of advanced medullary thyroid cancer. Onco Targets Ther lung cancer. Cancer Discov 2014;4:662–73. 2013;6:1–7. 28. Matsubara D, Kanai Y, Ishikawa S, Ohara S, Yoshimoto T, Saka- 7. Drilon A, Wang L, Hasanovic A, Suehara Y, Lipson D, Stephens P, et al. tani T, et al. Identification of CCDC6-RET fusion in the human Response to cabozantinib in patients with RET fusion-positive lung lung adenocarcinoma cell line, LC-2/ad. J Thorac Oncol 2012;7: adenocarcinomas. Cancer Discov 2013;3:630–5. 1872–6. 8. Guo T, Agaram NP, Wong GC, Hom G, D'Adamo D, Maki RG, et al. 29. Gozgit JM, Chen TH, Clackson T, Rivera V. Ponatinib is a highly potent Sorafenib inhibits the imatinib-resistant KITT670I gatekeeper mutation inhibitor of activated variants of RET found in MTC and NSCLC. Cancer in gastrointestinal stromal tumor. Clin Cancer Res 2007;13:4874–81. Res 2013;73 (8 Suppl):Abstract nr 2084. 9. Ou SH. Crizotinib: a novel and first-in-class multitargeted tyrosine 30. Arkell RS, Dickinson RJ, Squires M, Hayat S, Keyse SM, Cook SJ. kinase inhibitor for the treatment of anaplastic lymphoma kinase DUSP6/MKP-3 inactivates ERK1/2 but fails to bind and inactivate rearranged non-small cell lung cancer and beyond. Drug Des Devel ERK5. Cell Signal 2008;20:836–43. Ther 2011;5:471–85. 31. Chu Y, Solski PA, Khosravi-Far R, Der CJ, Kelly K. The mitogen- 10. Yasuda H, de Figueiredo-Pontes LL, Kobayashi S, Costa DB. Preclin- activated protein kinase phosphatases PAC1, MKP-1, and MKP-2 ical rationale for use of the clinically available multitargeted tyrosine have unique substrate specificities and reduced activity in vivo kinase inhibitor crizotinib in ROS1-translocated lung cancer. J Thorac toward the ERK2 sevenmaker mutation. J Biol Chem 1996;271: Oncol 2012;7:1086–90. 6497–501. 11. Bergethon K, Shaw AT, Ou SH, Katayama R, Lovly CM, McDonald NT, 32. Pratilas CA, Taylor BS, Ye Q, Viale A, Sander C, Solit DB, et al. et al. ROS1 rearrangements define a unique molecular class of lung V600EBRAF is associated with disabled feedback inhibition of RAF- cancers. J Clin Oncol 2012;30:863–70. MEK signaling and elevated transcriptional output of the pathway. 12. Davies KD, Le AT, Theodoro MF, Skokan MC, Aisner DL, Berge EM, Proc Natl Acad Sci U S A 2009;106:4519–24. et al. Identifying and targeting ROS1 gene fusions in non-small cell lung 33. Ben-Ami I, Armon L, Freimann S, Strassburger D, Ron-El R, Amster- cancer. Clin Cancer Res 2012;18:4570–9. dam A. EGF-like growth factors as LH mediators in the human corpus 13. Shaw AT, Camidge DR, Engelman JA, Solomon BJ, Kwak EL, Clark luteum. Hum Reprod 2009;24:176–84. JW, et al. Clinical activity of crizotinib in advanced non-small cell lung 34. Suzuki M, Makinoshima H, Matsumoto S, Suzuki A, Mimaki S, Mat- cancer (NSCLC) harboring ROS1 gene rearrangement. J Clin Oncol sushima K, et al. Identification of a lung adenocarcinoma cell line with 2012;30:863–70. CCDC6-RET fusion gene and the effect of RET inhibitors in vitro and in 14. Kamb A, Wee S, Lengauer C. Why is cancer drug discovery so difficult? vivo. Cancer Sci 2013;104:896–903. Nat Rev Drug Discov 2007;6:115–20. 35. Okamoto K, Okamoto I, Okamoto W, Tanaka K, Takezawa K, Kuwata 15. Gild ML, Bullock M, Robinson BG, Clifton-Bligh R. Multikinase inhibi- K, et al. Role of survivin in EGFR inhibitor-induced apoptosis in non- tors: a new option for the treatment of thyroid cancer. Nat Rev small cell lung cancers positive for EGFR mutations. Cancer Res Endocrinol 2011;7:617–24. 2010;70:10402–10. 16. Robinson ES, Matulonis UA, Ivy P, Berlin ST, Tyburski K, Penson RT, 36. Sequist LV, Waltman BA, Dias-Santagata D, Digumarthy S, Turke AB, et al. Rapid development of hypertension and proteinuria with cedir- Fidias P, et al. Genotypic and histological evolution of lung cancers anib, an oral vascular endothelial growth factor receptor inhibitor. Clin J acquiring resistance to EGFR inhibitors. Sci Transl Med 2011;3: Am Soc Nephrol 2010;5:477–83. 75ra26. 17. Sakamoto H, Tsukaguchi T, Hiroshima S, Kodama T, Kobayashi T, 37. Cortes J, Jabbour E, Kantarjian H, Yin CC, Shan J, O'Brien S, et al. Fukami TA, et al. CH5424802, a selective ALK inhibitor capable of Dynamics of BCR-ABL kinase domain mutations in chronic myeloid blocking the resistant gatekeeper mutant. Cancer Cell 2011;19:679–90. leukemia after sequential treatment with multiple tyrosine kinase 18. Seto T, Kiura K, Nishio M, Nakagawa K, Maemondo M, Inoue A, et al. inhibitors. Blood 2007;110:4005–11. CH5424802 (RO5424802) for patients with ALK-rearranged advanced 38. Lovly CM, Pao W. Escaping ALK inhibition: mechanisms of and non-small-cell lung cancer (AF-001JP study): a single-arm, open-label, strategies to overcome resistance. Sci Transl Med 2012;4:120ps2. phase 1-2 study. Lancet Oncol 2013;14:590–8. 39. Matsui J, Yamamoto Y, Funahashi Y, Tsuruoka A, Watanabe T, 19. Li B, Dewey CN. RSEM: accurate transcript quantification from RNA- Wakabayashi T, et al. E7080, a novel inhibitor that targets multiple Seq data with or without a reference genome. BMC Bioinformatics kinases, has potent antitumor activities against stem cell factor pro- 2011;12:323. ducing human small cell lung cancer H146, based on angiogenesis 20. Houvras Y. Completing the Arc: targeted inhibition of RET in medullary inhibition. Int J Cancer 2008;122:664–71. thyroid cancer. J Clin Oncol 2012;30:200–2. 40. Trudel S, Li ZH, Wei E, Wiesmann M, Chang H, Chen C, et al. CHIR-258, 21. Elisei R, Cosci B, Romei C, Bottici V, Sculli M, Lari R, et al. RET exon 11 a novel, multitargeted tyrosine kinase inhibitor for the potential treat- (G691S) polymorphism is significantly more frequent in sporadic ment of t(4;14) multiple myeloma. Blood 2005;105:2941–8. medullary thyroid carcinoma than in the general population. J Clin 41. Carlomagno F, Salvatore D, Santoro M, de Franciscis V, Quadro L, Endocrinol Metab 2004;89:3579–84. Panariello L, et al. Point mutation of the RET proto-oncogene in the TT 22. Schulte KM, Machens A, Fugazzola L, McGregor A, Diaz-Cano S, Izatt human medullary thyroid carcinoma cell line. Biochem Biophys Res L, et al. The clinical spectrum of multiple endocrine neoplasia type 2a Commun 1995;207:1022–8. caused by the rare intracellular RET mutation S891A. J Clin Endocrinol 42. Bentzien F, Zuzow M, Heald N, Gibson A, Shi Y, Goon L, et al. In vitro Metab 2010;95:E92–7. and in vivo activity of cabozantinib (XL184), an inhibitor of RET, MET, 23. Chen J, Jiang C, Wang S. LDK378: a Promising anaplastic lymphoma and VEGFR2, in a model of medullary thyroid cancer. Thyroid kinase (ALK) inhibitor. J Med Chem 2013;56:5673–4. 2013;23:1569–77. 24. Solomon B, Wilner KD, Shaw AT. Current status of targeted therapy for 43. Rikova K, Guo A, Zeng Q, Possemato A, Yu J, Haack H, et al. Global anaplastic lymphoma kinase-rearranged non-small cell lung cancer. survey of phosphotyrosine signaling identifies oncogenic kinases in Clin Pharmacol Ther 2013;95:15–23. lung cancer. Cell 2007;131:1190–203.

2918 Mol Cancer Ther; 13(12) December 2014 Molecular Cancer Therapeutics

Alectinib Shows Potent Antitumor Activity against RET-Rearranged Non −Small Cell Lung Cancer

Tatsushi Kodama, Toshiyuki Tsukaguchi, Yasuko Satoh, et al.

Mol Cancer Ther 2014;13:2910-2918. Published OnlineFirst October 27, 2014.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-14-0274

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2014/11/01/1535-7163.MCT-14-0274.DC1

Cited articles This article cites 42 articles, 17 of which you can access for free at: http://mct.aacrjournals.org/content/13/12/2910.full#ref-list-1

Citing articles This article has been cited by 16 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/13/12/2910.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://mct.aacrjournals.org/content/13/12/2910. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.