Activity of Panitumumab Alone Or with Chemotherapy in Non-Small Cell Lung Carcinoma Cell Lines Expressing Mutant Epidermal Growth Factor Receptor

Published OnlineFirst June 9, 2009; DOI: 10.1158/1535-7163.MCT-08-0978

1536

Activity of panitumumab alone or with chemotherapy in non-small cell lung carcinoma cell lines expressing mutant epidermal growth factor receptor

Daniel J. Freeman, Tammy Bush, protein kinase (pMAPK) staining in both xenografts com- Selam Ogbagabriel, Brian Belmontes, Todd Juan, pared with control. Docetaxel enhanced panitumumab Cherylene Plewa, Gwyneth Van, Carol Johnson, activity in NCI-H1650 xenografts (decreased Ki-67 and and Robert Radinsky pMAPK staining by >60%) when compared with either agent alone. Panitumumab inhibits ligand-induced EGFR Amgen Inc., Thousand Oaks, California phosphorylation, tumor growth, and markers of proliferation alone or with docetaxel in NSCLC cell lines that express clinically observed EGFR kinase domain mutations, Abstract including the small-molecule tyrosine kinase inhibitor- Epidermal growth factor receptor (EGFR) kinase domain resistant T790M mutation. [Mol Cancer Ther mutations cause hyperresponsiveness to ligand and hyper- 2009;8(6):1536–46] sensitivity to small-molecule tyrosine kinase inhibitors. However, little is known about how these mutations respond to antibodies against EGFR. We investigated the Introduction activity of panitumumab, a fully human anti-EGFR mono- The epidermal growth factor receptor (EGFR) is a trans- clonal antibody, in vitro in mutant EGFR-expressing non- membrane tyrosine kinase receptor expressed in many small cell lung carcinoma (NSCLC) cells and in vivo with different tumor types. The role of EGFR-mediated signal- chemotherapy in xenograft models. Mutant EGFR-expres- ing pathways in the progression of tumor growth and sing NSCLC cells (NCI-H1975 [L858R+T790M] and NCI- metastases has been well-established (1, 2). EGFR consists H1650 [Δ746-750]) and CHO cells were treated with of an extracellular ligand-binding domain, a transmem- panitumumab before EGF stimulation to assess the inhibi- brane domain, and an intracellular kinase domain with tion of EGFR autophosphorylation. Established tumors tyrosine autophosphorylation activity. Activation of EGFR were treated with panitumumab (25, 100, or 500 μg/ by EGF or transforming growth factor-α induces receptor mouse twice a week) alone or with docetaxel (10 or heterodimerization or homodimerization and tyrosine ki- 20 mg/kg once a week) or cisplatin (7.5 mg/kg once a nase phosphorylation of downstream signaling molecules week). Antitumor activity and levels of proliferation mar- that promote cell proliferation, survival signals, and me- kers were analyzed. Treatment of mutant EGFR-expressing tastases (2). CHO and NSCLC cells with panitumumab inhibited ligand- Panitumumab is a fully human monoclonal antibody dependent autophosphorylation. In NCI-H1975 and NCI- (mAb) that binds to EGFR with high affinity (5 × 10-11 H1650 xenografts, treatment with panitumumab alone mol/L), prevents ligand-induced activation, and arrests or with cisplatin inhibited tumor growth compared with tumor cell proliferation (3). In preclinical studies, treatment control (P < 0.0003). With panitumumab plus docetaxel, with panitumumab resulted in inhibition of tumor growth enhanced antitumor activity was seen in both xenografts and eradication of tumors in some animal models (3, 4). versus panitumumab alone. Panitumumab treatment Because panitumumab is a fully human mAb, it may have alone decreased Ki-67 and phospho- mitogen-activated greater specificity than EGFR small-molecule tyrosine kinase inhibitors (SM TKI; ref. 5), and fewer immunogenic reactions in patients compared with chimeric or humanized EGFR mAbs (3, 4, 6). In clinical studies, patients with solid Received 6/20/08; revised 2/20/09; accepted 3/25/09; published tumors, including colorectal, renal cell, and head and neck, OnlineFirst 6/9/09. and non-small cell lung carcinoma (NSCLC), the use of anti- Grant support: Amgen Inc. EGFR agents have shown encouraging efficacy and safety The costs of publication of this article were defrayed in part by the – payment of page charges. This article must therefore be hereby marked (7 12). More recently, greater efficacy has been observed advertisement in accordance with 18 U.S.C. Section 1734 solely to in patients stratified by tumor biomarkers (13, 14). indicate this fact. Mutations in the intracellular kinase domain of EGFR Note: D.J. Freeman and T. Bush contributed equally to this study. have been identified in patients with NSCLC but not All authors are employees and stock shareholders of Amgen Inc. other tumor types (15–17). Although findings from clinical Requests for reprints: Daniel J. Freeman, Department of Oncology studies suggest that patients harboring the EGFR muta- Research, Amgen Inc., One Amgen Center Drive, MS 15-2-A, Thousand Oaks, CA 91320. Phone: 805-447-0812; Fax: 805-498-8674. tions may account for the majority of observed clinical re- E-mail: [email protected] sponses to small-molecule inhibitors of EGFR (18, 19), Copyright © 2009 American Association for Cancer Research. the correlation of these mutations with survival benefit doi:10.1158/1535-7163.MCT-08-0978 is unclear (14, 20–22). Current data suggest that patients

Mol Cancer Ther 2009;8(6). June 2009

Molecular Cancer Therapeutics 1537

with EGFR mutations may also have an amplified EGFR placed with 50% of the medium containing 6 mg/mL gene (22–24), which may be correlated with a survival ben- polybrene (Sigma-Aldrich) and 50% of the medium contain- efit in patients with or without EGFR mutations. Preclinical ing 0.5 × multiplicity of infection of EGFR lentivirus. Mul- data suggest that the EGFR mutations render the receptor tiplicity of infection was calculated as the number of seeded hyperresponsive to ligand in both intensity and duration cells × 0.5 × 0.1[factor accounted for the difference in ex- of signal, resulting in increased mitogen-activated protein pression levels between the neo gene (expressed from IRES kinase (MAPK), phosphatidylinositol 3-kinase, and STAT5 EMCV virus) and the EGFR gene (expressed from the CMV signaling. This suggests that patients harboring the EGFR promoter)]. Fresh medium containing 500 μg/mL geneticin mutations may be more resistant to chemotherapy treatment (Life Technologies) was used for selection for 14 days. (25). Preclinical data also suggested that cells expressing the Investigational Antibody and Chemotherapy EGFR kinase domain mutations are hypersensitive to inhibi- Panitumumab was provided as a 20 mg/mL stock solution with SM TKIs (17–19, 25) and down-regulated by anti- tion in 50 mmol/L acetate, 100 mmol/L NaCl (pH 5.8; Am- EGFR antibodies (26, 27). gen) in 10 mL sterile vials and diluted in sterile PBS. Human In addition to the EGFR mutations found in patient IgG2 control antibody was provided as a 40 mg/mL stock samples, several human NSCLC cell lines, including solution (Amgen) and diluted in sterile PBS. Commercially NCI-H1650 and NCI-H1975, have been found to express available docetaxel and cisplatin were purchased as a stock clinically observed EGFR mutations. Both cell lines have concentration of 10 mg/mL for docetaxel in 13% ethanol/ an increased sensitivity to EGF-induced phosphorylation water (Aventis Pharmaceuticals) and 1mg/mL (aqueous in- at two COOH-terminal phosphorylation sites, Y992 and jectable; Bedford Labs) for cisplatin. Docetaxel and cisplatin Y1068, as well as an increased phosphorylation of AKT were dosed based on the weight of the mice receiving either and STAT5 (25). treatment before each dosing regimen. in vitro Our objectives were to determine the activity of EGFR Autophosphorylation Assays and Western Blot panitumumab against clinically observed EGFR mutations Analysis and the in vivo activity of panitumumab, either alone or Cells lines were seeded in 6-well plates at ∼85% confluency with docetaxel or cisplatin, in NSCLC cell lines expressing in the appropriate cell culture medium to determine the abil- mutant EGFR. These chemotherapy agents represent the ity of panitumumab to inhibit ligand-induced phosphoryla- standard of care for patients with advanced NSCLC. We tion. Cells were serum starved overnight before treatment found that panitumumab inhibits ligand-induced EGFR with panitumumab at 2, 20, 200, or 2,000 nmol/L for 3 h fol- phosphorylationinanisogenicChinesehamsterovary lowedbysubsequentEGFtreatmentat100ng/mLfor (CHO) cell system and in NSCLC cell lines expressing 30 min. NCI-H1975 and NCI-H1650 xenografts (∼300 mm3) both wild-type and clinically observed EGFR kinase domain t t mutations including the SM TKI-resistant T790M mutation. were treated at = 0 and = 72 h with 0.5 mg panitumumab Panitumumab inhibits ligand-induced EGFR phosphoryla- via an i.p. injection before a 30 min stimulation with an i.p. μ tion, tumor growth, and markers of proliferation alone or injection of 100 g EGF on day 5 (28). Total cell lysates were with docetaxel in wild-type and mutant-expressing NSCLC prepared, loaded on a 4% to 20% Tris-glycine gel, transferred μ cell lines. Our results may support further evaluation of onto 0.45 m nitrocellulose membrane, and blocked in 3% panitumumab in NSCLC patients, especially if subsets of bovine serum albumin in TBS (Invitrogen). The membranes patients more likely to benefit can be identified based on were then incubated with an anti-phospho-Y1068 EGFR or tumor molecule phenotype. an anti-total EGFR antibody (Cell Signaling Technology). Ex- pression was detected by enhanced chemiluminescence (Amersham Biosciences) and images were captured with Materials and Methods the Versa Doc 5000 imaging system (Bio-Rad). The IC50 Cell Lines and Reagents was determined using the concentration response data to Human NSCLC cell lines expressing wild-type EGFR fit a sigmoid function using nonlinear regression analysis (A549 and SK-MES-PD) or mutant EGFR (NCI-H1650 and (Prism; Graph Pad). AKT levels were measured as a loading NCI-H1975) were purchased from the American Type Cul- control using an anti-total AKT antibody (Cell Signaling ture Collection. Cells were cultured in RPMI 1640 or Ham's Technology). F-12 (Life Technologies) supplemented with 10% fetal Xenograft Models in Athymic Nude Mice bovine serum (Hyclone), 2 mmol/L L-glutamine, and All animal studies were conducted under an internal In- 0.1mmol/L nonessential amino acids (Life Technologies). stitutional Animal Care and Use Committee protocol and EGFR-expressing CHO cell lines were cultured in DMEM satisfied all Association for Assessment and Accreditation high-glucose (Life Technologies) supplemented with 10% of Laboratory Animal Care specifications. Athymic nude fetal bovine serum (Hyclone), 1× glutathione, 1× hypoxan- mice (Harlan-Sprague-Dawley; female, 5-6 weeks old) were thine, and 1× thymidine (Life Technologies). EGFR muta- housed 5 per cage, supplied with food and water ad libitum, tions, L858R, T790M, and Δ750-759 [consensus sequence and maintained under aseptic conditions in a ventilated from the EGFR deletion mutations reported by Paez et al. rack system. Human NSCLC cancer cells (5 × 106) were in- (19)] were introduced into CHO cells. CHO cells were seed- jected s.c. in the right flank area of the mice. After tumors ed into T-25 flasks; on the following day, all media were re- were allowed to establish and grow to an average size of

Mol Cancer Ther 2009;8(6). June 2009

1538 Panitumumab in Mutant EGFR-Expressing NSCLC Xenografts

∼200 mm3,micewererandomizedintoexperimental positive cells is defined as [1* (percentage of cells staining groups (n = 10), according to tumor volume, and then treat- at 1)] + [2 * (percentage of cells staining at 2)] + [3 * (percent- ment began on the following day. Tumors were measured age of cells staining at 3)] = long H score. twice a week with a digital caliper using the following equa- The group means were calculated with a n = 4 tumors per tion: [(W2 × L) / 2], where width (W) is defined as the smal- group, except in the NCI-H1975 control group where the ler of the two measurements and length (L) is defined as the n was 3 tumors. larger of the two measurements. Body weights were re- corded twice a week. Percent inhibition was calculated using the final average tumor volume measurement at the end Results of the study, [T / C × 100], where T is the treatment group Panitumumab Inhibits Ligand-Induced Autophos- and C is the control group. phorylation in Wild-type and Mutant EGFR-Expressing Cells In vitro Statistical Analysis To determine if panitumumab inhibits phosphorylation All continuous measurements were expressed as mean ± and activation in vitro, CHO cells were transduced with a SE and plotted as a function of time for xenograft data. multiplicity of infection of 0.5 with lentiviral constructs con- Statistical significance at the P < 0.05 level of observed dif- taining either wild-type or mutant EGFR sequences (L858R, ferences between growth curves was evaluated by repeated- Δ750-759, or T790M). Fluorescence-activated cell sorting measures ANOVA followed by Scheffé post hoc testing for analysis revealed that the expression level of cell-surface multiple comparisons. Nonstatistical differences compared EGFR in the ligand-inducible CHO cells was similar to that with the control are indicated (P = NS). seen in both EGFR-expressing NSCLC cell lines (data not Immunohistochemistry shown). Subconfluent cells were serum-starved, treated NCI-H1975 and NCI-H1650 xenografts were collected, with panitumumab or control IgG2 at varying concentra- fixed in buffered zinc formalin, and paraffin embedded. tions, before EGF stimulation. Treatment with panitumu- Deparaffinized tissue sections were pretreated with Antigen mab inhibited ligand-induced autophosphorylation of Retrieval Citra (BioGenex). The sections were blocked with EGFR in CHO cells that overexpressed either wild-type or CAS Block (Zymed Laboratories) and incubated with rabbit mutant EGFR in vitro (Fig. 1A). polyclonal anti-Ki-67 at a dilution of 1:2,000 (Novocastra Because the majority of EGFR mutations have been Laboratories) or anti-phospho-p44/42 MAPK at a dilution shown to occur in NSCLC patients, we wanted to determine of 1:50 (Cell Signaling). The antibody was detected with if panitumumab can inhibit ligand-induced phosphorylation a biotinylated goat anti-rabbit immunoglobulin (Vector in NSCLC cell lines that express either wild-type or mutant Laboratories) for anti-Ki-67 or with horseradish peroxi- EGFR. Similar to CHO cells, panitumumab treatment inhib- dase-conjugated goat anti-rabbit (Jackson ImmunoResearch ited EGF-induced phosphorylation in SK-MES-PD, A549, Laboratories) for anti-phospho-MAPK. Slides were NCI-H1650, and NCI-H1975 cells (Fig. 1B). However, we quenched with 3% hydrogen peroxide and then incubated noted that there was some basal level of phosphorylation with avidin-biotin complex (Vector Laboratories). Reaction in the mutant EGFR-expressing cell lines (NCI-H1650 and sites were visualized with 3,3′-diaminobenzidine (DAKO) NCI-H1975). and counterstained with hematoxylin. The IC50 for ligand-induced autophosphorylation by pa- Tumor sections stained for Ki-67 photographed in three nitumumab in CHO cells expressing wild-type or mutant widely spaced images from the peripheral viable regions EGFR (L858R, Δ750-759, or T790M) was 0.23, 0.18, 0.17, of the tumor mass using a ×10 objective on a Nikon FXA and 0.23 nmol/L, respectively. In the wild-type (A549 and upright microscope using standard bright-field illumina- SK-MES-PD) and mutant (NCI-H1975 and NCI-H1650) ex- tion. The number of Ki-67-positive cells in each field were pressing NSCLC lines, the IC50 with panitumumab was counted using the Visiomorph Integrator image analysis 0.34, 0.44, 0.92, and 0.25 nmol/L, respectively. system (Visiopharm), and the individual field areas were Panitumumab Inhibits Ligand-Induced Autophos- measured. The total positive cells and the total field areas phorylation in Mutant EGFR-Expressing NSCLC were summed, and the positive cell density was calculated. Xenografts In vivo The results were expressed as the percent of Ki-67-positive To test if panitumumab can inhibit EGFR autophosphor- cells versus the control treated group. ylation in vivo, mice with established NCI-H1650 or NCI- Tumor sections stained for phospho-MAPK were scored H1975 xenograft tumors of ∼300 mm3 were treated at t = t μ semiquantitatively for the percentage cells with cytoplasmic 0 and = 72 h with 500 g panitumumab or control IgG2. staining on a scale 0 (negative) to 3. Apoptotic or necrotic cells At day 5, based on previous work (28), mice were treated appeared to differentially take up stain and were excluded. with 100 μg EGF 30 min before tumor excision and anal- Because of uneven distribution of apparently viable and ysis. Similar to the in vitro results, panitumumab inhibited nonviable cells, it was not possible to perform automated ligand-induced EGFR autophosphorylation in NCI-H1650 morphometry. Three ×40 high-power fields were examined established tumor xenograft tissues as detected by immu- for each tumor and the percentage of cells at each intensity noblotting with anti-phospho-Y1068 EGFR and anti-EGFR was scored. A long H score was calculated by combining antibodies (Fig. 2A). Panitumumab treatment also de- the components of staining intensity and the percentage of creased basal EGFR phosphorylation when compared with

Mol Cancer Ther 2009;8(6). June 2009

Molecular Cancer Therapeutics 1539

the unstimulated control (Fig. 2A). Additionally, panitumu- Panitumumab plus Cisplatin Treatment Delayed mab treatment of NCI-H1975 tumor xenografts inhibited Tumor Growth In vivo ligand-induced phosphorylation (Fig. 2B). Interestingly, Because platinum-based chemotherapies are the standard treatment of NCI-H1975 tumors also reduced total EGFR of care in patients with NSCLC, we assessed whether pani- compared with untreated or control treated tumors (Fig. 2B). tumumab plus cisplatin could enhance antitumor efficacy. Panitumumab Treatment Alone Delayed Tumor In previous experiments, cisplatin treatment alone did not Growth In vivo in a Dose-Dependent Manner significantly inhibit tumor growth at any doses studied To determine if the in vivo inhibition of ligand-induced (2.5, 5.0, or 7.5 mg/kg) in either NCI-H1650 or NCI- phosphorylation observed could result in an inhibition of H1975 xenograft models compared with the control groups tumor growth, NCI-H1650 and NCI-H1975 NSCLC cells (data not shown). Panitumumab or IgG2 control antibody were grown as tumor xenografts. Mice were treated with was administered 500 μg twice a week for 3 weeks, with cis- 0, 25, 100, or 500 μg/mouse panitumumab or 500 μg/mouse platin treatment of 7.5 mg/kg weekly for 3 weeks the follow- IgG2 by i.p. injection twice a week for 3 to 4 weeks. In ing day. Treatment of panitumumab plus cisplatin NCI-H1650 xenografts, panitumumab treatment at 25, 100, significantly delayed tumor growth in NCI-H1650 xeno- and 500 μg significantly delayed tumor growth by 21% (P = grafts (51%, P < 0.0001versus control and 36%, P = 0.0419 0.3388), 47% (P < 0.0011), and 56% (P < 0.0001), respectively, versus cisplatin alone; Fig. 4A). In NCI-H1975 xenografts, tu- when compared with the control group (Fig. 3A). Similarly, mor growth delay was also observed (63%, P < 0.0001versus panitumumab treatment of NCI-H1975 xenografts at the control and 58%, P = 0.009 versus cisplatin alone; Fig. 4B). same doses significantly delayed tumor growth by 45%, However, the combination of panitumumab plus cisplatin 54%, and 55% (all treatments; P < 0.0001), respectively, did not enhance tumor growth delay when compared with when compared with the control group (Fig. 3B). panitumumab alone (P > 0.9009 versus panitumumab).

Figure 1. Panitumumab activity on ligand-induced EGFR phosphorylation in vitro in (A) CHO cells expressing wild-type or mutant EGFR or (B) NSCLC cell lines expressing wild-type or mutant EGFR. Cells were treated with panitumumab (or control IgG2) at varying concen- trations for 3 h followed by treatment with 100 ng/mL EGF for 30 min. Total EGFR and EGFR autophosphorylation levels were detected by Western blot analysis using anti-EGFR and anti-phospho-Y1068 EGFR antibodies, respectively. Each experiment was done in duplicate.

Mol Cancer Ther 2009;8(6). June 2009

1540 Panitumumab in Mutant EGFR-Expressing NSCLC Xenografts

Panitumumab plus Docetaxel Treatment Enhanced group in the NCI-H1650 xenograft model (Fig. 6A). Quanti- Tumor Growth Inhibition In vivo tative analysis in the NCI-H1650 tumors determined the per- The antitumor activity of panitumumab plus docetaxel cent reduction in Ki-67 and phospho-MAPK relative to the was also assessed in both mutant EGFR-expressing NSCLC control treated group was 12% and 5%, respectively, for pa- cell lines. In previous experiments, docetaxel treatment at 20 nitumumab alone and was 14% and 29%, respectively, for and 30 mg/kg statistically significantly inhibited tumor panitumumab plus docetaxel (Fig. 6A). However, treatment growth in established NCI-H1650 xenografts, by 29% (P = of NCI-H1975 xenograft tumors with panitumumab reduced 0.0067) and 43% (P < 0.0001), respectively, whereas treat- Ki-67 and phospho-MAPK immunostaining at day 18 post- ment of NCI-H1975 xenograft with 10, 20, and 30 mg/kg in- treatment compared with the control treated group, suggest- hibited tumor growth by 53% (P = 0.0215), 73.2% (P < ing that panitumumab-mediated inhibition of EGFR activity 0.0001), and 70% (P = 0.0005), respectively, compared with results in a decrease in both cellular proliferation and down- IgG control (data not shown). To determine the effect of stream kinase signaling in this model (Fig. 6B). In the NCI- treatment with panitumumab plus docetaxel versus either H1975 tumors, quantitative analysis determined the percent agent alone, docetaxel was used at a suboptimal dose in reduction in Ki-67 and phospho-MAPK relative to control the combination studies. The combination of panitumumab was 19% and 34%, respectively, for panitumumab alone (500 μg) plus docetaxel (20 mg/kg) significantly enhanced and was 35% and 53%, respectively, for panitumumab plus tumor growth inhibition in NCI-H1650 xenografts (76%, docetaxel. This correlated with the increased efficacy seen P < 0.0001versus control, P = 0.0003 versus docetaxel, and with the combination treatment versus either single agent P = 0.0135 versus panitumumab alone; Fig. 5A). Panitumu- alone in the xenograft models. mab (500 μg) plus docetaxel (10 mg/kg) also enhanced tumor growth inhibition in NCI-H1975 (87%, P < 0.0001 versus control, P < 0.0001versus docetaxel, and P = 0.0672 Discussion panitumumab alone; Fig. 5B). These data indicate that Understanding the different mechanisms of EGFR tyro- panitumumab plus docetaxel treatment significantly en- sine kinase inhibitors and mAbs may lead to insight into hanced the antitumor activity of panitumumab. Importantly, the EGFR pathway and the treatment modality likely to no negative interaction with this combination treatment was result in clinical benefit. The data here show that panitu- observed as determined by the efficacy and lack of signifi- mumab can inhibit ligand-induced EGFR autophosphory- cant weight loss (<10%; data not shown) in the animals. lation, tumor growth, and levels of proliferation markers Panitumumab Treatment Decreased Levels of in mouse models of human NSCLC expressing both wild- Proliferation Markers in Tumor Xenografts type and mutant EGFR, including the SM TKI-resistant Ligand-induced activation of the EGFR can induce cellu- T790M mutation. lar proliferation (1). To examine if panitumumab can inhibit We show that panitumumab inhibits ligand-induced markers of cellular proliferation in vivo, mice with estab- EGFR autophosphorylation in NSCLC cell lines and in iso- lished NCI-H1650 or NCI-H1975 tumor xenografts were genic CHO cells that expresses physiologically relevant le- treated for 18 days with panitumumab or docetaxel alone vels of mutant or wild-type EGFR. The difference in or the combination of the two agents at the same dose methodology may explain the differences in the activity of and schedule as the xenograft studies (Fig. 6). Fixed tissue the two EGFR inhibitors. In other studies, EGFR mutants sections from these tumors were measured for changes in were overexpressed using transient transfections (16–18, Ki-67 or anti-phospho-MAPK by immunohistochemistry. 25, 29), which resulted in nonphysiologic levels of the EGFR Inhibition of Ki-67 and phospho-MAPK with panitumumab and therefore ligand-independent phosphorylation of the alone was not significantly different from the control treated receptor. Thus, our approach was to transduce CHO cells

Figure 2. Panitumumab inhibited ligand-induced EGFR autophosphorylation in vivo in (A) NCI-H1650 or (B) NCI-H1975 xenograft tumors. Tumors were treated with 500 μg panitumumab or control IgG2 and then stimulated with 100 μg EGF. Total EGFR and EGFR autophosphorylation levels were detected by Western blot analysis using anti-EGFR and anti-phospho-Y1068 EGFR antibodies, respectively. As a loading control, AKT levels were detected with anti- total AKT antibody. Each experiment was done in duplicate.

Mol Cancer Ther 2009;8(6). June 2009

Molecular Cancer Therapeutics 1541

Figure 3. Panitumumab treatment of (A) NCI-H1650 and (B) NCI-H1975 mutant EGFR-expressing NSCLC xenografts delayed tumor growth. Mice were treated with panitumumab at 25, 100, or 500 μg/mouse i.p. twice a week once tumors reached 200 mm3. Tumor sizes were measured at prespecified time points. Mean ± SD (n = 10). Statistical analysis was done using StatView 5.0.1 using Scheffé's ANOVA.

with a more physiologically relevant number of EGFR on tributed to the mutation. Fluorescence in situ hybridization the cells and to evaluate panitumumab activity in a system analysis indicated that the mutant, but not the wild-type, that is dependent on EGF for EGFR phosphorylation. EGFR-expressing cell lines also contain an amplified EGFR Importantly, panitumumab also inhibited ligand-induced loci,1 which is consistent with what is observed in patient EGFR phosphorylation in cells expressing the T790M muta- samples (23). Amplification of EGFR has been shown to pre- tion. This mutation was identified as an acquired resistance dict response and survival in patients with NSCLC treated mutation in patients with NSCLC that initially responded to with SM TKIs to EGFR (14), which may explain the antitu- SM TKIs (16–18). Although panitumumab has yet to show mor activity observed with panitumumab in the NCI-H1975 clinical utility in patients with NSCLC, its activity against and NCI-H1650 mutant EGFR-expressing models. In con- the T790M mutation may suggest a potential role for pani- trast, amplification of EGFR has not consistently shown tumumab in either patients with de novo SM TKI resistance utility as a biomarker for anti-EGFR antibodies (24, 30). mutations, as seen in the NCI-H1975 NSCLC cell line (17), A challenge for antibody-targeted therapies is the require- or in patients who acquire a resistance mutation subsequent ment to overcome tumor penetration obstacles, such as hy- to an initial response to a SM TKI (16). drostatic microvascular pressure and interstitial fluid Although a basal level of phosphorylation was observed in the mutant versus wild-type EGFR-expressing NSCLC cells, this difference in phosphorylation may not only be at- 1 Freeman et al., unpublished data.

Mol Cancer Ther 2009;8(6). June 2009

1542 Panitumumab in Mutant EGFR-Expressing NSCLC Xenografts

pressure (31–33). Although these pressures were not directly NSCLC cells are addicted to the EGFR signaling pathway measured in our experiments, panitumumab treatment such that any perturbation of the proliferation or survival alone effectively inhibited ligand-induced EGFR autopho- signals results in cell death (36). This hypothesis is potential- sphorylation and reduced the basal level of phosphorylation ly supported by data from a large number of patients that in vivo in both NCI-H1650 and NCI-H1975 established xeno- harbor EGFR mutations that have shown clinical responses grafts. Of note, there was a measurable level of residual to SM TKIs (19, 22, 25, 29, 37) as well as responses seen in EGFR phosphorylation after treatment with panitumumab colorectal cancer patients with wild-type EGFR who have re- presumably because of the high number of receptors that sponded to antibodies to the receptor (3, 9, 24, 38, 39). A sec- may result in ligand-independent signaling (Fig. 2). ond hypothesis is that long-term panitumumab treatment Although treatment of established tumors with panitumu- results in internalization or down-regulation of EGFR ex- mab did not result in a complete inhibition of basal phos- pression, which ultimately results in a reduction in EGFR- phorylation at the day 8 time point, biweekly treatment dependent oncogenic signaling (40, 41). alone was associated with significant inhibition of tumor Although EGFR inhibitors have shown single-agent activ- growth in both xenograft models. This is significant given ity in an unselected patient population to date, in the future that the NCI-H1975 cells harbor a T790M mutation that is they are more likely to be used in a selected patient popu- completely resistant to EGFR TKIs in preclinical studies lation and in combination with chemotherapy in earlier and thought to be responsible for clinical resistance to gefi- lines of treatment (42). Currently, platinum-based che- tinib and erlotinib (16, 34, 35). This result may be explained motherapies are used as the standard of care in patients by two different hypotheses. First, the "oncogenic addic- with NSCLC, so we assessed whether panitumumab with tion" hypothesis asserts that the mutant EGFR-expressing cisplatin could enhance antitumor efficacy over either agent

Figure 4. Panitumumab plus cisplatin treatment did not enhance tumor growth delay versus panitumumab alone in xenografts of (A) NCI-H1650 and (B) NCI-H1975 NSCLC cells expressing mutant EGFR. Mice were treated with panitumumab at 25, 100, or 500 μg/mouse i.p. twice a week once tumors reached 200 mm3. Tumor sizes were measured at prespecified time points. Mean ± SD (n = 10). Statistical analysis was done using StatView 5.0.1 using Scheffé's ANOVA.

Mol Cancer Ther 2009;8(6). June 2009

Molecular Cancer Therapeutics 1543

Figure 5. Panitumumab plus docetaxel treatment enhanced tumor growth inhibition versus panitumumab alone in xenografts of (A) NCI-H1650 and (B)NCI- H1975 NSCLC cells expressing mutant EGFR. Mice were treated with panitumumab at 25, 100, or 500 μg/mousei.p.twiceaweek when tumors reached 200 mm3. Tumor sizes were measured at prespecified time points. Mean ± SD (n = 10). Statistical analysis was done using StatView 5.0.1 using Scheffé's ANOVA.

alone. Treatment of panitumumab with cisplatin did not re- from a first-line therapy trial of erlotinib and chemotherapy sult in an increase in antitumor activity over panitumumab in NSCLC suggest that adding a SM TKI EGFR inhibitor alone consistent with other anti-EGFR antibodies in similar before treating with a chemotherapy agent may negate preclinical models (43). These data are not completely sur- the effects of the chemotherapy (46), but our data showed prising for several reasons. Although platinum-based ther- additive effects with panitumumab and docetaxel, suggest- apies have activity in human patient populations, in our ing that the agents are not antagonistic when administered studies, the maximum tolerated dose of cisplatin was lower simultaneously. than the efficacious dose. Also, because the NSCLC cell lines Activity of EGFR mAbs in combination with platinum- have increased signaling through the AKT and STAT surviv- based chemotherapy has been studied in the first-line al pathways, they might be more resistant to treatment with setting of NSCLC. Clinical benefit has been shown with ce- chemotherapy agents (25). Of note, although the addition of tuximab when combined with cisplatin and vinorelbine in a cisplatin to panitumumab did not have an increased antitu- phase III study (FLEX) in NSCLC (48). Although no change mor activity versus panitumumab alone, no negative inter- in progression-free survival was detected (4.8 months for action occurred between the two agents. This observation is each arm, cetuximab plus chemotherapy or chemotherapy important given the results observed with erlotinib, which alone), overall survival benefit was statistically significant showed a negative interaction between the SM TKI and che- in favor of the cetuximab-containing arm versus chemo- motherapy in clinical trials for first-line treatment of NSCLC therapy alone (11.3 versus 10.1 months; P = 0.044). Panitu- (20, 44–47). mumab has also been evaluated in a randomized phase II Additionally, we also tested the antitumor activity of first-line NSCLC study in combination with paclitaxel and panitumumab with docetaxel, another agent approved carboplatin (49). This study failed to show an improve- for first-line treatment of NSCLC. We showed that treatment in time to progression or overall survival when ment of panitumumab plus docetaxel has enhanced anti- panitumumab was added to chemotherapy in this unse- tumor activity compared with either agent alone. Results lected patient population. Similarly, a phase III study

Mol Cancer Ther 2009;8(6). June 2009

1544 Panitumumab in Mutant EGFR-Expressing NSCLC Xenografts

Figure 6. In vivo effect of panitumumab, docetaxel, and panitumumab plus docetaxel on Ki-67 and MAPK proliferation markers in (A) NCI-H1650 and (B)NCI- H1975 xenograft tumors. Xeno- graft sections were prepared as described in Materials and Meth- ods and then incubated with either anti-Ki-67 or anti-MAPK antibodies. Bar, 100 μm.

comparing carboplatin and paclitaxel with cetuximab ed to correlate responsiveness to panitumumab to markers versus chemotherapy alone did not show a significant that identify NSCLC populations that would likely respond difference in progression-free survival (the primary end- to therapy. point of study; ref. 50). Overall, the clinical data suggest that there may be a role for EGFR mAbs in NSCLC but Disclosure ofPotential ConflictsofInterest that improved patient selection will be required. One of the major challenges for targeted therapeutics All of the authors are employees with ownership interest, Amgen, Inc. is to associate treatment with a pharmacodynamic read- out of efficacy. Our data indicated that the inhibition of Acknowledgments in vivo phosphorylation and the antitumor activity corre- We thank Mee Rhan Kim, Ph.D. (Amgen), for assistance with the writing lated with a reduction in markers of proliferation and and preparation of this article. downstream kinase signaling, confirming that panitumumab inhibits ligand-induced activation of EGFR, resulting References in an inhibition of the MAPK pathway and associated 1. Baselga J. Why the epidermal growth factor receptor? The rationale for proliferation. cancer therapy. Oncologist 2002;7:2–8. In conclusion, our studies show that panitumumab has 2. Yarden Y, Sliwkowski MX. Untangling the ErbB signalling network. Nat Rev Mol Cell Biol 2001;2:127–37. antitumor activity both alone and with docetaxel in NSCLC 3. Foon KA, Yang XD, Weiner LM, et al. Preclinical and clinical evaluations cell lines and xenografts expressing mutant EGFR. These of ABX-EGF, a fully human anti-epidermal growth factor receptor antibody. results provide evidence for the potential role of panitu- Int J Radiat Oncol Biol Phys 2004;58:984–90. mumab in the treatment of NSCLC. Future studies are need- 4. Yang XD, Jia XC, Corvalan JR, Wang P, Davis CG, Jakobovits A.

Mol Cancer Ther 2009;8(6). June 2009

Molecular Cancer Therapeutics 1545

Eradication of established tumors by a fully human monoclonal antibody treatment in colorectal cancer: a cohort study. Lancet Oncol 2005;6: to the epidermal growth factor receptor without concomitant chemo- 279–86. – therapy. Cancer Res 1999;59:1236 43. 25. Sordella R, Bell DW, Haber DA, Settleman J. Gefitinib-sensitizing 5. Pal SK, Pegram M. Epidermal growth factor receptor and signal transduc- EGFR mutations in lung cancer activate anti-apoptotic pathways. Science tion: potential targets for anti-cancer therapy. Anticancer Drugs 2005;16: 2004;305:1163–7. – 483 94. 26. Doody JF, Wang Y, Patel SN, et al. Inhibitory activity of cetuximab on 6. Chung CH, Mirakhur B, Chan E, et al. Cetuximab-induced anaphylaxis epidermal growth factor receptor mutations in non small cell lung cancers. and IgE specific for galactose-α-1,3-galactose. N Engl J Med 2008;358: Mol Cancer Ther 2007;6:2642–51. – 1109 17. 27. Perez-Torres M, Guix M, Gonzalez A, Arteaga CL. Epidermal growth 7. Baselga J, Trigo JM, Bourhis J, et al. Phase II multicenter study of the factor receptor (EGFR) antibody down-regulates mutant receptors and antiepidermal growth factor receptor monoclonal antibody cetuximab in inhibits tumors expressing EGFR mutations. J Biol Chem 2006;281: combination with platinum-based chemotherapy in patients with plati- 40183–92. num-refractory metastatic and/or recurrent squamous cell carcinoma of 28. McDorman K, Freeman DJ, Bush T, et al. ABX-EGF tumor penetration the head and neck. J Clin Oncol 2005;23:5568–77. and EGFR saturation correlate with pharmacokinteic, pharmacodynamic, 8. Burtness B, Goldwasser MA, Flood W, Mattar B, Forastiere AA. Phase and antitumor activity in an A431 xenograft model system [abstract LB- III randomized trial of cisplatin plus placebo compared with cisplatin plus 240]. Proc Am Assoc Cancer Res 2004;45. cetuximab in metastatic/recurrent head and neck cancer: an Eastern Co- 29. Mukohara T, Engelman JA, Hanna NH, et al. Differential effects of ge- operative Oncology Group study. J Clin Oncol 2005;23:8646–54. fitinib and cetuximab on non-small-cell lung cancers bearing epidermal 9. Cunningham D, Humblet Y, Siena S, et al. Cetuximab monotherapy and growth factor receptor mutations. J Natl Cancer Inst 2005;97:1185–94. cetuximab plus irinotecan in irinotecan-refractory metastatic colorectal 30. Lievre A, Bachet JB, Le Corre D, et al. KRAS mutation status is pre- cancer. N Engl J Med 2004;351:337–45. dictive of response to cetuximab therapy in colorectal cancer. Cancer Res 10. Rowinsky EK, Schwartz GH, Gollob JA, et al. Safety, pharmacokinet- 2006;66:3992–5. ics, and activity of ABX-EGF, a fully human anti-epidermal growth factor 31. Boucher Y, Jain RK. Microvascular pressure is the principal driving receptor monoclonal antibody in patients with metastatic renal cell cancer. force for interstitial hypertension in solid tumors: implications for vascular J Clin Oncol 2004;22:3003–15. collapse. Cancer Res 1992;52:5110–4. 11. Shepherd FA, Rodrigues Pereira J, Ciuleanu T, et al. Erlotinib in previ- 32. Tong RT, Boucher Y, Kozin SV, Winkler F, Hicklin DJ, Jain RK. Vascular ously treated non-small-cell lung cancer. N Engl J Med 2005;353:123–32. normalization by vascular endothelial growth factor receptor 2 blockade induces a pressure gradient across the vasculature and improves drug pen- 12. Van Cutsem E, Peeters M, Siena S, et al. An open-label, randomized, etration in tumors. Cancer Res 2004;64:3731–6. phase 3 clinical trial of panitumumab plus best supportive care versus best supportive care in patients with chemotherapy-refractory metastatic colo- 33. Gerber HP, Ferrara N. Pharmacology and pharmacodynamics of beva- rectal cancer. J Clin Oncol 2007;25:1658–64. cizumab as monotherapy or in combination with cytotoxic therapy in preclinical studies. Cancer Res 2005;65:671–80. 13. Amado RG, Wolf M, Peeters M, et al. Wild-type KRAS is required for panitumumab efficacy in patients with metastatic colorectal cancer. J Clin 34. Pao W, Ladanyi M, Miller VA. Erlotinib in lung cancer. N Engl J Med Oncol 2008;26:1626–34. 2005;353:1739–41; author reply 41. 14. Tsao MS, Sakurada A, Cutz JC, et al. Erlotinib in lung cancer—molec- 35. Kwak EL, Sordella R, Bell DW, et al. Irreversible inhibitors of the EGF ular and clinical predictors of outcome. N Engl J Med 2005;353:133–44. receptor may circumvent acquired resistance to gefitinib. Proc Natl Acad Sci U S A 2005;102:7665–70. 15. Calvo E, Rowinsky EK. Effect of epidermal growth factor receptor mutations on the response to epidermal growth factor receptor tyrosine ki- 36. Weinstein IB. Cancer. Addiction to oncogenes—the Achilles heal of nase inhibitors: target-based populations for target-based drugs. Clin cancer. Science 2002;297:63–4. Lung Cancer 2004;6 Suppl 1:S35–42. 37. Pao W, Wang TY, Riely GJ, et al. KRAS mutations and primary re- 16. Kobayashi S, Boggon TJ, Dayaram T, et al. EGFR mutation and resis- sistance of lung adenocarcinomas to gefitinib or erlotinib. PLoS Med tance of non-small-cell lung cancer to gefitinib. N Engl J Med 2005;352: 2005;2:e17. 786–92. 38. Folprecht G, Lutz MP, Schoffski P, et al. Cetuximab and irinotecan/5- 17. Pao W, Miller VA, Politi KA, et al. Acquired resistance of lung adeno- fluorouracil/folinic acid is a safe combination for the first-line treatment of carcinomas to gefitinib or erlotinib is associated with a second mutation in patients with epidermal growth factor receptor expressing metastatic co- the EGFR kinase domain. PLoS Med 2005;2:e73. lorectal carcinoma. Ann Oncol 2006;17:450–6. 18. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epi- 39. Tyagi P. Recent results and ongoing trials with panitumumab (ABX- dermal growth factor receptor underlying responsiveness of non-small-cell EGF), a fully human anti-epidermal growth factor receptor antibody, in me- – lung cancer to gefitinib. N Engl J Med 2004;350:2129–39. tastatic colorectal cancer. Clin Colorectal Cancer 2005;5:21 3. 19. Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung cancer: cor- 40. Perera RM, Narita Y, Furnari FB, et al. Treatment of human tumor xe- relation with clinical response to gefitinib therapy. Science 2004;304: nografts with monoclonal antibody 806 in combination with a prototypical 1497–500. epidermal growth factor receptor-specific antibody generates enhanced antitumor activity. Clin Cancer Res 2005;11:6390–9. 20. Eberhard DA, Johnson BE, Amler LC, et al. Mutations in the epidermal growth factor receptor and in KRAS are predictive and prognostic 41. Friedman LM, Rinon A, Schechter B, et al. Synergistic down-regulation of receptor tyrosine kinases by combinations of mAbs: implications indicators in patients with non-small-cell lung cancer treated with chemo- – therapy alone and in combination with erlotinib. J Clin Oncol 2005;23: for cancer immunotherapy. Proc Natl Acad Sci U S A 2005;102:1915 20. 5900–9. 42. Hirsch FR, Herbst RS, Olsen C, et al. Increased EGFR gene copy number detected by fluorescent in situ hybridization predicts outcome in non- 21. Mitsudomi T, Kosaka T, Endoh H, et al. Mutations of the epidermal growth factor receptor gene predict prolonged survival after gefitinib treat- small-cell lung cancer patients treated with cetuximab and chemotherapy. J Clin Oncol 2008;26:3351–7. ment in patients with non-small-cell lung cancer with postoperative recur- rence. J Clin Oncol 2005;23:2513–20. 43. Steiner P, Joynes C, Bassi R, et al. Tumor growth inhibition with cetuximab and chemotherapy in non-small cell lung cancer xenografts 22. Takano T, Ohe Y, Sakamoto H, et al. Epidermal growth factor receptor expressing wild-type and mutated epidermal growth factor receptor. Clin gene mutations and increased copy numbers predict gefitinib sensitivity in Cancer Res 2007;13:1540–51. patients with recurrent non-small-cell lung cancer. J Clin Oncol 2005;23: 6829–37. 44. Fuster LM, Sandler AB. Select clinical trials of erlotinib (OSI-774) in non-small-cell lung cancer with emphasis on phase III outcomes. Clin Lung 23. Hirsch FR, Varella-Garcia M, McCoy J, et al. Increased epidermal Cancer 2004;6 Suppl 1:S24–9. growth factor receptor gene copy number detected by fluorescence in situ hybridization associates with increased sensitivity to gefitinib in patients 45. Giaccone G, Herbst RS, Manegold C, et al. Gefitinib in combination with bronchioloalveolar carcinoma subtypes: a Southwest Oncology with gemcitabine and cisplatin in advanced non-small-cell lung cancer: a Group study. J Clin Oncol 2005;23:6838–45. phase III trial-INTACT 1. J Clin Oncol 2004;22:777–84. 24. Moroni M, Veronese S, Benvenuti S, et al. Gene copy number for epi- 46. Herbst RS, Prager D, Hermann R, et al. TRIBUTE: a phase III trial of dermal growth factor receptor (EGFR) and clinical response to anti-EGFR erlotinib hydrochloride (OSI-774) combined with carboplatin and paclitaxel

Mol Cancer Ther 2009;8(6). June 2009

1546 Panitumumab in Mutant EGFR-Expressing NSCLC Xenografts

chemotherapy in advanced non-small-cell lung cancer. J Clin Oncol 2005; 49. Crawford J, Swanson P, Prager D, et al. Panitumumab, a fully human 23:5892–9. antibody, combined with paclitaxel and carboplatin versus paclitaxel and 47. Herbst RS, Sandler AB. Overview of the current status of human epi- carboplatin alone for first line advanced non-small cell lung cancer dermal growth factor receptor inhibitors in lung cancer. Clin Lung Cancer (NSCLC): a primary analysis [abstract 1123]. Eur J Cancer Suppl 2005; 2004;6 Suppl 1:S7–19. 3:324. 48. Pirker R, Szczesna A, von Pawel J, et al. FLEX: A randomized, multi- 50. Lynch TJ, Patel T, Dreisbach L, et al. A randomized multicenter phase center, phase III study of cetuximab in combination with cisplatin/vinorel- III study of cetuximab (Erbitux®) in combination with taxane/carboplatin bine (CV) versus CV alone in the first-line treatment of patients with versus taxane/carboplatin alone as first-line treatment for patients with ad- advanced non-small cell lung cancer (NSCLC) [abstract 3]. J Clin Oncol vanced/metastatic non-small cell lung cancer (NSCLC): B3-03. J Thoracic 2008;26. Oncol 2007;2 Supplement 4:S340–1.

Mol Cancer Ther 2009;8(6). June 2009

Activity of panitumumab alone or with chemotherapy in non-small cell lung carcinoma cell lines expressing mutant epidermal growth factor receptor

Daniel J. Freeman, Tammy Bush, Selam Ogbagabriel, et al.

Mol Cancer Ther 2009;8:1536-1546. Published OnlineFirst June 9, 2009.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-08-0978

Cited articles This article cites 48 articles, 27 of which you can access for free at: http://mct.aacrjournals.org/content/8/6/1536.full#ref-list-1

Citing articles This article has been cited by 2 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/8/6/1536.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 Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/8/6/1536. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.