Original Study

Cetuximab Inhibits T790M-Mediated Resistance to Epidermal Growth Factor Receptor Tyrosine Kinase Inhibitor in a Lung Adenocarcinoma Patient-Derived Xenograft Mouse Model Petra Martin,1,2 Erin Stewart,1,3 Nhu-An Pham,1 Celine Mascaux,1,2 Devang Panchal,1 Ming Li,1 Lucia Kim,1 Shingo Sakashita,1,4 Dennis Wang,1 Jenna Sykes,1 Thomas Friess,5 Frances A. Shepherd,1,2 Geoffrey Liu,1,2 Ming-Sound Tsao1,3,4 Abstract The T790M (amino acid substitution at position 790 in EGFR from threonine [T] to methionine [M]) mutation in nonesmall-cell lung cancer results in resistance to epidermal growth factor receptor (EGFR)-tyrosine kinase inhibitor therapy. Using a patient-derived xenograft model containing an EGFR exon 19/T790M mutation, we showed that treatment with cetuximab induced a dramatic tumor response. EGFR ligand expression suggested a role for an autocrine feedback loop in the mutant EGFR signaling pathway. Background: The epidermal growth factor receptor (EGFR) kinase domain T790M (amino acid substitution at position 790 in EGFR from threonine [T] to methionine [M]) mutation in nonesmall-cell lung cancer (NSCLC) results in resistance to EGFR tyrosine kinase inhibitors (TKIs). We used a patient-derived tumor xenograft (PDX) model containing an EGFR exon 19 deletion/T790M mutation to assess response to the EGFR-directed antibody cetuximab. Changes in the EGFR signaling pathway and ligand expression after treatment were investigated. Methods: PDX were randomized into control and treatment arms. Pharmacodynamic studies were performed at 2 and 24 hours and at 4 days after a single administration of cetuximab, erlotinib, or dacomitinib. Changes in the EGFR signaling pathway were assessed using Western blot analysis, and baseline mRNA expression of EGFR ligands using microarray analysis. Relative changes after treatment were assessed using quantitative polymerase chain reaction. Results: The xenograft showed a dramatic response to cetuximab. A complete reduction of total EGFR and phosphorylated EGFR occurred after cetuximab treatment. The PDX had increased baseline levels of heparin-binding epidermal growth factor-like growth factor (HBEGF) compared with other PDX models with or without EGFR mutations. Amphiregulin was significantly reduced 2 hours after treatment with cetuximab. Compared with control mice, cetuximab- and EGFReTKI-treated mice had significantly reduced HB-EGF gene expression at 2 hours, however, by day 4 the level of HB-EGF expression was higher. The effect of cetuximab compared with EGFR TKI on HB-EGF gene expression levels differed significantly at 2 and 24 hours but not at 4 days. Conclusion: We showed a dramatic tumor response with cetuximab in an exon 19 deletion/ T790M EGFR mutant lung adenocarcinoma PDX model, which suggests a role for the autocrine feedback loop in the mutant EGFR signaling pathway. Further investigation using cetuximab in NSCLC with T790M mutation is warranted. Clinical Lung Cancer, Vol. -, No. -, 1-9 ª 2016 Elsevier Inc. All rights reserved. Keywords: Cancer, EGFR, Ligand, NSCLC, Signaling

Petra Martin and Erin Stewart are coprimary authors. 1

Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada 2 Department of Medicine, University of Toronto, Toronto, Ontario, Canada 3 Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada 4 Department of Laboratory Medicine and Pathobiology, University of Toronto,

1525-7304/$ - see frontmatter ª 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.cllc.2016.01.002

Toronto, Ontario, Canada 5 Department of Pharmacology, Roche Diagnostics GmbH, Mannheim, Germany Submitted: Oct 13, 2015; Revised: Jan 7, 2016; Accepted: Jan 12, 2016 Address for correspondence: Ming-Sound Tsao, MD, FRCPC, University Health Network, 200 Elizabeth St, Toronto, Ontario M5G 2C4, Canada Fax: 416-340-5517; e-mail contact: [email protected]

Clinical Lung Cancer Month 2016

-1

Effect of Cetuximab on an EGFR Exon 19 Del/T790 PDX Introduction

2

-

In the past decade, therapeutic strategies targeting the epidermal growth factor (EGF) receptor (EGFR) in nonesmall-cell lung cancer (NSCLC) with EGFR tyrosine kinase inhibitors (TKIs) have improved clinical outcomes in patients with activating EGFR mutations. Activating EGFR mutations occur in up to 10% to 24% of Caucasian patients and 30% to 50% of Asian patients with NSCLC.1-3 Mutations in EGFR can lead to the constitutive activation of the receptor and promote tumorigenesis. The most commonly observed mutations involve exon 19 deletions (5-6 amino acids) and point mutations in exon 21 (substitution of leucine with arginine at codon 858 [L858R]). Response rates of up to 70% with EGFR TKIs have been seen in NSCLC patients with EGFR-activating mutations. Therefore these mutations have been established as predictors of response to EGFR TKIs1,4-6 and EGFR TKIs are recommended in the first-line treatment of patients with advanced NSCLC who harbor EGFR-activating mutations.2-4,7 However, resistance to these agents invariably develops after a median of 10 to 16 months, most commonly the result of a secondary EGFR mutation, T790M (amino acid substitution at position 790 in EGFR from threonine [T] to methionine [M]).4,8-10 The T790M mutation is responsible for up to 50% of acquired resistance and has also been reported as a de novo mutation in up to 25% to 35% of advanced NSCLC patients before treatment.1,9 Substitution of threonine by the larger methionine impedes binding of the EGFR-TKI to the EGFR kinase domain while maintaining the catalytic activity.11 The T790M mutation causes drug resistance by increasing the affinity for adenosine triphosphate.12 Second-generation EGFR TKIs, which form a covalent bond with EGFR and are therefore considered irreversible inhibitors, have been shown to overcome T790M-mediated resistance in vitro.13,14 As a result, second-generation EGFR TKIs have been investigated in clinical trials involving patients who have had a durable response with EGFR TKIs but whose disease subsequently progressed. In the LUXLung 1 trial (afatinib vs. placebo for patients with advanced, metastatic non-small-cell lung cancer after failure of erlotinib, gefitinib, or both, and one or two lines of chemotherapy), in 585 patients who progressed after at least 12 weeks of EGFR-TKI, overall survival (OS) was not different between the 2 arms (P ¼ .74); however, progression-free survival (PFS) was 2.2 months longer in the afatinib group (P < .001).15 No information was provided on the 8 patients enrolled in the study who harbored a T790M mutation on archival tissue collected at diagnosis. No trials to date have shown improved outcomes with second-generation EGFR TKIs in patients who harbor T790M mutations compared with non-T790M.16 Thirdgeneration EGFR TKIs have shown efficacy after disease progression with an EGFR TKI in 2 studies.17,18 Sequist et al17 treated 130 patients with advanced EGFR-mutant NSCLC with rociletinib, a small-molecule, orally available, mutant-selective covalent inhibitor of commonly mutated forms of EGFR, including exon 19 deletions, L858R, and T790M, but not exon 20 insertions after disease progression after treatment with a first- or second-generation EGFR TKI. Among the 46 patients with T790M-positive tumors, 27 (59% [95% CI, 45%-73%]) patients demonstrated an overall response rate (ORR); 43 (93%) patients had disease control, and PFS was 13.1 months (95% CI, 5.4-13.1 months). 5 of the 17

Clinical Lung Cancer Month 2016

(29% [95% CI, 8 to 51]) patients whose tumors were T790Mnegative had an ORR 10 (59%) patients had disease control, and PFS 5.6 months (95% CI, 1.3-not reached). In another study, Janne et al18 investigated the EGFR inhibitor AZD9291, an oral, potent, irreversible EGFR TKI that is selective for EGFR TKIsensitizing mutations and the T790M resistance mutation in 253 patients who had disease progression after previous treatment with at least 1 EGFR TKI. The EGFR T790M mutation was detected in 138 of the 222 patients (62%) in the 5 expansion cohorts in which all patients were required to undergo a tumor biopsy after disease progression during the most recent line of therapy to test for T790M. Of the 239 evaluable patients, 123 (51%; 95% [CI, 4558) had an overall ORR and 78 of the 127 patients with T790M mutations; 95% CI, 52 to 70 (61%) had an ORR. The median PFS was 9.6 months (95% CI, 8.3-not reached) in EGFR T790Mpositive patients but only 2.8 months (95% CI, 2.1-4.3) in EGFR T790M-negative patients. Recently, Goss et al19 pooled 2 phase II studies in patients with pretreated EGFR-mutant advanced NSCLC with centrally confirmed T790M positive status who were treated with AZD9291 at 80 mg once daily until disease progression. In the 397 evaluable patients, AZD9291 showed a high ORR using independent central review (> 50%) across all subgroups. We previously showed that a primary human lung cancer (PHLC) patient-derived tumor xenograft (PDX) model that harbored a T790M mutation in conjunction with a rare EGFR exon 19 deletion (L747_T751 deletion) was resistant to erlotinib, showed only transient growth inhibition to second-generation EGFR TKIs, but responded to cetuximab alone.20 Our current study shows the molecular mechanisms underlying the observed response to cetuximab treatment in this PHLC PDX model. To understand these molecular mechanisms, we assessed the EGFR ligands and signaling pathway after cetuximab treatment in the tumor lysates.

Patients and Methods Patient Characteristics The PHLC 164 lung PDX has a de novo mutation in EGFR exon 19 deletion (L747_T751 deletion) and exon 20 T790M missense mutation.20 The tumor originated from a 63-year-old woman, an exsmoker, who underwent surgical resection for stage IIB NSCLC adenocarcinoma histology. Two years later, her disease relapsed with metastatic disease to the liver, bone, and lymph nodes and was treated with chemotherapy followed by erlotinib. She had no response and died within 1 month of starting erlotinib. The patient’s tumor and PDX tissue showed high EGFR polysomy, high EGFR protein expression, high cMET (hepatocyte growth factor receptor) polysomy but moderately low cMET protein expression.

PHLC 164 PDX Establishment The tumor model was established from a tumor fragment of a surgically resected early-stage NSCLC, using a protocol approved by the University Health Network Research Ethics Board and Animal Care Committee, as described in John et al.21 In brief, fresh tumor samples were propagated in nonobese diabetic severe combined immune deficient mice by implanting 5-mm section fragments immersed in 10% Matrigel at 4 C (BD Biosciences) and grown to a maximum diameter of 1.5 cm. The mice were housed under sterile conditions, and given autoclaved food and water ad libitum.

Petra Martin et al For the current studies, we used cryopreserved tumor from early passage donor (fifth passage) tumor fragments and expanded for experiments starting at passage 11 onward. Mouse body weight and tumor growth were assessed twice weekly. Tumor volume was calculated using the formula: length (mm)  width (mm)  width (mm)  0.52.

Cetuximab Screening Twelve mice were randomized into control and cetuximabtreated groups (6 mice per group), cetuximab 50 mg/kg was administered weekly intraperitoneally. Mice were sacrificed using CO2 asphyxiation when tumors measured 1.5 cm in diameter or after 30 days after treatment, whichever occurred first. At the end of all experiments, tumor tissues were harvested and divided for formalin fixation, or immediate snap freezing in liquid nitrogen and stored at 80 C for biomarker evaluation.

Pharmacodynamic Study Mice were implanted with treatment-naive tumors; when tumor growth reached approximately 1000 mm3 the mice were randomized into treatment groups (n ¼ 9 per group): control, cetuximab (50 mg/kg), dacomitinib (3 mg/kg), and erlotinib (50 mg/kg). Dacomitinib and erlotinib were administered via oral gavage. All 9 mice in each drug-treated group were given a single dose of their respective drugs at the same time. At 2 hours, 24 hours, and 4 days after injection, 3 mice from each group were sacrificed and their tumors were harvested. Tissue samples were stored at 80 C until further analysis.

Immunoblot Analysis Protein was extracted using tissue homogenization in radioimmunoprecipitation assay buffer (Sigma) and complete protease inhibitor cocktail tablets (Roche). Protein concentration was measured using the Pierce bicinchoninic acid assay protein quantification kit. Lysates were prepared in 1  gel loading buffer, and boiled. Western blot analyses were conducted after separation on 4% to 20% gradient pre-cast sodium dodecyl sulfate-polyacrylamide gel and transferred to nitrocellulose membranes. After protein transfer, the membranes were blocked in 5% nonfat dry milk with 0.05% Tween-tris buffered saline for 1 hour and incubated overnight at 4 C with the individual primary antibodies. After washing for 1 hour, the membranes were incubated for 1 hour with horseradish peroxidase-linked secondary antibodies. Glyceraldehyde-3phosphate dehydrogenase (GAPDH) was used as a housekeeping reference. Equal amounts of protein (20 mg total protein) were subjected to Western blot analysis and were probed with the primary antibodies, EGFR (cat#12020; Transduction Laboratories, Lexington, KY), antibodies from Cell Signaling Technologies (Danvers, MA) including phosphorylated (p)EGFR (cat#2236), protein kinase B (AKT) (cat#4685), pAKT (cat#4060), extracellular-signal-regulated kinase (ERK) (cat#4695), pERK (cat#4370), ribosomal protein S6 (S6) (cat#2217), pS6 (cat# 4858), and GAPDH (cat#3683). All primary antibodies were diluted 1:1000 in 5% bovine serum albumin solution made from wash buffer except for S6, pS6, and GAPDH, which were diluted 1:2000. Secondary antibody was diluted 1:4000 in a similar solution.

Microarray Quantification of Baseline EGFR Ligand Gene Expression Microarray (Human HT-12-V4 BeadChip, Illumina, with cDNA-mediated annealing, selection, extension and ligation assay [DASL] kit) was performed on a cohort (n ¼ 130) of baseline PHLC PDXs and reported previously.20 In brief, raw microarray data were preprocessed using Rv3.0.3 ‘lumi’ package (version 2.18) whereby log2 transformation and quantile normalization were applied. EGFR-family receptor ligand gene expressions were reported across all PDXs excluding the EGFR-mutant models, across all the EGFR-mutant models, and in model 164 alone.

Real-Time Reverse Transcription-Polymerase Chain Reaction Analysis to Assess EGFR Ligand Expression The relative mRNA-expression of EGFR ligands, EGF, transforming growth factor alpha (TGF-a), heparin-binding EGF-like growth factor (HB-EGF), amphiregulin (AR), were analyzed in tumor tissue samples from the pharmacodynamic study. Total RNA was extracted from the tumors using trizol reagent (Invitrogen, Karlsruhe, Germany) according to the manufacturer’s specifications. For each sample, reverse transcription polymerase chain reaction (RT-PCR) was performed in a volume of 10 mL containing cDNA transcripts equivalent to 50 ng RNA. RT-PCR reactions were carried out in triplicate using an ABI PRISM 7900 using SYBR Green Gene Expression Assay (Applied Biosystems). Cycling conditions were 95 C for 20 seconds, followed by 40 cycles at 95 C for 1 second and 60 C for 20 seconds. The threshold cycle (CT) value for each gene was determined using SDS software version 1.2 (Applied Biosystems). The fold change (2DDCT) was used for relative mRNA quantification and delta-CT was modeled as a linear function of time and treatment to investigate the effect of the treatments through time, and how they differed. All 3 EGFR TKIs were combined and considered as 1 single treatment group.

Results Cetuximab Screening The response of PHLC 164 to erlotinib, dacomitinib, or cetuximab have been reported previously.20 Briefly, this xenograft model failed to respond to erlotinib, with progression of tumor volume throughout the treatment course (Supplemental Figure 1 in the online version). Dacomitinib treatment resulted in transient stabilization of tumor growth for approximately 2 weeks, followed by disease progression. Treatment with cetuximab resulted in tumor mass reductions that were sustained throughout the month in which the monoclonal antibody was administered.

Effect of Individual Treatments on EGFR Signal Transduction To gain a better understanding of the underlying processes observed in the cetuximab study, compared with the previously reported response to the EGFR TKIs erlotinib and dacomitinib, we examined the phosphorylation state of EGFR and downstream signaling mediators in tumor lysates derived from the single-agent treated xenografts. Changes in phosphorylation levels of EGFR signal transduction pathway mediators were assessed in pharmacodynamics experiments at 2 hours, 24 hours, and 4 days after single administration of drugs with immunoblot analysis and compared

Clinical Lung Cancer Month 2016

-3

Effect of Cetuximab on an EGFR Exon 19 Del/T790 PDX Figure 1 Phosphorylation Status of the Epidermal Growth Factor Receptor (EGFR) Pathway During the Pharmacodynamic Study With Cetuximab Using Western Blot Analysis. Biological Triplicates Were Performed Per Time Point, in the Control (Ctrl) and Cetuximab (Cet)-Treated Arms. Cetuximab Decreased Total EGFR Receptor Which Led to Decreased EGFR Pathway Signaling

with time-matched controls. Cetuximab treatment resulted in decreased levels of phosphorylated EGFR and total EGFR at 24 hours and at 4 days (Figure 1). Total EGFR was downregulated to a greater extent than phosphorylated EGFR, which led to decreased EGFR pathway signaling with a decrease in pERK and pS6, but not pAKT. Treatment with erlotinib caused no major changes in the phosphorylation levels of the EGFR pathway mediators other than a slight decrease in pEGFR at 2 hours (Supplemental Figure 2 in the online version). Treatment with dacomitinib caused a noticeable decrease in pEGFR after 24 hours, and significant decreases in pERK and pS6 at 2 hours and 24 hours. Neither TKI caused a decrease in total EGFR protein.

EGFR Ligand Expression at Baseline The gene expression microarrays were established at baseline in the following groups: all PHLC PDXs (n ¼ 130 excluding the EGFRmutant models), all EGFR-mutant PHLC PDXs (n ¼ 5 excluding PHLC 164 PDX), and the 164 model (Figure 2). High baseline HBEGF was identified in the 164 model compared with all NSCLC xenograft models and EGFR-mutant NSCLC xenografts.

EGFR Ligand Expression After Treatment

4

-

Treatment with cetuximab caused a significant decrease in AR at 2 hours after treatment (at 2 hours P ¼ .04; Table 1, Figure 3, Supplemental Figure 3 in the online version). At 24 hours and at 4 days, AR had decreased further compared with the control; however, this decrease was not significantly different from the initial effect at 2 hours (P ¼ .10 and P ¼ .07, respectively). Whereas mice treated with cetuximab showed a further decrease in AR at 24 hours compared with 2 hours after treatment, the difference in AR between mice treated

Clinical Lung Cancer Month 2016

Figure 2 Expression of ERBB Family Receptors and Their Ligands, cDNA Mediated Annealing, Selection,Extension and Ligation Assay (DASL) Platform. Of the 3 Figures Per Shaded Area, the Top Represents the Gene Expression of All NoneSmallCell Lung Cancer (NSCLC) Patient-Derived Tumor Xenografts (PDXs), the Second Is the Gene Expression of All Epidermal Growth Factor Receptor (EGFR)-Mutant NSCLC PDXs, and the Red Diamond Is the Gene Expression of PDX 164

with EGFR-TKIs and control mice increased at 24 hours after treatment. This change in treatment effect at 24 hours differed significantly between the 2 treatment arms (P ¼ .02; Table 1, Figure 3, Supplemental Figure 3 in the online version). Cetuximab treatment caused a significant decrease in HB-EGF at 2 hours after treatment (at 2 hours P  .001, Table 1, Figure 3), however at day 4 the HB-EGF gene expression was increased compared with the control (P < .001). A similar pattern was observed in mice treated with EGFR-TKIs; however, the effect was significantly more pronounced in the cetuximab treated group at 2 hours (at 2 hours P ¼ .04), and the EGFR-TKIs showed a much larger increase between 2 hours and 24 hours, than mice treated with cetuximab, but this was of borderline significance only (P ¼ .05). Changes in EGF and TGF-a expression from baseline were not significantly different in linear modeling of treatment groups over time.

Discussion In this study we used a PDX model with a T790M mutation and a rare exon 19 L747_T751 deletion mutation, which constitutes 3.7% of exon 19 deletions,22 to assess response to cetuximab and to investigate the molecular mechanism underlying the observed response. NSCLC containing the T790M mutation has posed a particular challenge with regard to treatment, because it is responsible for half of the acquired mutations causing resistance to gefitinib and erlotinib. In addition, it has recently been recognized that de novo T790M mutations are more common than previously recognized, with rates of 25% to 35% being reported using the more sensitive

Petra Martin et al Table 1 Treatment Effect of CET and ERLO/DAC on AR, EGF, HBEGF, and TGF-a Expression P

Estimate AR CET Treatment effect (2 hours)

1.1

.04

Change in treatment effect (24 hours)

1.16

.1

Change in treatment effect (4 days)

1.39

.07

0.37

.4

ERLO/DAC Treatment effect (2 hours)

0.37

.6

0.19

.8

Treatment effect (2 hours)

0.74

.1

Change in treatment effect (24 hours)

1.52

.02

Change in treatment effect (4 days)

1.2

.06

Change in treatment effect (24 hours) Change in treatment effect (4 days) Difference between CET and ERLO/DAC

EGF CET Treatment effect (2 hours)

0.6

.4

Change in treatment effect (24 hours)

0.85

.4

Change in treatment effect (4 days)

0.93

.3

Treatment effect (2 hours)

0.09

.9

Change in treatment effect (24 hours)

0.43

.6

0.37

.7

ERLO/DAC

Change in treatment effect (4 days) Difference between CET and ERLO/DAC Treatment effect (2 hours)

0.7

.2

Change in treatment effect (24 hours)

0.43

.6

Change in treatment effect (4 days)

1.31

.1

HBEGF CET Treatment effect (2 hours)

1.63

4.00E-05

Change in treatment effect (24 hours)

0.01

1

Change in treatment effect (4 days)

1.77

9.00E-04

ERLO/DAC Treatment effect (2 hours)

1.04

.001

Change in treatment effect (24 hours)

0.81

.06

Change in treatment effect (4 days)

1.17

.008

Difference between CET and ERLO/DAC Treatment effect (2 hours)

0.59

Change in treatment effect (24 hours)

0.8

Change in treatment effect (4 days)

0.6

.04 .05 .1

TGF-a CET Treatment effect (2 hours) Change in treatment effect (24 hours) Change in treatment effect (4 days)

1.01

.08

0.57

.5

0.73

.4

0.14

.8

ERLO/DAC Treatment effect (2 hours) Change in treatment effect (24 hours) Change in treatment effect (4 days)

0.71

.3

0.51

.5

Clinical Lung Cancer Month 2016

-5

Effect of Cetuximab on an EGFR Exon 19 Del/T790 PDX Table 1 Continued P

Estimate Difference between CET and ERLO/DAC Treatment effect (2 hours)

0.87

.07

Change in treatment effect (24 hours)

0.14

.8

Change in treatment effect (4 days)

0.22

.7

Bold values are statistically significant (P < .05). Abbreviations: AR ¼ amphiregulin; CET ¼ cetuximab; DAC ¼ dacomitinib; EGF ¼ epidermal growth factor; ERLO ¼ erlotinib; HBEGF ¼ heparin-binding epidermal growth factor-like growth factor; TGF ¼ tumor growth factor.

detection method of matrix-assisted laser desorption ionization time of flight mass spectrometry.1,9 Therefore, patients who harbor a T790M mutation encompass a significant subset of NSCLC cases. Recently, third-generation EGFR-TKIs have shown promising efficacy against T790M-postivie EGFR-mutant NSCLC.17,18 Only recently, the US Food and Drug Administration (FDA) approved third-generation EGFR TKI osimertinib (AZD9291) for the treatment of patients with metastatic T790M mutation-positive NSCLC, whose disease has progressed during or after EGFR TKI therapy (http://www.fda.gov/newsevents/newsroom/pressannouncements/ ucm472525.htm). Before third-generation TKIs, it was hoped that irreversible second-generation EGFR TKIs would have the capacity to overcome T790M-mediated resistance. However, results to date have been disappointing. Perhaps this is not surprising because preclinical work showed limited activity of second-generation EGFR TKIs on their own in the treatment of T790M-positive NSCLC.23-25 Furthermore, clinical trials attempting to select for acquired mutations by including patients who had developed resistance to EGFRTKIs, have shown only modest activity.15,26 The most promising

result to date for EGFR-targeted therapy currently approved by the FDA, in the preclinical setting, was shown in xenografts that harbored the T790M mutation and a mutation in exon 21 (L858R), which responded to afatinib and cetuximab.23,27 Preclinical work in transgenic mouse lung cancer models showed that the combination of afatinib with EGFR-specific antibody cetuximab induced dramatic shrinkage of erlotinib-resistant tumors that harbored EGFRT790M and EGFRL858RþT790M mutations compared with either drug alone.23 In addition, the combination of cetuximab and afatinib was also superior to either agent alone in a separate in vivo model with mice bearing xenografts of H1975 lung adenocarcinoma cells that harbored EGFRL858RþT790M, that are resistant to erlotinib, in vitro.23 Therefore, this combination is increasingly being investigated in clinical trials.28 However, these findings are not fully consistent in PDXs established from tumor biopsies of patients whose disease progressed after a durable response to erlotinib treatment. In a study of afatinib, cetuximab, and the combination of both agents in PDX models with and without T790M mutations, only afatinib showed modest tumor growth inhibition in the T790M-positive model.24 There was no cetuximab-mediated

Figure 3 Fold Changes in Epidermal Growth Factor Receptor (EGF)-Ligand Gene Expression Over Time According to Treatment (Trt) Group (Cetuximab [CET] vs. EGF Tyrosine Kinase Inhibitors Erlotinib [ERLO] and Dacomitinib [DAC])

6

-

Abbreviations: AR ¼ amphiregulin; HBEGF ¼ heparin-binding EGF-like growth factor; TGFa ¼ tumor growth factor alpha.

Clinical Lung Cancer Month 2016

Petra Martin et al synergy with afatinib and no single-agent cetuximab activity. In contrast to the PDX study, cetuximab inhibited growth of a xenograft derived from the H1975 lung cancer cell line, which expressed L858R/T790M EGFR and is resistant to EGFR TKIs.27 Preliminary results from clinical trials in patients treated with cetuximab and afatinib who have acquired resistance to EGFR TKIs are somewhat promising. A phase Ib study combining afatinib and cetuximab in heavily pretreated patients with advanced EGFRmutant lung cancer and acquired resistance to erlotinib/gefitinib, showed a similar objective response rate (overall 29%) in T790Mpositive and T790M-negative tumors (32% vs. 25%; P ¼ .341). PFS was similar for T790M-negative and T790M-positive patients (4.6 vs. 4.8 months; P ¼ .643).28 To our best knowledge, our study is the first to show the benefit of single-agent cetuximab treatment in a T790M-positive PDX model. As such, we set out to investigate potential underlying mechanisms behind this response. In our study, levels of total EGFR decreased to a greater extent than levels of Tyr1068 phosphorylation of EGFR after cetuximab treatment suggesting that the efficacy of cetuximab was a result of receptor degradation. Downstream signaling via phosphorylated ERK was inhibited at 4 days after treatment, demonstrating its long half-life. Whereas erlotinib and dacomitinib also caused decreases in pEGFR, pERK, and pS6, they did not affect total EGFR protein expression, further suggesting that the decrease in total protein expression might explain cetuximab’s efficacy against T790M-mutant EGFR. In previous studies involving NSCLC cell lines resistant to first generation EGFR TKIs, downregulation of Erb-B2 receptor tyrosine kinase 3 (ErbB3)/phosphoinositide 3-kinase/AKT signaling was required for gefitinib to cause growth inhibition and apoptosis.29-32 In addition, cell line work using H1975 and H3255GR (cell lines that contain EGFR L858R and T790M mutations) and H820 (EGFR exon 19 [DEL 747-751,Ins S] and a T790M mutation),30 showed resistance in all 3 lines with gefitinib, but dacomitinib effectively inhibited the growth of these cell lines. Examination of the phosphorylation state of EGFR and downstream signaling molecules, showed that dacomitinib had greater inhibition than gefitinib of EGFR, ERBB3 and AKT phosphorylation. Furthermore, in HCC827(del E746_A750) Del/T790M cells, gefitinib was unable to inhibit EGFR, ERBB3, and AKT phosphorylation (except minimally at clinically unachievable doses; ie, 10 mmol/L), whereas dacomitinib inhibited the phosphorylation of these proteins starting at 10 nmol/L, showing its dramatic efficacy on growth of this cell line. In vivo investigation showed that dacomitinib was highly effective in a xenograft model with HCC827 Del/T790M-mediated acquired resistance to gefitinib, with dacomitinib treatment lasting for 95 days. In our PDX model, treatment with dacomitinib resulted in decreased pEGFR and downstream effector proteins (Supplemental Figure 2 in the online version); however, this did not translate into a sustained response to the TKI. Our findings might suggest that inhibition of the phosphorylation of EGFR and thus the signaling of the EGFR pathway might not always be the only mechanism important for shutting down the pathway, and that ligand-receptor autocrine loops might also play a major role. Autocrine/paracrine production of EGFR ligands and aberrant EGFR expression are 2 mechanisms implicated in cancer development.33,34 Ligand binding results in EGFR activation and cancer

cell proliferation. There are 7 ligand families associated with EGFREGF, AR, HB-EGF, TGF-a, epiregulin, betacellulin, and epigen.35 Cetuximab, a monoclonal immunoglobulin G1 antibody that binds specifically to EGFR, has shown ligand binding affinity to mutant EGFR.27 Cetuximab competed with labeled TGF-a in 2 EGFR activating mutant lung cancer cell lines and in a cell line containing wild type EGFR.27 Previous work on the effect of cetuximab treatment on circulating ligand (TGF-a, AR, EGF, HBEGF, betacellulin, and vascualr endothelial growth factor) changes in a number of cancer cell lines (epidermoid, colorectal cancer, glioma, breast, renal, pancreatic, prostate, and lung carcinoma cell lines), showed a rapid increase in the level of TGF-a in conditioned media after incubation for 24 hours with cetuximab, but not with small-molecules inhibitors.36 However, no upregulation of TGF-a mRNA was seen after treatment. This suggested that the increase in TGF-a did not occur as a compensatory response to treatment at the transcriptional level. In addition, AR also showed a doseelevated response in one cancer cell line after treatment. A positive feedback autocrine loop has been described after administration of an anti-EGFR monoclonal antibody 225 in A431 squamous carcinoma cells.37 This antibody binds extracellular EGFR and blocks binding of the ligand to the receptor. Treatment with the antibody decreased the resting mRNA levels of pro-TGFa, the transmembrane precursor of TGF-a, after 4 hours.37 EGFR ligands have been investigated for their prognostic and predictive significance in a number of cancers. In the National Cancer Institute of Canada BR.21 trial in which erlotinib was compared with placebo, patients in the placebo arm with higher plasma AR levels had poorer OS than those with low levels.38 TGF-a did not have prognostic significance, but higher TGF-a level predicted lack of benefit from erlotinib compared with low TGF-a.38 In a further study, pretreatment serum in 93 patients with NSCLC who received gefitinib reported that increased TGF-a and AR levels were associated with shorter survival times after gefitinib treatment.39 Similarly, another study on TGF-a and AR in 50 patients with advanced NSCLC before gefitinib treatment reported that high TGF-a level was associated with shorter survival times.40 Furthermore, increased TGF-a, HB-EGF, and AR have been associated with prognosis in a number of cancers including ovarian, endometrial, bladder, breast, gastric, colon, pancreatic and prostate cancer, renal cell carcinoma, glioblastoma, and malignant melanoma.33,41 HB-EGF expression levels have been shown to be increased in ovarian, gastric, and breast cancers, and melanoma and glioblastoma.42 In a study on cell lines transfected with small interfering RNAs dominant expression of HBEGF increased the number of apoptotic cells and suppressed the expression of EGFR and ERK. Therefore, HB-EGF plays important roles in cancer cell proliferation. We identified increased baseline HB-EGF mRNA expression in the 164 model, suggesting a receptorligand feedback loop which might play a role in driving and maintaining this tumor. HB-EGF was significantly decreased at 2 hours after treatment and was increased above control levels at 4 days. To our knowledge, no assessment of these ligands as predictive markers in patients with T790M mutations has been performed. Further investigation is required to confirm the use of EGFR ligands as predictive markers and therapeutic targets for cancer treatment. In our study, we showed significant growth inhibition with cetuximab in an exon 19 deletion/T790M mutation PDX model;

Clinical Lung Cancer Month 2016

-7

Effect of Cetuximab on an EGFR Exon 19 Del/T790 PDX however, to our knowledge, no clinical trial to date has specifically enrolled patients with T790M mutation. Rather, the focus has been on enrolling patients whose disease has progressed during treatment with an EGFR-TKI, and therefore have a greater likelihood of having this mutation. In a small trial of 18 patients with advanced NSCLC whose disease had progressed during erlotinib or gefitinib treatment and subsequently given cetuximab, no patients had a clinical response, but 5 had confirmed stable disease.43 Three patients were identified as having an activating EGFR mutation. One patient with an exon 19 deletion and previous complete response to erlotinib lasting 10 months achieved stable disease with cetuximab treatment for 6 months, and repeat biopsy of this patient’s tumor after TKI therapy had shown an acquired T790M resistance mutation. It is difficult to draw any conclusions from these trials regarding the efficacy of cetuximab because of the limited number of patients.

Conclusion We showed a sustained clinical response with cetuximab treatment in a PDX model containing T790M with exon 19 deletion mutation. This was associated with downregulation of the EGFR signaling pathway via decreased total EGFR. We also identified increased baseline HB-EGF mRNA expression in the 164 model, suggesting that an autocrine feedback loop might drive this tumor. Further investigation of single-agent cetuximab in the subgroup of patients with an activating mutation in conjunction with the T790M mutation is warranted.

Clinical Practice Points  The EGFR kinase domain T790M mutation in NSCLC results

in resistance to EGFR TKIs.  We used a PDX model containing an EGFR exon 19 deletion/

 

 

T790M mutation to assess response to the EGFR-directed antibody cetuximab. The xenograft showed a dramatic tumor response to cetuximab. A complete reduction of total EGFR and phosphorylated EGFR occurred after cetuximab treatment. The PDX had increased levels of HB-EGF mRNA expression at baseline. The results suggest a role for an autocrine feedback loop in the mutant EGFR signaling pathway. Further investigation using cetuximab in NSCLC with the T790M mutation is warranted.

Acknowledgments This work was supported by the Ontario Research Fund Research Excellence RE-03-020 Award, Canadian Cancer Society (CCS) Research Institute grants 020527 and 701595, and CCS and Rachelle Archambault Innovation grant 701637.

Disclosure The authors have stated that they have no conflicts of interest.

Supplemental Data Supplemental figures accompanying this article can be found in the online version at http://dx.doi.org/10.1016/j.cllc.2016.01.002.

8

-

Clinical Lung Cancer Month 2016

References 1. Su KY, Chen HY, Li KC, et al. Pretreatment epidermal growth factor receptor (EGFR) T790M mutation predicts shorter EGFR tyrosine kinase inhibitor response duration in patients with nonesmall-cell lung cancer. J Clin Oncol 2012; 30:433-40. 2. Lindeman NI, Cagle PT, Beasley MB, et al. Molecular testing guideline for selection of lung cancer patients for EGFR and ALK tyrosine kinase inhibitors: guideline from the College of American Pathologists, International Association for the Study of Lung Cancer, and Association for Molecular Pathology. J Thorac Oncol 2013; 8:823-59. 3. Hammerman PS, Janne PA, Johnson BE. Resistance to epidermal growth factor receptor tyrosine kinase inhibitors in nonesmall-cell lung cancer. Clin Cancer Res 2009; 15:7502-9. 4. Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med 2009; 361:947-57. 5. Mitsudomi T, Morita S, Yatabe Y, et al. Gefitinib versus cisplatin plus docetaxel in patients with nonesmall-cell lung cancer harbouring mutations of the epidermal growth factor receptor (WJTOG3405): an open label, randomised phase 3 trial. Lancet Oncol 2010; 11:121-8. 6. Jackman DM, Miller VA, Cioffredi LA, et al. Impact of epidermal growth factor receptor and KRAS mutations on clinical outcomes in previously untreated nonesmall-cell lung cancer patients: results of an online tumor registry of clinical trials. Clin Cancer Res 2009; 15:5267-73. 7. Sequist LV, Martins RG, Spigel D, et al. First-line gefitinib in patients with advanced nonesmall-cell lung cancer harboring somatic EGFR mutations. J Clin Oncol 2008; 26:2442-9. 8. Janne PA, Wang X, Socinski MA, et al. Randomized phase II trial of erlotinib alone or with carboplatin and paclitaxel in patients who were never or light former smokers with advanced lung adenocarcinoma: CALGB 30406 trial. J Clin Oncol 2012; 30:2063-9. 9. Rossell R, Molina M, Sanchez J, et al. De novo EGFR T790M mutation modifies outcome to second-line erlotinib in nonesmall-cell lung cancer (NSCLC) according to metastatic site and upfront chemotherapy. Cancer Res 2011; 71(8 suppl 1), Abstract 4112. 10. Rosell R, Moran T, Queralt C, et al. Screening for epidermal growth factor receptor mutations in lung cancer. N Engl J Med 2009; 361:958-67. 11. Godin-Heymann N, Ulkus L, Brannigan BW, et al. The T790M “gatekeeper” mutation in EGFR mediates resistance to low concentrations of an irreversible EGFR inhibitor. Mol Cancer Ther 2008; 7:874-9. 12. Yun CH, Mengwasser KE, Toms AV, et al. The T790M mutation in EGFR kinase causes drug resistance by increasing the affinity for ATP. Proc Natl Acad Sci U S A 2008; 105:2070-5. 13. Kobayashi S, Ji H, Yuza Y, et al. An alternative inhibitor overcomes resistance caused by a mutation of the epidermal growth factor receptor. Cancer Res 2005; 65: 7096-101. 14. Kwak EL, Sordella R, Bell DW, et al. Irreversible inhibitors of the EGF receptor may circumvent acquired resistance to gefitinib. Proc Natl Acad Sci U S A 2005; 102:7665-70. 15. Miller VA, Hirsh V, Cadranel J, et al. Afatinib versus placebo for patients with advanced, metastatic nonesmall-cell lung cancer after failure of erlotinib, gefitinib, or both, and one or two lines of chemotherapy (LUX-Lung 1): a phase 2b/3 randomised trial. Lancet Oncol 2012; 13:528-38. 16. Sun JM, Ahn MJ, Choi YL, et al. Clinical implications of T790M mutation in patients with acquired resistance to EGFR tyrosine kinase inhibitors. Lung Cancer 2013; 82:294-8. 17. Sequist LV, Soria JC, Goldman JW, et al. Rociletinib in EGFR-mutated nonesmall-cell lung cancer. N Engl J Med 2015; 372:1700-9. 18. Janne PA, Yang JC, Kim DW, et al. AZD9291 in EGFR inhibitor-resistant nonesmall-cell lung cancer. N Engl J Med 2015; 372:1689-99. 19. Goss GD, Yang JC, Ahn MJ, et al. AZD9291 in pre-treated patients with T790M positive advanced nonesmall-cell lung cancer (NSCLC): pooled analysis from two phase II studies. In proceedings at the European Cancer Congress, 25-29 September 2015, Vienna, Austria. ABSTRACT 3113, 365. 20. Stewart EL, Mascaux C, Pham NA, et al. Clinical correlation of patient-derived xenografts and response to epidermal growth factor receptor (EGFR) inhibitors in EGFR mutant lung adenocarcinoma. J Clin Oncol 2015; 33:2472-80. 21. John T, Kohler D, Pintilie M, et al. The ability to form primary tumor xenografts is predictive of increased risk of disease recurrence in early-stage nonesmall-cell lung cancer. Clin Cancer Res 2011; 17:134-41. 22. Chung KP, Wu SG, Wu JY, et al. Clinical outcomes in nonesmall-cell lung cancers harboring different exon 19 deletions in EGFR. Clin Cancer Res 2012; 18:3470-7. 23. Regales L, Gong Y, Shen R, et al. Dual targeting of EGFR can overcome a major drug resistance mutation in mouse models of EGFR mutant lung cancer. J Clin Invest 2009; 119:3000-10. 24. Mack P, Goodwin N, Holland W, et al. Divergent activity of afatinib (AFAT) and cetuximab (CET) in patient-derived xenograft (PDX) models of acquired erlotinib resistance. Presented at the 15th World Conference on Lung Cancer, Sydney, Australia. October 27-30, 2013. 25. Li D, Ambrogio L, Shimamura T, et al. BIBW2992, an irreversible EGFR/HER2 inhibitor highly effective in preclinical lung cancer models. Oncogene 2008; 27: 4702-11. 26. Katakami N, Atagi S, Goto K, et al. LUX-Lung 4: a phase II trial of afatinib in patients with advanced nonesmall-cell lung cancer who progressed during prior treatment with erlotinib, gefitinib, or both. J Clin Oncol 2013; 31:3335-41.

Petra Martin et al 27. Perez-Torres M, Guix M, Gonzalez A, et al. Epidermal growth factor receptor (EGFR) antibody down-regulates mutant receptors and inhibits tumors expressing EGFR mutations. J Biol Chem 2006; 281:40183-92. 28. Janjigian YY, Smit EF, Groen HJ, et al. Dual inhibition of EGFR with afatinib and cetuximab in kinase inhibitor-resistant EGFR-mutant lung cancer with and without T790M mutations. Cancer Discov 2014; 4:1036-45. 29. Bianco R, Shin I, Ritter CA, et al. Loss of PTEN/MMAC1/TEP in EGF receptorexpressing tumor cells counteracts the antitumor action of EGFR tyrosine kinase inhibitors. Oncogene 2003; 22:2812-22. 30. Engelman JA, Zejnullahu K, Gale CM, et al. PF00299804, an irreversible panERBB inhibitor, is effective in lung cancer models with EGFR and ERBB2 mutations that are resistant to gefitinib. Cancer Res 2007; 67:11924-32. 31. Engelman JA, Janne PA, Mermel C, et al. ErbB-3 mediates phosphoinositide 3kinase activity in gefitinib sensitive nonesmall-cell lung cancer cell lines. Proc Natl Acad Sci U S A 2005; 102:3788-93. 32. Engelman JA, Mukohara T, Zejnullahu K, et al. Allelic dilution obscures detection of a biologically significant resistance mutation in EGFR-amplified lung cancer. J Clin Invest 2006; 116:2695-706. 33. Miyamoto S, Fukami T, Yagi H, et al. Potential for molecularly targeted therapy against epidermal growth factor receptor ligands. Anticancer Res 2009; 29:823-30. 34. Hynes NE, Lane HA. ERBB receptors and cancer: the complexity of targeted inhibitors. Nat Rev Cancer 2005; 5:341-54. 35. Yoshida M, Shimura T, Sato M, et al. A novel predictive strategy by immunohistochemical analysis of four EGFR ligands in metastatic colorectal cancer treated with anti-EGFR antibodies. J Cancer Res Clin Oncol 2013; 139:367-78.

36. Mutsaers AJ, Francia G, Man S, et al. Dose-dependent increases in circulating TGF-alpha and other EGFR ligands act as pharmacodynamic markers for optimal biological dosing of cetuximab and are tumor independent. Clin Cancer Res 2009; 15:2397-405. 37. Baselga J, Mendelsohn J, Kim YM, et al. Autocrine regulation of membrane transforming growth factor-a cleavage. J Biol Chem 1996; 271:3279-84. 38. Addison CL, Ding K, Zhao H, et al. Plasma transforming growth factor alpha and amphiregulin protein levels in NCIC Clinical Trials Group BR.21. J Clin Oncol 2010; 28:5247-56. 39. Masago K, Fujita S, Hatachi Y, et al. Clinical significance of pretreatment serum amphiregulin and transforming growth factor-alpha, and an epidermal growth factor receptor somatic mutation in patients with advanced non-squamous, nonesmall-cell lung cancer. Cancer Sci 2008; 99:2295-301. 40. Ishikawa N, Daigo Y, Takano A, et al. Increases of amphiregulin and transforming growth factor-alpha in serum as predictors of poor response to gefitinib among patients with advanced nonesmall-cell lung cancers. Cancer Res 2005; 65: 9176-84. 41. Salomon DS, Brandt R, Ciardiello F, et al. Epidermal growth factor-related peptides and their receptors in human malignancies. Crit Rev Oncol Hematol 1995; 19: 183-232. 42. Yotsumoto F, Sanui A, Fukami T, et al. Efficacy of ligand-based targeting for the EGF system in cancer. Anticancer Res 2009; 29:4879-85. 43. Neal JW, Heist RS, Fidias P, et al. Cetuximab monotherapy in patients with advanced nonesmall-cell lung cancer after prior epidermal growth factor receptor tyrosine kinase inhibitor therapy. J Thorac Oncol 2010; 5:1855-8.

Clinical Lung Cancer Month 2016

-9

Effect of Cetuximab on an EGFR Exon 19 Del/T790 PDX Supplemental Figure 1 Growth Curves of 164 Control, Cetuximab, Erlotinib, and Dacomitinib Arms

Abbreviation: Treat ¼ treatment.

Supplemental Figure 2 Phosphorylation Status of the Epidermal Growth Factor Receptor (EGFR) Pathway During the Pharmacodynamic Study With Erlotinib (Erlo) and Dacomitinib (Dac) Using Western Blot Analysis. Biological Triplicates Were Performed Per Time Point, in the Erlo and Dac Treatment Arms. Western Blot Analysis Showed Reduction in phosphorylated (p)EGFR at 24 Hours and Inhibition of pERK and pS6 at 2 Hours and 24 Hours

9.e1

-

Clinical Lung Cancer Month 2016

Petra Martin et al Supplemental Figure 3 Fold Changes in Epidermal Growth Factor Receptor (EGFR)-Ligand Gene Expression Over Time According to Treatment (Trt) Group

Abbreviations: AR ¼ amphiregulin; CET ¼ cetuximab; DAC ¼ dacomitinib; ERLO ¼ erlotinib; EGF ¼ epidermal growth factor receptor; HBEGFR ¼ heparin-binding epidermal growth factor-like growth factor receptor; TGFa ¼ tumor growth factor alpha.

Clinical Lung Cancer Month 2016

- 9.e2