AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics: Discovery, Biology, and Clinical Applications: Abstracts
aacr.org
A125
Identification of EGFr and K-RAS Somatic Gene Mutations from a Panitumumab Phase 2 NSCLC Clinical Trial: Association of Mutations, Preclinical Characterization, and Clinical Outcome. Todd Juan*,1 Dan Freeman*,1 Ildiko Sarosi,1 Jeffrey Crawford,2 Alan Sandler,3 Joan Schiller,4 Diane Prager,5 David Johnson,3 Michael Wolf,1 Robert Radinsky.1 Amgen Inc.,1 Thousand Oaks, CA, Duke University Medical Center,2 Durham, NC, Vanderbilt-Ingram Cancer Center,3 Nashville, TN, University of Wisconsin,4 Madison, WI, UCLA Medical Center,5 Los Angeles, CA.
Background: Mutant epidermal growth factor receptor (EGFr) and mutant K-RAS have been associated with non-small cell lung cancer (NSCLC) responsiveness, or lack there of, to small molecule tyrosine kinase inhibitors. Unlike clinical associations that have been made with gefitinib and erlotinib, little is known about how patients with EGFr and K-RAS mutations respond to antibodies against EGFr. Panitumumab, a fully human monoclonal antibody directed against the EGFr extracellular domain, is currently being studied in a NSCLC trial comparing chemotherapy (carboplatin/paclitaxel) vs chemotherapy plus panitumumab (n=175; primary endpoint = time to progression). Study objectives were to associate EGFr and K-RAS mutations with patient clinical outcome and to determine in vitro if panitumumab has differential activity versus gefitinib against mutant EGFr.
Methods: Genomic DNA was isolated from dissected FFPE tumor sections (pretreatment); PCR was performed on EGFr gene exons 18, 19, 20, 21, and 23 and K-RAS exon 2. PCR products were subcloned, and >30 colonies per exon were sequenced. After mutations were found in tumors, genomic DNA was sequenced to confirm the existence of somatic mutations. These data were linked to clinical outcome data (investigator assessed per RECIST every 6 weeks). To determine inhibition of EGF-induced EGFr autophosphorylation in vitro, WT and mutant EGFr-overexpressing CHO cells were treated with 0-2 µM of either panitumumab or gefitinib prior to EGF stimulation. Results: Of 59 samples tested, 5 different somatic mutations in EGFr and K-RAS were identified in 17 patients (table). The respective IC50 for EGF-induced autophosphorylation of EGFr WT, (750-759), L858R and T790M cells were 14.6, 1.4, 3.2 and >2000nM with gefitinib and 0.23, 0.17, 0.18 and 0.23nM with panitumumab.
Conclusion: Eight patients harbored EGFr somatic gene mutations, whereas 10 patients harbored K-RAS mutations. In EGFr-overexpressing CHO cells, panitumumab inhibited EGF-induced EGFr autophosphorylation regardless of mutational status. A de novo EGFr exon 20 gate-keeper mutation was found in this study. This exon 20 EGFr mutation, previously shown to be resistant to gefitinib in vitro and in vivo, was sensitive to panitumumab in vitro. The patient carrying this exon 20 mutation experienced stable disease for 12 weeks (2 assessments) in response to treatment with panitumumab plus chemotherapy.
B43 Panitumumab Induces Internalization of the Epidermal Growth Factor Receptor (EGFr). Ian N. Foltz,1 Chadwick T. King,1 Meina Liang,2 Kenneth A. Schooley, 3 Douglas P. Cerretti,3 Dan Freeman,4 Robert Radinsky,4 Xiao-Dong Yang.2 Abgenix Biopharma Inc.,1 Burnaby, BC, Canada, Abgenix Inc.,2 Fremont, CA, Amgen,3 Seattle, WA, Amgen,4 Thousand Oaks, CA.
PURPOSE: To determine if panitumumab, a fully human monoclonal antibody to EGFr, is internalized upon binding to EGFr.
METHODS: Five human cancer cell lines were used: SKMES, A431, A549, H1975 and HeLa. Confocal microscopy was used to directly visualize the internalization of EGFr in cells treated with either panitumumab (0, 0.5, 6.5, 30.5 or 54.5 hrs) or EGF (0.5 hr). The cells were then fixed, permeabilized, and stained with a non-competitive murine antibody against EGFr conjugated to FITC to detect total EGFr and an anti-human IgG conjugated with phycoerythrin to detect panitumumab. As an alternative approach to quantitate internalization of panitumumab, EGFr-expressing cells were sequentially treated with a saturating concentration of panitumumab and an anti-human IgG-Fc Fab conjugated to a disulfide-linked fluorochrome at 4C. The cells were then incubated at either 4C or 37C for various times followed by treatment with or without a non-cell permeable reducing reagent to remove any extracellular disulfide-linked fluorochrome from the cell. The geometric mean fluorescence of each sample was determined using FACS and used to quantitate internalization. RESULTS: Panitumumab and EGF were found to be internalized on SKMES, A431, and A549 cells using confocal microscopy. Treatment with EGF led to a rapid internalization of EGFr with nearly complete removal of cell surface EGFr by 0.5 hr. While the internalization of EGFr in response to panitumumab treatment was visually less than that induced by EGF, EGFr and panitumumab were clearly co-localized inside the cell from 0.5 hr to 54.5 hrs. Of note, the internalization induced by EGF was observed as punctate perinuclear staining, while internalization induced by panitumumab was diffuse throughout the inside of the cell. By FACS, it was observed that panitumumab was internalized at a constant rate over the first hour in A431 cells. The percent of panitumumab internalized after 1 hour was 20%, 36%, 74% and 28% for A431, A549, H1975 and HeLa cells, respectively. This internalization was shown to be an energy dependent process as it was completely inhibited by sodium azide pre-treatment of cells. To correlate panitumumab-induced internalization with EGFr expression in vivo, H1975 tumor xenografts were treated with panitumumab or an isotype control antibody for three weeks. Western analysis indicated that panitumumab significantly reduced total EGFr expression, but not total AKT expression, as compared to the control antibody.
CONCLUSIONS: These results using two distinct approaches demonstrate that panitumumab is internalized upon binding to EGFr expressed on tumor cells. However, the internalization of EGFr induced by panitumumab varied among the different cell lines tested, displayed slower kinetics than EGF, and may be mechanistically distinct from that induced by EGF. Interestingly, the strong internalization observed in H1975 cells correlates with the down-regulation of EGFr in H1975 tumor xenografts, and suggests that the internalization of EGFr may represent a mechanism of action of panitumumab in certain tumors. B44 Effects of an Antibody to VEGFR-3, mF4-31C1,
B64 Activity of Panitumumab Alone and in combination with Chemotherapy Against Mutant Epidermal Growth Factor Receptor (EGFr)- Expressing Non- small Cell Lung Carcinoma (NSCLC) Cell Lines and Xenografts. Tammy Bush*,1 Dan Freeman*,1 Selam Ogbagabriel,1 Brian Belmontes,1 Carl Kozlosky,2 Angelo Baher, 1 Carol Johnson,1 Gwyneth Van,1 Doug Cerretti,2 Robert Radinsky.1 Amgen Inc.,1 Thousand Oaks, CA, Amgen Washington,2 Seattle, WA.
Background: EGFr kinase domain mutations render the receptor hyperresponsive to ligand and hypersensitive to small molecule tyrosine kinsase inhibitors. However, little is know about how these mutations respond to antibodies against EGFr. Panitumumab, a fully human monoclonal antibody, binds to EGFr with high affinity (5x10- 11M), prevents ligand-induced activation, and results in arrest of tumor cell proliferation. The primary study objectives were to measure inhibition of ligand-dependent EGFr phosphorylation by panitumumab in vitro in mutant EGFr-expressing NSCLC cells and to assess the anti-tumor activity of panitumumab alone and in combination with chemotherapy in xenograft models in vivo. The secondary objective was to address the ability of panitumumab to inhibit phosphorylation of the mutant EGFr in liganddependent and -independent cell lines.
Methods: Mutant EGFr expressing NSCLC (NCI-H1975 [L858R+T790M], NCI-H1650 [ 746-752]), ligand-independent 293T, and ligand-dependent CHO cells were treated with panitumumab prior to EGF stimulation to determine the inhibition of EGFr autophosphorylation. The inhibition of specific EGFr phosphorylation sites (Y845, Y1045, Y1068, Y1086, and Y1173) was assayed by western blot or immunoassay. Established tumors (200 mm3) were treated with panitumumab (25, 100 or 500 µg/mouse 2Xweek) alone or in combination with docetaxel (10 or 20 mg/kg 1Xweek) or cisplatin (7.5 mg/kg 1Xweek). FACS and immunohistochemistry were performed to evaluate EGFr levels, panitumumab penetration, and proliferation levels. Results: Treatment of mutant EGFr-expressing NSCLC cells with panitumumab resulted in inhibition of ligand-dependent autophosphorylation at all 5 tyrosine phosphorylation sites. In 293T (ligand independent) cells, no significant inhibition of autophoshorylation was observed. Treatment of established NCI-H1975 and NCI-H1650 xenografts with panitumumab resulted in statistically significant antitumor effect compared to IgG2 treatment (50% inhibition, p<0.0001). Combination treatment of panitumumab plus docetaxel enhanced antitumor efficacy against NCI-H1975 and NCI-H1650 xenografts compared to either agent alone (70% inhibition, p=0.1390 and 0.0534, respectively). In contrast, combination treatment of panitumumab plus cisplatin did not enhance antitumor efficacy of panitumumab. FACS analysis revealed that EGFr levels were 6-fold higher in NCI-H1650 vs NCI-H1975 while immunohistochemistry demonstrated tumor penetration by panitumumab in both tumors. With combination therapy of panitumumab plus docetaxel, Ki67 and pMAPK staining decreased > 60% compared to either agent alone in the NCI-H1650 xenografts, but not in the NCI-H1975 xenografts, on day 12.
Conclusions: Panitumumab effectively inhibited ligand-dependent phosphorylation of mutant EGFr. Treatment with panitumumab as monotherapy or in combination with docetaxel resulted in a significant growth inhibition of established tumors expressing mutant EGFr. Panitumumab was detected in tumor tissues, which correlated with a reduction in tumor cell proliferation. These data support the potential clinical application of panitumumab alone and in combination with chemotherapy for the treatment of NSCLC patients with EGFr kinase domain mutations.
B72 Antitumor Efficacy of Panitumumab Alone or in Combination with Chemotherapy against Human Pancreatic Carcinoma Xenografts. Tammy Bush*, Dan Freeman*, Selam Ogbagabriel, Beth Ziegler, Brian Belmontes, Gwyneth Van, Carol Johnson, Robert Radinsky. Amgen Inc., Thousand Oaks, CA.
Background: The prognosis for pancreatic cancer patients is poor and emphasizes the need for new, effective pancreatic cancer therapies. Panitumumab, a fully human monoclonal antibody, binds to the epidermal growth factor receptor (EGFr) with high affinity (Kd=5x10-11 M), prevents ligand-induced activation, and results in arrest of tumor cell proliferation. Because many primary pancreatic carcinomas express EGFr, targeting this receptor has the potential to effectively treat pancreatic cancer. The objective of this study was to assess the antitumor activity of panitumumab alone or in combination with chemotherapy against 2 established xenograft models of pancreatic cancer.
Methods: Inhibition of ligand-induced autophosphorylation by panitumumab was determined both in vitro and in vivo. In vitro, BxPC-3 and MiaPaCa-2 cells were treated with 0.5, 2, and 10 g/ml of panitumumab for 1 hour prior to 100 ng/ml EGF stimulation. In vivo, mice bearing established tumors were treated with 500 µg panitumumab or control 96 hours prior to stimulation with 100 µg of EGF. To determine in vivo efficacy, mice bearing established BxPC-3 and MiaPaCa-2 tumors (200 mm3) were treated with panitumumab (100, 200, or 500 g) alone or in combination with gemcitabine (80 mg/kg once weekly). Immunohistochemistry was performed to evaluate EGFr levels, panitumumab tumor penetration, and proliferation levels at day 18 post-treatment. Results: Treatment of BxPC-3 and MiaPaCa-2 cells in vitro and of xenograft tumors in vivo with panitumumab resulted in greater than 90% inhibition of ligand-induced autophosphorylation of EGFr. Monotherapy treatment with either panitumumab or gemcitabine resulted in significant growth inhibition of Mia- PaCa-2 (36% inhibition, p = 0.0009 and 38% inhibition, p = 0.0012, respectively) and BxPC-3 (44% inhibition, p < 0.0001 and 25% inhibition, p = 0.0007, respectively) xenograft tumors when compared to the control group. Combination therapy with panitumumab plus gemcitabine resulted in enhanced antitumor efficacy against MiaPaCa-2 (58% inhibition, p = 0.04 vs either monotherapy) and against BxPC-3 (61% inhibition, p = 0.1575 vs panitumumab monotherapy and p < 0.0001 vs gemcitabine monotherapy) xenograft tumors compared with either agent alone. EGFr levels and tumor penetration by panitumumab were detected by immunohistochemistry in both tumor models. Combination treatment with panitumumab plus gemcitabine resulted in a greater than 60% reduction in Ki67 and BrdU staining at day 18 post-treatment compared to either agent alone.
Conclusions: Panitumumab inhibited ligand-induced EGFr autophosphorylation in vitro and in vivo in both BxPC-3 and MiaPaCa-2 pancreatic models. Panitumumab is present in the tumor tissues and correlates with a significant reduction in Ki67 levels and cell proliferation when combined with gemcitabine, resulting in growth inhibition of established models of pancreatic cancer. These data provide preclinical evidence for the potential clinical application of panitumumab alone or in combination for the treatment of pancreatic cancer.
B73 Antitumor Activity of AMG 706, A Novel Tyrosine Kinase Inhibitor, in Combination with Panitumumab, A Fully Human Antibody Targeting the EGF Receptor, Against Multiple Established Human Tumor Xenograft Models in Nude Mice. C Starnes, D Freeman, T Bush, J Leal, K McDorman, A Coxon, V Patel, T Polverino, R Kendall, R Radinsky. Amgen Inc., Thousand Oaks, CA.
Background: AMG 706 is a potent, oral, small molecule, multi-kinase inhibitor with both antiangiogenic and direct antitumor activity achieved by selectively targeting VEGF, PDGF, and Kit receptors.
AMG 706 demonstrated potent antitumor activity against multiple established human tumor xenograft models in nude mice. Panitumumab is a fully human IgG2 antibody that specifically targets the human EGF receptor with high affinity and inhibits ligand-induced receptor tyrosine phosphorylation, resulting in antitumor activity against various xenograft models. The present studies examined 3 xenograft models (with varying degrees of responsiveness to either monotherapy) for their response to combination treatment with both targeted agents.
Methods: For each model, female CD1 nu/nu mice were challenged subcutaneously with 1x107 tumor cells on day 0. Treatment (range 10-75 mg/kg AMG 706/dose; 20-500 µg/mouse 2x/wk panitumumab) was initiated after the tumors became established (average tumor volume range ~200-800 mm3) and continued for approximately 4-5 weeks. Tumor size was measured with a caliper and reported as a function of time. Bodyweights were recorded as an index of toxicity. Individual groups are compared by p values calculated from Repeated Measures Analysis of Variance (RMANOVA) followed by Scheffe post hoc testing for multiple comparisons. Results: In each of the 3 models, the combination of panitumumab + AMG 706 resulted in a therapeutic effect that was statistically superior to that seen with either single agent alone at the given dose levels employed (A431 epidermoid carcinoma, combination vs. AMG 706 alone, p < 0.0001; vs. panitumumab alone, p = 0.0003; HT29 colorectal carcinoma, p = 0.0013 and p < 0.0001, respectively; Calu-6 non-small cell lung cancer, p = 0.0204 and p < 0.0001, respectively). No adverse drug-drug interactions were observed.
Conclusions: A beneficial therapeutic effect of the combination of AMG 706 and panitumumab was observed in 3 different xenograft models with no apparent adverse effects. These results support the potential of combining targeted therapies that interfere with different pathways to provide additive clinical benefit. Specifically, the combined use of AMG 706 and panitumumab in clinical trials is warranted. |