If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Anti–Epidermal Growth Factor Vaccine Antibodies Enhance the Efficacy of Tyrosine Kinase Inhibitors and Delay the Emergence of Resistance in EGFR Mutant Lung Cancer Cells
Corresponding author. Address for correspondence: Miguel Ángel Molina-Vila, PhD, Laboratory of Oncology/Pangaea Oncology S.L., Quirón-Dexeus University Institute, C/Sabino Arana 5, 08023 Barcelona, Spain.
Clinical and Translational Oncology Group, Institute of Oncology, Fundacion Santa Fe de Bogota, Bogota, ColumbiaFoundation for Clinical and Applied Cancer Research (FICMAC), Bogota, Columbia
Instituto Oncológico Dr. Rosell (IOR), Quirón-Dexeus University Institute, Barcelona, SpainInstituto Oncológico Dr. Rosell (IOR), Sagrat Cor Hospital, Barcelona, Spain
Instituto Oncológico Dr. Rosell (IOR), Quirón-Dexeus University Institute, Barcelona, SpainCatalan Institute of Oncology and Institut d’Investigació en Ciències de la Salut Germans Trias i Pujol, Badalona, Spain
Mutations in EGFR correlate with impaired response to immune checkpoint inhibitors and the development of novel immunotherapeutic approaches for EGFR mutant NSCLC is of particular interest. Immunization against epidermal growth factor (EGF) has shown efficacy in a phase III trial including unselected NSCLC patients, but little was known about the mechanisms involved in the effects of the anti-EGF antibodies generated by vaccination (anti-EGF VacAbs) or their activity in tumor cells with EGFR mutations.
Methods
The EGFR-mutant, NSCLC cell lines H1975, and PC9, together with several gefitinib and osimertinib-resistant cells derived from PC9, were treated with anti-EGF VacAbs and/or EGFR tyrosine kinase inhibitors (TKIs). Cell viability was analyzed by proliferation assays, cell cycle by fluorescence-activated cell sorting analysis, and levels of RNA and proteins by quantitative retro-transcription polymerase chain reaction and Western blotting.
Results
Anti-EGF VacAbs generated in rabbits suppressed EGF-induced cell proliferation and cycle progression and inhibited downstream EGFR signaling in EGFR-mutant cells. Sera from patients immunized with an EGF vaccine were also able to block activation of EGFR effectors. In combination, the anti-EGF VacAbs significantly enhanced the antitumor activity of all TKIs tested, suppressed Erk1/2 phosphorylation, blocked the activation of signal transducer and activator of transcription 3 (STAT3) and downregulated the expression of AXL receptor tyrosine kinase (AXL). Finally, anti-EGF VacAbs significantly delayed the emergence in vitro of EGFR TKI resistant clones.
Conclusions
EGFR-mutant patients can derive benefit from immunization against EGF, particularly if combined with EGFR TKIs. A phase I trial of an EGF vaccine in combination with afatinib has been initiated.
Several antitumor therapies target the EGFR pathway. EGFR tyrosine kinase inhibitors (TKIs) are the standard treatment in advanced NSCLC with EGFR mutations.
Erlotinib versus standard chemotherapy as first-line treatment for European patients with advanced EGFR mutation-positive non–small-cell lung cancer (EURTAC): a multicentre, open-label, randomised phase 3 trial.
Monoclonal antibodies against the extracellular portion of the receptor, such as cetuximab or panitumumab, have been approved for KRAS-wild-type colorectal cancer. In advanced NSCLC, the addition of cetuximab to chemotherapy or chemoradiation has not improved progression-free survival or overall survival in phase III trials, although some benefit was reported in tumors of squamous histology with EGFR amplification.
Standard-dose versus high-dose conformal radiotherapy with concurrent and consolidation carboplatin plus paclitaxel with or without cetuximab for patients with stage IIIA or IIIB non-small-cell lung cancer (RTOG 0617): a randomised, two-by-two factorial phase 3 study.
Cetuximab and first-line taxane/carboplatin chemotherapy in advanced non–small-cell lung cancer: results of the randomized multicenter phase III trial BMS099.
Herbst R, Redman M, Kim E. A randomized, phase III study comparing carboplatin/paclitaxel or carboplatin/paclitaxel/bevacizumab with our without concurrent cetuximab in patients with advanced non–small cell lung cancer (NSCLC): SWOG S0819. Presented at the 2015 World Conference on Lung Cancer, Denver, Colorado.
In the case of EGFR-mutant (EGFR-mut) tumors, a phase Ib study combining afatinib and cetuximab in 126 patients with acquired resistance to gefitinib/erlotinib showed a 29% objective response rate with a 44% of grade 3 adverse events.
Continued use of afatinib with the addition of cetuximab after progression on afatinib in patients with EGFR mutation-positive non-small-cell lung cancer and acquired resistance to gefitinib or erlotinib.
A phase III clinical trial of the epidermal growth factor vaccine CIMAvax-EGF as switch maintenance therapy in advanced non-small cell lung cancer patients.
The first of such vaccines, named CIMAvax-EGF and composed of human EGF conjugated to a P64K carrier protein, has been evaluated in a phase III clinical trial including 405 unselected stage IIIB/IV NSCLC patients. After first-line chemotherapy, they were randomized to CIMAvax-EGF or best supportive care. Median survival was 12.4 months for patients receiving at least 4 vaccine doses, versus 9.4 months for the control arm; and the clinical benefit of the CIMAvax-EGF was superior in patients with higher concentration of EGF in serum baseline. Despite these promising results, little is known about the mechanisms of action of the anti-EGF antibodies raised by vaccination or their differential activity in tumors with EGFR mutations or other genetic alterations.
In the current study, we analyzed the effects of antibodies against human EGF generated by vaccination (anti-EGF VacAbs) in EGFR-mut NSCLC cell lines. We found that they reversed the effects of EGF and potentiate the antitumor activity of EGFR TKIs, further inhibiting phosphorylation of Erk1/2 and other markers and enhancing cell cycle arrest and apoptosis. In addition, anti-EGF VacAbs significantly downregulated AXL receptor tyrosine kinase (AXL) expression and delayed the emergence of in vitro acquired resistance to afatinib and gefitinib. Our results provide the rationale for a phase I clinical trial of afatinib in combination with anti-EGF vaccination in advanced, EGFR-mut NSCLC patients.
Materials and Methods
Rabbit Antibodies and Human Sera
Anti-EGF VacAbs were obtained by immunizing rabbits with four injections of a human recombinant EGF (Scotia Biologics Ltd., Aberdeen, United Kingdom) combined with Montanide adjuvant (Seppic, Paris, France). Pre-immunization sera from the same animals were collected and purified to be used as control antibodies (C-Abs). Sera were purified by Melon gel and treated by caprylic acid to remove contaminants. Because the purification process dilutes the antibodies, the final preparation was re-analyzed by enzyme-linked immunosorbent assay (ELISA) revealing a titer of approximately 1:1,000; significantly below the 1:10,000 to 1:64,000 titers usually achieved in patients vaccinated with EGF. Sera from patients enrolled in the BV-NSCLC-001 clinical trial (NCT01444118) were kindly provided by Bioven Europe Ltd. The BV-NSCLC-001 is a randomized trial to study the safety and efficacy of EGF cancer vaccination in late-stage (IIIB/IV) NSCLC patients who are immunized with cyclophosphamide and the human rEGF-P64K/Montanide ISA 51 vaccine. All patients provided written, informed consent.
Materials and Cell Lines
TKIs were purchased from Selleck Chemicals (Houston, Texas), human EGF from Cell Signaling Technologies (Beverly, Massachusetts), and antibodies for Western blotting from Cell Signaling Technology or Sigma-Aldrich (St. Louis, Missouri). All tissue culture materials were obtained from Invitrogen (Paisley, Scotland, United Kingdom).
Six EGFR-mut cell lines with different molecular alterations were used in the study (Supplementary Table 1). NCI-H1975 cells were purchased from the American Type Culture Collection (ATCC, Manassas, Virginia), whereas PC9 cells were provided by F. Hoffman-La Roche Ltd (Basel, Switzerland) with the authorization of Dr. Mayumi Ono (Kyushu University, Fukuoka, Japan). PC9-GR2 and GR4 are gefitinib-resistant whereas PC9-OR4 and PC9-GR4AZD1 are osimertinib-resistant cells; the four lines were generated in our laboratory by exposing PC9 cells to increasing concentrations of the drug (Supplementary Fig. 1A).
The four PC9-derived cells over-express AXL, PC9-GR2 cells also present hepatocyte growth factor receptor (MET) activation. Cell lines were maintained in Roswell Park Memorial Institute (RPMI) medium supplemented with 10% fetal bovine serum (FBS), 50 μg/mL penicillin-streptomycin, 2 mmol/L L-glutamine, and 2.5 μmol/L gefitinib (PC9-GR2, PC9-GR4) or osimertinib (PC9-OR4, PC9-GR4AZD1) in a humidified atmosphere with 5% CO2 at 37ºC. Cells were weekly tested for mycoplasmas and authenticated by monthly genotyping for EGFR and tumor protein p53 (TP53) mutations and a panel of four polymorphisms. After no more than 15 passages, cells were discarded and a new, low-passage vial was thawed.
Cell Growth, Viability, and Emergence of Resistant Assays
Cell viability was assessed by the thiazolyl blue tetrazolium bromide (MTT; Sigma) assay. Cells were seeded at 2000 (PC9), 4000 (PC9-GR4, PC9-OR4), or 6000 (H1975) per well in 96-well plates, allowed to attach for 24 hours in RPMI + 10% FBS, washed twice with phosphate-buffered saline (PBS), and treated with EGF (10 ng/mL), antibodies, EGFR TKIs, or combinations for 72 hours in RPMI + 0.5% human serum (HS). After treatment, cells were incubated with medium containing MTT (0.75 mg/mL) for 1 to 2 hours at 37ºC. Culture medium was removed; formazan crystals reabsorbed in DMSO (Merck, Darmstadt, Germany) and cell numbers were estimated using an Anthos 2020 microplate reader (Biochrom Ltd., Cambridge, United Kingdom). Data were derived from at least three independent experiments. To study the acquisition of resistance to TKIs, we seeded 300 PC9 cells per well in 96-well plates, using two plates per treatment (120 wells). Cells were allowed to attach and the treatments were started after 24 hours in RPMI + 10% FBS. Media were changed every week, plates were inspected thrice a week under the microscope, and wells >50% confluent were scored as positive.
Cells were treated with EGF (10 ng/mL), C-Abs, anti-EGF VacAbs, EGFR TKIs, or combinations, for 2 to 24 hours in RPMI + 0.5% HS. After washing twice with cold PBS, cultures were scraped into radioimmunoprecipitation assay buffer (20 mmol/L Tris-HCl pH 7.5, 150 mmol/L NaCl, 1 mmol/L Na2EDTA, 1 mmol/L EGTA, 1% NP-40, 1% sodium deoxycholate, 2.5 mmol/L sodium pyrophosphate, 1 mmol/L β-glycerophosphate, 1 mmol/L Na3VO4, 1 μg/mL leupeptin (Cell Signaling Technologies), 2 mmol/L phenylmethylsulfonyl fluoride (PMSF) and protease inhibitor cocktail (Roche Diagnostics, Mannheim, Germany) and passed through an insulin syringe. Lysates were incubated on ice for 15 minutes, centrifuged for 10 minutes at 14,000 rpm, and immediately analyzed or frozen at –80 ºC. Protein extracts (25 μg) were boiled in Laemmli buffer (NuPAGE- LDS sample buffer 4X; Invitrogen. Carlsbad, California), resolved in sodium dodecyl sulfate –polyacrylamide gels and transferred to polyvinylidene difluoride membranes (Bio-Rad, Hercules, California). Membranes were incubated for 1 hour in Odyssey blocking buffer (Li-Cor Biosciences, Lincoln, Nebraska) or Phosphoblocker reagent (Cell Biolabs Inc., San Diego, California), cut, incubated with primary antibodies o/n at 4 ºC, washed three times with PBS-TWEEN 0.1% and incubated for 2 hours with a secondary antibody (anti-rabbit or anti-mouse immunoglobulin G [IgG] horseradish peroxidase-conjugated secondary antibody; GE Healthcare, New York, New York). Finally, membranes were washed with PBS-TWEEN 0.1% and revealed with Supersignal Chemiluminescence substrate (Thermofisher, Waltham, Massachusetts).
Flow Cytometry
Cells were treated under the same conditions described for Western blotting, trypsinized and centrifuged. For cycle analyses, cells were resuspended in PBS, fixed in 70% ethanol, incubated o/n at –20 ºC, centrifuged at 1200 rpm for 5 minutes at 4 ºC, incubated for 1 hour at 37 ºC in 250 μL of PBS with 50 μg/mL RNAse A (Sigma-Aldrich) and stained with propidium iodide (PI) (Roche Diagnostics) for 30 minutes at room temperature in the dark. For cell death analyses, the Annexin-V-FLUOS staining kit (Roche Diagnostics) was used according to the manufacturer’s instructions. Stained cells were analyzed with a FACSCanto II cytometer (BD Biosciences, Franklin Lakes, New Jersey) using the FACSDiva software version 6.1.2. In the case of cell death analyses, the combination of annexin V and PI was used to differentiate four cell populations; namely, viable cells (An -/PI-), early apoptotic (An+/PI-), necrotic (An-/PI+), and later apoptotic/necrotic (An+/PI+).
mRNA Analysis
RNA was isolated from cell lines as previously described.
Primer and probe sets were designed using Primer Express 3.0 Software (Applied Biosystems, Foster City, California) according to their reference sequences (http://www.ncbi.nlm.nih.gov/gene). Quantification of gene expression was performed using the ABI Prism 7900HT Sequence Detection System (Applied Biosystems). Expression levels were calculated according to the comparative ΔΔCt method. Commercial RNA controls were used as calibrators (Liver and Lung; Stratagene, La Jolla, California). For each cell line, three independent experiments were performed.
Results
Anti-EGF VacAbs Suppress the Proliferative Effects of EGF and the Activation of EGFR Downstream Pathways
We first assayed the effects of anti-EGF VacAbs generated in rabbits on four EGFR-mut NSCLC cell lines. Two of the lines, NCI-H1975 and PC9, derive from untreated, EGFR-mut tumors. The PC9 cells harbor an exon 19 deletion and are sensitive to all EGFR TKIs whereas the NCI-H1975 cells carry p.L858R and p.T790M mutations and are intrinsically resistant to first-generation but sensitive to third-generation TKIs. The two other cell lines were generated in our laboratory. PC9-GR4 is a gefitinib-resistant, osimertinib sensitive cell line carrying the p.T790M whereas PC9-OR4 cells are resistant to all EGFR TKIs, including osimertinib (Supplementary Table 1, Supplementary Fig. 1A). Dose-response experiments revealed that the anti-EGF VacAbs were able to suppress the stimulating effects of human EGF (hEGF) on cell proliferation in the four cell lines, whereas C-Abs had no activity (Figs. 1A and B, Supplementary Fig. 1B and C). Cell cycle analysis of PI-stained PC9 cells showed a significant increase in the number of cells in S + G2/M phase after 72 hours of incubation with EGF, which was blocked when anti-EGF VacAbs were added (Fig. 1C).
Figure 1Effects of EGF and anti-EGF VacAbs in EGFR-mut cell lines. Antibodies were added at the final dilutions indicated and EGF at 10 ng/mL. Medium was radioimmunoprecipitation assay + 0.5% human serum in all cases. (A,B) Results of 72-hour proliferation assays of anti-EGF VacAbs in EGFR-mut cell lines. (C) Percentage of cells in S + G2 phase by flow cytometry. Incubation time with anti-EGF VacAbs was 24 hours. (D,E) Western blot analysis of selected markers, incubation time 2 hours. (F–I) Effects of EGF on the sensitivity of EGFR-mut cells to gefitinib, erlotinib, osimertinib, and afatinib. Results of 72-hour proliferation assays. Data were pooled from at least three different experiments and presented as mean ± SEM. *p < 0.05. EGF, epidermal growth factor; VacAbs, antibodies created by vaccination; C-Ab, control antibodies; Ab, anti-EGF VacAbs.
Western blotting experiments revealed that hEGF triggered the phosphorylation of EGFR, Erk1/2, and Akt in PC9 and PC9-OR4 cells. The presence of anti-EGF VacAbs blocked the hEGF-induced activation of EGFR and Erk1/2 in a dose-dependent manner, whereas C-Abs had no effect (Fig. 1D, Supplementary Fig. 1D and E). In the NCI-H1975 cells, which carry the p.T790M mutation, anti-EGF VacAbs failed to block EGFR phosphorylation but nonetheless suppressed Erk1/2 activation (Fig. 1E). In the case of PC9-GR4 cells, hEGF did not induce a detectable activation of EGFR, Akt, or Erk1/2 at the incubation time tested (2 hours), but the anti-EGF VacAbs were still able to reduce the levels of pErk1/2 (Supplementary Fig. 1F). Control antibodies showed no effect in any of the cell lines.
Anti-EGF VacAbs Significantly Enhance the Antitumor Effects of EGFR TKIs on EGFR-mut Cells
We next investigated whether hEGF altered the antiproliferative effects of first-, second-, and third-generation EGFR TKIs on EGFR-mut NSCLC cells. We found that the addition of hEGF to the culture medium significantly reduced the activity of gefitinib, erlotinib, afatinib, and osimertinib on PC9 cells (Fig. 1F–I). Similar results were obtained in H1975, PC9-OR4, and PC9-GR4 cells treated with osimertinib (Supplementary Fig. 2A–C). In consequence, we hypothesized that the addition of anti-EGF VacAbs might potentiate the activity of EGFR TKIs. A series of experiments were performed to validate this hypothesis.
First, we used MTT assays to determine the effects of anti-EGF VacAbs in combination with TKIs. All assays were performed in presence of hEGF and incubation time was 72 hours. In the PC9 cell line, the addition of anti-EGF VacAbs significantly enhanced the antiproliferative activity of all EGFR TKIs tested, namely, gefitinib, erlotinib, afatinib, and osimertinib. Consequently, the concentrations that inhibit 50% of the EGFR TKIs were reduced 2- to 7-fold in presence of the anti-EGF VacAbs. The same potentiating effect of anti-EGF VacAbs was observed in H1975 and PC9-OR4 cells treated with osimertinib and, to a more modest extent, in PC9-GR4, where it did not reach statistical significance (Fig. 2A–G, Supplementary Table 2). Control antibodies did not show activity in any case. Cell cycle analyses in PC9 showed that gefitinib, osimertinib, and afatinib decreased the percentage of S + G2/M cells, and this effect was enhanced when they were combined with anti-EGF VacAbs. Also, annexin V assays revealed 8% to 10% of apoptotic PC9 cells under basal conditions, which decreased to <5% if EGF was present. The addition of gefitinib, afatinib, osimertinib, or a combination with anti-EGF VacAbs reversed the anti-apoptotic effect of EGF (Fig. 2H and I).
Figure 2Effects of anti-EGF VacAbs in combination with EGFR TKIs in EGFR-mutant cell lines. Antibodies were added at a 1/50 dilution and EGF at 10 ng/mL. Medium was radioimmunoprecipitation assay + 0.5% human serum in all cases. In flow cytometry assays, final concentrations of EGFR TKIs were 40 nmol/L gefitinib, 25 nmol/L osimertinib, and 1 nmol/L afatinib. (A-G) Results of 72-hour proliferation assays. *p < 0.05. (H) Percentage of cells in S + G2 phase by flow cytometry. *p < 0.05 compared to EGF treated cells. (I) Percentage of apoptotic cells by annexin V analysis. *p < 0.05 compared to EGF treated cells. Data were pooled from at least three different experiments and presented as mean ± SEM. EGF, epidermal growth factor; VacAbs, antibodies created by vaccination; TKI, tyrosine kinase inhibitor; C-Ab, control antibodies; Ab, anti-EGF VacAbs.
Combination Treatment With Anti-EGF VacAbs and EGFR-TKIs Effectively Blocks the Activation of EGFR, Erk1/2, and Akt
Next, we used Western blotting analyses of key signal transduction molecules of the EGFR pathway to gain insight into the mechanisms responsible for the improved efficacy of the combination therapy in EGFR-mut cells. In the case of PC9, after a 2-hour incubation period, we observed that gefitinib consistently reduced the phosphorylation of EGFR while the inhibitory effect of the drug on Erk1/2 phosphorylation was less pronounced if hEGF was present (Fig. 3A, compare lanes “Gefitinib” and “Gefitinib + EGF”). In contrast, the inhibitory effect on Erk1/2 activation of the anti-EGF VacAbs was of the same magnitude (Fig. 3A, compare lanes “Ab” and “Ab + EGF”). When combined in presence of EGF, gefitinib and the anti-EGF VacAbs strongly inhibited the phosphorylation of EGFR, Akt and Erk1/2 (Fig. 3A, compare the three right lanes). The combination was particularly effective in reducing the levels of pErk1/2 if compared to gefitinib or anti-EGF VacAbs alone. Similar results were obtained when treating PC9 cells with osimertinib, erlotinib, or afatinib, together with anti-EGF VacAbs, with an effective inhibition of EGFR and Akt phosphorylation and a particularly significant effect of the combined therapy on pErk1/2 (Fig. 3B and Supplementary Fig. 2D).
Figure 3Effects of anti-EGF VacAbs in combination with EGFR TKIs in EGFR-mutant cell lines. Western blot analysis of selected markers in PC9 and other cell lines. Antibodies were added at a 1/50 dilution and EGF at 10 ng/mL. Medium was RPMI + 0.5% HS in all cases. Final concentrations of EGFR TKIs were 40 nmol/L gefitinib (A), 25 nmol/L osimertinib (B,C), 2.5 μmol/L gefitinib (E), and 1 μmol/L osimertinib (D–F). Results shown are a representative of three different experiments. Medium was RPMI + 0.5% HS, incubation time 2 hours. EGF, epidermal growth factor; VacAbs, antibodies created by vaccination; TKI, tyrosine kinase inhibitor; C-Ab, control antibodies; Ab, anti-EGF VacAbs; RPMI, radioimmunoprecipitation assay; HS, human serum.
Next, we studied PC9-GR4 and PC9-OR4 cells treated with osimertinib. Again, hEGF reduced the inhibitory effects of the EGFR TKI on pErk1/2 (Fig. 3C and D, compare lanes “Osimertinib” and “Osimertinib + EGF”). Also, in presence of the growth factor, the phosphorylation Erk1/2 or EGFR was completely or almost completely suppressed only when osimertinib was combined with anti-EGF VacAbs (Fig. 3C and D, compare the three right lanes).
Two additional EGFR-mut cell lines, PC9-GR2 and PC9-GR4AZD1, were tested to validate these relevant findings. PC9-GR2 is a p.T790M-negative resistant cell line with AXL overexpression, whereas PC9-GR4AZD1 is an osimertinib-resistant line derived from PC9-GR4 (Supplementary Table 1). Western blot analyses showed that the VacAbs were able to significantly enhance the effects of osimertinib and gefitinib on EGFR and Erk1/2 activation also in these cell lines (Fig. 3E and F).
Sera From Patients Immunized With an EGF Vaccine Suppress the Activation of EGFR Downstream Pathways
The results described so far had been generated with anti-EGF VacAbs raised in rabbit. To determine whether sera from patients immunized with an anti-EGF vaccine have similar in vitro biological activities, a series of Western blot experiments were performed. First, we tested sera collected from patients before immunization; a representative result is presented in Figure 4. Pre-immunization sera were unable to block the EGF-induced phosphorylation of EGFR or Erk1/2 in PC9 cells and strongly induced Akt activation, being significantly more potent than EGF itself. In contrast, the four sera tested from patients immunized with the anti-EGF vaccine were able to suppress pErk1/2 in a dose-dependent way, although we observed a significant inter-individual variability (Fig. 4). Inhibition of EGFR activation was also apparent in most cases, again with significant differences among patients that did not correlate with those observed for pErk1/2 levels. Some sera from immunized patients induced Akt phosphorylation, although the extent of this effect was significantly lower than in the case of the control sera. Finally, the titers of the anti-EGF antibodies were measured by ELISA in the four immunized patients. They ranged from 1/16,000 to 1/64,000 and did not associate with the biological effects of the corresponding sera on the EGFR pathway.
Figure 4Effects of sera of patients in EGFR-mutant cell lines. Western blot analysis of selected markers. (A,B) Sera from patients immunized with an anti-EGF vaccine. Titres of anti-EGF antibodies, as quantified by enzyme-linked immunosorbent assay, were as follows: patient 2, 1/64000; patient 4, 1/16000; patient 3, 1/64000; and patient 7, 1/32000. (C) Representative serum from a control individual. (D) Quantification of the bands of phosphorylated proteins. Medium was radioimmunoprecipitation assay, concentration of EGF 10 ng/mL, incubation time 2 hours. EGF, epidermal growth factor.
Anti-EGF VacAbs Inhibit the Expression or Activation of Proteins Related to Resistance to EGFR TKIs
Anti-EGF antibodies have been shown to persist at significant titers for months in the blood of vaccinated patients. To gain insight into the effects of long-term exposure at the molecular level, EGFR-mut cells were treated with rabbit anti-EGF VacAbs, EGFR TKIs, or a combination, for 24 hours, and signal transducer and activator of transcription 3 (STAT3) was added to the panel of proteins analyzed. Western blotting of cell extracts did not reveal down-regulation of total EGFR, Erk1/2, or Akt in any case, but showed continued inhibition of pEGFR, pAkt, and pErk1/2 with the combined treatment. Regarding STAT3, gefitinib, erlotinib, and osimertinib were shown to induce the phosphorylation of this protein, but this effect was consistently suppressed by the addition of anti-EGF VacAbs in PC9, PC9-GR4, and PC9-OR4 cells (Fig. 5A–D).
Figure 5Effects of anti-EGF VacAbs in combination with EGFR TKIs in EGFR-mutant cell lines. (A–D) Western blot analysis of EGFR, Erk1/2, Akt, and STAT3 after a 24-hour incubation. (E) Western blot analysis of markers related to emergence of resistance in PC9. Results shown are a representative of three different experiments. Medium was RPMI + 0.5% HS, incubation time 24 hours, EGF was added at 10 ng/mL. Final concentrations of EGFR TKIs were 40 nmol/L gefitinib (A,E), 25 nmol/L erlotinib (B), 25 nmol/L osimertinib (C), and 1 μmol/L osimertinib (D). EGF, epidermal growth factor; VacAbs, antibodies created by vaccination; TKI, tyrosine kinase inhibitor; C-Ab, control antibodies; Ab, anti-EGF VacAbs; RPMI, radioimmunoprecipitation assay; HS, human serum.
We next investigated the effect of anti-EGF VacAbs in other proteins involved in the emergence of resistance to EGFR-TKIs. When treating PC9 cells for 24 hours, anti-EGF VacAbs induced a reduction in the levels of NOTCH-3 cleaved, hes family bHLH transcription factor 1 (HES-1), BMI1 proto-oncogene, polycomb ring finger (BMI1), and, particularly, AXL, that was maintained when the cells were incubated with the combination of anti-EGF VacAbs + gefitinib (Fig. 5E). No changes were observed in other proteins such as mammalian target of rapamycin, phospho–mammalian target of rapamycin, or aldehyde dehydrogenase 1 family member A1 (not shown). Also, C-Abs did not have any effect on AXL after 24 hours of incubation (Fig. 6A). Additional Western blotting and gene expression analyses showed a consistent decrease of AXL levels in PC9 and PC9-GR4 cells after 24 to 48 hours of treatment with anti-EGF VacAbs, which was maintained when the VacAbs were combined with gefitinib, afatinib, or osimertinib (Fig. 6B,Supplementary Fig. 3A–C). In the case of the PC9-OR4 and H1975 cells, the effect of VacAbs on AXL levels was less pronounced or nonexistent, although down-regulation or HES1 and/or BMI1 could be observed in some cases (Supplementary Fig. 3D and E).
Figure 6Effects of anti-EGF VacAbs in on emergence of resistance to EGFR TKIs. (A) Western blot analysis of AXL; medium was RPMI + 0.5% HS, incubation time was 24 hours. (B) Results of the AXL mRNA expression analysis, incubation times were 24 and 48 hours. *p < 0.05. (C) Hypothetical pathway explaining AXL downregulation by anti-EGF VacAbs. (D,E) Emergence of resistant colonies to gefitinib and afatinib in PC9 under different conditions. Medium was RPMI + 10% fetal bovine serum. EGF, epidermal growth factor; VacAbs, antibodies created by vaccination; TKI, tyrosine kinase inhibitor; C-Ab, control antibodies; Ab, anti-EGF VacAbs; RPMI, radioimmunoprecipitation assay; HS, human serum.
Anti-EGF VacAbs Delay In Vitro the Emergence of Resistance to EGFR-TKIs
The results obtained in the previous experiments prompted us to investigate if the anti-EGF VacAbs could prevent the emergence of resistance to EGFR TKIs. In a first, preliminary experiment, low confluence PC9 cultures were treated in 96-well plates with gefitinib (40 nmol/L), anti-EGF VacAbs (1/25 dilution), or a combination, and wells reaching 50% confluence were scored three times per week. Gefitinib-resistant colonies began to appear after 1 week and were present in all wells after 7 weeks. When anti-EGF VacAbs were added, the first colony-positive wells emerged after 3 weeks and the appearance of resistant colonies in all wells took 20 weeks. In a second experiment, we used afatinib, anti-EGF VacAbs, C-Abs, and combinations at two different concentrations. We found that anti-EGF VacAbs were very effective in preventing the emergence of resistance to afatinib (Fig. 6D and E, Supplementary Fig. 4). At 25 nmol/L, resistant colonies had appeared in half of the wells after 5 weeks, compared to the 11 and 14 weeks needed when anti-EGF VacAbs were present at 1/25 and 1/10 dilutions, respectively. C-Abs at 1/10 had the opposite effect and accelerated the appearance of colony-positive wells. The differences were statistically significant in a long rank test.
Discussion
NSCLC patients with EGFR mutations are less likely to derive benefit from immune checkpoint inhibitors. In three recent clinical trials of second-line docetaxel versus nivolumab, pembrolizumab, and atezolizumab, the inhibitors failed to improve the overall survival of EGFR-mut patients progressing to cytotoxic chemotherapy or EGFR TKIs.
Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non–small-cell lung cancer (KEYNOTE-010): a randomised controlled trial.
Atezolizumab versus docetaxel in patients with previously treated non–small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised controlled trial.
This inferior response can be a consequence of the un-inflamed microenvironment present in EGFR-mut tumors, with low programmed death ligand 1 expression and lack of T cell infiltration.
In consequence, the development of new immunotherapeutic approaches for EGFR-mut NSCLC patients can be of particular interest.
Vaccination against EGF constitutes a novel strategy that, contrary to programmed death 1 blockade, is not intended at reversing tumor-induced immunosuppression by activating the T cells. Instead, it aims to stimulate B cells to produce neutralizing antibodies that sequester circulating EGF, thus preventing its binding to EGFR. Vaccination against EGF, also referred to EGF–pathway targeted immunization, is well tolerated, generates few cases of severe adverse effects, and has shown promising results in two trials enrolling unselected advanced NSCLC patients.
A phase III clinical trial of the epidermal growth factor vaccine CIMAvax-EGF as switch maintenance therapy in advanced non-small cell lung cancer patients.
However, little was known about the molecular and cellular mechanisms involved in the effects of anti-EGF antibodies besides their capability to block ligand binding and phosphorylation of EGFR, or about their differential activity in tumors with EGFR mutations or other genetic alterations.
Effective inhibition of the epidermal growth factor/epidermal growth factor receptor binding by anti-epidermal growth factor antibodies is related to better survival in advanced non-small-cell lung cancer patients treated with the epidermal growth factor cancer vaccine.
Our study shows that anti-EGF VacAbs raised in rabbits suppress the effects of EGF on cell proliferation, cell cycle, and signal transduction pathways in EGFR-mut NSCLC cell lines, particularly in those derived from untreated patients. The concentrations of EGF used in our experiments (10 ng/mL) are close to those reported in HS, which have a median around 1 ng/mL and show a significant inter-individual variability.
A phase III clinical trial of the epidermal growth factor vaccine CIMAvax-EGF as switch maintenance therapy in advanced non-small cell lung cancer patients.
Remarkably, the anti-EGF VacAbs were also found to consistently reduce the levels of pErk1/2 in absence of exogenous EGF not only in PC9 cells (Fig. 3A and B, compare lanes “C” and “Ab”), but also in PC9-GR4 cells, where the growth factor did not show significant effects (Fig. 3C, compare lanes “C” and “Ab”). One of the possible explanations for this observation could be the existence of receptor/ligand feedback loops in the cell lines used. Sera from patients immunized with an anti-EGF vaccine were also shown to efficiently block the activation of pErk1/2 by EGF. Control sera from nonimmunized patients had little effect on Erk1/2 but strongly activated Akt, indicating that the blood of healthy individuals contains growth factors that specifically trigger pAkt in EGFR-mut cells. In contrast, the sera from the four immunized patients analyzed were less active in inducing Akt phosphorylation (Fig. 4D). Significant differences were observed in the potency of the sera from vaccinated individuals to block Erk1/2 phosphorylation and, to a lesser extent, to activate Akt (Fig. 4D, compare patients #2 and #7). Being a new therapeutic approach with only a phase III trial completed, the availability of samples from patients vaccinated against EGF is limited, and we could not gain access to additional sera to further explore this variability or to correlate it with clinical outcomes.
Next, we discovered that EGF significantly reduced the antiproliferative effects of gefitinib, erlotinib, afatinib, and osimertinib in several EGFR-mut NSCLC cells, both sensitive and resistant to EGFR TKIs. This finding correlated with the results of Western blotting experiments where the levels of pErk1/2 in cells treated with EGFR TKIs were significantly higher if EGF was present (Fig. 3A–D, compare lanes “TKI” with “TKI+EGF”). To the best of our knowledge, this is the first time that the effects of EGF on the sensitivity of EGFR-mut cells to TKIs are investigated. In the initial report describing the development of gefitinib, EGF was found to have the opposite effect in the two EGFR-wild-type (wt) cell line models used, the human umbilical vein endothelial cell endothelial and the KB squamous carcinoma cells, where the concentration that inhibits 50% of the drug decreased approximately 100 times in the presence of EGF.
Our finding suggests that EGFR-mut patients with high EGF levels might have worse outcomes to EGFR TKIs, and additional studies are warranted to validate this hypothesis. Increased serum levels of two EGFR ligands, transforming growth factor alpha and amphiregulin, have been reported to correlate with worse responses to EGFR TKIs in unselected NSCLC patients.
Ligands of epidermal growth factor receptor and the insulin-like growth factor family as serum biomarkers for response to epidermal growth factor receptor inhibitors in patients with advanced non–small cell lung cancer.
Increases of amphiregulin and transforming growth factor-alpha in serum as predictors of poor response to gefitinib among patients with advanced non–small cell lung cancers.
Regarding EGF, in the only study published so far, which included 11 EGFR-mut and 21 EGFR-wt NSCLC patients treated with erlotinib, those with high concentration of EGF in serum had shorter progression-free survival, although the differences did not reach statistical significance.
Pretreatment levels of the serum biomarkers CEA, CYFRA 21-1, SCC and the soluble EGFR and its ligands EGF, TGF-alpha, HB-EGF in the prediction of outcome in erlotinib treated non–small-cell lung cancer patients.
The reduction in the efficacy of EGFR TKIs induced by EGF provided us with a rationale to test the effects of the addition of anti-EGF VacAbs. The combination of gefitinib, erlotinib, afatinib, and osimertinib with anti-EGF VacAbs showed a stronger antiproliferative effect than the EGFR TKIs alone in the EGFR-mut cell lines tested (Fig. 2), a finding that correlated with a consistent decrease in pErk1/2 (Fig. 3 and Supplementary Fig. 2D, compare lanes TKI + EGF with TKI + Ab + EGF). The combination was also superior in blocking the anti-apoptotic and G2/M stimulating effects of EGF.
The anti-EGFR monoclonal antibody cetuximab has also been tested in EGFR-mut cell line models. Similarly to anti-EGF VacAbs, cetuximab blocks ligand binding in vitro and has been shown to prevent ligand-induced EGFR, Erk1/2, and Akt phosphorylation in PC9 and H1975 cells.
Regarding receptor down-regulation, there seems to be significant differences between the two antibodies. Cetuximab has been shown to markedly decrease the levels of total EGFR after 1 to 2 hours of incubation in EGFR-mut cells such as PC9, H1975 or H3255, whereas anti-EGF VacAbs did not induce any significant down-regulation of the receptor after 24 hours of incubation (Fig. 5).
Finally, although cetuximab has been reported to amplify the induction of apoptosis and tumor regression in EGFR-wt, head and neck cancer cell lines, and subcutaneous tumors, it failed to enhance the effects of gefitinib in PC9 xenografts.
In contrast, we found that the anti-EGF VacAbs potentiated the antiproliferative activity of EGFR TKIs in PC9 and the rest of EGFR-mut cell lines tested. This potentiating effect reached statistical significance in all cases, with the only exception being PC9-GR4 cells (Fig. 2). The fact that the anti-EGF VacAbs target the ligand instead of the receptor and do not induce EGFR down-regulation might offer a possible explanation for the differences found between the effects of cetuximab and anti-EGF antibodies. One of the limitations of our work was that we could not perform xenograft studies, which would have offered valuable information about the effects of the anti-EGF VacAbs in vivo. Xenografts models involve athymic mice, which cannot be easily vaccinated against EGF due to their defective immune system.
Finally, we examined the effects of prolonged exposure of EGFR-mut cells to TKIs and anti-EGF VacAbs. We found that the addition of the anti-EGF VacAbs significantly delayed the appearance of clones resistant to gefitinib and afatinib in the PC9 cell line. Cetuximab has also been reported to delay the emergence of resistance to erlotinib and afatinib in animal models of EGFR-mut lung adenocarcinoma but, again, we found some differences between the mechanisms of action of the two antibodies.
Afatinib plus cetuximab delays resistance compared to single-agent erlotinib or afatinib in mouse models of TKI-naive EGFR L858R-induced lung adenocarcinoma.
The addition of anti-EGF VacAbs consistently suppressed this TKI-induced STAT3 activation. In contrast, pSTAT3 has been shown to be elevated in head and neck human tumors progressing to cetuximab, suggesting that STAT3 activation is involved in resistance to this drug.
We also found that the anti-EGF VacAbs reduced the total levels of other proteins associated with the emergence of resistance to EGFR TKIs, such as AXL, BMI1, HES1, or NOTCH-3 cleaved in PC9 cells.
Cetuximab resistance in NSCLC is characterized by upregulation of AXL and EGFR ligands. In cetuximab-resistant cells, increased EGFR ligand production leads to AXL and EGFR heterodimerization which, in turn, induces the transcription of the AXL gene via Erk1/2.
In our case, because anti-EGF VacAbs capture EGF, preventing its binding to EGFR, we can hypothesize that they also block the formation of EGFR-AXL heterodimers and, consequently, the expression of AXL (Fig. 6C). Regarding HES1 and NOTCH-3 cleaved, they have been described to be elevated in EGFR-mut cells after EGFR TKI treatment and this event has been related to an increase in stem-like cells and emergence of resistance.
In summary, we have shown that anti-EGF VacAbs suppress the effects of EGF and significantly enhance the antitumor activity of EGFR TKIs in EGFR-mutated NSCLC cell lines. They also block STAT3 activation, reduce AXL expression, and delay the acquisition of resistance. Based on these findings, a phase I trial of an anti-EGF vaccine in combination with afatinib as a first-line therapy in EGFR-mut NSCLC patients has been initiated (EPICAL study).
Erlotinib versus standard chemotherapy as first-line treatment for European patients with advanced EGFR mutation-positive non–small-cell lung cancer (EURTAC): a multicentre, open-label, randomised phase 3 trial.
Standard-dose versus high-dose conformal radiotherapy with concurrent and consolidation carboplatin plus paclitaxel with or without cetuximab for patients with stage IIIA or IIIB non-small-cell lung cancer (RTOG 0617): a randomised, two-by-two factorial phase 3 study.
Cetuximab and first-line taxane/carboplatin chemotherapy in advanced non–small-cell lung cancer: results of the randomized multicenter phase III trial BMS099.
Herbst R, Redman M, Kim E. A randomized, phase III study comparing carboplatin/paclitaxel or carboplatin/paclitaxel/bevacizumab with our without concurrent cetuximab in patients with advanced non–small cell lung cancer (NSCLC): SWOG S0819. Presented at the 2015 World Conference on Lung Cancer, Denver, Colorado.
Continued use of afatinib with the addition of cetuximab after progression on afatinib in patients with EGFR mutation-positive non-small-cell lung cancer and acquired resistance to gefitinib or erlotinib.
A phase III clinical trial of the epidermal growth factor vaccine CIMAvax-EGF as switch maintenance therapy in advanced non-small cell lung cancer patients.
Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non–small-cell lung cancer (KEYNOTE-010): a randomised controlled trial.
Atezolizumab versus docetaxel in patients with previously treated non–small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised controlled trial.
Effective inhibition of the epidermal growth factor/epidermal growth factor receptor binding by anti-epidermal growth factor antibodies is related to better survival in advanced non-small-cell lung cancer patients treated with the epidermal growth factor cancer vaccine.
Ligands of epidermal growth factor receptor and the insulin-like growth factor family as serum biomarkers for response to epidermal growth factor receptor inhibitors in patients with advanced non–small cell lung cancer.
Increases of amphiregulin and transforming growth factor-alpha in serum as predictors of poor response to gefitinib among patients with advanced non–small cell lung cancers.
Pretreatment levels of the serum biomarkers CEA, CYFRA 21-1, SCC and the soluble EGFR and its ligands EGF, TGF-alpha, HB-EGF in the prediction of outcome in erlotinib treated non–small-cell lung cancer patients.
Afatinib plus cetuximab delays resistance compared to single-agent erlotinib or afatinib in mouse models of TKI-naive EGFR L858R-induced lung adenocarcinoma.
30592-6&id=gr1.jpg){kind=link}
30592-6&id=gr2.jpg){kind=link}
30592-6&id=gr3.jpg){kind=link}
30592-6&id=gr4.jpg){kind=link}
30592-6&id=gr5.jpg){kind=link}
30592-6&id=gr6.jpg){kind=link}
30592-6&id=figs1.jpg){kind=link}
30592-6&id=figs2.jpg){kind=link}
30592-6&id=figs3.jpg){kind=link}
30592-6&id=figs4.jpg){kind=link}