Purpose:

EGFR tyrosine kinase inhibitors (EGFR-TKI) benefit patients with advanced lung adenocarcinoma (ADC) harboring activating EGFR mutations. We aimed to identify biomarkers to monitor and predict the progression of patients receiving EGFR-TKIs via a comprehensive omic analysis.

Experimental Design:

We applied quantitative proteomics to generate the TKI resistance–associated pleural effusion (PE) proteome from patients with ADC with or without EGFR-TKI resistance. Candidates were selected from integrated genomic and proteomic datasets. The PE (n = 33) and serum (n = 329) levels of potential biomarkers were validated with ELISAs. Western blotting was applied to detect protein expression in tissues, PEs, and a cell line. Gene knockdown, TKI treatment, and proliferation assays were used to determine EGFR-TKI sensitivity. Progression-free survival (PFS) and overall survival (OS) were assessed to evaluate the prognostic values of the potential biomarkers.

Results:

Fifteen proteins were identified as potential biomarkers of EGFR-TKI resistance. Cadherin-3 (CDH3) was overexpressed in ADC tissues compared with normal tissues. CDH3 knockdown enhanced EGFR-TKI sensitivity in ADC cells. The PE level of soluble CDH3 (sCDH3) was increased in patients with resistance. The altered sCDH3 serum level reflected the efficacy of EGFR-TKI after 1 month of treatment (n = 43). Baseline sCDH3 was significantly associated with PFS and OS in patients with ADC after EGFR-TKI therapy (n = 76). Moreover, sCDH3 was positively associated with tumor stage in non–small cell lung cancer (n = 272).

Conclusions:

We provide useful marker candidates for drug resistance studies. sCDH3 is a survival predictor and real-time indicator of treatment efficacy in patients with ADC treated with EGFR-TKIs.

Translational Relevance

Mutated EGFR is a predictor of the initial treatment efficacy of EGFR–tyrosine kinase inhibitors (TKI) in advanced lung adenocarcinoma (ADC). Unfortunately, neither EGFR mutations nor molecular markers can accurately monitor cancer progression or predict patient survival. EGFR-TKI susceptibility associated biomarkers to benefit patients with lung ADC are in high demand. Pleural effusion (PE) has been recognized as a promising source for biomarker discovery. By integrating PE proteomic and tissue genomic analyses, we herein identified soluble cadherin-3 (sCDH3) as a marker for monitoring first/second-generation EGFR-TKI resistance in real time. Cadherin-3 (CDH3) knockdown enhances EGFR-TKI sensitivity in ADC cells. Importantly, the baseline serum sCDH3 level is associated with progression-free survival in patients with ADC who received EGFR-TKI therapy and with overall survival in patients with non–small cell lung cancer (NSCLC). Our results establish a biomarker dataset for EGFR-TKI resistance and prognosis. This study significantly advances the development of a liquid biopsy to benefit patients with NSCLC and provides new insights into EGFR-TKI resistance in lung ADC.

Lung cancer is the leading cause of cancer-related death worldwide (1). Non–small cell lung cancer (NSCLC) represents 80% of all lung cancers, and adenocarcinoma (ADC) is the most common subtype of NSCLC (2). EGFR mutations specifically target patients with lung ADC (∼60%) in East Asian countries, including Taiwan (3, 4). EGFR-tyrosine kinase inhibitors (EGFR-TKI), such as the first-generation TKIs erlotinib and gefitinib and the second-generation TKI afatinib, have been determined to benefit patients with ADC with EGFR-activating mutations (5). These response evaluations of targeted therapy are based on the RECIST guidelines (6), and an EGFR mutation is recognized as a good predictor for the initial efficacy of EGFR-TKI treatments, in which the response rates and disease control rates are 60%–70% and 85%–95%, respectively. These TKIs are associated with a median progression-free survival (PFS) time of 9–14 months compared with 5–7 months for platinum-based chemotherapy (7). Unfortunately, patients treated with first/second-generation EGFR-TKIs eventually develop acquired resistance to these agents after a median of 10–16 months that most commonly (∼50%–60% of cases) results from a secondary mutation at position 790 in EGFR (T790M; refs. 8, 9). Osimertinib, a third-generation irreversible EGFR-TKI that selectively inhibits both EGFR-TKI–sensitizing and EGFR T790M–resistance mutations (10), obtained FDA-accelerated approval in November 2015. Recent studies have also shown that osimertinib improves both the median PFS (18.9 vs. 10.2 months) and overall survival (OS; 38.6 vs. 31.8 months) compared with first-generation EGFR-TKIs in patients with treatment-naïve, EGFR mutation–positive (exon 19 deletion or L858R) advanced NSCLC (11, 12), suggesting that osimertinib will be a new standard of care in first-line EGFR-mutated NSCLC. T790M can be evaluated via either a tissue rebiopsy or circulating cell-free tumor DNA in the plasma, and its evaluation is currently mandatory after progression to first/second-generation EGFR-TKIs (13). However, neither cancer progression nor an improvement in the patient's OS is correlated with EGFR mutations. This suggests that even in patients with the same EGFR mutation status, the event of acquiring resistance and the timing of resistance are heterogeneous and unpredictable. To improve the prediction of the response to EGFR-TKI–targeted therapy in patients with ADC as well as their prognosis, certain serum molecules, such as carcinoembryonic antigen (CEA), CA125, CA19-9, CA27-29, and CYFRA 21-1, and cell-free circulating tumor DNA have been reported as potential biomarkers (14–16). Because of their limited sensitivity/specificity and sample size, the clinical significance of these potential markers is underevaluated.

Pleural effusion (PE), a tumor-proximal body fluid, is associated with lung malignancy and serves as a promising source for biomarker discovery (17, 18). Previously, we generated the differential PE proteomes from six types of exudative PEs via a label-free quantitative proteomics approach to facilitate the discovery of lung ADC biomarkers (19). This study aimed to search for serum biomarkers that could be used to monitor the treatment efficacy and prognosis of patients with advanced ADC with first/second-generation EGFR-TKI therapy in real time. We then established a drug resistance–associated PE database via isobaric tags for relative and absolute quantitation (iTRAQ)-based quantitative technology combined with high-throughput tandem mass spectrometry. We selected potential biomarkers by integrating our quantitative proteomic database with Gene Expression Profiling Interactive Analysis (GEPIA, http://gepia.cancer-pku.cn/) and the Human Protein Atlas (https://www.proteinatlas.org/) database. Cadherin-3 (CDH3), a promising marker candidate, was selected for further verification and validation using lung tissues, PEs, a cell line, and a large cohort of serum samples. Our results support that the baseline soluble CDH3 (sCDH3) level is a noninvasive biomarker for the diagnosis of late-stage NSCLC and a prognostic marker for patients with ADC receiving first/second-generation EGFR-TKI therapy.

Patient populations and clinical specimens

The acquired clinical samples were approved by the Institutional Review Board for Research Ethics at the Chang Gung Memorial Hospital (Linkou, Taoyuan, Taiwan; approval number: 99-3738B and 99-3671B). This study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice. Written informed consent was received from all patients before collection. For the discovery phase, malignant pleural effusion (MPE) from 23 patients with advanced lung ADC with EGFR mutations who received EGFR-TKI–targeted therapy with or without following chemotherapy regimens were included. PE from 10 patients with tuberculosis (TB) represents benign pulmonary disease. These 33 patients' PEs were divided into four groups (Fig. 1A; Supplementary Table S2): MPE obtained prior to treatment (before treatment, BT group) from 11 patients who achieved partial response (PR) for at least 6 months, MPE obtained on the day of failure from 8 patients who received EGFR-TKI treatment followed by chemotherapy [progressive disease; PD (T+C)], MPE obtained on the day of failure from 4 patients who received EGFR-TKI treatment [PD (T)], and PE from 10 patients with TB. For the verification phase, these 33 PEs used in the discovery phase were included in ELISA. To determine the sCDH3 level via ELISA, serum samples were obtained from 272 treatment-naïve NSCLC patients (baseline) and 57 healthy controls (Table 1). Specifically, 23 MPE and 76 serum samples from patients with ADC with mutated EGFR and who received first/second-generation EGFR-TKIs were collected between 2009 and 2016. Among these 272 patients with NSCLC, serum samples collected 1 month (29 ± 8 days) after EGFR-TKI treatment from 43 patients with ADC (28 PR and 15 non-PR) with EGFR mutations were included to examine alterations in sCDH3 levels upon EGFR-TKI treatment. To detect CDH3 expression in tissues via Western blotting, cancerous and adjacent normal tissues from 11 ADC patients was included (Supplementary Table S3). The PE and serum samples were centrifuged at 2,000 × g for 15 minutes at 4°C. The supernatants were transferred to a new tube and stored at −80°C for further analysis or verification. The tissue samples were collected during surgical resection and stored at −80°C until analysis.

Figure 1.

Experimental design and workflow used for the discovery of EGFR-TKI resistance–associated biomarkers in lung ADC harboring EGFR mutations. A, Allotment and grouping of patients used in the discovery phase. MPE from ADC and PE from TB were divided into four groups: 11 MPEs with PR collected before treatment, eight MPEs with ongoing PD (T+C), four MPEs with ongoing PD (T), and 10 PEs with TB. The MPE and PE of each group were pooled and prepared for iTRAQ labeling. B, MPE and PE samples allotted as shown in (A) were used for proteomic analysis, which comprises prefractionation by the removal of high-abundance proteins, chemical labeling by iTRAQ, and 2D LC/MS-MS analysis. Differentially expressed proteins were selected by integrating the proteomic and genomic datasets. The cancer tissue, PE, and serum samples were included in the verification and validation phase via immunodetection methods. A survival analysis was applied to evaluate the clinical significance of the potential biomarkers. ADC, adenocarcinoma; (T), EGFR-TKI; (T+C), EGFR-TKI and chemotherapy; Tx, treatment.

Figure 1.

Experimental design and workflow used for the discovery of EGFR-TKI resistance–associated biomarkers in lung ADC harboring EGFR mutations. A, Allotment and grouping of patients used in the discovery phase. MPE from ADC and PE from TB were divided into four groups: 11 MPEs with PR collected before treatment, eight MPEs with ongoing PD (T+C), four MPEs with ongoing PD (T), and 10 PEs with TB. The MPE and PE of each group were pooled and prepared for iTRAQ labeling. B, MPE and PE samples allotted as shown in (A) were used for proteomic analysis, which comprises prefractionation by the removal of high-abundance proteins, chemical labeling by iTRAQ, and 2D LC/MS-MS analysis. Differentially expressed proteins were selected by integrating the proteomic and genomic datasets. The cancer tissue, PE, and serum samples were included in the verification and validation phase via immunodetection methods. A survival analysis was applied to evaluate the clinical significance of the potential biomarkers. ADC, adenocarcinoma; (T), EGFR-TKI; (T+C), EGFR-TKI and chemotherapy; Tx, treatment.

Close modal
Table 1.

Demographics of 329 patients used to detect sCDH3 levels in serum via ELISA.

Healthy controlADCNon-ADC NSCLC
Variablen = 57n = 206n = 66
Sex 
 Female 26 (45.6%) 91 (44.2%) 10 (15.2%) 
 Male 31 (54.4%) 115 (56.8%) 56 (84.8%) 
Age 58.80 ± 15.82 61.60 ± 11.79 64.57 ± 10.56 
Smoke 
 No 42 (73.7%) 152 (73.8%) 16 (24%) 
 Yes 15 (26.3%) 54 (26.2%) 50 (76%) 
Stage 
 Early  77 (37.4%) 21 (31.8%) 
 Late  129 (62.6%) 45 (68.2%) 
EGFR status in late-stage ADC 
 WT  53 (41.1%)  
 Mutant  76 (58.9%)  
Healthy controlADCNon-ADC NSCLC
Variablen = 57n = 206n = 66
Sex 
 Female 26 (45.6%) 91 (44.2%) 10 (15.2%) 
 Male 31 (54.4%) 115 (56.8%) 56 (84.8%) 
Age 58.80 ± 15.82 61.60 ± 11.79 64.57 ± 10.56 
Smoke 
 No 42 (73.7%) 152 (73.8%) 16 (24%) 
 Yes 15 (26.3%) 54 (26.2%) 50 (76%) 
Stage 
 Early  77 (37.4%) 21 (31.8%) 
 Late  129 (62.6%) 45 (68.2%) 
EGFR status in late-stage ADC 
 WT  53 (41.1%)  
 Mutant  76 (58.9%)  

Removal of high-abundance proteins from PE

The PE from each group of patients was pooled and subjected to the depletion of 14 high-abundance proteins using a Multiple Affinity Removal System (MARS) Affinity Column (Agilent Technologies) via ÄKTA purifier 10 fast performance liquid chromatography according to the manufacturer's instructions (GE Healthcare). Unbound fractions were collected, concentrated, desalted, and concentrated by centrifugation in Amicon Ultra-4 Tubes (Millipore). The resulting protein concentration was quantified with a Bradford Protein Assay (Bio-Rad Laboratories, Inc.).

In-solution digestion of proteins and iTRAQ labeling

The PE proteins were reduced with 5 mmol/L tris-(2-carboxyethyl) phosphine at 60°C and then alkylated with 10 mmol/L iodoacetamide for 30 minutes at 37°C. Proteins were digested overnight at 37°C with trypsin (Promega), and the peptides were extracted for further analysis. For iTRAQ labeling, the tryptic peptides were reconstituted in iTRAQ reagent buffer, and the four groups of PE samples were separately labeled with four different iTRAQ labeling reagents, as shown in Fig. 1A [114, BT; 115, PD(T+C); 116, PD(T); and 117, TB] according to the manufacturer's instructions (AB Sciex, Inc.).

Two-dimensional LC/MS-MS and protein database search

The two-dimensional (2D) LC/MS-MS analysis was performed as described previously (20, 21). The resulting MS-MS spectra were searched against the Swiss-Prot human sequence database using the Mascot search engine (Matrix Science, version 2.2.04). To quantify the identified proteins, the raw spectrometry data, including the reporter ions quantifier node for iTRAQ, were analyzed using Proteome Discoverer Software (version 1.4, Thermo Fisher Scientific).

ELISA

The sCDH3 protein levels in PE or serum samples were determined with a Sandwich ELISA Kit according to the manufacturer's instructions (R&D Systems). The dilution factors used for PE and serum samples in the ELISA were 1:50 and 1:25, respectively. The serum CEA level was detected with an ELISA Kit purchased from GenWay (GWB-BQK050) at a 1:3 dilution. All ELISAs were performed in duplicate for each sample.

Western blot analysis

Proteins extracted from tissue samples were separated by SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membranes. The CDH3 protein was detected using a monoclonal mouse antibody (R&D Systems, MAB861) followed by incubation with a horseradish peroxidase–conjugated secondary antibody and developed using a chemiluminescent substrate (Merck Millipore).

Cell culture and cell viability assay

The PE089 cell line was derived from a female patient with advanced ADC and an EGFR exon 19 deletion. This cell line was established and kindly provided by Professor Ko-Jiunn Liu (National Institute of Cancer Research, National Health Research Institutes, Tainan, Taiwan; ref. 22). Cells were cultured in Minimum Essential Media (Invitrogen) supplemented with 10% FBS under a humidified atmosphere of 95% air/5% CO2 at 37°C. Mycoplasma free in cell cultures was determined by DNA fluorochrome staining with Hoechst 33258 bisbenzimide. For CDH3 gene knockdown, 19-nucleotide RNA duplexes targeting human CDH3 were synthesized and annealed by Dharmacon (Thermo Fisher Scientific). Cells were transfected with a control siRNA and a pooled siRNA targeting CDH3 (UGAAUCAGCUCAAGUCUAA; GUGACAACGUCUUCUACUA; GAAAUCGGCAACUUUAUAA; and GAGGGUGUCUUCGCUGUAG) using Lipofectamine RNAiMAX Reagent (Invitrogen) according to the manufacturer's protocol. After 24 hours, the cells were reseeded into a 24-well plate with or without gefitinib (Tocris), and a cell viability assay was performed using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenytetrazolium bromide colorimetric growth assay as described previously (21).

Statistical analysis

All data were processed using SPSS 12.0 (SPSS, Inc.). All continuous variables are presented as mean ± SD. The nonparametric Mann–Whitney U test or Student t test was used to compare the protein levels in the serum between two groups. To compare more than two continuous variables, one-way ANOVA was applied. Survival rates were obtained using the Kaplan–Meier method and were compared using the log-rank test via Prism software (GraphPad). A P value less than 0.05 was considered statistically significant.

Supplementary data

Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).

Generation of drug resistance–associated proteomic datasets from lung ADC MPE

On the basis of the experimental design shown in Fig. 1, we depleted the abundant proteins from the four groups of PE samples with a MARS affinity column. The depletion efficacy and differential protein pattern of the fractionated PE samples were examined by SDS-PAGE with silver staining (Supplementary Fig. S1). The PE proteins were digested and labeled with iTRAQ reagent followed by 2D LC/MS-MS. The reporter ion intensities of 114 (BT), 115 [PD(T+C)], 116 [PD(T)], and 117 (TB) represent the expression levels of the identified proteins in each group of PE samples. Accordingly, 694 and 658 quantified proteins with high-confidence identification were obtained from two independent analyses (Supplementary Fig. S2A). In total, 561 quantified proteins were repeatedly identified from the two independent proteomics analyses. The reproducibility of protein quantification obtained from these two experiments was confirmed by Pearson correlation (Supplementary Fig. S2B). Detailed information for proteins identified in duplicate experiments of four PE samples is shown in Supplementary Table S1. To verify the differential protein levels in these four PE groups, HGFAC and CDH3 were selected for quantification by ELISA (Supplementary Table S2). We observed that the relative protein levels detected by ELISA were consistent with those obtained from our mass spectrometry analysis, supporting the reliability of our proteomic datasets.

Selection of potential EGFR-TKI resistance–associated biomarkers

To select potential EGFR-TKI resistance–associated biomarkers, 98 differentially expressed proteins with a 116/114 iTRAQ ratio greater than 1.5 were selected. By integrating these 98 candidates with 113 malignancy-associated proteins (117/114 >1.5), we chose 15 proteins for further analysis (Supplementary Fig. S3A). We next searched for the gene expression profiles of these 15 candidate proteins based on the public GEPIA dataset, and we gained seven candidates with a 1.5-fold increase at the mRNA level in tumor (T) tissue compared with normal (N) lung tissues (Table 2; Supplementary Fig. S3B). Notably, CDH3 was the most upregulated gene, with the highest T/N ratio of 11.34. We also found that six of these 15 candidate proteins, namely, CDH3, MET, CDH1, FAM3C, HSP90B1, and NPEPPS, exhibited high to moderate expression patterns in lung cancer tissues via IHC (Table 2). We then applied Western blot analysis to confirm that CDH3 was indeed overexpressed in lung ADC tissues compared with their adjacent normal tissues (Fig. 2A). Consistent with our proteomic discovery, Western blot analysis revealed a differential level of sCDH3, a protein band with an appropriate molecular weight of 80 kDa in different groups of PE samples (Fig. 2B). Therefore, the combined database analysis allowed us to narrow down the biomarker candidates efficiently, and sCDH3 was selected as our top priority for further validation.

Table 2.

List of EGFR-TKI resistance–associated proteins identified in this study.

Protein nameGene symbol116/114a114/117aT/N mRNA ratiobProtein levelc
Cadherin-3 CDH3 5.84 1.62 11.34 
Hepatocyte growth factor receptor MET 1.84 3.23 4.23 
Cadherin-1 CDH1 1.67 1.88 4.11 
Protein FAM3C FAM3C 2.03 1.69 1.53 
Endoplasmin HSP90B1 1.64 1.5 1.42 
Puromycin-sensitive aminopeptidase NPEPPS 2.45 2.02 0.78 
Alpha-amylase 1 AMY1A 1.54 3.92 4.00 − 
Fibrinogen beta chain FGB 3.21 1.87 3.67 − 
Fibrinogen gamma chain FGG 3.73 2.09 2.21 − 
Multiple EGF-like domains protein 8 MEGF8 1.95 2.01 0.84 − 
Deleted in malignant brain tumors 1 protein DMBT1 11.24 2.28 0.54 − 
Coagulation factor XIII A chain F13A1 2.21 1.77 0.47 − 
Complement component C8 beta chain C8B 1.91 2.13 0.18 − 
Apolipoprotein A-I APOA1 1.51 1.58 0.1 − 
Apolipoprotein B-100 APOB 2.95 2.79 N.D.d − 
Protein nameGene symbol116/114a114/117aT/N mRNA ratiobProtein levelc
Cadherin-3 CDH3 5.84 1.62 11.34 
Hepatocyte growth factor receptor MET 1.84 3.23 4.23 
Cadherin-1 CDH1 1.67 1.88 4.11 
Protein FAM3C FAM3C 2.03 1.69 1.53 
Endoplasmin HSP90B1 1.64 1.5 1.42 
Puromycin-sensitive aminopeptidase NPEPPS 2.45 2.02 0.78 
Alpha-amylase 1 AMY1A 1.54 3.92 4.00 − 
Fibrinogen beta chain FGB 3.21 1.87 3.67 − 
Fibrinogen gamma chain FGG 3.73 2.09 2.21 − 
Multiple EGF-like domains protein 8 MEGF8 1.95 2.01 0.84 − 
Deleted in malignant brain tumors 1 protein DMBT1 11.24 2.28 0.54 − 
Coagulation factor XIII A chain F13A1 2.21 1.77 0.47 − 
Complement component C8 beta chain C8B 1.91 2.13 0.18 − 
Apolipoprotein A-I APOA1 1.51 1.58 0.1 − 
Apolipoprotein B-100 APOB 2.95 2.79 N.D.d − 

a114, 115, 116, and 117 represent the reporter ions quantifier node for iTRAQ.

bmRNA levels in tumor (T) and normal (N) tissues derived from the GEPIA database.

cPlus symbol (+) and minus symbol (−) represent proteins with and without high to moderate expression patterns (high + medium > 50%) in lung cancerous tissues based on IHC results, respectively. This information was obtained from the Human Protein Atlas database.

dThe fold change of mRNA levels in tumor and normal parts were unable to be estimated on the basis of GEPIA database.

Figure 2.

Validation of serum sCDH3 as an early indicator of treatment efficacy and a potential prognostic marker in patients with ADC receiving EGFR-TKI–targeted therapy. A, CDH3 was overexpressed in lung ADC tissues. Proteins from lung ADC tissues (pooled samples from 2 or 3 patients) and their adjacent normal tissues were prepared and subjected to Western blot analysis using an anti-CDH3 antibody. Actin was used as a loading control. E, early-stage; L, late-stage; N, adjacent normal; T, tumor. B, Detection of sCDH3 in PE samples via Western blot analysis. The crude and pooled MPE or PE samples used in the discovery phase were subjected to SDS-PAGE, transferred to PVDF membranes, and stained with Fast Green FCF dye as a loading control (left). The proteins were then detected with an anti-CDH3 antibody. *, albumin in crude PEs. C and D, Validation of sCDH3, but not CEA, as an early indicator of EGFR-TKI efficacy by ELISA. Serum samples were collected prior to EGFR-TKI treatment (before treatment, BT) and approximately 1 month (29±8 days) after treatment (AT). Compared with baseline, sCDH3 was significantly decreased in PR patients (n = 28) but not non-PR (n = 15) patients. The CEA level was not correlated with treatment efficacy. N.S., not significant; ***, P value less than 0.0001 indicates significance based on the paired t test. E, CDH3 knockdown promoted gefitinib sensitivity in PE089 cells. PE089 cells were transfected with control siRNA or CDH3 siRNA. After 24 hours, cells were reseeded onto a 12-well plate and subjected to Western blot analysis at different time points to examine gene knockdown efficacy. Simultaneously, cells were subjected to a viability assay via the colorimetric growth method. EGFR-TKI sensitivity was determined after incubation with gefitinib for 5 days. F and G, A high baseline level of sCDH3 was positively associated with poor PFS in patients with lung ADC receiving EGFR-TKI–targeted therapy. The serum sCDH3 levels obtained from the ELISA were used to perform the survival analysis of 76 patients with ADC with EGFR mutations who received EGFR-TKI therapy (F) and 53 patients with ADC with EGFR WT who received chemotherapy (G). H, A high baseline sCDH3 level was significantly associated with poor OS in patients with NSCLC. The serum sCDH3 levels obtained from the ELISA were used to perform the survival analysis of 272 patients with NSCLC. The patients were stratified into two groups (high vs. low CDH3 level) using a protein concentration of 12 μg/mL as the cut-off value (F–H). P < 0.05 obtained from the log-rank test indicated statistical significance.

Figure 2.

Validation of serum sCDH3 as an early indicator of treatment efficacy and a potential prognostic marker in patients with ADC receiving EGFR-TKI–targeted therapy. A, CDH3 was overexpressed in lung ADC tissues. Proteins from lung ADC tissues (pooled samples from 2 or 3 patients) and their adjacent normal tissues were prepared and subjected to Western blot analysis using an anti-CDH3 antibody. Actin was used as a loading control. E, early-stage; L, late-stage; N, adjacent normal; T, tumor. B, Detection of sCDH3 in PE samples via Western blot analysis. The crude and pooled MPE or PE samples used in the discovery phase were subjected to SDS-PAGE, transferred to PVDF membranes, and stained with Fast Green FCF dye as a loading control (left). The proteins were then detected with an anti-CDH3 antibody. *, albumin in crude PEs. C and D, Validation of sCDH3, but not CEA, as an early indicator of EGFR-TKI efficacy by ELISA. Serum samples were collected prior to EGFR-TKI treatment (before treatment, BT) and approximately 1 month (29±8 days) after treatment (AT). Compared with baseline, sCDH3 was significantly decreased in PR patients (n = 28) but not non-PR (n = 15) patients. The CEA level was not correlated with treatment efficacy. N.S., not significant; ***, P value less than 0.0001 indicates significance based on the paired t test. E, CDH3 knockdown promoted gefitinib sensitivity in PE089 cells. PE089 cells were transfected with control siRNA or CDH3 siRNA. After 24 hours, cells were reseeded onto a 12-well plate and subjected to Western blot analysis at different time points to examine gene knockdown efficacy. Simultaneously, cells were subjected to a viability assay via the colorimetric growth method. EGFR-TKI sensitivity was determined after incubation with gefitinib for 5 days. F and G, A high baseline level of sCDH3 was positively associated with poor PFS in patients with lung ADC receiving EGFR-TKI–targeted therapy. The serum sCDH3 levels obtained from the ELISA were used to perform the survival analysis of 76 patients with ADC with EGFR mutations who received EGFR-TKI therapy (F) and 53 patients with ADC with EGFR WT who received chemotherapy (G). H, A high baseline sCDH3 level was significantly associated with poor OS in patients with NSCLC. The serum sCDH3 levels obtained from the ELISA were used to perform the survival analysis of 272 patients with NSCLC. The patients were stratified into two groups (high vs. low CDH3 level) using a protein concentration of 12 μg/mL as the cut-off value (F–H). P < 0.05 obtained from the log-rank test indicated statistical significance.

Close modal

Validation of serum sCDH3 as an early indicator of treatment efficacy and a potential prognostic marker in patients with ADC receiving EGFR-TKI–targeted therapy

To develop a noninvasive method for detecting the treatment efficacy of EGFR-TKI as early as possible, we determined the serum levels of sCDH3 at baseline and 1 month after EGFR-TKI treatment from 43 patients with ADC with EGFR mutations. Figure 2C shows that the serum levels of CDH3 in 28 PR patients were significantly reduced after EGFR-TKI treatment for 1 month (9.49 ± 4.55 vs. 5.54 ± 2.67; P < 0.0001), but no significant change was observed in 15 non-PR patients (15.43 ± 16.74 vs. 15.6 ± 20.24). Simultaneously, we observed a similar CEA level in these 28 PR patients' paired samples (Fig. 2D; Supplementary Table S4). This result indicated that serum sCDH3 could be a potential marker for the early monitoring of EGFR-TKI treatment efficacy. To test the potential role of CDH3 in EGFR-TKI resistance, we examined cell viability and EGFR-TKI sensitivity in CDH3-knockdown PE089 cells, an ADC cell line with mutated EGFR (exon 19 deletion). CDH3 knockdown promoted EGFR-TKI sensitivity (Fig. 2E; P = 0.009), but it did not affect cell viability or the phosphorylated AKT level in PE089 cells. Next, we performed a survival analysis to evaluate the prognostic significance of baseline sCDH3 in 76 patients with late-stage ADC harboring EGFR mutations. These 76 patients were stratified into two groups (high vs. low sCDH3 level) using a protein concentration of 12 μg/mL as the cut-off value. The survival analysis revealed that patients with a high serum sCDH3 level exhibited a shorter PFS than those with a low sCDH3 level (P = 0.0220). Moreover, a high level of sCDH3 was significantly associated with poor OS (P = 0.0408; Fig. 2F; Supplementary Table S5). Notably, we also observed that the baseline serum sCDH3 level in 53 patients with late-stage ADC with wild-type (WT) EGFR and undergoing chemotherapy was associated with OS (P = 0.0093) but not PFS (P = 0.1566; Fig. 2G; Supplementary Table S6). The statistical analysis showed that baseline sCDH3 was not correlated with sex, age, smoking status, TKI response, or EGFR-TKIs. Consistent with these results, a high CDH3 level was significantly associated with fewer treatment days (Table 3; P = 0.036). These results suggest that the baseline sCDH3 level could be used as a marker for the early detection of treatment efficacy and for monitoring cancer progression in patients with ADC with EGFR mutations. The baseline sCDH3 level can also be used to predict OS in all patients with ADC regardless of their EGFR gene mutation status.

Table 3.

Relationships between baseline sCDH3 serum levels and clinical characteristics in 129 patients with late-stage ADC.

VariableNumbersCDH3 (ng/mL)Pa
Sex 
 Female 64 13.34 ± 14.89 0.341 
 Male 65 14.22 ± 12.09  
Age 
 <60 62 12.71 ± 15.54 0.241 
 ≥60 67 11.93 ± 11.90  
Smoke 
 No 90 14.01 ± 11.33 0.390 
 Yes 39 13.26 ± 17.69  
Treatment 
 EGFR-TKI 76 12.84 ± 7.71 0.663 
 Chemotherapy 53 15.14 ± 18.97  
EGFR-TKI response 
 PR 54 13.01 ± 7.49 0.481 
 Non-PR 22 12.43 ± 8.41  
EGFR-TKI drug 
 Gefitinib 44 12.44 ± 6.65 >0.05b 
 Erlotinib 23 11.81 ± 7.30 >0.05b 
 Afatinib 17.43 ± 12.08 >0.05b 
TKI treatment days 
 <1 year 40 14.53 ± 8.65 0.036c 
 ≥1 year 36 10.91 ± 6.03  
Chemotherapy 
 PR 18 10.70 ± 5.33 0.241 
 Non-PR 35 17.42 ± 22.81  
Chemotherapy treatment days 
 <180 days 35 17.59 ± 22.88 0.280 
 ≥180 days 18 10.42 ± 4.24  
VariableNumbersCDH3 (ng/mL)Pa
Sex 
 Female 64 13.34 ± 14.89 0.341 
 Male 65 14.22 ± 12.09  
Age 
 <60 62 12.71 ± 15.54 0.241 
 ≥60 67 11.93 ± 11.90  
Smoke 
 No 90 14.01 ± 11.33 0.390 
 Yes 39 13.26 ± 17.69  
Treatment 
 EGFR-TKI 76 12.84 ± 7.71 0.663 
 Chemotherapy 53 15.14 ± 18.97  
EGFR-TKI response 
 PR 54 13.01 ± 7.49 0.481 
 Non-PR 22 12.43 ± 8.41  
EGFR-TKI drug 
 Gefitinib 44 12.44 ± 6.65 >0.05b 
 Erlotinib 23 11.81 ± 7.30 >0.05b 
 Afatinib 17.43 ± 12.08 >0.05b 
TKI treatment days 
 <1 year 40 14.53 ± 8.65 0.036c 
 ≥1 year 36 10.91 ± 6.03  
Chemotherapy 
 PR 18 10.70 ± 5.33 0.241 
 Non-PR 35 17.42 ± 22.81  
Chemotherapy treatment days 
 <180 days 35 17.59 ± 22.88 0.280 
 ≥180 days 18 10.42 ± 4.24  

aMann–Whitney U test.

bOne-way ANOVA.

cA P value less than 0.05 indicates significance based on Mann-Whitney U test.

Serum sCDH3 is positively associated with tumor stage and poor OS in NSCLC

To further explore the clinical application of baseline sCDH3 in NSCLC diagnosis and prognosis, we detected baseline sCDH3 in the sera from 200 additional individuals, including 57 healthy controls, 77 patients with early-stage ADC, and 66 patients with non-ADC NSCLC (Table 1). The relationships between sCDH3 levels and clinical characteristics are summarized in Table 4. The sCDH3 levels were higher in 272 patients with NSCLC than in 57 healthy controls (10.65 ± 10.95 ng/mL vs. 6.63 ± 2.84 ng/mL; P = 0.0007). The sCDH3 levels in 174 patients with late-stage NSCLC were significantly higher than those in 98 patients with early-stage NSCLC (12.91 ± 12.10 ng/mL vs. 6.63 ± 6.94 ng/mL; P < 0.001). The survival analysis showed that higher sCDH3 levels were significantly correlated with poorer OS (Fig. 2H; P < 0.0001). The multivariable analysis further confirmed that the sCDH3 level is an independent predictive factor of OS in NSCLC (Supplementary Table S7). Consistent with these findings, the sCDH3 level was not associated with the PFS of patients with NSCLC who received chemotherapy (Supplementary Table S8). These results collectively support the potential clinical applications of sCDH3 in NSCLC diagnosis and prognosis.

Table 4.

Relations between baseline sCDH3 serum levels and clinical characteristics in 57 healthy controls and 272 patients with NSCLC.

VariableNumberCDH3 (ng/mL)aP
Sex 
 Female 156 9.31 ± 10.96 0.292 
 Male 173 10.53 ± 9.32  
Age 
 <60 143 9.50 ± 8.58 0.939 
 ≥60 186 10.30 ± 11.20  
Smoke 
 No 210 9.69 ± 9.18 0.065 
 Yes 119 10.42 ± 11.67  
Healthy vs. NSCLC 
 Healthy control 57 6.63 ± 2.84 0.0007 
 NSCLC 272 10.65 ± 10.95  
Lung cancer histology 
 Healthy 57 6.63 ± 2.84  
 NSCLC–non ADC 66 9.85 ± 7.24 <0.05b 
 ADC 206 10.90 ± 11.90 <0.01c 
Cancer stage 
 Healthy 57 6.63 ± 2.84  
 Early stage (stage I/II) 98 6.63 ± 6.94 <0.001d 
 Late stage (stage III/IV) 174 12.91 ± 12.10 <0.001e 
VariableNumberCDH3 (ng/mL)aP
Sex 
 Female 156 9.31 ± 10.96 0.292 
 Male 173 10.53 ± 9.32  
Age 
 <60 143 9.50 ± 8.58 0.939 
 ≥60 186 10.30 ± 11.20  
Smoke 
 No 210 9.69 ± 9.18 0.065 
 Yes 119 10.42 ± 11.67  
Healthy vs. NSCLC 
 Healthy control 57 6.63 ± 2.84 0.0007 
 NSCLC 272 10.65 ± 10.95  
Lung cancer histology 
 Healthy 57 6.63 ± 2.84  
 NSCLC–non ADC 66 9.85 ± 7.24 <0.05b 
 ADC 206 10.90 ± 11.90 <0.01c 
Cancer stage 
 Healthy 57 6.63 ± 2.84  
 Early stage (stage I/II) 98 6.63 ± 6.94 <0.001d 
 Late stage (stage III/IV) 174 12.91 ± 12.10 <0.001e 

aThe data are presented as the mean ± SD.

bThe P value represents the difference between healthy controls and patients with NSCLC–non ADC.

cThe P value represents the difference between healthy controls and patients with ADC.

dThe P value represents the difference between patients with early-stage and late-stage cancer.

eThe P value represents the difference between healthy controls and patients with late-stage cancer.

Previously, we used a high-abundance protein removal system combined with a gel-based and label-free proteomic analysis to identify 363 common PE proteins in all six types of exudative PEs (TB, pneumonia, lung ADC paramalignant PE, lung ADC MPE, breast cancer MPE, and gastric cancer MPE; ref. 19). Taking advantage of the isobaric labeling–based quantification methodology used for unbiased untargeted biomarker discovery, this study established, for the first time, a PE proteome with 561 reproducibly quantified proteins related to first/second-generation EGFR-TKI resistance from four groups of PEs via high-abundance protein depletion technology combined with an iTRAQ-based quantitative proteomics approach. To the best of our knowledge, this is the most comprehensive quantitative PE dataset. Mundt and colleagues identified 386 overlapping proteins in two pooled PE samples from three types of PEs (lung ADC, malignant mesothelioma, and benign mesothelioma; ref. 23). Recently, Shi and colleagues identified 432 proteins from lung MPE and TB (24). We herein integrated genomic and proteomic approaches to identify 15 potential EGFR-TKI resistance–associated protein markers. MET amplification, MET protein hyperactivation, and MET mutation are recognized as key mechanisms underlying acquired resistance to EGFR-TKIs (25). Notably, soluble MET and soluble cadherin-1 (CDH1) in the serum have been reported as lung cancer biomarkers by us and others (19, 26). MUC1 was recently identified as an effective sensor for monitoring the effects of erlotinib treatment (27). Consistent with this observation, we also detected an increase in the MUC1 level (1.44-fold; Supplementary Table S1) in patients resistant to first/second-generation EGFR-TKIs. This study verified and validated sCDH3 as a novel biomarker of EGFR-TKI resistance in MPE, and serum samples further supported the utility of our proteome datasets for the prediction, monitoring, and prognosis of patients with lung ADC with targeted therapy.

CDH3, encoded by the CDH3 gene, is a classical cadherin expressed in epithelial cells and was first found in the mouse placenta in 1986 (28). It has been reported that CDH3 is involved in the maintenance of tissue architecture, organ morphogenesis, cell differentiation, cellular movement, and stem cell biology (29–31). Several studies have proposed that CDH3 acts as an oncogene and a tumor suppressor gene, in which its role in oncogenesis is cell content and cancer type dependent (30). CDH3 overexpression is detected in breast, ovarian, prostate, endometrial, skin, gastric, and pancreatic cancers. However, CDH3 is downregulated in oral squamous cell carcinoma, melanoma, and hepatocellular carcinoma. Both the oncogenic and suppressive roles of CDH3 in one particular cancer type have also been observed (30, 32). Specifically, a tumor suppressive role of CDH3 in lung cancer was proposed according to an IHC analysis of 28 NSCLC cancer tissues in 1999 (33). Recently, Imai and colleagues reported that CDH3 was overexpressed in NSCLC tissues and correlated with a poor prognosis in NSCLC (32). The discrepancy regarding the role of CDH3 in these two NSCLC-related studies might have resulted from the different histologic types, subcellular localizations, and even antibodies used in the IHC analysis. Importantly, we herein identified sCDH3 as an EGFR-TKI resistance–associated biomarker via MS-MS, a highly accurate protein identification method based on unique peptide sequences. The specificity of the antibodies used in this study was also confirmed via Western blot analysis using lung cancer tissues and PE089 cell line with CDH3 gene knockdown (Fig. 2). We further provide evidence that CDH3 is overexpressed in ADC cancer tissues and that baseline sCDH3 is positively associated with late-stage cancer and OS in patients with NSCLC regardless of their EGFR gene status. Therefore, we herein propose an oncogenic role for CDH3 in ADC progression.

CDH3 is a Ca2+-dependent cell–cell adhesion glycoprotein comprised of five extracellular cadherin repeats, a transmembrane region, and a highly conserved cytoplasmic tail. It is well documented that both matrix metalloproteases (MMP) and metalloendopeptidases (ADAM) are involved in the shedding of the ectodomains of membrane-anchored growth factors, cytokines, and receptors (34). The extracellular cleavage and shedding of CDH3 is regulated by MMPs such as MMP-1 and MMP-2 in breast cancer cells (35). The sCDH3 level was significantly increased in patients with late-stage NSCLC with or without EGFR mutations (Table 3). However, the sCDH3 level was associated with the PFS of patients with ADC receiving EGFR-TKI therapy but not chemotherapy, supporting a specific role for CDH3 in EGFR-TKI resistance (Fig. 2). Although a proinvasive role of sCDH3 in breast cancer has been reported (35, 36), this study, for the first time, reports the clinical significance of CDH3 in EGFR-TKI resistance and NSCLC. CDH3 cross-talks with EGFR through the ligand-dependent signaling of EGFR and overexpressed CDH3 may prolong activation of the MAP and AKT signaling pathways in certain types of cancer cells (30, 37). Notably, a disintegrin and metalloproteinase-17 (ADAM17) is involved in the cleavage of the ectodomain of EGFR ligands such as amphiregulin, TGFα, and EGF. ADAM17 plays an undisputable role in ADC oncogenesis and has been recognized as the center of EGFR signaling (38). Although CDH3 and EGFR ligands are processed by MMPs and ADAM17, respectively, these two types of metalloproteinases are significantly dysregulated and contribute to lung cancer invasion. Interestingly, high levels of amphiregulin and TGFα in the serum predict a poor response to the EGFR-TKI gefitinib in patients with advanced NSCLC (39), suggesting potential interplay between MMPs and ADAM17 in the modulation of EGFR-mediated lung cancer progression and EGFR-TKI resistance.

Epithelial–mesenchymal transition (EMT), a temporary and reversible phenomenon in which adherent epithelial cells undergo morphologic changes accompanied by a gain in migratory capacity and invasion and vice versa, is a critical process in cancer metastasis (40). Cadherin switching, a process in which cells shift cadherin isoform expression from CDH1 to CDH2 and/or CDH3, is a hallmark of EMT in certain types of cancer (e.g., prostate, ovarian, and bladder cancers; ref. 41). The relationship between CDH3- and EMT-related genes is complicated and even controversial in different types of cancer. Overexpression of the mesenchymal markers SNAI2 (also known as Slug) and ZEB1 represses CDH1 expression, promoting the migration of melanoma cells, but decreases adhesion to human keratinocytes. SNAI2 overexpression also suppresses CDH3 expression (42). Interestingly, SNAI2 is a direct transcriptional activator at the ZEB1 promoter (43), indicating that SNAI2 and ZEB1 cooperatively regulate the EMT process by suppressing the expression of CDH1 and/or CDH3 in melanoma cells. However, CDH3 overexpression promotes the expression of three members of the Snail family (SNAI1, SNAI2, and SNAI3) and ZEB1 but does not regulate the level of TWIST1 or vimentin in oral squamous carcinoma SCC22A cells (37). In this study, CDH3 knockdown induced the expression of the epithelial marker CDH1, but also the expressions of mesenchymal marker ZEB1 and SNAI2 in lung ADC cells (Supplementary Fig. S4). These findings suggest that the role of CDH3 in the modulation of EMT remains an object of debate. On the other hand, EMT is strongly associated with acquired resistance to EGFR-TKIs (44). Recent studies provide evidence that EMT is a mechanism for escape from EGFR-targeted therapy in lung cancer (45). Furthermore, both EMT and resistance to the EGFR-TKI erlotinib are associated with overexpression of the mesenchymal marker AXL (46), supporting a positive correlation between a high level of sCDH3 and a poor outcome in EGFR-mutated patients treated with first/second-generation EGFR-TKIs.

The clinical significance of and novel oncogenic molecular pathways mediated by CDH3 are partially uncovered, providing the basis for the development of personalized medicine targeting CDH3 in invasive and/or metastatic diseases, including certain types of human cancer (30, 47). For example, the fully humanized anti-CDH3 (P-cadherin) IgG1 mAb PF-03732010 (48), the extended half-life dual-affinity retargeting bispecific molecule against CDH3 and CD3 (PF-06671008; ref. 49), and the yttrium Y 90 anti-CDH3 mAb FF-21101 (50) were developed to overcome CDH3-mediated pathogenesis, and these antibody-based inhibitors of CDH3 are currently in clinical trials. CDH3 knockdown enhanced gefitinib sensitivity, but CDH3 knockdown did not reduce cell viability or AKT activity in PE089 lung ADC cells (Fig. 2), supporting the theory that CDH3 may act as a novel regulator in EGFR-TKI resistance. Our future work will focus on the underlying mechanism of CDH3 in EGFR-TKI resistance and the potential application of targeting CDH3 to overcome the EGFR-TKI resistance in the clinic.

The limitation to this study is the lack of complete information on the T790M mutational status of patients with ADC recruited between 2009 and 2016. Nine of 76 patients (7 PR and 2 non-PR) were diagnosed with T790M-positive advanced ADC, and the serum levels of sCDH3 in 7 PR patients with the T790M mutation were significantly reduced after EGFR-TKI treatment for 1 month (9.68 ± 4.42 vs. 5.40 ± 2.12, P = 0.0043), but no significant change was observed in 2 non-PR patients (7.78 ± 2.13 vs. 6.22 ± 3.42). Future work will be interesting and necessary to investigate the potential application of sCDH3 in monitoring the efficacy of osimertinib and even the T790M mutation in advanced ADC.

In conclusion, our study established a comprehensive quantitative PE proteome useful in EGFR-TKI resistance studies, and we validated serum sCDH3 as a potentially novel biomarker for monitoring EGFR-TKI resistance. We also show the diagnostic and prognostic value of baseline sCDH3 in NSCLC. Together, our studies provide a new prospect for classic cadherin CDH3 and for its clinical application. Further investigations of sCDH3 and CDH3 should focus on its drug-resistant molecular mechanisms in detail.

No potential conflicts of interest were disclosed.

Conception and design: T.-F. Hsiao, C.-L. Wang, H.-Y. Lin, C.-J. Yu

Development of methodology: T.-F. Hsiao, H.-Y. Lin

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): T.-F. Hsiao, C.-L. Wang, Y.-C. Wu, Y.-C. Chiu, K.-J. Liu, G.-C. Chang, K.-Y. Chien, J.-S. Yu

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): T.-F. Hsiao, C.-L. Wang, H.-P. Feng, Y.-C. Chiu, H.-Y. Lin, C.-J. Yu

Writing, review, and/or revision of the manuscript: T.-F. Hsiao, C.-L. Wang, C.-J. Yu

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): C.-L. Wang, Y.-C. Wu, J.-S. Yu, C.-J. Yu

Study supervision: C.-L. Wang, C.-J. Yu

Other (mass spectrometric data acquisition): K.-Y. Chien

This work was financially supported by grants from Chang Gung Memorial Hospital (Taoyuan, Taiwan; CORPD1J0091, CMRPD1H0641-2, CMRPD1H0081-3, and BMRP894 to C.-J. Yu; CMRPG3A0661 to C.-L. Wang; and CLRPD190019); the Ministry of Science and Technology, Taiwan (105-2320-B-182-035-MY3 and 108-2320-B-182-007-MY3 to C.-J. Yu); the Molecular Medicine Research Center, Chang Gung University (Taoyuan, Taiwan); and The Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education, Taiwan.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Bray
F
,
Ferlay
J
,
Soerjomataram
I
,
Siegel
RL
,
Torre
LA
,
Jemal
A
. 
Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries
.
CA Cancer J Clin
2018
;
68
:
394
424
.
2.
Sun
S
,
Schiller
JH
,
Gazdar
AF
. 
Lung cancer in never smokers–a different disease
.
Nat Rev Cancer
2007
;
7
:
778
90
.
3.
Seo
JS
,
Ju
YS
,
Lee
WC
,
Shin
JY
,
Lee
JK
,
Bleazard
T
, et al
The transcriptional landscape and mutational profile of lung adenocarcinoma
.
Genome Res
2012
;
22
:
2109
19
.
4.
Hsu
KH
,
Ho
CC
,
Hsia
TC
,
Tseng
JS
,
Su
KY
,
Wu
MF
, et al
Identification of five driver gene mutations in patients with treatment-naive lung adenocarcinoma in Taiwan
.
PLoS One
2015
;
10
:
e0120852
.
5.
Fukuoka
M
,
Wu
YL
,
Thongprasert
S
,
Sunpaweravong
P
,
Leong
SS
,
Sriuranpong
V
, et al
Biomarker analyses and final overall survival results from a phase III, randomized, open-label, first-line study of gefitinib versus carboplatin/paclitaxel in clinically selected patients with advanced non-small-cell lung cancer in Asia (IPASS)
.
J Clin Oncol
2011
;
29
:
2866
74
.
6.
Eisenhauer
EA
,
Therasse
P
,
Bogaerts
J
,
Schwartz
LH
,
Sargent
D
,
Ford
R
, et al
New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1)
.
Eur J Cancer
2009
;
45
:
228
47
.
7.
Thomas
A
,
Liu
SV
,
Subramaniam
DS
,
Giaccone
G
. 
Refining the treatment of NSCLC according to histological and molecular subtypes
.
Nat Rev Clin Oncol
2015
;
12
:
511
26
.
8.
Oxnard
GR
,
Arcila
ME
,
Chmielecki
J
,
Ladanyi
M
,
Miller
VA
,
Pao
W
. 
New strategies in overcoming acquired resistance to epidermal growth factor receptor tyrosine kinase inhibitors in lung cancer
.
Clin Cancer Res
2011
;
17
:
5530
7
.
9.
Oxnard
GR
,
Arcila
ME
,
Sima
CS
,
Riely
GJ
,
Chmielecki
J
,
Kris
MG
, et al
Acquired resistance to EGFR tyrosine kinase inhibitors in EGFR-mutant lung cancer: distinct natural history of patients with tumors harboring the T790M mutation
.
Clin Cancer Res
2011
;
17
:
1616
22
.
10.
Mok
TS
,
Wu
YL
,
Ahn
MJ
,
Garassino
MC
,
Kim
HR
,
Ramalingam
SS
, et al
Osimertinib or platinum-pemetrexed in EGFR T790M-positive lung cancer
.
N Engl J Med
2017
;
376
:
629
40
.
11.
Soria
JC
,
Ohe
Y
,
Vansteenkiste
J
,
Reungwetwattana
T
,
Chewaskulyong
B
,
Lee
KH
, et al
Osimertinib in untreated EGFR-mutated advanced non-small-cell lung cancer
.
N Engl J Med
2018
;
378
:
113
25
.
12.
Ramalingam
SS
,
Vansteenkiste
J
,
Planchard
D
,
Cho
BC
,
Gray
JE
,
Ohe
Y
, et al
Overall survival with osimertinib in untreated, EGFR-mutated advanced NSCLC
.
N Engl J Med
2020
;
382
:
41
50
.
13.
Rolfo
C
,
Mack
PC
,
Scagliotti
GV
,
Baas
P
,
Barlesi
F
,
Bivona
TG
, et al
Liquid biopsy for advanced non-small cell lung cancer (NSCLC): a statement paper from the IASLC
.
J Thorac Oncol
2018
;
13
:
1248
68
.
14.
Phallen
J
,
Leal
A
,
Woodward
BD
,
Forde
PM
,
Naidoo
J
,
Marrone
KA
, et al
Early noninvasive detection of response to targeted therapy in non-small cell lung cancer
.
Cancer Res
2019
;
79
:
1204
13
.
15.
Takeuchi
A
,
Oguri
T
,
Sone
K
,
Ito
K
,
Kitamura
Y
,
Inoue
Y
, et al
Predictive and prognostic value of CYFRA 21–1 for advanced non-small cell lung cancer treated with EGFR-TKIs
.
Anticancer Res
2017
;
37
:
5771
76
.
16.
Noonan
SA
,
Patil
T
,
Gao
D
,
King
GG
,
Thibault
JR
,
Lu
X
, et al
Baseline and on-treatment characteristics of serum tumor markers in stage IV oncogene-addicted adenocarcinoma of the lung
.
J Thorac Oncol
2018
;
13
:
134
38
.
17.
Andrews
CO
,
Gora
ML
. 
Pleural effusions: pathophysiology and management
.
Ann Pharmacother
1994
;
28
:
894
903
.
18.
Yu
CJ
,
Wang
CL
,
Wang
CI
,
Chen
CD
,
Dan
YM
,
Wu
CC
, et al
Comprehensive proteome analysis of malignant pleural effusion for lung cancer biomarker discovery by using multidimensional protein identification technology
.
J Proteome Res
2011
;
10
:
4671
82
.
19.
Liu
PJ
,
Chen
CD
,
Wang
CL
,
Wu
YC
,
Hsu
CW
,
Lee
CW
, et al
In-depth proteomic analysis of six types of exudative pleural effusions for nonsmall cell lung cancer biomarker discovery
.
Mol Cell Proteomics
2015
;
14
:
917
32
.
20.
Chen
CD
,
Wang
CL
,
Yu
CJ
,
Chien
KY
,
Chen
YT
,
Chen
MC
, et al
Targeted proteomics pipeline reveals potential biomarkers for the diagnosis of metastatic lung cancer in pleural effusion
.
J Proteome Res
2014
;
13
:
2818
29
.
21.
Hsu
CH
,
Hsu
CW
,
Hsueh
C
,
Wang
CL
,
Wu
YC
,
Wu
CC
, et al
Identification and characterization of potential biomarkers by quantitative tissue proteomics of primary lung adenocarcinoma
.
Mol Cell Proteomics
2016
;
15
:
2396
410
.
22.
Hsu
LH
,
Liu
KJ
,
Tsai
MF
,
Wu
CR
,
Feng
AC
,
Chu
NM
, et al
Estrogen adversely affects the prognosis of patients with lung adenocarcinoma
.
Cancer Sci
2015
;
106
:
51
9
.
23.
Mundt
F
,
Johansson
HJ
,
Forshed
J
,
Arslan
S
,
Metintas
M
,
Dobra
K
, et al
Proteome screening of pleural effusions identifies galectin 1 as a diagnostic biomarker and highlights several prognostic biomarkers for malignant mesothelioma
.
Mol Cell Proteomics
2014
;
13
:
701
15
.
24.
Shi
J
,
Li
P
,
Zhou
L
,
Qi
S
,
Wang
B
,
Li
D
, et al
Potential biomarkers for antidiastole of tuberculous and malignant pleural effusion by proteome analysis
.
Biomark Med
2019
;
13
:
123
33
.
25.
Drilon
A
,
Cappuzzo
F
,
Ou
SI
,
Camidge
DR
. 
Targeting MET in lung cancer: will expectations finally be MET?
J Thorac Oncol
2017
;
12
:
15
26
.
26.
Fu
L
,
Guo
W
,
Liu
B
,
Sun
L
,
Bi
Z
,
Zhu
L
, et al
Shedding of c-Met ectodomain correlates with c-Met expression in non-small cell lung cancer
.
Biomarkers
2013
;
18
:
126
35
.
27.
de Boer
HR
,
Pool
M
,
Joosten
E
,
Everts
M
,
Samplonius
DF
,
Helfrich
W
, et al
Quantitative proteomics analysis identifies MUC1 as an effect sensor of EGFR inhibition
.
Oncogene
2019
;
38
:
1477
88
.
28.
Nose
A
,
Takeichi
M
. 
A novel cadherin cell adhesion molecule: its expression patterns associated with implantation and organogenesis of mouse embryos
.
J Cell Biol
1986
;
103
:
2649
58
.
29.
Tepass
U
,
Truong
K
,
Godt
D
,
Ikura
M
,
Peifer
M
. 
Cadherins in embryonic and neural morphogenesis
.
Nat Rev Mol Cell Biol
2000
;
1
:
91
100
.
30.
Vieira
AF
,
Paredes
J
. 
P-cadherin and the journey to cancer metastasis
.
Mol Cancer
2015
;
14
:
178
.
31.
Ribeiro
AS
,
Paredes
J
. 
P-cadherin linking breast cancer stem cells and invasion: a promising marker to identify an "intermediate/metastable" EMT state
.
Front Oncol
2014
;
4
:
371
.
32.
Imai
S
,
Kobayashi
M
,
Takasaki
C
,
Ishibashi
H
,
Okubo
K
. 
High expression of P-cadherin is significantly associated with poor prognosis in patients with non-small-cell lung cancer
.
Lung Cancer
2018
;
118
:
13
19
.
33.
Smythe
WR
,
Williams
JP
,
Wheelock
MJ
,
Johnson
KR
,
Kaiser
LR
,
Albelda
SM
. 
Cadherin and catenin expression in normal human bronchial epithelium and non-small cell lung cancer
.
Lung Cancer
1999
;
24
:
157
68
.
34.
Peschon
JJ
,
Slack
JL
,
Reddy
P
,
Stocking
KL
,
Sunnarborg
SW
,
Lee
DC
, et al
An essential role for ectodomain shedding in mammalian development
.
Science
1998
;
282
:
1281
4
.
35.
Ribeiro
AS
,
Albergaria
A
,
Sousa
B
,
Correia
AL
,
Bracke
M
,
Seruca
R
, et al
Extracellular cleavage and shedding of P-cadherin: a mechanism underlying the invasive behaviour of breast cancer cells
.
Oncogene
2010
;
29
:
392
402
.
36.
Ribeiro
AS
,
Nobre
AR
,
Mendes
N
,
Almeida
J
,
Vieira
AF
,
Sousa
B
, et al
SRC inhibition prevents P-cadherin mediated signaling and function in basal-like breast cancer cells
.
Cell Commun Signal
2018
;
16
:
75
.
37.
Lysne
D
,
Johns
J
,
Walker
A
,
Ecker
R
,
Fowler
C
,
Lawson
KR
. 
P-cadherin potentiates ligand-dependent EGFR and IGF-1R signaling in dysplastic and malignant oral keratinocytes
.
Oncol Rep
2014
;
32
:
2541
8
.
38.
Saad
MI
,
Rose-John
S
,
Jenkins
BJ
. 
ADAM17: an emerging therapeutic target for lung cancer
.
Cancers
2019
;
11
:
pii: E1218
.
39.
Ishikawa
N
,
Daigo
Y
,
Takano
A
,
Taniwaki
M
,
Kato
T
,
Hayama
S
, et al
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
.
Cancer Res
2005
;
65
:
9176
84
.
40.
Yilmaz
M
,
Christofori
G
. 
EMT, the cytoskeleton, and cancer cell invasion
.
Cancer Metastasis Rev
2009
;
28
:
15
33
.
41.
Siret
C
,
Dobric
A
,
Martirosyan
A
,
Terciolo
C
,
Germain
S
,
Bonier
R
, et al
Cadherin-1 and cadherin-3 cooperation determines the aggressiveness of pancreatic ductal adenocarcinoma
.
Br J Cancer
2018
;
118
:
546
57
.
42.
Fenouille
N
,
Tichet
M
,
Dufies
M
,
Pottier
A
,
Mogha
A
,
Soo
JK
, et al
The epithelial-mesenchymal transition (EMT) regulatory factor SLUG (SNAI2) is a downstream target of SPARC and AKT in promoting melanoma cell invasion
.
PLoS One
2012
;
7
:
e40378
.
43.
Wels
C
,
Joshi
S
,
Koefinger
P
,
Bergler
H
,
Schaider
H
. 
Transcriptional activation of ZEB1 by slug leads to cooperative regulation of the epithelial-mesenchymal transition-like phenotype in melanoma
.
J Invest Dermatol
2011
;
131
:
1877
85
.
44.
Uramoto
H
,
Iwata
T
,
Onitsuka
T
,
Shimokawa
H
,
Hanagiri
T
,
Oyama
T
. 
Epithelial-mesenchymal transition in EGFR-TKI acquired resistant lung adenocarcinoma
.
Anticancer Res
2010
;
30
:
2513
7
.
45.
Tulchinsky
E
,
Demidov
O
,
Kriajevska
M
,
Barlev
NA
,
Imyanitov
E
. 
EMT: a mechanism for escape from EGFR-targeted therapy in lung cancer
.
Biochim Biophys Acta Rev Cancer
2019
;
1871
:
29
39
.
46.
Byers
LA
,
Diao
L
,
Wang
J
,
Saintigny
P
,
Girard
L
,
Peyton
M
, et al
An epithelial-mesenchymal transition gene signature predicts resistance to EGFR and PI3K inhibitors and identifies Axl as a therapeutic target for overcoming EGFR inhibitor resistance
.
Clin Cancer Res
2013
;
19
:
279
90
.
47.
Yu
W
,
Yang
L
,
Li
T
,
Zhang
Y
. 
Cadherin signaling in cancer: its functions and role as a therapeutic target
.
Front Oncol
2019
;
9
:
989
.
48.
Park
J
,
Park
E
,
Han
SW
,
Im
SA
,
Kim
TY
,
Kim
WH
, et al
Down-regulation of P-cadherin with PF-03732010 inhibits cell migration and tumor growth in gastric cancer
.
Invest New Drugs
2012
;
30
:
1404
12
.
49.
Root
AR
,
Cao
W
,
Li
B
,
LaPan
P
,
Meade
C
,
Sanford
J
, et al
Development of PF-06671008, a highly potent anti-P-cadherin/anti-CD3 bispecific DART molecule with extended half-life for the treatment of cancer
.
Antibodies
2016
;
5
:
pii: E6
.
50.
Hosono
M
,
Ikebuchi
H
,
Kinuya
S
,
Yanagida
S
,
Nakamura
Y
,
Yamada
T
, et al
Manual on the proper use of yttrium-90-labeled anti-P-cadherin antibody injection for radionuclide therapy in clinical trials
.
Ann Nucl Med
2019
;
33
:
787
805
.