Virtual Library

 x 
My Library Virtual Library FAQ

Start Your Search

M.S. Tsao

Moderator of

  • +

    WS08 - Special Session: CAP/IASLC/AMP Guidelines for Molecular Testing in Lung Cancer (ID 481)

    • Event: WCLC 2016
    • Type: Workshop
    • Track:
    • Presentations: 1
    • +

      Special Session: CAP/IASLC/AMP Guidelines for Molecular Testing in Lung Cancer (ID 7220)

      07:30 - 08:30  |  Author(s): Y. Yatabe

      • Abstract
      • Presentation
      • Slides

      Abstract not provided

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.



Author of

  • +

    MA17 - Genetic Drivers (ID 409)

    • Event: WCLC 2016
    • Type: Mini Oral Session
    • Track: Biology/Pathology
    • Presentations: 1
    • +

      MA17.06 - Landscape of Somatic Mutations Involving Lung Cancer Associated Genes in Non-Small Cell Lung Cancer (NSCLC) Patient-Derived Xenografts (ID 6084)

      14:56 - 15:02  |  Author(s): M.S. Tsao

      • Abstract
      • Presentation
      • Slides

      Background:
      Patient-derived tumor xenografts (PDXs) have high fidelity to their histological origins, and maintain the molecular heterogeneity and genetic aberrations of the donor patient tumors more faithfully than established in non-small cell lung cancer (NSCLC) cell lines. This study evaluated whether our panel of PDX models recapitulate known cancer-related gene mutations.

      Methods:
      Whole-exome sequencing was completed on 103 NSCLC PDX models, 47 adenocarcinoma (AdC) and 56 squamous (SqCC), with a mean coverage of 84x. After filtering for contaminating mouse reads, the exome data were aligned using the Burrows-Wheeler Aligner, processed using the standard GATK pipeline, and mutations were identified using MuTect. Additional filtering using dbSNP, ExAC and ESP was performed for cases without corresponding normal adjacent lung exome data (n = 80). The identified mutations were compared to 1260 frequently mutated cancer-related genes, which were compiled from a panel of cancer-related mutated genes (555) and a panel of lung cancer-specific mutated genes (1082).

      Results:
      High rates of somatic mutations were observed in both AdC (mean of 12.4 mutations/megabase) and SqCC (mean of 11.7 mutations/megabase) PDX models. Compared to the rates observed in primary lung cancers in The Cancer Genome Atlas studies (mean of 8.9 mutations/megabase in AdC; 8.1 mutations/megabase in SqCC), these values appear higher, but may be inflated due to the lack of data from corresponding normal tissues. AdC models had a total of 953 mutated genes (median: 57 genes/model; range: 5-307), while SqCC models were characterized by 1007 mutated genes (median: 55 genes/model; range: 21-354). Specific mutation frequencies were compared to those determined in a recent study involving genomic alterations in human primary lung AdC and SqCC (Nature Genetics 2016; 48; 607–616). This comparison, based on mutated genes common in both studies, demonstrated significant correlation of the frequencies in 791 genes in AdC (ρ=0.78; p<2.2×10[-16]), as well as in 799 genes in SqCC (ρ=0.73; p<2.2×10[-16]). Three genes that were reported as significantly mutated in both AdC and SqCC primaries, and had higher mutation frequencies in SqCC, were also observed to be higher in our SqCC PDX models (TP53: 48.9% in AdC vs. 55.4% in SqCC; CDKN2A: 4.3% vs. 7.1% and PIK3CA: 2.1% vs. 23.2%); however, the statistical significance of these differences needs to be tested.

      Conclusion:
      Mutation landscapes in cancer genes are recapitulated in AdC and SqCC PDX models. The fidelity of these landscapes in matched patient primary tumour samples is being investigated.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

  • +

    MTE12 - Clinically Relevant Signal Transduction Pathways (Ticketed Session) (ID 306)

    • Event: WCLC 2016
    • Type: Meet the Expert Session (Ticketed Session)
    • Track: Biology/Pathology
    • Presentations: 1
    • Moderators:
    • Coordinates: 12/06/2016, 07:30 - 08:30, Schubert 4
    • +

      MTE12.01 - Clinically Relevant Signal Transduction Pathways (ID 6561)

      07:30 - 08:00  |  Author(s): M.S. Tsao

      • Abstract
      • Presentation
      • Slides

      Abstract:
      Up to a decade ago, the main non-surgical treatment modalities in oncology have been cytotoxic chemotherapy and/or radiation therapy. These therapies are aimed at inducing DNA damage, thus selectively killing the highly proliferative cancer cells. More recently, new therapies are targeting signaling pathways that are critical to support cancer cell proliferation and/or survival, including micro-environmental factors that sustain tumor. The first part of our presentation will review pathways operating mainly in the tumor cells, and how they constitute targets for lung cancer therapies. The second part will focus on the vascularization mechanisms in primary and metastatic lung tumors, antivascular drugs, potential biomarkers and on mechanisms through which tumors can become resistant to antivascular drugs. DNA Repair Pathway: Genomic DNA encodes all biochemical processes that drive cellular function and biology. Extensive damage to DNA encoding proteins/enzymes involved in cell proliferation will result in cell cycle arrest and cell death. DNA damage may also induce replicative errors and mutations, which leads to constitutive activation of oncogenes, or inactivation of tumor suppressor genes. DNA repair mechanisms are crucial for mitigating catastrophic chromosomal damage during DNA synthesis and replication, thus allowing tumor cells to survive chemotherapy or radiotherapy. New targeted anti-cancer agents being developed include those that inhibit the activity of critical molecules involved in DNA repair, or inhibit cell cycle checkpoint proteins that allow DNA repair mechanisms to occur. EGFR and downstream pathways: The proliferation of epithelial cells depends on growth stimuli arising from either factors produced by the tumor cells themselves (autocrine), factors produced by cells from distant organs (endocrine), or factors from neighboring tumor or non-tumor cells in the tumor microenvironment (paracrine). For lung epithelial cells, a major growth stimulating pathway involve the epidermal growth factor receptor (EGFR) family members. EGFR (HER1) is highly expressed in >90% of squamous cell carcinoma and in 60-80% of adenocarcinoma. EGFR has many ligands, including EGF, TGF-a, amphiregulin, HB-EGF, etc. Binding of the ligands to the EGFR induces homo or hetero dimerization of EGFR and its family members, activates the cytoplasmic tyrosine kinase of the receptor, and promotes auto-phosphorylation. This sequentially leads to binding of SOS1, activation of downstream RAS, RAF, MEK and ERK/MAPK. Targeting EGFR by monoclonal antibodies and small molecule kinase inhibitors have demonstrated clinical efficacy in subpopulation of NSCLC patients. Targeted agents against KRAS, BRAF and MEK are in clinical trials. MET, ALK, and ROS1 pathways: Other tyrosine kinase receptors (RTKs) that may play important role in lung cancers include hepatocyte growth factor (HGF) receptor MET and fibroblast growth factor receptor (FGFR) family members. In contrast to EGFR, the ligands for MET and FGFRs appear to be produced by the tumor stromal fibroblasts. While attempts to inhibit MET signaling pathway by neutralizing antibody have not been successful, more recent data suggest that MET kinase inhibitors may be highly effective in patients with MET exon 14 splice site mutations. Such mutations cause the loss of exon 14, which encode the Cbl binding site of the receptor, a crucial domain required for the degradation of MET receptor. The RTKs with close homology to MET are ALK and ROS1. Constitutive activation of ALK and ROS1 occurs by formation of new chimeric protein through translocation involving these genes. Inhibitors to ALK and ROS1 are now clinically approved for treatment of lung cancers that express fusion proteins resulting from the rearrangement of the ALK and ROS1 genes. PI3K/AKT/mTOR pathway: Aside from activating the MAPK pathway, tyrosine kinase receptors may also activate the PI3K/AKT/mTOR pathway, which plays a crucial role in the survival of lung cancer cells. This pathway is commonly activated in NSCLC through amplification or activating mutation of the PIK3CA gene, or inactivation of PTEN by gene deletion, mutation or methylation. While there is intense research to develop targeted therapies that inhibit this important survival pathway, the efforts have so far met little success, revealing the complexity of this pathway. There is also evidence that alternative RTK and PI3K signaling play an important role as bypass mechanisms for the development of resistance to kinase inhibitor therapies. Angiogenesis pathways: Because an adequate blood supply is regarded as essential for tumor development, there had been overwhelming optimism initially that blocking angiogenic pathways would represent an effective treatment strategy in solid malignancies, including primary and metastatic lung tumors. However, clinical trials investigating antivascular drugs have been both encouraging and disappointing. Success with antivascular strategies therefore requires a deeper knowledge of the clinical significance of the different angiogenic machineries that control lung tumors. VEGF (vascular endothelial growth factor) is the key molecular regulator of new tumor blood capillary formation (i.e. angiogenesis) and its high expression is associated with poorer survival in NSCLC. Bevacizumab, a humanized monoclonal anti-VEGF antibody, is currently approved for the first-line treatment of advanced stage non-squamous NSCLC in combination with chemotherapy. Ramucirumab (a fully human monoclonal antibody against VEGFR2) has been approved for use in combination with docetaxel for the treatment of metastatic NSCLC patients who progressed after platinum-based chemotherapy. Nintedanib (an oral RTK inhibitor against VEGFRs, platelet-derived growth factor receptors (PDGFR) and FGFRs in combination with chemotherapy has been approved by the EMEA in NSCLC patients with locally advanced, metastatic or locally recurrent lung adenocarcinoma after first-line chemotherapy. Additional anti-vascular strategies including vascular disrupting agents (VDAs) to destroy the established tumor vasculature and other investigational antiangiogenic antibodies and small molecule RTK inhibitors are also under clinical testing for NSCLC therapy, though enthusiasm is tempered by short disease control and modest overall survival benefit. Angiogenesis Resistance Mechanisms and Biomarkers: Unfortunately, resistance against antivascular therapies is poorly understood. The possible resistance mechanisms include increased intratumoral hypoxia, the activation of compensatory angiogenic machineries, the release of myeloid or endothelial progenitor cell populations, the downregulation of target receptors in endothelial and/or tumor cells, limited tumor tissue drug penetration, and also a switch to an alternative vascularization mechanism such as intussusceptive angiogenesis or vessel-cooption. Reliable biomarkers for the prediction of response to antivascular drugs are also yet to be identified and clinically validated.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

  • +

    OA19 - Translational Research in Early Stage NSCLC (ID 402)

    • Event: WCLC 2016
    • Type: Oral Session
    • Track: Early Stage NSCLC
    • Presentations: 1
    • +

      OA19.01 - A Standardized and Validation of Prognostic Gene Expression Signatures for Squamous Cell Lung Carcinoma by the SPECS Lung Consortium (ID 4329)

      11:00 - 11:10  |  Author(s): M.S. Tsao

      • Abstract
      • Presentation
      • Slides

      Background:
      High-throughput gene expression profiling led to proposal of multiple expression-based prognostic signatures for squamous cell lung carcinoma (SCC), but none has been validated. A multi-institutional squamous lung cancer consortium of investigators is developing prognostic signatures through the US NCI Lung SPECS (Strategic Partnership for Evaluation of Cancer Signatures) program. Six institutions contributed tumor specimens and published/unpublished expression-based prognostic signatures for validation using standardized sample cohorts (a primary validation cohort comprising institutional cases, and additional validation cohorts from two prospective cooperative group studies) and quality controlled assessment in independent laboratory and statistical cores. Here, we report the results of the primary validation.

      Methods:
      Cases of primary SCC (by central pathology review) meeting clinical (Stage I-II; surgical treatment only; 3-year followup) and specimen quality criteria (Tumor cellularity >= 50%; necrosis <= 20%) were submitted. Clinical, pathological and outcome data were uploaded to a central database. Frozen tumor samples underwent centralized mRNA extraction (Qiagen Symphony), quality control (RIN >= 6.0) and microarray profiling (Affymetrix U133) in core labs. An independent statistical core assessed validation of 7 pre-existing mRNA signatures and generated new models using MCP clustering.

      Results:
      Among 250 cases meeting entry criteria, median age was 70 (43-92), 161 (65%) were male, and most were former (70%) or current (28%) smokers. Surgery was pneumonectomy: 5%; bilobectomy: 2%; lobectomy: 74%; sublobar: 18%. Pathologic staging was T1: 49%; T2: 50%; T3: 1%; N0: 88%; N1: 12%, and grade was G1: 4%; G2: 50%; G3: 44%. At followup, 148 (59%) were deceased. Three mRNA signatures demonstrated significant univariable association with OS and added independent prognostic value (see Figure) to a multivariable model accounting for age, sex and stage (c-index = 0.641).

      Conclusion:
      The validated signatures, along with two novel signatures generated from the current dataset, are currently undergoing further validation studies using two prospective co-operative group cohorts. Figure 1



      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

  • +

    OA21 - Palliative and Supportive Care for Lung Cancer Patients (ID 405)

    • Event: WCLC 2016
    • Type: Oral Session
    • Track: Palliative Care/Ethics
    • Presentations: 1
    • +

      OA21.02 - ALK-Rearranged Non-Small Cell Lung Cancer is Associated with a High Rate of Venous Thromboembolism (ID 4290)

      11:10 - 11:20  |  Author(s): M.S. Tsao

      • Abstract
      • Presentation
      • Slides

      Background:
      Patients with lung cancer are at increased risk for venous thromboembolism (VTE), particularly those receiving chemotherapy. It is estimated that 8-15% of patients with advanced non-small cell lung cancer (NSCLC) experience a VTE in the course of their disease. The incidence in patients with specific molecular subtypes of NSCLC is unknown. We undertook this review to determine the incidence of VTE in patients with ALK-rearranged NSCLC.

      Methods:
      We identified all patients with ALK-rearranged NSCLC, diagnosed and/or treated at the Princess Margaret Cancer Centre (PM CC) in Canada between July 2012 and January 2015. Retrospective data were extracted from electronic medical records. We then included a validation cohort comprising all consecutive patients with ALK-rearranged NSCLC treated in two tertiary centers in Israel.

      Results:
      Within the PM CC cohort, of 55 patients with ALK-rearranged NSCLC, at a median follow-up of 22 months, 23 (42%) experienced VTE. Patients with VTE were more likely to be Caucasian (p=0.006). The occurrence of VTE was associated with a trend towards worse prognosis (overall survival HR=2.88, p=0.059). Within the validation cohort (N=43), VTE rate was 28% at a median follow-up of 13 months. Combining the cohorts (N=98) the VTE rate was 36%. Patients with VTE were younger (age 52 vs 58, p=0.04) and had a worse ECOG performance status (p=0.04). VTE was associated with shorter OS (HR=5.71, p=0.01)Figure 1.



      Conclusion:
      We found the rate of VTE in our ALK-rearranged cohort is 3-5-fold higher than previously reported for the general NSCLC population. This warrants confirmation in larger cohorts.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

  • +

    P1.02 - Poster Session with Presenters Present (ID 454)

    • Event: WCLC 2016
    • Type: Poster Presenters Present
    • Track: Biology/Pathology
    • Presentations: 1
    • +

      P1.02-036 - An EGFR Tyrosine Kinase Inhibitor Sensitive Patient-Derived Lung Cancer Xenograft Model without Classical Sensitizing Mutations (ID 5398)

      14:30 - 14:30  |  Author(s): M.S. Tsao

      • Abstract

      Background:
      Mutations in the tyrosine kinase (TK) domain of EGFR are oncogenic driver in 10-20% of lung adenocarcinoma (AdC) patients in Western countries. Approximately 90% of EGFR-TK inhibitor (TKI) sensitizing mutations occur as small in-frame deletions in exon 19 or L858R point mutations in exon 21. Recently, novel driver mutations in EGFR with oncogenic and TKI sensitizing activity have been reported. We present here an AdC patient-derived xenograft (PDX) model (PDX12) that is highly sensitive to EGFR-TKI, yet failed to demonstrate classical TKI sensitizing mechanisms.

      Methods:
      Comprehensive genomics profiling was used to characterize the genotype of PDX12, which was established from a resected stage IIIA AdC patient grafted in NGS mouse. The primary human lung cancer cell line (PHLC12) was extracted from its PDX model (PDX12). Aberrant EGFR cell lines used were H3255 (L858R), H2935 (exon 19 deletion), H1975 (L858R and T790M), and H1944 (wild type). Cell viability was assessed after erlotinib treatment at 1nM - 2μM for 72 hours using MTS assay. Levels of EGFR activation in both pre- and post-treatment by Western blot analysis.

      Results:
      PDX12 model had no known oncogenic mutations (EGFR wild type) on exons 18-21 by next-generation sequencing, RT-qPCR, and SISH, but was highly sensitive to EGFR-TKI. The IC50 to erlotinib treatment at 72 hr was 67.13 ± 7.63 nM for PHLC12, compared to 9.70 ± 2.64 nM for H3255, 64.88 ± 8.49 nM for HCC2935, > 2 μM for H1975, and > 2 μM for H1944 EGFR mutant or wild type cells, respectively. Western blot analysis demonstrated a relatively higher molecular weight band for EGFR protein with high expression level in PHLC12 when compared to other lung cancer cell lines. Using RT-qPCR, relative expression level of each EGFR domain (extracellular, tyrosine kinase, and c-terminal domain) in PHLC12 showed no difference compared to EGFR wild type. Phosphorylation status of EGFR in PHLC12 was similar in activity as compared to erlotinib sensitive cell lines.

      Conclusion:
      PHLC12 represents an enigmatic EGFR TKI sensitive lung PDX model without classical TKI sensitizing aberrations. Additional potential mechanisms of EGFR dependency including exon duplication, or post-translational modification of EGFR protein are being investigated.

  • +

    P1.03 - Poster Session with Presenters Present (ID 455)

    • Event: WCLC 2016
    • Type: Poster Presenters Present
    • Track: Radiology/Staging/Screening
    • Presentations: 1
    • +

      P1.03-041 - Do Several Rounds of Negative Screening Low Dose CT Scans Influence the Risk to Develop Lung Cancer? (ID 5373)

      14:30 - 14:30  |  Author(s): M.S. Tsao

      • Abstract

      Background:
      The purpose of this study was to assess whether several years of negative screening low-dose computed tomography (LDCT) scans predict a subsequent lower risk of developing lung cancer. This would have implications for recommended intervals and duration of LDCT lung cancer screening.

      Methods:
      The cohort was an at-risk population who had previous negative screening LDCTs and had not been screened for at least 5 years. Between 2003 and 2009, 4782 individuals had been enrolled in a lung cancer screening study based on age and smoking alone. At this time, their risk was re-calculated using a multifactorial assessment model, and they were contacted in decreasing order of their re-calculated risk. An initial phone interview assessed interim history, general health, interim diagnosis of lung cancer or interim chest CT. Those participants without lung cancer or recent CT were invited for a single LDCT (40mA, 135kV, 1mm axial reconstructions). Subsequent investigation was recommended depending on the LDCT findings: negative, no new or growing nodules (no further recommendation), positive, low suspicion for malignancy (follow up CT in 3-6 months) or positive, high suspicion for malignancy (referral to the local lung cancer rapid diagnostic assessment program).

      Results:
      To date, 361 individuals or family members have been contacted. Fifty-five individuals had passed away (20 from lung cancer), 24 were alive with lung cancer. 129 did not qualify for a LDCT scan (declined participation, or recent CT). A total of 153 have attended for LDCT, on average 7 years after their last LDCT. Ninety-one (59%) studies were reported as negative. Fourty-five (29%) LDCTs were positive with low suspicion and a follow up scan was recommended; in 13 cases nodules had resolved on follow up imaging, the remaining 32 are awaiting surveillance LDCTs. Seventeen (11%) LDCTs were reported as positive with high suspicion; 11 of those have a subsequently biopsy proven lung cancer and 6 are currently undergoing further investigations or LDCT surveillance. All lung cancers diagnosed were either stage I or II. Of the 11 individuals with biopsy proven cancers, 7 had normal previous CTs, 4 had a pre-existing groundglass nodules in the tumor location on the most recent exam. The overall prevalence of lung cancer in this cohort is 15.2% (55/361) and it may increase. The detection rate of LDCT to date is 7.2% (11/153).

      Conclusion:
      Lung cancer risk remains high despite several negative annual screening LDCT scans. Continued screening beyond three years is recommended in high risk individuals.

  • +

    P1.05 - Poster Session with Presenters Present (ID 457)

    • Event: WCLC 2016
    • Type: Poster Presenters Present
    • Track: Early Stage NSCLC
    • Presentations: 3
    • +

      P1.05-001 - Creation and Early Validation of Prognostic miRNA Signatures for Squamous Cell Lung Carcinoma by the SPECS Lung Consortium (ID 6088)

      14:30 - 14:30  |  Author(s): M.S. Tsao

      • Abstract

      Background:
      Despite overall favorable prognosis for operable early stage non-small cell lung cancer, predicting outcome for individual patients has remained challenging. Small retrospective studies have reported potential non-coding micro(mi)RNAs that might have prognostic significance; however, these studies lacked statistical power and validation. To refine these initial findings to clinical application, the investigators have undertaken a collaborative, structured evaluation of multiple signatures putatively prognostic for lung squamous cell carcinoma (SCC) under a NCI/SPECS (Strategic Partnerships fo Evaluating Cancer Signatures) award. The study design specifies a primary validation cohort comprising institutional cases, and additional validation cohorts of Cooperative Group cases, all profiled via a common pipeline.

      Methods:
      Completely resected SCC (confirmed by central pathology review) meeting clinical (Stage I-II; complete 3-year follow-up) and specimen quality criteria (Tumor cellularity ≥ 50%;necrosis ≤ 20%) were submitted by 6 institutions. Clinical, pathological and outcome data were uploaded to a central database. Lysates from 5 um sections of FFPE SSC tumor samples were run on the HTG EdgeSeq Processor (HTG Molecular Diagnostics, Tucson, AZ) using the miRNA whole transcriptome assay in which an excess of nuclease protection probes (NPPs) complimentary to each miRNA hybridize to their target. S1 nuclease then removes un-hybridized probes and RNA leaving behind only NPPs hybridized to their targets in a 1-to-1 ratio. Samples were individually barcoded (using a 16-cycle PCR reaction to add adapters and molecular barcodes), individually purified using AMPure XP beads (Beckman Coulter, Brea, CA) and quantitated using a KAPA Library Quantification kit (KAPA Biosystems, Wilmington, MA). Libraries were sequenced on the Illumina HiSeq platform (Illumina, San Diego, CA) for quantification. Standardization and normalization was provided to the project statistical core for validation of two pre-existing signatures and generation of new models (MCP clustering).

      Results:
      Among 224 cases with miRNA data, median age was 70 (43-92), 143 (64%) male, with 67% former (67%) and current (26%) smokers. All patients were completely resected stage I or II. . At follow-up, 59 (26%) had documented recurrence and 129 (58%) were deceased. To date, we have been unable to validate the previous models, but have created a novel signature of three miRNAs (see Figure) that is being validated in the second phase of the project using an independent, blinded multi-institutional cohort.

      Conclusion:
      The Squamous Lung Cancer SPECS Consortium has established well-annotated and quality-controlled resources for validation of prognostic miRNA signatures. A new candidate 3-miRNA signature has been identified for further development as a clinically useful biomarker.

    • +

      P1.05-006 - Identification of miRNAs and mRNAs Associated with Metastasis in Early-Stage Non-Small Cell Lung Cancer (NSCLC) (ID 5829)

      14:30 - 14:30  |  Author(s): M.S. Tsao

      • Abstract

      Background:
      Early-stage NSCLC patients whose tumours can form primary xenografts (XG) in immune deficient mice have significantly shorter disease-free survival and are at a greater risk of early metastasis compared with patients whose tumours do not form xenografts (non-XG). Genomic and proteomic characterization of XG and non-XG-forming primary patient tumours may reveal clinically relevant genetic aberrations that are associated with early metastasis.

      Methods:
      miRNA-seq and RNA-seq data of 100 early-stage NSCLC patients with known engraftment status were acquired. The cohort includes 62% adenocarcinoma (ADC) and 38% squamous cell carcinoma (SQCC). Least absolute shrinkage and selection operator (LASSO) was applied to identify features associated with XG status using integrated miRNA and mRNA abundance profiles. Gene Ontology (GO) annotation was subsequently performed to elucidate biological processes that may be altered between the two patient groups.

      Results:
      Using miRNA and mRNA data alone, ADC patients were classified as XG and non-XG with 88.7% and 95.2% accuracy. The integration of these two data types classified the patients with 100% accuracy using 20 features (7 miRNAs and 13 mRNAs). While less is known regarding the roles of the identified miRNAs in lung ADC, several of the genes have been suggested to affect the metastatic ability of lung cancer cells; these include PITX1, GPNMB and KRT14. In SQCC, both the miRNA and mRNA data alone and the integrated profiles were able to classify patients into XG and non-XG-forming groups with 100% accuracy. However, the roles of the selected features (1 miRNA and 11 mRNAs) in the metastasis of SQCC are not well defined. GO annotation of the identified mRNAs in ADC revealed enrichment of biological processes related to B cell differentiation, wound healing and regulation of the immune response and signalling pathway, while catabolic and metabolic processes were enriched in SQ.

      Conclusion:
      The use of single-dimensional data to classify patients into different prognostic groups may not be sufficient in the presence of heterogeneous patient populations. Integrative analysis of multi-omic data can provide greater insights into clinically relevant genetic aberrations, which can be used to improve the molecular classification of NSCLC.

    • +

      P1.05-027 - Novel Prognostic Gene Expression Signatures for Squamous Cell Lung Carcinoma: A Study by the SPECS Lung Consortium (ID 4490)

      14:30 - 14:30  |  Author(s): M.S. Tsao

      • Abstract

      Background:
      A multi-institutional squamous lung cancer consortium of investigators is developing prognostic signatures through the US NCI Lung SPECS (Strategic Partnership for Evaluation of Cancer Signatures) program. Six institutions contributed tumor specimens and published/unpublished expression-based prognostic signatures for validation using standardized sample cohorts (a primary validation cohort comprising institutional cases, and additional validation cohorts from two prospective cooperative group studies) and quality controlled assessment in independent laboratory and statistical cores. Here, we report on de novo prognostic signatures derived using the pooled institutional dataset.

      Methods:
      Highly quality-controlled cases of primary SCC from the pooled cohort (N=249) were analyzed to generate de novo prognostic signatures from among the 147 genes comprising pre-existing signatures, and from among all profiled genes. Minimax Concave Penalty (MCP) selection and Ward’s minimum variance clustering yielded survival analyses with 2 clusters that were evaluated using Cox regression and bootstrap cross validation (bCV; 500 iterations).

      Results:
      Two significantly prognostic models were generated (see Figure): Pooled Model A (PMA) was the optimal 2-cluster model using probesets representing 6 genes selected from components of pre-existing signatures: CASP8, MDM2, SEL1L3, RILPL1, LRR1, COPZ2. Pooled Model B (PMB) was the optimal 2-cluster model using probesets representing 6 genes selected from among all those profiled: SSX1, DIAPH3, LOC619427, CASP8, EIF2S1, HSPA13. PMA and PMB each remained independently prognostic in multivariable analyses incorporating an a priori baseline model (age, sex, stage; c-index = 0.641).

      Conclusion:
      Two de novo prognostic signatures were derived using a pooled multi-institutional cohort of SCC assembled for validation of pre-existing signatures. PMA and PMB were each found to be independently prognostic, accounting for established clinical predictors. Both now move forward, along with validated pre-existing signatures, to additional assessment of discrimination, calibration and clinical usefulness using additional independent prospective US co-operative group cohorts of cases. Figure 1



  • +

    P2.01 - Poster Session with Presenters Present (ID 461)

    • Event: WCLC 2016
    • Type: Poster Presenters Present
    • Track: Biology/Pathology
    • Presentations: 1
    • +

      P2.01-015 - Differentially Expressed microRNAs in Lung Adenocarcinoma Invert Effects of Copy Number Aberrations of Prognostic Genes (ID 4771)

      14:30 - 14:30  |  Author(s): M.S. Tsao

      • Abstract

      Background:
      Across multiple cancer histologies, many significantly down-regulated genes reside within chromosomal regions with increased number of copies, and vice versa. These “paradoxical genes” have been usually ignored as a noise, but could be a consequence of epigenetic regulatory mechanisms, including microRNA-mediated control of mRNA transcription.

      Methods:
      To identify paradoxical genes in lung adenocarcinoma (LUAD) we curated and analyzed gene expression and copy number aberrations across 1,064 LUAD samples, including newly-generated aCGH data from 65 samples. We then analyzed 9 LUAD microRNA expression studies to compile a list of consistently deregulated microRNAs. Finally, using microRNA:gene networks from mirDIP we examined possible association between microRNAs and paradoxical genes.

      Results:
      We identified 85 genes whose differential expression consistently contrasts the aberrations of their copy numbers. 70 genes were validated using TCGA-LUAD data. We showed that paradoxical expression of these genes is associated with 19 microRNAs, whose significant deregulation in LUAD has been consistently reported. Importantly, these genes form a clinically significant prognostic signature.Figure 1Figure 2





      Conclusion:
      Paradoxical gene expression, caused by microRNA deregulation, is preserved across patient cohorts, and forms a prognostic LUAD signature.

  • +

    P2.03b - Poster Session with Presenters Present (ID 465)

    • Event: WCLC 2016
    • Type: Poster Presenters Present
    • Track: Advanced NSCLC
    • Presentations: 4
    • +

      P2.03b-070 - Establishment of Organoid Cell Lines from Lung Squamous Cell Carcinoma (ID 5362)

      14:30 - 14:30  |  Author(s): M.S. Tsao

      • Abstract

      Background:
      The limited number of available lung squamous cell carcinoma (LUSC) cell lines poses significant challenge for biological, experimental therapeutic and biomarker research in LUSC. Novel approaches to establish new preclinical models are urgently needed. We have previously established patient-derived xenografts (PDX) from resected tumours of LUSC patients and characterized them on the genomic, transcriptomic, and proteomic levels. We have used these PDX models to develop a method for establishment of 3D organoid cultures and cell lines as new in vitro preclinical models of LUSC.

      Methods:
      PDX models were established and propagated from resected primary non-small cell lung cancer (NSCLC) in NOD/SCID mice; they were molecularly profiled by exome sequencing, SNP array for copy number analysis, and immunohistochemistry (IHC). PDX tissue harvested from mice was dissociated into single cells and plated in 100% matrigel dome, with overlaying media on top. Organoids were characterized by H&E, and IHC of p63, CK5/6, TTF-1, and CK7. Organoid growth rate and drug screening were assessed using Celltiter glo cell viability assay.

      Results:
      A total of 17 LUSC PDX models have been used for this study. All organoids were able to initiate in culture at passage 1, and the organoid establishment rate (beyond passage 4) is 50% (6/12). 4/12 (33%) LUSC organoids were able to be propagated beyond 10 passages for over 60 days with an average doubling rate of 2-3 days. Organoid tumour cells recapitulated the histological features of LUSC and were positive for p63 and CK5/6, and negative for TTF-1 and CK7 by IHC. Molecular characterization of LUSC PDX models revealed PIK3CA mutations, amplifications, and PTEN loss. Over 40% (4/9) of PI3K altered LUSC organoids were sensitive to PI3K inhibitor BKM120.

      Conclusion:
      LUSC organoids can be established for long term culture and recapitulate the phenotypic features of the PDX. The culture protocol is currently being tested on primary patient LUSC tumours. Organoid cultures and cell lines may be useful as additional preclinical models for functional validation of novel therapeutic targets in LUSC.

    • +

      P2.03b-071 - Therapeutic Targeting of the Phosphatidylinositol-3 Kinase Pathway in Lung Squamous Cell Carcinoma (ID 5369)

      14:30 - 14:30  |  Author(s): M.S. Tsao

      • Abstract

      Background:
      The phosphatidylinositol-3 kinase (PI3K) belongs to a family of lipid kinases involved in the regulation of cell proliferation and survival and is often dysregulated in cancer. Comprehensive molecular profiling by The Cancer Genome Atlas (TCGA) has identified PIK3CA mutations, amplifications, and the tumor suppressor PTEN loss in 30-40% of lung squamous cell carcinoma (LUSC) patients. Inhibitors of PI3K such as BKM120 have been initiated in BASALT-1 trial (NCT01820325) of PI3K activated LUSC, however with modest response rate (40% of patients with stable disease and 3.3% with partial response). We aim to assess the efficacy of PI3K inhibition in LUSC patient-derived xenografts (PDX) harboring different PI3K pathway alterations to identify potential mechanisms of innate resistance.

      Methods:
      PDX models were established from early stage LUSC patients and molecularly characterized via exome sequencing, SNP array for copy number variation (CNV) and gene expression analysis. PIK3CA mutations were validated by direct sequencing, amplifications by fluorescence in situ hybridization (FISH), and PTEN loss by immunohistochemistry (IHC). For in vivo drug screening, each PDX model was implanted in two mice; one treated with BKM120 (50mg/kg) and the other with vehicle control by daily oral gavage. Tumors were monitored twice weekly with caliper measurement. A responder is a tumor that regresses completely, shrinks more than 30%, or remains a stable size according to the RECIST criteria.

      Results:
      Of the 75 LUSC PDX models that our laboratory has established, 11 (14%) harbored PIK3CA E545K and E542K mutations, 36 (47%) harbored PIK3CA amplifications, and 23 (30%) showed loss of PTEN protein expression. Using the RECIST criteria, BKM120 screening in selected PDX models revealed stable disease and progressive disease in 4/9 (46%) and 5/9 (54%) of the PDX models, respectively, after 21 days of treatment. Of the 9 PDX models tested, 3/5 PIK3CA mutant models were responsive to BKM120, whereas none of the other 4 PIK3CA amplified and/or PTEN deleted models were responsive to BKM120. Additionally, downregulation of pErk1/2 and pS6 in a responder model and no change in phosphorylated proteins in non-responding models were observed. Pharmacodynamics studies, validation of responders with more mouse replicates, and testing on the remaining models are ongoing and the results will be reported.

      Conclusion:
      60% of LUSC PDXs with PIK3CA mutation demonstrate high sensitivity to pan-PI3K inhibitor. Understanding innate resistance mechanisms of PI3K inhibition may provide important insights on tractable targets and therapeutic strategy for LUSC patients with aberrant PI3K pathway.

    • +

      P2.03b-077 - EGFR/ALK+ Patient-Derived Xenografts from Advanced NSCLC for TKI Drug Selection & Resistance Development: The REAL-PDX Study (ID 6081)

      14:30 - 14:30  |  Author(s): M.S. Tsao

      • Abstract

      Background:
      Lung cancer patient-derived xenografts (PDX) have shown to be representative models for individual patient tumors. Theoretically, such models could inform the choice of subsequent lines of therapy, since PDX development, TKI resistance induction, and subsequent drug-screening can be completed before TKI resistance develops in the patient. The goal of Resistance modeling in EGFR and ALK Lung cancer (REAL)-PDX is to develop PDX models for real-time treatment selection of subsequent lines of therapy in advanced-stage NSCLC patients.

      Methods:
      Since August 2015, Princess Margaret Cancer Centre patients with EGFR/ALK+, as well as lifetime never-smoking lung cancer patients with unknown mutation status, were consented to have additional tumor sampling for PDX development during routine- or trial-related biopsies. Tumor sufficiency was confirmed prior to implantation into non-obese severe combined immunodeficient (NOD-SCID) mice, with successful engraftment defined as propagation beyond first passage; unsuccessful implantations had no palpable tumor after 6 months.

      Results:
      72/82 (88%) approached patients consented; 49/72 (68%) had adequate tumor tissue for implantation (71% stage III/IV): 46 adenocarcinomas, 2 squamous cell carcinoma, 1 LCNEC. 36/49 (73%) were lifetime never smokers. Patients received adjuvant chemotherapy (3), TKI therapy (15), both (5), or no treatment (26) prior to sampling. Tumor samples were taken from surgically resected lung (18), metastatic adrenal (1) and brain (2), CT-guided lung biopsies (5), endoscopic ultrasound-guided (EBUS) biopsies (6), and thoracentesis pleural fluid (17) specimens. Twenty-eight implanted tumors were EGFR+ (12 exon19 deletions, 2 exon19 deletion/T790M, 1 exon19 del/exon18 mutation, 12 L858R, and 1 L858R/T790M); 7 had ALK-rearrangements, and 1 had ROS1-rearrangement. Engraftment rates of 31 assessable implanted tumors were as follows: lung resections 12/12 (100%), metastatic resections 2/3 (67%), CT- or EBUS-guided biopsies 1/5 (20%), and pleural fluid 2/11 (18%); Engraftment rate was associated with no prior treatment (14/17 no treatment vs 3/14 any treatment, p=0.001). Of 17 assessable tumors with EGFR activating mutations, 9 engrafted (53%). Of 3 assessable tumors with ALK-rearrangement, 1 was successful (33%).

      Conclusion:
      PDX development of EGFR/ALK+ models for testing with novel therapeutics from various tumor biopsy sites is feasible and will provide valuable real-time information for subsequent treatment decisions in advanced NSCLC patients. Updated engraftment and drug screening data will be presented.

    • +

      P2.03b-089 - CD1C in Lung Adenocarcinoma: Prognosis and Cellular Origin (ID 4809)

      14:30 - 14:30  |  Author(s): M.S. Tsao

      • Abstract

      Background:
      Adaptive immune response is critical for cancer surveillance and elimination. Dendritic cells (DC) arise from a hematopoietic lineage distinct from other leukocytes which play a central role in adaptive immunity. CD1C is expressed in DC, presenting exogenous lipid antigens to T cell receptor to activate “unconventional” T cells. This study aims to evaluate the cellular expression and prognostic value of CD1C.

      Methods:
      The study used 5 gene expression datasets: UHN181 [lung adenocarcinoma (ADC, n=128), squamous cell carcinoma (SqCC, n=43)], GSE30219 (ADC n=81, non-ADC n=138)], and 3 integrated cohorts [non-SqCC NSCLC (n=1106), PRECOG (39 types of cancer, n=~18,000), and TCGA (33 types of cancer, n=11,000)]. Cancer Cell Line Encyclopedia (CCLE) data were used to determine if CD1C was expressed by cancer cell lines. CIBERSORT algorithm was used to estimate immune cell fraction and Cox proportional model was used to evaluate the association of CD1C expression with survival. Immunohistochemistry (IHC) was used to measure protein expression of CD1C.

      Results:
      Except for hematopoietic and lymphoid cancer cell lines, all CCLE cell lines lack CD1C expression. CIBERSORT analysis together with Pearson correlation analyses on the ADC cases in UHN181, the integrated cohort, and GSE30219 showed that CD1C was expressed by DC. IHC showed staining with a dendritic cell shape pattern. However, the staining of CD1C did not overlapped with CD11c staining, suggesting a specific DC subtype. Cox proportional regression revealed that CD1C was significantly prognostic in the UHN181 ADC cohort (HR=0.75, p=0.05) as the training set. When CD1C expression was categorized into 3 equal groups, the risk of death was reduced in high compared to low CD1C expression group (HR=0.55, 95%CI 0.28-1.07, p=0.07). CD1C is protective only in PD-L1 low expression group (n=108, HR=0.37, 95%CI 0.15-0.89, p=0.026). The favorable prognosis associated with CD1C expression was validated in the integrated cohort of non-SqCC NSCLC (HR=0.55, 95% CI 0.43-0.72, p<0.0001), and in GSE30219 ADC cohort (HR=0.30, 95% CI 0.11-0.84, p=0.02). In PRECOG and TCGA datasets, high CD1C expression is significantly good prognostic in all cancer types (p<1×10[-7] and p<0.001, respectively), suggesting a universal protective role of CD1C expression in cancers. CD1C IHC score was highly correlated with CD1C mRNA expression in ADC patients of UHN181 and was prognostic (HR=0.46, 95%CI 0.22-0.96, p=0.039).

      Conclusion:
      CD1C preferentially is expressed on a subset of DCs and higher expression of CD1C is significant protective factor in all cancer types, especially in lung adenocarcinoma

  • +

    P3.01 - Poster Session with Presenters Present (ID 469)

    • Event: WCLC 2016
    • Type: Poster Presenters Present
    • Track: Biology/Pathology
    • Presentations: 2
    • +

      P3.01-019 - Desmoplasia is Associated with Poor Prognosis and Carcinoma-Associated Fibroblast Heterogeneity in Non-Small Cell Lung Cancer (ID 5281)

      14:30 - 14:30  |  Author(s): M.S. Tsao

      • Abstract

      Background:
      Cancer-associated fibroblasts (CAFs) are known to influence tumor development, progression and metastasis. Their characteristics and prognostic role in non-small cell lung cancer (NSCLC) patients have been recognized. However, the functional heterogeneity of CAFs between patients and its genetic basis is less understood.

      Methods:
      Two pathologists scored for desmoplasia on Hematoxylin-Eosin stained sections of resected lung tumors from two patient cohorts: one consisting of 171 NSCLC patients (128 adenocarcinoma, 43 squamous carcinoma) and the second of 24 primary cultures of CAFs. Percent area of desmoplasia among total tumor stroma was used to define high desmoplasia (HD) versus low desmoplasia (LD). The desmoplasia and survival analysis were assessed for 171 NSCLC patients. Gene expression data on RNA extracted from CAFs in contracted gels following 24 hours incubation was obtained using Illumina Human HT-12v4 Bead Chips array and was preprocessed and normalized using RMA and values were log2 transformed. Significant genes whose expression is strongly correlated (Spearman correlation coefficient and p value) with percent of desmoplasia were identified in both cohorts. The gene set enrichment analysis (GSEA) was applied to test for the enrichment of CAF cohort significant genes in 171 NSCLC cohort. Additionally, CAF significant genes were subjected to pathway enrichment analysis using Pathway Data Integration Portal ver. 2.5 (http://ophid.utoronto.ca/pahtDIP).

      Results:
      The prognosis of adenocarcinoma patients with HD had poorer outcome in comparison to the patients with LD (disease free survival at 3 years 34% vs. 75% p=0.00045 and relapse rate 41% vs. 14%, p=0.0051). In the CAF cohort, the number of genes that are significantly associated with desmoplasia for enrichment are 356. Using GSEA, these genes were enriched in 171 NSCLC cohort (with a p value of 0.045). Protein-protein interaction (PPI) partners of these 356 genes were acquired using Integrated Interaction Database – IID (version 2016-03, http://dcv.uhnres.utoronto.ca/iid/). Obtained genes were then ranked according to their degree, i.e., number of PPIs. Top 44 (top 1%) of the genes were then selected to pathway enrichment analysis using pathDIP version 2.5. 245 pathways that significantly enriched by these 44 genes (FDR < 0.01) were obtained. Many of these pathways are known to be involved in lung cancer.

      Conclusion:
      We demonstrated that the prognosis of lung adenocarcinoma patients with HD had poorer outcome in comparison to the patients with LD. Furthermore, PPI analysis of CAF genes associated with HD reveals enrichment of many cancer-related pathways, suggesting their high relevance to lung cancer.

    • +

      P3.01-021 - Reproducibility of Comprehensive Histologic Assessment and Refining Histologic Criteria in P Staging of Multiple Tumour Nodules (ID 5365)

      14:30 - 14:30  |  Author(s): M.S. Tsao

      • Abstract
      • Slides

      Background:
      Multiple tumor nodules (MTNs) are being encountered, with increasing frequency with the 8[th] TNM staging system recommending classification as separate primary lung cancers (SPLC) or intrapulmonary metastases (IM). Pathological staging requires assessment of morphological features, with criteria of Martini and Melamed supplanted by comprehensive histologic assessment of tumour type, predominant pattern, other histologic patterns and cytologic features. With publication of the 2015 WHO classification of lung tumours, we assessed the reproducibility of comprehensive histologic assessment and also sought to identify the most useful histological features.

      Methods:
      We conducted an online survey in which pathologists reviewed a sequential cohort of resected multifocal tumours to determine whether they were SPLC, IM, or a combination. Specific histological features for each nodule were entered into the database by the observing pathologist (tumour type, predominant adenocarcinoma pattern, and histological features including presence of lepidic growth, intra-alveolar cell clusters, cell size, mitotic rate, nuclear pleomorphism, nucleolar size and pleomorphism, nuclear inclusions, necrosis pattern, vascular invasion, mucin content, keratinization, clear cell change, cytoplasmic granules¸ lymphocytosis, macrophage response, acute inflammation and emperipolesis). Results were statistically analyzed for concordance with submitting diagnosis (gold standard) and among pathologists. Consistency of each feature was correlated with final determination of SPLC vs. IM status (p staging) by chi square analysis and Fisher exact test.

      Results:
      Seventeen pathologists evaluated 126 tumors from 48 patients. Kappa score on overall assessment of primary v. metastatic status was 0.60. There was good agreement as measured by Cohen’s Kappa (0.64, p<0.0001) between WHO histological patterns in individual cases with SPLC or IM status but proportions for histology and SPT or IM status were not identical (McNemar's test, p<0.0001) and additional histological features were assessed. There was marked variation in p values among the specific histological features. The strongest correlations (<0.05) between p staging status and histological features were with nuclear pleomorphism, cell size, acinus formation, nucleolar size, mitotic rate, nuclear inclusions, intra-alveolar clusters and necrosis pattern. Correlation between lymphocytosis, mucin content, lepidic growth, vascular invasion, macrophage response, clear cell change, acute inflammation keratinization and emperipolesis did not reach a p value of 0.05.

      Conclusion:
      Comprehensive histologic assessment shows good reproducibility between practicing lung pathologists. In addition to main tumour type and predominant patterns, nuclear pleomorphism, cell size, acinus formation, nucleolar size, and mitotic rate appear to be useful in distinguishing between SPLC and IM.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

  • +

    P3.02b - Poster Session with Presenters Present (ID 494)

    • Event: WCLC 2016
    • Type: Poster Presenters Present
    • Track: Advanced NSCLC
    • Presentations: 1
    • +

      P3.02b-028 - Characterizing Residual Erlotinib-Tolerant Population Using EGFR-Mutated NSCLC Primary Derived Xenografts: The Last Holdouts (ID 5455)

      14:30 - 14:30  |  Author(s): M.S. Tsao

      • Abstract

      Background:
      Three generations of epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) have led to multi-fold improvements in progression free survival of advanced stage non-small cell lung cancer (NSCLC) patients carrying EGFR kinase domain mutations. However, cure is not yet achievable with any EGFR TKI monotherapy, as patients will eventually progress due to acquired resistance. In vitro evidence suggests that minor populations of epigenetically modified drug tolerant cells (DTCs) may be one important mechanism for tumor cells surviving the TKI. We hypothesize that characterizing the genomic and epigenomic alterations observed in DTCs in vivo and comparing them to the bulk tumour will delineate a number of mechanisms of tolerance exhibited by DTCs.

      Methods:
      DTCs were induced via chronic erlotinib treatment of a lung adenocarcinoma primary derived xenograft (PDX) harbouring an erlotinib sensitive exon 19 deletion. Molecular profiles of DTCs are compared to untreated controls via immunohistochemistry (IHC) and gene expression array. We are now undertaking exome-sequencing, assay for transposase-accessible chromatin with high-throughput sequencing (ATAC-seq), methylated DNA immunoprecipitation and sequencing (MeDIP-seq).

      Results:
      When compared to untreated tumours, DTCs exhibit decreased apoptosis (CC3 IHC) and proliferation (Ki67 IHC). DTCs maintained strong signaling via the EGFR pathway (pERK, pAKT, pS6). DTCs exhibited 2437 significantly differentially expressed genes (DEGs; >1.5-fold change and adjusted p-value <0.05) including multiple cancer stem cell markers (ALDH1A1, ALDH1A3, CD44). DEGs also were involved in vesicle-mediated transport (including lysosomes, exosomes and endosomes), autophagy, stress/unfolded protein response, cytoskeleton organization, chromatin organization, ion pumps and transporters, cell adhesion, WNT, NOTCH, PI3K and MAPK pathways. DTCs remained resistant to three cycles of cisplatin/vinorelbine either alone or when combined with erlotinib. Genomic and epigenomic profiling are on-going and results will be presented.

      Conclusion:
      DTCs may be a major impediment to cure by single-agent EGFR targeted therapies. Understanding the mechanisms and developing strategies to overcome DTCs may give insights on therapeutic strategy to further improve the survival of EGFR-mutated NSCLC patients.