K.J. O'Byrne

PARP inhibition induces BAX/BAK-independent synthetic lethality of BRCA1-deficient non-small cell lung cancer.

Evasion of apoptosis contributes to both tumourigenesis and drug resistance in non‐small cell lung carcinoma (NSCLC). The pro‐apoptotic BCL‐2 family proteins BAX and BAK are critical regulators of mitochondrial apoptosis. New strategies for targeting NSCLC in a mitochondria‐independent manner should bypass this common mechanism of apoptosis block. BRCA1 mutation frequency in lung cancer is low; however, decreased BRCA1 mRNA and protein expression levels have been reported in a significant proportion of lung adenocarcinomas. BRCA1 mutation/deficiency confers a defect in homologous recombination DNA repair that has been exploited by synthetic lethality through inhibition of PARP (PARPi) in breast and ovarian cells; however, it is not known whether this same synthetic lethal mechanism exists in NSCLC cells. Additionally, it is unknown whether the mitochondrial apoptotic pathway is required for BRCA1/PARPi‐mediated synthetic lethality. Here we demonstrate that silencing of BRCA1 expression by RNA interference sensitizes NSCLC cells to PARP inhibition. Importantly, this sensitivity was not attenuated in cells harbouring mitochondrial apoptosis block induced by co‐depletion of BAX and BAK. Furthermore, we demonstrate that BRCA1 inhibition cannot override platinum resistance, which is often mediated by loss of mitochondrial apoptosis signalling, but can still sensitize to PARP inhibition. Finally we demonstrate the existence of a BRCA1‐deficient subgroup (11–19%) of NSCLC patients by analysing BRCA1 protein levels using immunohistochemistry in two independent primary NSCLC cohorts. Taken together, the existence of BRCA1‐immunodeficient NSCLC suggests that this molecular subgroup could be effectively targeted by PARP inhibitors in the clinic and that PARP inhibitors could be used for the treatment of BRCA1‐immunodeficient, platinum‐resistant tumours. Copyright

7: Normal bronchial epithelial cells exhibit altered cellular behaviour when exposed to lung cancer cells in a co-culture system

Introduction: Non-small cell lung cancer is a complex solid tumour displaying significant heterogeneity between patients. Cancer cells may influence the behaviour of adjacent normal cells and this may contribute to the complexity of this disease. The objective of this study was to improve our understanding of this relationship. n nMethods: A normal bronchial epithelial cell line (HBEC4) was exposed to adenocarcinoma (A549), large cell (H460) and squamous (SK-MES-1) non-small cell lung cancer cell lines for a period of seven days in a transwell co-culture system. Subsequently, the HBEC4 cells were assessed for alterations in cellular behaviour (viability/apoptosis/cell cycle/exosome response) and changes in gene expression (RT-PCR/miRNA). n nResults: Following seven days of co-culture, the HBEC4 cells displayed changes in cell number, cellular viability and cell cycle. No significant differences in apoptosis were detected. Gene expression profiles were altered in inflammatory and stem cell pathways, with no changes in NSCLC subtype markers detected. In addition, a unique pattern of miRNA expression was observed in the HBEC4 cells. This varied depending on NSCLC exposure. n nConclusion: Lung cancer cells can alter the behavior of normal bronchial epithelial cells through multiple mechanisms. This may contribute to the heterogeneity observed in this disease.

6 Cisplatin induces the emergence and expansion of a distinct cancer stem cell (CSC) population in NSCLC

line expression of the miRNAs, the miR-30 family members and miR34a-5p were up-regulated in the CisR xenograft FFPE tissue relative to PT. Conclusion: A novel miRNA signature associated with cisplatin resistance was identified in vitro, genetic manipulation of which enhanced the cytotoxic effects of cisplatin. The 5-miR signature shows both diagnostic and prognostic biomarker potential across a number of biological mediums. Disclosure: All authors have declared no conflicts of interest.

4 Identification and targeting of the DNA repair gene, XRCC6BP1, in cisplatin resistant NSCLC

A. Pender1, S. Rana1, E. Izquierdo Delgado2, P. Proszek2, I. GarciaMurillas3, J. Bhosle4, M. O’Brien4, J.F. Palma5, N.C. Turner6, S. Popat6, J. Downward1, D. Gonzalez2. 1Lung Cancer Group, The Institute of Cancer Research, London, United Kingdom; 2The Centre for Molecular Pathology, The Royal Marsden NHS Foundation Trust, Sutton, United Kingdom; 3Molecular Oncology Group, The Institute of Cancer Research, London, United Kingdom; 4Department of Medicine, The Royal Marsden NHS Foundation Trust, London, United Kingdom; 5Genomics & Oncology, Roche Molecular Systems, Inc, Pleasanton, AL, United States of America; 6Medicine, Royal Marsden Hospital, London, United Kingdom

10 Development and characterization of a panel of NSCLC cell lines resistant to PI3K/mTOR inhibitor GDC-0980

analyses were undertaken to measure predictive performance of mutation testing in CTCs. Results: The DNA extracted from CTCs, were of sufficient quality to allow mutation analyses to be successfully performed in 100%, 99%, and 98% of samples for EGFR, KRAS, and BRAF genes, respectively. In CTC DNA, the KRAS mutation rate (codons 12/13 and 61) was 9.1% and concordance with the primary tumour was 78.8%. Six mutations were detected in CTCs, but not in primary tumours, and 13 mutations in primary tumours were not detected in corresponding CTC samples. Three mutations were detected in matched CTC and primary tumour specimens. One mutation in EGFR was detected in CTC DNA and 3 mutations were detected in primary tumours. In all cases, the mutations were detected in discordant specimens. The concordance between mutations detection in CTCs and primary tumours was 95.8%. BRAF V600E mutation was not detected in any sample. In general, the results suggested low sensitivity but high specificity (Table). Due to low number of EGFR mutations detected, test performance results require further validation.

8 PIK3CA mutations and response to PI3K and MEK targeted therapies in NSCLC

isolating CTCs are still in their infancy. One such method is the use of the modified invasion assay, VitaAssay. This technique uses CAM (Collagen Adhesion Matrix) coated plates to capture CTCs with an invasive phenotype. Another method of CTC detection is ScreenCell. This technique uses a polycarbonate filtration membrane containing multiple tiny pores. When blood flows across the membrane, tumour cells are captured due to their greater size. Method: Peripheral blood samples were obtained from patients with advanced NSCLC. In addition healthy blood samples spiked with NSCLC cell line cells were also analysed using both VitaAssay and Screencell. VitaAssay: Peripheral blood mononuclear cells (PBMCs) were obtained by Ficoll density centrifugation. PBMCs were seeded onto VitaAssay plates and cultured for 12 18 hrs. The supernatant is removed and the remaining captured cells are enriched for CTCs. Captured cells can be fixed and stained using immunocytochemistry. ScreenCell: Peripheral blood is diluted with a specified buffer and filtered across the Screencell filter. The filter with captured cells is mounted on a slide for fixation and staining. Fixed CTCs on the filters are then stained using H&E and/or immunocytochemistry. Results: Cells captured from NSCLC patients using VitaAssay were stained for EpCAM/pan-Cytokeratin and CD45. EpCAM/Pan-CK positive, CD45 negative cells were classed as CTCs. In healthy blood samples spiked with A549 and H2228 cells, approximately 20% (range 9 26.4%) of spiked cells were recovered using VitaAssay. In NSCLC patients an average of 30.67 CTCs per ml of blood were identified (range 14 52, n = 6). Using the ScreenCell technique CTCs were identified by size & morphology using H&E staining. In cell spiking experiments, approximately 70% of spiked cells were recovered using Screencell. Also, CTCs were detected in 70% of patient samples (n = 10). Numbers of CTCs detected ranged from 6 82 per ml of blood. In addition, clumps of tumour cells or Circulating Tumour Microemboli were detected in 50% of patient samples. Conclusions: The VitaAssay and ScreenCell techniques both appear to be viable method of isolating CTCs in NSCLC, in both model cell-spiking experiments and in clinical samples as determined by morphology and antigen expression detected with immunocytochemistry.

9 Elucidating the role of the NFkB pathway in cisplatin resistant NSCLC

isolating CTCs are still in their infancy. One such method is the use of the modified invasion assay, VitaAssay. This technique uses CAM (Collagen Adhesion Matrix) coated plates to capture CTCs with an invasive phenotype. Another method of CTC detection is ScreenCell. This technique uses a polycarbonate filtration membrane containing multiple tiny pores. When blood flows across the membrane, tumour cells are captured due to their greater size. Method: Peripheral blood samples were obtained from patients with advanced NSCLC. In addition healthy blood samples spiked with NSCLC cell line cells were also analysed using both VitaAssay and Screencell. VitaAssay: Peripheral blood mononuclear cells (PBMCs) were obtained by Ficoll density centrifugation. PBMCs were seeded onto VitaAssay plates and cultured for 12 18 hrs. The supernatant is removed and the remaining captured cells are enriched for CTCs. Captured cells can be fixed and stained using immunocytochemistry. ScreenCell: Peripheral blood is diluted with a specified buffer and filtered across the Screencell filter. The filter with captured cells is mounted on a slide for fixation and staining. Fixed CTCs on the filters are then stained using H&E and/or immunocytochemistry. Results: Cells captured from NSCLC patients using VitaAssay were stained for EpCAM/pan-Cytokeratin and CD45. EpCAM/Pan-CK positive, CD45 negative cells were classed as CTCs. In healthy blood samples spiked with A549 and H2228 cells, approximately 20% (range 9 26.4%) of spiked cells were recovered using VitaAssay. In NSCLC patients an average of 30.67 CTCs per ml of blood were identified (range 14 52, n = 6). Using the ScreenCell technique CTCs were identified by size & morphology using H&E staining. In cell spiking experiments, approximately 70% of spiked cells were recovered using Screencell. Also, CTCs were detected in 70% of patient samples (n = 10). Numbers of CTCs detected ranged from 6 82 per ml of blood. In addition, clumps of tumour cells or Circulating Tumour Microemboli were detected in 50% of patient samples. Conclusions: The VitaAssay and ScreenCell techniques both appear to be viable method of isolating CTCs in NSCLC, in both model cell-spiking experiments and in clinical samples as determined by morphology and antigen expression detected with immunocytochemistry.

1 Elucidating the role of inflammatory mediators in cisplatin resistant NSCLC

1 Elucidating the role of inflammatory mediators in cisplatin resistant NSCLC A.M. Baird1,2 *, P. Godwin1,2, S. Heavey1,2, K. Umezawa3, M.P. Barr2,4, K.J. O’Byrne2,4, K. Gately2,4. 1Dept. of Clinical Medicine, Trinity College Dublin, Ireland, 2Thoracic Oncology Research Group, Institute of Molecular Medicine, St. James’s Hospital, Dublin, Ireland, 3Dept. of Molecular Target Medicine Screening, Aichi Medical University, Aichi, Japan, 4HOPE Directorate, St. James’s Hospital, Dublin, Ireland

3 The establishment of an isogenic model of cisplatin resistance and the identification of a potential microRNA signature in non-small cell lung cancer

Introduction: Hypoxia and chronic inflammation are key triggers in the transformation process with at least 20% of all malignancies initiated or exacerbated by inflammation. The aim of this study was to examine inflammation in the process of lung carcinogenesis with a particular emphasis on TNF-a, IL-1b and hypoxia. This study developed an isogenic cell line model in which to examine their effects on normal bronchial epithelial cells, with the ultimate aim to modify the cells’ phenotype from normal to malignant. Method: A normal bronchial epithelial cell line (HBEC4) was modified to functionally over-express TNF-a and IL-1b (alone or in combination) and were continuously cultured for three months under normoxic or hypoxic (0.5% oxygen) conditions. Subsequently a range of experimental assays were carried out to assess functional cell change. These included: soft agar, proliferation, invasion, transformation, migration and angiogenesis. Results: Although transformation was not evident by soft agar, the expression levels of c-Myc and p53 have changed. The adhesion potential of normoxic and hypoxic clones has amplified and the growth rate has increased over time. Under normoxia the IL-1b and the TNF-a/IL-1b clones displayed an increased invasive capacity compared with the empty vector control (p < 0.05). Differences were also detected in the gene expression profile implicated in other pathways involved in the hallmarks of cancer cell signalling, apoptosis, angiogenesis and cell cycle regulation. Conclusions: Although malignancy is not yet evident in these cells, there are distinct indications that phenotypic changes occurred within the three month time frame. As pro-longed chronic exposure to inflammation is a pre-requisite for many disease states, these results warrant extended growth studies to further delineate the complex roles of TNF-a, IL-1b and hypoxia in the process of carcinogenesis. This will assist in the development of novel cancer target therapeutics and chemo-preventive agents in lung cancer. 2 IL-23A is epigenetically regulated in non small cell lung cancer A.-M. Baird1,2 *, E. Dockry2, J. Leonard2, L. Kilmartin2, K. O’Byrne2,3, S. Gray2,3. 1Dept. of Clinical Medicine, Trinity College Dublin, Dublin, Ireland, 2Thoracic Oncology Research Group, Institute of Molecular Medicine, St. James’s Hospital, Dublin, Ireland, 3HOPE Directorate, St. James’s Hospital, Dublin, Ireland