Virtual Library

Background:
The use of new, emerging techniques in the search of tailored patient therapies is rapidly becoming a reality. Here we describe the optimization and implementation of next generation sequencing for treatment predictive mutation screening in parallel with gene fusion status of ALK, RET and ROS1 in non-small cell lung cancer (NSCLC) patients.

Methods:
The Illumina TruSight tumor 26-gene NGS panel was validated in 81 clinical routine FFPE or cytology specimens and implemented in 533 diagnostic NSCLCs during one year of clinical analysis. In parallel, a RNA-based NanoString method was evaluated in 169 cases for gene fusion status of ALK, RET and ROS1.

Results:
We have successfully established a streamlined workflow with a 5-day turnaround time from specimen arrival to mutation report. The concordance in the validation cohort was 99% for comparable variants. In the 533 diagnostic samples, 1-2 variants were detected in 79% of the cases. Most frequently mutated genes included TP53, KRAS, EGFR, STK11, and BRAF, all with differences in mutational patterns between histological subgroups. The RNA-based NanoString assay was successfully established and validated. The success rate in the 169 cases was 80% and 10 gene fusions were found (five ALK fusions, three RET fusions and two ROS1 fusions) all in adenocarcinomas. Integration of mutation and gene fusion status revealed that 68% of adenocarcinomas, 13% of SqCCs and 56% of NSCLC-NOS harbored ≥1 actionable alteration ALK, RET, ROS1, EGFR, KRAS, PIK3CA, BRAF, NRAS, MAP2K1, ERBB2 or AKT1. Specifically, in 13.2% of the adenocarcinomas where no EGFR or ALK alteration was detected emerging targeted therapy may be considered in addition to the 15.3% of patients that was eligible for EGFR or ALK inhibitors. The corresponding proportions for SqCCs were 5.5% in addition to the 2.2%, and for NSCLC-NOS 2.5% in addition to the 11.2% eligible for EGFR or ALK inhibitors.

Conclusion:
Next generation sequencing in combination with the NanoString technology is time- and cost efficient in the diagnostic routine for treatment predictive mutation screening and gene fusion status detection. The techniques represent valuable tools for pinpointing patients eligible to standard targeted therapies in addition to new emerging therapies.

Background:
Our laboratory has a unique cohort of patients with pre-invasive lung squamous cell carcinoma (SqCC) lesions, within which there is a clear discrepancy between the prevalence of pre-invasive lesions and the incidence of lung cancer, suggesting that not all pre-invasive lesions progress to cancer. Using gene expression microarrays we identified 1846 genes significantly differentially expressed between progressive and regressive pre-invasive SqCC lesions. The macrophage metalloelastase MMP12 gene was found to be highly expressed in progressive lesions, and we hypothesised that it plays a role in epithelial-to-mesenchymal transition (EMT). Conversely, the actin binding protein LIM-domain only 7 (LMO7) gene was highly expressed in regressive lesions, and we postulated that it may be protective against EMT due to its role in the maintenance of epithelial architecture. Initial studies using three SqCC cell lines (A431, H357 and H376) with MMP12-shRNA knockdown showed a significant decrease in migration and invasion compared to non-silencing shRNA controls. LMO7-shRNA knockdown in HBECs was found to significantly increase migration. The aim of this study is to further characterise the function and signalling of MMP12 and LMO7 in lung SqCC development.

Methods:
Eight-week-old NOD/SCID mice were used for tumorigenesis experiments. A431 and H357 MMP12-shRNA knockdown and non-silencing shRNA cells were injected in a suspension of one million cells in a total of 200μl, subcutaneously in the right and left flank, respectively. Tumours were measured every 2–5 days. Adhesion assays were carried out to assess the roles of MMP12 knockdown or LMO7 overexpression on cell adhesion. Cell signalling mechanisms were assessed using western blotting, qPCR and immunostaining.

Results:
We observed that MMP12 knockdown decreases tumorigenicity in an immunocompromised mouse model. Both A431 and H357 MMP12 knockdown cells produced significantly smaller tumours compared with non-silencing shRNA cells. We found that MMP12 knockdown decreases cell adhesion, which is currently being further investigated along with effects on integrin signalling pathways. Levels of EMT markers were assessed in MMP12 knockdown and LMO7 overexpressing cells using qPCR, western blotting and immunostaining. Results indicate that higher MMP12 expression is associated with a mesenchymal phenotype, whereas higher LMO7 expression is associated with an epithelial phenotype.

Conclusion:
Our results suggest that MMP12 is a key driver of migration and invasion in SqCC and its high expression may contribute to EMT, whereas LMO7 is a putative tumour suppressor with a crucial role in maintaining epithelial cell architecture. MMP12 and LMO7 may be potential therapeutic markers for lung cancer at an early stage.

Abstract not provided

Background:
Patients with COPD and/or emphysema have an increased risk of non-small cell lung cancer (NSCLC). COPD and lung cancer are both characterised by increased oxidative stress associated with mitochondrial dysfunction. We hypothesise that mitochondrial dysfunction is a driving mechanism for the increased risk of NSCLC in COPD. We determined whether there is dysregulated expression of mitochondrial-related proteins in NSCLC arising in COPD, and if so, their clinical significance.

Methods:
To determine the clinical relevance of mitochondrial related gene expression, we examined a database containing transcriptomic data of more than 1, 000 human NSCLC samples and with survival outcomes (https://precog.stanford.edu/). Immunohistochemistry for PGAM5 and FUNDC1 was performed on cancer and background (‘normal’) tissue from lung cancer resections from non-smokers, healthy smokers (without COPD) and COPD/ emphysema patients. Protein expression was assessed using a semi-quantitative immunohistochemical scoring system (H score). Specific gene expression was further correlated with outcome in dataset GSE 72194, containing transcriptomic data of NSCLC cases and patient survival.

Results:
25 mitochondrial-related genes were linked to survival in NSCLC. Of those 25, we chose to study further the expression of PGAM5 and FUNDC1, which are regulators of mitochondrial degradation (mitophagy). In background lung tissue, PGAM5 and FUNDC1, only expressed in alveolar macrophages, were most highly expressed in COPD (H score: 180 ± 58 and 23 ± 9, respectively) compared to healthy smokers (146 ± 58 and 20 ± 8) and non-smokers (68 ± 48 and 3.3 ± 1.4) (p<0.05). In cancerous tissue, only the malignant epithelial cells and associated macrophages, at the periphery of the cancer, expressed PGAM5 and FUNDC1. PGAM5 was also expressed in pre-neoplastic epithelium (squamous dysplasia and carcinoma in situ). There was no difference in expression across the 3 groups, although the macrophages, at the edge of cancer, from COPD patients tended to show higher expression of PGAM5 and FUNDC1, compared to those from the other groups. When the expression of PGAM5 was compared with that of 50 known macrophage transcriptomic signatures within NSCLC samples, there was a positive correlation between PGAM5 and 9 macrophage signatures (r= 0.27 - 0.44, p<0.05), with one a determinant of patient survival.

Conclusion:
PGAM5 expression in pre-neoplastic tissue and NSCLC, but not in normal epithelium, suggests it plays a role in the transformation of malignant epithelial cells. PGAM5 and FUNDC1 may contribute to the pathogenesis of both COPD and NSCLC, possibly through mitophagic processes.

Background:
We and others have reported the prognostic impact of tumor spread through air spaces (STAS) in lung adenocarcinomas. The goal of this study is to investigate preoperative predicting factors for STAS and to determine whether STAS can be detected by intraoperative frozen section analysis.

Methods:
In a cohort of 874 patients with small (≤2cm) stage I adenocarcinoma (1995-2012), we reviewed preoperative computed tomography (CT) and positron emission tomography (PET) scans. According to the 2016 Fleischner Society’s criteria, radiological whole tumor size, consolidation size, as well as C/T ratio (consolidation/whole tumor diameter) were determined using thin slice (<3mm) CT scans where available (n=174). Clinico-radiological prediction of STAS was evaluated by logistic regression model. Using the frozen section slides with adequate adjacent lung parenchyma surrounding tumor without artifact (n=48), the presence of STAS was evaluated by five pathologists who are unaware of the radiological findings or the pathological information on permanent slides. The kappa statistic was calculated to measure the agreement between two pathologists.

Results:
In univariable model for predicting STAS, current smoker, larger consolidation tumor size, C/T ratio, and SUVmax were significant variables. In multivariable model, current smoker and C/T ratio were independent risk factors for the presence of STAS (p=0.027 and p<0.001, respectively; Table 1a). The sensitivity and the specificity of frozen section for prediction of STAS were 71% (95% confidence interval: 52-91%), 92.4% (81-100%) respectively, and the accuracy was 80% (71-89%). The kappa statistics were 0.40-0.74 (Table 1b) with 8/10 being moderate or substantial agreement.

Conclusion:
Smoking status and C/T ratio were independent predictors for the presence of STAS in patients with small lung adenocarcinomas. Frozen section prepared with adequate surrounding normal lung tissue may help identify STAS intraoperatively. Figure 1

Background:
Tumor spread through air spaces (STAS) is a recently recognized pattern of invasion in lung adenocarcinoma, however, the incidence of and prognostic importance of STAS have not yet been defined in squamous cell carcinoma (SCC).

Methods:
In a cohort of 445 patients with p-stage I-III lung SCC, cumulative incidence of recurrence and lung cancer-specific death (LCSD) was evaluated by competing risks analysis and overall survival (OS) by Cox models.

Results:
76% of patients were >65 years of age. 273 patients died during follow up, one third (91, 33.3%) died of lung cancer whereas two thirds died of competing events or unknown cause. STAS was present in 132 (30%). The cumulative incidence of any, distant, and locoregional recurrence as well as LCSD were significantly higher in patients with STAS compared to those without STAS (Figure), whereas there was no statistically significant difference in OS. STAS was an independent predictor for both recurrence and LCSD in multivariable analysis (p=0.034 and 0.016, respectively, Table).

Conclusion:
STAS was present in one third of resected lung SCC and it was an independent predictor of recurrence and LCSD, supporting our proposal that STAS is a clinically important pattern of invasion and not an artifact. Figure 1 Figure 2

Abstract not provided

Background:
In early-stage lung adenocarcinomas, high-grade micropapillary (MIP) and solid (SOL) predominant pathology is known to be associated with worse prognosis. The aim of this study is, in addition to predominant patterns, to investigate clinical impact of the presence of small amounts (≥5%) as well as increasing percentage of high-grade patterns.

Methods:
Invasive tumors from early-stage lung adenocarcinoma patients who underwent curative-intent resection with no induction therapy were investigated (N=2017; 1995-2012) (8[th] edition TNM pStage I=1390, II=357, III=270). In 388 cases, synchronous lymph node (LN) metastases were available. Histological subtype (lepidic [LEP], acinar [ACI], papillary [PAP], MIP, or SOL) percentages were stratified into 4 groups; 0-4%, 5-24%, 25-49%, and 50-100%. The association between increasing percentage of patterns of primary tumor and the incidence of lymphatic/vascular invasion, necrosis, tumor spread through air spaces (STAS) as well as estimated 5-year cumulative incidence of recurrence (CIR) were analyzed. The differences in distribution of each pathological variable between 4 groups was analyzed by Chi-square test. The percentages of histological pattern were compared between primary tumor and LN metastasis.

Results:
Increasing percentage of MIP pattern is associated with increasing incidence of lymphatic/vascular invasion, STAS, as well as 5-year CIR (Figure 1a, p<0.001). Increasing percentage of SOL pattern is associated with increasing incidence of necrosis and 5-year CIR (p<0.001). Presence (≥5%) of SOL pattern is associated with higher incidence of lymphatic/vascular invasion and STAS (p<0.001) compared to the absence (<5%) of SOL pattern, but no significant relationship between lymphatic/vascular invasion and proportion of SOL pattern. The percentage of SOL pattern in LN metastasis is higher than that in synchronous primary tumors (Figure 1b).

Conclusion:
In early-stage lung adenocarcinomas, presence (≥5%) of MIP or SOL patterns as well as increasing percentages is associated with poor prognostic clinicopathological variables and incidence of recurrence. Figure 1

Background:
Clinical significance of 2015 WHO classification histological subtype of early-stage lung adenocarcinoma (LADC) has been well documented; the incidence and significance of histological subtypes in autologous metastatic tumors is unknown.

Methods:
Histological subtyping was performed on paired primary and metastatic LADC tumor samples from patients who underwent resection of metastases (N=203, 1996-2012). 57 cases with inadequate tumor specimen and 4 cases diagnosed as local recurrence were excluded.

Results:
Location of metastatic sites were – brain 51 (35.9%), lung 48 (33.8%), lymph node 14 (9.9%), pleura 10 (7.0%), and adrenal gland 5 (3.5%). Metastatic tumors demonstrated more frequent solid histological pattern than primary tumors (first predominance: 51% vs. 24%; second predominance 29% vs. 17%, Figure 1). Among all histological subtypes, solid subtype showed the highest concordance between primary and metastatic tumors (Figure 2). In addition, analysis of all available clinicopathological factors showed significantly higher percentage of solid subtype in both primary and metastatic tumors was observed in patients with smoking history (p=0.003 and p=0.004, respectively).

Conclusion:
Analysis of a large cohort of primary and autologous metastatic LADC tumors demonstrated a higher percentage of solid histological pattern metastases, even in cancers with a low solid component in the primary site of disease. Figure 1Figure 2

Abstract not provided

Background:
The goal of this study is to investigate comprehensive comparative pathological analyses of both primary tumor and metastatic lymph node (LN) and correlate with lung cancer-specific death (LC-death) in patients with LN-positive lung adenocarcinoma.

Methods:
PN1/2 lung adenocarcinoma patients who underwent R0 resection without induction therapy (n=402, 2000-2012) were included in the study. In primary tumor, lymphatic/vascular/pleural invasion, necrosis, tumor spread through air spaces (STAS), as well as histologic subtypes according to 2015 WHO classification were evaluated. In metastatic LN, metastatic tumor size, extracapsular invasion, histologic subtypes were evaluated. Recurrence and LC-death were analyzed by Cox model.

Results:
Micropapillary and solid predominant subtypes were more frequent in LN metastases than in primary tumors (Figure). In multivariable analyses, adjuvant chemotherapy, pleural invasion, extracapsular invasion of LN metastasis, micropapillary predominant subtype in LN metastasis were independent factors for recurrence; adjuvant chemotherapy, pleural invasion, tumor STAS, and extracapsular invasion were for LC-death (Table).

Conclusion:
In lung adenocarcinoma lymph node metastases, predominant micropapillary pattern and extracapsular invasion indicate high risk for recurrence and lung cancer-specific death. Figure 1 Figure 2

Background:
The interaction of immune cells and cancer cells shapes the immunosuppressive tumor microenvironment. For successful cancer immunotherapy, comprehensive knowledge of anti-tumor immunity as a dynamic spacio-temporal process is required for each individual patient. To this end, we developed an "immunogram for the cancer-immunity cycle" using next-generation sequencing.

Methods:
Whole-exome sequencing and RNA-Seq were performed in 20 non-small cell lung cancer patients (12 adenocarcinoma, 7 squamous cell carcinoma, and 1 large cell neuroendocrine carcinoma). Mutated neoantigens and cancer-germline antigens expressed in the tumor were assessed for predicted binding to patients’ HLA molecules. The expression of genes related to cancer-immunity was assessed and normalized; immunogram was drawn in a radar chart composed of 8 axes reflecting 7 steps of cancer-immunity cycle.

Results:
Distinctive patterns of immunogram were observed in lung cancer patients: T-cell-rich and T-cell-poor. Patients with T-cell-rich pattern had gene signatures of abundant T cells, Tregs and MDSCs, checkpoint molecules and immune-inhibitory molecules in the tumor, suggesting the presence of counter regulatory immunosuppressive microenvironment. Unleashing of counter regulations, i.e. checkpoint inhibitors, may be indicated for these patients (Figure A). Immunogram of T-cell-poor phenotype reflected lack of anti-tumor immunity, inadequate DC activation, and insufficient antigen presentation in the tumor (Figure B). When the immunograms were overlaid within each tumor histology, no typical pattern was elucidated. Both T-cell-rich and T-cell-poor phenotypes were present in each histology, suggesting that histology cannot necessarily reflect the cancer-immunity status of the tumor (Figure C,D). These results were consistent with previous studies showing that clinical responses of checkpoint blockade were not easily predicted by the histology. Figure 1



Conclusion:
Utilizing the immunogram, the landscape of the tumor microenvironment in each patient can be appreciated. Immunogram for the cancer-immunity cycle can be used as an integrated biomarker and thus may become a helpful resource toward optimal personalized immunotherapy.

Background:
Recent evidence indicates that efficacy and durability of responses to immune checkpoint inhibitors in lung carcinomas correlate with increased nonsynonymous mutation (NSM) burden, putative neoantigen number, and in some tumor types, PD-L1 protein expression. In this study, we retrospectively analyzed the relationship of lung carcinoma mutation burden, PD-L1 expression and immune infiltrates with clinical response in patients receiving immune checkpoint blockade.

Methods:
Tumor nonsynonymous mutation data derived from clinical targeted next generation sequencing (309 genes) of lung carcinomas from 94 patients treated with immune checkpoint inhibitors was correlated with clinical outcomes, including durable clinical benefit (DCB; >6 months partial or stable response) and progression-free survival (PFS). PD-L1 immunohistochemistry (clone E1L3N, Cell Signaling Technology, Envision+ detection, Dako) was considered positive if ≥1% of tumor cells and/or tumor-infiltrating immune cells (IC) stained. PU.1, CD3, and FOXP3 immunohistochemistry was used to highlight tumor-associated macrophages and non-regulatory and regulatory T cell populations, which were manually quantified per mm[2].

Results:
The mean patient age was 62 years (range: 32-91 years). Lung tumor types included 69 adenocarcinomas, 11 squamous cell carcinomas, 5 small cell carcinomas, and 9 of other/combined histology. Therapies included PD1 inhibitors (82), a PD-L1 inhibitor (5) and multiple agents (7). Across all tumor types, patients with DCB had a significantly higher number of NSM (range: 1-42) than patients who showed no durable benefit (NDB) [DCB: 12; NDB: 8, p = 0.0027]. Patients with greater than the median number of NSM (9) had significantly longer PFS than those with ≤9 (p = 0.015). Increasing smoking history correlated with higher mutation load (p = 0.047) and patients with a longer smoking history tended to have longer PFS although this trend did not reach statistical significance (p = 0.07). Expression of PD-L1 in either tumor cells or ICs was not associated with NSM burden (p = 0.47) or PFS (p = 0.92). PD-L1 expression in the tumor microenvironment was associated with increased numbers of tumor-associated macrophages (p = 0.0002), and non-regulatory and regulatory T cells (p = 0.0038 and 0.01 respectively).

Conclusion:
The non-synonymous mutation burden in lung carcinoma as assessed by targeted next generation sequencing is associated with increased PFS and durable clinical benefit to immune checkpoint inhibitors. In this limited cohort, PD-L1 expression using clone E1L3N does not predict response to these therapies. We add to growing evidence that increased somatic mutations in carcinomas influence response to immune checkpoint blockade.

Background:
Immune checkpoint, PD-1, inhibitors, have been approved to treat advanced NSCLC patients without oncogenic driver in the second-line setting based on durable clinical benefit. It has been demonstrated that the overall mutational burden in tumor tissue was significantly associated with progression free survival (PFS) of advanced NSCLC patients treated with PD-1 inhibitor. However, tumor tissue may not be available from all patients at any time during PD-1 blockade therapy. Therefore, the purpose of this study was to explore the predictive value of mutation/neoantigen burden from ctDNA on efficacy of PD-1 inhibitors.

Methods:
We treated advanced NSCLC patients without oncogenic drivers with PD-1 inhibitor in the second or more line setting. The whole-exome of tumor tissues and ctDNA at baseline and ctDNA at every time of efficacy evaluation from these patients were sequenced by NGS. The hybrid-capture-enriched libraries were sequenced on the Illumina HiSeq 4000 platform with 75-base paired-end reads, sequencing depth was 300 for ctDNA whole-exome sequencing. We compared the results of whole-exome sequencing between patients who achieved objective response to PD-1 inhibitor and patients who experienced disease progression. Besides, we also compared the results of whole-exome sequencing between baseline ctDNA and ctDNA extracted at efficacy evaluation.

Results:
Up to now, a total of 23 patients treated with PD-1 inhibitor received efficacy evaluation at least once in this study. Of them, 4 patients achieved partial response (PR), 3 patients achieved stable disease (SD). Of 4 patients with PR, 3 patients were found to harbor high mutation burden (more than 400 nonsynonymous mutations) from ctDNA and only 1 patient harbored mutation burden of less than 100 from ctDNA at baseline. We found the mutation or neoantigen burden from ctDNA changed during PD-1 blockade therapy. The efficacy of PD-1 inhibitor appeared to be more significantly associated with neoantigen burden rather than mutation burden. Only one ctDNA sample was found positive for MSH6 mutation (C1337X) and all baseline ctDNA samples were negative for microsatellite instability (MSI) status.

Conclusion:
Evaluating nonsynonymous mutation burden/neoantigen burden from ctDNA was feasible in advanced NSCLC patients treated with PD-1 inhibitors. The predictive value of neoantigen burden from ctDNA on the efficacy of PD-1 inhibitor may be better than that of mutation burden in advanced NSCLC. It may not be feasible to determine the status of mismatch-repair deficiency and MSI using ctDNA samples in advanced NSCLC. An expanded study is ongoing. More details will be presented in the conference.

Abstract not provided

Background:
Anti-PD1 (programmed cell death 1) therapeutic antibodies have recently become available as a promising option in the treatment of patients with NSCLC in Austria. Several clinical studies suggested PD-L1 (programmed cell death ligand 1) protein expression in tumor cells to be a useful prognostic biomarker using several antibodies and different cutoffs. We studied PD-L1 expression in our NSCLC patient cohort and compared the performance of different antibodies. Furthermore we aimed to investigate the value of PD-L1 expression as a biomarker in a subset of patients treated with Anti-PD1 immunotherapy.

Methods:
PD-L1-Imunohistochemistry (IHC) was performed in 437 lung cancer specimens (316 adenocarcinomas, 77 squamous cell carcinomas and 44 NSCLC NOS) using the clones SP263 (Ventana), 28.8 (Abcam) and EL1L3N (Cell Signaling) on the VENTANA IHC platform. The percentages of tumor cells (TC) with membranous staining were determined - irrelevant of staining intensity; TC-counts of less than 1 % were interpreted as negative. Staining with at least two of the antibodies was available in 378 specimens (SP263/28.8 in 320 and 28.8/E1L3N in 117). 60 specimens were stained with three antibodies. From 58 patients receiving Nivolumab clinical information about response to therapy was available.

Results:
PD-L1 was expressed in 244 specimens (54.84%). 112 (25.63%) showed TC counts ≥50%, and 132 (30.21%) were <50%. 193 (44.16%) were negative. SP263 showed stronger staining intensity than 28.8 and E1L3N. Differences in TC-percentage were seen in 67 of 378 specimens, with major changes in 16 specimens (negative to positive in 4 and <50% to ≥50% in 12 cases). Higher TC percentages were seen with SP263. In the 58 treated patients complete remission was seen in 6 (4 ≥50%, 2 negative), partial remission in 14 (10 ≥50%, 3 <50%, 1 negative), stable disease in 4 (2 <50%, 2 negative), paradox reaction in 7 (1 ≥50%, 3 <50%, 3 negative) and progressive disease in 27 (4 ≥50%, 14 <50%, 9 negative).

Conclusion:
PD-L1 is expressed in the majority of NSCLC patients. Despite minor differences in expression levels all three tests provided reliable results. Furthermore PD-L1-IHC showed to be a useful biomarker in NSCLC especially concerning the good response to Anti-PD1 therapy in tumors with PD-L1 expression ≥50%. However as some PD-L1 negative tumors also responded, negative test results cannot definitely exclude patients from immunotherapy.

Background:
Diverse factors have been associated with clinical benefit to PD-1 axis blockers in NSCLC including PD-L1 protein expression by immunohistochemistry and increased mutation load/predicted class-I neoantigens. However, the association and predictive value of the tumor genomic landscape, composition of the tumor immune microenvironment and T-cell function remain unclear.

Methods:
We performed whole exome DNA sequencing and multiplexed quantitative immunofluorescence (QIF) for T-cells in pre-treatment FFPE samples from 45 NSCLC patients treated with PD-1 axis blockers (alone or in combination) in our institution. Genomic analysis was used to evaluate the mutational load and predicted class-I neoantigens. Multiplexed QIF-based immunoprofiling was used to measure the level of CD3+ tumor infiltrating lymphocytes (TILs), in situ T-cell proliferation (Ki-67 in CD3+ cells) and T-cell activation (Granzyme-B in CD3+ cells). We studied the association between the tumor somatic mutations, predicted neoantigens, T-cell infiltration/function and clinical benefit /survival.

Results:
Increased mutational load was positively associated with predicted class-I neoantigens, variants in DNA-repair genes, smoking and absence of activating mutations in EGFR; but not associated with the level of CD3+ T-cells, T-cell proliferation (Ki-67 in CD3+ cells) and function (Granzyme-B in CD3+ cells). Increased mutations and candidate class-I neoantigens were significantly associated with response to therapy (P=0.02 and 0.03, respectively), but not with overall survival at 3-years (median cut-point, log rank P=0.92 and 0.80, respectively). Higher CD3 positivity was not associated with response to therapy (P=0.17), but was significantly associated with overall survival (median cut-point, log rank P=0.03). Regardless of the mutational load and candidate neoantigen content, elevated CD3 with low Ki-67/Granzyme-B in CD3 predicted longer survival after PD-1 axis blockade than high CD3/high Ki-67/Granzyme-B in CD3, or low T-lymphocyte infiltration.

Conclusion:
Increased somatic mutations are associated with smoking and response to PD-1 agents, but not with tumor T-cell infiltration/activation and overall survival. Regardless of the mutational load, increased T-cell infiltration using QIF is significantly associated with longer survival after PD-1 axis blockade in NSCLC. The subgroup of NSCLC with the highest potential of benefit to immune reinvigoration using PD-1 axis blockade comprise tumors with elevated lymphocyte infiltration but low in situ activation/proliferation.

Background:
Non-small cell lung cancer (NSCLC) make up the majority of all lung cancer cases and is associated with very poor prognosis. Immune checkpoint blockers have now been shown to induce unprecedented durable response in a fraction of NSCLC patients with pre-existing T cell infiltration within their tumor. However in order to improve their efficacies beyond this subset of patients, a detailed molecular characterization to identify factors associated with lack of T cell infiltration is needed. A recent analysis in metastatic melanoma identified Wnt/B-catenin pathway activation as a mechanism for lack of T cell infiltration. We pursued similar analyses of immunologic gene signatures and molecular associations in squamous cell lung cancer (SCC) and lung adenocarcinoma (LA).

Methods:
We analyzed RNAseq data from two lung cancer datasets of The Cancer Genome Atlas (TCGA) (N = 499 for SCC and N = 514 for LA). Samples were categorized into non-T cell inflamed and T cell-inflamed groups using unsupervised consensus clustering based on the expression of 160 immune-related genes. Ingenuity pathway analysis was utilized to identify molecular pathways activated in non-T cell-inflamed tumors.

Results:
A similar proportion of non-T cell-inflamed tumors were identified in the two cohorts (SCC: 34%; LA: 31%). 47% of the SCC tumors were identified as T cell-inflamed, as compared to 37% in LA. A positive correlation was observed between CD8A and PD-L1, IDO1, LAG3 and TIM3 (p<0.00001). Total of 1,216 genes are significantly up-regulated in non-T cell-inflamed SCC tumors and 596 in LA with at least 1.5-fold change and FDR-adjusted p<0.05. Among these, a total of 194 genes are up-regulated in both SCC and LA, with the rest being specific for each subtype (SCC: 84%; LA: 67%). Pathway analysis suggested 35 upstream regulators were activated in SCC and 32 in LA (activation z-score≥2.0). Among these, 10 upstream regulators are activated in both datasets (ATF4, CTNNB1, KAT6A, KLF4, MYC, NFE2L2, PI3K, SCAP, SP1, SREBF2). Finally, we performed the same gene expression analysis on RNAseq data from matched normal tissues (N = 51 for SCC and N = 59 for LA) and confirmed that the T-cell inflamed gene signature is a property of the tumor rather than normal lung tissue.

Conclusion:
Our analyses successfully identified genes and associated pathways that are enriched in NSCLC subtypes with no immune infiltration. Rational strategies to improve the efficacy of immune checkpoint blockers beyond the current subset of responders should be based on targeting these pathways.

Abstract not provided

Background:
Nivolumab was approved in Japan on December 17, 2015 for previously treated non-small cell lung cancer (NSCLC). The expression of programmed death-ligand 1 (PD-L1) in tumor tissue is considered a predictive factor for clinical response to nivolumab. However, in Japan, there are no commercially available diagnostic kits for evaluating PD-L1 expression. In addition, little is known regarding other predictive factors of response to nivolumab monotherapy in patients with NSCLC. Therefore, we examined the relationships between the response to nivolumab monotherapy and clinical parameters in patients with NSCLC.

Methods:
Between December 2015 and April 2016, we performed a retrospective analysis of 50 patients with NSCLC treated with nivolumab monotherapy (3 mg/kg, every 2 weeks) at our Institution in the clinical setting.

Results:
Baseline characteristics of patients who received nivolumab monotherapy were: median age, 65 years [range:39–76]; 60% male; 26% ECOG-PS 0, 64% ECOG-PS 1; 38% smoker; 58% stage Ⅳ disease, 22% postoperative recurrence; 80% non-squamous (SQ) NSCLC; 36% non-SQ NSCLC patients had active EGFR mutations; 20% second-line, 18% third-line. The objective response rate (ORR) for all patients treated with nivolumab monotherapy was 18% (95%CI 10–31). Univariate analysis revealed that predictive factors of response to nivolumab monotherapy were associated with “SQ”, “response to the treatment immediately before nivolumab monotherapy”, “therapeutic line of nivolumab (second-line and third-line treatment)” and “smoker” categories (Table 1). In the multivariate logistic regression analysis, the independent predictive factors were “SQ” (P = 0.0069) and “response to the treatment immediately before nivolumab monotherapy” (P < 0.0001) (Table 1). Figure 1



Conclusion:
“Response to the treatment immediately before nivolumab monotherapy”, other than “SQ” histology may be predictors of clinical response to nivolumab in patients with NSCLC.

Background:
Although clinical studies have shown promise for targeting programmed cell death protein-1 (PD-1) and ligand (PD-L1) signaling in non-small cell lung cancer (NSCLC), the factors that predict which subtype patients will be responsive to checkpoint blockade remains elusive. This study was sought to identify the potential biomarkers that predicted response to PD-1 blockade immunotherapy in lung adenocarcinoma.

Methods:
We performed an integrated analysis on the multiple-dimensional data types including genomic, transcriptomic, proteomic and clinical data from cohorts of both lung adenocarcinoma public database including The Cancer Genome Atlas (TCGA), GEO repository (GSE72094) and Broad dataset, and clinical immunotherapeutic patients in our center. Gene Set Enrichment Analysis (GSEA) was used to determine potentially relevant gene expression signatures between specific subgroups.

Results:
We observed distinct function of TP53 and KRAS mutation in regulating immune tumor microenvironment (TME). It is TP53 mutation but not KRAS mutation in lung adenocarcinoma that significantly increased expression of immune checkpoints, facilitated CD8+T cell infiltration and activated T-effector and interferon-γ (IFN-γ) signature. Interestingly, TP53 and KRAS co-mutated subgroup manifested exclusive increased expression of PD-L1 and a highest proportion of PD-L1+/CD8A+. More importantly, TP53 or KRAS mutated tumors showed prominently increased mutation burden and specifically enriched in the transversion-high (TH) cohort. Further analysis focused on the potential molecular mechanism revealed that TP53 or KRAS mutation altered a group of genes involved in cell cycle regulating, DNA replication and damage repair. Finally, clinical immunotherapeutic data were further confirmed that TP53 or KRAS mutation lung adenocarcinoma patients, especially those with co-occurring TP53/KRAS mutations, showed remarkable clinical benefit to PD-1 blockade immunotherapy. Figure 1



Conclusion:
This work provides evidence that TP53 and KRAS mutation in lung adenocarcinoma may be served as a pair of potential predictive factors in guiding PD-1 blockade immunotherapy.

Background:
Immunotherapies that target PD-1/PD-L1 exploit the primary roles of cytotoxic agents in lung cancers. However, tyrosine kinase inhibitors (TKIs) are still considered to be the first choice in lung cancer patients with EGFR mutations. Although immunotherapies may be applied as second line or later therapeutic approaches in these patients, after acquisition of resistance to EGFR-TKIs, it is unclear if acquired resistance mechanisms alter PD-L1 expression status that is employed as an important predictive biomarker for PD-1/PD-L1 targeting agents.

Methods:
Lung cancer cell lines with EGFR mutations (HCC827, HCC4006, PC9, and H1975) and their isogenic descendants with acquired resistance to various EGFR-TKIs were examined in this study. The resistance mechanisms of descendants include T790M secondary mutation, MET gene amplification, epithelial to mesenchymal transition (EMT), and loss of amplified EGFR mutant allele. PD-L1 expression status was analyzed by immunohistochemistry (IHC) and immunoblotting. Effects of acquired resistance mechanisms on PD-L1 expression were also evaluated by shRNA mediated knockdown of candidate molecules, and co-localization analysis using fluorescent imaging. IFN-gamma was used to mimic immune cell attack. Published microarray data of cells with acquired resistance to EGFR-TKIs were also employed to evaluate our findings.

Results:
PD-L1 expression was upregulated in several resistant cells and correlated with EGFR activation. In addition, we found that the phosphorylation of EGFR tyrosine (Y) 992 site, but not Y845, Y1068, or Y1173, was correlated with increased expression of PD-L1. We also observed that TKI-resistant cells with marked E-cadherin downregulation (HCC4006 erlotinib resistant cells and H1975 osimertinib resistant cells), one of hallmarks of EMT, showed decreased expression of PD-L1. However, one cell line (853#10), displaying EMT-like phenotype but only slight E-cadherin downregulation, showed PD-L1 upregulation. Published microarray data from three TKI-resistant lines with EMT-like features also support the correlation of low E-cadherin and reduced PD-L1 expression. ShRNA mediated knockdown of E-cadherin decreased the expression of PD-L1 in parental cell lines. IFN-gamma treatment upregulated PD-L1 expression in both parental and in resistant cells with E-cadherin downregulation, however PD-L1 expression in resistant cells was still lower and localized mainly in the cytoplasm rather than the cell membrane.

Conclusion:
We observed a dramatic change of PD-L1 expression status in lung cancers with EGFR mutation after acquisition of resistance to EGFR-TKIs, depending on the resistance mechanisms. These results support the importance of re-biopsy after acquisition of resistance to EGFR-TKIs, not only for the resistance mechanisms but also for the evaluation of PD-L1 expression status.

Abstract not provided

Abstract:
Tobacco use is the single most preventable cause of death and disease, especially of lung cancer. Approximately 1.6 million people in the WHO European Region die of tobacco-related diseases every year and the Region has the highest proportion of deaths (16%) attributable to tobacco use. Globally, Europe also has the highest prevalence of tobacco smoking among adults (28%), including one of the highest smoking prevalence rates among women (19%). In the meantime it is wellknown that a wide range of political decisions can stop the tobacco epidemic and reduce substantially smoking. According to the first health treaty in history, the WHO Framework Convention on Tobacco Control, following strategies are the most effective: raising cigarette prices through higher cigarette taxe and combating illicit trade of cigarettes, protecting from secondhand smoke through comprehensive smoke-free air laws, enforcing bans on advertising, promotion and sponsorship, placing warnings (pictures and text as big as possible) on tobacco packages and communicate the warnings through media/educational programmes and finally offering greater access to smoking cessation services. WHO Europe has established a database on countries of the WHO European region showing the effects of the reduction in smoking prevalence as a result of implementing tobacco control policies. This presentation will analyse the current situation in this Region and showing the lessons learnt from past years with regards to future prevention of smoking.

Abstract:
Up to now strategies of tobacco control, which were successful in Australia, North America and Western Europe, have been introduced only in few Central European countries. Implementing the EU Tobacco Product Directive Austria extended existing smoking bans, advertising bans and mailing bans for tobacco products to e-cigarettes. Still missing is an enforced smoking ban without exceptions in the hospitality industry, an advertising ban and display ban at point of sale, ban of vending machines, increase of age limit for buying any cigarettes and tobacco products to 18 years, enforcement of age control by regular test purchases, ban of free cigarettes (still allowed for introduction of new sorts), extension of warnings and ban of aromas to cigars and pipes, smoke-free hospitals and health care centers, smoke-free school premises without exceptions, smoke-free playgrounds and cars carrying children, smoke-free public transportation including stations, increase of tobacco tax earmarked for tobacco prevention (up to now only the quitline has regular funding), stop of state funding for media violating article 13 FCTC, enforcement of article 5.3 FCTC (transparency law), obligatory TV air time (e.g. 90 min/month like in Turkey) for promotion of non-smoking, inclusion of smoking prevention and cessation in the curricula of health professionals, covering of counseling for smokers by health insurance, more frequent surveys on smoking prevalence, including cotinine tests for risk groups (pregnant women) and opinion leaders (journalists, health care workers), scientific evaluations of efficacy and effectiveness of smoking prevention and smoking cessation programs. Other Central European countries are facing similar problems. 13 of 16 federal states of Germany, the Czech Republic, Slovakia and most cantons of Switzerland did not succeed to pass and enforce smoking bans without exceptions in the hospitality industry. Germany is still violating the EU tobacco advertising ban and the Czech Republic recently failed to ban the use of water-pipes and e-cigarettes in enclosed spaces. Austria was the only Central European state, which ratified the FCTC smuggling protocol up to now. Switzerland did not even ratify FCTC and was abused by the tobacco industry as a base to sue and intimidate small countries like Uruguay for their progress in tobacco control. The Swiss government prepared amendments for the tobacco law without considering the EU tobacco directive. Tighter restrictions on tobacco advertising will probably be eliminated from the Swiss law. A ban of sponsoring by tobacco companies was not even considered within Switzerland. Lobbies in the Federal parliament and the government were able to block also comprehensive smoking bans. Fortunately, the country offers other political tools to progress, namely that counties can go for better policies, and Federal “initiatives” can be launched to bring issues to the ballots. In fact, 8 out of 26 Swiss Counties successfully adopted comprehensive smoking bans with no exemption. The Swiss parliament, however, was not even enabled to raise minimum prices of tobacco. If the Track-and-Tracing-System of TPD-II is not joined, Switzerland might become a platform for international tobacco smuggling. In restaurants smaller than 80 square meters employers can still choose to permit smoking. The largest progress of tobacco control in Switzerland was due to a tobacco prevention fund, financed from tobacco taxes (2.6 Rappen per pack of cigarette sold) since 2004. This fund finances also tobacco monitoring and smoking cessation (training of professionals). Swiss health insurance covers both counseling and pharmaceutical support of smoking cessation. The largest progress of tobacco control in Central Europe was seen in Hungary, which received the WNTD reward by WHO in 2013 and improved its ranking by the European Cancer Leagues from place 27 in 2010 to place 11 in 2013. Only in Hungary there is a total ban on smoking in all enclosed public places (except prisons and psychiatric wards) since 2012, and tobacco shops (reduced from 40 000 to 6 045) must not be entered below age 18. There is no tobacco advertising outside these shops, no vending machine for cigarettes and plain packaging was introduced in 2016. The main obstacle against improvement of tobacco control in the EU is corruption of certain politicians and media by the tobacco industry and the subsidiarity principle in public health. The Green Paper of the EU “Towards a Europe free from tobacco smoke” showed policy options. 24 ministers of health voted on November 30, 2009 in favor of implementing the WHO- FCTC until 2012, but not the representatives of Austria, Czech Republic and Slovak Republic. Up to now the implementation of FCTC and the EU Council Recommendation on Smoke-free Environments is voluntary. All countries of Central Europe except Switzerland ratified the FCTC treaty, but nonsmoker’s protection follows article 8 only in Hungary. In Central and Eastern Europe tobacco taxes and cigarette prices are much lower than in Northern and Western Europe. In a European ranking according to tobacco price increase by taxes, smoking restrictions at work and in public places, consumer information, tobacco advertising bans, health warnings and access to smoking cessation therapy, Austria, Germany, Cyprus, Czech Republic, Greece and Lithuania had the poorest score and would need help by more advanced EU members to reach Western standard. In summary, tobacco control in Central Europe needs enforcement of FCTC, in particular article 5.3, to stop interference of tobacco industry; application of strategies formulated by WHO (Monitor tobacco use & prevention, Protect from passive smoking, Offer help to quit, Warn about dangers, Enforce bans on ads & promotion, Raise tobacco tax) and the World Bank (Curbing the Epidemic: Governments and the Economics of Tobacco Control); financing of tobacco prevention and cessation by tobacco taxes; comprehensive bans on advertising, promotion and sponsorship of tobacco products and e-cigarettes (including ban of vending machines, display ban at point of sales, plain packaging, stop of state funding for media violating article 13 FCTC); enforcement of smoke-free public places (without exceptions for hospitality industry), workplaces, schools, kindergartens, playgrounds, public transportation and private cars carrying minors and promotion of nonsmoking by schools and media campaigns.

Abstract:
Lung cancer is the leading cause of cancer death worldwide and cigarette smoking is a major causative environmental factor. Some unique biologic profiles are associated to tobacco-induced lung cancer, including clinical, pathological and genetic features. Lung cancer in never smokers (up to 20% of cases worldwide) has been suggested to represent a distinct disease, compared to tobacco-induced lung cancer. Cigarette smoke is a mixture of more than 5000 chemical compounds, among which more than 60 are recognized to have a specific carcinogenic potential. Carcinogens and their metabolites (i.e., N-nitrosamines and polycyclic aromatic hydrocarbons, among others) can activate multiple pathways, contributing to pulmonary cell transformation in different ways. Nicotine, originally thought to be responsible for tobacco addiction, only, is also involved in tumor promotion and progression with anti-apoptotic and indirect mitogenic properties (Tonini et al. Future Oncol 2013;9:649-55). Preclinical models were employed to define epigenomic alterations and gene expression profiles in respiratory epithelia exposed to cigarette smoke condensate. In a study, smoke condensate significantly repressed miR-487b, that directly targets several genes, including SUZ12, BMI1, WNT5A, MYC, and KRAS. Such repression correlated with overexpression of the above targets in lung cancer and coincided with DNA methylation within the miR-487b genomic locus, indicating this molecule as a tumor suppressor microRNA silenced by epigenetic mechanisms during tobacco-induced pulmonary carcinogenesis. These findings may potentially pave the way for DNA demethylating agent treatment, in order to re-activate miR-487b in lung cancer therapy (Xi et al. JCI 2013; 123:1241-61). Among other effects of cigarette smoking, a synergy was described with the aryl hydrocarbon receptor (AHR), which is partially responsible for tobacco-induced carcinogenesis through incompletely understood mechanisms. It was reported that smoking induces AHR activating ligands, which in turn induced adrenomedullin both in vitro and in vivo, thus significantly contributing to the carcinogenicity of tobacco-activated AHR. These effects were not reproduced in fibroblasts and mice lacking the aryl hydrocarbon receptor (Portal-Nuñez et al. Cancer Res 2012; 72:5790-800). Genetic factors involved in tobacco-induced lung cancers have been widely investigated to determine the genetic susceptibility to lung cancer, including epigenomic alterations (Fujimoto et al. PlosOne 2010;5:e11847. Liu et al. Oncogene 2010;29:3650-64). In addition, tobacco-induced lung cancer is characterized by a deregulated inflammatory microenvironment (Spitz et al. Cancer Epidemiol Biomark Prev 2012; 21:1213-21). Therefore variants in inflammation pathway associated genes, as well as a number of genetic polymorphisms have been identified as putative candidates predisposing to lung cancer development. The effects of single polymorphisms on lung cancer development risk have been investigated, with inconsistent results. Most currently identified polymorphisms involve genes encoding proteins associated with the metabolic processing of tobacco smoke carcinogens and the repair of mutations induced by those carcinogens. Polymorphisms on chromosomes 5p15.33, 6p21, and 15q24-25.1 were identified, being the former specifically associated to a higher risk for adenocarcinoma (Yokota et al. Adv Cancer Res 2010;109:51-72). Regarding inflammation genes, analyzing a comprehensive panel of over 11,000 inflammation pathway single-nucleotide polymorphisms (SNP), six SNPs were significantly (p < 0.05) associated to a higher risk of lung cancer development, including two SNP variants in former smokers (BCL2L14) and in current smokers (IL2RB) (Spitz et al. Cancer Epidemiol Biomark Prev 2012; 21:1213-21). The above genetic alterations are observed in all histological subtypes of lung cancer with several differences, especially between small cell lung carcinoma (and the other neuroendocrine tumors) and non-small cell lung cancers. Though all lung cancers are generally tobacco related, changes of incidence of different histological types (with an increase of adenocarcinoma in both sexes) are well known, reflecting modifications of smoking habits, cigarette types, filter types and content of tar, among others. Wide sequencing of single cancer histotypes has provided a relatively complete map of most common alterations in each tumor. In adenocarcinoma, a mean exonic somatic mutation rate of 12.0 events/megabase was identified, which included most previously reported genes in adenocarcinoma as significantly mutated, as well as recurrent mutations in U2AF1, RBM10 and ARID1A genes, and structural rearrangements within EGFR and SIK2 kinases (Imielinski et al. Cell 2012; 150, 1107–1120). Regarding squamous cell carcinoma, the Cancer Genome Atlas Network profiled 178 tumors and found complex genomic alterations, with a mean of 360 exonic mutations, 165 genomic rearrangements, and 323 segments of copy number alteration per tumor. Recurrent mutations were found in 11 genes, with TP53 mutations occurring in nearly all specimens and novel alterations affecting a proportion of cases, including HLA-A class I major histocompatibility gene, NFE2L2, KEAP1, phosphatidylinositol-3-OH-kinase pathway genes, CDKN2A, RB1 and specific squamous differentiation genes (Cancer Genome Atlas Res Network. Nature 2012;489:519-525). As far as small cell lung cancer is concerned, high mutation rates (up to 8.6 non-synonymous mutations per million base pairs) were identified. Of these, up to 28% were found to be C:G>A:T transversions, a type of alteration associated to heavy smoking, although the smoking history was not correlated with the type and number of mutations. Other genes exhibiting mutations and inactivating translocations included the histone acetyltransferase genes CREBBP and EP300, genes with functional roles in the centrosome (ASPM, ALMS1 and PDE4DIP), in the RNA-regulating gene XRN1 and the tetraspanin gene PTGFRN. Damaged genes were commonly found, including the known TP53, RB1, but also TP73, CREBBP and COL22A1, as well as FMN2 and NOTCH family genes (mostly inactivation in the latter) (George et al. Nature 2015; 524: 47–53). Whether the identified genetic signatures and peculiar biological features will produce a corresponding reproducible therapeutic “signature” is still not the case, but the way is paved for stratifying patient groups based on their unique pathological and genetic tumor characteristics, among different histotypes and also within individual neoplastic variants. The future challenge will be to define the biological profile of immunocheckpoint molecule expression in tobacco related lung cancers, in order to identify a reliable predictive marker of response to treatments targeting PD1 or PDL1, in relationship with the different mutational burden and immunological status of individual cases.

Abstract:
Non small cell lung cancer(NSCLC) with sensitive epidermal growth factor receptor (EGFR) mutations invariably develop resistance to EGFR tyrosine kinase inhibitors (TKIs). 20%-30% of NSCLC patients haboring sensitive mutations have no good initial clinical response to EGFR-TKIs, which is defined as having intrinsic resistance to EGFR-TKIs; while the rest of patients with activating mutations who are initially responsive to EGFR-TKIs eventually develop acquired resistance after 10–12 months of consistent clinical beneft, followed by disease progression. The drug resistance is a really tough and urgent clinical problem. Part of resistant mechanisms have been reported, including BIM deletion polymorphism, combined with other bypass signal pathway activation, epithelial-mesenchymal transition (EMT) for primary resistance; T790M, cMET amplification, SCLC transformation for acquired resistance. However, partial resistant mechanisms still unknown. In contrast to acquired resistance to EGFR-TKIs, intrinsic resistance is more complicated. Next-generation sequencing (NGS) is a promising tool for analysis of tumor mutations. We aimed to investigate the intrinsic resistant mechanisms to EGFR-TKIs by NGS, further to optimize treatment strategies and improve clinical outcome in EGFR activating mutant patients having intrinsic resistance to EGFR-TKIs. At present, the study is underway, and the results will be presented at the 2016 WCLC.

Background:
About 10% of non-small cell lung cancer (NSCLC) patients with EGFR mutations will develop leptomeningeal metastasis (LM) either at initial diagnosis or during treatment. LM is a devastating complication of NSCLC associated with poor prognosis. The median overall survival is 4.5-11 months, with ~60% death due to LM or LM together with systemic lesions. However, the underlying mechanisms of the metastasis process are still poorly understood.

Methods:
we performed next-generation panel sequencing of primary tumor tissue, cerebrospinal fluid (CSF) and matched normal controls from 11 EGFRm+ NSCLC patients with LM. Among them, 2 patients had LM at initial diagnosis, and 8 patients developed LM during 1[st] generation EGFR-TKI treatment, while such clinical information was missing for 1 patient.

Results:
The status of EGFR active mutations was the same in the primary tumor and CSF of all the patients, except one whose EGFR mutation was undetectable in the primary site probably due to low sequence coverage. In total, there were 8 patients with EGFR L858R, 1 with 19Del, and 2 with L858R & 19Del dual mutation. One patient also had de novo EGFR T790M in the primary site. None of the CSF samples showed EGFR T790M mutation, suggesting that it was not the resistance mechanism for the 8 patients who developed LM during TKI treatment. PIK3CA E545K and H1047L, and PTEN R130Q were identified in primary site and/or CSF of 6 patients. Although with small sample size, this ratio is much higher than what was reported in general EGFR L858R or 19Del positive lung adenocarcinoma patient population (~2% from 4 datasets), implicating that alternations in PI3K pathway may associate with LM risk. Interestingly, in 9 of the 11 patients, only 0.9%-7.8% of variants in CSF samples overlapped with those in primary site, suggesting tumor heterogeneity, divergence and clonal evolution during LM development. Moreover, when we cataloged the recurrent CSF-unique somatic genomic alterations existing in >5 patients, we identified genes involved in DNA repair pathway, cell cycle regulation and epigenetic reprogramming (NPM1, RAD50, MRE11A, POLE, CHEK1, XPC, KMT2B, KMT2C, KMT2D, and ATRX).

Conclusion:
In summary, our study has shed light on the genomic variations of LM and paved the way for potential therapeutic approaches to this unmet medical need.

Background:
Leptomeningeal metastases (LM) occur in almost 5% of non-small cell lung cancer (NSCLC) patients (pts) and are associated with a poor prognosis. To date, no prospective study has identified active chemotherapy for NSCLC pts with LM. In retrospective studies, EGFR-TKI treatment is reported to be effective in the treatment of LM. We conducted a multi-center, single-arm phase II trial to evaluate the efficacy of erlotinib in pts with LM.

Methods:
NSCLC pts with cytologically confirmed LM were eligible and received erlotinib 150mg daily. Overall cytological response rate (ORR; defined “number of pts who achieves complete remission in CSF / number of all pts”), time to LM progression (TTP), overall survival (OS) and pharmacokinetics were analyzed. Under the null hypothesis, the regimen would be rejected if confirmed ORR was 5% or less. This study was closed because of low accrual with only 21 of required 32 pts (66 %) accrued.

Results:
From Dec 2011 to May 2015, 21 pts (17 pts with EGFR mutation) were enrolled. CSFs available for EGFR mutation analysis (N=17) were all EGFR T790M negative. ORR was 30 % (95%CI 12 -54 %). Median TTP was 2.3 months. Median OS was 3.1 months. Significantly longer TTP and OS were observed in EGFR-mutant than in EGFR-wild type (P=0.0054 and P<0.0001, respectively). Seven pts survived longer than 6 months. CSF penetration rate (Mean + SD) was 3.3 + 0.8 %. There was no correlation between CSF concentration and clinical efficacy.

Conclusion:
Erlotinib treatment for LM is active, especially in EGFR-mutant. Our findings suggest that erlotinib could represent a treatment option for EGFR mutated pts. CSF penetration in LM patients is equivalent to those in previous reports. Table1. Summary of ORR, TPP and OS

	ORR (%)	mTTP (M)	mOS (M)
All (N=20)	30	2.3	3.1
EGFR mutant (N=17)	35	2.7	4.0
EGFR wild (N=3)	0	0.7	0.8
P value (mt vs. wt)	-	0.0054	<0.0001

Background:
Central nervous system (CNS) metastasis, such as brain metastasis and leptomeningeal carcinomatosis (LMC), occurs frequently in EGFR mutant lung cancer. EGFR-TKIs are generally effective to CNS metastasis in EGFR mutant lung cancer patients who are naïve to TKI treatment. Nevertheless, progression of CNS lesions are frequently observed during EGFR-TKI treatment. Brain metastases are manageable by concomitant use of EGFR-TKI and local intervention, including whole brain irradiation and stereotactic radiotherapy. There is, however, no established therapy for LMC, which is resistant to first and second generation EGFR-TKIs. Therefore, novel and effective therapies need to be developed for managing LMC in EGFR mutant lung cancer patients who become refractory to these EGFR-TKIs. The purpose of this study is to clarify the mechanism of EGFR-TKI resistance in LMC and establish novel therapeutic strategy.

Methods:
We examined EGFR mutations, including T790M gatekeeper mutation, in 32 re-biopsy specimens from 12 LMC and 20 extracranial lesions (e.c., lung metastasis and malignant pleural effusions) of EGFR mutant lung cancer patients who became refractory to EGFR-TKI treatment. To clarify molecular mechanisms of acquired EGFR-TKI resistance in LMC, we utilized in vivo imaging model of LMC with EGFR mutant lung cancer cell line PC-9/ffluc and induced acquired resistance to gefitinib by continuous oral treatment.

Results:
We found that all 32 re-biopsy specimens had the same baseline EGFR mutations and that T790M was less frequent in LMC specimens than extracranial specimens (8% vs 55%). Compared with subcutaneous tumors, T790M was less frequent in LMC which acquired resistance to gefitnib. We further established PC-9/LMC-GR cells from the gefitinib-resistant LMC model and found that PC-9/LMC-GR cells were intermediately resistant to gefitinib and osimertinib (3[rd] generation EGFR-TKI). While EGFR-T790M was negative, MET copy number gain associated MET activation was involved in the gefitinib resistance in PC-9/LMC-GR cells. Moreover, combined use of EGFR-TKI and crizotinib, having inhibitory activity against MET, dramatically regressed LMC which already acquired resistance to gefitinib or osimertinib.

Conclusion:
These findings suggest that combined use of MET inhibitors may be promising for controlling LMC which acquires resistance to EGFR-TKIs including osimertinib.

Abstract not provided

Background:
Afatinib (AFA) is an effective treatment in advanced non-small-cell lung cancer (NSCLC) patients harboring epidermal growth factor receptor (EGFR) mutation. However, there were few reports about the cerebrospinal fluid (CSF) penetration rate and the efficacy for central nervous system (CNS) metastasis. Therefore, we conducted the study to evaluate the CSF penetration rate and efficacy of AFA in NSCLC patients harboring EGFR mutation with leptomeningeal carcinomatosis (LC).

Methods:
Eligibility criteria included performance status (PS) 0-3, aged 20 years or older, pathologically proven NSCLC, harboring EGFR mutation, with LC, adequate organ function, and written informed consent. Patients received AFA (40mg/body every day). We analyzed the blood and CSF level of AFA before administrating AFA on the eighth day. The primary endpoint was the CSF penetration rate. Secondary endpoints included objective response rate (ORR), progression free survival (PFS), overall survival (OS), and safety profile.

Results:
A total of 11 patients were enrolled. And we could analyze the blood level in10 patients and the CSF level in 8 patients. Median patients-age was 66 years old. All patients were adenocarcinoma. In EGFR mutation status, 5 patients had exon 19 deletion, 3 had L858R and 3 had minor (exon18) mutation. The patients with PS 2 were 3 patients and PS 3 were 4 patients. Almost all patients received AFA after third-line or more line chemotherapy. The blood level was the Median 88.2 ng/ml (range: 30.4-373), the CSF level was Median 1.4 ng/ml (range: 0.39-2.85) and the CSF penetration rate was Median 1.65% (range: 0.1-9.25). The ORR is 27.3%, and two of three petients with exon 18 mutation showed the partial response. Median OS was 3.8 months (95%CI: 1.1-13.1) and median PFS was 2.0 months (95%CI: 0.6-5.8). Hematological toxicity was mild, however we have to take care of severe diarrhea and skin toxicities, especially in patients with poor PS.

Conclusion:
The median CSF penetration rate (1.65%) of AFA were higher than the penetration rate in previous case report. Although the efficacy for EGFR mutation positive patients with LC was moderate, the promising efficacy for the patient with LC harboring EGFR exon 18 mutation was demonstrated. And we have to take care of diarrhea and skin toxicities, especially in the patients with poor PS.

Background:
Patients treated with crizotinib often experience disease progression in the brain. Brigatinib, an investigational next-generation ALK inhibitor, is being evaluated in an ongoing phase 1/2 trial (Ph1/2) and an ongoing pivotal phase 2 trial (ALTA).

Methods:
In Ph1/2, patients with advanced malignancies, including ALK+ NSCLC, received 30–300 mg brigatinib per day. In ALTA, patients with crizotinib-resistant advanced ALK+ NSCLC received 90 mg qd (arm A) or 180 mg qd with a 7-day lead-in at 90 mg (arm B). Efficacy (in both trials) and safety (in ALTA) are reported for ALK+ NSCLC patients with brain metastases at baseline.

Results:
In Ph1/2 and ALTA, 50/79 (63%; IRC-assessed) and 154/222 (69%; investigator-assessed) of ALK+ NSCLC patients, respectively, had baseline brain metastases. In Ph1/2 (n=50), median age was 53 years, 76% received prior chemotherapy, and 8% were crizotinib-naive. In ALTA (n=154), median age was 52 years; 75% received prior chemotherapy. As of November 16, 2015, 25/50 (50%) patients were receiving brigatinib in Ph1/2; as of February 29, 2016, 101/154 (66%) patients were receiving brigatinib in ALTA. For patients with measurable lesions, confirmed iORR was 53% in Ph1/2 and 42%/67% in ALTA A/B (Table). Among patients with only nonmeasurable lesions (Ph1/2, n=31; ALTA A/B, n=54/n=55), 35% had confirmed complete resolution of lesions in Ph1/2; 7%/18% had confirmed complete resolution in ALTA A/B. For all evaluable patients with baseline brain metastases, median intracranial PFS was 15.6 months in Ph1/2 (n=46) and 15.6/12.8 months in ALTA A/B (n=80/n=73). Most common treatment-emergent adverse events in ALTA in patients with baseline brain metastases (n=151 treated): nausea (A/B, 32%/43%), headache (30%/30%), diarrhea (18%/36%), cough (21%/30%), vomiting (25%/26%); grade ≥3 (excluding neoplasm progression): increased blood CPK (1%/11%), hypertension (4%/7%), increased lipase (3%/3%), pneumonia (1%/4%).

Conclusion:
Brigatinib has demonstrated substantial clinical activity in ALK+ NSCLC patients with brain metastases in both Ph1/2 and ALTA.

IRC-Assessed Confirmed Intracranial Response Rates for Patients With Measurable Brain Metastases at Baseline

		Any	No rad/active[a]
Ph1/2[b]		n=15	n=9
	iORR	8(53)	6(67)
	iDCR	13(87)	8(89)
ALTA[c]
Arm A		n=26	n=19
	iORR	11(42)	8(42)
	iDCR	22(85)	16(84)
Arm B		n=18	n=15
	iORR	12(67)	11(73)
	iDCR	15(83)	14(93)
Data are n(%) iDCR=intracranial disease control rate iORR=intracranial objective response rate IRC=independent review committee [a]No prior brain radiotherapy (Ph1/2); active (untreated or treated and progressed) brain lesions (ALTA) [b]NCT01449461; last scan date: October 8, 2015 [c]NCT02094573; last scan date: April 14, 2016

Background:
BRAF mutations occur in around 3% of non-small cell lung cancers (NSCLC) and V600E accounts for 50%. BRAF V600E is an attractive molecular target for cancer tyrosine kinase treatment, but ideal treatment is still not defined.

Methods:
A case of a patient with BRAF-mutated non–small cell lung cancer (NSCLC) detected by NGS Ion torrent technology who presented with LM disease and was treated with the selective BRAF inhibitor vemurafenib is described.

Results:
58 years old, female, non-smoker, who presented in the emergency department with pericardial effusion. Pericardial fluid cytology confirmed adenocarcinoma TTF1 positive. Multi-organ metastatic disease was diagnosed (bone, lung and thyroid) without EGFR mutation or ALK-EML4 translocation. Four cycles of chemotherapy with pemetrexed and carboplatin were done. She started with refractory headache and vomiting, brain CT and MRI showed no evidence of metastasis, a lumbar puncture confirmed malignant cells in the cerebrospinal fluid. A BRAF V600E was detected by NGS, Ion Torrence PGM technology in the initial tumour sample and in plasma circulating free DNA. An off-label treatment with vemurafenib 960 mg q12hr was offered to the patient, with clinical improvement and radiologic lung stability. At month 2 of treatment, the patient developed respiratory insufficiency with lung infiltrates and Influenza A virus was identified in a nasal swab. Vemurafenib was temporary suspended and re-introduced until 720 mg q12h and maintained until disease progression (large volume pleural effusion with positive cytology), at month 6 of vemurafenib tretament. Third line treatment is being planned.

Conclusion:
The authors highlight the importance of using a multiplex screening strategy to detect targetable mutations in advanced lung cancer patients. The application of next generation sequencing to the tumour and plasma cfDNA allowed the detection of a BRAF-V600E mutation. The improvement of neurologic symptoms and disease control achieved with vemurafenib supports vemurafenib´s efficacy. Care should be taken to the possibility of occurring lung toxicity.

Abstract not provided

Background:
Delivering high-quality cancer care at an affordable cost is the main challenge for health care professionals and policy makers. Immunotherapy achieved encouraging results in NSCLC. PD-L1 expression is being studied as a predictive biomarker. The objective of our study is to assess the economical impact of NIVO and PEMBRO with and without the use of PD-L1 as a biomarker in the US.

Methods:
We developed a decision-analytic model to determine the cost-effectiveness of PD-L1 assessment and second-line treatment with NIVO or PEMBRO versus docetaxel. The model used outcomes data from RCTs and costs from the US. We included the costs of adverse events and post-progression therapies. Thereafter, we used American epidemiology data to estimate the impact of the treatment.

Results:
We included three RCTs (two with NIVO and one with PEMBRO). The estimated number of cases eligible was 37,638. Treating all patients with NIVOLUMAB would cost 1.6 billion dollars each year, increasing total oncology drug expenditure in the US by 4%. Treating only patients with PD-L1 > 1% with NIVOLUMAB would cost US$ 850 million each year and would increase total oncology drug expenditure by 2%. However, with such patient selection up to 46% of cases would not be treated and 2,509 fewer life-years would be saved. The cost of each year-of-life saved was improved by PD-L1 selection (from US$ 223,000 to US$ 186,000 thousand). Table 1 summarizes our findings. Results were similar with NIVOLUMAB and PEMBROLIZUMAB.

Scenario	QALY gain	ICER U$	Life-Years Saved	Years of life not saved	Not Treated	%	Total Cost U$	Impact on Total Cancer Drug Expenditure	Cost/LYS U$
					37,638	100
NIVO ALL COMERS	0.148	124K	7,043	0	0	0	1.6 bi	4%	223K
NIVO PD-L1 > 1%	0.201	91K	4,534	2,509	17,389	46	850 mi	2%	186K
PEMBRO PD-L1 > 1%	0.138	116K	5,302	NA	12,685	34	971 mi	2%	183K
NIVO ALL SQ/>1% NSQ	0.216	93K	5,868	1,175	13,303	35	1 bi	3%	178K
PEMBRO PD-L1 > 50%	0.164	97K	2,270	NA	26,912	72	420 mi	1%	184K

Conclusion:
The use of PD-L1 expression as a biomarker for treatment with immunotherapy decreases the overall economic impact and the cost per life-year saved. Nevertheless, the number of life-years saved with this strategy would be significantly smaller than if we choose to treat all patients. Further study and societal discussion is warranted in order to find the optimal strategy for patient selection.

Background:
Annual LDCT screening for individuals 55-74 yrs with >30 pack-year smoking history is supported by evidence from the NLST but has led to questions of implementation. Compared to organized screening (ORG), opportunistic screening (OPP) may utilize broader entry criteria and not include smoking cessation.

Methods:
Health and economic impacts of ORG using NLST entry criteria were modelled using population microsimulation (OncSim – formerly Canadian Risk Management Model v 2.3) and compared to OPP scenarios. We modeled ORG at a participation rate of 30 and 60%, with and without smoking cessation, compared to various plausible OPP scenarios: younger individuals (40-74 yrs); lesser smoking histories (10 or 20 pack-yrs). Outcomes projected to 20 years included incidence, mortality, number of scans, invasive diagnostics for false positives, and screening and treatment costs. A lifetime horizon and 3% discounting were used to estimate the incremental cost-effectiveness ratio (ICER) from a health system perspective. All costs are in 2016 CAD.

Results:
A large number of outputs can be presented. At a participation rate of 30%, average annual incremental incident cases of lung cancer with OPP for 40-74 yr-olds with 10 pack-yr histories are higher by 254 over ORG without cessation, and there would be an average 135 fewer deaths annually. However, the annual number of CT scans would increase by 433,000 on average and diagnostic tests for false positive results would increase by 1540. Average annual costs would increase by $141 M compared with ORG without cessation, resulting in an ICER of $133,000/QALY. OPP with 40-74 yr-olds having 20 and 30 pk-yr histories result in $92,147 and 74,978/QALY respectively. In all cases of OPP compared to ORG with cessation there are net losses of QALY. Notably, ORG with smoking cessation compared to ORG without yields an ICER of $2800/QALY.

Conclusion:
OPP screening results in more incident cases and fewer deaths but more cost from over-diagnosis and false positives. In Canada, an annual screening program with strict adherence to NLST entry criteria could be highly cost-effective. Jurisdictions will have to weigh the benefits and risks of LDCT scanning beyond the currently available evidence.

Background:
Effect of insurance type on lung cancer diagnosis, treatment and survival is still under debate in Asian patients living in United States.

Methods:
A total of 447,167 patients (18 to 113 years), diagnosed with lung cancer between 2004 and 2013 in the Surveillance, Epidemiology, and End Results database were analyzed. Patient demographics and clinical characteristics were compared between Asian and Non-Asian patients. In Asian patients, patient demographics and characteristics were compared among insurance types. Multivariable logistic regression analysis was performed to identify the effect of insurance types on stage at diagnosis and treatment modalities. Multivariable cox’s regression analysis was performed to identify the effect of insurance type on cancer-specific death.

Results:
Asian were significantly more frequently males (56.7% vs. 53.1%), married (62.2% vs. 50.2%), with Medicaid (17.4% vs. 8.7%), living in rural area (93.6% vs. 86.9%), in a low income county (26.3% vs. 13.4%), and stage 4 at time of diagnosis (51.1% vs. 48.0%) than non-Asian patients (all p-value < 0.001). Among 26,884 Asian lung cancer patients, uninsured were significant younger (61.1±10.8 years) than non-Medicaid (69.1±11.9 years) and Medicaid (70.7±11.7 years), p <0.001, more likely single (18.9 % vs. 8.8% vs. 13.0%); living in a high income county (41.8% vs. 30.5% vs. 38.6%); more likely to be stage IV (63.7% vs. 50.0% vs. 51.2%); and not undergo surgery (86.2% vs. 75.4% vs. 82.6%), [all p-value < 0.001). Localized disease was more frequent in non-Medicaid (21.2%) and Medicaid (17.3%) compared to uninsured (9.0), (p < 0.001).At multivariable analyses, insurance type was not associated with cancer-directed surgery and radiotherapy. Insurance was significantly associated with cancer-specific death (uninsured HR 1.37 95%CI 1.07-1.75; non-Medicaid HR 1.17 95% CI 1.07-1.28 vs Medicaid).Figure 1



Conclusion:
Insurance type affects stage at diagnosis and cancer-specific death but not surgical treatment and radiotherapy in Asian lung cancer patients.

Abstract not provided

Background:
Lung cancer has a high incidence in Brazil; approximately thirty-four thousand new cases are diagnosed each year. In Brazil, as in other countries, the majority of patients diagnosed with lung cancer are elderly. There are few studies that evaluate demographic and clinical characteristics, disease staging, treatment modalities and survival in young patients, mostly carried out in developed countries. This study aimed to describe these aspects in patients with non-small cell lung cancer (NSCLC) according to age.

Methods:
Retrospective cohort consisted of patients diagnosed with NSCLC followed in a referral hospital in São Paulo. During the monitoring the survival time was evaluated. Survival functions were calculated using the method of Kaplan-Meier. The survival stratified by age was also obtained, according to distribution of percentages (less than 55; between 55 and 72 years; older than 72 years). Differences between survival curves were determined using the log-rank test.

Results:
From January 2000 to July 2015 790 patients were followed, 165 aged less than 55 years, 423 between 55 and 72 years and 202 older than 72 years. Higher incidence of adenocarcinoma was seen at the groups up to 72 years. 575 (73%) patients with advanced disease (IIIB-IV stages) were observed. The median five-year survival was 12 months [46-4]. The survival of patients in different age groups was not different.

	<55 165	>55<72 423	>72 202	p
Male n(%)	87 (53)	279 (66)	127 (63)	0.012*
Smoke n(%)	136 (82)	363 (86)	165 (82)	0.34*
Male	78 (90)	263 (94)	121 (95)	0.21*
Female	58 (74)	100 (69)	44 (59)	0.10*
Histological type n(%)				0.13*
Adenocarcinoma	92 (56)	216 (51)	91 (45)
Squamous Cell Carcinoma	52 (32)	170 (40)	91 (45)
Staging n(%)				0.057*
IA/IIIA	34 (21)	127 (30)	52 (26)
IIIB/IV	131 (79)	294 (70)	150 (74)
Deaths n (%)	83 (50)	232 (55)	105 (52)	0.56*
Follow-up (months) Median[IIQ]	4.9 [1.3-13.2]	6.5 [2.0-16.3]	4.4 [1.4-12.9]	0.07†

*Chi-square test; † Kruskal-Wallis (Duncan test); ‡oneway ANOVA (Bonferroni test).

Conclusion:
In the age group of younger patients (<55) women predominated, histological type adenocarcinoma was more frequent, and there were more patients with advanced stage at the diagnosis and a higher percentage of smokers in both genders.

Background:
Globally, 600,000 non-smokers die due to tobacco related diseases and health care cost for tobacco related diseases are soaring especially in developing countries.Cigarette smoke contains 69 known carcinogens out of 7000 chemicals causing various health problems in children and infants including ear infections, bronchitis, pneumonia, frequent and severe asthma attacks, sudden infant death syndrome and cancer. Secondhand smoke can cause coronary heart disease, stroke and various kinds of cancer including lung cancer in adult. 2006 U.S. Surgeon General Report clearly confirmed that there is no safe level of exposure to secondhand smoke. Therefore, there is a strong need for comprehensive smoke free law in reducing the burden of tobacco use in the world.

Methods:
"Section not applicable"

Results:
Iceland is a high income country with a long history of tobacco control. The comprehensive tobacco control law was passed in 1984 which included restriction on smoking in service areas of public and private buildings, premises of health care facilities, schools, workplaces, and in public transport whereas Thailand is an upper middle income country and once had a very high smoking rate (70 percent for male and around 5 percent for female) in early 1970. Smoking ban in movie theaters and buses ordinance was issued by Bangkok Metropolitan Administration I n 1976 followed by the comprehensive interventions including smoke-free areas by the royal Thai government in 1991. As a result, the smoking prevalence went down to 32 percent in Thailand but this still demonstrated that one in three adults in 1991. It was found that the role of government and different civil society organizations in implementing smoke free policies was significant in both countries in this review paper. Although both countries issued smoke free policies, there are still challenges on implementation. Iceland try to overcome the challenges by developing comprehensive system including investigation after complaint and allocating budget for enforcement. Thailand also implemented 100% Smoke-Free Hospitals with the compliance rate of 86.4 percent according to a research in 2010. Thailand established the Thai Health Promotion Foundation (ThaiHealth) in 2001 which allowed Thailand to implement comprehensive tobacco control measures in a sustainable way. ThaiHealth used knowledge generation, social mobilization and policy advocacy called tri-power strategy in achieving the success.

Conclusion:
This paper concluded that civil society initiative and continuing efforts on tobacco control which leads to adoption of comprehensive tobacco control law is the heart of the success of tobacco control in both countries.

Background:
Despite guideline recommendations on time intervals in the care of a lung cancer patient, delays are often experienced. The goal of this study was to quantify time intervals and identify delays in the workup to treatment of non-small cell lung cancer (NSCLC) at our institution.

Methods:
A retrospective review of all NSCLC cases in the Tumor Registry at a tertiary military medical center diagnosed and treated between July 2011 and July 2014 was performed. Dates of radiographic identification of a suspicious pulmonary lesion, tissue diagnosis, evaluation by the treating specialist, and initial treatment (whether surgery, radiation, chemotherapy, or best supportive care/palliative care) were recorded. Time intervals were calculated from these dates; if any interval was more than 60 days, reasons for delays were recorded.

Results:
The median time from the identification of a suspicious pulmonary lesion to the treatment of NSCLC was 74 days (range 5-557 days) and the median time from tissue diagnosis to first treatment was 33 days (range 1-252 days) for the 148 patients included in the analysis. Even after excluding outliers, the adjusted median time from the identification of a suspicious pulmonary lesion to the treatment was 71 days. The most common reasons for treatment delay were waiting for consultant evaluations, staging procedures, repeat biopsies, or additional studies (pre-operative risk stratification, molecular testing). Only 1 patient was upstaged from time of tissue diagnosis to first treatment (from IA to IIB following resection).

Table 1: Critical time intervals

Time interval	Median	Range
Days from suspicious imaging study to first treatment	74	5-557
Days from suspicious imaging study to tissue diagnosis	34	1-556
Days from suspicious imaging study to evaluation by treating specialist	50	1-553
Days from tissue diagnosis to first treatment	33	1-252
Days from tissue diagnosis to evaluation by treating specialist Cardiothoracic surgery Radiation-oncology Medical-oncology Palliative care	20 21 21 18 15	1-157 10-56 1-157 1-64 4-48
Days from evaluation by treating specialist to first treatment Surgery Radiation Chemotherapy Best supportive care	16 23 14 16 1	1-261 6-261 1-88 1-28 1-28

Conclusion:
Several time intervals identified in the workup to initial treatment of NSCLC patients at our institution exceeded recommendations from various guidelines. These delays could be addressed with process and quality improvement projects involving a standardized NSCLC clinical care pathway.

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided