Virtual Library

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Background:
Extracellular vesicles are small vesicles released from all cell types which can be used as a form of cell to cell communication. Recently these extracellular vesicles have been shown to play a key role in cancer development, growth, progression and angiogenesis. These extracellular vesicles are loaded with functional mRNAs, miRNAs and proteins which can be transferred from one cell to another. Extracellular vesicles have been known to enter neighboring cells including the surrounding stroma, and even enter biofluids. Our research shows that miRNAs transferred from lung adenocarcinoma cells through extracellular vesicles influence gene expression in endothelial cells and enhance their ability to form new blood vessels.

Methods:
Using 5 lung adenocarcinoma cell lines (H1395, H1437, H2073, H2228 and H2347) we isolated extracellular vesicles using differential ultracentrifugation. RNA was extracted from the extracellular vesicles as well as the cells from which they were derived and profiled for 742 miRNAs using the miRCURY LNA[TM] Universal RT miRNA PCR system (Exiqon) to identify miRNAs that were enriched by at least 4-fold in the extracellular vesicles. Tube formation assays were conducted on a commonly used endothelial cell line HMEC-1.

Results:
We found an enrichment of a select set of miRNAs within lung adenocarcinoma extracellular vesicles. These miRNAs have previously been identified as tumor suppressors: miR-142-3p, miR-143-3p, miR-144-3p, miR-145-5p, miR-150-5p, miR-223-3p, miR-451a, miR-486-5p, miR-605-5p in various cancer types. When extracellular vesicles are isolated from miR-143 and miR-145 over expressing adenocarcinoma lines they contain an increase in their over expressed miRNAs. When these miRNA enriched exosomes were incubated with HMEC-1 cells, we observed an increase in their ability to form new blood vessels and a decrease in the expression of CAMK1D in the endothelial cells. miR-143-3p and miR-145-5p were also found to be enriched in serum samples draining directly from lung adenocarcinoma tumors compared to arterial serum.

Conclusion:
Extracellular vesicles originating from lung adenocarcinoma cells can enter into endothelial cells and increase their ability to form new blood vessels through extracellular vesicle transfer of miR-145/miR-143 suggesting that this form of communication increases angiogenesis within lung adenocarcinoma tumors.

Background:
MicroRNAs are a class of small non-coding RNAs that range in size from 19-25 nucleotides. They have been shown to regulate a number of processes within tumour biology, including metastasis, invasion and angiogenesis. More recently, miRNAs have been linked to chemoresistance in solid tumours, including lung cancer.

Methods:
MicroRNA expression within an isogenic panel of age-matched parent (PT) and cisplatin resistant (CisR) NSCLC cell lines was profiled using the 7[th] generation miRCURY LNA arrays (Exiqon). Significantly altered miRNAs within the CisR sublines were manipulated using antagomirs (Exiqon) and Pre-miRs (Ambion) and functional studies were carried out in the presence and absence of cisplatin. To examine the translational relevance of these miRNAs, their expression was examined in a cohort of chemo-naïve patient-matched normal and lung tumour tissue and serum from NSCLC patients of different histologies. To create a xenograft model of cisplatin resistance 1x10[3 ]cells H460 PT or CisR cells were injected into 5-7week old NOD/SCID mice. Tumour volume was measured over time and harvested once the tumour mass measured 500mm[3] and formalin-fixed and paraffin embedded (FFPE). Expression of the 5-miR signature was analysed within FFPE murine tumours and cisplatin resistance was investigated relative to cisplatin sensitive controls.

Results:
Profiling and subsequent validation revealed a 5-miR signature associated with our model of cisplatin resistance (miR-30a-3p, miR-30b-5p, miR-30c-5p, miR-34a-5p, miR-4286). Inhibition of the miR-30 family and miR-34a-5p reduced clonogenic survival of CisR cells when treated cisplatin. Expression of the miRNA signature was significantly altered in both adenocarcinoma (AD) and squamous cell carcinoma (SCC) relative to matched normal lung tissue and between SCC and AD tissue. miR-4286 was significantly up-regulated in SCC sera compared to normal control and AD sera. Similarly to the cell line expression of the miRNAs, the miR-30 family members and miR-34a-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 altered clonogenic response to cisplatin. The 5-miR signature shows both diagnostic and prognostic biomarker potential across a number of diagnostically relevant biological mediums.

Background:
Fetal and tumour development share striking similarities, such as intense cell proliferation, angiogenesis, increased cell motility, and immune evasion. Molecular regulators, including microRNAs (miRNAs), play important roles in both fetal lung development and in the malignant transformation of adult lung cells. Consequently, investigation of lung tumour biology in the context of lung development may reveal key regulatory mechanisms that tumours hijack from normal development, which potentially play critical roles in the pathology of lung cancer.

Methods:
131 pairs of non-small cell lung cancer (NSCLC) tumour and non-malignant lung tissues and 15 human fetal lung tissue samples were profiled by miRNA-sequencing. Genes controlled by the oncofetal miRNAs identified were first investigated by miRNA-Data-Integration-Portal (mirDIP) prediction, followed by luciferase-reporter assays. Associations between patient survival and mRNA expression of oncofetal miRNA-gene targets were evaluated in independent samples (>1,400 cases) across multiple NSCLC cohorts. Immunohistochemical analysis of oncofetal miRNA targets was performed on 96 lung adenocarcinoma (LUAD) specimens.

Results:
We describe for the first time a comprehensive characterization of miRNA expression in human fetal lung tissue, and identified numerous miRNAs that recapitulate their fetal expression patterns in NSCLC. Nuclear Factor I/B (NFIB), a transcription factor essential for lung development, was identified as being frequently targeted by these oncofetal miRNAs. Overexpression of the oncofetal miRNA miR-92b-3p, significantly reduced NFIB levels in vitro. Concordantly, analysis of NFIB expression in multiple NSCLC cohorts revealed its frequent underexpression in tumours (~40-70%). This is in contrast with its recurrent oncogenic overexpression recently reported in SCLC. Low expression of NFIB was significantly associated with poorer survival in LUAD patients but not in squamous cell carcinoma patients, consistent with the functional role of NFIB in distal lung cell differentiation (i.e., precursor cells of LUAD). Furthermore, an NFIB-related gene signature was identified in LUAD tumours, comprising several well-known lung differentiation markers (e.g., TTF-1, SFTPB, ABCA3). The underexpression of NFIB protein was ultimately validated in LUAD specimens, which also revealed that tumours presenting lower levels of this transcription factor are associated with higher grade, biologically more aggressive LUAD (invasive mucinous, micropapillary and solid subtypes).

Conclusion:
This work has revealed a prominent mechanism for the downregulation of NFIB, a transcription factor essential for lung differentiation, which we found to be associated with aggressive phenotypes of LUAD and consequently, poor patient survival. Restoration of NFIB expression, specifically in LUAD, has the potential to induce lung cell differentiation and thereby reduce tumour aggressiveness.

Abstract not provided

Background:
Non-small cell lung cancer (NSCLC) accounts for 80% of all lung cancers and adenocarcinoma (ADC) represents the most common histological type with a heterogeneous pattern of growth classified as lepidic, acinar, papillary, solid, and micropapillary. For ADC there are restricted available therapeutic options except for patients that could benefit from target therapy. A valuable therapeutic strategy is represented by angiogenesis inhibitors such as bevacizumab that has been approved for the treatment of NSCLC patients. However, there are concerns about its treatment-related toxicity and the identification of new reliable biomarkers to stratify patients who can really benefit from antiangiogenic drugs is urgently needed. Using miRNA target prediction tools, we selected and investigated the expression level of a panel of miRNAs togheter with their mRNA target involved in the angiogenesis pathway.

Methods:
We designed a custom codeset including probes for six genes (VEGF-A, FLT1, KDR, FLT4, PDGFRa and PDGFRb) and sixteen miRNAs. The expression analysis was performed by the nCounter System® (NanoString Technologies) directly on RNA, enriched of small RNA, purified from the formalin­-fixed and paraffin­-embedded tumor tissues of 80 ADC patients. Of these 25 were predominatly lepidic (31.25%), 24 were predominatly solid (30%), 20 were predominatly acinar (16%), 11 were predominatly papillary (13.75%).

Results:
Comparing the expression levels of mRNAs with the different histological ADC subtypes we found a significant higher expression of VEGF-A in papillary than in other subtypes (p=0.02). In contrast PDGFRa and PDGFRb were upregulated in lepidic and downregulated in papillary subtypes (both p=0.03). Among 16 miRNAs that target the angiogenic mRNA, 6 were significantly downregulated in papillary compared to other groups.

Conclusion:
Our data suggest a distinct angiogenic miRNA-mRNA expression profile among the subtypes of ADC. The higher level of VEGF-A in papillary than in lepidic subtypes could represent a useful biomarker to stratify patients who can effectively treated with bevacizumab, which is directed against VEGF. Moreover, the regulation of angiogenic mRNA factors by miRNAs could provide a novel therapeutic approach based on their expression pattern specific for distinct ADC subtypes. Further studies are nedeed in a larger cohort of patients to confirm our results and to investigate whether different rates of response to treatment are observed among patients stratified according to the proposed biomarkers.

Abstract not provided

Background:
Non-small-cell lung cancer (NSCLC) is one of the most common and high mortality rate carcinoma in China which biomarkers for diagnosis are limited. Therefore, novel biomarkers and methods with increased specificity for diagnosis are explored and required. For now, liquid biopsy for lung cancer oncogenes and next generation sequencing technique are extensive employed in NSCLC. However, increasing evidence illustrates that exosomal microRNAs in circulating fluids provide a promising way as biomarkers for noninvasive cancer diagnosis. Exosomes are 30–150 nm particles which are released from cells into the extracellular environment and stable miRNAs have been identified in plasma exosomes, which play important role in cell communication. Furthermore, exosomal miRNAs present different profiles between patients with cancer and healthy individuals. Whether exosomal miRNAs could benefit NSCLC patient diagnosis remains to be explored.

Methods:
Blood samples were collected from 40 NSCLC patients and 24 healthy volunteers matched with age, gender and blood collection time. Plasma exosomes were accessed by 110,000×g ultracentrifugation and visualized by NS300 equipment. The raw data of exosomal miRNA profiles of NSCLC patients and healthy individuals were generated by NGS around 400× read depth and its expression were measured by Taqman probe quantitive PCR

Results:
In the present study, we revealed that nearly half of exosome RNA was miRNA and NSCLS patients expressed a set of exosomal miRNAs with specificity compared with healthy volunteers. We demonstrated that miR-126-5p and miR-21-3p were down-regulated in NSCLC patients. In addition, we showed that the expression level of miR-124-3p and miR-99a-3p in NSCLC patients was higher than that of healthy individuals. Furthermore, we found miR-99a-3p was clinical stages related in NSCLC patient plasma and miR-375-3p was a potential biomarker for diagnosis and prognosis in NSCLC.

Conclusion:
Exosomal miRNAs in plasma could indicate the progress of NSCLC and a combination of the explored miRNA could serve as a promising biomarker for NSCLS diagnosis and prognosis.

Background:
Lung cancer remains the cause of the most cancer-related deaths each year, with a 5 year survival rate of less than 17%. Targeted therapeutics have been developed against drivers of the lung adenocarcinoma (AC) subtype, but are relevant only to the proportion of patients harbouring these genetic aberrations, emphasizing the need to explore alternative mechanisms of AC development. Natural antisense transcripts (NATs) are long non-coding RNA (lncRNA) products expressed from the opposite strand of coding mRNAs. NATs can function in cis or trans to regulate the transcriptional activity of their cognate gene partner in either a positive or negative fashion. Here we take a novel approach to identify cis- NATs deregulated in lung AC, and explore the function of these genes with regards to their protein coding partner genes.

Methods:
We performed RNA-sequencing on a set of 36 lung AC and matched non-malignant lung tissues. A sign-rank test was used to identify NATs and partner genes with significantly altered expression between tumor and matched normal tissues. These findings were validated in an external dataset of 50 lung AC tumors with matched non-malignant tissue obtained from The Cancer Genome Atlas (TCGA). Survival analysis was performed using a Cox Proportional hazard model, as well as the log-rank method.

Results:
Analysis of Illumina Hi-seq data from TCGA revealed the majority (79%) of deregulated sense-antisense partnerships observed in AC displayed concordant regulation. However, several discordant cis-NAT pairs were identified including an antisense to OPA INTERACTING PROTEIN 5 (OIP5), a gene required for chromatin segregation, as well as an antisense to HIGH MOBILITY GROUP A1 (HMGA1) a gene involved in the metastatic progression of many cancer types. Both the antisense to OIP5 (OIP5-AS1) as well as the antisense to HMGA1, (HMGA1-AS1) were significantly underexpressed in AC, while we find the overlapping protein coding partner genes to be significantly overexpressed, suggesting that these genes may negatively regulate their sense counterparts. In addition both OIP5 and HMGA1 are significantly associated with 5-year survival. Patients with higher expression levels of either of these genes had a significantly shorter overall survival time than patients with low expression levels, highlighting the potential clinical importance of these genes.

Conclusion:
This study characterizes the landscape of antisense expression in AC and highlights novel mechanisms of oncogene regulation through natural antisense transcripts. Characterizing these oncogene regulatory mechanisms could uncover therapeutic intervention points and further our understanding of AC biology.

Background:
The advent of next generation sequencing has lead to the discovery of the functional importance of non-coding RNAs (ncRNAs) in a wide variety of cellular processes, and these genes can be exploited by tumours to drive the hallmarks of cancer. Pseudogenes are DNA sequences that are defunct relatives of their functional parent genes but retain high sequence homology. Long non-coding RNAs (lncRNAs) have been shown to regulate protein-coding genes; however, complex folding patterns make lncRNA function difficult to predict. Several lncRNAs expressed from pseudogene loci have been shown to regulate the protein-coding parent genes of these pseudogenes in trans due to sequence complementarity. The biological impact of this mechanism has not been investigated in lung adenocarcinoma (LUAD). We hypothesize that expression changes in lncRNAs expressed from pseudogene loci can affect the expression of corresponding protein-coding parent genes in trans, and that these events provide an alternative mechanism of cancer gene deregulation in LUAD tumourigenesis.

Methods:
We analysed RNA-seq data from 50 LUAD with matched non-malignant tissue obtained from the TCGA for both protein-coding and non-coding gene expression. Significantly differentially expressed lncRNAs located within pseudogene loci were identified by sign-rank test (p<0.001). Mann Whitney U-tests were used to identify lncRNA-parent gene pairs which significantly correlated expression, and survival analysis was performed using a Cox proportional hazard model.

Results:
Our analysis has identified 172 lncRNAs expressed from pseudogene loci that were significantly deregulated in LUAD. Remarkably, many of these lncRNAs were expressed from the loci of pseudogenes related to known cancer genes. One of these lncRNAs, CTD-2583A14.8, was expressed from a pseudogene to ubiquitin-conjugating enzyme E2C (UBE2C), which regulates tumor growth, apoptosis, and angiogenesis through phospho-ERK1/2. We find CTD-2583A14.8 as well as the UBE2C parent gene to be significantly upregulated in LUAD tumours compared to matched normal tissue. Furthermore, tumours with higher levels of CTD-2583A14.8 have significantly higher levels of UBE2C expression than tumours with low levels of CTD-2583A14.8, indicating that CTD-2583A14.8 may positively regulate UBE2C in trans.

Conclusion:
Here we show expression of lncRNAs within pseudogene loci is deregulated in LUAD, and can correlate with the expression of their protein-coding counterparts. Many of these genes associated with this putative lncRNA-pseudogene-protein-coding axis have previously been implicated in cancer. Therefore, this represents an alternative mechanism of cancer gene deregulation, and may represent novel therapeutic intervention points for the treatment of LUAD.

Abstract not provided

Background:
The new 8[th] AJCC/UICC lung cancer staging system was developed and validated using the International Association for the Study of Lung Cancer (IASLC) database, which contains 94,708 lung cancer patients worldwide, but only 5% of patients in this database came from North America. The goal of this study is to validate the prognostic performance of this new staging system, focusing on the upgraded "T" and “M” parameters, in North American lung cancer patients.

Methods:
We analyzed 1,163,465 non-small cell lung cancer (NSCLC) cases collected from 2004 to 2013 in the United States in the National Cancer Database (NCDB). After excluding patients with more than one malignant primary tumor or tumor size larger than 10 cm, 545,776 NSCLC patients were included in the final data analysis. We defined 8[th] T and M parameters according to the primary coding guidelines of the Collaborative Staging Manual and Coding Instructions for the new 8[th] AJCC/UICC lung cancer staging system. Kaplan-Meier survival curves and log-rank tests were used to compare survival difference among different stage groups, and Cox regression models were used for multivariate analysis adjusting for potential confounders.

Results:
We validated that the new staging system can provide better survival prognosis for NSCLC patients in the NCDB cohort than the existing 7[th] staging system. The median survival time for T1a is 58 months (N=15,860), for T1b is 47 months (N=78,379), and for T1c is 25 months (N=79,828) (p<2e-16). The median survival time for T2a is 19 months (N=111,925), for T2b is 12 months (N=54,601), for T3 is 10 months (N=105,234), and for T4 is 7 months (N=99,949) (p<2e-16). And the median survival time for M0 is 25 months (N=411,048), for M1a is 8 months (N=49,352), for M1b is 5months (N=42,224), and for M1c is 3 months (N=15,926 cases) (p<2e-16). Multivariate analysis showed that these staging parameters are significantly associated with survival when adjusting other factors.

Conclusion:
Both upgraded “T” and “M” parameters of the 8[th] AJCC/UICC lung cancer staging systems are significantly associated with NSCLC patient survival outcomes using data from the NCDB, indicating a good validation performance in patients from North America.

Background:
The new classification for lung cancer refines the T-descriptor criteria into more categories. We examined whether this affects the accuracy of clinical staging, and how this affects the final stage of patients.

Methods:
71 patients underwent resection for primary lung cancer from January 2014 to December 2014. T-component was measured based on the maximum tumour size on CT, PET-CT and histology report. The possible effect on staging based on T-component was compared between both TNMs.

Results:
PET-CT more accurately estimates the pathological size of the tumor (mean difference from histology: CT 3mm (range -1.6 to 2.6cm) and PET-CT 1.3mm (-2 to 2.5cm). Discordance between radiological and pathological T-stage was higher with the 8th edition (7th edition concordance CT 42(59%) and PET-CT 31(43%), 8th edition CT 31(44%) and PET-CT 29(41%) (CT p=0.01; PET-CT p=0.7)). The final stage groupings was also more discrepant in the 8th edition. Concordance was for CT 7th Edition 37(54%) vs 8th Edition 21(31%) (p<0.001), and for PET-CT 34(48%) vs 19(28%) (p<0.001). The discrepancy in stage grouping is contributed significantly by T-stage discordance. In the 30 patients who were not upstaged pathologically by pleural invasion or nodal staging, there is a over 50% increase in inaccuracy of clinical staging in the 8th edition. The CT concordance was 7th edition 24(80%), 8th edition 13(43%) (p<0.001); and for PET-CT 23(77%) vs 10(33%) (p<0.001).

Conclusion:
We showed that the 8th edition TNM lung cancer staging system was associated with a significant increase in discordance between clinical and pathological staging due to differences in measurement of tumour size and consequently T-stage groupings by different modalities. This has implications for prognostication and clinical trial interpretation especially in patients who do not undergo surgery for pathological stage confirmation.Figure 1

Background:
The IASLC lung cancer staging project recently published recommendations for the 8th edition of the TNM classification of lung cancer. This recommends the same N descriptors be used, however, further analysis of an alternative system (proposed 9[th] Edition) was recommended. The aim of this retrospective study was to assess the utility of this proposed nodal staging system at a large UK tertiary lung cancer centre.

Methods:
Patients who underwent surgical resection for non-small cell lung cancer between 2011-2014 (allowing minimum of 2 years follow-up) were identified from a prospective database (n=1308). Stratification of pathological N-stage as per the IASCLC proposal was performed: N0, single station N1 (N1a), multi-statin N1 (N1b), single station N2 without N1 involvement – skip metastases - (N2a1), single station N2 with N1 involvement (N2a2), multi-station N2 (N2b) and N3. Survival data was obtained from national death registries.

Results:
There a significant effect of N-stage on mortality using Cox proportional hazards regression analysis, using pN0 (n=848) as the reference group and adjusting for sex, age and histology (table 1). There appears to be similar survival outcomes between multi-station N1 (pN1b) and single station N2 skip metastases (pN2a1), and single station N2 with N1 involvement (pN2a2) and multi-station N2 (pN2b).

Table 1.

N-stage	Hazard Ratio	95% CI	p-value
pN1a (n=146)	1.68	(1.26, 2.26)	0.001
pN1b (n=55)	2.25	(1.49, 3.39)	<0.001
pN2a1 (n=50)	2.24	(1.46, 3.45)	<0.001
pN2a2 (n=81)	2.94	(2.15, 4.03)	<0.001
pN2b (n=67)	2.99	(2.09, 4.27)	<0.001

Conclusion:
The proposed 9[th] edition N-staging classifications appear to add additional insight into prognosis. However, interpretation is limited by the small numbers of patients within the pN1/pN2 sub-groups. 65% of this large cohort were pN0 and acted as the reference group and a further 5% were pNx as no nodes were submitted. Furthermore, the accuracy of pN-staging is reliant on the quality of intra-operative lymph node sampling. Although significant improvements have been made in this timeframe at our centre (published previously), any sub-optimal performance has the potential to affect the validity of the results, particularly if multi-station N2 disease is missed.

Abstract not provided

Background:
In lung adenocarcinomas, the histologic lepidic growth pattern tends to correlate with the ground glass opacity (GGO) component, while solid components correspond with invasive adenocarcinoma. The Eighth edition of the TNM staging system suggests that the tumor size be determined according to the invasive size excluding the lepidic component. However, this new concept causes fatal confusion, i.e., tumors are classified into a same T category despite the part-solid or pure-solid appearances provided they showed a same solid component size.

Methods:
Between 2008 and 2012, we retrospectively evaluated 719 surgically resected cN0 lung adenocarcinomas that measures 30mm or less in total dimension to assess the prognostic impact on the presence of GGO among the Eighth TNM classification. According to the new T category, it was defined based on the solid component size as follow: Tis; 0 cm (pure-GGO), T1mi; ≤ 5 mm, T1a; 6-10 mm, T1b; 11-20 mm, T1c; 21-30mm. Furthermore, all tumors were classified into 2 groups, i.e., GGO or Solid arms based on the presence of GGO component.

Results:
Of the cases, 133 (18%) were categorized in Tis, 88 (12%) in T1mi, 121 (17%) in T1a, 244 (34%) in T1b and 133 (19%) in T1c, respectively. Multivariate analysis revealed that both a presence of GGO and solid component were independently significant prognostic factors (p=0.007, 0.002). The 5y-overall survival (OS) was 99.2% in Tis, 95.8% in T1mi, 96.5% in T1a, 81.8% in T1b and 66.4% in T1c (p=0.038) with a median follow-up period of 56 months. When we evaluated the impact of T category based on GGO presence, the 5y-OS was significantly different between GGO and Solid arm in each T categories (T1a; 99.0% vs. 95.7%, p=0.045, T1b; 89.8% vs. 73.3%, p=0.004, T1c; 90.0% vs. 62.6%, p=0.046). Furthermore, clinical T categories significantly separated the OS in Solid arm (p=0.015) (T1a vs. T1b; p=0.090, T1b vs. T1c; p=0.037). In contrast, the 5y-OS was approximately 90% or more in GGO arm despite their T categories. Moreover, regarding radiological and pathological correlations, the rates of AIS was only 65% in Tis, and 51% showed invasive adenocarcinoma even in T1mi.

Conclusion:
Clinical T category should be considered based on the presence of GGO on thin-section CT, and tumor size should be applied exclusively to radiological solid lung cancer. In contrast, oncological outcomes of the tumor with GGO component were excellent despite their T categories, which should be described as Tis for pure-GGO, and T1a for part-solid tumor.

Background:
Central tumour location is considered as an independent risk factor for occult mediastinal metastases in patients with non­-small cell lung cancer (NSCLC) after negative computerized tomography (CT) and integrated positron emission tomography/CT. However, the distinction between a central and a peripheral tumour has not been codified, some authors consider any tumour in the inner­third of the hemithorax to be central, and others in the inner two­thirds. The objective of this study is to identify the best centrality tumour definition for detecting occult mediastinal metastasis.

Methods:
We retrospectively reviewed our thoracic surgery database for cases between January 2011 and December 2015. It was identified patients with potentially operable NSCLC screened by CT and PET/CT and they were classified according to tumour location in the inner third, middle third or outer third. The prevalence of occult mediastinal lymph node metastases was analysed in relation to tumour location. Statistical analysis for best centrality definition was performed. Univariable analysis was performed using the Fisher exact test and multivarible analysis using logistic regression.

Results:
A total of 359 patients with clinical operable NSCLC were included in our study. Seventy­-five (20.9%) tumours were located in the inner­third, 137 (38.2%) in the middle­third and 147 (40.9%) in the outer­third. It was detected 23 patients with N2 disease and negative TC and PET/CT, 8/38 (21.1%) in the inner­third, 6/121 (5.0%) in the middle­third and 9/122 (7.4%) in the outer­third. Defining centrality as tumour located in the inner­third of the hemithorax the incidence of occult N2 was 21.1% and 6.2% for central and peripheral tumour respectively. And defining centrality as tumour located in the inner two­thirds the incidence of occult N2 was 8.8% for central tumours and 7.4% for peripheral. Univariable analysis shows statistical differences in occult N2 involvement between central or peripheral defining central lesion as innerthird (p=0.002), but not for central definition of innternal twothirds (p=0.651). In multivariable analysis considering centrality possible defenition, histology, Clinical T factor and Clinical N1 affection, only innerthird as centrality definition was statistical significance predictor factor for occult mediastinal lymph node involvement (p= 0.027)

Conclusion:
Considering the results of our study the best definition for central tumour location is limited to the inner­third of the hemithorax. Given the low rate of occult N2 disease in the middle third location, the systematic evaluation of the mediastinum by ecobronchoscopy and/or mediastinoscopy in this group of patients is not justified.

Background:
False-negative 18F-fluorodeoxyglucose (FDG) uptake can be divided into those cases related to technological limitations of positron-emission tomography (PET) and others related to inherent properties of neoplasms. We aim to clarify possible factors causing false-negative (FN) PET results in solid-type pulmonary adenocarcinomas (PAs).

Methods:
From 01/2007 to 12/2014, among 255 Stage-I NSCLCs we retrospectively review PET/CT-records, clinical information, preoperative thin-section CT-images, and pathological features (classified by the IASLC/ATS/ERS subtyping criteria) of 94 consecutive solid-type Stage-I PA undergone surgical resection at Our Institution. Univariate and multivariate logistic analysis were used to identify and weigh the independent predictors of PET-findings: body weight, blood glucose level, tumor-size, and histological classification.

Results:
There were 58 males and 36 females (mean age= 68.7 yrs, range 42-85). Seventeen lesions (18.1%) were judged as PET-negative and 77 lesions (81.9%) as PET-positive. Overall, mean SUVmax was 8.0 (range 0-35) with higher SUVmax-values (p<0.001) in PA>2cm (mean SUVmax=10.6) than PA<2cm (mean SUVmax=4.8). PET false-negative (FN) results were also differently distributed (27.9% in PA <2cm vs 9.8% in PA>2 cm, p=0.023). When clustering the PA in 2 histological classes (Class-A [“colloid/mucinous/lepidic”] vs Class-B [“micropapillary/solid/acinar/papillary”]), the radiometabolic patterns were significantly different [mean SUVmax 3.8 in Class-A vs 9.9 in Class-B, p<0.001], as reported in Figure 1. Similarly, a different distribution of PET FN-cases was observed (38.7% FN in Class-A vs 7.9% FN in Class-B, p=0.001). Table 1 shows the results of multivariate logistic analysis. Both the tumor-size (cut-off=2cm) and IASLC/ATS/ERS aggregated clusters were clinically relevant factors for determining whether PET results were negative or positive, but only histology was statistically significant (OR:6.1, 95%CI: 1.85-20.15, p=0.003). Figure 1



Conclusion:
Among solid-type lung adenocarcinoma, tumor-size and histopathological findings were significantly associated with FDG-uptake. In particular, it warrants attention that lesions ≤2cm and “colloid/mucinous/lepidic” adenocarcinomas have a tendency for negative PET-findings.

Abstract not provided

Background:
Immune check-point inhibitors (ICPIs) exert their activity by blocking inhibitory signaling and therefore enhancing T-cell activity against tumor cells; however, this peculiar mechanism of action might lead to many difficulties in evaluating clinical response with the usual CT imaging due to inflammatory patterns that could confuse the evaluation. The aim of this study was to assess the role of FDG-PET to support clinical decision based on CT scan.

Methods:
From May 2015 to April 2016, 74 patients with advanced pretreated NSCLC received at least one dose of nivolumab (3 mg/Kg every 14 days) within a single-institutional translational research trial. Among these, 58 patients were evaluable for response assessment. The patients underwent CT scan and FDG-PET every four cycles and, in case of progressive disease, an additional evaluation was performed after two further cycles in order to confirm it. We evaluated the response to treatment by CT scan with RECIST criteria, Immuno-related Response Criteria (irRC), WHO criteria and immunoRECIST criteria, while the metabolic response has been determined with PERCIST criteria. Finally, we determined the concordance in terms of response between CT evaluation criteria and metabolic response obtained with PERCIST; concordance was calculated with kappa value.

Results:
Our findings showed a low concordance of all CT scan evaluation criteria to PERCIST, the best concordance being between PERCIST and RECIST (K=0.500) and the worst agreement being between PERCIST and irRC (K=0.295) . In particular, PERCIST seems to underestimate the progressive disease (PD). In fact, between 46% and 55% of patients, defined in progression with CT evaluation criteria were considered in stable metabolic disease (SMD) by PERCIST; among these, 50% of patients in the RECIST PD group and 80% of RECIST SD patients were alive at 6 months. Furthermore, in our sample, between 9% and 18% of patients were considered in progression with CT evaluation criteria when they were in partial response with PERCIST; these patients were still alive with a survival similar to those who defined in partial response with RECIST (>9 months).

Conclusion:
FDG-PET evaluation by PERCIST could not be helpful when SMD is reported, in fact, patients that have a RECIST PD maintain a poor prognosis compared to RECIST SD between the patients define as SMD. Conversely, PERCIST evaluation could be informative when it define a partial response, specially when RECIST criteria show a PD.

Background:
Nivolumab is approved for treatment of squamous and non-squamous advanced NSCLC. Since nivolumab restores antitumor immunity, it is not clear whether 18F-FDG-PET/CT is able to distinguish response from tumor progression. We evaluated early metabolic patterns of response to nivolumab in advanced NSCLC patients.

Methods:
We retrospectively reviewed PET/CT scans and paired CT scans from 22 patients with advanced NSCLC who received nivolumab 3mg/kg every 2 weeks and performed PET/CT before and after 4 infusions of nivulomab. Total Lesion Glycolysis (TLG) and Metabolic Tumor Volume (MTV) of every lesion up to 5 per patient were measured on baseline and follow-up PET/CT. Percentage changes in MTV (ΔMTV) and TLG (ΔTLG) between the two PET/CT were calculated. Patients were classified into responders (nivolumab for >6 months), non responders (nivolumab ≤6 months) or having pseudo-progression (PP, nivolumab and clinical benefit >6 months despite initial progressive disease according to RECIST criteria)

Results:
Among 22 patients, 6 (27%) were responders, 15 (68%) were non-responders and 1 (4.5%) had PP. Baseline MTV and TLG were significantly lower in responders than in non-responders (medians 27 vs. 63 mL, p=0.03 and 124 vs. 254 g, p=0.04, respectively). After 4 infusions of nivolumab, metabolic parameters were significantly lower in responders than in non-responders (median MTV : 2 vs. 148 mL, p=0.001 and median TLG : 6 vs. 835 g, p=0.002). Mean ΔMTV and ΔTLG were both -88% in responders, and +236% and +312% respectively in non-responders, which was significantly different (p=0.0005). The only patient with PP had lower ΔMTV (+11%) and ΔTLG (+41%) than non-responders patients.

Conclusion:
In NSCLC, objective response and disease progression upon nivolumab usually translate into early and clear-cut patterns of change in PET/CT. Early PET/CT may help to distinguish progression from pseudo-progression.

Background:
Immune check-point inhibitors have dramatically changed the management of advanced non-small cell lung cancer (NSCLC); however, their mechanism of action creates concerns on the most appropriate method to determine radiological responses to this drug class. The aim of this study is to compare a set of different evaluation criteria for patients receiving nivolumab for advanced NSCLC.

Methods:
Patients with pre-treated advanced NSCLC were enrolled in a single-institutional translational research study in the San Martino Hospital – National Institute for Cancer Research, Genova, Italy and received nivolumab (3 mg/kg every 14 days). Computed tomography (CT) was performed at baseline and after every 4 administrations. The assessments were performed according to Immune-related response criteria (irRC), response evaluation criteria in solid tumors (RECIST 1.1), World Health Organization (WHO), and immune-related RECIST (irRECIST), which are recently proposed based on the original RECIST with the following differences derived by irRC: 1) new lesions do not automatically define progressive disease (PD), but are added to the target lesions count; 2) PD has to be confirmed with a subsequent CT-scan after 2 additional cycles. The concordance among the different criteria was determined with Cohen’s kappa coefficient (K).

Results:
Fifty-two patients were eligible: median age= 70 years (44-85); male/female: 70%/30%; current or former smokers= 87%; non-squamous/squamous histology= 79%/21%; median number of cycles= 6 (4-29). The following responses were observed:

	Partial Response	Stable Disease	Progressive Disease
	First evaluation (4 cycles)
RECIST 1.1	4 (7.7%)	19 (36.5%)	29 (55.8%)
irRC	3 (5.8%)	23 (44.2%)	26 (50%)
WHO	3 (5.8%)	20 (38.5%)	29 (55.8%)
irRECIST	4 (7.6%)	24 (46.2%)	24 (46.2%)
	Best Response
	Partial Response	Stable Disease	Progressive Disease
RECIST 1.1	9 (17.3%)	14 (26.9%)	29 (55.8%)
irRC	8 (15.4%)	19 (36.5%)	25 (48.1%)
WHO	7 (13.5%)	17 (32.7%)	28 (53.8%)
irRECIST	11 (21.2%)	18 (34.6%)	23 (44.2%)

Generally, the concordance between first evaluation and best response was good for all the criteria (K ranging from 0.783 to 0.839); the concordance between irRECIST and irRC was high (K= 0.828) and RECIST 1.1 had a good concordance with IRC (K= 0.734), irRECIST (K= 0.767), and WHO (0.766).

Conclusion:
The different response assessment methods were generally concordant. Since response is more easily assessed with irRECIST than with irRC, the former might be proposed as an appropriate method of response evaluation.

Abstract not provided

Abstract not provided

Abstract not provided

Background:
Malignant pleural mesothelioma (MPM) is a highly aggressive, asbestos-related malignancy characterized by poor outcome and limited therapeutic options. Fibroblast growth factor (FGF) signals play important roles in mesothelioma cell growth and malignant behavior and their inhibition leads to reduced tumor growth. MicroRNAs (miRNAs) are conserved noncoding RNAs controlling gene expression via translational repression of target mRNAs. The miR-15/16 family is downregulated in MPM and has tumor suppressor functions. Several FGFs/FGFRs are predicted miR-15/16 targets. The aim of this study was to explore the link between the miR-15/16 and the FGF/R family in MPM.

Methods:
Gene and microRNA expression was determined by RT-qPCR or Taqman Low Density Arrays (TLDAs). Mimics were used for restoring microRNA expression. Stimulation or inhibition of FGF signals or bcl-2 was achieved by recombinant FGF2, siRNAs, or small-molecule inhibitors, respectively. A SYBR green-based proliferation assay and colony formation assays were used to monitor effects on cell growth.

Results:
Expression analysis showed a consistent downregulation of target FGF/FGFR genes after transfection with miRNA mimics. Restoration of miR-15/16 led to dose-dependent growth inhibition, which significantly correlated with sensitivity to the specific FGFR1 inhibitor PD166866. Re-expression of microRNAs in combination with FGFR knock-down or pharmacological inhibition resulted in reduced activity, indicating target competition. Combined inhibition of the FGF-axis and bcl-2, another established target of miR-15/16, resulted in enhanced activity. Treatment with recombinant FGF2 further reduced mature as well as pri-microRNA levels and also could prevent/reduce growth inhibition by mimics, but only when added within 24 hours after transfection. TLDA screens after stimulation/inhibition of FGF signals identified regulation of several other miRNAs involved in pathways relevant for tumour growth and aggressiveness.

Conclusion:
Our data shows that the post-transcriptional repression of FGF-mediated signals contributes to the tumour-suppressor function of the microRNA-15/16 family. Impairing hyperactivated FGF signals as well as the anti-apoptotic protein bcl-2 through the restoration of this miRNA family might serve as a novel therapeutic strategy in mesothelioma.

Background:
Malignant mesothelioma is an aggressive cancer that develops from mesothelial cells, most often in the pleural lining of the lung. We have previously shown that the BMP inhibitor protein gremlin-1 is highly expressed in mesothelioma and induces a mesenchymal and chemoresistant phenotype in mesothelioma cells. Since mesothelioma tumors are locally highly invasive, we analyzed the role of gremlin-1 in mesothelioma cell migration and invasive growth.

Methods:
Primary mesothelioma cells isolated from patient pleural fluid as well as mesothelioma cell lines were used for in vitro studies. Cells were transfected with siRNAs or transduced with lentiviral expression vectors. Invasive growth was analyzed in 3D matrigel or collagen I matrices. mRNA expression was analyzed using a commercial PCR array and quantitative RT-PCR. Migration assays were performed using scratch wound assay or Transwell migration assay with fibronectin or collagen coating. TGF-β and BMP signaling activity was measured with reporter-luciferase assays. For in vivo mouse xenograft experiment cells were additionally transduced to express a luciferase marker. Subcutaneous cell injection with matrigel matrix was performed in the flank of nude mice.

Results:
Mesothelioma cells expressing gremlin-1 showed invasive sprouting when tumor cell spheroids were imbedded into 3D collagen matrix. Silencing of gremlin-1 expression significantly reduced invasive growth. In addition, cells overexpressing gremlin-1 gained invasive growth ability. This was associated with increased mRNA expression levels of Slug and matrix metalloproteinases (MMP) as well as reduced expression of E-cadherin. The cells were more migratory and exhibited increased expression of certain integrins, especially the α~v~ subunit. Gremlin-1 induced invasive growth was dependent on MMP activity and associated with increased TGF-β activity. Intrapleural injection of gremlin-1 overexpressing mesothelioma cells isolated from a patient with epithelioid mesothelioma, produced tumors in 2/4 mice over 4 months after injection. Cells transduced with vector only did not produced tumors (0/4). When cells were injected subcutaneously together with matrigel gremlin-1 overexpressing tumors appeared more slowly, but exhibited comparable luciferase signal 2.5 months after injection. However, gremlin-1 tumors showed more local spreading and in contrast to control tumors some also developed metastasis (2/6 mice).

Conclusion:
Mesothelioma invades locally and has poor prognosis. We have identified gremlin-1 as an important regulator of mesothelioma chemoresistance and invasive growth behavior. Blocking gremlin function may overcome drug resistance and reduce invasion of mesothelioma.

Background:
Malignant pleural mesothelioma (MPM) is a rare but devastating malignancy. Despite the search for new promising treatment approaches, the outcome for most MPM patients remains dismal. Therefore, the identification of novel biomarkers is urgently needed in order to identify patients with a better prognosis and to support personalized therapeutic decisions. In our previously published study, we were able to show that fibroblast growth factor 18 (FGF18) is overexpressed in MPM tissue specimens and cell models. The objective of this study was the evaluation of FGF18 as a circulating biomarker in MPM.

Methods:
Plasma was collected from 107 MPM patients at the time of diagnosis or before surgical resection. Samples were included from the Medical University of Vienna, University Hospital Center in Zagreb and from The Concord Repatriation General Hospital and Strathfield Private Hospital in Sydney. Samples from 49 healthy volunteers and from 8 patients with non-malignant pleural diseases served as controls. Circulating FGF18 was measured by enzyme-linked immunosorbent assay and correlated to clinical, pathologic and radiologic parameters.

Results:
Plasma FGF18 level was significantly elevated in MPM patients vs. healthy controls (P<0.0001). A slight increase of circulating FGF18 level was also detected in patients with pleuritis or fibrosis (vs. control, P=0.0067). Sarcomatoid (n=7) morphology was associated with high FGF18 levels when compared to the epithelioid (n=77) histology (P=0.0064). Importantly, MPM patients presenting with FGF18 levels below the median had a significantly longer overall survival when compared to those with high FGF18 levels (median survival 625 versus 382 d, P=0.0038). Data on multivariate analysis, disease-free survival, correlation with other biomarkers and tumor volume will be presented at the conference.

Conclusion:
Our findings reveal that FGF18 is a promising blood-derived candidate biomarker in MPM. Furthermore FGF18 may support the histological classification of MPM and the identification of MPM patients with poor prognosis. .

Abstract not provided

Background:
Malignant pleural mesothelioma (MPM) is an aggressive cancer caused by asbestos exposure with limited therapeutic options. Dysregulated microRNAs play an important role in MPM biology and candidate microRNAs have been investigated as diagnostic and prognostic biomarkers or as a potential treatment targets. The role of miR-223 has previously been investigated in MPM tumour cells and was shown to act as a tumour suppressor by regulating cell mobility. Previous research indicated miR-223 to be primarily expressed by myeloid progenitor derived cells during differentiation of granulocytes and monocytes. This suggests miR-223 might have a more significant role in the inflammatory response during tumourigenesis. In this study we aimed to investigate the origin of miR-223 using mesothelioma xenograft and syngraft models.

Methods:
Human and mouse mesothelioma cell line-derived xenograft (MSTO-H211 and H226) and syngraft (AB1) models were established. MicroRNA profiles of xenografts were compared against profiles of their corresponding in vitro cultured cells to determine candidates. RT-qPCR using TaqMan MicroRNA assays was used to validate expression levels of miR-143-3p, miR-214-3p and miR-223-3p in tumour xenografts and syngrafts with those in corresponding cell lines in vitro. Species-specific ddPCR analysis was performed on RNA from xenograft tumours to determine the expression of human and mouse pri-miR-223.

Results:
MicroRNA profiles of xenograft tumours showed significant upregulation (p < 0.05) of miR-143-3p, miR-214-3p and miR-223-3p compared to corresponding in vitro mesothelioma cell lines. Only miR-223 showed significant upregulation in both xenograft and syngraft tumours compared to corresponding in vitro mesothelioma cell lines (>10000-fold increase). Other microRNAs were not significantly different between cell lines and tumours. RNA isolated from xenograft tumours contained significantly more mouse pri-miR-223 than human pri-miR-223 (p < 0.001), with only minimal expression levels of human tumour pri-miR-223 within xenograft tumours.

Conclusion:
Mature miR-223 is significantly overexpressed in xenograft tumours compared to corresponding in vitro mesothelioma cell lines suggesting stromal contribution. Species-specific pri-miRNA confirmed miR-223 is almost exclusively expressed by the mouse stromal cells in xenograft tumours. Ultimately, localising the expression of miR-223 to specific cell types (such as myeloid derived cells) through in situ hybridisation should help identify a more biologically relevant role for miR-223 in the tumour microenvironment.

Background:
To overcome tumor-mediated inhibition of chimeric antigen receptor (CAR) T cells, we herein investigated the impact of tumor PD-L1 upregulation on CAR T-cell exhaustion and anti-tumor efficacy, and further developed clinically translatable T-cell extrinsic as well as intrinsic strategies to overcome PD-L1 inhibition in models of lung cancer (LC) and malignant pleural mesothelioma (MPM).

Methods:
Human T cells were transduced with MSLN-specific CAR with CD28 and CD3zeta domains (M28z) were tested in vitro and in clinically-relevant LC and MPM mouse models by bioluminescence imaging, BLI of tumor burden progression. To counteract PD-1/PD-L1 inhibition in vivo, we evaluated the efficacy of PD-1 blocking antibody or cell-intrinsic genetic-engineering strategies by cotranducing M28z CAR T cells with a PD-1 dominant negative receptor (PD1-DNR) or with PD-1/4-1BB fusion protein.

Results:
A single, low-dose of M28z CAR T cells is able to resist the progression of established tumor for 40 days, but mice eventually died with progressing tumor. Tumor harvest analysis demonstrated the PD-1 and PD-L1 upregulation on CAR T cells and tumor cells (Figure panel A). We then confirmed in vitro that PD-L1 inhibits M28z T-cell effector functions (proliferation, cytotoxicity and cytokine secretion). The addition of PD-1 blocking potentiates CAR T-cell therapy in vivo but its efficacy requires multiple injections (Panel B). In contrast, a single dose of M28z T cells coexpressing PD1-DNR restore effector functions, enhance tumor burden control (Panel C) and prolong median survival (56 vs 82 days, p=0.001). Converting PD-L1 inhibition into a positive costimulatory signal by PD-1/4-1BB construct cotransduction into M28z CAR T cells enhanced cytokine secretion and T-cell accumulation (Panel D). Figure 1



Conclusion:
Our results demonstrate the therapeutic benefit of providing optimal costimulation and coinhibitory blockade to counteract PD-L1/PD-1 immunosuppression, thus potentiate CAR T-cell therapy for lung cancer and mesothelioma.

Background:
Malignant pleural mesothelioma (MPM) is an aggressive cancer with a poor prognosis and an increasing incidence, for which novel therapeutic strategies are urgently required. Since the immune system has been described to play a role in protection against MPM, characterization of its tumor immune microenvironment (TME) and immune checkpoints might help to identify new immunotherapeutic targets and their predictive and/or prognostic value.

Methods:
Immunohistochemistry (IHC) was performed on tissue samples of untreated (n=40) and chemotherapy-pretreated (n=14) MPM patients. Different subsets if immune cells were identified based on staining for CD4, CD8, FoxP3, CD68, CD45RO and granzyme B. The expression of the immune checkpoints TIM-3, LAG-3, PD-1 and its ligand PD-L1 was also investigated. The relationship between the immunological parameters and survival, as well as response to chemotherapy was analyzed using the R statistical software.

Results:
All patients had CD8+ tumor infiltrating lymphocytes (TILs), CD68+ histiocytes and macrophages and CD45RO+ T cells in their stroma, with CD8+ TILs being the predominant cell type of the immune infiltrate. Stromal CD4+ TILs were found in 75% of the untreated and 71% of the pretreated samples. A subset of those cells was also FoxP3+ and these CD4+FoxP3+ cells were positively correlated with stromal CD4 expression (p<0.001). Less than half of the samples showed positivity for granzyme B. Both, untreated and pretreated patients had PD-1+ TILs, while only 10% of the untreated patients also had PD-1+ tumor cells. PD-L1 positivity on lymphocytes and/or tumor cells was observed for more than half of the patients, with significant differences according to the histological subtype (p<0.001). Patients with a sarcomatoid histology showed the most PD-1 expression. TIM-3 was expressed in tumor cells, stromal lymphocytes and plasma cells, less often in pretreated samples compared to untreated samples. All samples were negative for LAG-3. After multivariate analysis stromal CD45RO expression was found to be an independent negative predictive factor for response to chemotherapy (p=0.017) and expression of CD4 and TIM-3 in lymphoid aggregates were good prognostic factors (p=0.008; p=0.001).

Conclusion:
Our data reveal the diversity of immune cells present in MPM and point to TIM-3 as a new target in mesothelioma. Administering chemotherapy before or together with PD-1/PD-L1/TIM-3 blocking agents may not be the best combination sequence and further research on the predictive value of CD45RO in the stroma might guide patient selection for chemotherapy.

Abstract not provided

Background:
Overall survival (OS) is considered as the gold standard for evaluating efficacy of antineoplastic treatments, including chemotherapy and targeted therapies. In randomized trials, allowing patients to cross-over to the other arm usually prevents demonstration of a survival benefit. However, it may provide important information with clinical relevance.

Methods:
The phase III IFCT-1503 ULTIMATE study compared weekly paclitaxel and bevacizumab (wPB) vs. docetaxel (DOC) as second- or third-line therapy in non-squamous NSCLC. At progression, patients were allowed to cross over to the other arm. Date of progression was collected for patients who crossed over to the other arm and for those who did not cross over but received a post-discontinuation treatment within 60 days following progression. Post-discontinuation progression-free survival (PFS2) and OS2 were calculated from day 1 of post-discontinuation treatment.

Results:
The study met its primary endpoint, PFS, which was significantly improved in the wPB arm (medians 5.4 vs. 3.9 mo, hazard ratio (HR) 0.62, p=0.006). No overall survival was observed (medians 9.9 vs. 11.4 mo, HR 1.18, p=0.4). Out of patients treated with DOC (n=55), those who crossed over to wPB (n=21, 38.2%) had a median PFS2 of 4.9 mo [3.1-6.2] and a median OS2 of 12.5 mo (7.0-NR), whereas those who did not cross over but received a post-discontinuation treatment (n=13, 23.7%) had a median PFS2 of 1.7 mo [1.1-2.2] and a median OS2 of 4.1 mo [2.1-5.9]. Out of patients treated with wPB (n=111), median PFS2 was 1.9 mo [1.2-2.2] for those who crossed over to DOC (n=9, 8.3%) and median PFS2 and OS2 were 1.9 mo [1.7-2.6] and 5.0 m [3.4-9.0] for those who did not cross over but received a post-discontinuation treatment (n=57, 52.3%).

Conclusion:
Allowing patients to cross over to the other arm demonstrated benefit of wPB following progression on docetaxel and explains the absence of OS benefit.

Background:
Targeting the tumor microenvironment, such as tumor angiogenesis, has led to the successful development and approval of a number of targeted therapies thereby changing the standard of care for many types of cancer. However, treatment options are limited in third-line non-small cell lung cancer (NSCLC) patients. Fruquintinib is a potent and highly selective oral kinase inhibitor targeting vascular endothelial growth factor receptors and is currently in late stage development for multiple cancers. This Phase II study was designed to evaluate the efficacy and safety of fruquintinib in third-line NSCLC patients (NCT02590965).

Methods:
A total of 91 patients were randomized to receive best supportive care (BSC) plus fruquintinib or BSC plus placebo in a 2:1 ratio from 12 Chinese clinical centers. Fruquintinib initial dose was 5 mg once daily and treatment was given in every 4-week cycle (3 weeks treatment followed by 1 week off). The primary objective was to compare progression free survival (PFS) between the two treatment groups. Secondary efficacy parameters included objective response rate (ORR), disease control rate (DCR), overall survival (OS). Tumor response was assessed per RECIST 1.1.

Results:
As of August 7, 2015, median PFS was 3.8 months for the fruquintinib group comparing with 1.2 months for the placebo group (hazard ratio=0.27, p<0.001). The ORR was 16.4% for the fruquintinib group comparing with 0% for the placebo group (p=0.02). The DCR of the fruquintinib group was significantly higher than that of the placebo group with a difference of 53.8% (36.3, 71.4; 95% CI, p<0.001). OS was not mature and initial analysis revealed 3- and 6-month OS rates of 90.2% and 68.3% for the fruquintinib group, and 73.3% and 58.2% for the placebo group, respectively. Adverse event was reported in 68.9% and 60.0% patients in fruquintinib and placebo group, respectively. The incidence of serious adverse events was 3.3% in the fruquintinib group and 6.7% in the placebo group.

Conclusion:
Fruquintinib in third-line NSCLC met the primary efficacy endpoint of PFS and demonstrated superiority in the secondary endpoints of ORR and DCR as compared with placebo. OS has yet to mature. Fruquintinib was generally well tolerated and safety profile consistent with previously reported. These results support further development of fruquintinib in third-line NSCLC patients. A randomized, double-blind, multi-center Phase III registration study was initiated in December 2015 (NCT02691299). Clinical trial information: NCT02590965.

Abstract not provided

Background:
RET rearrangements are actionable drivers found in 1-2% of non-small cell lung cancers. We previously reported the efficacy and safety of the multikinase RET inhibitor cabozantinib in 16 patients with RET-rearranged lung cancers in the first stage of our Simon two-stage phase 2 clinical trial (overall response rate 38%; Drilon, ASCO 2015). This study has since completed accrual of both stages, now with 26 patients treated with cabozantinib.

Methods:
This was an open-label, single center, phase 2 trial (NCT01639508). Eligibility criteria: stage IV pathologically-confirmed lung cancers, presence of a RET rearrangement, KPS >70%, and measurable disease. RET rearrangements were detected by FISH or next-generation sequencing. Cabozantinib was administered in tablet form at 60 mg daily until progression of disease or unacceptable toxicity. The primary objective was to determine the overall response rate (ORR, RECIST v1.1). Secondary objectives included determining progression-free survival (PFS), overall survival (OS), and toxicity. 5 responses in 25 response-evaluable patients were required to meet the primary endpoint (Simon two-stage minimax design: H~0~ 10% vs H~A~ 30% ORR). All patients who received at least one dose of cabozantinib were evaluable for toxicity.

Results:
26 patients with RET-rearranged lung adenocarcinomas were treated with cabozantinib. KIF5B-RET was the predominant fusion type identified in 16 (62%) patients. The median number of prior chemotherapy lines was 1 (0-5). One patient who discontinued therapy in cycle 1 and did not undergo a response assessment was not response-evaluable as per protocol. The study met its primary endpoint with confirmed partial responses observed in 7 (ORR 28% [95% CI 12-49%]) of 25 response-evaluable patients. The median PFS was 5.5 months (95% CI 3.8-8.4). The median OS was 9.9 months (95% CI 8.1-not reached). Response by RET fusion partner: Unknown (FISH+) 2/6 (33%), KIF5B 3/15 (20%), CLIP1 1/1, TRIM33 1/1, CCDC6 0/1, ERC1 0/1. In 26 patients evaluable for toxicity, the most common all-grade treatment-related adverse events were increased alanine aminotransferase in 25 (96%) patients, increased aspartate aminotransferase in 19 (73%) patients, hypothyroidism in 18 (69%) patients, diarrhea in 16 (62%) patients, and palmar plantar erythrodysesthesia in 15 (58%) patients. Nineteen (73%) patients required dose reduction.

Conclusion:
This study met its primary endpoint. Cabozantinib is an active agent in patients with RET-rearranged lung cancers. An improved understanding of tumor biology and novel therapeutic approaches will be required to improve outcomes with RET-directed therapy.

Background:
Increasing studies have shown that anti-angiogenic therapy targeting VEGF/VEGFR2 axis are furnishing demonstrable therapeutic effect on lung cancer，but the treatment benefit is transitory in clinic, generally followed by restoration of tumor growth and disease progression. Blockade of VEGF/VEGFR2 pathway can not only induce anti-vascular effect, but also remodel the immunosuppressive tumor microenvironment probably due to promoting suppressive cells infiltration and enhancing PD-L1 expression, resulting in impairing antitumor immunity. Therefore, the current study aimed to investigate whether combining anti-angiogenic and anti-PD-L1 treatments can induce synergistic antitumor effect by enhancing antitumor immune response in murine lung cancer.

Methods:
We evaluated the antitumor effects of anti-VEGFR2 agent (apatinib) as monotherapy or in combination with anti-PD-L1 monoclonal antibody in a murine lung cancer model using Lewis lung cancer cells (LLCs). The changes of immune components in tumor and spleen were dynamically tested in different treatment groups and time points by flow cytometry and immunohistochemistry.

Results:
The results showed that VEGF/VEGFR2 blockade could retard tumor growth and inhibit tumor neovascularization via eradicating Foxp3[+ ]regulatory T cells (Tregs) and myeloid derived suppressive cells (MDSCs) and reducing the density of microvessels in the first two weeks of treatment. On the third week of apatinib monotherapy, the number of Foxp3[+ ]Tregs and MDSCs had increased again. Although VEGF/VEGFR2 blockade induced more tumor infiltrating lymphocytes (TILs), especially CD8[+] T cells, infiltrating into the tumor mass than control group (P < 0.01), the expression of PD-1 and PD-L1was also significantly upregulated than that control group (P < 0.01). Compared to apatinib monotherapy, combining treatment demonstrated that anti-VEGFR2 plus anti-PD-L1 therapy could significantly inhibit tumor growth (P < 0.01) by persistently eliminating Foxp3[+ ]Tregs and MDSCs. Furthermore, combining anti-VEGFR2 and anti-PD-L1 therapy could not only dramatically increase TILs infiltration, especially CD8[+] T cells, but also significantly reduce the expression of PD-1 and PD-L1.

Conclusion:
Simultaneous blockade of VEGF/VEGFR2 and PD-1/PD-L1 pathways induced a synergistic anti-tumor effect in-vivo, possibly through eliminating immunosuppressive components including Tregs and MDSCs and enhance antitumor immune response.

Abstract not provided

Background:
Ramucirumab, an antiangiogenic IgG1 VEGFR2-targeted monoclonal antibody, and erlotinib, an EGFR tyrosine kinase inhibitor, are both active in advanced NSCLC. This global phase 1b/3 study (NCT02411448) will assess safety, tolerability and efficacy of the combination of ramucirumab with erlotinib in previously untreated patients with EGFR mutation-positive stage IV NSCLC. Here we report phase 1b safety results.

Methods:
Eligible patients with ECOG PS 0-1, an activating EGFR mutation, and previously untreated stage IV NSCLC received ramucirumab 10 mg/kg intravenously on day 1 of repeating 14-day (± 3 days) cycle and erlotinib 150 mg orally daily. Treatment continued until disease progression or unacceptable toxicity. The primary objective of part A was to assess the safety and tolerability, in terms of dose limiting toxicities (DLT), of adding the recommended dose of ramucirumab for phase 3 (part B) to standard dose erlotinib. Data were analyzed separately for Japan (JP) (cohort 1) and US/EU (cohort 2). The DLT assessment occurred during the first 2 cycles (approximately 28 days).

Results:
As of Dec 16th, 2015, 14 patients were treated in the phase 1b part of this trial and 12 were DLT evaluable (6 JP; 6 US/EU). Overall, 6 grade (Gr) 3 treatment-emergent adverse events (TEAE) were noted, with at least one TEAE in 5 patients; no serious adverse events or Gr 4-5 TEAEs occurred. In the JP cohort the median age was 73 (64-79), 57% had ECOG PS 1 and 29% had a history of smoking. Four patients (57%) experienced a Gr 3 TEAE, of which one was a DLT (elevation of alanine aminotransferase) while the others (hypertension [n=2], dermatitis acneiform, and diarrhea) were not DLTs. In the US/EU cohort the median age was 71 (31-83), 86% had ECOG PS 1, and no patients had a history of smoking. One patient experienced Gr 3 TEAE of rash; no DLTs were observed in this cohort.

Conclusion:
Enrollment on the phase 1b portion of this trial is complete and the safety results were consistent with previous combinations of antiangiogenic/erlotinib in this patient population. No unexpected toxicities were identified. Phase 3 enrollment has been initiated maintaining the dose of ramucirumab at 10 mg/kg Q2W.

Background:
There is growing evidence that tumor-associated fibroblasts (TAFs) play a major role in critical steps of tumor progression in solid tumors including NSCLC. However the role of TAFs in regulating the response to targeted therapies is poorly understood. One of such targeted therapies is nintedanib (NTD), a multi-kinase inhibitor of VEGF, FGF and PDGF receptors that has been recently approved to treat advanced lung adenocarcinoma (ADC) patients. Although the therapeutic effects of NTD in lung cancer have been associated with its anti-angiogenic functions, NTD has also been shown to exhibit anti-fibrotic effects in patients with idiopathic pulmonary fibrosis. Since lung fibrosis is largely driven by activated fibroblasts/myofibroblasts, and TAFs are positive for myofibroblasts markers, it is conceivable that NTD anti-tumor effects may be additionally driven through its direct action on lung TAFs. The main goal of this study was to analyze the latter hypothesis.

Methods:
Patient derived lung TAFs from ADC and SCC patients as well as paired control fibroblasts from non-malignant pulmonary tissue were exposed to increasing concentrations of NTD and analyzed for growth and activation upon stimulation with growth factors and TGF- β1, respectively. Activation markers included alpha-smooth muscle actin and collagen-I.

Results:
We found that NTD exhibited a dual inhibitory role in TAFs in terms of growth and TGF-β1-induced activation in a subtype-specific fashion. Specifically, NTD-mediated growth inhibition was larger in SCC-TAFs than in ADC-TAFs, which correlated with the larger Erk signaling previously reported by our group in SCC-TAFs in the absence of mitogenic stimuli. Conversely, inhibition by NTD of TGF-β1-mediated activation was larger in ADC-TAFs than SCC-TAFs. Likewise, NTD inhibited the growth and invasive advantages of ADC cancer cells in vitro elicited by the conditioned medium of ADC-TAFs treated with TGF-β1 compared to those advantages elicited in the absence of NTD. These results reveal for the first time that the pro-tumorigenic effects of ADC-TAFs in vitro are markedly reduced in the presence of NTD.

Conclusion:
TAFs in vivo are largely activated and quiescent, and TGF-β1 is a potent fibroblast activator that is frequently upregulated in lung cancer and associated with poor prognosis. Based on these previous observations, we argue that our new findings strongly suggest that the selective therapeutic advantage observed for NTD in ADC patients may be in part related to its selective inhibition of TGF-β1-dependent activation of ADC-TAFs. These findings provide novel mechanistic insights on the subtype-specific therapeutic effects of NTD in NSCLC.

Background:
The objective of this study is to confirm limited resection efficacy as radical surgery in patients with minimally invasive lung cancer as indicated by high-resolution (HR) computed tomography (CT), and to confirm intraoperative cytology as a negative margin indicator and reliable margin non-recurrence predictor.

Methods:
Enrollment required patients with a tumor ≤ 2 cm in diameter, diagnosed or suspected as a clinical T1N0M0 carcinoma in the lung periphery based on a CT scan. They had to have a HRCT scan indicating a sub-solid nodule with tumor disappearance ratio; TDR ≥ 0.5. (TDR = 1- DM/DL; DM: maximum tumor diameter on mediastinal settings, DL: maximum tumor diameter on lung settings). Patients unfit for lobectomy and systematic lymph node dissection were excluded. We performed a wedge or segmental resection. The used stapling cartridges were washed with saline, which was cytologically evaluated. If cytology was cancer positive, additional margin was resected, and cytologic examination repeated. If the second exam was positive, a routine lobectomy and systematic lymph node dissection was performed. We aimed at enrolling 100 patients. The primary endpoint is 10-year local recurrence free survival rate.

Results:
This prospective study started in November 2003, and 101 patients were enrolled in 6 years. Of them, 99 were eligible for analysis. The mean age was 62 years (range: 30-75), and 60 were women. There were 11 Noguchi type A tumors, 54 type B tumors, 26 type C tumors, one type D tumor, one malignant lymphoma, 3 hyperplastic lesions, and 3 inflammatory fibroses. None of the 93 malignant nodules showed any vessel invasion. Although no positive cytology results were obtained, pathologically positive margin was reported after surgery in one type C patient. He later underwent a routine lobectomy and systematic lymph node dissection. There was no clear correlation between tumor size, TDR, and Noguchi subtype. No mortality occurred, but one patient developed postoperative pneumothorax and pneumonia, and another hemorrhagic gastric ulcer. With a median follow-up period of 105 months (range: 72−129) as of June 2016, there have been no recurrences, but one patient died for unspecified cause.

Conclusion:
We have repeatedly warned that delayed cut-end recurrence is possible following limited resection even for small sub-solid lung cancers. So far, however, HRCT scans appear to predict non- or minimally invasive sub-solid lung cancers with high reliability, warranting limited resection as curative surgery in this cohort. Intraoperative cytology reliably indicated negative margins and seems to predict freedom from local recurrence.

Background:
Lobectomy remains the standard treatment for early-stage non-small cell lung cancer (NSCLC).In practice, however, sublobar resection has been selectively offered for patients with clinical Stage IA NSCLC as curative treatment. To seek optimal surgical procedure for early stage lung cancer, we carried out retrospective analyses of 2122 patients who had undergone limited resection for c-T1N0M0 NSCLC from 26 institutions of Japanese association for chest surgery.

Methods:
A total of 1963 patients with lobectomy tolerance were eligible for survival analysis. We retrospectively categorized patients of these nodules on numbers of criteria for CT findings; scores were added according to the dominance of ground glass appearance (GGA); >75% = 0, <75% =1, and size of tumor; T1a =0, T1b =1. Statistical analyses were carried out using propensity-matching and Kaplan-Myer with log-rank testing.

Results:
We analyzed 1:1 matched 731 patients for segmentectomy and wedge resection with propensity matching.The overall survival (OS) for score 0 group was 90.2% in segmentectomy (n=419) and 94.7% in wedge resection (n=451) (p=0.0351). The disease free survival (DFS) for score 0 group was 90.2% in segmentectomy and 92.7% in wedge resection (p=0.0645). The OS for score 1 group was 93.6% in segmentectomy (n=278) and 80.4% in wedge resection (n=246)(P<0.001)(Fig. 1). The DFS for score 1 group is 94.1% in segmentectomy and 75.3% in wedge resection (P<0.001). The OS for scores 2 was 79.1% in segmentectomy (n=34) and 69.2% in wedge resection (n=34) (p=0.109). The DFS for score 2 group was 87.0% in segmentectomy and 58.1% in wedge resection (p=0.581). Figure 1



Conclusion:
This study showed that GGA dominant T1a may be treated by wedge resection where possible. The consolidation dominant T1b did not benefit from sublobar resection. In patients with GGA dominant T1b or consolidation dominant T1a, anatomical segmentectomy with curative intension may provide better prognosis.

Background:
The prognosis after standard lobectomy for non-small cell lung cancer (NSCLC) with interstitial lung disease (ILD) is poor. This study aimed to compare the prognosis after lobectomy and sublobar resection for early NSCLC with ILD.

Methods:
Among 794 consecutive patients with clinical stage I NSCLC who underwent complete resection, 107 patients with ILD on high-resolution computed tomography (HRCT), which was defined according to the American Thoracic Society, European Respiratory Society, Japanese Respiratory Society, and Latin American Thoracic Association classification, were identified.

Results:
Overall survival (OS) was significantly worse for patients with possible usual interstitial pneumonia (UIP) or UIP pattern than those with inconsistent with UIP pattern (3-year OS, 64.5% vs. 82.1%, respectively; P = 0.031). No significant difference existed in OS between lobectomy and sublobar resection for all patients with ILD (3-year OS, 67.1% vs. 81.9%, respectively; P = 0.14). Although in patients with inconsistent with UIP pattern, OS was similar between lobectomy and sublobar resection groups (3-year OS, 81.1% vs. 83.6%, respectively; P = 0.87), OS was better for patients who underwent sublobar resection than lobectomy in patients with possible UIP or UIP patterns (3-year OS, 81.0% vs. 50.5%, respectively; P = 0.069). Multivariate Cox analysis demonstrated that preoperative diffusing capacity of the lung for carbon monoxide (P = 0.018), not the surgical procedure (P = 0.14), was an independent prognostic factor for OS.

Conclusion:
Sublobar resection can be an alternative choice for clinical stage I NSCLC with ILD especially for UIP or possible UIP patterns on HRCT.

Abstract not provided

Background:
The identification of an accurate segment is essential for successful anatomic pulmonary segmentectomy. We have previously developed a new fluorescence technique using a PDD endoscope system[TM] and vitamin B2 for identification of pulmonary segments in animal experiments. In this study, we applied this technique clinically to examine the efficacy and safety in anatomical pulmonary segmentectomy and sub-segmentectomy for pulmonary malignancies.

Methods:
Our technique requires two key instruments, a PDD endoscope system[TM](KARL STORZ GmbH and Co., Tuttlingen, Germany) as a fluorescence sensing device and vitamin B2 solution as a fluorescent substance. 17 patients with small lung nodules were enrolled in this study. Regarding our surgical technique, after identification of the target segmental or sub-segmental bronchus, vitaminB2 solution is injected via the bronchus. The target segment is identified as a fluorescent segment with the PDD endoscope system[TM]. The identified segment is resected with an electric cautery, stapling devices, or combination of them. In case patient’s lung has severe abnormal change such as emphysema or fibrosis, another technique is indicated. After ligation of the target segmental or sub-segmental artery, vitaminB2 solution is systemically administrated with intravenous injection. The target segment is identified as a defect of fluorescence with the PDD endoscope system[TM]. Following outcomes were collected; success rate of identifying the pulmonary segments, pathological evaluation of dissection margin, duration of chest drainage, and perioperative complications.

Results:
A total of 18 procedures were performed using this technique. Performed segmentectomy or sub-segmentectomy were as follows; Right S1, S2, S3, S2a+3b, S6, S9, Left S1+2, S3, S4+5, S6, S8a+9b, S9+10. Resected nodules were 14 primary lung cancers, 1 MALT-lymphoma, 1 metastatic lung cancer, and 2 benign lung nodules. Histology of primary lung cancer was adenocarcinoma in all 14 cases. Pathological stage of lung cancer was 12 stageIA (pT1a; 10, pT1b; 2), 1 stageIIA (pT1aN1), and 1 stageIIIA (pT1aN2). The success rate of identifying pulmonary segments was 100%. Dissection of segmental border was performed with only electric cautery in 12 procedures, and with both of electric cautery and stapling device in 6 procedures. In all cases, no cancer cells were found on the resection margin pathologically. Mean drainage time was 1.7 days (1-4 days). Regarding perioperative complications, veno-vagal reflex was occurred after systemic injection of vitaminB2 in one case, and 1 delayed pneumothorax was found in one case.

Conclusion:
Our novel fluorescence technique involving a PDD endoscope system[TM] and vitaminB2 allowed performing accurate and safe pulmonary segmentectomy and sub-segmentectomy.

Background:
Although the confirmation of an appropriate resection margin from the tumor is crucial for reducing the risk of local recurrence after lung segmentectomy for pulmonary malignancies, there has been no method of measurement. We established a novel approach for performing segmentectomy by using an infrared thoracoscopy with transbronchial instillation of indocianine green (ICG), and improved this method by adding an advanced computer technology via lung volume analyzer for obtaining an appropriate resection margin.

Methods:
Preoperatively, each patient underwent multislice enhanced computed tomography (CT) using 320-slice scanners for pulmonary angiography and virtual bronchoscopy, and to create several virtual segmentectomies by using Volume Analyzer Synapse VINCENT (Fujifilm co., Tokyo, Japan). We measured the shortest distance from the tumor to the resection margin in each simulated segmentectomy and selected the most appropriate area of sublobar resection based on the adequate resection margin of approximately 2 cm from the tumor. We prospectively performed segmentectomy in 17 patients and compared between simulated distance and actual distance measured from the specimen.

Results:
The average number of created patterns of virtual segmentectomy in each case was 4.1 ± 1.0. The mean distance of resection margin in selected virtual segmentectomy was 19.3 ± 9.7 mm. On the other hand, actual shortest distance in resected specimen was 25.4 ± 8.1 mm, which was significantly longer than simulated distance (p=0.027). There was no tumor recurrence in all patients.

Conclusion:
Lung volume analyzer was an excellent tool for selecting an ideal area of sublobar resection with an appropriate resection margin.

Background:
The standard surgical treatment of stage I non-small cell lung cancer is anatomical lobectomy. However, in some cases, small peripheral lung cancer (≤2cm) is treated by sublobar resection. The purpose of this study was to define the necessity of completion lobectomy when the tumor was revealed as non-small cell lung cancer with pleural invasion or lymphovascular invasion after sublobar resection.

Methods:
We retrospectively reviewed 271 consecutive patients who underwent curative resection for stage I non–small cell lung cancer of 2 cm or less. We analyzed clinicopathological findings and survival between two groups with either invasion-positive tumor (tumor with visceral pleural invasion or lymphovascular invasion) or invasion-negative tumor (tumor without visceral pleural invasion and lymphovascular invasion): sublobar resection group and lobectomy group.

Results:
Except for age and pulmonary function, there were no differences in clinocopathological characteristics between sublobar resection group and lobectomy group with invasion-positive tumor or invasion-negative tumor. There was no difference in the 5-year recurrence-free survival rate between two groups in the invasion-positive tumor and invasion-negative tumor (78.9% vs 79.8%; p=0.928, 80.2% vs 85.4%; p=0.505). In the multivariate analysis, only number of dissected lymph nodes was a significant recurrence-related factor of stage I invasive-positive non-small cell lung cancer (hazard ratio 0.914, 95% confidence interval 0.845-0.988, p=0.023). Sublobar resection was not a risk factor for recurrence.

Conclusion:
The survival between sublobar resection group and lobectomy group in small (≤2cm) non-small cell lung cancer with visceral pleural invasion or lymphovascular invasion were not different. Completion lobectomy is not necessary in small lung cancer after sublobar resection whether the tumor has visceral pleural invasion or lymphovascular invasion.

Abstract not provided

Background:
In NSCLC, atezolizumab (anti-PDL1) efficacy correlates with PD-L1 expression on tumor cells (TC) and tumor-infiltrating immune cells (IC). Here we examined the association between atezolizumab efficacy and TMB assessed by FoundationOne (F1) sequencing panel.

Methods:
Pretreatment tumor specimens from 102 1L and 465 2L+ NSCLC patients enrolled on three Ph 2 atezolizumab monotherapy trials (POPLAR: randomized 2/3L trial comparing atezolizumab vs docetaxel; BIRCH/FIR: single-arm, 1L/2L+ PD-L1‒selected trials) were available for targeted genetic sequencing using the F1 panel of 315 cancer-related genes. TMB was quantified using an updated TMB algorithm and efficacy was assessed in groups defined by the 75th (high) and 50th (median) percentile of each study-specific TMB. Atezolizumab efficacy was examined at Dec 1, 2015 (POPLAR and BIRCH); and Jan 7, 2015 (FIR) data cutoffs.

Results:
Across samples, median TMB was similar in 1L and 2L+ patients (9/MB and 9.9/MB, respectively). In 1L and 2L+ PD-L1–selected patients, atezolizumab benefit was increased in those with ≥ TMB cut-offs (Table). In unselected 2L+ patients from POPLAR, the OS, PFS, and ORR benefits of atezolizumab vs docetaxel were also enhanced in patients with increased TMB. TMB and PD-L1 expression were independently associated with improved atezolizumab efficacy. TMB associations with PD-L1 expression, tumor-infiltrating lymphocyte infiltration and T-effector cell gene expression will be presented.

Conclusion:
For the first time, we demonstrate that TMB assessed with F1 targeted sequencing is associated with improved atezolizumab outcomes in 1L and 2L+ NSCLC. Moreover, this is the first study demonstrating the association of TMB with improved anti-PD-L1/PD-1 efficacy in a randomized trial. Importantly, the association between TMB and atezolizumab efficacy occurred in both unselected and PD-L1-selected patients. Therefore, in addition to PD-L1, TMB may be an independent predictor of improved responsiveness to atezolizumab in advanced NSCLC.

Atezolizumab efficacy by TMB subgroups
	PD-L1‒selected
BIRCH+FIR	1L n=102		2L+n=371
	Median (≥9/MB)	High (≥13.5/MB)	Median (≥9.9/MB)	High (≥17.1/MB)
OS,HR[a] (95% CI)	0.79 (0.39-1.58)	0.45 (0.17-1.16)	0.87 (0.65-1.16)	0.7 (0.49-1.00)
PFS,HR[a] (95% CI)	0.58 (0.36-0.94)	0.54 (0.3-0.97)	0.64 (0.5-0.8)	0.5 (0.38-0.67)
ORR,above/below cutoff	28%/13%	25%/20%	25%/14%	29%/16%
POPLAR	2L+ unselected n=92
	Biomarker- evaluable population	Median (≥9.9/MB)		High (≥15.8/MB)
OS,HR[b ] (95% CI)	0.65 (0.38-1.12)	0.48 (0.23-1.04)		0.5 (0.15-1.67)
PFS,HR[b] (95% CI)	0.98 (0.63-1.53)	0.49 (0.25-0.93)		0.49 (0.19-1.3)
ORR,atezolizumab/docetaxel	13%/15%	20%/4%		20%/8%
[a]HR:efficacy-evaluable patients, atezolizumab at/above cutoff vs below.[b]HR:efficacy-evaluable patients, atezolizumab vs docetaxel at/above cutoff.

Background:
Anti-PD-1 immunotherapy has resulted in durable clinical responses in heavily pretreated patients with non-small cell lung cancer (NSCLC). While NSCLC is typically seen as non-immunogenic, there is a 15-20% objective response rate and a median duration of response of 17 months in patients treated with the PD-1 inhibitor, nivolumab. This duration of response has not been reported with other systemic therapies in advanced NSCLC. While tumor PD-L1 expression may be a biomarker of sensitivity to anti-PD-1 therapy, and the number of somatic mutations may play a role in PD-1 upregulation on T cells, the mechanisms underlying response vs. progressive disease have yet to be fully elucidated.

Methods:
Whole exome sequencing and mutation-associated neoantigen (MANA) prediction was performed on tumor sections from two advanced NSCLC patients with complete response to nivolumab. Peptides representing MANAs were synthesized and tested against PBMC in a 10-day cultured IFNg ELISpot assay. Reactive MANAs were assessed in binding and stability assays. TCR sequencing was performed on reactive cell cultures and on DNA obtained from tumor resections to match MANA-reactive TCR clones with clones that were infiltrating the tumor.

Results:
The mean mutational burden in NSCLC as reported previously is 360 sequence alterations. In our study, patient 1 had 30 sequence alterations and patient 2 had 314. Despite the difference in mutational load, MANA reactivity was detected in peripheral blood of both patients >1 year after being declared cancer-free. TCR clones of MANA-reactive peripheral T cells were detected in tumor resections and were expanded in MANA-stimulated T cell cultures. Binding and stability assays confirmed that these MANAs bind to their cognate HLA with high affinity and stability.

Conclusion:
These findings show that NSCLC tumors with differential mutational burden can show regression following checkpoint blockade and suggest that the quality of mutations may be more influential in immunogenicity than the overall quantity of mutations. Our data also show that MANA reactivity may be the underlying mechanism by which T cells eliminate tumor following anti-PD-1 immunotherapy. Additional studies should evaluate mechanisms of enhancing MANA reactivity in patients who do not respond to checkpoint blockade and should further validate the link between MANA reactivity and clinical response to anti-PD-1.

Abstract not provided

Background:
The clinical efficacy observed with PD-1/PD-L1 inhibitors in non-small cell lung carcinoma (NSCLC) has prompted to characterize the immune response in lung tumors treated with chemotherapy. Our goal was to determine the characteristics of immune microenvironment of localized, surgically resected, NSCLCs from patients who received and did not receive neo-adjuvant chemotherapy. Using multiplex immunofluorescence (mIF) and image analysis, we investigated PD-1/PD-L1 expression, and quantified tumor infiltrating lymphocytes (TILs) and tumor associated macrophages (TAMs).

Methods:
We studied formalin-fixed and paraffin embedded (FFPE) tumor tissues from 111 stage II and III resected NSCLC, including 61 chemonaïve (adenocarcinoma, ADC=33; squamous cell carcinoma, SCC=28) and 50 chemotherapy-treated (ADC=30; SCC=20) tumors. mIF was performed using the Opal 7-color fIHC Kit™ and analyzed using the Vectra™ multispectral microscope and inForm™ Cell Analysis software (Perkin Elmer, Waltham, MA). The markers studied were grouped in two 6-antibody panels: Panel 1, AE1/AE3 pancytokeratins, PD-L1 (clone E1L3N), CD3, CD4, CD8 and CD68; and Panel 2, AE1/AE3, PD1, Granzyme B, FOXP3, CD45RO and CD57.

Results:
Positive PD-L1 expression (>5%) in malignant cells (MCs) was detected in 48% (n=53/111) of NSCLCs. Overall, chemotherapy-treated tumors showed significantly higher percentages of MCs expressing PD-L1 (median, 18.2%) than chemo-naïve cases (median, 1.8%; P=0.033). Higher densities of inflammatory cells expressing granzyme B (P=0.036), CD57 (P=0.001) and PD-1 (P=0.016) were detected in chemotherapy-treated NSCLCs compared with chemo-naïve tumors. In contrast, lower densities of FOXP3-positive regulatory T cells were detected in chemotherapy-treated tumors when compared with chemo-naïve cases (P=0.032). Following chemotherapy ADCs exhibited significantly higher levels of CD57-positive cells (P<0.0001) and lower density of FOXP3-positive cells (P=0.002) than chemo-naïve tumors. Chemotherapy-treated SCCs demonstrated higher density of PD-1-positive cells than chemo-naïve tumors (P=0.004). In chemotherapy-treated cancers, lower levels of CD4 helper T positive cells and tumor associated macrophages (TAMs) CD68-positive cells were associated with worse overall survival (OS; P=0.04 and P=0.005, respectively) in univariate analysis. In chemotherapy-treated ADC patients, lower levels of CD68-positive (P=0.010) and higher levels of FOXP3-positive cells correlated with worse OS (P=0.044).

Conclusion:
We developed a robust mIF panel of 10 markers to study inflammatory cells infiltrates in FFPE NSCLC tumor tissues. Chemotherapy-treated NSCLCs exhibited higher levels of PD-L1 expression and T cell subsets compared to chemo-naïve tumors, suggesting that chemotherapy activates specific immune response mechanisms in lung cancer. (Supported by CPRIT MIRA and UT Lung SPORE grants, and MD Anderson Moon Shot Program).

Background:
Previous attempts to define tumor and stromal immunologic environment in non-small cell lung cancer (NSCLC) utilized archival tissue. We established prospective comprehensive immonogenomic profiling protocol in NSCLC (ICON Project). The goal is to integrate immunomic, genomic, transcriptomic, proteomic, demographic, clinical, pathologic, and outcome data from 100 surgically resected early stage NSCLC.

Methods:
Tumor and normal lung tissue are collected at the time of surgery, blood samples before and after surgery. Tumor samples are processed for tumor infiltrating lymphocyte (TILs) isolation and expansion; development of patient derived xenografts (PDX), immunohistochemical immune markers, and immunopeptidome profiling. Blood samples are analyzed with flow cytometry.

Results:
57 patients with median age of 65 years (27 males) have been enrolled within 5 months, of which 33 (66%) contributed samples to the study. Four were never smokers, with others being former or current smokers. Majority (N=27) had adenocarcinoma, 4 squamous cell carcinoma, and 2 pleomorphic carcinoma. 15 patients had stage I, 11 stage II, and 7 stage III disease; 5 patients received induction chemotherapy. Median tumor size was 3.5 cm and 29 underwent R0 and 4 R1 resection. Pre-REP TIL expansion was successful in the majority of samples (68.2%, n=22). Twelve PDX models with a take rate of 40% have been generated. Interim analysis of tumor samples by IHC demonstrated higher median distribution of all cell types: CD3+ T cells, cytotoxic T cells CD8+, PD1+ cells, tumor associated macrophages (TAM) CD68+, TAM CD68+PD-L1+, CD20+B cells, memory T cells CD45R0, natural killer cells CD57+, regulatory FOXP3+ T cells, and cytotoxic granzyme B cells (cells/mm[2]) in the stroma as compared to the tumor compartment. Intra-tumoral regulatory FOXP3+T cells were more abundant in squamous cell carcinomas compared to adenocarcinomas (median 312 vs 51 cells/mm[2], p = 0.05). Higher concentration of intra-tumoral CD68+PD-L1+ expressing cells was observed following neoadjuvant chemotherapy (median 97 vs 60 cells/mm[2] no chemo; p= 0.077), as was the concentration of memory T cells CD45R0 (median 129 vs 30 cells/mm[2], no chemo; p = 0.077). Mass spectrometry-based immunopeptidome analysis identified several thousand peptides, of which 4 promising antigens have been chosen for further development as immunotherapeutic T-cell targets.

Conclusion:
The ICON is an ongoing, ambitious prospective project that aims to define the baseline immunologic characteristics of surgically resectable NSCLC. The rapid enrollment illustrates the enthusiasm for tumor immunoprofiling amongst patients and physicians alike. Data from this patient cohort will serve as a baseline comparison for upcoming neoadjuvant immunotherapy trials.

Background:
Immunotherapy with antibodies against B7/CD28 family members, including PD-1, PD-L1, and CTLA-4 has shifted the treatment paradigm for non-small-cell lung carcinoma (NSCLC) with improved clinical outcome. HHLA2 is a newly discovered member of the family. By regulating T-cell function, HHLA2 could contribute to tumor immune suppression and thus be a novel target for cancer immunotherapy. There is limited information and critical need to characterize its expression profile and clinical significance in NSCLC.

Methods:
We performed immunohistochemistry with an HHLA2-specific antibody (clone 566.1) using tissue microarrays constructed from 679 NSCLC tumor tissues, including 392 cases in the discovery set and 287 cases in the validation cohort. We also studied clinico-pathological characteristics of these patients.

Results:
Overall, HHLA2 was not detected in most of normal lung tissue but expressed in 66% of NSCLC across different subtypes. In particular, EGFR–mutated NSCLC was significantly associated with higher tumor HHLA2 expression in both discovery (EGFR vs. WT: 76% vs. 53%, P=0.01) and validation cohorts (89% vs. 69%, P=0.01). In one of the two cohorts, HHLA2 expression was higher in lung adenocarcinoma as compared to squamous and large cell histology, non-Hispanic White vs. Hispanics, and tumors with high tumor infiltrating lymphocyte (TIL) density. In the multivariate analysis, EGFR mutation status and high TIL intensity were independently associated with HHLA2 expression in lung adenocarcinoma.

Conclusion:
HHLA2 is widely expressed in NSCLC and is associated with EGFR mutation and high TILs in lung adenocarcinoma. It is potentially a novel target for lung cancer immunotherapy.

Abstract not provided

Background:
IL-7/IL-7 receptor (IL-7R) interactions have been shown to prevent apoptosis in lung cancer cells and promote stromal pro-tumor immune cell homing and differentiation. The aim of this study is to investigate the correlation between tumoral IL-7R expression and stromal pro-tumor immune cells (FoxP3+ Tregs and CD163+ M2 macrophages) and to determine prognostic impact of the combination of these markers in lung adenocarcinomas.

Methods:
In resected stage I lung adenocarcinoma (n=913; 1995-2009), antigen expression of IL-7R, FoxP3 and CD163 was evaluated by immunohistochemistry (IHC) using tissue microarrays and mRNA expression was quantified by RT-PCR. Prognosis was analyzed by both recurrence free probability (RFP) and lung cancer-specific survival (LCSS).

Results:
In IHC analysis, high tumoral IL-7R, stromal FoxP3, and stromal CD163 expression were individually associated with lymphatic/vascular invasion, and increasing percentage of solid histological patten. A correlation was seen between IL-7R, FoxP3 and CD163 expression by mRNA and IHC analyses (Figure1). The co-existence of high expression of these 3 markers was found in 16% of patients and was associated with worse outcomes (Figure2). In multivariable analysis, triple marker co-existence was an independent risk factor for RFP (p=0.004) and LCSS (p=0.008).

Conclusion:
Tumoral IL-7 receptor is a potential target for lung adenocarcinoma immunotherapy. Figure 1 Figure 2

Background:
Monoclonal Antibodies (mAbs) against the Epidermal Growth Factor Receptor (EGFR) in association with platinum-based doublet chemotherapy have emerged as a potential first-line treatment option for advanced non-small cell lung cancer (NSCLC). This study was conducted to systematically review available data and evaluate the efficacy and toxicity of anti-EGFR mAbs plus chemotherapy vs chemotherapy alone for advanced NSCLC.

Methods:
We carried out a search on network databases and oncology conference abstracts for studies between 1990 and January 2016. Only prospective randomized clinical trials were included. Primary endpoints were overall survival (OS) and toxicity frequency. Secondary endpoints were progression-free survival (PFS) and overall response rate (ORR). Subgroup analysis was performed assessing histological subtypes, EGFR protein expression by immunohistochemistry (IHC), EGFR gene copy number by fluorescence in-situ hybridization (FISH), EGFR mutation status, and smoking status.

Results:
Seven studies (2 with necitumumab and 5 with cetuximab) were included with 5,057 patients. Compared to chemotherapy alone, significant benefits were demonstrated by the addition of anti-EGFR mAb to chemotherapy in OS (HR 0.90; 95%CI 0.84-0.95), PFS (HR 0.93; 95%CI 0.87-0.98), and ORR (OR 1.27; 95%CI 1.06-1.51). In subgroup analyses, the association of anti-EGFR mAb was associated with improved OS among patients with squamous histology (HR 0.84; 95%CI 0.76-0.92), tumours with high EGFR expression by IHC (HR 0.83; 95%CI 0.70-0.98), and smokers (HR 0.87; 95%CI 0.79-0.96). Patients with squamous histology and high EGFR expression by IHC achieved the highest benefit with the association (HR 0.71; 95%CI 0.59-0.86). The OS with the association also seemed to be higher in EGFR FISH negative and in EGFR wild-type tumours, but without statistical significance. Chemotherapy plus anti-EGFR mAb caused more grade 3 or worse adverse events (OR 1.73; 95%CI 1.50-2.00), remarkedly these known to be associated with anti-EGFR therapy, such as acne-like rash (OR 34.13; 95%CI 16.40-71.00) and hypomagnesemia (OR 6.23; 95%CI 3.04-12.77).

Conclusion:
Anti-EGFR therapy plus platinum-based doublet chemotherapy as first-line treatment demonstrated significant efficacy benefits with acceptable toxicity for advanced NSCLC. This benefit is more expressive among squamous histology with high EGFR expression. EGFR protein expression by IHC seems to be a predictive marker for survival in the association group. Further research is needed to corroborate these findings.

Background:
SQUIRE (NCT00981058) demonstrated adding necitumumab (N) to gemcitabine/cisplatin (GC) improved survival in patients with Stage IV squamous NSCLC (SQ-NSCLC). Retrospective analysis revealed consistent treatment effect in favor of patients receiving N monotherapy as continuation after chemotherapy (CT) (GC+N continuation patients) versus continuation therapy-eligible GC arm patients (GC non-progressors). In the EU, N is approved for patients with EGFR-expressing tumors. We repeated the analysis in this patient population.

Methods:
Patients with Stage IV SQ-NSCLC were randomized 1:1 for ≤6 cycles of G (1250 mg/m[2] iv, Days [d] 1,8) and C (75 mg/m[2] iv, d1) either with or without N (800 mg iv, d1,8). Patients in GC+N without progression continued N until progressive disease (PD). SQUIRE included mandatory tissue collection. EGFR protein expression was assessed by IHC in a central lab (Dako EGFR PharmDx kit). Analyses were done in EGFR-expressing patients (EGFR >0). Patients who received ≥4 cycles of CT without PD were included. Overall survival (OS) and progression-free survival (PFS) were calculated by Kaplan-Meier method. 95% CIs and hazard ratios estimated using stratified Cox proportional hazards model.

Results:
Of 1093 patients (ITT population), 982 patients (89.8%) had evaluable IHC assay results; 935/982 (95.2%) had EGFR>0. GC+N arm continuation therapy patients included 228 patients with EGFR>0 and 194 patients (EGFR>0) were GC arm non-progressors. Baseline characteristics were similar except gender (Males: 81% in GC+N vs 91% in GC arm). CT exposure was balanced. Median OS from randomization in GC+N vs GC was 16.1 vs 14.9 months; HR 0.76 (95% CI, 0.61, 0.95). Median PFS in GC+N vs GC was 7.4 vs 6.9 months; HR 0.81 (95% CI, 0.66, 1.00). Figure 1



Conclusion:
In patients with EGFR-expressing tumors, a consistent treatment effect in favor of GC+N continuation maintenance compared to GC non-progressors was observed, similar to ITT population with no unexpected increases in AEs.

Background:
Squamous cell carcinoma (SCC) of the lung is a genetically complex and difficult-to-treat cancer. In LUX-Lung 8, afatinib (40mg/day) significantly improved OS (median 7.9 vs 6.8 months, HR=0.81 [95% CI, 0.69‒0.95], p=0.008), PFS (2.6 vs 1.9 months, HR=0.81 [0.69‒0.96], p=0.010) and DCR versus erlotinib (150mg/day) in patients with relapsed/refractory SCC of the lung (n=795). Notably, 12-month (36 vs 28%; p=0.016) and 18-month survival (22 vs 14%; p=0.016) was significantly higher with afatinib than erlotinib, indicating that some patients derive prolonged benefit from afatinib. Here, we present post-hoc analysis of baseline characteristics and efficacy/safety of afatinib in long-term responders (treatment for ≥12 months). Hypothesis-generating analysis of archived tumor samples and blood serum was undertaken to identify possible molecular/clinical biomarkers.

Methods:
Tumor samples were retrospectively analyzed using FoundationOne[TM] next-generation sequencing (NGS); EGFR expression was determined by immunohistochemistry. Pre-treatment serum samples were analyzed with VeriStrat[®], a MALDI-TOF mass spectrometry test, and classified as VeriStrat-Good or VeriStrat-Poor-risk.

Results:
15/398 patients treated with afatinib were long-term responders. Median duration of treatment was 16.6 months (range: 12.3‒25.8). Patient characteristics were similar to the overall dataset (median age: 65 years [range: 54‒81]; male: 80.0%; Asian: 13.3%; ECOG 0/1: 40.0%/60.0%; best response to chemotherapy CR or PR/SD: 53.3%/46.7%; current and ex-smokers: 80.0%). Median PFS was 16.2 months (range: 2.8‒24.0); median OS was 23.1 months (range: 12.9‒31.5). The most common treatment-related AEs (all grade/grade 3) were: diarrhea (73.3%/6.7%); rash/acne (66.7%/6.7%); stomatitis (13%/7%). AEs generally occurred soon after treatment onset (median onset, days [range]: diarrhea 11 [5‒48]; rash/acne 17 [9‒107]; stomatitis 15 [11‒19]). Four patients required a dose reduction to 30mg/day due to treatment-related AEs (diarrhea, rash, stomatitis, diarrhea/rash). NGS was undertaken in 9 patients and details will be presented at the meeting. Genomic aberrations in the ErbB/FGF gene families were identified in 44.4%/55.6% of long-term responders (overall dataset: 29.4%/58.0%). Of 14 patients assessed by VeriStrat, 85.7% were VeriStrat-Good (overall dataset: 61.6%). Immunohistochemistry data was available for two patients; one overexpressed EGFR (≥10% positive cells; H-score ≥200)

Conclusion:
Baseline characteristics of long-term responders to afatinib were similar to the overall dataset. In this sub-group, afatinib conferred a survival benefit of nearly 2 years. Afatinib was well tolerated with predictable and transient AEs that occurred soon after treatment onset. The dataset was too small to identify any clear NGS/VeriStrat predictive signals. Further studies are required to predict long-term response to afatinib.

Abstract not provided

Background:
The irreversible ErbB family blocker afatinib and the reversible EGFR TKI gefitinib are approved for first-line treatment of advanced EGFRm+ NSCLC. This Phase IIb trial prospectively compared afatinib versus gefitinib in this setting.

Methods:
LUX-Lung 7 assessed afatinib (40 mg/day) versus gefitinib (250 mg/day) in treatment-naïve patients with stage IIIb/IV NSCLC harbouring a common EGFR mutation (Del19/L858R). Co-primary endpoints were PFS (independent review), time to treatment failure (TTF) and OS. Other endpoints included ORR and AEs. In case of grade ≥3/selected grade 2 drug-related AEs the afatinib dose could be reduced to 30 mg or 20 mg (minimum). The primary analysis of PFS/TTF was undertaken after ~250 PFS events. The primary OS analysis was planned after ~213 OS events and a follow-up period of ≥32 months.

Results:
319 patients were randomised (afatinib: 160; gefitinib: 159). At the time of primary analysis, PFS (HR [95% CI] 0.73 [0.57‒0.95], p=0.017), TTF (0.73 [0.58‒0.92], p=0.007) and ORR (70 vs 56%, p=0.008) were significantly improved with afatinib versus gefitinib. The most common grade ≥3 AEs were diarrhoea (13%) and rash/acne (9%) with afatinib and elevated ALT/AST (9%) with gefitinib. 42% of patients treated with afatinib had ≥1 dose reduction due to AEs; dose reductions were more common in females than males (77%/23%) and non-Asians than Asians (64%/36%). Dose reduction of afatinib did not negatively impact PFS (<40mg vs ≥40mg; HR [95% CI]: 1.34 [0.90‒2.00]) but reduced incidence and severity of drug-related grade ≥3 AEs. Treatment discontinuation due to drug-related AEs was the same in each arm (6%). The data cut-off for primary OS analysis occurred on 8 April 2016. At this time, median treatment duration (range) was 13.7 (0‒46.4) versus 11.5 (0.5‒48.7) months with afatinib and gefitinib. 25% (afatinib) and 13% (gefitinib) of patients received treatment for >24 months. 73% and 77% of patients in the afatinib and gefitinib arms had ≥1 subsequent systemic anti-cancer treatment, with 46% and 56% receiving a subsequent EGFR-TKI including osimertinib (14%)/olmutinib (14%). OS data, including subgroup analysis with respect to subsequent therapy will be presented at the meeting.

Conclusion:
Afatinib significantly improved PFS, TTF and ORR versus gefitinib in EGFRm+ NSCLC patients, with a manageable AE profile and few drug-related discontinuations. Dose adjustment of afatinib reduced drug-related AEs without compromising efficacy. Primary OS analysis will be reported.

Background:
Since the epidermal growth factor receptor-tyrosine kinase inhibitor (EGFR-TKI) was launched in Japan, the survival periods of advanced/recurrent EGFR mutation positive (EGFR m+) non-small cell lung cancer (NSCLC) patients have been getting longer. However, clinical factors which contributed to the extension of survival periods of these patients remain unclear. We investigated overall survival, prognostic factors and treatments patterns of EGFR m+ NSCLC patients in real-world clinical practice.

Methods:
This is a multi-center, observational, retrospective study. Histologically or cytologically diagnosed EGFR m+ NSCLC patients who were started first-line treatment from 1/1/2008 to 31/12/2012 were enrolled. The primary objective was the estimated OS. The secondary objectives were to determine prognostic factors, real-world treatment patterns.

Results:
1,660 EGFR m+ NSCLC patients were enrolled from 17 hospitals in Japan (median age 67.0 years, female 64.8%, 38.9% had smoking history, ECOG-performance status 0, 1, 2, 3, 4 were 39.5%, 41.1%, 7.1%, 4.9%, 0.7%, respectively, adenocarcinoma 95.2%, 50.1% exon 19 deletion, 66.7% at stage IV). Median estimated OS was 29.7 months. Cox regression analysis revealed age, smoking history, performance status, histological diagnosis, EGFR mutation type and clinical stage were independently associated with OS. Five year survival rate of stage IV patients was 13.8%. The median number of treatment regimens was two. EGFR-TKI and platinum-doublet chemotherapy were most frequently used as first- and second-line treatments.

Conclusion:
Real world treatment of the large data-set of 1,660 EGFR m+ NSCLC patients was retrospectively investigated. Median OS was 29.7 months and EGFR-TKIs are major components of the treatment regimens for these patients in Japan. (NCT0247520)

Background:
In 2011, following gefitinib (IRESSA[®]) NDA voluntary withdrawal, US patients benefiting from gefitinib were eligible to continue gefitinib through the IRESSA Clinical Access Program (ICAP), an IRB-approved protocol. A subset of ICAP investigators subsequently collected additional retrospective data on their ICAP patients through another IRB-approved project (“chart-review subset”).

Methods:
For all enrolled ICAP patients, demographic and serious adverse event (SAE) reports were reviewed. All ICAP investigators were invited to participate in chart review; 47 accepted and collected data on patient/tumor characteristics and safety/tolerability of prolonged gefitinib therapy among their 79 ICAP patients.

Results:
Across 137 US sites, 191 patients enrolled in ICAP. As of September 2016, 75 (39%) remain on gefitinib; discontinuations were due to progression (36%), death (34%), AEs (13%), or other (17%). Sixty-four (34%) patients reported 162 SAEs; 5 (2.6%) patients had 12 SAEs considered to be gefitinib-related by investigators. The chart-review subset included 79 (41%) patients with median age of 69 years at ICAP enrollment, who were predominantly female (70%) and white (84%); 95% had a confirmed NSCLC diagnosis. Due to the evolving understanding of genetic mutations in NSCLC at the time of gefitinib initiation, the majority of patients (79%) never had EGFR sequencing performed. Although tissue is not available for EGFR status confirmation, we assume these patients are nearly exclusively EGFR mutation-positive. Median total length of gefitinib was 11.1 years (6.5-15.1; Table). Long-term gefitinib was well-tolerated; 5% discontinued due to a gefitinib-related AE. Ten-year survival rate from first-ever initiation of gefitinib was 86% and 15-year was 59%. Table. Gefitinib treatment patterns and tolerability among ICAP chart-review patients.

Parameter n, % Observed Population (N=79) Total time on gefitinib, prior to and during ICAP Median duration, y, range 11.1 (6.5-15.1) Prior to ICAP Median duration, y, range 7.8 (5.4-10.9) Starting dose 250 mg/day 67 (84.8) No dose changes due to AEs 75 (94.9) During ICAP Median duration, y, range 3.5 (0.04-4.7) Dose: 250 mg/day 76 (96.2) Treatment-related AEs Grade 1-2 Grade ≥3 Grade unknown 13 (16.5) 1 (1.3) 2 (2.5) Dose reductions due to treatment-related AEs 1 (1.3) Discontinuations due to treatment-related AEs 4 (5.1) Discontinuations due to progressive disease 11 (28.9)

Conclusion:
The majority of this subset of patients who participated in ICAP based on long-term clinical benefit from gefitinib continue to do well with gefitinib, demonstrating good tolerance of therapy and survival for a median duration of more than 10 years.

Abstract not provided