Virtual Library

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Background:
Large cell carcinoma (LCC) is an uncommon lung tumor that arises predominantly in smokers and shares many features of solid adenocarcinoma (ADC). 40% of LCC/solid ADC harbor mutations in KRAS; EGFR and ALK alterations are rare in this tumor type. The majority of these tumors, however, lack one of the commonly queried oncogenic driver alterations, thus therapeutic options are limited for patients with this tumor type. Immunomodulatory therapies, including targeting PDL1, have shown promise in a variety of tumor types. Tumor neo-antigens, including those induced by smoking, are associated with mutational burden and may predict susceptibility to cytolytic immune response; in addition, high PDL1 expression in non small cell lung carcinoma has been associated with response to anti-PDL1 drugs. Given the high prevalence of smoking in patients with LCC and solid ADC, we hypothesize that these tumors may be amenable to immunomodulatory therapy and sought to define the frequency of PDL1 expression in tumors lacking an oncogenic driver mutation.

Methods:
This study was restricted to 27 LCC and solid ADC known to be negative for KRAS, EGFR, ALK and ROS1 alterations. Hybrid capture targeted next generation sequencing (NGS) on an Illumina HiSeq 2500 was performed using a cancer genomic assay to detect mutations, copy number variations (CNVs) and structural variants. The assay captures exonic sequences of 275 cancer genes and 91 introns across 30 genes for rearrangement detection. Findings were compared to an institutional cohort of 732 consecutive lung tumors sequenced on the same platform. Immunohistochemistry for PDL1 was performed using a rabbit monoclonal antibody (Cell Signaling Technologies) at 1:100 dilution following pretreatment with citrate buffer/pressure cooker and detected using the Envision + polymer system (DAKO). Immunostaining was considered positive in the tumor component or the inflammatory component if ≥5% of the cells showed membranous staining.

Results:
Of the 27 tumors tested, 26 were resected from smokers. NGS revealed an average of 14.9 mutations per case for LCC/solid ADC cohort versus 8.1 mutations in the overall cohort of lung tumors (p<0.0001). 11 cases (41%) were positive for PD-L1. 7 cases (26%) showed strong, diffuse staining (≥70% of cells) for PD-L1. The inflammatory component was positive for PD-L1 in 25 cases (93%). Two cases with strong expression of PD-L1 by immunohistochemistry (>90% of cells) showed focal amplification of CD274 by NGS.

Conclusion:
LCC and solid ADC are strongly associated with a smoking history and harbor a significantly higher average mutational burden than other lung tumors. 41% of LCC/solid ADC are positive for PDL1 by immunohistochemistry with 26% showing very strong PDL1 expression and nearly all cases showing some degree of positivity in the associated inflammatory infiltrate. In some cases, high PDL1 expression is associated with focal amplification of CD274, the gene encoding PDL1. These findings suggest that LCC/ solid ADC is likely to have smoking-associated neo-antigen expression and that PDL1-directed immunotherapies may be a promising therapeutic approach in this otherwise poorly-characterized lung tumor.

Background:
Inhibition of PD-L1 can lead to reactivation of tumor immunity and assist in cancer therapy. PD-L1 expression in tumor cells has been reported to correlate with clinicopathological parameters and prognosis in a variety of cancers including lung adenocarcinomas (ADC). However, it has not been well studied whether PD-L1 expression is altered along with tumor progression. In addition, little is known about the role of PD-L1 expression in predicting response to chemotherapy in ADC. Thus, we sought to compare PD-L1 expression in the main tumor and lymph node metastases of stage II and III ADC, and correlate PD-L1 expression with survival in patients who underwent adjuvant chemotherapy.

Methods:
The study cohort consisted of 109 ADC who underwent curative resection without neoadjuvant therapy and were diagnosed to have stage II or III disease. Of those, 60 cases received platinum-based adjuvant therapy and were followed at our institution. Immunohistochemistry for PD-L1 (E1L3N, 1:200, CST) was performed on sections of the primary tumor and/or metastatic lymph nodes and the primary tumor sections were also stained with CD8 (4B11, RTU, Leica Bond). Membranous staining of any intensity present in 5% or more of the tumor cells was deemed positive for PD-L1 expression. CD8+ tumor infiltrating lymphocytes (TILs) were evaluated using a 4-tier grading system (0-3). The PD-L1 expression in the primary tumor was correlated with that in lymph node metastases as well as clinicopathological parameters, including CD8+ TILs, and recurrence free survival (RFS).

Results:
Of the 109 cases, 53 (48.6 %) exhibited PD-L1 expression in the primary tumor, which was significantly associated with smaller tumor size, lower pT stage, nodal disease, solid-predominant pattern, the presence of tumor islands, necrosis and lymphovascular invasion, and increased CD8+ TILs (grade 2-3). Upon multivariate analysis, only increased CD8+ TILs remained significant (p=0.039). As for the primary – lymph node correlation, PD-L1 expression was seen in 57.6% of 59 N1 nodes, 53.1% of 32 N2 nodes, and 100% of one N3 node available for evaluation. The PD-L1 expression status was the same between the primary tumor and nodal metastases in the majority (76.3 % of N1 nodes, and 75.0% of N2; p<0.001 and p=0.005, respectively), and the upregulation of PD-L1 expression (positive expression was present in nodal metastasis with negative primary) was seen in only small fractions of the cohort (6.8% of N1 nodes and 9.3% of L2 nodes). Interestingly, PD-L1 expression in the primary tumor was associated with longer RFS in patients who underwent platinum-based adjuvant therapy (mean 84 months vs. 41 months in PD-L1 negative patients, p=0.016), but not in those without adjuvant therapy.

Conclusion:
PD-L1 expression in the primary tumor was associated with prominent CD8+ TILs as well as several adverse clinicopathological parameters including nodal disease, but PD-L1 expression in the nodal metastasis was similar to that in the primary tumor in the majority of cases. Although the evaluation was limited due to a small size of the cohort, PD-L1 expression in the primary tumor appears to be predictive of response to platinum-based adjuvant therapy.

Background:
Lung cancer is the leading cause of cancer death worldwide and most of the patients are diagnosed with advanced disease. Inflammatory components play a key role in tumor progression and survival. Neutrophils are increased in blood of patients with lung cancer and they are associated with poor clinical outcomes. CD47 is a protein which control cell communication, apoptosis, adhesion and proliferation and it has been found increased in cancer and related with phagocytosis evasion mechanism.The aim of this study was to evaluate CD47 expression levels on peripheral neutrophils, also assess the phenotype, apoptosis, activation state, reactive oxygen species production of neutrophils between patients with Non-Small Cell Lung Cancer (NSCLC) and healthy subjects.

Methods:
Fifty NSCLC patients (stage IIIB and IV) naive to treatment and 25 healthy subjects were analized for: CD47 peripheral blood expression, neutrophils phenotype and activation state, evaluation of apoptosis and phagocytosis by flow cytometry. Reactive oxygen species (ROS) production by circulating neutrophils upon stimulation with PMA was assessed by flow cytometry. For the phagocytosis assay, PMNC were labeled with CMFDA and were cultured in RPMI for 24 hrs. To obtain apoptotic target cells, 24h PMNC were labeled with Annexin-V. For the evaluation of phagocytosis, the neutrophils from NSCLC patients were co-cultured with THP-1 cells. The percentage of phagocytosis was assessed by flow cytometry.

Results:
Our results showed a lower percentage of total CD47 in peripheral blood cells in NSCLC patients compared to controls [P=0.042]. Mean Fluorescence Intensity (MFI) of CD47 was higher in patients [P<0.001]. The percentage of CD66b+ cells characterized as neutrophils was higher in patients as well as their MFI of CD47 [P< 0.001]. MFI of CD66b was higher in patients [P< 0.0178]. This would be related with a more activated state. We found that a higher disease stage (IIIB vs. IV) associated with a higher MFI of CD47 [P=0.020]. Plasma pro-inflammatory cytokines, was increased in patients compare to controls IL-6 (P<0.002), IL-8 (P<0.001), IL-12p70 (P<0.008), TNF (0.010) and IFN-g (P<0.001). MFI of CD47 >1635.5 was associated with a higher median Overall survival (P= 0.007). We found a decrease of AnnexinV+/7AAD+ in neutrophils of patients [P=0.0317]. Caspases 3 and 7 were found decreased in neutrophils of patients [P= 0.049]. Oxygen species (ROS) production of neutrophils upon PMA stimulation was increased in patients [P=0.029], suggesting it might play a role in immune effector function. Phagocytosis of apoptotic neutrophils by differentiated THP-1 cells was decrease in cancer patients cells (P=0.0445). Mean fluorescence Intensity of CD47 was increased after 24 hrs in patients [P=0.0408]. This result suggests that neutrophils from patients avoid being engulfed and this may be associated with overexpression of CD47.

Conclusion:
Taken together, these findings suggest that these are altered mechanisms by which neutrophils evade anti-tumor immune response and their increased expression of CD47 is a potential therapeutic target for NSCLC.

Background:
Recent evidence suggests that PD-L1 expression can be induced with radiotherapy and may be a mechanism for resistance to radiotherapy and immunotherapy. Sequentially assessing PD-L1 expression on cancer associated cells in circulation during treatment regimens may be a way to assess the efficacy of radiotherapy and immunotherapy in clinical trials. For this feasibility study, we evaluated the association of RAD50 induction, and PD-L1 expression, on CTCs and Cancer Associated Macrophage-Like Cells (CAMLs) in lung cancer patients (pts) before and during radiotherapy to determine expression changes of these markers.

Methods:
Eleven pts with stage I-IV lung cancer were included in this pilot study. Three pts received Stereotactic Body Radiation Therapy (SBRT) for stage I disease and 8 other pts received chemoradiation for stage II-IV disease. Baseline blood samples (7.5 ml) were drawn prior to the start of radiotherapy (T0) and a second blood sample was drawn at a follow up visit during radiotherapy; or for three pts, after completing SBRT (T1); for a total of 22 samples. Blood was processed using CellSieve™ microfiltration (Creatv Microtech), stained for cytokeratin 8, 18 & 19 and CD45, and imaged. Using the QUAS-R (Quench, Underivatize, Amine-Strip and Restain) technique to remove fluoresce signal, all cells were restained for RAD50-AlexaFluor550 and PD-L1-AlexaFluor 488, along with DAPI nuclear stain. The RAD50 foci numbers within nuclear regions were quantified. PD-L1 pixel intensity was measured by the ZenBlue software and grouped into 4 IHC groups: 0-negative (pixel average 0-215), 1-low (pixel average 216-300), 2-medium (pixel average 301-750), and 3-high (pixel average 751+).

Results:
There was at least one cytokeratin positive cell (i.e. CTC or CAMLs) found in each of the samples. Specifically CTCs were found in 82% of T0 and 64% of T1 samples, and CAMLs were found in 91% of T0 and 100% of T1 samples. RAD50 foci ranged from 0-16 per cell, with an average of 0.69 at T0 that increased to 3.46 at T1 (p=0.002) during radiotherapy. Distinctively, there were 6 pts with greater than 2 fold RAD50 foci increase at T1 and 5 pts with ≤ 2 fold induction. PD-L1 expression ranged from 34-2004 pixel intensity, with an average of 170 at T0 and 336 at T1 (p=0.08). Interestingly, 4 pts had no PD-L1 expression at T0 but an increase to 2 to 3+ at T1, 4 pts with low/no PD-L1 expression remained low at T1, and 3 pts had high PD-L1 expression that remained high or decreased at T1. There was no correlation between RAD50 induction and PD-L1 expression.

Conclusion:
Both RAD50 foci and PD-L1 expression were quantifiable in both CTCs and CAMLs, and had variable responses to radiotherapy +/- chemotherapy. These data suggest that sequential tracking of CTCs or immune-related cells from the primary lung tumor is feasible using microfiltration and potentially can serve as predictive biomarkers for cancer therapy.

Abstract not provided

Background:
Preclinical modeling of immunotherapeutics in PDX-bearing mice has been limited by the absence of a relevant immune microenvironment, as a highly immunosuppressive environment is often required for the implanted tumor to grow. Checkpoint inhibitors including anti-PD-1 and anti-PD-L1 antibodies (mAbs) are promising new treatments in non-small cell lung cancer (NSCLC). The creation of a PDX model system that supports human tumor growth and recapitulates the relevant genomics in NSCLC while providing the immune microenvironment necessary for anti-PD-1 and anti-PD-L1 mAb activity is critical for validation of combination checkpoint inhibitor strategies in NSCLC.

Methods:
Hematopoietic CD34+ progenitor stem cells (CD34+ HPC) were engrafted into the tail vein of sublethally irradiated NSG mice. A KRAS G12D PDX was assayed for PD-L1 expression by FACS (Biolegend; clone 29E. 2A3, San Diego CA) and implanted into Hu-CD34 NSG mice with > 25% Hu-CD45+ cells 12 weeks post CD34+ HPC injection. Multilineage engraftment of immune cell subsets was assayed in peripheral blood, spleen and tumor by FACS (CD45, CD3, CD4, CD8, CD19). PDX were treated with vehicle Q5D x 6, pembrolizumab (Merck, Whitehorse Station PA) 5 mg/kg Q5D x 6, and combination pembrolizumab and docetaxel (Hospira, Lake Forest) 10 mg/kg Q7D x4 at the same single agent dosages. Body weight and tumor growth were assessed twice weekly.

Results:
Hu-CD45+ cells were detected in peripheral blood, spleen and tumor by flow cytometry on single cell suspension. The majority of Hu-CD45+ cells were T-cells: CD3CD4+ (mean blood 50%, spleen 53%, tumor 52%) and CD3CD8+ (mean blood 14%, spleen 15%, tumor 39%). KRAS G12D tumor had 89% surface expression of PD-L1. No significant change in Hu-CD45+ cell composition was noted between the different treatment groups. Pembrolizumab both alone and in combination with docetaxel showed activity in KRAS G12D PDX with substantial tumor growth inhibition and decreased mean tumor volume at day 24 post-treatment.

Conclusion:
Multilineage engraftment of relevant immune cell subsets for PD-1 inhibition is present in the humanized immune-mouse (Hu-CD34 NSG). PD-1 inhibition in a KRAS G12D Hu-CD34 NSG with high PD-L1 expression demonstrated substantial tumor growth inhibition both alone and in combination with chemotherapy. Additional studies are underway exploiting the Hu-CD34 NSG mouse model for study of anti-PD-1/PD-L1 therapies in KRAS mutant and other important molecular subsets of NSCLC.

Background:
The WHO estimated that 1.6 million people died of lung cancer in 2012. Nivolumab, an anti-PD-1 immune checkpoint inhibitor, was FDA approved on March 4, 2015 for platinum-refractory, metastatic, squamous-cell NSCLC, based upon a RR to single agent nivolumab of ~15% and improved OS. Combinatorial strategies may enhance these outcomes. Increased tumor expression of T cell chemokines, such as CCL5 and CXCL10, is associated with a better response to immunotherapy. Furthermore, expression of T cell chemokines is strongly and positively associated with increased T cell infiltration and improved patient survival. Therefore, enhancement of expression of T cell chemokines may augment response to PD-1 blockade immunotherapy.

Methods:
FDA-approved oncology agents were utilized from the Approved Oncology Drugs Set (97 agents) from the Developmental Therapeutics program of NCI. LKR cells were plated in 96-well plates, and a viability assay was performed 48 hours after drug administration (Cell Counting Kit-8, Dojindo Laboratories). Mice were bred and housed in the animal facility at Moffitt Cancer Center. Cells were harvested in logarithmic growth phase after being cultured for less than 2 weeks. 1x10[6] LKR or 344SQ cells were injected s.c. and tumors were monitored for growth by measurements 2-3 times per week. Romidepsin was injected i.p. (2mg/kg) on days 14,16, and 18 after tumor inoculation. Anti-PD-1 was injected i.p. (300μg/mouse) on days 15, 17, and 19 after tumor inoculation. Relative tumor size between treatment groups was analyzed using the t test with Welch’s correction.

Results:
Histone deacetylase inhibitors (HDACi), including vorinostat, emerged as the only class of agents in a 97-drug screen capable of inducing expression of multiple T cell chemokines, including CCL5, CXCL9, and CXCL10, in mouse and human lung cancer cell lines and primary tumors. HDACi’s ability to induce T cell chemokine expression was dependent on both JAK-STAT and NF-kB pathways. HDACi (romidepsin) treatment of mice bearing LKR tumors did not substantially cause tumor shrinkage but significantly reduced growth (p<0.0001; final tumor volume). This effect of HDACi was completely T cell dependent. LKR tumor cells had low cell surface expression of PD-L1 but which was substantially increased by IFN-g. PD-1 blockade with mAb reduced tumor growth but rarely induced rejection. However, when PD-1 blockade was combined with HDACi, 9 out of 11 mice demonstrated complete tumor rejection. HDACi anti-tumor response correlated with T cell chemokine induction in tumors and greater presence of tumor-infiltrating lymphocytes (TILs). We next used a mouse tumor model (344SQ) that was relatively resistant to anti-PD-1 treatment. PD-1 blockade combined with HDACi significantly reduced growth of these tumors compared to untreated (p=0.0003), anti-PD-1 alone (p=0.01), or HDACi (p=0.004) alone treated mice.

Conclusion:
HDACi not only enhanced anti-tumor response against PD-1 blockade sensitive tumors (LKR), but also induced response against PD-1 blockade resistant tumors (344SQ). HDACi induces JAK-STAT and NF-kB dependent chemokine expression and may induce tumor-infiltrating lymphocytes. Thus, a Phase I/randomized Phase II clinical trial of vorinostat, an orally active, small molecule HDACi, plus pembrolizumab, an anti-PD-1 humanized monoclonal IgG4-kappa antibody, is planned in patients with immune therapy naïve and pre-treated metastatic NSCLC.

Background:
Myeloid cell infiltration of the tumor microenvironment is associated with decreased overall survival in multiple tumor types, including lung cancer. This myeloid cell infiltration represents a tissue component of the heterogeneous group of cells termed myeloid derived suppressor cells (MDSC), which can inhibit the endogenous anti-tumor response, direct angiogenesis, and promote tumor progression. In several studies of patients with non-small cell lung cancer (NSCLC), there is wide variation in the presence of myeloid cells in the tumor with increasing levels peripheral blood MDSC associated with poor survival. We have previously shown that CD14+ monocytes can be converted by the tumor microenvironment to an immune suppressive phenotype in non-Hodgkin lymphoma and glioblastoma. In this work, we expand on our earlier observations to include recruitment of CD14+ cells by lung cancer cell lines and their conversion to an immune suppressive phenotype. While most models of myeloid cells in the microenvironment describe the effects of these cells on non-malignant systems, we show that myeloid cells may have profound direct effects on the tumor.

Methods:
Human lung cancer cell lines were cultured and supernatants collected for ELISA. CD14+ cells were isolated from the peripheral blood of healthy volunteers using anti-CD14 immunomagnetic beads. Lung cancer cell lines and CD14+ cells were cocultured under a variety of low serum conditions with or without cisplatin. Changes in CD14+ cell HLA-DR expression and tumor cell survival were measured by flow cytometry. CD14+ cell migration through a permeable transwell membrane was measured in real time with live cell imaging.

Results:
Under normal culture conditions, 7 of 8 human lung cancer cell lines secreted detectable levels of CCL2, a major chemoattractant for monocytes, ranging from 30 to 10,000 pg/ml of CCL2 found in culture supernatants. CD14+ cells more robustly migrated towards cell lines with higher production of CCL2. The coculture system showed a differential impact on monocytes by the tested lung cancer cell lines which either reliably upregulated or downregulated CD14+ cell expression of HLA-DR. In 3 of 8 lung cancer cell lines, CD14+ cell HLA-DR was downregulated in a manner expected to promote local immune suppression. Under serum starvation conditions, one lung cancer cell line showed improved survival when cocultured with CD14+ cells. Similarly, coculture with CD14+ cells enhanced tumor survival of two cell lines after exposure to cisplatin.

Conclusion:
The studied lung cancer cell lines differ in the degree of CD14+ cell recruitment, CD14+ cell HLA-DR expression after coculture, and level of conferred survival benefit under stressful conditions. Taken together these results suggest that the variable myeloid involvement in lung cancer patients can be modeled using lung cancer lines. In addition, we have identified that for some tumors, monocytes confer a significant survival advantage that is not associated with immune or angiogenic responses. Future work is needed to explore the impact of CD14+ cells on lung tumor invasiveness, angiogenesis, and the mechanisms underlying these pro-tumor effects.

Background:
Immune checkpoint pathways including the PD-1/PD-L1 pathway are involved in tumor evasion from the immune system. Elevated PD-L1 expression in tumor cells inhibits tumor-infiltrating T cell function and may be associated with poor prognosis in lung cancer patients. There is increasing interest in developing immunotherapies that block the immunosuppressive effects of checkpoint pathways such as PD-L1, and identifying patients who may benefit from PD-L1 blockade. Activating KRAS mutations are common driver mutations in non-small cell lung carcinoma. Patients with mutated KRAS demonstrate less benefit from adjuvant chemotherapy and resistance to tyrosine kinase inhibitors. The effect of cancer cell driver mutations on immune checkpoint immune regulation is poorly understood. While recent clinical trials have suggested better response to PD-1 blockade in KRAS mutation subjects, it is unclear if this clinical finding is directly driven by KRAS regulating the PD-1/PD-L1 pathway with resultant improved efficacy to anti-PD-L1 immunotherapy or if the presence of a KRAS mutation is merely a surrogate marker of the overall mutational load and tumor immunogenicity. KRAS mutations are known to activate the RAF-MEK-ERK pathway. We hypothesize that KRAS mutation directly regulates the PD-1/PD-L1 pathway through ERK activation.

Methods:
Immortalized human bronchial epithelial cells (HBEC-vector control), KRAS–mutated (KRAS[v12]) HBEC cells (HBEC-KRAS), p53 knockdown HBEC cells (HBEC-p53), and p53 knockdown/KRAS mutated cells (HBEC-p53/KRAS) were used to assess mRNA and/or surface protein expression levels of immune checkpoints including Lag-3, Tim-3, PD-L1 and PD-L2 by real time-qPCR (RT-qPCR) and flow cytometry, respectively. HBEC-vector and HBEC-KRAS cells were treated with MEK (ERK kinase) inhibitor (PD0325901) at 1µM for 24hrs and evaluated for mRNA and surface protein expression of PD-L1. The premalignant HBEC cell lines were used instead of human lung cancer cell lines in order to assess the role of KRAS mutation in isolation without other mutations.

Results:
PD-L1 and PD-L2 mRNA levels increased 2.4 fold (p<0.001) and 3.6 (p<0.001) fold in comparing HBEC-KRAS to HBEC-vector (wild-type) cells, while Lag-3 and Tim-3 mRNA expression levels were unchanged. Based on mean fluorescence intensity on flow cytometry, cell surface PD-L1 protein expression level was 2.2 and 1.6 fold higher in HBEC-KRAS and HBEC-p53/KRAS, respectively, compared to HBEC-vector cells. There was no increase in surface PD-L1 expression in HBEC-p53 cells compared to HBEC-vector control, suggesting that p53 mutation did not alter PD-L1 expression in HBEC-p53/KRAS cells. With MEK inhibition, PD-L1 mRNA levels decreased 10 and 11 fold in HBEC-vector and HBEC-KRAS cells, respectively. Analogously, PD-L1 surface protein levels were reduced 2.7 fold in HBEC-vector and HBEC-KRAS cells, respectively. These findings suggest that ERK activation mediates intrinsic expression and KRAS mutation mediates over-expression of PD-L1 mRNA and protein.

Conclusion:
Here, we demonstrate that PD-L1 expression is elevated in premalignant KRAS mutated human bronchial epithelial cells, and ERK activation mediates constitutive and KRAS mutation driven up-regulation of PD-L1 in these cells. Our findings suggest that KRAS mutation may directly regulate the PD-1/PD-L1 immune checkpoint pathway. Further understanding of KRAS driven molecular pathways that modulate immune checkpoints may elucidate therapeutic targets for potential combinational drugs to PD-L1 inhibition.

Abstract not provided

Background:
As one of the novel therapy strategies, PD-L1 has been shown the function of down-regulating T-cell activation through receptor PD-1. Moreover, prognosis of cancer patients are based not only on tumor-related factors but also on host-related factors, particularly systemic inflammatory response. As significant indicators of patients’ inflammation status, circulating monocyte count, neutrophil ratio and lymphocyte ratio were proved as predictors of prognosis in various cancers. Squamous non-small cell lung cancer (NSCLC) revealed to be divergent clinical and molecular phenotypes compared with non-squamous NSCLC. Significantly, combining application of appropriate biomarkers in prognosis prediction is emerging its high importance in cancer research.

Methods:
Chart review was performed on 1286 consecutive patients, 156 of these patients were enrolled in the final analysis. Patients with squamous NCSLC were randomly assigned (2:1) centrally by computer into training group and validation group. Monocyte ratio, Neutrophils to Lymphocytes Ratio, PD-L1 immunostaining score and PD-1-positive stained tumor-infiltrating lymphocytes counts were assessed by Fisher’s linear discriminant analysis to discriminate if OS would exceeding 5 years. The final model was used to calculate the discriminant score in each study participant. And this prediction model was validated in a second set of squamous NCSLC patients. We internally validated the model using a cross-validation procedure.

Results:
4 independent predictors of OS were identified by using FLDA with stepwise variant-selection. The clinical classifying model was described by the following equation: Y = −1.212 + 0.211 × NLR ratio + 0.437 × monocyte ratio - 0.390 × PD-L1 + 0.035 × PD-1 (eigenvalue 0.673, canonical correlation 0.634, P < 0.001). In this equation, PD-L1 represented PD-L1 immunostaining score; and PD-1 represented PD-1 positive TILs counts. For the training set of 104 leave-one-out-cross-validated cases, 27 of 29 OS > 5 years (93.1% sensitivity) and 61 of 75 OS <= 5 years (81.3% specificity) were correctly classified with an overall accuracy of 84.6% (88 of 104) and an AUC of 0.938 [P < 0.001, 95% confidence interval (CI) 0.864–1] Next, the predicting model consisting of the 4 predictors (NLR ratio, monocyte ratio, PD-L1 and PD-1) were applied to the validation set of 52 patients (14 OS > 5 years and 38 OS <= 5 years). A survival prediction for 38 of the 52 patients (73.1%) with an AUC of 0.908 (P < 0.001, 95% CI 0.806–1) was achieved. Also, 12 of 14 OS > 5 years (85.7% sensitivity) and 26 of 38 OS <= 5 years (68.4% specificity) were correctly identified.

Conclusion:
The analysis of a set of immunological markers could effectively and reproducibly classify patients with squamous NCSLC according to their overall survival. Further prospective validation in larger independent cohorts of patients with similar or different regimens is warranted to fully assess its predictive power. The 4-immunological-marker model offers a novel tool for survival prediction and could have important clinical implications for the consideration of differential treatment strategies in patients with squamous NCSLC, thus providing a framework for future individualized therapy.

Background:
Inducible nitric oxide synthase (iNOS) and reactive nitrosylation are important mediators of tumor immunosuppression by myeloid-derived suppressor cells (MDSCs). However, the role of CD33+ peritumoral PMN-MDSCs in these pathways remains unclear. We conducted a retrospective cohort study of NSCLC subjects treated with surgery, with the primary objective to determine the association of MDSC biomarkers with time to progression (TTP) and overall survival (OS).

Methods:
Inclusion criteria: Surgically treated NSCLC of all stages at a single institution between 1996 and 2010. Somatic mutations tested by PCR. Anti-human antibodies optimized for immunohistochemistry. Samples scored by blinded pathologist based on intensity and percentage of peritumoral cells. Peritumoral nitrotyrosine (NT) and iNOS used Allred scoring. Time to progression (TTP) defined as time from resection to progression event or censored at last evaluation. Overall survival (OS) defined as time from resection to death.

Results:
Of 458 tumor samples, 366 lung primaries, 38 soft tissue metastases, and 39 brain metastases. Demographics: median age 67 yrs, 54% female, 96% Caucasian. Of 151 tested for somatic mutations, 36% KRASm, 8.6% EGFRm, 25% p53m, respectively. Histology: adenocarcinoma 76%, squamous 10%. Higher % CD3+ tumor infiltrating lymphocytes (TILs) and CD33+ myeloid cells were observed in tumors than normal tissue (p < .0001 and p = .002, respectively). More CD3+ TILs observed in soft tissue metastases than primary lung tumors (p < .0001). No difference in iNOS expression between tumor and normal lung tissue. More CD3+ TIL was observed in p53 mutant tumors (p=.03). iNOS was positively correlated with CD3+ TIL (p < .001) and CD73+ epithelial cells (p <.001), but not CD33+ myeloid cells. NT expression correlated with the absence of CD3+ TIL (p = .02), consistent with its putative immunosuppressive activity. Median TTP: 10.4 months; 320 (69.7%) events. Median OS: 35.4 months; 353 (77.1%) events. Expectedly, presence of CD3+ TIL was associated with favorable OS; HR 0.5 [0.4 – 0.7], p < .0001, and TTP; HR 0.7 [.5 – .9], p =.009. CD33+ myeloid cells were associated with favorable OS; HR 0.6 [0.5-0.8], p = .0002. Presence of peritumoral iNOS trended toward favorable OS; HR 0.81 [0.6-1.0], p = .07. Peritumoral iNOS was not associated with TTP. Figure 1



Conclusion:
Increased presence of TILs in p53 mutant tumors has been reported in other cancers, and may be related to somatic mutational load. An inflamed tumor phenotype was associated with improved overall survival. Unexpectedly, iNOS was positively correlated with both CD3+ infiltration and overall survival.

Background:
Cancer development and biology is influenced by the host immune system. Emerging data indicate that the context of immune cell infiltrates within the tumor is associated with cancer prognosis. Both the activation state and type of immune cells present can provide mediators that either promote or inhibit tumor growth. While the presence of activated cytotoxic T lymphocytes (CTL) may correlate with better patient survival, the presence of tumor associated macrophages and effector lymphocytes that lack cytotoxic properties may promote tumor growth. Thus, in established tumors, a balance between pro and anti-tumor mediators are present, but as advanced tumors rarely regress without therapeutic intervention, this balance is likely skewed towards tumor-promoting inflammation. In attempts to gain insight into the immune networks that regulate tumorigenesis, we used genome wide gene expression datasets of primary lung cancer tissues to identify and functionally validate immune related genes that are associated with patient survival.

Methods:
Gene expression analysis was conducted on microarray datasets from 128 early-stage NSCLC resected tumor samples. Limiting analysis to immune-related gene sets curated by NIAID ImmPortal, we identified a minimum gene set using MAximizing R Square Algorithm (MARSA) that selected for the greatest separation between good and poor prognostic patient subgroups. The prognostic value of this gene signature was validated in nine additional independently published microarray datasets of NSCLC. From the gene signature, we functionally characterized the potential role of the soluble protein LAIR2 in immune regulation of lung cancer.

Results:
We identify a 9-gene signature that separate patients into high and low-risk subgroups for 5-year cancer-free survival (hazard ratio 10.26, 95% confidence interval 4.32-24.34, p<0.0001). The prognostic accuracy of this signature was validated in additional NSCLC datasets (total 1095 patients without adjuvant treatment). Amongst the 9-genes, the gene encoding the soluble protein LAIR2 was highly expressed within the high-risk patient subgroup. Functionally, we found that addition of recombinant LAIR2 resulted in increased NK cell expression of cytotoxicity receptors and secretion of pro-inflammatory cytokines in the presence of lung cancer cell lines.

Conclusion:
By limiting gene expression analysis to immune related genes, we identify a 9-gene prognostic immune signature in resected early stage NSCLC patients. The signature identifies a role for the soluble protein LAIR2 in the modulation of immune cell activation during lung cancer development and may suggest that LAIR2 induce a pro-inflammatory microenvironment which promote tumorigenesis and poor patient outcome.

Abstract not provided

Background:
Treatment of epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) is the standard therapy for advanced non-small cell lung cancer (NSCLC) with common EGFR mutations such as exon 19 deletions or L858R mutations. However, the efficacy of EGFR-TKIs in patients with uncommon EGFR mutations remains unclear.

Methods:
We have retrospectively surveyed a consecutive database of NSCLC patients with EGFR mutations at Kinki-chuo Chest Medical Center, Osaka Prefectural Medical Center for Respiratory and Allergic Diseases. We analyzed the collect data of NSCLC patients with uncommon mutations including single or complex (uncommon plus uncommon mutations, or uncommon plus common mutations) mutations, treated with gefitinib or erlotinib between July 2007 and September 2014.

Results:
Forty-one patients who had any EGFR uncommon mutations were analyzed in this study. By the Response Evaluation Criteria in Solid Tumors criteria, the overall response rate (RR) was 22.0% with 9 partial response (PR) in all patients with uncommon mutations. Among uncommon single mutations, RR was 12.5% with 3 PR in patients with G719X mutation and 33.3% with 2 PR in patients with L861Q mutation. As for complex mutations, there were no patients in PR with uncommon plus uncommon mutations but RR was 50.0% with 4 PR in patients with uncommon plus common mutations. Median progression-free survival (PFS) was 3.5 months in all patients with uncommon mutations. Among uncommon single mutations, PFS in patients with G719X (median PFS: 1.8 months) was shorter than PFS in patients with L861Q mutation (median PFS: 7.6 months). Furthermore, there was a difference in the efficacy of EGFR-TKIs among patients with each G719X mutation (median PFS in G719A: 8.2 months, median PFS in G719C: 1.1 months, median PFS in G719S: 1.7 months).Figure 1



Conclusion:
First generation EGFR-TKIs are less effective in NSCLC patients with uncommon mutations than in those with common mutations. However, they had favorable response in patients with L861Q or G719A mutations, compared with G719C or G719S mutations.

Background:
Erlotinib, gefitinib and afatinib, tyrosine kinase inhibitors directed at EGFR signalling (EGFR-TKI), are currently used for the treatment of patients with advanced-stage non-small cell lung cancer (NSCLC). A considerable progress in the field of molecular oncology and cancer genomics in recent years has let to identification of several gene alterations predicting clinical outcome of patients treated with EGFR-TKIs. Activating EGFR mutations are widely recognized predictors of good response to EGFR-TKI treatment. While the predictive role of common EGFR mutations (exon 19 deletions and exon 21 L858R point mutation) is well described, very little clinical evidence data exist on the role of rare EGFR mutation types. The aim of this study was to assess the distribution of common and rare EGFR mutations in patients with NSCLC and to evaluate the efficacy of EGFR-TKIs for patients harboring rare and common EGFR mutations.

Methods:
Clinical data of 305 patients with advanced-stage NSCLC (IIIB or IV) treated with EGFR-TKIs having EGFR mutation positive primary tumors at the time of diagnosis were evaluated in a retrospective setting. The therapy included erlotinib, gefitinib or afatinib as recorded in a Czech national lung cancer registry – TULUNG. The relative frequency and survival data (PFS and OS) were evaluated for individual EGFR mutation types.

Results:
The common activating EGFR mutations (exon 19 deletion and exon 21 L858R point mutation) were found in a total of 249 (81.6%) patients. Rare EGFR mutations were found in 56 (18.4%) patients, the most frequent of which was exon 18 - G719X mutation found in 29 patients (9.5%), followed by mutations in exon 20 found in 28 patients S768I in 3 patients (0.98%) and insertion 3 mutations in 16 patients (5.2%). Patients with exon 19 deletion had median median OS 11.0 months, patients with exon 21 point mutation L858R median OS 9.4 months,respectively. Patients with rare EGFR mutations median OS 12.5 months.Comparing frequent and rare mutations, there were no differences in sex, age, PS, stage of disease and adverse effects of first line gefitinib therapy, the group of patients with rare mutations were more frequently smokers, duration of gefitinib therapy was shorter, response rate and PCR, PFS and OS were worse than in patients having frequent EGFR mutations. There were no significant differences in characteristics, PFS and OS between exon 19 deletion and exon 21 L858R point mutation tumour patients.

Conclusion:
While patients with frequent EGFR sensitive mutations have significant benefit from gefitinib therapy, patients with G719X mutation on exon 18 have marginal PFS and OS benefit, while pagtients with exon 20 insertion mutations have no demonstrable benefit from targeted therapy.Next generation tyrosinkinase inhibitors may prolong survival in some of rare EGFR mutated tumour patients.

Background:
Rociletinib (CO-1686) is a novel, oral, irreversible mutant selective tyrosine kinase inhibitor for the treatment of patients with mutant epidermal growth factor receptor (EGFR) non-small cell lung cancer (NSCLC). Rociletinib has demonstrated efficacy against activating mutations (L858R and Del19) and the dominant acquired resistance mutation (T790M), while sparing wild-type EGFR. A maximum tolerated dose was not identified in Phase 1 with 1000 mg BID the highest dose studied. Here we assess the efficacy and safety of the three doses of rociletinib (500 mg BID, 625 mg BID and 750 mg BID) selected for Phase 2 study.

Methods:
TIGER-X (NCT01526928) is a Phase 1/2 open-label, safety, pharmacokinetics and preliminary efficacy study of rociletinib in patients with advanced EGFR mutant NSCLC progressing after ≥1 EGFR tyrosine kinase inhibitor (TKI). Efficacy is assessed using RECIST. Safety is evaluated using standard adverse event (AE) reporting.

Results:
As of April 2015, a total of 231 central T790M positive patients were evaluable for efficacy and 343 for safety (any T790M). All patients were enrolled in the USA (85%), Europe (9%) and Australia (6%). Baseline characteristics were similar in each dose group. The median number of prior therapies was 2. 85% had EGFR TKI as their most recent prior therapy and 10% had a history of diabetes/hyperglycemia. Immature ORRs are 53% (500 mg BID), 52% (625 mg BID) and 43% (750 mg BID), with disease control rates of 89% (500 mg BID), 87% (625 mg BID) and 82% (750 mg BID). The most common ≥grade 3 treatment-related AE was hyperglycemia [16% (500 mg BID), 25% (625 mg BID) and 35% (750 mg BID)] which was managed with oral hypoglycemic agents. Only one patient discontinued the study for hyperglycemia. Grade 3 QTc prolongation was uncommon, occurring in 2% (500 mg BID), 7% (625 mg BID) and 10% (750 mg BID) of patients, and demonstrated a relationship to dose. There were no clinically relevant cutaneous toxicities with 7 cases of grade 1 rash and 4 cases of grade 1 stomatitis (no dose relationship) and no paronychia.

Conclusion:
All 3 Phase 2 doses of rociletinib are active and well tolerated in a Western patient population with advanced NSCLC. The lack of cutaneous toxicities confirms the selectivity of rociletinib for mutant forms of EGFR and is an important contributor to QOL and maintaining dose intensity (Lacouture et al. 2011). Overall, the adverse event frequency appears to be related to dose, but antitumor activity does not, thus the risk/benefit profile may be optimal at the lowest dose studied.

Background:
Rociletinib (CO-1686) is a novel, oral, irreversible tyrosine kinase inhibitor for the treatment of patients with mutant epidermal growth factor receptor (EGFR) non-small cell lung cancer (NSCLC) with activity against the activating mutations (L858R and Del19) and the dominant acquired resistance mutation (T790M), while sparing wild-type EGFR. TIGER-X (NCT01526928) is a Phase I/II open-label, safety, pharmacokinetics and preliminary efficacy study of rociletinib in patients with advanced EGFR mutation-positive NSCLC with progressive disease after ≥1 EGFR tyrosine kinase inhibitor (TKI). An overall response rate of 67% has previously been reported in this trial for T790M positive patients treated with the 500 and 625 mg BID doses (Soria 2014). Here we provide preliminary data on the activity of rociletinib in the subgroup of patients with a history of CNS disease.

Methods:
Patients with a history of CNS disease were permitted if asymptomatic and stable, as defined by steroid requirements. The primary activity endpoint was RECIST overall response rate. However, patients who developed progressive disease (PD) while on study treatment were allowed to continue therapy with rociletinib if deemed clinically beneficial by the investigator.

Results:
As of 16 March 2015, a total of 401 patients received therapeutic dose levels of rociletinib (500, 625 and 750 mg BID) including 170 (42%) patients with a history of CNS metastases. Based on this interim analysis, the RECIST overall response rate among these patients with a history of CNS disease is 41%. To date, 42 patients with a history of CNS disease have continued therapy with rociletinib post-progression. Of those who continued for at least 14 days the average treatment duration beyond PD was 89 days (range: 14 - 336 days). Twenty-two of the 42 patients with a history of CNS disease with PD also received brain radiation and continued rociletinib treatment for an average of 120 days (range: 22 – 336 days) after PD. Rociletinib is held on radiation days only. Progression-free survival data for these subgroups is not yet mature. The three most common adverse events in the patient population with a history of CNS disease are similar to those found in the general TIGER-X patient population: hyperglycemia, diarrhea and nausea.

Conclusion:
In patients with a history of CNS disease, a factor associated with poor prognosis, rociletinib is active with a RECIST response rate of 41%. Local CNS radiation has been administered safely with rociletinib held on radiation days and continued afterwards. Prolonged use of rociletinib post CNS radiation suggests ongoing systemic benefit is still experienced by these patients. The role of rociletinib in NSCLC patients with CNS involvement is being further explored in the ongoing TIGER clinical development program.

Abstract not provided

Background:
AZD9291 is an oral, potent, irreversible epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI), selective for both EGFR-TKI-sensitizing (EGFRm) and T790M resistance mutations. The Phase I AURA study was a dose escalation/expansion study in patients with EGFRm positive advanced non-small cell lung cancer (NSCLC) who had progressed after EGFR‑TKI treatment. The 80 mg once daily (qd) dose was chosen for further evaluation in a Phase II extension cohort of the AURA study, and in an additional Phase II study (AURA2). Here we report efficacy and safety of AZD9291 from the AURA study Phase II extension cohort (NCT01802632) in patients pre-treated with EGFR-TKI and with centrally confirmed T790M positive advanced NSCLC.

Methods:
Eligible patients had measurable disease, World Health Organization performance status (WHO PS) 0 or 1, and acceptable organ function; stable brain metastases were allowed. A mandatory tumor sample was taken after disease progression on the most recent line of therapy, for prospective confirmation of T790M positive status by central laboratory testing (cobas™ EGFR Mutation Test). Patients received AZD9291 at 80 mg qd until disease progression. The primary endpoint was objective response rate (ORR) according to RECIST 1.1 (assessed by independent central review, ICR). Secondary objectives included disease control rate (DCR), duration of response (DoR), progression-free survival (PFS), investigator-assessed ORR, and safety. Planned enrollment was 175 patients to give an estimate of the ORR with 95% CI within ±8%. Data cut-off was January 9, 2015 after all patients should have undergone the second tumor assessment.

Results:
201 patients were dosed in the extension cohort of the study; two patients without measurable disease at baseline by ICR were excluded from the evaluable-for-response set. By central testing, EGFR mutation subtypes were: T790M, 98%; Ex19del, 71%; L858R, 25%; other, 3%. Median age was 62 years; female, 66%; Asian, 57%; WHO PS 0/1/2, 34%/66%/1%; second/≥third-line, 30%/70%. At the data cut-off, median treatment exposure was 4.9 months and 168 patients remain on treatment. ORR by ICR was 58% (115/199; 95% CI 51, 65) and DCR was 92% (95% CI 87, 95). ORRs were similar across lines of therapy (second-line, 59.0% [36/61] vs ≥third-line, 57.2% [79/138]). Investigator-assessed ORR was 68% (137/201; 95% CI 61, 75). Median DoR and median PFS have not been reached (maturity 2% and 21%, respectively). The most common all-causality adverse events (AEs) were diarrhea, 41% (0.5% Gr≥3) and grouped rash terms 37% (0.5% Gr≥3); 42 (21%) patients experienced Gr≥3 AEs. Interstitial lung disease grouped terms were reported in five (2.5%) patients, one of which was fatal (0.5%) and considered possibly causally related to AZD9291 by the investigator. Eight patients (4%) discontinued treatment due to an AE. Updated results from a later data cut-off will be available for presentation.

Conclusion:
In the AURA study Phase II extension cohort, AZD9291 80 mg qd demonstrates clinical activity, manageable tolerability, and a low discontinuation rate in patients with centrally confirmed EGFR T790M positive advanced NSCLC that has progressed on or after EGFR‑TKI treatment. These data provide further validation of the results from the Phase I study cohorts.

Background:
AZD9291 is an oral, potent, irreversible epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI) selective for both EGFR-sensitizing (EGFRm) and T790M resistance mutations. It has shown anticancer activity and manageable tolerability in patients with EGFRm advanced NSCLC that had progressed after EGFR‑TKI treatment.

Methods:
In this first-line expansion cohort (AURA, NCT01802632), patients received AZD9291 at 80 or 160 mg/day, in sequential dose groups. EGFRm status was determined locally and/or by central testing using the cobas EGFR Mutation Test. Other inclusion criteria included measurable disease, World Health Organization performance status (WHO PS) 0 or 1, and acceptable organ function; stable brain metastases were allowed. Safety, tolerability, and anticancer activity were assessed in these cohorts, to evaluate AZD9291 in the first-line treatment setting. The data cut-off was December 2, 2014.

Results:
Sixty treatment-naïve patients were enrolled; 30 patients in each dose group (80 or 160 mg/day). By central testing, EGFR mutation subtypes were: L858R 40%; exon 19 deletion, 37%; other EGFR-sensitizing mutations, 3%; and T790M, 8%. Baseline median age was 63.5 years; 25% of patients were male; 57%/43% had WHO PS 0/1, respectively; 72% were Asian and 23% Caucasian. Median treatment exposure at the 80 and 160 mg dose levels was 260 and 171 days, respectively. Fifty-two out of 60 patients remained on treatment at the data cut-off. Anticancer activity of AZD9291 is shown in Table 1. One-third (33%) of patients experienced Grade ≥3 adverse events; two patients had Grade 3 diarrhea and one patient had Grade 3 skin rash. New data from a 2015 data cut of the AURA first-line expansion will be available for presentation.

Table 1. Anticancer activity findings in AURA first-line expansion

Endpoint	Finding
Objective response rate:
Overall	70% (95% CI 57, 81)
AZD9291 80 mg/160 mg	60%/80%
Disease control rate:
Overall	97% (95% CI 89, 100)
AZD9291 80 mg/160 mg	93%/100%
Progression-free survival:
Median	Not yet reached
3-month/6-month	93%/87%
Events to date	7/60 (12% mature)

Conclusion:
AZD9291 has a manageable tolerability profile and is associated with promising anticancer activity in treatment-naïve patients with EGFRm advanced NSCLC. A Phase III study (FLAURA, NCT02296125) has been initiated to assess the efficacy and safety of AZD9291 in comparison with a standard-of-care EGFR-TKI (gefitinib or erlotinib) in the first-line setting.

Background:
The epidermal growth factor receptor (EGFR) T790M mutation is found in about half of patients who have developed resistance to EGFR-tyrosine kinase inhibitors (TKIs), gefitinib or erlotinib. AZD9291 is an oral, potent, irreversible EGFR-TKI selective for both EGFR-sensitizing (EGFRm) and T790M resistance mutations. In the Phase I AURA study, AZD9291 80 mg (dose selected for further evaluation) was found to be clinically active, with an acceptable tolerability profile. This ongoing AURA2 Phase II study (NCT02094261) investigates the efficacy and safety of AZD9291 80 mg once daily after previous EGFR-TKI treatment in patients with EGFRm and T790M positive advanced NSCLC.

Methods:
AURA2 (NCT02094261) is a global, open-label, single-arm Phase II study. To be eligible, all patients had a mandatory tumor sample taken after disease progression on the most recent line of therapy, for confirmation of T790M positive status by central laboratory testing using the cobas™ EGFR Mutation Test. Further inclusion criteria included measurable disease, World Health Organization performance status (WHO PS) 0 or 1, and acceptable organ function; stable brain metastases were allowed. Patients receive AZD9291 at 80 mg once daily until disease progression. The primary endpoint was objective response rate (ORR) according to RECIST 1.1 (assessed by independent central review, ICR). Secondary objectives included disease control rate (DCR), duration of response (DoR), progression-free survival (PFS), and safety. Planned enrollment was 175 patients to give an ORR with 95% confidence interval (CI) within ±8%. The data cut-off was January 9, 2015.

Results:
Recruitment is complete and 210 patients were enrolled; 12 patients did not have measurable disease at baseline by ICR and are excluded from the evaluable-for-response set. By central testing, in addition to T790M, patients had background EGFR mutation: Ex19del, 65%; L858R, 32%; other, 3%. Baseline characteristics: median age, 64 years; female, 70%; WHO PS 0/1, 40%/60%; Asian, 63%; second-/≥third-line, 32%/68%. Median treatment exposure was 4.0 months and 183 patients remain on treatment at the data cut-off. ORR by ICR was 64% (127/198; 95% CI 57, 71) and DCR was 90% (95% CI 85, 94). Investigator-assessed ORR was 64% (135/210; 95% CI 57, 71). Median DoR and median PFS have not been reached (maturity 6% and 20%, respectively). The estimated proportion of patients who are alive and progression free is 82% and 70% at 3 and 6 months, respectively. The most common all-causality adverse events (AEs) were diarrhea, 34% (1% Gr≥3) and grouped rash terms 40% (0.5% Gr≥3); 38 (18%) patients experienced Gr≥3 AEs. Interstitial lung disease grouped terms were reported in four (1.9%) patients, one of which was fatal (0.5%) and considered possibly causally related to AZD9291 by the investigator. Eight patients (4%) discontinued treatment due to an AE. Updated results from a later data cut-off will be available for presentation.

Conclusion:
AZD9291 80 mg once daily demonstrates clinical activity and manageable tolerability in patients with EGFRm, T790M mutation positive advanced NSCLC that has progressed on or after EGFR‑TKI treatment. AZD9291 is being investigated in the randomized AURA3 Phase III study (NCT02151981) in comparison with platinum-based doublet chemotherapy.

Background:
Epidermal growth factor receptor (EGFR)-mutant non-small cell lung cancer (NSCLC) exhibits sensitivity to EGFR tyrosine kinase inhibitors (TKIs) such as erlotinib and gefitinib; however, acquired resistance eventually develops in most patients. The most common mechanism of TKI resistance is a second-site mutation in the EGFR kinase domain, T790M. AZD9291 is an oral, potent, irreversible EGFR-TKI with potency against both T790M resistance and sensitizing EGFR mutations. In the ongoing Phase I AURA study (NCT01802632), AZD9291 induced durable responses in patients with acquired resistance to EGFR-TKIs. We report results of paired biopsy cohorts of the AURA trial, reviewing modulation of key molecular biomarkers of AZD9291 activity in patient tumor samples.

Methods:
Two cohorts of patients on the AURA trial were consented for collection of paired tumor biopsies. These patients had a pre-study tumor biopsy with T790M positive tumor status confirmed by central laboratory EGFR testing (Cobas™ EGFR Mutation Test). Following 8 to 15 days of once daily AZD9291 treatment (80 or 160 mg), a post-dose tumor biopsy was obtained. Baseline and post-dose tumor tissue was processed for routine histology and pathologic evaluation. More than 100 viable tumor cells per sample were required for subsequent biomarker scoring. Formalin-fixed paraffin-embedded tumor biopsies were profiled by immunohistochemistry with a suite of key pathway and tumor-relevant markers (phospho[p]-EGFR, pERK, pAKT, pS6, PD-L1, CD8). Matching plasma pharmacokinetic samples were also obtained for PK-PD correlations.

Results:
As of February 2015, 58 potential patients with an evaluable baseline biopsy were identified as candidates for post-dose biopsy collection. Sixteen of these patients did not proceed to an on-study biopsy as the identified lesions had regressed too substantially or were no longer considered suitable for re-biopsy, one patient was medically excluded from re-biopsy, and one patient’s sample was not available. In total, 40 patients supplied matched pre- and on-treatment biopsies. As of March 2015, paired tumor samples were available for QC from 26 of these 40 patients. Ten of these 26 biopsy specimens subsequently failed QC due to inadequate tumor content, leaving 16 paired tumor samples available for biomarker analyses, of which five have thus far been evaluated. AZD9291 treatment resulted in the inhibition of EGFR pathway components in the majority of post-treatment tumor biopsies. Tissue biomarker analyses are ongoing and updated data on evaluable biopsy pairs will be reported at the time of the congress.

Conclusion:
The completion of a paired biopsy cohort within the AURA trial was challenging due to the rapid onset of anti-tumor effects of AZD9291. Approximately 29% (17/58) of potentially eligible patients were unsuitable for the post-dose biopsy procedure due to tumor regression and 38% (10/26) of available post-dose biopsies were found to contain too little tumor for analysis. In the evaluable tumor pairs, pharmacodynamic modulation of the EGFR pathway was evident. Further biomarker analyses, including evidence of modulation of immune system markers, may help inform future combination strategies.

Abstract not provided

Background:
Afatinib, an irreversible tyrosine kinase inhibitor (TKI), has shown clinical benefits and prolonged progression free survival in EGFR mutated patients. HGF, a ligand of c-MET, may be involved in resistance to EGFR-TKIs.

Methods:
A total of 66 patients with advanced lung adenocarcinoma (stage IIB and IV) and documented progression to first-line chemotherapy were enrolled to receive afatinib 40 mg/day. Mutational EGFR and HER-2 status were assessed by RT-PCR. HER2 amplification was evaluated by FISH. Plasma HGF levels were measured by ELISA before and 2 months after the start of treatment with afatinib. We assessed the change in plasma HGF levels and the association with objective response rate (ORR), progression free survival (PFS) and overall survival (OS). The protocol is registered in ClinicalTrials.gov (NCT01542437).

Results:
We identified 2 patients carrying a HER2 mutation and both presented stable disease (SD). HER2 amplification was not detected. HGF-positive plasma reduction status had a significant higher ORR (75.0% vs 44.1% p= 0.011), and was strongly associated with longer PFS (HR 0.40 [95% CI 0.18 - 0.87], p= 0.02) and OS (HR 0.31 [0.13 - 0.71] p=0.006). A stratified multivariate analysis in EGFR mutated patients showed that the HGF plasma levels reduction remains as a significant and independent factor associated with longer PFS (HR 0.34 [95% CI 0.13 - 0.89] p= 0.04) and OS (HR 0.34 [95% CI 0.13 - 0.88] p= 0.02).

Conclusion:
HGF plasma levels reduction is strongly related to better outcomes with afatinib therapy, irrespective of EGFR mutation status. The lack of reduction might allow the identification of a subgroup of patients who will not expected to respond and could benefit with the use of drugs targeting the HGF-c-Met axis. Further studies are warranted.

Background:
The phenomenon of small cell lung cancer(SCLC) transformation in EGFR mutated adenocarcinoma had been previously identified as a resistant mechanism. However, this phenomenon was only reported in single case and a repeat biopsy patient cohort. Moreover, the underlying molecular mechanism remains unclear. Previous study found that the inactivation of TP53 and Rb1 could efficiently transform neuroendocrine and alveolar type 2 cells into SCLC. We inferred that TP53 and Rb1 might also play an important role in SCLC transformation. So we use the sh-RNA mediated depletion of RB1 adenocarcinoma cell line, that also have TP53 inactivation in nature, to investigate the molecular mechanism of SCLC transformation.

Methods:
Both primary and metastatic tissue were analyzed on 3 SCLC transformation patients by whole genome sequencing (WGS). We knock down RB1 in TP53 inactivation cell lines, PC-9, HCC-827 and H1975. Western blot and immunohistochemistry (IHC) were used to confirm RB1 knock down and expression of neuroendocrine (NE) markers. Trans well cell invasion assay and softer agar clone formation test were investigated the change of invasion and migration. CCK8 kit was used to evaluate relative viability of cells after RB1 knock down. Cell cycle and apoptosis were determined by folw cytometry. And we use balb/c mice for cell line tumorgenesis.

Results:
① Pathological analysis of the 3 patients’ primary lesion and the consistent EGFR mutation status confirmed the phenomenon of SCLC transformation. WGS showed the copy number variation of primary tumor and transformed metastasis was distinct. RB1 is lost in 100% of the three transformed cases but occur in one patient’s primary tissue in extremely low frequency(<5%). ② PC-9, HCC827 and H1975 cell line showed up-regulation of NE markers after sh-RNA mediated RB1 depletion, which presented more capable of invasion and migration. Cell folw cytometry showed more cells was in G2 and S phase after RB1 knock down. The expression of Bik and puma that belong to Bcl-2 family was up-regulated after RB1 inactivation compared with the control group.

Conclusion:
The NE differentiation and changes in invasion, migration, apoptosis and cell cycle indicated that the loss of TP53 and RB1 promote the process of SCLC transformation. TP53 and RB1 deficiency may be a necessary event for SCLC transformation to emerge, but is still insufficient to induce SCLC transformation.

Background:
For non-small-cell lung cancer (NSCLC) harboring epidermal growth factor receptor (EGFR) mutations, preclinical data showed the superiority of exon 19 mutations to exon 21 mutations in both response to EGFR tyrosine kinase inhibitors (TKIs) and survival. Meanwhile, retrospective studies demonstrated that erlotinib was significantly superior to gefitinib in progression-free survival (PFS) for advanced NSCLC patients with EGFR mutations. However, no randomized controlled trials compared erlotinib to gefitinib in advanced NSCLC patients with EGFR exon 19 or 21 mutations.

Methods:
We conducted a randomized controlled trial (CTONG 0901；NCT01024413) comparing erlotinib to gefitinib in advanced NSCLC harboring EGFR exon 19 or 21 mutations from July 2009 to July 2014. Eligible patients were randomized to receive erlotinib (150 mg, qd) or gefitinib (250 mg, qd) at the ratio of 1:1 in any line settings. The primary endpoint was PFS, and the secondary endpoints included overall survival (OS), objective response rate (ORR), post-progression survival (PPS), and toxicities.

Results:
The last follow-up was on March 30, 2015. Totally, 256 patients (148 with exon 19 mutations and 108 with exon 21 mutations), of whom 165, 83 and 9 were in the first, second or further-line settings respectively, were randomized to receive erlotinib or gefitinib. Median PFS was 12.4 (95%CI: 10.6~14.1) months in erlotinib arm and 10.4 (95%CI: 8.8~11.9) months in gefitinib arm, HR=0.80 (0.61~1.05), p=0.100; ORR, median PPS and OS were 56.3% versus 52.3% (p=0.530), 6.9 (95%CI: 4.3~9.5) versus 6.9 (95%CI: 4.5~9.2) months (p=0.784), and 22.4 (95%CI: 17.9~27.0) versus 20.5 (95%CI: 17.1~23.8) months (HR=0.90 [0.67~1.22]; p=0.496) respectively. There were no significant differences in toxicities between the two arms, p＞0.05. In the four subgroups (the first-line, second or further-line setting, exon 19 and 21 mutations), except for median PFS being 11.4 versus 7.9 months (HR=0.58 [0.37~0.90], p=0.015) in the second or further-line setting, no significant differencs were observed in median PFS and OS respectively between the two arms, p＞0.05. Receiving erlotinib or gefitinib treatment, EGFR exon 19 mutant patients were superior to those with exon 21 mutations in terms of ORR (62.2% versus 43.5%, p=0.003), median PPS (9.1 [95%CI: 7.0~11.2] versus 4.6 [95%CI: 3.4~5.8] months, p=0.011 ) and OS (24.8 [95%CI: 20.9~28.8] versus 17.7 [95%CI: 15.1~20.3] months, HR=0.66 [0.48~0.89], p=0.006) respectively, even though there was no significantly difference in median PFS (11.4 [95%CI: 9.6~13.2] versus 11.1 [95%CI: 9.4~12.9] months, HR=0.80 [0.60~1.05], p=0.101). Multivariant Cox regression analysis showed that subsequent EGFR TKIs, combination of subsequent EGFR TKIs and local treatment, as well as subsequent chemotherapy were prognostic factors for OS, p＜0.05.

Conclusion:
Erlotinib was not significantly superior to gefitinib in advanced NSCLC with either exon 19 or 21 mutations in response and survival, with similar toxicities. However, EGFR exon 19 mutant patients had remarkably increased ORR, PPS and OS than those with exon 21 mutations after taking erlotinib or gefitinib. Subsequent treatments after failure to EGFR TKIs were significantly prognostic factors for OS.

Background:
Patients with EGFR-mutant lung cancers treated with EGFR tyrosine kinase inhibitors (TKI) develop clinical resistance, often associated with acquisition of EGFR T790M. Upregulation of JAK/STAT signaling is involved in resistance to EGFR TKIs and JAK inhibition is a proposed treatment strategy in the setting of acquired resistance by restoring sensitivity to erlotinib. Ruxolitinib is an FDA-approved oral JAK1/2 inhibitor given at 20mg twice daily for hematologic malignancies with a largely non-overlapping toxicity profile with erlotinib.

Methods:
We evaluated the toxicity and efficacy of once daily oral erlotinib and twice daily oral ruxolitinib in patients with EGFR-mutant lung cancers and acquired resistance to erlotinib therapy (NCT02155465). Using a 3+3 dose escalation, we assessed escalating doses of ruxolitinib (10mg BID, 15mg BID, 20mg BID) with erlotinib 150mg daily for 21 day cycles. Response was evaluated by RECIST 1.1. Tissue and peripheral blood samples were obtained; exosomes will be extracted from peripheral blood and molecular and proteomic analyses will be performed.

Results:
From May 2014 to February 2015, 12 patients (pts) were enrolled. Median age: 60; Women: 7 (58%); never-smokers: 6 (50%); EGFR L858R=4 (33%) and Exon 19 deletion=8 (67%). Two of twelve (17%) were EGFR T790M positive at rebiopsy at the time of acquired resistance. Of 12 pts treated, 3 received ruxolitinib 10mg BID, 3 received 15mg bid and 6 received 20mg BID with erlotinib 150mg daily. No dose limiting toxicities were seen. The recommended phase 2 dose is ruxolitinib 20mg BID with 150mg erlotinib daily. Treatment-related AEs were all grade 1-3. The most frequent treatment related clinical adverse events (all grade 1-3) were anemia (25%), diarrhea (25%), rash (25%), pain (17%), fatigue (8%), and pneumonitis (8%). The most frequent treatment-related laboratory adverse events (all grade 1-2) were anemia (33%), elevated ALT (17%), elevated AST (17%), and hyperbilirubinemia (8%). Of the 12 pts treated, 2 (17%) required a dose reduction of erlotinib for treatment emergent toxicities; both subjects were on lower doses of erlotinib than 150mg daily prior to study enrollment. There were no dose reductions of ruxolitinib. Of 12 evaluable patients, no partial responses were seen. The median-progression free survival is 3 months. Two patients remain on study. One patient has been on study for 10 months with ongoing stable disease. Nine patients (75%) came off study for progression, 1 (8%) for toxicity. One person discontinued treatment on study for grade 3 pneumonitis, possibly related to the combination of erlotinib and ruxolitinib. The symptoms resolved with discontinuation of erlotinib and ruxolitinib.

Conclusion:
Combination erlotinib and ruxolitinib is well-tolerated. The phase 2 dose of ruxolitinib is 20mg BID in combination with erlotinib. There were no partial responses, but durable disease control was seen in some patients. The phase 2 study of erlotinib and ruxolitinib in this population is ongoing.

Abstract not provided

Background:
Reovirus is a naturally occurring virus which preferentially infects and causes oncolysis in tumor cells with a Ras-activated pathway. Cells that express high levels of EGFR are also susceptible to reovirus infection. In preclinical studies, reovirus induces host immunity and cell cycle arrest, acting synergistically with standard cytotoxic agents. Its adverse effects are mild to moderate flu-like symptoms. We have hypothesized those patients with EGFR-mutated, EGFR-amplified, or Kras-mutated NSCLC through a common downstream activated Ras pathway should be susceptible to treatment with reovirus

Methods:
We designed a Fleming, single-arm, phase II study to evaluate the objective response rate (CR + PR RECIST, or >40% PET SUV decrease) of reovirus in combination with paclitaxel-carboplatin as first-line therapy in patients with metastatic NSCLC. Secondary endpoints included progression free and overall survival. Eligible patients had ECOG PS 0-2, adequate organ function, no prior systemic chemotherapy for metastatic disease, and tumors with the specified genotype, as per CLIA certified testing. Adjuvant chemotherapy, or erlotinib/gefitinib for pts with EGFR mutant tumors was permitted.

Results:
Thirty-seven patients were enrolled. Molecular tumor demographics included 20 pts with Kras mutations; 10 with EGFR amplification alone; 3 patients with EGFR mutations and four patients with BRAF V600E mutations. Overall, 258 cycles (median 4, range 1-47) were administered. Initial doses used were C AUC 6 on day 1, and P 200 mg/m[2],on day 1 of each 21-day cycle. Due to unacceptable toxicities (grade 3 diarrhea and febrile neutropenia [1 each]) in the first two patients, doses were reduced to P 175 mg/m-m[2] and C AUC 5.. Common toxicities considered at least possibly related to the therapy included fatigue (30 pts); diarrhea (21 pts); nausea (19 pts); arthralgia-myalgia (15 pts); and anorexia (9 pts). Grade 3-4 adverse events included neutropenia (7 Gr3, 1 Gr4), anemia (2 Gr3), fatigue (9 Gr3), diarrhea (3 Gr3), nausea/vomiting (3 Gr3) and a single case of sepsis. Response evaluation showed 11 PR (5 Kras mutant), 20 SD, 4 PD and 2 NE patients by RECIST (ORR: 31%, 90% one-sided lower CI: 21%). Four of the SD patients had >40% PET SUV reductions after two cycles. Three patients opted to switch to pemetrexed maintenance after 4 cycles without disease progression or moderate/severe toxicity. Median PFS, OS and 12 month overall survival rates were: 4 months (95% CI: 2.9-6.1), 13.1 months (95% CI: 9.2-21.6) and 57% (95% CI: 39-72%), respectively. Seven patients are alive after a median follow up of 34.2 months (range: 26.9-71.5), including two patients with no evidence of disease progression to date (50 and 37 months).

Conclusion:
Oncolytic reovirus administration in combination with paclitaxel and carboplatin was well tolerated. The RECIST response rate (11/35 [31%]; 28% of Kras mutants)(15/35; 43% if PET is considered) is not conclusive, nor excludes additional benefit of the reovirus to chemotherapy. However, the number of patients surviving longer than 2 years (11; 30%) is substantial, suggesting either effect of second/third line post paclitaxel/carboplatin/reolysin treatment or perhaps the triggering of an immune response following tumor reovirus infiltration. The latter concept merits further investigation.

Background:
KRAS mutations, which occur in approximately 30% of lung adenocarcinoma cases, represent a major unmet clinical need in thoracic oncology. Preclinical studies have demonstrated that KRAS mutant NSCLC cell lines and xenografts with additional alterations in either p53 or INK4a/Arf (CDKN2A) are sensitive to FAK inhibition. Defactinib (VS-6063) is a selective oral inhibitor of FAK. This trial examined the effect of FAK inhibition in patients with KRAS mutant NSCLC and various permutations of p53 and CDKN2A alterations.

Methods:
This multi-center, non-randomized, open-label, multi-cohort trial enrolled patients with advanced KRAS mutant NSCLC who had received at least one prior (platinum-based chemotherapy doublet) line of therapy. The primary endpoint was progression-free survival (PFS) at 12 weeks. Patients were enrolled into one of four cohorts defined by INK4a/Arf and p53 status. In all cohorts, patients received defactinib 400 mg orally BID until disease progression.

Results:
Fifty-three patients with KRAS mutant NSCLC were enrolled across 9 US sites as of the data cut-off date (13-Mar-2015). Forty-seven patients were enrolled to one of the four molecularly defined cohorts. The median age was 62 years (range 33-80); 48% were female. The median number of prior lines of therapy was 3 (range 1-8) 15 (28%) pts met the 12 week PFS endpoint, with one patient achieving a PR. Median PFS was 46 days (range 12-205 days). Eight patients remained on study as of the data cut-off date. Clinical efficacy did not correlate with secondary mutation status across this KRAS mutant population. Adverse events considered at least possibly related to defactinib were experienced by 35 pts (76%). The majority of these were grade 1 or 2. 11 patients (24%) experienced at least possibly related grade 3-5 events, including 2 grade 5 respiratory failure events. Underlying disease was a confounding factor in many pts. The most commonly reported treatment emergent adverse events of any grade were fatigue (24%) and increased bilirubin (24%).

Conclusion:
In pretreated pts with KRAS mutant NSCLC defactinib demonstrates promising clinical activity with disease control rates comparable to other molecularly targeted agents for this pt population. Defactinib was generally well tolerated. Further development is warranted. Clinical trial: NCT01778803.

Background:
Tobacco-related non-small cell lung cancer (NSCLC) is associated with reduced survival and greater genomic instability. Veliparib (V) is a PARP inhibitor that augments platinum-induced DNA damage in preclinical studies, and a recent Phase 2 trial of advanced NSCLC trended to improved survival (HR 0.80; CI 0.54–1.18) when V was added to carboplatin (C) and paclitaxel (P). Here we report outcomes based on smoking status from this randomized Phase 2 study of CP with either V or placebo in advanced NSCLC.

Methods:
Patients with previously untreated advanced/metastatic NSCLC were randomized 2:1 to CP with either V at 120mg BID or placebo (randomization stratified by histology and smoking history). Cotinine was measured in patients’ plasma samples as an index of recent tobacco use.

Results:
Of 158 patients, 68% were male, and 49% had squamous NSCLC. At study entry, 60% pts were self-reported current smokers, 27% former smokers, and 13% never smoked. There were no significant differences in veliparib pharmacokinetic parameters between cotinine-high and low. Grade 3/4 AEs were elevated in current-smokers treated with VCP vs CP (66% vs. 40%, p=0.026); all-grade AEs and SAEs were similar between the two groups. The most common AEs in current-smokers were neutropenia (41% VCP; 27% CP), alopecia (36%; 33%), and anemia (31%; 40%). Figure 1 A sensitivity analysis of heavy vs light-smokers (≥ vs <39 pack-years, current or former smokers) showed advantage of veliparib in heavy-smokers: median PFS [HR(95% CI)] for VCP/CP was 7.0 vs 3.5 [0.43(0.20–0.94)] for heavy-smokers and 4.4 vs 4.2 [0.97(0.49–1.92)] for light-smokers; median OS was 12.6 vs 8.8 [0.52 (0.27–1.02)] for heavy-smokers and 9.9 vs 8.8 [0.92(0.53–1.61)] for light-smokers. A cotinine sensitivity analysis found that outcomes in cotinine-high were similar to current-smokers: PFS, cotinine-high HR was 0.38 (0.19–0.73) and cotinine-low was 0.97 (0.51–1.87); OS, cotinine-high HR was 0.52 (0.29–0.92) and cotinine-low was 1.07 (0.63–1.81). In univariate analyses assessing the influence of baseline characteristics and treatment on outcomes, smoking status and treatment had a significant interaction (p=0.0301 PFS, p=0.0118 OS). Additionally, multivariate analysis including all factors also identified current smoking as predictive of improved outcomes with VCP.



Conclusion:
Smoking status was a strong predictor of efficacy for veliparib-chemotherapy combination in advanced NSCLC. No differences in pharmacokinetics of V were seen based on plasma cotinine; toxicity of VCP was acceptable regardless of smoking history. A Phase 3 study has been initiated in patients with smoking history (M14-359).

Background:
Cabozantinib (C) is a small molecule inhibitor of multiple receptor tyrosine kinases, including MET, VEGFR2 & RET. MET is involved in tumor differentiation & VEGFR2 is a mediator of angiogenesis. Erlotinib (E) is FDA approved for the treatment of NSCLC.

Methods:
The primary objective of this randomized phase 2 study was to compare progression-free survival (PFS) of pts treated with E vs. C, & E vs E+C; each comparison had 91% power to detect a PFS hazard ratio (HR) of 0.5 with a 1-sided 0.10-level test stratified on prior number of therapies & ECOG PS. Secondary objectives included overall survival (OS), RECIST 1.1 response & CTCAE v4 toxicity. Pts were selected with previously treated (1-2 regimens) metastatic non-squamous EGFR wt NSCLC. Submission of archival tissue for central MET IHC testing was required. Oral daily dosing was: E-150 mg; C-60 mg; E+C-150 mg E, 40 mg C. Imaging was performed every 8 weeks. Pts optionally crossed over to E+C following progression on E or C.

Results:
125 pts were enrolled, of which 115 were eligible & treated (E, n=39; C, n=39; E+C, n=37). Pt characteristics were balanced between arms except for lower rate of brain mets history on E (p=0.02). Median follow up is 8.5 m. Compared with E (median 1.9 m), PFS was significantly improved on C (3.9 m, HR 0.33, p=0.0002, 80% CI 0.22-0.49) & E+C (4.1 m, HR 0.31, p=0.0002, 80% CI 0.21-0.46). Similarly, compared with E (median 4.0 m), OS was significantly improved on C (HR 0.52, p=0.02) & E+C arm HR 0.50, p=0.02). Grade 3-4 treatment-related hypertension & mucositis were higher on C and grade 3-4 diarrhea was higher on E+C. Overall worst grade toxicities were also significantly higher on C and E+C. MET IHC results were available on 88 patients from the primary analysis & 85% were positive (1-3+ membrane or cytoplasm staining with MET4 antibody). There was no correlation between MET status and PFS.

Conclusion:
C & C+E significantly improved PFS over E alone in pts with EGFR wt NSCLC. Cabozantinib-based regimens are promising for further investigation in this patient population. Funded by ECOG-ACRIN and NCI Contract No. HHSN261200800001E.

Abstract not provided

Background:
EGFR exon 20 insertions (ins20) represent a rare subtype (4%) of EGFR mutations and are refractory to EGFR-specific tyrosine kinase inhibitors (TKIs). No effective targeted therapies exist for patients (pts) with ins20; median PFS on the irreversible EGFR TKI Afatinib is 2.8 months (mos). Based on a durable RECIST partial response (PR) to AUY922, a Heat Shock Protein 90 (Hsp90) inhibitor, observed in an EGFR ins20 patient in a previous study (NCT01124864), we designed a phase II investigator-initiated trial to assess the activity of AUY922 in NSCLC pts with EGFR ins20. Since pts with these mutations are rare, we identified other international investigators who have treated ins20 patients with AUY922. Here, we present the results of a pooled international experience of 21 patients with EGFR ins20 treated with AUY922 in the United States, Taiwan and the Netherlands.

Methods:
A total of 21 patients with EGFR in20 are included in this analysis. 14 were treated on a single-arm, multi-center, open-label study of AUY922 in advanced NSCLC pts with EGFR ins20 mutations in the US (NCT01854034). Five were treated on a multicenter Taiwanese trial of AUY922 across a variety of molecular NSCLC subtypes (NCT01922583) and two were treated on a compassionate-use basis in the Netherlands. The starting dose of AUY922 was 70mg/m2 IV weekly for all patients.

Results:
21 pts, including 14 females and 7 males, average age 55 (range, 27-75) were included in this analysis. The median number of prior therapies was 2 (range, 1-6.) 6 pts received a prior EGFR TKI; none responded to TKI monotherapy. The most common AUY922-related toxicities were grade 1-2 visual changes (18/21; 86%) diarrhea (18/21; 86%) and fatigue (15/21; 71%). The only treatment-related grade 3 toxicities was hypertension (2/21; 1%) and AST elevation (1/21; 0.5%). There was one death on study, related to pre-existing comorbidity/unrelated to AUY922. Among the 21 patients treated, 5 achieved a partial response by RECIST 1.1 (ORR 24%) (Figure 1.) The median PFS estimate is 3.9 mos (95% CI, 2.9 to 10.7.) 6 patients remain on treatment at the time of abstract submission. Updated results and correlation with specific ins20 mutations will be presented. Figure 1



Conclusion:
This international experience suggests that AUY922 may be an active therapy for advanced NSCLC pts with EGFR ins20 mutations with an ORR 24% and median PFS 3.9 mo. AUY922 is generally well-tolerated, though reversible low-grade ocular toxicity is common. Further study of AUY922 in this population is warranted.

Background:
To avoid uncontrolled off-label use and allow for a nationwide safe access to crizotinib (crz) for patients (pts) with an ALK, MET or ROS1 positive (+) tumor, the French National Cancer Institute (INCa) launched the AcSé program, funding both access to tumor molecular diagnosis and an exploratory multi-tumor 2-stage design phase II trial. We report the preliminary results of the ROS1+ NSCLC cohort.

Methods:
ROS1 status was assessed in 28 regional INCa molecular genetic centers by break-apart FISH assays in tumor samples showing an IHC score of ≥1+. Pts with ROS1 rearrangements, progressing after at least one standard treatment (including a platinum-based doublet, unless pts were considered as unfit for chemotherapy) were proposed to receive crz 250 mg BID. Responses were centrally assessed using RECIST v1.1. The objective response rate (ORR) and disease control rate (DCR) were assessed every 8 weeks.

Results:
From Aug. 5, 2013 to Mar. 1, 2015, 39 pts with ROS1+ NSCLC were enrolled. 37 pts had received crz, leading to 37 pts with clinical information. Median age: 62 years (range 33–81), 70% females, 95% non-squamous histology, and 94% metastatic disease at study entry. Median number of prior treatments: 2 (range 1 –7). Twenty four pts were still on treatment at the cut-off date, 13 have stopped crz (8 PD, 3 adverse events (AEs), 2 deaths). Among the 27 pts evaluable for response at 8 weeks, we observed 16 PR, 7 SD and 4 PD, leading to ORR=59% [95% CI:39-78], and DCR=85% [66-96]. DCR at 6 months was 57% (disease control was achieved in 12/21 evaluable pts). Crz was well tolerated with only 4 grade ≥3 (1 AE + 3 SAEs) and 9 grade 1-2 SAEs. Most common AEs, mainly grade 1, were visual disorders (54% of pts), peripheral edema (51%), diarrhea (48%), nausea (46%), and elevated transaminases (43%).

Conclusion:
Crz was well tolerated and achieved a robust treatment response rate in ROS1+ NSCLC. These results underline the interest of integrating ROS1 in biomarkers routine screening. Survival data and duration of response will be presented.

Background:
Non-small cell lung cancer (NSCLC) patients with ROS1 chromosomal rearrangement benefit from treatment with the ROS1 inhibitor crizotinib with remarkable response rates and durable disease control. Similar to ALK and EGFR mutant NSCLC treated with targeted kinase inhibitors, disease progression inevitably occurs due to acquired resistance either by mutation within the kinase domain of ROS1 or via bypass signaling. However, limited data exists on the spectrum of resistance mechanisms in ROS1+ NSCLC. Here report on a novel bypass mechanism for ROS1 resistance discovered in a ROS1+ tumor sample from patient with acquired resistance to crizotinib in which an activating mutation in the KIT receptor (p.D816G) desensitize ROS1 cells to crizotinib inhibition.

Methods:
Patients with ROS1+ NSCLC treated with crizotinib who developed acquired resistance underwent biopsy of a progressing tumor. Tumor samples were analyzed for potential resistance mechanisms. Assessment of mutations within the ROS1 kinase domain was accomplished by direct sequencing of exons 35 thru exon 42 of ROS1 from genomic DNA isolated from FFPE tissue. The SNaPshot® Multiplex System was used to profile additional tumor related genes for mutations. The ROS1 rearranged cell lines, HCC78 and CUTO-2, were transduced with lentivirus to generate ectopic expression of the KIT[D816G] cDNA. Cell proliferation was assessed by an MTS assay and cellular signaling was measured by western blot analysis.

Results:
Sequencing of the patient’s post crizotinib sample showed no mutation in the ROS1 kinase domain. Additional mutational profiling by SNaPshot® revealed the acquisition of a KIT[D816G] mutation in the post-crizotinib sample that was not present in the pre-crizotinib tumor sample. HCC78 and CUTO-2 ROS1+ cell lines expressing the KIT[D816G] mutation were refractory to crizotinib by both cell proliferation assays and analysis of downstream signaling pathways. Both ROS1 and KIT activity had to be inhibited in order to suppress downstream signaling and proliferation in these cells.

Conclusion:
Activation of KIT by a gain-of-function mutation is a novel mechanism of resistance to crizotinib in ROS1 rearranged NSCLC. This bypass-signaling pathway serves as a ROS1 independent mechanism of progression, similarly to previously identified EGFR or RAS signaling pathways, and can potentially be targeted by KIT inhibitors.

Background:
Chromosomal rearrangements involving neurotrophic tyrosine kinase 1 (NTRK1) occur in less than 1% of NSCLCs. Cell-based assays have demonstrated that NTRK1 rearrangement leads to expression of an oncogenic TrkA fusion protein. While inhibition of TrkA in preclinical models reduces TrkA auto-phosphorylation and cell proliferation, the clinical activity of TrkA inhibitors in NSCLCs harboring an NTRK1 fusion is not known. Entrectinib (RXDX-101) is an orally available tyrosine kinase inhibitor of TrkA, TrkB, TrkC, ROS1, and ALK, with IC50 values for kinase inhibition ≤ 2 nM.

Methods:
We used an anchored multiplex polymerase chain reaction (AMP) assay to screen for NTRK1 rearrangements (Zheng et al., Nature Medicine 2014). Among over 663 NSCLC cases screened, we identified one positive case in which the 3’ end of SQSTM1 exon 6 was fused to the 5’ end of NTRK1 exon 10, leading to an SQSTM1-NTRK1 fusion transcript. We enrolled the patient onto the Phase 1 dose escalation study of entrectinib in adult patients with locally advanced or metastatic tumors (NCT02097810). The dose of entrectinib was 400 mg/m[2] (750 mg) once daily. We assessed safety of entrectinib and response to treatment using RECIST 1.1.

Results:
The patient is a 46 yo male with a 30 pack year smoking history who was first diagnosed with metastatic NSCLC in November 2013. Prior therapies included carboplatin/pemetrexed, pembrolizumab, docetaxel, and vinorelbine. At the time of study enrollment, the patient had an ECOG performance status of 2 and required supplemental oxygen at a rate of 3 liters per minute by nasal cannula. He reported significant pain and dyspnea due to widely metastatic disease, including a large left hilar mass narrowing the left upper lobe bronchus and obstructing the left lower lobe bronchus, extensive and palpable neck and chest lymphadenopathy, and a palpable expansile left chest wall mass. Staging head CT also revealed numerous (15 to 20) asymptomatic brain metastases measuring up to 1.7 cm that had not been previously treated. The patient was started on entrectinib and tolerated the study medication well, with one adverse event of grade 1 dysgeusia, which resolved after two weeks. Within three weeks of starting treatment, the patient reported resolution of dyspnea and pain, and improvement in energy and appetite. He no longer required supplemental oxygen and all sites of palpable disease had improved or resolved. At four weeks of treatment, restaging CT scans demonstrated a partial response by RECIST of -47%, with significant regression or resolution of lymphadenopathy, reduction in size of the chest wall mass, and marked reexpansion of the left lung. Restaging of the CNS by head CT demonstrated near complete resolution of previously visualized brain metastases.

Conclusion:
In a heavily pre-treated patient with NSCLC harboring an NTRK1 gene fusion, entrectinib therapy resulted in rapid clinical improvement and a radiologic partial response at 4 weeks with minimal toxicity. This preliminary report suggests that entrectinib may be an effective therapy for patients with NTRK1-rearranged NSCLC.

Abstract not provided

Background:
FGFR1 amplification is one of the most common potential driving oncogenes in squamous cell carcinoma (SCC), which accounts for 20% of non-small cell lung cancer (NSCLC) squamous cell carcinoma. This phase II study evaluated the efficacy and toxicity profile of dovitinib, an orally active FGFR (fibroblast growth factor receptor) inhibitor, in advanced SCC patients.

Methods:
Patients with histological confirmed advanced squamous cell NSCLC and previously treated with at least one cytotoxic chemotherapy were enrolled from April 2013 to December 2014. All patients had FGFR1 gene amplification more than 5 copies by fluorescent in situ hybridization (FISH). Each 7-day treatment cycle consisted of dovitinib 500mg orally administration on days 1 to 5 and 2 days off. Primary endpoint was overall response rate and secondary endpoints included PFS, OS and toxicity.

Results:
All 26 patients were male with the median age of 68 years (range, 52 – 80). Most patients were ever smokers (96%) and had good ECOG (0-1) performance status (85%). The median number of dovitinib treatment cycles administered was 2.5 (range, 1-12). The overall response rate (ORR) was 11.5% (95% CI, 0.8 – 23.8) and disease control rate (DCR) was 50% (95% CI, 30.8 – 69.2). There were three partial responses (PR) and ten stable diseases (SD). Duration of response in 3 patients who achieved PR was 4.5+, 5.1+ and 6.1months. After the median follow-up duration of 15.7 months (range, 5.8 – 25.6 ), the median overall survival (OS) was 5.0 months (95% Confidential Interval, 3.61 – 6.39) and progression-free survival (PFS) was 2.9 months (95% CI, 1.54 – 4.26). Grade 1/2 fatigue (69%) and anorexia (85%) were most commonly reported adverse events and 12 patients (46%) required dose reduction of dovitinib.

Conclusion:
Dovitinib treatment a showed modest efficacy in advanced squamous cell lung cancer patients with FGFR1 amplification. Further studies to evaluate other biomarkers correlated with the efficacy of dovitinib in SCC should be warranted.

Background:
In non-squamous non-small cell lung cancer (NSCLC), PI3-kinase (PI3K) pathway activation, including downregulation of phosphatase and tensin homolog (PTEN) expression, may promote cell survival and enhance chemotherapy resistance. Additionally, mutations in KRAS have been shown preclinically to confer resistance to PI3K inhibition. The pan-PI3K inhibitor pictilisib potentiates the activity of taxanes, platinum agents, and antivascular endothelial growth factor therapy in preclinical models of NSCLC. This phase II hypothesis-generating study (NCT01493843) evaluated the safety and efficacy of pictilisib in combination with carboplatin, paclitaxel, and bevacizumab in patients not treated for advanced or recurrent non-squamous NSCLC.

Methods:
Overall, 158 patients were randomized to receive carboplatin (area under the curve [AUC] = 6 mg/ml/min), paclitaxel (200 mg/m[2]), and bevacizumab (15 mg/kg) every 3 weeks (q3w) with 340 mg oral pictilisib (n=79) or placebo (n=79) daily in the first 2 weeks of each cycle for a total of 4 cycles. Bevacizumab q3w with daily pictilisib or placebo was continued until disease progression or unacceptable toxicity. Stratification factors included Eastern Cooperative Oncology Group performance status and smoking status. The primary endpoint was progression-free survival (PFS) in the intention-to-treat (ITT) population and in patients with PTEN null/low expression (assessed by immunohistochemistry). Overall survival (OS), objective response rate (ORR), and safety were secondary endpoints. Pre-planned exploratory analyses included efficacy in the KRAS-wildtype subgroup. Tumor assessment was based on RECIST v1.1. Safety analyses were performed on patients who received at least one dose of study drug.

Results:
Median PFS in the ITT population was 6.9 months in the pictilisib arm and 5.9 months in the placebo arm (HR 0.82; 90% CI 0.59–1.13), while median OS was 13.6 months (pictilisib arm) versus 16.1 months (placebo arm) (HR 1.12; 90% CI 0.79–1.59). In patients with PTEN null/low expression, median PFS was 5.9 months (pictilisib arm) and 5.7 months (placebo arm) (HR 0.74; 90% CI 0.41–1.32). In the KRAS-wildtype subgroup, median PFS was 9.7 months (pictilisib arm) versus 5.7 months (placebo arm) (HR 0.70; 90% CI 0.45–1.09); median OS was 14.5 months in both arms. ORR in the ITT population was 37% (pictilisib arm) versus 29% (placebo arm). In the pictilisib arm, common grade ≥3 adverse events (AEs) included neutropenia (23%), rash (20%), thrombocytopenia (8%), febrile neutropenia (5%), and hyperglycemia (5%). AEs led to higher rates of discontinuation in the pictilisib arm (26% versus 16% in the placebo arm), particularly during the first 4 cycles. However, the proportion of AE-related deaths was higher in the placebo arm (9 [12%] versus 5 [6%] in the pictilisib arm).

Conclusion:
This phase II trial of first-line pictilisib plus chemotherapy and bevacizumab in patients with non-squamous NSCLC showed a modest trend for improved PFS, with additional toxicity and no OS benefit. The safety profile was consistent with other pictilisib trials. PTEN null/low expression was not a predictive biomarker, although its prognostic value cannot be excluded. A trend for improved PFS, but not OS, was observed in the KRAS-wildtype subgroup, especially during the maintenance phase of treatment.

Background:
In squamous non-small cell lung cancer (NSCLC), the PI3-kinase (PI3K) pathway may be activated via several mechanisms including PIK3CA amplification and downregulation of phosphatase and tensin homolog (PTEN) expression; activation of this pathway can promote cell survival and enhance chemotherapy resistance. Pictilisib, a pan-PI3K inhibitor, potentiates the activity of taxanes and platinum agents in preclinical NSCLC models. This phase II, hypothesis-generating study (NCT01493843) evaluated the safety and efficacy of pictilisib in combination with carboplatin and paclitaxel in patients with advanced or recurrent squamous NSCLC.

Methods:
Overall, 160 patients were randomized to receive carboplatin (target area under the curve [AUC] = 6 mg/ml/min) and paclitaxel (200 mg/m[2]) every 3 weeks with 340 mg oral pictilisib (n=81) or placebo (n=79) daily in the first 2 weeks of each cycle for a total of 4 cycles. Pictilisib or placebo was continued daily until disease progression or intolerable toxicity. Stratification factors included Eastern Cooperative Oncology Group performance status and smoking status. The primary endpoint was progression-free survival (PFS) in the intention-to-treat (ITT) population and in patients with PIK3CA amplification (assessed by chromogenic in situ hybridization [CISH]). Overall survival (OS), objective response rate (ORR), safety, and PFS in the PTEN null/low subgroup were secondary endpoints. Tumor assessment was based on RECIST v1.1. Safety analyses were performed on patients who received at least one dose of study drug.

Results:
Median PFS in the ITT population was 5.6 months in the pictilisib arm and 5.5 months in the placebo arm (HR 0.82; 90% CI 0.60–1.12). Median OS was 11.7 months in the pictilisib arm and 12.2 months in the placebo arm (HR 1.10; 90% CI 0.77–1.57). PFS and OS analyses in patients with PIK3CA amplification will be presented. Median PFS for the PTEN null/low subgroup was 6.7 months in the pictilisib arm and 5.5 months in the placebo arm (HR 0.69; 90% CI 0.42–1.13). ORR in the ITT population was 28% in the pictilisib arm and 34% in the placebo arm. Common grade ≥3 adverse events (AEs) included neutropenia (18%), rash (8%), and thrombocytopenia (7%). AEs led to higher proportion of discontinuations (22% in the pictilisib arm vs. 15% in the placebo arm) and AE-related deaths in the pictilisib arm (12 [14%] vs. 2 [3%] in the placebo arm). Deaths were due to disease progression or AEs typically reported in lung cancer. No unexpected safety signals were identified for pictilisib.

Conclusion:
In this first phase II trial of a PI3K inhibitor in first-line squamous NSCLC, the combination of pictilisib with chemotherapy introduced additional toxicity with a minimal PFS improvement and no OS benefit in the ITT population. The safety profile was consistent with other pictilisib trials. PTEN null/low expression did not identify a subgroup with significantly improved efficacy, although the prognostic value of PTEN as a biomarker in squamous NSCLC cannot be excluded. Efficacy analysis in the PIK3CA amplification subgroup is ongoing and will be presented at the conference.

Background:
This abstract is under embargo until September 9, 2015 and will be distributed onsite on September 9 in a Late Breaking Abstract Supplement.

Methods:


Results:


Conclusion:

Abstract not provided

Abstract not provided

Abstract not provided

Abstract:
Cancer immunotherapy in a broad sense is any interaction with the immune system to treat cancer. One approach is non-antigen-specific modulation of the immune system. Historical examples with e.g. BCG, interferon, interleukins, were disappointing in lung cancer. More recently, specific antibodies against the Programmed Death 1 (PD-1) receptor or its ligands (PD-L1) have delivered exciting results, with major patient benefits in randomised controlled trials (RCTs) in relapsing NSCLC {Brahmer, 2015 19949 /id}. Antigen-specific immunotherapy aims at specific priming of immune system to recognize the tumour as foreign, thereby generating specific antibodies and/or cytotoxic T cells. This is “therapeutic cancer vaccination (TCV)”. Conditions for optimal TCV are: 1/ specificity (well-defined target antigen(s) in the tumour, not in other tissues); 2/ selectivity (use in the population expressing the target); 3/ immunogenicity (interaction with antigen leads to effective humoral and/or cellular response); 4/ tumour sensitive to immune kill in order to obtain improvement in patients’ outcome. Better knowledge of tumour immunity has led to encouraging data in phase II RCTs with several TCVs, which then have entered large phase III trials. Examples are the MAGE-A3 vaccine studied in resected early stage NSCLC, the BLP-25 vaccine in locally advanced NSCLC after chemoradiotherapy, and e.g. belagenpumatucel-L and the TG4010 vaccine in advanced stage NSCLC. The MAGE-A3 protein is totally tumour-specific and present in about 35% of early stage NSCLC. In the hypothesis generating double-blind, randomized, placebo-controlled phase II study, 182 patients with completely resected MAGE-A3-positive stage IB-II NSCLC received recombinant MAGE-A3 protein combined with an immunostimulant (13 doses over 27 months) or placebo (2). No significant toxicity was observed. There was a 24% – non-significant – improvement in disease-free survival (DFS, HR 0.76; 95% CI 0.48 to 1.21). The ensuing large phase III study MAGRIT (MAGE-A3 as Adjuvant Non-Small Cell LunG cancer ImmunoTherapy) was reported at the ESMO 2014 meeting (3). MAGE-A3 positive patients with completely resected stage IB-II-IIIA NSCLC and adjuvant chemotherapy as clinically indicated, were randomly 2:1 assigned to receive MAGE-A3 vaccine or placebo. Almost 14,000 surgical patients were screened, 4210 patients were MAGE-A3 positive (33%), and 2312 patients were randomised. The median DFS (primary endpoint) was slightly better with MAGE-A3 (60.5 versus 57.9 months), but the difference was unfortunately not significant (Hazard Ratio, HR, 1.02, 95%CI: 0.89, 1.18, P=0.74). No subgroups with potential benefit could be identified. Based on this disappointing result, further development of the MAGE-A3 vaccine in NSCLC has been abandoned. Mucins like the MUC1 protein are present in many epithelia, but MUC1 expression is altered (mainly by aberrant glycosylation) in many cancer types, including NSCLC. The tandem repeat MUC1-peptide liposomal vaccine BLP-25 has been studied in patients with stage IIIB-IV NSCLC (4). While overall survival (OS) was not significantly different in the total group, a challenging effect was observed in stage IIIB patients (HR 0.524; 95%CI 0.261-1.052). This led to START (Stimulating Targeted Antigenic Responses to NSCLC Trial), a phase III, double blind, RCT comparing maintenance therapy with Tecemotide (n=829) or placebo (n=410) in patients with unresectable stage III NSCLC who did not progress after sequential or concurrent chemo-radiotherapy (5). The primary endpoint – OS – was not significantly different between the vaccine and placebo group (25.6 and 22.3 months). However, pre-planned subgroup analysis showed that the patients treated with concurrent chemoradiotherapy (N=829) had a 10.2-month improvement in OS (30.8 versus 20.6 months, adjusted HR 0.78, P=0.016). The consequential trial was START 2, a similar large RCT in patients who completed concurrent chemoradiotherapy for unresectable stage III NSCLC (NCT02049151). However, this RCT and further development of Tecemotide was abandoned after disappointing results of a smaller trial in Japanese patients with stage III NSCLC and concurrent chemoradiotherapy. Belagenpumatucel-L is a vaccine based on a mixture of allogeneic tumour cells with TGF-β2 antisense blockade as adjuvant. A phase III trial in patients with stage III-IV NSCLC in disease control after first-line therapy was reported at the 2013 ESMO meeting (STOP, NCT00676507) (6). Patients without progression after 1[st] line chemotherapy, were randomly assigned to intradermal belagenpumatucel-L (N=270) versus placebo (N=262)for 24 months. Median OS was 20.3 months with belagenpumatucel-L versus 17.8 months with placebo (HR 0.94, p=0.594). In subgroup analysis of patients randomized <12 weeks after the last chemotherapy, the HR of the median OS was 0.77 (P=0.092). For patients enrolled within 12 weeks and treated with previous radiotherapy, the HR was HR 0.45 (P=0.014). The vaccine was well tolerated with mainly mild local administration side-effects. TG4010 is a vaccine based on a recombinant viral vector (attenuated strain of vaccinia virus) expressing both the tumour-associated antigen MUC1 and interleukin-2. This vaccine is explored in the phase IIB/III RCT TIME trial (NCT01383148). This double-blind, placebo-controlled trial evaluates standard first-line chemotherapy with or without TG4010 in MUC1-positive stage IV NSCLC patients. In the phase IIB part, the predictive value of activated NKs (TrPAL: triple positive activated lymphocytes) was evaluated based on a PFS endpoint, and reported in an interim report at the 2014 ESMO meeting (7). Based on a Bayesian analysis, the predefined endpoint of a HR <1 in the patients with low level of NK cells was met. The PFS was not significantly different between vaccine and placebo (HR 0.78, 95%CI 0.55-1.10]. In subgroup analyses, the effect was more pronounced in patients with non-squamous NSCLC (HR 0.71, 95CI 0.51-0.97) than in squamous histology. Therefore, a decision was made to continue the phase III part of the trial in non-squamous NSCLC only, with OS a the primary endpoint. References 1. Brahmer J, Reckamp KL, Baas P et al. Nivolumab versus docetaxel in advanced squamous cell non-small cell lung cancer. N Engl J Med 2015; on-line May 31. 2. Vansteenkiste J, Zielinski M, Linder A et al. Adjuvant MAGE-A3 immunotherapy in resected non-small cell lung cancer: Phase II randomized study results. J Clin Oncol 2013;31:2396-2403. 3. Vansteenkiste JF, Cho BC, Vanakesa T et al. MAGRIT, a double-blind, randomized, placebo-controlled phase III study to assess the efficacy of the recMAGE-A3 + AS15 cancer immunotherapeutic as adjuvant therapy in patients with resected MAGE-A3-positive non-small cell lung cancer (NSCLC). Ann Oncol 2014; 25 Suppl 4: abstract 1173O. 4. Butts C, Murray N, Maksymiuk A et al. Randomized phase IIB trial of BLP25 liposome vaccine in stage IIIB and IV non-small cell lung cancer. J Clin Oncol 2005;23:6674-6681. 5. Butts C, Socinski MA, Mitchell PL et al. Tecemotide (L-BLP25) versus placebo after chemoradiotherapy for stage III non-small-cell lung cancer (START): A randomised, double-blind, phase 3 trial. Lancet Oncol 2014;15:59-68. 6. Giaccone G, Bazhenova L, Nemunaitis J et al. A phase III study of belagenpumatucel-L therapeutic tumor cell vaccine for non-small cell lung cancer (NSCLC). Eur J Cancer 2013; 47 Suppl 2: abstract LBA 7081. 7. Quoix E, Losonczy G, Forget F et al. TIME, a phase 2B/3 study evaluating TG4010 in combination with first-line therapy in advanced non-small lung cancer (NSCLC). Phase 2B results. Ann Oncol 2014; 25 Suppl 4: abstract 1055PD.

Abstract:
The traditional approaches to lung cancer therapy have focused on treating the malignant epithelial cancer cells within the tumor. However, it is now realized that in most cases, most of the tumor consists of “supporting cells” that include endothelium, pericytes, fibroblasts, and a variety of innate and acquired (B cells and T cells) immune cells. Thus, targeting these non-tumor cells could be an alternative therapeutic strategy. This concept is already being used in clinical practice. One example is targeting the endothelial cells within the tumor using an anti-VEGF antibody (bevacizumab). Another example are the checkpoint inhibitors (anti-CTLA4 and anti-PD1 antibodies) that target endogenous T cells. However, it may also be possible to attack other targets such as macrophages, Tregs, neutrophils or cancer-associated fibroblasts (CAFs). Tumor-associated macrophages represent one target. These cells take on a tumor-supportive phenotype and produce anti-inflammatory cytokines/chemokines (i.e. TGFbeta, PGE2, IL10, VEGF), arginase (which inactivates T cells), and angiogenic factors. This has led to the hypothesis that changing the state of the macrophage to an anti-tumor phenotype in which immune-activating mediators would be made and antigen-presentation could be enhanced would have direct anti-tumor activities and would allow endogenous T cells to kill tumor cells. Macrophage activation has been attempted for many years using agents such as bacterial endotoxin, TNF, liposomal-encapsulated muramyl tripeptides, lipopeptides or oligonucleotides/agents that activate toll-like receptors. To date, however, this approach has not been very successful, primarily due to lack of specificity for tumor infiltrating macrophages resulting in intolerable systemic toxicity. Our group explored the use of a cell permeable flavonoid compound called DMXAA for this purpose. Administration of DMXAA causes activation of tumor-associated macrophages via multiple pathways with release of cytokines and chemokines resulting in hemorrhagic tumor necrosis, a subsequent inflammatory/immuno-permissive tumor environment, and ultimately attracts CD8 T cells into tumors (Jassar et al. 2005). Although intra-tumoral treatment of both large and small lung cancers in mouse models led to striking tumor regression, there was a major problem in translating this work- DMXAA does not react with human macrophages. Since we did not know how DMXAA was working (i.e. what was the DMXAA receptor that triggered macrophage activation) progress was stalled. This changed recently, when it was discovered that DMXAA worked by binding to a newly described intracellular sensor of cytosolic DNA (working through binding to cyclic dinucleotides) called STING (stimulator of Interferon Genes). STING activates innate immunity by signaling through the TBK/IRF3 axis, NF-kB and STAT6 pathways. Interestingly, it was noted that DMXAA bound well to mouse STING but NOT to human STING (explaining its lack of efficacy in humans). A company (Aduro Biotech) has designed a compound that binds to human STING and thus activates human macrophages like DMXAA activates mouse macrophages. Their lead compound has strong in vivo anti-tumor activity (much like DMXAA) and clinical trials using intra-tumoral injections are about to start (Corrales et al., 2015). Another potential target in the tumor microenvironment is the cancer-associated fibroblasts (CAFs). Fibroblasts and their associated stroma promote tumor growth through multiple mechanisms, including suppression of anti-tumor immunity, supporting angiogenesis, as a depot for growth factors/ cytokines/chemokines, modulating the inflammatory response, and shielding the tumor from infiltrating cells. Our group at Penn has been developing genetically altered T cells that can be targeted to any expressed surface antigen by transducing autologous T cells with a chimeric antigen receptor (CAR). A CAR is composed of a single chain antibody fused to the cytoplasmic sequences from the CD3zeta chain and a co-activating receptor (41BB/CD137). This construct combines antibody specificity with the ability to activate the killing machinery of T cells. Our lead CAR T cell target is CD19 to treat B cell malignancies, however, we are also testing CARs targeted to mesothelin (mesothelioma, lung cancer, pancreas cancer, ovarian cancer) and other solid tumor cell targets. We hypothesized that we could use this approach to eliminate CAFs. To do so, we identified Fibroblast Activation Protein (FAP) as a target antigen for CAFs. In epithelial-derived tumors, FAP is selectively expressed by cancer-associated stromal cells It is highly expressed in the stroma of lung (and many other) cancers, but not in benign tumors or normal adult quiescent tissues (although it is upregulated in wounds and fibrotic tissues). We thus produced T cells expressing anti-mouse FAP CARs. The FAP CAR T cells selectively killed FAP-expressing cells. Immune-competent C57BL/6 mice bearing large established subcutaneous murine lung cancers and human A549 tumors in immune-deficient mice were treated. FAP-CAR T cells reduced the number of FAP+ cells, markedly reduced the amount of tumor matrix and limited tumor growth in all three lung cancer models (Wang et al., 2014; Lo et al., 2015). We hope to move this approach forward to clinical trials in lung cancer and mesothelioma. We will likely combine anti-fibroblast therapy with chemotherapy, vaccines, or other types of immunotherapy. In summary, a new therapeutic paradigm is now emerging based on therapy aimed at the non-malignant host cells, NOT directly targeting the cancer cells. Examples include antibodies targeting endothelial cells and checkpoint inhibitors that target T cells. An advantage of this approach is that stromal cells are more genetically stable compared with tumor cells and they are unlikely to lose their antigen(s) and become invisible to T cells. Another advantage is the same targets could be used in multiple tumors. Future applications will likely include activation or elimination of TAMS, targeting fibroblasts, and deletion of T-regulatory cells. References: Corrales L, et al. Direct Activation of STING in the Tumor Microenvironment Leads to Potent and Systemic Tumor Regression and Immunity. Cell Reports 2015. 11:1-13. Jassar, A., et al. Activated Tumor-Associated Macrophages and CD8[+] T-cells are the Key Mediators of Anti-tumor Effects of the Vascular Disrupting Agent DMXAA in Murine Models of Lung Cancer and Mesothelioma”. Cancer Research 2005. 65:11752-11761. Lo A, et al. Tumor-promoting desmoplasia is disrupted by depleting FAP-expressing stromal cells. Cancer Res. 2015 May 15. [Epub ahead of print] Wang LC, et al. Targeting Fibroblast Activation Protein in Tumor Stroma with Chimeric Antigen Receptor T Cells Can Inhibit Tumor Growth and Augment Host Immunity Without Severe Toxicity. Cancer Immunology Research 2014. 2:154-166.

Abstract:
In the wake of the demonstrated 20% mortality reduction benefit reported from the randomized National Lung Screening Trial (NLST), the United States Preventive Services Task Forces gave a “B” recommendation for low dose CT screening of for lung cancer in high risk populations (1). This favorable endorsement in turn led the Centers for Medicare and Medicaid Service to fully reimbursement the cost of providing this service by federal and commercial insurers for high risk smokers between the ages of 55-77 who have been smoking within the last 15 years. With these provisions, national lung cancer screening is now being implemented in the United States. The protocol and screening process used for the NLST was fixed at the time of study initiation in 2002 when 4-detector scanners were the default CT device and screening management was delivered based on the existing community standard (2). In the time since the NLST was conducted there have been a number of developments that have improved the process of lung cancer screening services (3). These innovations range from the introduction of more capable CT scanners, lower medical radiation scanning protocols, more effective and efficient diagnostic work up approaches, as well as improved and more tailored surgical approaches. The aggregate effect of all of these advances is that the cost efficiency of this process is also improving (4). Further improvements with clinical management may occur as the use of quantitative CT imaging allows for more consistent measurement of suspicious pulmonary nodules, as size criteria is emerging as a key determinant guiding invasive screening work-up (5). However implementing national CT screening to ensure delivery of high quality, best-practice early lung cancer detection in the target population of tobacco-exposed individuals constitutes a profound challenge. Still the public health impact of tobacco-exposure is singularly lethal. In the United States alone over 438,000 annual deaths are related to tobacco-exposure with lung cancer being the most common cause of tobacco-related death approaching 30% of this total mortality burden. Advocacy groups have worked with academic medical centers as well as community hospitals to address this implementation challenge by creating a consortium of institutions that are conducting screening programs to systematically adopt best standard of screening practice for all components of clinical management (6). A critical aspect of the “Framework” process includes the expectation that participating institutions will prospectively acquire clinical follow-up information so that the outcomes of lung cancer screening efforts can be accessed and reported (6). This effort builds on previous models of cooperative research such as with using institutional feedback to accelerate learning curve in allowing new screening institutions to rapidly implement effective screening process. The best example to date of this approach with screening is the use of the recent I-ELCAP screening outcomes to evaluate the most favorable pulmonary nodule size to use as a threshold for a more invasive diagnostic work-up (7). Increasing the nodule size as the threshold for further diagnostic work-up markedly improves the efficiency of the screening management while reducing the rate of false positive work-ups, cost and morbidity (7). An indispensible element in the national implementation of screening is the simultaneous provision of best practice smoking cessation services for those individuals that continue to smoke. Pyenson and co-workers have reported that routine integration of smoking cessation with the annual CT screening process can improve the cost utility ratio of quality adjusted life years by close to 40% (4). Indisputably implementing national annual CT screening in high risk populations is a significant societal cost. However there are attractive opportunities to leverage this new pattern of care to further benefit health outcomes in this at-risk cohort. For example, the annual CT visit provides a scaffolding to support more intensive research to define better smoking cessation measures. In asymptomatic tobacco-exposed individuals, a growing body of research suggests that the CT scan done to evaluate for early lung cancer also commonly finds individuals with evidence of asymptomatic COPD/emphysema or coronary artery disease (8, 9). These diseases along with lung cancer account for close to 70% of the excess mortality in heavily tobacco-exposed populations. Lung cancer screening will permit additional research opportunities in this tobacco-exposed cohort including catalyzing the development for more effective drugs to manage the early stage of lung cancer. With screening, the frequency of finding early stage lung cancer is greatly increased and focusing on these early stage patients could allow for much more rapid evaluation of new targeted therapeutic agents compared to the current setting. For the same reason, lung cancer screening will also find many more early asymptomatic COPD patients and quantitative CT provides an economical biomarker to allow much more efficient COPD drug development research than is currently possible. Particular classes of drug targets such as immunomodulators could conceivably show benefit in arresting the progression of both early lung cancer and COPD. This time of initial US national screening dissemination is allowing a full national discussion not only about how to provide high quality lung cancer screening services, but also about how to thoughtfully leverage this newly reimbursed screening service to extend the utility of the thoracic imaging encounter and greatly accelerate progress with improving health outcomes in heavily tobacco-exposed populations. At the very least, evidence for one or more of these additional diseases on annual screening may heighten a smoker’s motivation to stop that habit. Other life style interventions such as diet modification and exercise are being successfully employed to manage the consequence of asymptomatic coronary calcification. Life style counseling could also emerge as integral part of the annual CT evaluation as these interventions can have markedly positive impact for a range of tobacco-dependent conditions. The emergence of lung cancer screening as a public health tool has evoked a lively global debate regarding its potential merits. While this healthy debate should continue, there are potentially unprecedented opportunities arising with this new approach to the lethality of chronic tobacco exposure that merit serious consideration. References: 1) Moyer VA. Screening for lung cancer: U.S. preventive services task force recommendation statement. Ann Intern Med. 2013 Dec 31. 2) Aberle D, Adams A, Berg C et al. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011; 365(5): 395-409. 3) Mulshine JL, D'Amico TA. Issues with implementing a high-quality lung cancer screening program. CA Cancer J Clin. 2014 Jun 27. 4) Villanti AC, Jiang Y, Abrams DB, Pyenson BS. A cost-utility analysis of lung cancer screening and the additional benefits of incorporating smoking cessation interventions. PLoS One. 2013 Aug 7; 8(8): e71379. PMCID: PMC3737088. 5) Mulshine JL1, Avila R, Yankelevitz D et al. Lung Cancer Workshop XI: Tobacco-Induced Disease: Advances in Policy, Early Detection and Management. J Thorac Oncol. 2015 May;10(5):762-7. doi: 10.1097/JTO.0000000000000489. 6) Rights and expectations for excellence in lung cancer screening and continuum of care.[homepage on the Internet]. Available from: http://www.screenforlungcancer.org/national-framework/. 7) Henschke CI. Definition of a positive test result in computed tomography screening for lung cancer: A cohort study. Ann Intern Med. 2013; 158(4): 246-252. 8) Zulueta J, Wisnivesky J, Henschke C, et al. Emphysema scores predict death from COPD and lung cancer. Chest. 2012; 141(5): 1216-1223. 9) Htwe Y, Cham MD, Henschke CI, et al. Coronary artery calcification on low-dose computed tomography: comparison of Agatston and Ordinal Scores. Clin Imaging. 2015 Apr 18. pii: S0899-7071(15)00098-4. doi: 10.1016/j.clinimag.2015.04.006. [Epub ahead of print]

Abstract:
Background: lung cancer mortality is still the leading cause of cancer death worldwide[(][1][)]. The majority of patients present with advanced disease and the current 5-year survival is only 15%. Despite treatment advances, there is little improvement of prognosis. As the American National Lung cancer Screening trial (NLST) showed that low-dose CT-scan (LDCT) screening can reduce lung cancer mortality in high risk subjects[(][2][)], the United States Preventive Task Force (USPTFS) recommends annual LDCT screening in adults who have a smoking history of at least 30 pack-years, and smoke now or have quit within the past 15 years and are between 55 and 80 years old[(][3][)]. However, there are still some important challenges, f.e. high prevalence of false-positives, overdiagnosis and the optimal screening strategy . In Europe, the sufficiently powered Dutch-Belgian lung cancer screening trial (NELSON) is still ongoing. This trial is currently in the final phase of follow-up prior to definitive analysis and reporting. Results: the NELSON trial has been setup in 2003, in which subjects at high risk for lung cancer were selected from the general population[(][4][)]. After informed consent, 15,791 participants were randomised (1:1) to the screen arm (n=7,900) or the control arm (n=7,891) (Figure 1). The screen arm participants received LDCT screening at four times: at baseline, after one year, after two years and after two and a half years, whereas the control arm participants received usual care (no screening). According to size and volume doubling time (VDT) of the nodules, three screen results were possible: negative (invitation for the next screening round), indeterminate (an invitation for a follow-up scan) or positive (referred to the pulmonologist because of suspected lung cancer). Main results of the first three screening rounds showed a favorable cancer stage distribution of the screen-detected lung cancers detected in the NELSON trial compared to the other trials and was more favorable (p<0.001) than in the NLST[(][5][)]. More than half of the screen-detected lung cancers were adenocarcinomas (51.2%) and a large proportion was localized in the right upper lobe (45.0%). Women diagnosed with lung cancer were significantly younger (58.0 vs. 62.0 years; p=0.03), had a lower BMI (23.8 vs 25.9;p=0.003) and were diagnosed at a more favorable cancer stage (p=0.028) than the men diagnosed with lung cancer. From the first screening round up to two years of follow-up after the third round scan, 34 participants were diagnosed with an interval lung cancer[(][6][, ][7][)]. Retrospectively, two-thirds of these interval cancers were visible on the last screening CT; detection errors, interpretation errors, and human errors were identified as the main causes of the failure in half of the interval cancers. Interval cancers were diagnosed at more advanced stages (p<0.001 ), and were more often small cell carcinoma (p=0.003) and less often adenocarcinomas (p=0.005) than screen-detected lung cancers. For the first three rounds combined, sensitivity was 84.6% (95% CI 79.6-89.2%), specificity was 98.6% (95% 98.5%-98.8%), positive predictive value was 40.4% (95% CI 35.9-44.7%), and negative predictive value was 99.8% (95% CI 99.8%-99.9%)[(][6][)]. For the first screening round, the sensitivity was the same, but the specificity was higher in the NELSON trial relative to the NLST: 98.3% vs. 73.4%. The positive predictive value was in our trial (40.4%) substantially higher than in other trials: f.e. 3.8% in the NLST[(][2][)]. Furthermore, our findings showed that the 2 year-probability of developing lung cancer for all included participants was 1.3% (1.2-1.5)[(][8][)]. For screened participants without any nodules this probability (more than half of the included participants) was 0.4%, which suggests that a screening interval of at least two years might be safe to apply in these individuals. In all participants with CT-detected nodules, lung cancer probability was 2.5% (2.1-2.9) but individuals’ probabilities depended strongly on nodule volume, diameter and VDT. New data: the last screening round, which took place 2.5 years after the third round, showed 46 screen-detected lung cancers, of which 58.7% were diagnosed at stage I and 23.8% at stage III/IV. More squamous-cell carcinomas (21.7% vs. 16.3%), small cell carcinomas (6.5% vs. 3.8%) and bronchioalveolar carcinomas (8.7% vs. 5.3%) were detected compared to the first three screening rounds. However, relative to the first three rounds the lung cancer detection rate was lower (0.80 vs 0.80-1.1) and lung cancer was detected at a more advanced stage (stage III/IV; 23.8% vs 8.1). Currently, we are working on the review of blinded medical files of the deceased participants to determine the cause of death, and we are collecting medical data of control arm participants. Figure 1 References 1. Siegel R, Ma J, Zou Z, Jemal A. Cancer statistics, 2014. CA Cancer J Clin. 2014 Jan-Feb;64(1):9-29. 2. National Lung Screening Trial Research T, Aberle DR, Adams AM, Berg CD, Black WC, Clapp JD, et al. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011 Aug 4;365(5):395-409. 3. Moyer VA, Force USPST. Screening for lung cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2014 Mar 4;160(5):330-8. 4. van Klaveren RJ, Oudkerk M, Prokop M, Scholten ET, Nackaerts K, Vernhout R, et al. Management of lung nodules detected by volume CT scanning. N Engl J Med. 2009 Dec 3;361(23):2221-9. 5. Horeweg N, van der Aalst CM, Thunnissen E, Nackaerts K, Weenink C, Groen HJ, et al. Characteristics of lung cancers detected by computer tomography screening in the randomized NELSON trial. Am J Respir Crit Care Med. 2013 Apr 15;187(8):848-54. 6. Horeweg N, Scholten ET, de Jong PA, van der Aalst CM, Weenink C, Lammers JW, et al. Detection of lung cancer through low-dose CT screening (NELSON): a prespecified analysis of screening test performance and interval cancers. Lancet Oncol. 2014 Nov;15(12):1342-50. 7. Scholten ET, Horeweg N, de Koning HJ, Vliegenthart R, Oudkerk M, Mali WP, et al. Computed tomographic characteristics of interval and post screen carcinomas in lung cancer screening. Eur Radiol. 2015 Jan;25(1):81-8. 8. Horeweg N, van Rosmalen J, Heuvelmans MA, van der Aalst CM, Vliegenthart R, Scholten ET, et al. Lung cancer probability in patients with CT-detected pulmonary nodules: a prespecified analysis of data from the NELSON trial of low-dose CT screening. Lancet Oncol. 2014 Nov;15(12):1332-41.

Abstract:
Future implementation of lung cancer screening programmes will require accurate identification of the population who will benefit the most, to ensure that the benefits outweigh the harms [1]. In the USA, the current criteria for Medicare reimbursement [2], for screening are: age 55 to 77, a smoking history of 30 pack-years or more and smoking within 14 years of entry [3]. However, an in-depth analysis of the NLST showed that there was marked variation in individual risk of lung cancer death, with some screened that had only a low chance of benefit: 20% of participants at lowest risk accounted for only 1% of prevented lung-cancer deaths). [4]Conversely, 88% of the prevented deaths were in the 60% of participants that were at highest risk. The only risk prediction model so far utilised in the recruitment of participants into a CT Lung Cancer Screening RCT, is the LLP~v2~ risk model in the pilot UK lung cancer screening trial (UKLS) [5]. The Liverpool Lung Project (LLP) risk model was based on a case-control study [6]. The LLP~v1~ model utilised conditional logistic regression to develop a model based on factors that were significantly associated with lung cancer (smoking duration, prior diagnosis of pneumonia, occupational exposure to asbestos, prior diagnosis of cancer family history of lung cancer (early onset <60 years) and exposure to asbestosis [6]. The multivariable model was combined with age-standardised incidence data to estimate the absolute risk of developing lung cancer. The discrimination of the LLP was evaluated and demonstrated its predicted benefit for stratifying patients for CT screening by using data from three independent studies from Europe and North America [7]. The LLP~v2~ was used to select subjects with ≥5% risk of developing lung cancer in the next five years for UKLS [8]. This method may improve cost-effectiveness by limiting screening to high-risk individuals. The UKLS approached 247,354 individuals in the two pilot sites, 75,958 people (30.7%) responded positively to the screening invitation. Demographic factors associated with positive response were: higher socioeconomic status, age 56-70 years, and ex-smokers. Those from lower socioeconomic groups and current smokers were less likely to respond. 8,729 (11.5%) positive responders were calculated as high risk of lung cancer. The high risk individuals were more often elderly, current smokers, of lower socioeconomic status and males (2.4x females). 4,055 were randomised into the UKLS. Forty two UKLS participants have been diagnosed with confirmed lung cancer, 34 of these were detected at baseline or three months, giving a baseline prevalence of 1.7% which is significantly higher than that reported by the NLST[9]or NELSON [10]trials. To date, 2.1% of all individuals screened have been diagnosed with lung cancer. 36/42 (85.7%) of the screen-detected cancers were identified at stage 1 or 2. Of those with a confirmed cancer, 17/42 (40.5%) were from the most deprived Index of Multiple Deprivation (IMD) quintile. Figure 1 Figure 1: Percentage of UKLS positive responders (n=75,958) with an LLP risk of >5%, by individual year of age. The positive response rate increased steadily with higher socioeconomic status: 21.7% of individuals in the lowest (most deprived) IMD quintile gave a positive response compared with 39.7% in the highest quintile (p<0.001;) (Figure 2). The proportion of individuals with a high LLP risk score decreased with higher socioeconomic status; ranging from 18.2% in the most deprived quintile to 8.3% in the least deprived quintile (p<0.001;). LLP risk were offset by, the socio-demographic spectrum of the individuals attending the clinic, which was in proportion to that of the original approached sample. People recruited into the UKLS trial therefore spanned all IMD quintiles in roughly equal numbers, including a representative proportion from more deprived postcodes. However, in the high risk sub group of individuals invited for screening, there was a trend towards individuals of higher socioeconomic status being more likely to consent to participate in the trial. Figure 2 Figure 2: Impact of socioeconomic status upon initial response rate (lower line), LLP risk (bars) and trial consent rate (upper line). The demographic and response data from the UKLS pilot trial enable specific recommendations to be made regarding the implementation of any future UK-wide lung LDCT screening programme. Such a programme would need to target those most at risk who may be least likely to take up offers of screening (i.e. the most deprived, current smokers, and the over 70s), and women.

Abstract:
Background Lung cancer is the leading cause of cancer death in Japan as well as western countries. To decrease the lung cancer mortality, lung cancer screening using low-dose thoracic computed tomography (CT) may be a promising measure. Several randomized controlled trials (RCTs) are being conducted in the US and Europe to evaluate the efficacy of lung cancer CT screening, and one of those trials, National Lung Screening Trial (NLST), recently reported favorable results. However, the focus of all trials has been the efficacy in smokers.　Adenocarcinomas have increased worldwide in non-smokers, especially in Japan and other Asian countries, and a powerful lung cancer screening modality for non-smokers is also desired. In a recent Japanese cohort study, mortality reduction by thoracic CT screening was suggested even in non-smokers/smokers under 30 pack-years (personal communication). Therefore, we are now conducting the JECS Study (The Japanese randomized trial for evaluating the Efficacy of low-dose thoracic CT Screening for lung cancer in non-smokers and smokers under 30 pack-years). Methods The aim of the JECS study is to assess the efficacy of lung cancer screening tests using low-dose thoracic CT once every five years, compared with chest roentgenography (XP), in people aged 50–64 with a smoking history under 30 pack-years. This study is a multi-regional prospective randomized controlled trial (RCT). The design of the RCT was described elsewhere (Jpn J Clin Oncol 42: 1219-21, 2012). Briefly, participants were recruited from people who ranged from 50-64 years old with smoking history under 30 pack-years, and underwent regular lung cancer screening using chest x-ray in the previous year (this latter requirement may be deleted after 2015). A letter for recruitment to participate in the trial was mailed to each citizen in the target municipalities, who was 50-64 years old with a smoking history under 30 pack-years. The letter explained the eligibility criteria, randomization, follow-up, possible benefit and harm including false-positive, radiation exposure and overdiagnosis. Several meetings were held for those who were interested in the trial for further explanation. People with a history of lung cancer or other malignant diseases as well as a history of thoracic CT screening were excluded. Appropriate written informed consent was completed by each participant who chose to take part in the trial. The participants were randomly assigned into one of 2 groups, a CT group and an XP group. The duration of this screening-follow-up period is 10 years. For the intervention arm, low-dose thoracic CT is provided for each participant in the first year and the sixth year. For the control arm, chest XP is provided for each participant in the first year. All of the participants are encouraged to receive annual routine lung cancer screening using chest XP in the other years. Thoracic CT findings were interpreted by two physicians, based upon the “Low-dose CT Lung Cancer Screening Guidelines for Pulmonary Nodules Management” established by the Japanese Society of CT Screening. A positive rate under 5% is preferred. The interpretation of chest XP findings is performed according to “The Manual of the Lung Cancer Screening” section in the “General rule for clinical and pathological record of lung cancer” published by the Japan Lung Cancer Society. The lung cancer incidence and mortality would be compared. The design of the trial was approved by the Institutional Review Board of the Kanazawa Medical University in 2009, and was registered on the University Hospital Medical Information Network Clinical Trial Registration (UMIN-CTR), Japan in 2011 (registration number: UMIN000005909). The sample size, 17,500 subjects for each arm, is required to detect a 60% mortality reduction after 10 years. At the first step, 3,000 subjects are needed for evaluating QOL and value of contamination of the study. Results As of March 1, 2015, local governments of 20 municipalities in 7 prefectures in Japan permitted that we sent invitation letters for the JECS Study to residents. A letter for recruitment was mailed to each of 9,268 people who were 50-64 years old with a smoking history under 30 pack-years and underwent regular lung cancer screening in the previous year. Of them, 1,812 people (19.6%) sent us a reply letter and wanted to attend one of the meetings for further explanation. One thousand five hundred people actually attended one of the meetings. Finally 1,458 people participated in the JECS Study (15.7% of people who was invited and 97.2% of people who attended a meeting). Of them, 720 people were assigned to CT group and remaining 738 people were assigned to XP group. The screening results of 48 of the 720 people who received low-dose thoracic CT screening (6.7%) were positive, whereas 20 of 738 people who underwent chest XP screening (2.7%) were positive. Until now, three lung cancer patients were diagnosed and 22 patients were under follow-up for the suspicion of having lung cancer in this whole cohort. Comments The results of the NLST demonstrated the efficacy of thoracic CT screening in smokers. However, the efficacy in non-smokers is still completely unknown. Therefore, we started to conduct a randomized trial, the JECS study, to evaluate the efficacy of low-dose thoracic CT screening for lung cancer in non-smokers/smokers under 30 pack-years. This is a first RCT in the world for not heavy-smokers. The preliminary results of this study indicated that 15.7% of the people who received the recruitment letter participated in the RCT. The compliance rate was high in comparison to that in the PLCO or the ITALUNG trial (0.3-7.2%). Over 97% of the 1,500 people who attended a meeting participated in the RCT, which was extremely high. This indicated that the letter for recruitment was effective both for excluding ineligible subjects and for explaining the contents of the trial. The compliance in the preliminary results for this study was very high, and the above RCT appears to be feasible in Japan, if the sufficient budget is obtained. This study was supported in part by the Health and Labour Sciences Research Grant from the Ministry of Health, Labour and Welfare, Japan.

Abstract:
With the ever increasing acceptance of CT screening the need to now minimize harms becomes even greater. One of the harms which occurs with the greatest frequency are “false positives” as they can lead to unnecessary additional work up, sometimes invasive, added cost, and cause anxiety for the person being screened. The term “false positive” is somewhat confusing and is defined differently by different groups. In the computer aided diagnosis domain, it refers to a finding that does not represent a nodule and is mistakenly labelled by the computer to represent nodule. Most frequently this is a blood vessel. Thus, positive results are nodules (often described as actionable based on a size criteria) and false positives are findings not representing nodules. In the clinical domain, when interpreting a CT scan, a positive finding is something that meets a specified definition to be considered a positive result. A positive finding is not something that is inherent to the image but requires certain criteria to be met. Thus, a nodule by itself is not necessarily a positive finding, but must meet certain criteria to be considered positive. Typically it is a non-calcified nodule of a specific size. Thus, in the National Lung Screening Trial the cutoff was at 4 mm, while in I-ELCAP it was at 5 mm for non-calcified nodules. Given a positive result, the confusion now occurs in terms of whether the nodule actually turns out to be a cancer or not. Some prefer to call these cases “false positive” even though they are truly nodules and positive in the sense that they meet the definition of positive based on the CT criteria. Others merely refer to the rate at which positive results occur considering them all positive regardless of their final disposition with the view that imaging does not determine malignancy. Regardless of the linguistics and their potential for causing some confusion, the main concern is to limit the excess amount of work up on those cases which are not cancer. This can be accomplished primarily in two ways. First, to be certain that the population being screened is at high risk for cancer, and secondly, to identify those criteria most associated with cancer and use that in the definition of a positive result. By far, the most dominant of those criteria is size defined either volumetrically or by diameter. An important consideration when defining size cutoffs for positive results, is that the frequency of nodules decreases with increasing size, and the frequency of cancer increases with increasing size. Also, with increasing size of the cancer, the chance for cure decreases. The extent to which all this occurs is not fully known and has many additional considerations. As a start however, and especially in the era of increased scanner resolution, the frequency of positive results would approach 100% if the size criteria is made small enough and the overwhelming majority would be benign. One approach to determining an optimal size criteria is to perform a sensitivity analysis on a screening population balancing the positive rate against what might be considered an acceptable “miss” rate. Using the I-ELCAP database, the frequency of positive results in the baseline round using the 5 mm size cutoff for positive result (any parenchymal, solid or part-solid, noncalcified nodule ≥5.0 mm) was 16% (3396/21 136). When alternative threshold values of 6.0, 7.0, 8.0 and 9.0 mm were used, the frequencies of positive results were 10.2% (95% CI, 9.8% to 10.6%), 7.1% (CI, 6.7% to 7.4%), 5.1% (CI, 4.8% to 5.4%), and 4.0% (CI, 3.7% to 4.2%), respectively. Use of these alternative definitions would have reduced the work-up by 36%, 56%, 68%, and 75%, respectively. Concomitantly, lung cancer diagnostics would have been delayed by at most 9 months for 0%, 5.0% (CI, 1.1% to 9.0%), 5.9% (CI, 1.7 to 10.1%), and 6.7% (CI, 2.2% to 11.2%) of the cases of cancer, respectively. This type of analysis was also performed on the NLST data which using their 4 mm size cutoff had reported a 26.6% positive rate on baseline. The frequency of positive results using the definition of a positive result of any parenchymal, solid or part-solid, noncalcified nodule of 5.0 mm or larger was 15.8%. Using alternative thresholds of 6.0, 7.0, 8.0, and 9.0 mm, the frequencies of positive results were 10.5% (2700 of 25 813, 7.2% , 5.3% , and 4.1% , respectively, and the corresponding proportional reduction in additional CT scans would have been 33.8% , 54.7% , 66.6% , and 73.8% , respectively. Concomitantly, the proportion of lung cancer diagnoses determined within the first 12 months would be delayed up to 9 months for 0.9% (two of 232), 2.6% (six of 232), 6.0% (14 of 232), and 9.9% (23 of 232) of the patients, respectively. The use of the 6 mm size threshold has now gained widespread acceptance in the context of screening having been endorsed by the NCCN, Lung-Rads and I-ELCAP. Nevertheless, it must still be recognized that the tradeoff is the delay in diagnosis of some small cancers for an additional nine months when the next annual screen would occur. While these cancers are unlikely to substantially change in size, the potential for progression is still present and this is the main consideration in balancing against the decrease in positive rate. While size does remain the dominant feature in defining a positive result in this high risk population, there are other approaches that consider additional features of the nodules that also have prognostic significance and may be useful in defining positive results.

Abstract:
With the advent of lung cancer screening (LCS) with low-dose chest CT images, the attention for computer aided tools advances from proof of concept and validation studies to clinical utility. Computer aided image-based diagnosis tools (CAD) for LCS on the initial CT have a primary objective of improved decision making for follow up actions. There are four roles for CAD tools in this context: nodule detection, nodule characterization, nodule growth-rate measurement for malignancy status, and companion diagnostics. The special low-dose CT scan acquired as the primary test in LCS is of lower quality than a traditional clinical CT scan and, consequently, presents a higher challenge to computer analysis methods. Computer aided nodule detection systems address the critical screening task of identifying pulmonary nodules in low-dose CT images. These systems typically identify the location of nodule candidates in the CT images. In general, they detect small sphere like high intensity image regions that correspond to the most common and important finding in LCS. Their performance is related to size and most evaluations are focused on nodules of 4-5 mm or larger. For smaller nodules the false positive rate is much higher. The first of such systems received FDA approval in 2004. There has been significant technology improvement since then with sensitivities in research systems higher than 90% reported in 2007 [1]. In 2012 Zhao et al [2] reported on a study using commercial software on 400 randomly selected cases from the NELSON study. They found that the CAD system could obtain 96.7% sensitivity on nodules greater than 50 mm[3] (4.6 mm) with only 1.9 false positives per scan. In contrast, the double reading achieved 78.1% sensitivity. While the benefit of using computer detection for LCS has been clearly demonstrated and good commercial products are available, there has been little adoption of these methods in recent LCS studies. The second area in which the computer may by useful is in analyzing the images of pulmonary nodule candidates especially with respect to the critical issue of malignant or benign. The classical approach here is to generate some diagnostic features from the appearance of the nodule images and to perform classification from these to determine malignancy. A number of research studies have shown encouraging results; however, these studies have either used non-screening nodules and images, which have a vastly larger size and higher quality or did not separate out the contribution of nodule size, which is highly predictive of malignancy in LCS populations, from the other image features. A recent study [3] has shown that after compensating for size, for LCS CT images, the other image features provide only a moderate amount of additional information. This information is insufficient for a diagnosis by itself but may be used to refine follow up decisions. The measurement of nodule growth rate from two or more CT scans has been shown to be highly predictive of nodule malignancy status [4]. Since at least a second scan is required this method should be considered as a follow up procedure among other clinical follow up methods. The main barrier to clinical implementation of this method is that it requires the computing of the difference of the two CT scans, which is highly dependent on the geometric image quality of each scan. Unfortunately, there exists no agency or process by which this quality is monitored or measured on current scanners and without any scanner calibration imprecise results may occur. Correct use of this method requires careful attention to details. CT scans acquired for LCS also image other critical organs that are at risk for the screening population. Companion diagnostics refers to computer analysis for conditions other than lung cancer from the periodic LCS CT images. Conceptually, this is similar to a blood test where additional conditions may be evaluated from a single patient interaction. Therefore, the automatic risk factor assessment of these additional regions provides additional benefit without requiring additional imaging for the LCS population. Work in this area is still at an early stage. Research targets for automated evaluation reported in the literature include: lung (emphysema and COPD), cardiac (coronary artery calcium, aorta profile and calcium), breast (density assessment), and bone (vertebral body density evaluation). Computer aided methods will inevitably make major contributions to increasing the efficiency and benefit of LCS as they transition from research prototypes to clinical practice. More sophisticated computer algorithms and modern machine learning techniques will greatly improve CAD performance; however, such methods require very large training image datasets. Research studies to date typically involve 100 images examples or less; future algorithm development can greatly benefit by the millions of images that will be acquired with LCS practice. References [1] Enquobahrie A A, Reeves A P, Yankelevitz D F and Henschke C I, “Automated Detection of Small Pulmonary Nodules in Whole Lung CT Scans”, Acad Radiol, 14(5): 579-593, 2007. [2] Zhao Y, de Bock G H, Vliegenthart R, van Klaveren R J, Wang Y, Bogoni L, de Jong P A, Mali W P, van Ooijen P M A and Oudkerk M, “Performance of computer-aided detection of pulmonary nodules in low-dose CT: comparison with double reading by nodule volume”, Eur Radiol, 22(10): 2076-2084, 2012. [3] Reeves A P, Xie Y and Jirapatnakul A, “Automated pulmonary nodule CT image characterization in lung cancer screening”, IJCARS, doi: 10.1007/s11548-015-1245-7, 2015. [4] Reeves A P, “Measurement of Change in Size of Lung Nodules”. In Li Q, Nishikawa R M (ed) Computer-Aided Detection and Diagnosis in Medical Imaging, Taylor & Francis, Chapter 11, 2015.

Abstract:
Effective screening programs should detect all cancers and reduce as much as possible the probability of false-positive results, not representing malignant disease. In lung cancer screening, false-positive low-dose computed tomography (LDCT) results are even more crucial than in other fields, because of the magnitude of risks and costs related to invasive diagnostic examinations, and the need of potentially harmful surgical procedures. Long-term follow-up of nodules ≤ 5 mm at baseline CT has proven that these nodules don’t require additional workup, but for non-calcified nodules between 5 and 10 mm, surveillance of growth is mandatory to identify the relatively few malignant lesions. With the NLST diagnostic algorithm, based on diameter measurement, 24% of subjects had a positive LDCT but 96% of them proved to be false positives, with a positive predictive value (PPV) of only 3.6% at baseline, 2.4 first repeat and 5.2% at second repeat [1,2]. On the contrary, the diagnostic algorithm of Nelson trial, based on the automated assessment of 3D volumetry and doubling time, obtained a 36% PPV and a 99.9% negative predictive value (NPV) [3]. However, in the Nelson trial, where positron emission tomography (PET) was not included in the diagnostic algorithm, the frequency of invasive procedures for benign disease proved to be quite high (27%), and similar to the one observed in NLST trial (24%) [4]. Large meta-analyses have demonstrated the clinical value of PET in the differential diagnosis of undetermined pulmonary nodules detected by spiral CT, with a sensitivity rate of 96-97%, a specificity of 78-82% [5], and accuracy rate reaching 92% with the CT/PET fusion machine [6]. In 2000, our pilot study in Milan was the first screening protocol to include selective use of PET in the diagnostic algorhitm, thus showing that PET may be helpful in the management of CT detected nodules ≥ 7 mm. In the first five years of screening, PET was applied to only 1.4% of spiral CTs, with an overall sensitivity rate of 94%, specificity of 82%, and an accuracy rate of 88% [7,8]. In the Milan pilot trial, the cumulative frequency of surgical procedures for benign disease at 5 years was 15%. The MILD randomized trial has obtained similar results, in terms of frequency and diagnostic accuracy. From 2005 to 2015, a total of 113 PET were applied to 2376 individuals and 12,314 LDCTs, representing 4.8% of all screened individuals in 10 years, and 0.93% of all LDCTs. Excluding lung cancer cases, where PET would have been applied later for staging purposes, the true excess of PET examinations for screening purposes only reached a total 33 exams (1.4% of subjects, 0.3% of LDCTs). The sensitivity rate was 85%, specificity 80%, accuracy 83%, PPV 89% and NPV 74%. Of interest, only 3 patients underwent pulmonary resection for benign disease, out of 66 surgical procedures (5%) performed in the MILD trial. Such a low benign resection rate, is not only due to selective use of PET, but also to the active surveillance programme applied to non-solid lesions in the MILD trial. Beyond differential diagnosis, PET may play a role in prediction of outcome, and identification of indolent lung cancer. We have demonstrated in a previous paper, based on 34 lung cancer patients from the first pilot trial, that PET-SUV value can accurately predict long term survival and identify individuals with 100% 5-year survival [9]. In the MILD trial we have confirmed the value of metabolic profile as a predictor of outcome. The following figure illustrates the 5-year survival of 95 patients, from pilot and MILD trials. Figure 1 The possibility to combine metabolic profile with other biomarkers, such as circulating miRNAs [10], to identify indolent disease will require future investigations, to improve performance and reduce over-diagnosis of LDCT screening. 1 Aberle DR, Adams AM, Berg CD, et al. The National Lung Screening Trial Research Team (2011). Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med 365:395-409. 2 Aberle DR, DeMello S, Berg CD, et al. (2013) Results of the Two Incidence Screenings in the National Lung Screening Trial. N Engl J Med 369:920-31. 3 van Klaveren RJ, Oudkerk M, Prokop M, et al. (2009) Management of lung nodules detected by volume CT scanning. N Engl J Med 361:2221-9. 4 Kramer BS, Berg CD, Aberle DR, Prorok PC. Lung cancer screening with low-dose helical CT: results from the National Lung Screening Trial (NLST). J Med Screen. 2011;18:109-111. 5 M.K. Gould, C.C. Maclean, W.G. Kuschner, et a. l(2001). Owens, Accuracy of positron emission tomography for diagnosis of pulmonary nodules and mass lesions: a meta-analysis. JAMA 285: 914–924. 6 Kim SK, Allen-Auerbach M, Goldin J, et al. (2007) Accuracy of PET/CT in characterization of solitary pulmonary lesions. J Nucl Med 48:214–220. 7 Pastorino U (2010) Lung Cancer Screening. British Journal of Cancer 102: 1681–1686 8 Veronesi G, Bellomi M, Veronesi U, et al. (2007) Role of positron emission tomography scanning in the management of lung nodules detected at baseline computed tomography screening. Ann Thorac Surg 84:959-66 9 Pastorino U, Landoni C, Marchianò A, et al. (2009) Fluorodeoxyglucose (FDG) uptake measured by positron emission tomography (PET) and standardised uptake value (SUV) predicts long-term survival of CT screening-detected lung cancer in heavy smokers. J Thor Oncol 11:1352-6 10. Sozzi G, Boeri M, Rossi M, et al: Clinical utility of a plasma-based miRNA signature classifier within computed tomography lung cancer screening: A correlative MILD trial study. J Clin Oncol 32:768-773, 2014

Abstract:
The complexity of biomarker discovery is amplified by the multitude of platforms on which the biomarker is discovered (mutational sequencing, fluorescence in situ hybridization (FISH), single-nucleotide polymorphisms (SNPs), copy-number variation (CNV) of chromosomes, immunohistochemistry, epigenetics including methylation studies, or microRNA ), and by the material used (tissue, plasma, serum, urine, breath, sputum, effusion). The aim is to define these biomarkers in a way whereby their use is contingent on maximal accuracy, which depends on the ability of biomarker researchers to not only put forth markers with the greatest sensitivity and specificity, but also to be able to validate these biomarkers in a methodologic algorithm that will satisfy regulatory bodies including the Food and Drug Administration (FDA) in the United States as well as other agencies abroad. This lecture will concentrate on novel biomarkers for lung cancer being investigated by the Lung Group and industrial members of the Early Detection Research Network. These biomarkers include autoantibodies, MRM proteomics, micro and lncRNAs, SomaMers, and airway transcriptomics.

Background:
The interaction of PD-1, with its ligand, PD-L1 induces apoptosis of T cells and inhibits cytokine production, allowing tumor cells to bypass immune surveillance. PD-L1 expression on tumor cells can be upregulated via interferon gamma that is secreted by CD8+ cytotoxic T lymphocytes (CTLs) and/or Th1 pathway activation, counterbalancing the Th1/CTL microenvironment. Blockade of the PD-1/PD-L1 immune checkpoint in solid tumors has resulted in durable responses in early phase clinical trials. Moreover, protein expression of PD-L1 by immunohistochemistry (IHC) reportedly predicts patient response to anti-PD-1/PD-L1 therapies. Multiple studies have reported associations of PD-L1 expression with clinicopathological variables in lung adenocarcinomas (ADC), but such studies have produced conflicting results, possibly due to use of different antibody clones and cutoffs and possibly different ethnicities of the cohort. Thus, we correlated PD-L1 expression with clinicopathological and molecular profiles including subtypes of tumor infiltrating lymphocytes (TILs) in a large lung ADC cohort using a cut-off commonly used in clinical trials.

Methods:
PD-L1 (E1L3N, 1:200, CST), CD8 (4B11, RTU, Leica Bond), T-bet (Th1 transcription factor, D6N8B, 1:100, CST), and GATA3 (Th2 transcription factor, L50-823, 1:250, Biocare) IHC were performed on tissue microarrays constructed of 242 resected lung ADC. All cases underwent detailed histological analysis and a subset (n=128) of cases underwent clinical molecular testing. Membranous expression (regardless of intensity) in 5% or more tumor cells was deemed positive for PD-L1 expression. CD8+, T-bet+ and GATA3+ tumor infiltrating lymphocytes (TILs) were evaluated using a 4-tier grading system (0-3).

Results:
Our study cohort consisted of 242 patients with a pathologic stage of 0 in 1 case, I in 188, II in 37, III in 9, and IV in 7. Among those, 38 (15.7%) exhibited PD-L1 expression which was significantly associated with smoking history (p=0.008), large tumor size (p=0.007), solid predominant pattern (p<0.001), high nuclear grade (grade 3, p<0.001), vascular invasion (p=0.012), increased T-bet+ TILs (grade 2, p<0.001) and CD8+ TILs (grade 2, p<0.001), and KRAS mutations (p=0.001). High nuclear grade (p=0.011), KRAS mutations (p=0.004), and increased CD8+ TILs (p=0.005) remained significant predictors of PD-L1 expression in multivariate analysis, while advanced stage (II or higher vs. I, p=0.056) showed a trend towards PD-L1 expression. There was no difference in the 5-year progression free survival (PFS) between the PD-L1 positive and negative patients. In contrast, increased CD8+ TILs showed a borderline significance with favorable outcome (p=0.082), with the 5-year PFS being 87% for the CD8 positive group and 68% for the CD8 negative group, but neither PD-L1 nor CD8+ TILs was a significant predictor of survival by the cox proportional-hazards regression model.

Conclusion:
PD-L1 expression in ADC significantly correlates with KRAS mutations and several clinicopathological signatures of KRAS-mutants, including significant smoking history. The latter may have resulted in development of multiple passenger mutations that serve as neoantigens promoting the Th1/CTL microenvironment. These results suggest that blockade of the PD-1/PD-L1 axis may be a promising treatment strategy to reinstitute the Th1/CTL microenvironment for patients with KRAS-mutated ADC, in which there are currently no available treatment options.

Background:
Programmed cell death protein 1 (PD-1) receptors are members of the B7:CD28 family that interact with PD-1 ligands PD-L1 and PD-L2 to regulate cytotoxic T cell (CTL) tolerance (Freeman, J Exp Med. 2000; Latchman, Nat Immunol. 2001). Successful evasion of transformed cells from host defense is a feature of cancer (Hanahan, Cell 2011). Immune evasion can occur via the engagement of PD-1 with PD-L1 or PD-L2 (Dong, Nature Med 2002). In metastatic non-small cell lung cancer (NSCLC), PD-L1 expression has been associated with increased response to inhibitors of PD-1 (Garon, NEJM 2015). Current adjuvant cytotoxic approaches are associated with a real but small survival increases and significant toxicity. Characterization of PD-L1 expression in resected tumors could guide development of immune checkpoint based adjuvant trials.

Methods:
Microarray analyses were performed to assess gene expression for 320 NSCLC and 15 normal lung resection specimens profiled on the Agilent Whole Human Genome 4x44K 2-color platform. The reference sample used in the experiments was an equal mixture of 258 of the 320 NSCLC samples included in the study. Microarray data was imported into Rosetta Resolver for analysis. The Rosetta Similarity Tool (ROAST) was utilized to find genes correlated to PD-L1 expression. Both PD-L1 and the target gene had to be differentially expressed for sample to be included in computation of correlation. Cosine correlation was used as the similarity metric. Functional genomic analysis on the list of PD-L1 correlated genes was performed using tools available with the DAVID Bioinformatics resources (david.abcc.ncifcrf.gov) Survival analyses based on PD-L1 expression were performed using the Kaplan-Meier method and compared using the log-rank test. Samples with PD-L1 log(ratio) > 0 and p-value < 0.01 were classified as upregulated, samples with p-value>0.01 were classified as unchanged, and sample with log(ratio) < 0 and p-value <0.01 were classified as downregulated.

Results:
The reference level of PD-L1 expression among the subset of normal lung and NSCLC tissue samples was higher compared to levels seen in 503 breast cancer and 149 endometrial cancer tissue samples. Within the 320 NSCLC tissue samples, 174 unique genes are highly correlated with PD-L1 expression (r range= 0.692-0.904). 80 tissue samples (25%) had a PD-L1 log ratio > 0, and 63 tissue samples had large sets of highly correlated genes, a similar prevalence to membranous staining in half the cells in metastatic NSCLC (Garon, NEJM 2015). Functional analyses revealed that the genes significantly correlated with PD-L1 expression were involved in immune and inflammatory response. No significant difference in overall survival was noted (p=.661), but increased PD-L1 expression was clearly not associated with better outcomes.

Conclusion:
Within the NSCLC cohort, there is a group of patients with high expression for PD-L1 and related genes. This group does not have a better prognosis in comparison to those with typical or decreased PD-L1 expression. Due to the relationship between PD-L1 expression and response to anti-PD-1 therapy in metastatic NSCLC, this data and its correlation with other clinical characteristics of the patients can guide the design of adjuvant approaches based on immune checkpoint inhibitors.

Background:
PD-L1 is one of the immune-checkpoint molecules that regulates Th1 immune responses and mediates cancer immune evasion. PD-L1 can be expressed on tumor cells (TC) or tumor-infiltrating immune cells (IC) and expression in both cell types can negatively regulate T-cell function in the tumor microenvironment. The goal of this study was to evaluate the intra-patient heterogeneity and temporal changes in PD-L1 expression and overall immune phenotype in NSCLC using paired synchronous and metachronous tumor specimens.

Methods:
Thirty-nine patients (pts) with NSCLC treated at Memorial Sloan Kettering Cancer Center were evaluated as part of an IRB approved project. Most were former/current smokers (n=30, 77%) and had adenocarcinoma histology (n=36, 92%). 17 pts were KRAS mutant (45%), and 5 were EGFR mutant (13%). Paired synchronous samples were collected from 17 pts with stage IIIA-N2 resected primary lung and metastatic lymph node (met LN) tissue. Paired metachronous samples were collected from 23 pts (including one patient also with synchronous tissue) with at least two metachronous primary/metastatic (n=14) or metastatic/metastatic tissues (n=9). In pts with metachronous samples, 14 (61%) had systemic intercurrent anti-cancer therapy and 9 (39%) had none. PD-L1 expression was assessed by IHC (clone SP142) on TC and IC. CD8 expression was evaluated by IHC using the C8/144 clone. In addition, expression of ~600 immune genes was analyzed by iChip.

Results:
Twenty-five out of 39 tissue pairs were evaluable by PD-L1 IHC (14/17 synchronous, 11/23 metachronous). Among pts with synchronous samples, in the primary tumor, PD-L1 was expressed in <1% of TC or IC in 6 pts, in 1-4% of cells in 5 pts, and in ≥5% of cells in 3 pts. Among those with metachronous samples, in the first collected sample, the PD-L1 expression in <1% of TC or IC was detected in 6 pts, in 1-4% of cells in 2 pts, and in ≥5% of cells in 3 pts. PD-L1 expression was similar across all paired tissues. PD-L1 status at the TC or IC 5% cut-off remained unchanged in all evaluable paired specimens and at the TC or IC 1% cut-off remained unchanged in 80% (11/14 synchronous and 9/11 metachronous) pairs. In both synchronous and metachronous samples, CD8 expression was also similar across paired specimens. The median inter-sample difference in CD8+ T-cell infiltration was 0.5% (95% CI: -0.6% - 3.4%) in synchronous pairs; three pts had a difference >5%. In metachronous pairs, the median difference was -0.4% (95% CI: -1.4% - 0.1%); one pt had a >5% change in CD8+ T-cell infiltration.

Conclusion:
In this study, there was a high agreement in PD-L1 expression and CD8+ T-cell infiltration in both paired synchronous and metachronous NSCLC specimens. The low intra-patient heterogeneity of PD-L1 and CD8 expression in this study suggests any available tissue (e.g. primary or met) may be reliable to assess these markers in NSCLC. Overall immune characterization by gene expression analysis in paired tumor specimens will be presented.

Abstract not provided

Background:
Lung cancer is the largest cause of cancer-related mortality in the developed world. Advances in molecular targeted therapies have led to improved survival in a subset of non-small cell lung cancer (NSCLC) patients. Recently, inhibitors of the programmed cell death receptor 1 (PD1) have proven clinical efficacy in NSCLC. Only a subset of patients respond to PD1 inhibitors, likely reflecting variation in tumor-expression of the PD1 ligand (PD-L1). Many clinical trials have evaluated PD-L1 as a possible predictive biomarker for immune therapy; however several parallel and uncoordinated efforts have led to a high amount of heterogeneity, uncertainty, and ambiguity in the literature around PD-L1 and its use as a biomarker. We aim to investigate the feasibility of PD-L1 biomarker testing in NSCLC using immunohistochemistry (IHC).

Methods:
Cases of stage II, surgically resected NSCLC, adenocarcinoma were identified retrospectively from the archives of the British Columbia Cancer Agency. A tissue microarray (TMA) was constructed with matched primary and metastatic lung tumors. IHC directed towards PD-L1 was performed with 3 different primary antibody clones: E1L3N (Cell Signaling Technology), SP142 (Spring Bioscience), and 28-8 (Dako), each stain was prepared using a unique protocol. Additional cases of NSCLC with available whole-genome sequence were also stained. Staining results were reviewed and scored by intensity of staining and the percentage of positive tumor cells. Cases with positive staining of any intensity in greater than 1% of tumor cells were considered positive (H score > 1). Clinical, pathological, and genomic features of PD-L1 positive cases were reviewed.

Results:
Eighty cases of NSCLC were identified and used in TMA construction. 78 cases had matched lymph node metastases included in the TMA. 29 cases (36%) were positive by the SP142 clone, 19 (24%) by E1L3N, and 27 (34%) by the 28-8 clone. The 3 clones showed concordant results in 61 (76%) of cases, 15 (19%) discordant cases showed low level staining with SP142/28-8 and no staining with E1L3N, 2 (2.5%) cases showed no staining by 28-8 with moderate staining by SP142/E1L3N. Lymph node metastases showed a concordant PD-L1 score in 65 (83%) cases, with no detectable trend in the discordance. Comparison of primary antibodies showed a high rate of concordance (κ=0.68). Exploratory analysis of 6 additional cases with whole-genome and transcriptome data showed no statistical correlation between PD-L1 IHC and tobacco-induced hypermutation signature (p=0.22), or PD-L1 mRNA expression (R[2] = 0.35) by linear regression.

Conclusion:
PD-L1 IHC is reproducible in the setting of an academic reference laboratory. There are small, but potentially clinically relevant, differences between commercially available PD-L1 diagnostic antibodies. Primary tumor PD-L1 status is generally reflective of metastatic tumor PD-L1 status. Molecular correlates of PD-L1 positive cases remain to be elucidated and warrant further investigation.

Background:
PDL-1 expression in NSCLC is frequently associated with response to PD-1 pathway inhibitor therapy. However, it is unclear whether PDL-1 expression status is maintained in nodes and distant metastases and further information is needed on the relationship between expression and patient and tumour characteristics and prognosis

Methods:
TMAs were constructed using 1mm cores (triplicate) of FFPE primary tumour from patients undergoing surgery with curative intent, from N2 nodal tumour (triplicate) and from metastatic NSCLC tumour (duplicate cores or single small sections). PDL-1 protein expression was measured using a validated, automated immuno-histochemical assay using the 28-8 monoclonal antibody (Dako, Carpinteria, CA), with samples categorised as positive when tumour cell membranes were stained to any intensity in 5% of assessable tumour in any core.

Results:
57 paired primary–metastasis cases were analysed: median age 64 years (33-56); 30 male (53%); adenocarcinoma 27 (47%) and squamous cell 15 (26%). Metastatic sites were: brain 27; trachea/bronchus/lung/pleura 17; chest wall/skin 5; lymph nodes 6. Seven cases were synchronous (6 brain) while the median interval between primary and metastasis for other cases was 1.3 years (range 0.2-8.5). Primary and metastatic tumour samples were PDL1 positive for 13 (23%) and 14 (25%) cases respectively and for 44 cases (77%) expression was concordant. Discordance with negative primary and positive metastasis was seen in 7 cases (12%), while 6 cases (11%) were positive in primary and negative in metastasis. Using assay cut offs of 1% and 50%, concordance was 63% and 89% respectively. Eight cases had more than one metastasis analysed and the primary and all metastases were concordant for 6 cases while 2 cases were positive in primary but negative in one of the metastases. TMAs from 518 primary cases were also analysed and data on 123 cases are currently available. Histology was adenocarcinoma 58 (47%) and squamous 38 (31%). Thirty seven cases (30%) were PDL1+. Of the 13 never or light (≤10 PY) smokers, only 2 (9%) were positive. Further data on all cases and matched primary-nodes will be presented.

Conclusion:
PDL1 expression in ≥5% tumour cell was seen in 30% of cases. Concordance of expression in matched primary and metastasis was seen in 77% of cases. These data suggest that if PDL-1 expression status is critical in the decision to treat metastatic NSCLC with a PD-1 pathway inhibitor, then re-biopsy of a metastasis may be warranted.

Background:
Promising results in the treatment of NSCLC have been seen with immunomodulatory agents targeting immune checkpoints, such as programmed cell death 1 (PD-1) or programmed cell death 1 ligand (PD-L1). However, only a select group of patients respond to these interventions. The identification of biomarkers that predict clinical benefit to immune checkpoint blockade is critical to successful clinical translation of these agents. Epithelial-mesenchymal transition (EMT) is a key process driving metastasis and drug resistance. Previously we have developed a robust EMT gene signature, highlighting differential patterns of drug responsiveness for epithelial and mesenchymal tumor cells.

Methods:
We conducted an integrated analysis of gene expression profiling from three independent large datasets, including The Cancer Genome Atlas (TCGA) of lung and two large datasets from MD Anderson Cancer Center, Profiling of Resistance patterns and Oncogenic Signaling Pathways in Evaluation of Cancers of the Thorax (named PROSPECT) and the Biomarker-integrated Approaches of Targeted Therapy for Lung Cancer Elimination (named BATTLE-1). Comprehensive analysis of mRNA gene expression, reverse phase protein array (RPPA), immunohistochemistry, in vivo mouse models and correlation with clinical data were performed.

Results:
EMT is highly associated with an inflammatory tumor microenvironment in lung adenocarcinoma, independent of tumor mutational burden. We found immune activation co-existent with elevation of immune checkpoint molecules, including PD-L1, PD-L2, PD-1, TIM-3, BTLA and CTLA-4, along with increases in tumor infiltration by CD4+Foxp3+ regulatory T cells in lung adenocarcinomas that displayed an EMT phenotype. Similarly, IL-6 and indoleamine 2, 3-dioxygenase (IDO) were elevated in these tumors. We demonstrate that in murine models of lung adenocarcinoma, many of these changes are recapitulated by modulation of the miR-200/ZEB1 axis, a known regulator of EMT. Furthermore, B7-H3 is found to negatively correlate with overall survival and recurrence free survival, indicating a potential new therapeutic target in lung adenocarcinoma in the future.

Conclusion:
EMT, commonly related to cancer metastasis and drug resistance, is highly associated with an inflammatory tumor microenvironment with elevation of multiple targetable immune checkpoints and that is regulated at least in part by the miR-200/ZEB1 axis. These findings suggest that EMT may have potential utility as a biomarker selecting patients more likely to benefit from immune checkpoint blockade agents and other immunotherapies in NSCLC and possibly a broad range of other cancers.

Abstract not provided

Background:
Clinical studies with anti-EGFR agents demonstrate that EGFR TKIs play critical roles in the treatment of non-small cell lung cancer, especially in patients with positive EGFR mutation. Icotinib is an oral, selective EGFR TKIs. Phase 3 study showed that icotinib is non-inferior to gefitinib in treating unselected or EGFR-mutated advanced NSCLC patients as second-line therapy but better safety profile, which provide a rationale to examine icotinib in first-line setting. The objective of this study is to evaluate progression-free survival (PFS), overall survival (OS) and safety of icotinib in chemotherapy naïve NSCLC patients with EGFR mutation.

Methods:
In this phase 3, open-label, randomized study (CONVINCE, NCT01719536), 285 patients (pathologically confirmed NSCLC, positive 19/21 EGFR mutation, treatment naive) will be 1:1 randomized to receive oral icotinib (125 mg, three times daily) or cisplatine (intravenous [IV], 75 mg/m2, day 1) plus pemetrexed (IV, 500 mg/m2, day 1), patients achieving disease control after 4-cycle chemotherapy continue to receive single pemetrexed (IV, 500 mg/m2, day 1) as maintenance therapy until progression. Randomization will be stratified by performance status (0-1/2), smoking status (smoker/non-smoker), disease stage (IIIB/IV), and mutation type (19/21). A total of 228 events would provide 90% power to detect an HR for PFS of 1 at 2-sided significance level of 0.05. Response will be reviewed by both investigator and independent data monitoring committee using Response Evaluation Criteria In Solid Tumors (RECIST version 1.1). Progression Between January, 2013 and August, 2014, 285 patients were randomized and treated at 18 centers from 13 cities in China. The data cut-off was planned at October, 2015 when 228 PFS events were observed in full analysis set (80% maturity). Final results were expected on December, 2015.

Results:
Not applicable

Conclusion:
Not applicable.

Background:
Pemetrexed (P) is the standard of care for non-squamous non-small cell lung cancer (NS NSCLC), whereas epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs), such as gefitinib (G), are the standard of care for advanced NSCLC with EGFR mutations. Clinical and nonclinical studies have demonstrated synergistic effects of EGFR TKIs and P. Based on these observations, the efficacy and safety of G+P was compared with G monotherapy in patients with NS NSCLC positive for activating EGFR mutations.

Methods:
The primary objective of this randomized, multicenter, open-label, parallel-arm, phase 2 East-Asian study was to assess whether G+P prolongs progression-free survival (PFS) versus G alone. Secondary endpoints included overall survival (OS), overall response rate, disease control rate, time to progressive disease, duration of response, and treatment-emergent adverse events (TEAEs). Eligible patients had stage IV NS NSCLC with activating EGFR mutations, were chemonaïve, and had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1. Patients were randomized in a 2:1 ratio (G+P:G). Dosing schedule was concurrent G (250 mg/day) and P (500 mg/m[2] every 3 weeks) in the G+P arm and G monotherapy (250 mg/day) in the G arm. Treatment continued until progression or unacceptable toxicity. The primary endpoint was analyzed after 144 events, which provided 70% power at a 1-sided 20% significance level, assuming a true hazard ratio (HR) of 0.79.

Results:
Between February 2012 and August 2013, 191 patients were randomized and treated (G+P: N=126; G: N=65). Patients were mostly female (64.4%) with a mean age of 62 years; most were never-smokers (67.0%), had confirmed stage IV disease (84.8%), and ECOG PS of 1 (68.6%). Overall, 55.0% had exon 19 deletions, 39.3% had exon 21 L858R mutations, and 5.8% had other activating EGFR mutations. Baseline characteristics were balanced between treatment arms. Patients in the G+P arm received 96.3% and 92.9% of the planned mean dose of G and P, respectively; patients in the G arm received 97.9% of the planned mean dose of G. Median PFS for G+P (15.8 months) was significantly longer than for G (10.9 months); HR=0.68; 95% confidence interval 0.48, 0.96; 1-sided P=0.014; 2-sided P=0.029. OS data are immature and will be reported at study completion. The incidence of grade 3/4 study drug-related TEAEs was significantly higher for G+P (42.1%) than for G (18.5%); P=0.001. The most common study drug-related TEAEs for G+P were diarrhea (44.4%), aspartate aminotransferase increased (41.3%), and dermatitis acneiform and alanine aminotransferase increased (38.1% for each), and for G were diarrhea (47.7%), dermatitis acneiform (43.1%), and dry skin (35.4%). The proportion of treatment discontinuations due to TEAEs was 16.7% in the G+P arm and 9.2% in the G arm; 2 patients (G+P arm) died due to study drug-related adverse events.

Conclusion:
The combination of G+P led to a significant improvement in PFS compared with G monotherapy for East-Asian patients with EGFR mutation-positive NS NSCLC, and met the primary study endpoint. The incidence of grade 3/4 study drug-related AEs was higher for G+P than for G. ClinicalTrials.gov identifier: NCT01469000.

Background:
Bevacizumab (B), an anti-vascular endothelial growth factor (VEGF) monoclonal antibody has been proven to provide additional efficacy benefit in combination with platinum-based chemotherapy for 1[st] line therapy of non-squamous non-small cell lung cancer (NSCLC). In JO25567 study, we observed that bevacizumab in combination with epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor, erlotinib (E) also provided additional 6.3 months median progression free survival (PFS) in advanced EGFR mutation-positive non-squamous NSCLC. To try to understand this additional effect of bevacizumab, we investigated the predictive biomarkers related to angiogenesis comprehensively in JO25567 study. Clnical trials registry number: JapicCTI-111390

Methods:
We evaluated the biomarkers in blood and tissue samples. All samples were collected before E+B or E treatment in JO25567 study. Angiogenesis related ligands and soluble receptors in serum were analyzed by multiplex, bead-based suspension array. Single nucleotide polymorphisms (SNPs) or variable number of tandem repeats (VNTR) of angiogenesis related genes were analyzed by direct sequencing or electrophoresis after PCR for blood sample. VEGF-A concentration in plasma were analyzed by Immunological Multi-Parametric Chip Technique (IMPACT) assay. Messenger RNA of genes related to angiogenesis in tumor tissue were quantitated by multiplex TOF-mass spectrometry (MassARRAY). Immunohistochemistry of neuropilin and exploratory proteomics analysis were planned for surgically resected tumor tissues. PFS were used as an efficacy variable of prediction. Multivariate Fractional Polynomial (MFP) and Subpopulation Treatment Effect Pattern Plot (STEPP) were used for biomarker screening.

Results:
One hundred fifty-two patients were treated with E+B or E in JO25567 study. We analyzed 26 ligands or soluble receptors in 134 serum samples. Follistatin and leptin were identified as potential biomarkers by MFP. The interaction p-value with adjustment of covariates for biomarker and efficacy was 0.0168 for follistatin and 0.0049 for leptin. STEPP suggested that high follistatin related to limited bevacizumab efficacy and low leptin related to higher bevacizumab efficacy. SNPs could be analyzed in 135 blood samples. In 12 SNPs and 1 VNTR of 8 genes, no gene related to bevacizumab efficacy. Plasma samples were collected from 105 patients. Median VEGF-A concentration of E+B group and E group were 18.0 pg/mL and 18.8 pg/mL respectively and was one sixth or more lower than previously reported breast and gastric cancers. Hazard ratio of E+B comparing with E for was 0.23 (95% CI: 0.09-0.60) for low plasma VEGF and was 0.56 (95% CI: 0.26-1.25) for high plasma VEGF. This trend was not consistent with previously reported studies. We analyzed mRNA expression from 24 surgical resected tumors and no predictive value was observed. Because of limited number of surgically resected tumors obtained, we couldn’t proceed exploratory proteomics analysis nor evaluate predictive value of neuropilin expression.

Conclusion:
In this comprehensive predictive biomarker analysis, follistatin and leptin in blood were identified as potential biomarker candidates for E+B therapy.

Abstract not provided

Background:
The most described EGFR mutations are deletions in exon-19 and L858R in exon-21. They constitute approximately 50-90% of total EGFR mutations. Their clinical characteristics and sensitivity to EGFR tyrosine kinase inhibitors (EGFR-TKI) are well-known, given that they are described as classic/sensitizing mutations. Recently, despite the lower frequency G719X in exon-18, T790M and insertions in exon-20, and L861Q in exon-21 were described as uncommon EGFR mutations with known clinical significance having distinct features and EGFR-TKI sensitivity. Meanwhile, several other mutations have been reported and characterized as novel mutations. However, the characteristics and clinical benefit of EGFR-TKI in patients with these rare mutations remains unclear. This study aims to describe the epidemiology and the clinical outcomes of patients with rare EGFR mutations.

Methods:
We retrospectively analyzed 287 patients with advanced NSCLC tested for EGFR mutations at the Brazilian National Cancer Institute from May-2011 to May-2014. Del 19 and L858R were described as classic EGFR mutations (CM). All other EGFR mutations, excluding uncommon mutations with known clinical significance (G719X, T790M, insertions in exon-20, and L861Q), were considered rare EGFR mutations (RM). All samples were formalin-fixed paraffin embedded (FFPE). The best response was considered as the best outcome the assistant physician registered in the medical chart. Time for Treatment Failure (TTF) was considered from the beginning of the treatment until suspension by the medical staff. Overall Survival (OS) was measured from diagnosis to death.

Results:
Of the 287 tested patients, 40 (14%) harbored CM, and 32 (11%) had RM. Only 7 patients harbored uncommon mutations with known clinical significance (1 with G719X and 6 with insertions in exon-20). Of the RM, 18 (56%) were women, 22 (69%) were ever-smokers and only 10 (31%) were never-smokers. RM were associated with smoking as compared to CM (p=0.04). OS was increased in the CM patients (26.4 v 13.4 v 13.7 months,p<0.001; for CM, RM and wild-type, respectively). Sixty patients received EGFR-TKI treatment with 17 harboring RM. Of these 17 RM treated patients, 5 received Erlotinib as first-line, 8 as second/third-line, and 4 as maintenance. When EGFR-TKI was started most patients had PS≤2 (89%). The best response documented was stable disease in 4 (24%) cases. All other 13 (76%) cases had progressive disease. Only three patients received Erlotinib for more than 6 months. Median-TTF was 3.4 months. Median-OS was 17.2 months. In seven cases the mutations have never been described before. In the Erlotinib-treated cohort, RM were associated with worse outcomes (TTF: 13.9 v 3.4 v 3.9 months,p<0.001; OS: 62.9 v 17.2 v 25months,p=0.002; for CM, RM and wild-type, respectively).

Conclusion:
Clinical characteristics of rare EGFR mutant patients differ from classic EGFR mutant. Rare EGFR mutations also conferred little clinical benefit and short TTF with EGFR-TKI treatment. The TTF and OS in rare EGFR mutations were similar to EGFR wild-type patients. Thereby, in this subset of patients the indiscriminate use of EGFR-TKI should be abandoned.

Background:
MET dysregulation is one mechanism responsible for EGFR-TKI (epidermal growth factor receptor-tyrosine kinase inhibitor) resistance in patients (pts) with EGFR mutated lung cancer. INC280 is a potent oral small molecular inhibitor of the c-MET kinase. We conducted a phase I/II study of INC280 plus erlotinib to determine the maximum tolerated dose (MTD), dose limiting toxicity (DLT), pharmacokinetics (PK) and antitumor activity of this combination. Tumor analysis of the EGFR and MET pathways was exploratory.

Methods:
Using a 3 + 3, dose escalation design, INC280 was increased over 5 dose levels (DL) from 100 - 600 mg po bid. Daily erlotinib was given at 100 mg in DL1 and 150 mg in DL 2- 6. DL 6 is a transition cohort from INC280 capsules (600 mg) to tablets (400 mg). Both agents were given for 28 days (1 cycle). Key eligibility included: lung adenocarcinoma with MET expression by a CLIA certified lab, age > 18, ECOG PS of < 2, acceptable organ function, and > 1 systemic therapy for advanced disease.

Results:
18 pts were treated on 6 dose levels. Pt characteristics: median age 59 (range 52-78), M/F (7/11), ECOG 0-1/2 (16/2), MET expression by IHC/FISH/RT-PCR/NGS (6/2/9/1), EGFR mutated tumors (9) and previously treated with erlotinib (12). 17 patients completed at least 1 cycle. One DLT (grade 3 neutropenia) occurred in DL 5 (Table 1). Common drug-related adverse events (AE) of any grade were rash (50%) and diarrhea (45%), fatigue (39%), anorexia and nausea (28% each) and increased alkaline phosphatase, hypoalbuminemia and paronychia (22% each). Drug-related grade 3/4 AE were anorexia, increased amylase or lipase and neutropenia (all 6%). PK analysis revealed that INC280 exhibited a linear PK and no interaction with erlotinib. Of the 17 evaluable patients, 3 (18%) patients had partial responses, 10 (59%) had stable disease, 3 of whom had a minor response (10-29% decrease in target lesion) (Table 1). Eight pts have received treatment for >3 months. Figure 1



Conclusion:
In patients with MET-expressing lung adenocarcinoma, INC280 plus erlotinib is feasible, tolerable and demonstrates anti-tumor activity. The recommended phase 2 doses are INC280 400 mg (tablets) bid plus erlotinib 150 mg daily. Three expansion cohorts have been initiated: 1 - EGFR mutated tumors refractory to an EGFR-TKI, 2 - EGFR-TKI naïve in the first line setting and 3 - WT EGFR that are EGFR-TKI naïve as second or third line therapy. Updated trial results from the expansion cohorts will be presented. NCT01911507

Background:
AZD9291 is an irreversible, mutant-selective epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI) developed to have potency against both EGFR-sensitizing mutations and T790M. In the ongoing Phase I study of AZD9291 (AURA, NCT01802632), the response rate in patients with T790M positive lung cancer with disease progression on previous EGFR-TKI was >60%, with a preliminary median progression-free survival of >10 months. The molecular mechanisms underlying acquired resistance to AZD9291 are currently under investigation.

Methods:
Plasma genotyping was performed on patients from AURA who had progressed on AZD9291 if they had detectable T790M pre-AZD9291, as assessed by tumor or plasma genotyping, and if they had plasma collected at progression available for analysis. Cell-free DNA (cfDNA) was extracted from plasma taken at progression. Droplet digital PCR (ddPCR) was performed for EGFR exon 19 deletions, L858R, T790M, and C797S. For further exploration, next-generation sequencing (NGS) of an amplicon panel was performed on available progression cfDNA. Lastly, targeted NGS was performed on available resistance biopsy specimens.

Results:
Plasma specimens were available following disease progression on AZD9291 from 40 patients with tumors positive for T790M through tumor (33) or plasma genotyping (7). Twenty-six progression cfDNA specimens were positive for an EGFR-sensitizing mutation by ddPCR, and were deemed eligible for initial resistance analysis. Of these, 12 (46%) had no detectable T790M in plasma despite presence of the EGFR-sensitizing mutation, suggesting overgrowth of an alternate resistance mechanism. Seven patients had detectable C797S on ddPCR (27%), all with detectable T790M; of 14 with detectable T790M at resistance, C797S was only detected with EGFR exon 19 deletions (7/9) and not L858R (0/5, p=0.02). Plasma NGS was performed on 12 cases with acquired resistance that were T790M positive pretreatment. Exon 19 deletion/T790M/C797S were detected in four cases, with two of these harboring two different DNA mutations encoding for C797S. One case lost T790M and exhibited HER2 copy number gain (6.3 copies); a tumor biopsy from a separate case underwent aCGH at Institute Gustave Roussy and was also found to have focal HER2 amplification with loss of T790M. Targeted NGS was performed on resistance biopsies from a total of 10 patients from four centers with T790M positive biopsies pre-AZD9291. Six cases maintained T790M, with three harboring exon 19 del/T790M/C797S. In four cases with loss of T790M, one harbored BRAF V600E and one harbored PIK3CA E545K.

Conclusion:
Complementary genomic analysis of plasma and tumor DNA provides insight into the diverse molecular mechanisms of acquired resistance to AZD9291 in EGFR-mutant lung cancer. Our studies show that a majority of cases maintained T790M at resistance, at times acquiring a new C797S mutation in those with EGFR exon 19 deletion. Loss of T790M at progression may be mediated by overgrowth of cells harboring HER2 amplification, BRAF V600E, or PIK3CA mutations. These data highlight the need for investigation of combination therapies to effectively prevent or treat the complexity of drug resistance in EGFR-mutant lung cancer.

Background:
Exon 20 T790M mutation is the most common cause of acquired resistance to first-line epidermal growth factor receptor tyrosine kinase inhibitors (EGFR TKIs). The IMPRESS study (NCT01544179; Phase III, double-blind IRESSA[TM ]Mutation Positive Multicentre Treatment Beyond ProgRESsion Study; Lancet Oncology: in press) reported no statistically significant difference in progression-free survival (PFS; primary endpoint) between gefitinib plus cisplatin/pemetrexed (cis/pem) (G) vs placebo plus cis/pem (P) in patients with acquired resistance to first-line gefitinib (hazard ratio [HR] 0.86; 95% confidence interval [CI] 0.65–1.13; p=0.273; median PFS 5.4 months in both arms) and other secondary endpoints. Among the subgroup analyses performed for IMPRESS, the most noticeable difference was observed by T790M status as tested via plasma circulating free tumor-derived DNA (ctDNA).

Methods:
Patients (age ≥18 years [Japan ≥20 years], chemotherapy-naïve, locally advanced/metastatic NSCLC with an activating EGFR mutation, prior disease progression on first-line gefitinib) from 71 centers (Europe/Asia Pacific) were randomized to G or P (gefitinib 250 mg/day or placebo, plus cis 75 mg/m[2]/pem 500 mg/m[2]). For biomarker analysis, consenting randomized patients provided 10-mL blood samples (at Visit 1 [baseline], 4, 6; then every 6 weeks and at discontinuation) from which to obtain ctDNA. ctDNA levels of EGFR mutations, including T790M, were detected using a quantitative emulsion (BEAMing) digital PCR assay (Sysmex[®]) conducted at a central laboratory (positivity defined as ≥0.02% mutant DNA fraction).

Results:
Data are reported for plasma samples from baseline visits (serial data will be available in the future). Blood samples were available for all 261 randomized patients, of whom T790M status was known for 247 (93.2%): T790M mutation-positive n=142 (57.5%; G=81, P=61) and T790M mutation negative n=105 (42.5%; G=46, P=59). Median PFS for the T790M mutation-positive subgroup was 4.6 vs 5.3 months for G and P, respectively (HR 0.97, 95% CI 0.67 to 1.42, p=0.8829). Median PFS for the T790M mutation-negative subgroup was 6.7 vs 5.4 months for G and P, respectively (HR 0.67, 95% CI 0.43 to 1.03, p=0.0745). See Table for additional study endpoints.

Conclusion:
Following acquired resistance to first-line gefitinib, these data suggest there were two distinct patient populations defined by T790M genotype. For plasma T790M-positive, gefitinib should not be continued when platinum-based doublet chemotherapy is used as second-line therapy. For plasma T790M-negative, continuation of gefitinib in combination with platinum-based doublet chemotherapy may offer clinical benefit, which would require further confirmation in a prospective randomized study.

	IMPRESS subgroup populations (plasma)
	T790M mutation-positive N=142		T790M mutation-negative N=105
ORR, % (G vs P)	28.4 vs 39.3	p=0.282	37.0 vs 27.1	p=0.171
DCR, % (G vs P)	81.5 vs 77.0	p=0.5175	93.5 vs 83.1	p=0.0895
OS, HR (95% CI)*	2.16 (1.26, 3.82)	p=0.0067	0.83 (0.36, 1.85)	p=0.6644
	Plasma BEAMing PCR (compared with tumor), % (n/N)
	Exon 19 Deletions		L858R
Sensitivity	73.8 (124/168)		81.6 (62/76)
Specificity	96.7 (89/92)		95.3 (161/169)
Concordance	81.9 (213/260)		91.0 (224/247)
*OS immature, follow up ongoing G: gefitinib plus cisplatin/pemetrexed; P: placebo plus cisplatin/pemetrexed ORR, objective response rate; DCR, disease control rate; OS, overall survival

Abstract not provided