Virtual Library

Abstract not provided

Abstract not provided

Abstract:
Stereotactic body radiotherapy (SBRT) is a technique, introduced in the late 1990s. SBRT is a method of using single 10-20Gy of high dose and hypofractionated radiotherapy. Recently, many papers have been published on its clinical results, especially in early stage lung cancer. In Japan as JCOG 0403 clinical trial, between July 2004 and November 2008, 169 patients from 15 institutions were registered. 100 inoperable and 64 operable in total 164 patients were eligible. Of the inoperable 100 patients, the % 3-year OS was 59•9% (95% CI: 49•6% - 68•8%). Grade 3 and 4 toxicities were observed in 10, and 2, respectively. No grade 5 toxicity was observed. Of the 64 operable patients, the % 3-year OS was 76•5% (95% CI: 64•0% - 85•1%). Grade 3 toxicities were observed in 5. No Grades 4 and 5 toxicities were observed.　SBRT for stage I NSCLC is effective with low incidences of severe toxicity. This treatment can be considered a standard treatment for inoperable stage I NSCLC. This treatment is promising as an alternative to surgery for operable stage I NSCLC. The current ongoing protocols for lung cancer as JCOG 1408 comparing two different doses、42Gy in 4 fractions and 55Gy in 4 fractions for T1N0M0 and clinically diagnosed lung cancer will be presented. In the world, RTOG 0618 showed a good result of SBRT in the treatment of patients with operable stage I/II NSCLC. RTOG 0915 showed no difference in survival between 48Gy in 4 fractions and 34Gy in a single fraction. In central hilar lung cancer patients, additional attentions are required to avoid serious complications. RTOG 0813 is a Phase I/II study for finding an optimal dose. 55Gy in 5 fractions are their recommended dose. JROSG 10-1 recommended 60Gy in 8 fractions. Optimal dose and fractions are still unknown for centrally located lung tumor. Most updated status of SBRT will also be reviewed.

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract:
Summary of the Lecture “Serendipities of acquired immunity” In 1992, we started working on PD-1 and found that this acts as a brake in the immune system. Then, in 2002, we discovered that PD-1 inhibition could be effective in treating cancer in animal models. After 22 years of study, this idea has borne fruit in a new, breakthrough immunotherapy that is being hailed as a 'penicillin moment' in cancer treatment. I believe that, just as a number of antibiotics developed in the wake of the discovery of penicillin now protect humans against threats of infectious diseases, this discovery will play a leading role in advancement of cancer immunotherapy so that in the future the fear of dying from cancer will cease to exist. Through evolution, vertebrate animals have developed immunity against infection by microorganisms. In the process, they incidentally acquired a sophisticated system for diversifying genomic information by combining gene fragments. It was doubly fortunate that the success in cancer treatment via PD-1 inhibition brought the realization that immunity, a “weapon” against infectious diseases, could also serve as a “shield” against cancer. It has been said that, whereas humankind’s greatest enemies in the 20th century were infectious diseases, cancer is the major foe in the 21st century. It is a pleasant surprise to discover that the acquired immunity system holds the keys to overcoming both of these difficult medical challenges.

Abstract not provided

Abstract:
PD-L1 immunohistochemistry (IHC) has been established as companion or complementary diagnostic assays, each having been developed as a predictive biomarker for specific anti-PD-1/PD-L1 immunotherapies.[1] The Blueprint phase 1 was conducted as a feasibility study to assess the staining (analytical) comparability of four PD-L1 IHC assays (22C3, 28-8, SP142, and SP263) that were developed for their respective immune checkpoint inhibitor therapies.[2] Without correlation with treatment outcome, the study also assessed the putative diagnostic performance of these assays through comparisons of PD-L1 status classification above and below selected expression cutoffs associated with the clinical use of various assays. Serial sections from paraffin blocks of 39 resected non-small cell lung cancers (NSCLC) were stained using assays that were used in the clinical trials, and three experts in interpreting the four respective assays independently assessed the percentages of tumor and immune cells staining positive at any intensity. The results demonstrated that three PD-L1 assays (28-8, 22C3, SP263) showed comparable analytical performance for assessment of PD-L1 expression on tumor cells, while the SP-142 PD-L1 assay appeared to stain less tumor cells compared to the other assays.[2] In contrast, all assays stained tumor infiltrating immune cells, but with poor concordance between assays. The phase 1 study had several limitations: (1) samples were obtained from a commercial source and did not necessarily reflect the real-world samples tested clinically, and (2) the number of pathologists involved in the scoring was small. In addition, a fifth PD-L1 assay (73-10) has since been developed as a potential biomarker for avelumab (EMD Serono/Merck KGaA/Pfizer). The goals of Blueprint phase 2 are: (1) to validate the assay comparability results obtained in Blueprint phase 1 study using real world clinical lung cancer samples and all five clinically used PD-L1 assays (28-8, 22C3, SP142, SP263, and 73-10), (2) to assess the comparability and heterogeneity of PD-L1 assay results in surgical tumor resection, core needle and FNA samples prepared from same tumor, and (3) to assess the concordance of PD-L1 scoring by pathologists from around the world using standard light microscopy vs. digital images accessed by a web-based system. In blueprint phase 2A, 18 participating pathologists, with respective institutional research ethics board approval, contributed unstained serial sections from altogether 81 lung cancer cases that came through routine clinical practice. These included 40 adenocarcinomas, 25 squamous cell carcinomas, 5 poorly differentiated non-small cell carcinoma and 11 small cell carcinomas. The cases included resected tumor (n=20), core/bronchial biopsies (n=20), tumor positive lymph node biopsy/resection (n=20) and cytology cell block (n=21) samples. In blueprint phase 2B, 9 pathologists prepared from 30 freshly resected NSCLC specimens, paraffin blocks of matched resection, core needle and fine needle aspiration samples. Each slide set of 81 cases from phase 2A were stained with the FDA-cleared (28-8, 22C3, SP142) or clinical trial (SP263 and 73-10) PD-L1 assays, in a CLIA-approved immunohistochemistry laboratory. The slides were scored by 24 experienced pulmonary pathologists (IASLC Pathology committee Blueprint phase 2 members),[4] all having received group training on scoring the PD-L1 IHC on tumor and immune cells. PD-L1 stained tumor cells were scored as continuous number (0% to 100%), and placed into 1 of 7 categories (<1%, 1-4%, 5-9%, 10-24%, 25-49%, 50-79%, 80-100%). These categories represent cut-offs that have been used in various immune checkpoint inhibitor trials. All assays were also scored for immune cell PD-L1 staining based on the scoring system developed for the SP-142 assay. As only one set of glass slides is available for each assay, each pathologist was randomly assigned to conduct the scoring using microscope (2 glass assays) or by web-based digital images (3 digital assays). The inter-assay concordance of PD-L1 staining on tumor cells and tumor infiltrating immune cells will be assessed using the mean scores from all pathologists. The large sample size scores should provide more reliable data on their analytical comparability. Inter-pathologist concordance results should provide evidence on reliability of scoring with different cut-points. Importantly, the above concordance results across different sample types should also provide insights on potential variability and feasibility in PD-L1 scoring across different sample types, especially cytology samples. This may then allow for a broad implementation strategy on PD-L1 testing in clinical practice. The results of phase 2A will be presented at the meeting.IASLC Pathology Committee Blueprint phase 2 members: Mary-Beth Beasley, Alain Borczuk, Johan Botling, Lukas Bubendorf, Gang Chen, Lucian Chirieac, Teh-Ying Chou, Jin-Haeng Chung, Sanja Dacic, Fred R. Hirsch, Keith M. Kerr, Mari Mino-Kenudson, Sylvie Lantuejoul, Andre Moreira, Andrew Nicholson, Masayuki Noguchi, Guiseppe Pelosi, Claudia Poleri, Prudence Russell, Jennifer Sauter, Erik Thunnissen, William D. Travis, Ming S. Tsao, Ignacio Wistuba, Murry Wynes, Yasushi Yatabe, Hui Yu. References: IASLC ATLAS of PD-L1 Immunohistochemistry Testing in Lung Cancer. M.S.Tsao, K.M. Kerr, Y. Yatabe, S. Dacic, F.R. Hirsch (Editors), International Association for Study of Lung Cancer (IASLC) Press, 2017 Hirsch FR, McElhinny A, Stanforth D, et al. PD-L1 Immunohistochemistry Assays for Lung Cancer: Results from Phase 1 of the "Blueprint PD-L1 IHC Assay Comparison Project". J Thorac Oncol. 2017 Feb;12(2):208-222. Feng Z, Schlichting M, Helwig C, et al. Comparative study of two PD-L1 expression assays in patients with non-small cell lung cancer (NSCLC). J Clin Oncol 35, 2017 (suppl; abstr e20581)

Abstract:
The therapeutic approach to overcome Tumor-induced suppression of specific T-cell activation by using specific antibodies, which are interacting with regulating pathways, the immune check-points has substantially changed treatment of patients with advanced NSCLC. In particular the development of antibodies inhibiting cytotoxic T-lymphocyte-associated antigen 4 (anti CTL4), programmed death receptor 1 (anti PD-1) or programmed death receptor ligand 1 (anti PDL-1) have shown efficacy in NSCLC. In five randomized trials anti PD-1/anti PD-L1 antibodies have shown significant improvement in overall survival (OS) compared to chemotherapy and an improved safety profile. Efficacy was correlated to PD-L1 expression on tumor cells. However confirmed activity has also been documented in patients with PD-L1 negative tumors. Besides harmonization of different existing tests for assessment of PD-L1 expression identification of novel markers like tumor mutational burden (TMB) might help to identify best benefitting patients. In selected patients with high PD-L1 expression on tumor cells (TPS =/> 50%) monotherapy with the anti PD1 antibody pembrolizumab has demonstrated superiority in response, progression free survival and OS compared to platinum based chemotherapy in first-line treatment. Current randomized trials evaluate the activity of combinations of chemotherapy with checkpoint inhibitors or of immunotherapy combinations compared to chemotherapy and will provide important information about the next steps in the development of immunotherapy in lung cancer. For the future the development of novel immunotherapeutic agents either as monotherapy or in combination with checkpoint inhibitors as well as the understanding of resistance mechanisms and strategies to overcome resistance will be of paramount importance.

Background:
The treatment for stage IV non-small cell lung cancer relies on molecular diagnosis on tumor DNA for guiding systemic therapy. Liquid biopsy provides a timely and non-invasive way of detecting potential therapeutic targets and spares the need of performing a confirmatory biopsy procedure in patients with a positive result.

Method:
We report local experience on epidermal growth factor receptor (EGFR) mutations detected by two platforms (droplet digital polymerase chain reaction (ddPCR) technique and Cobas EGFR v2) on cell free tumor DNA from radiological stage IV lung cancer patients, guiding first- and second-generation EGFR tyrosine kinase inhibitor (TKI) therapy. All these patients who had either plasma EGFR platforms done from Nov 2015 to May 2017 are retrospectively identified and analyzed.

Result:
97 patients identified. 11 out of 39 (28.2%) and 16 out of 58 (27.6%) has detectable EGFR mutation in plasma by using ddPCR and Cobas techniques respectively. Much higher detection rate is noted in never smokers (23/46, 50.0%) compared with chronic/ever smokers (4/41, 9.8%) and is highly statistically significant for difference (p-value: <0.001 ). The types of EGFR mutations detected are as follows: ddPCR method: 4 (36.4%) del 19, 7 (63.6%) L858R. Cobas method: 5 (31.3%) del 19, 9 (56.3%) L858R, 2 (12.5%) G719X. 20 (74.1%) patients were eventually put on 1[st] or 2[nd] EGFR-TKI as first line treatment and, among those, 17 (85%) attained partial response as the best treatment response and the median PFS and OS of is not reached. Among the 56 patients who obtained tissue biopsy/cytology along the course of disease, 6 (10.7%) of which carries histology (3 squamous cell carcinoma, 1 peripheral nerve sheath tumor, and 1 small cell carcinoma and 1 large cell tumor neuroendocrine tumor) that's not normally sent for EGFR molecular testing, as well as 11 (19.6%) patients having NSCLC-NOS histology only.

Conclusion:
ddPCR and Cobas EGFR v2 platforms have similar detection rate and chance of avoiding the need of a confirmatory tissue biopsy. A simple clinical history of smoking status determines the yield of a positive result. Plasma liquid biopsy provides a non-invasive and promising way for treatment initiation in radiological lung cancer.

Background:
Almost all EGFR-mutant lung cancers develop resistance to EGFR tyrosine kinase inhibitors (EGFR-TKIs). Several mechanisms for this acquired resistance have been identified, including development of an EGFR T790M mutation, MET amplification, hepatocyte growth factor (HGF) overexpression, loss of PTEN expression, epithelial to mesenchymal transition (EMT) and transformation to small cell lung cancer.

Method:
We presented a lung cancer patient with EGFR exon 19 deletion who was resistant to EGFR TKI treatment during disease progression. The original lung tumor started to enlarge and bred a secondary growing nodule adjacently. A left upper lobectomy plus systematic mediastinal lymphadenectomy by VATS was performed after a whole body evaluation which had no sign of extrapulmonary abnormalities. The pathological examination, genetic sequencing, and histological staining of E-cadherin and Vimentin were performed on the TKI resistant postoperative specimens.

Result:
The original lung tumor was confirmed to be adenocarcinoma according to the typical morphology and positivity of TTF1 and CK7. The secondary lung tumor has sarcomatoid histology showing a more spindle-like mesenchymal morphology with CK and Vimentin positivity. Next generation sequencing (NGS) test confirmed that the original lung adenocarcinoama retained EGFR exon 19 deletion but was also found T790M mutation in EGFR exon 20. The secondary sarcomatoid carcinoma retained EGFR exon 19 deletion as well. Besides, sarcomatoid carcinoma had genetic mutation on FANCL and BCL2L2, and amplification on CDK4, MDM2, APFRP1, GNAS, CIC, FANCE, Notch4 and AKT2. In addition, in comparison with adenocarcinoma from biopsy and postoperative specimens, the second growing sarcomatoid tumor is strongly positive for Vimentin and nearly negative for E-cadherin, indicating that sarcomatoid histology probably underwent EMT.

Conclusion:
The clinical implication of this case provides significant insights into our understanding that two or more mechanisms might be involved simultaneously in the EGFR TKI resistance. Therefore, if possible, it is necessary to perform tumor biopsies after the development of resistance to identify the best treatment options for patients.

Background:
To investigate the relationship between EGFR mutation status and bone metastasis in lung adenocarcinoma.

Method:
We collected the data of 331 patients with advanced lung adenocarcinoma, and the status of EGFR mutation was detected by RT-PCR. Follow-up these patients was performed.

Result:
331 cases of advanced lung adenocarcinoma, EGFR mutation rate was 52.2%. The incidence of bone metastases (54.9%) and brain metastases (29.4%) with EGFR mutant patients was higher than that wild type. Compared with the number of metastases, EGFR mutant bone metastases were more than 2 and more, and the proportion of wild type single-site metastasis was more. The median survival of bone metastases in advanced lung adenocarcinoma was 14.7 months. EGFR gene mutant group, bone metastasis group median survival in 18.2 months, no bone metastasis group median survival in 21.8 months. In the wild-type group, the median survival of the bone metastases group was 11.5 months, and the median survival in the non-bone metastases group was 15.5 months.

Conclusion:
EGFR mutations in patients with advanced lung adenocarcinoma are more likely to develop bone metastasis and brain metastasis.

Background:
The aim of the current study was to examine the clinical application of BWF samples in detecting epidermal growth factor receptor (EGFR) mutations in a large sample set of NSCLC patients.

Method:
In diagnostic bronchoscopic examinations, before or after biopsy to target lesions, subsequent bronchial washing by saline was performed. Thereafter, EGFR mutation testing for both supernatant and sediment of BWF and histologic tissues was performed via amplification refractory mutation system real-time PCR (ARMS RT-PCR) assay.

Result:
A total of 127 cases of histologic and corresponding BWF samples of patients underwent bronchoscopy for suspected lung malignant tumor lesions on chest radiography were successfully obtained. Of these, 72 cases were pathologically confirmed to be NSCLC based on forceps biopsy samples and EGFR mutations were identified in 26 cases. In 70 of 72 cases, the results of EGFR mutation status were concordant for BWF and histologic samples, and the concordance rate was 97%. In 13 cases that were not pathologically diagnosed with NSCLC with forceps biopsy samples but other samples, five cases (38.46%) with EGFR mutated-type were detected by BWF. The overall EGFR mutation concordance rate between supernatant and sediment specimens was 100%. The detection of EGFR mutations with supernatant/sediment of BWF samples showed a sensitivity of 86.5%, a specificity of 100%.

Conclusion:
This study demonstrates a clear comparison of supernatant/sediment of BWF samples and histologic tissues for EGFR mutation testing with largest clinical samples to date. Both supernatant and sediment of BWF samples showed high credibility and concordance via highly sensitive PCR analysis. BWF is considered a simple, rapid and effective alternative for histologic samples in EGFR mutation testing.

Background:
Osimertinib is a third-generation, CNS active EGFR-TKI that potently and selectively inhibits both EGFR-sensitizing and EGFR T790M resistance mutations in non-small cell lung cancer (NSCLC). We report interim clinical and molecular diagnostic testing results from a predefined interim analysis of the ongoing ASTRIS study (NCT02474355).

Method:
Patients (pts) received osimertinib 80 mg once daily. Eligible pts had advanced NSCLC that had progressed on prior EGFR-TKI therapy and with a T790M mutation determined by local validated molecular test, WHO performance status (PS) 0−2, acceptable organ and bone marrow function and no history of interstitial lung disease or QTc prolongation. Asymptomatic, stable CNS metastases were permitted. The primary efficacy outcome was overall survival; other outcomes included local test methods, specimen type, EGFR mutations identified, investigator-assessed response rate (RR), progression-free survival and time to treatment discontinuation. Safety data are also reported.

Result:
From 18 Sept 2015 to the planned 3 Nov 2016 data cut-off (DCO), 1217 pts received osimertinib 80 mg once-daily across 14 countries with a median age 64 yrs (27–92 yrs), 67% female, 61% White, 37% Asian, 87% WHO PS 0/1, 44% prior chemotherapy, 45% prior radiotherapy. All pts tested positive for T790M; T790M was reported alone in 185 pts (15%). The most common testing methods were PNA-Clamp 317 pts (27%), Qiagen therascreen 254 pts (22%), and Roche cobas 204 pts (17%). Exon 19 deletion was the most common co-occurring mutation with T790M (57%), followed by L858R (27%). Tissue or cytology specimens were used in 720 pts (59%), plasma in 433 pts (36%), and other specimens in 64 pts (5%). At DCO, the median duration of exposure was 3.8 months (<1–13.2 months) with a median follow-up time of 4.1 months (<1−14 months). In pts evaluable for response, the investigator-assessed RR was 64% (569/886; 95% CI 61, 67). Adverse events (AEs) leading to dose modification and treatment discontinuation were reported in 122 pts (10%) and 54 pts (4%), respectively. Serious AEs were reported in 165 pts (14%) and AEs leading to death in 28 pts (2%).

Conclusion:
ASTRIS is the largest reported global study of osimertinib in pts with T790M-positive NSCLC identified by a wide array of molecular testing methods and from various specimen types. Considering this breadth of T790M testing, the clinical activity of osimertinib is like that observed in the clinical trial program and no new safety signals were identified.

Background:
Leptomeningeal carcinomatosis (LC) is a detrimental complication of non-small cell lung cancer (NSCLC). Osimertinib is the current standard therapy for pretreated EGFR T790M-positive NSCLC patients. However, the efficacy of osimertinib for these patients with LC is unknown.

Method:
Retrospective case series of 5 patients with pretreated EGFR T790M-positive NSCLC who developed LC and received osimertinib therapy in an Expanded Access Program was reviewed. We evaluated the clinical outcomes of these patients.

Result:
Four female patients and one male patient (age, range 51-67) with EGFR T790M-positive NSCLC and LC received osimertinib therapy at a starting dose of 80 mg/day. EGFR T790M mutation was detected in three re-biopsied specimens and two plasma samples. Four patients had Eastern Cooperative Oncology Group performance status (PS) ≧ 2. One patient received whole-brain radiotherapy after commencing osimertinib therapy. Osimertinib dose escalation to 160 mg/day or 160 mg every other day was administered to 3 patients who did not respond to standard dose therapy. Radiologically decreased leptomeningeal enhancement was seen in 3 out of 4 evaluable patients, and improvement of clinical symptoms was recorded in 2 patients. Two patients died of aspiration pneumonia, and one died of hypoxic respiratory failure of unknown cause. Osimertinib therapy is ongoing in two patients at 80 mg/day for 9 and 10 months, respectively, with good tolerability.

Conclusion:
Osimertinib is well tolerated even in patients with poor PS. Clinical benefits were seen in some patients, and the optimal dose should be explored.

Background:
Patients with epidermal growth factor receptor (EGFR) gene-mutated NSCLC initially show substantial clinical benefit from EGFR Tyrosine-Kinase Inhibitors (TKIs), but will ultimately develop resistance with a median progression-free survival of 9-15 months. However, the type and timing of TKI-resistance cannot be predicted and several mechanisms may occur simultaneously/subsequently during TKI-treatment. We present a patient case with advanced non-small cell lung cancer (NSCLC) of adenocarcinoma subtype (ADC) with EGFR-mutation and Erlotinib-treated Different mechanisms of TKI-resistance were detected in tumor biopsies during treatment time.

Method:
The patient was a 49 year-old, previously healthy, Caucasian male with metastatic ADC. A diagnostic biopsy from hepatic metastasis, cytology from metastatic pleural effusion at first progression during TKI-therapy and a biopsy from new liver metastasis at second progression were analyzed by histology, immunohistochemistry and targeted next-generation sequencing (NGS) of hot-spot mutations in 50 cancer-related genes (Ion AmpliSeq Cancer Hotspot Panel v.2, Ion Torrent, Thermo Fisher Scientific).

Result:
CT-scans revealed tumor in the right upper lobe with mediastinal infiltration and multiple pulmonary and hepatic metastases, stage T4N2M1b. Diagnostic liver biopsy revealed ADC (mucin-producing, CK7- and TTF1-positive epithelial acinar structures), which concomitantly harbored an EGFR exon 19-mutation (p.E746_A750delELREA) and a previously unreported 2 bp microdeletion in the fibroblast growth factor receptor 3 (FGFR3; p.D785fs*31) gene. The patient received first-line Erlotinib but progressed after 7 weeks with metastatic pleural effusion, in which transformation to small cell lung cancer (SCLC) and maintenance of EGFR-mutation and FGFR3-mutation was identified. The progression was treated with standard Carboplatin-Etoposide regimen for SCLC together with Erlotinib continuation. The second progression 7 months later was a new liver-metastasis with persistence of the original EGFR- and FGFR3-mutated ADC-phenotype and additional emergence of the Erlotinib-resistant T790M EGFR-mutation. The patient rapidly deteriorated and deceased.

Conclusion:
Thus, in this advanced EGFR-mutated NSCLC rapid onset and heterogeneous mechanisms of TKI-resistance occurred at different times of metastatic disease: 1. Concomitant FGFR3-mutation prior to and during TKI-treatment as potential intrinsic resistance-mechanism; 2. Transformation to SCLC at 1st progression during TKI-therapy; 3. Acquisition of T790M EGFR-mutation at 2nd progression. This suggests a continuous variation of TKI-resistant cells’ genetic and phenotypic behavior. Therefore, re-biopsies are important to provide the current status of the disease and better define subsequent treatment options. “Liquid biopsies” may potentially help identify heterogeneous genetic resistance-mechanisms; however assessment of mechanisms such as SCLC-transformation needs tissue biopsies

Background:
Advanced non-small cell lung cancer(NSCLC) studies indicated a potential association between chemotherapy efficacy and circulating tumor cells (CTC) counts in peripheral blood. The icotinib efficacy and circulating tumor cells (CTC) counts in non-small cell lung cancer remain unknown. The aim is to investigate association between the icotinib efficacy and CTC counts in advanced NSCLC patients.

Method:
A total of 74 advanced NSCLC patients consented to provide 5ml of peripheral blood before systematic therapy, and divided into two groups (those with high CTC counts and those with low CTC counts) based on the patients′ median CTC counts. All the patients were treated with icotinib, and the icotinib efficacy and prognoses were compared.

Result:
The treatment efficacies were 46.88% (15/32) and 23.81% (10/42) for the low CTC group and the high CTC group, respectively. The median overall survival was 22.0 months (95%CI: 19.6-28.7 months) for the low CTC group and 18.3 months (95% CI: 15.3-25.4 months) for the high CTC group. The median progression-free survival was 11.6 months (95% CI: 8.7-15.6 months) and 7.2 months (95% CI: 3.4-8.7 months) for the low group and the high CTC group, respectively.

Conclusion:
The CTC counts can be used as a important biomarker for therapy monitoring the icotinib effect on patients with advanced NSCLC. Efficacy and prognosis of icotinib and CTC counts were considered important, and the CTC counts could be used to predict the efficacy of icotinib and prognosis of advanced NSCLC. The change in CTC count levels can be used as a biomarker for evaluating the prognosis of patients with advanced NSCLC.

Background:
EGFR exon 18 E709X mutation is only reported in small case numbers of non-small cell lung cancer (NSCLC) in the literature, and their influences on the effectiveness of icotinib have not been fully understood. The study of this aim is to investigate the efficacy of icotinib in patients with NSCLC that carrying EGFR exon 18 E709X mutation.

Method:
Three cases of EGFR exon 18 E709X mutations were retrospectively analyzed until the progress of the disease or the emergence of the side effects and clinical efficacy was observed after months of followed-up.

Result:
The median progression-free survival (PFS) of three cases with EGFR exon 18 E709X (E709_T710>D, E709A, E709K plus L858R) was 3.1 months, patients with complex mutations showed a better PFS than those with single mutations (7.2 months vs. 2.7 months, P=0.225). Clinical efficacy of icotinib with advanced NSCLC harboring EGFR exon 18 E709X mutation (ORR: 66.67%, DCR: 66.67%).

Conclusion:
Icotinib is effective in patients with exon 18 E709X mutations. Patients with complex mutations benefited more from icotinib than those with single mutations.

Background:
Lung cancer is the leading cause of cancer death in men and women worldwide. Brain metastasis (BMs) of NSCLC is the most important cause of death. This study aimed to explore the association of epidermal growth factor receptor (EGFR) mutations and brain metastases (BMs) in non-small cell lung cancer (NSCLC).

Method:
We analyzed 50 NSCLC patients with BMs and 50 match-paired NSCLC patients with no brain metastases (NBMs). The EGFR mutation status of primary lesions was detected using the amplification refractory mutation system (ARMS) polymerase chain reaction.

Result:
The BMs patients had a higher frequency of EGFR mutations than the NBMs patients (52.0% vs. 22.0% respectively, P＜0.001), in both adenocarcinoma (60.5% vs 30.6%, P=0.003) and squamous carcinoma (37.5% vs. 0%, P=0.04). The incidence of BMs in patients with EGFR mutations was higher than in patients with wild-type EGFR (70.3% vs. 38.1%, P=002).

Conclusion:
NSCLC patients with BMs had a higher incidence of EGFR mutations and those with mutant EGFR had a higher frequency of BMs. EGFR mutations may promote brain metastasis growth of NSCLC.

Background:
Brain metastases are often seen in Eastern-Asian lung adenocarcinoma in which most common oncogenic driver mutations are EGFR sensitive mutations or EML4- ALK fusions, occurring in about 60% patients. Elucidation of EGFR and ALK-driven mutation prognostic factors in patients with brain metastases has important clinical implications.

Method:
Lung adenocarcinoma harboring EGFR mutations or ALK fusions were studied for their occurrence of brain metastases and relevant prognostic factors.

Result:
Of 602 lung patients with lung adenocarcinoma, 516 had EGFR mutations and 86 had EML4-ALK fusions. EGFR-mutation patients had much more brain metastases than ALK-fusion patients, whether at first diagnosis (40.1% vs 28%, P=0.03) or followed up two years (63.8% vs 42.7%, P＜0.01). Brain metastasis is a significant indicator for poor survival in the EGFR-mutation patients (P＜0.001), but not in the ALK-fusion patients (P=0.79). In the EGFR-mutation patients, late occurred brain metastases （after one year） predicted better survival (P=0.001). Meanwhile, female or young age was associated with better survival in the brain metastatic patients with ALK fusions (P＜0.05). Moreover, multivariate analysis indicated that tyrosine kinase inhibitor (TKI) treatment, no symptomatic brain metastases and cranial radiotherapy were the significant indicators for better survival in both EGFR-mutation and ALK-fusion patients (P＜0.05). Although there was no difference in overall survival between EGFR mutation and ALK fusion patients with brain metastases (P=0.23), in the TKI-treating subgroup ALK-fusion patients had longer survival time than EGFR-mutation patients(NR vs. 36 months, P=0.15). Furthermore, in the EGFR-mutation patients with brain metastases erlotinib was associated with better survival compared with gefitinib as first-line treatment (38 months vs 34 months, P=0.03)

Conclusion:
EGFR-mutation and ALK-fusion lung adenocarcinoma patients with brain metastases have distinct clinical characteristics and prognostic factors. However, while doing targeted treatment ALK-fusion indicated better survival than EGFR-mutation in the patients with brain metastases.

Background:
Patients with non-small-cell lung cancer (NSCLC) experience high disease burden due to many cancer-related symptoms (eg, cough, dyspnea, pain, and fatigue). Decreasing tumor burden may reduce/delay symptoms and favorably impact global health-related quality of life (HRQoL). Dacomitinib is an irreversible, small-molecule inhibitor of EGFR/HER-1, HER-2, and HER-4 tyrosine kinases. In a global, multicenter, randomized, open-label phase 3 study (NCT01774721) for first-line treatment of NSCLC, dacomitinib improved the primary objective of progression-free survival per independent radiologic review (median, 14.7 vs 9.2 months; hazard ratio, 0.59; 95% confidence interval [CI], 0.47–0.74; P<0.0001) over gefitinib. Median duration of treatment was longer with dacomitinib than with gefitinib (67 vs 52 weeks, respectively).[1] A secondary objective was to explore HRQoL. Here, we report the impact of dacomitinib and gefitinib treatment on core lung cancer symptoms.

Method:
Patients were randomized 1:1 to receive oral dacomitinib (45 mg) or gefitinib (250 mg) once daily. Disease-related symptoms were measured using the European Organisation for Research and Treatment of Cancer Quality-of-Life Questionnaire–Core 30 (QLQ-C30) and Lung Cancer 13 (QLQ-LC13). Scores were summarized by the mean and 95% CI for each group and plotted over 30 cycles and at the end of treatment; the number of cycles (n=30) chosen for this analysis was not prespecified. Mean changes from baseline (cycle 1, day 1 [C1D1]) were reported for each group.

Result:
Between 9-May-2013 and 20-March-2015, 452 patients were randomly assigned to dacomitinib (n=227) or gefitinib (n=225). Baseline scores were similar between treatment arms. On-study completion rates were high, with >90% of patients answering all questions for most treatment cycles. Statistically significant improvements from baseline (95% CI excludes 0; no adjustment for multiplicity) for most cycles were seen in fatigue, pain, dyspnea, and cough in both arms. Improvements were reported as early as C1D8. Clinically meaningful improvements (≥10 points score change) were recorded for pain in chest (23/30 cycles) and cough (28/30 cycles) with dacomitinib and for cough (22/30 cycles) with gefitinib; hence, improvements appear to be more frequent with dacomitinib. Symptom burden at end of treatment was generally higher than during treatment. As treatment duration was longer with dacomitinib, key lung cancer symptom improvements were seen for a longer time in patients treated with dacomitinib.

Conclusion:
Dacomitinib, along with gefitinib, demonstrated favorable clinical benefit and improvements in key NSCLC symptoms. These findings are important when considering choice of therapy. Reference 1. Mok T, et al. J Clin Oncol. 2017;35(Suppl):abstract LBA9007.

Background:
EGFR mutation positive (EGFRm) NSCLC patients commonly progress in the CNS. We reviewed CNS disease development and its impact on resource utilization and outcomes in EGFRm patients who received first-line EGFR TKI.

Method:
Methods: A retrospective review was completed of all advanced EGFR+ patients referred to the BC Cancer Agency and treated with a first/second-generation EGFR TKI from 2010-2015. Baseline characteristics, systemic treatment and CNS management was collected. Comparison was made between the CNS positive (CNS+) and negative (CNS-) patients’ health resource utilization from median time of CNS+ diagnosis to death/last follow-up (8.9 m) and for no CNS metastases group, 9 months preceding death/last follow-up, using the Chi squared test and t-test.

Result:
499 patients were identified. Baseline characteristics: Female 68%, median age 66 (30-90), adenocarcinoma 89%, Asian 51%, never/former/current smoker 67/24/9%, exon 19/21/other/not specified 57/37/3/3%. 229/499 patients (46%) developed CNS+; 39% at diagnosis, 61% during the course of disease. Systemic treatment: first-line EGFR TKI 95%, first-line platinum doublet 5%; 40% (202/499) second-line EGFR TKI 24%, second -line platinum doublet 56%, single agent chemo 13%, osimertinib 7%, third-line therapy 47%. CNS+ management: surgery+/-WBXRT 13%, WBXRT alone 73%, SRS+/-WBXRT 5%, no CNS directed therapy 9%. Median time from diagnosis to CNS+ was 7.6 m. Median time from development of CNS+ diagnosis to death was 8.9 m. Median OS was 24m in CNS+ versus 33m in CNS- (p<0.001).

Events in 9m preceding death or last follow-up (consistent with median time from CNS+ to death)	No CNS Metastases n=270	CNS Metastases n=229	p value
Average number of clinic visits	8.53	12.71	<0.001
Average number of hospitalizations	0.43	0.76	<0.001
Average number of CNS imaging investigations	0.52	2.65	<0.001
Average number of ER visits	0.03	0.14	0.001
Palliative Care Unit admission	22 (8%)	22 (10%)	0.64
Hospice admission	9 (3%)	43 (19%)	<0.001

Conclusion:
The incidence of CNS+ in EGFRm patients is high and associated with increased Health Resource Utilization. Prevention or delay of CNS+ with newer systemic therapy options may result in decreased interactions with health care providers, which may translate into lower resource utilization and cost savings.

Background:
Both first and second generation EGFR tyrosine kinase inhibitors (TKIs) have efficacy in NSCLC with activating EGFR mutations (EGFRM+). Previous studies showed a differential benefit of second generation TKIs based on mutational subtypes but failed to show a difference in overall survival (OS). We aimed to characterize the patterns of use and outcomes of first and second generation TKIs and describe any differences with mutation subtype in the real world setting.

Method:
A retrospective review of all advanced EGFRM+ NSCLC patients treated with TKIs between the years 2010-2015 at the British Columbia Cancer Agency was performed. All time to event analyses were performed from date of diagnosis of metastatic disease. Multivariate regressions were performed to examine for associations of OS, treatment and mutation subtypes.

Result:
500 patients were eligible for analysis: 283 patients had an exon 19 deletion (del19), 185 had an exon 21 L858R mutation and 32 were not specified or have mutational variants such as G719var. Patient characteristics in the del19 vs. L858R group were similar: 69%/66% were female, 66%/71% were never smokers, 90%/89% were adenocarcinoma, 20%/20% had CNS metastases at diagnosis, 85%/84% had de novo metastatic disease and 41%/37% received ≥2 lines of therapy (all p>0.05). The del19 cohort had less Asians (46% vs 58%, p=0.02) and were younger (median age 63 vs. 69, p=0.02) compared to L858R group. In the del19/L858R cohorts, 81%/19% and 84%/16% received a first and second generation TKI respectively. 43% of patients receiving a second generation TKI required a dose reduction to manage the toxicity and one patient discontinued the medication. OS in the entire cohort was 26 months, with the del19 group surviving longer compared to the L858R cohort (27 vs. 22 months, p<0.01). In multivariate analysis, factors associated with improved OS were del19 (HR0.7, p<0.01 95%CI0.6-0.9), treatment with a second generation TKI (HR0.6, p=0.01 95%CI0.5-0.9) and absence of CNS metastases (HR0.7, p<0.01 95%CI0.5-0.9). First line treatment with a second generation TKI was associated with better OS compared to a first generation TKI (HR0.6, p=0.04 95%CI0.3-1.0). This was statistically significant only in the del19 subgroup (HR0.4, p=0.04 95%CI0.2-1.0).

Conclusion:
Use of a second generation TKI in EGFRM+ advanced NSCLC is associated with improved OS in multivariate analyses controlling for other prognostic factors. This was significant in the entire group and in the del19 cohort, supporting the use of mutational subtype to guide therapy decisions.

Background:
First-line epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKI) offer an advantage compared to doublet chemotherapy in progression free survival, tolerability, and quality of life (QOL) in EGFR-mutated advanced non-small cell lung cancer (NSCLC) patients. In Taiwan, gefitinib, erlotinib and afatinib are all reimbursed as first-line therapy. It provides a rare opportunity to investigate factors associated with the extent of symptoms and QOL improvement in real-world patient population.

Method:
We conducted a multicenter, prospective, observational study to evaluate the QOL and disease-related symptoms at baseline, 2, 4, and 12 weeks in EGFR-mutated advanced NSCLC patients with first-line EGFR TKI treatment. QOL was assessed by the instrument of Functional Assessment of Cancer Therapy-Lung questionnaire (FACT-L) and Treatment Outcome Index (TOI) derived from FACT-L. Symptoms assessment was evaluated by the Lung Cancer Subscale (LCS). The mean change from baseline of QoL and LCS score was analyzed by paired t-test.

Result:
The average age was 65.1± 12.5 (range 31.4–92.9) years old, with a larger proportion of females (62.6%) than males, and more never-smokers (74.0%) than ever-smokers. Patients were treated with gefitinib 250 mg (72.4%), erlotinib 150 mg (18.9%) or afatinib 40 mg (8.7%). For FACT-L, the total score was increased by 4.0 ± 15.49 at week 2, 4.9 ± 18.31 at week 4, and 4.1 ± 20.44 at week 12 (all p<0.001). Similarly, increased TOI of 2.4 ± 11.61 (p<0.001), 3.1 ± 13.48 (p<0.001), and 2.4 ± 14.35 (p=0.009) were observed at week 2, 4, and 12, respectively. For LCS, it was slightly increased by 1.7 ± 4.59 at week 2, 2.0 ± 5.48 at week 4, and 1.9 ± 5.35 at week 12 (all p<0.001). In general, subgroup analyses indicate that patients with more than 2 metastatic sites and ex-smokers were associated with clinically meaningful improvement in terms of LCS (change in LCS ≥ 2 points). Other subgroup analyses show that patients with characteristics such as at least 3 metastatic sites, ex-smoker, PS of 1, and treatment with gefitinib group, were associated with improved QOL in terms of TOI and FACT-L.

Conclusion:
In EGFR mutated NSCLC patients, first-line EGFR-TKI treatment was associated with improvement in disease-associated symptoms and QOL. Patients with 2 or more metastatic sites and ex-smokers were associated with symptoms and QOL improvement. In addition, PS of 1 and treatment with gefitinib were associated with clinical meaningful improvement in global QOL.

Background:
Osimertinib (AZD9291) is a third-generation EGFR TKI specific against T790M resistance mutations in patients with metastatic EGFR-mutant (EGFRm+) NSCLC after prior first-line TKI therapy. We aimed to evaluate the clinical efficacy of osimertinib in patients treated under the AZD9291 Early Access Program (EAP) at our local institution.

Method:
This retrospective study included 57 patients who were enrolled on the AZD9291 EAP between Jul 2015 and Nov 2016 after being tested T790M+ on tumor (by direct Sanger sequencing or Roche COBAS EGFR mutation test v2) and/or plasma (cfDNA) specimens (by Lung Colon Panel v2 or ARMS PCR). Of these patients, 52 were treated with osimertinib. Tumor responses were independently assessed by radiologists using RECIST 1.1 criteria. DOR, PFS and OS were estimated by Kaplan-Meier method.

Result:
The median age at diagnosis was 58 years (range: 35-76), 80.7% patients were non-smokers, 89.5% had ECOG 0-2, 96.5% had adenocarcinoma subtype and 87.8% had either EGFR exon 19/ exon 21 mutations. Median line of therapy when osimertinib was administered was third-line (range 2nd – 9th), and 30 (53%) had brain metastasis at osimertinib initiation. RR by RECIST 1.1 was 46% (95% CI 32.2 – 60.5%) (4 CR + 20 PR) with median DOR of 8.7 months. With median follow-up of 6.2 months from osimertinib initiation, median PFS was 10.3 months (95% CI 7.52 to 15.87 months). For the 52 patients treated with osimertinib, EGFR T790M mutation was tested on the following specimens: tumor-only (n=43), plasma-only (n=25), and both (n=17). In patients with paired tumor/plasma T790M testing, 4/17 had concordant results (RR 75%), while 13 patients with discordant results [T790M+ in 8 tumor-only: RR 25% (95% CI 3.2% - 65.1%) or in 5 plasma-only: RR 40% (95% CI 5.3% - 85.3%)] had overall RR 31% (95% CI 9.1% – 61.4%). ECOG status was associated with PFS by univariable analysis, with higher ECOG 2-4 associated with shorter PFS (HR=6.54, 95% CI: 2.48 to 17.26; p<0.001) than ECOG 0-1. Line of osimertinib treatment and presence of brain metastases at osimertinib initiation were not associated with clinical outcome.

Conclusion:
Osimertinib is effective in patients with advanced EGFR T790M+ NSCLC after progression on prior EGFR TKI, regardless of presence/ absence of CNS metastasis or line of therapy. Notwithstanding ongoing optimization of plasma-based assays, T790M tumor-plasma discordance in our patient cohort is likely a reflection of overall burden of T790M subclones, and may represent a potential negative predictive biomarker of response to osimertinib.

Background:
To detect the mutation abundance and sites of epidermal growth factor receptor (EGFR), and to investigate their influence on the therapeutic efficacy of EGFR-tyrosine kinase inhibitors (EGFR-TKIs) in patients with advanced non-small cell lung cancer (NSCLC).

Method:
A total of 2,417 NSCLC patients with EGFR gene mutations were retrospectively analyzed using an amplification refractory mutation system (ARMS). Of 861 patients, 407 patients had exon 19 deletions and 454 patients had exon 21 (L858R) site mutations of the EGFR gene. Next we analyzed the mutation abundance of lung adenocarcinoma patients who received EGFR-TKI therapy and complete follow up. 194 patients diagnosed with stage IIIB or stage IV lung adenocarcinoma with an ECOG score of 0-3 were enrolled. The primary endpoint was to determine associations between progression-free survival (PFS) and mutation abundance or mutation sites after EGFR-TKI therapy. The secondary endpoint was to evaluate the objective response rate and effects when EGFR-TKI was administered as the first-line treatment, as well as risk factor analysis for PFS.

Result:
Of the 194 enrolled patients, the median PFS was 9.3 months (95% CI, 8.2–10.8 months). The PFS was significantly different with EGFR gene mutation abundance after EGFR-TKI therapy (P = 0.014). The median PFS was significantly longer when the cut-off value of EGFR mutation abundance of exon 19 or exon 21, exon19 was > 26.7% and 61.8%, respectively. For patients who received EGFR-TKI as first-line treatment, the median PFS was significantly longer in the high mutation abundance group than in the low mutation abundance group (12.7 m vs 8.7 m, P = 0.002). However, there was no significant correlation with gender, age or smoking status and PFS.

Conclusion:
The PFS benefits were greater in patients with a higher abundance of exon 19 mutations in the EGFR gene after EGFR-TKI treatment.

Background:
Plasma detection of EGFR T790M mutations in circulating tumour DNA (ctDNA) of advanced lung cancer patients with acquired resistance to EGFR tyrosine kinase inhibitor (TKI) has been proposed as alternative to tumor re-biopsy. This national validation study across Canadian centres aimed to establish the sensitivity and specificity of plasma detection of T790M as a clinical test using digital droplet (dd)PCR and next generation sequencing (NGS) assays.

Method:
Canadian patients at 7 centres undergoing screening for ASTRIS (NCT02474355) were invited to participate in this companion blood-based study. Patients with acquired resistance to EGFR TKI consented to collection of blood samples and demographic data. Samples were analysed using ddPCR and/or NGS platforms available at 4 molecular diagnostic laboratories across Canada. Concordance between the results of plasma T790M assayed in these 4 laboratories with reference tissue/plasma testing conducted for ASTRIS was assessed.

Result:
63 patients participated; the median age was 64 years (range 31-87), 69%(40/58) were Asian; 55%(33/60) were male. All patients received prior EGFR TKI, 17%(10/60) also received prior chemotherapy. Reference testing for EGFR T790M for ASTRIS eligibility identified positive T790M(+) results for 31(49%), negative(-) for 30(48%) and indeterminate(i) results for 2(3%) patients. One laboratory tested all 63 patient samples using both ddPCR and NGS (Oncomine Lung cfDNA assay), another laboratory tested 18 samples using ddPCR and NextSeq, a third tested 10 samples using ddPCR and COBAS EGFRv2, and a fourth tested 6 samples using Ion Torrent PGM. A total of 188 tests were performed including 91 by ddPCR, 87 NGS and 10 COBAS assays. Combining test results for each patient, 60%(38/63) of patient plasma samples were T790M+, 23(37%) were T790M-, and 2(3%) were inconclusive. Of 31 patients with reference T790M+ results from ASTRIS, 23(74%) had T790M detected in plasma, 6(19%) did not (T790M-), and 2(7%) had indeterminate (T790Mi) plasma results. For 30 patients with T790M- reference results from ASTRIS, 13(43%) had plasma T790M+ and 17 plasma T790M- results. The 2 patients with T790Mi by reference testing both had T790M+ results from plasma. Altogether, 47%(15/32) of patients deemed to have T790M-/i tumours by reference testing were found to have T790M+ results by plasma in this multicentre study. Combining results from both tissue and plasma testing, 73%(46/63) of study patients had T790M+ results.

Conclusion:
Plasma ctDNA testing in this multicentre Canadian study identified a significant number of additional patients eligible for osimertinib therapy beyond routine biopsy tissue testing for EGFR T790M.

Background:
Brunei is indifferent to its neighbouring South East Asia countries whereby the highest cancer mortality rate is attributed to lung cancer. EGFR mutations are present in a subset of patients with non-squamous NSCLC that predict a favourable response to EGFR tyrosine kinase inhibitor (TKI) therapy. Thus far, the prevalence of EGFR mutations in Brunei patients with NSCLC remains unknown. This study was conducted to characterise non-squamous NSCLC patients with EGFR mutations in Brunei.

Method:
Retrospective data collection from clinical case notes of patients with non-squamous NSCLC diagnosed from January 2010 to March 2017 was undertaken at The Brunei Cancer Centre to determine the frequency of EGFR mutations and correlation with clinico-pathological variables. The progression-free survival (PFS) of EGFR mutated patients on EGFR TKI and the overall survival (OS) of patients with mutated and wild type EGFR were estimated and compared using Kaplan-Meier curves and log-rank tests.

Result:
EGFR mutation status was evaluable in 71 out of 191 lung adenocarcinoma cases. The overall mutation frequency was 43.7 % with a mean age of 63.3 years. EGFR mutations were significantly more common in female (71%) and never smokers but 33.3% of patients with EGFR mutations were current smokers. The most prevalent EGFR mutation was exon 21 point mutation (L858R) (43.3%) followed by exon 19 deletion (40.0%). 87% of patients with EGFR mutations received EGFR TKI therapy and the objective response rate (ORR) was 65.2%. The median PFS for patients on EGFR TKI was 7 months and there was no significant differences between the two most common mutations. The median PFS of patients treated with EGFR TKI upfront and after first line chemotherapy was 8 and 6 months (p=0.045) respectively. Although not statistically significant (p=0.232), our result showed a trend of improvement in median OS for EGFR mutated patients (29 months) compared to wild type (17 months).

Conclusion:
This is the first study to reveal the clinical characteristics of patients with EGFR mutated non-squamous NSCLC in Brunei. The prevalence of EGFR mutation in our study is high and comparable to other Asian NSCLC patients. Interestingly, PFS rate on EGFR TKI observed in our population was inferior to published studies despite the high ORR and this warrants further exploration. Nevertheless, EGFR mutation analysis should be performed in all patients with non-squamous NSCLC to identify the subset of patients who may benefit from EGFR TKI.

Background:
EGFR mutation plays a dominant role in the precise treatment of non-small cell lung cancer (NSCLC), and EGFR-TKIs has been recommended for patients with positive EGFR-sensitive mutation as a standard regimen in clinical practice. In China, application of EGFR-TKIs without knowing EGFR mutation status has been a common phenomenon due to various reasons including the vast territory, uneven distribution of medical resources, differences level of testing technology and others. Therefore, we prospectively conducted a real-world investigation to understand the actual situation of EGFR testing in Northern China, and identify the underlying causes affecting EGFR detection, in order to provide references to improve the standardized treatment.( NCT02620657)

Method:
The patients with IIIB/IV NSCLC who were firstly diagnosed or postoperative recurrence between 2014-1-1 and 2014-12-31 in 28 research centers of Northern China were analyzed. The primary endpoint was testing rate，the secondary endpoints were factors affecting EGFR testing, EGFR mutation status, detection methods and the survival outcomes of patients.

Result:
Among 2809 patients, 2250(90.78%) were adenocarcinoma, 208(7.40%) were squamous carcinoma, 51(1.82%) were other pathologic types. Testing rate was 42.54%(1195/2809) and was significantly related to city level (first-tier cities vs. new first-tier cities vs. second-tier cities vs. third-tier and above cities : 69.04% vs. 38.08% vs. 34.05% vs. 14.11%, P < 0.001), smoking status (never smoking vs. ever smoking vs. smoking: 45.42% vs. 51.10% vs. 33.37%, P<0.001), ECOG PS(0 vs.1vs.2vs.≥3:47.93%vs. 44.48vs.34.89%vs.20.37%, P=0.011), pathological type (adenocarcinoma vs. squamous carcinoma: 44.94% vs.19.23%, P=0.003) and medical insurance situation (social basic medical insurance vs. new rural cooperative medical insurance vs. own expense: 44.98% vs. 36.49% vs. 29.55%, P=0.001). EGFR sensitive mutation rate was 46.44%, the most common subtype was 19Del(42.16%), followed by L858R(40.00%), Exon 20 insertions(1.62%) and other subtypes(16.20%). The most common methodology is ARMS(63.77%), the second common one is DNA sequencing(5.36%). The 1-year and 2-year survival rate in patients receiving EGFR testing was 73.6%and 51.9%, compared with 64.3% and 43.7% respectively in patients without EGFR testing.

Conclusion:
There were regional differences in EGFR testing rates among IIIB/IV NSCLC patients in Northern China. The intention of doctors and patients, medical insurance coverage and differences technical level are major factors affecting the testing rate of EGFR. Approaches should be taken to improve the situation, such as strengthening the training, expanding the coverage of medical insurance, and relying on commercial gene detection companies, and further standardize the molecularly pathological diagnosis and treatment of NSCLC.

Background:
Sensitizing Epidermal Growth Factor Receptor (EGFR) mutations confers a better prognosis and predicts a favorable response to treatment with EGFR tyrosine kinase inhibitors (TKIs) in patients with advanced non-small cell lung cancer (NSCLC). We sought to determine the prognostic utility of a comprehensive panel of clinical, immune and biochemical markers in EGFR-mutant advanced NSCLC patients treated with first-line EGFR TKIs.

Method:
A retrospective analysis was conducted on a cohort of NSCLC patients treated at our institution. Only patients who met the following criteria were included: NSCLC with EGFR mutations, diagnosed with Stage 4 disease at initial diagnosis or incurable disease recurrence, and received first-line EGFR TKI treatment. Statistical analysis on descriptive and survival data was performed using IBM SPSS Statistics, version 20.

Result:
We identified 74 patients based on predefined criteria. The mean age of the patients was 63.8 years with 38 (51.4%) female patients and 36 (48.6%) male patients, 51 (68.9%) were never smokers and 39 (52.7%) harbored EGFR exon 19 deletion. 15 (22.7%) out of 66 patients with available data had high serum LDH (median: 456 U/L; normal range: 250-580 U/L). In addition, 19 (25.7%) had brain metastasis, 12 (16.2%) had liver metastasis and 12 (16.2%) had adrenal metastasis at stage 4 diagnosis. These represent the descriptive data of the variables that were later identified to carry prognostic significance. The median PFS and OS for this cohort of patients were 12 and 30 months respectively. On our multivariable Cox-PH regression analysis, old age (>=60; HR = 2.35; 95% CI = 1.18-4.69, p-value = 0.015), female gender (HR = 2.05; 95% CI = 1.05-4.00, p-value = 0.036) and high LDH levels (HR = 2.45; 95% CI = 1.04-5.81, p-value = 0.041) retained their association with unfavorable PFS and only high LDH levels (HR = 2.84; 95% CI = 1.25-6.49, p-value = 0.013) is an independent prognostic indicator of unfavorable OS. Total leukocyte count, hemoglobin levels, absolute neutrophil, lymphocyte and platelet counts at diagnosis were not statistically significant for PFS or OS.

Conclusion:
We identified old age, female gender and high serum LDH levels as independent prognostic predictors of progression-free survival on multivariable analysis. The strong association between elevated LDH and unfavorable overall survival is in line with recent published studies. Further studies confirming the prognostic utility of these markers are necessary to potentially develop a prognostic scoring system that can guide risk stratification in patients with EGFR mutations treated with EGFR TKIs.

Background:
The efficacy of first-line afatinib was demonstrated for epidermal growth factor receptor (EGFR)-mutant non-small cell lung cancer (NSCLC) in large randomized trials, and it has been approved and reimbursed in South Korea since year 2014. This study evaluated clinical outcomes of afatinib in real world practice.

Method:
Patients treated with first-line afatinib for advanced EGFR-mutant NSCLC in Samsung Medical Center (Seoul, South Korea) from October 2014 to December 2016 were included into the analyses.

Result:
In total, 165 patients were analyzed. One hundred fourteen had deletion in exon 19, 37 L858R, and 14 patients had uncommon EGFR mutations including 4 de novo T790M. Median progression-free survival (PFS) was 19. 1 months (95% CI: 12.3- 25.9 months). There was difference in median PFS according to EGFR mutation type: deletion in exon 19 (19.1 months), L858R (15.8 months), de novo T790M (4.7 months), and uncommon EGFR excepting for T790M (not yet reached) (P = 0.01). Though 112 patients (67.8%) had to reduce afatinib dose from 40 mg per day into 30 mg or 20 mg per day due to adverse events, it did not impair its efficacy in terms of PFS (23.5 months in a reduction group vs. 12.4 months in no reduction group). Among 29 patients with evaluable follow-up brain MRI for non-irradiated brain metastatic lesions, significant response was documented in 22 cases (75.9%). Out of 20 patients who were biopsied again at disease progression (excluding one case with de novo T790M), T790M appeared in 7 the repeat-biopsied specimens (35.0%).

Conclusion:
In the real practice in South Korea, first-line afatinib showed comparable or better efficacy data compared with previous clinical trials.

Background:
Not all the patients with EGFR-mutant non-small cell lung cancer (NSCLC) could benefit from EGFR-TKIs and both the predictive and prognostic markers remain undetermined. Previous study demonstrated that EGFR activation could up-regulate the PD-L1 expression and then contribute to immune escape, indicating the critical role of PD-L1 in EGFR-driven lung tumors. Therefore, we aimed to investigate the predictive and prognostic significance of PD-L1 expression in EGFR-mutant NSCLC treated with EGFR-TKIs. Characterization of PD-L1 expression in this populations were also explored.

Method:
We analyzed a cohort of 73 NSCLC patients with EGFR mutations. PD-L1 expression were assessed by immunohistochemistry and staining > 5% of tumor cells were considered as positive. Published studies that assessed the predictive or prognostic value of PD-L1 expression in EGFR-mutant NSCLC patients treated with EGFR-TKIs were included.

Result:
19 (26.0%) patients had positive PD-L1 expression. Positive PD-L1 expression was significantly associated with non-adenocarcinoma histology in this population (P = 0.028). Two cases with positive PD-L1 expression had KRAS mutation while none of those with negative PD-L1 expression harbored KRAS mutation (11.8% vs. 0%, P = 0.109). Five publications had reported both the predictive and prognostic value of PD-L1 expression in EGFR-mutant NSCLC patients treated with EGFR-TKIs. The pooled analysis including the current study showed that overall survival (OS) was significantly shorter in PD-L1 positive group than in PD-L1 negative group (HR, 1.92; 95% CI, 1.15-3.22; P = 0.01). However, positive PD-L1 expression was not correlated with progression-free survival (PFS: HR, 0.82; 95% CI, 0.51-1.30; P = 0.39).

Conclusion:
Positive PD-L1 expression tends to correlate with non-adenocarcinoma histology and KRAS mutation in EGFR-mutant NSCLC. Positive PD-L1 expression was significantly associated with better OS instead of PFS in EGFR-mutant NSCLC patients treated with EGFR-TKIs.

Background:
Limited data are available on treatment experience of advanced lung adenocarcinoma with unknown EGFR gene status (UN-EGFR-GS). We studied the demographic profile and treatment outcomes of advanced lung adenocarcinoma patients, which the EGFR gene status was unknown.

Method:
Retrospective study of patients with UN-EGFR-GS advanced lung adenocarcinoma over a 5-year period at a tertiary hospital in China. Patients diagnosed with stage IIIB or IV were included for analysis during 2010 and 2014.

Result:
In total, 140 patients were included, females and males constituted 48.6% (n=68) and 51.4% (n=72), respectively. The mean age was 58y. Among the 113 patients, 79 were non-smokers and 61 were smokers. Majority of patients had stage IV disease (90.0%), only 14 patients had stage IIIB disease. Most of the patients performance status (PS) score were 0 or 1 (n=128). Ninety-two patients were advanced stage when diagnosed and 48 patients were relapsed disease once received surgical resection. Nineteen patients received adjuvant chemotherapy, which were not relapsed in 6 months after finishing last cycle. In the140 patients, eighty-six had EGFR-TKIs ever, the rest were had chemotherapy only and never received EGFR-TKIs ever. The common regimens of first-line treatment were pemetrexed plus platinum (n=44), gemcitabine plus platinum (n=40), and paclitaxel plus platinum (n=18). Other drugs included docetaxel, novelbine. Twenty patients received EGFR-TKIs as first-line treatment. The commonest second-line treatment was oral EGFR-TKIs (n=47). Nineteen patients received EGFR-TKIs as third-line treatment and four received EGFR-TKIs as fourth-line treatment. At the end of follow-up (2016-7), 104 patients were dead and 36 patients were alive or lost follow up. The median survival of this whole cohort was 19.9m.Those who had a chance taken EGFR-TKIs lived longer than never; the median overall survival was 20.5 months and 16.4 months, respectively. EGFR-TKIs was associated with an improvement of overall survival, respectively in univariate analysis (p=0.027) and multvariate analysis (p=0.007). PS score was also associated with survival, respectively in univariate analysis (p＜0.01) and multvariate analysis (p＜0.01). The overall survival was not associated with sex (p=0.336), and smoking (p=0.414), TNM stage (p=0.565), age (p=0.100). The overall survival was also not associated with the chemotherapy drugs used in the first-line treatment.

Conclusion:
EGFR-TKIs have efficacy improved overall survival in patients with UN-EGFR-GS advanced lung adenocarcinoma, compared with conventional chemotherapy only. Oral EGFR-TKIs appear to be useful for this group of patients.

Background:
In patients with advanced EGFR mutation-positive (EGFRm+) NSCLC, first-line afatinib significantly improved PFS and objective response rate (ORR) versus platinum-doublet chemotherapy in the phase III LUX-Lung (LL) 3 and LL6 studies, and PFS, time-to-treatment failure (TTF) and ORR versus gefitinib in the phase IIb LL7 study. Here, we present post-hoc analyses of efficacy, safety and patient-reported outcomes (PROs) in afatinib long-term responders (LTRs) in LL3/6/7.

Method:
Treatment-naïve patients with stage IIIb/IV EGFRm+ NSCLC who were randomized to 40mg/day afatinib in LL3/6/7 and remained on treatment for ≥3 years were defined as LTRs. In these patients, we assessed efficacy and safety outcomes, as well as PROs measured using the European Organization for Research and Treatment of Cancer (EORTC) Quality of Life (QoL) Questionnaire and the EQ-5D™ health status self-assessment questionnaire; these included scores on the EORTC Global Health [GH]/QoL scale (0–100), EORTC Performance Functioning scale (PF; 0–100), EQ Visual Analogue Scale (VAS; 0–100) and EQ-5D UK utility scale (EQ UK utility; 0–1).

Result:
In LL3/6/7, there were 24/229 (10%)/ 23/239 (10%)/ 19/160 (12%) afatinib-treated LTRs; 6/9/14 remained on treatment at time of analysis. Baseline characteristics were similar to the overall study populations, except for the proportion of women (LL3/6 only [LTRs versus overall]: 92/78% vs 64/64%) and Del19+ patients (LL3/6/7: 63–79% vs 49–58%). In LL3/6/7, 4–11% of LTRs had brain metastases at enrolment. Median (range) duration of treatment in LL3/6/7 LTRs was 50 (41–73)/56 (37–68)/42 (37–50) months. Due to few deaths, median OS could not be estimated. Median follow-up for OS in LL3/6/7 was 64.6/57.0/42.1 months. ORR among LTRs in LL3/6/7 was 70.8% (complete response: 4.2%; n=1)/78.3% (13.0%; n=3)/89.5% (5.3%; n=1). The frequency of afatinib dose reductions due to treatment-related AEs, and the frequency/duration of subsequent treatments were similar to the overall LL3/6/7 populations. In afatinib-treated LTRs in LL3/6/7, PROs appeared stable between ~Week 24 to ~Week 160, with slight improvements after ~3 years afatinib treatment versus scores at the start of treatment.

Conclusion:
In LL3/6/7, 10%–12% of afatinib-treated patients were LTRs. Afatinib was well tolerated among these patients. Long-term treatment was independent of tolerability-guided dose adjustment or presence of brain metastases at time of enrolment, and had no detrimental impact on subsequent treatment. In afatinib-treated LTRs, PROs were not negatively affected by long-term treatment, and were slightly improved after ~3 years of treatment versus scores at treatment initiation.

Background:
Overcoming acquired resistance to EGFR TKIs remains challenging, identification of actionable genetic alterations conferring drug-resistance can be helpful for guiding the subsequent treatment decision. In our previous study, we performed mutational profiling in a cohort of 83 NSCLC patients using targeted next generation sequencing (NGS) and identified TET2 mutations in 12% of patients. This study aims to further explore the role of TET2 mutation in acquired EGFR-TKIs resistance in NSCLC cell lines with mutant EGFR.

Method:
CRISPR/Cas9 system was used to knock out TET2 gene in NSCLC cell line PC-9. Quantitative real-time PCR was performed to detect mRNA levels of TET2 gene, and western bot was performed to detect the expression levels of TET2 protein, thus cell line of TET2 silencing can be selected; positive for Annexin V and/or propidium iodide by flow cytometry was used to determine apoptotic rate. Bisulfite sequencing PCR (BSP) and real-time PCR for detection of EGFR promoter methylation status and mRNA expression.

Result:
By applying combined analyses of gene expression, apoptotic rate, and biochemical analyses of EGFR inhibitor responsiveness, we identified homozygous loss of TET2 to segregate EGFR-dependent cells. We show that in EGFR-dependent cells, TET2 loss partially uncouples mutant EGFR from downstream signaling and activates EGFR, thereby contributing to erlotinib resistance. Figure 1



Conclusion:
Our study is expected to provide new ideas and put forward corresponding relevant tactics for patients with secondary resistance to the first-generation EGFR TKIs.

Background:
CNS (central nervous system) involvement is common in advanced NSCLC (non-small cell lung cancer). Osimertinib has shown activity in CNS in preclinical studies and phase II, III trials (AURA2, AURA3). We reported the efficacy of CNS metastases from an open label, multinational, multicenter, real world treatment study (ASTRIS, NCT02474355) in patients with T790M-positive advanced NSCLC who have progressed on or after prior epidermal growth factor receptor-tyrosine-kinase inhibitors (EGFR-TKI) therapy.

Method:
Patients with T790M-positive (from tissue, plasma or other fluids) advanced NSCLC received osimertinib 80mg once daily. Of the 88 patients who were enrolled in ASTRIS at Yonsei Cancer Center, 10 patients who did not have baseline brain workup and 15 patients without CNS metastases at the beginning of study were excluded from this analysis. A subgroup analysis was conducted in patients with CNS metastases at the baseline, as assessed by neuroradiologist, to define CNS overall response rate (ORR), duration of response (DOR), and progression-free-survival (PFS) by RECIST(Response Evaluation Criteria in Solid Tumors) v1.1. The CNS full analysis set (cFAS) included patients with ≥ 1 measurable and/or non-measurable CNS metastasis present on baseline scan; the CNS evaluable for response set (cEFR) included only patients with ≥ 1 measurable CNS metastasis.

Result:
Among the 63 patients who had CNS metastases at baseline, fifty-four (61.4%) patients were included in the subgroup analysis as cFAS, except for the 9 patients who did not follow up brain image during ASTRIS. In patients without brain metastases at the time of initiation of osimertinib (n=15), no experienced CNS progression during ASTRIS. CNS ORR was 81.3% (95% confidence interval [CI] 73.2-89.4) in the cEFR and 40.7% (95% CI 30.4-50.9) in the cFAS. In the cEFR and cFAS, median CNS DOR was “not reached” vs 40.1 weeks (95% CI 36.95-43.25). The median CNS PFS was not reached in both cFAS and cEFR. CNS ORR of 33.3%(95% CI 11.7-64.9) and 42.2% (95% CI 28.9-56.7) were observed for patients with CNS metastases within 3 months brain radiation and without prior radiation or ≥ 3months brain radiation, increasing to 75.0% (95% CI 28.9-96.6) and 83.3%(95% CI 54.0-96.5) respectively, for patients with measurable CNS disease only. CNS ORR of T790M-positive patients in tissue and plasma were 37.5%(95% CI 21.1-57.4) and 46.4% (95% CI 29.5-64.2) in the cFAS, vs 100%(95% CI 55.7-100.0) and 70.0%(95% CI 39.2-89.7) in the cEFR.

Conclusion:
Osimertinib had good CNS efficacy irrespective of radiation history in T790M-positive advanced NSCLC.

Background:
Osimertinib is an oral, potent, irreversible 3[rd] generation EGFR tyrosine kinase inhibitor(TKI) approved for the treatment of T790M positive non-small cell lung cancer (NSCLC) patients who failed 1[st] or 2[nd] generation EGFR TKIs. Interstitial lung disease (ILD) is a rare complication with osimertinib, accounting for 1-3%. Recently, relatively high incidence of transient asymptomatic pulmonary opacities (TAPOs) which are different from ILD has been described (Noonan et al, JTO 2016). However, its clinical implication has not been fully determined yet.

Method:
We retrospectively analyzed 75 EGFR mutant NSCLC patients treated with osimertinib at Samsung Medical Center. Serial CT findings were reviewed by radiologist (Dr. HY Lee) and TAPO was classified according to its radiologic pattern. We also analyzed the correlation of TAPO with clinical outcomes.

Result:
Among 74 patients, TAPO was found in 15 (20.3%). The median time to TAPOs development was 23.5 weeks (1 – 72 weeks) and the median duration of TAPOs was 6.0 weeks (5 – 24 weeks) during continued osimertinib treatment. The most common radiological patterns of TAPO include cryptogenic organizing pneumonia and/or simple eosinophilic pneumonia (SEP). There was no significant difference in patient characteristics between TAPO positive and negative group. The duration of exposure to osimertinib is longer in TAPO positive than negative group (25.2 months vs 14.0 months, p=0.016 ). The progression free survival (PFS) and overall survival (OS) was numerically longer in patients with TAPO positive than negative group (PFS : 15.0 m vs 12.5 m, p= 0.201/ OS : 37.0 m vs 24 m, p=0.155)

Conclusion:
TAPOs are frequently observed with osimertinib treatment and may be mistaken for isolated pulmonary progression or ILD. Given the lack of serious clinical deterioration, it is reasonable to continue osimertinib with regular CT scan follow-up. For further clinical validation of TAPOs, long-term and large studies are warranted.

Background:
The overall survival (OS) of patients with EGFR mutation positive (EGFRm) lung cancer has changed dramatically due to the combined benefit of targeted systemic therapy, local management of oligometastatic disease and incorporation of judicious use of radiotherapy. We reviewed EGFRm NSCLC patients who were diagnosed with CNS metastasis to determine the prognostic importance of the initial CNS presentation.

Method:
A retrospective review was conducted of EGFR+ referred to the BC Cancer Agency between 2010 and 2015, treated with a first/second-generation EGFR TKI who developed CNS metastases (CNSm). Baseline characteristics, presenting symptoms and CNS-targeted treatment data was collected. Cox regression was conducted to determine the prognostic implications of the most common clinical presentations on OS.

Result:
229 patients were identified; 90 presented with CNSm and 139 developed CNSm during the course of their disease. Method of CNSm detection: CT only 61%, MRI only 8%, CT and MRI 30%, PET 1%. 80% of patients were symptomatic at CNSm diagnosis. Baseline characteristics: female 66%, median age 62 (34-90), Asian 51%, exon 19/exon 21/rare mutation/not specified 56/39/3/2%. CNS management: 13% surgery+/- whole brain radiotherapy (WBRT), 73% WBRT alone, 5% stereotactic radiosurgery (SRS)+/-WBRT, 9% no CNS directed therapy. OS was 21.6 months in patients who presented with CNSm vs. 27.1 months in those who developed CNSm during the course of disease (p=0.39). On multivariate analysis, the only presenting symptom associated with increased risk of death was cognitive dysfunction.

	Frequency of symptom at presentation	Univariate Analysis		Multivariate Analysis
		HR	P value	HR	P value
Cognitive Dysfunction	19%	1.26	0.01	1.21	0.04
Motor Dysfunction	18%	1.04	0.69
Balance and Ataxia	11%	1.04	0.46
Cranial Nerve Changes	8%	0.94	0.65
Headache and Dizziness	30%	0.94	0.41
Nausea and Vomiting	13%	0.92	0.48
Visual Disturbance	10%	0.99	0.97
Speech and Aphasia	10%	1.19	0.16
Seizures	7%	0.75	0.09	0.78	0.13
Leptomeningeal disease	12%	1.25	0.04	1.20	0.10

Conclusion:
The most common symptoms at initial presentation of CNSm in EGFRm patients were headache and dizziness, cognitive dysfunction and motor dysfunction. In multivariate analysis, only cognitive dysfunction was associated with poorer survival. Clinicians should have a low threshold for CNS screening based on the varied clinical symptoms experienced by patients

Background:
EGFR mutation in plasma circulating free tumor-derived DNA (ctDNA) by ARMS methods has been widely used in clinical settings in China. However, the prevalence of EGFR mutations in ctDNA was still unknown in the real world. This large-scale study (NCT02623257) aimed to explore the prevalence of EGFR mutations and determine the correlation of EGFR mutation status with clinical characteristics.

Method:
Plasma DNA samples from 721 patients with stage III/IV NSCLC who received ≤1 line chemotherapy were collected from 65 hospitals. EGFR mutations were tested by ARMS method. The EGFR mutation rate was calculated and the associations between EGFR mutant and patients’ demographic data, disease status as well as treatment pattern were explored.

Result:
EGFR mutations were detected in 176 of 721 (24.4%) patients, 165 of 620 (26.6%) in adenocarcinoma and 8 of 85 (9.4%) in squamous carcinoma. 138 (19.1%) harbored sensitizing mutations (66 19del, 62 L858R, 7 G719X, 2 L861Q, and 1 S768I) alone, 20 (2.8%) had resistance mutations (13 T790M, 7 Ins20) alone, 2 (0.3%) had a combination of activating mutations, and 16 (2.2%) had a combination of activating and resistance mutations. Twenty-eight (3.8%) patients were detected de novo T790M mutation either existed alone or combination, but no difference of clinical characteristics was observed. Higher detection rate was observed in 566 chemotherapy-naïve patients than in 155 patients received 1[st] line chemotherapy (27.2% versus 14.2%; p<0.001). Of which, the mutation rate of exon 19 deletion was 11.3% for naïve patients and 8.4% for the patients with 1[st] line chemotherapy; while the mutation rate of L858R decreased most obviously from 11.9%(naïve) to 1.9%(1[st] line). We also noticed 117 patients had ≥ 2 organ metastases and the mutation rate was 35.0% in these patients. Multivariate analysis showed female, chemotherapy native, or patients with ≥2 metastatic organs had higher percentage of EGFR mutation. Additionally, in 194 patients who had the follow-up treatment records, 34 of 49 patients (69.4%) with sensitive EGFR mutations received EGFR-TKI, 96 of 136 (70.6%) patients without sensitive EGFR mutation received chemo±radiation.

Conclusion:
Using plasma samples to detect EGFR mutation is feasible. ctDNA based EGFR mutation test could be a surrogate when tissue biopsy is not available due to limited tissue availability and procedural feasibility.

Background:
De-novo EGFR T790M mutations are rare, occurs less than 5% in the world and resistant to EGFR-TKI treatment. In Persahabatan Hospital, patients with De novo EGFR T790M mutations are given systemic chemotherapy with or without radiotherapy. No data on clinical profiles and 6 months survival rate in this group. the purpose of this preliminary study is to evalaute clincal profiles and 6 months survival of EGFR-T790M mutation in lung adenocarcinoma in Persahabtan Hospital Jakarta Indonesia.

Method:
Subjects was recruited consecutively from lung cancer Adenocarcinoma with De-novo EGFR T790M mutations. EGFR mutation is routinely assesed in lung adenocarcinoma in Persahabatan Hospital using PCR-High Resollution Melting, RFLP Analysis and direct sequencing. Subjects' clinical characteristics (race, age, gender), smoking habits, family history of malignancy, occupation, date at diagnosis (through anatomic pathology examination), complete diagnosis, stage, performance status, metastasis, chemotherapy and radiotherapy, RECIST evaluation, date of death were recorded.

Result:
Fourteen subjects were elligible and included in this study of which 9 subjects were male (64,2%), 5 subjects female (35,7%). Smoking status were 8 subject were smoker (57,1%), never smoke (35,7%), ex smokers (7,14%). Two subjects ( 14.2%) has family history of malignancy. All subjects were stage 4 with 42.8% pleural effusion (6 subjects), 21.4% pleural effusion and bone metastasis (3 subjects), 21.4% ( 3 subjects) with distant lymph nodes metastasis, 7.2% brain metastasis ( 1 subjects), 7.14% brain and bone metastasis. Mutation in ONLY exon 20 T790M 85,7% ( 12 subjects); double mutation did exist with Exon 19 ins/del and exon 20 T790M (1 subject); exon 19 ins/del and exon 20 T790M (1 subject); exon 21 L861Q and exon 20 T790M in 1 subject. Systemic chemotherapy were done in 9 subjects (64,3%), systemic chemotherapy and radiotherapy were done in 5 subject (35,7%). Six months survival rates in this group was 50%; and 1 year survival were 21,42%; Interestingly 4 subjects (28,6%) survived more than 1 year.

Conclusion:
De-novo EGFR T790M mutations are rare (< 5%). The presence double mutation ie. De-novo EGFR T790M mutations and other activating EGFR mutation did exist. these group has variable clinical responses with chemotherapy and warrant further investigation.

Background:
Personalized treatment based on the molecular markers in tissue of patients have crucial role in clinical practice. However, Obtaining Tumor tissue samples are not always available, and rebiopsy is not easy to do. The liquid biopsy with blood samples will be a good alternative diagnostic method. We compared the detection rate of EGFR mutation detection techniques between matched tumor tissue and peripheral blood sample in patients with lung adenocarcinoma.

Method:
We collected the paired samples from plasma and paraffin-embedded tumor tissue in patients before EGFR-TKIs treatment or after progression of EFGR-TKIs treatment. EGFR mutation analysis was done by two detection methods. One is the PNAClampTM (Clamp) which is the PNA-based PCR clamping that selectively amplifies only the mutated target DNA sequence as minor portion in mixture with the major wild type DNA sequences. The other is the PANAMutyperTM EGFR kit (Mutyper), which use PNA clamping-assisted fluorescence melting curve analysis to perform mutation detection and genotyping. The tissues were tested by Clamp method, and blood was tested in two ways. We compared the sensitivity of EGFR mutation detection techniques from tumor tissue and plasma circulating tumor DNA in patients.

Result:
Total patients were thirty two (17 male, 15 female) with advanced stage, mean age was 62.8 years-old, all patients were adenocarcinoma, and fifteen patients were never smoker. EGFR mutation positive rate of tissue was 31.3%. In plasma samples, there were 21.9% in Clamp method and 31.3% in Mutyper method. EGFR plasma detection rate of Mutyper was higher than Clamp. Two patients who showed wild type in tissue Clamp method have EGFR 19 deletion by Mutyper method. Both patients were female never smoker. In two cases, T790M was detected only in tissue but not in liquid biopsy. The overall concordance and degree of agreement between two samples were better in Mutyper (75%, gamma=0.872, p<0.001) than Clamp (71.3%, gamma=0.824 p=0.003).

Conclusion:
The plasma detection rate and the degree of agreement of Mutyper test were better than Clamp test. This method can be useful to detect EGFR mutation in circulating cell-free DNA sample. Liquid biopsy is an excellent resource, and had additional effect in choosing available drug for EGFR rich lung cancer population.

Background:
Pleural effusion is a highly efficient sample for liquid biopsy due to cancer cell enriched components. Liquid biopsy for EGFR genotyping is mostly being done using cell-free DNA. Extracellular vesicles (EVs) are known to carry oncogenic double stranded DNA that is considered as a noble biomarker. We set up to investigate the liquid biopsy using cell-free (cf) DNA and extracellular vesicular (EV) DNA of pleural effusion for EGFR genotyping in pulmonary adenocarcinoma patients.

Method:
Forty nine pleural effusion samples of pulmonary adenocarcinoma patients were evaluated. Non-cellular components after removing cell pellets by centrifuge (400g, 10 min, 4[0]C) were used for liquid biopsy and EVs were isolated by ultracentrifuge method (200,000g, 1 hr, 4[o]C). EV DNA and cf DNA were extracted separately and EGFR genotyping was done by PNA-clamping method. For the analysis of T790M detection, cell block slides were used as rebiopsy sample, when compared with liquid samples.

Result:
Among 31 EGFR-TKI naïve patients with known tissue EGFR genotyping, liquid biopsy using effusion EV DNA showed 100% matching with tissue EGFR genotyping in 19 EGFR mutant cases and detected 3 more EGFR mutant cases in tissue wild type (WT) patients, while liquid biopsy using effusion cf DNA missed 2 cases of tissue-based EGFR mutant patients and found 2 more EGFR mutant cases in tissue WT patients. In 18 patients with acquired resistance to EGFR-TKI, EGFR genotyping using effusion EV DNA detected T790M mutation in 13 of 18 (72.2%) patients, while 11 of 18 (61.1%) patients were detected by using effusion cf DNA, respectively. In contrast, only 3 patients were found to have T790M when using cell block slides.

Conclusion:
Liquid biopsy using pleural effusion is particularly effective for EGFR genotyping than conventional cytology or cell block sample. Liquid biopsy using effusion EV DNA is highly promising for EGFR genotyping, especially detecting T790M mutation, when compared with cf DNA.

Background:
This is a prospective, post-marketing surveillance study (NCT02131259) to evaluate safety and effectiveness of the irreversible ErbB family blocker, afatinib, which is approved in Japan for the treatment of inoperable/recurrent EGFRm+ NSCLC.

Method:
Patients with inoperable/recurrent EGFRm+ NSCLC received afatinib at the approved dose (20–50 mg/day) and were observed following treatment initiation for 52 weeks/until premature discontinuation. Data were included for all patients who received afatinib during the investigational period of this study, thus minimizing patient selection bias. The incidence/severity of adverse drug reactions (ADRs)/serious ADRs (sADRs) was the primary endpoint. Other endpoints included effectiveness (objective response rate [ORR]) and the incidence/severity of ADRs of special interest (diarrhea, rash/acne, nail effects [NEs] and interstitial lung disease [ILD]).

Result:
As of February 2017, 1,602 patients were included in the analysis (59% female, 81% aged <75 years, 86% ECOG PS 0–1, 83% BMI <25 kg/m[2]). 97% of patients had adenocarcinoma, and 64%/26% had EGFR Del19/L858R mutations. 70% had ≥1 line of prior chemotherapy; 48%/30% had prior gefitinib/erlotinib. Afatinib starting dose was 40 mg in 77% of patients. 95% had ADRs (36% grade ≥3). The most frequently reported ADRs (all grade/grade 3–4) were diarrhea (78%/15%), rash/acne (59%/6%), stomatitis (31%/4%), and NEs (38%/4%). ILD (all grade/grade 3–4/grade 5) occurred in 4%/2%/1% of patients. Median (range) time to initial onset was 5.0 (1–316) days for diarrhea, 11.0 (1–406) days for rash/acne, 9.0 (1–327) days for stomatitis, 38.0 (1–526) days for NEs, and 35.5 (3–329) days for ILD. Four patients (<1%) had creatinine elevation following grade ≥3 diarrhea. Dose reductions/permanent discontinuations occurred in 8%/7% of patients following diarrhea, 6%/4% following rash/acne, 3%/2% following stomatitis, 5%/2% following NEs, and <1%/4% following ILD. 33% of patients experienced sADRs. ADR frequency was associated with starting dose (96%/91% with 40/<40 mg afatinib), but was not unfavorably impacted by age, ECOG PS, number of prior chemotherapies, or previous EGFR TKIs. ORR with afatinib was higher in EGFR TKI-naïve patients than those who had previously been treated with EGFR TKIs (68% versus 21%).

Conclusion:
Consistent with previous studies, afatinib was effective in inoperable/recurrent EGFRm+ NSCLC, particularly as first-line targeted treatment (ORR ~70%). ADRs were predictable and generally manageable. ADR frequency was not notably affected by age, ECOG PS or number of previous therapies. In clinical practice, patients should be closely monitored and ADRs, particularly diarrhea and ILD, treated early to prevent sADRs.

Background:
In the Phase III LUX-Lung (LL) 3 and LL6 trials, first-line afatinib significantly improved PFS vs platinum-doublet chemotherapy in patients with EGFRm+ NSCLC (independent review; LL3: 11.1 vs 6.9 months, HR=0.58; p=0.001; LL6: 11.0 vs 5.6 months, HR=0.28; p<0.0001). In the Phase IIb LL7 trial, afatinib significantly improved PFS and TTF vs gefitinib in patients with EGFRm+ NSCLC harboring common EGFR mutations (PFS, independent review: 11.0 vs 10.9 months, HR=0.73; p=0.017; TTF: 13.7 vs 11.5 months, HR=0.73, p=0.0073). Here we report interim analysis results of a large Phase IIIb study of afatinib in a broad population of EGFR TKI-naïve patients with EGFRm+ NSCLC.

Method:
In this Phase IIIb trial with a similar setting to ‘real-world’ practice, EGFR TKI-naïve patients with locally advanced/metastatic EGFRm+ NSCLC were recruited from centers in China, Hong Kong, India, Singapore and Taiwan and received afatinib 40mg/day until investigator-assessed progression or lack of tolerability. Primary endpoint: number of patients with serious adverse events (SAEs). Secondary endpoints included: number of patients with afatinib-related AEs; time to symptomatic progression (TTSP). Other assessments included PFS (investigator review).

Result:
At data cut-off (13 Feb 2017) 479 patients were treated with afatinib (median age: 59.0 years; female: 52.4%; common [(Del19 and/or L858R) with or without uncommon]/uncommon only EGFR mutations: 86.0%/14.0%; ECOG PS 0/1: 19.8%/78.1%; brain metastases: 19.2%; 0/1/≥2 lines of prior chemotherapy: 59.7%/30.1%/10.2%. 24.8% of patients required dose reductions to 30mg; 6.1% had further reductions to 20mg. Median (range) treatment time was 9.7 months (0.2–38.6). SAEs were reported in 115 (24.0%) patients and afatinib-related SAEs in 29 (6.1%) patients. Grade ≥3 afatinib-related AEs occurred in 122 (25.5%) patients; diarrhea (n=50; 10.4%) and rash/acne (n=38; 7.9%) were the most common (≥5%). 18 (3.8%) patients discontinued treatment due to afatinib-related AEs. Median TTSP (15.3 months [95% CI: 13.4–17.5]) was 3 months longer than PFS (12.1 months [10.8–13.7]), suggesting afatinib may be continued beyond progression, and both were longer in patients with common (with/without uncommon) vs uncommon only EGFR mutations (PFS: 12.6 vs 9.1; TTSP: 15.8 vs 10.0 months).

Conclusion:
The safety data of afatinib from this interim analysis of a large-scale population of EGFR TKI-naïve EGFRm+ NSCLC patients are consistent with LL3/6/7 and confirm that most afatinib-related AEs are manageable and result in few treatment discontinuations. Afatinib also demonstrated encouraging efficacy in patients with common and uncommon EGFR mutations. Data from larger patient populations will be evaluated in further analyses of this trial.

Background:
Epidermal growth factor receptor (EGFR) T790M mutation is the most common mechanism for resistance to first- and second-generation EGFR tyrosine kinase inhibitors (TKI). Osimertinib has been demonstrated to overcome EGFR T790M non-small cell lung cancer (NSCLC). However, most of these patients eventually developed resistance after 10 months. Here we performed a comprehensive next generation sequencing (NGS) using circulating tumor DNA (ctDNA) to study resistance mechanisms in patients with advanced NSCLC who had developed resistance to osimertinib, and to provide potential opportunities for treatment.

Method:
10 advanced NSCLC patients were enrolled in this study after progression from first-generation EGFR-TKI treatment. Patients received osimertinib with 80mg daily, the response rate and progression-free survival (PFS) was assessed during treatment. 10ml peripheral blood was collected from patients after progression from osimertinib, and ctDNA genotyping was performed by next-generation sequence (NGS).

Result:
There were 3 patients received gefitinib and 7 patients received erlotinib before osimertinib therapy. All patients confirmed a partial response (PR) on osimertinib, the median PFS was 13.3 months. The initial gene mutation patterns before osimertinib therapy could be classified into two groups: L858R/Exon 19 Deletion (19Del) +T790M, and L858R/19Del+T790M unknown or wild-type. However, after progression from osimertinib, mutation patterns varied from groups. In the L858R/19Del+T790M group, two of six patients detected with L858R/19Del+T790M+C797S, two patients with L858R (+T790M) +HER2 amplification, and two patients with only 19Del+T790M; In the L858R/19Del+T790M unknown or wild-type group, there was only one patient detected with L858R/19Del+C797S mutation, the other three patients detected with only L858R mutation. For the C797S-mutated patients, it showed that T790M and C797S mutation presented in the same allele (cis) in two patients, while in both trans and cis in one patient. Other EGFR mutations were also detected, such as A750P, L792F, E709K, A1013V, L718V, and EGFR amplification.

Conclusion:
Genotyping of ctDNA in osimertinib resistant patients showed that the resistance might be related to specific gene mutation, e.g., EGFR C797S and HER2 amplification. Our findings provide insight that resistance to osimertinib might be different between T790M-mutated patients and T790M wilt-type patients. Combination therapy of osimertinib with other agents may be candidate to overcome the acquired mutation.

Background:
EGFR TKI treatment is standard of care in patients with metastasized NSCLC carrying an activating EGFR mutation. Targeted therapies achieve a higher ORR, OS, PFS and a better quality of life than chemotherapy. With the advent of 2[nd] and 3[rd] generation TKI´s effective in 1[st] generation TKI resistant tumors, we wanted to study the impact of these drugs on the outcome of patients in a real life setting in 3 lung cancer centers.

Method:
1473 patients from three cancer centers diagnosed with non-squamous cell NSCLC stage IV (UICC 7) were examined. Methods for the mutation testing was performed according to the German Oncopedia guidelines using either Sanger Sequencing or COBAS ® or Next Generation Sequencing.

Result:
964/1473 (65%) consecutive patients with non-squamous cell NSCLC were studied for the presence of tumor mutations. The EGFR mutation rate was 16% (147/943), and the ALK-translocation rate 4% (26/695). Median OS in EGFR mt+ patients was 27 months (n=147) compared to 11 months (n=796) in patients with EGFR WT (p<0.000). Median OS in EGFR mt+ patients depending on the center was 27 (n=95) vs. 28 (n=38) vs. 16 (n=14) months respectively (center 1 vs. center 2 vs. center 3). Median OS in ALK mt+ patients was 24 months (n=19) in center 1 and 11 months (n=5) in center 2 (p<0.025). The ORR in the CR/PR group was 54.2% for patients treated with chemotherapy and 77% for patients treated with TKI on 1[st] line therapy. The chance to reach a CR/PR is 2.83 higher for patients on TKI than for patients on chemotherapy (p<0.02). The use of 3[rd] generation TKI Osimertinib (n=19) lead to a significantly higher OS (n=19, median OS 67 months) than the use of only 1[st] and 2[nd] generation TKI (n=119, median OS 23 months, p<0.000). The hazard ratio for patients treated without Osimertinib was 4.66 [95% CI 2.006-10.81] (p<0.000). Patients treated with 3[rd] gen TKI had significantly longer PFS (11 months, n=7) than patients treated without (5 months, n=20) (p<0.037). Similarly, use of 2[nd] and 3[rd] generation ALKi impacted significantly on median OS: Crizotinib alone (n=8) 17 months, Crizotinib followed by Ceritinib and/or Brigatinib/Alectinib (n=12) median OS not reached and 3 months for other therapies (n=6) (p<0.000).

Conclusion:
Small differences in OS were observed, depending on the treatment centers, but the use of multiple EGFR and ALK-I impacted highly significantly on the outcome of patients with EGFR and ALK-alterations in a real life setting.

Background:
Clinical trials have demonstrated the efficacy and safety of EGFR tyrosine kinase inhibitors (TKIs), as first-line therapy in patients with metastatic non-small-cell lung cancer (NSCLC) whose tumours harbour activating mutations in the EGFR gene. However, there are limited data in the real-world setting describing TKI use among patients with EGFR-mutated NSCLC.

Method:
A retrospective non-interventional study was conducted using the Flatiron Health database, a longitudinal database containing electronic health record data, of patients seen in over 265 United States cancer clinics (~800 sites of care). Adult patients with histologically confirmed stage IIIB or IV NSCLC, with documented EGFR-mutated disease (del 19 or L858R), and who had received the first sequence of systemic therapy between January, 2011 and March, 2016 were included. Patient demographic and clinical disease characteristics were analysed and time-to-next treatment (TTNT) was used as a surrogate measure for progression-free survival. If validated in time, overall survival data will be presented.

Result:
Overall 20,924 adult patients with advanced NSCLC were identified, of whom 12,148 (58.1 %) were tested for EGFR mutations. Among these, 1,919 (15.8 %) patients had a positive EGFR result with 1,412 patients carrying del 19 or exon 21 mutations. Of these, 886 patients were treated with regimens in the first-line setting. Median age was 69.0 years; 67.7% were female; 55.5% were Caucasian; 54.3% were non-smokers; 97.5% had non-squamous histology and 57.3% received at least one subsequent treatment. Most patients received a TKI (71.8%) and 28.2% received other therapy. There were no differences observed in a patient’s type of insurance and the type of treatment (TKI or chemotherapy) that they were likely to receive. Patients treated with an EGFR TKI had a significantly longer median TTNT [erlotinib (13.2 months [95% CI 12.2–14.4; p=<0.001]) afatinib (12.6 months [95% CI 8.6–16.2; p=<0.001]), gefitinib (not evaluable; due to small sample size n=9) than those treated with other anti-cancer agents (4.6 months [95% CI 4.0–5.4]).

Conclusion:
These real world data mirror the results shown in randomised clinical trials, with the exception of patients being generally older. Importantly, patients who did not receive a TKI had significantly shorter TTNT. A large proportion of patients (42%) are not being tested for EGFR mutations, although there was an improvement over time. The reasons for not testing all eligible patients for EGFR mutations should be investigated further.

Background:
The analysis of EGFR mutations in plasma cell-free DNA (cfDNA) may become an auxiliary tool in the molecular diagnosis of advanced NSCLC patients from whom sufficient tumor tissue/cells specimens cannot be obtained. We evaluated the diagnostic performance of the two most common real-time PCR-based platforms for detection of EGFR mutations in plasma cfDNA that we use in our routine laboratory practice.

Method:
EGFR mutations were analyzed in plasma cfDNA collected from 107 NSCLC patients (non-SQC type, stages I-IV) before treatment (n=80) or at the time of progression on EGFR-TKIs (n=27) using two IVD-marked assays: ‘Cobas EGFR Mutation Test v2’ (Roche) and ‘Therascreen EGFR Plasma RGQ PCR Kit’ (Qiagen).

Result:
The overall concordance between EGFR mutation status in plasma cfDNA and paired tumor tissue samples was 62% (46% for Cobas and 32% for Therascreen test) for all patients regardless the NSCLC stage. The concordance increased to 88% (83% for Cobas and to 56% for Therascreen) when only samples from advanced NSCLC patients were evaluated. Accordingly, the Cobas test showed higher positive percent agreement between cfDNA and tissue results (PPA=83%) than did the Therascreen test (PPA=56%). Both test showed 100% negative percent agreement. The T790M mutation was detected in 26% cfDNA samples from patients upon progression on EGFR-TKIs. In 12 (44%) cases with progressive disease, the plasma cfDNA was the only specimen available for EGFR mutation analysis.

Conclusion:
In our laboratory setting, Cobas test demonstrated superior diagnostic performance over Therascreen assay in detection of EGFR mutations in plasma cell-free DNA from advanced NSCLC patients. Since analysis of mutated EGFR cfDNA in plasma is troublesome laboratory procedure, we recommend to validate the true diagnostic performance of each commercial assay in a specific laboratory conditions despite its IVD certification. The study is on-going.

Background:
The Compassionate-Use-Program (CUP) was available in Thailand for requesting afatinib during 2012-2014 for patients whom had failed at least one-line of platinum-based-chemotherapy and progressed following at least 6 months on 1[st]-generation EGFR-TKIs. This is a multicenter-retrospective-study in Thailand aimed to explore the efficacy and tolerability of afatinib in this group of patients.

Method:
Full medical-records of 79 patients from 7 institutions were reviewed. Clinical and tumor characteristics were analyzed using descriptive statistics. The efficacy in brain metastases was explored. Survival curves were constructed using the Kaplan-Meier method. All analysis was done in Stata version14.

Result:
Sixty-eight percent of patients were younger than 65 years old, 60% were female, and 67% received more than 2 of prior-regimen. EGFR-mutation was tested in 75% of patients; comprise of 86% common-mutations, 14% uncommon-mutations. Eleven patients had T790M acquired-resistance in combination with sensitive-mutation before receiving afatinib. One patient had De-novo-T790M. The mOS, mPFS, and mTTF were 11.5, 3.9, and 5.1 months, respectively. Eighteen patients had brain-metastases at enrollment and 6 patients had new brain-metastases during afatinib treatment. The mOS, mPFS, and mTTF were not statistically different among new brain-metastases or pre-existing brain-metastases. There was no dose-reduction in 38%, 1 dose-reduction 44%, 2 dose-reductions 12%, and 3 dose-reductions 6% of patients. The mean dose for every patient was 35 mg. Time-to-first-dose-reduction significantly affected the mPFS and mTTF as shown in Table. Furthermore, number of prior-treatment more than 2 was the significant factor causing first-dose-reduction within 1 month and age younger than 65 years-old was the significant factor causing first-dose-reduction within 2 and 3 months in multivariate and univariate model, respectively.

Time to 1[st] dose reduction	OS		PFS		TTF
	Hazard ratio (95%CI)	P-value	Hazard ratio (95%CI)	P-value	Hazard ratio (95%CI)	P-value
> 1 month <= 1 month	1 1.59 (0.74-3.41)	0.23	1 1.52 (0.82-2.82)	0.19	1 1.45 (0.77-2.74)	0.25
>2 month <= 2 months	1 1.54 (0.72-3.3)	0.27	1 2.11 (1.14-3.92)	0.02	1 1.63 (0.88-3.04)	0.12
>3 months <= 3 months	1 2.31 (0.98-5.45)	0.06	1 2.67 (1.38-5.20)	0.004	1 2.12 (1.09-4.12)	0.03

Conclusion:
The mOS, mPFS, and mTTF of our study were comparable with LUX-Lung1 study. Number of prior-treatment and age were the significant factors causing dose-reduction. Taken together with time-to-first-dose-reduction also affected the survival of patients.

Background:
In LUX-Lung 8 (LL8), second-line afatinib (an irreversible ErbB family blocker) significantly improved OS (median 7.9 versus 6.8 months; HR [95% CI]: 0.81 [0.69‒0.95]; p=0.0077), and PFS (2.6 versus 1.9 months; 0.81 [0.69‒0.96]; p=0.0103) versus erlotinib in lung SCC (N=795). Comprehensive genetic analysis in LL8 patients assessed whether afatinib efficacy varied according to genetic aberrations in cancer-related genes, including ErbB family mutations.

Method:
Tumor genetic analysis (TGA) was performed using Foundation Medicine FoundationOne™ next-generation sequencing (NGS). The cohort was enriched for patients with PFS >2 months, reflecting a range of responsiveness to EGFR-TKIs. EGFR expression was assessed by immunohistochemistry (IHC) in a largely separate cohort. Cox regression analysis correlated PFS/OS with genetic mutations (individual/grouped) and EGFR expression.

Result:
Of 440 patients selected for TGA (PFS >2 months: n=320; ≤2 months: n=120), samples from 245 were eligible (afatinib: n=132; erlotinib: n=113). In the selected TGA population, PFS/OS outcomes were improved in the afatinib versus erlotinib arm. Baseline characteristics were similar in TGA and IHC cohorts and LL8 overall. In the TGA subset, 53 patients (21.6%) had ≥1 ErbB family mutation (EGFR: 6.5%; HER2: 4.9%; HER3: 6.1%; HER4: 5.7%). Beyond the benefit seen for afatinib in the overall population, in afatinib-treated patients, PFS/OS were longer when ErbB mutations were present (PFS: 4.9 versus 3.0 months; OS: 10.6 versus 8.1 months). Conversely, survival outcomes in erlotinib-treated patients were similar with/without ErbB mutations. Presence of HER2 mutations predicted favorable PFS/OS with afatinib versus erlotinib. The Table shows outcomes in patients with/without ErbB family mutations, and by EGFR expression levels (afatinib: n=157; erlotinib: n=188).

Conclusion:
These data are provocative and suggest that NGS may enable identification of lung SCC patients who would derive additional clinical benefit from afatinib. Differential outcomes with respect to ErbB mutations for afatinib and erlotinib are hypothesized to reflect afatinib’s broader mechanism of action.

Subgroup	n	Afatinib vs erlotinib
		PFS		OS
		HR (95% CI)	p~interaction~	HR (95% CI)	p~interaction~
ErbB mutation-positive ErbB mutation-negative	53 192	0.56 (0.29–1.08) 0.70 (0.50–0.97)	0.718	0.62 (0.35‒1.12) 0.76 (0.56‒1.03)	0.683
EGFR mutation-positive EGFR mutation-negative	16 229	0.64 (0.17–2.44) 0.67 (0.50–0.91)	0.981	1.01 (0.32–3.16) 0.72 (0.54–0.95)	0.529
HER2 mutation-positive HER2 mutation-negative	12 233	0.06 (0.01–0.59) 0.72 (0.54–0.97)	0.006	0.06 (0.01–0.57) 0.76 (0.58–1.00)	0.004
HER3 mutation-positive HER3 mutation-negative	15 230	0.52 (0.16–1.72) 0.69 (0.51–0.94)	0.692	0.84 (0.27–2.59) 0.73 (0.56–0.97)	0.998
HER4 mutation-positive HER4 mutation-negative	14 231	0.21 (0.02–1.94) 0.67 (0.50–0.91)	0.909	0.22 (0.05–1.04) 0.75 (0.56–0.99)	0.272
EGFR IHC positive EGFR IHC negative	292 53	0.74 (0.56–0.97) 0.76 (0.41–1.40)	0.985	0.82 (0.63–1.06) 0.75 (0.41–1.40)	0.882
EGFR amplification present EGFR amplification absent	17 228	0.72 (0.18–2.90) 0.68 (0.50–0.92)	0.994	0.50 (0.15–1.65) 0.76 (0.58–1.00)	0.413
HER2 amplification present HER2 amplification absent	9 236	0.94 (0.20–4.38) 0.68 (0.50–0.91)	0.861	1.10 (0.27–4.48) 0.72 (0.54–0.94)	0.388

Background:
In EURTAC, ENSURE and OPTIMAL, first-line erlotinib was associated with higher response rates and improved PFS versus chemotherapy in EGFR Mut+ NSCLC. OS benefit was not observed, possibly due to high crossover. We present statistical analyses of OS adjusting for crossover in the three studies.

Method:
Rank Preserving Structural Failure Time (RPSFT) analysis is a retrospective statistical methodology that estimates the effect of active treatment and then removes that effect from the control patients, during the period of active treatment. Separately, post-hoc OS analyses were conducted on non-randomized subgroups of patients who received a single line of therapy, chemotherapy only or erlotinib only; or a sequence, chemotherapy followed by TKI or erlotinib followed by chemotherapy. Impact of sequencing on OS was evaluated (Kaplan-Meier methodology).

Result:
RPSFT analyses (Table) show OS was numerically longer with erlotinib versus chemotherapy in EURTAC and ENSURE. OS was not longer with erlotinib in OPTIMAL, possibly due to low treatment adherence in the erlotinib arm, where >50% of patients refused chemotherapy or did not complete four cycles. In the non-randomized analyses, patients receiving erlotinib as the sole treatment had longer OS versus patients treated with only chemotherapy, in all three studies. Patients who received erlotinib followed by chemotherapy had longer OS versus patients who received chemotherapy followed by erlotinib in EURTAC and ENSURE. Patients who received chemotherapy followed by erlotinib in OPTIMAL had longer OS than patients who received erlotinib then chemotherapy, again possibly due to erlotinib patients refusing the full extent of post-progression chemotherapy.

Conclusion:
RPSFT analysis showed increased OS between erlotinib and chemotherapy in EURTAC and ENSURE versus the original ITT analysis. Similar impact in OPTIMAL was not apparent. Post-hoc analysis of non-randomized subgroups (patients treated with a single therapy) showed longer OS for erlotinib versus chemotherapy, notwithstanding caveats of such an approach.

OS (RPSFT analysis)
Study	Treatment Group	N	Median OS, months	HR (95% CI)
EURTAC	Erlotinib	86	22.9	0.64 (0.44, 0.95)
	Chemo	88	15.4
ENSURE	Erlotinib	110	26.3	0.47 (0.32, 0.69)
	Chemo	107	17.1
OPTIMAL	Erlotinib	82	22.7	1.78 (1.21, 2.64)
	Chemo	72	30.5
Influence of sequencing on OS in patients who crossed over (Kaplan-Meier analysis)
Study	Treatment Group	N	Patients crossing over, %	Median OS, months
EURTAC	Erlotinib/chemo	49	57	24.9
	Chemo/TKI	72	82	22.6
	Erlotinib alone	37		17.2
	Chemo alone	16		1.6*
ENSURE	Erlotinib/Chemo	69	63	27.0
	Chemo/TKI	89	83	25.6
	Erlotinib alone	41		23.2
	Chemo alone	18		22.9
OPTIMAL	Erlotinib/ Chemo	49	60	26.8
	Chemo/TKI	52	72	31.4
	Erlotinib alone	33		18.6
	Chemo alone	20		11.2
*There were 5 deaths and 5 censored observations that occurred in the first two months

Background:
Circulating-free tumor DNA (ctDNA) has emerged as a sensitive and feasible non-invasive blood-based approach alternative to tissue biopsies to screen for genetic drivers in advanced non-small cell lung cancer (NSCLC) patients. Recently, the COBAS Mutation Test has been FDA-approved for the detection of EGFR mutation (EGFRm) in peripheral blood but other approaches are currently used in clinical practice. Here, we aimed to correlate two different platforms for EGFRm monitoring ctDNA in an enriched cohort of tissue-genetically phenotyped EGFRm advanced NSCLC patients.

Method:
Blood samples were prospectively obtained from an enriched cohort of EGFR+ advanced NSCLC patients. Formalin-fixed paraffin-embedded (FFPE) tumor was characterized by NGS (Ion AmpliSeq Lung Cancer Research Panel v.2) in all patients at diagnose. Plasma ctDNA derived from peripheral whole blood was evaluated by COBAS EGFR Mutation Test v2 and a Peptide Nucleic Acid (PNA) probe–based real-time polymerase chain reaction blinded to baseline tumor genotype. Diagnostic accuracy and concordance of both blood techniques was used for direct comparison with respect to the molecular status of FFPE tissue.

Result:
A total of 80 matched pairs of peripheral blood samples from 40 patients were collected. Baseline tissue NGS reported mutations at exon 19 del (n=23), exon 21 (n=10), exon 18 (n=2), exon 20 (n=2) and T790M (n=3). Four wild-type EGFR tumors were used as controls. Blood samples were obtained at diagnose (n=12) and during tyrosine-kinase inhibitor (TKi) treatment for monitoring (n=28). Overall, concordance between both blood-based techniques with respect tissue-NGS was 100% (4/4) for negative controls and 55% (20/36) for positive tissue-NGS samples. Detection based on PNA and COBAS was negative in 60% (24/40) and 32.5% (13/40) patients respectively. Among 19 samples negative by PNA at monitoring, COBAS allowed plasma EGFRm detection in 11 patients. At baseline, the only two negative samples patients by both techniques were found in patients with localized brain disease. Six patients had detectable driver T790M mutation; among three patients with T790M+ in tissue, COBAS allowed detection in plasma in one patient whereas none was identified with PNA. The other three patients had acquired T790M mutations identified only in blood, all by COBAS.

Conclusion:
In this prospective blinded validation cohort, both methods retained high specificity. However, major differences between techniques were observed for longitudinal monitoring of EGFRm in blood.

Background:
Detection of actionable mutations using circulating tumor DNA (ctDNA) isolated from patient plasma is now accepted as clinical practice in NSCLC. Nevertheless, the full extent to which longitudinal plasma analysis can be utilized to guide clinical decision-making has yet to be realized. We prospectively incorporated serial next-generation sequencing (NGS) of ctDNA into the ongoing SWOG S1403 clinical trial (NCT02438722) of afatinib+cetuximab vs afatinib in treatment-naïve NSCLC patients with EGFR-mutant tumors.

Method:
Time points for specimen collection were pre-treatment, after two months of therapy on Cycle 3 Day1 (C3D1) and at progression. Objectives were to: 1) determine the prognostic and predictive significance the EGFR mutant allele frequency (MAF) at each time point; 2) correlate changes in MAF over time with regard to patient outcome, and 3) identify putative emergent resistance mechanisms and companion mutations. Specimen analysis was conducted using the Guardant360 73-gene digital NGS panel.

Result:
To date, 53 patients with advanced EGFR-mutant NSCLC have contributed baseline samples. Of these, 46 had ctDNA detectable at baseline (87%). 39 of these 46 (85%) had detectable, tissue-identical EGFR mutations, for an overall EGFR detection rate of 74% (39/53). A positive finding for EGFR amplification (Amp) in plasma correlated with high ctDNA MAF: median for Amp 16.9 vs nonAmp 0.9 (range/n: 11.6-43.7/10 vs 0.11-7.7/17; p<0.0001). Of patients with detectable EGFR mutation at baseline, 27 had analyzable ctDNA collected at C3D1. Of these, 26/27 showed decreasing MAFs on-treatment (mean for baseline: 9.8 vs C3D1: 0.14; p<0.0001), with 20 cases having no detectable EGFR mutation at C3D1 (mean of 7 positives at C3D1: 0.55). At progression, samples were collected from 14 patients and 10 had EGFR mutations detectable, with T790M present in 3. Another patient had an FGFR3 fusion at PD, but no previous draws were available to determine if it was emergent.

Conclusion:
Longitudinal analysis of plasma ctDNA in S1403 patients demonstrated significant treatment-induced changes in mutation burden and identified resistance mechanisms at progression. EGFR gene amplification, as assessed in plasma, was significantly associated with increased ctDNA MAFs. Patients showed a significant, one-to-two orders of magnitude decline in EGFR MAF after two months of therapy, with 74% dropping below detectable levels. At progression, EGFR mutation detection rates increased, often concomitantly with a putative emergent resistance factor. Accrual to S1403 is ongoing and patient treatment and outcomes remain blinded. The prognostic and predictive utility of baseline and therapy-induced changes in ctDNA MAF kinetics will be determined at study unblinding.

Background:
Erlotinib is a thyrosine kinase inhibitor which prevents epidermal growth factor receptor(EGFR) by blocking intracellular phosphorylation. Various cutaneous side effects have been associated with EGFR inhibitors. We present a case with ectropion and conjunctivitis followed by erythemato papulopustular skin rash.

Method:
Figure 1A 72 year old male who has suffered from non small cell carcinoma (adenocarcinoma)of lung stage IV since one year ago came to our clinic with bilateral ectropion and conjunctivitis. He has received erlotinib 150 mg/d since three to four months ago.



Result:
Figure 1We prescribed supportive care and symptomatic treatment and discontinued erlotinib for one week due to the severity of eye disease. It resolved gradually after two weeks but erythemato papulr rash appeared



Conclusion:
EGFR inhibitors are one of the most important treatment modalities among patients with lung adenocarcinoma.Cutaneous side effects of these agents are reported frequently. Eye disease including ectropion is a rare adverse effect of these factors. We describe this case to consider the newly diagnosed eye edverse effect of this important agents.

Background:
Third generation EGFR inhibitors are efficacious in patients with NSCLC harboring EGFR sensitizing mutations. Acquired resistance includes alternative mechanisms of EGFR activation and activation of other signaling pathways. Here we report a patient with a CBL mutation as a potential mechanism of acquired resistance to 3[rd] generation EGFR TKI effectively treated with sitravatinib, a spectrum-selective RTK inhibitor of TAM receptors (comprised of TYRO3, AXL, and MER), PDGFRA, KIT, and MET.

Method:
In Study 516-001, a phase 1/1b study of sitravatinib in patients with advanced solid tumor malignancies, patients are selected based on mutations in targeted pathways. Eligible genetic alterations include inactivating mutations of CBL, which encodes an E3-ubiquitin ligase that regulates degradation of activated RTKs that include EGFR, TAM receptors, PDGFRA, KIT, and MET and which account for ~2-3% of NSCLC. Based on overlap of sitravatinib and CBL RTK targets and supporting cancer cell line and tumor model data, we hypothesize that inactivating CBL mutations predict response to sitravatinib.

Result:
After Phase 1, patients were enrolled by mutational profile, and here we report on a patient enrolled into the CBL mutation cohort. The patient is a 77 year-old female, lifelong non-smoker with metastatic lung adenocarcinoma characterized by EGFR exon19del initially treated with erlotinib with a partial response (PR) of 9 months. Upon disease progression (PD), a new EGFR T790M mutation led to treatment with rociletinib (an experimental inhibitor of EGFR T790M) with stable disease for 11 months. Upon PD she was treated with osimertinib without response. She then received carboplatin/pemetrexed with a PR. Brief erlotinib re-challenge resulted in PD, with post-progression NGS revealing loss of EGFR T790M, presence of the original EGFR exon19del, and a new CBL C384R, a mutation located in the RING domain and predicted to result in the loss of CBL ligase adaptor function resulting in sustained activation of relevant RTKs. The patient then started sitravatinib treatment in Study 516-001. After 36 days on-study, restaging demonstrated PR, which was later confirmed with a maximum decrease in unidimensional target lesion measurement of 77%. She remains on-study.

Conclusion:
Loss of function mutations in CBL represent a unique class of mutations that may be responsible for acquired resistance to 3[rd] generation EGFR TKI. Sitravatinib has demonstrated clinical activity in a patient with NSCLC characterized by EGFR exon19del and CBL C384R mutations. The clinical trial with sitravatinib is currently enrolling patients with CBL loss of function mutations. (NCT02219711).

Background:
All EGFR mutated lung cancers will eventually develop acquired resistance to first-line EGFR tyrosine kinase inhibitors (TKi). The most common cause is the EGFR T790M point mutation. Recent studies demonstrated the utility of liquid biopsies (LB) as initial detection strategy. The mutation may be indicative of a more indolent disease and plasma positivity seems to be related to more extensive disease.

Method:
Retrospective evaluation of T790M status and clinical characteristics of patients with advanced or recurrent non-small cell lung cancer (NSCLC) who progressed under EGFR-TKi was performed at our institution. The T790M mutation was assessed in circulating cell-free DNA (cfDNA) in plasma as initial strategy; if negative, a tumor biopsy (TB) was obtained. Samples were analyzed using Cobas® EGFR Mutation Test v2. The purposes of this study were: to explore the feasibility of T790M mutation detection in our practice; to evaluate objective response rate (ORR) and progression-free survival (PFS) on first-line TKi therapy according to T790M status and clinical characteristics.

Result:
We identified 29 patients. Regarding the pre-treatment activating EGFR mutations, 59% (n=17) harbored an exon 19 deletion, 38% (n=11) the L858R mutation and one case had both mutations (two primaries). All patients were treated with EGFR-TKi, 90% (n=26) in the first-line treatment setting. The T790M mutation was positive in 18 patients (12 plasma+ and 6 plasma-/tissue+). The remaining 11 negative results, 7 were assessed only by plasma and the other 4 were confirmed on TB. The resistance mutation was more frequent in patients with EGFR exon 19 deletions (71%, 12/17). In patients with a positive T790M result, those with extra-thoracic disease had detectable T790M mutation in plasma in the majority of cases (86%, 12/14), while all patients with exclusive intra-thoracic disease (n=4) had negative result in plasma. Patients who developed a T790M mutation had a longer PFS under first-line TKi, but not statistically significant (14 months in T790M+ [CI 95% 7.8-20.2] vs. 9 months [CI 95% 6.8-11.2] in T790M-, p=0.201) with similar ORR (55.6% in T790M+ vs. 45.5% in T790M-, p=0.597).

Conclusion:
The evaluation of EGFR T790M mutation is feasible in LB (cfDNA) in patients progressing under EGFR-TKi, but TB is recommended in the negative cases. The likelihood of detection of T790M in plasma is greater in patients with extra-thoracic disease, probably related with higher disease burden. Tumors which develop T790M tend to have a more indolent course, but the conclusion is limited due to the small sample size.

Background:
ASTRIS (NCT02474355) is an open-label, single-arm, multination, real world treatment study, investigating the safety and efficacy of osimertinib in patients with T790M-positive advanced non-small cell lung cancer (NSCLC), who have previously received EGFR-TKI. We report the first results of Korean subset from ASTRIS which is the largest real world treatment study of osimertinib to date.

Method:
Eligible patients had advanced NSCLC harbouring a T790M mutation determined by local validated molecular tests, received prior EGFR-TKI therapy, acceptable organ and bone marrow function and no history of interstitial lung disease (ILD) or QTc prolongation. Enrollment of patients with asymptomatic, stable CNS metastases were permitted. Patients received osimertinib 80 mg once daily. The primary efficacy outcome was overall survival; other outcomes included investigator-assessed response rate (RR), progression-free survival (PFS) and time to treatment discontinuation (TTD). Safety assessment was also conducted. Data cut-off (DCO) was 3 November 2016; results from 1,217 patients in the global study have been presented previously (ASCO 2017 Abstract 9036).

Result:
A total of 371 patients received at least one dose of osimertinib from 30 Korean sites (full analysis set); at DCO, 319 patients (81.4%) were ongoing and median follow-up time was of 3.1 (0–8) months. Baseline patients’ characteristics were median age 61.1 (27–85) years old, female 65.5%, PS 0/1 88%, prior chemotherapy 47%, prior radiotherapy 48%. Tissue was the most common specimen source to test T790M mutation as well as other EGFR mutations (287/371, 77.4%) and plasma was the next (39/371, 13.1%). Fifty two patients (13.3%) had discontinued treatment; median duration of exposure 3.3 (0–7) months, 30 pts (7.7%) had disease progression and 24 patients (6.5%) died. In patients evaluable for response, defined as at least one dose of osimertinib and one response assessment, the investigator-assessed RR was 72.1% (212/294; 95% CI 66.6 – 77.2). Due to limited follow-up period, OS, PFS, and TTD were immature to analyze. Adverse events (AEs) leading to dose modification and treatment discontinuation were reported in 26 patients (7%) and 14 patients (3.8%), respectively. Serious AEs were reported in 50 patients (13.5%) and AEs leading to death in 8 patients (2.2%). ILD/pneumonitis-like events were reported in 9 patients (2.4%), and QTc prolongation (>470ms) in 5 patients (1.3%).

Conclusion:
At DCO for the 1[st] interim analysis of ASTRIS, Korean subgroup results demonstrated similar clinical activities (RR) to that observed in the osimertinib clinical trial program with no new safety signals.

Background:
T790M inhibitor, a third-generation epithelial growth factor receptor tyrosine kinase inhibitor (EGFR-TKI), was significantly superior to chemotherapy for EGFR-mutant patients with advanced non-small-cell lung cancer (NSCLC) harboring T790M mutation after resistance to the first-generation EGFR-TKIs. However, for those without acquired T790M mutations or MET amplification, few data showed whether the second-generation EGFR-TKI could overcome resistance to the first-generation EGFR-TKIs.

Method:
EGFR mutation was detected by amplification refractory mutation system (ARMS) technology. Overexpression of MET was determined by immunohistochemistry (IHC), and MET amplification was detected by FISH. Next generation sequencing (NGS) was used to test tumor tissues before and after resistance to TKIs.

Result:
Patient 1 was a 53-year-old Asian male, who underwent surgery followed by adjuvant chemoradiotherapy with an 11 months of disease-free survival. A biopsy was performed and revealed an activating deletion mutation in exon 19 of the EGFR. He started gefitinib 250mg once daily and achieved an initially good partial response (PR) followed by continuously stable disease for a progression-free-survival (PFS) about 29 months. A rebiopsy showed that exon 19 deletion mutation still existed without an acquired T790M mutation or MET amplification. He refused to receive chemotherapy so we prescribed afatinib 40mg daily. He then achieved a dramatic PR 23 days later which was confirmed 1.5 months after that. We sent tumor tissue of the patient before and after TKI therapy for NGS, the result showed a decreasing abundance of EGFR exon 19 deletion without T790M mutation, MET amplification and Her2 mutation/amplification. Patient 2 was a 63-year-old Asian female harboring EGFR exon 20 insertion (A763-Y764 ins), who received first and second lines of chemotherapy with short PFSs. He was enrolled in the trial of allitinib (AST-1306) as the third line therapy with the best response of stable disease(SD) and the PFS of 7 months. A later rebiopsy showed the strong overexpression of MET, but the forth-line therapy crizotinib failed. He then received chemotherapy as the-fifth line therapy, 4 cycles later he progressed. So he started afatinib 40mg daily. His symptoms were relieved significantly 5 days later, and PR was confirmed 1.5 months later.

Conclusion:
Afatinib might overcome resistance to the first-generation EGFR-TKIs for EGFR-mutant patients with advanced NSCLC without acquired T790M mutation or MET amplification. Further investigations are warranted.

Background:
The prevalence of EGFR mutation has been well elucidated in different ethnicities. Recently, increasing attention has been given to dual EGFR mutations. However, less attention has been invested in dual in cis EGFR mutations. Until now, none of retrospective or prospective research has focused on dual in cis EGFR mutations except case reports.

Method:
In this real world study, we performed capture-based ultra-deep targeted sequencing on circulating tumor DNA to identify and investigate the prevalence and genotype distribution of dual in cis EGFR mutations in 3,000 Chinese advanced NSCLC patients. This cohort consisted of both treatment-naïve and previously treated patients. Ten milliliter of peripheral blood was collected from every patient and a minimum of 50ng of ctDNA was needed for library construction. The panel covered critical exons and introns of 168 genes (160kb of human genomic regions).

Result:
1,266 patients harbored EGFR mutant in this cohort; among them, 501 patients harbored 19 deletions, 489 harbored L858R, and the remaining harbored other EGFR mutations. We identified 1.5% patients (19/1,266) harboring dual in cis EGFR mutations. Among them 37% (7/19）carried two rare EGFR mutations and the remaining 63% (12/19) carried EGFR L858R in combination with a rare mutation. No patient carried EGFR 19 del in combination with other rare mutations was identified in this cohort, suggesting EGFR 19del is a stronger oncogenic driver than EGFR L858R (p=0.000197, Fisher’s exact test). For patients carried two rare mutations, both mutations were either located on exon 18 or exon 21. The allelic fractions (AF) of both mutations were similar. The AF of either EGFR mutations was the maximum AF in all patients, demonstrating the clones harboring EGFR mutations were major clones. Interestingly, 1 patient carried additional KRAS mutation and 2 patients had EGFR amplification.

Conclusion:
In cis dual EGFR mutation was rare (1.5%) in EGFR mutant Chinese advanced NSCLC patients. EGFR L858R was significantly more likely to couple with a rare in cis dual mutation than 19 del. EGFR 19del might be a stronger oncogenic driver than EGFR L858R.

Background:
Epidermal growth factor receptor (EGFR) mutation can predict the response to EGFR tyrosine kinase inhibitor (TKI) in non small cell lung cancer (NSCLC). In this study, we determined the limit of detection of different methods, such as sequencing, restriction frament length polymorphism (RFLP), high resolution melting (HRM) and reversed dot blot (RDB) hybridization to detect for EGFR mutation detection.

Method:
Mutation detections of exons19 and 21 in EGFR gene were performed using sequencing, restriction fragment length polymorphism (RFLP) and high resolution melting (HRM) analysis method. Moreover, we also developed Reversed Dot Blot(RDB) method hybridization as an alternative procedure. Genomic DNA of H1975, HCT116 cell line as well as specially designed oligonucleotides were used to determine the limit of detection of these procedures. We also determine the sensitivity and specificity of RDB method compared to existing methods.

Result:
Limit of detection of each method was determined by titration of mutant allele in wildtype background. HRM analysis and RFLP were able to detect mutation in samples containing 6.25% to 12.5% mutated DNA.Limit of detection of RDB method was around 12.5%. Analysis of 40 patients had shown that the specificity for to detect mutations in exon 19 and 21 was 100% and 89.5% and the sensitivity for both exon was 62.5%.

Conclusion:
RDB method maybe used as an alternative method to standard procedure such as sequencing. This method is specific, but need further improvement to increase its sensitivity.

Background:
The existence of circulating tumor DNA (ct-DNA) in urine as a noninvasive and alternative sample to tissue biopsy has been promising. However, it still has some challenges. Common mutations of epidermal growth factor receptor (EGFR), such as L858R in exon 21 mutation has been used to predict lung cancer treatment response to tyrosine kinase inhibitors (TKI). The study aimed to demonstrate that DNA from urine can be detected using simple methods with high sensitivity, enabling early Non-Small Cell Lung Cancer (NSCLC) detection from these noninvasive samples.

Method:
Cytological smear and urine samples from 59 NSCLC patients had been collected and processed to obtain genomic DNA. EGFR common mutations in exon 21 were analyzed using polymerase chain reactions (PCR) and restriction fragment length polymorphism (RFLP) methods that having analytical sensitivity of detecting 3% EGFR mutant alleles. Diagnostic sensitivity and specificity test was used to evaluate the feasibility of urine as source of noninvasive samples compared with cytological samples.

Result:
We were able to detect EGFR exon 21 L858R mutations in 17 of 59 (28,81%) urine samples while 27 of 59 (45,76%) cytological samples were EGFR positive mutations. Agreement between urine and cytological slide was 45.76%. Diagnostic sensitivity and specificity were 22,22% and 65.62% respectively.

Conclusion:
EGFR mutation of lung cancer patients were detected from urine using RFLP methods. In this study, mutation rate in urine (28,81%) was similar to mutation rate from plasma (22%) of Asian Lung cancer patients as described Han, B et al 2015. However, poor sensitivity (22.22%) and specificity (65.62%) of urinary ctDNA L858R analysis may not lead to treatment decision. More in-depth studies focusing on urine collection techniques and more sensitive technology may lead to useful application of urine as an alternative noninvasive liquid biopsy sample for NSCLC detection.

Background:
Non-small-cell lung cancer (NSCLC) patients with activating mutation of epidermal growth factor receptor (EGFR) gene inevitably develop disease progression to EGFR-tyrosine kinase inhibitors (TKIs). T790M gatekeeper mutation accounts for approximately 60% of the acquired resistance, and osimertinib, third generation EGFR-TKI, is effective against such tumors. Demonstration of T790M requires second biopsy, which is inconvenient and sometimes hazardous. Liquid biopsy of circulating tumor DNA (ctDNA) in the plasma is a non-invasive alternative, but clinical relevance of this method is yet to be fully elucidated.

Method:
NSCLC patients with EGFR mutation undergoing 1st- or 2nd-generation EGFR-TKI treatment are monthly or bi-monthly monitored for plasma ctDNA EGFR mutations, including T790M, by Cobas EGFR mutation test® via 10 ml blood sampling. The treatment could be changed on the attending physicians’ discretion, including osimertinib in patients with documented T790M mutation, with continuation of the monitoring. The primary endpoints are T790M positivity rate of ctDNA among patients with tissue-confirmed T790M mutation, and T790M positivity rate of ctDNA at landmark points, such as radiological progression, clinical deterioration or second biopsy of the tumor. The secondary endpoints include EGFR mutation detection rate of plasma ctDNA at landmark points, the interval of tissue and plasma T790M detections, and response rate and progression-free survival in patients treated with osimertinib according to plasma/tissue T790M status.

Result:
From Oct 2016 to Mar 2017, 121 eligible patients were enrolled. The median age 73 years (range, 42-92), 42 male (34.7%), PS 0, 1 and 2 were 64 (52.9%), 54 (44.6%) and 3 (2.5%) respectively. EGFR mutation types were 61 patients (50.4%) del 19, 55 (45.5%) L858R and 5 (4.0%) others. 80 (66.1%) were never smokers. At the time of EGFR-TKI initiation, 82 (67.8%) had advanced and 39 (32.2%) had recurrent diseases. 65 (53.7%) had any prior therapy, including 21 (14.8%) with prior cytotoxic chemotherapy. Used EGFR-TKIs were gefitinib 50 (41.3%), erlotinib 40 (33.1%), and afatinib 31 (25.6%). Newly-diagnosed/on TKI therapy at enrollment was 18 (14.9%)/103 (85.1%). The median follow-up is 12[鈴木1] months. So far, plasma T790M was detected in 8 patients.

Conclusion:
This observational study will elucidate the clinical usefulness and limitations of monitoring of ctDNA for T790M mutation in the real-world setting.

Background:
Osimertinib is an effective, irreversible EGFR-TKI selective for EGFR-TKI-sensitizing (EGFRm) and T790M[+] resistance mutations. Intracranial activity of osimertinib has been reported in T790M[+] patients; however its activity in EGFR-TKI naïve setting has not been investigated yet. Here, we present preliminary data of a phase II study assessing the intracranial activity of osimertinib in EGFRm NSCLC patients with asymptomatic brain metastases as 1[st] line therapy for naïve patients and as 2[nd] line therapy for EGFRm patients who progressed on EGFR TKI and harbor T790M[+]. Dose escalation from 80 mg QD to 160 mg QD was allowed in cases of brain progression.

Method:
A phase II, open label, two arm study is presented. Treatment-naïve (arm A) advanced NSCLC with sensitizing EGFRm and previously treated (arm B) with 1[st] or 2[nd]-generation EGFR TKIs (gefitinib, erlotinib or afatinib) in whom T790M[+] was diagnosed were enrolled to this study. Intracranial response was assessed by brain MRI scans every 6 weeks and systemic evaluation by PET-CT scan every 3 months unttil progression. The study plans to enroll 20 patients in each arm aiming to reach a statistical power of 0.8.

Result:
As of May 31 2017, 22 patients were enrolled; 15 patients in arm A and 7 patients in arm B, age 67.8±10.4 years, 14 females and 8 males. Preliminary outcome analysis is presented for 16 patients with a median follow up of 6.1 months (range 1.4-11.4 months). On May 31[st] 2017 data cut-off was performed. The Intracranial response rate (ICRR) was 81% (13/16 pts); 82% (9/11 pts) in arm A and 80% (4/5 pts) in arm B. Time to response was 6 weeks (1[st] MRI) in all responding patients. Intracranial disease control rate (IDCR) was 81% (13/16 pts) for the current follow up time of 6.1 months; median PFS not reached. Dose escalation was performed in 2 cases and data is not mature yet. Toxicity profile is similar to previous reported data.

Conclusion:
Osimertinib shows a promising intracranial response rate in naïve EGFRm NSCLC patients (82% ICRR) and in 2[nd] line T790M (+) setting with 80% ICRR. This study is still recruiting.

Background:
There has been a marvelous paradigm shift in the diagnosis and management of lung cancer, with the discovery of driver mutations that can be targeted by specific inhibitors. Personalized therapy is driving the demand for mutation testing in cancer. Especially, molecular testing of lung adenocarcinoma for the epidermal growth factor receptor (EGFR) is now considered one of the most important standard of care with other driver mutations. However, sampling tumor tissue other than surgical resection has inevitable limitations. It may be difficult to obtain enough DNA for EGFR mutation test if biopsy tissue was inadequate. There is a promising ctDNA detection technology, that is PANA Mutyper technology, which is a novel technology that integrates PNA clamp and PANA S-melting. We investigated effectiveness of both methods for the detection of EGFR mutation.

Method:
Ninety patients with malignant pleural effusion of lung cancer were enrolled. EGFR mutations were assessed by PANA Mutyper and PNA clamping using tumor tissues, cell blocks, pleural effusions, and bloods separated by serum and plasma. The diagnostic performance of both methods was investigated.

Result:
Among the 90 patients with malignant pleural effusion of lung cancer, 56 patients were adenocarcinoma of lung, 11 were squamous carcinoma of lung, and 17 were small cell lung cancer. PANA Mutyper using bloods showed that 70% sensitivity, 100% specificity, 100% positive predictive value and 83% negative predictive value comparing to the combination of PANA Mutyper and PNA clamping through tumor tissues, cell blocks and pleural effusions (p=0.014).

Conclusion:
PANA Mutyper using bloods had a diagnostic performance for the detection of EGFR mutations in NSCLC that was comparable to that of tumor tissues, cell blocks and pleural effusions. The diagnostic performance of PANA Mutyper was better than that of PNA clamping. This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (No. 2014R1A2A1A11052422).

Background:
To perform a retrospective analysis of patients with brain metastases (BM) from epidermal growth factor receptor (EGFR)-mutant non-small cell lung cancer (NSCLC) to evaluate the preferred treatment timing of EGFR-tyrosine kinase inhibitors (TKIs) in this population.

Method:
Of 94 initial or recurrent patients diagnosed BM from EGFR-mutant NSCLC between Jan 1, 2012, and Jun 31, 2016, 30 received upfront brain RT followed by EGFR-TKI, 39 EGFR-TKI combined with concurrent brain RT, and 25 upfront EGFR-TKI then brain RT. Disease-specific-graded prognostic assessment was similar among all 3 groups. The primary endpoint was overall survival (OS), and the second endpoint was intracranial progression-free survival.

Result:
Although the median OS was not significantly different among the upfront RT, concurrent treatment, and upfront EGFR-TKI groups (29.0 vs 24.0 vs 17.0 months; P=0.186), however, it was longer in the upfront RT group compared with the upfront EGFR-TKI group (P=0.035). On subgroup analysis, the exon 19 deletions patients had longer OS than the exon 21 mutations patients in upfront EGFR-TKI group (23.0 vs 15.0 months; P=0.048), but the upfront RT (34.0 vs 23.0 months; P=0.186) and the concurrent treatment groups (24.0 vs 13.0 months; P=0.827) did not. According to multivariate COX analysis, KPS (≥ 70) and intracranial metastasis alone was associated with a longer OS. The median intracranial progression-free survival was significantly improved in patients receiving upfront RT compared with those receiving concurrent treatment or upfront EGFR-TKI (not reached vs 40.0 vs 9.0 months; P=0.003).

Conclusion:
The present study suggests that the use of upfront brain radiotherapy, and the deferral of EGFR-TKI may result in superior OS in patients with brain metastases from EGFR-mutant NSCLC. Also, upfront brain radiotherapy management could significantly reduce the risk of intracranial progression. A prospective, multi-institutional, randomized trial of upfront EGFR-TKI then RT versus upfront RT followed by EGFR-TKI is urgently needed, especially based on different subgroup population.

Background:
EGFR T790M mutation is responsible for 50 to 60% cases of resistance to first and second generation epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKI) in patients who have lung cancer with an activating EGFR mutation.

Method:
We administered osimertinib 80 mg once daily in 11 patients with advanced lung cancer who had radiologically documented disease progression after previous treatment with first and second-line EGFR tyrosine kinase inhibitors.

Result:
We treated 11 patients with osimertinib with stage IV lung adenocarcinoma. Three patients were males and eleven were females, median age 64 (raging from 54 to 82). Seven patients were never smoker, and four were ex-smokers. Ten patients had initially deletion 19 in EGFR gene and one had L858R in exon 21 and then developed T790M mutation. In all patients T790M was proven from tumor tissue. Majority of patients were ECOG 1. All patients were previously treated with first or second line EGFR TKIs (erlotinib, gefitinib or afatinib) and had radiologically documented disease progression. Two patients received osimertinib in 2[nd] line setting, three patients in third line setting, 3 in fourth, one in fifth, one in seventh and one even in tenth line setting. Median time to response was 4 weeks (raging from 3 to 7). All11 patients had partial response (PR) with progressive disease in only one patient after 12 months of treatment. Median duration of response is at the moment 12.5 months, (4 to 60). No significant side effects were observed.

Conclusion:
Osimertinib is highly active in patients with lung adenocarcinoma which harbor EGFR T790M mutation who had had disease progression during prior therapy with EGFR tyrosine kinase inhibitors. Efficacy in Croatian population is similar to previously observed. There were no serious side effects of treatment.

Background:
Next-generation sequencing (NGS) of cell-free circulating tumor DNA (ctDNA) enables a non-invasive option for comprehensive genomic analysis of lung cancer patients. In this study, we evaluated the impact of ctDNA sequencing on outcomes and treatment strategy.

Method:
In this retrospective study, data was collected from files of advanced non-small cell lung cancer (NSCLC) patients at the Thoracic Cancer Unit, Rabin Medical Center, Israel, between 2014-2017. Plasma samples were analyzed by a commercial test (Guardant360[TM]), using massively parallel paired-end synthesis to sequence a targeted 68-73 gene panel.

Result:
116 consecutive NSCLC patients were included in this study. Median age at diagnosis was 63 years, 74% had adenocarcinoma. 41% performed ctDNA analysis before 1st line therapy (Group A) and 59% on progression (Group B), among them 41% after progression on EGFR TKIs (Group B1) and 59% on other treatments (Group B2). ctDNA analysis yielded actionable mutations (EGFR, ALK, RET, BRAF, MET, ERBB2 (HER2)) in 40.5%: 31% in group A, 47% in group B, 71% in group B1 and 30% in group B2. Treatment decision was taken toward targeted therapy subsequent to NGS in 26%.

Genetic alterations frequencies among groups A, B, B1 and B2

Group A 19 individual mutations	Group B 52 individual mutations	Group B1 34 individual mutations	Group B2 18 individual mutations
EGFR Sensitizing 52.5% (10/19)	EGFR Sensitizing 42% (22/52)	EGFR Sensitizing 59% (20/34)	MET 55.6% (10/18)
MET 16% (3/19)	MET 27% (14/52)	EGFR T790M 23% (8/34)	ERBB2 16.7% (3/18)
ERBB2 10.5% (2/19)	EGFR T790M 15% (8/52)	MET 12% (4/34)	RET 16.7% (3/18)
BRAF V600E 10.5% (2/19)	ERBB2 8% (4/52)	ERBB2 3% (1/34)	EGFR Sensitizing 11.1% (2/18)
RET 10.5% (2/19)	RET 6% (3/52)	ALK 3% (1/34)
	ALK 2% (1/52)

Response assessment (RECIST) to targeted therapy showed complete response in 4%, partial response in 44%, stable disease in 37% and progressive disease in 15%. Response rate was 44% for group A, 50% for group B, 60% for group B1, 37.5% for group B2. Total objective response rate was 48% and disease control rate was 85%. Overall survival was evaluated for 40%, median was 14.4 months for patients who received targeted therapy vs 13.6 months for patients who received standard treatment.

Conclusion:
Comprehensive ctDNA testing revealed treatment options for 40.5% of patients analyzed. The highest impact was seen in progressors on EGFR therapy. These positive results emphasize the utility of liquid biopsy analysis to guide clinicians to select the most efficacious therapy for each patient.

Background:
To investigate the efficacy of gefitinib combined with recombinant human endostatin (endostar) as a salvage therapy in patients who failure to benefit from gefitinib in previous treatment of advanced non‐small cell lung cancer (NSCLC) due to resistance with T790M mutant negative or unknown.

Method:
Patients with pathologically confirmed stage IV NSCLC who had failed of disease control by gefitinib were retrospectively reviewed. In total, twenty patients were enrolled incluing two patients with squamous disease, eighteen patients were adenocarcinoma. The median age was 61.6, ranging from 43 to 75. After acquired gefitnib treatment resistance, six patients were tested their tumor samples by re‐biopsy and confirmed T790M negative. Fourteen patients refused to re-biopsy, therefore the T790M status were unknow among them. In our analysis, from September 2009 to May 2014, all patients met the selection standards above received endostar (15 mg/day, i.v. day 1‐7, 21 days per cycle) additionally and keep taking gefitinib after preious treatment failure. The efficacy and toxicities of the combined treatment were observed.

Result:
Two patient achieved partial response (PR) (10%) , twelve patients had stable disease (SD) (60%) , six of those patient had progressive disease (PD) (30%) . The objective response rate (ORR) was 10% and the disease control rate (DCR) was 70%. Median progression free survival was 4.2 months (95% CI, 3.21 ‐ 5.19 months) , whereas median OS time was 8.0 months (95% CI, 4.96 ‐ 11.04 months). The results of efficacy accessed in terms of DCR, PFS and OS were not correlated with the sex, age, performance status (PS) score, histologic types or the treatment duration of gefitinib.

Conclusion:
After the failed attempting of gefitnib treatment, the combination of Endostar with gefitinib treatment could stabilize disease in T790M negative or unknown patients, and prolonged their survival time with tolerable toxicities.

Background:
Plasma detection of circulating tumour DNA (ctDNA) with T790M mutation in the context of EGFR tyrosine kinase resistance has been shown to have high concordance with tissue biopsy specimens. In a public healthcare system, patients’ perceived value of a test and willingness to pay can inform policy decisions regarding implementation and funding of a novel technology.

Method:
As part of screening for the ASTRIS clinical trial (NCT02474355), Canadian patients were invited to participate in a national validation study of blood-based ctDNA T790M testing. Eligible patients had acquired resistance to EGFR TKI and consented to collection of blood samples, demographic data, and completion of a structured interview measuring their perceived value of blood-based ctDNA testing as an alternative to tumour biopsy. They were asked about their willingness to pay for testing using both open-ended and iterative bidding approaches. The study was supported by a grant from AstraZeneca.

Result:
60 patients were accrued to the study. Median age of the cohort is 64 years (range 31-87); 69% are Asian (40/58); 55% (33/60) are male. All patients had received prior EGFR kinase inhibitor treatment, with 67% (45/60) receiving gefitinib. 17% of patients also received chemotherapy (10/60). A median of 1 prior line of therapy had been received (range 1-6). All patients preferred to have the blood test over repeat tumour biopsy. Patients estimated a mean reasonable price to pay for the test of $954; median $300 (range 0-10,000; IQR 150-800). Patients were personally willing to pay a mean of $281; median $100 (range 0-2500; IQR 33-350).

Conclusion:
In a public health system that covers the cost of standard diagnostic tests, Canadian patients indicated a willingness to pay out of pocket for peripheral blood detection of ctT790M. Patients have high perceived value of ctDNA and prefer it to tumor biopsy.

Background:
The concomitance of epidermal growth factor receptor (EGFR) mutation and anaplastic lymphoma kinase (ALK) rearrangement defines a new molecular subtype of non-small-cell lung cancer (NSCLC). We investigated the factors associated with the efficacy of targeted therapy and resistance mechanisms in EGFR/ALK co-altered NSCLCs.

Method:
EGFR mutation was identified with direct sequencing or Scorpion amplification refractory mutation system (ARMS). Relatively low EGFR-mutant abundance is considered as sequencing (-)/ARMS (+), while high abundance as sequencing (+). ALK-positive is assessed with any of the 3 methods: fluorescence in situ hybridization (FISH), rapid amplification of cDNA ends -coupled polymerase chain reaction and sequencing or Ventana immunohistochemistry (IHC). Next-generation sequencing was employed to analyze genetic profiles in patients with specimens before and after targeted therapy.

Result:
From December 2011 to December 2016, sixteen patients were identified with concomitant EGFR/ALK co-alterations, accounting for 0.6% (16/2632) in NSCLC patients, 1.8% (16/867) in EGFR-mutant and 8.6% (16/185) in ALK-positive patients. Five ALK-IHC (-)/FISH (+)/EGFR (+) patients with EGFR-TKIs experienced 3 PR, 1 SD and 1 waitiing for response evaluation, with median PFS of 11 months. Three with relatively low EGFR-mutant abundance achieved PR with crizotinib, while three with relatively high EGFR-mutant abundance obtained 2 PR and 1 SD with EGFR-TKIs. Spatial and inter-tumoral heterogeneity was observed in one EGFR/ALK co-altered patient. (Figure 1B) Two patients with T790M, one with Met pathway activation and two with loss of EGFR mutation were found after resistance to EGFR-TKIs. One with KRAS mutation was found pre- and post-EGFR-TKIs. ALK_F1174C mutation was observed in one patient after progression to crizotinib and ALK_G1202R mutation after resistance to ceritinib.Figure 1



Conclusion:
ALK protein expression, EGFR mutation abundance and tumor heterogeneity were associated with efficacy of targeted treatment for EGFR/ALK co-altered patients. Most mechanisms resistance to EGFR-TKIs and crizotinib were similar to those in typical EGFR mutation and ALK rearrangement respectively.

Background:
Detection of EGFR mutations in circulating tumor DNA using plasma samples is non-invasive and safety method. Cobas EGFR mutation test v2 with tumor tissue has been approved for detection of EGFR mutations to offer EGFR-TKI therapy in Japan. Recently, cobas test v2 using plasma samples at disease progression after prior EGFR-TKI therapy was approved for detection of T790M mutation. However, diagnostic accuracy of this test using plasma samples has not yet known in clinical practice.

Method:
This study is a single-center retrospective study. Between May 2016 and June 2017, we obtained 70 plasma samples from 53 advanced NSCLC patients harboring EGFR mutation in our hospital, and evaluated concordance rate of tissue and plasma EGFR mutation status using sensitivity and specificity. All plasma samples were tested using cobas EGFR mutation test v2.

Result:
70 samples were classified into three cohorts, diagnostic cohort(n=22), continuing response cohort(n=10), and acquired resistance cohort(n=38) by clinical setting. In diagnostic cohort, sensitivity for major mutations was 89%(ex19=100%, L858R=83%, respectively). In all cohorts, sensitivity for major mutations was 70%(ex19=80%, L858R=65%, respectively). T790M detection was evaluated in acquired resistance cohort, sensitivity was 67%, and specificity was 74%, respectively.

Conclusion:
The cobas EGFR mutation test v2 using plasma samples had good accuracy for detection of EGFR mutation in diagnostic setting. On the other hand, there was moderate accuracy for T790M detection in acquired resistance setting.

Background:
Along the past 10 years determination of EGFR mutational status has becoming part of the standard clinical practice for NSCLC patients. It is also acknowledged that median age of NSCLC patients is progresively increasing and nowadays a significant percentage are elderly when diagnosed. The aim of this paper is to analize the characteristics of NSCLC-EGFR-mutated patients that were 75 years or older at diagnosis and assess their survival.

Method:
EGFR-mutated patients with advanced NSCLC treated betweeen June/2010 and June/2016 in our Department were analyzed. The subgroup of those >75 years have been reviewed for assessment.

Result:
Out of the 73 patients with EGFR mutation, 22 (30.1%) were 75 years or older. Median age was 80.5 y (75-86). Performance status: 0 3p, 1 8p, 2 10p, 3 1p. Gender: male 36.4%, female 63.6%. Stage IIIb 18.2%, IV 81.8%. 20p (90.9%) received therapy, all with a TKI (erlotinib, gefitinib or afatinib). The other 2p refused therapy. Median survival was 16 moths (m) (range 1-31). It was greater for patients with PS 0-1 (31m) and for females (32m) vs male (8m). No significant differences existed by stage or TKI. Lnadmark 2-year survival was 18.2% and 3-year 4.5%. These figures were lower than the corresponding for younger patients (median 21 months, 2-y 35.3%, 3-y 13.7).

Conclusion:
A significant percentage of NSCLC with EGFR mutation were older than 75 years. They tended to be more females and slightly worse PS than youger patients. Survival was also lower. These patients merit an specific approach, careful evaluation and follow-up.

Background:
The incidence of CNS metastases among patients (pts) with pulmonary adenocarcinoma harboring EGFR-activating mutations seems to be increasing, probably related to their prolonged overall survival. A multidisciplinary approach, including radiation therapy, surgery and EGFR-TKIs, is absolutely essential for the management of these pts. We aimed to study the effectiveness of this multidisciplinary approach of CNS metastases in pts not previously treated with EGFR-TKis.

Method:
It is a retrospective, uni-institutional study, and all pts were consecutively identified and treated between 2009 and 2017. Eligible pts had to be diagnosed with pulmonary adenocarcinoma harboring EGFR-activating mutations (Sanger sequencing performed in either archived, or new, formalin-fixed and paraffin-embedded samples). CNS involvement was diagnosed based on CT scans, MRI and/or cerebrospinal fluid findings. All studied pts were treated with EGFR-TKIs (erlotinib or gefitinib) and other multimodality therapies including radiation therapy (whole brain or stereotactic radiosurgery), metastasectomy and/or intrathecal methotrexate, based on the decision of the multidisciplinary team.

Result:
35 pts were studied, with a median age 63 y.o. (35-90), being 26 (74%) female, 19 (54%) never-smokers, and 26 (74%) ECOG-PS 0-2. All pts received an EGFR-TKI (erlotinib or gefitinib), and 26 also were treated with radiation therapy, 11 were submitted to surgery, 2 to stereotactic radiosurgery and 1 received IT-MTX; 4/35 were treated with an EGFR-TKI only. Median treatment time with an EGFR-TKI was 7.6 mo. Median progression-free survival (mPFS) was 8.2 mo and the median overall survival (mOS) was 11.9 mo. Among those 25 pts whose tumors were diagnosed with an exon 19 deletion, mPFS was 8.2 mo. and mOS 11.9 mo., and among those 9 pts with L858R mutation, mPFS was 10.8 mo. and mOS 13.8 mo.

Conclusion:
Those pts diagnosed with pulmonary adenocarcinoma harboring EGFR-activating mutations treated with a multidisciplinary approach (including an EGFR-TKI) may derive promising PFS and OS results. The approach must be individualized, considering patient characteristics, tumor biology aspects and also the available healthcare resources.

Background:
Activating mutations in the epidermal growth factor receptor (EGFR) are strongly predictive of EGFR-tyrosine kinase inhibitor (TKI) activity in non-small cell lung cancer (NSCLC). However, primary resistance to EGFR-TKIs occurs in approximately 20-30% of NSCLC patients with EGFR mutations, acquired resistance is inevitable. The aim of study is to discover unknown resistant mechanisms and contribute to more precisely administrate advanced and metastatic NSCLC with EGFR mutations.

Method:
60 NSCLC patients with EGFR sensitive mutation were enrolled this study. All of patients received EGFR-TKI treatment. 21 of patients were primary resistance and 39 acquired resistance according to Jackman standard. Tumor tissues of all of patients were collected before EGFR-TKIs treatment, and rebiopsy tissues were gained after acquired resistance in 39 NSCLC patients. Whole exome sequencing were performed in Illumina HiSeq2000 platform. The captured sequencing data was further processed to identify somatic mutations, including single nucleotide variants (SNVs), short insertions/deletions (indels) and copy number variations (CNVs).

Result:
In primary resistance patients, 13 patients occurred rapid progress (PFS ≤60 days) were put into group 1, and other 8 patients with PFS within 90-180 days were into group 2; in acquired resistance patients, 9 patients were observed long-term clinical benefit (PFS≥540 days) were into group 3; remaining 30 patients with PFS between 180 to 540 days were into group 4. Median PFS were 29, 95, 761 and 311 days from group 1 to 4, respectively. More signaling pathways were activated in group 1, relative to other groups, including bypass activation, downstream signal activation, apoptotic pathways disturbance and EMT activation. Meanwhile, the activation of more signaling pathways were found in samples after acquired resistance compared with paired baseline samples. In all of baseline samples, 60.0% patients harbored TP53 mutations, and these mutations distributed in exon 2,4,5,6,7,8 and 11, respectively. Interestingly, TP53 mutations of 23% patients were in exon 6 in group 1, mutations in exon 5 occurred in 33.3% patients with long-term clinical benefit (group 3). Patients with exon 6 mutation had more shorter PFS than those with exon 5 mutation (105 days vs 284 days).

Conclusion:
For patients resistant to EGFR-TKI, more signal pathways were activation, and the heterogeneity of tumor cloning were complicated. TP53 mutations in different exons may have distinct effect on response to EGFR-TKI of patients with EGFR sensitive mutation.

Background:
IGNITE was a large multi-national, non-comparative, non-interventional diagnostic study whose objective was to in part determine the concordance in EGFR mutation status between tissue/cytology and plasma testing in the real world setting[1]. Locally advanced or metastatic NSCLC patients were enrolled from 9 countries in Asia-Pacific including Russia. A very significant difference was observed in performance of plasma testing in Russia (PPV=39.3%) vs. Asia Pac (PPV=92.5%) vs tissue results with only 51/84 ctDNA EGFR mutation positive results from Russian labs confirmed by the corresponding tissue result. No single lab appeared to contribute predominantly to the questionable overall performance. 1. Han, B et al.; ‘Determining the prevalence of EGFR mutations in Asian and Russian patients (pts) with advanced non-small-cell lung cancer (aNSCLC) of adenocarcinoma (ADC) and non-ADC histology: IGNITE study’ (ELCC 2015 Abstract No. 96O).

Method:
142 of the 941 evaluable plasma samples from Russia were selected to include originally discordant samples and a similar number of concordant positive and concordant negative samples. All samples were shipped to an experienced lab and re-tested for TKI sensitizing mutations using the QIAamp Circulating Nucleic Acid Kit and the Therascreen® EGFR Plasma RGQ PCR test (Qiagen). The central laboratory was blinded to the origin of the samples and original EGFR mutation result.

Result:
Adhering to package insert defined criteria for the internal control, 38 of 112 (34%) of cases were re-classified (19 to pos. and 19 to neg.). By independent clinical laboratory interpretation of the results, 55 of 142 (39%) cases were re-classified (25 to pos. and 30 to neg.). A marked improvement was seen in concordance with the local tissue result after re-testing from 38% to 69% with the false positive rate decreasing from 18.8% to 3.6%. Re-classification of 17 (81%) original ‘false positive’ and 17 (35 %) original ‘false negative’ plasma results validates the original tissue results and appropriate collection and handling of plasma.

Conclusion:
Low concordance between original tissue and plasma EGFR status for the TKI sensitizing mutations observed in the Russian cohort from the IGNITE study was primarily due analytical issues in plasma test methodology. The contribution of the specific analytical process (extraction, testing or analysis) could not be determined. This follow-up study has been communicated back to the Russian laboratories resulting in the appropriate validation and proficiency programs being put in place to ensure quality plasma EGFR ctDNA testing.

Background:
Brain metastases (BM) are a common cause of morbidity and mortality in patients with non-smallcell lung cancer (NSCLC). EGFR-TKIs, whole-brain radiotherapy (WBRT), stereotactic radiosurgery (SRS) and chemotherapy are treatment options available. But the optimal management, the combination or sequence, remains undefined. This study evaluated the efficacy of EGFR-mutant NSCLC patients with BM receiving multiple regimens and analyze the prognostic factors.

Method:
We retrospectively studied 45 EGFR-mutated NSCLC patients with BM, who had received EGFR TKIs、radiation therapies (WBRT or SRS) and pemetrexed based chemotherapy successively between 2010 and 2015 at Zhejiang Cancer Hospital. Patient follow-up by telephone was done until January 2016. Treatment response was evaluated and survival data were collected and analyzed.

Result:
Among the 45 patients, 22(48.9%) had an EGFR exon 19 deletion and 23(51.1%) had an EGFR exon 21 L858R mutation. Thirty (66.7%) patients received upfront radiation therapies (RT). 28 patients (62.2%) were treated with pemetrexed based chemotherapy as the first line treatment while 17 (37.8%) received first line EGFR TKIs instead. The median overall survival (OS) from diagnosis of BM was 28.0 months for the whole cohort (95% CI, 17.3-38.7 months). Patients with the exon 19 deletion had a longer median OS than patients with the exon 21 L858R mutation (not reached vs. 23.2 months, P=0.031). There was no difference in OS between the upfront RT group and the deferral group (26.5 vs. 28.0 months, P=0.57), and similar results were found between the first line chemotherapy group and the TKIs group (28.0 vs. 23.2 months, P=0.499). Patients with a higher GPA score showed better survival. In multivariate analysis, the prognosis was correlated with EGFR mutation type (P=0.017). Patients with extracranial metastases showed the strongest trend toward decreased OS, although it did not reach statistical significance (P = 0.070).

Conclusion:
Applying three treatments can significantly improve OS of patients, especially those with EGFR exon 19 deletion. Different sequences of treatments and timing of RT need further studies to identify the optimal treatment model.

Background:
Seveal studies have demonstrated targeting EGFR mutations and tumor angiogenesis simultaneously has synergistic effect in the 1[st] line setting in EGFR mutant NSCLC. However, in JO25567 trial, the ≥grade 3 hypertension incidence with combination therapy was much higher (60%) when compared to historic incidence of hypertension with bevacizumab (10-15%). Considering relatively shorter half-lives for small molecule tyrosine kinase inhibitors, it might be a better choice to combine EGFR TKI and VEGFR TKI when it comes to hypertension management. Fruquintinib is a potent and highly selective oral kinase inhibitor targeting VEGFR and it has demonstrated favorable benefit-to-risk profile in third line treatment in NSCLC patients.Thus it is important to assess safety, tolerability and efficacy of this new combination in the 1[st] line setting in EGFR mutant NSCLC patients. NCT02976116

Method:
This is a single arm, open-label, multi-center study. All patients will receive gefitinib continuously at 250 mg qd. Fruquintinib will be given at 4 mg as starting dose for 3 weeks followed by 1 week off in the first 4-week cycle. Fruquintinib dose can be escalated to 5 mg with the same 4-week cycle if no ≥grade 3 adverse event (AE) or ≥grades 2 liver dysfunction occurs in the first cycle. Treatment continues until disease progression, unacceptable toxicity, or patient withdrawal. The primary objective is to assess the safety and tolerability of this combination. Key eligibility criteria include: histologically or cytologically confirmed NSCLC, ECOG PS 0-1, no prior systematic treatment, no brain metastasis. Key exclusion criteria include: known T790M mutation and bleeding history within 1 month before enrollment.

Result:
As of Jun 20, 2017, 9 patients have been enrolled and received at least one dose of fruquintinib and gefitinib. Six patients had L858R mutations, and three patients had exon 19 deletions. All patients reported AEs, but only one patient (11.1%) had grade 3 proteinuria. No SAE was reported. The most common AEs were increased ALT (3 [33.3%] patients), increased AST (3 [33.3%] patients), increased TBIL (3 [33.3%] patients), proteinuria (3 [33.3%] patients) and rash (3 [33.3%] patients). Fruquintinib dose reduction occurred in 3 patients due to grade 3 proteinuria, grade 2 increased ALT and grade 2 hemoptysis, respectively.

Conclusion:
The study is ongoing. As of the cut-off date, no unexpected toxicities were identified. The combination of fruquintinib and gefitinib showed an expected and manageable preliminary safety profile. Additional patients and follow-up data are required to further confirm the full potential of this combination treatment.

Background:
Despite the likelihood of an initial response to an EGFR TKI, NSCLC patients with an activating EGFR mutation eventually develop disease progression. Anti-angiogenic agents in combination with an EGFR TKI may provide additional benefit in EGFR-mutant NSCLC, but additional studies are needed to confirm the benefit.

Method:
This ongoing phase 1b/3 study (NCT02411448; RELAY) enrolled patients with previously untreated stage IV NSCLC, ECOG PS 0-1, and an activating EGFR mutation (exon 19 deletion or exon 21 L858R substitution). In part A (phase 1b), patients received ramucirumab (anti-VEGFR2 antibody) 10 mg/kg intravenously on day 1 of a 14-day cycle and oral erlotinib at 150 mg/day. Treatment continued until disease progression or unacceptable toxicity. The primary objective of part A was to assess safety and tolerability, in terms of dose-limiting toxicities (DLTs) during the first two cycles of therapy, and to determine the recommended dose for Part B (phase 3). Data cutoff was 31-May-2017.

Result:
Fourteen patients were enrolled and treated in part A. Two patients discontinued prior to completion of cycle 2, due to non-DLT adverse events of grade 2 interstitial pneumonia and grade 1 hemoptysis and were therefore not eligible for DLT assessment. Of the 12 DLT-evaluable patients (Japan, n=6; US/Europe, n=6), median age was 72 years (range 51-83), 83% were female, and 75% had ECOG PS of 1. Median duration of therapy was 64.3 weeks (interquartile range [IQR] 19.5-89.0) with ramucirumab and 68 weeks (IQR 44-95) with erlotinib. Treatment-emergent adverse events (TEAEs) occurred in all patients, most commonly rash (100%), diarrhea (92%), paronychia (67%), hypertension (58%), and dry skin (58%). Ten (83%) patients experienced grade 3 TEAEs (hypertension [n=4], rash [n=3], diarrhea [n=2], neutropenia, conjunctivitis, elevation of alanine aminotransferase [DLT; resolved within 4 days], and elevation of aspartate aminotransferase). No serious or grade 4-5 adverse events occurred. Median PFS was 17.1 months (95% CI 8.8-NR; 50% censored) and the PFS survival rate at 21-months was 46.9%. Five patients remain on study treatment.

Conclusion:
Ramucirumab plus erlotinib demonstrated encouraging clinical activity with no unexpected toxicities in Part A. Randomization into Part B began in January-2016, maintaining the dose of ramucirumab at 10 mg/kg Q2W with oral erlotinib at 150 mg/day.

Background:
Second-generation EGFR tyrosine-kinase inhibitor dacomitinib has shown encouraging activity as first-line therapy in patients with EGFR-activating mutation-positive (EGFR[+]) advanced NSCLC. We performed the first randomized, open-label phase 3 trial comparing dacomitinib with gefitinib as first-line therapy (NCT01774721) which demonstrated a clinically meaningful and statistically significant benefit of dacomitinib versus gefitinib (PFS per IRC: HR, 0.59 [95%CI, 0.47–0.74]; 1-sided P<0.0001; median PFS, 14.7 vs 9.2 months). We present results from Japanese patients enrolled in this ongoing study.

Method:
Patients with newly diagnosed stage IIIB/IV recurrent NSCLC harboring an EGFR-activating mutation (exon 19 deletion or exon 21 L858R ± exon 20 T790M) were randomized 1:1 to once-daily oral dacomitinib 45 mg or gefitinib 250 mg until disease progression or discontinuation. Patients with CNS mets excluded. Stratification was by race and EGFR mutation subtype. The primary endpoint was progression-free survival (PFS) per blinded independent review committee (IRC).

Result:
Among 452 patients enrolled in ARCHER 1050, 81 were Japanese. Slight imbalances in baseline characteristics were observed (Table). PFS and duration of response improvement in Japanese patients was consistent with global results.

Japanese Intention-to-Treat Population
	Dacomitinib (n = 40) n (%)	Gefitinib (n = 41) n (%)	Unstratified HR [95% CI] 1-sided p-value
Male	15 (37.5)	20 (48.8)
Age, years <65 ≥65	19 (47.5) 21 (52.5)	15 (36.6) 26 (63.4)
Smoking status Never smoked Ex-smoker Smoker	19 (47.5) 20 (50.0) 1 (2.5)	24 (58.5) 16 (39.0) 1 (2.4)
ECOG PS 0 1	28 (70.0) 12 (30.0)	21 (51.2) 20 (48.8)
	Median, months	Median, months
PFS per IRC	18.2 (95% CI, 11.0–31.3)	9.3 (95% CI, 7.4–14.7)	0.54 (95% CI, 0.31–0.95) P=0.0141
PFS per INV	18.3 (95% CI, 14.6–22.1)	10.2 (95% CI, 7.3–16.9)	0.61 (95% CI, 0.36–1.04) P=0.0334
DoR per IRC in responders	# of responders=30 17.5 (95% CI, 10.2–34.3)	# of responders=31 8.3 (95% CI, 5.6–12.9)	0.44 (95% CI, 0.22–0.84) P=0.0056
CI, confidence interval; DoR, duration of response; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; INV, investigator assessment.

Objective response rates per IRC were similar (dacomitinib, 75.0% [95%CI, 58.8–87.3]; gefitinib, 75.6% [95%CI, 59.7–87.6]; 2-sided P=0.9579). Overall survival data are not mature. All 81 patients received study treatment. No grade 4/5 adverse events (AEs were observed with dacomitinib, while 3 grade 4 AEs and 1 grade 5 AE (disease progression) occurred with gefitinib. The most common grade 3 AEs were dermatitis acneiform (27.5%) and paronychia (22.5%) with dacomitinib and alanine aminotransferase increased (12.2%) and abnormal hepatic function (7.3%) with gefitinib. No new safety signals were identified.

Conclusion:
Dacomitinib significantly improved PFS and duration of response over gefitinib in first-line treatment of Japanese patients with advanced EGFR[+] NSCLC, with a manageable safety profile.

Background:
Overexpression of several receptor tyrosine kinases (RTKs) substitutes EGFR signaling in EGFR-mutant NSCLC. The MET ligand hepatocyte growth factor (HGF) provides an alternative signaling mechanism for EGFR by inducing inter-receptor cross talk with EphA2, CUB domain-containing protein-1 (CDCP1) or AXL. SHP2, a non-receptor protein tyrosine phosphatase is central in signal transduction downstream of RTK signaling and in Src activation. We previously demonstrated that STAT3 and Src-YAP1 signaling limits EGFR TKI efficacy in EGFR-mutant NSCLC. We are now exploring the possibility of multiple RTK activation through a Src-YAP1-mediated transcriptional program. We are evaluating whether combined EGFR inhibition with TPX-0005, a novel orally available multikinase inhibitor and potent Src/FAK and JAK2 inhibitor, can be more efficient than EGFR inhibition alone in EGFR-mutant NSCLC cells.

Method:
We studied the mRNA expression levels of stromal HGF and tumor RTKs, AXL, CDCP1, MET, and EphA2, as well as SHP2, and clinical outcome in baseline samples of 64 EGFR-mutant NSCLC patients treated with first-line EGFR TKI. We combined in vitro approaches to explore whether gefitinib or osimertinib combined with TPX-0005 can abolish STAT3 and Src-YAP1 and downregulate the expression of RTKs.

Result:
High levels of AXL, CDCP1 and SHP2 mRNA expression were associated with worse outcome to EGFR TKI in 64 EGFR-mutant NSCLC patients. Median progression-free survival (PFS) was 14.5 and 23.4 months for patients with high and low AXL mRNA, respectively (p=0.0359). Median PFS was 9.1 and 20.2 months for patients with high and low CDCP1 mRNA, respectively (p=0.0179). Tumoral EPHA2 and MET or stromal HGF levels did not affect PFS. Median PFS was 11.4 and 24.1 months for patients with high and low SHP2 mRNA, respectively (p=0.0094). The combination of gefitinib/osimertinib with TPX-0005 resulted in highly synergistic suppression of cell viability and reduced colony formation in two EGFR-mutant cell lines. The combination abolished the EGFR inhibition-induced STAT3 and YAP1 phosphorylation, as confirmed by western blotting and immunofluorescence. The results of TaqMan quantitative-PCR assay revealed that gefitinib/osimertinib plus TPX-0005 reduced the mRNA levels of AXL, CDCP1 and MET, an effect that could not be obtained with EGFR inhibition alone. In vivo experiments are ongoing.

Conclusion:
AXL and CDCP1 are adverse predictive markers of PFS in EGFR-mutant NSCLC patients. STAT3 and Src-YAP1 signaling limits the efficacy EGFR TKI. Combined EGFR inhibition with TPX-0005 (currently in phase I clinical testing) is a particularly attractive strategy

Background:
Osimertinib is a third-generation, central nervous system active epidermal growth factor receptor (EGFR) – tyrosine kinase inhibitor (TKI) that potently and selectively inhibits both EGFR-sensitising and EGFR T790M resistance mutations. Osimertinib is approved for EGFR mutation positive non small cell lung cancer (NSCLC) patients who develop EGFR T790M resistant mutation and resistant to prior EGFR TKI. Osimertinib resistance pattern and clinical outcome after osimertinib treatment are undergoing intensive investigation.

Method:
Seventy-one EGFR-TKI resistant patients received osimertinib in the AURA study in one medical center. We excluded patients treated as first-line or who do not have detectable T790M mutation. We collect available data of pre-osimertinib treatment plasma and tissue and post-osimertinib plasma, tissue samples and tested for EGFR, HER2, K-ras, B-raf, mutations, ALK fusion and cMET or HER2 gene amplification. Clinical and pathological characteristics before and after osimertinib treatment were collected.

Result:
Of the 53 T790M-positive patients, 6 did not progress. Three and 18 patients discontinued osimertinib due to side effect or progression, respectively; 26 received osimertinib beyond progression (1.1 to 20.5 months); 7 patients received osimertinib combination after progression. Fourteen patients are still alive. Median progression-free survival(PFS), overall survival(OS) and post-progression survival (PPS) were 11.1 months, 16.9 months and 5.0 months, respectively (only progression patients). In 47 progressive patients, post progression EGFR plasma tests were available in 40 patients. T790M was detected by BEAMing in 12 patients (4 patients combined with C797S) and not detected in 28 patients (70%). OS and PPS was longest for patients with no detectable EGFR activating mutation and T790M in the plasma before and/or after osimertinib treatment. Patients who lost detectable T790M but maintained activating EGFR mutation in the plasma had shortest osimertinib PFS. Post progression tissue sample or pleural effusion tumor cells were available in 22 patients. Two patients developed small cell transformation, one patient developed squamous cell carcinoma. Post progression tissue or effusion genomic tests were performed (N= tested patient number) and showed T790M+ in 9 patients(N=18), C797S in 2 (N=12), cMET amplification in 5 (N=10), B-Raf V600 mutation in 1 (N=13), K-ras mutation in 1 (N=13) and no ALK, ROS1 and RET fusions.

Conclusion:
Heterogeneous resistance mechanisms develop after osimertinib treatment, in tumors retain T790M or losing T790M. Patients who have no detectable activating EGFR mutations in the plasma had best survival outcomes. Loss of T790M but maintainance activating EGFR mutations in the plasma correlated with short osimertinib PFS.

Background:
Afatinib 40mg/day is approved globally for first-line treatment of EGFR mutation-positive (EGFRm+) NSCLC. Afatinib is available in several tablet strengths (20/30/40/50mg), and tolerability-guided dose adjustment schemes are well established. Here, we evaluate the impact of afatinib dose reduction on safety (AEs), pharmacokinetics, PFS and patient-reported outcomes (PROs) in the Phase III LUX-Lung (LL) 3 and 6 trials.

Method:
Treatment-naïve patients with stage IIIB/IV EGFRm+ NSCLC in LL3/6 received either 40mg/day afatinib or chemotherapy. In case of any treatment-related grade ≥3 AEs or selected prolonged grade 2 AEs, afatinib dose was reduced by 10mg decrements (minimum dose 20mg/day). In this post-hoc analysis of all afatinib-treated patients in LL3/6 (n=229/n=239), we compared incidence and severity of common AEs before and after dose reduction, afatinib plasma concentrations in patients who reduced to 30mg versus those remaining on 40mg, and PFS in patients with/without dose reductions in the first 6 months of treatment. PROs were measured using the European Organization for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire and the EQ-5D™ health status self-assessment questionnaire, and pooled data from both trials were assessed before/after dose reduction; these included scores on the EORTC Global Health/Quality of Life scale (GH/QoL; 0–100), EORTC Performance Functioning scale (PF; 0–100), EQ Visual Analogue Scale (VAS; 0–100) and EQ-5D UK utility scale (EQ UK utility; 0–1).

Result:
Dose reductions occurred in 122/229 (53.3%) patients in LL3 and 67/239 (28.0%) in LL6; >80% of dose reductions occurred in the first 6 months of treatment. Dose reductions decreased the incidence of treatment-related AEs (grade ≥3 AEs before/after dose reduction: LL3, 73%/20%; LL6, 81%/12%), and were more likely among patients who had higher afatinib plasma concentrations prior to subsequent dose reduction (Day 22). On Day 43, geometric mean afatinib plasma concentrations were comparable between patients who had dose reduced (n=59; 23.3ng/mL) and patients who remained on 40mg (n=284; 22.8ng/mL). Median PFS was comparable between patients with or without dose reductions in the first 6 months (LL3: 11.3 versus 11.0 months; HR [95% CI] 1.25 [0.91–1.72]; p=0.175; LL6: 12.3 versus 11.0 months; 1.00 [0.69–1.46]; p=0.982). There were no clinically meaningful changes in PROs following afatinib dose reduction: GH (40/30mg: 59.1/66.9; n=136); PF (79.4/83.0; n=136); EQ VAS (70.1/75.1; n=135); EQ UK utility (0.70/0.78; n=135).

Conclusion:
Tolerability-guided dose adjustments effectively reduced afatinib-related AEs without negatively affecting therapeutic efficacy and PROs.

Background:
In an open-label, Phase 3 randomized study (Clinical trial registry of India: CTRI/2015/08/006113), Gefitinib was found to be superior to Pemetrexed-Platinum in terms of progression-free survival in patients with EGFR mutated NSCLC. In this analysis, we aimed to assess the benefit of gefitinib over Pemetrexed platinum using quality-adjusted time without symptom or toxicity analysis method.

Method:
In this phase III, randomized study EGFR activating mutated patients were randomized in 1:1 fashion to either receive pemetrexed-carboplatin followed by maintenance pemetrexed or gefitinib till progression. Patients post progression received the treatment received in other arm if they were fit. These patients were followed up until death. Toxicity during the whole course of treatment was documented in accordance with CTCAE version 4.02 criteria. For QTWiST analysis, the overall survival and progression-free survival were documented. The overall survival was partitioned in 3 health states. These were TOX, TWiST, and REL. TOX state comprised of time in months spent by the patient in grade 2 or above toxicity post randomization but before the first progression. TWiST state comprised of time in months spent by the patient post randomization but before the first progression without grade 2 or above toxicity. REL state was defined as the time in months spent post the first progression until death. RStudio will be used for analysis. The data will be censored for this analysis on 30th June 2017. The restricted mean of all three health states would be calculated using non-parametric 500 boot straps. The time spent in each state will be weighted by a corresponding health-state utility coefficient for QTWiST calculation. A threshold utility analysis will be performed using utility values between 0-1. Where 0 represents a health state similar to death and 1 represents a perfect health state.

Result:
The restricted mean health state duration in months for each state with its 95% CI for each arm, the difference between the 2 arms for each health state with its 95%CI and corresponding p-value will be provided. The results of threshold utility analysis with the corresponding QTWiST difference between the 2 arms with p-value would be presented.

Conclusion:
LBA: Not applicable

Abstract not provided

Background:
Methylnaltrexone Bromide (MB) is a selective antagonist of opioid binding at the mu-receptor (μ or MOR receptor). Constipation is a quite common side effect in Non-Small-Cell-Lung-Cancer (NSCLC) patients receiving opioids for chronic pain, usually due to skeletal metastases. We set out to investigate if MB is effective in those patients who received opioids and complained of constipation.

Method:
Sixty-eight NSCLC patients with a life expectancy of at least three months were recruited for our study after providing written consent. All patients received either fentanyl as a transdermal patch, in its inhaled form or per os. Patients were randomized (1:1) to four weeks of treatment with either MB 12mg/0.6ml (n=22) administered subcutaneously (sc) or a placebo, on alternate days. We recorded the number of patients who defecated within four hours of the MB or placebo injection, and the number of patients needing a second dose of MB or placebo within six hours from the first dose. We recorded the side effects of this treatment. Patients were not allowed to use other laxatives.

Result:
In the MB group after one injection, fifty one patients (75%) had a bowel movement within four hours compared to nine placebo patients (13%), p=0.02. Fifteen patients in the MB group had a bowel movement after two or more doses of MB, raising the percentage of patients who responded to MB to 96%. The more severe the constipation, the higher the response with MB. The overall rate of adverse events was similar in the MB (43%) and placebo groups (42%). In the MB group, the most commonly reported adverse events were abdominal pain (16%), flatulence (16%), vomiting (11%), and nausea (13%). Most treatment related adverse events were rated as mild or moderate by the patients. Discontinuation due to adverse events occurred in 5% and 6% of patients in the MB and placebo groups, respectively.

Conclusion:
Methylnaltrexone Bromide has been shown to be superior to placebo in achieving defecation within a short time, in NSCLC patients with opioid-induced constipation, The more severe constipation, the higher the response with MB. There were no serious adverse events. We conclude that Methylnaltrexone Bromide is effective, safe and superior to other commonly used laxatives.

Background:
The optimal treatment strategy for Stage IIIB NSCLC patients with a T4N0-1 tumor is a matter of debate. In prospective combined modality series including surgery, the median overall survival (OS) is approximately 24 months. We hypothesized that results comparable to regimens including surgery can be achieved with concurrent chemoradiation in this patient group.

Method:
In our retrospectively collected database of NSCLC patients, all patients with T4 (mediastinal invasion) N0-1 NSCLC receiving concurrent chemoradiation were included. One patient had a recurrence after previous pneumonectomy. All patients were given 3 cycles of chemotherapy (cisplatin and etoposide). Radiotherapy (RT) was started at the 2nd course of chemotherapy. OS was calculated from date of diagnosis (Kaplan-Meier method). Toxicity was scored according to CTCAEv3.0.

Result:
42 patients (8 female, 34 male) with a median age of 62.5 ± 9 years (44-80 years) were included from January 2005 until December 2009. Stage distribution: 86% T4N0 (n=36), 14% T4N1 (n=6). The maximal tumor dose was 66 Gy using conventional fractionation. The median prescribed mean lung dose was 15 ± 4.4 Gy (5.03 -19.9 Gy). Acute toxicity: 1 patient experienced grade 3 dyspnea during RT. Grade 3 dysphagia occurred in 5 patients (12%) during RT requiring tube feeding in 3 of these patients (7%). Dysphagia persisted later than 1 month after RT in 1 patient (2%). Grade 3 dysphagia only occurred in patients treated concurrently. Grade 3 cough occurred in 1 patient during RT, no patient experienced grade 3 cough 1 month after RT. 2 patients died within 3 months after start of RT, one due to myocardial infarction, one of unknown causes. Severe late toxicity was not present: no grade 3 complications more than 3 months after the end of radiotherapy. With a median follow-up of 42 months, the median OS for the whole group is 34 months (95% CI 24-43 months). 2-year OS survival is 55%.

Conclusion:
Concurrent accelerated chemoradiation using an individualized dose prescription is a valid treatment strategy for stage IIIb, T4N0-1 NSCLC patients yielding very promising OS results with low toxicity.

Background:
The survival benefit of systemic chemotherapy has been demonstrated for treating patients with stage IIIa (N2+) lung cancer. The NCCN guideline recommends induction chemotherapy with or without irradiation followed by surgery for those patients if no disease progression was noted after induction therapy. However, there are also studies revealed the survival benefit of adjuvant chemotherapy for patients with N2+ IIIa disease. The current study compared the survival results of neoadjuvant (before surgery, BS) and adjuvant (after surgery, AS) chemotherapy plus surgical resection for the patients with non-small cell lung cancer with N2+ stage IIIa disease.

Method:
There were 217 patients with Stage IIIa N2+ who ercieved surgery resection in the recent decade in our hospital, with a mean follow-up duration of 44 months. The overall survival time was evaluated and compared between these three groups of patients

Result:
Figure 1There were 62, 44 and 111 patients without chemotherapy(C/T) (Nil) or C/T given as neoadjuvant (BS) and adjuvant (AS) setting respectively. There were more patients with advanced age in the Nil and AS groups and more patients with AS group have received sublobal resection (p<0.01 respectively) as compared to the patients of BS group. The mean survival duration after surgery for the patients of AS and BS groups was 57.6 and 50.4 months respectively which was signinficantly longer than those patients of Nil group (MST: 26.4 months : p<0.001 respectively). Multivariate analysis revealed the addition of chemotherapy as a single prognostic factor of the patients. However, there was no significant difference of survival duration between the patients of AS and BS groups.



Conclusion:
Chemotherapy given both as adjuvant or neoadjuvant setting can provide a survival benefit for the patients with stage IIIa N2+ non-small cell lung cancer after surgery. No statistical difference was observed about the survival duration for these two groups of patients.

Background:
Despite complete pre-operative staging, incidental N2 disease is still found during surgical resection of NSCLC. Proceeding with resection versus aborting the operation to treat with definitive chemotherapy and radiotherapy is controversial. The aim of the current study was to evaluate survival of pathological N2 disease after complete resection.

Method:
The Institut de Cardiologie et Pneumologie de Quebec Biobank was queried for all patients with pathological N2 NSCLC who underwent complete (R0) surgical resection either by lobectomy, bilobectomy or pneumonectomy between January 2000 and February 2017. Survival was examined using the Kaplan-Meier method with log rank analysis. Significance was set at p≤0.05.

Result:
We identified 224 eligible patients; 119 (53%) were male, mean age was 63±9, there were 143 (64%) adenocarcinoma and 60 (27%) squamous cell carcinoma. Regarding surgical modality, 156 (70%) patients underwent lobectomy or bilobectomy and 68 (30%) pneumonectomy. The 30-day mortality of the cohort was 3% (5 pneumonectomy and 2 lobectomy). During 17 years of follow-up, 142 (63%) patients died, including 87 (61%) in the lobectomy/bilobectomy group and 55 (39%) in the pneumonectomy group. Among all deaths, 105 (74%) were cancer-related. Median OS of the entire cohort was 2.6 y (CI 1.9-4.4). In the univariate analysis cox model, median OS was shorter for pneumonectomy than lobectomy/bilobectomy (2,1 years [1,6-2,6) vs 4,4 years [2,2-5,8]), HR 1.54 (CI 1.09 – 2.16, p=0.01; figure 1). However, when only considering cancer-related deaths, the difference was not statistically significant (p=0.95). Figure 1: Overall survival of pathological N2 according to type of resection. Figure 1



Conclusion:
In our institutional database study, median OS after surgical resection of N2 NSCLC was 2.6y. Pneumonectomy is indicated as cancer cure treatment however, major efforts should be made to decrease peri-operative morbidity and mortality.

Background:
According to current guidelines, invasive pre-operative staging should be performed with endoscopic ultrasound in NSCLC in suspected N2 disease. Due to its higher negative predictive value, in case of PET positive, CT enlarged mediastinal lymph nodes or central tumors, mediastinoscopy remains indicated when EBUS staging is negative. The aim of the current study was to evaluate OS of patients with locally advanced NSCLC who underwent surgical resection after negative EBUS and mediastinoscopy.

Method:
The Institut de Cardiologie et Pneumologie de Quebec Biobank was queried for all patients with high probability of N2 disease or central tumors with negative EBUS and mediastinoscopy that underwent complete surgical resection (R0) between March 2009 and February 2017. Survival was examined using the Kaplan-Meier method with log rank analysis. Significance was set at p≤0.05.

Result:
We identified 88 eligible patients (Table 1); 56 (64%) were male, mean age was 65±9 and 50% of the cases were adenocarcinoma. Regarding surgical modality, 1 (1%) patient underwent sublobar resection, 65 (74%) lobectomy or bilobectomy and 22 (25%) pneumonectomy. Among these, there were 11 (13%) pathological N2 cases. There was no 30-day mortality. During 8 years of follow-up, 30 patients died, including 20 (31%) in the lobectomy/bilobectomy group compared to 9 (41%) in the pneumonectomy group. We then identified 16 (80%) cancer-related deaths in the lobectomy/bilobectomy and 7 (78%) in the pneumonectomy group. Median OS of the entire cohort was 5.7 years, with no difference between groups (HR 1.29, CI 0.58-2.87, p=0.53). Table 1: Demographics of Locally advanced NSCLC Cohort

Characteristic	Lobectomy/Bilobectomy (n=65) N(%)	Pneumonectomy (n=22) N(%)	Total (87) N (%)	p value
Sex Male Female	44 (68) 21 (32)	11 (50) 11 (50)	55 (63) 32 (37)	0.200
Histology Squamous Adenocarcinoma Other	21 (32) 34 (53) 10 (15)	10 (45.5) 10 (45.5) 2 (1)	31 (36) 44 (50) 12 (14)	0.549
Mean age (years)	65.8 ± 8.2	63.4 ± 6.8	65.4 ± 8.1	0.216
Pathological stage IA IB IIA IIB IIIA IIIB	10 (15) 13 (20) 12 (18) 10 (15) 19 (30) 1 (2)	0 4 (18) 3 (14) 9 (40) 5 (23) 1 (5)	10 (11) 17 (20) 15 (17) 19 (22) 24 (28) 2 (2)	0.075

Conclusion:
In our institutional database, patients locally advanced NSCLC had 13% incidence of pathological N2 disease and the OS was 5.7y. Our data supports surgical complete resection either by lobectomy or pneumonectomy in this group of patients with locally advanced disease.

Background:
To optimize the personalized medicine for advanced non-small cell lung cancer (NSCLC), sufficient tumor tissue is mandatory to analyze molecular and genetic profile. The demand for repeat biopsy in NSCLC is increasing, it is more difficult to obtain specimen after initial treatment. The aim of this study was to evaluate the impact of surgical rebiopsy in advanced NSCLC.

Method:
From Jan 2014 to Mar 2017, 146 consecutive patients underwent surgical rebiopsy for NSCLC which was resistant to prior chemotherapy. Their medical record were reviewed retrospectively.

Result:
There were 60 male and 86 female patients with mean age of 57 years (range 30-83). Adenocarcinoma was most common histologic type (n=142, 93%). Among them, 107 patients represent EGFR mutation before chemotherapy, deletion in exon 19 (n=73) was most frequently observed. Before surgical rebiopsy, 121 patients (83%) were treated with EGFR-TKIs. The mean number of change in chemotherapy regimen was 2 (range 1-6) and 24% of patients underwent more than 3 different chemotherapy before rebiopsy. The median time between initial treatment and rebiopsy was 17.4 months (IQR 9-25). Surgical rebiopsy was possible in all cases. One hundred and seven patients (73%) underwent pleura biopsy, 22 underwent lung resection and 12 patients underwent both pleural and lung resection. Most procedure underwent video-assisted thoracic surgery (n=136, 93%), 10 patients required mini-thoracotomy. Median postoperative hospital stay was 4 days (IQR, 3-6) and the 30-day mortality was 2.7%. All specimens were confirmed as NSCLC and adequate for mutational and genetic analysis except 2 patients. One patient was failed to mutational analysis, other patients was failed to genetic sequencing due to low tumor volume. After surgery, 129 patients can resume chemotherapy. Of those, 85 patients were enrolled clinical trial or treated with new target agent. Thirty nine patients were treated with cytotoxic chemotherapy and 5 patients continued with prior target agent.

Conclusion:
Surgical rebiopsy can detect changes in cancer characteristics and may be used in therapeutic decision making in advanced NSCLC resistant to previous treatment.

Background:
Multiple randomized clinical trials have demonstrated that epidermal growth factor receptor (EGFR) exon 19 deletion (19Del) and exon 21 L858R mutation (L858R) are highly correlated with sensitivity to epidermal growth factor receptor tyrosine kinase inhibitorsinhibitor (EGFR-TKIs)TKI) treatment in non-small -cell lung cancer (NSCLC). A mutation in exon 20 (T790M) is reportedly associated with resistance to EGFR-TKIs. However, there are few studies have focused on those patients harboring double mutations in these three3 mutation sites.

Method:
Between March 2007 toand November 2016, 2546 Chinese patients with non-small cell lung cancer,NSCLC underwent EGFR mutation detections in tumor tissues at three3 medical institutions and were enrolled in this retrospectivelyretrospective study. Clinical characteristics of these patients, including the response to EGFR-TKIs and progression -free survival outcomesoutcome for EGFR-TKIsTKI treatment (PFS-TKIs) of these patientsTKI), were analyzed.

Result:
Forty-five patients (45/2546, 1.7%) harbored double mutations of 19Del, L858R, and T790M. Patients with EGFR double mutations were more likely to be non-smoker, PS smokers, have a performance status of 0-1, have adenocarcinoma, and be at stage III-IV. Twenty-four patients with EGFR double mutations received EGFR-TKIsTKI therapy. The objective response rate (ORR), disease control rate (DCR)), and median PFS-TKIsTKI were 25% (6/24), 62.5% (15/24) and 5.95 months, respectively. The ORR, DCR, and median PFS-TKIsTKI in patients with double mutations of 19Del and T790M were 50% (4/8), 75% (6/8)), and 16.5 months, respectively. In those patients with 19Del and L858R, the ORR, DCR, and median PFS-TKIsTKI were 18.2% (2/11), 45.5% (5/11)), and 3.3 months, respectively. The ORR, DCR, and median PFS-TKIsTKI were 0% (0/5), 80% (4/5)), and 3.0 months, respectively, in patients with concomitant mutations of L858R and T790M.

Conclusion:
The ORR, DCR, and median PFS-TKIsTKI in patients harboring EGFR double mutations were lower than in patients with a single EGFR -activating mutation. The differences ofin ORR and DCR were statistically insignificant between the three3 groups. Patients with double mutations of 19Del and T790M had longer PFS-TKIs than patients in the other two2 groups.

Background:
Activated circulating endothelial cells (aCECs) have been indicated as a potential biomarker for cancerous angiogenesis in varieties of malignancies. Furthermore, several studies have exhibited aCECs were related with progression-free survival (PFS) and overall survival (OS) in anti-angiogenesis therapy. Anlotinib is a TKI of VEGFR1/2/3, FGFR1-4, PDGFR α/β, and c-Kit. As a third-line and above treatment on advanced NSCLC, Anlotinib has shown an affirmatory efficacy in ALTER0303 controlled trial. Herein we investigated the connection between aCECs and PFS, OS and metastases burden in the trial.

Method:
Blood samples were collected at baseline (pre-therapeutically), the 7[th], 15[th], 21[th], 42[th,] 63[th] day of Anlotinib or placebo. aCECs was measured by Flow Cytometry. Receiver-operating characteristics (ROC) analysis was used to determine a cutoff value of baseline aCECs counts to divide them into high and low groups. The predicting value of aCECs for PFS was investigated by univariate survival analysis. Chi-square test for baseline aCECs counts and patients’ clinical characteristics before Anlotinib or placebo treatment was performed.

Result:
aCECs were obtained in 78 patients (Anlotinib n=49). No significant difference in baseline characteristics was found between two arms (P﹥0.05). High baseline aCECs count was statistically in connection with more metastatic lesions (﹥3) (c[2]= 4.905,P=0.027). 49 Anlotinib treated patients were divided into 35 and 14 according to the ratio of minimal aCECs counts at every time point and baseline (aCECs min/baseline), as <1 and ≥1. Median follow-up was 8.6 months. Patients with min/baseline<1 had longer median PFS than ones with min/baseline>1 (193 vs.124 days, HR=0.439, 95%CI 0.211-0.912, P=0.023. shown in Table1). However, no significant relation between PFS and aCECs min/baseline was found in control arm.

Table 1.Comparison of Progression Rate in Various aCECs min/baseline
	N	Progression Rate %			Log Rank χ2	P-value
		3 months	6 months	9 months
aCECs min/baseline≥1	14	36.5	52.4	84.1	5.149	0.023
aCECs min/baseline<1	35	20.4	26.7	47.3

Conclusion:
Decreased aCECs during an initial period of Anlotinib therapy may predict longer PFS and baseline aCECs count may be related with the number of metastatic lesions.

Background:
Apatinib exerts anti-tumor effects by selectively inhibiting VEGFR-2. A single-arm Phase 2 study of apatinib monotherapy in advanced non-squamous NSCLC patients showed promising response across multiple therapy lines (P3.02C-025, WCLC 2016 Abstracts). Here we report the updated efficacy and safety data, as well as the clinical benefit of continuing apatinib beyond initial progression.

Method:
Forty patients with previously heavily treated advanced non-squamous NSCLC were enrolled to receive apatinib until progression, unacceptable toxicity, withdrawal or death. After study discontinuation, apatinib monotherapy or combined therapy was allowed for patients on disease progression at the discretion of the investigators and under the consent of patients.

Result:
The cutoff date was March 12, 2017. The treatment duration of apatinib was 82 (43, 127) days with a mean dosage of 477.0 ± 85.3 mg/day. Thirty-eight patients were available for tumor response evaluation, and the best overall response rate (ORR) and disease control rate (DCR) were 21.1% and 76.3%, respectively. The median progression-free (PFS) and overall survival (OS) were 3.32 (95% CI, 2.37–4.86) and 9.26 (95% CI, 5.36–not estimable) months, respectively. Common adverse events (AEs) were hand-foot-skin reaction (HFSR) (30.0%), proteinuria (27.5%), hypertension (17.5%) and aphthous stomatitis (22.5%). Severe AEs included Grade 3 HFSR (5%), hypertension (5%) and thrombocytopenia (5%). Results of preliminary subgroup analyses indicated that age, gender, PS score, treatment line and having a driver gene mutation had no significant effects on ORR, DCR and survival. After initial progression following apatinib treatment, 9 patients received apatinib alone or combined therapy with docetaxel, gefitinib or erlotinib (Table). One PR and 6 SD were achieved. Encouragingly, 8 patients had treatment duration over 4 months.

Table: Patients continued apatinib alone or combined therapy after progression

No.	Regimens after progression	Dose of apatinib (mg)	Best efficacy	Treatment duration (months)	Reason for discontinue treatment
1	Apatinib plus Docetaxel	500	PR	13.11	Second progression
2	Apatinib plus Gefitinib	250	SD	7.98	Lost to follow-up
3	Apatinib plus Gefitinib	375	SD	7.82	Death
4	Apatinib plus Gefitinib	500	SD	5.82	Lost to follow-up
5	Apatinib plus Erlotinib	500	SD	4.27	Lost to follow-up
6	Apatinib	375	SD	5.13	Second progression
7	Apatinib	500	NE	4.44	Second progression
8	Apatinib	250	SD	4.24	Death
9	Apatinib	500	NE	0.33	Death

Conclusion:
This updated analysis further confirmed the efficacy and safety of apatinib for heavily treated advanced non-squamous NSCLC. Continuing apatinib monotherapy or combined therapy could bring clinical benefit.

Background:
The CAP/IASLC/AMP molecular testing guidelines recommend ALK and EGFR testing on patients with lung adenocarcinoma, regardless of clinical characteristics. While PD-L1 testing has recently become available, a potential challenge to implement this testing is limited tissue availability. NGS may address this issue but there are limited published data assessing the impact of PD-L1 testing on NSCLC biomarker testing and treatment patterns using large real-world data sources. The objective of this study is to describe real-world biomarker testing and treatment patterns in the United States (US).

Method:
Flatiron Health’s database is a longitudinal, demographically and geographically diverse database containing EHR data. The database includes over 265 cancer clinics (~800 sites of care) representing more than 1.7 million active US cancer patients. Patients with ≥2 visits after Jan 1, 2011, ≥18 years of age, treated in first line (1L), and stage IIIB/IV NSCLC diagnosis from Jan 2012 - Mar 2017 were included in this analysis. Results were stratified by year of diagnosis (2012-2015 vs. 2016+).

Result:
Of 21,514 patients identified, the majority (80%) were diagnosed with de novo disease and 76% presented with non-squamous histology. PD-L1 testing rates were higher in those diagnosed in 2016+ compared to 2012-2015 (36% vs. 7%). A larger proportion of patients were tested for at least one biomarker in 2016+ (75%) vs. 2012-2015 (65%). The use of NGS also doubled (10% in 2012-2015 vs. 21% in 2016+) during this time period. For all patients, biomarker positivity rates varied by biomarker (PD-L1: 34%, EGFR: 17%, ALK: 4%, ROS1: 2%, KRAS: 30%) and by histology with the exception of PD-L1. The percentage of patients who initiated 1L systemic therapy, prior to receiving their first positive biomarker test results, ranged from 17% to 27% depending on the biomarker test.

Conclusion:
The introduction of PD-L1 testing has coincided with an increase in the proportion of patients being tested for a biomarker, as well as an increase in NGS. NGS has previously been shown to be associated with the longest turn-around time, and up to 27% of patients initiate systemic therapy prior to receiving positive biomarker test results. Additional research to understand the resource implications and clinical outcomes of initiating systemic therapy prior to test results (rather than delaying therapy) is underway.

Background:
Anlotinib hydrochloride is a novel TKI targeting the VEGFR, FGFR, PDGFR and c-Kit. With the capability of inhibiting the tumor angiogenesis and tumor cell itself, anlotinib had showed significantly improvement in OS (9.63 vs. 6.30 months, HR=0.68, 95%CI 0.54-0.87, p=0.0018) and PFS (5.37 vs. 1.40 months, HR=0.26, 95%CI 0.21-0.33, p<0.0001) in ALTER 0303 study for refractory cancer, a randomized, double-blind, placebo-controlled Phase Ⅲ trial in China. Here, we report the main impact factors affecting the efficacy and prognosis of anlotinib based on the data from ALTER0303 to elucidate the most benefit population.

Method:
Analyzed data were collected from 294 patients that were enrolled in ALTER0303 trial and received anlotinib treatment between 4[th] March 2015 and 15[th] August 2016. The statistical analysis was conducted using SPSS19.0 software, in which the measuring and enumeration materials were described with Mean±SD and frequency/percentage respectively, Kaplan-Meier method was used for survival curves in survival analysis. Independent impact factors of OS and PFS were identified by univariate and multivariate analysis in Cox proportional hazards regression model (Significant level, α=0.05).

Result:
Several factors were discovered to be associated with the efficacy of Anlotinib treatment. The impact factors were presented in Tab1.

Tab1. Impact factors for PFS and OS analyzed by Cox proportional hazards regression model
	Independent risk factor	Independent protective factor
PFS	Ratio of granulocytes to lymphocytes at PD (HR=1.07, 95%CI 1.041-1.100, p<0.0001) Elevated ALP level (HR=1.553, 95%CI 1.142-2.112, p=0.005) Baseline sum of longest diameters of target lesions (HR=1.004, 95%CI 1.001-1.006, p=0.007)	Elevated TSH level (HR= 0.555, 95%CI 0.422-0.730, p<0.0001) Hypercholesteremia (HR=0.720, 95%CI 0.534-0.971, p=0.031) Hypertension (HR=0.482, 95%CI 0.370-0.628, p<0.0001) Hand-foot skin reaction (HR=0.489, 95%CI 0.373- 0.643, p<0.0001) Elevated LDL level (HR=0.630, 95%CI 0.437-0.909, p=0.014) Age (HR=0.987, 95%CI 0.975-0.999, p=0.039)
OS	Ratio of granulocytes to lymphocytes at PD (HR=1.116, 95%CI 1.081-1.151, p<0.0001) Baseline sum of longest diameters of target lesions (HR=1.006, 95%CI 1.003-1.008, p<0.0001) ECOG PS≥2 at PD (HR=2.245, 95%CI 1.704- 3.508, p<0.0001)	Elevated TSH level (HR=0.725, 95%CI 0.524- 1.005, p=0.053) Hypertriglyceridemia (HR=0.601, 95%CI 0.440-0.821, p<0.0001) Rash (HR=0.581, 95%CI 0.369-0.916, p=0.019) Female (HR=0.713, 95%CI 0.533-0.953, p=0.022)

Conclusion:
This analysis explored the possible impact factors of PFS and OS in Anlotinib treatment. Moreover, we provide real data for the prediction of Anlotinib efficacy and most benefit population through the baseline characteristics and variety of clinical index. However, further analysis in the larger scale study is still looking forward.

Background:
Treatment of non-small cell lung cancer (NSCLC) is quickly evolving. New biomarkers and targeted agents have been approved at a rapid pace. It is of high interest to patients, physicians and reimbursement institutions to investigate molecular testing and subsequent treatment decisions in routine practice.

Method:
The prospective, national clinical research platform CRISP recruits patients in up to 150 cancer centres in all therapeutic sectors in Germany. Patients will be followed until death or for a maximum of 3 years. Raw data from 717 patients recruited by 78 centers by April 03[rd] was analysed regarding molecular testing and 1[st]-line treatment.

Result:
Data on histology was available for 635 patients, 71% non-squamous carcinoma (nsqc), 18% squamous carcinoma (sqc). Median age was 67 years and 61% of patients were male. 11% of patients were never smokers. In patients with nsqc (n=507) molecular test rates for EGFR, ALK, ROS-1 and PD-L1 were 69%, 65%, 51% and 26%, respectively. The overall PD-L1 test rate (nsqc & sqc) was 21% in 2016 and has been 27% so far in 2017. Of patients for whom test results were available at time of analysis 58% (n=87) were PD-L1 positive (nsqc 60%, n=76 and sqc 46%, n=11). An EGFR alteration was detected in 16% (n=57), an ALK alteration in 8% (n=25), and a ROS-1 alteration in 4% (n=9) of nsqc patients. Overall, 53% of patients received a carboplatin-based, 22% a cisplatin-based, and 7% a platin-free chemotherapy, while 14% received targeted and 4% other (experimental) therapies. Patients with EGFR alterations (n=57) were most frequently treated with 1[st]-line afatinib (33%), erlotinib (12%), or gefitinib (11%). Patients with ALK alterations (n=25) were most frequently treated with 1[st]-line crizotinib (48%). Patients with PD-L1 positive tumours were most frequently treated with platinum based doublet therapies (carboplatin combined with gemcitabine/taxane/pemetrexed or cisplatin combined with pemetrexed) or with pembrolizumab. The use of 1[st]-line pembrolizumab increased from 7% to 16% from 2016 to 2017. An update with data collected until October 2017 (>1,100 patients expected) will be presented.

Conclusion:
For the first time, CRISP presents real life data from all therapeutic sectors in Germany. Patients are frequently tested for molecular alterations and more than half of the patients with molecular alterations can start 1[st]-line treatment with a targeted therapy. Supported by AstraZeneca, Boehringer Ingelheim, BMS, Celgene, Lilly, MSD, Novartis, Pfizer and Roche.

Background:
Nanoparticle albumin-bound paclitaxel (nab-paclitaxel) is an albumin-bound formulation of paclitaxel. This single-arm, phase II trial evaluated nab-paclitaxel monotherapy in pretreated patients with advanced non-small cell lung cancer (NSCLC).

Method:
In this multicentre, single arm phase II trial, we enrolled patients with advanced NSCLC who had previously treated more than one chemotherapy regimen. Patients received nab-paclitaxel 80 mg/m[2] days 1, 8, and 15 (21-day cycle). The primary endpoint was investigator-assessed overall response rate (ORR); secondary endpoints included overall survival (OS), progression-free survival (PFS), the disease control rate (DCR), and safety. The planned enrollment was 30 patients by Simon 2-stage minimax design.

Result:
We enrolled 30 patients. We analyzed endpoints about initially enrolled 24 cases that were available for the evaluation now. Sixty-three % of patients had previous treatment more than 2 regimens. The ORR and DCR were 25% (95% CI 8-42%) and 75%, respectively. Median PFS and OS were 5.8 months and 9.8 months, respectively. No new safety signals were reported; the most common grade ≥3 adverse events included neutropenia (54%), leukopenia (9%), and infection (13%).

Conclusion:
In patients with heavily advanced NSCLC, nab-paclitaxel demonstrated promising antitumor activity; further assessment of nab-paclitaxel monotherapy in this population of patients is supported.

Background:
Phase 3 trial has been mandatory to establish new treatment. However, molecular targeted agents were often approved based on phase 2 trials. There have not been fully investigated whether efficacy data in phase 2 would be replicated in phase 3.

Method:
We extracted phase 2 and 3 trials for advanced non-small cell lung cancer (NSCLC) using platinum doublet (Plt) or EGFR-tyrosine kinase inhibitor (TKI) monotherapy, published between 2005 and 2015. Overall response rate (ORR) and median progression-free survival (PFS) in each study were collected. We compared these data between phase 2 and 3.

Result:
155 phase 2 trials and 13 phase 3 trials were adopted as Plt trials, while 21 phase 2 trials and 6 phase 3 trials were adopted as TKI trials. Plt trials had larger sample size (median number of patients: 47 in phase 2, and 203 in phase 3) than TKI trials (median number of patients: 29 in phase 2, and 101.5 in phase 3). In Plt trials, median ORR and median of median PFS were 31% and 5.2 months in phase 2, while 27% and 4.7 months in phase 3. There was statistically significant difference between phase 2 and phase 3 in ORR and mPFS (p = 0.03 and 0.03, respectively). In TKI trials, median ORR and median of median PFS were 64.0% and 9.7 months in phase 2, while 64.5% and 10.9 months in phase 3. There were not significant difference between phase 2 and phase 3 either in ORR and mPFS (p = 0.88 and 0.31, respectively). Among TKI trials, equivalence of efficacy data between phase 2 and phase 3 was also investigated in ORR and mPFS, but not proved (p = 0.30 and 0.45, respectively).

Conclusion:
Efficacy of TKI in phase 2 trial was well replicated in phase 3 trial.

Background:
Bone metastasis is the most frequent complication in NSCLC resulting in osteolytic lesions. During the bone metastasis, there is an increase on the osteoclastogenesis in detriment of the bone formation processes. Epidermal growth factor receptor (EGFR) pathway is constitutively activated in NSCLC and it is known that EGFR binds Amphiregulin (AREG) that is overexpressed in several cancers such as colon, breast and lung. Moreover, its levels in plasma derived from NSCLC patients correlate with poor prognosis. AREG was recently described as a signaling molecule in exosomes derived from cancer cell lines. Exosomes have a key role in the cell-cell communication and they were indicated as important actors in metastatic niche preparation. For this reason, in the present work, we hypothesize a role of AREG carried by exosomes derived from NSCLC cells and plasma of NSCLC patients, in osteoclast differentiation.

Method:
Exosomes were isolated from CRL-2868 cells, by ultracentrifugation and characterized by Western Blotting (WB) and electron microscopy analysis. AREG expression and EGFR phosphorylation was evaluated by WB in, CRL-2868 cells, exosomes and exosomes isolated from plasma derived from NSCLC patients. The osteoclasts morphology was assessed by confocal microscopy and RANKL, MMP9 and TRAP mRNA expression were measured by Real time PCR and RANKL and MMP9 secretion was evaluated by ELISA. The human biological material used in this publication was provided by Biobank@UZA (Antwerp, Belgium; ID: BE71030031000) and Banco de muestras biologicas Centro de investigacion Medica Aplicada (CIMA) Universidad de Navarra.

Result:
Exosomes derived from NSCLC plasmacontains AREG that induces EGFR pathway activation in pre-osteoclasts increasing the expression of RANKL which is able to induce the expression of proteolytic enzymes, MMP9 and TRAP, well-known markers of osteoclastogenesis. AREG function has been confirmed by loss and gain experiments with recombinant and neutralazing AREG, furthermore, knockdown-AREG exosomes do not induce osteoclast differentiation. To conclude, exosomes released in plasma of NCSLC patients, contain AREG, and induce osteoclasts differentiation of human primary osteoclasts.

Conclusion:
Exosomal AREG induces EGFR pathway activation that can induce RANKL expression that in turn, increases the expression of MMP9 and TRAP initiating an osteolytic bone metastasis.

Background:
Tumor Treating Fields (TTFields) are a non-invasive, anti-mitotic treatment modality. TTFields disrupt the formation of the mitotic spindle, and dislocation of intracellular constituents. TTFields significantly extend the survival of newly diagnosed glioblastoma patients when combined with temozolomide. Efficacy of TTFields in NSCLC has been shown preclinically and in a phase I/II pilot study with pemetrexed, where overall survival (OS) improved by > 5 months vs historical controls. We hypothesize that adding TTFields to 2nd line therapies in advanced NSCLC will increase OS.

Method:
Patients (N=512) with squamous or non-squamous NSCLC are enrolled in this Phase 3 study LUNAR [NCT02973789]. Patients are stratified by 2[nd] line therapy (PD-1 inhibitor or docetaxel), histology (squamous vs. non-squamous) and geographical region. Key inclusion criteria are 1st disease progression (RECIST 1.1), ECOG 0-1, no prior surgery or radiation therapy, no electronic medical devices in the upper torso, and absence of brain metastasis.Docetaxel or PD-1 inhibitors (either nivolumab or pembrolizumab) are given at standard doses. TTFields are applied to the upper torso for at least 18 hours/day, allowing patients to maintain daily activities. TTFields are continued until progression in the thorax and/or liver according to the immune-related response criteria (irRC). Follow up is performed once q6 weeks, including CT scans of the chest and abdomen. On progression in the upper torso, patients are followed monthly for survival. The primary endpoint is superiority in OS between patients treated with TTFields in combination with either docetaxel or PD-1 inhibitors, compared to docetaxel or PD-1 inhibitors alone. A co-primary endpoint compares the OS in patients treated with TTFields and docetaxel to those treated with PD-1 inhibitors alone in a non-inferiority analysis. Secondary endpoints include progression-free survival, radiological response rate based on the irRC, quality of life based on the EORTC QLQ C30 questionnaire and severity & frequency of adverse events. The sample size is powered to detect a HR of 0.75 in TTFields-treated patients versus control group.

Result:
Section not applicable

Conclusion:
Section not applicable

Background:
TTFields is an antimitotic cancer treatment that utilizes alternating electric fields in the intermediate frequency range . TTFields are approved for treating Glioblastoma Multiforme (GBM), and a pivotal trial testing the efficacy of TTFields for treating brain metastases (METIS) is currently underway. TTFields are delivered in two orthogonal directions using 2 pairs of transducer arrays placed on the patient’s scalp. The field distribution within the brain depends on the position of the arrays on the head. Therefore, personalizing array placement to optimize field delivery to the tumor requires a deep understanding of how brain anatomy, tumor position and array position influence the field distribution within the brains of patients. Here we present an overview of computational studies investigating TTFields distribution within the brain

Method:
In order to simulate the delivery of TTFields to the head realistic computational models are constructed by segmenting MRI datasets of both healthy individuals and brain tumor patients. Both healthy and pathological tissues are identified and assigned appropriate dielectric properties. Virtual transducer arrays are placed on scalps of the models and TTFields are created within the models.

Result:
Studies show that the field distribution within the brain is heterogeneous and depends on the anatomy of the model and the location of the arrays on the scalp. By shifting the arrays on the head it is possible to increase the field intensity in the tumor bed by a factor of two or more relative to a generic layout in which arrays are geometrically centered on the head. Optimizing array position, it is possible to guarantee that field intensities within the tumor bed and large portions of the brain exceed the therapeutic threshold of 1 V/cm. One particular layout worth noting is a layout in which each array of one pair are laterally placed superficially to the lower region of the occipital lobe, and the two arrays of the second pair are placed on the calvarium and the superior aspect of the neck. This layout delivers fields above the therapeutic threshold to both the supratentorial and infratentorial regions of the brain, making it suitable for treating multi-focal disease with tumors or metastases in both these regions.

Conclusion:
Optimal delivery of TTFields is layout dependent. The study suggests that TTFields can be used to treat brain tumors and metastases throughout the brain, as well as multi-focal disease encompassing both the supratentorial and infratentorial regions of the brain.

Background:
Formalin-fixed paraffin-embedded tissue (FFPET) samples are invaluable sources for both clinical research and in vitro diagnostics. However, accurate detection of genetic mutations in FFPET is a major challenge due to artifactual results, due to sample age and quality.

Method:
In a pre-clinical study, we used 315 non-small cell lung cancer (NSCLC) FFPET-derived DNA (FFPET-DNA) samples to establish sample criteria reflecting the minimum DNA quality suitable for PCR by comparing the results of droplet digital PCR-based mutation test (ddEGFR test) and qPCR-based EGFR mutation test (cobas[®] EGFR test). Using this criteria, we conducted a retrospective comparative clinical study of the ddEGFR and cobas EGFR tests of 171 NSCLC FFPET-DNA samples.

Result:
Based on the pre-clinical study, the FFPET-DNA sample criterion was established as internal quality control (iQC) index ≥ 0.5 (iQC copies ≥ 500, using 3.3 ng [1000 genome equivalents] of FFPET-DNA), indicating that more than half of the input DNA was amplifiable. Based on this iQC index, an independent clinical study revealed that both tests were significantly concordant (overall percent agreement (OPA) = 92.98%). Discordants not detected by the cobas EGFR test were detected using the ddEGFR test, indicating that the higher sensitivity of the ddEGFR test is due to its lower limit of detection.

Conclusion:
iQC index is a reliable indicator of the quality of FFPET-DNA and could be used to prevent incorrect diagnoses arising from low-quality samples. Compared with the cobas EGFR test, the ddEGFR test exhibited superior analytical performance and at least equivalent clinical performance.

Background:
The treatment choice for lung cancer patients is dependent on the expertise of the treating physician and the tumor board meetings (TBM) in which a treatment plan is discussed multidisciplinary. However, despite availability of a national guideline, the treatment selection criteria for NSCLC patients are not very explicit. Consequently, this may result in a variation of treatment recommendations across TBM. In this study, we investigated the variation in treatment recommendations by TBM within the Netherlands for 3 patients. Furthermore, patient characteristics associated with the treatment choice were explored.

Method:
A qualitative study was performed across 9 TBM at general hospitals in the Netherlands of which three hospitals had a radiotherapy(RT) department. Three existing patient cases were selected. The full medical history including imaging examinations was provided. Patient 1: 92years, cT3-4N0-1M0 (stage III). Patient 2: 73years, cT3NxM1b oligometastatic disease with a single adrenal metastasis (stage IV). Patient 3: 66years, cT2b-3N2M0 (stage III). All patients had a good performance score (WHO 0-1). The discussions and treatment plans during TBM were audiotaped. Analysis of the tapes was performed using open and axial coding techniques.

Result:
For patient 1 the recommended treatment was radical RT (45%), palliative RT (33%), surgery + RT (11%) and restaging before any treatment recommendation (11%). For patient 2 the recommended treatment was concurrent chemoradiotherapy (CRT) (45%), concurrent CRT + adrenal resection (33%), palliative RT (11%) and sequential CRT (11%). For patient 3 the recommended treatment was concurrent CRT (89%) and sequential CRT (11%). The patient characteristic ‘age’ was quoted in 96% of the treatment discussions, followed by ‘performance score’(89%) and ‘medical history’(85%). Figure 1



Conclusion:
In this qualitative study, a large variation in recommended treatment across the Netherlands was observed. The most extreme variation was seen in the treatment recommendation for the elder patient (92years), with treatment plans ranging from palliative RT to radical surgery.

Background:
Patients with epidermal growth factor receptor (EGFR) mutant advanced non-small cell lung cancer receiving first-line EGFR-tyrosine kinase inhibitor (TKI) inevitably developed disease progression after 9-13 months.

Method:
Before 1[st] January 2017, patients were investigated for resistance mechanisms upon failure of first-line EGFR-TKI by tissue re-biopsy. Liquid biopsy to detect secondary T790M mutation in plasma cell free tumor-DNA (cfDNA) was only performed if tissue re-biopsy was not feasible. Starting 2017, liquid biopsy was the first-choice of investigation. Tumor re-biopsy was only perfomed when cfDNA was negative for T790M mutation or if the patients had rapidly enlarging tumors.

Result:
Of 45 patients who were tested, 31(68.9%) underwent tissue re-biopsy and 14 (31.1%) underwent liquid biopsy as the first investigation to determine the presence of T790M mutation. For the latter group, 4 (8.9%) patients subsequently also had tumor re-biopsy. T790M mutation was detected in 30 (66.7%) of the 45 patients. C-Met amplification was tested in 7 T790M mutation-negative patients for possible recruitment into a clinical trial with 4 of them showing c-Met amplification. Small cell lung cancer transformation and ALK rearrangement were detected in 2 (5.7%) and in 1 (2.9%) of the re-biopsy tissue specimens, respectively. The resistance mechanisms in 8 patients (17.8%) was unknown. In short, two-third (66.7%) of our patients had T790M mutation upon first-line EGFR-TKI failure; while another one-third (33.3%) failed first-line EGFR-TKI due to other resistance mechanisms. The demographic, clinical and treatment characteristics were equally distributed among these 2 groups of patients. (Table 1)

Table 1. Demographic, clinical and treatment characteristics of 45 patients with first-line EGFR-TKI treatment failure

Characteristic	Total patients (n = 45)	T790M mutation (n = 30)	Other resistance mechanims (n =15)	p value of univariate analysis
Gender, No (%). Male Female	18 (40.0) 27 (60.0)	13 (43.3) 17 (56.7)	5 (33.3) 10 (66.7)	0.780
Smoking history, No (%). Never Ex or current smoker	36 (80.0%) 9 (20.0%)	22 (73.3) 8 (26.7)	14(93..3) 1 (6.7)	0.963
EGFR mutation subtype, No (%) Exon 19 deletion Exon 21 L858R Others Unknown	26 (57.8) 16(35.6) 2 (4.4) 1(2.2)	17 (56.7) 12 (40.0) 1 (3.3) 0	9 (60.0) 4 (26.7) 1 (6.7) 1 (6.7)	0.999
Treatment received, No (%) EGFR-TKI EGFR-TKI followed by chemotherapy	34 (75.5) 11 (24.5)	22 (73.3) 8 (26.7)	12 (80.0) 3 (20.0)	0.484
Best tumor response, No (%) Partial response Stable disease Disease progression	38 (84.4) 6 (13.3) 1 (2.2)	27 (90.0) 2 (6.7) 1 (3.3)	11 (73.3) 4 (26.7) 0	0.419
Initial progression free survival, months Median	13.0	13.0	11.7	0.538

Conclusion:
T790M mutation is the commonest acquired resistance mechanism causing first-line EGFR-TKI treatment failure. There was no difference in the clinical and treatment characteristics between patients with and without acquired T790M mutation as causes of resistance to first-line EGFR-TKI treatment.

Background:
There remains an unmet need for additional treatment options in patients with advanced non-small cell lung cancer (NSCLC) refractory to chemotherapy and immunotherapy. The insulin-like growth factor (IGF) axis and cyclin D-cyclin-dependent kinase (CDK) 4/6-retinoblastoma pathway have been implicated in the pathogenesis and resistance mechanisms of a variety of cancers, including NSCLC. Binding of IGF-I and -II to the IGF receptor results in upregulation of cyclin D1, and subsequent progression through the cell cycle, thus providing rationale for the simultaneous inhibition of IGF-I and -II and CDK4/6. This trial assesses the maximum-tolerated dose (MTD)/recommended phase II dose (RP2D), safety and preliminary efficacy of the IGF-ligand-neutralizing antibody, xentuzumab, in combination with abemaciclib, a selective, small-molecule inhibitor of both CDK4 and 6, in patients with solid tumors. One of two expansion cohorts (Cohort E) will further characterize the safety, tolerability, pharmacokinetics (PK) and preliminary efficacy of the combination in patients with NSCLC.

Method:
Study BI 1280.18 (NCT03099174) is a phase Ib multicenter, non-randomized, open-label, dose escalation trial with four dose-finding cohorts (Cohorts A–D) followed by two expansion cohorts (Cohorts E, F). For the NSCLC cohort (Cohort E), eligible patients include adults ≥18 years (≥20 for Japan), with measurable or evaluable disease, adequate organ function, ECOG PS≤1, and CDK4/6 inhibitor-naïve stage IV NSCLC after 1–2 lines of therapy and failure after platinum-based chemotherapy and immunotherapy. Patients with NSCLC (Cohort E) will receive the combination at the RP2D determined in patients with solid tumors (Cohort A) who will receive xentuzumab (starting dose 1,000 mg weekly i.v.) plus abemaciclib (starting dose 150 mg every 12 hours). The primary endpoint of Cohort E is the objective response (OR) in patients with pre-treated advanced NSCLC; disease control, duration of disease control, time to OR, duration of OR, and progression-free survival (PFS) are secondary endpoints. This study will be conducted in the US, Europe and Japan. Patient screening started in May 2017. Target enrolment is ~88 patients, including ~20 patients with stage IV NSCLC refractory to platinum-based chemotherapy and immunotherapy.

Result:
Section not applicable

Conclusion:
Section not applicable

Background:
Accurate illness understanding is important for the delivery of effective care in patients with advanced cancers. However, substantial proportions of them and their caregivers are prone to mistakenly understand their prognoses and the goals of therapy. The aims of this study were to explore prognostic understanding at diagnosis in both patients with advanced lung cancer and their caregivers and to investigate how their understandings change with the laps of time after diagnosis.

Method:
A total of 245 patients with newly diagnosed advanced lung cancer (clinical stage IIIB or IV) and their 208 caregivers were recruited at Keio University and its 16 affiliated hospitals (the Keio Lung Oncology Group) in Japan between December 2013 and March 2016. We assessed their perceptions of prognosis and goals of therapy, and examined associations with their clinical status, sociodemographic characteristics, mood symptom, status of insurance, and self-reported quality of life (QOL), as well as the status of disclosure of information by their treating physicians. Participants were asked to complete the questionnaires at diagnosis and at multiple time points after diagnosis.

Result:
At the time of diagnosis, 21.7% of patients and 17.8% of caregivers mistakenly believed that the patients’ cancer was “completely curable.” Levels of depression in both patients and caregivers were significantly higher compared with those who had accurate perception of prognosis. After 3 and 6 months from the diagnosis, 18.4% and 20.0% of patients, and 17.2% and 13.4% of caregivers still believed completely curable cancer, respectively. Patients with sustained misunderstanding at 3 months after diagnosis showed significantly high functional well-being score (p =0.0077), indicating they believe they are still physically, socially, and emotionally in good condition. Of the patients and caregivers with prognostic misunderstanding at diagnosis, there were 12 patients and 11 caregivers who turned to recognize accurate perception of prognosis in 3 months later. These patients showed significantly low functional well-being score compared with those who had sustained prognostic misunderstanding(p=0.002), whereas they did not show any significant difference in performance status and the efficacy of treatment. In caregivers, there was no factor associated with accurate perception. Patients with accurate perception at 6 months after diagnosis significantly showed low physical well-being score (p=0.041).

Conclusion:
Over 15% of patients and caregivers misunderstood their prognosis. Accurate prognostic understanding at diagnosis was associated with poorer psychological status. These misunderstanding are sustained even after 6 months of diagnosis, unless their physical, social, and emotional conditions changed.

Background:
Osimertinib is an essential drug for the treatment of non-small cell lung cancer (NSCLC) with epidermal growth factor receptor (EGFR) T790M mutation and rebiopsy is recommended for detecting T790M. However, significance of repeating rebiopsy in NSCLCs that were T790M negative with first rebiopsy remains unclear. Here, we sought to clarify this issue using a retrospective cohort.

Method:
We reviewed the medical records of patients with consecutive advanced NSCLC harboring activating EGFR mutations who underwent EGFR tyrosine kinase inhibitor (TKI) at Okayama University Hospital between Jan 2015 and Jan 2017.

Result:
In total, 104 patients were included in this study, and 47 patients underwent rebiopsy after acquiring resistance to prior EGFR TKIs. Preexisting activating EGFR mutations were found in all the 47 rebiopsied samples. Nineteen of them were T790M positive (40%). In the remaining 28 patients (T790M negative with first rebiopsy), 18 patients underwent additional rebiopsies following to interval therapies. Eleven (61%) of them were T790M positive with 2nd/3rd rebiopsy (10 with 2nd rebiopsy and 1 with 3rd rebiopsy). In majority of the 11 patients, rebiopsied samples were obtained from different lesions between first and 2nd/3rd rebiopsy (8/11, 73%). We also evaluated the efficacy of osimertinib in the 11 patients who needed 2nd/3rd rebiopsy for detecting T790M. Osimertinib showed good activity with the objective response rate 56% and the median progression free survival 5.5 months (95% confidence interval 4.1-6.9), though it is worse compared to with historical control osimertinib therapy.

Conclusion:
T790M could be found even in T790M negative NSCLCs with first rebiospy. Data will be updated at the meeting.

Background:
Primary pulmonary lymphoepithelioma-like carcinoma (LELC) is a rare subtype of non-small-cell lung cancer (NSCLC) and an Epstein-Barr virus (EBV)-associated epithelial neoplasm. We investigated the clinical significance of plasma concentrations of EBV DNA in pulmonary LELC patients.

Method:
Two independent sets of plasma samples from a total of 429 patients with pulmonary LELC patients (287 initial and 142 confirmatory) were available for EBV DNA determination. Plasma samples from the patients were subjected to a real-time quantitative polymerase chain reaction (qPCR) before treatment and three months after radical resection. Cutoff points were determined for pretreatment plasma EBV DNA (low, < 4000 copies/mL; high, ≥ 4000 copies/mL) on the basis of a measure of heterogeneity with the log-rank test statistic with respect to overall survival. Kaplan-Meier method and Cox regression were used to evaluate the relationship between plasma EBV DNA concentrations and clinical outcome. Among advanced stage pulmonary LELC patients who underwent sequential blood draws, we evaluated the relationship between change in disease status and change in EBV DNA concentrations using nonparametric tests.

Result:
High EBV DNA concentration was associated with poor OS in the initial, confirmatory, and combined data sets (combined data set: hazard ratio (HR), 3.67; 95% CI, 2.72 to 4.38; P < 0.001). These findings persisted after multivariable adjustment. Compared with low EBV DNA concentration, high EBV DNA concentration was associated with poor OS in patients with any stage. High EBV DNA concentration was also associated with poor disease-free survival (DFS) in patients with stage I/II disease. Patients with persistently detectable plasma EBV DNA had significantly poor OS (P < 0.001) and DFS (P < 0.001) than patients with undetectable EBV DNA three months after radical resection. In patients who underwent sequential evaluation of EBV DNA, an association was identified between an increase in EBV DNA concentration and a poor response to treatment and disease progression of pulmonary LELC.

Conclusion:
High baseline EBV DNA concentration is an independent poor prognostic marker in pulmonary LELC patients. These results should be confirmed in larger prospective trials.

Background:
Treatment of EGFR-mutated NSCLC patients with acquired resistance to EGFR tyrosine kinase inhibitors (TKI) relies on identification of the mechanism of resistance.

Method:
We performed a retrospective multicentric study in order to investigate use and results of liquid (circulating free DNA) and solid rebiopsies in routine practice.

Result:
We included 95 EGFR-mutated NSCLC patients with at least one rebiopsy after treatment with EGFR TKI (mostly 1[st] generation TKI). 87 solid rebiopsies and 71 liquid rebiopsies were performed. The number of liquid biopsies increased over time, from 1/y (6,6% of all rebiopsies in 2014) to 53/y (70,6% of all rebiopsies in 2016). The proportion of liquid biopsies increased with the number of rebiopsies per patient, from 35,5% for the first rebiopsy to 83,3% for the third rebiopsy. The rebiopsy identified a mechanism of acquired resistance in 48 patients (50,5%), including 43 patients with a T790M mutation (45,2%), 2 patients (2,1%) with MET amplification, 1 patient (1%) with small cell lung cancer transformation, 1 patient (1%) with a C797S mutation and 1 patient (1%) with a KRAS mutation. The initial EGFR mutation was found in 74 solid rebiopsies (85%) and 43 liquid rebiopsies (60%). The T790M mutation was found in 32 solid rebiopsies (36,8%) and 18 liquid rebiopsies (25,3%). Among 44 patients having both liquid and solid rebiopsies performed at the same time, an EGFR mutation was found by both techniques in 26 cases (59,1%) and the overall concordance was 88,6%. Analysis of cerebrospinal fluid was positive in 10 patients (100%) for the initial EGFR mutation and 1 patient (10%) for the T790M mutation. A T790M mutation was identified in 37,3% of all first rebiopsies. Repeated rebiopsies when the first biopsy was negative identified the T790M mutation in 29,6% of cases.

Conclusion:
In our series representative of daily practice, rebiopsies of EGFR-mutated NSCLC patients with acquired resistance to EGFR TKI identified a mechanism of resistance in half of the cases. Repeating rebiopsies increased the chance of detecting a T790M mutation.

Background:
Galectin-3 is a β-galactoside binding lectin, which is associated with cell proliferation, cell adhesion, apoptosis, angiogenesis and metastasis in several types of malignancies. However, the role of galectin-3 in the prognosis of lung cancer remains to be elucidated. In this study, we examined whether tissue expression and serum level of galectin-3 could serve as a prognostic marker in NSCLC.

Method:
Galectin-3 expression was analyzed by immunohistochemistry for 42 NSCLC patients who had undergone the radical surgery between 2014 and 2016. Serum galectin-3 level was assessed before surgery and 1 month after surgery by the enzyme-linked immunosorbent assay method.

Result:
Galectin-3 expression was observed in 14 patients (33.3%), and significantly associated with the incidence of recurrences (p<0.001) and the relapse-free survival time (p<0.001). Serum galectin-3 level was not reduced after radical resection, and there was no significant correlation between tissue expression and serum level of galectin-3.

Conclusion:
These data suggest that galectin-3 expression on tumor cells would serve as a prognostic marker in NSCLC, but serum galectin-3 level has no prognostic role on recurrence.

Background:
MET pathway is dysregulated in several cancers, having three types of alterations: overexpression, amplification and mutation. MET amplification has been reported in 2-4% of NSCLC while MET exon 14 skipping mutation has been found in 3-4% of lung adenocarcinoma. Increased MET expression is associated with higher malignancy, and acts as a negative prognostic factor. To date, several MET inhibitors are in development. Hence, the importance of understanding MET biology, its incidence, patterns of appearance, and which type of genomic abnormality. Is MET a trunchal genomic abnormality or does it appear as a result of mechanism of resistance of tumor cells once exposed to therapeutic agents? In the past, MET inhibitors have failed to improve OS in clinical trials; lack of benefit from this approach was the fact that there was no standard method for detecting MET related anomalies that allowed an adequate patient (pt) selection.

Method:
A search for MET alterations was performed in a retrospective analysis from November 2014 until May 2017. A total of 142 liquid biopsies (LBx) from pts with advanced NSCLC using NGS were identified. Once pts with a MET genomic alteration were identified by LBx, we proceed to review their initial tumor biopsy (TBx) genetic analyses and assess for the presence of MET alterations. Data regarding histologic subtypes, therapy received in each case, and timing of LBx analysis were collected.

Result:
A total of 142 LBx were identified in our cohort of patients with advanced NSCLC from November 2014 until May 2017. These samples came from 127 pts; 20 pts (15.7%) had MET alterations identified by LBx. Characteristics of these 20 pts were: median age was 72 (range, 62-85); 15 pts (75%) were females; all patients had adenocarcinoma histology. Genomics alterations in LBx were distributed as: 5 amplifications and 15 mutations. Genetic analyses from the TBx of these 20 pts done at initial diagnosis were reviewed. There were 8 pts who had TBx and LBx done at the same time and at initial diagnosis; only 2 pts out of these 8 had MET alterations in both tests (amplifications). The other 6 pts (with both TBx and LBx done simultaneously) had MET alterations identified only by LBx despite other genetic abnormalities concordance found between TBx and LBx in 5/6 pts. In these pts, median number of alterations identified by LBx and TBx were 7 and 0.5, respectively. There were 12 pts who had LBx done after progression of disease. All these pts had MET alterations identified in LBx however, none of them had MET genomic abnormalities in their TBx at initial diagnosis. One pt who had MET exon 14 skipping mutation in LBx (at progression of disease) did not have TBx genetic analysis due to insufficient tumor tissue.

Conclusion:
In this pt sample, we were able to identify more sensitivity from LBx than TBx for detection of MET alteration at initial diagnosis when both tests were done simultaneously. Our pts had metastatic disease with lesions outside the lung parenchyma. This will increase the diagnostic yield of LBx. Also, tumor heterogeneity plays a major role for discrepancies between TBx and LBx. The median number of genomic alterations identified by LBx was also higher than the one reported by TBx. Interestingly, 12 pts with MET alterations at progression of disease did not have these anomalies at TBx (initial diagnosis). Many questions remain unanswered. A prospective clinical trial using NGS in both tissue and blood at initial diagnosis will help us to identify and understand the biology of MET. It seems that LBx could be a complimentary tool at initial assessment of pt with metastatic NSCLC.

Background:
Identifying tumor-specific mutations in plasma from cancer patients serves as a non-invasive supplement to taking biopsies. Targeted sequencing of the circulating cell-free DNA (cfDNA) is an efficient method, for screening for a number of relevant mutations. Different approaches of targeted sequencing have been optimised for clinical use on FFPE, e.g. the Ion Torrent Colon and Lung panel. The size of DNA extracted from FFPE tissue is comparable with that from cfDNA. We therefore investigated the performance of the clinically relevant Ion Torrent Colon and Lung panel on cfDNA.

Method:
We used the Horizon multiplex cfDNA standard with eight known mutations at concentrations of 5 % (5-6.3 %), 1 % (1-1.3 %) , 0.1 % (0.1-0.13 %) and no mutations (wild type), respectively, to test the reproducibility of the panel. We obtained plasma from healthy donors from the danish Blood Bank to set a baseline for the panel. Lastly, the panel was tested on 52 patient samples. Patient plasma samples are from a previously collected cohort of EGFR wild-type non-small cell lung cancer patients (: NCT02043002) All samples were sequenced using the Ion Torrent Oncomine Solid Tumor DNA kit (Colon and Lung panel) from Thermo Fisher. Sample preparation was performed using the Ion Torrent Chef and sequencing was performed on the Personal Genome Machine (PGM) system. Data was analyzed using the Torrent Suite software, and variants called by Ion Reporter.

Result:
No somatic mutations were identified in neither the Horizon multiplex wild type nor the cfDNA from healthy donors. The Horizon multiplex samples were sequenced three times in different runs, to test the reproducibility of the panel. For the 5 % sample all mutations were detected in all runs. For the 1 % sample the four mutations at 1.3 % where detected in all runs, while two out of three runs missed one mutation at 1 %. In both cases the mutation could be identified by visualization of the reads, but was not called. For the 0.1 % sample, no mutations were detected. After finishing the validation the panel was used for sequencing patient samples. Of the 52 samples, 47 were successfully sequenced (90 %), and COSMIC-verified mutations were identified in 32 samples.

Conclusion:
The panel reliably and reproducibly detects mutations down to 1.3 %. Mutations present in lower concentration can also be detected, but for reliable detection higher coverage is needed. Sequencing was successfully performed on a range of patient samples.

Background:
No molecular markers are available so far to predict the efficacy of treatment and to monitor outcome in patients with EGFR-ALK-ROS1 wild-type advanced non-small cell lung cancer (wtNSCLC). Detection of genetic alterations in plasma may allow to follow biological effects of treatment without invasive procedures. Aim of this study (validation cohort) is to evaluate the detection rate and to monitor tumor-associated mutations in plasma during treatment in wtNSCLC.

Method:
Advanced wtNSCLCs referring to our Institution and undergoing anti-cancer systemic treatment are prospectively enrolled. Tissue genetic alterations are screened by Sequenom massarray platform (Diatech) or next-generation sequencing (NGS) (Illumina). Plasma samples are collected at baseline, after the first cycle of treatment, at first radiological restaging and at radiological progression. DNA is extracted and analyzed for genetic alterations previously detected in tissue by using droplet digital PCR (ddPCR,Biorad) or real-time PCR (Diatech). Semi-quantitative index of fractional abundancy of mutated allele is used. Primary end-point of MAGIC1 is to assess the sensitivity of technologies to detect genetic alterations in plasma (target: 60% within +/-15%).

Result:
Since January 2017 we have prospectively enrolled 56 patients treated with platinum-based chemotherapy (28), single-agent chemotherapy (9) or immunotherapy (18). 21 tissue samples have been analyzed (by Sequenom) in order to detect mutations in EGFR, HER2, KRAS, NRAS, BRAF, ALK, MET, PI3K, DDR2: ten KRAS mutation (including 3 G12D, 4 G12C, 2 G13D, 1 G12A mutations), one BRAF V600E and three EGFR exon 20 mutations were detected. At baseline, a concordance plasma/tissue was confirmed in 5/10 KRAS, 0/1 exon 20 EGFR, and 1/1 BRAF mutations. Longitudinal exploratory analysis was performed in 8 patients and 4/5 mutations found at baseline were not detected at subsequent evaluations in patients not experiencing systemic progression. The only patient with persistent KRAS mutation in serial plasma samples experienced radiological progression. In 1/5 patients negative at baseline, one sub-clonal KRAS mutation emerged at the third evaluation when progression was recorded. NGS tissue analysis is ongoing and ddPCR will be used to detect and monitor other genetic alterations such as STK11/LKB1, TP53, AKT mutations. The validation cohort will include 68 patients. MAGIC2 will include an expansion cohort to monitor genetic alterations in plasma during immunotherapy and assess timing of modifications in plasma according to clinical benefit.

Conclusion:
Preliminary data indicate the feasibility of detecting KRAS and BRAF mutations in plasma in clinical practice and potential usefulness in tracking biological changes in tumors.

Background:
Non small cell lung cancer (NSCLC) is the primary cause of cancer-related death, and 5-years survival rate remains below 16% mainly because of disseminated disease, also in fully resected early stages. Biomarkers identifying patients with a higher risk of relapse could be very useful. Circulating microRNAs (miRNAs), represent promising markers in this setting.

Method:
A case series of 182 resected early stage (IA-IIIA) NSCLC, of which 99 adenocarcinoma (ADC) and 83 squamous cell carcinoma (SCC), was analyzed. Peripheral blood samples were collected from each patient before surgical resection and serum was obtained after centrifugation and stored at -80°C until miRNA extraction. A panel of 84 circulating miRNAs was analyzed by Real Time PCR. Data were normalized by means of an external spike in, cel-miR-39, and the mean of two most stable endogenous housekeeping chosen separately for ADC and SCC samples. miRNA expression was analyzed in relation to disease-free survival (DFS) through Cox regression model. Results are reported as hazard ratios (HRs) and 95% confidence intervals (CIs).

Result:
Of the 99 ADC, 45 (45.5%) had a relapse during the follow-up whereas among the 83 SCC patients, 50 relapses (60.2%) were observed. The minimum follow-up time was three years for both groups of patients. In the group of ADC patients, stage was significantly associated with DFS (HR stage II-IIIA vs stage I = 4.94 , 95% CI [2.71 - 9.02]). Multiple statistical analysis methods were used to analyze miRNA expression data. At univariate analysis, two miRNAs (miR-222-3p and miR-22-3p) were significantly associated with time to relapse (p = 0.033 and p = 0.041, respectively). The significance was not maintained after adjustment for multiple testing. In the group of SCC patients, stage of disease was significantly associated with DFS (HR stage II-IIIA vs stage I = 3.31, 95% CI [1.74 - 6.33]). Five miRNAs (let-7a-5p, miR-126-3p, miR-26a-5p, miR-130b-3p, miR-21-5p) were found significantly associated with DFS even after adjustment for multiple testing false discovery rate q-value <0.001.

Conclusion:
Pre-surgery circulating levels of let-7a-5p, miR-126-3p, miR-26a-5p, miR-130b-3p and miR-21-5p seem to be significantly correlated with prognosis in resected early stage SCC patients.

Background:
It remains a great challenge to differentiate solitary pulmonary nodules which might cause excessive medical care and unnecessary psychological burden. Disadvantages of previous image-based or invasive methods call for better diagnostic tools .

Method:
Based on several cohorts, we performed methylation profiling by high throughput bisulfite DNA sequencing in tissue samples to learn methylation patterns that differentiate cancerous from benign lesions by in-depth data mining. Then we filtered out methylation patterns exhibiting high background in cfDNA for plasma sample classification. Notably, given the usual low amount of ctDNA in plasma, we developed an ultra-sensitive library preparation method (AnchorIRIS[TM]) to perform targeted bisulfite DNA sequencing from as low as 1 ng of cell free DNA (cfDNA) or 1 mL of plasma.

Result:
In the training set (n = 230) which includes 129 malignant specimens with different subtypes (invasive adenocarcinoma, MIA, AIS, squamous carcinoma, small cell and large cell lung cancer) as well as 101 benign specimens in different categories (hamartoma, granuloma/tuberculosis, inflammation and infections), we were able to achieve a preliminary sensitivity of 94.3% ± 4.3% for identification of maglignacies, with a preliminary specificity of 93.6% ± 4.9% against all benign specimens. From an independent validation set of 145 plasma samples, this assay obtained a preliminary sensitivity of 78.5% and a preliminary specificity of 83.3% for differentiating patients with malignant tumor (n = 79) from patients with benign lesions and asymptomatic normal individuals (n = 66). Specifically, our assay is demonstrated to be highly sensitive towards early-stage lung cancer detection, with a preliminary sensitivity of 71.1% in 38 patients with stage Ia lung cancer and 87.5% in 16 patients with stage Ib lung cancer. To further confirm the clinical application of this model, an additional validation study was carried out in an independent center. Our assay obtained an overall sensitivity of 89.2% for 37 malignant tissue samples, particularly with a sensitivity of 100% for adenocarcinoma, and an overall specificity of 85.7% for 21 benign tissue samples. For plasma samples, our assay achieved a sensitivity and a specificity of 80% and 66.7% respectively for 20 malignant and 3 benign samples.

Conclusion:
We have developed a highly sensitive blood based non-invasive diagnostic assay for differentiation of solitary pulmonary nodules, which can aid clinical decisions for patients with a CT scan positive for lung nodules. To the best of our knowledge, this is the first study to examine the diagnostic value of high throughput targeted DNA methylation sequencing for early stage lung cancer.

Background:
Cell-free DNA (cfDNA) is an alternative non-invasive source to assess gene mutations which are necessary for precision medicine in cancer patients. Next-generation sequencing (NGS) can list multiple gene mutations in single testing and require low input of DNA. We evaluated NGS application to detect cfDNA mutations in lung adenocarcinoma (LAC).

Method:
Retrospectively, cfDNA was isolated from 12 LAC patients in Hiroshima University Hospital (first group). Then another cfDNA was isolated from 30 LAC patients (second group). Ion PGM AmpliSeq system was applied using two panels, Cancer Hotspot (50 genes) for first group and Colon and Lung Cancer Research (22 genes) for second group. Variants were manually reviewed. EGFR mutation in cfDNA and tumor DNA (tDNA) were matched. EGFR mutations in tDNA, derived from either cytology or biopsy specimen, were previously analysed using PNA-clamp PCR.

Result:
Stage were 1 IIA, 1 IIIA and 10 IV in first group and 3 IA, 7 IB, 3 IIIA and 17 IV in second group. Mean coverage in first and second group were 863 and 2915, respectively. Mutation were found in TP53 (33/42), EGFR (8/42), STK11 (5/42), PTEN (3/42), ERBB4 (2/42), SMAD4 (2/42), MET (2/42), KDR (8/12), APC (4/12), RB1 (2/12), GNAQ (1/12), KIT (1/12), MLH1 (1/12), RET (2/12), VHL (1/12), KRAS (1/42). KRAS (A146V) mutated in one patient with no EGFR mutation detected in tDNA. In first group, EGFR mutations from tDNA were identified in 5 of 12 patients. Concordant EGFR mutations between cfDNA and tDNA were found in 1 of 5 patients (Exon 19 deletion, Allele Frequency, AF 9.1% (11/121)). In second group, EGFR mutations from tDNA were identified in 15 of 30 patients. Concordant EGFR mutations between cfDNA and tDNA were found in 7 of 15 patients. AF in EGFR exon 19 deletion were 0.1% (3/2424), 0.2% (6/2973), 8% (272/3394) and 24% (470/1960) and in exon 21 mutations were 2% (61/3884), 0.3% (5/1702) and 57% (1711/3025). The sensitivity and specificity in first and second group were 25%, 100% and 40%, 100%, respectively. In both group, patients whose EGFR mutations were detected in both tDNA and cfDNA were all in stage IV.

Conclusion:
NGS using cfDNA is less invasive method to detect various mutations simultaneously, especially in advance stage. EGFR mutation detection in cfDNA by NGS achieved a high specificity. Reducing target genes or deep sequencing may increase the sensitivity of detecting mutations.

Background:
Exracellular vesicles(EVs) are endosome-derived nano-size (30-1000 nm) vesicles released from many cell types including cancer cells. Previous researches have demonstrated that the level of EVs is increased in cancer patients than healthy controls. This study was conducted to evaluate the variation of EVs-count in the proximal tumor-drainage vessel and peripheral vessels during surgery for VX2 rabbit lung cancer model and primary lung cancer patients

Method:
A total of 6 rabbits were used in this study (3 in normal group, 31 in lung cancer group). Rabbit VX2 lung cancer model was made by real-time computed tomographic guided inoculation of VX2 cancer. Two weeks after injection, blood was collected from rabbit peripheral vessel and pulmonary vein (proximal tumor-drainage vein). A total of 3 healthy controls and 31 patients with primary lung cancer who had pT2aN0 (stage IB) and underwent lobectomy were selected. For each patient, 3 ml of blood was sampled from the radial artery (peripheral vessel) before surgery and from pulmonary vein of the primary tumor site (proximal tumor-drainage vein) during surgical procedure. Healthy controls blood were collected from peripheral vessels. EVs were isolated by serial centrifugation followed by ExoQuick[TM] and quantitative analysis was performed by NanoSight and western blotting.

Result:
EVs-count was not different in normal rabbit model according to blood sampling sites (peripheral vessel: 2.78 x 10[8] particles/ml, pulmonary vessel: 2.64 x 10[8]; p = 0.104). However, in rabbit lung cancer model, EVs were increased by 623.5% in peripheral vessels (1.73 x 10[9 ]particles/ml; p = 0.003) and 787.9% in proximal tumor-drainage vein (2.08 x 10[9] particles/ml; p = 0.001) comparing to those of normal rabbits. Moreover, we confirmed that EVs-count in VX2 lung cancer model was increased by 120.0% (p = 0.05) on the proximal tumor-drainage vein than peripheral vessel. In human blood samples, peripheral blood derived EVs were increased by 181.6% in lung cancer patient in comparison with healthy controls (2.44 x 10[8] particles/ml in healthy control, 4.43 x 10[8] particles/ml in lung cancer patients; p = 0.04). And, EVs were significantly increased by 700.8% in pulmonary vein of the primary tumor site (1.71 x 10[9] particles/ml; p = 0.0001) comparing to peripheral vessels in lung cancer patients

Conclusion:
The increase of EVs was more prominent in tumor-drainage veins than peripheral vessels in animal cancer models and lung cancer patients. We suggest that increase of EVs from tumor-drainage veins may provide more relevant prognostic information of the lung cancer patients comparing to those from peripheral vessel after surgery.

Background:
In advanced non-small cell lung cancer (NSCLC), circulating tumor DNA (ctDNA) can be used to identify clinically actionable mutations when tissue is insufficient or unobtainable for genotyping. In early-stage NSCLC, the persistence of ctDNA after complete resection may suggest the presence of minimal residual disease and predict an increased risk of recurrence; however, data on the longitudinal use of ctDNA is very limited. The aim of this study is to assess the clinical feasibility of ctDNA assessment for the prediction of lung cancer recurrence following surgical resection.

Method:
Patients undergoing curative-intent surgery for NSCLC were prospectively enrolled. Peripheral blood samples were collected at pre-specified time points immediately prior to surgery and 2-3 weeks after surgery. Plasma cell-free DNA samples were analyzed with a novel 21-gene Digital Sequencing NGS panel (Guardant Health, Inc.) with a theoretical genomic sensitivity of > 90% for NSCLC. For each patient, the identities and quantities of somatic alterations identified in ctDNA were correlated to clinical outcomes.

Result:
Of the 55 pair-matched patients, 33 were evaluable at the time of abstract submission. The median age was 61 (18 men). The histologic type was adenocarcinoma in 19 (58%), squamous cell carcinoma in 14 (42%), and small cell lung cancer in 2 (6%). The pathologic stage was I in 20 (61%), II in 3 (9%), and IIIA in 10 (30%). Median clinical follow up was 13 months. Of the 4 patients with detectable post-operative ctDNA, 3 recurred within the follow-up period. Of the 8 total recurrences observed, the most common site was the brain (4) and lung (3). One of 4 adenocarcinoma recurrences, 1 of 2 squamous cell carcinoma recurrences, and 1 of 2 small cell recurrences were associated with detectable post-operative ctDNA. By stage, post-operative ctDNA was detected in 1 of 1 stage II and 3 of 7 stage IIIA patients. Recurrences not associated with detectable post-operative ctDNA were enriched in oligometastatic recurrence (4 of 5 unpredicted recurrences were in isolated lymph nodes or the brain).

Conclusion:
In this small pilot cohort, ctDNA detected at 2 weeks post resection was associated with recurrence (RR 9.38, p=0.038), while the absence of detectable ctDNA was not significantly associated with lack of recurrence (RR 0.65, p=0.12). Oligometastatic disease, especially in the brain, was a primary risk factor for ctDNA-negative recurrence. These findings establish proof of concept for the use of ctDNA diagnostics as a risk stratification tool in resected lung cancer.

Background:
Reliable EGFR mutation testing techniques are required to identify eligible patients for EGFR-TKI treatment. Nowadays, surgically resected tissues or biopsy specimens are mainly used for the molecular testing. However, the biopsy samples sometimes have a certain limitation and so the cytology specimens are chosen for EGFR testing instead. Plasma sample has also become an option in EGFR-TKI resistant cases in which often have difficulties to obtain the inadequate tumor yield. In this study, we evaluated the feasibilities of using cytology samples and plasma specimens the EGFR molecular testing.

Method:
Cytology samples were obtained from biopsy and cells were suspended into liquid-based cytology (LBC) media. Tumor contents in the samples were confirmed with Papanicolaou stained slides. Plasma samples were also collected from patients shortly before the tissue biopsy. EGFR mutations in these samples were analyzed by cobas EGFR Mutation Test v2. Also, EGFR testing result of tissue specimens of the patients corresponded were collected from the medical records measured by cobas EGFR Mutation Test v2 as references.The feasibilities of both cytology and plasma specimen were evaluated comparing the results with the tissue samples.

Result:

EGFR mutation rate of tissue, plasma and LBC

	Tissue	Plasma	LBC
EGFR mutation +	18	9	14
EGFR mutation -	42	51	43
Invalid	0	0	3
Total	60	60	60
Positive rate (%)	30.0	15.0	23.3

One-hundred seven patients were registered to this study. 60 patients were enrolled to this study. EGFR mutation rates in tissue, cytology, and plasma were 30.0, 23.3 and 15.0 %, respectively. Concordance analysis was performed comparing cytology specimens and plasma specimens to the tissue samples. Overall concordance EGFR mutation status was 85.0 and 81.7%, respectively. A total 7 cytology specimens had discordances and 3 were invalid results. For plasma samples, discordants were found in11 samples but no invalids.

Conclusion:
Plasma was easy to obtain sample, but rate of detection was low. If a cancer cell is detected enough, LBC is useful for the examination for EGFR. Preservation of sample is easy, and re-useful for the examination, repeatedly.

Background:
Peripheral blood biomarkers can provide valuable information in a relatively non-invasive manner. In solid tumors, it has been suggested that ctDNA mutant allele frequency (MAF) and immune cell counts from peripheral blood complete blood count (CBC) at baseline may be associated with survival outcome. Here, we investigated the role of ctDNA MAF and immune cells as prognostic biomarkers that may predict overall survival in patients with NSCLC.

Method:
A retrospective cohort of 128 patients with ctDNA sample testing performed by ctDNA NGS test (Guardant360) were selected. ctDNA MAF of the dominant clone was collected. CBC’s drawn within a 1-2 week window (baseline) of ctDNA were reported for absolute neutrophil count (ANC), absolute lymphocyte count (ALC), absolute monocyte count (AMC), neutrophil lymphocyte ratio (NLR) and platelet count. Platelets and MAF were analyzed by quartiles (lower 75% vs highest 25%). Survival analyses and Cox regression analyses were performed on these variables.

Result:
A significant association was found for ANC (hazard ratio (HR) = 1.17, p<0.001), AMC (HR=1.98, p=0.037), MAF (HR=2.53, p=0.005), NLR>5(HR=2.98, p<0.001), and platelet counts (HR=2.49, p=0.006) with overall survival (OS), but not ALC. In multivariate analyses adjusting for clinical variables including age, sex, smoking status, histology, disease stage, number of prior lines of treatment, prior radiation, history of other cancers and ALK/EGFR mutation status, ANC remained as an independent predictor of OS(HR= 1.19, p<0.001). Figure 1



Conclusion:
Higher ANC, AMC, NLR, platelet counts and MAF were associated with poor overall survival. Further studies are required to validate our findings in patients with NSCLC.

Background:
While several studies have evaluated hybrid-capture NGS for cfDNA genotyping, amplicon-based NGS is an attractive alternative with the potential to be faster and less expensive. We performed a blinded evaluation of this approach for the characterization and monitoring of the molecular profile of advanced NSCLC during genotype-directed therapy.

Method:
Plasma samples from patients with advanced NSCLC and a known targetable genotype (EGFR, BRAF, MET, HER2 mutations; ALK, ROS1 rearrangements) were collected and analyzed, blinded to tumor genotype, with IRB approval. Up to 4 specimens were collected for each patient: baseline, initial 2 follow-ups, and progression. Plasma NGS was performed using enhanced tagged amplicon sequencing of hotspots and coding regions from 36 genes. A novel approach was used to detect ALK/ROS1 fusions using amplicon sequencing in cfDNA. Diagnostic accuracy was compared to plasma ddPCR and tumor genotype (including NGS when available).

Result:
A total of 146 specimens from 49 patients were studied. Testing was completed for 115 specimens at the time of analysis. Matched plasma NGS and ddPCR were available across 95 samples and revealed high concordance of allelic fraction (AF). At baseline, sensitivity of plasma NGS for the detection of the driver was 100% (26/26) for EGFR (88.5% ddPCR sensitivity). Sensitivity for the detection of ALK/ROS1 fusions was 89% (6/7 ALK, 2/2 ROS1). Rare instances of plasma NGS-positive/tissue NGS-negative discordance were seen across 13 cases with match tumor NGS (3/442 genes sequenced) and appear related to resistance heterogeneity, clonal hematopoiesis, and low tumor content of biopsy. Among patients with acquired T790M and available specimens at osimertinib resistance (n=21), 11 resistance mechanisms could be detected including tertiary EGFR mutations (e.g. C797S), mutations in BRAF, PIK3CA, or KRAS, and amplification of MET, HER2, or FGFR1. 4 were detected pre-osimertinib.

Conclusion:
This blinded analysis demonstrates for the first time the ability of amplicon-based plasma NGS to detect a full range of targetable genotypes in NSCLC. This approach has attractive sensitivity and specificity and deserves further study as an alternative to better-established hybrid capture approaches. Serial plasma NGS can detect competing resistance mutations in patients with TKIs resistance, highlighting the pitfalls of PCR-based plasma assays in patients with heterogeneous resistance and paving the way towards combination therapies.

Background:
To investigate the diagnostic value of folate receptor (FR)-positive circulating tumor cells (CTCs) detected by ligand-targeted polymerase chain reaction (LT-PCR) in distinguishing lung cancer from lung benign diseases in patients with solitary pulmonary nodules (SPN), and to identify the tumor invasiveness in lung adenocarcinomas of 3cm or smaller.

Method:
We enriched FR[+]-CTCs by immunomagnetic depletion of leukocytes and labeling with a conjugate of a tumor specific ligand folic acid and a synthesized oligonucleotide. The bounded conjugates were then analyzed by quantitative PCR (qPCR). Blood samples were collected from 319 lung cancer patients (257 adenocarcinoma patients), 120 benign patients. Including 257 lung adenocarcinoma patients, 41 were adnocarcinoma in situ (AIS), 53 were microinvasive adnocarcinoma (MIA), and 163 were invasive adnocarcinoma (INV). The clinicopathological characteristics were analyzed from medical records.

Result:
Patients were randomly assigned to a training set and a validation set. FR[+]-CTC levels of patients with lung cancer were significantly higher than patients with benign lung diseases (p=0.000). With a threshold of 8.74 CTC units, FR[+]-CTC showed a markedly sensitivity (training set, 72.05%; validation set, 74.36%) and specificity (training set, 80.9%; validation set 77.42%) in the diagnosis of lung cancer. When compared with established clinical biomarkers including carcinoembryonic antigen (CEA), cytokeratin 19 fragment (CYFRA21-1), CA-125, squamous cell related antigen (SCC), and neuron-specific enolase (NSE), FR[+]-CTC showed the highest area under the receiver operative characteristic curve (training set, 0.754; validation set 0.765). With a threshold of 9.051 CTC units, FR[+]-CTC showed a comparable sensitivity (training set, 77.50%; validation set, 79.07%) and specificity (training set, 61.24%; validation set 60%) to CEA and CA-125 in distinguishing the AIS from INV. Notably, the combination of FR[+]-CTC with established clinical biomarkers including CEA, CYFRA21-1, CA-125, SCC and NSE could significantly improve the diagnostic efficiency.

Conclusion:
LT-PCR based FR[+]-CTC detection platform is reliable to be used as a biomarker for the diagnosis of lung cancer in patients with SPN, moreover, the FR[+]-CTC level might be a promising biomarker to identify the tumor invasiveness in lung adenocarcinomas of 3cm or smaller.

Background:
Postoperative hyperthermia has been applied as an important adjuvant therapy to enhance the efficacy of traditional cancer treatment. To provide an update on the recent knowledge about the mechanisms of microwave (MW) hyperthermia on NSCLC cells using special MW device in vitro.

Method:
The various NSCLC cells (H460, PC-9 and H1975) were exposed to hyperthermia treatment using water bath or MW applicator. Cell survival was determined by CCK-8 assay. Cell apoptosis and cell cycle distributions were performed by Flow cytometry. Western blot assay was used to detect cell cycle arrest related molecular changes. Assessment of intracellular ROS changes in cells was performed using fluorescence probes DCFH-DA staining.

Result:
The MW hyperthermia can significantly inhibit cell growth in 3 cell lines, compared with water bath heating system. Flow cytometry results showed that the apoptotic cells increased significantly in MW hyperthermia treatment group, mean value of percent apoptosis treated by the MW hyperthermia was 12.45%, 40.00%, 37.70% in H460, PC-9, and H1975 cells respectively. The cell cycle distribution analysis showed that MW hyperthermia induced G2/M phase arrest. Western blot results suggested that MW thermal down-regulated the expressions of cyclinB1/cdc2, cdc 25c and p-cdc-25c, up-regulated expression of p21. Assessment of intracellular ROS change showed that MW hyperthermia induced production of ROS in NSCLC cells.

Conclusion:
Exposure of NSCLC cells to 433 MHz microwave hyperthermia induces ROS production and cell cycle arrest, which in turn induces cyclinB1/cdc 2 complex inactivation through inhibiting expressions of cdc 25c and increasing expression of p21, and finally promotes cell apoptosis. The heat treatment by water bath has not the same effect. This study suggests that microwave hyperthermia may be a potential therapeutic option for treating NSCLC in future.

Background:
Monoclonal antibodies targeting the programmed cell death 1 (PD-1) receptor and its ligand (PD-L1) have been showing promising results in advanced non-small cell lung cancer (NSCLC). Here, we investigated PD-L1 messenger RNA (mRNA) expression by a novel RNA in situ hybridization technique and compared it with PD-L1 protein expression in NSCLC.

Method:
Primary NSCLC specimens of 687 patients (476 adenocarcinoma and 211 squamous cell carcinoma) were constructed into tissue microarrays. PD-L1 mRNA in situ hybridization was performed with RNAscope[®] assay and classified using two independent scoring systems (RNA scope score and RNA proportion score). We also performed immunohistochmisty (IHC) using the Dako 22C3 pharmDx assay for evaluating PD-L1 protein expression.

Result:
PD-L1 mRNA expression was detected in 11.9% by RNA scope score and 8.3% by RNA proportion score. PD-L1 protein expression showed a positivity of 25.2% for the 1% cut-off, and 9.9% for the 50% cut-off. The two RNA scoring systems showed a linear correlation to each other (r = 0.83, p < 0.01), as well as to PD-L1 IHC repectively (p < 0.0001). The cut-off value that best correlated with PD-L1 protein expression was “1” for both RNA scope score (κ = 0.43 for the 1% and 0.58 for the 50% IHC cut-off) and RNA proportion score(κ = 0.36 for the 1% and 0.60 for the 50% IHC cut-off). Applying this criteria, discordant cases were found in 20% and 9% with 1% and 50% IHC cut-offs, respectively.

Conclusion:
PD-L1 mRNA expression, evaluated either as RNA scope score or as RNA proportion score, showed promising statistical results in predicting PD-L1 protein levels. We could thereby set out the RNA scoring system for PD-L1 in NSCLC. Discordant cases with positive mRNA levels and negative immunohistochemical results may indicate that PD-L1 mRNA in situ hybridization could be a more sensitive marker than immunohistochemistry. To verify the predictive role of PD-L1 mRNA expression in immunotherapy, however, therapeutic response to anti-PD-1/PD-L1 inhibitors should be investigated in future studies.