Virtual Library

Background:
Analysis of circulating tumor DNA (ctDNA) represents a potential strategy for the early detection of lung cancer. Despite significant interest, few studies have evaluated ctDNA levels in early stage lung cancer patients and the feasibility of ctDNA-based screening remains unclear.

Method:
We applied lung cancer-focused Cancer Personalized Profiling by deep Sequencing (CAPP-Seq) to assess ctDNA levels in 55 localized lung cancer patients treated with curative intent (stage I: n=22, stage II: n=7, stage III: n=26) and 50 healthy controls. Histological subtypes included: adenocarcinoma (n=30), squamous cell carcinoma (n=19), NSCLC NOS (n=4), and small cell lung cancer (n=2). Sensitivity and specificity of ctDNA detection were evaluated in all patients and in a subset of NSCLC patients with node negative (N0) stage I-II disease. Additionally, for patients with stage I adenocarcinoma in whom ctDNA was not detectable using the standard population-based CAPP-Seq approach, we designed personalized CAPP-Seq assays covering a median of 320 mutations/patient based on tumor exome sequencing from the respective patients.

Result:
We detected ctDNA in the pre-treatment plasma of 43/55 (78%) patients at a median allele fraction (AF) of 0.48% (range: 0.004%-26.1%). ROC analysis revealed an area under curve of 0.91, with sensitivity and specificity of 78% and 98%, respectively. Among patients with non-adenocarcinoma histologies, 92% (23/25) had detectable ctDNA (median AF: 0.90%), compared to 67% of patients with adenocarcinoma (20/30; median AF: 0.23%; P=0.046). However, tumor volumes were significantly smaller in adenocarcinomas (P=0.01) and in multivariate analysis ctDNA detection was significantly associated with tumor volume (P=0.01) but not histological subtype (P=0.16). In N0 stage I-II NSCLC patients (n=22), ctDNA was detected in 64% of patients (7/14 adeno vs 7/8 non-adeno) with a specificity of 98% and median AF of 0.022% (median AF of 0.018% vs 0.030% in adeno vs non-adeno patients, respectively). Using personalized CAPP-Seq assays, we detected ctDNA in 3/4 patients with stage I adenocarcinoma in whom ctDNA was not detected using our standard lung-cancer focused CAPP-Seq assay. In these 3 patients, tumor volumes ranged from 11.6-14.7 mL and the ctDNA AF ranged from 0.0014%-0.003%. Taken together, we detected ctDNA in 17/22 (77%) N0 stage I-II tumors.

Conclusion:
These data suggest tumor volume is a stronger determinant of ctDNA levels than histology in localized lung cancers. Additionally, our findings suggest that the majority of localized lung cancers shed ctDNA and that ultra-sensitive assays will be required for early detection of lung cancer using ctDNA

Background:
Recently, anti PD-1/PD-L1 immunotherapies have yielded promising outcomes in advanced squamous NSCLC. Several studies have suggested that tumor PD-L1 protein expression status might correlate with outcome and response to treatment. The aim of this study is to identify mRNA gene signatures and microRNAs associated with tumor PD-L1 expression in early-stage lung squamous cell carcinoma (SCC).

Method:
Early stage (I-II) SCC resected patient tumors were collected from 6 cancer centers as part of the SPECS II program. Gene expression profiling was performed on the specimens. PD-L1 protein expression was evaluated by immunohistochemistry on SCC FFPE tissue using the Dako 22C3 PD-L1 antibody. The tumor proportion score (TPS) for PD-L1 protein expression was compared with comprehensive clinicopathological, mRNA and miRNA data.

Result:
The prevalence of PD-L1 expression in this cohort of 255 Stage I-II SCC patients was 46.7% with a TPS cutoff of ≥ 1%, and 9.8% with a cutoff of ≥ 50%. Among 202 cases with available clinical and expression data, no significant association was observed between PD-L1 expression and clinical outcome. We identified a 12-gene signature from mRNA microarray using the Minimax Concave Penalty (MCP) regression method with an AUC of 0.92 at ≥ 5% TPS cutoff. A subset of 138 miRNAs was shown to be significantly differentially expressed between PD-L1 positive and PD-L1 negative groups at false discovery rate (FDR) of 0.05 with TPS cutoffs of ≥ 1%, ≥ 5% and ≥ 10%. No miRNAs were found to be significantly differentially expressed between the groups using a TPS cutoff of ≥ 50%. Gene Set Enrichment Analysis (GSEA) identified two pathways with gene sets that were significantly enriched (FDR < 0.05) in the PD-L1 negative group. No significant association was found between tumor mutation burden and PD-L1 expression level.

Conclusion:
PD-L1 expression prevalence is lower in early-stage lung SCC than in advanced NSCLC. No significant association was found between PD-L1 expression and prognosis in this cohort. Both mRNA gene signatures and miRNAs were identified to be predictive of PD-L1 expression. Through GSEA, two distinct gene sets were identified with expression correlated to PD-L1, one comprising genes related to ovary and another related to collagens and extracellular matrix (ECM). No significant association was found between tumor mutation burden and PD-L1 expression level. Following validation, these predictive signatures could be used to select patients with positive PD-L1 expression who may benefit from immunotherapy.

Background:
Adjuvant chemotherapy (ACT) for ES-NSCLC has a modest improvement in survival but it is often associated with serious adverse effects. Thus, identifying subgroups of ES-NSCLC patients who may benefit from ACT is of high clinical relevance. We evaluated the prognostic and predictive role of quantitative spatial profiling of PD-1/PD-L1 interaction in the tumor cells of ES-NSCLC patients.

Method:
451 whole tissue sections of formalin-fixed, paraffin embedded surgical resection specimens from ES-NSCLC patients with/without ACT were tested with a multiplexed fluorescence immunohistochemistry assay to detect PD-1, PD-L1, cytokeratin and DAPI labeling. Fluorescence Images were acquired on the Perkin Elmer Vectra platform and analyzed with AQUA® algorithms to determine the percent positivity of each biomarker as well as the co-localization of PD-1 and PD-L1 (the Interaction Score).

Result:
High PD-1/PD-L1 Interaction Scores correlated with improved progression-free and overall survival for ES-NSCLC patients receiving ACT after surgery (p = 0.01) whereas no difference in survival was observed for patients who received surgery alone (p = 0.9) (Figure 1). Interestingly, the levels of PD-1 or PD-L1 alone did not demonstrate any difference in survival for surgery + ACT or surgery alone patient populations. Figure 1



Conclusion:
PD-1/PD-L1 Interaction Score is predictive of benefit from ACT in patients with ES-NSCLC. Future studies will determine if this tool can be used to select patients that may be spared chemotherapy without compromising outcome.

Background:
Stereotactic body radiation therapy (SBRT) is currently the standard of care for inoperable patients with early stage non-small cell lung cancer (NSCLC). Despite this, ≈20% will relapse at 2 years. While adjuvant chemotherapy is recommended for surgically resected patients with early stage NSCLC (IB-IIIA), data on the role of adjuvant systemic therapy following SBRT for early stage NSCLC are sparse. The goal of this study was to evaluate the role of adjuvant chemotherapy following SBRT in early-stage, inoperable NSCLC.

Method:
Adults diagnosed with early-stage (clinical stage I and II) between the years of 2004 and 2013 were identified from the National Cancer Database (NCDB). Variables abstracted included: age, gender, clinical stage, race, comorbidity, insurance status, treating facility, treatment received and survival. Chi-square tests were used to compare clinical characteristics by therapy type. Kaplan-Meier, Cox regression, and propensity score analyses were employed for survival analyses.

Result:
Data from 12,414 patients with early-stage NSCLC were analyzed. Of these, 75.6% and 25.4% had clinical stage I and II disease, respectively. A total of 9,164 (73.6%) patients received SBRT alone and 3,268 (26.4%) had SBRT followed by chemotherapy. Among patients with clinical stage I, 83.5% received SBRT alone and 16.5% received SBRT followed by chemotherapy. Among those with clinical stage II, 43% received SBRT alone while 57% received SBRT followed by systemic therapy. On multivariate analysis, increasing age, male gender and stage II disease were associated with worse overall survival (OS). There was evidence of a clinical stage by treatment interaction (p <0.001). When treatment effect was analyzed by stage after adjusting for age and gender, patients with stage I treated with SBRT alone had a better median OS, 26.2 months compared to 22.4 months in the combined arm (HR=0.78; p<0.001; CI: 0.73-0.83). In contrast, among patients with stage II NSCLC, median OS was 15 months in the SBRT compared to 20.2 months in the combined group (HR=1.3; p<0.001; CI: 1.22-1.44).

Conclusion:
SBRT should be the sole modality treatment for patients with inoperable stage I NSCLC. However, patients with stage II disease appear to benefit from adjuvant chemotherapy. Randomized trials are needed in this area to answer this question conclusively.

Abstract not provided

Background:
Stereotactic body radiation therapy (SBRT) provides over 90% local tumor control, is a treatment of choice in patients with early stage medically inoperable non-small cell lung cancer (NSCLC). However, long-term overall survival after SBRT remain suboptimal, with 5-year rates rendering less than 50%, which cannot be fully explained by comorbidity, distant tumor progression or commonly known toxicities. We hypothesize that doses of radiation to normal lung and heart contribute to poor survival in these patients.

Method:
Patients with T1-T2 NSCLC received more than 100 Gy BED and with retrievable RT plan were eligible. The primary endpoint was overall survival, calculated from the start of SBRT. Clinical factors included age, gender, race, tobacco history, respiratory and cardiovascular comorbidity, tumor lobar location, histology, T stage, gross tumor volume (GTV), planning target volume (PTV), and prescription dose. Heart and lung were contoured consistently by one radiation oncologist according to the RTOG atlas. Dosimetric factors of the total lung were computed with biocorrection of fractionation effect.

Result:
A total of 280 patients with T1-3N0 met criteria. The median follow-up were 47 months. The median survival time was 33 months (95% CI: 25-42 months). The 2-year, 3-year and 5-year survival rates were 63%, 53% and 45%, respectively. Univariate analysis demonstrated that age (HR=1.02, p=0.04), gender (HR=0.75 for female, p=0.07), tumor T stage (HR=1.3 for T2, 2.5 for T3, using T1 as the reference, p=0.10), GTV (HR=1.01, p<0.001), PTV (HR=1.01, p<0.001), mean lung dose (HR=1.2, p<0.001), V5 (HR=1.02, p=0.03), V10 (HR=1.03, p=0.01), V20 (HR=1.1, p<0.001) of total lung and mean heart dose (HR=1.001, p=0.029) were associated with survival probability. The median mean lung and heart doses were 4.1Gy (range 0.8-11.2) and 0.99 Gy (range 0.3-9.7), respectively. Presence of radiation pneumonitis was not significant (p>0.1). Multivariate analysis was not performed as the dosimetric factors were correlated with each other. Among the risk factors, lung dosimetric factors were most significant. Increase in dose or volume of lung was significantly associated with poorer overall survival. A 1 Gy increase in mean lung dose corresponded to a 12 % increase in the risk of death, or 5% reduction in 5-year survival.

Conclusion:
This study demonstrated at the first time that doses to lung and heart are significant for overall survival after SBRT, while radiation pneumonitis was not. This suggests that the suboptimal survival after SBRT may be improved with plan optimization. This data also challenges the current practice of toxicity based normal tissue dose tolerance policy.

Background:
NRG Oncology RTOG 0915/NCCTG N0927 was a randomized lung stereotactic body radiotherapy (SBRT) trial of 34 Gy in 1 fraction (arm 1) versus 48 Gy in 4 fractions (arm 2) designed to select the better of the 2 regimens by comparing them at 1 year (yr): first by rates of pre-specified protocol-specified adverse events (psAEs), then by primary tumor control for each arm. 34 Gy emerged as the least toxic yet equally efficacious regimen. Herein, we update those results with long-term follow-up.

Method:
This phase II North American multicenter study of patients aged 18 yrs or older with medically inoperable non-small cell lung cancer with biopsy-proven peripheral (≥2 cm from the central bronchial tree) T1 or T2, N0 (clinically node negative by positron emission tomography), M0 tumors was designed to detect 1-yr psAEs rates >17% as primary endpoint. Primary tumor failure (PTF) (either infield or marginal failure) and local failure (either infield, marginal, or involved lobe failure) [with death without failure considered as a competing event]; overall survival (OS); disease-free survival (DFS) and progression-free survival (PFS) were secondary endpoints, but the study was not designed for statistical comparisons of these outcomes. The study opened in September 2009 and closed in March 2011. Updated data were analyzed through November 14, 2016.

Result:
Ninety four patients were accrued, with 86 eligible for analysis: 41 in arm 1 and 45 in arm 2, after 8 cases were excluded. Median follow-up time was 3.8 yrs for all patients, and 5.1 yrs for those alive at analysis. The grade 3 and higher treatment-related toxicity profile was unchanged since previous report, with specifically no new high grade chest wall or grade 5 events. Four of 48 Gy patients had subsequent grade 3 changes in spirometry since meeting the primary endpoint. Medians (in yrs) for 34 Gy and 48 Gy were: 4.1 vs. 4.0 for OS, and 2.6 vs. 2.8 for DFS, respectively. Five-yr outcomes as % (95% CI) for 34 Gy and 48 Gy were: PTF rate of 7.9 (2.0, 19.5) vs. 6.8 (1.7, 16.9); OS of 28.8 (15.4, 43.8) vs. 40.2 (24.9, 55.0); PFS of 19.1 (8.5, 33.0) vs. 31.8 (18.6, 45.9); and second primary rate of 15.5 (6.1, 28.9) vs. 13.3 (5.3, 25.1), respectively. Distant failure as the sole failure or a component of first failure was numerically higher in the 34 Gy arm (7 (46.7%)), but in the 48 Gy arm, rate of second primary development was higher (7 (43.8%)). Approximately 1/3 of patients’ causes of death was unknown, and another 1/3 was related to causes other than cancer or treatment.

Conclusion:
No excess in late-appearing toxicity was seen in either arm. Primary tumor control rates at 5 yrs were similar by arm. Median survival times of 4 yrs for each arm suggest similar efficacy pending any larger studies appropriately powered to detect survival differences.

Background:
There is a paucity of data on the long-term outcomes following lung SBRT. This impacts our understanding of late toxicity, relapse patterns and rates of second lung cancers. We report our multi-institutional outcomes of those who survived ≥5 years from lung SBRT treatment.

Method:
1192 patients were treated for primary non-small cell lung cancer, T1/2N0 from 5 international institutions. For those who survived ≥5 years from lung SBRT treatment details of patient factors, treatment and outcome factors were extracted from the multi-institutional database. All events were calculated from the end of radiotherapy. Local (LR), regional (RR), and distant metastases (DM) and toxicity events after 5 years are reported. New cases of metachronous lung cancers after 5 years are reported. Univariable analyses was performed to determine factors associated with survival ≥5 years.

Result:
Of 1192 patients there were 182 (14%) ≥5 year survivors. Only 52 (8%) survived ≥ 7 years and 2 (0.2%) ≥10 years. Those surviving ≥5 years were younger (74.3 vs 71.5 years; p<0.01) and had better FEV1 (55% vs 65%; p<0.01) than those that did not survive 5 years. The Charleston Comorbidity Score was 1.4 (0-12) vs 2.1 (0-7) (p<0.001) in those who survived < vs ≥5 years. Those who survived <5 years had a trend to larger tumors (2.4cm vs 2.3 cm; p=0.07). Of the 182 patients 23 (13%) were operable, 67 (37%) were <70 years old at treatment, 115 (63%) were ECOG 0-1 and 156 (86%) had peripheral tumor location at time of initial diagnosis. 84 (46.2%) were T1a, 60 (33%) T1b, 33 (18.1%) T2a and 5 (2.7%) T2b. After 5 years there were 13 new events of grade ≥ 2 toxicity. These toxicities were 4 grade 2 fatigue, 1 grade 2 rib fracture, 7 grade 2 chronic myositis and 1 patient with grade 2 chronic myositis and fatigue. In this 182 patient cohort, after 5 years, there were 3 local recurrences, 2 regional failures and 5 distant failures. After 5 years follow-up there were 22 (12%) new primary lung tumors in the 182 patients (and 20 had a subsequent lung SBRT treatment).

Conclusion:
Late survivors after SBRT require dedicated follow-up as they remain at risk for second lung cancer, tumor recurrence and toxicity. Second cancers in these late survivors can be considered for SBRT treatments.

Abstract not provided

Background:
The regularity of intrapulmonary lobar and segmental lymph node (LSN) metastasis in cT1N0M0 stage lung adenocarcinoma remains unclear. Thus, segmentectomy with uncertain LSNs metastatic status remains a potential oncological risk. We aimed to facilitate more accurate determination of N staging and identification of more suitable cases for segmentectomy.

Method:
A prospective study was performed from March 2014 to March 2016. A total of 156 patients diagnosed with cT1N0M0 stage lung adenocarcinoma received lobectomy and mediastinal lymph node dissection. The intrapulmonary LSNs were dissected and classified as adjacent LSNs or isolated LSNs. The metastatic status of the LSNs together with the TNM stage were analyzed. A comparison of the metastatic probability of isolated LSNs was carried out considering imaging features, serum carcinoembryonic (CEA) levels, pathological subtypes, size of the lesions, and metastatic status of adjacent LSNs.

Result:
Among the 156 cases enrolled, 129 were confirmed as pN0, 21 as pN1, 5 as pN1+N2, and 1 as skip pN2. When the LSNs had not been dissected, the false negative rate for N staging was 5.1% (7/136). Patients with a pure ground-glass-nodule had a lower isolated LSN metastasis rate (p = 0.027). Non-lepidic predominant invasive adenocarcinoma (p = 0.003), the cT1c group (p = 0.020), and those with adjacent LSN metastasis (p < 0.001) were detected with a higher isolated LSN metastasis rate. No significant difference in isolated LSN metastasis rate was found between groups with different serum CEA levels (p = 0.121).

Conclusion:
Dissection of intrapulmonary LSNs reduces the false negative rate of lymph node metastasis. Partial solid or solid lung adenocarcinoma, non-lepidic predominant invasive adenocarcinoma, and cT1c lung adenocarcinoma might not be suitable for segmentectomy. The lymph node sampling area during segmentectomy should include adjacent LSNs of the target segment. When metastasis to the adjacent LSNs is confirmed by fast frozen pathology, segmentectomy would not be suitable.

Background:
Sublobar resection is widespread for selected patients with small sized non-small cell lung cancer (NSCLC). Segmentectomy has been considered superior to wedge resection, however well-balanced comparative studies are lacking. We compared oncologic outcomes between wedge resection and segmentectomy for patients with less than 2cm of NSCLC according to parenchymal safety margin

Method:
A retrospective review of a prospective database was performed (2003-2015), excluding patients with poor lung function (FEV~1~ or DLCO <50%), neoadjuvant therapy, previous lobar resection for primary lung cancer, and multiple primary other cancer. Demographic, clinical, and pathological data were reviewed. Overall survival (OS), cancer-specific survival (CSS) and recurrence-free survival (RFS) were estimated using the Kaplan-Meier method and differences compared using log-rank test. Lymph nodes (LN) were evaluated preoperatively by PET/CT and EBUS-FNA if chest CT showed more than 1cm of LN short diameter

Result:
Two hundred nighty-six patients met our selection criteria, including wedge resection (W) in 188 and segmentectomy (S) in 108. There was no difference in smoking history, comorbidity, pulmonary reserve, cell type, and postoperative complication rate. All enrolled patients had clinically negative LN and 3 patients (1 in W, 2 in S) had pathologically node positive LN. The segmentectomy had more likely to have larger tumor (1.26 vs 1.36, p=0.045), more total nodes resected (1 vs 6, p<0.001), more distance of safety margin (median 0.5 vs 1.5, p<0.001). There was no statistical difference in OS (p=0.897), CSS (p=0.844), and RFS (p=0.763) between two groups. Patients were stratified by safety margin into ≤ 5mm in 118 (w101 vs s17), 5~10mm in 65 (w48 vs s17), 11~15mm in 45 (w23 vs s22), 16~20mm in 23 (w9 vs s14), and >20mm in 45 (w7 vs s38). Recurrence was developed in 8 (7%, w5 vs s3), 2 (3%, w2 vs s0), 1 (2%, s1), 0 (0%), 1 (2%, s1) of each safety margin. There was no significant difference of RFS between wedge resection and segmentectomy in each safety margin

Conclusion:
Wedge resection and segmentectomy showed comparable oncologic outcomes for carefully staged less than 2cm NSCLC patients. After thorough preoperative LN evaluation, wedge resection of sufficient safety margin is good surgical option

Background:
Spread through air spaces (STAS), defined as “micropapillary clusters, solid nests, or single cells beyond the edge of the tumor into air spaces in the surrounding lung parenchyma” (Travis WD, et al. WHO Classification of Tumours of the Lung, Pleura, Thymus and Heart, 2015), has been recognized as a pattern of tumor invasiveness. Because complete lymph node dissection and sufficient surgical margin lengths are difficult to obtain, wedge resection is associated with worse surgical outcomes than those of lobectomy. Because of the insufficient margin length associated with wedge resection, we speculated that STAS has a prognostic impact in wedge resection cases compared with segmentectomy cases. The aim of this study was to clarify the prognostic impact of STAS in patients with non-small cell lung cancer (NSCLC) who underwent wedge resection.

Method:
This was a retrospective study using our prospectively collected institutional database, established in May 2004. From May 2004 to May 2017, 1071 patients with NSCLC underwent complete resection. After excluding patients with pure ground glass opacity or multiple lung cancers and those who underwent lobectomy or preoperative therapy, we evaluated 196 patients with clinical stage IA cancer who underwent segmentectomy or wedge resection. TNM staging was performed according to the seventh edition. We assessed the prognostic impact of STAS on stage IA lung cancer cases who underwent wedge resection compared with segmentectomy.

Result:
Segmentectomy was performed in 110 patients and wedge resection in 86. The wedge resection cases were older (p<0.001) and had a higher CEA level (p=0.023). The frequencies of STAS were 14.5% and 18.6% in the segmentectomy and wedge resection cases, respectively (p=0.447). STAS was a significant prognostic factor for overall survival in the wedge resection cases on univariate (p=0.003) and multivariate (p=0.013) analyses, but it was not significant in the segmentectomy cases (p=0.597). STAS was a significant prognostic factor for disease-free survival in the wedge resection cases on univariate (p<0.001) and multivariate (p<0.001) analyses, but this was not the case in the segmentectomy cases (p=0.108). STAS was a significant prognostic factor for the recurrence-free rate in the wedge resection cases on univariate (p<0.001) and multivariate (p<0.001) analyses, but it was not significant in the segmentectomy cases (p=0.205).

Conclusion:
STAS is a prognosticator of poor survival outcomes in NSCLC patients who underwent wedge resection, but not in those who underwent segmentectomy. We speculate that NSCLC with STAS tends to have invasive characteristics, and wedge resection is not sufficient to improve survival outcomes in such patients.

Background:
The early detection and improved survival of resected non-small cell lung cancer (NSCLC) may increase the number of patients who eventually undergo subsequent pulmonary resection. We investigated the surgical outcomes and survival of patients following second and third pulmonary resections for NSCLC.

Method:
Patients who underwent second or third pulmonary resections without induction therapy for synchronous or metachronous NSCLC (511 patients, 535 procedures, 2000-2014) were included in the analysis.

Result:
Among 535 operations, 361 (67%) were sublobar resection and 103 (19%) were performed by minimally invasive approach, with the proportion of minimally-invasive procedures increasing in recent years (Fig. 1). The majority of re-resections were performed within 4 years of the previous resection (Fig. 2). Risk regression analysis demonstrated that predicted postoperative (ppo) FEV1 (p<0.001) and same side operation (p=0.002) were independent risk factors for severe complications (CTCAE grade ≧ 3; N=45). Multivariable Cox regression analysis revealed that age at subsequent surgery, male sex, ppoDLCO, interval from prior surgery, and tumor stage were independently associated with overall survival.

Conclusion:
In this large cohort of pulmonary re-resections for NSCLC, predicted postoperative pulmonary function tests were indictive of major complications and overall survival. Figure 1 Figure 2

Abstract not provided

Abstract not provided

Abstract:
It is widely recognized that patient participation in clinical research is a significant challenge, with only less than 5% of adult patients with cancer entering on clinical trials. This issue is becoming exacerbated in the field of lung cancer. With the advent of new therapeutic options and strategies accompanied by endless potential combinations of these approaches, there are more important trials and research studies that need to be done than patients available to participate in them. To address this problem, Lung Cancer Alliance developed a unique, innovative program known as LungMATCH that combines direct patient services with increasing enrollment to clinical studies. LungMATCH provides patient education and navigation through the changing treatment landscape in lung cancer while informing and educating patients about research opportunities that they can discuss with their treatment team. There are multiple programs under the LungMATCH umbrella. These include a molecular testing service where patients can initiate the process of molecular testing themselves, a novel patient-friendly online clinical trial matching widget, and phone-based personalized treatment and trial specialists who are knowledgeable about the state of lung cancer today and the many clinical trials available. These specialists can empower a patient or caregiver to have an informed conversation with their treatment team. Notably, while providing the community with much-needed navigation services, the program also aims to accelerate research discoveries through patient enrollment on a registry protocol for the molecular testing component and increasing patient accrual to clinical trials. In the first six months since launch, there have been barriers to overcome and lessons learned but many successes. Over 50 patients have been referred to the molecular testing program. More than 10 times as many users are viewing clinical trial search results compared to a previous clinical trial matching tool. Patients who speak with the trial navigators report having conversations about clinical trials with their healthcare providers much more frequently than other callers. These initial data indicate that the program is filling a need in our community and can help provide support to patients and caregivers while accelerating lung cancer research.

Abstract:
Background: Decision makers in research, industry, policy and health-care settings are actively seeking robust sources of patient data to inform their practices. Patient registries are ideally positioned to capitalize on this growing interest by building high-impact source of patient information. Registries give patients a direct means to participate in the care continuum, leveraging their input and insights to focus priorities and improve outcomes. Definition: Patient registries are organized systems designed to directly involve patients in the collection of their disease data and engage the medical community in the analysis and use of this data to improve lung cancer patient care. By creating a centralized registry of robust, on-going data to which patients, health-care professionals, researchers, industry, and policy makers have open access, registries provide data-driven, longitudinal information to: • support and facilitate improved disease management and standard of care; • assess individual patients to help determine the best line(s) of treatment; • measure patient outcomes to assess the quality and efficacy of health care provided; • use aggregate data as a tool to find patterns (for example, by ethnicity, geographic location, sex, age, stage of diagnosis, treatment protocols, health care facility/clinic, etc.) that could better predict the medical future and lead to improved outcomes and quality of life for patients, both individually and as a group, over time. • eventually allow us to evaluate the clinical and cost effectiveness of different diagnostic tools, treatments, and services. Current Landscape: Launched in 2012, The Registry of Patient Registries run by the Agency for Healthcare Research and Quality[1] currently list over 3,100 registries on its site covering a wide variety of registry types, but with most being research or institution generated. According to the National Institutes of Health, “Registries can be used to recruit patients for clinical trials to learn about a particular disease or condition; to develop therapeutics or to learn about population behavior patterns and their association with disease development; developing research hypotheses; or for improving and monitoring the quality of health care.”[2] In more recent times we have seen a rise in the development of Patient Powered Registries which “are somewhat indistinguishable from traditional registries, with one exception: In patient-powered patient registries, patients and family members “power” the registry by managing or controlling the collection of the data, the research agenda for the data, and/or the translation and dissemination of the research from the data.[3] Like researcher-generated patient registries, there is no single complete listing or documented number of PPRs in the United States however one study of 201 disease advocacy organizations found that forty-five percent had supported a research registry or biobank.[4] With the definition, purpose and current landscape being covered, we can now look at how registries continue to evolve both to meet the most common concerns and to greatly impact a patient’s care and quality of life. Multiple data sources: By leveraging the power of digital technology many patient registries now supplement patient reported data with clinical data through integration with Electronic Heath Record (EHR) systems. This integration allows for the capture of additional data points and to verify patient reported data, specifically when the patient may be unsure of the answer. In addition, natural language processing now provides the capability to “pull” information from uploaded documents and reports such as CT scans, lab reports, x-rays, genomic tests. Having multiple data sources allows researchers to paint a more detailed picture of the patients within the registry and more accurately identify patterns and eventually interventions. Clinical trial matching functionality: As with the integration of EHR, digital technology is now facilitating clinical trial matching functionality directly into patient registries. This service now allows patients, based on the information provided to not only receive a list of potential matches, but even allows for direct communication to the investigators involved in the study. Building patient communities: Unlike research or institution generated studies, patient powered registries may serve as a tool for building connected patient communities particularly when matched with general social media platforms such as Facebook and LinkedIn or more specific patient communities such as Patients Like Me and Inspire. Because patient registries now ask participants to consent not only to sharing their data, but to be contacted by the provider, registries can quickly bring together patients from across the globe for a variety of purposes, including driving research into their specific disease. This is particularly important in lung cancer which we now know is not one disease, but a series of “rare diseases” based on genetic differentiation. Educating patients: Being able to communicate directly with patients not only helps facilitate the building of communities and clinical trial matching, it also allows the registry provider to disseminate customized, personalized education materials directly to the participant. Multiple studies have shown how access to such information improves not only quality of life, but satisfaction with care. Conclusion: Well-designed patient registries are based on the idea that to improve outcomes and quality of life, no silos within the care community should exist, collaboration is critical, and there must be open and shared access to all information. In addition, today’s registries should leverage technology to allow for multiple data sources, longitudinal data collection, two-way communication and clinical trial matching to better serve its participants. By putting these tools to work and providing accurate, digestible information in the hands of industry, health IT companies, physicians and other health care practitioners, registries will organize and engage the entire lung cancer community around the common goal of better care for patients. 1. 4. Landy D, Brinich M, Colten M, et al. How disease advocacy organizations participate in clinical research: A survey of genetic organizations. Genetics in Medicine. 2012;14(2):223–8.

Abstract:
INTRODUCTION: Patient-centricity means much more than episodic patient involvement in health care. A Patient-centric culture requires that patients play an active role in all health care subareas from identifying unmet needs, pre-clinical research, clinical research, drug authorization, drug reimbursement to drug prescription. This study focus on patient-centricity in clinical trials. While in former days a patient participating in a clinical trial very often felt like a “laboratory mouse” only being instructed what to do and often without any understandable information nowadays those responsible for research and drug development discover the advantages of patient involvement such as better recruitment, patient friendly protocols, a better retention rate etc. Till now the main focus was effectiveness and safety of drugs to be shown by cost-intensive clinical trials. Nowadays value of treatments and relevance of outcomes to patients are of special interest. Patient Advocacy organizations help with recruitment by informing their members about new clinical trials. Crowdsourcing techniques addressing different communities like medical experts, patients and researchers are used to provide responses to survey questions. Apart from the arising enthusiasm about the doubtlessly potential opportunities the necessary requirements and their costs such as patient education costs, efforts to establish long-term partnership with patient advocacy groups, implementing patient advisory boards, legal framework conditions which protect patients but do not hinder their active participation as stakeholders etc. should not be underestimated. Additionally, pharmaceutical companies need to implement patient-centricity as a strategy across their functional departments. Successful patient-centric clinical trials require the willingness of patients especially of those who experienced clinical trials to provide their input in order to support improvement. METHOD: An anonymous survey has been designed and submitted to patients who already participated in a clinical trial. As patient-centricity in clinical trials is not just a lung cancer but a general issue Patient Advocacy Groups for different diseases such as lung cancer, breast cancer, cancer in general, lung fibrosis, arthritis, HIV etc. have been involved in US and Europe and further countries aiming to learn more about patient experience in clinical trials. As there is an increasing number of industry-funded clinical trials[1] another online survey was submitted to pharmaceutical companies in order to to find out more about the daily practice. RESULTS: 40 of 61 participants who already have been on a clinical trial completed the survey. Almost 50% recently (2016 or 2017) participated in a clinical trial. 16 patients live in US, followed by 8 patients from Austria and single patients from different European countries, one from Australia and one from Israel. Almost 40% of the surveyed have not been asked for feedback concerning the procedure during or after participation in the trial. 87% of patients would take part again in a clinical trial due to reasons like getting effective treatment, new treatment, helping others etc. Previous studies have found that trial participants do not receive the trial results.[2 ]This study replicated this finding. Almost 50% have not been informed about the final trial results although the WMA Declaration of Helsinki 2013 requested that all medical research subjects should be given the option of being informed about the general outcome and results of the study.[3] There is a high rate of willingness (82%) to provide input from patient perspective to improve the design of clinical trials. The analysis of the survey did not show any relevant differences concerning geographic area. 7 research-based pharmaceutical companies completed the second survey. Asking in which parts of the trial process (clinical question, clinical trial design, patient recruitment, trial management) patients are actively involved two companies indicated that patients are not involved in any of the mentioned parts. About 70% of the surveyed request feedback from patients about the procedure of clinical trials (30% do not). 3 of 7 companies do not see any hurdles for driving forward patient-centricity, 3 companies mentioned hurdles like e.g. substantial differences in the patients´ attitudes / concerns / current management across different countries/regions, added complexity, clarity on compliance rules, contractual framework and often patient report disease impact is difficult to quantify (e.g. I´m feeling tired). The most important measures in the pharmaceutical companies are understanding of patient needs, inclusion of patient reported outcomes, patient satisfaction, safety of IMP (investigational medical product) and patient selection biomarkers. The 2 companies who do not actively involve patients in the different parts of the trial process mentioned patient selection biomarkers, safety of IMP, safeguarding of patient’s rights, data quality and integrity as important measures. Both did not see any hurdles to drive forward patient-centricity in clinical trials. CONCLUSION: Although the number of participants in both surveys is quite low, the informational results gained by the patient survey can shed some light to the patient experience concerning clinical trials, the willingness as well as the reasons to participate more than once in a clinical trial. “Effective treatment” as most frequently mentioned reason to participate in a trial indicates a lack of information about clinical trials. There is a vast majority of patients with clinical trial experience being willed to provide their input in order to improve the design of new clinical trials. On the other hand there are researching companies who do not involve patients/patient advocates in the clinical trial process at all and the results of the survey provides an indication of the inconsistent proceedings within the pharmaceutical industry. In order to increase patient-centricity in clinical trials it is necessary that all relevant stakeholders (patients, patient advocates, regulators, medical experts, pharmaceutical companies etc.) compile consensual guidelines that ensure inclusion of relevant and adequately validated PRO, accurate patient information about clinical trials (in different languages), the option to be informed about the trial results for trial participants, information and education about clinical trials and collection and usage of patient feedback. [1]Ehrhardt S, Lawrence J. Trends in National Institutes of Health Funding for Clinical Trials Registered in ClinicalTrials.gov. JAMA. 2015;314(23):2566-2567 [2 ]Partridge AH, Winer EP. Informing clinical trial participants about study results. JAMA. 2002;288(3):363-365 [3]https://www.wma.net/policies-post/wma-declaration-of-helsinki-ethical-principles-for-medical-research-involving-human-subjects/ (August 2[nd], 2017)

Abstract not provided

Abstract:
Major advances in SCLC include improvements in RT techniques, the use of prophylactic cranial irradiation (PCI) for all stages of SCLC and the improved combination of chemotherapy and RT (Califano). The role of thoracic radiotherapy is well established in the management of stage I-III SCLC (Pignon). There is increasing evidence in the literature in favour of early concurrent chemo-radiotherapy, and a standard of care for patients with a good performance status is twice-daily thoracic radiotherapy (45 Gy in 3 weeks) with concurrent cisplatin and etoposide (Turrisi). Although current clinical trials are exploring the efficacy of new chemotherapeutic strategies, essential questions related to the optimisation of thoracic radiotherapy and the benefit of chemo-radiotherapy in sub-groups of patients remain unanswered, e.g. the elderly (Blackhall). The CONVERT trial (twice-daily (BD) versus once daily (OD) radiotherapy given concurrently with chemotherapy in stage I-III SCLC) was reported at ASCO 2016 (Faivre-Finn). OD RT did not result in a superior survival or worse toxicity than BD RT (two-year survival was 56% (95% CI 50-61) vs 51% (95% CI 45-57), p= 0.14). The survival for both regimens was higher than previously reported and using modern RT techniques radiation toxicities were lower than expected. In the stage IV setting the CREST trial as shown that the addition on thoracic RT to chemotherapy and PCI leads to a significant reduction in intrathoracic recurrence and, despite the lack of a significant benefit in overall survival at 1 year, there were significant improvements in overall survival at 2 years (Slotman). It is crucial that patients with SCLC are given the opportunity to participate in clinical research in order to continue to improve the survival of this disease. Molecular studies are ongoing aiming to gain improved insight into the molecular biology of SCLC, discover and/or validate candidate biomarkers for response, resistance to or toxicity of systemic treatment and radiation. It is expected that this understanding will lead to the development of targeted therapies that will not only prove efficacious but also less toxic than more conventional chemotherapy treatments. In combination with advanced RT techniques and better imaging, it is hoped that the rates of long term survivors will increase significantly in the future. We will address the following controversial questions • Should BDRT be considered standard of care in stage I-III SCLC? • Role of CTRT in stage I-II SCLC • Role of CTRT in elderly patients with LS-SCLC • Should thoracic RT be given to all patients with stage IV SCLC? References Califano R, et al. Management of small cell lung cancer: recent developments for optimal care. Drugs. 2012 5;72(4):471-90 Pignon JP, et al. A meta-analysis of thoracic radiotherapy for small-cell lung cancer. N Engl J Med 1992; 327:1618-1622. Turrisi AT, et al. Twice daily compared to once-daily thoracic radiotherapy in limited-stage small-cell lung cancer treated concurrently with cisplatin and etoposide. N Engl J Med 1999; 340: 265-271. Auperin A, et al. Prophylactic cranial irradiation for patients with small-cell lung cancer in complete remission. Prophylactic Cranial Irradiation Overview Collaborative Group. N Engl J Med 1999;341(7):476-84. Blanchard P, et al. Prophylactic cranial irradiation in lung cancer. Curr Opin Oncol. 2010; 22(2):94-101 Blackhall F et al. Treatment of limited small cell lung cancer: an old or new challenge? Curr Opin Oncol. 2011; 23(2):158-62 Faivre-Finn C, et al. Concurrent once-daily versus twice-daily chemoradiotherapy in patients with limited-stage small-cell lung cancer (CONVERT): an open-label, phase 3, randomised, superiority trial. Lancet Oncol. 2017;18(8):1116-1125 Slotman BJ, et al.Use of thoracic radiotherapy for extensive stage small-cell lung cancer: a phase 3 randomised controlled trial. Lancet. 2015 3;385(9962):36-42 Slotman BJ, et al. Which patients with ES-SCLC are most likely to benefit from more aggressive radiotherapy: A secondary analysis of the Phase III CREST trial. Lung Cancer. 2017; 108:150-153

Abstract:
The brain is a well recognized sanctuary site for micrometastases in patients with small cell lung cancer (SCLC) treated with chemotherapy. The administration of prophylactic cranial irradiation (PCI) in patients responding to chemotherapy reduces the incidence of clinically detectable brain metastases in patients with both limited [1] and extensive disease [2, 3]. In patients with limited disease, this translated into a survival benefit in a meta-analysis of 987 complete responders , but in patients with extensive disease, one trial of 286 responders to chemotherapy showed a survival benefit [2], whilst in another of224 patients it did not [3]. The design of this latter trial suggested that regular surveillance with MRI and treatment at the time of metastasis detection might be an equally effective strategy. This could reduce the incidence of the most serious toxicity of PCI which is delayed neurotoxicity. Limiting the total dose is another possible means of reducing neurotoxicity. A randomized trial of three dose prescriptions restricted to patients with limited disease showed no difference in rate of brain metastases between 25 Gy in 10 fractions versus two higher dose schedules , but there was a higher mortality associated with the higher doses, due to an unexplained increase in disease progression [4]. In spite of this high level evidence there are still numerous uncertainties facing the clinician when deciding whether to recommend PCI. 1. Is the classification limited versus extensive disease still appropriate when selecting patients? 2. Are there subgroups of patients with metastatic disease more likely to benefit from PCI than others because their overall prognosis is better? [5] 3. Should there be an upper age limit? 4. Are pre-existing neurologic conditions or paraneoplastic syndromes contraindications? 5. In patients with extensive disease, is pretreatment MRI required? 6. Is MRI surveillance cost effective? 7. Is there a role for hippocampal sparing techniques or neuroprotective agents? 8. Is there a place for lower dose/shorter fractionation schedules? 9. In patients with extensive disease having consolidative chest irradiation, is there any difference between giving PCI simultaneously versus sequentially? 10. What will be the role of PCI in the era of immunotherapy? Question 7 is currently the subject of an actively recruiting randomized trial (NRG-CC003), but for the other questions, physician and patient discretion will be required for the forseeable future. References 1. Auperin, A., et al., Prophylactic cranial irradiation for patients with small-cell lung cancer in complete remission. Prophylactic Cranial Irradiation Overview Collaborative Group. N Engl J Med, 1999. 341: 476-84. 2. Slotman, B., et al., Prophylactic Cranial Irradiation in Extensive Small-Cell Lung Cancer N Engl J Med, 2007. 357: 664-672. 3. Takahashi, T., et al., Prophylactic cranial irradiation versus observation in patients with extensive-disease small-cell lung cancer: a multicentre, randomised, open-label, phase 3 trial. Lancet Oncol, 2017. 18: 663-671.. 4. Le Pechoux, C., et al., Standard-dose versus higher-dose prophylactic cranial irradiation (PCI) in patients with limited-stage small-cell lung cancer in complete remission after chemotherapy and thoracic radiotherapy (PCI 99-01, EORTC 22003-08004, RTOG 0212, and IFCT 99-01): a randomised clinical trial. Lancet Oncol, 2009. 10:467-74. 5. Eberhardt, W.E., et al., The IASLC Lung Cancer Staging Project: Proposals for the Revision of the M Descriptors in the Forthcoming Eighth Edition of the TNM Classification of Lung Cancer. J Thorac Oncol, 2015. 10:1515-22.

Abstract:
Difficulties of surgery after induction chemoradiotherapy (CRT) Surgery after induction therapy can be very challenging. From my own observations, this may particularly be an issue when the patient had nodal involvement at presentation and experienced a good response to the induction treatment (induction CRT or even induction C alone) where fibrosis of the responding lymph nodes may make the hilar vascular mobilization more difficult. The following lines will describe some tips on how to minimize the risks of such resections. Preoperatively: One should always take advantage of the induction therapy period (12 weeks or more) to achieve absolute smoking cessation in these patients. After induction therapy, one should obtain a fresh set of PFTs including a DCO measurement as both radiation and some of the induction chemotherapy agents may have caused significant pneumonitis with resulting altered lung function. If there is a possibility that a pneumonectomy will be required, prepare yourself. (QVQ, stress echo). Radiation esophagitis may have brought some nutritional issues, consider an alimentary “boost” when a significant weight loss has occurred during the induction phase of therapy. If there may be a risk that the SVC will be clamped, repaired or replaced, discontinue the port-o-cath preoperatively. Finally, make sure to review all imaging, particularly the pre-induction therapy CT. Surgery Personally, I approach these cases via open thoracotomy, though some have reported successful resections on MIS platforms. All patients get an epidural catheter preop. If one anticipates, issues with the SVC with a R upper lobar resection, get infra diaphragmatic IV access preop. Emphasize w anesthesia, in the preoperative area, the absolute need to keep these patients dry: an irradiated mediastinum cannot handle excess fluid. At entry, I prepare/ harvest the intercostal muscle, without dividing it. (I now keep the omentum for “later”, if needed to manage for a complication, and spare the serratus anterior in all cases). If accessible, I usually clear zone 7 first, with frozen section read. (The frozen section information/ feedback during this type of surgery can be very useful to have when one gets into a tough surgical corner…) If one is attempting a lobectomy and anticipates possible difficulties in the hilum, one should consider obtaining early circumferential control of the main veins and main PA (if possible), and even do so intrapericardially if needed. Similarly, early division of the azygos vein on the right, may help when one anticipates difficulties accessing the main R PA and surrounding structures. One may have to alter their “routine” sequence of dividing the central structures. With both upper lobes, when bulky fibrotic changes are present in the suprahilar areas, proceeding retrograde, from the fissures up is often very helpful. As well, dividing open the upper lobe bronchi from the back can help access and control the often fibrosed and foreshortened proximal lobar PA branches. On the left, I will identify the vagus nerve low down, away from the fibrotic field and follow it up into the AP window to minimize the risks of inadvertently injuring the recurrent nerve. During the case, I constantly remind anesthesia to keep these patients dry. For pneumonectomies, we give a steroid bolus before dividing the main PA , similarly we initiate amiodorone prophylaxis by infusion. In a R pneumonectomy, when we have stapled the bronchial stump, I add an additional non resorbable monofilament stitch to reinforce both ends of the bronchial stapled line. Finally, we cover all of our bronchial stumps with the IC muscle flap. POSTOP Remember to keep the patient dry. We give amidorone prophylaxis for 72 hours after pneumonectomies’ or other patients considered at elevated risk of going into atrial fibrillation post op. Beware of possible left vocal cord palsy after L pneumonectomy and left upper lobectomy: we keep these patients NPO until we are confident that their cords are OK. If there are any concerns, we get an immediate laryngoscopy. If the left cord is found to be paretic, we immediately get VC medialization (not negotiable)

Abstract:
Surgery for Stage IIIA NSCLC - Surgeon's Perspective Gail E Darling MD FRCSC, University Health Network, University of Toronto, Toronto Canada Outline Indications for operating on IIIA Choice of Induction regimen: chemotherapy vs chemoradiation Restaging the mediastinum: why or why not? how (EBUS vs med, redo med?) , Factors in decision making whether or not to proceed with surgery after induction Salvage resection after definitive chemoradiation Indications for Surgery for Stage IIIA NSCLC 1.T3N1M0, T4N0M0, T4N1M0: if primary is resectable with R0 resection 2.T1-3N2M0: single station microscopic N2, resectable with R0 resection, selected multistation N2 if microscopic N2 and nodal disease is resectable with R0 3.Patient is fit for planned resection Contraindications to surgery for Stage IIIA NSCLC 1. Extracapsular N2 2. Fixed bulky nodes 3. Inability to achieve an R0 resection 4. T4N2 5. Patient not fit for planned resection (right pneumonectomy required?) Choice of Induction: Chemotherapy alone vs Chemoradiation 1. Intergroup 0139 trial: concurrent cisplatin + etoposide + 45 Gy improved disease free survival over surgery alone 2. SAKK trial: RCT Chemo vs chemoRT ( sequential) →No difference in survival 3. Concurrent chemoRT higher rates of mediastinal downstaging compared to chemotherapy alone 4. Radiation may be undesirable if sleeve resection of bronchus or artery is planned Restaging after Induction therapy 1. CT scan: to assess response but primarily to confirm resectability and exclude progression 2. PET: to rule out progression; assess response to determine if surgery is required? 3. Mediastinum: a) restage only if surgeon will not operative if persistent N2. Intergoup 0139 trial reported best survival if mediastinum was sterilized by induction but even if persistent disease survival was better with surgery than chemoRT alone. B) Restaging: if first mediastinal staging was performed by mediastinoscopy PET is most accurate; EBUS/EUS not reliable (sampling error); redo mediastinoscopy difficult, not reliable. If first staging was performed by EBUS/EUS then mediastinoscopy is appropriate and reliable for restaging. Factors in Decision Making to Proceed with Surgery after Induction 1. Patient is fit ( recovered fully from induction, good performance status) 2. Cancer has not progressed 3. Lack of radiographic response is not a contraindication to resection 4. R0 resection is possible Salvage Surgery after Definitive Chemoradiation 1. Patient is fit with good performance status 2. Primary site is only site of disease

Abstract:
Thymic malignacies are distintive by being relatively rare and by their generally more indolent behavior; nevertheless all thymomas are malignant tumors. Progress in the understanding of these tumors has been limited until recently. The advent of several regional and international collaborative groups (e.g. the Japanese Association for Research on the Thymus [JART] and the International Thymic Malignancy Interest Group [ITMIG] has provided an infrastructure to facilitate progress, and the field has advanced significantly. The level of understanding available now makes the previously common approach of empiric management of thymic malignancies inappropriate. Three major categories of thymic malnignancies with different biologic behavior are distinguished: thymoma, thymic carcinoma (TC) and Neuroendocrine tumors of the thymus (NETT, also known as thymic carcinoid). Thymomas are subdivided into 5 histotypes with slightly different clinical features, but these clinical differences are small and mixed histotypes are frequent. Through international collaboration, the first fromal stage classification system has been developed. Many patients with thymic malignancies have associated autoimmune conditions, most prominently myathenia gravis. Chest CT with intravenous contrast is the primary imaging modality for patient evaluation. A combination of patient demographics, clinical and imaging features allows a reliable clinical diagnosis of a thymic malignancy to be made in most patients; this can be supplemented with a needle biopsy if needed. However, the ability of imaging to accurately define the degree of local invasion or pleural dissemination is limited. In general, surgical exploration is warranted (in a facility prepared to handle resection of all possibly involved structures) unless imaging clearly demonstrates unresectable invasion. A complete resection of all involved structures is the cornerstone of treatment. The role of adjuvant radiotherapy after an R0 resection remains unclear, especially after resection of more invasive tumors or int he case of TC/NETT. It is easier to argue for adjuvant RT after an R1 resection or an R0 resection with close margins. Surgeons should be prepared to resect involved structures; a planned subtotal (debulking) resection is not recommended. When more invasive thymic tumors raise questions about the ability to achieve an R0 resection, preoperative chemo- or chemoradiotherapy is typically given. Recurrences are relatively low after complete resection. Longer-term (5- and 10-year) survival rates with aggressive multimodality treatment are fairly good, evin in tumors that invade major mediastinal structures or demonstrate pleural dissemination. When resection is not possible, management with palliative chemotherapy (± RT) can be useful. Several regimens have been shown to be active, and an optimal chemotherapy regimen has not been identified.

Abstract:
Extragonadal germ cell tumors (EGCTs) are among the most perplexing and embryologically intriguing entities in clinical oncology. They represent only 2-5% of adult germ cell malignancies. Their histology is identical to germ cell tumors (GCTs) derived from primary gonadal sites. However, their biology, especially primary mediastinal non-seminomatous GCT (PMNSGCT), is substantially different. PMNSGCT represents a clinically and biologically distinct subset of EGCT. It carries a poor prognosis with 40-50% overall survival after platinum-based chemotherapy and surgery (1-3). The cure rate is even worse if there is metastatic disease of the lung, liver, or supraclavicular lymph nodes with only 25% overall survival. In contrast, mediastinal seminomas have a good prognosis with 88-90% overall survival (4,5). Survival outcome for PMNSGCT is dependent on successful chemotherapy and expert thoracic surgery to remove residual disease when feasible. Patients presenting with anterior mediastinal tumor do not have a testis primary and there is no need to evaluate with testis ultrasound. The significant elevation in tumor markers including human chorionic gonadotrophin (hCG) > 1,000 U/1 and/or any elevation in alphafetoprotein (AFP) confirms the diagnosis of NSGCT and no biopsy is needed. The patient should be treated with 4 courses of platinum-based triple chemotherapy. We recommend using etoposide (VP-16), ifosfamide and cisplatin (VIP x 4) as the standard chemotherapy for PMNSGCT followed by surgical resection (2). Patients with normal tumor markers or mild elevation in hCG require a biopsy to establish the diagnosis of pure seminoma vs. NSGCT vs. other etiology such as thymic malignancy or lymphoma. Between 1980 and 2013, 221 patients with PMNSGCT who underwent postchemotherapy surgery at Indiana University were retrospectively reviewed. Our initial experience was entirely with BEP x 4. Because of our clinical observation of unacceptable pulmonary toxicity, we subsequently substituted ifosfamide for bleomycin. One hundred sixty-six patients received BEP and 55 received VIP. There were 11perioperative deaths when bleomycin was part of the preoperative chemotherapy (6.6%). By contrast, we saw no perioperative deaths on the VIP arm. Furthermore, 30 patients (18.1%) had prolonged ventilator use for greater than 48 hours. Only 2 patients (3.6%) had this complication on the VIP arm (6). A phase III intergroup study of 181 patients with poor-risk germ cell tumors, including patients with PMNSGCT, demonstrated equivalent survival in patients with VIP compared to BEP. Given the high rate of postoperative pulmonary failure and mortality after BEP, we feel these results strongly support our present policy of preferring to use VIP chemotherapy in patients with PMNSGCT before major thoracic surgical procedure. Surgical resection of residual disease after chemotherapy is an integral part of the management of patients with PMNSGCT. Surgical resection after chemotherapy serves to assess response, remove chemotherapy-resistant disease and direct additional chemotherapy. Viable teratoma or germ cell cancer is present in 30-47% of patients who undergo resection post-chemotherapy (7,8). Chest CT scan every 4 months for 2 years, then every 6 months for years 3-5 is recommended if teratoma or malignant transformation of teratoma is present in the surgical specimen. However, if there is persistent germ cell cancer in the surgical specimen, then patients should receive 2 post-operative courses of etoposide and cisplatin (EP). Surgical treatment for relapsed PMNSGCT in the presence of rising tumor markers after chemotherapy is controversial. Our preferred treatment option is surgery, if feasible, especially if carried out by experienced thoracic surgical oncologists. We recently reported our results with surgery alone in 35 patients with relapsed PMNSGCTs with rising serum AFP (32 patients) or hCG (3 patients). Seven (20%) of 35 patients remained continuously disease-free, with median followup of 64 months (range, 25-220 months) (10). If surgery is not feasible, then high dose chemotherapy in an experienced center is a reasonable approach. Mature teratoma is the most common GCT in the mediastinum. Patients present with large circumscribed anterior mediastinal mass with normal serum hCG and AFP. The treatment of mature teratoma is surgical resection and there is no role for chemotherapy unless elevated tumor markers are present. Mediastinal pure seminomas comprise 30-40% of malignant mediastinal GCT. Mediastinal seminoma carries an excellent prognosis and should be treated with good-risk chemotherapy regimen (3 cycles of bleomycin, etoposide, and platinum or 4 cycles of EP) with no surgical resection needed. References: International Germ Cell Consensus classification: a prognostic factor-based staging system for metastatic germ cell cancers. International Germ Cell Cancer Collaborative Group. J Clin Oncol 1997; 15:594-603. Kesler KA, Rieger KM, Hammoud ZT, et al.: A 25-year single institution experience with surgery for primary mediastinal non-seminomatous germ cell tumors. Ann Thorac Surg 2008; 85:371-378. Rivera C, Arame A, Jougon J, et al.: Prognostic factors in patients with primary mediastinal germ cell tumors, a surgical multicenter retrospective study. Interact Cardiovasc Thorac Surg 2010; 11:585-589. Bokemeyer C, Nichols CR, Droz JP, et al.: Extragonadal germ cell tumors of the mediastinum and retroperitoneum: results from an international analysis. J Clin Oncol 2002; 20:1864-1873. Hartmann JT, Nichols CR, Droz JP, et al.: Prognostic variables for response and outcome in patients with extragonadal germ cell tumors. Ann Oncol 2002; 13:1017-1028. Ranganath P, Kesler KA and Einhorn LH: Perioperative morbidity and mortality associated with bleomycin in primary mediastinal non-seminomatous germ cell tumor. J Clin Oncol 2016; 34:4445-4448. 2016. Vuky J, Bains M, Bacik J, et al.: Role of postchemotherapy adjunctive surgery in the management of patients with non-seminoma arising from the mediastinum. J Clin Oncol 2001; 19:682-688. Kesler KA, Rieger KM, Ganjoo KN, et al.: Primary mediastinal non-seminomatous germ cell tumors; the influence of postchemotherapy pathology on long-term survival after surgery. J Thorac Cardiovasc Surg 1999; 118:692-700. Hinton S, Catalano PJ, Einhorn LH, et al.: Cisplatin, etoposide , and either bleomycin or ifosfamide in the treatment of disseminated germ cell tumors. Final analysis of an intergroup trial. Cancer 2003; 97:1869-1875. Radaideh SM, Cook VC, Kesler KA, Einhorn LH: Outcome following resection for patients with primary mediastinal non-seminomatous germ cell tumors and rising serum tumor markers postchemotherapy. Ann Oncol 2010; 21:804-807.

Background:
TET have been associated with autoimmune disorders (AID) in up to 30 % of patients. However, there have been wide variations in the reported prevalence of TET associated disorders based mostly on small single center series. RYTHMIC (Réseau tumeurs THYMiques et Cancer) is a French network mandated to systematically discuss every case of TET. Using our database, we aimed to describe the prevalence of AID in a large French population with TET.

Method:
RYTHMIC database prospectively includes all consecutive patients with a diagnosis of TET discussed in our national tumor board. We calculated the prevalence and described epidemiologic, clinical and pathological characteristics of patients with TET’s related autoimmune diseases.

Result:
From January 2012 to May 2017, 1693 patients were included in the registry. Of these, 200 patients (11.8%) had autoimmune disorder. The mean age at diagnosis of TET was 54 years old and 52% were male. 149 had myasthenia gravis (75.3%), 15 Good syndrome (7.6%), 14 pure red cell aplasia (7.1%), 12 systemic erythematous lupus (6.1%) and 12 thyroiditis (6.1%). Some patients (14.5%) eventually developed more than 1 AID. Diagnosis of AID was mostly done at the same time of TET diagnosis (54.6%) but some patient had their AID diagnosed before (19.8%) or during follow up (13.4%). Masaoka Koga stages were overall well balanced with 16.5% stage III, 16% stage IIb, 13.5% stage I, 13% stage IIa and IV. Histologic subtype distribution was in order of frequency; B2 (37%), AB (14.5%), B3 (14%), B1 (10.5%), thymic carcinoma (4.5 %) and A (4%).

Conclusion:
In our registry of TET, the prevalence of autoimmune diseases was 11.8% and most diagnosis were established at the same time as TET. The extent of disease, measured by Masaoka Koga staging, does not seem correlated.

Background:
Thymic epithelial tumours (TETs) are rare and under-researched intrathoracic cancers. So far the only significant finding is a recurrent (43%) missense mutation in GTF2I. In addition to validating this finding, we set out to expand our understanding of the molecular changes underlying TETs through whole exome sequencing (WES) and detection of copy number alterations (CNAs) following SNP genotyping.

Method:
WES was performed on 17 TETs (2AB, 1B1, 3B2, 2B3, 6CA and 3NETT) and matched normal tissue. Somatic single nucleotide variants (SNVs) were identified with the GATK HaplotypeCaller and annotated for impact prediction (SnpEff 3.6b and SnpSift 1.3.4b) and population frequency (SnpSift 1.3.4b). The frequency of the GTF2I mutation was assessed with Sanger sequencing with semi-nested primers on DNA from 144 TETs of all subtypes. SNP genotyping was performed on 100 TETs of all subtypes with matched normal tissue in most cases, to identify somatic CNAs. Analysis was performed with ASCAT (v2.4.4) and copy number segments were annotated with Bedtools (v2.26.0).

Result:
WES confirmed a low mutation burden for TETs. No highly recurrent mutations were found. Hotspot mutations in NRAS and KRAS were seen, as was a hotspot mutation in TP53 in a NETT. A high impact frameshift MSH6 somatic mutation was noted in one of the squamous cell carcinomas (SCCs). Another SCC had a germline BAP1 mutation and a family history of other cancers, suggesting a BAP1 familial cancer predisposition syndrome in this individual. The GTF2I mutation was seen in 48 of 141 evaluable TETs (34%) and was present more commonly in type A (90%) and AB (69%) thymomas. The frequency decreased to 16%, 6% and 13% in B1, B2 and B3 thymomas respectively and was not seen in any squamous (0/12) or neuroendocrine carcinomas (0/6). Overall, the most frequent copy number gains in TETs involved chromosomes 7q (22%), 1q (17%) and 11q (17%). The commonest gain was in a gene not previously found to be amplified in solid tumours. The most frequent copy number losses were in chromosomes 6p (40%), 2q (37%) and 7q (32%). Gains and losses demonstrated distinct patterns between aggressive versus indolent subtypes.

Conclusion:
The mutation in GTF2I remains the single most frequently recurrent mutation in TETs. We are in the process of establishing a clinical use for this finding. Results from WES and CNA through SNP genotyping have provided important insight into other potential key players in the aetiology of this intriguing malignancy.

Background:
The mainstay of treatment for thymoma is surgery with neoadjuvant chemotherapy recommended to patients with locally advanced disease. EGFR is overexpressed in thymoma. Clinical responses to single-agent cetuximab have been reported in patients with advanced cetuximab. We conducted this two-site prospective phase II trial of cetuximab combined with a standard induction chemotherapy regimen of cisplatin, doxorubicin and cyclophosphamide (PAC) in patients with locally advanced thymoma prior to surgical resection.

Method:
Patients with clinical Masaoka stage III-IVA thymoma were treated with cetuximab (250mg/m[2] weekly x 4 weeks) followed by cetuximab (250 mg/m[2] weekly) combined with cisplatin (50mg/m[2]), doxorubicin (50 mg/m[2]) and cyclophosphamide (500mg/m[2]) 3 weeks x 4 cycles). Radiographic response was assessed by CT using RECIST 1.1 and FDG-PET using PERCIST. All patients went on to surgery after completion of induction therapy. The primary endpoint was major pathologic response (MPR, >90% treatment effect). Planned enrollment was 18 patients in first stage of a two stage design. If 1 MPR was observed, then enrollment would expand to 28 patients.

Result:
Eighteen patients were enrolled: 8 women, median age 53 (range 32-73). WHO Histologic subtype A: 2, AB: 3, B1: 3, B2: 7, B3: 3. Final Masaoka stage I: 2, II: 2, III: 5, IVA: 9. There were no responses to cetuximab alone by RECIST criteria, although 1 patient had a 25% reduction in indicator lesions. Response rate (CR+PR), in evaluable patients after complete treatment course was 50% (8/16, 95% CI 28-72%). Partial responses by PERCIST criteria were seen on PET in 11/18 (61%) evaluable patients. There were no MPRs. R0 resection was obtained in 7 patients; 5 had R1 and 6 had R2 resections.

Conclusion:
The addition of cetuximab to PAC chemotherapy did not lead to pathologic complete responses in the neoadjuvant setting. Cetuximab alone appears to have little effect during 4 weeks of treatment. There was no apparent increase in radiographic response rate with the addition of cetuximab to PAC chemotherapy compared to historical series.

Background:
No standard treatment exists for patients with thymic epithelial tumor (TET) who progress after platinum-containing chemotherapy. We conducted a phase II study of pembrolizumab in patients with refractory or relapsed TET to evaluate the efficacy and safety.

Method:
Between March 2016 and June 2017, patients with histologically confirmed TET who progressed after at least one platinum-containing chemotherapy were eligible. Patients were excluded if they had an active autoimmune disease requiring systemic treatment within the past one year or documented history of clinically severe autoimmune disease. Patients received 200mg of pembrolizumab intravenously every 3 weeks until tumor progression or unacceptable toxicity. Response was assessed every 9 weeks by investigator. The trial was registered with ClinicalTrials.gov, number NCT02607631.

Result:
Thirty-three patients were enrolled, 26 with thymic carcinoma and 7 with thymoma. 19 (57.3%) patients received two or more prior lines of systemic chemotherapy. Median number of cycles was 8 (ranges, 1-22) and median follow-up was 11.8 months (ranges, 1.6-14.9 months). Of 33 patients, eight (24.2%) achieved partial responses, 17 (51.5%) stable disease, and 8 (24.2%) progressive disease as best response, resulting in overall response rate of 24.2% (7 confirmed PR). The median progression-free survival was 6.1 months for both of thymoma and thymic carcinoma. The most common adverse events of any grade include dyspnea (11 [33.3%] of 33 patients), chest wall pain (10 [30.3%]), anorexia (7 [21.2%]) and fatigue (7 [21.2%]). Treatment-related adverse events ≥ grade 3 associated with immune related adverse events (irAE) include hepatitis (4 [12.1%]), myocarditis (3 [9.1%]), myasthenia gravis (2 [6.1%]), thyroiditis (1 [3.0%]), ANCA-associated rapidly progressive glomerulonephritis (1 [3.0%]), colitis (1 [3.0%]), and subacute myoclonus (1 [3.0%]) except anemia (1 [3.0%]). Eight (24.2%) patients (5 thymoma, 3 thymic carcinoma) discontinued study treatment due to irAE, whereas irAEs were manageable with immediate administration of high dose corticosteroid and other immunosuppressive agents in most of patients (7 of 8 [87.5%]). In 18 (54.5%) patients who had tumor specimens available for correlative biomarker analysis, all of four patients achieved partial response had 50% or more proportion score of PD-L1 immunostaining and higher PD-L1 RNA expression compared with non-responders (p=0.0471).

Conclusion:
Pembrolizumab showed promising antitumor activity in patients with refractory or relapsed TET. Given the relatively high incidence of irAEs especially in thymoma, early detection and management of autoimmune toxicity is essential to ensure feasibility of pembrolizumab treatment in patients with TET.

Abstract not provided

Background:
Paragangliomas (PGLs) are rare neuroendocrine neoplasms arising from paraganglia of the autonomic nervous system. Only about 2% of all PGLs are found in the mediastinum; therefore they are not thoroughly investigated. Our study aims to characterize the clinicopathologic, immunophenotypic and genetic features of mediastinal PGLs.

Method:
Surgical files of Mayo Clinic Rochester and an institutional database (1973-2015) were searched for mediastinal PGLs. Thirteen patients were previously reported (Brown ML et al, Ann Thorac Surg 2008). All cases were reviewed by a thoracic pathologist to confirm the diagnosis. Immunohistochemistry was performed using antibodies to SDHB, Ki-67, ATRX, p53, PD-L1 (clone SP263) and ASCL1. Ki-67 labelling index (LI) was calculated as the mean percent Ki-67-positive cells per 3 HPF. Next generation sequencing (NGS) used a 50-gene neuro-oncology panel. Clinical data were retrieved from medical records. Statistical analysis was performed.

Result:
Twenty-four patients with mediastinal PGLs (7 men, 29.2%) had a median age at time of surgery of 44.6 years (19.8-72.2). Twenty-one (87.5%) paragangliomas were completely resected, 3 had an incomplete resection. The median tumor size was 5.1 cm (1.6-14). The median Ki-67 LI was 4.1% (0.5-29.7). PD-L1 was expressed in 10% (n=2), 1% (n=4) or 0% (n=17) of tumor cells. ASCL1 was focally expressed in 2 of 23 (8.7%) cases. Diffuse loss of SDHB expression was seen in 17 of 22 (77.3%) cases. ATRX was expressed in all 22 cases tested, but its expression was patchy in 1 case that showed ATRX duplication by NGS. NGS, performed in 19 cases, revealed a single pathogenic mutation in 10 cases including SDHB (n=3), SDHD (n=6) and ATRX (n=1); and two or more mutations in 2 cases including SDHC+TERT (n=1) and SDHB+ATRX+TP53 (n=1). Two additional patients were found to have an SDHB mutation. Ten of 21 patients with available SDHx mutation status (47.6%) had no SDHx mutation. Median follow up, available in 21 patients, was 4.7 years (0.8-11.5). Five patients (2 of which with SDHx mutation) developed metastases and/or recurrence; 4 patients (2 without SDHx mutation, 2 with unknown SDHx mutation status) died (1 perisurgically, 3 of unknown cause at 2.3, 5.8, and 10.6 years after surgery each). Statistical analysis of clinicopathologic features based upon SDHx mutation could not be performed due to the small number of cases.

Conclusion:
Our study demonstrates that the majority of mediastinal PGLs harbors an SDHx gene mutation. Alterations in ATRX gene might represent another genetic event in mediastinal PGLs.

Background:
Nomogram has demonstrated its capability in individualized survival estimates. Some nomogram studies on esophageal squamous cell carcinoma (ESCC) have been reported, but accuracy of those nomograms is not high enough. Also, the role of ESCC nomogram in adjuvant chemotherapy remains unclear. Therefore, clinicopathological prognostic nomogram was developed and validated for patients with ESCC in this study. Its role in personalized adjuvant chemotherapy was investigated as well.

Method:
Data were retrieved about 1042 ESCC patients undergoing right transthoracic radical esophageactomy at Sun Yat-sen University Cancer Centre from January 1997 to December 2013. Of 1042 patients, 886 without adjuvant chemotherapy were divided into training set (Group A, n=590) and validation set (Group B, n=296) in a 2:1 ratio. Group C (the remaining 156 patients with adjuvant chemotherapy plus Group B, n=452) was used to evaluate the role of nomogram in personalized adjuvant chemotherapy. Cut-off points of continuous variables were established by X-tile. Survival and univariate analyses were calculated by Kaplan–Meier method. R software and "rms" package were used to perform Cox proportional hazard (CPH) regression model, plot nomogram, compute C-index, compare models, validate model, and plot calibration curve. The "survivalROC" package was used to plot time-dependent receiver operating curve (ROC).

Result:
The 1-, 3- and 5-year overall survival (OS) for the entire cohort was 88.3%, 64.5% and 54.8%, respectively. Univariate analyses were performed for 13 potential clinicopathological factors. Significant variables were analyzed for CPH regression model using R software and “rms” package. A prognostic nomogram including 8 factors (grade, location, T, N, resected negative nodes, length, gender, drinking history) was developed. C-index of the model was 0.739 (95% CI 0.719-0.759), higher than that of the 7[th] TNM staging system (0.696, 95% CI 0.676-0.716), p < 0.001. The calibration curve and time-dependent ROC showed this nomogram was superior to the 7[th] TNM staging system. The cut-off of prognostic score was 160, by which we grouped patients into low and high prognostic risk subgroup. In Group C, 209 patients belonged to high prognostic risk subgroup. Among them, patients receiving adjuvant chemotherapy had longer OS (p = 0.035). The remaining 243 patients belonged to low prognostic risk subgroup. Adjuvant chemotherapy didn’t improve OS (p = 0.799) in this subgroup.

Conclusion:
An accuracy clinicopathological prognostic nomogram was developed and validated for ESCC patients undergoing right transthoracic radical esophagectomy. The nomogram provided individual prediction of survival. Risk group stratification based on nomogram prognostic score successfully guided personalized adjuvant chemotherapy for ESCC patients.

Abstract not provided

Background:
Experimental studies have shown different pathways of lymphatic drainage from particular pulmonary lobes. Especially important is lymphatic drainage from the left lower lobe to the contralateral mediastinal nodes. The aim of this study was to analyse the impact of bilateral mediastinal lymphadenectomy (BML) on survival in non small-cell lung cancer (NSCLC) patients.

Method:
Prospective, randomised trial including patients with proven or suspected NSCLC, stage cI-IIIA. Randomisation was performed in ratio 1:1. In the BML group, systematic lymph node dissection (SLND) was supplemented with contralateral mediastinal lymphadenectomy via additional cervical incision. In the SLND group, standard lung resection with SLND was performed.

Result:
102 patients were enrolled. 13 of them met the exclusion criteria, and data of 89 patients were analysed: 40 in the BML group and 49 in the SLND group. There were no significant differences between groups regarding age, sex, Thoracoscore, Revised Cardiac Risk Index, dyspnoea, lobar location of the tumour, histology and cTNM. Mean follow-up time was 66.5 months. In the whole group, the 4-year survival rate was significantly higher in the BML group than in the SLND group (72.5% vs 51%, p=0.039). Separate comparisons were performed for different lobar locations of the tumour. There was no significant difference in 4-year survival rates and mean survival time between both groups for tumours located in the right lung and those located in the left upper lobe. For the left lower lobe, the 4-year survival rate, and mean survival time was significantly higher in the BML group (90.9% vs 25%, p=0.003, and 1923 vs 1244 days, p=0.027, respectively). Also, analysis of the survival curves (Figure) has shown significant difference (p=0.018.).Figure 1



Conclusion:
For NSCLC located in the left lower lobe, removal of the contralateral mediastinal lymph nodes is associated with survival benefit. These results should be confirmed in larger studies.

Background:
Induction chemoradiation (ICR) for advanced non-small cell lung caner is often limited to 50Gy or less to avoid perioperative complications. Pulmonary resection following definitive chemoradiotherapy (DCR) has been an alternative approach for locally advanced lung cancer.

Method:
In this study, we compared pulmonary resection following ICR and DCR. From 1997 to 2016, we had 31 pulmonary resections following CR. There were 13 ICR and 18 DCR. Intercostal muscle flaps were used in 7 ICR and 2 DCR. Omental flaps were used in 12 DCR. Pericardial fat pad was used in 1 DCR．There was no mortality in any groups.

Result:
In comparison with ICR and HCR, operation time (min, interquartile range) were 344 (283-513) and 418 (563-572) (p = 0.057), estimated blood loss (ml) were 440 (225-575) and 525 (323-1313) (p =0,262), morbidity (%), 69 and 28 (p = 0.021). Bronchopleural fistula developed in one case of DCR who used intercostal muscle flap. Post operation hospital stay (days) was 21 (13.5-26.5) in ICR, 14.5 (13-20) in DCR (P = 0.221). Although operation time was longer and there were more blood loss in DCR, there was no significant increase of peri- and post-operative complications. 2- and 5-year over all survival rates (%) were 50 and 42 in ICR, and 68 and 51 in DCR (p=0.73, log-rank test).

Conclusion:
As a conclusion, high dose ICR may contribute to better local control and longer survival. Pulmonary resection after DCR is as safe as that following ICR.

Background:
To evaluate (1) the potential effect of primary tumor resection, an aggressive local consolidative therapy, for patients with oligometastatic NSCLC on 3 year overall survival; (2) the surgical outcomes in the treatment of patients with oligometastatic NSCLC; (3) the potential clinical factors predicting survival in order to better select patients for surgery.

Method:
According to the extent of pulmonary resection, the patients were divided into two subgroups. A. intent to cure (ITC: removal of total or primary pulmonary lesions); B. intent to biopsy (ITB: preservation of major lesions, only diagnostic biopsy via minimally invasive approach). M stage classified based on 8th UICC/AJCC TNM M categories.

Result:
From Jan 2002 through Dec 2015, a total of 115 consecutive metastatic NSCLC patients were enrolled from Peking University Cancer Hospital. The 3-year overall survival (OS) of ITC and ITB were 64.3% and 34.9% (log-rank p = 0.0009), respectively. Multivariate cox proportional regression analysis identified multiple station lymph nodes (LN) and bone involvement may be prognostic indicators. Figure 1Figure 2





Conclusion:
The current findings suggest that aggressive surgical therapy can extend the survival in selected stage IV NSCLC patients, and should be further explored in phase 3 trials as a standard treatment option in this clinical scenario.

Abstract not provided

Background:
The confirmation of an appropriate resection margin from the tumor is crucial for reducing the risk of local recurrence after lung segmentectomy for pulmonary malignancies. And the precise anatomical segmentectomy is also important for preserving pulmonary function. We evaluated intermediate results of tumor recurrence after anatomical ICG segmentectomy based on the virtual segmentectomy simulation.

Method:
From August 2014 to May 2017, forty-five patients underwent pulmonary segmentectomy under the guidance of ICG fluorescence. Before operation, several types of virtual segmentectomy were created by using Volume Analyzer Synapse VINCENT (Fujifilm co., Tokyo, Japan). We measured the shortest distance from the tumor to the resection margin in each simulated segmentectomy and selected the most appropriate area of sublobar resection based on the adequate resection margin of approximately 2 cm from the tumor. After this virtual segmentectomy, we performed segmentectomy by using an infrared thoracoscopy with transbronchial ICG instillation. Before operation, 10ml of 10-fold diluted ICG with autologous blood and 400ml of air were instilled into each associated subsegmental bronchus. Segmentectomy was performed under ICG visualization. We evaluated tumor recurrence and survival after the operation.

Result:
Thirty-seven patients were primary lung cancer and eight patients were metastatic lung tumor. Active limited resection was done in 28 patients, passive limited resection was in nine and metastatic lung tumor resection was in eight. Subsegmental resection was done in five, segmental resection in 22 and extended segmentectomy, which indicates resection of several segments with adjacent subsegment(s), was 18. The average shortest distance from the tumor to the resection margin in simulation and resected specimen were 22.5+/-11.7 mm and 24.1+/-7.3 mm, respectively (p=0.405). Postoperative complications were prolonged air leak longer than seven days in two cases and atrial fibrillation in one. In terms of the recurrence and survival after ICG segmentectomy, although the mean duration of follow-up was still short (530+/-349 days), no cancer recurrence in the ipsilateral lung was identified in lung cancer patients. In particular, no recurrence was found in the lung as well as lymph node in active segmentectomy patients.

Conclusion:
The combination of lung volume analyzer and ICG segmentectomy was an excellent tool for precise anatomical segmentectomy with an appropriate resection margin and excellent control of tumor recurrence.

Background:
A 1995 survey of Society of Thoracic Surgeons (STS) members revealed widespread variation in post resection lung cancer surveillance practices as well as pessimism regarding any survival benefit. We sought to compare contemporary practice patterns and attitudes toward post-resection surveillance among members of STS and of European Society of Thoracic Surgeons (ESTS).

Method:
A survey identical to one conducted in 1995 was administered via mail or electronically to surgeon members of the STS and ESTS requesting response from those treating NSCLC. Demographic information, type, frequency and duration of post-resection testing were collected. Goodness of fit tests were used to compare profiles of respondents and attitudes toward testing between groups. Analyses were performed using SAS 9.4 (SAS Institute, Cary, NC).

Result:
Between 8/2016-10/2016, 2978 STS members (response rate 7.8%, n=234) and 1450 ESTS members (response rate 8.4%, n=122) were surveyed. ESTS and STS respondents were similar in their frequent use of history and physical examination (75% vs 78%, p=0.26) and CT chest (71% vs 73%, p=0.61) although the timing of testing was widely variable. Compared to STS members, ESTS members reported more frequent use of testing not recommended by guidelines (for asymptomatic patients) including CXR (46% vs 34%, p=0.02) bronchoscopy (10% vs 1% p<0.01), bone scan (5% vs 0, p<0.01), brain CT (6% vs 0, p<0.01), and brain MRI (3% vs 0%, p=0.01). Compared to STS members, ESTS surgeons were more likely to “agree” or “strongly agree” that routine testing for NSCLC recurrence results in potentially curative treatment (86% vs 70%, p<0.01). Similarly, ESTS respondents were more likely to believe surveillance would identify a curable second primary NSCLC (94% vs 84%, p<0.01). More ESTS than STS respondents believed that current literature documents definitive survival benefits from routine follow-up testing (57% vs 30%, p<0.01), a misconception reported by only 26% STS respondents in the identical 1995 survey.

Conclusion:
Our survey suggests significant differences between ESTS and STS members regarding the method, frequency and attitudes regarding post-resection surveillance for NSCLC. European surgeons report a more optimistic belief in significant survival benefit from early detection of both recurrent and second primary NSCLC thus adopting more aggressive surveillance practices. This is in spite of a lack of definitive evidence-based literature underscoring the need for both better prospective studies and joint recommendations to standardize practice.

Background:
Tubeless technique, defined as non-intubated anesthesia and omitting chest tube after lung resection surgery, is a new concept to further minimize surgical trauma of thoracoscopic surgery. However, the feasibility and safety have been less investigated. Here we set up a protocol to prevent postoperative pneumothorax after tubeless single-port thoracoscopic surgery with the aid of digital chest drainage system (DCS).

Method:
From Nov. 2016 to Jun. 2017, 34 consecutive non-intubated single-port thoracoscopic surgery were performed in patients with pulmonary nodules. After excluding patients with nodule≧2 cm, intrapleural adhesion, and FEV1< 1.5 L/sec., 21 patients were selected to enter the tubeless protocol. At the end of the procedure, a single 16-Fr. catheter was placed into the pleural cavity and connected to a DCS which pressure was set at −15 cmH2O. Then the single incision was closed continuously. If the air flow reached zero after completion of wound closure, the catheter will be removed immediately; otherwise the catheter will be kept for drainage. The clinical characteristics and perioperative outcomes of patients were presented. Figure 1



Result:
Among tubeless protocol cases, 3 patients were detected to have minor air leak by DCS and were converted to have intrapleural drainage remained after closure of surgical wound. Among the remaining 18 patients, in whom the DCS showed no air leak, the intrapleural drainage catheter was not placed. Immediate postoperative chest roentgenogram showed full expansion in all protocol patients without pneumothorax. Only 5 (23.8%) patients developed minor subclinical pneumothorax on the first postoperative day. All patients were discharged uneventful without the need of intervention.

Conclusion:
Our tubeless protocol utilizes DCS to select patients for omitting intrapleural drainage after non-intubated single-port thoracoscopic surgery. With objective parameters by DCS, we believe that this is an effective way to reduce the rate of pneumothorax after tubeless single-port thoracoscopic surgery in selected patients after lung resection.

Background:
Lobe-specific lymph node dissection (L-SND) was proposed for clinical T1a-2b N0-1 non-small cell lung cancer (NSCLC), however, the benefit of this approach is still uncertain, especially for pathological nodal staging. In this study, we evaluated the percent detection of pN2 disease in L-SND and in systematic lymph node dissection (SLND).

Method:
From 2010 to 2016, 166 patients with cT1a-T2b N0-1 NSCLC underwent a lobectomy with SLND at Chiang Mai University Hospital. The pathologic results of the lymph nodes dissected in each station were extracted form medical records. Patients who underwent a SLND when then reclassified as L-SLN according to the site of the primary tumor; right upper lobe (station 2R-4R), left upper lobe (station 4L-6), and both lower lobes (station 7-9). Percent detection of pN2 disease was compared between L-SLN and SLND.

Result:
The rate of detection of pN2 in the SLND was higher than in the L-SLD, but the difference was not statistically significant (27.0 %versus 23.6%, p=0.474). The overall percent of cases upstaged to pN2 was only 4.4% after SLND; it was 4.4% in right upper lobe, 3.4 % in left lower lobe, 3.2 % in right lower lobe, and 1.9 % in left upper lobe (p=0.904). The pN2 disease detection agreement between L-SND and SLND was high (kappa=0.911 (95% CI; 0.784 - 0.957). Table 1 pathological N2 status in lobe-specific dissection and systematic lymph node dissection

Location of primary tumor	Clinical N0-1 status		Pathologic N2 status (lobe-specific dissection)		Pathologic N2 status (Systematic dissection)		% upstaged to N2 disease
	Negative	Positive	Negative	Positive	Negative	Positive
RUL (n=67)	57 (85.1)	10 (14.9)	52 (77.6)	15 (22.4)	49 (73.1)	18 (26.9)	4.5
RLL (n=31)	29 (93.65)	2 (6.5)	24 (77.4)	7 (22.6)	23 (74.2)	8 (25.8)	3.2
LUL (n=51)	43(84.3)	8 (15.7)	37 (72.5)	14 (27.5)	36 (70.6)	15 (29.4)	1.9
LLL (n=29)	24 (82.8)	5 (17.2)	23 (79.3)	6 (20.7)	22 (75.9)	7 (24.1)	3.4

Conclusion:
These results suggest that lobe-specific lymph node dissection is as adequate as SLND for pathological N2 staging in clinically early stage NSCLC surgery. However, systematic lymph node dissection achieved a higher detection of pN2 disease. Further studies with a larger sample size are warranted to confirm these results.

Abstract not provided

Background:
Large-scale genomic characterization of large-cell neuroendocrine carcinoma (LCNEC) has revealed several putative oncogenic drivers. There are, however, little data to suggest that these alterations have clinical relevance.

Method:
We performed comprehensive genomic profiling of 68 stage IV LCNECs of the lung (including next-generation sequencing) and analyzed differences in the clinical characteristics of two major LCNECs subtypes: KRAS mutation and PIK3CA mutation. In order to better understand the divergence that might exist between brain metastases and their lung primaries, we performed whole-exome sequencing of paired lung primaries and brain metastases from four lung LCNEC patients.

Result:
Patients with PIK3CA mutation tumors had aggressive disease marked by worse survival (median OS 7.9 vs. 18.6 mo, P = 0.002), higher metastatic burden (> 3 organs 15.2% vs. 4.7%, P = 0.029), and greater incidence of brain metastases (19.0% vs.2.3% in others, P = 0.001). Whole-exome and RNA sequencing on paired brain metastases and primary LCNECs of the lung revealed that LCNEC primaries that gave rise to brain metastases harbored PIK3CA mutation. Significant tumor growth inhibition with GDC0941 was observed exclusively in the LCNEC patient-derived xenograft model that harbored PIK3CA mutation.

Conclusion:
PIK3CA mutation defines a distinct disease phenotype characterized by brain metastasis in LCNEC of the lung. The result may be relevant for targeted therapy and prophylaxis of NSCLC brain metastases.

Background:
Rovalpituzumab tesirine is a promising first-in-class DLL3-targeted antibody-drug conjugate for the treatment of HGNECs. In clinical practice, biopsies are often rendered for diagnoses of HGNECs before treatment. We tested DLL3 in paired biopsy and surgical specimens, aiming to assess the reliability of the scoring system in biopsy specimens and the correlation with HGNEC clinical characteristics and prognoses.

Method:
A total of 378 patients with de novo HGNECs, including 43 LCNECs and 335 SCLCs, were recruited between 2006 and 2015. All 43 LCNECs and 42 of 335 SCLCs had paired biopsy and surgical excision specimens, and the remaining 293 SCLCs had only biopsies. Immunohistochemical evaluation of DLL3 expression was performed using anti-DLL3 antibody (Abcam, ab103102) and was determined using immunohistochemical H score (HS).

Result:
No significant differences of DLL3 expression levels were observed in paired biopsy and excision specimens of LCNECs and SCLCs (Figure B-C). Discordant DLL3 results (high, HS > 150 vs low, HS ≤ 150) in paired specimens were observed in none of LCNECs and 2 of 42 SCLCs. DLL3 levels were significantly higher (p = 0.015) in all SCLCs (n = 335, median HS 200, IQR 100-300) than in LCNECs (n = 43, HS 160, IQR 100-200; Figure D). SCLCs with high DLL3 levels were more frequently male (p = 0.037), smokers (p = 0.019), and TTF-1 positive (p = 0.005) than SCLCs with low DLL3. SCLCs with low DLL3 experienced a superior overall survival compared with SCLCs with high DLL3, with the difference, however, not reaching statistical significance (p = 0.077; Figure F). Figure 1



Conclusion:
Biopsy specimen is a reliable material for evaluating DLL3 expression, which is equivalent to surgical specimen in a large percentage of HGNECs. High DLL3 level in SCLCs demonstrate a correlation with smoking history, TTF1 (neuroendocrine differentiation) and a trend of poor survival.

Background:
Small cell lung cancer (SCLC) shows an aggressive course with early metastases to distant organs. The main treatment regimen for SCLC patients involves platinum based chemotherapy (cisplatin or carboplatin) and etoposide. Genetic alternations, as single-nucleotide polymorphisms (SNPs) in TOP2A (topoisomerase II alpha) and in ERCC1 (endonuclease non-catalytic subunit) genes, were tested as a prognostic and predictive factors in patients with non-small cell lung cancer (NSCLC) treated with chemotherapy. However, there is limited data about clinical relevance of these genetic alternations in SCLC. Therefore, we undertook the present retrospective study to determine the influence of SNPs in TOP2A (rs34300454; rs13695; rs11540720) and ERCC1 (rs11615; rs3212986) genes on efficiency and toxicity of chemotherapy with platinum and etoposide in patients with SCLC.

Method:
The studied group included 103 Caucasian SCLC patients (65 male, 38 female, median age 65 ± 7,5 years). We collected detailed clinical-demographical data including: smoking history, environmental/occupational exposure to carcinogens, performance status, disease stage, and the presence of distant metastases. The response to the treatment was carefully monitored according to RECIST criteria, as well as side effects as anemia, neutropenia or weight loss were noted. For SNPs genotyping, we used DNA isolated from peripheral blood leukocytes using Qiamp DNA Mini Kit (Qiagen, Germany) and TaqMan hydrolyzing probes (Applied Biosystem, USA) in real-time PCR technique on Eco Illumina (Illumina, USA) device.

Result:
Patients with C/C genotype in rs13695 of TOP2A gene had significantly lower risk of neutropenia during chemotherapy than C/T heterozygous patients (p=0,01894; χ²=5.51; OR=2,676; 95% CI=1,165-6,143). Patients harbouring homozygous C/C genotype in rs3212986 of ERCC1 gene had significantly higher risk of anaemia during chemotherapy than heterozygous C/A patients (p=0,04531; χ²=4,01; OR=0,417; 95% CI=0,175-0,991). Furthermore, homozygous A/A genotype in rs11615 of ERCC1 gene was associated with significant prolongation of overall survival (12 vs. 9 months) compared to heterozygous G/A genotype of this gene (p=0,0120; χ²=6,3063; HR=1,657; 95% CI=1,0710-2,5633).

Conclusion:
SNPs in ERCC1 and TOP2 genes are associated with the toxicities and overall survival of SCLC patients treated with platinum and etoposide based chemotherapy.

Abstract not provided

Background:
Small cell lung cancer (SCLC) accounts for 15-30% of newly diagnosed lung cancers. Although chemotherapy and radiation play a vital role in the initial management of limited-stage SCLC, rates of combined modality utilization in the United States have not been comprehensively studied. As such, the National Cancer Database (NCDB) is a valuable resource to understand patterns of care for limited-stage SCLC, as it captures the majority of newly diagnosed thoracic malignancies in the United States.

Method:
All cases of IASLC defined limited-stage SCLC in the United States National Cancer Database (NCDB) were identified from 2004 to 2013. Rates of chemotherapy and radiation utilization were determined along with key factors associated with their use. Kaplan-Meier analysis and multivariable analysis were used to determine factors independently associated with overall survival.

Result:
From 2004 to 2013, there were 70,247 cases with analyzable data in the NCDB that met IASLC criteria for limited-stage SCLC. Of these cases, 40% did not receive radiation and 20% received neither chemotherapy nor radiation. For the irradiated group, the mean radiation dose was 52.8 Gy with a 16.2 Gy interquartile range. On multivariable analysis, being uninsured (OR 0.75, 95% CI 0.67-0.85, p < 0.001), Medicaid (OR 0.79, 95% CI 0.72-0.87, p < 0.001), and Medicare (OR 0.86, 95% CI 0.82-0.91, p < 0.001) were independently associated with a lower likelihood of radiation delivery in comparison to private/managed care insurance (after adjusting for age, tumor stage, and co-morbidity score). The irradiated group had significantly better median survival than the non-radiated group (33 vs. 17 months, p < 0.001). Radiation (HR 0.62, 95% CI 0.6-0.63, p < 0.001) and chemotherapy (HR 0.55, 95% CI 0.54-0.57, p < 0.001) delivery were both independently associated with better survival on multivariable analysis. Adjusted analysis showed that non-academic programs (HR > 1, p < 0.001) and non-private/managed care insurance (HR > 1, p < 0.001) was independently associated with a survival detriment.

Conclusion:
This is the most comprehensive study currently available describing the utilization of combined modality therapy in the initial management of limited-stage SCLC in the United States. A remarkable number of patients received neither radiation nor chemotherapy as part of their initial oncologic treatment. Insurance status was a key determinant of radiation and chemotherapy delivery even after adjusting for potentially confounding factors. Our findings highlight substantial barriers to quality care delivery and challenges in accrual seen for cooperative group clinical trials for limited-stage SCLC.

Background:
Veliparib, a potent inhibitor of Poly (ADP) ribose polymerase (PARP) enzyme potentiates standard chemotherapy against small cell lung cancer (SCLC) in preclinical studies. The combination of veliparib (V) with cisplatin/etoposide (CE) doublet as first-line therapy of extensive stage SCLC (ES-SCLC) showed significant signal of efficacy with adjusted PFS HR: 0.63 1-sided p=0.01. There was strata by treatment interaction indicating different efficacy benefit in patient subsets (adjusted treatment HR comparing CE+V: CE: 0.34; 80% CI: 0.22 - 0.51; 1-sided p<0.001 for male patients with high tumor burden versus adjusted HR: 0.81 80% CI: 0.60 - 1.09; 1-sided p=0.18 for other patients subsets). We explored clinical and tissue-based biomarkers as predictors of benefit from this treatment strategy.

Method:
Post-hoc analysis of clinical data was conducted to identify clinical differences in patients who derived significant benefit from the experimental therapy. Clinical differences were compared between patients in the control and experimental arms within the patient stratum with significant clinical benefit. Similarly, comparison was performed between the strata. Archival tissue samples collected from patients with ES-SCLC enrolled and treated on E2511 study was employed for biomarker analysis using immunohistochemistry to assessSLFN11 and DNA-PK expression. The study has 88% power to detect a PFS hazard ratio of 0.5 comparing biomarker positive to negative patients using a one-sided 0.025 level logrank test.

Result:
There was an imbalance between control and experimental arms in the Male/abnormal LDH stratum (in strata) with respect to Age: p=0.006; malignant pleural effusion: p=0.095 and T stage: p=0.02. Median PFS was 5.1 mos on CE (95% CI 4.1-6.1) vs. 6.2 mos on CE+V (95% CI 5.9-8.8); HR=0.32, p=0.002 (unadjusted); median OS on CE was 8.8 mos (95% CI 6.6-11.1) vs. 9.5 mos on CE+V (95% CI 7.8-12.8); HR=0.76, p=0.39. Mutivariable analysis controlling for these imbalances still showed a benefit of veliparib (HR=0.26, p=0.001). Comparison of “in strata” group (N=46) to the “not in strata” group (N=82) showed significant imbalance in pleural effusion (p=0.058); elevated AST (p=0.0099) and bilirubin (p=0.0447). Median PFS was identical at 5.9 mos for both groups while median OS was 10.7 mos (95% CI 8.9-13.2) for “not in strata” subsests vs. 8.8 mos (95% CI 7.8-10.8) for “in strata” with a HR of 1.57 (p=0.027) comparing “in strata” to “not in strata”. Outcome differences based on SLFN11 and DNA-PK expression will be presented at the meeting.

Conclusion:
PFS benefit of PARP inhibitor therapy in extensive stage SCLC patients with elevated LDH and male gender was not associated with any other clinical characteristics.

Background:
We investigated whether pazopanib maintenance following first-line chemotherapy would improve survival in patients with extensive disease small-cell lung cancer (ED-SCLD).

Method:
This study is a randomized, placebo-controlled, phase II study that enrolled ED-SCLC patients who had not progressed after four cycles of etoposide plus platinum combination therapy. Eligible patients were randomly assigned (1:1 ratio) to either placebo or pazopanib 800 mg per day until progression or unacceptable toxicity. The primary end point was progression-free survival (PFS).

Result:
Ninety-seven patients were enrolled and randomly assigned; 2 patients did not receive study drugs. In total, 95 patients received maintenance therapy (pazopanib, n = 48; placebo, n = 47) and were included into the analyses. Grade 3 toxicities for pazopanib maintenance included thrombocytopenia (10.4%, including 1 case with grade 4 toxicity), liver enzyme elevation (10.4%), fatigue (6.3%), and hypertension (6.3%). Median PFS was 3.7 months for pazopanib maintenance and 1.8 months for placebo (Hazard ratio [HR] 0.44, 95% CI: 0.29 – 0.69, p < 0.0001). Median PFS longer than 6 months were achieved by 9 patients (18.8%) in pazopanib arm and 2 (4.3%) in placebo. Median overall survival for the pazopanib and placebo arms were 10.6 months and 12.9 months, respectively (HR 1.14, 95% CI: 0.74 – 1.76, p = 0.54).

Conclusion:
Though this study met the primary endpoint of PFS, it failed to translate into improvement of overall survival with pazopanib maintenance. Given the unneglectable toxicity profiles, relevant biomarkers to select patients for benefit from pazopanib should be further investigated.

Abstract not provided

Background:
Radiologic assessments of the baseline and post-treatment tumor burden are subject to measurement variability, but the impact of this variability on response categorization and the resulting overall response rate (ORR) in a specific trial has been practically unpredictable.

Method:
We built up a hierarchical model of measurement variability using a clinical trial dataset of CT scans. Simulations were then performed using the model 1) to establish the behaviour of differences between the first and the hypothetical second assessments of percent change of tumor burden in various scenarios, 2) to elaborate on the probabilistic nature of decisions about categorization, and 3) to estimate the variation in the ORR due to measurement variability.

Result:
The extent of the discrepancies between assessments of the percent change depended on the baseline burden. Smaller differences were associated with larger shrinkage of tumor burdens. The simulated probability for a specific categorization (-30% or 20%) to result from reassessment had a sigmoid shape depending on the percent change in the first set of readings, inflecting at the cutoff point for the categorization. In 3 virtual trials having the same baseline burden and the same ORR of 50%, the presence of fewer percent changes around the cutoff in a trial resulted in a more reproducible ORR (95% central range, 35%-65% vs. 40%-60% vs. 45%-60%). Figure 1



Conclusion:
Since determinations of partial response or progression are probabilistic outcomes due to measurement variability, quantification of the variation in the ORR by potential measurement variability is essential and will help inform decisions made on the basis of trial data.

Background:
►►OS assessment requires high follow-up times and patient numbers and is impacted by crossover (CO). OS hazard ratios (HRs) are generally inferior to PS HRs due to impact of post-progression survival (PPS) and CO. Some authors propose that absolute OS gains (ΔOS) should be similar to those in PFS (ΔPFS). Hence, ΔPFS might be a useful OS surrogate (Clin Cancer Res 2013;19:2646; Ann Oncol 2016;27:373).

Method:
To assess this further, we reviewed Journal of Clinical Oncology and New England Journal of Medicine 01/01/2012-06/12/2017 for randomized drug trials in incurable solid tumors. We extracted data for PFS and OS medians and HRs, calculated ΔPFS and ΔOS (experimental medians minus control medians), and did paired comparisons between 2-6 different arms in each study (245 comparisons across 201 trials).

Result:
Mean ΔOS across studies (1.03 months) was similar to mean ΔPFS (1.06 months) (n=201 evaluable, p=0.88). ΔOS correlated with ΔPFS (r=0.50, p<0.0001). With CO in <20% of patients or unstated %CO (n=144), mean ΔOS and ΔPFS were 0.93 and 0.92 months, respectively. With CO in >20% of patients (n=57), mean ΔOS and ΔPFS were 1.29 and 1.41 months, while with CO>50% (n=20), they were 1.4 and 1.9 months. OS HRs (mean=0.92) were inferior to PFS HRs (mean=0.82, n=196, p<0.0001), although OS and PFS HRs correlated with each other (r=0.64, p<0.0001). With CO<20% or unstated (n=135), mean OS and PFS HRs were 0.93 and 0.83, while with CO>20% (n=61), they were 0.90 and 0.80, and with CO>50% (n=20), they were 0.94 and 0.71.

Conclusion:
OS HRs were inferior to PFS HRs, probably due to PPS, competing causes of death and CO. The better mean gains and HRs in high vs low CO trials may be due to more frequently allowing CO in trials with more effective therapies. This increases risk of false-negative OS results with effective therapies if CO is permitted, but it is potentially unethical to withhold CO of effective therapies. With PFS, clinically insignificant gains may be statistically significant. Since ΔOS and ΔPFS are similar, an alternate approach would be a primary study outcome requiring PFS HR to be statistically significant and ΔPFS 95% CIs in a range considered clinically relevant for OS gains. To better understand the limitations of this approach, we are analyzing examples with minimal OS gains despite ΔPFS>2 months and examples of ΔOS>2 months but no gain in PFS, and have formulated a potential biological/statistical explanation for the latter.

Background:
Clinical trials (CTs) are recognized as a key component of a quality cancer care system. When funding for systemic therapy (ST) services in Ontario transitioned in 2014/15 from a one-time payment for new cases to bundled payments for specific evidence-informed regimens, stakeholders expressed concern that the funding model could exclude patients from participation in CTs as treatment facilities would only receive funding when evidence-informed regimens were used.

Method:
A CT policy was implemented to enable public funding through the ST funding model for older and inexpensive drugs and their administration within randomized CTs at the level of the standard of care. Non-randomized CTs were to be funded at the level of best supportive care or other appropriate funding level. New and expensive drugs in a CT could be funded through a separate provincial drug reimbursement program if used according to public funding criteria with administration costs covered by the ST funding model. Each new CT is now assessed to determine the level of public funding possible. The funding model can now capture data on phase of trial, disease type, treatment regimen, trial purpose (adjuvant, palliative) and patient accrual by treatment facility

Result:
During 2015/16 and 2016/17, 43 and 44 lung CTs, respectively, were assessed and activated in Ontario. Trial accrual increased by 33% (from 311 to 413 patients) over the two years since the introduction of the funding model. Accrual varied by facility. In 2016/17, it ranged from a low of 0.25% to 37.5% of the new lung cancer (LC) patients seen at individual facilities. For the five largest cancer centres in Ontario, the percentage of patients recruited ranged from 5% to 18.3% and total accruals from these centres (n=257) comprised 63% of provincial LC trial recruitment. Five immuno-oncology trials accrued 183 patients and made up 44% of total LC trial accruals. Public funding through the ST funding model amounted to $415,000 in 2015/16 and increased to $815,000 in 2016/17.

Conclusion:
The new bundled payment system for ST and the CT policy have enabled public funding to support lung CTs. The ST funding model has facilitated the capture of CT data and trends not previously available for LC and other tumours. The new provincial CT policy and payment system do not appear to have negatively impacted participation in lung cancer CTs. The variance in trial accrual between centres warrants further study.

Abstract not provided

Background:
Human epidermal growth factor receptor 2 (HER2, ERBB2) mutation and amplification each occurs in 2% of lung cancers, resulting in receptor dimerization and oncogenic signaling with in vitro sensitivity to trastuzumab. Ado-trastuzumab emtansine is a HER2 targeted antibody drug conjugate linking trastuzumab with the anti-microtubule agent emtansine.

Method:
Patients with advanced HER2 mutant or amplified lung cancers were enrolled into separate cohorts in a basket trial of ado-trastuzumab emtansine, treated at 3.6mg/kg IV every 3 weeks. The primary endpoint was overall response rate (ORR) using RECIST v1.1. A separate cohort included patients with HER2 mutant lung cancers assessed using PERCIST, with pre-treatment 89Zr-trastuzumab PET as correlative imaging. A Simon two stage optimal design was used with type I error rate under 2.7% (and a family wise error rate across baskets under 10%), power of 89%, H0 10%, H1 40%. Other endpoints include progression-free survival (PFS) and toxicity. HER2 testing was performed on tumor tissue by next generation sequencing (NGS), fluorescence in situ hybridization (FISH) and immunohistochemistry (IHC).

Result:
A total of 33 patients were identified by NGS and enrolled. The first HER2 mutant cohort completed accrual of 18 patients with ORR of 44% (8/18 confirmed, 95% CI 22-69%), rejecting null hypothesis. The median PFS was 4 months, and median PFS for responders was 6 months (range 4-11 months). The PERCIST measured HER2 mutant cohort accrued 9 patients, there were 2 confirmed partial responses (PR) in 5 patients evaluated. The HER2 amplified cohort accrued 6 patients, with 3 confirmed PR observed including 1 with concurrent EGFR sensitizing mutation and resistance to erlotinib. Toxicities included grade 1 or 2 including infusion reaction, thrombocytopenia and transaminitis, there were no treatment related deaths. Of the 27 patients in the HER2 mutant cohorts, there were 18 (67%) exon 20 insertions and 9 (33%) point mutations; responders were seen across mutation subtypes (A775_G776insYVMA, G776delinsVC, V659E, S310F, L755P). Concurrent HER2 amplification was observed in 4 of 27 patients by either NGS or FISH. IHC ranged from 0 to 3+ and did not predict response. Of the 6 patients in the HER2 amplified cohort, 2 had concurrent HER2 mutation and 1 had concurrent EGFR L858R mutation.

Conclusion:
Ado-trastuzumab emtansine is active and well tolerated in patients with advanced HER2 mutant or amplified lung cancers as identified by NGS. While cohort expansion is ongoing, this study has met its primary endpoint in patients with HER2 activating mutations. Further development is warranted.

Background:
Entrectinib is a potent, investigational, CNS-active, oral inhibitor of ROS1 with a biochemical IC~50~ (0.2 nM) ~30 times more potent than crizotinib, the only agent approved for the treatment of ROS1-positive NSCLC. Previously, we reported an objective response rate of 85% in 13 ROS1 inhibitor-naïve NSCLC patients who were treated in Phase 1 studies (Drilon and Siena et al, Cancer Discov 2017), including 2 of 3 (67%) patients with CNS disease. Responses were durable, with 1 patient remaining on study for more than 3 years. Entrectinib was well tolerated, with predominantly Grades 1 or 2 adverse events that were reversible with dose modification.

Method:
Patients with ROS1 inhibitor-naïve NSCLC were enrolled across Phase 1 and 2 studies of entrectinib. Patients were screened for ROS1 gene fusions either locally or centrally at Ignyta’s diagnostic laboratory using next generation sequencing. Entrectinib was administered orally at 600 mg once-daily in 4-week cycles. Safety was assessed by monitoring adverse events, laboratory tests, and clinic visits. Tumor assessments were performed at the end of Cycle 1 and every 8 weeks thereafter. All scans were read locally (INV) and by blinded independent central review (BICR) using RECIST v1.1. INV results will be presented except where noted.

Result:
As of 24 May 2017, a total of 32 patients were evaluable for response (median age 52 years, 72% female). At a median follow-up of 12 months, objective responses were observed in 24 of 32 (75% [95% CI: 56.6, 88.5]; 3 complete responses) patients, including 7 of 11 (64% [95% CI: 30.8, 89.1]) patients with CNS disease at baseline. Five of 7 patients with evaluable CNS lesions by BICR experienced confirmed RECIST intracranial responses, for a CNS response rate of 71% (95% CI: 29.0, 96.3). With 19 (59%) patients remaining on study, the median duration of response was 17.2 months (95% CI: 6.5, 36.0) and progression-free survival was 19.1 months (95% CI: 6.5, 36.6). The most common (>15%) treatment-related adverse events were fatigue/asthenia (34%), dysgeusia (34%), dizziness (24%), weight increase (21%), paresthesia (19%), nausea (18%), constipation (18%), and diarrhea (16%). All data will be updated at the time of presentation.

Conclusion:
Entrectinib is well tolerated and has shown promising antitumor activity in ROS1 inhibitor-naïve NSCLC, including patients with CNS disease. Patients with ROS1+ NSCLC and other tumor types continue to be enrolled in STARTRK-2 (NCT02568267) in order to support a potential regulatory filing for entrectinib in this population.

Background:
Lung-MAP (S1400) is a master umbrella protocol designed to establish genomic screening for previously treated squamous cell lung cancer patients (SqCCA), and independently evaluate targeted therapies with matching biomarkers and alternative therapies (designated non-match therapy) in patients without putative markers. The protocol opened June 16, 2014 with four biomarker-driven sub-studies and one non-match sub-study.

Method:
Eligibility stipulated advanced SqCCA, progressing after at least one prior platinum-based chemotherapy, PS 0–2, and EGFR/ALK wild-type. Tumor samples were required and analyzed for gene alterations by FoundationOne NGS assay (Foundation Medicine). The original biomarker and non-match studies were: S1400B evaluating taselisib for PI3K mutations, S1400C evaluating palbociclib for cell cycle gene alterations (CCGA), S1400D evaluating AZD4547 for FGFR mutations, S1400E evaluating rilotumumab and erlotinib for c-MET positive tumors, and S1400A evaluating durvalumab in patients with no matching biomarkers. The original design included randomization to a control arm, but was amended to a single-arm phase 2 design. The primary endpoint for each modified sub-study was response.

Result:
As of June 16, 2017 all original sub-studies have been closed to accrual; 1298 patients registered to the screening component of the trial and 486 patients have registered to a sub-study. Two new sub-studies have been launched and are currently accruing. Details of the completed sub-studies are included in the table.

Sub-study	Final Accrual	Biomarker prevalence/% of sub-study registrations	Closure Date	Response to investigational therapy N (%)	Status
S1400A (non-match)	Total: 116 Durvalumab: 78 Docetaxel: 38	NA/59%	12/18/15 Docetaxel arm closed: 4/22/15	11 (16%)	Administratively closed to enable activation of new non-match study.
S1400B PI3K	Total: 39 taselisib: 31 Docetaxel: 8	8%/9%	12/12/16 Docetaxel arm closed: 12/18/15	1 (4%)	Closed at interim futility analysis.
S1400C (CCGA+)	Total: 54 Palbociclib: 37 Docetaxel: 17	19%/15%	09/01/16 Docetaxel arm closed: 12/18/15	2 (6%)	Closed at interim futility analysis.
S1400D (FGFR+)	Total: 45 AZD4547: 35 Docetaxel: 10	16%/12%	10/31/16 Docetaxel arm closed: 12/18/15	2 (7%)	Closed at interim futility analysis.
S1400E (MET+)	Total: 9 R+E: 4 E: 5	N/A (closed too early)	11/26/2014	N/A	Closed d/t discontinuation of development of rilotumumab

Conclusion:
Lung-MAP as a master genomic screening protocol has demonstrated feasibility with respect to accrual and evaluation of targeted therapies in lower prevalence patient populations. This dynamic, centralized, single-IRB platform is well positioned to efficiently assess multiple novel therapeutics for advanced SqCCA patients.

Abstract not provided