Scientific Program

Filter Results:

Show Only CME Accredited Sessions

  • +

    SH03 - Highlight of the Previous Day Sessions

    • Type: Highlight of the Day Session
    • Track:
    • Moderators:
    • +

      SH03.01 - Thymoma & Outcomes

      07:00 - 07:12  |  Presenting Author(s): Shirish Gadgeel

      • Abstract

      Abstract not provided

    • +

      SH03.02 - Targeted

      07:12 - 07:24  |  Presenting Author(s): Alice T. Shaw

      • Abstract

      Abstract not provided

    • +

      SH03.03 - Biology

      07:24 - 07:36  |  Presenting Author(s): Glen Reid

      • Abstract

      Abstract not provided

    • +

      SH03.04 - Screening

      07:36 - 07:48  |  Presenting Author(s): Stephen Lam

      • Abstract

      Abstract not provided

    • +

      SH03.05 - Q&A

      07:48 - 08:00

      • Abstract

      Abstract not provided

  • +

    SH04 - Highlight of the Previous Day Sessions

    • Type: Highlight of the Day Session
    • Track:
    • Moderators:
    • +

      SH04.01 - SCLC/NET

      07:00 - 07:12  |  Presenting Author(s): Melissa L. Johnson

      • Abstract

      Abstract not provided

    • +

      SH04.02 - Surgery

      07:12 - 07:24  |  Presenting Author(s): Biniam Kidane

      • Abstract

      Abstract not provided

    • +

      SH04.03 - Locally Advanced/IO

      07:24 - 07:36  |  Presenting Author(s): Edgardo S. Santos

      • Abstract

      Abstract not provided

    • +

      SH04.04 - Survivorship & Radiation

      07:36 - 07:48  |  Presenting Author(s): Chad Rusthoven

      • Abstract

      Abstract not provided

    • +

      SH04.05 - Q&A

      07:48 - 08:00

      • Abstract

      Abstract not provided

  • +

    PL03 - Call to Action - Challenges Ahead

    • Type: Plenary Session
    • Track:
    • Moderators:
    • +

      PL03.01 - From Bench to Bedside

      08:15 - 08:30  |  Presenting Author(s): Charles Swanton

      • Abstract

      Abstract not provided

    • +

      PL03.02 - Host Immunity - No Longer a Passive Bystander

      08:30 - 08:45  |  Presenting Author(s): Julie R. Brahmer

      • Abstract

      Abstract not provided

    • +

      PL03.03 - Beaming into the Future

      08:45 - 09:00  |  Presenting Author(s): David Palma

      • Abstract

      Abstract not provided

    • +

      PL03.04 - A Cut Above

      09:00 - 09:15  |  Presenting Author(s): Yolonda Colson

      • Abstract

      Abstract not provided

    • +

      PL03.05 - It's All In The Tissue

      09:15 - 09:30  |  Presenting Author(s): Keith M Kerr

      • Abstract

      Abstract not provided

    • +

      PL03.06 - The Future of Targeted Therapy

      09:30 - 09:45  |  Presenting Author(s): Tetsuya Mitsudomi

      • Abstract

      Abstract not provided

  • +

    Media Availability

    • Type: Press Conference
    • Track:
    • Moderators:
  • +

    EX04 - Mini Oral Abstract Session - MA08.06, MA18.02, MA19.02, MA20.11

    • Type: Exhibit Showcase
    • Track: Advanced NSCLC
    • Moderators:
    • +

      EX04.01 - Outcomes Among Patients with EGFR-Mutant Metastatic NSCLC with and without Brain Metastases

      09:55 - 10:00  |  Presenting Author(s): Janet L Espirito  |  Author(s): Eric Nadler, Melissa Pavilack, Bismark Baidoo, Ancilla W. Fernandes

      • Abstract

      Background

      Brain metastases (BM) in patients with non-small cell lung cancer (NSCLC) are common, with reported frequencies of up to 44%. The epidermal growth factor receptor-mutant (EGFRm) subtype of NSCLC is known to have specific pathologic features that may influence patterns of metastases and outcomes. The prevalence of BM in patients with EGFRm NSCLC is unknown; however, incidence is expected to increase as new treatments emerge that prolong survival. Therefore, an understanding of the impact of BM and burden of illness in this population is needed.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      A retrospective observational matched cohort study of adults with EGFRm metastatic NSCLC, with BM and without BM (NBM), between January 1, 2014, and July 31, 2016, with follow-up to March 31, 2017. Data were extracted from the iKnowMedTM electronic health record database of patients in the US Oncology Network of community oncology practices, supplemented with chart review. Patients were matched 1:1 by age categories and sex. Patients enrolled in interventional clinical trials or with other cancer diagnoses were excluded. Categorical variables were analyzed using chi-squared tests; time to treatment failure (TTF) and overall survival (OS) were assessed using the Kaplan-Meier method starting from the matched line of therapy for TTF and from NSCLC diagnosis for OS.

      4c3880bb027f159e801041b1021e88e8 Result

      A total of 402 (BM, 201; NBM, 201) patients were included; median age was 70 and 77 years, respectively (p<.05). Patients in both cohorts were predominantly female (65%), Caucasian (69%), non-smokers (42%), and had adenocarcinoma (92%). Over 90% of patients were treated in the first-line setting. Median TTF in the BM and NBM cohorts following initial treatment was 10.9 months (95% confidence interval [CI], 9.5-12.0) and 10.4 months (95% CI, 8.9-12.2), respectively. Median OS for the BM and NBM cohorts was 11.9 months (95% CI, 9.7-13.4) and 16.0 months (95% CI, 9.1-20.6), respectively (log-rank p=.017). Median OS from onset of BM in the BM cohort was 10.0 months (95% CI, 7.4-11.2). While CNS symptoms were present in both cohorts, they were significantly higher in the BM cohort; patients in the BM cohort had significantly greater use of home healthcare, physical therapy, and social work services (p<.05).

      8eea62084ca7e541d918e823422bd82e Conclusion

      Median TTF was similar in patients with EGFRm metastatic NSCLC with BM and NBM; however, OS was significantly worse in the BM cohort. Symptom burden and healthcare resource utilization were also higher in the BM cohort. These findings highlight an unmet treatment need for patients with EGFRm NSCLC with BM.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      EX04.02 - The Impact of Treatment Evolution in NSCLC (iTEN) Model: Development and Validation

      10:00 - 10:05  |  Presenting Author(s): Manjusha Hurry  |  Author(s): Daniel Moldaver, Diana Tran, William Kenneth Evans, Parneet Kaur Cheema, Randeep Sangha, Ronald Burkes, Barbara Melosky, Erik Orava, Daniel Grima

      • Abstract

      Background

      Background: The iTEN model was developed to estimate the survival impact of new treatments for advanced NSCLC (aNSCLC) patients. The structure and key assumptions of the iTEN model and outputs validated against published real-world survival data are presented.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Methods: The iTEN model is a discrete event simulation of aNSCLC patients’ treatment plans. Individual patient characteristics (histology, molecular subtypes (EGFR, ALK, ROS1, BRAF, PD-L1), and performance status) are generated by random sampling from Canadian prevalence estimates. All Health Canada approved agents for treatment of aNSCLC are included. Simulated patients start on first-line therapy and move to subsequent lines of therapy in modelled progression events. Up to six-lines of therapy can be included. Time-of-event for progression or death for each patient is calculated based on random probabilities and progression-free survival (PFS) and overall survival (OS) curves modelled from published clinical trials. For example, a simulated ALK+ patient might receive first-line crizotinib, followed by second-line ceritinib and BSC, based on PFS/OS data from PROFILE 1014 and ASCEND-5. Predicted OS is calculated as the cumulative time spent on active therapy and BSC. PFS/OS data were extrapolated using best practices. Treatment on previous therapies was assumed to have no impact on the efficacy of subsequent therapies. Model survival predictions were validated against published real-world estimates from the Ontario Cancer and Austrian (TYROL) registries, by reproducing the same treatment mix in the simulated patients as in the publications.

      4c3880bb027f159e801041b1021e88e8 Result

      Results: iTEN estimated two- to five-year survival rates were similar to those reported by the Ontario Cancer and TYROL registries.

      abstract1 image.png

      8eea62084ca7e541d918e823422bd82e Conclusion

      Conclusions: While further analyses are required, the iTEN model produces survival estimates comparable to published real-world data; therefore, the iTEN model may be a valid tool to estimate aNSCLC patient survival.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      EX04.03 - Prior Therapy and Increased Expression of PD-L1 in NSCLC Tumor Samples

      10:05 - 10:10  |  Presenting Author(s): Anne-Marie Boothman  |  Author(s): Marietta Scott, Marianne Ratcliffe, Jessica Whiteley, Phillip A. Dennis, Catherine Wadsworth, Alan Sharpe, Naiyer A Rizvi, Marina Chiara Garassino, Jill Walker

      • Abstract

      Background

      Tumor PD-L1 expression has been shown to enrich for response to immunotherapy in several indications, including advanced NSCLC. However, the stability of PD-L1 expression over time and its relationship with non-immunotherapy cancer treatment is currently uncertain. We hypothesized that prior chemotherapy or radiotherapy would increase PD-L1 expression.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      In the Phase 2, open-label, single-arm durvalumab ATLANTIC study (NCT02087423), patients who had received ≥2 prior systemic regimens in the treatment of Stage IIIB or IV NSCLC were screened for tumor PD-L1 expression by immunohistochemistry using the VENTANA PD-L1 (SP263) Assay (25% tumor cell [TC] cutoff). PD-L1 expression was assessed using either a recent (<3 months) or archival sample; a subset of patients provided both. The relationship between non-immunotherapy cancer treatment and prevalence of tumor PD-L1 expression ≥25% (TC≥25%) was assessed in patients who received therapy prior to sample acquisition versus those who did not. Pearson’s chi-squared test was used to examine the differences between patient subgroups.

      4c3880bb027f159e801041b1021e88e8 Result

      Of the patients screened for participation in ATLANTIC, 1590 were successfully assessed for PD-L1 expression. PD-L1 TC≥25% prevalence was higher in patients who had received prior radiotherapy or chemotherapy before sample acquisition, with prevalence noticeably higher in those who had received ≥2 lines of prior chemotherapy. Prior EGFR inhibitor treatment did not have any noticeable relationship to TC≥25% prevalence (Table). In the subset of patients with paired recent and archival samples, TC≥25% prevalence remained the same in 74% of cases, increased over time in 19.5%, and decreased in 6.5%.

      Treatment regimen

      Subgroup (n)

      PD-L1 TC≥25% prevalence (%)

      P-value

      Prior tyrosine kinase inhibitor (TKI)

      No prior TKI (607)

      39.9

      0.947

      Prior TKI (411)

      39.7

      Prior EGFR inhibitor (379)

      38.5

      0.154

      Prior ALK inhibitor (15)

      60.0

      Prior chemotherapy

      No prior chemotherapy (145)

      29.0

      0.004

      Prior chemotherapy (873)

      41.6

      Number of lines of prior chemotherapy

      0 (155)

      29.0

      0.031

      1 (10)

      30.0

      2 (138)

      42.8

      >2 (725)

      41.5

      Prior radiotherapy

      No prior radiotherapy (599)

      37.1

      0.034

      Prior radiotherapy (419)

      43.7

      8eea62084ca7e541d918e823422bd82e Conclusion

      PD-L1 expression may increase in response to chemotherapy or radiotherapy and is unlikely to decrease over time. Re-biopsy may provide a more accurate assessment of current tumor PD-L1 expression status when a low/negative result is seen in an archival sample, particularly if the patient has received multiple lines of intervening radiotherapy or chemotherapy.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      EX04.04 - Automatic Nodule Size Measurements Can Improve Prediction Accuracy Within a Brock Risk Model

      10:10 - 10:15  |  Presenting Author(s): Timor Kadir  |  Author(s): Carlos Arteta, Nick Dowson, Petr Novotny, Catarina Santos, Lyndsey Pickup, Sarim Ather, Jerome Declerck, Fergus V Gleeson

      • Abstract

      Background

      The intrinsic variance of manually measured nodule diameters may limit the predictive accuracy of the Brock University Cancer Prediction Model, especially given the relative weight of its coefficient within the model. Size measurements that are automatically derived may improve this prediction performance. This study aims to examine whether automatic nodule segmentation can improve the predictive efficacy of the Brock model.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      A retrospective analysis was performed on all images from the third annual screening study (T2) of the US National Lung Cancer Screening Trial (NLST). Following BTS guidelines, all nodules >=5mm were selected with no other exclusions whether on type, attenuation margin, or otherwise. This resulted in 7551 benign and 314 malignant nodules from 5373 patients with mean age 62±5 years and 3180 males. An automatic segmentation method, based on Deep Learning, was used to segment the nodule volume using a single point placed within the nodule on the CT image as an initialization. The nodule volume, V, was calculated from the segmentation and then converted to an equivalent spherical diameter, Dsph=∛(6/π V).

      We evaluated four implementations: 1) the original Brock model using as input the manual nodule sizes provided in the NLST dataset (BaselineManual); 2) the original Brock model using as input automatically calculated equivalent spherical diameter, Dsph, (BaselineAuto); 3) a Brock model re-fitted to the NLST dataset, using the manual nodule sizes (OptimManual); 4) a Brock model re-fitted to the NLST dataset, using the automatic nodule sizes (OptimAuto).

      Statistics were computed by bootstrapping across 1000 draws without replacement with 70%/30% training/testing per-patient splits. The performance of each combination was measured using Area-Under-the-Receiver-Operating-Curve (AUC-ROC) of predicting nodule malignancy. The relative weighting of size coefficients was also compared.

      4c3880bb027f159e801041b1021e88e8 Result

      The AUC-ROC was 86.5% (95% confidence interval (CI): 83.2, 89.7) for BaselineManual, 87.4% for OptManual (84.3, 90.5), 88.5% for BaselineAuto (85.7, 91.4), and 89.0% for OptAuto (86.2, 91.8). The relative absolute weight of the size coefficient is 0.46 in OptManual (95% CI: 0.42,0.49) increasing to 0.53 (0.49, 0.57) in OptAuto, an increase of 0.07 (0.06, 0.09)

      8eea62084ca7e541d918e823422bd82e Conclusion

      Automatic segmentation appears to improve the prediction accuracy of both the original Brock model and the NLST optimized version. The benefits of repeatable measurements from automatic segmentations are apparent even as direct replacements within the original model, i.e. without re-fitting. However, re-fitting the parameters increases the relative weight of the nodule size coefficient and further improves performance.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      EX04.05 - Q&A

      10:15 - 10:25

      • Abstract

      Abstract not provided

  • +

    ES06 - Oligometastatic Disease

    • Type: Educational Session
    • Track: Oligometastatic NSCLC
    • Moderators:
    • +

      ES06.01 - Current Clinical Trials in Oligometastases

      10:30 - 10:50  |  Presenting Author(s): Anne-Marie C. Dingemans

      • Abstract

      Abstract not provided

    • +

      ES06.02 - Integrating New Systemic Therapy into Trials in OMD

      10:50 - 11:10  |  Presenting Author(s): Sanjay Popat

      • Abstract

      Abstract not provided

    • +

      ES06.03 - Developments in SBRT in the Oligometastatic Paradigm in NSCLC

      11:10 - 11:30  |  Presenting Author(s): Patrick Cheung

      • Abstract

      Abstract not provided

    • +

      ES06.04 - Surgical Considerations in OMD

      11:30 - 11:50  |  Presenting Author(s): Jessica Donington

      • Abstract

      Abstract not provided

    • +

      ES06.05 - Q&A

      11:50 - 12:00

      • Abstract

      Abstract not provided

  • +

    MA23 - Early Stage Lung Cancer: Present and Future

    • Type: Mini Oral Abstract Session
    • Track: Treatment of Early Stage/Localized Disease
    • Moderators:
    • +

      MA23.01 - Buffy Coat Immunooncologic Diagnosis and Prognosis of pStage I Lung Adenocarcinoma

      10:30 - 10:35  |  Presenting Author(s): Harvey Pass  |  Author(s): Wenjie Xu, Chandra Goparaju

      • Abstract

      Background

      Accurate diagnosis of CT detected pulmonary nodules is crucial to decrease futile minimally invasive resections. We hypothesize that the circulating cellular microenvironment (buffy coat [BC]) containing granulocytes, mononuclear cells, and platelets presents immunologic transcriptional profiles for distinguishing adenocarcinoma from granulomas, and can prognosticate pStage I adenocarcinomas.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      BC total RNA was extracted from 48 pStage I adenocarcinomas having R0 resection and 23 granuloma patients found at surgery. Of the cancers, 23 did not recur either systemically or locoregionally and were matched by age, gender, and pack-years to progressors. RNA was profiled using Nanostring PanImmune Oncology panel for 770 immune genes and cytokines. A machine learning algorithm based on Elastic Net using 1000 loops of cross validation was run to estimate AUC, and logistic regression models used for prognostic classification.

      4c3880bb027f159e801041b1021e88e8 Result

      wclc 18 final.pngThe figure depicts the results of the diagnostic profiling in the upper panel (A-C), and prognostication in lower panel (D-F). Unsupervised heat map separated adenocarcinomas (grey) from granulomas, and among the gene set analysis, differences in interleukins, microglial function, antigen processing, cytokines and macrophage functions were the most significant for cohort discrimination. A 33 gene set signature was able to separate granuloma from cancer with mean AUCs of 0.982. Expresion of BC immune related genes were remarkably different between the patients who recurred, and there were differences in profiles of locoregional recurrences compared to systemic recurrences (data not shown).Patients with elevated T-regs but decreased T-cells were characterized by distant progression, while decreased mast cells and increased CD-8 and total T-cells were associated with local recurrence. A three step prognostication index using 19 genes could stratify these patients by time to progression (E,F).

      8eea62084ca7e541d918e823422bd82e Conclusion

      These encouraging data offer the potential for immune transcriptional signatures to better select patients for diagnosis/management of early stage adenocarcinoma. Further validation studies are underway for possible presentation at WCLC18.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA23.02 - Circulating Tumor DNA Analysis with a Novel Variant Classifier for Recurrence Detection in Resected, Early-Stage Lung Cancer

      10:35 - 10:40  |  Presenting Author(s): Vincent K Lam  |  Author(s): Hai T Tran, Mayra E Vasquez, Kitty Li, Kenneth Yuen, Feng Vang, Ariel Jaimovich, Drew Kennedy, Justin Odegaard, Stefanie Mortimer, Steven R. Olsen, Victoria M. Raymond, Ara A Vaporciyan, Mara B Antonoff, Garrett L Walsh, Emily Roarty, Lara Lacerda, Jack A Roth, Stephen Swisher, Chantale Bernatchez, Boris Sepesi, Don Lynn Gibbons, Jianjun Zhang, John V Heymach

      • Abstract

      Background

      ctDNA is a blood-based biomarker with promising potential in lung cancer for minimal residual disease (MRD) assessment and early detection of recurrence. However, data regarding feasibility are limited, especially for stage I-II disease.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      We performed longitudinal plasma ctDNA profiling of early-stage lung cancer patients (pts) that underwent resection at MD Anderson Cancer Center from Apr 2016 to Jan 2017. Plasma ctDNA was analyzed from pre-operative and multiple post-operative time points until disease recurrence. ctDNA profiling was performed using a 30kb Digital Sequencing panel (Guardant Health) covering SNVs in 21 genes and indels in 9 genes that are commonly present in lung cancer. ctDNA profiles from ~30,000 lung cancer pts were used to train a classifier to exclude non-tumor related mutations.

      4c3880bb027f159e801041b1021e88e8 Result

      A total of 40 pts were included in this analysis, comprised of the first 17 pts with recurrence in the longitudinal study and 23 consecutive pts without recurrence. This cohort was primarily stage I and II (15 [38%], 16 [40%]). Histology included adenocarcinoma (29 [73%]), SCC (6 [15%]), and SCLC (2 [5%]). 58% had adjuvant therapy. Median follow-up was 17.7 (3.4 – 24.5) months and median time to recurrence was 7.1 (3.4 – 16.5) months in this selected cohort. At least one ctDNA alteration was detected in 55% (21/38) of pts with evaluable pre-op samples and in 22% (8/37) of pts at 4 weeks post-op. Presence of ctDNA at 4 weeks post-op heralded eventual recurrence with 43% sensitivity and 91% specificity (75% PPV, 73% NPV) and was significantly associated with worse recurrence free survival (p=0.022, HR 6.52; 95% CI 1.3 – 32.6), while also accounting for stage. In the absence of the variant classifier, an additional 7/37 pts had non-tumor alterations detected at 4 weeks post-op with a recurrence sensitivity and specificity of 57.1% and 69.6%. ctDNA was identified in 76% (13/17) of pts prior to or at the time of recurrence. The median interval between ctDNA detection and radiographic recurrence was 91 days.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Detection of post-op ctDNA, as early as 4 weeks after resection of early-stage lung cancer, is associated with significantly increased risk of recurrence. Accurate detection of ctDNA in this MRD setting is enabled by a highly sensitive sequencing platform that incorporates a novel variant classifier to enhance clinical specificity.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA23.03 - Risk Assessment for Indeterminate Pulmonary Nodules Using a Novel, Plasma-Protein Based Biomarker Assay

      10:40 - 10:45  |  Presenting Author(s): Anil Vachani  |  Author(s): Amanda L Fish, Mike Beggs, Luis Carbonell, Zaid Haddad, Alice Juang, Sandy Kamer, Bhavin Patel, Heng Yu, Alan Wu, Pierre P Massion, Mehrdad Arjomandi, James Brown, Neil Trivedi, Tess Rubenstein

      • Abstract

      Background

      To reduce overdiagnosis and overtreatment of non-cancerous pulmonary nodules found on CT scans, a noninvasive and easily administered test is needed to assess clinically significant disease risk. Such an assay should also accurately inform whether additional aggressive evaluation, including lung biopsy or thoracic surgery, is warranted.

      Objective: To determine the performance of a novel, plasma-based multiplexed protein test model when compared to the Veterans Affairs Clinical Factors Model (VA model) for discriminating between a lung cancer diagnosis established pathologically and an Indeterminate Pulmonary Nodule (IPN) found to be clinically and radiographically stable for at least one year.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      The protein biomarker-based risk model had been trained and tested with a cohort of 277 subjects at high risk of lung cancer, aged 25-85, who were current smokers with an indeterminate lung nodule 4-30mm in diameter (121 subject training set; 59 subject test set) from eight medical centers across the US. Using retrospective plasma samples, we compared the protein biomarker model results with the malignant or benign outcomes in an independent validation cohort comprised of 97 subjects from the Vanderbilt University medical center.

      4c3880bb027f159e801041b1021e88e8 Result

      Among the 97 validation study subjects (average age 60.1 years, range 42-83; average nodule size 16.1mm), the protein biomarker model correctly identified as benign or malignant an additional 44 of the 68 (65%) indeterminate pulmonary nodules classified as having intermediate risk by the VA model. Negative predictive value was 0.94. Only three patients with malignant disease were missed (94% sensitivity) while an additional 28 intermediate risk samples (41%) were properly classified as true positive, thus potentially avoiding aggressive interventions in those subjects with benign disease.

      8eea62084ca7e541d918e823422bd82e Conclusion

      This study evaluated a novel plasma protein biomarker assay model as a noninvasive risk assessment aid for characterizing indeterminate pulmonary nodules. When the model results are combined with the VA model, risk stratification for benign nodules is improved compared to current methods in clinical practice. We hypothesize patients with benign disease may benefit the most from this assay by avoiding unnecessary lung biopsy and subsequent overtreatment, while improving patient quality of care and reducing risks from these procedures. Providers and their patients in whom they suspect lung cancer may consider using this novel assay prior to proceeding with more aggressive interventions.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA23.04 - Discussant - MA 23.01, MA 23.02, MA 23.03

      10:45 - 11:00  |  Presenting Author(s): Scott Bratman

      • Abstract

      Abstract not provided

    • +

      MA23.05 - Post-Operative Radiation Improves Overall Survival in Patients with Node-Positive Non-Small Cell Lung Cancer Undergoing Sublobar Resections

      11:00 - 11:05  |  Presenting Author(s): Karl Fabian Lim Uy  |  Author(s): Rickie P Voland, John Michael Varlotto, Malcolm Decamp, Kerri McKie, Debra Maddox, Paul Rava, TJ Fitzgerald, Jennifer Toth, Paulo Oliveira, Michael F Reed, Chandra Belani, Jennifer Baima, Jianying Zhang, William Vincent Walsh, Monali Patel, Max Rosen, Lacey McIntosh, Negar Rassaei, John Flickinger

      • Abstract

      Background

      The incidence and prognosis associated with patients undergoing sub-lobar resections and having positive lymph nodes(PLN) has been rarely studied. Our investigation will retrospectively review this topic.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      The National Cancer Database(NCDB) was queried during the years 2004-2014 to assess patients undergoing sub-lobar resection (wedge, segmentectomy, and sub-lobar-NOS, N = 38,599) and specifically the patients with PLN (N = 5484). Patients were excluded who had any pre-op chemotherapy and/or radiation, had follow-up of less than 3 months, had stage IV disease or >1 tumor nodule. Multi-variable modeling(MVA) was used to determine factors for overall survival (OS). Propensity score matching(PSM) was used to determine pre-operative risk factors for PLN in patients having at least one node examined(N=22712) (matched by median number of nodes examined) and to assess the role of radiation in those patients with node positive disease (matched by age, sex, stage, chemotherapy, and number of nodes positive).

      4c3880bb027f159e801041b1021e88e8 Result

      The incidence of PLN decreased progressively during our study from 17.9% in 2004 to 9.4% in 2014 (N1 8.3-5.0% and N2 9.6-4.4%). A lower risk of PLN was noted for squamous cell carcinomas, bronchoalveolar (minimally invasive) adenocarcinomas, and right upper lobe locations; but the risk increased with age, tumor size and clinical stage. In the node positive group, MVA demonstrated that OS was worse with males, older ages, non-Hispanic Whites (compared to Asian and Hispanic Whites), lowest income quartile, Charlson co-morbidity > 0, grade, tumor size, number of positive nodes, positive surgical margins, length of stay, and not receiving chemotherapy or radiation. PSM demonstrated that radiation increased OS in patients having PLN regardless of margin status or N level involvement.

      8eea62084ca7e541d918e823422bd82e Conclusion

      The incidence of node positive sub-lobar resections has decreased during the years of our study, but still can be found in nearly 10%. In both MVA and PSM, post-operative radiation improves OS.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA23.06 - Small Residual Setup Errors After Image-Guided Radiotherapy Affect Heart Dose and Are Linked to Overall Survival

      11:05 - 11:10  |  Presenting Author(s): Alan McWilliam  |  Author(s): Corinne Johnson, Gareth Price, Corinne Faivre-Finn, Marcel Van Herk

      • Abstract

      Background

      There is limited evidence of the effect of radiotherapy image guidance on survival. This work investigates the relationship between small residual set-up errors following IGRT and overall survival in lung cancer patients (mostly with significant comorbidities), and explores which anatomy may be responsible for observed differences.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Residual setup errors of 546 NSCLC patients treated with an off-line 5mm action threshold correction protocol for bony anatomy were summarized per patient as the mean and standard deviation for each axis, as well as the vector magnitude in a direction from tumour towards the heart, and included in multivariate Cox regression. Delivered dose distributions including residual setup errors were estimated and the difference between the delivered and planned dose was compared for patients who did/did not survive longer than 1 year. Permutation testing (n=1000) assessed significance.

      4c3880bb027f159e801041b1021e88e8 Result

      Residual setup errors were not correlated with any pre-treatment clinical variable. Patients with a residual shift towards the heart (mean ~2 mm, max 5mm) have significantly worse overall survival (hazard ratio 1.310, p = 0.001). The average dose in the heart region changes linearly with the residual shift magnitude towards the heart (~0.8Gy/mm). A higher delivered dose than planned in a region at the heart base (Figure 1, arrow) is associated with poorer survival in multivariate analysis (hazard ratio 1.214/Gy, p<0.001).

      figure1.png

      8eea62084ca7e541d918e823422bd82e Conclusion

      Small residual shifts after IGRT are strongly associated with overall survival in NSCLC patients, with shifts of the high dose region towards the heart leading to worse survival. The most likely cause of shorter survival is a corresponding increase in dose to the heart base. This analysis provides direct evidence of the importance of accurate patient positioning and highlights the significance of the heart base as a dose sensitive organ in thoracic radiotherapy patients with early effects on survival.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA23.07 - Defining the Role of Adjuvant Therapy for Early Stage Large Cell Neuroendocrine Lung Cancer

      11:10 - 11:15  |  Presenting Author(s): Elliot Wakeam  |  Author(s): Sean Stokes, Alexander Adibfar, Natasha B Leighl, Meredith Giuliani, Thomas K. Varghese, Gail Elizabeth Darling

      • Abstract

      Background

      Large cell neuroendocrine lung cancer (LC-NEC) is a rare, high-grade neuroendocrine tumor. Patterns of adjuvant treatment after surgical resection have not been well defined.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Patients with a pathologic diagnosis of LC-NEC were identified in the National Cancer Database 2004-2014. Patient demographics, tumor and treatment characteristics were examined. Survival differences in patients receiving chemotherapy were evaluated using Kaplan-Meier curves, and multivariate hierarchical Cox models were constructed to evaluate the impact of patient, histologic, tumor, treatment and hospital characteristics on overall survival (OS). A conditional landmark of 90-day postoperative survival was used to address immortal time bias and propensity-matching was used to address imbalance in covariates between groups.

      4c3880bb027f159e801041b1021e88e8 Result

      1,793 patients were identified with pathologically stage I LC-NEC, of which 482 (26.9%) received adjuvant chemotherapy. Use of adjuvant chemotherapy remained similar across the study period. Patients receiving adjuvant chemo were younger, less comorbid and more likely to have T2 tumors. Significantly longer survival was observed with the receipt of adjuvant chemotherapy (5-year OS 59.2% vs. 45.3%), which persisted after adjustment in multivariable Cox models (HR 0.69, 95%CI0.58 – 0.82, p<0.0001). Adjuvant chemotherapy was associated with longer survival in patients with tumors 2-3cm (60.4% vs. 41.8%; HR 0.64, 95%CI 0.46-0.89, p<0.0001), and T2 tumors (59.8% vs. 42.1%; HR 0.63, 95%CI 0.50-0.81, p<0.0001), but no differences were observed for LC-NEC patients with tumor size <2cm. Adjuvant chest radiotherapy was not associated with improved survival. T-stage specific propensity-matching confirmed these findings, however the association between survival and adjuvant chemotherapy for patients with tumors 2-3cm was no longer significant.

      slide1.jpg

      8eea62084ca7e541d918e823422bd82e Conclusion

      In this national study of LC-NEC, adjuvant chemotherapy was associated with significantly longer survival in Stage I tumors greater than 2cm. Adjuvant radiation was not associated with survival. A randomized trial of stage T2-4N0 LCNEC is needed to clarify the role of adjuvant chemotherapy in this population.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA23.08 - Discussant - MA 23.05, MA 23.06, MA 23.07

      11:15 - 11:30  |  Presenting Author(s): Kevin Franks

      • Abstract

      Abstract not provided

    • +

      MA23.09 - Minimally Invasive Approaches Do Not Compromise Outcomes for Pneumonectomy, a Comparison Utilizing the National Cancer Database

      11:30 - 11:35  |  Presenting Author(s): Mark Hennon  |  Author(s): Abhinav Kumar, Harshita Devisetty, Thomas A. D'Amico, Todd Demmy, Adrienne Groman, Sai Yendamuri

      • Abstract

      Background

      Minimally invasive approaches are increasingly being used for the conduct of complex surgical procedures. Whether the benefits of minimally invasive approaches compared to thoracotomy for sublobar and lobar lung resection for nonsmall cell lung carcinoma are realized for patients undergoing pneumonectomy is not clear.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      The National Cancer Database was queried for patients who underwent pneumonectomy for NSCLC from 2010-2014. Those who underwent resection by a minimally invasive approach (MIS) were compared with those who were done by thoracotomy (Open) in an intent-to-treat analysis. Associations between potential covariates and treatment were analyzed using the Pearson Chi-square test for categorical variables and Wilcoxon Rank Sum test for continuous variables. Univariable and multivariable logistic models and proportional hazards model were used to assess the effect of surgical approach on 30 day and 90 day mortality and overall survival. Relative prognosis was summarized using odds ratios (OR) and hazards ratios (HR) estimates and 95% confidence limits.

      4c3880bb027f159e801041b1021e88e8 Result

      A total of 4,938 patients underwent pneumonectomy during the study period, of which 755 (15.3%) were completed by minimally invasive approaches (MIS). No difference was noted in 30 and 90-day mortality rates for MIS compared to Open approaches (6.8% and 12.3% vs 6.7% and 11.9% respectively, p = 0.9 and 0.86). Tumor histology and stage characteristics were similar between the two groups. Mean lymph nodes examined was higher in the MIS group compared to Open (17.1± 0.4 vs 16.1 ± 0.2, p=0.034). Surgical approach was not associated with any difference in perioperative mortality with univariable or multivariable analysis. MIS was associated with improved overall survival on univariable analysis, but this was not evident with multivariable analysis.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Pneumonectomy performed by minimally invasive approaches does not compromise perioperative mortality or long term outcomes. Further investigation into the impact of minimally invasive approaches on perioperative outcomes for whole lung resection is warranted.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA23.10 - Cone-Beam Computed Tomography-Guided Microcoil Localization of Pulmonary Nodules During Video-Assisted Thoracic Surgery

      11:35 - 11:40  |  Presenting Author(s): Hideki Ujiie  |  Author(s): Alexander Gregor, Tomonari Kinoshita, Kosuke Fujino, Chang Young Lee, Yamato Motooka, Terunaga Inage, Tatsuya Kato, Elsie Nguyen, Kazuhiro Yasufuku

      • Abstract

      Background

      The standard procedure at our institution for intraoperative localization of non-palpable small lung nodules is computed tomography (CT)-guided microcoil placement prior to video-assisted thoracic surgery (VATS). Typically, microcoil placement is performed in the radiology suite followed by transfer to the operation room (OR). Our institution has built the Guided Therapeutics (GTx) OR, which includes a robotic cone-beam CT (CBCT). The GTx OR allows imaging and therapy to occur in one location. This can improve workflow and reduce patient transportation, which may increase the risk for microcoil dislodgement or the development of pneumothorax/hemothorax. Our objective was to determine the safety and efficacy of CBCT-guided microcoil placement for nodule localization during VATS.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      This is a single center phase I clinical trial (NCT02496624). Patients with small lung nodules who were candidates for standard CT-guided microcoil localization were enrolled. CBCT was used to generate a 3D reconstruction. The lesion was then segmented using Syngo iGuide software. This reconstruction was next integrated into the digital workspace and automatically registered onto the fluoroscopic images, creating ‘augmented fluoroscopy’. The microcoil was placed percutaneously using ‘augmented’ guidance, proximal to the lesion, using local anesthetic. Patients were subsequently induced into general anesthesia, intubated, and positioned for VATS. Minimally invasive resection of the nodule together with the microcoil was performed under standard fluoroscopic guidance.

      4c3880bb027f159e801041b1021e88e8 Result

      A total of 11 patients were enrolled (mean age 70 ± 11SD). The average tumor size on CT was 1.3 cm (range 0.9-1.7). The average deepest depth from the pleural surface was 2.3 cm (1.3-3.8). The average CBCT-guided intervention time was 39 minutes (25-54), and VATS procedural time was 54 minutes (14-78). We were able to detect and successfully resect all nodules. Average total radiation dose was in an acceptable low range (8307 μGy*m2, range, 2402–18,371). There were no intraoperative complications. Average post-operative length of stay was 1.8 days. A pathological diagnosis was made for all patients: 8 primary lung cancers and 3 lung metastases. All surgical margins were negative on final pathology.

      8eea62084ca7e541d918e823422bd82e Conclusion

      CBCT-guided microcoil insertion followed by VATS was safe, with short operative times, short length of stay and 100% diagnostic yield. With the GTx OR’s real-time guidance capabilities, surgeons can operate with increased confidence of finding and removing the target lesion. This technique will become increasingly important in the future with growing numbers of small nodules being detected on CT by lung cancer screening programs.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA23.11 - Lobe-Specific Nodal Dissection for Clinical Stage I and II Non-Small Cell Lung Cancer: Japanese Multi-Institutional Retrospective Study

      11:40 - 11:45  |  Presenting Author(s): Tomoyuki Hishida  |  Author(s): Etsuo Miyaoka, Kohei Yokoi, Hisao Asamura, Katsuyuki Kiura, Kazuhisa Takahashi, Hirotoshi Dosaka-Akita, Hideo Kobayashi, Hiroshi Date, Hirohito Tada, Meinoshin Okumura, Ichiro Yoshino

      • Abstract

      Background

      Systematic nodal dissection (SND) is an international standard of lymph node dissection for non-small cell lung cancer (NSCLC). Recently, lobe-specific patterns of mediastinal lymph node metastases have been recognized, and lobe-specific nodal dissection (LSD) has been proposed for early-stage NSCLC. The purpose of this study was to assess the surgical outcomes according to the extent of mediastinal lymph node dissection for patients with NSCLC by using a nationwide registry database.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      From among 11,663 patients in a Japanese lung cancer registry study for 2004, 5392 patients with clinical stage (c-stage) I or II NSCLC that was completely resected by lobectomy and either SND or LSD were enrolled. Patients who received preoperative therapy or had middle lobe tumor were excluded. In the LSD group, inferior mediastinal (subcarinal) nodes were not dissected for upper lobe tumors, and superior mediastinal nodes were not dissected for lower lobe tumors. To reduce the selection bias, an inverse probability of treatment weighting (IPTW) method using a propensity score was implemented.

      4c3880bb027f159e801041b1021e88e8 Result

      LSD and SND were performed in 1,268 (23.5%) and 4,124 (76.5%) patients, respectively. LSD group included more c-IA and upper lobe tumors relative to SND group, although there was no significant differences in age and preoperative comorbidity. There was no significant difference in postoperative morbidity and mortality between 2 groups. Extended pathological N2 disease outside LSD area was found in 3.2% of the SND group, but recurrences were not different between 2 groups (all recurrences: 22.0% in LSD, 26.9% in SND; local recurrence: 6.1% in LSD, 7.7% in SND; p=0.788). The 5-year overall survival (OS) was 81.5% in LSD and SND in 75.9%. An IPTW–adjusted Cox model showed that LSD did not have a negative prognostic impact and instead was associated with favorable survival (hazard ratio: 0.68, 95% confidence interval: 0.60-0.77).

      8eea62084ca7e541d918e823422bd82e Conclusion

      This retrospective registry study suggested that LSD is an alternative to SND for selected patients with c-stage I or II NSCLC. Future prospective studies are warranted to determine whether LSD is applicable and provides clinical benefit for the general population of patients with cstage I or II NSCLC.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA23.12 - Discussant - MA 23.09, MA 23.10, MA 23.11

      11:45 - 12:00  |  Presenting Author(s): Giulia Veronesi

      • Abstract

      Abstract not provided

  • +

    MA24 - Genomic Evolution, KEAP 3 and More Non-Coding RNA

    • Type: Mini Oral Abstract Session
    • Track: Biology
    • Moderators:
    • +

      MA24.01 - Genomic Evolution Trajectory Depicts Invasiveness Acquisition from Pre-invasive to Invasive Adenocarcinoma

      10:30 - 10:35  |  Presenting Author(s): Chao Zhang  |  Author(s): Zhi Xie, Fang-ping Xu, Jian Su, Song Dong, Qiang Nie, Yang W. Shao, Qing Zhou, Jin -Ji Yang, Xue-ning Yang, Xu-Chao Zhang, Yi-Long Wu, Wen-Zhao Zhong

      • Abstract

      Background

      Accumulation of molecular abnormalities may depict evolution trajectories of tumor initiation and development. However, the genomic profile of early stage adenocarcinoma and molecular mechanism of invasiveness acquisition from pre-invasive to invasive adenocarcinoma remains barely explored.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      We simultaneously collected 20 patients with adenocarcinoma in situ (AIS) (n=5), minimally invasive adenocarcinoma (MIA) (n=5) and stage IA adenocarcinoma (lepidic/acinar predominant) (n=10). Whole exon sequencing (WES) was performed in pre-invasive adenocarcinoma with multi-region specimens and stage IA adenocarcinoma. Analysis of genomic alteration among different pathological status was performed and tumor mutation burden (TMB) was calculated as well as six mutation types individually. Enriched pathways of each pathology were measured through KEGG analysis.

      4c3880bb027f159e801041b1021e88e8 Result

      Baseline characteristics was generated through heatmap with smokers (2/20, 10%) and EGFR mutation (13/20, 65%) among whole population. AIS/MIA indicated much lower number of mutations than invasive adenocarcinoma (IAC) while TMB revealed the same trend without statistical significance. Multi-region sequencing showed high heterogeneity of single nucleotide variation (SNV) in AIS and MIA. Unique SNV presented dominant proportion in initial status. Cluster analysis showed higher copy number variation in AIS/MIA than IAC with cell adhesion molecules (CAMs) enriched in AIS/MIA while variety pathway enrichment in IAC through KEGG analysis. C>A transversions held major proportion in early stage adenocarcinoma and a significant increase in the proportion of C>T and C>G mutation was exhibited when evolving into IAC.

      图片1.png

      8eea62084ca7e541d918e823422bd82e Conclusion

      Intratumor heterogeneity may occur in the very beginning of adenocarcinoma. High copy number variation was dominant event for AIS/MIA while higher tumor mutation burden was seen in IAC. Tobacco signature encompassing C>A transversions dominates the early development of adenocarcinoma and APOBEC signature may play a potential role in acquisition of cancer invasiveness.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA24.02 - Genomic Alterations in Lung Adenocarcinoma Precursor Lesions

      10:35 - 10:40  |  Presenting Author(s): Dennis Wigle  |  Author(s): Michael K. Asiedu, Nanette Reed, MC Aubry, Anja C Roden

      • Abstract

      Background

      Adenocarcinoma in situ (AIS) and minimally invasive adenocarcinoma (MIA) are thought to be precursor lesions of invasive disease. Genomic alterations in these lesions have not been described.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Genomic analysis including whole genome and exome sequencing, and SNP array analysis were performed on 9 AIS and 18 MIA pathologically confirmed samples to identify single nucleotide variants (SNVs), structural variations and copy number variations. Mutation significance and signature analysis were determined by MutSig and NMF analyses. Pathway analysis was performed using ingenuity IPA.

      4c3880bb027f159e801041b1021e88e8 Result

      The range of mutation burden for AIS and MIA was 0.7 to 12/Mb with a median of 1.7/Mb. This compared to a mean of 7.2/Mb for invasive lung adenocarcinoma. Significantly mutated genes identified in AIS and MIA were ELAVL4, LIN37, XCL1, ELK3, RPS9, FBXO2, HLA-B and MYOG, which affected pathways regulating ESR1, ELAVL1 and TP53. Genes with recurrent mutations included MTPN, CDC27, GGT2, CTBP2, EGFR, NCOR1 and TGIF1 and implicated EGFR, MYC and MAPK1 pathways. Somatic mutations were characterized by C>T and T>C transition signature, whereas CNV analysis found high concentrations of copy number amplifications at 6p21.3 to 6p22.1, 8q24.12 to 8q24.3 and at 21q22.3. There were comparable structural variations in the AIS cases compared to MIA.

      8eea62084ca7e541d918e823422bd82e Conclusion

      In contrast to hypothesized models of tumor progression, AIS and MIA can harbor significant genomic alterations and tumor mutation burden. These observations challenge the notion of accumulating mutation burden during the progression to invasive disease. The finding of high mutation burden in some of these precursor lesions also suggests the intriguing concept of immunotherapeutic options for either treatment or chemoprevention.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA24.03 - Biologic Profiling of Pre-Metastatic Niche in Completely Resected Pathological Stage I Non-Small Cell Lung Cancer

      10:40 - 10:45  |  Presenting Author(s): Tomohito Saito  |  Author(s): Hironori Ryota, Mitsuaki Ishida, Kento J Fukumoto, Hiroshi Matsui, Yohei Taniguchi, Hiroaki Yanagimoto, Koji Tsuta, Tomohiro Murakawa

      • Abstract

      Background

      Despite refinement of treatment strategy for non-small cell lung cancer (NSCLC), pathological Stage I NSCLC still develops recurrent disease in approximately 20% of patients even after complete resection. Recently, tumor microenvironment which promotes distant metastasis, or 'pre-metastatic niche', has been indicated to play pertinent roles in postoperative recurrence of cancer. Our aim is to investigate biologic profiles of pre-metastatic niche in pathological Stage I NSCLC.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Eighteen (12.7%) of 141 patients with pathological Stage IA or IB NSCLC who underwent R0 lobectomy between Jan. 2008 and Dec. 2013 developed distant metastasis postoperatively. From archived formalin-fixed paraffin-embedded specimens, these 18 cases of postoperative distant metastasis and matched cases were selected for total RNA extraction. To overcome inherent bias in selecting control patients, one-to-one matched pairs were created using propensity score matching of which model included age, sex, smoking history, and pathological stage. The samples with inadequate mRNA quality/ quantity were excluded. Gene expressions were detected by nCounter (NanoString Technologies, WA, USA) with PanCancer Immune Profilling Panel and PanCancer Progression Panel. Detected expressions were then analyzed and compared between the two groups by nCounter Advanced Analysis (version 2.0.115). Genes with unadjusted P-value < 0.01 were regarded as candidates for further investigation

      4c3880bb027f159e801041b1021e88e8 Result

      From preliminary comparative study on 6 paired cases, distant metastasis group showed upregulations of TPM2, CCL21, SOX2, CXCL12, EGFL7, PTGDS, BGN, PS8L1, ID4 and TGFB, whereas it showed downregulations of MTOR and CCL8 compared to recurrence-free group.

      In LATE-BREAKING ABSTRACT, we will report following results:

      1) Complete dataset of the comparative analysis including all pairs with adequate mRNA.

      2) Immunohistochemstry and/or in situ hybridization of the candidate genes/proteins on pathological sections.volcano plotnew.annotationearly.rec.png

      8eea62084ca7e541d918e823422bd82e Conclusion

      Biologic profilings of brain metastasis from NSCLC may subsequently help to understand underlying mechanism of postoperative distant metastasis and ultimately lead to novel targeted therapy.

      Futher details will be added in LATE-BREAKING ABSTRACT.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA24.04 - Discussant - MA 24.01, MA 24.02, MA 24.03

      10:45 - 11:00  |  Presenting Author(s): Udayan Guha

      • Abstract

      Abstract not provided

    • +

      MA24.05 - Baseline Spatial Heterogeneity of T790M in Tyrosine Kinase Inhibitor Naïve EGFR-Mutant Lung Adenocarcinomas

      11:00 - 11:05  |  Presenting Author(s): Michael Cabanero  |  Author(s): James Kuo, Ni Liu, Ming Sound Tsao

      • Abstract

      Background

      Despite a good initial response, most epidermal growth factor receptor EGFR-mutant lung adenocarcinomas develop resistance after treatment with 1st and 2nd generation tyrosine kinase inhibitors (TKIs). Approximately 50-60% of resistance can be attributed to the EGFRT790M mutation, which provides a higher affinity ATP binding site that outcompetes TKIs and restore constitutive kinase function. Although classically thought of as de novoacquisition, the presence of T790M has been shown to exist in some tumors before TKI-therapy. Obtaining a spatial map of T790M in TKI-naïve tumors can provide insight into development of this type of resistance.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Eight cases of TKI-naive primary lung adenocarcinoma resections that were later found to be positive for T790M post-treatment with TKI were collected from 2004 to 2012. Initial pre-treatment tumor surgical resections were used for DNA extraction. Two tumor sections per case were divided into equal grid-like regions of approximately 8x8 mm2. The number of grids ranged from 16 to 32 per case, based on tumor size (0.8 cm to 6.5 cm). Digital droplet polymerase chain reaction was used to genotype the sensitizing EGFR-mutations and T790M mutations at each region. Allelic frequencies (AF) of the mutations were measured. Recurrence free interval, defined as surgical resection date to date of recurrence detection, and total duration of TKI-therapy were extracted from medical records.

      4c3880bb027f159e801041b1021e88e8 Result

      All eight cases of TKI-naïve EGFR-mutant lung adenocarcinomas were positive for T790M at baseline. T790M tumor burden, defined as the mean allelic frequency of T790M, ranged from 0.17% to 40.15% across the tumors. Three main patterns of distribution were observed. In two cases, T790M was present at low level (<1% AF) prevalence throughout the entire tumor. Five cases were characterized by the presence of distinct sub-clonal region, defined as T790M AF high in one or adjacent regions surrounded by regions with low or zero T790M AF. In one case, T790M was the predominant clone with T790M AF closely matching sensitizing EGFR-mutation AF. T790M tumor burden was not associated with either tumor size or recurrence free interval.

      8eea62084ca7e541d918e823422bd82e Conclusion

      T790M tumor cells exist prior to TKI-therapy in a majority, if not all, EGFR-mutated lung adenocarcinoma that developed T790M mutation as the resistance mechanism to EGFR TKI therapy, rather than by de novo acquisition during TKI-therapy exposure. However, pre-treatment T790M tumor burden did not appear to be associated with recurrence free survival, although this requires more cases for confirmation.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA24.06 - Long Non-Coding Rna Expression Patterns Delineate Infiltrating Immune Cells in the Lung Tumour Microenvironment

      11:05 - 11:10  |  Presenting Author(s): Adam P Sage  |  Author(s): Kevin W. Ng, Erin A Marshall, Katey S.S. Enfield, Greg L. Stewart, Spencer D. Martin, Brenda C. Minatel, Carolyn J Brown, Ninan Abraham, Wan Lam

      • Abstract

      Background

      The tumour microenvironment is characterized by complex interactions between different cell types, including immune cells that may exhibit pro- or anti-tumour effects. Sequencing and deconvolution techniques present opportunities to identify immune cell composition of bulk tumour data; similarly, these have renewed an interest in the non-coding transcriptome and its regulation of immune- and tumour-biology. Numerous long non-coding RNAs (lncRNAs; >200nt) have emerged as regulators of tumour initiation, progression, and metastasis. Additionally, several immune-related lncRNAs mediate fine-level regulation to balance pro- and anti-inflammatory phenotypes; yet, the landscape of lncRNA expression in human immune cells remains uncharacterized. Thus, delineating these multifaceted regulatory networks is critical to cancer immunology, particularly in immunogenic malignancies such as lung cancer.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      RNA-sequencing data of purified immune-cell subsets (CD8+ T, CD4+ T, B, Monocytes, Neutrophils, and Natural Killer) obtained from flow-sorted healthy peripheral blood samples were probed for lncRNA expression. Sequencing reads were aligned to the hg38 reference genome and quantified to Ensembl v89, yielding 4919 expressed lncRNAs. These immune-associated lncRNAs were correlated with immune cell infiltrate in tumour and paired non-malignant lung adenocarcinoma samples (n=54, The Cancer Genome Atlas) as estimated by the proportion of consistent immune-associated methylation profiles, denoted by leukocytes unmethylation for purity (LUMP) scores.

      4c3880bb027f159e801041b1021e88e8 Result

      We observed that lncRNA expression patterns display a greater degree of cell-type specificity than protein-coding genes in immune cells. In fact, 676 lncRNAs had detectable expression in exclusively one cell type. We uncovered previously-uncharacterized lncRNAs that have expression patterns suggestive of immune-regulatory roles. Compared with lung tumour samples, 19 immune-associated lncRNAs were significantly negatively correlated with LUMP scores (r<-0.400, BH-p<0.0100), 17 of which were also strongly positively correlated with CD45 gene expression (r>0.400, BH-p<0.0100) suggesting expression from immune rather than tumour cells. For instance, the lncRNA USP30-AS1 is significantly downregulated in tumours (average fold-change=2.96, BH-p=6.88*10-13), suggesting its relevance to tumour biology; however, higher transcript expression is correlated with decreased LUMP score (r=-0.685, BH-p=1.02*10-4), illustrating its specificity to immune cells.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Here, we present an atlas of cell-type specific lncRNAs in human immune cells. Our data suggest a functional relevance of lncRNAs to the biology of the tumour microenvironment, and the necessary consideration of tumour purity when examining non-coding RNA expression in order to avoid conclusions confounded by immune cells in bulk tumour data. Thus, we provide a resource for further elucidation of genomic links between immune and malignant cells, which may aid the development of future prognostic and therapeutic strategies.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA24.07 - A Novel cis-Acting lncRNA Controls HMGA1 Expression in Lung Adenocarcinoma

      11:10 - 11:15  |  Presenting Author(s): Adam P Sage  |  Author(s): Greg L. Stewart, David A. Rowbotham, Katey S.S. Enfield, Erin A Marshall, Victor D Martinez, Christine Anderson, Wan Lam

      • Abstract

      Background

      High mobility group A1 (HMGA1) chromatin remodeling protein is enriched in several aggressive cancer types, including NSCLC, where mRNA and protein expression are markedly increased. Additionally, high HMGA1 expression has been associated with poor overall survival and chemotherapy resistance. While HMGA1 is deregulated in lung cancer, the mechanisms that mediate its expression are only beginning to emerge. Long non-coding RNAs (lncRNAs), are a class of transcripts have been implicated in the onset of cancer-associated phenotypes in tumourigenesis and metastasis. Recently, an emerging class of lncRNAs - cis-acting - has been shown to regulate the expression of neighbouring protein-coding genes, including oncogenes and tumour suppressor genes. Thus, lncRNAs may represent novel actionable therapeutic intervention points in known cancer driving pathways. Here we investigate the role of a cis-acting lncRNA, RP11.513I15.6, its deregulation in NSCLC, and its relationship with HMGA1.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      LncRNA transcriptomes were deduced from RNA-sequences of 36 microdissected tumour and matched non-malignant tissues. Normalized sequence read counts were used to identify transcripts with significantly deregulated expression (Wilcoxon Signed-Rank Test, BH-p<0.05). Sequencing data obtained from The Cancer Genome Atlas were analyzed to validate these results. SiRNA-mediated knockdown of lncRNA candidates identified in these analyses were performed in a non-malignant lung epithelial cell line (BEAS-2B). Quantitative real-time PCR quantified the effects of lncRNA knockdown on the expression of neighbouring cancer-associated protein-coding genes.

      4c3880bb027f159e801041b1021e88e8 Result

      Our analyses identified RP11.513I15.6, an undescribed lncRNA neighbouring HMGA1, to be significantly downregulated in adenocarcinoma (>2-fold downregulation in 81.5% of cases). This observation was confirmed in our validation cohort. HMGA1 expression was found to be anticorrelated with RP11.513I15.6, as tumours with downregulated RP11.513I15.6 displayed significant overexpression of HMGA1. This suggested that this lncRNA may be a key negative regulator of HMGA1. In vitro experiments demonstrated siRNA-mediated inhibition of RP11.513I15.6 in immortalized lung epithelial cells resulted in a significant increase in the expression of HMGA1 mRNA and protein. Taken together, our results suggest that RP11.513I15.6 is a novel cis-acting lncRNA that negatively regulates HMGA1, and may contribute mechanistically to the maintenance of cancer phenotypes.

      8eea62084ca7e541d918e823422bd82e Conclusion

      We have discovered a novel, 18-fold downregulated transcript that is anti-correlated with expression of HMGA1, a well established oncogene. In vitro studies support the hypothesis that this transcript, RP11.513I15.6, is a cis-acting lncRNA as siRNA-mediated inhibition led to upregulation of neighbouring HMGA1. Characterizing this oncogene regulatory mechanism will not only further our understanding of cancer biology, but could uncover a novel therapeutic intervention point.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA24.08 - Discussant - MA 24.05, MA 24.06, MA 24.07

      11:15 - 11:30  |  Presenting Author(s): Alice Berger

      • Abstract

      Abstract not provided

    • +

      MA24.09 - Synergy Between the KEAP1/NRF2 and PI3K Pathways Drives Non-Small Cell Lung Cancer with an Altered Metabolism

      11:30 - 11:35  |  Presenting Author(s): Sarah Ann Best  |  Author(s): David P Desouza, Ariena Kersbergen, Antonia N Policheni, Saravanan Dayalan, Dedria Tull, Vivek Rathi, Daniel H Gray, Matthew E Ritchie, Malcolm J McConville, Kate D Sutherland

      • Abstract

      Background

      The lung is a highly oxidative environment, tolerated through the engagement of tightly controlled stress response pathways. A critical stress response mediator is the transcription factor Nuclear Factor Erythroid-2-Related Factor 2 (NFE2L2/NRF2), which is negatively regulated by Kelch-like ECH-Associated Protein 1 (KEAP1). Alterations in the KEAP1/NRF2 pathway have been identified in 23% of lung adenocarcinomas, suggesting that deregulation of the pathway is a major driver in lung cancer.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      We generated a novel genetically engineered mouse model (GEMM) whereby Keap1 (Keap1f/f) and Pten (Ptenf/f) were conditionally deleted in the lung, utilising intranasal inhalation of Adenovirus-Cre. The effects on lung pathology were investigated using histopathology, metabolomics and flow cytometry.

      4c3880bb027f159e801041b1021e88e8 Result

      We found that, while loss of Keap1 alone displayed no abnormalities in the lung, loss of Keap1 combined with loss of the tumour suppressor Pten, promoted malignant transformation. We further monitored tumour progression and immune infiltration in the lung, and metabolite profile changes in the serum of the Keap1f/f/Ptenf/f mouse model. Notably, a tumour-specific metabolite signature was identified in the plasma of Keap1f/f/Ptenf/f tumour-bearing mice, which indicated that tumourigenesis is associated with metabolic reprogramming. Furthermore, the immune milieu was dramatically changed by Keap1 and Pten deletion, and tumour regression was achieved utilising immune checkpoint inhibition.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Our study highlights the ability to exploit both metabolic and immune characteristics in the detection and treatment of lung adenocarcinomas harbouring KEAP1/NRF2 pathway alterations.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA24.10 - Interrogating the Metabolic Effects of Keap1 Inactivation in Adenocarcinoma

      11:35 - 11:40  |  Presenting Author(s): Sarah Ann Best  |  Author(s): Sheryl Ding, Ariena Kersbergen, Boris Reljic, Kate D Sutherland

      • Abstract

      Background

      The most frequently altered gene in lung adenocarcinoma (ADC) is the KRAS oncogene. There are currently no effective treatments to target KRAS-mutant ADC. Loss of function in Kelch-like ECH-associated protein 1 (KEAP1) is co-mutated in 18% of KRAS-mutant ADC and is mutually exclusive with inactivating mutations in Tumour protein 53 (TP53). KEAP1 is a negative regulator of the transcription factor NRF2, which regulates cellular antioxidant and metabolic pathways. Metabolic dysregulation is considered a hallmark of cancer cells, which utilize anaerobic glycolysis and anabolic glucose metabolism in preference to aerobic oxidative phosphorylation in a phenomenon termed the Warburg effect.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      We interrogated the consequences of Keap1 loss in KrasG12D-induced ADC using conditional genetically engineered mouse models (GEMMs). To examine metabolic features unique to Keap1-mutant ADC, GEMMs with KrasLSL-G12D/+ alone or with either the p53flox/flox allele (Kras/p53) or Keap1flox/flox allele (Kras/Keap1) were investigated. Gene and protein expression of key enzymes involved in glucose metabolism were measured in spontaneous lung tumors. Glycolytic functional assays were performed in live FACS-isolated tumor cells using a novel protocol.

      4c3880bb027f159e801041b1021e88e8 Result

      Major alterations in glycolytic function were identified as unique features of Keap1 inactivation in lung adenocarcinomas carrying oncogenic activation of Kras. Loss of Keap1 function in Kras-mutant ADC therefore created a pro-oncogenic metabolic environment to drive lung tumourigenesis.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Targeting metabolic dependency in KRAS-mutant tumors may provide a unique method of treating this aggressive subset of lung ADC.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA24.11 - Loss of Tumour Suppressors is Adequate and Sufficient to Drive Lung Cancer in CRISPR/Cas9 Mice.

      11:40 - 11:45  |  Presenting Author(s): Paola A Marignani

      • Abstract

      Background

      In non-small cell lung carcinoma (NSCLC), loss-of-function (LOF) mutations are found in tumour suppressors, highlighting the importance of these genes in the aetiology of lung cancer. The major tumour suppressors (TS) associated with the development of lung cancer are p53, and the kinase LKB1. Unlike oncogenes that have been successfully exploited therapeutically, LOF alterations in TS are difficult to exploit. The goal of our research is to understand how the loss of TS function allows for metabolic and epigenetic adaptation that favour conditions for tumour growth.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      We developed a CRISPR/Cas9 mouse model of lung cancer representative of tumour suppressors lost in NSCLC. Since LKB1 and p53 are two of the most common LOF tumour suppressors found in NSCLC along with activating mutations in Kras, we used a Cre-dependent Cas9 mouse to create mice that simultaneously lack pulmonary Lkb1 and p53 along with activating KRasG12D. Here, we evaluated the development of lung cancers in the context of metabolism and epigenetic marks.

      4c3880bb027f159e801041b1021e88e8 Result

      The metabolic profile of lung tumours harvested from CRISPR/Cas9 mice was significantly different from the mTOR and metabolic profile of lungs harvested from control mice, favouring a switch from glycolysis to mitochondrial metabolism. Epigenetic marks; acetylation and methylation modifications to histones 3 (H3) and H4 were significantly different compared to control mice, indicative of poor prognosis.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Our study is the first to characterize a new CRISPR/Cas9 mouse model for lung cancer that leverages the loss of two common tumour suppressor proteins associated with lung cancer; Lkb1, and p53. We show that the simultaneous loss of these TS leads to activation of mTOR and pro-survival pathways sustained by metabolic adaptation. Further to this, we observe epigenetic marks that agree with poor prognosis. We conclude from our study that patients with lung cancers that lack expression of LKB1 are likely to respond favourably to interventions that simultaneously target aberrant metabolism with modifiers of tumour epigenetic landscape. Our findings suggest that loss of LKB1 expression serves as a marker for NSCLC.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA24.12 - Discussant - MA 24.09, MA 24.10, MA 24.11

      11:45 - 12:00  |  Presenting Author(s): Vladmir Cláudio Cordeiro De Lima

      • Abstract

      Abstract not provided

  • +

    MS22 - Biology of the Lung and Lung Cancer

    • Type: Mini Symposium
    • Track: Biology
    • Moderators:
    • +

      MS22.01 - Lung Development and Stem Cells

      10:30 - 10:45  |  Presenting Author(s): Samuel Rowbotham  |  Author(s): Joo-Hyeon Lee, Carla Kim

      • Abstract

      Abstract not provided

    • +

      MS22.02 - Epigenetic Alterations in Lung Cancer Development

      10:45 - 11:00  |  Presenting Author(s): Wan Lam

      • Abstract

      Abstract not provided

    • +

      MS22.03 - Non-Coding RNA in Lung Cancer

      11:00 - 11:15  |  Presenting Author(s): Sven Diederichs

      • Abstract

      Abstract not provided

    • +

      MS22.04 - Genomics of Resistance and Response in Lung Cancer

      11:15 - 11:30  |  Presenting Author(s): Marc Ladanyi

      • Abstract

      Abstract not provided

    • +

      MS22.05 - Inflammation in Lung Cancer

      11:30 - 11:45  |  Presenting Author(s): Katerina Politi

      • Abstract

      Abstract not provided

    • +

      MS22.06 - Transcriptomic-Metabolomic Reprograming as Resistant Mechanism

      11:45 - 12:00  |  Presenting Author(s): Patrick C Ma  |  Author(s): XIaoliang Wu, Zuan-Fu Lim, Satoshi Komo

      • Abstract

      Abstract not provided

  • +

    MS23 - What's New in Targeted Therapy?

    • Type: Mini Symposium
    • Track: Targeted Therapy
    • Moderators:
    • +

      MS23.01 - Novel Combinations of Targeted Therapies

      10:30 - 10:45  |  Presenting Author(s): Pasi A Jänne

      • Abstract

      Abstract not provided

    • +

      MS23.02 - Emerging or Rare Targets - Fusions, Mutational Burden Etc

      10:45 - 11:00  |  Presenting Author(s): Luis Paz-Ares

      • Abstract

      Abstract not provided

    • +

      MS23.03 - Management of CNS Mets in the Era of CNS-Penetrant TKIs (Med Onc and Rad Onco Perspectives)

      11:00 - 11:15  |  Presenting Author(s): Marina Chiara Garassino  |  Author(s): Giulia Galli

      • Abstract

      Abstract not provided

    • +

      MS23.04 - Targeting Driver Oncogenes in Stage 1-3 NSCLC

      11:15 - 11:30  |  Presenting Author(s): Helena Yu

      • Abstract

      Abstract not provided

    • +

      MS23.05 - Role of Immunotherapy in Patients with Molecular Driven NSCLC

      11:30 - 11:45  |  Presenting Author(s): Benjamin Besse

      • Abstract

      Abstract not provided

    • +

      MS23.06 - Q&A

      11:45 - 12:00

      • Abstract

      Abstract not provided