Scientific Program

Filter Results:

  • +

    ES01 - Advances in Lung Cancer Screening Through Imaging

    • Type: Educational Session
    • Track: Screening and Early Detection
    • Moderators:
    • +

      ES01.01 - Image Quality Characteristics and Nodule Growth Measurement, Medical Physics and Machine Parameters

      13:30 - 13:50  |  Presenting Author(s): Ricardo S Avila

      • Abstract

      Abstract

      Modern CT scanners are routinely being used to determine the malignancy potential of small sub-centimeter pulmonary nodules. Increasingly, this involves CT scanning and quantitative volume measurement of lung nodules over short time intervals (e.g. 3 or 6 months) to determine whether a change in nodule size consistent with malignant growth has occurred. Although it may appear that current CT scanners are more than capable of reliably performing these quantitative measurements with high quality due to their ability to obtain sub-millimeter resolution lung images, many clinical sites are not taking the steps needed to achieve consistent high quality small lung nodule measurement results. A study of volume measurement performance in a phase II clinical trial observed multiple clinical sites using CT scanners which resulted in errors in volume change measurements as high as 43% [1]. In addition, a 2016 crowd-sourcing study of CT scanner image quality performance using the site’s low dose CT lung cancer screening acquisition protocol revealed that 37% of sites used insufficient slice thickness (<= 1.25mm slice thickness is needed) and only 19% of sites used the needed slice thickness and a reconstruction kernel that avoided excessive smoothing and avoided high levels of edge enhancement [2]. Poor CT image acquisition performance has the potential to result in poor lung nodule volume measurement performance which can negatively impact patient care by contributing to unnecessary biopsies and delays in early lung cancer diagnosis.

      To address these issues the RSNA’s Quantitative Imaging Biomarkers Alliance (QIBA®) has developed the QIBA CT Small Lung Nodule Profile that provides a comprehensive set of specifications to ensure that a clinical site attains a minimum level of quantitative imaging performance necessary to achieve a specified lung nodule volume measurement accuracy. The Small Lung Nodule Profile outlines six fundamental image quality characteristics that are needed throughout the full scanner field of view to support precise volumetric measurement of small lung nodules. These characteristics are (1) Edge Enhancement, (2) Three-Dimensional Resolution, (3) Resolution Aspect Ratio, (4) CT linearity, (5) Spatial Warping, and (6) Noise. In general, CT scanners achieve highest fundamental image quality performance at scanner iso-center with some scanners and image acquisition protocols exhibiting large losses in image quality performance as a function of distance from scanner iso-center [3]. These fundamental image quality properties can now be quickly and easily measured by a technologist at any clinical site using a new image quality measurement phantom and fully automated and cloud-based phantom analysis software.

      To determine the clinical impact of achieved CT image quality performance, a new set of modeling and simulation tools has been developed that can create simulated CT images given the image quality characteristics for a CT scanner and image acquisition protocol [4]. Quantitative measurement software can then be applied to these images resulting in expected measurement performance for a clinical task, such as the bias and precision of solid lung nodule volume change measurement for virtual lung nodules of different sizes. Having these estimates of a CT scanners performance can further be used to one day quantitatively determine the minimum time interval needed in order to be able to distinguish malignant nodule volume growth from a stable lung nodule. Regularly performing these measurements also has the potential to offer numerous advantages to lung cancer screening sites including the ability to determine if scans from two different CT scanner models will produce sufficiently similar image quality and measurement performance.

      In summary, a new set of phantoms and cloud-based software tools is available that enables more careful control and optimization of CT lung cancer imaging performance based on fundamental image quality properties. These new tools provide several new opportunities for clinical sites to more precisely perform CT lung cancer imaging studies and measurements.

      References

      [1] Henschke CI, Yankelevitz DF, Yip R, Archer A, Zahlmann G, Krishnan K, Helba B, Avila R, “Tumor volume measurement error using computed tomography imaging in a phase II clinical trial in lung cancer.” Journal of Medical Imaging 3(3), 035505 (Jul–Sep 2016).

      [2] Avila R, Yankelevitz D, Yip R, Henschke C, “P1.03-021 Initial Results from A Novel and Low Cost Method For Measuring CT Image Quality,” January 2017. Journal of Thoracic Oncology 12(1):S554-S555.

      [3] Avila R, Subramaniam R, Henschke C, Yankelevitz D, “Hot Topic: Clinical Implications of CT Image Quality Variation in Low Dose Lung Cancer Screening Scans,” 4th World Congress of Thoracic Imaging Proceedings, Journal of Thoracic Imaging, accepted for oral presentation, 2017.

      [4] Avila R, Jirapatnakul A, Subramaniam R, Yankelevitz D, “A new method for predicting CT lung nodule volume measurement performance,” SPIE Medical Imaging Proceedings, 2017.

      e353dbe42c8654f33588d4da0b517469

    • +

      ES01.02 - Image Interpretation and Advances from the Perspective of the Radiologist

      13:50 - 14:10  |  Presenting Author(s): David F Yankelevitz

      • Abstract

      Abstract

      To a large extent the success of the screening process depends on the algorithm used to manage the findings. Therefore, even though the image production might be identical for two different screening programs, if they differ in the way they manage findings, especially nodules, the comparative results between them would be quite different. It is with this in mind that screening programs develop and choose their workup algorithm. In regard to how these algorithms are chosen, several features must be balanced including the rate of false positives and also potential delays in diagnosing lung cancer. Each of these is associated with a potential cost, the fewer the false positives might be associated with an increase in the number of cases where lung cancer diagnosis is delayed. As we have moved from 5 mm size thresholds to 8 mm size thresholds on the baseline scan we can clearly see how these factors are balanced against each other. The number of false positives dramatically declines as the number of cancers where diagnosis will be delayed by nine months will increase. While it is generally assumed that there is a substantial downside to delaying diagnosis, the challenge is understanding how much the delay actually costs in terms of decrease in the curability of the cancer and this is then considered in terms of how often this might occur.

      Another major feature of management protocols is their dependence on change in size over time to guide the management protocol. Different protocols apply different criteria for measuring change. One of the main differences is that some recommend the use of 3D volumetric analysis while others still rely on 2D measurements. As a general rule, the 3D approach has inherent advantages in that boundaries for the nodule are automatically chosen by the computer, asymmetric growth can be more easily recognized, and the proportional change for a given amount of change is far greater for volume then for diameter measurements. Nevertheless, there may still be circumstances where volume measures can still be misleading and the radiologist still has a very important role in visually inspecting the nodule to confirm whether change has occurred. Along with the measurement of size change, the time interval between measurements is also important in determining whether growth is meaningful. It is not simply enough to say that a nodule is growing, but rather the intent is to understand its growth rate, and this depends on time, with shorter time intervals between scans introducing greater uncertainty.

      There are currently many algorithms that have been developed. Some focus solely on screening such as Lung-RADS, while others are designed primarily for the incidentally detected nodules. Differences between the algorithms focus primarily on the size thresholds used to define a positive result, the time intervals between repeat scans, the choice of management for a positive finding, differences in the management of nodule subtypes (solid, part-solid, nonsolid), and differences between baseline rounds and repeat rounds. These different algorithms will be compared and data will be presented in terms of the influence on the rate of positive results. An additional consideration here is also how we define a positive result. Some algorithms define the positivity based on a size threshold, whereas others consider this based on a growth threshold or a combination of size and growth. When these growth thresholds are used, the rate of positive results dramatically decreases.

      In addition to the finding of lung nodules there are many other findings that commonly occur on the scans such as micro-nodules, areas of atelectasis, perifissural nodules, waxing and waning nodules, endobronchial nodules, presumed pneumonias, that might be found by the radiologist but there are no specific guidance rules for management. Here again the radiologist is confronted with the challenge of attempting to balance excess workup against obtaining a firm clinical diagnosis. While many of these examples have no authoritative guidelines as to how they should be managed, some practical guidance is presented.

      e353dbe42c8654f33588d4da0b517469

    • +

      ES01.03 - Deep Machine Learning for Screening LDCT

      14:10 - 14:30  |  Presenting Author(s): Bram Van Ginneken

      • Abstract

      Abstract

      The first computer algorithms to automatically detect pulmonary nodules in CT scans, based on classical machine learning approaches, were developed almost two decades ago. These systems appeared in commercially available computer-aided detection packages. However, a recent study concluded that such older software systems fail to flag a substantial number of cancerous lesions and have a fairly high false positive rate.

      Recently, algorithms based on deep learning, in particular, convolutional neural networks, have been developed that report high sensitivity with low false positive rates. Similar deep learning algorithms have been successful in classifying nodules as solid, subsolid or part-solid with accuracy comparable to radiologists, and in estimating the probability of malignancy of nodules.

      The 2017 Kaggle Data Science Bowl combined these tasks into a single challenge where 2000 teams developed methods to predict, on the basis of a single screening CT scan, whether a patient would be diagnosed with lung cancer within one year of the date of the scan. The 10 best performing solutions are now available under an open source license and form the basis of commercial solutions that show, in recent validation studies, a performance comparable to radiologists.

      Thorough validation studies are now needed to investigate if the good performance of these deep learning systems can be replicated, independent of CT parameters, and how such systems can be implemented in a lung cancer screening setting. Possibilities include the use of AI software as a second reader, as a concurrent reader, or even a stand-alone reader for a fraction of the cases, when widespread implementation of screening will put a too large burden on scarce radiological resources.

      In this lecture, I will review the currently available computer solutions and discuss their validation and integration into CT lung screening worksflows.

      e353dbe42c8654f33588d4da0b517469

    • +

      ES01.04 - Multi-Phasic Screening - Can We Address Competing Causes of Morbidity * Mortality Such as Coronary Artery Disease and COPD

      14:30 - 14:50  |  Presenting Author(s): Rozemarijn Vliegenthart

      • Abstract

      Abstract

      Lung cancer, chronic obstructive pulmonary disease (COPD) and cardiovascular disease (CVD) are highly prevalent in the Western population (annual incidences in the Netherlands: lung cancer n=12,200, COPD n=53,300, and CVD n=101,700). This results in a high burden on the health care system and associated costs, with annual costs of 10 billion euros in the Netherlands alone. Furthermore, lung cancer, COPD and CAD are responsible for a high burden of morbidity, with disability adjusted live year reduction of 2.9, 3.4 and 5.0, respectively. For these so-called Big-3 diseases, treatment is most often initiated at late stages due to late diagnosis after development of symptoms. Early detection and treatment will cure many patients in time, and delay or stop disease progression. Therefore, prevention and early detection are crucial.

      Currently, no screening is performed for the Big-3. The impact of low-dose computed tomography (CT) lung cancer screening on lung cancer stage shift and reduction of lung cancer mortality has been demonstrated.(1,2) These results have led to recommendations to implement CT screening in high-risk individuals in the USA. Population-based studies have shown the strong relationship between CT-derived extent of CAD and COPD, and mortality, also in lung cancer screening setting.(3-8) However, there is as yet no evidence from randomized controlled trials regarding benefit of CT screening for COPD or CAD. As the high-risk population for the Big-3 is comparable (namely long-term [ex-]smokers), combining imaging biomarkers will likely improve CT screening efficiency.

      Technological developments in CT allow the determination of early imaging biomarkers for the Big-3, namely lung nodule volume, coronary artery calcium score and lung density/ bronchial wall thickness with low-dose CT (see Figure). Combined evaluation of early signs of the Big-3 diseases has not been extensively explored yet. Major advantages of integrated Big-3 screening can be anticipated due to shared risk factors (in particular long-term smoking) and thus overlapping at-risk population, simultaneous presence of B3 diseases, and the health economic yield compared to a single disease. However, at this moment there is no single CT acquisition that allows for accurate assessment of all Big-3 biomarkers. In particular, calcium scoring based on low-dose chest CT, while providing a good correlation on a population basis, is inaccurate for determining the score on an individual basis.(9)

      biomarkers.jpg

      Furthermore, there are several challenges that need to be addressed in the preparation and establishment of a B3 screening program. These include the need for evidence of morbidity/mortality reduction for screening of COPD and CAD. Also, B3 imaging biomarkers, particularly for COPD, need validation and standardization. Another hurdle is the labour-intensive work required to obtain B3 imaging biomarkers. Also, education and training for evaluation of B3 CT screening examinations is lacking. Finally, the cost-efficiency of integral B3 screening has not been established.

      The presentation includes discussion of the background of interest in Big-3 screening, estimated health economic consequences of Big-3 screening, status of imaging biomarker development for the Big-3 diseases, screening population and CT scan protocol, and impediments to Big-3 screening implementation.

      References:

      1. Aberle DR, Adams AM, Berg CD, et al. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med 2011;365:395–409.

      2. van Klaveren RJ, Oudkerk M, Prokop M, et al. Management of lung nodules detected by volume CT scanning. N Engl J Med 2009;361:2221-9.

      3. Oudkerk M, Stillman AE, Halliburton SS, et al; European Society of Cardiac Radiology; North American Society for Cardiovascular Imaging. 2008. Coronary artery calcium screening: current status and recommendations from the European Society of Cardiac Radiology and North American Society for Cardiovascular Imaging. Eur Radiol. 18:2785-807.

      4. Hecht HS, Cronin P, Blaha MJ, et al. 2016 SCCT/STR guidelines for coronary artery calcium scoring of noncontrast noncardiac chest CT scans: A report of the Society of Cardiovascular Computed Tomography and Society of Thoracic Radiology. J Cardiovasc Comput Tomogr. 2017;11:74-84.

      5. Mets OM, Vliegenthart R, Gondrie MJ, et al. Lung cancer screening CT-based prediction of cardiovascular events. JACC Cardiovasc Imaging. 2013;6:899-907.

      6. Oelsner EC, Smith BM, Hoffman EA, et al. Prognostic Significance of Large Airway Dimensions on Computed Tomography in the General Population. The Multi-Ethnic Study of Atherosclerosis (MESA) Lung Study. Ann Am Thorac Soc. 2018;15:718-27.

      7. Mets OM, Buckens CF, Zanen P, et al. Identification of chronic obstructive pulmonary disease in lung cancer screening computed tomographic scans. JAMA. 2011;306:1775-81.

      8. Oelsner EC, Carr JJ, Enright PL, et al. Per cent emphysema is associated with respiratory and lung cancer mortality in the general population: a cohort study. Thorax. 2016;71:624-32.

      9. Xie X, Zhao Y, de Bock GH, et al. Validation and prognosis of coronary artery calcium scoring in nontriggered thoracic computed tomography: systematic review and meta-analysis. Circ Cardiovasc Imaging. 2013 Jul;6(4):514-21.

      e353dbe42c8654f33588d4da0b517469

    • +

      ES01.05 - Q&A

      14:50 - 15:00

      • Abstract

      Abstract not provided

  • +

    MA04 - Novel Approaches with IO

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

      MA04.01 - Cemiplimab, a Human Monoclonal Anti-PD-1, Alone or in Combination with Radiotherapy: Phase 1 NSCLC Expansion Cohorts

      13:30 - 13:35  |  Presenting Author(s): Victor Moreno  |  Author(s): Marta Gil-Martin, Melissa L. Johnson, Raid Aljumaily, Maria Pilar Lopez-Criado, Donald W Northfelt, Marka Crittenden, Salma Jabbour, Lee Rosen, Emiliano Calvo, Kyriakos P Papadopoulos, Pilar Garrido, Asuncion Hervás Morón, Petra Rietschel, Kosalai Mohan, Jingjin Li, Elizabeth Stankevich, Minjie Feng, Israel Lowy, Matthew Fury

      • Abstract

      Background

      Cemiplimab (REGN2810), a human monoclonal anti-PD-1, has exhibited substantial antitumor activities in patients with advanced malignancies in a first-in-human study. We report interim results of the Phase 1 expansion cohorts (ECs 1 and 2) of cemiplimab, alone or plus radiotherapy, in advanced NSCLC (NCT02383212).

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Patients with advanced NSCLC who had relapsed after, or were refractory to, at least, first-line therapy received cemiplimab 200 mg Q2W in EC 1, or cemiplimab 3 mg/kg Q2W plus radiotherapy (9 Gy × 3 times/week 1 week after first dose of cemiplimab) to a single lesion in EC 2. For EC 2, patients were required to have NSCLC for which palliative radiation therapy was indicated. Planned treatment duration was up to 48 weeks in both ECs. The co-primary objectives were to evaluate the safety, tolerability, and efficacy of cemiplimab, alone or plus radiotherapy. Tumor measurements (of non-irradiated lesions) were performed by RECIST 1.1 Q8W.

      4c3880bb027f159e801041b1021e88e8 Result

      As of Sept 1, 2017, 20 patients (13 M/ 7 F; median age 64.0 years [range, 50–82]) and 33 patients (22 M/ 11 F; median age 67.0 years [range, 47–82]) were enrolled in EC 1 and EC 2, respectively. ECOG performance status 1 versus 0 was 80.0% versus 20.0% and 66.7% versus 30.3%, respectively, for ECs 1 and 2, and missing in one in EC 2; 75.0% (EC 1) and 48.5% (EC 2) had received prior radiotherapy. Investigator-assessed overall response rate (ORR; complete response [CR] + partial response [PR]) was 40.0% (1 CR and 7 PRs) and 18.2% (6 PRs) in EC 1 and EC 2, respectively. Disease control rate (ORR + stable disease [SD]) was 60.0% (1 CR + 7 PRs + 4 SDs) and 72.7% (6 PRs + 18 SDs) in EC 1 and EC 2, respectively. The most common treatment-emergent adverse events (TEAEs) of any grade were arthralgia, asthenia, cough, and dyspnea (each 20.0%) in EC 1, and decreased appetite (30.3%), fatigue (27.3%), cough (24.2%), asthenia and back pain (each 21.2%) in EC 2. Grade ≥3 TEAEs occurring in ≥2 patients were pneumonia (10.0%) in EC 1; and anemia (12.1%), hypophosphatemia and urinary tract infection (each 6.1%) in EC 2. One patient in EC 2 experienced TEAE of pneumonitis with an outcome of death, considered related to study drug.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Cemiplimab monotherapy demonstrated substantial antitumor activity in pretreated NSCLC patients. The safety profiles were comparable with other anti-PD-1 agents and radiotherapy.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +
      • Abstract

      Background

      Responses in NSCLC to agents targeting the PD-1/PD-L1 axis are correlated with PD-L1 expression by immunohistochemistry (IHC), tumor mutational burden (TMB), interferon-associated mRNA expression profile (GEP), and the absence of liver metastases. Anti-PD-1 impedes the inhibition of T cells while pegilodecakin (AM0010) stimulates the survival and expansion of intratumoral, antigen-activated CD8+ T cells (Mumm et al, 2010). This provides a rationale for combining anti-PD-1 agents with pegilodecakin.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Pretreated NSCLC subjects (N = 34) received pegilodecakin (10-20 µg/kg QD, SC) with pembrolizumab (2 mg/kg, Q3W, IV; n = 5) or nivolumab (3 mg/kg, Q2W, IV; n = 29). Median follow-up is 31.2 months (range, 28.3-33+ months) and 17.5 months (range, 8.3- 25.9+ months), respectively. Responses were assessed by irRC. Twenty subjects had sufficient tissue for PD-L1 testing with the 22C3 IHC assay (CLIA) and 10 subjects had sufficient tissue for TMB evaluation by whole exome sequencing (WES) and pretreatment GEP by NanoString.

      4c3880bb027f159e801041b1021e88e8 Result

      In 26 subjects evaluable for response, the ORR was 41% (11 PRs). Another 12 subjects (46%) had SD as best response. As investigators were asked to preferentially enroll PD-L1–negative patients, PD-L1 expression was <1% in 12 of 20 PD-L1–evaluable subjects with 4 achieving a PR. Ten subjects had sufficient tissue for TMB and GEP, including 6 PRs. Five of the 8 who tested low to intermediate for TMB (<243 mut) had a PR as did 2 of 6 GEP-negative subjects. In addition, 5 of 8 subjects with liver metastasis had a PR. The mPFS and mOS of the 5 NSCLC subjects (4/4 tested PD-L1 <1%) treated with pegilodecakin + pembrolizumab was 10.9 and 32.2 months, respectively. The mPFS and mOS for the pegilodecakin + nivolumab cohort (8/16 tested PD-L1 <1%) has not been reached.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Pegilodecakin, when added to anti-PD-1 therapy in advanced NSCLC patients, was associated with response rates and durability of benefit greater than has been seen with anti-PD-1 alone. Responses were seen in settings in which anti-PD-1 therapy has demonstrated limited benefit, such as absent PD-L1 expression, low TMB, and/or the presence of liver metastasis. These preliminary findings support further studies of pegilodecakin with anti-PD-1 therapies.

      6f8b794f3246b0c1e1780bb4d4d5dc53

      • Abstract

      Background

      Prospective data on immunotherapy for NSCLC with oncogenic driver mutations are limited. We recently reported first results from the global IMMUNOTARGET registry (Mazières, ASCO 2018). Here, we present new data for PD-L1 and mutation subgroups.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      In 2017, we started an international retrospective registry study ("IMMUNOTARGET") for patients with advanced NSCLC, known driver mutations (KRAS, EGFR, ALK, ROS1, BRAF, HER2, MET and RET) and PD-L1 immune checkpoint inhibitor therapy. The registry is approved by University of Toulouse and Swissethics, and funded by University of Toulouse and Cantonal Hospital of Lucerne. Anonymized real-world data submitted to the coordinating center include: patient and tumor characteristics, mutation test methods and results, systemic therapy lines, immune related adverse events, best response by RECIST, survival, and tumor PD-L1 expression (optional). Statistical calculations including best response, median PFS and OS are done at University of Toulouse.

      4c3880bb027f159e801041b1021e88e8 Result

      In April 2018, the registry included 551 pts from Europe, USA, Israel and Australia. Patients were 50% male/female, 28% current smokers, median age 60 years (range 28-83), 85% had PS0/1. Most (73%) tumors were stage IV at diagnosis, almost all (96%) were adenocarcinomas. Molecular classification by dominant driver mutation: KRAS=271 (49%), EGFR=125 (23%), BRAF=43 (8%), MET=36 (7%), HER2=29 (5%), ALK=23 (4%), RET=16 (3%), ROS1=7 (1%), 1 (0.2%) not classified (ALK+RET+MET). Most pts received nivolumab (466) or pembrolizumab (48) and were treated with immunotherapy in second or third line (67%). The median number of cycles was 5 (range 1-68). Fifty (11%) pts had grade 3-5 toxicity. Median OS from start of immunotherapy was 13.3 months, median PFS was 2.8 months. Best response was PR/CR in: KRAS=26%, BRAF=24%, ROS1=17%, MET=16%, EGFR=12%, HER2=7%, RET=6%, ALK=0%. Percentage of PD-L1 positive cells was available for 177 pts: 0%=71 (40%), 1-49%=46 (26%), 50-100%=60 (34%). Median % of positive cells was highest for ROS1 (90%), BRAF (50%), MET (30%) and RET (26%) mutant tumors. PD-L1 positivity was predictive for improved PFS in KRAS and EGFR mutant tumors. PD-L1 status was known in 18 tumors with ALK, ROS1 or RET rearrangements: 5 had 0%, 4 had 1-49% and 9 had 50%-100%. No tumor remissions were observed in this subgroup. The registry remains open, updated results will be presented at the conference.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Although response rates were lower than in KRAS mutant NSCLC, individual tumors with other driver mutations responded to immunotherapy. PD-L1 expression may not accurately predict clinical benefit from immunotherapy in some molecular subgroups, better markers are needed.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA04.04 - Discussant - MA 04.01, MA 04.02, MA 04.03

      13:45 - 14:00  |  Presenting Author(s): Jose Pacheco

      • Abstract

      Abstract not provided

    • +

      MA04.05 - Outcomes in NSCLC Patients Treated with First-Line Pembrolizumab and a PD-L1 TPS of 50-74% vs 75-100% or 50-89% vs 90-100%

      14:00 - 14:05  |  Presenting Author(s): Mark M. Awad  |  Author(s): Elizabeth Jimenez Alguilar, Justin F Gainor, Sasha Kravets, Sara Khosrowjerdi, Christine A Lydon, Anika Adeni, Safiya Subegdjo, Hira Rizvi, Matthew D. Hellmann

      • Abstract

      Background

      Among patients with NSCLC and a PD-L1 tumor proportion score (TPS) ≥50%, the response rate to the PD-1 inhibitor pembrolizumab is ~45%. Whether certain subsets of patients with a PD-L1 TPS ≥50% are more likely to benefit from treatment with a PD-1 inhibitor is currently unknown. We compared outcomes among NSCLC patients treated with first-line pembrolizumab and different PD-L1 TPS groupings: 50-74% vs 75-100% or 50-89% vs 90-100%.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      We retrospectively analyzed patients who received commercial pembrolizumab as first-line treatment for NSCLC with a PD-L1 TPS of ≥50% from the Dana-Farber Cancer Institute, Memorial Sloan Kettering Cancer Center, and Massachusetts General Hospital. Clinicopathologic characteristics and clinical outcomes were compared among patients with a PD-L1 TPS of 50-74% vs 75-100% or 50-89% vs 90-100%. Event-time distributions were estimated using Kaplan-Meier and compared with the log-rank test.

      4c3880bb027f159e801041b1021e88e8 Result

      172 patients were identified for inclusion in this study. In the entire cohort, the overall response rate (ORR) to pembrolizumab was 33.9%, median progression-free survival (mPFS) was 4.8 months, and median overall survival (mOS) was 20.6 months. Compared to patients with TPS 50-74% (N=68, 39.5%), patients with TPS 75-100% (N=104, 60.5%) had a significantly higher ORR (45.2% vs 20.6%, P=0.001), a significantly longer mPFS (5.3 vs 2.5 mo, HR=0.61 [95% CI: 0.41-0.90], P=0.008), and a trend towards improved mOS (33.6 vs 20.6 mo, HR=0.60 [95% CI: 0.34-1.04], P=0.056). Compared to patients with TPS 50-89% (N=99, 57.6%), patients with TPS 90-100% (N=73, 42.4%) had a significantly higher ORR (50.7% vs 24.2%, P<0.001), a significantly longer mPFS (6.4 vs 2.8 mo, HR=0.52 [95% CI: 0.36-0.76], P<0.001), and a significantly longer mOS (33.6 vs 18.0 mo, HR=0.46 [95% CI: 0.27-0.79], P=0.008). There were no significant differences in smoking history, histology, sex, and age between patients in each TPS cutoff group.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Among NSCLCs with a PD-L1 TPS ≥50% treated with first-line pembrolizumab, higher PD-L1 TPS levels above 75% and 90% are associated with improved clinical outcomes compared to NSCLCs with lower PD-L1 levels.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA04.06 - PD-1 Blockade Promotes Hyperprogressive Disease in NSCLC Through Macrophages Activation via Antibody-Fc/FcR Interaction

      14:05 - 14:10  |  Presenting Author(s): Gabriella Sozzi  |  Author(s): Michele Sommariva, Massimo Moro, Claudia Proto, Diego Signorelli, Monica Ganzinelli, Sabina Sangaletti, Mattia Boeri, Giuseppe Lo Russo, Simona Ferro, Elena Tassi, Veronica Huber, Lucia Sfondrini, Massimo Milione, Claudio Tripodo, Mario Colombo, Andrea Anichini, Andrea Balsari, Licia Rivoltini, Marina Chiara Garassino

      • Abstract

      Background

      In a subset of patients, named hyperprogressors (HPs), immunotherapy seems to paradoxically boost tumor growth. However, neither pathological and clinical features nor the underlying biological mechanism have been identified. We dissected the role of tumor-myeloid cells crosstalk as possible players.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      HPs were defined on the basis of clinical and radiological features. Baseline histological samples from patients treated with immune checkpoints inhibitors (ICI) were evaluated by immunohistochemistry for myeloid and lymphoid markers. We tested the effect on tumor growth of murine and human ICI in T-cell deficient mice injected with human lung cancer cell lines and PDXs bearing different genotypes (EGFR+, KRAS+, STK11+ and wt). Innate immune microenvironment was evaluated by FACS analysis and immunohistochemistry. In vitro studies of ICI binding functional modulation were performed in human myeloid cells from patients and healthy donors.

      4c3880bb027f159e801041b1021e88e8 Result

      In a clinical series of 187 patients treated with ICI, hyperprogression was observed in 40 (26.3%) cases. All available HP pre-treatment tissue samples (11 cases) showed CD163+CD33+PD-L1+Arginase-A1+ clustered epithelioid macrophages infiltrating the tumor foci also expressing FcRs including CD32b. No differences in T cell compartment were observed. Murine and human PD1 blocking mAbs induced a boost of tumor growth in H460 xenografs in imunocompromised mice. A similar effect was observed in EGFR+ but not in KRAS+ and wt PDXs treated with human anti-PD1. Notably, no hyperprogression was observed after treatment with murine and human anti PD-1 F(ab)2. Hyperprogressive tumors were enriched in arginase+ myeloid-macrophage cells and fibrotic features. ICI bind in vitro to human macrophages and monocytes via Fc/FcR interactions, likely involving CD32b (FcgRIIb) and triggering functional polarization.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Our results provide evidence that FcR triggering on macrophages by ICI delivers a signaling cascade promoting a functional reprogramming of these cells toward a more aggressive pro-tumorigenic behavior eventually inducing hyperprogression in a subset of patients with distinctive immune and genetic profile. A validation prospective study in ongoing.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA04.07 - MicroRNA-Based Liquid Biopsy Combines with PD-L1 Tumor Expression to Predict Response to Immunotherapy in Advance NSCLC Patients

      14:10 - 14:15  |  Presenting Author(s): Mattia Boeri  |  Author(s): Massimo Milione, Diego Signorelli, Claudia Proto, Giuseppe Lo Russo, Carlotta Galeone, Giovanni Centonze, Ugo Pastorino, Marina Chiara Garassino, Gabriella Sozzi

      • Abstract

      Background

      The advent of the new immune checkpoint inhibitors (ICIs) targeting the PD-1/L1 axis drastically improves survival of advance non-small-cell lung cancer (NSCLC) patients. However, only a limited subset of patients actually benefits of ICIs treatment and PD-L1 as predictive biomarker has a limited efficacy. We have previously identified a plasma microRNA-signature classifier (MSC) reflecting a circulating tumor-host interaction with diagnostic and prognostic value in low-dose computed tomography (LDCT) lung cancer screening trials.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      The tumor immune contexture of 40 LDCT-screening detected lung tumors was characterized by the “cell-type identification by estimating relative subsets of RNA transcripts” (CIBERSORT) software. In a consecutive series of 84 advanced lung cancer patients treated with ICIs, both plasma and tissue samples were collected and prospectively analyzed. Both 2-years progression free (PFS) and overall survival (OS) in strata of plasma MSC risk level alone or combined with tumor PD-L1 expression were evaluated in univariate and multivariate analysis by log-rank test and Cox proportional hazards models.

      4c3880bb027f159e801041b1021e88e8 Result

      A pro-tumorigenic immune contexture was identified in tumors of MSC high risk patients. Lower levels of cytotoxic CD8+ and CD4+ T cells and increased levels of Tregs, γδ T Cells, M2 macrophages characterized these tumors. In addition, genes differentially expressed according to MSC risk level (high vs. intermediate and low) were associated with 5-years OS in the screening series (p-values=0.02), as well as in additional 1000 cases from The Cancer Genome Atlas database (p-values<0.01). In the 84 advanced NSCLC patients treated with ICIs, the PFS hazard ratio ranged from 0.44 (95%CI: 0.25-0.75) of PD-L1 (adjusted p-value=0.005) and 0.38 (95%CI:0.2-0.73) of MSC (adjusted p-value=0.004) alone, to 0.25 (95%CI: 0.14-0.45) if combined (adjusted p-value<0.0001). In the subgroup of 45 patients with both plasma and tumor tissue available, the combination of MSC and PD-L1 stratified patients in three groups with 2-years PFS ranging from 25%to 10% and 0% (p-value=0.01) according to the presence of 2, 1 or 0 favorable markers, respectively. Similar results were obtained when considering OS, where the median survival time for patients with no favorable markers was 5.6 months (p-value<0.0001).

      8eea62084ca7e541d918e823422bd82e Conclusion

      Overall, these findings suggest that a circulating microRNA-based risk level, reflecting an altered tumor immune contexture, could implement PD-L1 tumor tissue expression as predictive biomarkers of response to immunotherapy.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA04.08 - Discussant - MA 04.05, MA 04.06, MA 04.07

      14:15 - 14:30  |  Presenting Author(s): Patrick M Forde

      • Abstract

      Abstract not provided

    • +

      MA04.09 - Neoadjuvant Atezolizumab in Resectable Non-Small Cell Lung Cancer (NSCLC): Updated Results from a Multicenter Study (LCMC3)

      14:30 - 14:35  |  Presenting Author(s): Valerie W Rusch  |  Author(s): Jamie E Chaft, Bruce E Johnson, Ignacio I. Wistuba, Mark G Kris, Jay M Lee, Paul A. Bunn, Jr., David J Kwiatkowski, Karen L. Reckamp, David J. Finley, Eric B. Haura, Saiama N. Waqar, Robert C. Doebele, Edward B Garon, Justin Blasberg, Alan Nicholas, Katja Schulze, See Phan, Mayank Gandhi, David P Carbone

      • Abstract

      Background

      Cisplatin-based chemotherapy, before or after surgery, provides only a 5% benefit in 5yr. OS in resectable NSCLC. A 20 patient study (NEJM April 2018) showed that preoperative immune checkpoint inhibitor therapy yielded a clinically meaningful major pathologic response rate (MPR ≤10% residual viable tumor cells) and did not delay or complicate surgery. This large multicenter trial measures MPR and biomarkers of benefit using neoadjuvant atezolizumab (atezo) [NCT02927301].

      a9ded1e5ce5d75814730bb4caaf49419 Method

      We planned 2 cycles of atezo (1200mg, days 1, 22) in patients with stages IB -selected IIIB resectable NSCLC prior to surgical resection (day 40 +/- 10). Chest CT, PET were planned pre-atezo and presurgery to assess response. Primary tumor +/- node biopsies and blood samples were obtained before atezo and presurgery for biomarker studies. The primary endpoint was MPR. Secondary endpoints included safety, response by PD-L1, OS, and DFS.

      4c3880bb027f159e801041b1021e88e8 Result

      For this updated efficacy and safety analysis (Feb’18 datacut), we report first 54 of 180 planned pts: 29 males, median age 65 yr, all ECOG 0-1; 17 current, 33 former smokers; 35 non-squamous NSCLC; clinical stages Ib/IIa/IIb/IIIa/IIIb = 5/11/13/20/5. Two pts received one dose of atezo due to treatment related AE (Gr 1 pyrexia, Gr 2 dyspnea) but underwent uncomplicated resection with MPR assessment. There was 1 unrelated Gr 5 AE (sudden cardiac death post surgical resection), 16 Gr 3-4 AEs (3 treatment related). Surgery was delayed in 1 pt due to Gr3 pneumonitis. By RECIST, 3 pts had PR, and 49 had SD. 50 pts underwent surgery and 47 pts had MPR assessment: 2 pts discontinued study preop due to radiographic PD and 2 discontinued due to other reasons; 3 pts had unresectable disease. MPR rate was 10/50 (20%, 95% CI 10-34%) including 3 pts who had pCR (no viable tumor cells) in the primary tumor. Excluding 5 pts who had known driver mutations (4 EGFR+, 1 ALK+), MPR rate was 10/45 (22%, 95% CI 11-37%). PD-L1 status was evaluable in 44/54 pts; 8/10 pts with MPR had PD-L1+ status and 2 had unknown PD-L1 status; 8/28 PDL-1 (+) patients had MPR (29%).

      8eea62084ca7e541d918e823422bd82e Conclusion

      In a multicenter study, neoadjuvant atezo was well tolerated. MPR rate is encouraging. Clinical and pathological responses are often discordant. Correlative analyses on pre- and post atezo tissues are ongoing. Preliminary correlative analyses in blood samples are included in a separate abstract.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA04.10 - Comprehensive Peripheral Blood Immunophenotyping and T-Cell Clonal Analysis During Neoadjuvant Immunotherapy with Atezolizumab in NSCLC

      14:35 - 14:40  |  Presenting Author(s): Filiz Oezkan  |  Author(s): Kai He, Dwight Hall Owen, Maciej Pietrzak, Rhonda Kitzler, Rebecca Pearson, Alan Nicholas, Paul A. Bunn, Jr., Mark G Kris, David J. Kwiatkowski, Bruce E Johnson, Fred R. Hirsch, Ignacio I. Wistuba, Valerie W Rusch, Jay M. Lee, Mayank Gandhi, Katja Schulze, David S. Shames, Gerard Lozanski, David P Carbone

      • Abstract

      Background

      Immune-checkpoint blockade targeting PD-L1/PD-1 to activate anti-tumor immunity is associated with improved response rates and survival compared to chemotherapy in selected metastatic NSCLC patients. Evaluation of the pre-therapeutic immune profile and its treatment-related evolution associated with clinical benefit will guide future immunotherapy development and support clinical decision-making. Here, we present an analysis of peripheral blood (PB) immunophenotyping and T-cell-receptor (TCR) clonality before and after immunotherapy from an ongoing 180-patient phase II study of atezolizumab as neoadjuvant therapy with stage IB-IIIB resectable NSCLC (NCT02927301; LCMC3).

      a9ded1e5ce5d75814730bb4caaf49419 Method

      As of February 5th datacut, the first 54 enrolled and dosed patients are presented. The biomarker evaluable population (BEP) further subset to patients with paired PB samples analyzed within 72 hours after collection and a major pathological response (MPR) assessment. Comprehensive immune cell phenotyping (10-color flow cytometry, IMMUNOME) and TCR-Vß-analysis by flow cytometry were performed. Immunoprofile analyses were correlated with atezolizumab treatment, pathological response and PD-L1 expression.

      4c3880bb027f159e801041b1021e88e8 Result

      In this ongoing analysis, BEP included 31 patients. 5 patients (16%, 95% CI (5%, 34%)) had a MPR; all of which stained positive for PD-L1 by IHC using 22C3 (TPS≥1%) and SP142 (PD-L1 expression on ≥1% tumor cells (TC) and/or tumor infiltrating immune cells (IC)) at baseline. We observed significant increases in natural killer (NK) cells (p=0.005) and CD8+ T-cells (p=0.031) and a Th1-response related dendritic cell (DC) subpopulation (p=0.031) and significant decreases in B-cells (p=0.015) after treatment.

      Patients who achieved MPR show lower baseline levels of degranulated CD8+ T-cells (p=0.015), late-activated NK-cells (p=0.043), memory CD4+ (p=0.048) and memory CD8+ T-cells (p=0.032); changes in PB NK-cells (p=0.041), a decrease in M-MDSCs and a Th-2 and Th-17-response related DC subpopulation (p=0.043) in response to treatment were noted in patients with MPR versus non-MPR.

      Among the 16 patients with TC/IC 1/2/3 (> 1% PD-L1 expression) the following significant differences were observed compared to TC0/IC0 (7 patients): higher levels of late-activated CD4+ T-cells (p=0.025) and mid-activated CD8+ T-cells (p=0.044) at baseline, decrease of senescent T-cells (p=0.041), monocytic myeloid-suppressor cell subpopulations (M-MDSCs) and an increase in a Th1-response related DC subpopulation (p=0.026) after treatment.

      TCR clonality analysis showed expansions in Vß-subtypes after atezolizumab treatment.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Immunophenotyping and TCR-Vß-repertoire analysis in peripheral blood samples from NSCLC patients treated with neoadjuvant atezolizumab show differences in immune cell subsets in baseline samples and changes after treatment.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA04.11 - Neoantigen Targeting and T Cell Reshaping in Resectable NSCLC Patients Treated with Neoadjuvant PD-1 Blockade

      14:40 - 14:45  |  Presenting Author(s): Kellie Nicole Smith  |  Author(s): Margueritta El Asmar, Jiajia Zhang, Justina X Caushi, Zhicheng Ji, Valsamo Anagnostou, Tricia R Cottrell, Hok Yee Chan, Prerna Suri, Haidan Guo, Kristen A. Marrone, Jarushka Naidoo, Taha Merghoub, Jamie E Chaft, Matthew D. Hellmann, Janis M Taube, Julie R. Brahmer, Patrick M Forde, Victor Velculescu, Drew M Pardoll, Hongkai Ji

      • Abstract

      Background

      PD-1 blockade is now standard treatment for advanced non-small cell lung cancer (NSCLC) and has recently shown impressive efficacy in promoting major pathologic response (MPR) and delaying relapse in the neoadjuvant setting. The role of tumor mutational burden, and specifically T cells targeting neoantigens derived from these mutations, in facilitating tumor clearance has been demonstrated in advanced NSCLC. However, it is unknown how neoadjuvant PD-1 blockade impacts the frequency and function of tumor specific T cells and their ability to promote major pathologic response, or how these factors may synergize to prevent or delay relapse after surgical resection.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Whole exome sequencing and neoantigen prediction was performed on pre-treatment tumor biopsies and matched normal tissue from 11 patients with resectable NSCLC treated with neoadjuvant nivolumab as part of a clinical trial (NCT02259621). T cell recognition of peptides representing candidate neoantigens was evaluated using the MANAFEST assay, which identifies T cell receptor clonotypes corresponding to antigen specificities. T cell receptor sequencing was additionally performed on serial peripheral blood T cells, pre-treatment tumor biopsies, and resected post-treatment tissues. A bioinformatic platform was developed to evaluate the dynamics of intratumoral T cell clonotypes, and more specifically neoantigen-specific clonotypes detected before, during, and after treatment and during long-term follow-up.

      4c3880bb027f159e801041b1021e88e8 Result

      High-magnitude, polyclonal neoantigen-specific T cell responses were detected in the peripheral blood and persisted for many months after surgical resection and cessation of treatment. Binding to and stability with cognate HLA I molecules was validated for reactive neoantigens. Significant treatment-induced systemic perturbations in the tumor-specific T cell repertoire and an influx of peripheral T cell clonotypes into tumor tissue and lymph nodes was observed in patients regardless of pathologic response, whereas peripheral clonotypic reshaping of the anti-tumor repertoire and intratumoral T cell clonality were associated with MPR status.

      8eea62084ca7e541d918e823422bd82e Conclusion

      We show significant and systemic alterations in the peripheral anti-tumor T cell repertoire in NSCLC patients treated with neoadjuvant anti-PD-1 regardless of MPR status. Notwithstanding, the impaired restructuring of the anti-tumor T cell repertoire in patients without MPR highlights a potential immunological deficiency to overcome in future therapeutic approaches aiming to increase the MPR rate in NSCLC patients treated with neoadjuvant PD-1 blockade.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA04.12 - Discussant - MA 04.09, MA 04.10, MA 04.11

      14:45 - 15:00  |  Presenting Author(s): Alex Adjei

      • Abstract

      Abstract not provided

  • +

    MA05 - Improving Outcomes in Locoregional NSCLC II

    • Type: Mini Oral Abstract Session
    • Track: Treatment of Locoregional Disease - NSCLC
    • Moderators:
    • +

      MA05.01 - E6508: Phase II Study of Immunotherapy with Tecemotide and Bevacizumab after Chemoradiation in Unresectable Stage III NS-NSCLC

      13:30 - 13:35  |  Presenting Author(s): Jyoti Patel  |  Author(s): Ju-Whei Lee, Henry Wagner Jr, David P Carbone, Anil Shanker, Leora Horn, Melissa L. Johnson, David E Gerber, Jane Jijun Liu, Millie S Das, Mohammad Ali Al-Nsour, Christopher S R Dakhil, Suresh S. Ramalingam, Joan Schiller

      • Abstract

      Background

      Chemoradiation (CRT) is standard of care for unresectable stage III NSCLC. Tecemotide is a MUC1 antigen-specific cancer immunotherapy. Bevacizumab is considered to have a significant role in immune modulation. Immunotherapy in combination with VEGF blockade was tested in this phase II trial combining tecemotide and bevacizumab in patients with stage III NS- NSCLC.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Subjects with stage III NS- NSCLC suitable for definitive CRT received carboplatin(C) AUC 2 + paclitaxel(P) 45 mg/m2 weekly + 66 Gy/33fx/6.5wk and consolidation C AUC 6 + P 225 mg/m2 q21 days x 2. Patients with CR/PR/SD were then registered onto Step 2 (S2). S2 was 6 weekly tecemotide injections followed by q6 weekly injections and bevacizumab 15 mg/kg q3 weeks for up to 34 doses. The primary endpoint was safety of tecemotide and bevacizumab after CRT and consolidation. The proportion of circulating dendritic cells and their expression of CD40, HLA-DR and CD123 (IL-3R) were analyzed by flow cytometry at various time points.

      4c3880bb027f159e801041b1021e88e8 Result

      70 patients were enrolled from Dec 2010 to Oct 2014; 68 started therapy, and 39 completed CRT and consolidation therapy. Reasons for discontinuation included progression (11) and toxicity (10). 33 patients were registered to S2. The median number of S2 cycles was 12 (range 2-34). S2 toxicity: gr 3 N=9 (6 hypertension), gr 4 N=1, gr 5 N=1. Among the treated and eligible patients (n=31), from study entry, the median PFS was 14.3 (95% CI 11.0-22.2), OS was 40.1 (95% CI 21.7-NA) months. A correlative trend of increased expression of CD40 and HLA-DR on CD11c+ cells was observed at cycle 7 (week 21) of S2.

      e6508.patel.png

      8eea62084ca7e541d918e823422bd82e Conclusion

      This cooperative group trial met its endpoint, demonstrating tolerability of tecemotide and bevacizumab after CRT and consolidation in NS-NSCLC pts. In this select group of patients, therapy with tecemotide and bevacizumab was associated with encouraging PFS and OS.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA05.02 - PACIFIC Subgroup Analysis: Pneumonitis in Stage III, Unresectable NSCLC Patients Treated with Durvalumab vs. Placebo After CRT

      13:35 - 13:40  |  Presenting Author(s): Johan F. Vansteenkiste  |  Author(s): Jarushka Naidoo, Corinne Faivre-Finn, Mustafa Özgüroğlu, Augusto Villegas, Davey Daniel, Shuji Murakami, Rina Hui, Ki Hyeong Lee, Byoung Chul Cho, Kaoru Kubota, Lynne Poole, Catherine Wadsworth, Phillip A. Dennis, Scott J Antonia

      • Abstract

      Background

      In the Phase 3 PACIFIC study of durvalumab versus placebo in patients with stage III, unresectable non-small cell lung cancer (NSCLC) after concurrent chemoradiotherapy (cCRT), on-treatment pneumonitis or radiation pneumonitis (‘pneumonitis’) occurred in both arms with similar rates of grade 3/4 pneumonitis (durvalumab, 3.4%; placebo, 2.6%). We performed exploratory analyses to further characterize time to onset and duration of pneumonitis and examine its relationship with underlying risk factors, including patient characteristics and prior CRT.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      PACIFIC (NCT02125461) was a randomized, double-blind study of patients with WHO PS 0/1 without progression after ≥2 cycles of platinum-based cCRT. Patients were stratified by age, sex, and smoking history and randomized (2:1) 1–42 days after completing cCRT to durvalumab 10 mg/kg IV Q2W or placebo up to 12 months. Potential associations between the presence of the AE pneumonitis (investigator assessed with review/adjudication by study sponsor) and baseline characteristics or patient disposition were investigated.

      4c3880bb027f159e801041b1021e88e8 Result

      As of Feb 13, 2017, 709 patients had received treatment; 33.6% on durvalumab and 24.9% on placebo had any-grade pneumonitis. Treatment exposure was similar in patients with or without pneumonitis across both arms. Median time to onset of pneumonitis from treatment start was the same for both durvalumab and placebo, 55.0 days (73.0 and 76.5 days from RT completion). Pneumonitis was self-limited, with median durations of 64.0 and 57.0 days, respectively. Patients with pneumonitis were more likely to be Asian (47.9% vs 17.6%) or have EGFR mutations (11.0% vs 3.8%); however, the proportions of patients with pneumonitis and these risk factors were numerically lower with durvalumab than with placebo (Asian: 44.4% [71/160] vs 57.6% [34/59]; EGFRm: 10.6% [17/160] vs 11.9% [7/59]), suggesting no apparent interaction with treatment. There were no apparent associations of pneumonitis with baseline respiratory disorders, prior RT dose, or prior cisplatin or carboplatin use. Previous induction CT was more commonly associated with the absence of pneumonitis in both treatment arms (durvalumab: 30.1% vs 17.5%; placebo: 31.5% vs 20.3%). The presence of pneumonitis was associated with greater discontinuation due to AEs (durvalumab: 25.6% vs 10.2%; placebo: 18.6% vs 6.8%) regardless of treatment.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Rates of pneumonitis were higher in Asian patients and those with EGFRm, as previously reported. Durvalumab did not increase pneumonitis in patients with these risk factors. There were no differences in treatment exposure in patients based on the presence/absence of pneumonitis. Multivariate analyses may further assist in the discernment of etiologic risks.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA05.03 - Immune Microenvironment and its Association with Adjuvant Chemotherapy Benefit in Locoregionally Advanced Lung Adenocarcinoma

      13:40 - 13:45  |  Presenting Author(s): Raj Ghanshyam Vaghjiani  |  Author(s): Takashi Eguchi, Navin Chintala, Xiaoyu Li, Rania G Aly, Katsura Emoto, Kay See Tan, David R. Jones, Prasad S. Adusumilli

      • Abstract

      Background

      The impact of the tumor immune microenvironment on the effectiveness of platinum-based adjuvant chemotherapy (ACT) in locoregionally advanced (stage II-III) lung adenocarcinoma (ADC) is unknown. We performed an analysis of the cellular components of the tumoral and tumor-associated stromal immune environment in stage II-III lung ADC and examined their association with ACT benefit.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Tissue microarrays (6 tumor and 3 stromal cores from each tumor) were constructed using resected tissue from patients with pT2-T4N1 lung ADC (n=500, 2000-2012) who did (n=225) and did not (n=214) receive ACT. Multiplex immunofluorescence was used to determine the quantity, localization, and colocalization of 21 types of immune cells and markers (including PD-1, PD-L1, CD3, CD20, CD68, CD163, MPO, and PanCK). The association between immune cell infiltration and recurrence free probability (RFP) was compared using Kaplan-Meier methods, and benefit from ACT by unsupervised hierarchical cluster modeling.

      4c3880bb027f159e801041b1021e88e8 Result

      Overall, increased tumoral infiltration of CD20+ B-cells and CD3+ and CD4+ T-cells was associated with an improvement in 5-yr RFP (CD20+ low vs high: 37% vs 49%, p=.03; CD3+: 39% vs 48%, p=.003; and CD4+: 39% vs 47%, p=.02, respectively) whereas increased stromal MPO+ neutrophil infiltration was associated with a worse 5-yr RFP (low vs high: 50% vs 38%, p=.003). Among patients who received ACT, cluster modeling revealed 5 risk groups (Groups A-E; Figure) with immune signatures including tumoral B-cells and CD163+PD-1+ macrophages as well as stromal CD57+ NK-cells and CD163+PD-L1+ macrophages that provided a progressive stratification of RFP following adjuvant treatment.

      vaghjiani.jpg

      8eea62084ca7e541d918e823422bd82e Conclusion

      Immune infiltration analysis can predict benefit from ACT and thereby provide a rationale to select patients for either chemotherapy, immunotherapy, or combination therapy following surgical resection.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA05.04 - Discussant - MA 05.01, MA 05.02, MA 05.03

      13:45 - 14:00  |  Presenting Author(s): Scott N. Gettinger

      • Abstract

      Abstract not provided

    • +

      MA05.05 - Photon-Based Cardiac Sparing Via Volumetric Modulated Arc Therapy in Thoracic Radiation Therapy for Locally Advanced Non-Small Cell Lung Cancer

      14:00 - 14:05  |  Presenting Author(s): Matthew J Ferris  |  Author(s): Katherine Sykes, Oluwatosin A Kayode, Jonathan Wolf, Robert H Press, Jeffrey M Switchenko, Walter John Curran, Jr., Kristin A Higgins

      • Abstract

      Background

      Increasing radiation dose to the heart is associated with worse survival in stage III non-small cell lung cancer. Techniques to reduce the dose to the heart, including proton beam therapy (PBT), are being evaluated in ongoing clinical trials. However, advanced technologies such as PBT are not readily accessible for most patients. We therefore sought to evaluate the efficacy of volumetric modulated arc therapy (VMAT), a readily available technology in the United States, to spare cardiac substructures and determine how a cardiac optimization treatment planning algorithm influences dose distribution to other thoracic organs at risk (OARs).

      a9ded1e5ce5d75814730bb4caaf49419 Method

      We selected stage III non-small cell lung cancer patients who were treated at our institution with VMAT to 60 Gy in 2 Gy fractions. Cardiac substructures were retrospectively contoured, and included: valves, atrioventricular node (AVN), coronary arteries (CA), chambers, and great vessels. New radiation treatment plans were created to spare these structures while preserving planning target volume (PTV) coverage and maintaining standard dose constraints to OARs. Dosimetry variables—maximum dose (Dmax), mean dose (Dmean), and common clinically relevant dose-volume relationships—for the new cardiac-sparing radiation treatment plans were compared via paired t-test to the original radiation treatment plans.

      4c3880bb027f159e801041b1021e88e8 Result

      Twenty-six patients, treated from July 2013 to September 2017, were included. Statistically significant improvements were demonstrated for all cardiac structures for the new cardiac-sparing plans compared to the original plans, while maintaining appropriate lung, esophagus, and spinal cord constraints, and PTV coverage goals, as demonstrated in Table 1 (significant P-values in bold).

      Table 1

      Dosimetry variable

      Cardiac-sparing plan (mean)

      Original plan (mean)

      P-value*

      Cardiac parameters

      Heart Dmax

      64.9 Gy

      63.6 Gy

      0.928

      Heart Dmean

      12.3

      16.1

      < 0.001

      Heart V5Gy

      55.4

      64.1

      0.003

      Heart V30Gy

      12.5

      18.7

      < 0.001

      Heart V40Gy

      7.9

      11.5

      < 0.001

      Heart V45Gy

      6.5

      11.5

      < 0.001

      Heart V60Gy

      2.7

      3.4

      0.001

      Aortic valve Dmax

      22.9

      31.7

      < 0.001

      Aortic valve Dmean

      11.4

      31.7

      < 0.001

      Mitral valve Dmax

      24.6

      29.4

      0.002

      Mitral valve Dmean

      11.2

      16.7

      < 0.001

      Pulmonic valve Dmax

      26.8

      35.4

      < 0.001

      Pulmonic valve Dmean

      14.1

      25.1

      < 0.001

      Tricuspid valve Dmax

      9.7

      16.6

      < 0.001

      Tricuspid valve Dmean

      5.6

      10.3

      < 0.001

      AVN Dmax

      13.4

      20.4

      < 0.001

      AVN Dmean

      8.1

      14.0

      < 0.001

      Left main CA Dmax

      26.4

      38.8

      < 0.001

      Left main CA Dmean

      16.4

      30.2

      < 0.001

      Left anterior descending CA Dmax

      27.4

      34.8

      < 0.001

      Left anterior descending CA Dmean

      14.4

      22.6

      < 0.001

      Left circumflex CA Dmean

      32.6

      36.8

      0.001

      Left circumflex CA Dmean

      19.3

      26.9

      < 0.001

      Right CA Dmax

      18.1

      26.1

      < 0.001

      Right CA Dmean

      9.4

      15.7

      < 0.001

      Left atrium Dmax

      51.8

      54.8

      0.091

      Left atrium Dmean

      17.5

      21.0

      < 0.001

      Left ventricle Dmax

      35.7

      40.1

      < 0.001

      Left ventricle Dmean

      8.3

      11.3

      < 0.001

      Right atrium Dmax

      31.4

      36.1

      0.004

      Right atrium Dmean

      11.1

      13.9

      < 0.001

      Right ventricle Dmax

      23.7

      33.2

      < 0.001

      Right ventricle Dmean

      6.9

      12.2

      < 0.001

      Aorta Dmax

      50.8

      55.3

      0.001

      Aorta Dmean

      20.4

      27.9

      < 0.001

      Pulmonary artery Dmax

      65.2

      65.1

      0.895

      Pulmonary artery Dmean

      32.3

      37.9

      < 0.001

      Superior vena cava Dmax

      47.4

      51.7

      0.002

      Superior vena cava Dmean

      29.4

      33.1

      0.006

      Other OAR parameters

      Lungs V5Gy

      56.5

      58.2

      0.121

      Lungs V20

      22.4

      23.3

      0.083

      Lungs Dmean

      13.6

      14.8

      0.012

      Spinal cord Dmax

      28.0

      31.1

      0.013

      Esophagus Dmean

      21.2

      22.1

      0.023

      PTV coverage parameters

      PTV Dmax

      65.5

      67.2

      0.189

      PTV minimum dose

      51.2

      52.7

      0.019

      PTV V100%

      95.2%

      95.4%

      0.195

      8eea62084ca7e541d918e823422bd82e Conclusion

      Dose to the heart and cardiac substructures can be substantially lowered using a cardiac-sparing optimization algorithm with VMAT, without increasing radiation dose other thoracic OARs or compromising PTV coverage. Though time-consuming, delineation of the full complement of cardiac substructures provides an effective means of improving the quality of radiation treatment plans with readily available technologies.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA05.06 - Locally Advanced Lung Cancer Radiotherapy in Deep Inspiration Breath Hold: Dosimetric Benefits from a Prospective Trial

      14:05 - 14:10  |  Presenting Author(s): Mirjana Josipovic  |  Author(s): Marianne C Aznar, Jonas Scherman Rydhög, Jakob Borup Thomsen, Sidsel Marie Skov Damkjaer, Lotte Nygård, Mette Pøhl, Seppo W Langer, Lena Specht, Gitte Fredberg Persson

      • Abstract

      Background

      Radiotherapy for locally advanced non-small cell lung (NSCLC) cancer is often complicated by treatment-related toxicity. A toxicity-reducing technique is deep inspiration breath hold (DIBH), where the lungs inflate and the heart is pushed downwards. DIBH is widely applied in breast radiotherapy, but only sporadically in NSCLC. We initiated the INHALE trial, investigating compliance and benefits of DIBH for NSCLC at a single academic institution.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Patients referred for definitive radiotherapy of locally advanced NSCLC (66Gy/33 fractions) were included from May 2015-Dec 2017. All patients underwent respiratory coaching for voluntary visually guided DIBH and were imaged with PET/CT, 4D-CT and DIBH-CT. Target volumes were defined according to national guidelines. PTV margins were patient- and modality-specific. For all patients, FB and DIBH plans were made with volumetric modulated arc therapy, with equal PTV coverage. The plan with the lowest lung and/or heart dose was chosen for treatment. Normal tissue complication probability for pneumonitis was calculated retrospectively based on a logistic dose response model.

      4c3880bb027f159e801041b1021e88e8 Result

      The treatment intent was maintained in 69 of included 88 patients (2 were downstaged, 12 upstaged, 2 withdrew consent, other causes in 3). 62/69 were DIBH compliant and in 61 patients a FB and a DIBH plan were made (in one patient, 4DCT image quality was not sufficient). In 54/61 patients, the DIBH plan was chosen for treatment. 3/54 patients lost DIBH compliance within the first few fractions.

      All data is presented as median (range), with p<0.001 (Wilcoxon signed rank). Lung volume increased in DIBH by 55% (20-168%). Compared to FB, DIBH reduced mean lung dose from 14.4Gy (1.2-25.3Gy) to 11.8Gy (1.0-20.4Gy), and lung V20 from 23.7% (1.5-47.8%) to 20.8% (1.2-39.7%). Reduced lung dose translated to reduced pneumonitis risk: from 8.6% (2.3-23.3%) to 6.5% (2.2-14.4%). Lung dose constraints were violated in 5/62 patients in FB and 1/62 patients in DIBH.

      Mean heart dose was reduced from 3.6Gy (0.1-25.8Gy) in FB to 2.4Gy (0.1-25.3Gy) in DIBH. DIBH reduced mean heart dose in 44/61 patients. The differences between FB and DIBH varied between – 6.6Gy and 8.9Gy, stressing the influence of tumour location on the potential of reducing heart dose with DIBH.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Benefits of changed anatomy with DIBH were reduced dose to lungs and, for most patients, to the heart. Curative treatment intent could be maintained in more patients. Risk of developing radiation pneumonitis was reduced. Continuous follow up of INHALE patients will reveal how the reduced risk is manifested clinically.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA05.07 - Dose Escalated Chemo-RT to 84 Gy in Stage III NSCLC Appears Excessively Toxic: Results from a Randomized Phase II Trial

      14:10 - 14:15  |  Presenting Author(s): Jan Nyman  |  Author(s): Stefan Bergström, Hedvig Björkestrand, Anna-Maja Svärd, Simon Ekman, Erik Lundin, Erik Holmberg, Mikael Johansson, Signe Friesland, Andreas Hallqvist

      • Abstract

      Background

      Concurrent chemoradiotherapy is the mainstay treatment for NSCLC stage III disease, however, with a rather high probability of locoregional and metastatic recurrence further treatment optimization is warranted. Based on previous one-armed trials with dose escalated radiotherapy, showing feasibility, the Swedish Lung Cancer Study Group aimed to investigate whether dose escalation based on individual normal tissue constraints could improve outcome in this randomized phase II trial.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      NSCLC patients with stage III disease, good performance status (0-1), adequate lung function (FEV1 > 1.0 L and CO diff. > 40%) received three cycles of cisplatin (75 mg/m2 day 1) and vinorelbine (25 mg/m2 day 1 and 8) every third week. The radiotherapy started concurrently with the second cycle, with either 2 Gy daily, 5 days a week, to a total dose of 68 Gy (standard arm A) or escalated therapy (B) based on constraints to the spinal cord, esophagus and lungs up to 84 Gy by adding an extra fraction of 2 Gy per week while keeping the total treatment time constant at seven weeks with the same dose to involved nodes and primary tumor.

      4c3880bb027f159e801041b1021e88e8 Result

      A pre-planned safety analysis revealed excessive toxicity and decreased survival in the escalated arm, and the study was stopped. Thirty-six patients were included during 2011-2013 (56% male, 78% with adenocarcinoma, 64% with PS 0 and 53% with stage IIIB). The median progression-free survival (PFS) and overall survival (OS) were 11 and 17 months in the dose escalated group compared to 28 and 45 months in the standard group. The 1-, 3- and 5-year survival rates were 56%, 33% and 17% in the escalated arm and 72%, 61% and 34% in the standard arm. There were four toxicity-related deaths due to esophageal perforations (one in arm A and three in arm B) and three deaths due to pneumonitis (one in arm A and two in arm B).

      8eea62084ca7e541d918e823422bd82e Conclusion

      Dose-escalated concurrent chemoradiotherapy to 84 Gy to primary tumor and nodal disease is hazardous, with a high risk of excessive toxicity, whereas modern standard dose chemoradiotherapy with proper staging given in the control arm shows a promising outcome with a median survival of 45 months and a 5-year survival of 34%. A possible step forward will be to improve systemic therapy, but future approaches with escalated radiotherapy may include boost techniques to remaining PET positive areas or different escalation schedules to the primary tumor and mediastinal nodes.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA05.08 - Discussant - MA 05.05, MA 05.06, MA 05.07

      14:15 - 14:30  |  Presenting Author(s): Benjamin H Lok

      • Abstract

      Abstract not provided

    • +

      MA05.09 - PFS and Cardiac-Toxicity-Adjusted-PFS As Predictors of OS in Locally Advanced NSCLC Treated with Concurrent Chemoradiation

      14:30 - 14:35  |  Presenting Author(s): Chen Hu  |  Author(s): Mitchell Machtay, James Dignam, Rebecca Paulus, Jeffrey Bradley

      • Abstract

      Background

      Overall survival (OS) is the gold standard for LA-NSCLC with chemoradiation (CCRT), while the complex relationships among RT dosimetry, systemic therapies, cardiopulmonary toxicity, progression (PD) and OS are also of increasing scientific and clinical interest.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      NRG Oncology RTOG 0617 (NCT00533949) was a randomized phase 3 trial comparing standard (SD, 60 Gy) versus high-dose (HD, 74 Gy) CCRT +/- cetuximab from 11/07-06/11. This secondary analysis includes 469 patients (pts) given ≥50 Gy. A PFS event was defined as the first occurrence of local, regional, distant PD or death w/o documented PD. A CTA-PFS event was the first occurrence of grade 2+ treatment-related cardiac toxicity event or a PFS event. Landmark analyses at 6mo and 12mo were used to minimize the immortal time bias. Cox model with PD or CT/PD as a time-dependent covariate was used to evaluate their predictive roles. Median f/u time for surviving pts was 5.1 years.

      4c3880bb027f159e801041b1021e88e8 Result

      As previously reported, pts treated with HD had significantly lower OS rates (HR=1.28, 95%CI: 1.04-1.58, p=0.018) and CTA-PFS rates (HR=1.24, 95%CI: 1.02-1.51, p=0.035), and marginally lower PFS rates (HR=1.21, 95%CI: 0.99-1.47, p=0.06) than pts treated with SD. Median survival time (MST) among pts having PD within 6mo versus not were 13.4mo (95%CI: 10.0-19.0mo) and 30.7mo (95%CI: 28.0-37.0mo) (p<0.001). MST for pts having PD within 12mo versus not were 20.6mo (95%CI: 18.8-25.0mo) and 60mo (95%CI: 47.6-74.5mo)(p<0.001). Results are similar when using CTA-PFS with 6mo or 12mo cutoff (p<0.001). RT dose was no longer significantly associated with OS (p=0.08 or p=0.15) when PD or CT/PD was included in multivariable analysis (p<0.001), suggesting OS differences in HD/SD may be partially captured by PFS or CTA-PFS.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Long-term survival results from RTOG 0617 suggest that PFS (or CTA-PFS) status at 6mo or 12mo predicts long-term OS, and may potentially be considered as a surrogate endpoint of OS in clinical trials. Pts who were progression-free at 12mo had a MST of 5 years. Further validation on external datasets and in the modern era of immunotherapy are needed.

      Funding: This project was supported by grants NCORP (UG1CA189867), NRG Operations (U10CA180868), NRG SDMC (U10CA180822), IROC (U24CA180803), and CTEP from the National Cancer Institute (NCI).

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA05.10 - The Pathologic Response of Locally Advanced NSCLC Treated with Concomitant Chemoradiation to 60 Gy in Image Guided Radiation Therapy (IGRT)

      14:35 - 14:40  |  Presenting Author(s): Sarit Appel  |  Author(s): Jair Bar, Damien Urban, Amir Onn, Marina Perelman, Yaacov Richard Lawrence, Alon Ben-Nun, Ory Haisraely, Zvi Symon, Tatiana Rabin El Ezra, Edith Marom, Sivan Liberman, Efrat Ofek

      • Abstract

      Background

      Neoadjuvant concomitant chemoradiation (NACCRT) was historically limited to 45 Gy. We recently published data on the safety of a higher radiation dose in this setting. Here we evaluate the pathologic response of locally advanced non small cell lung cancer (LANSCLC) treated with 60Gy NACCRT combined with modern IGRT.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Our cohort comprised patients that underwent NACCRT followed by surgery during August 2012-December 2017 at our institution. We retrospectively collected the demographic, stage, histology, and treatment details. Radiation was planned using eclipse system to deliver 2 Gy per fraction to a total of 60 Gy

      Treatment effect was determined from the pathologic specimen in accordance with College of American Pathologists recommendations, based on the modified tumor regression grading: Favorable pathologic responses included major tumor regression (MTR); we also evaluated the average percent of the residual tumor cells seen in the specimen. Statistical analysis was performed to analyze treatment effect on the pathologic response using spearman correlation and Kruskal-Wallis test with SPSS software v.24.

      4c3880bb027f159e801041b1021e88e8 Result

      Our cohort included 70 patients. Mean age was 63 years (range 45–79.7), men n=49 (70%), smoking status: never smokers n=11 (16.2%), past smokers n=10 (14.7%), current smokers n=47 (69.1%). Histology consisted adenocarcinoma n=42 (60%), squamous n=21 (30%) and other n=7 (10). Stage 2 were n=65 (78.3%) and stage 3 n=15 (21.4%). Chemotherapy consisted of platinum-doublet administered to 69 patients (98.5%). A mean radiation dose of 59 Gy (range 46-72 Gy) was delivered with IGRT prior to each fraction. Five patients received lower radiation doses due to toxicity or dose constraints. Surgery comprised of lobectomy n=50 (71.4%), chest wall resection n=9 (12.9%) or pneumonectomy n=11 (15.7%). Negative surgical margins were achieved in n=63 (90%) and positive margins in n=7 (10%). 30-day mortality was n=2 (2.8%) both cases after Right-sided pneumonectomy.

      MTR was observed in 45 cases (64.3%) including a pathological complete response in 25 (35.7%) and < 10% residual tumor in 20 cases (28.5%). The mean percent of residual tumor cells was 16% and the median 6.5%. Percent of residual tumor cells did not correlate to radiation dose (Rs=0.092), and not to the histology (p= 0.165), and not to type of chemotherapy (p=0.35).

      8eea62084ca7e541d918e823422bd82e Conclusion

      NACCRT delivered to 60 Gy with modern image-guided radiation therapy is safe. Two thirds of such patients achieve major tumor regression.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA05.11 - Radiomics Analysis Using SVM Predicts Mediastinal Lymph Nodes Status of Squamous Cell Lung Cancer by Pre-Treatment Chest CT Scan

      14:40 - 14:45  |  Presenting Author(s): Xing Wang, Wu Nan  |  Author(s): Shi Yan, Quanzheng Li, Ning Guo, Zhe Guo

      • Abstract

      Background

      Assessment of mediastinal lymph nodes (N2 station) is essential in staging patients with Non-small-cell lung cancer (NSCLC), for patients with preoperative confirmed N2 status should follow neoadjuvant therapy before surgery, and occult N2 status should be avoided. There are several invasive and non-invasive exams available for preoperative N staging, like EBUS-TBNA and PET-CT scan. Chest CT scan was the basic examination of every patient, while only the length of minor axis could be used to predict lymph node involvement, and the potential value of CT might be underestimated. In this study we aimed to explore the value of radiomics analysis with machine learning in differentiating N2 from N1/N0 subjects using pre-treatment chest CT.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Ninety-three patients with squamous cell lung cancer, who underwent pre-treatment CT scans were included in this study. By use of Laplacian of Gaussian (LoG) filter and matrix based radiomics models (e.g. gray-level co-occurrence matrix), comprehensive radiomics features were extracted from the regions of interest which were manually delineated on primary tumors. We performed radiomics analysis using support vector machine (SVM) to test texture and heterogeneity features derived from pre-treatment CT images as indicators for the staging of lymph node metastasis, especially N2. The gold standard of N staging is confirmed pathologically after systematic mediastinal lymphadenectomy (N2 subjects=31).

      4c3880bb027f159e801041b1021e88e8 Result

      For the performance evaluation of single image feature, there are 16 features able to differentiate N2 subjects from others (N0 and N1) with p value <0.05. Furthermore, SVM training and classification were performed using 5-feature combinations as inputs. With feature selection, the best performance of N2 prediction is 83% accuracy with 87% sensitivity and 81% specificity.

      figure.jpg

      8eea62084ca7e541d918e823422bd82e Conclusion

      Radiomics analysis using SVM training can successfully predict N staging by pre-treatment chest CT scan for NSCLC patients, which could diminish the odds of occult N2 status and provide unique information preoperatively for treatment planning.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA05.12 - Discussant - MA 05.09, MA 05.10, MA 05.11

      14:45 - 15:00  |  Presenting Author(s): Matthew Hatton

      • Abstract

      Abstract not provided

  • +

    MA06 - PDL1, TMB and DNA Repair

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

      MA06.01 - The Intrinsic PD-L1 Promotes Cellular Invasiveness Via their PD-1 Receptor in Lung Adenocarcinoma Cells

      13:30 - 13:35  |  Presenting Author(s): Wen-Pin Su  |  Author(s): Hung-Chang Wu, Shuen-Ru Yang, Jheng-Cheng Huang, Jing-Jou Yan, Wan-Chen Kao, Li-Chan Chang, Wu-Chou Su

      • Abstract

      Background

      Lung cancer is the most frequent cause of cancer death. Programmed death 1 (PD-1) in T cells and its ligand PD-L1 in tumor cells play a key role in immune checkpoint therapy and had applied to advanced stage lung cancer. Migration and invasion of tumor cells is a prerequisite for tumor cell metastasis. Since intrinsic PD-1 receptor functions promote tumor growth was reported, we will investigate the interaction between PD-1 and PD-L1 in lung adenocarcinoma cell lines, the impact on chemosensitivity, and clinical outcome.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      In vitro experiments, lung adenocarcinoma CL1-5 cells, derived from CL1-0 cells. We prepared PD-L1-overexpression human lung adenocarcinoma cell line, derived from CL1-0 cells (CL1-0-PD1). Migration and invasion ability were assessed by transwell assay; EMT marker and regulator were evaluated by Western blotting. We also observed the morphology of cells. To explore interaction between PD1 and PD-L1, we added anti-PD-1 antibody into CL1-0, CL1-5, and CL1-0-PDL1 cells, and then test migration, invasion and cellular morphology. We also suppressed PD-1 by siRNA to test whether PD-1/PDL-1 interaction contributed to the EMT change. Further, we evaluated cellular proliferation and chemosensitivity by MTT assay and colony formation assay. We will correlate PD-L1 expression in lung cancer cells with clinical outcome by IHC stain clinically.

      4c3880bb027f159e801041b1021e88e8 Result

      In CL1-5 cells, derived from CL1-0 cells, with high PD-L1 expression possessed higher cellular migration ability than the parental CL1-0 cells with less PD-L1 expression. CL1-0 cells with PD-L1 overexpression had more expression of EMT (epithelial mesenchymal transition) regulator and mesenchymal marker. We also observed that CL1-5 and CL1-0-PDL1, which had more PD-L1 expression, are shaped like spindles; while CL1-0 cells are more rounded. Therefore, PD-L1 up-regulated cell migration and invasiveness in human lung adenocarcinoma cells and promotes EMT.

      After adding anti-PD1 antibody in CL1-5, CL1-0, and CL1-0-PDL1 cells, migration and invasion ability decreased. These result indicated anti-PD-1 antibody block the link between PD-1 and PD-L1 in cancer cells. The phenomenon was confirmed by PD-1 siRNA. Therefore, PD-1/PD-L1 axis regulated cancer cells migration and invasiveness. PD-L1 expression also decreased cellular proliferation and had little influence on chemsensitivity. Finally, we found that higher PD-L1 expression was correlated with lymph node metastasis in clinical specimen.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Lung adenocarcinoma cells with higher PD-L1 expression promote cell migration, invasiveness, EMT, and little chemoresistance. PD-L1 expression lowers proliferation rate. PD-1 and PD-L1 interaction on lung adenocarcinoma cells contribute cellular migration and invasiveness.

      6f8b794f3246b0c1e1780bb4d4d5dc53

      • Abstract

      Background

      Our previous work has demonstrated that higher level of genomic complexity is associated with more heterogeneous neoantigen repertoire, suppressed T cell repertoire and postsurgical relapse in localized non-small cell lung cancers (NSCLC) highlighting the complex interaction of tumor molecular and immune landscape and their impact on cancer biology and patient survival. We launched the ICON Project (Immune Genomic Profiling of NSCLC) to prospectively delineate the molecular and immune landscape of early stage NSCLC and their impact on patient survival through a multidisciplinary approach. Here we report the updated genomic and immune analyses.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Surgical specimens from stage I-III NSCLC were subjected to whole-exome and RNA sequencing for mutational analysis, in silico neoantigen prediction and gene expression analysis as well as T cell receptor sequencing, cytometry by time-of-flight and multiplex immunofluorescence staining.

      4c3880bb027f159e801041b1021e88e8 Result

      From 2016-2018, 127 patients were accrued and 50 surgical samples have undergone WES, RNAseq, TCR sequencing and immune phenotyping. Median age is 66 yrs (range: 39-86), 52% (26/50) were female and 76% (38/50) former smokers. 76% (38/50) are non-squamous carcinomas and 24% (12/50) squamous cell carcinomas. 34% have stage I disease (17/50), 30% stage II (15/50), 34% stage III (17/50) and 2% stage IV (1/50). The majority of patients had upfront surgery (45/50; 90%). With median follow-up of 19 months, 15 patients have relapsed. Median tumor mutational burden is 7.8mut/Mb and predicted neoantigen burden was 10/sample (range: 0-250). Predicted neoantigen burden is significantly correlated with tumor mutational burden (r=0.41, p=0.002). The most commonly mutated genes are TP53, KRAS, CDKN2A, PIK3CA, EGFR, BRAF, GRIN2A and ATM. C->A transversions and C->T transitions were the most common mutational subtypes. PD-1 expression and regulatory T-cell (CD4+/FoxP3+) infiltration are significantly increased in tumor tissue compared to normal tissue (p=0.003 and p=0.02 respectively), while CD3, CD8, granzyme B and CD45RO are decreased in tumor tissue compared to normal lung.

      8eea62084ca7e541d918e823422bd82e Conclusion

      NSCLC tumors have an immunosuppressive microenvironment compared to tumor adjacent normal lung tissues. Clinical data will be adequate to conduct genomic and immune profiling comparisons across different clinical subgroups. Mutational and neoantigen profiling are consistent with previously reported studies and correlations between molecular and immune landscapes and its impact on patient survival are ongoing.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA06.03 - PD-1 and Id-1 Combined Blockade Impacts Tumor Growth and Survival Through PD-L1 Expression and Tumor Infiltration by Immune-Related Cells 

      13:40 - 13:45  |  Presenting Author(s): Ignacio Gil-Bazo  |  Author(s): Iosune Baraibar, Marta Roman Moreno, Ines Lopez, Jesus Corral, Juan Jose Lasarte, Alfonso Calvo, Silve Vicent, Daniel Ajona

      • Abstract

      Background

      PD-1/PDL-1 inhibitors are approved in advanced non-small cell lung cancer (NSCLC). Long-term survival rates associated to PD-1/PDL-1 blockade have changed treatment paradigm. However, many patients do not benefit from PD-1/PDL-1 blockade. New therapeutic combinations are under investigation. Id1 is involved in proliferation, angiogenesis and immunosuppression. We described Id1 as an independent prognostic factor in NSCLC (Ponz-Sarvise, Clin Cancer Res 2011) and more recently showed Id1’s role in lung cancer metastasis (Castanon, Cancer Letters 2017). Here we test a combined therapeutic strategy targeting PD-1 and Id1 in a murine lung cancer model.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Three in vivo studies evaluated the impact of Id1 inhibition in tumor cells, tumor microenvironment and in both, on tumor volumes and mice survival. A syngeneic tumor model using C57BL/6 and Id1-/- Id3+/- mice was created by subcutaneous injection of Lewis Lung Carcinoma (3LL) cells and Id1 silenced 3LL (Id1Sh) cells. After injection, mice were treated with an anti-PD-1 (RMP-1-14) monoclonal antibody or PBS. Tumor volumes according to mice strain, Id1 status in tumor cells and treatment were quantified. Mice's survival was calculated in those groups. Tumor CD8+ and CD3+ TILs and CD68+ cells were quantified by specific immunostainings.

      4c3880bb027f159e801041b1021e88e8 Result

      Id1 inhibition in the tumor environment and the injected tumor cells, combined with anti-PD-1 treatment, induced a significant tumor growth impairment (p < 0.0001) and increased survival (p = 0.0051). CD3+ and CD8+ TILs and tumor CD68 + cells were significantly higher in tumors from mice with the combined Id1-PD-1 blockade treated with the anti-PD-1 inhibitor compared to control animals suggesting that tumor increased immune-related cells infiltration exerts the effector phase of the antitumor immune response. Additionally, PD-L1 expression seemed to be higher when Id1 expression was absent in the immune microenvironment (p = 0.04). Additional data based on multiplexed immunohistochemistry results will be presented at the meeting.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Id1 and PD-1 combined blockade in our syngeneic murine lung cancer model significantly impaired tumor growth and increased survival. Increased tumor PD-L1 expression and CD3+ and CD8+ TILs and CD68+ cells may explain these findings.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA06.04 - Discussant - MA 06.01, MA 06.02, MA 06.03

      13:45 - 14:00  |  Presenting Author(s): Akihiko Yoshida

      • Abstract

      Abstract not provided

    • +

      MA06.05 - The Micro-Environmental Cross Talk Between Mast Cells and Lung Cancer Cells Through Cell-to-Cell Contact

      14:00 - 14:05  |  Presenting Author(s): Rachel Shemesh  |  Author(s): Yaara Gorzalczany, Smadar Geva, Laila C. Roisman, Ronit Sagi-Eisenberg, Nir Peled

      • Abstract

      Background

      Mast cells (MCs) are key effectors in allergic reactions, but are also involved in tissue remodeling, wound healing and protection against pathogens. MCs infiltrate tumors and their number within the tumor microenvironment in certain cancer types, such as lung cancer, have been correlated with poor prognosis. The nature of crosstalk between lung cancer and MCs remain poorly resolved. In this study, we investigated the activation patterns within the MCs following cell-to-cell contact with lung cancer cells showing CD73 involvement and implying metabolic changes.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Human MCs (HMC-1 and LAD-2) were exposed to Human lung cancer cells (H1299), derived membranes to recapitulate cell contact mediated activation. Lysates of MCs were tested for protein expression and posttranslational modifications (i.e. phosphorylation) by targeted western blotting. We unraveled the intracellular signaling molecules that are necessary for this signaling pathway by a pharmacological approach using several inhibitors. Each condition was repeated at least twice.

      4c3880bb027f159e801041b1021e88e8 Result

      H1299 membrane exposure activated the ERK 1/2 MAP kinases in HMC-1 and in LAD-2 cells. AKT signaling was also activated in LAD-2 cells as a result of this contact. CD73 dephosphorylates AMP to adenosine within the MCs. Interestingly enough, this ERK 1/2 activation was inhibited by CD73 inhibitor and A3 receptor antagonists in HMC-1 cells. ERK 1/2 activation was inhibited by A3 receptor antagonists and PI3K in LAD-2 cells. Furthermore, we discovered that protein kinase C (PKC) inhibitor augments the activation of ERK 1/2 in LAD-2 cells. In contrast, PKC inhibitor inhibits the activation of ERK 1/2 in HMC-1 cells. In addition, we discovered that the AKT activation was inhibited by A3 receptor and PI3K inhibitors but not by CD 73 inhibitors.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Our results suggest that H1299 membranes activate ERK 1/2 in HMC-1 cells by a mechanism that involves autocrine formation of adenosine and is mediated by CD 73 and A3 receptor. In addition, we discovered that there is an important difference between the ERK 1/2 MAP kinase signal transduction in HMC-1 and LAD-2 cells, PKC is an inhibitor of the H1299 activation of ERK 1/2 in LAD-2 cells. In contrast, the H1299 membrane activation of ERK 1/2 kinase in HMC-1 cells is mediated by PKC. Furthermore, we can conclude that H1299 membranes activate AKT in an A3 receptor dependent mechanism that is mediated by PI3K.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA06.06 - An Ex-Vivo Patient-Derived, Immunocompetent (PDI) Culture System to Evaluate Immunotherapeutic Agents’ Anti-Tumor Efficacy

      14:05 - 14:10  |  Presenting Author(s): Zachary E. Tano  |  Author(s): Stefan Kiesgen, Navin Chintala, Jordan Dozier, John Messinger, Kay See Tan, Prasad S. Adusumilli

      • Abstract

      Background

      Anti-tumor efficacy of human immunotherapeutic agents, such as antibodies, chimeric antigen receptor (CAR) and T-cell receptor transduced T cells, are currently being investigated in immunodeficient mice prior to clinical translation. We developed and optimized an ex-vivo culture system utilizing malignant pleural effusions (MPEs) to compliment these investigations in a human, immunocompetent, tumor-like environment. We hypothesized that CAR T cells’ cytotoxicity will vary by the different immune compositions in each MPE, which are conditions unavailable in current efficacy assays.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Mesothelin-targeted CAR T cells from multiple donors were exposed to MPEs derived from non-small cell lung cancer patients (n=15) and RPMI culture medium. Influence of the MPEs on CAR T-cell efficacy was evaluated by viability and phenotype (flow cytometry), cytotoxicity (chromium release assay), and gene expression (NanoString). Group-based trajectory modeling was used to stratify the inhibitory effect of MPEs. MPE composition (ELISA and Luminex assays) was evaluated to interpret its influence on CAR T cells.

      4c3880bb027f159e801041b1021e88e8 Result

      With the incorporation of our optimized protocols, T cells retain their viability, phenotype (CD4/CD8), and percentage of CAR expression when cultured in MPEs. MPE soluble factor levels remained stable over multiple freeze/thaw cycles. CAR T cells co-cultured in MPE exhibited variable antigen-specific cytotoxicity (Fig. A). MPE-induced T-cell inhibition was stratified into groups of strong, mild, or no inhibition. (Fig. B). Compared to MPEs with either mild or no inhibition, MPEs with strong inhibition had significantly higher levels of TGFβ-2 (average TGFβ-2 level in strong vs. mild inhibition: 402 vs. 50 pg/mL, p<0.05) (Fig. C), IL-6, RANTES, and IL-5.

      pdi culture system.jpg

      8eea62084ca7e541d918e823422bd82e Conclusion

      We present the first human immunocompetent culture system that can be used to evaluate immunotherapeutic agents’ efficacy prior to their clinical translation. Furthermore, analyses of the culture system’s soluble factors sheds light on their relative influence on T-cell efficacy.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA06.07 - Genetic and Epigenetic Alterations are Associated with Tumor Mutation Burden in Non-Small Cell Lung Cancer

      14:10 - 14:15  |  Presenting Author(s): Liang-Liang Cai  |  Author(s): Hua Bai, Zhi-Jie Wang, Shuhang Wang, Jian-Chun Duan, Shu-Geng Gao, Jie He, Jie Wang

      • Abstract

      Background

      Although several studies have indicated that tumor mutation burden (TMB) is associated with non-small cell lung cancer (NSCLC) development and clinical efficacy of immune checkpoint inhibitors (CPIs), identification of factors associated with TMB is still a major biological issue. It is well-known that DNA transcription can be regulated through methylation and demethylation, gene silencing caused by DNA hypermethylation is associated with cancer development. However, the relationship between DNA methylation and TMB in NSCLCs remains unclear.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      The landscape of DNA sequence in Chinese NSCLCs population were surveyed by using whole-exome sequencing (WES) by profiling 178 lung tissues (89 without any systemic anti-cancer therapy tumors and matched normal lung tissues). According to the 104 median-level of TMB in our cohort, high TMB (n=16, 252-465 range mutations per tumor) and low TMB (n=13, 57-79 range mutations per tumor) groups were divded. The NSCLC methylome between high and low TMB was characterized on a genome-wide scale using Illumina Infinium MethylationEPIC arrays combined with the WES data.

      4c3880bb027f159e801041b1021e88e8 Result

      The results show frequently aberrant DNA methylation, abundant chromosomal amplifications and deletions, and mutational signatures in high TMB lung cancer. Combining with clinical data, cigarette smoking associated with high TMB were observed in our cohort. Cancer-specific epigenetic alterations were observed in 294,141 CpG sites, comprising both tumor hyper- (769,38) and hypo- (217,203) methylation in high TMB lung cancer while none in low. These different methylations sites cover 1232 genes including 25 HOX genes.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Global DNA hypomethylation and TP53 mutation, associated with increased chromosomal instability, were associated with TMB in NSCLCs.The high TMB NSCLCs are characterized by numerous copy number alterations and aberrantly methylated sites and display distinct mutational signatures. 25 hypermethylated HOX genes can be potentially useful as DNA methylation markers for prediction of TMB level. The results provide insights into the epigenetic impact of TMB, which may contribute to improve precison management of NSCLCs.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA06.08 - Discussant - MA 06.05, MA 06.06, MA 06.07

      14:15 - 14:30  |  Presenting Author(s): Jyoti Patel

      • Abstract

      Abstract not provided

    • +

      MA06.09 - XRCC6BP1: A DNA Repair Gene in Cisplatin Resistant Lung Cancer Stem Cells That May Predict Survival Outcomes in Patients

      14:30 - 14:35  |  Presenting Author(s): Martin P Barr  |  Author(s): Robert Farrell, Saravjeet Singh, Emma Foley, Yuexi He, Lauren Brady, Vincent Young, Ronan Ryan, Siobhan Nicholson, Niamh Leonard, Sinead Cuffe, Stephen Finn

      • Abstract

      Background

      Alterations in the DNA repair capacity of damaged cells is now recognised as an important factor in mediating resistance to chemotherapeutic agents.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      DNA Repair Pathway RT2 Profiler Arrays were used to elucidate key DNA repair genes implicated in chemoresistant NSCLC cells using cisplatin resistant (CisR) and corresponding parental (PT) H460 cells. DNA repair genes significantly altered in CisR cells were validated at the mRNA and protein level. The translational relevance of differentially expressed genes was examined in a cohort of chemo-naïve matched normal and tumour lung tissues from NSCLC patients. Loss of function studies were carried out using siRNA technology. The effect of XRCC6BP1 gene knockdown on apoptosis was assessed by FACS. Cellular expression and localisation of XRCC6BP1 protein and γH2AX foci in response to cisplatin were examined by immunofluorescence (Cytell™). To investigate a role for XRCC6BP1 in lung cancer stem cells, Side Population (SP) studies were used to characterise stem-like subpopulations within chemoresistant cells. XRCC6BP1 mRNA analysis was also examined in ALDH1+ and ALDH1- subpopulations. Immunohistochemistry analysis was carried out in resected lung tumour tissues and XRCC6BP1 expression was correlated with survival in addition to a number of clinicopathological parameters such as tumour stage & grade, gender, smoking status and chemotherapy.

      4c3880bb027f159e801041b1021e88e8 Result

      We identified a number of critical DNA repair genes that are differentially regulated between PT and CisR NSCLC cells. XRCC6BP1 mRNA and protein expression was significantly increased H460 CisR cells relative to their PT counterparts. Relative to matched normal lung tissues, XRCC6BP1 mRNA was significantly increased in lung adenocarcinoma patients. Gene silencing of XRCC6BP1 induced significant apoptosis of chemoresistant cells and reduced their DNA repair capacity. Immunofluorescence studies showed an increase in XRCC6BP1 protein expression and gH2AX foci in CisR cells. SP analysis revealed a significantly higher stem cell population in resistant cells, while XRCC6BP1 mRNA expression was considerably increased in SKMES-1, H460 and H1299 CisR cells positive for ALDH1 activity (ALDH1+) compared to ALDH1- cells. IHC scoring of XRCC6BP1 demonstrated poor survival outcomes for NSCLC patients with high expression of this DNA repair gene.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Our data highlight the potential of targeting components of the DNA repair pathway, in particular XRCC6BP1, in chemoresistant lung cancer. Furthermore, XRCC6BP1 may play an important role in subsets of lung cancer stem cells which, at least in part, may be responsible for driving and maintaining the cisplatin resistant phenotype in NSCLC.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA06.10 - Germline Mutation in ATM Affect Lung Cancer Risk with High Effect

      14:35 - 14:40  |  Presenting Author(s): Xuemei Ji  |  Author(s): Dakai Zhu, Claudio Pikielny, Olga Gorlova, Maria Teresa Landi, John Kirkpatrick Field, Paul Brennan, Mattias Johansson, Rayjean J. Hung, James D McKay, Christopher Ian Amos

      • Abstract

      Background

      Genome wide association studies have identified several lung cancer susceptibility regions and common variants influencing lung cancer risk. However, few previous studies investigated the association between germline mutations and lung cancer risk.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      We analyzed data from a case-control study with 19053 lung cancer cases and 15446 healthy controls of European ancestry in a discovery phase and performed a validation analysis using a case-control study comprising 4261 lung cancer cases and 4152 healthy controls of European ancestry for replication. Logistic regression was used to identify germline mutations with high effect within exome regions associated with lung cancer risk.

      4c3880bb027f159e801041b1021e88e8 Result

      We found rs56009889 in ATM was statistically associated with lung cancer risk in the discovery set (OR = 3.05, P = 3.68 × 10−8) and was nonsignificantly associated with lung cancer risk in the validation set (OR = 1.83, P = 0.16). Stratified analyses by gender with adjustment for age and smoking status showed that females carrying at least one mutated allele of rs56009889 (T/C + T/T) had an increased risk of lung cancer with ORs being 7.77 (95% CI 3.45 - 17.47) in discovery and 6.73 (95% CI 1.46–30.98) in replication, compared to C/C homozygotes among females. Individuals carrying at least one T allele showed a significant 6.9-fold increased risk for lung adenocarcinoma in discovery (adjusted OR = 6.85; 95% CI 4.37 – 10.75) and approximately a 4.9-fold increased risk in replication (adjusted OR = 4.89; 95% CI 2.01 – 11.91). Never smokers with combined genotypes (T/C + T/T) had a greater than 8-fold increased risk of lung cancer in discovery (adjusted OR = 8.03, 95% CI 4.00 – 16.13), while smokers only showed a 2.13-fold increased risk (adjusted OR = 2.13, 95% CI 1.25 – 3.65). In replication, however, the risks from this variant were comparable between smokers and nonsmokers, although the sample size is small for nonsmokers (adjusted OR = 2.16; 95% CI 0.48 – 9.79 for never-smokers and adjusted OR = 2.07; 95% CI 0.66 – 6.52 for smokers). All the T/T homozygotes of rs56009889 developed lung adenocarcinoma in discovery (P = 0.036). The association exhibited a dose-response relationship between the number of T allele of rs56009889 and lung cancer risk in discovery (Ptrend = 1.07 x 10 -9).

      8eea62084ca7e541d918e823422bd82e Conclusion

      rs56009889 highly affected the risk of lung cancer, mainly of lung adenocarcinoma, primarily in women and never smokers. These germline mutations provide important insights for the prevention of lung cancer.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA06.11 - Distinct Origins of Lymphatic and Brain Metastasis in Lung Cancer

      14:40 - 14:45  |  Presenting Author(s): Tao Jiang  |  Author(s): Yan Yan, Caicun Zhou

      • Abstract

      Background

      Generally, distant metastases are seeded by lymph node metastases in most solid tumors. This concept provides a mechanistic basis for the TNM staging system and is the rationale for surgical resection of tumor-draining lymph nodes. However, a recent study found that lymphatic and distant metastases could arise from independent subclones in the primary colorectal cancer. The current study aimed to investigate the origins of lymphatic and brain metastasis in lung cancer.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      39 samples from twelve patients with primary lung cancer and brain metastases were identified. Three of them had the matched lymph node metastases. All tissues and matched peripheral blood samples were collected before any systemic treatment. Whole-exome (>150×) sequencing were conducted on these samples.

      4c3880bb027f159e801041b1021e88e8 Result

      Compared to the primary lesions, both brain and lymph node metastases had the significantly different patterns of somatic genome alterations. The mutational landscape of brain metastases was also distinctly different from matched lymph node metastases. Primary lesions, matched brain and lymph node metastases showed the similar mutation pattern in terms of transition and transversion, and all of samples displayed a higher percentage of C>T transition. Brain metastases had numerically higher tumor mutational burden (TMB) than primary lesions but it did not reach the statistical significance. Notably, we observed the totally distinct origins of lymphatic and brain metastasis in all three matched cases.

      8eea62084ca7e541d918e823422bd82e Conclusion

      The current evidence suggested that brain metastases and matched lymph node metastases had different mutational landscape in patients with lung cancer. Brain metastases had higher TMB than their primary lesions. Lymphatic and brain metastasis had distinct origins in lung cancer. These results had profound clinical implications for application of immunotherapy and improvement of prognosis in patients with lung cancer and brain metastases.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA06.12 - Discussant - MA 06.09, MA 06.10, MA 06.11

      14:45 - 15:00  |  Presenting Author(s): Rebecca Heist

      • Abstract

      Abstract not provided

  • +

    MA07 - Towards Survivorship: The Landscape, Supports and Barriers

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

      MA07.01 - No Longer Outliers: Understanding the Needs of Long-Term Lung Cancer Survivors

      13:30 - 13:35  |  Presenting Author(s): Maureen Rigney  |  Author(s): Jennifer C King, Andrew Ciupek

      • Abstract

      Background

      Lung cancer is the leading cause of cancer death in most developed and developing countries. But people do survive, sometimes for many years. Those diagnosed with lung cancer experience higher levels of distress and have greater unmet physical and emotional needs compared with other types of cancer. But what of long-term survivors?

      Globally, The Cancer Atlas reported an estimated 1,878,000 people were living with lung cancer in 2012. With the introduction of screening and rapid treatment advancements, that number is only expected to increase. Are we prepared to meet the long term and late effects of lung cancer? First, we must better understand the experiences and identified needs of long-term survivors.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      820 people responded to a 120 question online survey that was distributed via social media and targeted outreach. 471 identified as lung cancer patients/survivors and 349 as loved ones. 21% of survivor-respondents indicated they had been diagnosed 5+ years prior.

      Queried on treatment and smoking histories, long-term survivors identified their most prevalent and problematic symptoms and side effects experienced during treatment, shortly after treatment ended and at 5+ years post-diagnosis. They also answered questions regarding treatment decision-making and palliative care discussions and provision of post-treatment survivorship plans.

      4c3880bb027f159e801041b1021e88e8 Result

      74% of long-term survivors had surgery, 43% had experienced a recurrence and 5% had participated in a clinical trial. None were current smokers.

      The most common (and problematic) late and long term symptoms and side effects were shortness of breath (39%), fatigue (28%) and anxiety (24%). Memory problems were also rated as common (27%).

      Long-term survivors indicated that during treatment, physical side effects were most problematic but post-treatment and long-term, emotional effects were more difficult. Financial issues were also more problematic 5+ years after treatment compared with other time periods. Both discussions of palliative care and provision of survivorship care plans were rare.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Long-term lung cancer survivors were once considered ouliers but today those diagnosed are increasingly living five years and longer. How do the late and long-term physical effects of lung cancer and its treatments differ from survivors of other types of cancer? How do long-term survivors manage stigma and survivor guilt? What physical and emotional support and services do they need? This survey provides initial insights into the physical. emotional and financial effects of living longer with lung cancer but more research is needed to allow us to more fully understand how we can support our long-term survivors.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA07.02 - Line of Therapy and Patient Preferences Treating Lung Cancer: A Discrete-Choice Experiment

      13:35 - 13:40  |  Presenting Author(s): Andrea Ferris  |  Author(s): John F.P. Bridges, Upal Basu Roy, Ellen Janssen

      • Abstract

      Background

      Patient preferences now play an important role in cancer research, regulatory science, and value assessment. While there is a growing literature exploring the preference of patients with lung cancer, few studies have explored how preferences vary with patients’ treatment experience. We sought to quantify patient preferences for the benefits and risks of therapy and explore how they vary across line of treatment.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Preferences were estimated using a discrete choice experiment (DCE) developed in partnership with a patient and stakeholder advisory boards. A D-optimal experimental design was used to generate 3 blocks of 9 choice tasks spanning five attributes: progression-free survival (PFS), short-term side effects, long-term side effects, risk of developing late-onset side effects, and mode of administration – each defined across 3 relevant levels. A diverse sample was recruited via email sent to the LUNGevity lung cancer patient database and via social media. A choice mode was estimated use a conditional logistic regression where the dependent variable was the respondents preferred treatment in each profile. The relative attribute importance (conditioned on the chosen attribute levels) was then compared across the respondents’ self-reported line of treatment.

      4c3880bb027f159e801041b1021e88e8 Result

      In total we had 350 eligible respondents, of which 279 (80%) completed as least on DCE task of which 3% did not receive a pharmacotherapy, 39% received first line therapy, and 58% had two or more lines of theory. As with previous studies, PFS was the most important attribute for patients and was similarly valued (P=0.406) among first- and later (second lines and more) lines of treatment (33.4% v 33.8%). Patients on first-line treatment placed great emphasis (P<0.001) on long-term side (18.9% v 14.1%) and late onset side effects (15.3% v 10.3%), but less emphasis (P<0.001) on short-term side effects (27.8% v 29.8 %) and mode of administration (4.6% v 12.0%) than those on later lines.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Population estimate of patient preference remain important, but more effort is needed to understand how patient preference vary across patient with different backgrounds and treatment experiences. We show that line of treatment does not effect how patients value time, but their experience may have an impact on treatment characteristics. Latent class analysis may allow for the identification of groups with similar preferences that could allow for multivariate analyses to explain preference heterogeneity.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA07.03 - Attitudes to Lung Cancer in Europe: Findings from a Global Consumer Survey

      13:40 - 13:45  |  Presenting Author(s): Jesme Fox  |  Author(s): Aoife McNamara, Maureen Rigney, Greg Manuel, Sarah Winstone

      • Abstract

      Background

      If lung cancer is diagnosed early, patients’ chances of successful treatment are increased. Stigma
      around lung cancer, as a tobacco-related cancer, can discourage patients from talking to their doctor
      about potential symptoms. In 2017, the GLCC commissioned Populus to undertake an international
      consumer survey in each of the 25 countries of the GLCC members.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      1,000 adults, in 16 European countries, participated via an online survey in July 2017. To assess
      attitudes to lung cancer, they were told that lung cancer is mainly caused by smoking and other
      tobacco products. They were then asked the extent to which they agreed or disagreed with the
      statement: “I have less sympathy for people with lung cancer than for people with other cancers.”

      4c3880bb027f159e801041b1021e88e8 Result

      One in five (20%) people in Europe agreed that they have less sympathy for people with lung cancer
      than other forms of cancer (Chart 1). There was variation between countries with 30% of people in
      Portugal agreeing they have less sympathy in comparison to only 17% agreeing in Denmark, the
      Netherlands, Norway, Russia, Slovenia and Spain. Men in Europe are generally less sympathetic
      than women, and those aged over 55 are most sympathetic. In addition, there was a statistically
      significant correlation between those countries with lower cigarette consumption and people agreeing
      that they have less sympathy for people with lung cancer.

      Chart 1: European attitudes to lung cancer

      glcc - european attitudes - chart 1.png

      8eea62084ca7e541d918e823422bd82e Conclusion

      Everyone - no matter what the cause of their cancer - deserves to have high quality treatment and
      care. The persistent and varied levels of stigma associated with lung cancer across Europe needs to
      be addressed, so that people experiencing symptoms are not discouraged from seeking early
      intervention.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA07.04 - Discussant - MA 07.01, MA 07.02, MA 07.03

      13:45 - 14:00  |  Presenting Author(s): Kim Norris

      • Abstract

      Abstract not provided

    • +

      MA07.05 - Psychosocial Needs and Programs of Cancer Patients/Survivors and Their Relatives: Unmet Needs from an International Study

      14:00 - 14:05  |  Presenting Author(s): Csaba László Dégi  |  Author(s): Samantha Serpentini, Savita Goswami

      • Abstract

      Background

      In consideration of the dynamic nature of cancer patients’ needs, systematic understanding of their unmet needs from a socio-ecological perspective may be essential as the patients’ needs and available services are likely to vary by different healthcare systems in different countries. To investigate the role of geographical influence in cancer patients’ unmet needs, this study seeks to compare the unmet needs of and available programs for cancer patients/survivors and their family members by different types of healthcare systems across different countries.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      The IPOS Survivorship Online Survey is distributed to international and regional Psycho-Oncology organization members, which covers countries in six continents. Survey participants’ countries where they practice/research will be categorized into four groups by the types of healthcare system: Beveridge Model, Bismarck Model, National Health Insurance Model, and Out-of-Pocket Model.

      4c3880bb027f159e801041b1021e88e8 Result

      With estimated survey to be completed by August 30th, 2018, repeated measures ANOVA will be employed to test differences in patients’ unmet needs by the four healthcare system groups, separately for patients’ unmet needs and their family caregivers’. Differences by individual countries will also be explored.

      8eea62084ca7e541d918e823422bd82e Conclusion

      Findings will provide a global overview and a specific knowledge of the geografical differences in the psychosocial unmet needs and psycho-oncological programs for cancer patients/survivors and their family members/caregivers. Findings will also guide how to prioritize areas of cancer care that require improvement in psycho-oncology interventions and practices; and to highlight critical aspects for delivering quality care that vary by healthcare systems.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA07.06 - Telephonic Communication In Palliative Care For Better Management Of Terminal Cancer Patients In Rural India -  An NGO Based Approach. 

      14:05 - 14:10  |  Presenting Author(s): Nabanita Mandal

      • Abstract

      Background

      Due to financial incapability and absence of manpower poor families often fail to carry their advanced cancer patients to the nodal centres. This pilot study will explore whether communication by mobile phone can lessen this burden.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Initially a plan was generated regarding management of an advanced cancer patient in a nodal centre at District Head Quarter. Subsequently every two week a trained social worker attached to nodal centre will follow up and give necessary advice and emotional support to the patients and their families through their registered mobile phone number. Patient’s family were also encouraged to communicate with the team by phone in case of fresh complain and urgency in between.

      4c3880bb027f159e801041b1021e88e8 Result

      Since initiation in January 2017, 210 cancer patients were contacted by mobile phone every two weeks to enquire about their difficulties. In 76% of the situation trained social workers could give necessary advice by phone regarding management of their physical symptoms. Moreover patient’s family were really overwhelmed by the emotional support offered by the team over phone. Only 24% of cancer patients has to attend the nodal centre for expert advice from Palliative Care specialists.

      8eea62084ca7e541d918e823422bd82e Conclusion

      This novel approach helped
      * In providing regular physical and emotional support to the patients and their families.

      * In significantly reducing the financial and manpower problems of carrying patients to the nodal units.
      * In improve the quality of life of patients by continuous guidance.


      More and more team members can take help of this new strategy for better communication and uninterrupted care.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA07.07 - Identifying the Severity of Psychosocial Symptoms Among Patients Diagnosed with Lung Cancer. Do We Really Need Emotional Support Groups?

      14:10 - 14:15  |  Presenting Author(s): Arooj Fatima  |  Author(s): Syed Sammar Abbas Zaidi

      • Abstract

      Background

      Lung cancer is the second most common cancer among men and women. Most of the lung cancers are diagnosed at later stages among those patients who are underprivileged. The diagnosis and treatment of lung cancer is a continuous emotional distress for both patient and their family. We aim to identify the severity of depression, emotional distress, stress and mental fatigue among those patients who are diagnosed with lung cancer .

      a9ded1e5ce5d75814730bb4caaf49419 Method

      A cross sectional study was conducted in Shaukat Khanum Hospital, Lahore from March 2014 to April 2015. Exclusion and Inclusion criteria were made. 150 were enrolled in the study. Socio demographic characteristics were evaluated using Beck Depression Inventory and socio demographic form. Severity of depression was estimated by using Hamilton D (HAM-D). Various variables were analysed including parent’s age, level of education, socioeconomic status, gender and number of children.

      4c3880bb027f159e801041b1021e88e8 Result

      68% of the participants exhibited severe range of depression. 27% showed moderate depression where as 5% participants were showing the mild range of depression. An inverse co relation was found between educational status, occupational status (paid or unpaid), their marital status, socioeconomic family status and depression. Women 71% were found be more depressed than males.

      8eea62084ca7e541d918e823422bd82e Conclusion

      We concluded that majority of patients from psychosocial symptoms particularly depression and it is mainly associated with some factors. There is need to incorporate patients into the diagnosis and treatment process so that we can over come the effects of depression on the health outcomes of patients diagnosed with lung cancer. This can only be possible through appropriate education and emotional support programmes.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA07.08 - Discussant - MA 07.05, MA 07.06, MA 07.07

      14:15 - 14:30  |  Presenting Author(s): Jennifer C King

      • Abstract

      Abstract not provided

    • +

      MA07.09 - Willingness to Perform Multiple Biopsies to Improve Quality of Lung Cancer Care: Understanding the Oncologists’ Perspective

      14:30 - 14:35  |  Presenting Author(s): Upal Basu Roy  |  Author(s): Margery Jacobson, Andrea Ferris

      • Abstract

      Background

      Biomarker testing of advanced-stage non-small cell lung cancer (NSCLC) at the time of diagnosis is required to determine if a patient will benefit from a targeted therapy or immunotherapy. A patient may, however, need additional biopsies (rebiopsy) if the cancer recurs to determine the next line of therapy or to determine eligibility for a new drug or participation in a clinical trial. A LUNGevity study, conducted with 340 patients, revealed that patients were willing to undergo rebiopsies if that meant access to additional treatment options at the time of recurrence. However, only 36% of patients reported that their doctors recommended repeat biopsies at progression.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      To understand this patient-physician communications gap, we conducted an IRB-approved semi-structured survey-based study of 130 oncologists from academic research centers, community cancer centers, and private practice.

      4c3880bb027f159e801041b1021e88e8 Result

      Of the 130 oncologists surveyed,

      - Ninety percent of oncologists reported recommending a rebiopsy to their patients. However, when stratified by advanced-stage patient volume, oncologists with higher advanced-stage patient volumes reported higher rebiopsy and testing rates than those with low volumes (95% vs. 78%, p<0.05). Only 29% of the oncologists prescribed a rebiopsy in the past one year.

      - Major barriers to rebiopsy reported by oncologists included cost/reimbursement of a rebiopsy and treatment delay for 2nd- or subsequent lines of therapy

      - Among the types of biomarker testing performed at the time of progression, oncologists were more likely to prescribe testing for biomarkers with approved treatments (driver mutations – 94%, PD-L1 – 85%) unlike biomarkers for treatments in clinical development (43%) (p<0.05).

      - A forward linear regression analysis revealed that positive predictors of rebiopsy included treatment at a NCI Designated Cancer Center, while treatment at a community cancer center or private practice, presence of driver mutations at the time of diagnosis, and performance status of patient were negative predictors of rebiopsy

      - When presented with specific treatment scenarios for biomarkers (EGFR and ALK) that have 2nd-line treatment options, oncologists differed in their approach, suggesting a need for oncologist education about rebiopsying and subsequent biomarker testing

      8eea62084ca7e541d918e823422bd82e Conclusion

      Our study demonstrates that rebiopsy practices vary by practice settings and volume of advanced-stage lung cancer patients. Even when rebiopsies are prescribed, a comprehensive biomarker profile of the tumor may not be obtained, due to variations in tests requested. A major implication is the need for appropriate oncologists’ education to ensure practice change for delivery of optimal care to lung cancer patients.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA07.10 - Utilizing a Personalized Navigation Program to Identify Barriers and Increase Clinical Trial Participation Among Lung Cancer Patients

      14:35 - 14:40  |  Presenting Author(s): Andrew Ciupek  |  Author(s): Tara Perloff, Achintya Jaitly, Jennifer C King

      • Abstract

      Background

      Only about 5% of cancer patients participate in clinical trials. We previously conducted a survey of U.S. lung cancer patients and found that only 22% reported discussing clinical trials with their oncologist at the time of making treatment decisions. We hypothesized that a personalized navigation program could both increase rates of trial discussion and identify barriers to participation among lung cancer patients.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      We asked callers to Lung Cancer Alliance's 1-800 support line if they had considered clinical trial participation and referred willing callers to a navigator for further discussion. Navigators provided basic clinical trial education and a personalized list of trial matches. Patients were encouraged to discuss these trials with their treating oncologist. Navigators then regularly followed up with participants, via email or phone, at two to four-week intervals, to offer further support and collect outcomes information.

      4c3880bb027f159e801041b1021e88e8 Result

      We referred sixty callers to a navigator. Only 43% of callers reported a prior clinical trials conversation with their provider. Patients who had not started treatment or were on first-line treatment reported lower discussion rates (30%) than those on later treatment lines (60%). Among patients with follow up, 13 of 20 patients who had not discussed trials with their provider reported doing so after navigation. Ten of eleven patients that had a previous trial conversation initiated an additional one. Primary reasons given for not talking discussing after navigation were having stable disease on a current treatment or waiting for a clinical result. Ten patients reported contacting a trial. Primary reasons for not contacting a trial after discussion were disease progression, choosing a standard of care alternative, or waiting for a clinical result. Four patients have enrolled on a trial. Two patients were determined ineligible for a trial they approached for not meeting listed eligibility criteria and two for reasons not appearing in public trial information.

      8eea62084ca7e541d918e823422bd82e Conclusion

      We identified barriers throughout the clinical trials consideration and enrollment process. One set of barriers was related to care coordination, as exemplified by low rates of trial discussion during early stages of treatment and patient reports of delayed trial consideration when currently receiving treatment or waiting on a clinical result. Communication of trial information was another area presenting barriers, as exemplified by exclusion of patients from trials for reasons not readily apparent from public trial information. Improving integration of trial discussion during care and ensuring availability of accurate, updated trial information may be essential to increase trial participation.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA07.11 - Drug Price Comparison in Advanced Lung Cancer – High Cost Prices is Accompanied by Patient Benefits?

      14:40 - 14:45  |  Presenting Author(s): Luciene Bonan

      • Abstract

      Background

      In our recent decade we are seen new drugs coming up with high speed development to attend personalized conditions in lung cancer treatment. After the first TKI for EGFR mutation, many other target drugs such as TKI for ALK/ROS1 alteration, third-generation EGFR TKI, anti-PD-1/PD-L1 immunotherapies bring together an improvement in survival with better quality of life than chemotherapies. But this new specialty drugs are also testing the affordability of the market with new launched ceiling prices. Frequently, their prices have been settled down in a context of an unmet condition appeal rather than the truly health benefits. In pricing it is a common practice to use the external reference price between countries to align the prices based on international market. But if the first price is launched (frequently in USA) in countries that don’t use metrics based on evidence or clinical benefits, the price plateau could be replicated even without necessarily deserving this price. The objective of this presentation is to show the price comparison of drugs included in TKI class and immunotherapy class between high and middle-income countries. Then to compare the cost-treatment of therapies commonly used in advanced lung cancer and their magnitude of clinical benefit.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      All local currencies were converted to US dollars using PPP factor. The magnitude of effect was evaluated based on the ESMO Magnitude of Clinical Benefit Score.

      4c3880bb027f159e801041b1021e88e8 Result

      USA has the highest drug price followed by Brazil, especially in recent launched drugs. Costs of advanced lung cancer treatment significantly increase 5 times more when compared first-generated TKI and new generation TKI. Immunotherapy for second line costs 6 times more than first line with EGFR TKI and could cost more than 7 to 130 times the chemotherapy with docetaxel. Clinical benefits do not reach the same scale.

      8eea62084ca7e541d918e823422bd82e Conclusion

      The market of anticancer drug increasing 10% annually, but clinical benefits don’t advance in the same compass. Specialized drugs come into the market with pricing warrant of unmeet conditions, but if we think in precision medicine all new drug-target biomarker could be priced higher because it will cover a rare or unmet condition. In the context of precision medicine, is it fear a patient pays more because he has a different biomarker for the same clinical condition? If countries do not start to evaluate and pricing drugs based on value, market strategists will continue to test the ceiling price that health systems can(not) afford.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      MA07.12 - Discussant - MA 07.09, MA 07.10, MA 07.11

      14:45 - 15:00  |  Presenting Author(s): Govind Babu Kanakasetty

      • Abstract

      Abstract not provided

  • +

    MS05 - Diagnostic Dilemma in Lung Cancer

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

      MS05.01 - Staging and Pathology of Multiple Lung Nodules

      13:30 - 13:50  |  Presenting Author(s): Alain C. Borczuk

      • Abstract

      Abstract

      The evaluation of patients with multiple lung nodules, which include situations of synchronous as well as metachronous presentations, has become an area of focused interest. This is in part due to valuable outcome datasets that have evolved in conjunction with staging during the adoption of the AJCC 7th and now 8th edition for lung tumor staging. It has become evident that the evaluation requires an integration of clinical, radiologic, pathologic and molecular data to produce the most precise stage to guide patient management.

      In earlier staging editions, multiple nodules were largely assumed to represent advanced disease; however, the survival data pointed to a more heterogeneous patient group. Differences in histology were cited as reasons for a conclusion of synchronous primary tumors, but these differences needed to be relatively stark – for example adenocarcinoma versus squamous carcinoma. Three major shifts in knowledge impacted this field - 1) in pathology, movement away from the “lumping” of adenocarcinoma into mixed subtype with recognition of categories of early non-mucinous lepidic tumors 2) imaging advances, including the advent of successful lung cancer screening and imaging correlates of early lepidic pattern tumors and 3) the explosion of molecular pathology in lung cancer, particularly in adenocarcinoma.

      The result of these advances is a sophisticated approach that integrates these data into a stage as well as a conceptual conclusion regarding tumor biology and pathogenesis. Once lung malignancy is confirmed pathologically in a patient, the radiologist can use imaging features, especially ground glass nodules and the configuration of part solid nodules, to lead to clinically relevant conclusions about the likelihood of synchronous primaries when multiple nodules are encountered. In fact, other parameters, such as persistence, size, growth, imaging characteristics and evolution of a solid component in a part solid lesion may lead to relevant predictions in the absence of histologic confirmation. For the pathologist, non-mucinous lepidic tumors (adenocarcinoma-in-situ, minimally invasive adenocarcinoma and lepidic predominant adenocarcinoma) and the recognition of unusual patterns of invasive carcinoma, function well in the decision making process, as non-mucinous AIS, MIA and LPA are considered primary tumors and unusual patterns of invasive carcinoma (e.g. micropapillary or invasive mucinous) within the multifocal lesions are an important feature in recognizing intra-pulmonary metastasis. This area still requires refinement, especially among the LPA tumors and tumor in which common patterns such as acinar patterns predominate.

      Several molecular features can also be introduced into the discussion. Certain molecular events, such as particular patterns of copy number alterations, may occur in an individual patient’s tumor but would otherwise be uncommon in that tumor type overall. Such a feature or set of features can assemble to form a molecular fingerprint of that primary tumor that is likely to be preserved in metastatic foci but unlikely to occur by chance in a new primary. In a similar fashion early molecular drivers of adenocarcinoma such as EGFR mutation and KRAS mutation, most often persist in metastatic foci. Discordance in such drivers can be valuable evidence to support synchronous primary carcinoma when used in conjunction with other information. Use of a wider set of mutational alterations may lead to more accurate information with regard to the likelihood of tumor evolution from one primary (i.e. intrapulmonary metastasis).

      After integration of these data, the authors of the AJCC staging 8th edition recognize 4 disease patterns with associated imaging, pathology, TNM classification and conceptual viewpoint. Second primary lung cancer (unrelated tumors), multifocal ground glass or lepidic nodules (separate tumors, despite common non-mucinous lepidic morphology), pneumonic-type adenocarcinoma (often mucinous, single tumor with diffuse pulmonary involvement), and separate tumor nodule (single tumor with intrapulmonary metastasis). While data already exist suggesting the staging, often downstaging, of multiple pulmonary carcinomas, widespread use of the AJCC 8th edition and accrual of cases, with survival information, that fit the above conceptual approach are needed to support its biological significance.

      References:

      Detterbeck FC et al. The IASLC Lung Cancer Staging Project: Background Data and Proposed Criteria to Distinguish Separate Primary Lung Cancers from Metastatic Foci in Patients with Two Lung Tumors in the Forthcoming Eighth Edition of the TNM Classification for Lung Cancer. J Thorac Oncol. 2016 11(5):651-65.

      Detterbeck FC et al. The IASLC Lung Cancer Staging Project: Background Data and Proposals for the Application of TNM Staging Rules to Lung Cancer Presenting as Multiple Nodules with Ground Glass or Lepidic Features or a Pneumonic Type of Involvement in the Forthcoming Eighth Edition of the TNM Classification. J Thorac Oncol. 2016;11(5):666-80.

      Nicholson AG et al. Interobserver Variation among Pathologists and Refinement of Criteria in Distinguishing Separate Primary Tumors from Intrapulmonary Metastases in Lung. J Thorac Oncol. 2018;13(2):205-217.

      Girard N et al. Genomic and mutational profiling to assess clonal relationships between multiple non-small cell lung cancers. Clin Cancer Res. 2009;15(16):5184-90.

      Asmar et al, Use of Oncogenic Driver Mutations in Staging of Multiple Primary Lung Carcinomas: A Single-Center Experience. J Thorac Oncol. 2017;12(10):1524-1535.

      e353dbe42c8654f33588d4da0b517469

    • +

      MS05.02 - Defining Invasion in Minimally Invasive Adenocarcinoma

      13:50 - 14:10  |  Presenting Author(s): Masayuki Noguchi

      • Abstract

      Abstract

      For pathological diagnosis of early invasive forms of adenocarcinoma, such as adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), and early-stage lepidic adenocarcinoma, morphological criteria for “invasion” were defined by the recent WHO classification in 2015. However, the inter-observer agreement rate has not been examined in detail. The IASLC pathology committee has initiated a trial of the diagnostic reproducibility of morphological criteria for invasive lung adenocarcinoma, especially the minimally invasive form. Although the study is still ongoing, several interesting results have emerged. First, although total tumor size showed a good k-ratio, the extents of tumor invasion calculated by the participants revealed a very low k-ratio. Secondly, although the areas determined by the participants were similar, the criteria upon which their judgements were based varied, and the same morphological features were not always employed. These results suggest that the morphological criteria used for defining tumor invasion are very difficult to generalize.

      On the other hand, driver oncogene mutations such as EGFR, BRAF, and RAS, and activated fusion proteins such as ALK, ROS, and RET have been detected in invasive adenocarcinoma. Except for EGFR mutation, however, other genetic alterations have not been detected in AIS. My group has been investigating the expression and genetic profiles of AIS and early but invasive lung adenocarcinoma using a molecular biological approach, and characterized several genomic and epigenetic abnormalities in these tumors.

      Among them, stratifin (SFN) appears to be a promising biomarker of tumor malignancy at the early stage, its overexpression being very specific to the stage when AIS progresses to early invasive adenocarcinoma (Fig. 1). SFN, a member of the 14-3-3 protein family (14-3-3 sigma), is completely suppressed epigenetically in normal lung tissue. However, during the course of malignant progression, its promoter region becomes di-methylated and expression of SFN increases. SFN binds specifically to SKP1 and inhibits the function of E3 ubiquitin ligase, thus preserving the functions of specific proliferation-associated proteins such as p-cyclin E1 and p-c-Jun. Abnormally high expression of SFN is thought to be a biological marker of lung adenocarcinoma invasion. Interestingly, several inhibitors of SFN-SKP1 binding are very effective for suppressing tumor growth in nude mice.

      Another interesting biomarker of early adenocarcinoma is epithelial cell transforming sequence 2 (ECT2), a guanine nucleotide exchange factor physiologically localized in the nucleus and functioning to regulate cytokinesis. ECT2 is also localized in the cytoplasm of cancer cells. Aberrant cytoplasmic expression of ECT2 is thought to drive tumor growth and invasion. Cytoplasmic expression of ECT2 in 167 lung adenocarcinomas was evaluated by immunohistochemistry and its clinical significance was examined (Fig. 2). Cytoplasmic expression of ECT2 was found to increase during cancer progression. Cytoplasmic positivity for ECT2 was associated with a poor outcome in terms of both disease-free and overall survival (both P<0.001), and was an independent prognostic factor related to overall survival (P= 0.025).

      Reference

      Kosibaty Z, Dai T, Murata Y, Minami Y, Kano J, Nakagawa T, Sakashita S, Noguchi M. Cytoplasmic expression of epithelial cell transforming sequence 2 (ECT2) in lung adenocarcinoma and its implication for malignant progression. Lab Invest, 2018 (in press)

      Kim Y, Shiba-Ishii A, Nakagawa T, Iemura SI, Natsume T, Nakano N, Matsuoka R, Sakashita S, Lee S, Kawaguchi A, Sato Y, Noguchi M. Stratifin regulates stabilization of receptor tyrosine kinases via interaction with ubiquitin-specific protease 8 in lung adenocarcinoma. Oncogene, 2018. [Epub ahead of print]

      Shiba-Ishii A , YunJung Kim, Toshihiro Shiozawa, Shinji Iyama, Kaishi Satomi, Junko Kano, Shingo Sakashita, Yukio Morishita,Masayuki NoguchiStratifin accelerates progression of lung adenocarcinoma at an early stage.Molecular Cancer, 2015,14:142.

      Murata Y, Minami Y, Iwakawa R, Yokota J, Usui S, Tsuta K, Shiraishi K, Sakashita S, Satomi K, Iijima T, Noguchi M.ECT2 amplification and overexpression as a new prognostic biomarker for early-stage lung adenocarcinoma. Cancer Sci, 2014,105(4):490-7.

      Figure 1

      fig 1 mnoguchi.jpg

      Figure 2

      fig 2 mnoguchi.jpg

      e353dbe42c8654f33588d4da0b517469

    • +

      MS05.03 - Tumor Heterogeneity in Lung Cancer

      14:10 - 14:30  |  Presenting Author(s): Elisabeth Brambilla

      • Abstract

      Abstract

      Lung cancer are composed of populations of cells with distinct molecular genetics , epigenetics and phenotypic features : this phenomenon called intra-tumor heterogeneity (ITH) adds complexity to the already well known inter-tumor heterogeneity which is responsible for the huge number of types and subtypes in each main histological category type as defined in the 2015 WHO classification. Intra-tumor heterogeneity impacts on tumor biopsy strategy, characterization of actionable targets , treatment planning and drug resistance.

      Tumor types which display the highest histological heterogeneity are adenocarcinoma (ADC) , adenosquamous carcinoma, pleomorphic carcinoma and the high grade neuroendocrine tumor small cell lung carcinoma (SCLC) and large cell neuroendocrine(LCNEC) when combined . However intra-tumor heterogeneity is not restricted to histology since molecular heterogeneity at the levels of genetics (mutations , copy number alterations), DNA methylation , mRNA expression profiling and immune context /PDL1 expression show high intra-tumor variability in most conventional lung tumor histologies. Capturing the full molecular landscape of each tumor and choosing the right target is indeed a crucial clinical dilemma facing intra-tumor heterogeneity . Although histological main type correlates with characteristic genomics (1) , histological heterogeneity relies more on transcriptomic expressions profiles and functional pathways ( embryonal, stemcellness, EMT ..) than on driver mutations , challenging the mechanisms of tumor plasticity and resistance.(2)

      A -Histological Heterogeneity

      1-Adenocarcinoma (60% of lung cancer) present with up to 5 histological patterns declined as predominant or not , lepidic, papillary, acinar/cribriform, micropapillary and solid (WHO 2015) requiring resection samples to allow an accurate diagnosis of combinations. Advantages derive from this complexity: some patterns have high prognostic value : high grade patterns solid , micropapillary and cribriform (even not predominant >= 5%) are independent determinants of short survival, whereas the pure lepidic pattern (AIS) or minimally invasise MIA predict 100% 5 years survival and predominant lepidic invasive ADC have statistically better prognosis than all other ADC .Spatial heterogeneity is well conserved temporally allowing discrimination of multiple primaries versus lung metastases when different or same combinations are compared . Disadvantage resides in interpretation of small samples showing only a part of the patterns . At molecular levels (genomics, expression profiling, immune context) adenocarcinoma show extended intra-tumor heterogeneity. Unfortunately each pattern does no predict specific driver mutations (EGFR,KRAS, BRAF, gene amplification) .

      2- Adenosquamous carcinoma (2% of lung cancer) is typically composed of both squamous cell carcinoma (SQCC) and adenocarcinoma (ADC) components (at least 10%). Both components might be morphologically obvious (ADC pattern, keratinizing SQCC ) or one is obvious and IHC is necessary to identify the other as SQCC (P40+) or ADC (TTF1) .Diagnostic challenges include size of the sample, one component missing on small sample with its associated mutations ,and variable components in primary and metastases. Molecular features are characteristic of one or both components: 56% tyrosine kinase mutations (32% EGFR; 11% KRAS;1-4% others) in ADC or in both ,FGFR1 amplification in SQCC. Dual histospecific mutations indicate common cell of origin with early clonal trunk mutation maintained at progression (2)

      3-Pleiomorphic carcinoma : ( 0.5% of lung cancer) is the most heterogeneous entity , composed of one or several components of NSCLC ( ADC, SQCC, large cell ) with at least 10% of spindle or giant cell carcinoma a clear feature of EMT (epithelial mesenchymal transition). Diagnosis requires resection sample. In undifferentiated areas IHC (TTF1/P40) will identify the NSCLC component and predict mutations. Mutations regionality is predicted and parallels histological components, if ADC (Kras, EGFR,BRAF Met ,ALK rearrangement ) if SQCC FGFR1 amplification. The most frequent MET exon 14 skipping mutation (20%) is heterogeneously distributed. Tumors with SMARC4 mutations described also in adenocarcinoma with obvious EMT transformation and BRG1 negative IH are now considered as pleomorphic carcinoma.

      4.Combined high grade tumors : Combined SCLC small cell Lung Carcinoma (SCLC) and combined Large cell neuroendocrine carcinoma LCNEC account for 20% of each in resected samples ,less readily detectable on small samples ,10% SCLC or LCNEC are sufficient. The genetic alterations are not therapeutically actionable yet, excepted in trials (DLL3) .Occurrence of mutations in the non- neuroendocrine component deserve recognition since it can induce drug resistance and transdifferentiation. Transformation of SCLC to NSCLC has been reported after cytotoxic SCLC chemotherapy (E. Brambilla JCO 1991 ) suggesting intra-tumor heterogeneity or cell plasticity under therapeutic pressure, and a common initiating stem cell between SCLC and NSCLC . Inter-tumor heterogeneity at molecular level is better documented than is intra-tumor heterogeneity, in absence of multi-region sequencing. NOTCH and DDL3 therapeutic targets shows both inter and intra-tumor heterogeneity(7)

      B-Temporal intra-tumor heterogeneity

      Genetic dynamics characterizes tumor progression , well studied during ADC progression from AAH to invasive ADC. The conservation of driver clonal but not subclonal mutations occur in metastases vs primary. Transdifferentiation is a typical exemple of temporal heterogeneity where a tumor treated ,more rarely spontaneously, transforms in another tumor type , SCLC to NSCLC , EGFR mutant adenocarcinoma to SQC or SCLC or pleiomorphic carcinoma with EMT, showing that tumor plasticity conferred by EMT results in temporal intra-tumor heterogeneity with transdifferentition (2)

      C-Spatial intra tumor heterogeneity across histology

      All studies (2-3) establish the concept of trunk and branches in the phylogenetic tree , the clonal trunk mutations are present in all regions (EGFR MET BRAF TP53 ALK … )in 75% of tumors and additional variant subclonal mutations in some but not all regions.

      Lessons :

      -One biopsy may not capture the extent of landscape ITH but trunk mutation are homogeneously distributed whereas branched mutations are heterogeneouly distributed . Cell free DNA overtime better capture genetic landscape.

      References:

      1-A genomic-based classification of human lung tumors, Sci Transl. Med 2013 5;209,153

      2-Xue et al. Evolution from genetics to phenotypes : reinterpretationof NSCLC plasticity,heterogeneity,and drug resistance Protein Cell 2017;8 :178-190

      3-Zhang J et al. Intra-tumor heterogeneity in localized lung adenocarcinoma delineated by multiregion sequencing, Science 2014; 10 ;346 :256-9 . DeBruin E et al . Spatial and temporal diversity in genomic instability processes defines lung cancer evolution Science 2014 346:251 . Jamal Hanjani M et al. Tracking the evolution of NSCLC, NEJM 2017 ;376:2109

      e353dbe42c8654f33588d4da0b517469

    • +

      MS05.04 - Diagnosis and Classification in Biopsies

      14:30 - 14:50  |  Presenting Author(s): Andre Moreira

      • Abstract

      Abstract

      In the last decade, significant progress has been made in the field of thoracic oncology, mostly in the management of patients with non-small cell carcinoma (NSCLC). The determination of subtypes of NSCLC is now directly linked with options of chemotherapy regimens, and further screening for targetable molecular alterations. In the 2015 W.H.O Classification of Lung cancer, the importance of small biopsies has been highlight as well as the use of immunohistochemical stains as an ancillary test to separate adenocarcinoma from squamous cell carcinomas.

      Despite the use of IHC, the classification of NSCLC remain based on histologic features and IHC is recommended in cases with no clear evidence of differentiation between adenocarcinoma and squamous cell carcinoma. It is important to notice that the diagnosis of Large Cell Carcinoma is not recommended in small biopsy specimens. For the histological diagnosis of adenocarcinoma, the following features need to be presence in the biopsy: acinar, papillary, micropapillary and lepidic patterns, and for the histological diagnosis of squamous cell carcinoma, the presence of keratinization and intracellular bridges must be presence. Therefore, in a biopsy with solid growth pattern and no evidence of keratinization, the use of IHC is recommended. The use of IHC improved diagnostic accuracy in the lung carcinoma classification, but the interpretation can be challenging and the pathologists must be aware of many interpretation pitfalls that can involve antibody clones, interpretation of the staining pattern and when to add more stains to the panels.

      When confronted with a tumor of likely lung origin for which the main question is subtyping of adenocarcinoma versus squamous carcinoma, the recommendation is to use a limited panel that includes TTF1 and p40. The use of this panel can classify most tumors and saves tissue for biomarker testing. Napsin-A can be added to the panel in challenging cases but there is no clear evidence that the latter marker is superior to TTF-1. Differences in sensitivity and specificity for napsin-A depends on whether a monoclonal or polyclonal antibody is used. Keratin 7 is not useful to separate adenocarcinoma from Squamous cell carcinoma.

      According to the classification, a NSCLC with immunohistochemical evidence of adenocarcinoma differentiation (any nuclear TTF-1 positivity) should be diagnosed as NSCLC-favor adenocarcinoma, whereas a tumor with evidence of squamous differentiation (strong and diffuse p40 positivity) should be classified as NSCLC, favor squamous cell carcinoma. In tumors with inconclusive immunoprofile, the best diagnostic term is NSCLC-NOS. For tumors with double positivity (TTF1/p40), it is important to remember that double positivity in the same cells does not define an adenosquamous carcinoma. For the diagnosis of adenosquamous carcinoma each marker should be seen in different components/areas of the tumor. Double staining is seen in adenocarcinomas or in NSCLC-Nos that carries mutations similar to adenocarcinoma.

      For double negative tumors (TTF-1 neg/p40 neg) in patients with no other history of malignancy, the diagnosis of NSCLC-NOS can be accepted because approximately 20-30% of adenocarcinoma of the lung are negative for these markers. If clinical history is not known, however, other stains should be added to the panel to rule out metastatic disease. One important marker is keratin, because metastatic melanomas or other non-epithelial tumors may mimic carcinoma, especially in small biopsy samples. For patients with known history of other malignancies, histological comparison with prior tumors should be pursued, targeted IHC with organ specific markers (PAX-8, GATA-3, NKX3.1, etc.) provides strong support to this interpretation, particularly when previous materials are unavailable for review.

      The current WHO classification recommends that neuroendocrine (NE) markers should be performed only when NE morphologic features are present. IHC positivity for NE markers alone is not diagnostic of a NE tumor because positivity for NE markers may be encountered in approximately 10-30% of adenocarcinomas.

      A panel of chromogranin A, synaptophysin and CD56 is the best combination for the diagnosis of NE tumors. There is no consensus on how many markers should be used for the diagnosis of NE, most cases of NE tumors are positive for ≥2 out of the 3 NE markers. Positivity for at least one NE marker is necessary for the diagnosis of Large Cell Neuroendocrine Carcinoma (LCNEC) in association with histological features (chromatin pattern, palisading etc). The diagnosis of LCNEC in a biopsy or cytology specimen is not recommended but can be suggested if a combination of histological features and NE markers positivity is encountered.

      A proliferative marker such as Ki-67 is very useful for the classification of NE tumors in small biopsy specimens, especially in samples with significant crush artifact. In small biopsies Ki-67 stains can separate a low-grade NE tumor (carcinoid) from a high grade tumor such as small cell carcinoma. Ki-67 is not recommended in the classification of typical from atypical carcinoid in excision specimens. In cytology samples, the marker does not work well in alcohol-fixed specimen and can lead to misclassification of NE tumors.

      Suggested reading:

      Sauter JL, Grogg KL, Vrana JA, et al. Young investigator challenge: Validation and optimization of immunohistochemistry protocols for use on cellient cell block specimens. Cancer Cytopathol 2016;124:89-100

      Rekhtman N, Kazi S. Nonspecific reactivity of polyclonal napsin a antibody in mucinous adenocarcinomas of various sites: a word of caution. Arch Pathol Lab Med 2015;139:434-436.

      Rekhtman N, Ang DC, Sima CS, et al. Immunohistochemical algorithm for differentiation of lung adenocarcinoma and squamous cell carcinoma based on large series of whole-tissue sections with validation in small specimens. Mod Pathol 2011;24:1348-1359

      Thunnissen E et al. The use of Immunohistochemistry improves the diagnosis of Small Cell Lung Cancer and its Differential Diagnosis. An International reproducibility Study in a demanding set of cases. J Thorac Oncol. 2017;12:334-46.

      Moreira AL, Travis WD. Histologic Classification and its need for Treatment of Lung Cancer. In Diagnosing Non-Small Cell Lung Cancer in Small Biopsy and Cytology. Edited By Moreira AL and Saqi A. 2014. Springer Science Business Media, New York. Pages 1-14. Doi 10.1007/978-1-4939-1607-8_1.

      Moreira AL, Mino-Kenudson M. Update on Histologic Classification of Non-Small Cell Lung Cancer. Diagnostic Histopathology, 2014. 20: 385-91.

      e353dbe42c8654f33588d4da0b517469

    • +

      MS05.05 - Q&A

      14:50 - 15:00

      • Abstract

      Abstract not provided

  • +

    MS06 - Practical Issues in the Management of Oligometastatic NSCLC

    • Type: Mini Symposium
    • Track: Oligometastatic NSCLC
    • Moderators:
    • +

      MS06.01 - The Role of Radiation in Treating the Mets

      13:30 - 13:45  |  Presenting Author(s): Hak Choy

      • Abstract

      Abstract not provided

    • +

      MS06.02 - The Role of Interventional Pulmonology and Radiology

      13:45 - 14:00  |  Presenting Author(s): Kwun M Fong

      • Abstract

      Abstract

      The Role of Pulmonology and Interventional Radiology in the Management of Oligometastatic NSCLC

      Oligometastatic disease is a difficult management issue when encountered in NSCLC. Often defined as low metastatic burden and limited organ involvement disease, with clinical impact perceived as between truly localised potentially curable disease and a extensive incurable metastatic tumour, therefore with major implication for the patient in terms of treatment, with the potential to pivot the decision making from palliative to curative intent.

      In addition the site(s) of oligometastatic disease is heterogenous with potential to affect any site in the body, both intra and extra- thoracic, thus requiring a truly multidisciplinary approach to its management.

      It is essential to make an accurate diagnosis, to consider synchronous cancers and to rule out more extensive metastatic disease given the differing management strategies required.

      For the Pulmonologist and Interventional Radiologist, for intrathoracic oligometastatic disease, the issues are the diagnostic challenges and potential therapeutics. The use of image guided trans thoracic needle biopsy has long been used effectively by Interventional Radiology for the pathological diagnosis of suspected lesions. With modern endobronchial ultrasound, Pulmonologists are able to contribute more by accessing central and peripheral lung lesions by navigation and guided bronchoscopy e.g. ultrasound guided, electromagnetic navigation and transparenchymal approaches. With the increasing use of molecular techniques to distinguish cancers even of the same morphological appearance, such diagnostic approaches are of increasing clinical utility.

      For intrathoracic therapeutics, local Interventional Radiology ablative techniques such as radiofrequency ablation has long been used, with microwave ablation increasingly popular. In parallel with diagnostics, emerging Pulmonology techniques are also researching the potential application of these techniques applied endobronchially in addition to new modalities such as steam ablation.

      Interventional Radiologists also have a useful role to play in the management of extra-thoracic oligometastatic disease, as techniques such as radiofrequency and microwave ablation can be used such as for liver lesions.

      This session will review the emerging data for the key role of Pulmonology and Interventional Radiology in the Management of Oligometastatic NSCLC.

      e353dbe42c8654f33588d4da0b517469

    • +

      MS06.03 - The Role of Surgical Resection

      14:00 - 14:15  |  Presenting Author(s): Boris Sepesi

      • Abstract

      Abstract

      Lung cancer is an aggressive malignancy with low (< 10%) 5-year survival rates when metastases appear. Historically, surgical resection in metastatic setting has not been beneficial in term of survival; this was especially true when options for effective systemic therapy were limited. It is now recognized that up 50% of patients who presents with metastatic NSCLC have a limited (up to 5) number of metastases. Common sites of metastases include brain, lungs, adrenal glands, bones, or liver.

      In appropriately selected patients, surgical resection in metastatic lung cancer may be considered for the control of primary tumor, regional nodal disease, and metastases present within the same surgical field (example: metastases in ipsilateral ribs, or another lobe in the same lung as primary tumor). Surgical resection of selected extra-thoracic metastases is also possible; however, radiation is more commonly employed in those circumstances.

      The main goal of surgery in oligometastatic setting is to achieve complete disease control. Careful patient selection must account for the extent and the location of primary tumor and nodal disease, physiologic status and reserve of the patient, tumor histology, biomarkers, synchronous or meta-chronous presence of metastases and number of metastases, previous therapies and planned post-surgical systemic or other local therapies.

      Invasive mediastinal staging with either mediastinoscopy or endobronchial ultrasound is of utmost importance prior to proceeding with surgery. Long-term outcomes with positive ipsilateral mediastinal N2 or contralateral N3 lymph nodes are poor as compared to N0 or N1 nodal status; N2 and N3 disease should be considered contraindication for surgical therapy outside of a clinical trial. Lobectomy with mediastinal node dissection is a preferred procedure for primary tumor in oligometastatic circumstances. Pneumonectomy, especially on the right side should be avoided, and its need should be regarded as a relative contra-indication. If brain metastases are present, they should be treated first; other metastatic sites maybe addressed based on symptoms. Commonly, systemic platinum-based chemotherapy has been administered as the 1st line therapy for 3-4 cycles to sort out the cancer biology and response to chemotherapy. Stable or improved disease following chemotherapy makes proceeding with surgical resection of primary tumor and control of oligometastases reasonable. However, this paradigm continues to evolve in the era of immunotherapy and targeted therapy.

      There are a number of advantages to surgical resection of lung cancer in oligometastatic setting, as compared to other local modalities. Complete resection provides the best chance for prolonged local and regional disease control. It hypothetically removes all or at least the majority of cancer clones, it allows for detailed pathological analysis of the tumor and lymph nodes, and it provides adequate tissue for translational studies or personalized experimental treatment approaches. With current minimally invasive approaches such as video assisted or robotic assisted thoracic surgery recovery from surgical treatment of lung cancer can be expeditious. With the use of enhanced recovery pathways, even recovery after thoracotomy mimics recovery after minimally invasive thoracic surgery. Potential disadvantage of surgical therapy in oligometastatic lung cancer is the risk of post-operative cardiac or pulmonary complications, resolution of which may require time and may delay other necessary therapy. This again highlights the importance of patient selection for surgical therapy.

      Surgical Resection of Oligometastatic Lung Cancer in the Era of Targeted Therapy and Immunotherapy.

      Within the last year, immunotherapy has become the frontline treatment modality for metastatic lung cancer. Response rates, durability of responses, and toxicity profiles with immunotherapy are all better than with chemotherapy. Although imperfect, PD-L1 expression and tumor mutation burden serve as predictive biomarkers for immunotherapy. Many studies of combined chemo-immune regimens, or combination immune regimens are underway. These results are encouraging, however, as we improve systemic disease control, loco-regional lung cancer control will become even more important. The role and timing of local treatments following immunotherapy or targeted therapy in oligometastatic setting are currently under study. While there are reports of radiation therapy potentiating immunotherapy, evidence is not conclusive and biomarkers are still under investigation. Surgical therapy and radiation therapy should be considered complementary modalities in oligometastatic lung cancer. If lobectomy is possible for primary disease control in an acceptable-risk surgical candidate, it should not be discarded as an option over radiation. Considering current operative experience following prolonged immunotherapy, it is hypothesized that salvage surgical resections for local recurrence following immune-radio therapy may be difficult and may convert possible routine early lobectomy to an eventual salvage pneumonectomy.

      As the therapeutic options for oligometastastic lung cancer continue to evolve with much improved systemic options, effective local and regional therapy will become more important. Novel surgical techniques and enhanced recovery pathways have decreased the morbidity and sped up patients’ surgical recovery. Surgical resection should therefore continue to play a role in the multi-modality setting for oligometastatic lung cancer in appropriately selected patients.

      References:

      Gomez DR, Blumenschein GR Jr, Lee JJ, Hernandez M, Ye R, Camidge DR, Doebele RC, Skoulidis F, Gaspar LE, Gibbons DL, Karam JA, Kavanagh BD, Tang C, Komaki R, Louie AV, Palma DA, Tsao AS, Sepesi B, William WN, Zhang J, Shi Q, Wang XS, Swisher SG, Heymach JV. Local consolidative therapy versus maintenance therapy or observation for patients with oligometastatic non-small-cell lung cancer without progression after first-line systemic therapy: a multicentre, randomised, controlled, phase 2 study. Lancet Oncol. 2016 Dec;17(12):1672-1682. doi: 10.1016/S1470-2045(16)30532-0. Epub 2016 Oct 24.

      Patrini D, Panagiotopoulos N, Bedetti B, Mitsos S, Crisci R, Solli P, Bertolaccini L, Scarci M. Surgical approach in oligometastatic non-small cell lung cancer. Ann Transl Med. 2018 Mar;6(5):93. doi: 10.21037/atm.2018.02.16. Review.

      Stephens SJ, Moravan MJ, Salama JK. Managing Patients With Oligometastatic Non-Small-Cell Lung Cancer. J Oncol Pract. 2018 Jan;14(1):23-31.

      e353dbe42c8654f33588d4da0b517469

    • +

      MS06.04 - Systemic Therapy for Oligomets: Before, During, or After Local Therapies?

      14:15 - 14:30  |  Presenting Author(s): Ross Camidge

      • Abstract

      Abstract

      Oligometastatic disease reflecting the concept of limited metastatic spread may be discussed in the setting of the sites of disease at diagnosis of stage IV disease. However, an analogous situation may also occur at progression on a therapy controlling all other sites of disease – so called oligoprogressive disease. Both may be considered for local ablative therapies (usually radiation or surgery, but also some other techniques such as radiofrequency ablation). When considering systemic therapy this may be before the local therapy as induction treatment for initial oligometastatic disease; as part of the radical local therapy (for example, combination chemoradiation therapy); or as maintenance therapy after the local therapy. The latter could reflect continuation of a tyrosine kinase inhibitor, pemetrexed or bevacizumab and could also apply to the oligoprogressive scenario.

      The issues to address are the adequacy of assessing oligometastatic disease at the time local therapy is considered; the details that appropriately define oligometastatic or oligoprogressive disease suitable for local therapy; and the potential for a direct interaction with the local therapy.

      e353dbe42c8654f33588d4da0b517469

    • +

      MS06.05 - The Special Case of Brain Metastases: Systemic Therapy, Radiation or Both?

      14:30 - 14:45  |  Presenting Author(s): Laurie Gaspar

      • Abstract

      Abstract not provided

    • +

      MS06.06 - Q&A

      14:45 - 15:00

      • Abstract

      Abstract not provided

  • +

    MS07 - Antibody-Drug Conjugates in Advanced NSCLC

    • Type: Mini Symposium
    • Track: Advanced NSCLC
    • Moderators:
    • +

      MS07.01 - Basic Science

      13:30 - 13:50  |  Presenting Author(s): David E Gerber

      • Abstract

      Abstract

      While in some cases antibodies that specifically bind tumor surface antigens can have therapeutic effects, many unmodified (naked) antibodies lack anti-cancer activity. Conjugation to cytotoxic drugs, radionuclides, or toxins can expand the utility of monoclonal antibodies and improve their potency. In this way, antibodies are employed as a means to target and deliver a toxic payload to the selected tissue. Factors critical to the success of antibody drug conjugates include the target antigen, antibody, linker, and payload. The target antigen should be overexpressed in the tumor relative to healthy tissue. The antibody should have high affinity and avidity for the targeted tumor antigen. The linker should be stable in circulation but must efficiently release the payload after internalization within the cancer cell. The payload should be a highly potent (picomolar range) cytotoxic agent with efficacy against the cancer under treatment, with a reproducible and optimal drug-to-antibody ratio (usually 3-4 drug molecules per antibody molecule), that does not meaningfully compromise the antibody’s biophysical and pharmacokinetic properties. The steps entailed in the mechanism of antibody-drug conjugate action include: (a) antigen binding; (b) antigen-antibody-drug conjugate complex internalization into endosomal vesicles; (c) processing along the endosomal-lysosomal pathway; (d) degradation in this acidic and proteolytic-rich environment; (e) intracellular release of cytotoxic compound. Antibody-drug conjugates may also retain antibody interactions with host immune effector functions, including antibody-dependent cellular cytotoxicity (ADCC). Mechanisms of resistance include (a) reduced target gene expression or presence of increased antigen mutations resulting in reduced target antigen on cell surface; (b) reduced antibody-drug conjugate internalization due to reduced cell surface trafficking or recycling; (c) multidrug resistance ransporter efflux out of the targeted cell.

      In addition to differential expression on tumor cells, antibody-drug conjugate target antigens need to (1) have an extracellular epitope amenable to specific antibody binding, and (2) be able to undergo internalization via receptor-mediated endocytosis into target cells, where the drug payload can be released. High-level expression is advantageous but not necessary. The higher the cell surface density of the target antigen, the more antibody-drug conjugate can be taken up and metabolized by the cell, to release the active cytotoxic agent. Rapid uptake of antibody-drug conjugate not only enhances efficacy, but also reduces opportunity for extracellular payload release. Antigens that are shed should be avoided.

      Antibody-drug conjugate technology has improved considerably in recent years. The employed antibodies are now chimeric, humanized, or fully human, and therefore less immunogenic than early murine antibodies. Efficient linkers (disulfide, dipeptide, hydrazine) have improved stability in circulation as well as better release of active drug within the tumor cell. Finally, highly potent agents with IC50 in the subnanomolar range (maytansine derivatives [DM1, DM4], auristatin [MMAE, MMAF], both of which are microtubule disrupting agents) have replaced conventional chemotherapy drugs such as doxorubicin, vinca alkaloids, and methotrexate.

      Despite the targeted nature of antibody-drug conjugates, toxicities may occur through multiple mechanisms: (1) Non-specific systemic release of the cytotoxic drug payload; (2) Non-selective cytoxicity of target-negative cells in proximity to target-positive cells (bystander effect); (3) internalization of the antibody-drug conjugate by target-negative cells. Prior experience has shown that unrecognized expression of target antigen on healthy tissue can result in profound toxicities: LewisY antigen (gastric mucosaàhemorrhagic gastritis); CD446v6 (deep layers of skinàfatal exfoliation); CA9 (intestinal mucosaàfatal gastrointestinal toxicity). Further complicating the development of antibody-drug conjugates, preclinical models may not adequately predict clinical activity and tolerability. In mouse models, the target antigen is not commonly expressed in host tissues, resulting in misleadingly favorable preclinical activity.

      As evidence of the inherent challenges facing antibody-drug conjugates, only two have been approved by the U.S. Food and Drug Administration: ado-trastuzumab emtansine (anti-HER2; for HER2-positive breast cancer) and brentuximab vedotin (anti-CD30; for Hodgkin’s lymphoma and anaplastic large-cell lymphoma). Target antigens relevant to non-small cell lung cancer under investigation include mesothelin (antetumab ravtansine), folate receptor (mirvetuximab soravtansine), sodium-dependent phosphate transporter (NaPi2b) (lifatuzumab vedotin; DNIB0600A), glycoprotein nonmetastatic B (gpNMB; osteoactivin) (glemtumumab vedotin), and epidermal growth factor receptor (depatuxizumab mafodotin), Perhaps the most developed antibody-drug conjugate in thoracic oncology, rovalpituzumab tersirine (Rova-T) targets delta-like ligand 3 (DLL3), which is highly expressed in approximately two-thirds of small cell lung cancer, but not NSCLC.

      Ultimately, several factors may influence the efficacy of an antibody-drug conjugate, among them target expression, sensitivity to the payload, and aspects of target biology that may impact internalization and intracellular trafficking. Revisiting longstanding principles of cytotoxic therapy may improve the performance of antibody-drug conjugates. These principles include the importance of combination therapies and the recognition that not all tumors are sensitive to microtubule-disrupting agents.

      e353dbe42c8654f33588d4da0b517469

    • +

      MS07.02 - Pharmacology

      13:50 - 14:10  |  Presenting Author(s): Christian Rolfo

      • Abstract

      Abstract not provided

    • +

      MS07.03 - Clinical Data

      14:10 - 14:30  |  Presenting Author(s): Thomas E. Stinchcombe

      • Abstract

      Abstract

      Antibody drug conjugates (ADC’s) are currently being developed for several thoracic malignancies, and preliminary studies have revealed activity in patients with progressive disease after standard therapy for non-small cell lung cancer (NSCLC).

      Sacituzumab govitecan targets the Trop-2 which is overexpressed on many epithelial cancers compared to normal tissues and the antibody is bound to SN-38 which is the active metabolite of irinotecan.1,2 .A single arm phase 2 trial (n=54) investigated in patients with NSCLC with disease progression after at least one line of therapy. The ORR was 17%, median PFS was 5.2 months, and median OS was 9.5 months. The grade ≥ 3 adverse events observed were neutropenia, anemia, diarrhea, nausea, and fatigue. Trop-2 staining was not associated with benefit from therapy and the majority of patients had high Trop-2 staining.

      Telisotuzumab vedotin (ABBV-399) targets the c-Met protein and is conjugated to monomethyl auristatin E, a microtubule inhibitor.3 A phase 1 trial in metastatic solids tumor included an expansion cohort of in patients with NSCLC who were c-Met + (defined as IHC H-score of ≥ 150). An objective response was observed in three of 16 patients (18.8%) with c-Met positive NSCLC, and at week 12 six of 16 (37.5%) experienced disease control The treatment related adverse events observed (at all dose levels) were fatigue, nausea, neuropathy, decreased appetite, vomiting, and hypoalbuminemia. No responses were observed among patients with c-Met negative tumors.

      Ado-trastuzumab emtansine (T-DM1) is an ADC that uses the trastuzumab monoclonal antibody which targets the HER2 protein and delivers the maytansinoid antimicrotubule agent DM1. A single arm phase 2 trial investigated this agent in patients with HER2 positive NSCLC (defined as IHC score of 3+, IHC score or 2+ and FISH positivity, or HER2 mutation).4 Of the 15 patients enrolled 1 patient experienced a partial response (6.7%), the median PFS was 2.0 months, and median OS was 10.9 months. This study was terminated at the interim analysis since it did not meet the predefined efficacy criteria to proceed to the second stage. A phase 2 basket trial investigated T-DM1 in patients with HER2 mutant or amplified NSCLC.5 In the HER2 mutant cohort the ORR was 44% (8 of 18), median PFS was 5 months, and median OS was 11 months. In the HER amplified cohort the ORR was 50% (3 of 6), and the median PFS was 6 months, and the median OS was 12 months. The HER2 mutant cohort met the study’s primary end-point and the HER2 amplified cohort was expanded as per the Simon two-stage design. A phase 2 study investigated T-DM1 in patients with IHC 2+ or 3+ NSCLC, and each cohort was analyzed independently.6 Among the 29 patients enrolled on the IHC 2+ cohort no responses were observed, and the ORR in the IHC 3+ cohort was 20% (4 of 20). In the IHC 2+ cohort the median PFS and OS was 2.6 and 12.2 months respectively, and in the IHC 3+ cohort the median PFS and OS was 2.7 and 12.1 months, respectively. The grade ≥ 3 adverse events observed have been neutropenia, thrombocytopenia, anemia, fatigue, dyspnea, and hepatotoxicity, and have been consistent with the adverse events observed in breast cancer.

      These studies have revealed preliminary activity of ADC’s in patients NSCLC with disease progression after standard therapies. The rate of adverse events to the ADC’s are manageable. Refinement in the biomarker selection of patients for these therapies may improve the efficacy.

      1. Heist RS, Guarino MJ, Masters G, et al. Therapy of Advanced Non-Small-Cell Lung Cancer With an SN-38-Anti-Trop-2 Drug Conjugate, Sacituzumab Govitecan. J Clin Oncol. 2017;35(24):2790-2797.

      2. Gray JE, Heist RS, Starodub AN, et al. Therapy of Small Cell Lung Cancer (SCLC) with a Topoisomerase-I-inhibiting Antibody-Drug Conjugate (ADC) Targeting Trop-2, Sacituzumab Govitecan. Clin Cancer Res. 2017;23(19):5711-5719.

      3. Angevin E, Kelly K, Heist R, et al. First-in-human phase 1, dose-escalation and -expansion study of ABBV-399, an antibody-drug conjugate (ADC) targeting c-Met, in patients (pts) with advanced solid tumors. Annals of Oncology. 2016;27(suppl_6):371P-371P.

      4. Hotta K, Aoe K, Kozuki T, et al. A Phase II Study of Trastuzumab Emtansine in HER2-Positive Non-Small Cell Lung Cancer. J Thorac Oncol. 2018;13(2):273-279.

      5. Li BT, Shen R, Buonocore D, et al. Phase 2 Basket Trial of Ado-Trastuzumab Emtansine in Patients with HER2 Mutant or Amplified Lung Cancers. WCLC 2017: 18th World Conference on Lung Cancer in Yokohama, Japan. October 15-18. 2017.

      6. Stinchcombe T, Stahel RA, Bubendorf L, et al. Efficacy, safety, and biomarker results of trastuzumab emtansine (T-DM1) in patients (pts) with previously treated HER2-overexpressing locally advanced or metastatic non-small cell lung cancer (mNSCLC). Journal of Clinical Oncology. 2017;35(15_suppl):8509-8509.

      e353dbe42c8654f33588d4da0b517469

    • +

      MS07.04 - Future Directions

      14:30 - 14:50  |  Presenting Author(s): Silvia Novello  |  Author(s): Annapaola Mariniello

      • Abstract

      Abstract

      Despite the success achieved by molecularly targeted agents in oncogene-addicted NSCLC over the past decade, patients inevitably develop acquired resistance with subsequent disease progression.

      Moreover, most lung tumors still do not present actionable targets derived by driver genetic alterations.

      To circumvent these issues, alternative strategies to target tumor cells have been studied.

      Antibody drugs conjugates (ADC) represent a novel class of anti-cancer treatment designed with the aim to combine the cytotoxic activity of chemotherapeutics with the selectivity of target agents, thus increasing the therapeutic interval of available compounds.

      The paradigm of ADCs is to convey cytotoxicity selectively into tumor cells, by way of a monoclonal antibody (MoAb) bound to cytotoxic chemicals via synthetic linkers. This complex is able to recognize tumor antigens which serve as Trojan horse for ADC internalization into cancer cell.

      To date, two ADCs have entered clinical practice, brentuximab vedotin for hematological malignancies and trastuzumab emtansine in advanced breast cancer overexpressing human epidermal growth factor receptor 2 (HER2) [1].

      In the field of thoracic malignancies, ADC research has focused on rarer “orphan” diseases, for which neither target therapies nor valuable second-line treatment options are available. In fact, the two ADCs currently underway in most advanced phases of clinical trials are rovalpituzumab tesirine for SCLC and anetumab emtansine for unresectable mesothelioma; however, with suboptimal early results [2;3].

      In NSCLC, no ADC has been experimented yet in phase III trials. However, several basket trials are currently evaluating anti-cancer activity and safety of ADCs mostly targeting HER2, Trop2 and NaPi2b, among others [4-6].

      In NSCLC treatment, as well as in other malignancies, ADC development has been limited by several factors.

      First, to realize a suitable ADC numerous criteria must be fulfilled [7].

      About the cytotoxin (also known as warhead), this must be a small molecule showing high potency, relative hydrophilicity and a lack of susceptibility to P-glycoprotein, which is a very common resistance mechanism for ADCs. Linkers connecting the MoAb to the warhead must be stable in blood, but they are also required to be able to release the drug inside the target cancer cells.

      Apart from these technical issues, which are underway of further improvement, the major unaddressed concern for ADCs in clinics may be finding the optimal antigen target. Cancer antigen must be transmembrane and able to internalize into the target cell after MoAb binding. Importantly, the ideal antigen should be tumor specific or at least with a minimal expression on healthy tissues, to reduce off-target toxicity. However, the type of non-cancerous cells on which the target antigen is exposed may also play a role in predicting ADCs toxicity. Because of the cell cycle-dependent mechanism of action, non-highly proliferating cells may be less vulnerable to the ADC cytotoxic damage. [8]

      Despite these limitations, the fact that the tumor target antigen is not required to drive tumor growth, constitutes an appealing advantage for ADCs application in malignancies lacking acknowledged driver mutations. In the context of thoracic cancers, squamous lung carcinoma or less differentiated entities would be amenable for greater research efforts in finding an adequate ADC antigen target.

      A further barrier to the design of viable ADCs is the unavailability of a reproducible, standardized and economic technique to identify and validate valuable target antigens. So far, the techniques used are immunohistochemistry or - more recently - functional genomic mRNA profiling, with the latter failing to localize the predicted protein into cancer cells [1].

      In NSCLC setting, concern is raised by the ruthless competition with the numerous systemic treatments already available or on study. A smart approach to find a place for ADCs in the NSCLC treatment algorithm, may be to explore combination strategies with drugs already on market. In the era of immune-oncology we are living, great attention is dedicated to combination strategies with immune checkpoint inhibitors (ICI). ADCs in most advanced phases of clinical experimentation are currently being tested in association with ICIs. Notably, a phase I/II trial of anetumab plus atezolizumab in advanced NSCLC patients has already been planned [9]. The supporting rationale is based on the demonstrated immunogenic cell death properties displayed by ADCs. Moreover, additional ADC immunogenicity is warranted by their antibody-dependent cell-mediated cytotoxicity potential. [10]

      A last and unsolved question regards tumor heterogeneity, which makes it unlikely to eliminate a whole neoplasm by using a single ADC target. As a future perspective, we could envision extensive tumor sequencing that allows for efficient antigen prediction and for the design of tailored ADCs targeting, contemporarily, multiple tumor specific antigens.

      Long and winding is the path for the ADC entry in NSCLC clinical practice. However, the rapidly evolving landscape in tumor sequencing and proteomics along with the clinical research experiences, may pave the way to find the most appropriate setting for ADCs in terms of manageable side effects and NSCLC patients’ selection.

      1. Moek KL, de Groot DJA, de Vries EGE et al. The antibody–drug conjugate target landscape across a broad range of tumour types. Ann Oncol. 2017 Dec 1;28(12):3083-3091.

      2. NCT02610140.

      3. NCT03061812.

      4. Hamilton EP, Barve MA, Bardia A et al. Phase 1 dose escalation of XMT-1522, a novel HER2-targeting antibody-drug conjugate, in patients with HER2-expressing breast, lung and gastric tumors. J Clin Oncol 36, 2018 (suppl; abstr 2546).

      5. Sands JM, Shimizu T, Garon EB et al. First-in-human phase 1 study of DS-1062a in patients with advanced solid tumors. J Clin Oncol 36, 2018 suppl; abstr TPS2605.

      6. NCT03319628.

      7. Beck A, Goetsch L, Dumontet C et al. Strategies and challenges for the next generation of antibody-drug conjugates. Nat Rev Drug Discov. 2017 May;16(5):315-337.

      8. Damelin M, Zhong W, Myers J, Sapra P. Evolving Strategies for Target Selection for Antibody-Drug Conjugates. Pharm Res. 2015 Nov;32(11):3494-507.

      9. NCT03455556.

      10. Gerber, HP, Sapra P, Loganzo F et al. Combining antibody–drug conjugates and immune- mediated cancer therapy: what to expect? Biochem Pharmacol 2016. 102, 1–6.

      e353dbe42c8654f33588d4da0b517469

    • +

      MS07.05 - Q&A

      14:50 - 15:00

      • Abstract

      Abstract not provided

  • +

    MS08 - Lung Cancer in the Real World

    • Type: Mini Symposium
    • Track: Treatment in the Real World - Support, Survivorship, Systems Research
    • Moderators:
    • +

      MS08.01 - How Can Real World Data Improve Clinical Evidence Generation and Impact Regulatory Bodies - European Perspective

      13:30 - 13:45  |  Presenting Author(s): Yolande Lievens

      • Abstract

      Abstract

      Health technology assessment (HTA) evaluates the efficacy and effectiveness of new interventions, integrates these data with the costs to define efficiency and addresses future availability and distribution. As such, HTA focuses on accessibility, affordability and equity. Economic evaluations are central in this concept: weighing costs and effects, they support evidence-based decisions on reimbursement, thus endorsing the introduction of innovative healthcare interventions in daily practice.

      After more than 20 years of voluntary cooperation on HTA in Europe, more than 50 HTA bodies are currently operating in the European Union (EU), be it still fragmented with different systems, different procedures and different requirements regarding the type of clinical evidence. Efficacy, the outcome a new intervention provides in the well-defined circumstances of a randomised controlled trial (RCT) and typically the input used to derive cost-effectiveness evidence, may not provide the best insight into the impact of an intervention in daily clinical care. Moreover, in contrast to what is the case for pharmaceuticals, it is much more difficult to perform RCTs on radiotherapy or surgery, especially when it comes to evaluating the incremental evolution typical for new techniques and technologies, or the long-term benefits anticipated to follow more accurate treatment delivery. This has resulted in different regulatory systems for systemic and non-systemic treatment strategies, with attempts to come to a more homogenised HTA approach in the EU having so far failed.

      Real-world data are gradually expanding their role in the evidence generation of lung cancer radiotherapy. A first step moving away from RCTs is to adopt a more pragmatic approach to evidence generation, as is the case in the OligoCare project. This joint ESTRO-EORTC initiative is evaluating the outcome of oligometastatic patients, amongst others from primary lung cancer, treated with radical radiotherapy in a large prospective cohort study. Various patterns of care studies have generated clinical evidence on the uptake of new treatment techniques such as SBRT (Stereotactic Body Radiotherapy), on the value of multimodality treatments for locally-advanced non-small cell lung cancer (NSCLC) beyond the context of RCTs, or have allowed to develop prediction models to support shared decision-making. Furthermore, the wealth of data available in cancer registries and other nation-wide databases can be leveraged to learn more about the quality of care, actual access to different treatment strategies, geographical and institutional variations and their potential impact on lung cancer survival. Real-life data can also be used to generate the cost information necessary to perform cost-effectiveness evaluations. Besides the more frequent approach to derive this evidence from reimbursement data, actual resource costs can also be computed in daily practice. One such initiative was undertaken in Belgium, using Time-Driven Activity-Based Costing to compute real-life costs of radiotherapy, and more specifically of innovative radiotherapy techniques such as SBRT.

      Whereas all these examples provide interesting insight into the clinical and financial consequences of access to standard-of-care and innovative lung cancer radiotherapy, there is a dearth of information on how this evidence defines policy. One interesting approach to change practice is using coverage with evidence generation to provide early access to radiotherapy innovations, while stimulating the further generation of data. Such a programme of provisional financing has been set-up in Belgium to generate evidence from daily practice on SBRT for primary tumours, most typically early-stage NSCLC, and for oligometastatic disease. While this programme allowed radiation oncology centres to develop and provide SBRT without being financially penalised, it also generated the reassuring clinical evidence that will soon lead to the inclusion of SBRT in the formal national radiotherapy reimbursement system.

      e353dbe42c8654f33588d4da0b517469

    • +

      MS08.02 - How Can Real World Data Improve Clinical Evidence Generation and Impact Regulatory Bodies – US Perspective

      13:45 - 14:00  |  Presenting Author(s): Gideon Blumenthal

      • Abstract

      Abstract not provided

    • +

      MS08.03 - Sources of Real World Data: Research Designs, Statistical Modelling and Quality Assurance Requirements

      14:00 - 14:15  |  Presenting Author(s): Mary W. Redman

      • Abstract

      Abstract not provided

    • +

      MS08.04 - The ASCO Perspective

      14:15 - 14:30  |  Presenting Author(s): Bruce E Johnson

      • Abstract

      Abstract not provided

    • +

      MS08.05 - The ESTRO Perspective

      14:30 - 14:45  |  Presenting Author(s): Umberto Ricardi

      • Abstract

      Abstract not provided

    • +

      MS08.06 - Discussion

      14:45 - 15:00

      • Abstract

      Abstract not provided

  • +

    OA05 - Clinical Trials in IO

    • Type: Oral Abstract Session
    • Track: Advanced NSCLC
    • Moderators:
    • +

      OA05.01 - Efficacy/Safety of Entinostat (ENT) and Pembrolizumab (PEMBRO) in NSCLC Patients Previously Treated with Anti-PD-(L)1 Therapy

      13:30 - 13:40  |  Presenting Author(s): Matthew D. Hellmann  |  Author(s): Pasi A Jänne, Mateusz Opyrchal, Navid Hafez, Luis E Raez, Dmitry Gabrilovich, Fang Wang, Peter Ordentlich, Susan Brouwer, Serap Sankoh, Emmett Schmidt, Michael L Meyers, Suresh S. Ramalingam

      • Abstract

      Background

      Treatment options are limited for lung cancer patients whose disease has progressed on anti-PD-(L)1 therapy. HDAC inhibitors may synergize with PD-(L)1 inhibition to overcome resistance. We report the interim results of a Phase 2 trial of entinostat (ENT), a class I selective histone deacetylase (HDAC) inhibitor, plus pembrolizumab (PEMBRO) in patients with NSCLC previously treated with anti-PD-(L)1 therapy.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      ENCORE-601 is an open-label study evaluating the combination of ENT + PEMBRO in patients with recurrent or metastatic NSCLC and prior progression on anti-PD-1/PD-L1 therapy. Patients were eligible irrespective of histology or baseline PD-L1 expression. Patients were treated with ENT 5 mg PO weekly and PEMBRO 200 mg IV Q3W. The primary endpoint was ORR as assessed by irRECIST. Tumor biopsies and blood samples for immune correlates were taken prior to and during treatment in a subset of patients. A total of 70 patients will be enrolled.

      4c3880bb027f159e801041b1021e88e8 Result

      Of 57 patients with anti-PD-(L)1 resistant/refractory NSCLC, the confirmed objective response rate with ENT + PEMBRO was 11% (6 of 57, 95% CI: 4-21%). Of 49 patients with post-baseline tumor measurements, 47% had at least some reduction in tumor. Anti-PD-(L)1 therapy was the most recent line of therapy in 38 of 57 patients, and the median time from last dose of prior anti-PD-(L)1 to study entry was 67 days. The median duration of response with ENT + PEMBRO was 5 months, with the longest over 14 months. Of the 6 responders, four were PD-L1 negative at study entry. Response was associated with a higher median baseline level of peripheral classical monocytes (CD14+CD16-HLA-DRhi) with 16.9% of total live PBMCs in responders (n=6) compared to 8.2% in non-responders (n=45). 5 patients (8.8%) experienced Grade 3/4 related irAEs (2 events each of pneumonitis and colitis, 1 event of hyperthyroidism). In addition, 19 patients (33.3%) experienced other Grade 3/4 related AEs with only fatigue, anemia, hypophosphatemia, and hyponatremia occurring in more than 1 patient. Additional correlative analyses to identify biomarkers of response, including whole exome sequencing and RNAseq, are in progress.

      8eea62084ca7e541d918e823422bd82e Conclusion

      ENT + PEMBRO demonstrated anti-tumor activity and acceptable safety in patients with NSCLC who have progressed on prior PD-(L)1 blockade. Ongoing analysis of immune correlates may identify strategies for effective patient selection.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      OA05.02 - Epacadostat Plus Pembrolizumab in Patients with Non-Small Cell Lung Cancer: Phase 1/2 Results from ECHO-202/KEYNOTE-037

      13:40 - 13:50  |  Presenting Author(s): Liza Villaruz  |  Author(s): Bryan J Schneider, Todd M. Bauer, Alexander Spira, Gina D'Amato, Jeffery Wasser, Ani Balmanoukian, Primo Lara, Anthony Olszanski, Thomas Gajewski, Sandip Patel, Ahmad Tarhini, Joshua Michael Bauml, Emmett Schmidt, Jill Bowman, Jeannie Daniel, Sherry Owens, Tara C Mitchell

      • Abstract

      Background

      Epacadostat (E) is a potent, highly selective inhibitor of the indoleamine 2, 3-dioxygenase 1 (IDO1) enzyme. ECHO-202/KEYNOTE-037 is an open-label, phase 1/2 study of E + pembrolizumab (P) in patients (pts) with advanced tumors (NCT02178722). We report updated efficacy and safety data for the phase 1 and 2 non-small cell lung cancer (NSCLC) cohort as of 8 Jan 2018 data cutoff.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Adult pts with prior platinum-based therapy (tx), no prior immune checkpoint inhibitors, and those intolerant to EGFR-targeted therapy were eligible. Pts could receive E (25, 50, 100, or 300 mg twice daily [BID]) + P (2 mg/kg or 200 mg every 3 weeks [Q3W]) during phase 1; maximum tolerated dose was not achieved. E (100 mg BID) + P (200 mg Q3W) tx doses were selected for phase 2 evaluation. Efficacy was assessed by RECIST v1.1 criteria. Programmed death-ligand 1 (PD-L1) tumor proportion score (TPS) was evaluated using the 22C3 pharmDx assay. IDO1 status was measured by RNAScope. Safety was assessed in pts receiving ≥1 E + P dose.

      4c3880bb027f159e801041b1021e88e8 Result

      A total of 70 pts (phase 1, n=12; phase 2, n=58) were evaluated. Median age was 63 years and 57% of pts were female. EGFR mutation status (+/-/not available) was 10%/71%/19%; KRAS status was 19%/26%/56%, respectively. Most pts had adenocarcinoma (70%), a history of smoking (76%), and 0–1 prior lines of tx (64%). PD-L1 TPS results were available in 55/70 pts; IDO1 status (+/-) in 41/70 pts. Overall, the objective response rate (ORR; complete response [CR] + partial response [PR]) and disease control rate (CR + PR + stable disease [SD]) were 29% (20/70; 17 PR) and 50% (35/70; 15 SD), respectively. ORR was 22% (2/9) in pts with TPS ≥50% and 0–1 prior tx; for pts with TPS <50% and 0–1 prior tx, ORR was 22% (5/23). ORR for pts with IDO1+ and IDO1- tumors were 24% (6/25) and 19% (3/16), respectively. At data cutoff, 10/20 responses were ongoing. Median progression-free survival (PFS) was 4.0 mo (90% confidence interval, 2.1–6.2 mo); PFS rates at 6, 12, and 18 mo were 43% (33%–53%), 31% (22%–40%), and 17% (10%–26%), respectively. Grade ≥3 treatment-related adverse events (TRAEs) were reported in 27% of pts; 3 pts discontinued due to TRAEs; and no TRAEs led to death.

      8eea62084ca7e541d918e823422bd82e Conclusion

      The E + P combination was generally well tolerated and associated with promising responses in pts with NSCLC.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      OA05.03 - Safety and Clinical Activity of Adoptive Cell Transfer Using Tumor Infiltrating Lymphocytes (TIL) Combined with Nivolumab in NSCLC

      13:50 - 14:00  |  Presenting Author(s): Ben C. Creelan  |  Author(s): Jamie K Teer, Eric M Toloza, John E Mullinax, Ana M Landin, Jhanelle Elaine Gray, Tawee T Tanvetyanon, Matthew C Taddeo, David R Noyes, Linda L Kelley, Bin Fang, John M Koomen, Amod A Sarnaik, Sungjune Kim, Eric B. Haura, Scott J Antonia

      • Abstract

      Background

      Adoptive transfer of tumor infiltrating lymphocytes (TIL) can cause durable regression by recognition of neoantigens unique to the patient. NSCLC TIL has synergistic preclinical activity with nivolumab, and we hypothesized it may induce remissions in anti-PD1-refractory patients. We initiated a phase I trial with the primary objective to characterize the safety and preliminary activity of the combination.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Metastases from patients with Stage 4 NSCLC were resected, morselized, cultured, and tested for autologous reactivity. Reactive TIL fragments were pooled and cryopreserved. Patients received nivolumab over 8 weeks. Patients with progressive disease (PD) proceeded to lymphodepletion cyclophosphamide/fludarabine (Cy/Flu), TIL, and IL-2. Tumor whole exome sequencing, transcriptomics, and LC-MS/MS peptide sequencing was performed. TCR-Vß rearrangements were analyzed from tumor, TIL, and pre-/post-infusion peripheral lymphocytes.

      4c3880bb027f159e801041b1021e88e8 Result

      Of 14 patients enrolled to date, 13 had successful ex vivo TIL expansion from resected metastases. TIL had high proliferative capacity, expanding to median 81 billion CD3+ cells infused per patient (range 27–138 billion) and median 27% of fragments were autologously reactive (range 0-67%). Demographics: median age 54 (range 44-74), median TMB 4 mutations/MB (range 0.9–25), median PD-L1 proportion-score 0% (range 0–100%), and 4 had LKB1 allelic inactivation. Predicted neoantigens correlated with variants on proteomic sequencing. Outcomes: 9 patients had confirmed PD on nivolumab, and proceeded to receive Cy/Flu/TIL/IL-2. No unexpected serious adverse reactions (SUSARs) were identified. Of these 9 patients, 7 had reduction in sum of target lesions at Day+28 CT scan (Figure 1). Peripheral lymphocytes expanded at Days 2-7 in the majority of patients. In patients tested to date, TIL clonotypes persisted through Day+100, and CCR7+CD95+CD45RA+ stem cell-like memory (TSCM) cells were increased at post-infusion timepoints.

      abstract figure.png

      8eea62084ca7e541d918e823422bd82e Conclusion

      Adoptive cell transfer with TIL and nivolumab for NSCLC had acceptable toxicity and preliminary activity in this ongoing trial.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      OA05.04 - Discussant - OA 05.01, OA 05.02, OA 05.03

      14:00 - 14:15  |  Presenting Author(s): Scott Owen  |  Author(s): Martin J. Edelman

      • Abstract

      Abstract not provided

    • +

      OA05.05 - Avelumab vs Docetaxel for Previously Treated Advanced NSCLC: Primary Analysis of the Phase 3 JAVELIN Lung 200 Trial

      14:15 - 14:25  |  Presenting Author(s): Fabrice Barlesi  |  Author(s): Johan F. Vansteenkiste, David R. Spigel, Hidenobu Ishii, Marina Chiara Garassino, Filippo De Marinis, Mustafa Özgüroğlu, Aleksandra Szczesna, Andreas Polychronis, Ruchan Uslu, Maciej Krzakowski, Jong-Seok Lee, Luana Calabro, Osvaldo Arén Frontera, Barbara Ellers-Lenz, Marcis Bajars, Mary Ruisi, Keunchil Park

      • Abstract

      Background

      Avelumab is a human anti–PD-L1 IgG1 monoclonal antibody that is an approved treatment for metastatic Merkel cell carcinoma (various regions) and platinum-treated advanced urothelial carcinoma (US). We report findings from a global, open-label, phase 3 trial of avelumab vs docetaxel in patients with advanced NSCLC after platinum failure (NCT02395172).

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Patients with stage IIIB/IV or recurrent NSCLC with disease progression after platinum doublet therapy were randomized 1:1 to avelumab 10 mg/kg Q2W or docetaxel 75 mg/m2 Q3W, stratified by PD-L1 status (PD-L1+/PD-L1−) and histology (squamous/nonsquamous). The primary endpoint was overall survival (OS) in the PD-L1+ population (expression on ≥1% of tumor cells, assessed using the PD-L1 IHC 73-10 assay).

      4c3880bb027f159e801041b1021e88e8 Result

      Between April 2015 and February 2017, 792 patients were randomized to receive avelumab or docetaxel, including 264 and 265 with PD-L1+ tumors, respectively; 0.8% vs 7.5% did not receive study treatment. Median follow-up in the avelumab and docetaxel arms was 18.9 and 17.8 months; 15.5% vs 1.5% remained on treatment at data cutoff (November 22, 2017). In the avelumab and docetaxel arms, 39.8% vs 47.5% received subsequent anticancer therapy after discontinuation, including checkpoint inhibitors in 5.7% vs 26.4%, respectively. In the PD-L1+ population, median OS in the avelumab and docetaxel arms was 11.4 vs 10.3 months (hazard ratio [HR], 0.90 [96% CI, 0.72-1.12]; P=0.1627, 1-sided). Pre-planned exploratory analyses based on higher PD-L1 cutoffs showed increased OS with avelumab vs docetaxel, including PD-L1-high (≥80% cutoff, 29% of patients; 17.1 vs 9.3 months; HR, 0.59 [95% CI, 0.42-0.83]; P=.0022, 2-sided) and PD-L1-medium/high (≥50% cutoff, 40% of patients; 13.6 vs 9.2 months; HR, 0.67 [95% CI, 0.51-0.89]; P=0.0052, 2-sided) subgroups. In the PD-L1+ population (≥1% cutoff), ORR was 18.9% vs 11.7% (odds ratio, 1.76 [95% CI, 1.08-2.86]; P=0.0105, 1-sided); median duration of response was not reached with avelumab (95% CI, 9.9-not estimable [NE]) vs 6.9 months with docetaxel (95% CI, 3.5-NE). Overall rates of treatment-related adverse events (AEs) were lower with avelumab than docetaxel, including all grades (63.9% vs 85.8%) and grade ≥3 (9.9% vs 49.3%). Immune-related AEs occurred in 16.5% of avelumab-treated patients (grade ≥3 in 2.8%).

      8eea62084ca7e541d918e823422bd82e Conclusion

      Avelumab showed increasing clinical activity in patients who had platinum-treated NSCLC with higher tumor PD-L1 expression; however, the trial did not meet its primary objective of improving OS vs docetaxel in PD-L1+ tumors (≥1% cutoff). OS findings may have been confounded by subsequent checkpoint inhibitor therapy in the docetaxel arm.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      OA05.06 - CheckMate 227: Patient-Reported Outcomes of First-Line Nivolumab + Ipilimumab in High Tumor Mutational Burden Advanced NSCLC

      14:25 - 14:35  |  Presenting Author(s): Julie R. Brahmer  |  Author(s): Michael Schenker, Ki Hyeong Lee, Mariano Provencio, Makoto Nishio, Krzysztof Lesniewski-Kmak, Randeep Sangha, Samreen Ahmed, Judith Raimbourg, Kynan Feeney, Romain Corre, Fabio Andre Franke, Eduardo Richardet, John R. Penrod, Yong Yuan, Faith Nathan, Prabhu Bhagavatheeswaran, Michael De Rosa, Fiona Taylor, Rachael Lawrance, Martin Reck

      • Abstract

      Background

      The randomized, open-label, multipart phase 3 study CheckMate 227 (NCT02477826) demonstrated a significant progression-free survival benefit (co-primary endpoint) with first-line nivolumab+ipilimumab versus histology-based, platinum-doublet chemotherapy in patients with advanced non-small cell lung cancer (NSCLC) and high tumor mutational burden (TMB; ≥10 mutations/Mb). Patient-reported outcomes (PROs) for this population within CheckMate 227 are presented.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Eligible chemotherapy-naïve patients had stage IV or recurrent NSCLC, ECOG performance status 0−1, and no known sensitizing EGFR/ALK alterations. PROs were assessed as an exploratory endpoint; this analysis included patients with high TMB randomized to nivolumab+ipilimumab or chemotherapy; specific outcomes included proportion of patients with disease-related symptom deterioration by 12 weeks and time to deterioration in symptoms (by Lung Cancer Symptom Scale [LCSS] Average Symptom Burden Index [ASBI]), and assessment of quality of life and overall health status (by EuroQoL-5 Dimension [EQ-5D] utility index [UI] and visual analog scale [VAS]). PROs were evaluated each cycle (Q2W, nivolumab+ipilimumab; Q3W, chemotherapy) for the first 6 months, every 6 weeks thereafter during treatment, and at follow-up visits 1/2. EQ-5D was also assessed during survival follow-up.

      4c3880bb027f159e801041b1021e88e8 Result

      PRO completion rates were ~90% at baseline and >80% for nearly all on-treatment assessments. Among patients with high TMB, fewer patients in the nivolumab+ipilimumab (n=139) versus chemotherapy (n=160) groups reported symptom deterioration by week 12, irrespective of whether they were still on therapy or had discontinued (22.3% versus 35.0%; absolute risk reduction: 12.7% [95% CI: 2.4–22.5]). Time to first deterioration (TTD) using common assessment time points (on/off treatment) was delayed with nivolumab+ipilimumab versus chemotherapy for the LCSS ASBI (hazard ratio [HR]: 0.40; 95% CI: 0.26–0.63) and 3-Item Global Index (3-IGI; HR: 0.56; 95% CI: 0.38–0.82). The estimated benefit in TTD generally favored nivolumab+ipilimumab for individual symptoms in the ASBI and each item in the 3-IGI (HRs: 0.48–0.74), except for hemoptysis (HR: 1.20), which exhibited very low burden; an advantage for nivolumab+ipilimumab was also seen in the EQ-5D VAS (HR: 0.62; 95% CI: 0.42–0.92) and UI (HR: 0.50; 95% CI: 0.34–0.73). Mean changes from baseline with nivolumab+ipilimumab showed early and clinically meaningful improvements in LCSS ASBI/3-IGI and EQ-5D VAS/UI; for patients treated with chemotherapy, symptoms and quality of life remained stable (LCSS ASBI/3-IGI, EQ-5D UI) or improved following completion of chemotherapy (EQ-5D VAS).

      8eea62084ca7e541d918e823422bd82e Conclusion

      Nivolumab+ipilimumab demonstrated early and sustained improvements in health-related quality of life versus chemotherapy in patients with advanced NSCLC and high TMB.

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      OA05.07 - IMpower132: PFS and Safety Results with 1L Atezolizumab + Carboplatin/Cisplatin + Pemetrexed in Stage IV Non-Squamous NSCLC

      14:35 - 14:45  |  Presenting Author(s): Vassiliki A Papadimitrakopoulou  |  Author(s): Manuel Cobo, Rodolfo Bordoni, Pascale Dubray-Longeras, Zsuzsanna Szalai, Grigoriy Ursol, Silvia Novello, Francisco Orlandi, Simon Ball, Jerome Goldschmidt Jr., Rachel E Sanborn, Tien Hoang, Diana Mendus, Yu Deng, Marcin Kowanetz, Xiaohui Wen, Wei Lin, Alan Sandler, Makoto Nishio

      • Abstract

      Background

      In addition to the combination of atezolizumab (anti–PD-L1)+platinum+taxane±bevacizumab, non-squamous NSCLC patients may derive benefit from adding atezolizumab to platinum+pemetrexed. The randomized Phase III IMpower132 study (NCT02657434) evaluated first-line pemetrexed+carboplatin or cisplatin±atezolizumab in patients with stage IV non-squamous NSCLC without EGFR or ALK driver mutations.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      Patients were randomized 1:1 to receive 4 or 6 cycles of carboplatin AUC 6 mg/mL/min or cisplatin 75 mg/m2 + pemetrexed 500 mg/m2 Q3W alone (arm PP) or in combination with atezolizumab 1200 mg Q3W (arm APP), followed by pemetrexed (arm PP) or atezolizumab + pemetrexed (arm APP) maintenance. Atezolizumab could be continued beyond disease progression per protocol guidelines. PFS and OS were co-primary endpoints. Efficacy by PD-L1 expression was an exploratory endpoint. Here, we present investigator-assessed PFS per RECIST v1.1 (final analysis), OS (interim analysis) and safety data.

      4c3880bb027f159e801041b1021e88e8 Result

      292 pts were enrolled in arm APP and 286 pts in arm PP. 43% and 40% in arms APP and PP, respectively, had ECOG PS 0. At the data cutoff (22 May 2018), median follow-up was 14.8 mo overall. PFS analysis showed a statistically significant improvement between arm APP vs PP (median PFS 7.6 vs 5.2 mo; HR=0.596; 95% CI: 0.494, 0.719; P<0.0001). Interim analysis showed a numerical but non-statistically significant OS improvement (Table). Efficacy results were consistent across all key clinical subgroups. Grade 3-4 TRAEs occurred in 53.6% of patients (APP) vs 39.1% (PP). 7.2% of patients (APP) vs 5.1% (PP) had a Grade 5 AE, 3.8% (APP) vs 2.6% (PP) were treatment related. 48.5% of patients (APP) vs 38.0% (PP) experienced AEs of special interest specific to atezolizumab.

      8eea62084ca7e541d918e823422bd82e Conclusion

      IMpower132 met its co-primary endpoint of investigator-assessed PFS in the ITT. Interim OS data (not fully mature) showed numerical improvement. Atezolizumab+pemetrexed+carboplatin or cisplatin was well tolerated, and no new safety signals were identified.

      Table. IMpower132 Efficacy Analyses

      APP Arm
      (atezolizumab+pemetrexed+ carboplatin or cisplatin)
      PP Arm
      (pemetrexed+carboplatin or cisplatin)
      ITT n=292 n=286
      Median PFS (95% CI), mo 7.6 (6.6, 8.5) 5.2 (4.3, 5.6)
      HRa (95% CI; P value) 0.596 (0.494, 0.719; P < 0.0001)
      12-Month PFS (95% CI), % 33.7% (28.2, 39.2) 17.0% (12.6, 21.4)
      Median OS (95% CI), mo 18.1 (13.0, NE) 13.6 (11.4, 15.5)
      HRa (95% CI; P value) 0.813 (0.644, 1.025; P = 0.0797)
      12-Month OS (95% CI), % 59.6% (53.9, 65.3) 55.4% (49.5, 61.2)
      ORR (confirmed, inv-assessed), % 46.9% 32.2%
      DOR (95% CI), mo 10.1 (7.2, 13.3) 7.2 (5.7, 9.0)
      PD-L1–highb n=25 n=20
      Median PFS (95% CI), mo 10.8 (7.9, NE) 6.5 (2.4, 10.6)
      HR (95% CI; P value) 0.464 (0.224, 0.960; P = 0.0339)
      PD-L1–lowb n=63 n=73
      Median PFS (95% CI), mo 6.2 (4.4, 8.4) 5.7 (4.2, 7.9)
      HR (95% CI; P value) 0.804 (0.556, 1.163; P = 0.2462)
      PD-L1–negativeb n=88 n=75
      Median PFS (95% CI), mo 8.5 (6.1, 11.2) 4.9 (4.2, 5.8)
      HR (95% CI; P value) 0.448 (0.313, 0.642; P < 0.0001)

      DOR, duration of response; HR, hazard ratio; inv, investigator; ORR, objective response rate; OS, overall survival; PFS, progression-free survival.

      a Stratified. b Baseline tissue available in 60% of patients. PD-L1high (TC3/IC3): patients with PD-L1 expression in ≥50% of tumor cells or ≥10% of tumor-infiltrating immune cells; PD-L1low (TC12/IC12): patients with PD-L1 expression in ≥1% and <50% of tumor cells or ≥1% and <10% of tumor-infiltrating immune cells; and PD-L1negative (TC0/IC0): patients with PD-L1 expression in <1% of tumor cells and <1% of tumor-infiltrating immune cells.

      NCT02657434

      6f8b794f3246b0c1e1780bb4d4d5dc53

    • +

      OA05.08 - Discussant - OA 05.05, OA 05.06, OA 05.07

      14:45 - 15:00  |  Presenting Author(s): Hossein Borghaei

      • Abstract

      Abstract not provided