Scientific Program

Filter Results:

  • +

    MS14 - IO in Early Stage NSCLC

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

      MS14.01 - Basic Science Rationale of IO in Early Stage NSCLC?

      10:30 - 10:45  |  Presenting Author(s): Johan F. Vansteenkiste  |  Author(s): Christophe A. Dooms

      • Abstract

      Abstract

      The 5-year overall survival (OS) of patients with surgically treated early stage NSCLC ranges from 92% (stage IA) to 36% (stage IIIA). Numerous clinical trials over the last decades tried to improve on these results by either adding adjuvant therapy (i.e. in addition after a radical local therapy) or neoadjuvant therapy (i.e. cytoreductive before a radical local therapy).

      Adjuvant setting

      Adjuvant therapy for resected NSCLC came into place in 2004, when the large International Adjuvant Lung cancer Trial (IALT) demonstrated OS benefit with cisplatin-based chemotherapy. No further progress with other types of adjuvant therapy has been made since then. The progress in patients with advanced NSCLC in stage IV with pemetrexed, bevacizumab and EGFR-directed TKIs was not translated in the adjuvant setting [1].

      As for immunotherapy, the MAGRIT trial assessed antigen-specific immunotherapy directed to the MAGE-A3 antigen in completely resected MAGE-A3 positive NSCLC [2]. The adjuvant setting – where the aim is to eliminate sparse tumor cells remaining after surgery – was judged to be the ideal setting for this approach. Unfortunately, the trial did not improve survival. One of the hypotheses why is that the generated MAGE-A3 directed T-cells with cancer killing properties did not travel nor enter easily in tumors, and even if they did, they encountered checkpoints that blocked their potential benefit.

      Immune checkpoint inhibitor (ICI) immunotherapy, which revolutionized the approach to metastatic NSCLC, is now tested in the post-surgical setting with micrometastatic disease and low tumor burden. In contrast to directly cytotoxic chemotherapy, however, agents blocking the PD-1/PD-L1 axis require an interaction between antigen-presenting cells, CD8+ T cells, and tumor cells. It is far from certain that these intercellular interactions will occur sufficiently in patients with micrometastatic disease, where low tumor (mutation) burden may not sufficiently incite the tumor microenvironment. Because of the adjuvant setting, without detectable disease, no early effectiveness read-outs of efficacy nor translational tissue research is feasible. The results of the different ongoing trials with adjuvant ICI therapy in early stage NSCLC needs to be awaited.

      Neoadjuvant setting

      Because of the aforementioned limitations, the neoadjuvant use of PD-1/PD-L1 blockade is of greater interest for several reasons.

      From a biologic perspective, PD-1/PD-L1 blockade before surgery, when the tumor mass and intact locoregional lymph nodes with ongoing activation of T cells may be present, might be more rational. In a breast cancer model in mice, it has been suggested that neoadjuvant ICI is more effective than adjuvant ICI [3]. In the neoadjuvant setting, there was a stronger systemic antitumor T-cell response with maintenance of tumor specific CD8+ T cells in the blood early after immunotherapy. High levels of CD8+ T cells predicted long-term survival in this mouse model. Moreover, induction of this systemic immune response could lead to immune memory that may prevent metastatic relapse over time.

      From a clinical perspective, the neoadjuvant strategy has advantages, as we learned from the chemotherapy era. The systemic therapy has better distribution because of the intact vascularization, which may reduce the locoregional tumor extension. There is an early attack microscopic metastatic disease and the effectiveness can be evaluated in vivo with pre- and post-imaging.

      From a clinical trial perspective, a big advantage is the possibility to study surrogate endpoints. The crucial outcome in this setting is OS, but this endpoint requires many years of follow-up. Here as well, lessons learnt from neoadjuvant chemotherapy followed by surgical resection are of use. Tumor-free lymph nodes and pathological response in the primary tumor have been associated with significantly better outcome. In a dedicated study on the prognostic impact of morphometric tissue analysis of tumor regression, the latter was graded in I: no regression; II: remaining vital tumor tissue ≥10% (grade IIa) or <10% (grade IIb); and III: no evidence of vital tumor tissue [4]. Patients with tumors of regression grades IIb or III showed significantly longer OS than the others (3-year OS 52% vs 9%, P=0.02). These findings, however, do not help to select patients for surgery, as they are post-hoc analyses on the resection specimen. Ideally, one should have pre-surgical (pre-hoc analysis) predictors, but the standard clinical restaging with repeat CT after neoadjuvant only is a raw tool for this purpose. Our group first demonstrated that morphometric tissue analysis of mediastinal lymph nodes after induction was a strong prognosticator in that setting [5]. This was then validated in a prospective multi-center setting, using the first video-mediastinoscopy after neoadjuvant chemotherapy [6]. Patients with a grade IIA or III regression (<10% viable tumor) had much better outcome than the others: 5-year OS 43% versus 19%.

      A recent pilot study evaluated the clinical, pathological, and immunologic effects of short-term neoadjuvant PD-1 blockade in NSCLC [7]. Here again, major pathological response was defined as ≤10% residual viable tumor on sections of the resected tumor. With this criterion, 45% of the patients had a major pathological response, with clear infiltration of CD8+ T cells in regressing tumors. Moreover, a systemic immune response with T-cells of similar T-cell receptor repertoire as in the tissue were described. Even with its small sample (N=21), this strategy has a strong mechanistic principle, both in the tumor and on distant sites, and has resulted in a promising surrogate endpoint effect. Whether this will translated in better 5-year OS now needs to be defined in larger multi-center phase 3 trials, several of which are currently started.

      References

      1. Deslypere G. et al. Ther Adv Med Oncol 2018; 10: 1-11.

      2. Vansteenkiste J. et al. Lancet Oncol 2016; 17: 822-35.

      3. Liu J et al. Cancer Discov 2016; 6: 1382-99.

      4. Junker K et al. Chest 2001; 120: 1584-91.

      5. Dooms C et al. J Clin Oncol 2008; 26: 1128-34.

      6. Dooms C et al. J.Thorac.Oncol. 12 Suppl 1, 460S. 2017.

      7. Forde P et al. N Engl J Med 2018; 378: 1976-86.

    • +

      MS14.02 - Adjuvant / Neoadjuvant IO Therapy

      10:45 - 11:00  |  Presenting Author(s): Julie R. Brahmer

      • Abstract

      Abstract not provided

    • +

      MS14.03 - Concerns About IO for the Thoracic Surgeon?

      11:00 - 11:15  |  Presenting Author(s): Bernward Passlick

      • Abstract

      Abstract

      Perioperativ immunotherapy is a relative new field in thoracic oncology. Although several studies are ongoing, there are until now only two publications describing the effect of preoperative immunotherapy and the intraoperative and the postoperative course of patients with thoracic malignancies.

      The first study is a publication by Forde and colleagues (1), dealing with patients with resectable lung cancer who received one or two preoperative doses of PD-1-inhibitior Nivolumab. The immunotherapy was administered intravenously every 2 weeks, with surgery planned approximately 4 weeks after the first dose. The authors described an acceptable side-effect profile and no delays in surgery. Treatment-related adverse events of any grade occurred in 5 of 22 patients (23%) an only one event was a 3 grades side effect. This was a patient with a pneumonitis who could be resected thereafter that without any complications. The median interval between the administration of the second dose of Nivolumab and surgery was 18 days (range 11 to 29). Twenty of 21 eligible patients underwent a complete resection. The preoperative imaging showed that the majority of the patients (86 %) had stable disease.

      The other report (2) deals with the safety and feasibility of lung resection after immunotherapy for metastatic or unresectable tumors. This is a retrospective study on 19 patients treated between 2012 and 2016. The majority of them had lung cancer (47 %) followed by metastatic melanoma. Various types of immunotherapeutic agents were used. Anatomic resections (lobectomy or greater) were performed in 11 patients. Complications occurred in 32 % but the majority of them were again only minor. One patients developed postoperative pneumonitis in the contralateral lung approximately 2 weeks after VATS-resection on the other side.

      In summary, preoperative immunotherapy seems to be applicable with only minor side effects. However, there are only few reports and theoretically other complications might occur. Especially if the side effect spectrum in patients with metastatic disease and immunotherapy is considered (3). Frequent side effects are pneumonitis, hepatitis or even myocarditis. The development of such side effects should be excluded in each patient undergoing surgery after immunotherapy.

      References

      1. Forde PM, Chaft JE, Smithe KN et al. Neoadjuvant PD-1-Blockade in Resectable Lung Cancer. N Engl J MED 2018 378;21.

      2. Matthew JB, Cools-Lartigue J, Tan KS et al. Safety and Feasibility of Lung Resection after Immunotherapy for Metastatic or Unresectable Tumors. Ann Thorac Surg 2018; …..

      3. Hellmann MD, Ciuleanu TE, Pluzanski A, et al. Nivolumab plus Ipilimumab in Lung Cancer with a High Tumor Mutational Burden. N Engl J Med 2018 378;22.

    • +

      MS14.04 - IO - New Standards in Stage III NSCLC

      11:15 - 11:30  |  Presenting Author(s): Rafal Dziadziuszko

      • Abstract

      Abstract not provided

    • +

      MS14.05 - Do Biomarkers used in Metastatic Setting Apply in Earlier Stages

      11:30 - 11:45  |  Presenting Author(s): Naiyer A Rizvi

      • Abstract

      Abstract not provided

    • +

      MS14.06 - Q&A

      11:45 - 12:00

      • Abstract

      Abstract not provided

  • +

    MS15 - Disruptive Technology and Lung Cancer Risk

    • Type: Mini Symposium
    • Track: Prevention and Tobacco Control
    • Moderators:
    • +

      MS15.01 - Just Water Vapor? Toxicology Perspectives on Electronic Cigarettes

      10:30 - 10:45  |  Presenting Author(s): Maciej Lukasz Goniewicz

      • Abstract

      Abstract

      Electronic cigarettes (e-cigarettes) were introduced to the global market in the mid 2000’s. While it is common to hear the term “e-cigarettes”, this label is a broad term referring to a heterogeneous class of devices that differ in shape, size, and functional characteristics. Common features of e-cigarettes include a heating element that heats a propylene glycol and/or vegetable glycerin based solution (“e-liquid” /“e-juice”) that contains stabilizers, flavorings and often, nicotine. Numerous flavors are available, including tobacco, menthol, fruit, and sweet flavors. Heating of e-liquids produces an aerosol, which is then inhaled by the user. Due to their relatively short existence, data on the long-term health effects of e-cigarette use are not currently available. In the interim, evidence from animal studies, in vitro and in vivo laboratory studies, observational studies, and small-scale clinical trials may provide important information on the potential harms of e-cigarette use [1].

      Many studies conducted on e-cigarettes have focused on the measurement of potentially harmful chemicals that may be produced by these products. Chemicals identified in e-cigarette aerosol include nicotine, tobacco‑specific nitrosamines (TSNAs), metals, polycyclic aromatic hydrocarbons (PAHs), volatile organic compounds (VOCs), and aldehydes. Within these classes, there are several respiratory irritants and toxicants, as well as carcinogenic substances linked to the development of respiratory cancers. In our work, we have tested emissions from numerous e-cigarette brands [2] and identified the presence of formaldehyde, acetaldehyde, and acrolein in e-cigarette emissions. Overall, concentrations of toxicants identified in e-cigarette aerosols were 9 to 450 times lower than in tobacco smoke [2]. Concerns have also been raised about the presence of metal particles in e-cigarette aerosol (particularly nickel and chromium, two main elements in heating coils).The inhalation of these metals in larger quantities may cause respiratory diseases, bronchitis, and pneumonia [3], however, these effects have not been definitively elucidated.

      The size of particulate matter generated from e-cigarettes affects pulmonary nicotine absorption and determines settlement of particulate matter into various parts of the upper or lower airways. There is likely substantial variation across generations of e‑cigarette devices, and across brands. Results from laboratory studies indicate that e-cigarettes may expose users to small particles and lower amounts of particulate matter in general. While inhalation of high levels of particulate matter has been linked to greater mortality risk from cardiopulmonary illnesses, the available data indicate that e-cigarette particulate emissions expose users at a level akin to WHO guidelines and are far lower than those of conventional cigarettes.

      Issues raised about toxicological effects mostly question effects on cells, with a special interest in lung epithelial cells. For instance, many flavorings used in e‑cigarettes (e.g. cinnamaldehyde, benaldehyde, diacetyl) are already approved for use in food, yet their impact on respiratory health via repeated inhalation is currently unknown. Several studies have shown that e-cigarette flavorings could lead to lung cell damage (mostly by releasing free radicals) and inflammation in lung tissue [4]. Studies on cytotoxic effects of e-cigarette constituents have also identified negative effects on DNA. In one in vitro research, e-cigarette liquids aerosolized at biologically relevant doses induced increased DNA strand breaks and apoptosis while decreasing survival in both normal epithelial and head and neck squamous cell carcinoma cell lines [5]. Moreover, in experiments conducted by Yu et al. [6] e-cigarettes aerosol has shown cytotoxic effects on epithelial cell lines and acted as a DNA-breaking agent.

      Given multiple potential etiologic mechanisms related to incident case development coupled with the long latency period in developing illness, there is currently no definitive evidence to commenting on the role of e-cigarettes in increasing lung cancer risk. As an intermediate assessment, cross-sectional biomarker data can be suggestive of possible carcinogen exposures related to cancer development. For instance, Shahab et al. [7] examined a large panel of biomarker data among e-cigarette users, cigarette users, and users of both products (“dual users”). The e-cigarette–only users had significantly lower metabolite levels for tobacco-specific nitrosamines (TSNAs), particularly NNAL, a metabolite of potent lung carcinogen NNK. Several observational longitudinal studies also showed a substantial reduction in exposure to NNK and several VOCs, including respiratory toxicants like acrolein, acrylamide, acrylonitrile, 1,3-butadiene (human carcinogen), and ethylene oxide among smokers who switched to e-cigarette [8]. Although evidence from biomarker studies are insufficient to evaluate causative mechanisms, but show users of e-cigarettes display lower levels of exposure to biomarkers of lung carcinogens when compared to smokers, such as NNK.

      Since e-cigarettes have only been on the market for a decade, it is presently not possible to assess all potential long-term harmful effects of e-cigarette use. To date, findings from clinical studies have demonstrated that e-cigarettes are likely less harmful compared to conventional tobacco cigarettes, and any harmful side effects are noticeably milder compared with regular cigarettes. Furthermore, it is also clear that e-cigarette aerosols are not “a harmless water vapor”, as claimed by manufacturers and retailers, and potential health effects from vaping may emerge after long-term use.

      References

      1. National Academies of Sciences, Engineering and Medicine. Public health consequences of e-cigarettes. Washington, DC: The National Academies Press; 2018.

      2. Goniewicz et al. Levels of selected carcinogens and toxicants in vapour from electronic cigarettes. Tob Control. 2014;23:133-139

      3. Lerner et al. Environmental health hazards of e cigarettes and their components: oxidants and copper in e-cigarette aerosols. Environ Pollut. 2015;198:100-107

      4. Leigh et al. Flavourings significantly affect inhalation toxicity of aerosol generated from electronic nicotine delivery systems (ENDS). Tob Control. 2016;25(Suppl 2):ii81-ii87.

      5. Welz et al. Cytotoxic and genotoxic effects of electronic cigarette liquids on human mucosal tissue cultures of the oropharynx. J Environ Pathol Toxicol Oncol. 2016;35:343-354.

      6. Yu et al. Electronic cigarettes induce DNA strand breaks and cell death independently of nicotine in cell lines. Oral Oncol. 2016;52:58-65.

      7. Shahab et al. Nicotine, carcinogen, and toxin exposure in long-term e-cigarette and nicotine replacement therapy users: a cross-sectional study. Ann Intern Med. 2017;166:390-400.

      8. Goniewicz et al. Exposure to nicotine and selected toxicants in cigarette smokers who switched to electronic cigarettes: a longitudinal within-subjects observational study. Nicotine Tob Res. 2017;19:160-167.

    • +

      MS15.02 - PRO- Electronic Cigarettes: A Cessation Tool

      10:45 - 11:00  |  Presenting Author(s): K. Michael Cummings

      • Abstract

      Abstract

      Should clinicians recommend electronic cigarettes (e-cigarettes) to their lung cancer patients who continue to smoke? For many clinicians the answer to this question is a resounding NO WAY. However, I hope to persuade you that there are good reasons why clinicians should strongly consider recommending e-cigarettes to at least some of their patients who smoke. No one disputes the enormous health risks posed by cigarette smoking, or the need to find better treatments to help smokers overcome addiction to cigarettes. In theory, a product that can deliver nicotine like a cigarette without the toxins found in smoke could be used instead of cigarettes would be a welcome invention. E-cigarettes were first introduced into the marketplace in 2003 and were initially promoted as an aid to help smokers to stop smoking. However, concerns have been raised about whether e-cigarettes are an effective cessation aid or if they would actually reduce successful quitting by adult smokers, whether they are safe to use, and if they may be a gateway into smoking for youth. Admittedly, the science is in a state of flux and evolving rapidly. The National Academy of Sciences, Engineering, and Medicine (NASEM) report on the public health consequences of e-cigarettes, and a new evidence review on e-cigarettes and heated tobacco products commissioned by Public Health England (PHE) both come to similar conclusions. Both reports acknowledge that available evidence indicates that e-cigarette use is less risky than use of combustible tobacco cigarettes; that e-cigarettes may be helpful to smokers who are trying to stop smoking cigarettes, and that e-cigarettes should not be used by non-smokers, especially youth. However, both reports also state that e-cigarettes contain constituents that are not inert and are likely to have some negative health effects on their own. Given uncertainties regarding e-cigarettes, clinicians should advise cigarette smokers seeking to stop smoking to use evidence-based, FDA-approved, safe, and effective smoking cessation pharmacotherapies as first-line treatments in preference to e-cigarettes. So why consider e-cigarettes as a treatment option? First, many patients have already tried the evidence based methods and have found them not to be helpful. We need better treatments and e-cigarettes, in theory, might work for some patients. Second, nicotine seeking is the primary motivation for continued smoking so providing addicted smokers with an alternative that delivers nicotine without most of the harmful toxins in smoke makes sense. Nicotine is not the problem, it is the smoke. E-cigarettes are not lit, they do not burn, and do not produce cigarette smoke. Finally, many smokers simply prefer e-cigarettes over other aids to quitting. It is great to have government approved stop smoking treatments, but if smokers are unwilling to use them, they don’t do much good. Sales of e-cigarettes have grown exponentially over the past decade, and they have become the most popular quitting aid used by smokers in many countries. For those smokers and ex-smokers who are already using the e-cigarettes, clinicians need to be informed and prepared to answer questions regarding potential harms and benefits and to advise patients about use.

    • +

      MS15.03 - CON- Electronic Cigarettes: Not Evidence Based Cessation

      11:00 - 11:15  |  Presenting Author(s): Alison Wallace

      • Abstract

      Abstract not provided

    • +

      MS15.04 - The Rise of Heat-Not Burn Tobacco in Japan: A “Hot” Issue for Tobacco Control

      11:15 - 11:30  |  Presenting Author(s): Junji Yoshida

      • Abstract

      Abstract

      Just like in other developed countries, public health has been of major public interest in Japan. Reduced-risk tobacco alternative that exposes smokers who cannot quit and non-smokers around them to lower levels of harmful and potentially harmful constituents (HPHC) has long been a goal. In many countries, electronic nicotine delivery systems (ENDS) have been introduced into the market for that purpose. However, there is a high hurdle for ENDS in Japan. Nicotine is classified as a medicinal drug, and devices using nicotine require governmental approval under the pharmaceutical affairs law. Although Japanese citizens can privately import ENDS as an unapproved pharmaceutical product for private use, bureaucratic formalities are troublesome, and the monthly amount able to be imported is limited.

      In 1997, Japan Tobacco (JT) started to sell AIRS®, the R.J. Reynolds’ Premia® product but with a slight modification. AIRS® had a charcoal heat source at one end, and heated air was inhaled through tobacco leaves. This product did not sell well and was discontinued in 2004. Then in 2010, the next alternative was ZERO STYLE STIX®, also by JT. This product has no heat source. It is a snuff variant, and ambient air is simply inhaled through tobacco leaf powder. When its nicotine was measured by ISO metrics, the amount of inhaled nicotine was shown to be less than a half of low nicotine cigarette.

      Following these alternatives, heat-not burn tobacco (HNBT) came into the market. Tobacco manufacturers aimed at public tobacco harm reduction in total by reducing HPHC emissions from their products and encouraging public acceptance of such products. The tobacco manufacturers understood that Japanese value cleanliness and physical fitness. Most Japanese people are very anxious about their reputation within their community. An American anthropologist Ruth Benedict described it as “Shame Culture” in her book “The Chrysanthemum and the Sword” in 1946. Many Japanese smokers feel shame at releasing HPHC and ash. The manufacturers also know Japan is a country of electric gadget geeks.

      The first of the updated or more recent heated tobacco products was Ploom® from JT in 2013, which was updated to Ploom TECH® in 2016. These were bought from a Californian company Ploom Inc. They are not HNBT but aerosol-not burn or e-cigarette, to be precise. Glycerin, propylene glycol and water vapor passes through a tobacco capsule at the maximum temperature of 30. Philip Morris International introduced their true HNBT, iQOS®, to the Japanese market in 2015, and British American Tobacco product named glo® in 2016. All these HNBT’s were sold in limited areas of Japan at first and gradually extended to nation-wide sales, in response to the demand for HNBT, expanding their production lines. HNBT sales increased rapidly, and heated tobacco category usage among all tobacco users grew up to over 30% in early 2018 from about 4% in late 2015.

      Cigarette ads on TV, radio, or Internet have been allowed in Japan if it is technically possible to restrict access only to the adults, and this is true for the HNBT ads. According to the guidelines set down by the Ministry of Finance, access to HNBT advertising/information web sites and purchase at virtual/real shops have been strictly restricted, requiring proof of age such as a driving license. In their web sites and on product packages, there are always some messaging to the effect of reduced HPHC, but not necessarily guaranteeing less harm to health, which is in accordance with the Ordinance for Enforcement of the Tobacco Business Act.

      Reduction in HPHC exposure and lower levels of biomarker responses have been vigorously reported in peer-reviewed journals, but many of them are funded, studied and reported by the tobacco manufacturing companies themselves. There have been several contradictory articles and in media coverage suggesting that some HPHC are released more from HNBT than from cigarettes. However, manufacturers refute the findings by pointing out inappropriate methods, evaluation and data presentation in these reports. The truth will need to wait for the academic and public health community to sort out the reality of the potential for HNBT to reduce the morbidity and mortality caused by tobacco product use.

      Currently, it seems reasonable to understand that HNBT is likely to expose users and bystanders to lower levels of most HPHC. Although the extent of the reduction found so far varies between studies, reduced exposure to some HPHC is reported to be associated with less short-term harm. To this extent, HNBT might be beneficial for smokers who cannot quit and non-smokers around them. However, less is not zero. It will take 15-20 years to confirm long-term health harm caused by reduced HPHC, such as clinical carcinogenesis. It requires tremendous effort to maintain and watch a large cohort of HNBT users for that long period of time. It is questioned which party ought to do and publish the study. Another question is who pays the cost and whether the study is worth the cost. Because zero is zero, tobacco products including HNBT are desired to be completely abandoned in the future, eliminating the need for long-term harm studies.

    • +

      MS15.05 - Heat-Not-Burn Tobacco: Real Risk Reduction or Industry’s Next Promise

      11:30 - 11:45  |  Presenting Author(s): Carolyn Dresler

      • Abstract

      Abstract

      So, the question is: what are we to think about the products emerging from the global tobacco manufacterers that are being marketed in upper income countries? To be clear the topic is not about electronic nicoitne delivery systems (ENDS) that have been sweeping the world’s marketplaces - those arise from a multitude of mostly small companies. ENDS are an interesting topic that is quickly changing also - but, not the topic of this dissertation.

      Rather the topic for this paper will discuss the current market-disrupters which are the so-called ‘heat not burn’ cigarettes (HNB). The tobacco control community tends to call them ‘heated tobacco products’ (HTP), as some of them do come close to ‘burning’ (combusting) the tobacco, so, HTP is a more accurate, broad term for the category. The question to be addressed is whether they are less harmful than cigarettes or are they the usual ‘oversell’ or untrue marketing claims that have been the usual practice of the global tobacco manufacturers.

      A brief review: the global tobacco manufacturers are: Philip Morris, International. (PMI-Switzerland); British American Tobacco (BAT-UK); Imperial Tobacco (UK); Altria (USA) and Japan Tobacco International (JTI- Japan).(see: https://www.statista.com/statistics/259204/leading-10-tobacco-companies-worldwide-based-on-net-sales/). Altria is really just a spin-off from Philip Morris International who sells their products in the USA, including trying to do so for their HNB product (iQOS). Each of them have their own offerings in the HNB category: PMI/Altria is the current leader with iQOS, quipped as standing for: I quit ordinary smoking; BAT has ‘glo’; and Imperial has not embraced/marketed a HNB product but had thought to stay with their ENDS (blu) product. Reportedly, Imperial is considering getting their own HNB, as this category appears to be growing (commerically renumerative). JTI has PloomTECH.

      IQOS and glo use a lithium battery to heat compressed/powdered tobacco to around 300 Celsius which creates an aerosol for inhalation. PloomTech heats a liquid that has nicotine in it (closer to an ENDS)

      A recent paper by Farsalinos et, tested iQOS, an ENDS and a regular cigarette for emitted carbonyls, like formaldehyde and acrolein. (1) They found that iQOS emitted 82-98% less carbonyls compared to smoking 20 tobacco cigarettes. The ENDS emitted 92-99.8% less carbonyls. So, iQOS were better than cigarettes by alot, but the ENDS was better still.

      In a webpage hosted by RIVM (National Institute for Public Health and the Environment, misistry of Health, Welfare and Sport), their research to date is quoted as:

      “Heated tobacco products are newly available on the market. An example of such a product is the heatstick which is heated with an iQOS, a device that looks like an e-cigarette. Heating the heatstick leads to the formation of carcinogenic and other harmful substances. The use of heatsticks with the iQOS is harmful to health, but probably less harmful than smoking tobacco cigarettes. This is RIVM’s conclusion, based on its research into heated tobacco products (HTPs).”(2)

      In another recent paper from Japan, HTPs (iQOS, glo and Ploom TECH) were compared with ‘reference cigarettes’(research cigarettes that are well-defined and not for human consumption (as if any should ever be)). (3) Chemists interested in these findings should check out this paper. For the rest of us - it just demonstrates the tremendous variability in the products. Cigarettes deliver nicotine better than the other tested products, and iQOS does better than glo which does better than PloomTech. Since users are going for the nicotine, it seems that the order of preference will be cigarettes, iQOS, glo and PloomTech. However, the manufacturers are constantly changing their products, so they could be different next month. (This is the problem of unregulated products)

      Most of the published/available research is by the tobacco industry. For the hardy, there is a very large amount of scientific information from PMI on their iQOS submission to the USA FDA. (4) This is all publicly available information - and actually, a very interesting read. The question always remains when reviewing tobacco industry research - can it be verified?

      Overall, however, it does seem that heated tobacco products (or, ENDS) deliver lower levels of most toxicants to the human lung. Low enough to prevent disease is a different question. BUT, cigarettes kill more than half of the people who use them, cause a huge amount of morbidity and healthcare costs - and, are addictive. So, the question facing society - or, public health - is it possible or probable that if people who are already addicted to nicotine and using cigarettes switch (completely) to HTP products - would there be lower morbidity and mortality than we currently experience from cigarttes? No one knows the answer at present - and, we are unlikely to know for a few decades the real answer.

      However, are we horrified enough at the currently greater than 7 million deaths globally per year from cigarettes to radically alter our passionless inactivity and push, by regulation (and social pressure) for people to quit cigarettes. If they use HTPs (or ENDS) to quit the unacceptably deadly cigarettes, so be it. I don’t know how they could be worse than what we currently allow.

      {Disclaimer: I am still passionate about the standard tobacco control measures that are working in some countries, per the Framework Convention for Tobacco Control. I do not trust the global tobacco manufacturers that they are willing to help everyone in the world to quit cigarettes, particularly in low income countries, where these HTP products would be more expensive (unaffordable?) for their current customers. I believe the industry is still intersted in maintaining tobacco addiction - I want to ultimately reduce it as far as possible. First however, I want people who are nicotine addicted to stop dying from that addiction.}

      1. Farsalinos KE at al. Addiction 2019 June 19. doi: 10.1111/add 14365

      2. https://www.rivm.nl/en/Documents_and_publications/Common_and_Present/Newsmessages/2018/Addictive_nicotine_and_harmful_substances_also_present_in_heated_tobacco (accessed 6/29/2018)

      3. Uchiyama S. et al. Simple Determination of gaseous and particulate compounds generated from heated tobacco products. Chem. Res. Toxicol 2018.

      4. https://www.fda.gov/tobaccoproducts/labeling/marketingandadvertising/ucm546281.htm (accessed 6/29j/2018)

    • +

      MS15.06 - Q&A

      11:45 - 12:00

      • Abstract

      Abstract not provided

  • +

    OA10 - Right Patient, Right Target & Right Drug - Novel Treatments and Research Partnerships

    • Type: Oral Abstract Session
    • Track: Targeted Therapy
    • Moderators:
    • +

      OA10.01 - Patient Preferences for Tyrosine Kinase Inhibitor Treatments for EGFR Mutation Positive Metastatic NSCLC

      10:30 - 10:40  |  Presenting Author(s): John F.P. Bridges  |  Author(s): Marie De La Cruz, Melissa Pavilack, Emuella Flood, Ellen Janssen, Nabil Chehab, Ancilla W. Fernandes

      • Abstract

      Background

      EGFR mutation positive (EGFRm) NSCLC responds to EGFR-tyrosine kinase inhibitors (TKIs). First-, second-, and third-generation EGFR-TKI treatment attributes vary in efficacy, side effects, and dosing regimen. We used two different methods to explore treatment preferences among patients with EGFRm metastatic NSCLC.

      Method

      Patients with EGFRm metastatic NSCLC were recruited in the US to participate in an online survey containing two complimentary preference elicitation methods. First, preferences were assessed through direct elicitation exercises where participants chose between competing treatment profiles. The first exercise compared two profiles with a large difference in progression-free survival (PFS) (6 vs 18 months). The second exercise compared two profiles with a smaller difference (10 vs 18 months). Both exercises compared the same side effect risks (0–1% vs 2–16% risk). Second, a discrete choice experiment (DCE) was used to assess preferences for variation in treatments in terms of PFS, severity of side effects (mild/moderate/severe), and mode of administration. These attributes and levels were varied based on a D-efficient design. Participants completed 10 DCE choice tasks where they saw pairs of hypothetical treatments with different attribute levels and selected their preferred treatment. A choice model was estimated using conditional logit regression.

      Result

      Between 10/2017 and 03/2018, 90 patients with EGFRm metastatic NSCLC were recruited and completed the survey: 42% female; 79% white; 76% taking first-line or second-line EGFR-TKIs at time of survey. Median time since diagnosis: 2.1 years (inter-quartile range: 1.1–2.7). In the direct elicitation exercise, participants opted for shorter PFS in exchange for more favorable side effects, but were less likely to do so for a large difference in PFS (52% of participants) vs a smaller difference (66%; p<0.001). Participants who chose shorter PFS when difference in PFS was large were more likely to be taking EGFR-TKIs (odds ratio: 21.4; 95% confidence interval: 2.24, 204.88). No relationship between choice and treatment characteristic was observed when difference in PFS was small. In the DCE, conditional logit regression indicated that to avoid severe levels of nausea/vomiting, diarrhea, rash, or fatigue, participants on average would accept reductions in PFS of 13, 11, 9, and 8 months, respectively. Participants would accept reduction in PFS of 7 months for oral treatment taken with/without food vs IV.

      Conclusion

      In this online survey of patients with EGFRm metastatic NSCLC, some patients were willing to accept shorter PFS for a better safety profile and dosing convenience; however, PFS remained an important attribute in treatment choice.

    • +

      OA10.02 - Oncogene-Driven Patient Groups: A New Era For Research Partnerships

      10:40 - 10:50  |  Presenting Author(s): Janet Freeman-Daily  |  Author(s): Ivy Elkins, Laura Greco, Marcia K Horn, Bonnie Addario, David LeDuc, Robert C. Doebele, Christine Lovly

      • Abstract

      Background

      Genomic alterations drive more than 60% of non-small cell lung cancer (NSCLC). About 20% of NSCLC cases (EGFR, ALK, ROS1, BRAF) will have an oncogenic driver that can be treated with approved targeted therapy drugs, and more have clinical trial options. In some cases, patients on targeted therapies will have years of good quality of life. However, targeted therapies do not cure, and these patients will eventually see their cancer progress. Patients and caregivers dealing with cancers driven by EGFR, ALK, ROS1 and Exon20 oncogenes have organized globally into online groups and are building partnerships that seek to provide support, increase awareness and education, accelerate and fund research, and improve access to effective diagnosis and treatment.

      Method

      Each patient-caregiver group forms using social media, patient blogs, websites, fliers, newsletters, and contacts with clinicians. Each has a private Facebook group or other network to inform and educate patients and caregivers of approved and experimental treatments, share common experiences, provide information and tips from expert clinicians and researchers, and enable real-life connections. Each group sets its own priorities such as creating websites, raising funds for research, donating tissue, developing studies of demographics and treatment sequences, and creating biorepositories with annotated specimens that will be made widely available. These projects are accomplished in patient-driven partnership with researchers, clinicians, advocacy groups, and industry. None of the patient-caregiver groups are corporations or other types of legal entities.

      Result

      The ROS1ders focus on ROS1+ cancers of all types and include over 275 members from 21 countries across 8 cancer types. They have partnered in two studies to create ROS1+ cancer models. ALK Positive focuses on ALK+ NSCLC and other cancers, and includes over 1000 members from 38 countries. They have raised nearly $400,000 for research and will announce two grants in May. Exon 20 Group focuses on EGFR and HER2 insertions in Exon 20, and includes 110 members from 19 countries. They created an Exon 20 molecular tumor board and awarded two grants. The EGFR Resisters focus on EGFR+ NSCLC and cancers that develop resistance to EGFR targeted therapies, and include over 450 members from 20 countries.

      Conclusion

      Oncogene-driven patient-caregiver groups are creating a new paradigm in cancer research and have demonstrated that their partnerships with advocacy organizations, clinicians, researchers and industry, combined with social media outreach, can increase available patient data, specimens, cancer models and research funding for geographically distributed, oncogene-driven cancer populations.

    • +

      OA10.03 - Discussant - OA 10.01, OA 10.02

      10:50 - 11:05  |  Presenting Author(s): Zofia Piotrowska

      • Abstract

      Abstract not provided

    • +

      OA10.04 - Afatinib With or Without Cetuximab for EGFR-Mutant Non-Small Cell Lung Cancer: Safety and Efficacy Results from SWOG S1403

      11:05 - 11:15  |  Presenting Author(s): Sarah B Goldberg  |  Author(s): Mary W. Redman, Rogerio Lilenbaum, Katerina Politi, Thomas E. Stinchcombe, Leora Horn, Everett H. Chen, Sandeep Mashru, Scott N. Gettinger, Mary Ann Melnick, Jieling Miao, James Moon, Karen Kelly, David R. Gandara

      • Abstract

      Background

      Several EGFR tyrosine kinase inhibitors (TKIs) are used for the treatment of EGFR-mutant non-small cell lung cancer (NSCLC), however resistance inevitably develops. The combination of the irreversible ErbB family TKI afatinib and the EGFR monoclonal antibody cetuximab was previously shown to overcome resistance to first-line EGFR TKIs. To attempt to delay resistance, we conducted a randomized trial of afatinib plus cetuximab versus afatinib alone in treatment-naïve patients with advanced EGFR-mutant NSCLC (NCT02438722).

      Method

      Patients with previously-untreated EGFR-mutant NSCLC were randomized to afatinib 40mg PO daily plus cetuximab 500mg/m2 IV every 2 weeks or afatinib 40mg PO daily. The study was designed to accrue a total of 212 patients, comparing progression-free survival (PFS) between the arms at the 1-sided 0.025 level when 134 PFS events had been observed. Secondary objectives included comparison of overall survival (OS), time to treatment discontinuation (TTD), and toxicity. An interim analysis evaluating early stopping for futility occurred when at least 64 PFS events were reported.

      Result

      Between March 26, 2015 and April 23, 2018, 170 eligible patients were accrued: 86 to afatinib/cetuximab and 84 to afatinib. Median age was 66.4 years, 66% were female, 64% had an EGFR exon 19 deletion mutation and 36% had an L858R point mutation. With 109 events observed, there was no improvement in PFS with the combination compared to single-agent (HR 1.17, 95% CI 0.80-1.73, P = 0.42, median 10.6 months vs 13.1 months). OS was also not improved with the addition of cetuximab (HR 1.23, 95% CI 0.62-2.44, P = 0.55, median 26.9 months vs not reached). TTD was similar between the two groups (HR 0.95, 95% CI 0.64-1.39, P = 0.79, median 12.5 months vs 12.2 months). Grade > 3 treatment-related adverse events (AEs) were more common among patients treated with afatinib/cetuximab, and more patients in the combination arm required at least 1 dose reduction of afatinib (57% vs 26%). However, treatment discontinuations due to AEs were similar between the two groups (11.6% vs 10.7%).

      Conclusion

      There was no difference in PFS, OS or TTD with the addition of cetuximab to afatinib for treatment-naïve patients with EGFR-mutant NSCLC. The trial was closed to accrual at the interim analysis having met the criteria for futility. Correlative analysis of tumor tissue and blood from patients is ongoing.

    • +

      OA10.05 - An Open-Label, Multicenter, Phase II Single Arm Trial of Osimertinib in NSCLC Patients with Uncommon EGFR Mutation(KCSG-LU15-09)

      11:15 - 11:25  |  Presenting Author(s): Jang Ho Cho  |  Author(s): Jong-Mu Sun, Se-Hoon Lee, Jin Seok Ahn, Keunchil Park, Keon Uk Park, Eun Joo Kang, Yoon Hee Choi, Ki Hwan Kim, Ho Jung, An An, Hyun Woo Lee, Myung-Ju Ahn

      • Abstract

      Background

      Approximately 10% of EGFR mutants harbor uncommon mutations, which represent a heterogeneous group of rare molecular alterations within exons 18-21 and the sensitivity to EGFR TKIs is variable. Osimertinib is a potent irreversible inhibitor of both sensitizing EGFR mutation and T790M. In preclinical data, osimertinib was found to be active against most uncommon EGFR mutants, apart from the exon 20 insertion variant. Here we present the efficacy and safety of osimertinib in patients with uncommon EGFR mutation positive NSCLC.

      Method

      Patients with histologically confirmed metastatic or recurrent NSCLC with activating EGFR mutation other than exon 19 deletion, L858R, T790M and insertion in exon 20 were eligible. Patients received 80mg of osimertinib orally, once daily until progression or unacceptable toxicity. Response was assessed every 8 weeks by investigator. The trial was registered with ClinicalTrials.gov, number NCT03424759.

      Result

      Between Mar 2016 and Oct 2017, 35 patients were enrolled. Median age was 59, 63% male, 43% never smoker, 97% adenocarcinoma. 63% of patients were treated as first-line therapy. The mutations identified were G719A/C/D/S/X (19, 54.3%) followed by L861Q (9, 25.7%), S7681 (8, 22.9%), and others (3, 8.6%). The overall response rate was 50.0% (95% CI 32.8-67.2) and DCR was 88.9% (95% CI 78.1-99.7). Seven patients (77.8%) with L861Q mutation achieved partial response; 10/19 (52.6%) with G719A/C/D/S/X mutation; 3/8 (37.5%) with S768I mutation. At data cutoff (Apr, 2018), the median PFS was 8.2 months (95% CI 1.9- 14.4) and median duration of response was 9.8 months (95% CI 7.6-12.0). The most common adverse events were rash (n=11, 31.4%), anorexia (n=8, 22.9%), and diarrhea (n=7, 20.0%). Grade 3 or 4 AEs were reported in 8 of 35 patients (23%), but all of AEs were manageable.

      Conclusion

      Osimertinib showed highly active in NSCLC patients harboring uncommon EGFR mutation with manageable safety profile, consistent with previous reports. Further analysis will be provided.

    • +

      OA10.06 - A First-in-Human Phase 1 Trial of the EGFR-cMET Bispecific Antibody JNJ-61186372 in Patients with Advanced Non-Small Cell Lung Cancer (NSCLC)

      11:25 - 11:35  |  Presenting Author(s): Keunchil Park  |  Author(s): Myung-Ju Ahn, Se-Hoon Lee, Hye Ryun Kim, Min Hee Hong, Dawn Millington, Martin Curtis, Spyros Triantos, Sandra Chaplan, Nahor Haddish-Berhane, Roland Knoblauch, Zuleima Aguilar, Sylvie Laquerre, Matthew V. Lorenzi, Byoung Chul Cho

      • Abstract

      Background

      JNJ-61186372 (JNJ-372) is a bispecific antibody targeting both EGFR and cMET. In preclinical studies, JNJ-372 demonstrated efficacy in EGFR and cMET driven tumor xenograft models (including EGFR T790M and MET-amplified/HGF secretion), consistent with inhibition of ligand binding, receptor degradation, and ADCC activity. The goal of Part 1 of this study (reported here) was to assess the safety, pharmacokinetics (PK), and preliminary efficacy of JNJ-372 and to identify the recommended phase 2 dose(s) to be explored in Part 2.

      Method

      Patients with previously treated, advanced NSCLC were enrolled at two sites and treated with escalating doses of JNJ-372 administered IV weekly for the first 4-week cycle, then biweekly for each subsequent cycle. PK sampling was taken at multiple time points within cycle 1 and 2. Disease assessments were performed every 8 weeks. Tumors were characterized at baseline through next-generation sequencing of circulating tumor DNA (Guardant 360).

      Result

      25 patients were treated with JNJ-372 during dose escalation: 140mg (n=3), 350mg (n=3), 700mg (n=9), 1050mg (n=7), 1400mg (n=3). Median age was 63y, 48% were male, 100% were Asian, 84%/12%/4% had adenocarcinoma/squamous/other histology, and median prior therapies was 4. No dose-limiting toxicities were observed at any dose level tested. The most frequent treatment-emergent AEs were infusion-related reactions (76%), rash/acneiform dermatitis (40%), dyspnea (24%), paronychia (24%), pruritus (20%), fatigue (20%), and nausea (20%); incidence of peripheral edema (cMET-related toxicity) was 12%. Infusion-related reactions were grade ≤2 severity, observed primarily with the first dose. The worst severity of rash/acneiform dermatitis was grade 2 (16%). One treatment-related AE of grade ≥3 severity was reported (neutropenia grade 3, possibly related). JNJ-372 demonstrated linear PK at dose levels 350 mg and above with non-linear PK at lower concentrations, suggesting target-mediated drug disposition. Doses ≥700mg resulted in average steady-state concentrations at or above the preclinically established therapeutic target level. Preliminary evidence of efficacy (maximum change from baseline in sum of target lesion diameters) was observed in a patient with squamous cell carcinoma (-20%), a patient with wtEGFR adenocarcinoma (-20%), and 4 patients with EGFR-mutant adenocarcinoma (≥-30%).

      Conclusion

      JNJ-372 is a novel EGFR-cMET bispecific antibody. The manageable safety profile and preliminary evidence of clinical activity support active accrual of patients with previously treated EGFR-mutant NSCLC. The first recommended dose of 1050mg is being evaluated in Part 2.

    • +

      OA10.07 - Resistance Mechanisms of Osimertinib in Chinese Non-Small Cell Lung Cancer Patients: Analysis from AURA17 Trial

      11:35 - 11:45  |  Presenting Author(s): Caicun Zhou  |  Author(s): Min Hu, Xuehua Zhu, Yun Sun, Xuesong Lu, Jia Wang, Mengzhao Wang, Ying Cheng, Yuan Chen, Yanqiu Zhao, Yuan-Kai Shi, You Lu, Meiqi Shi, He-Long Zhang, Xiangning Huang, Zheng Wang, Zhijie Ding, Heyan Li, George Chen, Yi-Long Wu

      • Abstract

      Background

      Osimertinib is approved for metastatic NSCLC patients with EGFR T790M mutation after progression from TKI therapy. Despite impressive tumor responses, drug resistance usually develops. The resistance mechanisms of osimertinib are emerging but studies with large cohorts of Chinese patients and association with clinical outcomes are lacking. Here we report a biomarker study of osimertinib using plasma samples from 107 Chinese patients who had progressed by 24 months after LSFD (Oct. 2017) of AURA17 (NCT02442349), the 2nd-line pivotal trial in China.

      Method

      Serial plasma cell-free DNA (cfDNA) were collected from baseline until progressive disease (PD) by investigator assessment. Capture-based 75-gene NGS panel with unique molecular index (UMI) system was used to identify resistance mechanisms to osimertinib by comparing paired cfDNA at baseline and PD. Droplet digital PCR (ddPCR) was used to dynamically monitor EGFR mutation changes (L858R, Ex19Del, T790M and C797S) during treatment course. Association of cfDNA biomarkers based on valid test results with objective response rate (ORR), progression-free survival (PFS) and overall survival (OS) from DCO3 (Mar. 20, 2018) was analyzed.

      Result

      The 107 patients were with ORR of 68.2%, median PFS of 8.2 months, and median OS of 21.5 months. Eight-two had detectable EGFR sensitizing mutations (L858R or Ex19Del) in their PD cfDNA samples. Among them, 15 had acquired EGFR C797S, all in cis with T790M, and with no enrichment for L858R or Ex19Del (6 and 9, respectively). The median time of C797S detection from plasma was 2.8 (1.2-8.4) months prior to PD. EGFR L718Q, I744T, C775Y, G796S/D, T854I mutations, or amplification were found in 11 patients. Aberrations in bypass tracks including AKT2, ALK, DDR2, ERBB2/3, HRAS, JAK1/2, KRAS, MET, NTRK1, PIK3CA, RIT1, etc. were observed in 45 patients.

      Clearance of EGFR sensitizing mutations at weeks 3 of treatment was associated with favorable ORR (78.7% vs. 33.3%), PFS (9.6 vs. 4.0 months, p<0.001) and OS (21.5 vs. 11.7 months, p<0.001). Clearance of EGFR sensitizing mutations at weeks 6 of treatment was also associated with favorable ORR (80.0% vs. 36.8%) and PFS (8.3 vs. 4.2 months, p<0.001). Presence of T790M at PD was correlated with longer PFS (12.3 vs. 5.5 months, p<0.001) and OS (21.3 vs. 13.2 months, p=0.045). Acquired or enriched TP53 alterations at PD were associated with worse PFS (4.2 vs. 8.3 months, p=0.008).

      Conclusion

      Our study revealed diverse resistance mechanisms to osimertinib in Chinese NSCLC patients and urged for new drug discovery or combination strategies to overcome this clinical challenge.

    • +

      OA10.08 - Discussant - OA 10.04, OA 10.05, OA 10.06, OA 10.07

      11:45 - 12:00  |  Presenting Author(s): Linda Coate

      • Abstract

      Abstract not provided

  • +

    PC05 - Optimizing Clinical Trial Design in NSCLC

    • Type: Pro-Con Session
    • Track: Advanced NSCLC
    • Moderators:
    • +

      PC05.00 - Introduction with Poll Questions

      10:30 - 10:35  |  Presenting Author(s): Pilar Garrido, Luis Ubillos

      • Abstract

      Abstract not provided

    • +

      PC05.01 - Debate 1: Which is Most Important Efficacy Endpoint in First Line Trials in Advanced NSCLC PFS or OS - Point of View: PFS

      10:35 - 10:50  |  Presenting Author(s): Mary O’brien

      • Abstract

      Abstract

      This is an old chestnut. The debate has gone on for years and for first line trials in advanced NSCLC it is always OS that wins.

      So why do I think I will win this time?. Well maybe it is because (I was given the task and I like a challenge) we are in a new era of medicine and we are in such a hurry to get new drugs to our patients to prolong our lives and cure them.

      A PFS (DFS) endpoint in a clinical trial is all about the activity of that drug or intervention, it is reported at a point when crossover to other treatments has not happened. It is the standard endpoint for treatments for advanced breast cancer and other disease where there are many treatment options and in general patients will live to get all those treatments options.

      An OS endpoint in advanced lung cancer in the past has been quite similar to PFS – both of them short ie. less than one year. An OS endpoint suggests that a treatment has truly modified the natural history of a disease and therefore this new treatment should be given as early as possible.

      In this era of new treatments it is gratifying to see that an improved PFS has predicted an improved OS e.g. Keynote 024 and it is a gratifying to see that the trial did not report until PFS and OS were available at the same time as in coprimary endpoints. So in these trials of very active treatment PFS reported one year earlier than OS could have cut down the expense of the trial and got the drug to patients earlier.

      Looking at drugs that target the EGFR mutation gefitinib, erlotinib and Afatinib – all positive for PFS and all negative for OS because of crossover. If OS had been the only endpoint, all these drugs would still not be on our pharmacy shelves.

      So I believe if you have a great treatment that should activity very early on then PFS should be the endpoint.

      However if the benefit is small but still significant then OS must be the endpoint.

      But surely at this point in time we only want very active treatments that give us big differences in PFS as we know that these will be the treatments that can change lives.

    • +

      PC05.02 - Debate 1: Which is Most Important Efficacy Endpoint in First Line Trials in Advanced NSCLC PFS or OS - Point of View: OS

      10:50 - 11:05  |  Presenting Author(s): Yu Shyr

      • Abstract

      Abstract not provided

    • +

      PC05.03 - Discussion

      11:05 - 11:10

      • Abstract

      Abstract not provided

    • +

      PC05.04 - Debate 2: Patient Reported Outcomes Should Be the Most Important Endpoint in Trials in the Palliative Setting - YES

      11:10 - 11:25  |  Presenting Author(s): Maurice Pérol

      • Abstract

      Abstract

      The aim of an anticancer treatment in palliative setting, i.e. when there is no curative intent, is to improve the quantity and the quality of life. The evaluation of a treatment effect is mainly based on randomized clinical trials which are meant to provide a correct assessment of the effect size. Therefore, the selection of the primary endpoint is critical not only for the sample size calculation but mainly to judge the relevance and the meaning of the treatment results which define the treatment utility from both the patient perspective and the healthcare system in terms of cost-effectiveness. Two categories of endpoints are used in clinical trials: some are patient-centered, clinical endpoints including overall survival (OS), health-related quality of life (HRQoL) assessment and tolerance profile whereas other are tumor-centered as progression-free survival (PFS), response rate, disease control rate. OS is a universally accepted measure of the treatment benefit in oncology, directly reflecting the quantity of life gained and is still the preferred endpoint of Phase III clinical trials. However, OS as a primary endpoint has some limitations: it usually requires a longer follow-up and a large sample size leading to an increase in the cost of trials. Moreover, especially in cancer with an improved prognosis as for lung cancer depending on an oncogenic driver, OS can be confounded by the effect of crossover and subsequent treatments with small expected OS differences between treatment sequences. These reasons make increasingly difficult to demonstrate an OS benefit especially for first-line therapy. These limitations have led to consider tumor-centered endpoints, mainly PFS, as surrogate endpoints for patient benefit in a clinical trial. PFS provides results earlier than OS and is not influenced by subsequent treatments. However, selecting PFS as primary endpoint assumes that there is a statistical validation of PFS as a surrogate endpoint for OS, which is not the case in advanced NSCLC, especially in palliative setting: many trials have demonstrated a PFS benefit without improvement of OS. Furthermore, there is a lack of consensus for PFS definition or tumor progression (i.e. taking or not into consideration symptomatic progressions), blinding or independent review is required to provide more robust evidence of a PFS benefit and its assessment is dependent on the schedule of tumor evaluations. In this context, HRQoL may constitute an alternative endpoint, which ensures earlier assessment of direct clinical benefit for the patient and takes into account the impact of potential treatment-related toxicities; HRQoL is therefore more recognized as a component endpoint for assessment of the benefit coming from cancer therapy by the American Society of Clinical Oncology and the European Society of Medical Oncology and for drug approval by Regulatory Agencies. It appears obvious that a delay in the time to symptom deterioration, or improvements in symptoms and in HRQoL might be more meaningful to the patient with an advanced lung cancer than the RECIST criteria or PFS, not necessarily reflecting a clinically meaningful benefit.

      HRQOL is a multidimensional concept representing the patient's perception of the effect of illness and treatment on physical, psychological, and social aspects of life. QoL measurement in clinical trials is based upon patient reported outcome (PROs) which are reports of the status of a patient's health condition coming directly from the patient, without any interpretation by a clinician. Many validated self-completion QoL questionnaires for cancer patients are available, taking into consideration the multidomain aspect of QoL; the EORTC QLQ-C30 with its specific module for lung cancer (LC-13), the FACT-L are the most frequently used for assessment of patients HRQL, physical functioning, or tumor related symptoms in lung cancer. HRQoL evaluated with PROs has been shown to be an independent prognostic factor in advanced lung cancer. However, the complexity and heterogeneity of QoL data reports and numerous methodological issues for QoL analyses make the PROs difficult to understand for clinicians and limit their utility to help for therapeutic decisions. Indeed, HRQoL is challenging to assess and raises many methodological issues, including missing data and patient drop-out rate, multiplicity, longitudinal analysis and lack of standardization making impossible to perform cross trials comparisons. These issues as well as the usual deferred publication of QoL results after tumor-centered endpoints communication have led the scientific community and the clinicians to give little attention to PROs.

      Nevertheless, there is a greater awareness that a similar degree of scientific rigor should be applied to the PRO strategy to make it a key component of treatment effect, likely more sensitive to assess therapy's effect on the disease and the patient than conventional measures. Different approaches have been underscored to improve the use of PROs in clinical trials. Recent extensions of the CONSORT have suggested recommendations for PROs analysis and report in clinical trials, with the need to define PROs as a primary or secondary endpoint with a specific statistical defined hypothesis, including approaches to deal with missing data; it is then possible to consider HRQoL as a co-primary endpoint. New statistical methods have also been suggested for longitudinal analyses of PROs as linear mixed model for repeated measures or time until definitive deterioration for symptoms or QoL score deterioration. Other approaches suggest focusing on three well-defined concepts: symptomatic adverse events, physical function, and disease-related symptoms which are easier to assess and are key components of a treatment effect. Information after the treatment and during subsequent treatment should also be collected to take into consideration the impact of a delayed disease progression on QoL or those of subsequent treatment or residual toxicities.

      As tremendous advances have been made in the field of lung cancer treatment, especially in advanced disease with its transformation into a chronic disease for some patients, the treatment focus has shifted to improve or preserve the patients QoL. It is therefore mandatory to develop new and validated endpoints that better indicate a clinical benefit rather than small variations in tumor size. PROs should not be viewed anymore as a surrogate endpoint, but regarded as an important patient-centered endpoint, needing an optimization of trials design and methods to analyze HRQoL.

    • +

      PC05.05 - Debate 2: Patient Reported Outcomes Should Be the Most Important Endpoint in Trials in the Palliative Setting - NO

      11:25 - 11:40  |  Presenting Author(s): Rathi N Pillai

      • Abstract

      Abstract not provided

    • +

      PC05.06 - Discussion

      11:40 - 11:45

      • Abstract

      Abstract not provided

    • +

      PC05.07 - Let's Not Forget Chemo-Regret

      11:45 - 11:55  |  Presenting Author(s): Patricia J. Hollen

      • Abstract

      Abstract

      Background: Although thousands of patients annually receive treatment for advanced non-small cell lung cancer (NSCLC), little is known about their views on the decision to receive that treatment. Even with encouraging newer treatments, survival rates continue to be low in NSCLC. With potential treatment toxicities and with modest response rates, there are many risks for regret by both patients and their supporters. The highly symptomatic nature of NSCLC, coupled with pressures to decide rapidly on therapy, creates challenges that affect quality decision making. In a recent review of 59 studies dealing with regret (Becerra Perez et al, Med Decis Making; 2016;36:777-790), no study focused on lung cancer, even with 66% of the studies in oncology settings, none reported reduced regret over time during the study, and only 2 studies reported delayed regret several months after the decision. In this review, risk factors for regret seemed to be able to be affected by clinicians, according to the opinion of the authors of this systematic review. In a large survey (Morse et al, J Psychosocial Oncol, 2014;32:112-123) results including more than 3500 patients with cancer. More than 90% of these individuals rated "making decisions about care" as being important to them, out of 26 issues. The patients with lung cancer, ranked decision making as the third most important issue of the 26 surveyed. In a recent article in the New England Journal of Medicine (Groopman and Hartzband 2017;377:1507-1509), the authors state that the risk of regret is in almost every medical decision a patient makes and the impact of regret is underestimated in patient decision making. They observe that regret and its impacts are not sufficiently considered by physicians as part of the decision-making process. A profile of the extent of regret, and factors contributing to that regret, is lacking for those undergoing systemic therapy for lung cancer, and for their supporters. Identifying patients early and predicting patients at high risk for regret during treatment could assist in early intervention, which may decrease eventual regret.

      Methods: We sought to investigate prospectively the incidence, extent, and associated factors concerning regret in patients receiving systemic treatment for advanced NSCLC. All patients were enrolled into a phase III trial which included 164 patients. All patients were monitored using the LCSS PRO assessment questionnaire. Patients were randomized 1:1 to either usual care or to enhanced care. In the enhanced care group, all patients also completed the DecisionKEYS decision aid, and their oncologists were aware of the ongoing LCSS results. Also, in the enhanced care group, the results of the LCSS symptomatic and quality of life scores over time, and the results of the decision aid were discussed with the patient and supporter. The LCSS was completed every 3 weeks using the electronic LCSS (eLCSS-QL). The LCSS analysis included using the 3-Item Global Index (“3-IGI,” assessing: distress, activities, and quality of life). All patients, in both arms of the study, completed the Decision Regret Scale (Brehaut et al, Med Decis Making 2003;23:281-292) at 3 months after starting treatment, and at the conclusion of treatment.

      Results: 164 patients signed consent and 160 received treatment (the modified intent to treat group). Patient characteristics included: 43% women; 92% Stage IV; median PS = 1; mean age 63, and 46% of patients represented minority groups. 86% of patients alive at 3 months completed the Decision Regret Scale. Results combined the 2 top categories indicating the greatest extent of regret. By this criterion, 13% of patients (95% CI: 7.4% - 19.2%) expressed regret at the 3-month timepoint after starting chemotherapy. Of note, nearly all patients expressing regret at this point could be identified at the 6-week LCSS evaluation. Of the patients expressing regret at the 6- week mark after starting treatment, nearly all of the patients had a 20% decline or greater by 6 weeks in the 3-IGI score compared with baseline. Specifically, when asked if they would make the same decision again, only 1% not having the 20% 3-IGI decline had regret, as contrasted to 15% with 20% or greater 3-IGI decline (p = 0.01). Results were similar in both randomization groups. Additionally, the supporters of these patients were also asked to complete the Decision Regret Scale. Only a subset of the supporters filled out the regret scale; of those 41 supporters, 3 (7%, 95% confidence interval: 1.5% to 19.9%) expressed regret. This indicates a similar extent of regret among both patients and their supporters.

      Conclusions: Regret is not common in patients receiving systemic therapy for NSCLC or their supporters. It is a positive finding that regret is found in only about one of eight patients undergoing systemic treatment or their supporters; however, it should be acknowledged that this regret is unfortunate for those individuals, and steps taken to modify this regret could contribute to enhanced care. Our findings clearly demonstrate that nearly all patients ultimately having regret could be identified much earlier using the 3-IGI of the LCSS measure. Identifying patients at high risk for regret allows for interventions which could address potential regret in patients and their supporters.

      Support: NIH/NCI R01 CA157409

    • +

      PC05.08 - Discussion

      11:55 - 12:00

      • Abstract

      Abstract not provided

  • +

    ISS14 - Symposium Supported by Novartis: Expert Perspectives on the Management of NSCLC in Clinical Practice: Are Novel Approaches Realistic? (Not IASLC CME Accredited)

    • Moderators:
    • +

      Symposium Supported by Novartis: Expert Perspectives on the Management of NSCLC in Clinical Practice: Are Novel Approaches Realistic?

      12:00 - 13:30  |  Presenting Author(s): Daniel S.W. Tan, Scott A. Laurie, Laura Q Chow

      • Abstract

      Abstract not provided

  • +

    EXH02 - Thermo Fisher Exhibit Showcase Session - The Way We Test Matters for Patient Outcomes (Not IASLC CME Accredited)

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

      EXH02.01 - Thermo Fisher Exhibit Showcase Session - The Way We Test Matters for Patient Outcomes

      12:15 - 13:15  |  Presenting Author(s): Anagh Vora

      • Abstract

      Abstract not provided

  • +

    MA15 - Colliding Approaches - EGFR and Immunotherapy

    • Type: Mini Oral Abstract Session
    • Track: Targeted Therapy
    • Moderators:
    • +

      MA15.01 - Strong PD-L1 Expression Predicts Poor Response and de Novo Resistance to EGFR TKIs Among Non-Small Cell Lung Cancer Patients with EGFR Mutation

      13:30 - 13:35  |  Presenting Author(s): Shan Su  |  Author(s): Zhong-Yi Dong, Xie Zhi, Jian Su, Zhi-Hong Chen, Jin-Ji Yang, Hai-Yan Tu, Qing Zhou, Wen Zhao Zhong, Xu-Chao Zhang, Yi-Long Wu

      • Abstract

      Background

      This study evaluated whether tumor expression of programmed death-ligand 1 (PD-L1) predicted the response of EGFR-mutated non-small cell lung cancer (NSCLC) to epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs).

      Method

      We retrospectively evaluated patients who received EGFR-TKIs for advanced NSCLC between April 2016 and September 2017 at the Guangdong Lung Cancer Institute. None of them were enrolled in clinical studies, and their EGFR and PD-L1 statuses were simultaneously evaluated.

      Result

      Among the 101 eligible patients, strong PD-L1 expression significantly decreased the objective response rate (ORR) compared with those with weak or negative PD-L1 expression (35.7% vs 63.2% vs 67.3%, P=0.002) as well as shortened progression-free survival (PFS, 3.8months vs 6.0months vs 9.5months, P<0.001), regardless of EGFR mutation types (19del or L858R). Furthermore, positive PD-L1 expression was predominantly observed among patients with de novo resistance rather than acquired resistance to EGFR-TKIs (66.7% vs 30.2%, P=0.009). Notably, patients with de novo resistance had a high proportion of dual positive for PD-L1 and CD8 (46.7%, 7/15). Finally, one patient with de novo resistance to EGFR-TKIs and dual positivity for PD-L1 and CD8 experienced a favorable response to anti-PD-1 therapy.1.png

      Conclusion

      This study uncovered the adverse impact of PD-L1 expression on the efficacy of EGFR- TKIs, especially in those with de novo resistance NSCLC, which inclined to reshape an inflamed immune phenotype of dual positive for PD-L1 and CD8 and showed potential therapeutic sensitivity to PD-1 blockade.

    • +

      MA15.02 - Long-Term Safety and Clinical Activity Results from a Phase Ib Study of Erlotinib Plus Atezolizumab in Advanced NSCLC

      13:35 - 13:40  |  Presenting Author(s): Charles M. Rudin  |  Author(s): Andres Cervantes, Afshin Dowlati, Benjamin Besse, Brigette Ma, Daniel B Costa, Peter Schmid, Rebecca Heist, Victoria M. Villaflor, Indrani Sarkar, Genevive Hernandez, Paul Foster, Jessica Spahn, Carol O'Hear, Scott N. Gettinger

      • Abstract

      Background

      Patients with EGFR mutation–positive non-small cell lung cancer (NSCLC) achieve favorable outcomes with EGFR tyrosine kinase inhibitors (TKIs); however, the long-term efficacy of these agents is limited by development of acquired resistance. Atezolizumab (anti–PD-L1 mAb) monotherapy has shown tolerability and durable clinical activity in NSCLC. By selectively targeting PD-L1 to block its interaction with receptors PD-1 and B7.1, atezolizumab can reinvigorate anti-cancer T-cell activity. Thus, combining atezolizumab and erlotinib could result in improved anti-tumor immunity and durable anti-tumor effects. Safety and preliminary clinical activity from a Phase Ib study of erlotinib plus atezolizumab in locally advanced or metastatic NSCLC have been previously reported (Rudin, et al. WCLC 2016). Here we describe updated findings from this study (NCT02013219).

      Method

      EGFR TKI-naive patients with NSCLC were enrolled into a safety-evaluation stage (Stage 1) regardless of EGFR status, and an expansion stage (Stage 2) enrolled patients with EGFR-mutant NSCLC who were previously untreated or treated with 1 prior non–EGFR TKI therapy. A 7-day run-in period with erlotinib 150 mg PO QD was followed by addition of atezolizumab 1200 mg IV q3w. The primary endpoint was safety/tolerability of the combination; secondary endpoints included clinical activity per RECIST v1.1.

      Result

      At data cutoff (August 18, 2017), 28 patients (Stage 1, n = 8; Stage 2, n = 20) were evaluable for safety, and the median survival follow-up was 26.0 months (range, 7.8-32.9; Stage 2). The median age was 61 years (range, 47-84), and the most common EGFR mutation was exon 19 deletion (44%). Grade 3 treatment-related AEs (TRAEs) were reported in 43% of patients; no Grade 4 or 5 TRAEs occurred. The most common TRAEs were increased ALT, pyrexia, rash and diarrhea (2 patients each). Serious AEs occurred in 54% of patients; treatment-emergent AEs led to atezolizumab discontinuation in 18% and erlotinib discontinuation in 11%. Clinical activity was evaluated in Stage 2 patients. ORR was 75% (95% CI: 50.9, 91.3), with a median DOR of 16.7 months (range, 4.2-26.0+). Median PFS was 15.4 months (95% CI: 8.4, not estimable [NE]), and median OS was 32.7 months (95% CI: 32.7, NE).

      Conclusion

      Atezolizumab plus erlotinib demonstrated a tolerable safety profile and compared favorably with prior reports of efficacy with erlotinib monotherapy. OS data are expected to mature and improve with longer follow-up; updated clinical and biomarker data will be presented. Further investigation of the combination is warranted to assess its full potential.

    • +

      MA15.03 - PD-L1 Expression in Untreated EGFRm Advanced NSCLC and Response to Osimertinib and SoC EGFR-TKIs in the FLAURA Trial

      13:40 - 13:45  |  Presenting Author(s): Suresh S. Ramalingam  |  Author(s): Helen Brown, Johan F. Vansteenkiste, Kazuhiko Nakagawa, Manuel Cobo Dols, Thomas John, Craig Barker, Alexander Kohlmann, Alexander Todd, Matilde Saggese, Juliann Chmielecki, Aleksandra Markovets

      • Abstract

      Background

      In the Phase III FLAURA trial (NCT02296125), osimertinib significantly improved PFS relative to SoC EGFR-TKIs (gefitinib/erlotinib) in patients with untreated Ex19del/L858R positive (EGFRm) NSCLC. EGFRm NSCLC tumors can exhibit high PD-L1 expression, an important biomarker for immunotherapy treatment decisions. The frequency and clinical relevance of exhibiting both biomarkers prior to treatment are unclear. We report PD-L1 expression in patients with EGFRm advanced NSCLC and association with clinical outcomes following EGFR-TKI treatment.

      Method

      Tissue samples from 994 patients with advanced NSCLC were screened for EGFR Ex19del/L858R mutations for enrolment in FLAURA; 556 were randomized to treatment. 197 tissue-blocks from the screened population (including EGFR mutation-positive and -negative samples) were tested for PD-L1 using the SP263 (Ventana) immunohistochemical assay; positive tumour cell (TC) staining PD-L1 TC≥25% and TC≥1% thresholds were applied. PFS was investigator-assessed, per RECIST 1.1, according to PD-L1-expressers (TC≥1%) or -negatives (TC<1%) in randomized patients.

      Result

      193/197 blocks had sufficient tumor tissue for staining. 65/193 patients were EGFR mutation-negative. 128/193 patients were EGFR mutation-positive: 106/128 were randomized to treatment (osimertinib: 54; SoC: 52). The table presents PD-L1 expression according to EGFR mutation status. For PD-L1-expressers (TC≥1%), median PFS was 18.4 months for osimertinib and 6.9 months for SoC (HR 0.30 [95% CI 0.15, 0.60]). For PD-L1-negative patients (TC<1%), median PFS was 18.9 months for osimertinib and 10.9 months for SoC (HR 0.37 [95% CI 0.17, 0.74]).

      PD-L1 TC≥1%, n (%)

      PD-L1 TC≥25%, n (%)

      EGFR mutation-negative (n=65)

      Screened population (n=65)

      44 (68)

      23 (35)

      EGFR mutation-positive (n=128)

      Screened population (n=128)

      65 (51)

      10 (8)

      Randomized to treatment (n=106)

      52 (49)

      8 (8)

      Randomized to osimertinib (n=54)

      28 (52)

      3 (6)

      Randomized to SoC EGFR-TKI (n=52)

      24 (46)

      5 (10)

      Conclusion

      There was PFS benefit with osimertinib versus SoC regardless of whether tumors were PD-L1-expressers (TC≥1%) or -negatives (TC<1%). Using the TC≥25% threshold, PD-L1 prevalence was lower in EGFR mutation-positive than mutation-negative samples; there were insufficient patients with TC≥25% tumors for PFS assessment.

      These results support the efficacy of EGFR-TKIs, including osimertinib, as first-line treatment of EGFRm advanced NSCLC, irrespective of PD-L1 expression.

    • +

      MA15.04 - Discussant - MA 15.01, MA 15.02, MA 15.03

      13:45 - 14:00  |  Presenting Author(s): Adrian Sacher

      • Abstract

      Abstract not provided

    • +

      MA15.05 - The Mutational Profiles of EGFR 19 Exon Deletion and 21 Exon L858R Mutation and Their Association with Primary Response to EGFR-TKIs

      14:00 - 14:05  |  Presenting Author(s): Wenhua Liang  |  Author(s): Caichen Li, Yi Zhao, Jianxing He

      • Abstract

      Background

      In addition to the known resistant genetic alterations, such as de novo T790M and TP53 mutations, some recent studies suggested that other concomitant mutations might also compromise the efficacy of EGFR-TKIs. Meanwhile, it has been widely observed that EGFR 19 exon deletion (19del) was associated with better outcomes in treatments with EGFR-TKIs than L858R mutation (L858R). It is of interest to explore whether the respective mutational profiles of 19del and L858R may explain their different sensitivity to EGFR-TKIs.

      Method

      We obtained individual patient data from 7 studies (including 1 from our center) using next generation sequencing to comprehensively determine the mutational profile of EGFR mutated patients. As different panels were used by different studies, we analyzed T790M, TP53 and the loci that were reported by at least 4 studies (half of the included ones).

      Result

      A total of 250 19del and 279 L858R patients were included. We found similar prevalence of all co-mutations (19del 65.2% vs. L858R 64.2%), TP53 (19del 36.8% vs. L858R 40.9%), APC, BRAF, PIK3CA, STK11, PDGFRA, PTEN, AKT1 and KIT between the two types, except for de novo T790M (19del 34.4% vs. L858R 50%, P=0.07), CDKN2A (19del 10.5% vs. L858R 3.4%, P=0.03) and KRAS (19del 0.5% vs. L858R 4.0%, P=0.01). In addition, the number of co-mutations was equivalent between the two types (19del 0.95±0.95 vs. L858R 0.93±0.90, P=1.00). Through multivariate logistic regression, we found that EGFR mutation type, de novo T790M, TP53 and other co-mutations, all had independent adverse effects on primary response rate. Consistently, we found that in those without co-mutations, 19del still showed better outcomes compared with L858R (pure EGFR mutations: 19del 88.0% vs. L858R 68.4%), and that the response rates in both mutation types decreased in the same extent when contaminated with other mutations (EGFR mutations with co-mutations: 19del 53.1% vs. L858R 30.0%).

      Conclusion

      Our study demonstrated that the presence of T790M, TP53 and other concomitant mutations indicated poor response to EGFR-TKIs, but most of them showed similar prevalence between 19del and L858R. Patients with L858R showed significantly less response than those with 19del regardless of the co-mutation status, indicating that some intrinsic factors and potentially de novo T790M, rather than other co-mutations, might underlie the different sensitivity of 19del and L858R to EGFR-TKIs.

    • +

      MA15.06 - Circulating Tumor DNA Portrays the Resistance Landscape to a Novel Third Generation EGFR Inhibitor, AC0010

      14:05 - 14:10  |  Presenting Author(s): Yi-Chen Zhang  |  Author(s): Qing Zhou, Zhi-Hong Chen, Chong-Rui Xu, Jin -Ji Yang, Xu-Chao Zhang, Xiao-Yan Bai, Hong-Hong Yan, Zhi Xie, Wen -Feng Li, Shannon Chuai, Jun-Yi Ye, Han Zhang-Han, Zhou Zhang, Yi-Long Wu

      • Abstract

      Background

      In a Phase I/II dose-escalation and expansion study conducted at Guangdong Lung Cancer Institute, AC0010 demonstrated promising efficacy and good tolerability in advanced NSCLC patients with EGFR T790M-mediated resistance to previous EGFR TKIs, (NCT02330367). As disease progression (PD) with EGFR T790M-directed therapy also emerges over time, we investigated the resistance mechanisms to AC0010 in this study.

      Method

      Serial ctDNA samples obtained from patients who developed PD with AC0010 were analyzed using ultra-deep sequencing capturing 295 cancer-related genes. Alterations that were absent before treatment and acquired at PD or that increased in abundance during treatment were identified as putative resistance mechanisms.

      Result

      Longitudinal plasma samples were obtained from 23 patients who progressed on AC0010 (data cut-off October 14, 2016; figure1). Putative resistance mechanisms to AC0010 were identified in 19/23 patients (>1 putative resistance mechanism was detected in some patients). EGFR amplification was the predominant resistance mechanism (21.1% [4/19 patients]), followed by TP53 loss of heterozygosity (15.8% [3/19]). EGFR C797S mutation, Met amplification and mutations in the PI3KCA pathway each occurred in 10.5% of patients (2/19). SCLC transformation, ERBB2 amplification, CD79A_A32G mutation, CDKN2A_R80 mutation, CRLF2 amplification, MLH1 amplification, Rb1 loss, and concurrent rise in the allelic fraction of tumor suppressor gene TP53 and Rb1 were each detected in 5.3% of patients (1/19). In a patient with PD following single-agent AC0010 and EGFR amplification as the putative resistance mechanism to AC0010, subsequent treatment with AC0010 plus nimotuzumab (EGFR monoclonal antibody) successfully overcame resistance, resulting in a response that lasted for 7.7 months.

      fig 1_study profile.jpg

      Conclusion

      The resistance landscape to AC0010 appears to differ from that described previously with osimertinib. In this cohort of patients in China, EGFR amplification was the predominant resistance mechanism to AC0010 and could be potentially overcome by EGFR dual inhibition.

    • +

      MA15.07 - Different Responses to Osimertinib in Primary and Acquired EGFR T790M-Mutant NSCLC Patients

      14:10 - 14:15  |  Presenting Author(s): Shuyuan Wang  |  Author(s): Bo Zhang, Baohui Han

      • Abstract

      Background

      Primary EGFR T790M could be occasionally identified by routine molecular testing in tyrosine kinase inhibitor TKI-naive non-small cell lung cancer (NSCLC) patients. This study was aimed to compare clinical characteristics of primary and acquired T790M mutations and their responses to Osimertinib in NSCLC patients.

      Method

      We collected clinical characteristics of patients diagnosed with epidermal growth factor receptor (EGFR) mutation from 2012 to 2017 in Shanghai Chest Hospital. For patients with primary and acquired T790M mutations, the responses to Osimertinib were analyzed.

      Result

      Primary T790M was identified in 1.03% (61/5900) of TKI-naive patients. Acquired T790M was detected in 45.50% (96/211) of TKI-treated patients. T790M always coexisted with sensitizing EGFR mutations. Primary T790M was always coexisted with 21L858R (45/61) whereas acquired T790M was coexisted with 19del (61/96). Among them, 18 patients with primary T790M mutation acquired Osimertinib and 75 patients with acquired T790M mutation received Osimertinib. The median progression-free survival (mPFS) of Osimertinib in primary T790M group was greatly longer than that in acquired T790M group (18.0 months:95% CI:15.0-21.0 VS 10.0months:95% CI:8.3-11.7, P=0.016). The DCR of both groups were 89.3% and 100%. In primary T790M group, the mPFS of concomitant occurrence of 20 T790M and 21 L858R or 19del were 15.7m and 24.0 m, respectively. In acquired T790M group, the mPFS of concomitant occurrence of 20 T790M and 21 L858R or 19del were 11.0m and 10.0m, respectively.

      Conclusion

      Primary and acquired T790M-mutation patients showed different molecular characteristics. Both of them may respond to Osimertinib. However, primary T790M patients showed greater survival benefits from Osimertinib than acquired T790M patients.

    • +

      MA15.08 - Discussant - MA 15.05, MA 15.06, MA 15.07

      14:15 - 14:30  |  Presenting Author(s): Charu Aggarwal

      • Abstract

      Abstract not provided

    • +

      MA15.09 - Dynamic Monitoring and Predictive Value of Circulating Tumor Cells in EGFR Mutant Advanced NSCLC Patients Treated with First-Line EGFR-TKIs

      14:30 - 14:35  |  Presenting Author(s): Chunxia Su  |  Author(s): Tao Jiang, Jing Zhao, Xuefei Li, Caicun Zhou

      • Abstract

      Background

      We proposed a non-invasive, folate receptor (FR)-based circulating tumor cell (CTC) assay counts to predict and dynamically monitor the therapeutic response to first-line EGFR-TKIs in patients with EGFR-mutant non-small cell lung cancer (NSCLC).

      Method

      Eligible patients were enrolled and three milliliter (mL) of blood were obtained prior to initial treatment, after one month, and follow-up every two months hereafter. CTCs were isolated based on negative enrichment by immunomagnetic beads and detected by a ligand-targeted polymerase chain reaction (LT-PCR) method.

      Result

      232 patients with EGFR mutations treated with first-line EGFR-TKIs were included. The objective response rate (ORR) of patients with low baseline CTC level (<20.5 FU/3 mL) were significantly higher than those with high baseline CTC level (≥20.5 FU/3 mL) (55.8% vs 38.5%, P = 0.030). Moreover, patients with low baseline CTC level had a markedly longer progression-free survival (PFS) than those with high baseline CTC level (HR = 0.50, P < 0.001). This difference remained significant after multivariate analysis (P = 0.003). Dynamic change of CTC value was significantly associated with partial response (PR) (P = 0.042) and stable disease (SD)/progression disease (PD) (P = 0.032). Of note, dynamic monitoring of CTC provided evidence of resistance to EGFR-TKIs before CT scanning, median: 113 days; range: 45 to 169 days.

      Table 1. Clinical and molecular characteristics of included patients.

       

      CTC < 20.5 (n = 165)

      %

      CTC > 20.5 (n = 52)

      %

      P value

      Age, years

       

       

       

       

       

      Median

      61

       

      63

       

       

      Range

      27-85

       

      40-83

       

       

      Sex

       

       

       

       

       

      Male

      79

      47.88

      28

      53.85

      0.453

      Female

      86

      52.12

      24

      46.15

       

      Smoking status

       

       

       

       

       

      Never-smoker

      122

      73.94

      35

      67.31

      0.351

      Current/ever Smoker

      43

      26.06

      17

      32.69

       

      Pathological type

       

       

       

       

       

      ADC

      150

      90.91

      47

      90.38

      0.909

      ADS

      4

      2.42

      1

      1.92

       

      NOS

      13

      7.88

      4

      7.69

       

      Clinical stage

       

       

       

       

       

      Ⅲb

      9

      5.45

      3

      5.77

      0.794

      156

      94.55

      49

      94.23

       

      Distant metastases

       

       

       

       

       

      Brain

      43

      26.06

      15

      28.85

      0.953

      Bone

      78

      47.27

      21

      40.38

       

      Liver

      10

      6.06

      1

      1.92

       

      Other sites

      109

      66.06

      34

      65.38

       

      No metastases

      12

      7.27

      3

      5.77

       

      Mutation type

       

       

       

       

       

      19DEL

      76

      46.06

      24

      46.15

      0.012

      L858R

      79

      47.88

      19

      36.54

       

      Rare mutations

      10

      6.06

      9

      17.31

       

      Response rate

       

       

       

       

       

      Complete response

      0

      0.00

      0

      0.00

       

      Partial response

      92

      55.76

      20

      38.46

       

      Stable disease

      48

      29.09

      21

      40.38

       

      Progressive disease

      25

      151.50

      11

      21.15

       

      Disease control rate

      140

      84.85

      41

      78.85

      0.310

      Objective response rate

      92

      55.76

      20

      38.46

      0.030

      ADC, adenocarcinoma; ADS, adenosquamous carcinoma; CTC, circulating tumor cell.

      Conclusion

      The current evidences suggest that FR-positive CTCs can be used for both the dynamic monitoring and prediction of outcome in EGFR-mutant NSCLC patients treated with EGFR-TKIs, which could serve as an alternative or supplement to CT scanning.

    • +

      MA15.10 - Loss of T790M Mutation is Associated with Early Progression to Osimertinib in Chinese Advanced NSCLC Patients Harboring EGFR T790M

      14:35 - 14:40  |  Presenting Author(s): Sha Zhao  |  Author(s): Tao Jiang, Xuefei Li, Chao Zhao, Chunxia Su, Shengxiang Ren, Caicun Zhou

      • Abstract

      Background

      Osimertinib has demonstrated striking superior efficacy in non-small cell lung cancer (NSCLC) patients detected acuqired T790M mutation as resistant mechanism to upfront early-generation EGFR-TKIs. However, not all the T790M positive tumors are homogeneously sensitive to osimertinib, and the duration of response often varies. Previous studies suggest that loss of T790M mutation upon progression is related to decreased therapeutic benefit from osimertinib. The aim of this study is to investigate the association of T790M-mutant status and clinical outcomes after osimertinib treatment in Chinese NSCLC patients harboring acquired EGFR T790M mutation.

      Method

      We reviewed the electric medical records of all patients receiving osimertinib monotherapy after detected acquired T790M mutation in rebiopsy after resistant to prior EGFR-TKIs, and underwent re-rebiopsy again upon progression to osimertinib at our hospital. Detailed clinicopathologic characteristics and response data were collected for all patients.

      Result

      From January 2014 to December 2016, 230 patients were confirmed T790M mutation positive for acquired resistance to early-generation EGFR-TKIs (gefitinib, erlotinib, afatinib and icotinib). Among them, 90 patients received osimertinib monotherapy as subsequent treatment. Out of the patients, 84 (93.3%) were eligible for osimertinib-resistance analysis, and 31 (34.4%) patients underwent T790M detection in biopsy after disease progression to osimertinib. The most commonly used biopsy sample were tumor tissue, peripheral blood and hydrothorax. 16 patients remained T790M positive, while 15 patients lost T790M mutation in their re-rebiopsy samples. Loss of T790M upon progression was significantly associated with shorter duration of response to osimertinib (median time 5.93 vs 11.87 m, HR:0.325, 95%CI: 0.087 to 0.45, p=0.0005), while overall survival (OS) was not statistically different between T790M-loss and -remain groups. The objective response rates were also similar in two groups (85% and 100%, respectively). In multivariate analysis, T790M mutation loss remained significantly associated with early progression to osimertinib.

      Conclusion

      Loss of T790M mutation was associated early progression to osimertinib in Chinese NSCLC patients harboring acquired T790M mutation. Dynamic detection during osimertinib treatment may be a potential strategy to timely reveal disease progression.

    • +

      MA15.11 - Real World Biomarker Testing and Treatment Patterns in Patients with Advanced NSCLC Receiving EGFR-TKIs

      14:40 - 14:45  |  Presenting Author(s): Anne Chiang  |  Author(s): Ancilla W. Fernandes, Melissa Pavilack, Jennifer W. Wu, Francois Laliberté, Mei Sheng Duh, Nabil Chehab, Janakiraman Subramanian

      • Abstract

      Background

      In patients who progress on treatment with first- or second-generation EGFR-TKIs, 50–60% will have an EGFR T790M resistance mutation. Osimertinib, a third-generation EGFR-TKI, is FDA approved for use in patients with metastatic EGFR T790M-positive NSCLC and disease progression on or after prior EGFR-TKI therapy, and recently gained additional approval for first-line treatment for patients with EGFR Ex19del/L858R positive advanced NSCLC. We sought to observe how many patients in the real world underwent biomarker testing on progression and subsequently received osimertinib, when T790M positive.

      Method

      Flatiron Health EHR-derived database was used to identify adult patients with NSCLC treated with a first- or second-generation EGFR-TKI from 11/2015–09/2017, with the start of first EGFR-TKI defined as the index date. Patients were stratified by EGFR-TKI use as a first (1L) or later line (2L+) treatment. EGFRm status, including T790M testing and subsequent treatments received after initiating first- or second-generation EGFR-TKI, were described. Chart review was conducted on patients who received a subsequent therapy to confirm disease progression.

      Result

      Patients in this study (n=782; 1L: n=435; 2L+: n=347) had a median age of 69 years, 64% were female, 56% were white, 87% were seen in the community, with a median follow-up of 307 days. During the study period, 30% (235/782; 1L: 96/435 [22%]; 2L+: 139/347 [40%]) of patients died without receiving a subsequent therapy, and 38% (294/782; 1L: 160/435 [37%]; 2L+: 134/347 [39%]) received subsequent therapies. From post-index date to initiation of subsequent therapies, 30% (88/294; 1L 63/160 [39%]; 2L+ 25/134 [19%]) of patients were tested for EGFR mutations including T790M. Among patients who received subsequent therapies, treatments included chemotherapies (1L=23%; 2L+=37%), immunotherapies (1L=16%; 2L+=43%), and targeted therapies (1L=64%; 2L+=30%). On progression, 25% (1L 40/160) and 5% (2L+ 7/134) of the patients received osimertinib. Of those tested for EGFR mutation post-index date (n=88), 28% (n=25) were positive for T790M and 96% of these (n=24) received osimertinib. Most patients (251/294 [85%]; 1L: 136/160 [85%]; 2L+: 115/134 [86%]) on subsequent therapies were confirmed to have disease progression at chart review.

      Conclusion

      In this study, a third of patients were tested for subsequent EGFR mutations including EGFR T790M following treatment with first- or second-generation EGFR-TKI. Of these, a third were positive for T790M and nearly all of the T790M positive patients received osimertinib. These findings highlight low rates of biomarker testing at progression, and the need for optimal treatments that maximize benefits for patients with EGFRm NSCLC.

    • +

      MA15.12 - Discussant - MA 15.09, MA 15.10, MA 15.11

      14:45 - 15:00  |  Presenting Author(s): Joel W. Neal

      • Abstract

      Abstract not provided

  • +

    MA16 - Novel Mechanisms for Molecular Profiling

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

      MA16.01 - Frequency and Genomic Context of Emerging Markers for Molecular Testing in Lung Adenocarcinoma in Cell-Free DNA NGS Analysis

      13:30 - 13:35  |  Presenting Author(s): Vincent K Lam  |  Author(s): Lesli Ann Kiedrowski, Zofia Piotrowska, Anne S. Tsao, Ashley M. Wells, Richard B Lanman, Vassiliki A Papadimitrakopoulou, Rebecca J Nagy

      • Abstract

      Background

      The recently updated CAP/IASLC/AMP lung cancer molecular testing guideline (Lindeman et al 2018) recommends several genes be analyzed by next-generation sequencing (NGS) in lung adenocarcinoma (LUAD), including EGFR, ALK, BRAF, KRAS, and others. It also includes a list of 20 emerging markers (EMs) for molecular testing and suggests practitioners remain aware of these and other genes between guideline updates. We investigated the frequency of genomic alterations (GAs) in several of these EMs in a cohort of patients with advanced lung adenocarcinoma who underwent clinical cell-free DNA (cfDNA) NGS analysis and assessed co-occurrence with canonical driver GAs.

      Method

      Genomic data was reviewed from 6530 patients with at least one GA detected on clinical Guardant360 cfDNA NGS testing (Guardant Health, Inc) with an indicated diagnosis of lung adenocarcinoma from 11/25/16-3/1/18. Synonymous alterations were excluded from further analyses.

      Result

      2600 patients (40%) had at least one nonsynonymous alteration in the EM genes assessed; excluding GAs classified as variants of unknown significance (VUS), 1350 patients (21%) had at least one characterized alteration. Table 1 shows number and frequency of GAs observed per patient by gene and alteration type. Of EMs assessed, GAs were observed most commonly in NF1, PIK3CA, PDGFRA, KIT, and FGFR1-2. GAs in multiple EMs, including RIT1, NRAS, FGFR2-3, NTRK1, KIT, and AKT1, were observed co-occurring with established driver alterations, often in a genomic context consistent with resistance to targeted therapy at allelic fractions suggestive of subclonality.

      table 1.jpg

      Conclusion

      Effective therapies are continually emerging for a growing number of molecular biomarkers in lung cancer. Comprehensive genomic profiling with cfDNA NGS can identify GAs in both recommended and EM genes to guide therapeutic decision-making and catalyze clinical trial enrollment. Further investigation of mutual exclusivity and co-occurrence of established drivers and EMs may reveal novel resistance mechanisms and facilitate identification of rational combination therapeutic strategies.

    • +

      MA16.02 - Prospective Clinical Validation of the InVisionFirst™ ctDNA Assay for Molecular Profiling of Patients with Advanced NSCLC

      13:35 - 13:40  |  Presenting Author(s): Ramaswamy Govindan  |  Author(s): Michael A. Pritchett, Ross Camidge, Manu Patel, Jamil Khatri, Steven Boniol, Elke K. Friedman, Abderrahim Khomani, Samir Dalia, Katherine Baker-Neblett, Vincent Plagnol, Karen Howarth, Nitzan Rosenfeld, Clive Morris

      • Abstract

      Background

      Clinical practice guidelines advocate molecular profiling as a part of the evaluation of advanced NSCLC, with ctDNA based profiling being an option for those with insufficient tissue. Thorough prospective clinical validation studies of NGS based ctDNA assays are lacking. Here we report the multi-centered prospective clinical validation of a ctDNA NGS panel for stratification of patients with advanced untreated NSCLC.

      Method

      InVisionFirst™ (Inivata) is a ctDNA NGS assay for detection of genomic alterations in 36 genes commonly mutated in NSCLC and other cancers. 264 patients with untreated advanced NSCLC were prospectively recruited by 41 US centers. 178 patients had tumour tissue available for molecular profiling (predominantly by NGS) and the remaining 86 patients without tissue were included to compare ctDNA profiles obtained from patients with and without tissue for profiling.

      Result

      A total of 204 patients (77.3%) had detectable ctDNA alterations. Using tissue results as the reference, overall concordance for the full 36 genes in the InVisionFirst™ panel with matched tissue profiling was 97.8% with 82.9% PPV, 98.5% NPV, 70.6% sensitivity and 99.2% specificity. Considering a subgroup of 8 genes that can influence routine clinical patient management (EGFR, ALK, ROS1, ERBB2, MET, BRAF, KRAS, STK11) the PPV was 93.7%, 96.8% NPV, 72.4% sensitivity and 99.4% specificity. Excluding patients with undetectable ctDNA, these figures become 93.7% PPV, 98.4% NPV, 87.3% sensitivity and 99.3% specificity. The observed pattern of genomic changes seen in ctDNA was consistent across patients with and without tissue for profiling. Across the whole study, 44 patients with actionable alterations were identified by ctDNA testing compared to only 36 by tissue testing. 47% of patients tested by ctDNA had an actionable alteration or an alteration that is generally mutually exclusive for such actionable changes such as KRAS or STK11.

      Conclusion

      The InVisionFirst™ assay demonstrates excellent concordance with tissue profiling in this multi-centered prospective clinical validation study. The performance of this assay in terms of overall sensitivity and specificity appears comparable if not higher than other established commercial ctDNA assays. Utilization of InVisionFirst™ ctDNA testing led to the detection of 22% more actionable alterations than standard of care tissue testing in this study supporting its use for the molecular stratification of patients with advanced NSCLC. Further analyses on the features associated with detectable ctDNA signatures are ongoing.

    • +

      MA16.03 - Discussant - MA 16.01, MA 16.02

      13:40 - 13:55  |  Presenting Author(s): Geoffrey R. Oxnard

      • Abstract

      Abstract not provided

    • +

      MA16.04 - Clinical and Molecular Characteristics of EGFR Mutant Lung Cancers with Concurrent TP53 and RB1 Mutations.

      13:55 - 14:00  |  Presenting Author(s): Michael Offin  |  Author(s): Megan Tenet, Ronglai Shen, Natasha Rekhtman, Hira Rizvi, Matthew D. Hellmann, Maria E Arcila, Marc Ladanyi, Charles M. Rudin, Mark G Kris, Helena Yu

      • Abstract

      Background

      20% of patients with metastatic lung adenocarcinoma have activating EGFR-mutations. EGFR-mutant lung cancers can undergo histologic transformation to small cell lung cancer (SCLC) as a response to the selective pressure of EGFR-TKIs in <5% of patients after earlier-generation EGFR-TKIs and have been reported after osimertinib. SCLC nearly universally harbor TP53/RB1-alterations which are rarely seen in EGFR-mutant lung adenocarcinomas. We sought to identify this subset of patients, describe their clinical course and likelihood of SCLC transformation.

      Method

      Retrospective review of targeted next generation sequencing (NGS, MSK-IMACT) at Memorial Sloan Kettering (MSK) was performed to identify patients with concurrent EGFR-activating mutations and TP53/RB1-mutations within the same tumor sample from NGS between April 2014 to February 2018 with a data cutoff of March 2018. For comparison, consecutive patients with lung cancers harboring EGFR-mutations who were EGFR-TKI naïve and TP53/RB1-wildtype were also collected during that time-period.

      Result

      Of the 21% of lung cancer patients with activating EGFR-mutations (759/3662), 5% (40/759) had concurrent TP53/RB1-mutations. 43% (17/40) were female, 58% former-smokers (23/40, median pack-years: 8), and median age of 68 (range 25-86 years). 88% (35/40) were adenocarcinoma at diagnosis, of which 11% (4/35) transformed to SCLC during treatment; 10% (4/40) were de-novo SCLC at diagnosis, and 1 was large cell neuroendocrine. The transformation rate was significantly higher compared to previous work from MSK evaluating EGFR-mutant patients showing 4% (4/155) transformation (p=0.04). Concurrent PIK3CA mutations were more frequently seen in the EGFR/TP53/RB1 mutant group compared to the TP53/RB1-wildtype group (17% (n=6/35) vs 7% (n=4/60), p=0.11). 20 patients were EGFR TKI-naïve at the time of NGS; the median time on EGFR-TKI (ToT) was 7.6 months versus 14.2 months in the TP53/RB1-wildtype group (HR 4.48, p=0.0003). The overall survival (OS) of this cohort versus TP53/RB1-wildtype was not different (4.3 vs 4.1 years, HR 1.35, p=0.51). In the 4 patients with SCLC transformation, the median time to transformation was 2.4 years after a median of 1.5 EGFR-TKI therapies (range 1-5 lines). Median OS from time of transformation was 7 months. 63% (25/40) of the EGFR/TP53/RB1-mutant cohort had brain metastases during their disease course as compared to 50% (n=30) in the TP53/RB1-wildtype group (p=0.30).

      Conclusion

      SCLC transformation is enriched in EGFR/TP53/RB1-mutant lung cancers, occurring in 11% of patients. Once SCLC transformation occurs, overall survival is short. Patients with EGFR/TP53/RB1 have a shorter time on EGFR-TKI. Further investigation into optimal treatment for this subset of EGFR/TP53/RB1 mutant lung cancers is critical.

    • +

      MA16.05 - MET Kinase Domain Rearrangements (KDRE) in Non-Small Cell Lung Cancer (NSCLC) Identified Through Comprehensive Genomic Profiling (CGP)

      14:00 - 14:05  |  Presenting Author(s): Sai-Hong Ignatius Ou  |  Author(s): Alexa B Schrock, Dean Pavlick, Mark Rosenzweig, Rachel L Erlich, Lee A Albacker, Garrett M Frampton, Jeffrey S. Ross, Vincent A Miller, Siraj M Ali

      • Abstract

      Background

      MET is a known oncogenic driver in NSCLC, and activation via various means including gene amplification, exon 14 skipping, and fusion has been reported to be targetable in the clinical setting. However, the prevalence and characteristics of NSCLCs harboring MET kinase fusions as well as diverse KDRE have not been comprehensively assessed.

      Method

      Hybrid-capture based CGP was performed on 32,643 FFPE or 3,076 blood-based ctDNA NSCLC specimens during the course of clinical care. Tumor mutational burden (TMB) was determined on 0.83-1.1 megabases (Mb) of sequenced DNA and is reported as mutations/Mb.

      Result

      In 35,719 unique NSCLC patient cases assayed, RE retaining the MET kinase domain were detected in 63 (0.2%) cases. These included MET fusions with an identified 5’ fusion partner (n=8, 12.7%), probable MET fusions with an unidentified fusion partner (n=32, 50.8%), MET kinase domain duplications (KDD, n=7, 11.1%), MET exon 14 skipping rearrangements (n=6, 9.5%), and MET rearrangements of the 5’ terminus (n=10, 15.9%). Fusion partners identified include CAPZA2 (2 cases), CAV2, CD47, INPPL1, LHFPL3, PRKAR2B and SPECC1. MET kinase-inactivating rearrangements were also detected in the larger cohort (n=10, 0.03%). MET KDRE cases harbored at least one targetable NSCLC NCCN guidelines driver alteration in 27.0% (n=17) of cases including EGFR activating subs (L858R, L861Q) or exon 19 deletions (n=8), MET exon 14 skipping short variants (n=6), ALK fusions (n=2), and a BRAF V600E mutation. KRAS short variants were detected in 6 additional cases. MET amplification also co-occurred in 36.5% of cases with MET KDRE (n=23, median copy number 16). The median TMB in cases with MET KDRE was 6.1 mut/Mb, as compared to a median of 7.0 mut/Mb for NSCLC samples overall. Clinical treatment information for a subset of cases will be presented, including paired cases in which the MET KDRE may represent an acquired mechanism of resistance to EGFR- or ALK-targeted therapy.

      Conclusion

      Potentially targetable MET rearrangements retaining the kinase domain were observed in 0.2% of NSCLCs, which may be an underestimate considering that these assays sequence all MET exons, but do not specifically bait MET intronic breakpoints. These data suggest that MET KDRE, including, but not limited to fusions and KDD, may represent a distinct subset of driver alterations and potential resistance mechanisms. The value of testing for and potentially targeting these alterations as part of routine clinical care in NSCLC warrants further investigation.

    • +

      MA16.06 - EGFR Clonality and Tumor Mutation Burden (TMB) by Circulating Tumor DNA (ctDNA) Sequencing in Advanced Non-Small Cell Lung Cancer (NSCLC)

      14:05 - 14:10  |  Presenting Author(s): Xinghao Ai  |  Author(s): Yingcheng Lin, Jiexia Zhang, Congying Xie, Anwen Liu, Xiaohua Hu, Qiong Zhao, Yuansheng Zang, Chuangzhou Rao, Xiaohua Hu, Lianpeng Chang, Qin Li, Yanfang Guan, Rongrong Chen, Xin Yi, Shun Lu

      • Abstract

      Background

      TKI has significantly improved survival time of NSCLC pts with sensitive mutation. However, pts present different outcome while receiving TKI treatment. We conduct a prospective multicenter clinical trial to determine whether clonality of sensitive mutation is related to the efficacy of TKI. We also evaluate the consistency of TMB between tissue and blood in this cohort.

      Method

      Paired tumor and plasma samples at diagnosis were obtained from systemic treatment naïve pts with advanced NSCLC. DNA was sequenced by target-capture deep sequencing of 1021 previously annotated genes related to solid tumors. Clonal EGFR mutation was defined if EGFR mutation was in the cluster with the highest mean variated allele frequency with PyClone, and otherwise subclonal EGFR mutation. TMB of tissue (tTMB) and blood (bTMB) analysis interrogated single nucleotide variants, small insertion and deletion, with VAF ≥3 % and ≥0.5 %, respectively. TMB-high pts were identified with ≥9 mut/MB (upper quartile of data from geneplus).

      Result

      During February 2017 to April 2018, 127 advanced NSCLC pts were enrolled from 9 centers. A total of 653 somatic variations were detected in tissues. Mutations occurred most frequently in EGFR (57 %), TP53 (54 %), KRAS (9 %), ALK (8 %). In matched plasma, 405 (62 %) tumor-derived mutations were detected by pan-caner panel sequencing. A total of 90 EGFR mutations were detected in 73 pts, most of which occurred in tyrosine kinase domain (L858R, 41%; Ex19del, 33%). Most EGFR mutation were clonal in tissue and plasma, with a consistence of 83 % in paired samples. In addition, bTMB was significantly correlated to tTMB (Pearson r= 0.85, p-value= 1.8e-30), with a consistence of 89 %. Interestingly, high TMB was observed in a small fraction of patients (8 %) with driver mutations, such as mutations in EGFR, ALK fusion, ERBB2 and PIK3CA.

      Conclusion

      Deep sequencing with the pan-cancer panel can effectively detect mutations and evaluate TMB in both tissue and blood with high consistence. EGFR mutations can be clonal or subclonal in both tissue and blood. Prospective multicenter study is ongoing to determine the EGFR clonality as a predictive factor for the TKI efficacy in NSCLC (TRACELib-NSCLC, NCT03059641).

    • +

      MA16.07 - Discussant - MA 16.04, MA 16.05, MA 16.06

      14:10 - 14:25  |  Presenting Author(s): Ben J Solomon

      • Abstract

      Abstract not provided

    • +

      MA16.08 - Clinical Utility of Detecting ROS1 Genetic Alterations in Plasma

      14:25 - 14:30  |  Presenting Author(s): Ibiayi Dagogo-Jack  |  Author(s): Rebecca J Nagy, Lorin A Ferris, Jessica Jiyeong Lin, Justin F Gainor, Richard B Lanman, Alice T. Shaw

      • Abstract

      Background

      ROS1-rearranged lung cancer harbors an oncogenic fusion protein created by the juxtaposition of the ROS1 gene to various fusion partners. Due to the lack of a conserved breakpoint and inclusion of intronic segments, ROS1 rearrangements can be challenging to identify with DNA-based sequencing approaches. The feasibility and clinical utility of detecting ROS1 fusions in circulating tumor DNA is not well established.

      Method

      The Guardant360 de-identified database was queried to identify lung cancer cases with plasma ROS1 fusions and describe the molecular features of the ROS1-rearranged cohort. In addition, we performed longitudinal analysis of plasma specimens from four patients at our institution who were treated with next-generation ROS1 inhibitors after progressing on crizotinib.

      Result

      From review of 24,009 plasma specimens from lung cancer patients, we identified 56 patients with ROS1 fusions. CD74 was the most common of 7 identified fusion partners [n=35 (62%) CD74, n=7 (12.5%) SDC4, n=7 (12.5%) EZR, n=3 (5%) TPM3, n=2 (4%) TFG, and n=1 (2%) each of CCDC6 and SLC34A2]. ROS1 fusions commonly co-occurred with TP53 mutations (n=36, 64%) and genes involved in cell-cycle regulation (n=11, 20%) or the WNT/ß-catenin pathway (n=16, 29%). In 4 (80%) of 5 cases where plasma genotyping occurred at crizotinib progression, we identified a putative resistance mechanism, including a ROS1 resistance mutation in 3 patients (n=2 G2032R & n=1 L2026M) and a BRAF V600E mutation in 1 patient. We analyzed longitudinal plasma specimens from 4 patients with crizotinib-resistant lung cancer who were subsequently treated with a next-generation ROS1 inhibitor (n=3 lorlatinib, n=1 entrectinib). One patient treated with lorlatinib had a pretreatment ROS1 G2032R mutation (in plasma and tissue); plasma analysis revealed stability of the G2032R allelic fraction in the setting of primary progression of pleural disease. Of the 2 patients without pretreatment ROS1 mutations who received lorlatinib, one developed a ROS1 G2032R mutation after initial response to treatment. The second patient experienced primary progression and plasma genotyping revealed low level FGFR1 copy number gain (3.3 copies); pre-crizotinib plasma was not available for comparison. One patient had a plasma PIK3CA E545K mutation at the time of crizotinib progression, and did not respond to next-line entrectinib.

      Conclusion

      Next-generation sequencing can be used to detect ROS1 fusions and resistance mutations in plasma. Longitudinal plasma analysis may provide insight into the activity of investigational drugs against ROS1 mutations that mediate resistance to crizotinib.

    • +

      MA16.09 - Feasibility, Clinical Relevance of ALK/ROS1 Fusion Variant Detection by Liquid Biopsy in Advanced Non-Small Cell Lung Cancer

      14:30 - 14:35  |  Presenting Author(s): Aurélie Swalduz  |  Author(s): Laura Mezquita, Sandra Ortiz-Cuaran, Cecile Jovelet, Virginie Avrillon, David Planchard, Solène Marteau, Gonzalo Recondo, Séverine Martinez, Karen Howarth, Vincent Plagnol, Clive Morris, Emma Green, Luc Odier, Ludovic Lacroix, Stéphane Hominal, Etienne Rouleau, Claire Tissot, Caroline Caramella, Pierre Fournel, Luc Friboulet, Maurice Pérol, Benjamin Besse, Pierre Saintigny

      • Abstract

      Background

      Liquid biopsy offers an alternative non-invasive approach to reflect the tumor genomic landscape of NSCLC patients; however, the potential of liquid biopsies for ALK/ROS1 fusion detection is poorly described. Herein, we evaluated an amplicon-based NGS assay for ctDNA detection of ALK and ROS1 fusions in a large cohort of ALK and ROS1 NSCLC patients and correlation of variants with clinical data.

      Method

      ALK- and ROS1-translocated advanced NSCLC patients, were prospectively enrolled from October 2015 to April 2018 in 9 French institutions. ALK or ROS1 positivity was as confirmed by immunochemistry and FISH or RNAseq. ALK (EML4 variants v1, v2, v3), ROS1 (CD74, SLC34A2, SDC4 and EZR) fusions, and mutations in a panel of 36 NSCLC-associated genes were investigated in ctDNA using InVisionFirst™ (Plagnol V PLoS ONE, 2018).

      Result

      A total of 120 patients were included: 96 ALK and 24 ROS1. 30 samples were collected from patients who were TKI-treatment-naive, 257 during follow-up and 73 at progressive disease (PD) under TKI. The median age was 55 years-old (range 23-75); most patients were female (57%) and had a non-smoking history (59%). At diagnosis, 20% of patients presented with brain metastasis. All patients received at least 1 ALK-TKI (median: 1.6; range:1-6).

      Preliminary results are available for the first 54 patients: 21 at diagnosis and 33 at PD under TKI. ALK/ROS1 fusions were detected in 13/21 patients (62%) at diagnosis: 12/20 ALK-fusions (7 v1, 2 v2 and 3 v3) and in 1/1 ROS1-fusion (CD74-ROS1). No fusion was detected in 8 patients, which may be due to partner genes or variants not covered by this panel. However, 5 of these 8 patients had exclusive thoracic or brain PD.

      Liquid biopsies collected at the radiographic evaluation under therapy revealed complete ctDNA clearance of the fusion when patients experienced PR (n=4). In samples at PD, fusion was detected in 44% of patients (24/55) with evidence of acquired resistance in patients both positive and negative for fusion.

      Results for the remaining samples, correlation between fusion variant and survival, fusion variant and mechanism of resistance will be presented at the Congress.

      Conclusion

      Our results suggest that ctDNA profiling is a promising non-invasive tool for identification of ALK/ROS1 fusions and monitoring of response in advanced NSCLC patients. Systematic identification of the fusion partner may help to better understand the heterogeneity and evolution (sensitivity profile to targeted inhibitors and associated-mechanisms of resistance) of NSCLC driven by ALK and ROS1 rearrangement.

    • +

      MA16.10 - Clinical Utility of Cerebrospinal Fluid Cell-Free DNA for Clarifying Genetic Features of Leptomeningeal Metastases in ALK Rearrangement NSCLC

      14:35 - 14:40  |  Presenting Author(s): Meimei Zheng  |  Author(s): Yangsi Li, Ben-Yuan Jiang, Hai-Yan Tu, Xu-Chao Zhang, Jun-Yi Ye, Wen-Fang Tang, Shannon Chuai, Yi-Long Wu

      • Abstract

      Background

      Leptomeningeal metastases (LM) were associated with a poor prognosis in non small cell lung cancer (NSCLC). LM were much more frequent in EGFR mutant patients, and cerebrospinal fluid (CSF) cell-free DNA (cfDNA) has shown unique genetic profiles of LM in patients harboring EGFR mutations in our previous studies. However, studies in ALK positive NSCLC patients with LM are scarce.

      Method

      Lung cancer patients with ALK rearrangement were screened from Sept 2011 to Feb 2018 at our institute. Leptomeningeal metastases were diagnosed by MRI or CSF cytology or next-generation sequencing (NGS) of CSF cfDNAs. Paired plasma were also tested by NGS.

      Result

      LM were diagnosed in 22 (7.6%) of 288 ALK rearrangement patients with lung cancer. A total of 11 ALK positive patients with LM were enrolled with CSF cfDNA tested by NGS (one case used CSF precipitates instead of CSF cfDNA). Paired plasma were available in 11 patients. Driver genes were detected in 75.0% CSF samples and 45.5% plasma respectively (P=0.214). Max allele fractions were higher in CSF cfDNA than in plasma (40.8% versus 0%, P=0.021). ALK variant 1 (E13:A20) was detected in 3 cases of CSF and paired plasma, respectively. ALK variant 2 (E20:A20) was identified in 5 cases of CSF and 1 paired plasma. Multiple copy number variants (CNV) were mainly found in CSF cfDNA, including EGFR copy number gains. Resistance mutations including gatekeeper gene ALK G1202R was identified in CSF cfDNA with ALK variant 1 and ALK G1269A was detected in plasma. The detection rate of TP53 was 45.4% versus 27.3% in CSF cfDNA and plasma.

      figures.jpg

      Conclusion

      CSF cfDNA was more sensitive than plasma to reveal genetic features of ALK-fusion LM, confirming its role as a liquid biopsy medium for LM in driver gene positive NSCLC.

    • +

      MA16.11 - Discussant - MA 16.08, MA 16.09, MA 16.10

      14:40 - 14:55  |  Presenting Author(s): Robert C. Doebele

      • Abstract

      Abstract not provided

    • +

      MA16.12 - Q&A

      14:55 - 15:00

      • Abstract

      Abstract not provided

  • +

    MA17 - New Methods to Improve Lung Cancer Patients Outcomes

    • Type: Mini Oral Abstract Session
    • Track: Nursing and Allied Professionals
    • Moderators:
    • +

      MA17.01 - A Sense of Understanding and Belonging When Life Is at Stake – Operable Lung Cancer Patients’ Lived Experiences of Participation in Exercise

      13:30 - 13:35  |  Presenting Author(s): Malene Missel  |  Author(s): Mai Nanna Schoenau, Britt Borregaard

      • Abstract

      Background

      Exercise has been introduced to improve physical capacity and quality of life and to reduce symptoms and side effects of treatment in surgically treated non-small cell lung cancer (NSCLC) patients. The effects of an exercise programme for this patient group has been tested in a randomized controlled trial – the PROLUCA study. The questions though, of how patients experience participation in group-based exercise studies and the impact of the shared community with fellow patients has not been previously examined. The objective was to explore lived experiences and social benefits among patients with operable NSCLC who participated in an exercise programme (the PROLUCA study) post-surgery.

      Method

      Nineteen patients enrolled in an exercise intervention two weeks post-surgery participated in qualitative interviews at three time points. A phenomenological hermeneutical approach comprised the epistemological stance and the methodological basis was Ricoeur’s narrative philosophy. The goal of the analysis and interpretation was to provide descriptions that captured the meaning of the lived experiences of the patients.

      Result

      Patients included in this qualitative study had a mean age of 63 years (range 48-75), 58% were female, and 68% was retired. Eighty-four percent had performance status 0 (WHO) and almost all patients were used to some kind of physical activity. The analysis revealed social benefits of taking part in the group-based exercise intervention. The patients experienced themselves as part of a community, and the physical exercise intervention was significant in terms of the patients’ social capital. In this sense, patients gained access to resources that derived from human interaction in the exercise group, and their illness and treatment became easier to manage when shared with others in the same situation. The exercise intervention helped to create a community for patients after lung cancer surgery, and the patients experienced a feeling of belonging and equality with the other participants.

      Conclusion

      The group based exercise intervention created opportunities for mutual understanding between patients, making illness and treatment easier to manage. The patients experienced support to gain renewed balance in life during the exercise intervention in the interaction with peers in the group. It is relevant to inform operable NSCLC patients about the potential community of understanding and belonging in group-based exercise interventions.

    • +

      MA17.02 - Early Accrual to a Precision Lung Cancer Survivorship Intervention: The Kentucky LEADS Collaborative Lung Cancer Survivorship Care Program

      13:35 - 13:40  |  Presenting Author(s): Jamie L Studts  |  Author(s): Jessica L Burris, Michael A Andrykowski, Tara Schapmire, Barbara Head, Maureen Rigney, Angela Meredith Criswell, Susanne M Arnold, Allyson R Yates, Courtney Blair, Amy Christian

      • Abstract

      Background

      Recent advances in early detection and treatment of lung cancer have created a need for survivorship care interventions to reduce the psychosocial and symptom burden of lung cancer, but few interventions address the unique experience of lung cancer survivors and their caregivers. Leveraging shared decision making and motivational interviewing, the Kentucky LEADS Collaborative developed a precision psychosocial intervention addressing the unique experiences and challenges of individuals diagnosed with lung cancer and their caregivers. This sub-study describes the demographic, diagnostic, and psychosocial characteristics of the initial participants in the Kentucky LEADS Collaborative Lung Cancer Survivorship Care Program.

      Method

      Participants include 61 lung cancer survivors across 9 lung cancer care sites in Kentucky, USA. Data were drawn from baseline surveys of demographic characteristics, disease/treatment information, symptom burden, psychosocial functioning and quality of life administered to lung cancer survivors and caregivers enrolled in the single-arm intervention trial.

      Result

      Of the first 61 LC survivors enrolled, 32 had a caregiver join them as participants in the intervention (53%). Participants had a mean age of 62 years. Approximately 20% of LC survivors did not have a caregiver available to participate, and 27% declined to invite a caregiver join the program. Most participating caregivers were spouses (63%), but siblings (10%) and children (19%) were also included. Most survivors were female (66%), Caucasian (97%), and covered by health insurance (95%), and 59% were married or living in a committed relationship. Most participants had been diagnosed with non-small cell lung cancer (84%) and late-stage disease (IIIB-IV; 53%). Most participants had a history of smoking (95%); 30% had smoked within the past 30 days, and 29% were current smokers. Among current smokers, participants reported very high levels of quit planning (9.23±2.77) and quitting confidence (9.14±2.89). Finally, approximately 55% reported clinically significant distress, with a mean level of distress of 3.98 (2.99) on a scale from 0-10.

      Conclusion

      Early accrual to the trial has exceeded expectations. Most survivors had advanced disease and reported significant distress. A substantial minority continued to use tobacco. Data suggest that modifications made to the survivorship approach emphasizing empathy and patient preference may help improve intervention acceptability and feasibility. Subsequent analyses will evaluate the impact of the intervention on quality of life, psychosocial functioning, and symptom burden. Data will also be collected regarding acceptability of the intervention and potential program changes to optimize benefits.

    • +

      MA17.03 - Shared Decision-Making for Patients with Advanced Non-Small Cell Lung Cancer

      13:40 - 13:45  |  Presenting Author(s): Mette Kargo Jensen  |  Author(s): Karin Piil, Gitte Fredberg Persson, Seppo W Langer, Mette Vinter, Katrina Pitt Winther, Susanne Friis-Haché, Mette Pøhl

      • Abstract

      Background

      Lung cancer is the leading cause of cancer-related death in the world and more than half of the patients have metastatic disease at the time of diagnosis. Although, treatment options are developing rapidly, most patients are facing a poor prognosis. The role of 3rd or 4th line treatment with chemotherapy remains controversial with sparse evidence of efficacy. Therefore, the patient’s preferences become central. Shared decision-making enables the patients to be actively involved in choosing the treatment option that best reflects both medical evidence and individual preferences.

      This study examines how patients with lung cancer and their relatives are empowered and supported when they have to make informed choices regarding 3rd or 4th line of treatment. The aim was to develop a model for shared decision-making and to test decision aid tools that enable a collaborative process that takes into account the best available scientific evidence, as well as the patient's values and preferences.

      Method

      Patients diagnosed with advanced non-small cell lung cancer, their relatives and the health care professionals were involved in the process that included: 1) Multidisciplinary workshops and workshops with patients and relatives, 2) Training course in communication on existential issues and shared decision-making for health care professionals, 3) Designing and testing five decision aid tools, 4) Creating a Podcast and 5) Evaluation by patient satisfaction surveys.

      Result

      Three strategic focus areas were identified: 1) The meaningful service, 2) considerations in end-of-life care and 3) patient involvement in decision making. The patient reported quality of communication was increased during the study period. The patient satisfaction surveys (n=77 baseline) and (n=60 final evaluation) demonstrated statistical significant improvements from baseline to final evaluation in regard to:1) involving patients in the treatment decisions to the extent they prefer (Pearson Chi-Square, P=0.048) and 2) encouraging patients to ask questions (Pearson Chi-Square, P=0.008). The study improved the health care professionals understanding of the importance of incorporating patients in shared decision-making processes in clinical practice. However, some barriers for implementation were identified, such as changing established behaviour among health care professionals.

      Conclusion

      The findings indicate that decision aid tools are useful and related to significant changes in patient experience of the quality of communication. We suggest investigating the feasibility and potential concerns of integrating these tools to a larger extend in clinical practice.

    • +

      MA17.04 - Discussant - MA 17.01, MA 17.02, MA 17.03

      13:45 - 14:00  |  Presenting Author(s): Anne Fraser

      • Abstract

      Abstract not provided

    • +

      MA17.05 - Development of a Telephone Clinic for Patients Undergoing Long Term Follow-Up After Thoracic Surgery

      14:00 - 14:05  |  Presenting Author(s): Jenny Mitchell  |  Author(s): Calum Buchanan, Sarah Malone, Francesco Di Chiara, Dionisios Stavroulias, Elizabeth Belcher

      • Abstract

      Background

      Patients undergoing long term follow-up after lung cancer surgery in our institution follow an imaging based follow-up programme. Protocol led CT imaging followed by an out-patient appointment is undertaken every 6 months for two years after surgery then annually until year 5. Feedback from patients indicated they find two trips to the hospital burdensome and they frequently requested results of surveillance imaging over the telephone. Limited capacity in the thoracic surgery clinics led to long waits for an appointment to be informed of imaging results. To address these issues, we developed a model of nurse led telephone follow-up after surveillance imaging.

      Method

      A proposal to hold one telephone clinic per week was made to commissioners in the autumn of 2016. Following approval, the telephone clinic commenced in April 2017. Patients are triaged by the specialist nurse when CT results are available and allocated to the telephone clinic if appropriate. They are given a timed appointment and the telephone number they will be contacted on is confirmed prior to the appointment. A database is completed during the appointment, a record of the consultation is made in both paper and electronic patient records and a letter is sent to the GP and other teams who have contact with the patient. Patients with significant abnormalities on CT imaging are referred for discussion by the multidisciplinary team and seen in a face to face clinic.

      Result

      In the first twelve months (April 2017 to March 2018) there were 254 patient appointments in 51 telephone clinics. Average call length is 10 minutes with a range of 3 to 22 minutes. One patient scheduled for a telephone appointment was not contactable at the appointed time (0.4%). Satisfaction with the clinic is high with 98% of patients requesting their next follow-up appointment in the telephone clinic. Clinic capacity was increased at reduced cost to commissioners as a telemedicine appointment is charged at £25.34 compared to £70.16 for a face to face appointment.

      Conclusion

      Early results suggest nurse led telephone clinics are an effective way of providing follow-up to patients on an imaging based follow-up programme after surgery. They are well received by patients. We aim to introduce an online tool to objectively assess symptoms in this patient group. Further evaluation of patient experience in this clinic would be beneficial, along with an evaluation of the impact of introduction of telephone follow-up on the rest of the service.

    • +

      MA17.06 - The Specialist Lung Cancer Nurse and Self-Management for People Living with Lung Cancer: A Model of Engagement

      14:05 - 14:10  |  Presenting Author(s): Vanessa Nicole Brunelli  |  Author(s): Patsy Yates, Carol Windsor

      • Abstract

      Background

      There is increasing evidence that patient self-management strategies, used as adjuncts to traditional pharmacological interventions, can improve symptom control for people living with lung cancer. While it is acknowledged that the specialist lung cancer nurse (SLCN) is well positioned in the multidisciplinary team to facilitate patient self-management education (PSME), limited guidance is available to SLCNs on this role.

      The aim of this study was to understand the knowledge and skills required of SLCNs to facilitate PSME and how such skills might best be developed. The intent was to develop a pedagogy that enhances SLCN-patient interactions so that patients can be better supported to make self-care decisions and to act on these decisions.

      Method

      The epistemological lens of the study drew on the sociocultural works of Vygotsky and Leontiev. Fifteen participants were purposively recruited through the Australia and New Zealand Lung Cancer Nurses Forum. The sample comprised Australian registered nurses employed at the level of clinical nurse consultant. Eligibility for the study required participants have a minimum of 5 years lung cancer nursing experience. Of the sample, the average was 13 years’ experience.

      Through individual, face to face interviews, a biographical approach to data collection focused on participants’ work and non-work lives. The theoretically informed analysis generated understanding about the salient influences on SLCN learning and how and why these influences shaped PSME.

      Result

      PSME is an inherently complex activity. Although seeking to facilitate patient learning to empower patients to self-manage, the SLCN experiences challenges in the contemporary health care environment. Entrenched power relations, professional boundaries, minimal practice guidelines and issues of resourcing of lung cancer and lung cancer nursing are key factors that shape PSME.

      A model of engagement was designed to reflect the pedagogy that underpins optimal interactions between SLCNs and patients. The model brings forth the socially situated contexts of the SLCN and patient as central to the interaction. A reflective mode of practice creates a teaching and learning environment inclusive of sociocultural and individual processes on learning and thus the mechanisms of co-constructing knowledge for the purpose of shaping patient behaviour.

      Conclusion

      The study assumes the strategic importance of addressing how best the SLCN workforce can support people living with lung cancer to self-manage. A key strength of the research is the focus on understanding the individual and social interrelatedness of SLCN learning for the purpose of enhancing educational interactions in complex environments with people living with lung cancer.

    • +

      MA17.07 - Nurse-Led Telehealth Clinic in Treatment Monitoring and Follow Up

      14:10 - 14:15  |  Presenting Author(s): Sarah Louise Cubbin  |  Author(s): Michael Brada

      • Abstract

      Background

      A clinical consultation is usually physician led and traditionally carried out in a direct person-person interaction in an outpatient clinic. The alternative is replacing it through electronic means via telehealth. This has already been exploited as a phone consultation and is gaining momentum as a video based consultation although it has not been widely introduced into oncology.

      Method

      The aim was to ascertain the acceptance of the video consulting technology in real clinical settings and the effectiveness of video consultation in replacing conventional consultation.

      Eligible patients who attended a pre-existing nurse led clinic completed a checklist to ensure access to appropriate IT equipment. Those able to participate were asked to replace two nurse led appointments with nurse led teleheath appointments using iKonsult video consultation platform. After each consultation a satisfaction questionnaire was completed.

      Result

      Patients were recruited from a protocolised nurse led clinic for mutation driven NSCLC on oral tyrosine kinase inhibitors (TKI). 42 patients were approached over a three month period; only 6 agreed and were followed up via the telehealth platform. 4 patients considered it as a possibility, 4 did not feel confident and 3 did not have the correct equipment. The remaining patients cited numerous reasons for not taking up this service.

      In the satisfaction analysis of 17 initial telemedicine consultations 5/6 patients (81%) were very satisfied with telemedicine follow up. 4 patients (66%) found the platform extremely easy and 2 (34%) easy to use.

      Conclusion

      On treatment monitoring of oral TKI therapy could be effectively carried out using video consultation platform reducing the number of hospital visits. The consultations provided necessary information and allowed for adequate clinical assessment. However the initial take up rate is low mostly due to patient reluctance rather than unavailable technology. The overall feedback from participants was very positive and accepting of the service. The iKonsult video consultation is being introduced into other oncology settings.

    • +

      MA17.08 - Discussant - MA 17.05, MA 17.06, MA 17.07

      14:15 - 14:30  |  Presenting Author(s): Melissa Culligan

      • Abstract

      Abstract not provided

    • +

      MA17.09 - Remote Symptom Reporting for Tele-Nursing Team in Thoracic Oncology Clinics: Environmental Scan and Stakeholder Engagement

      14:30 - 14:35  |  Presenting Author(s): Simranjit Kooner  |  Author(s): Jennifer Harris, Suheon Lee, Catherine De Guia, Doris Howell, Geoffrey Liu

      • Abstract

      Background

      50+% of cancer-related toxicities are under-reported. A real-time Remote (i.e., at-home) Symptom Reporting (RSR) system could help patients seek help when symptoms exceed thresholds, mitigating unplanned clinic/emergency room visits. A RSR system for solid-tumor patients undergoing chemotherapy is associated with improved health-related quality of life and survival (Basch et al, 2017). Adapting RSR into the thoracic cancer clinic environment requires assessments of potential implementation barriers, and tailoring of the RSR-system.

      Method

      Over a five month period, we performed an environmental scan to determine readiness of RSR implementation in our comprehensive thoracic oncology outpatient clinic. A qualitative assessment of potential RSR integration into the telephone triage environment was performed through one-on-one interviews and focus groups, followed by thematic analysis. Discussions were held with multiple stakeholders; key implementation champions were identified. We utilized the Canadian Institutes of Health Research Knowledge-to-Action Framework, Steps 2-4 as our guide.

      Result

      In the environmental scan, 125 telephone triage calls were logged over randomly-chosen days in a 6-week period. The mean ± SEM call duration was 5.4 ± 0.62 minutes. Mean time until response was 44.4 ± 3.8 minutes. Nurses spent on average 2.7 ± 0.2 minutes documenting into the electronic-patient-record. The mean duration from initial contact to completion was 24.1 ± 4.5 minutes. Resolution of the triage calls involved telephone advice alone (87%; n=109), unplanned clinic visits (6%; n=8), and emergency visits (6%; n=7).

      In the qualitative analyses, top stakeholder-identified issues were: lack of assessment standardization; wasted time transcribing paper triage notes to electronic records; and a high patient/family burden in terms on understanding when to seek help. There was universal interest in adopting a RSR system from administrative assistants, nursing administration, clinic nurses, physicians and trainees. Perceived benefits of RSR were: standardized, focused telephone assessments; tailored symptom assessments in the thoracic setting (i.e., dyspnea, coughing, hemoptysis); patient empowerment; and improved efficiency in patient contact, intra-team communication, and documentation. Key stakeholder RSR features were: a phone/web application that assesses symptom severity and indicates when to contact the triage team; one-touch feature to reach team; longitudinal symptom trend display for tele-nursing team; and embedding of the COSTaRS framework to facilitate tele-nursing interventions and documentation.

      Conclusion

      Integration of a RSR system integration was perceived favorably by stakeholders to increase nursing efficiency and improve health related patient outcomes, but success hinges on an identified set of key requirements.

    • +

      MA17.10 - The Use of Technology in the Delivery of Supportive Care of Lung Cancer Patients After Treatment

      14:35 - 14:40  |  Presenting Author(s): Pamela Rose  |  Author(s): Heather Quail, John McPhelim, Mhairi Simpson

      • Abstract

      Background

      The NHS Lanarkshire Lung Cancer Project is part of the Transforming Care after Treatment (TCAT) programme.

      During phase one 58 patients participated in the project. 88% of patients opted for a telephone consultation, which was more time effective taking only on average 20 minutes compared to 48 minutes for a face to face consultation. 90% of patients rated the service as excellent and a review of additional Patient Reported Outcome Measures demonstarted an improvement in overall quality of life.

      Further funding was secured as part of Phase 3 of the TCAT programme allowing for continued testing.

      Method

      Between November 2017 – May 2018 lung cancer patients living in Lanarkshire were offered two monthly SPARC assessments on completion of treatment . This was followed up with and a telephone consultation from a lung cancer clinical nurse specialist, the provision of a personalised care plan and access to self-management information. The choice of an electronic or paper SPARC assessment was offered.

      To support the evaluation a Functional Assessment of Cancer Therapy – Lung (FACT-L), Memorial Symptoms Assessment Scale (MSAS) and Supportive Care Needs Survey Long-Form 59 (SCNS) were completed prior to their first and after their final assessment. A patient experience questionnaire was also provided on completion of their final assessment.

      Result

      24 patients participated in phase 3. 53% opted to complete their assessment electronically with 47% preferring the paper option.

      15 patients (63%) completed both assessments resulting in a total of 582 concerns being identified. Data analysis of these patients shows a 27% reduction in concerns with the number of high concerns falling by 62% between the first and second assessment. The average length of time for telephone review remained similar phase one at 22 minutes ranging from 7 minutes to 55 minutes.

      Patient satisfaction in the project continued to be high with 82% rating the service as excellent and 18% as good.. Data analysis for 15 patients’ who had completed 2 FACT-L, MSAS and SCNS yielded a significant reduction in symptom burden and psychological distress with a significant improvement in quality of life.

      Conclusion

      Findings from this project are encouraging that this model of working is not only acceptable to patients but time efficient and clinically effective. However, a limitation of this project is its small sample size. Therefore, further work is needed to explore its transferability and cost effectiveness to allow it to be considered for implementation in standard practice.

    • +

      MA17.11 - Multi-Centred, Prospective, Audit to Identify Readmission Causes and Complications Within 30 of Primary Lung Cancer Surgery

      14:40 - 14:45  |  Presenting Author(s): Amy Kerr  |  Author(s): Maureen King, Sandra Dixon, Sarah Taylor, Alison Smith, Charlotte Merriman, Jenny Mitchell, Verity Hunter

      • Abstract

      Background

      Surgery remains the first choice of curative treatment, for patients with non-small lung cancer, the proportion of patients undergoing surgery has risen in recent years. Post-operative complications are well recognised following curative lung cancer surgery but there is limited data on readmission rates and causes . The UK Thoracic Surgery Group (TSG), a subgroup of the National Lung Cancer Forum (NLCNF) conducted a multicentre audit to assess readmission potential causes and patient experience.

      Method

      The audit involved 6 UK thoracic surgical centres with prospective data collection over 3 months from primary lung cancer resection patients. Patients were contacted 1 month post discharge by telephone. Data collection included demographics, socioeconomic, smoking status, comorbidities, surgery, postoperative recovery, discharge satisfaction and readmission details.

      Result

      268 patients underwent thoracic surgery, the overall readmission rate was 11% (30), with variable readmission rate across the centres (range 3-24%), most readmission occurred within 7 days of discharge 47% (14) with patients being readmitted to a hospital that did not performed the procedure 43%(17). The most common cause of readmission was mainly pulmonary related with chest infections being largest cause, pain, wound infection and pneumothorax were also common. Length of stay following readmission was longer than initial surgical stay median 8 (range 0-94) vs 5 (range 2-27).Type of surgical approach had no impact on readmission. However readmission was associated with smoking, post-operative complications, discharge with drain, length of stay post-surgery and the patient’s readiness for discharge (see table 1).

      table 1.png

      Conclusion

      This audit provides a broad overview of the pattern and trend of readmissions rates within 30 days post discharge following lung cancer resection. Whilst not every readmission can be avoided, there is opportunity to identify and prevent patient readmission. Listening to patient’s assessment of their readiness for discharge is crucial to facilitating patient compliance with discharge and confidence in community carers.

    • +

      MA17.12 - Discussant - MA 17.09, MA 17.10, MA 17.11

      14:45 - 15:00  |  Presenting Author(s): Mary Duffy

      • Abstract

      Abstract not provided

  • +

    MA18 - Modelling, Decision-Making and Population-Based Outcomes

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

      MA18.01 - Non-Small Cell Lung Cancer Risk Assessment with Artificial Neural Networks

      13:30 - 13:35  |  Presenting Author(s): Tafadzwa Lawrence Chaunzwa  |  Author(s): Yiwen Xu, David C Christiani, Andrea Shafer, Nancy Diao, Michael Lanuti, Raymond Mak, Hugo Aerts

      • Abstract

      Background

      Lung cancer is a heterogeneous disease with many clinically important subtypes. Given the complexity of classification, there is room for innovative risk assessment tools to help ascertain prognosis and management. In this work we tested an Artificial Neural Network (ANN) to stratify patients into clinically significant low and high risk categories.

      Method

      CT imaging, survival, and cancer staging data was extracted for a sample of 311 patients with Stage-I (n = 186) and Stage-II (n = 125) non-small cell lung cancer (NSCLC) from the comprehensive Boston Lung Cancer Survival (BLCS) cohort. Median follow-up from time of diagnosis was 3.5 years, with 86% 2-year survival. A deep convolutional neural network pretrained on ImageNet was used, with fine-tuning of the last convolutional layers, dense layers, and softmax for stratification. Inputs of this model were 50 x 50 mm2 image patches. Training was performed on 182 labeled CT scans (112 Stage-I and 70 Stage-II). 46 cases were used for initial cross-validation, with an independent test set of 83 cases. The median prediction probability from the ANN was used as a cutoff to divide patients into low and high risk groups.

      Result

      The model was able to perform classification of cancer stage on the heterogeneous test set (AUC = 0.73, p< 0.0005). The test set was split evenly into low risk (n = 42) and high risk (n= 41) groups based on model predictions. There was statistically significant separation in the Kaplan Meier-estimates for survivorship in the two stratified groups (p < 0.02).

      ialsc_figure.png

      Conclusion

      ANNs can be effective tools for quantitative risk stratification in NSCLC. In addition to the potential for real-time clinical decision support, ANNs may also help create new paradigms in lung cancer risk assessment. The models have the capacity to perform suprahuman computations, which can help meet future demands of clinical practice, given expanding digital-imaging volumes.

    • +

      MA18.03 - How in the Real World Are Lung Cancer Patients Treated? The Ontario, Canada Experience

      13:35 - 13:40  |  Presenting Author(s): William Kenneth Evans  |  Author(s): William Flanagan, Cindy Gauvreau, Phongsack Manivong, Saima Memon, Natalie Fitzgerald, John R Goffin, Rochelle Garner, Edwin Khoo, Nicole Mittmann

      • Abstract

      Background

      Clinical trials define treatment recommendations but how patients are actually treated in the real world is poorly understood. The Canadian Partnership Against Cancer has developed a model of lung cancer (LC) management (OncoSim-lung) based on clinical trials data and expert advice. To credibly project the future clinical and economic impacts of cancer control measures using OncoSim, the model has been refined using real-world data.

      Method

      Treatment data by histology and stage were extracted from the Ontario Cancer Registry for LC cohorts diagnosed in 2010 and 2013. All incident cases that satisfied the IARC rule of a new primary were included. Missing or unknown stage cases were excluded. Clinical pathways were validated by oncologists from different disciplines across Canada.

      Result

      The 2013 cohort included 8,086 staged LC: NSCLC (n=7,143) Stage I 18.7%, II 8%, III/IIIa 11.4%, IIIb 4.9% IV 56.8%; SCLC (n= 943) limited 67.7%, extensive 32.3%. Of 1340 stage I NSCLC patients, 61% underwent surgery; 39% had no surgery and one third of these had no active treatment (NAT). 55% of those not receiving surgery underwent radical radiotherapy and 6% had palliative radiotherapy. Of 579 patients with stage II NSCLC, 60% underwent surgery and 47% of these received adjuvant chemotherapy; 40% had no surgery and 22% of these had NAT. Radical radiotherapy, radiotherapy plus chemotherapy or palliative radiotherapy were given in 33%, 19% and 18% of non-surgical cases, respectively. Of 813 stage III/IIIa patients, only 26% underwent surgery, 41% of whom received adjuvant chemotherapy or postoperative radical radiotherapy (16%); 13% received trimodality treatment. Of the 75% of Stage III not receiving surgery, 26% had NAT and 21% had palliative radiotherapy alone. Of those receiving active treatment, 20% received combined chemo +radiotherapy and 13% each had chemotherapy alone or radical radiotherapy alone. Of 356 stage IIIb patients, 17% had NAT, 28% received palliative radiotherapy and only 30% had chemo + radical radiotherapy. 18% had chemo alone. Of 4055 stage IV NSCLC, 47% had NAT, 24% received chemotherapy alone and 23% had palliative radiotherapy only. Of those who received first-line chemotherapy (n= 1059), 47% received second line chemotherapy and of those, 37% received third line therapy.

      Conclusion

      Compared to prior expert opinion, there was a higher use of radiotherapy for early stage disease, a lower frequency of chemo-radiotherapy in Stage III disease and a higher frequency of NAT across all stages of disease.

    • +

      MA18.04 - Discussant - MA 18.01, MA 18.03

      13:40 - 13:55  |  Presenting Author(s): Shalini K Vinod

      • Abstract

      Abstract not provided

    • +

      MA18.05 - Characteristics and Long-Term OS of Non-Small Cell Lung Cancer Patients Receiving EGFR Tyrosine Kinase Inhibitor Treatment

      13:55 - 14:00  |  Presenting Author(s): Michael Bergqvist  |  Author(s): Helene Nordahl Christensen, Fredrik Wiklund, Stefan Bergström

      • Abstract

      Background

      Epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) are important therapeutic agents in treatment of EGFR mutation-positive non-small cell lung cancer (NSCLC) patients. However, long-term follow-up and knowledge of clinical factors and TKI treatment patterns, which may be associated with longer OS, remains unclear. Using nationwide registry data, the aim was to investigate survival, prognostic factors for OS, and first line TKI treatment pattern of stage IIIB/IV NSCLC patients in Sweden.

      Method

      In this cohort study, data on all patients diagnosed with stage IIIB-IV NSCLC during 2010—2015 from the nationwide Cancer Registry of Sweden were linked with data on dispensed EGFR-TKI drugs, comorbidity, and mortality data from Swedish national health registries. OS was defined as the interval from date of diagnosis until date of death. Survival rates were estimated using the Kaplan-Meier method. Assessment of predictive factors for OS was performed in multivariable Cox regression.

      Result

      Of 9,992 stage IIIB/IV NSCLC patients (mean age 70 years, female 49%), 1419 (14%) received first-line TKI treatment. Overall, 59% of TKI treated patients (median age 68 years) were female, 44% had at least one comorbidity, 85% had adenocarcinoma, and 89% were stage IV. Median follow-up time was 15 months and median OS was 16 months; 1- and 3-years survival rates were 62% and 15%, respectively. Predictors of longer OS were younger age at diagnosis, adenocarcinoma, less advanced clinical stage, and less comorbid disease. Furthermore, patients included in the end of the period had a longer OS compared to earlier. TKI treatment switching/re-challenging, as well as prolonged TKI treatment, also predicted longer OS.

      Conclusion

      This is the first nationwide study on NSCLC patients receiving first-line EGFR TKIs in routine clinical practice in Sweden. In addition to the reported prolonged TKI treatment length and TKI switching/re-challenging during the observation period, improvements and extension of EGFR testing targeting the appropriate NSCLC patient population may further have contributed to the observed relatively long overall survival.

    • +

      MA18.06 - Patterns of Lung Cancer Care in the United States: Developments and Disparities

      14:00 - 14:05  |  Presenting Author(s): Erik Ferdinand Blom  |  Author(s): Kevin ten Haaf, Douglas Arenberg, Harry J De Koning

      • Abstract

      Background

      The level of adherence to lung cancer treatment guidelines is unclear. The aims of this current study were to provide an overview of current patterns of lung cancer care in the United States and to identify possible disparities in receiving standard of care.

      Method

      Using the National Cancer Database, we evaluated the first course therapy of 468,422 lung cancer cases diagnosed between 2010-2014. We used a series of multivariate logistic regression models to identify relationships between patient, tumor, and health care provider characteristics and receiving predefined stage-specific standards of care.

      Result

      Most common treatments were surgery only (15.2%), radiotherapy only (12.8%), chemotherapy only (13.5%), and radiotherapy and chemotherapy (26.2%). 22.1% of subjects received no treatment. Between 2010-2014, the use of Video-Assisted Thoracoscopic Surgery among surgically treated cases increased from 24.6% to 42.3%, while the rate of conversions to open surgery decreased from 18.3% to 10.4%. Among stage IA non-small cell lung cancer patients treated with thoracic radiotherapy, the use of Stereotactic Body Radiotherapy increased from 53.4% to 73.0%. Overall, only 63.3% of subjects received standard of care. Receiving surgery for early-stage non-small cell lung cancer was less likely with increasing age (for those 80 and over: odds ratio [OR], 0.08; 95% confidence interval [95%CI], 0.07-0.09), for non-Hispanic Blacks (OR, 0.59; 95%CI, 0.57-0.62), and for squamous cell histology (OR, 0.46; 95%CI, 0.45-0.47). These disparities were also present in other stages.

      Conclusion

      Particularly elderly lung cancer patients, non-Hispanic Blacks, and those with squamous cell histology are less likely to receive standard of care. These disparities may have consequences for lung cancer screening, as the effectiveness depends on adequate treatment of lung cancer.

      • Abstract

      Background

      Continued smoking after a LC diagnosis is associated with poorer cancer outcomes including increased risk of treatment-related side-effects, reduced treatment efficacy and poorer prognosis. Smoking cessation is an integral part of LC survivorship by improving both cancer and non-cancer outcomes. To enhance survivorship education, clinicians should understand patient awareness of the harms of continued smoking.

      Method

      LC survivors from Princess Margaret Cancer Centre, Toronto (2014-2017) were surveyed with respect to self-awareness of the harms of continued smoking on cancer-related outcomes. Univariable and multivariable logistic regression models assessed factors associated with awareness and whether awareness was associated with cessation among current smokers at diagnosis.

      Result

      Of 553 patients, 181 were lifetime never-smokers. Among those smoking during the peri-diagnosis period (n=177), 65% quit after diagnosis. Among all, few patients were aware that smoking negatively impacts treatment-related outcomes [complications from cancer surgery (only 41% aware), radiation side-effects (30%), quality-of-life on chemotherapy (44%) and treatment efficacy (36%)]; half were aware that smoking negatively impacts cancer prognosis (51% aware) and risk of developing second primaries (50%). Compared to ex-smokers/never-smokers at diagnosis, current smokers at diagnosis were less aware of the impact of smoking on radiation side-effects (22% vs 31% aware, P=0.01), prognosis (44% vs 55%, P=0.02) and risk of second primaries (42% vs 55%, P=0.007). Among sociodemographic variables, only those speaking English at home were consistently found more likely unaware that smoking negatively impacts these outcomes (ORs=1.52-2.20, P<0.04). Patients with early stage disease were more likely unaware that smoking negative impacts radiation side-effects (OR=1.60, 95%CI[1.09-2.35], P=0.02); while patients on curative treatment (OR=1.53[1.08-2.17], P=0.02) and those exposed to second-hand smoke (SHS) were more likely unaware that smoking impacts quality-of-life on chemotherapy (OR=1.64[1.05-2.58], P=0.03). Exposure to SHS, treatment intent and stage were not associated with awareness of impact on prognosis or second primaries (P>0.11). Among smokers in the peri-diagnosis period, awareness of the impact of smoking on surgical complications (aOR=2.09 [0.96-4.54], P=0.06), quality-of-life while receiving chemotherapy (aOR=2.60[1.17-5.79], P=0.02) and on treatment efficacy (aOR =2.24[0.97-5.20], P=0.06) were each associated with subsequent quitting, adjusted for marital status, pack-years, self-rated health and SHS exposure.

      Conclusion

      Many LC patients are unaware of the harms of continued smoking on cancer outcomes, particularly those smoking at diagnosis. Awareness of some of these outcomes was associated with subsequent tobacco cessation. Patient education on the health benefits of smoking cessation may increase quit rates and improve outcomes for LC patients.

    • +

      MA18.08 - Discussant - MA 18.05, MA 18.06, MA 18.07

      14:10 - 14:25  |  Presenting Author(s): Vera Hirsh

      • Abstract

      Abstract not provided

    • +

      MA18.09 - Predictors of Health Utility Scores (HUS) in Advanced EGFR-Mutated NSCLC.

      14:25 - 14:30  |  Presenting Author(s): Shirley Xue Jiang  |  Author(s): Manjusha Hurry, Katrina Hueniken, M Catherine Brown, Mindy Liang, Devalben Patel, Catherine Labbe, Lawson Eng, Hiten Naik, Penelope Bradbury, Natasha B Leighl, Frances A Shepherd, Wei Xu, Geoffrey Liu, Ryan N. Walton, Grainne M O'Kane

      • Abstract

      Background

      Advanced NSCLC patients with EGFR mutations (EGFRm) are currently treated with first - to third-generation tyrosine kinase inhibitors (TKIs). In the advanced setting, quality of life is an important goal; we therefore evaluated determinants of HUS in this population.

      Method

      In a prospective, observational study, patients with advanced EGFRm NSCLC completed EQ-5D surveys at outpatient visits generating HUS (range 0-1). Patients were allowed to enrol at any point in their disease course. Baseline clinical characteristics and outcome data were extracted from chart review. Patient imaging was reviewed and health states (stable/progressing) at each encounter recorded. Univariable analyses conducted using ANOVA and multivariable regression analyses with generalized estimating equations identified factors associated with HUS.

      Result

      From November 2014 to July 2017, 782 encounters (follow-up visits) were collected for 244 patients. Median age at first encounter was 64 years (range:29-96); 54% were female and 54% Asian. Median time from diagnosis of stage IV NSCLC to first encounter was 23 months (range:0-67). The median number of HUS collected per patient was 2 (range:1-14). For patients with multiple visits the median time between completed questionnaires was 1.8 months (1-18). 105 patients (43%) presented with or developed brain metastases during the study period. In a univariable analysis, regardless of treatment line, mean HUS (mHUS) on osimertinib was 0.85 (standard deviation (SD):0.15) (n=33 patients; 114 encounters) compared to mHUS=0.80 (SD:0.17) on gefitinib (n=147, 351 encounters); mHUS=0.72 (SD:0.16) on chemotherapy (n=32, 76 encounters); and mHUS=0.79 (SD=0.15) on other TKIs (n=49, 133 encounters); p<0.001. In a multivariable analysis, disease progression (p=0.04) and ECOG performance status >0 (p<0.001) were associated with lower HUS. In contrast, treatment with osimertinib (when compared to a reference group of first-generation TKIs, gefitinib/erlotinib) was associated with improved HUS (p=0.01), while line of therapy and number of metastatic sites of disease were not associated with HUS. In addition, brain metastases had no significant impact on HUS (p=0.33).

      Conclusion

      Progressive disease and worse performance status associate with lower HUS in patients with EGFRm NSCLC. Patients treated with osimertinib had the highest HUS when compared with a reference group of first-generation EGFR TKIs regardless of line of therapy. These results may help in the choice of EGFR-TKI, especially in patients with a poor performance status.

    • +

      MA18.10 - Evolving Immunotherapy Practice Patterns in Advanced NSCLC: Analysis of an Online Treatment Decision Tool

      14:30 - 14:35  |  Presenting Author(s): David R. Gandara  |  Author(s): Timothy A Quill, Martin J. Edelman, Suresh S. Ramalingam, Heather A Wakelee, Howard West, Kevin L Obholz

      • Abstract

      Background

      Checkpoint immunotherapy (IO) is revolutionizing NSCLC therapy. We have previously published results of an online decision support tool designed to provide clinicians with education and expert guidance (Chow et al: JTO 2015). Here we report an analysis of a recently updated version of this online tool, capturing the impact of emerging IO options.

      Method

      From June 2016 to July 2017, the NSCLC decision tool was updated to incorporate new treatment options for 280 different case scenarios. Briefly, oncologists entered patient and disease characteristics and then their planned treatment into the tool. Afterwards recommendations from 5 lung cancer experts were provided for that specific patient scenario.

      Result

      This analysis includes 1481 individual cases entered by 863 practicing oncologists between June 2016 and April 2018 (USA 19%, Europe 33%, Rest of World 48%). During this time, treatment choices for EGFR and ALK cancers by oncologists closely resemble those of experts. After approval of 1st-line pembrolizumab for patients with high PD-L1 expression, oncologists recommended pembrolizumab less often than experts (67% vs 95%). In the 2nd-line setting following platinum chemotherapy, both tumor histology and PD-L1 expression level impacted treatment recommendations (see Table). For PD-L1 expression < 1%, recommendations between oncologists and experts differed substantially.

      Second-line setting after platinum chemotherapy
      Participants' Treatment Choice Experts' Treatment Choice
      2016 2017 2016 2017
      Nonsquamous
      PD-L1 (1%)

      54% IO

      34% CT

      (n = 35)

      79% IO

      15% CT

      (n = 47)

      100% IO

      85% IO

      15% CT

      PD-L1 (< 1%)

      28% IO

      65% CT

      (n = 104)

      49% IO

      41% CT

      (n = 63)

      40% IO

      55% CT

      75% IO

      25% CT

      Squamous
      PD-L1 (1%)

      62% IO

      25% CT

      (n = 24)

      74% IO

      4% CT

      (n = 23)

      100% IO

      100% IO

      PD-L1 (< 1%)

      28% IO

      65% CT

      (n = 74)

      38% IO

      40% CT

      (n = 45)

      85% IO

      15% CT

      80% IO

      20% CT

      Conclusion

      This updated analysis of an online NSCLC decision-making tool integrates recent changes to the treatment landscape in 2017, capturing emerging patterns in IO therapy. Compared to earlier versions, practicing oncologist’s choice of 1st-line EGFR- and ALK- targeted therapy more closely tracked with experts during this period, while selection of IO differs from expert recommendations. A detailed analysis of expert versus online user data will be presented.

    • +

      MA18.11 - Implementing a Comprehensive National Audit of Lung Cancer Surgery: The English Lung Cancer Clinical Outcomes Publication (LCCOP) Project

      14:35 - 14:40  |  Presenting Author(s): Doug West  |  Author(s): Richard Page, Neal Navani, Susan V Harden, Aamir Khakwani, Richard Hubbard, Paul Beckett

      • Abstract

      Background

      We report the establishment of a national audit of outcomes after lung cancer resection (LCCOP) in the English National Health Service (NHS), a government healthcare system providing the great majority of lung cancer surgery. LCCOP is a compulsory audit commissioned by NHS England.

      Method

      Unusually, for a surgical audit, data is initially obtained from the cancer registry, and matched to national Hospital Episode Data (HES), before local validation by clinical teams. After case mix adjustment, unit level survival rates at 30, 60 and 90 days, and length-of-stay data are published online and in an annual report. The first annual report was released in 2014.

      Survival is adjusted for age, sex, performance status, stage, laterality, FEV1 percentage, comorbidity and socioeconomic status

      Result

      The number of resections rose by 21% between 2015-2017 (4892 to 5936). Median annual activity per surgeon rose from 30 to 49 cases between 2014-2017, a 63% increase. In 2015 survival at 30, 90 and 365 days was 98.1%, 96.3% and 87.9% respectively. Median length of stay was 6 days (IQR 4-9).

      In 2015, 43.9% of lobectomies were completed by VATS, 4.3% were started VATS and completed by open surgery and 0.7% completed by robotics.

      Adjusted 90 day survival by surgical unit: 2017 report (2015 data)

      90 day 2017(15).png

      Conclusion

      Using routinely collected NHS activity data for surgical audit is feasible, and reduces the data collection burden for hospital teams. Clinical validation remains important to correct discrepancies. Surgical activity has risen significantly. Increases in individual surgeon case volume may reflect increasing subspecialisation. Significant inter-provider variation remains, particularly in length of stay.

      More lung cancer surgery is being done in the English NHS. Surgeons are increasingly subspecialising, with higher case volumes. Local variation remains, particularly around length of stay. A mixed model of routinely collected data with local validation appears acceptable to clinical units.

    • +

      MA18.12 - Discussant - MA 18.09, MA 18.10, MA 18.11

      14:40 - 14:55  |  Presenting Author(s): Martin R Stockler

      • Abstract

      Abstract not provided

    • +

      MA18.12a - Q&A

      14:55 - 15:00

      • Abstract

      Abstract not provided