Virtual Library

Abstract
There is a long and distinguished history of clinical trials investigating the treatment of lung cancer dating back to the 1950’s. One of the first such trials to be published looked at the use of nitrogen mustard and DON (6-diazo-5-oxo-L-norleucine) in patients with various cancers including bronchogenic carcinoma[1], and probably the first published randomized trial specifically for lung cancer patients compared surgery and supervoltage radiotherapy[2]. Continued progress in treatment relies on randomized clinical trials and meta-analyses, and a review was undertaken in 2003 to assess the quality and quantity of all published randomized lung cancer treatment trials[3]. This showed that, of nearly 1000 trials, only 4 accrued more than 1000 patients, and that clinicians (and patients) could only rely on a few meta-analyses to guide treatment decision making. The review called for greater global collaboration to design and run large trials to improve the survival rates and quality of life of lung cancer patients. Ten years on, an update and extension of the 2003 review has been undertaken, to see what changes have occurred, what patterns have emerged, and whether the data will help the design of future trials. The Cochrane Library was used as the e-library, as it contains some abstracts from major meetings, as Song et al[4] estimate that up to 50% of all trials are not fully published. In addition to sample size, information was collected regarding the histological subgroup studied, the treatment modality investigated, the trial design, the outcome and the country of affiliation of the first author. A search of ‘(lung OR bronchus) AND (cancer OR carcinoma)’ in Sept 2012 produced 7130 reports (5218 full papers and 1912 meeting abstracts) classified by the Cochrane Library as ‘trials’. However, as the inclusion of meeting abstracts was found to be inconsistent, the results presented here are based only on the full publications. A total of 1792 randomized clinical trials of lung cancer were identified, of which 1677 included the sample size in the title or abstract, 49 did not, and in the remaining 66 the abstract could be accessed. An analysis of the 1677 trials with sample size stated showed that: There was an increasing number of randomized trials, mainly due to an increasing number of randomized phase II trials[5] Only 20 trials included >1000 patients, and 48% had less than 100. The median sample size has remained unchanged for 40 years (~100 patients) although the median sample size for those designated as randomized phase III trials has increased from ~200 to ~400 patients over the last 10 years In nearly half of the trials the primary research question addressed chemotherapy, 15% supportive care, 8% radiotherapy and only 1% surgery Of those that indicated the histology, trials of NSCLC cancer accounted for 73% The country producing the most trials (397) was the USA, but China is now the second country, and published the most trials over the last 10 years. Although there are limitations of this review (it is necessarily several years out of date, is only looking at the title and abstract of trials on the Cochrane Library, is complicated by multiple reporting of trials, etc), its strengths are the longitudinal overview and comprehensive inclusion criteria which substantially extend the results reported by Subramanian et al[6] who compared ongoing trials of medical treatment for NSCLC in 2012 with those in 2009. The need to improve the quantity and quality of trials in lung cancer has been highlighted for more than 20 years[7-10 ]as trials have not been sufficiently large, or poorly designed, or unnecessarily duplicated previous work[11-12]. In contrast, well designed trials make best use of researchers’ time, funders’ money and patients’ goodwill[13], and may inform future work. The results of the current review suggest that greater global collaboration is still required to run large trials that will produce reliable results and influence practice worldwide. Updated and more detailed results will be presented. References 1. Krantz S, et al for the Veterans Administration Cancer Chemotherapy Study Group. A Clinical Study of the Comparative Effect of Nitrogen Mustard and DON in Patients With Bronchogenic Carcinoma, Hodgkin's Disease, Lymphosarcoma, and Melanoma. JNCI 1959, 22, 433-9 2. Morrison R. The treatment of carcinoma of the bronchus. A clinical trial to compare surgery and supervoltage radiotherapy. Lancet 1963, 1, 683-4 3. Stephens R. The need for a world strategy for clinical trials. Lung Ca 2003, 41 (suppl 3), S96 (abs E-77) 4. Song F, et al. Dissemination and publication of research findings: an updated review of related biases. Health technology Assessment 2010, 14, No. 8 5. Turrisi AT. Creeping phase II-ism and the Medical Pharmaceutical Complex: Weapons of Mass Distraction in the War against Lung Cancer. JCO 2005, 23, 4827-9 6. Subramanian J, et al. Review of Ongoing Clinical Trials in NSCLC - a status report from the CinicalTrials.gov Web Site. J Thorac Oncol 2013, 8, 860-5 7. Nicolucci A, et al. Quality, evolution, and clinical implications of randomized, controlled trials on the treatment of lung cancer. A lost opportunity for meta-analysis. JAMA 1989, 262, 2101-7 8. Brundage MD, Mackillop WJ. Locally advanced non-small cell lung cancer: Do we know the questions? A survey of randomized trials from 1966-1993. J Clin Epidemiol 1996, 49, 183-92 9. Breathnach OS, et al. Twenty-two Years of Phase III trials for Patients With Advanced Non-Small Cell Lung Cancer: Sobering Results. J Clin Oncol 2001, 19, 1734-42 10. Macbeth F, et al. An open letter to all members of the IASLC. Lung Cancer 2004, 45, 119-20 11. Chalmers I. The lethal consequences of failing to make full use of all relevant evidence about the effects of medical treatments: the importance of systematic reviews. Rothwell P, ed. Treating individuals: from randomised trials to personalised medicine. London: Lancet, 2007: 37-58. 12. Young C, Horton R. Putting clinical trials into context. Lancet 2005, 366, 107-8 13. Chalmers I, Glasziou P. Avoidable waste in the production and reporting of research evidence. Lancet 2009, 374, 86-9

Abstract not provided

Abstract not provided

Abstract not provided

Abstract
For a long time, radiotherapy for lung cancer were considered in terms of being either curative, palliative or radical. Conventional radiotherapy for early stage NSCLC was general considered ‘ curative’, but the more modest expectation in patients with locally advanced NSCLC led to it being referred to as ‘radical’, as disease recurrences was considered very likely. In metastatic NSCLC, the role of radiotherapy was considered only in a palliative context, since it was generally accepted that in disseminated disease, local therapies could never be curative. These classifications have been reappraised in the light of recent data. In small cell lung cancer (SCLC), which has a very high risk of distant spread, recent studies have established that local treatments can extent survival and contribute to cure. This is not only the case for thoracic radiotherapy, but also for prophylactic radiotherapy to the brain (PCI; prophylactic cranial irradiation). Studies in patients with limited disease SCLC have shown a survival benefit of about 5% at 3 years,a benefit of similar magnitude to that for thoracic radiotherapy in this patient group. Even in patients with proven disseminated SCLC (extensive stage), the use of PCI has improved survival. In a randomized trial, patients who received PCI had a 1 year survival of 27% compared to 13% for patients who did not receive PCI [1]. A study on the use of thoracic radiotherapy after the completion of chemotherapy in patients with ES-SCLC has recently been closed and the results are awaited next year. With the advancements in radiotherapy techniques, we can now ablate tumors with very high doses, and with great precision using stereotactic techniques. Local control rates in excess of 90% are obtained and in fit, potentially operable patients, 5-year survival rates above 60 % are achieved [2,3]. These ablative radiation doses may also exert a benefical immunological effect both locally and systemically. The same principles of high-dose high-precision radiotherapy are being applied in patients with a limited number of metastases (oligometastases). This patient group is increasingly being identified due to improved imaging techniques, and interest is growing due to the availability of more effectivity of systemic therapy in subgroups of lung cancer. Aggressive treatment of metastases by surgery or ablative therapy, such as SBRT, might offer a real chance of cure for these patients. We have initiated a clinical study to evaluate the benefit of this approach [4]. References 1. Slotman BJ et al., New Engl J Med 357, 664-72, 2007. 2. Onishi H et al., Int J Radiat Oncol Biol Phys 75, 243-247, 2011. 3. Lagerwaard FJ et al., Int J Radiat Oncol Biol Phys 83, 348-53, 2012 [updated]. 4. Palma D et al., BMC Cancer 12, 305, 2012.

Abstract
Decision making in advanced NSCLC undoubtedly should primarily consider evidence based treatment standards recommended by international guidelines (1, 2) According to international guidelines, decision making in 2013 is characterized by customizing therapy in advanced non-small cell lung cancer, by selecting a specific therapeutic regimen based on tumor histology and molecular biology. This refers to first-line therapy in patients with good performance status, but also to subsequent lines of therapy since antineoplastic drugs and regimens used in induction therapy directly influence the selection of agents/regimens considered for second/third-line treatment. The availability of antineoplastic monoclonal antibodies, small molecules and newer cytotoxic such as pemetrexed, the antiangiogenic bevacizumab as well as the tyrosine kinase inhibitors erlotinib, gefitinib and crizotinib has recently changed the treatment algorithm of advanced NSCLC. Therapy optimization by modern medical therapy is characterized by treatment individualization based on predictive factors. This process seems to continue since other products holding considerable promise for the near future such as tumor vaccines and other immunotherapeutic approaches, anti-angeniogenic agents, and newer EGFR-targeting monoclonal antibodies have already reached the level of phase III-testing or the registration process. Today’s recommendation can be summarized as follows: First-line therapy: In patients with good performance status with non-squamous tumors haboring an activating EGFR-mutation, an EGFR tyrosine kinase inhibitor may be the leading option, not only because of being active but also because of its feasibility and improved toxicity profile (3-5). In patients with non-squamous cell tumors not expressing EGFR-mutations, combination therapy remains standard with pemetrexed as the preferred partner of cisplatin (6). Furthermore, it has been demonstrated that in non-squamous NSCLC the addition of bevacizumab to standard doublet therapy improves survival (7, 8). With regard to patients with squamous cell tumors gemcitabine, vinorelbine or taxanes in combination with platinum-based agents remain the chemotherapeutic standards. Bevacizumab and pemetrexed are not recommended in squamous cell tumors either because of the agents toxicity profile (bevacizumab) or because of being less effective in this particular histological subtype (pemetrexed). For elderly patients not being fit for standard doublet chemotherapy medical treatment requires modification. Here single agent therapy is widely considered to be the preferred option for maintaining quality of life, reducing tumor association symptoms and improving survival (9). Fit elderly patients benefit from combination chemotherapy as much as younger patients and platinum-based chemotherapy is recommended as well (10, 11). Combination therapy remains investigational in patients with poor performance status (≥ PS 2), and single agent chemotherapy is the preferred option (12). Second-line therapy: In patients with disease progression during or after completion of induction (first-line) chemotherapy second-line therapy is indicated if the patient remains in good clinical condition. Selection of drugs for second-line therapy is based on whether the drug has been used earlier, the toxicity profile or the patients wish. The approved options for second-line therapy are docetaxel and pemetrexed (in case of non-squamous histology), erlotinib and gefitinib (in tumors expressing activating EGFR-mutations) (13-16). Crizotinib has recently been registered for second-line therapy in ALK-positive tumors (17). Maintenance therapy: The prolongation of induction therapy represents an evidence based new option to improve outcome in advanced NSCLC, in general and with strong implications for second line therapy, in particular. Three randomized studies investigated pemetrexed and erlotinib as maintenance therapy following 4 cycles of chemotherapy (18-21). In these trials switch-early second-line- (pemetrexed, erlotinib) or continuation-true- (pemetrexed) maintenance therapy has significantly increased PFS and OS. Based on this given evidence these two compounds have been registered for maintenance therapy in patients with advanced NSCLC, not progressing following 4 cycles of first-line standard platinum-based therapy. According to the European Medicines Agency (EMA) pemetrexed is indicated as monotherapy for maintenance treatment of locally advanced or metastatic NSCLC other than predominantly squamous cell histology in patients who’s disease has not progressed immediately following platinum-based chemotherapy, including platinum-pemetrexed combinations, and erlotinib is indicated as monotherapy for maintenance treatment in patients with locally advanced or metastatic NSCLC with stable disease after 4 cycles of standard platinum-based first-line therapy. When prescribing erlotinib, clinical factors associate with prolonged survival should be taken into account. There is clearly more which must be considered and which would lead to stop or even not to initiate a modern medical therapy. Even if national and international treatment recommendations undoubtedly represent the backbone of the decision making process, the management of patients with advanced non-small cell lung cancer often clinically requires modification for a number of reasons not necessarily considered by guidelines.This refers to first and subsequent-line treatment as well. The majority of lung cancer patients are older patients and often express specific age related treatment expectations. Other relevant factors for decision making in a non-curative setting are patient social environment and patient psychological status, regulations of national health care providers and reimbursement systems, drug availability, diagnostic and health care infrastructure. These circumstances modify decision making and lead to the use of older agents and regimens, to changes in dosing and scheduling of newer agents and regimens, to single agent therapy, to a reduction of treatment cycles and to a more free indication of best supportive care measures. Our increased understanding of the molecular biology of lung cancer and the change of its epidemiology has opened up venues for more rational treatment strategies and at the same time has challenged the additional diagnostic algorithms, the established health care financing and the complex process of drug development. Customized therapy and therapeutic targeting were made possible through the identification of new treatment predictors and the development of a number of antineoplastic agents which have shown clinical evidence for being more beneficial then the treatment standard in selected patients. Consequently, therapy individualization by histology and molecular markers has significantly influence the work of pathologists around the globe and the process of obtaining a therapeutically relevant tumor diagnosis. Not only histological sub-typing becomes clinical relevant but molecular information is also of increasing importance for treatment selection. Routine molecular testing in certified laboratories must be established, and this diagnostic process should ideally performed under the guidance of evidence based recommendation such as the recently published guidelines from the College of American Pathologists, the International Association for the Study of Lung Cancer, and the Association for Molecular Pathology for the molecular testing to select patients for EGFR-ALK-thyrosine kinase inhibitors (22). This process requires advanced diagnostic techniques and expertise. Investigating and implementing medical targeting in lung cancer because of its large dimension is costly and characterize by the limitation of financial and clinical resources, and currently not every where available. Since the majority of large randomized phase III trials during the last decade turned out negative in their primary endpoints, the time of conducting trials on large unselected patient population seems definitely over and the classical investigational strategy must probably be replaced by smaller trials in selected patients, and trials using study endpoints which can function as substitutes for the traditional overall survival endpoints (23). This inevitable change in the altitude of conducting clinical trials will increase the need for patients treated in clinical trials and, therefore, has a significant impact in decision making in advanced NSCLC. Another hurdle for making promising therapies difficult to be generally prescribed are that clinical evidence of being beneficial and the scientific rational do not necessarily equal clinical practicability and reimbursement in a given cultural and economic system. For Europe and specifically Germany it is important that an agent has reached a positive vote by the registration agency EMA and that it find acceptance by the national health care providers responsible for reimbursement. For this reasons the clinical evidence of being beneficial must be well documented and strong. Information about treatment selection by predictive factors and treatment restriction to patients who benefit the most is very welcome. In addition, if the treatment selection requires advanced diagnostic testing generally accepted and validated methods should easy to be reached and should provide reliable and reproducible results by at the same time being cost effective. References: 1. Azzoli et al, J Clin Oncol 29, 3825-3831, 2011 2. Peters et al, Ann Oncol 23 (suppl. 7), 56-64, 2012 3. Mok et al et al, N Engl J Med 361, 947-957, 2009 4. Zhou et al, Lancet Oncol 12, 735-742, 2011 5. Rosell et al, Lancet Oncol 13, 329-346, 2012 6. Scagliotti et al, J Clin Oncol 26, 3543-3551, 2008 7. Sandler et al, N Engl J Med 355, 2542-2550, 2006 8. Reck et al, J Clin Oncol 27, 1227-1235, 2009 9. Gridell et al, J Natl Cancer Inst 95, 362-372, 2003 10. Langer et al, J Clin Oncol 22 (suppl), 639 (abstr. 2571), 2003 11. Quoix et al, Lancet 378, 1079-1088, 2011 12. Gridelli et al, Ann Oncol 15, 419-426, 2004 13. Shepherd et al, J Clin Oncol 18, 2095-2103, 2000 14. Hanna et al, J Clin Oncol 22, 1589-1597, 2004 15. Thatcher et al, Lancet 366, 1527-1537, 2005 16. Douillard et al, J Clin Oncol 28, 744-752, 2010 17. Shaw et al, N Engl J Med 368, 2385-2394, 2013 18. Capuzzo et al, Lancet Oncol 11, 521-529, 2010 19. Ciuleanu et al, Lancet 374, 1432-1400, 2009 20. Paz-Ares et al, Lancet Oncol 13, 247-255, 2012 21. Paz-Ares et al, J Clin Oncol 31, 2895-2902, 2013 22. Lindeman et al, J Thoracic Oncol 8, 823-859, 2013 23. Pilz and Manegold, Memo 6, 92-97, 2013

Abstract
Advance care planning refers to a “process of discussion, reflection, understanding and communication between a patient, their family and health providers for the purpose of clarifying values, treatment preferences and goals of end of life care” (1). Advance care planning provides a formal means of ensuring health professionals and family members are aware of the patient’s wishes for care if they were to become too unwell to speak for themselves in the future. This process may include the patient completing an Advance Care Directive, which documents their wishes and/or the appointment of a substitute decision maker. Advance care planning is a patient-centred initiative that promotes shared-decision making and supports substitute decision-making, where appropriate, and aims to achieve good end of life care. The potential benefits of advance care planning for patients with advanced lung cancer have been highlighted in recent literature and clinical practice guidelines (2-4). However, there are many barriers to implementing advance care planning in this setting including patient, family, health professional and system related factors. In particular doctors and nurses caring for patients with lung cancer may be reluctant to raise the topic of advance care planning for fear of upsetting patients or they may lack confidence in knowing how to discuss it. This talk will focus on evidence-based strategies and practical tips for health professionals when having advance care planning discussions with patients with advanced lung cancer and their families. Recommendations from Australian “Clinical practice guidelines for communicating prognosis and end-of-life issues with adults in the advanced stages of a life-limiting illness, and their caregivers” will be described (5). The key recommendations are for health professionals to: Prepare for the discussion, where possible; Relate to the person; Elicit patient/caregiver understanding and information preferences; Provide information, tailored to the individual needs of both patients and their families; Acknowledge emotions and concerns; (foster) Realistic hope; Encourage questions and further discussions; and Document what had been discussed. (PREPARED). Sample phrases and useful questions for facilitating advance care planning discussions will be presented. Further resources regarding advance care planning will be provided (6-9). References: (1) Royal Australasian College of General Practitioners Advance Care Planning Definition http://www.racgp.org.au/your-practice/business/tools/support/acp/ (2) Mack JW, Cronin A, Keating NL et al. Associations between end-of-life discussion characteristics and care received near death: a prospective cohort study. Journal of Clinical Oncology 2012; 35: 4387-4395 (3) (USA) National Comprehensive Cancer Network: Practice guidelines in oncology: Palliative care. http://www.nccn.org/professionals/physician_gls/f_guidelines.asp (4) Smith, T, Clayton, J, Michael, N. What is the role of advance care planning and timing of referral for patients with lung cancer? [Version URL: http://wiki.cancer.org.au/australiawiki/index.php?oldid=47787, cited 2013 Aug 26]. Available from http://wiki.cancer.org.au/australia/Clinical_question:What_is_the_role_of_advance_care_planning_and_timing_of_referral_for_patients_with_lung_cancer%3F. In: Cancer Council Australia Lung Cancer Guidelines Working Party. Clinical practice guidelines for the treatment of lung cancer. Sydney: Cancer Council Australia. Available from: http://wiki.cancer.org.au/australia/Guidelines:Lung_cancer (5) Clayton JM, Hancock KM, Butow PN, Tattersall MHN, Currow DC. Clinical practice guidelines for communicating prognosis and end-of-life issues with adults in the advanced stages of a life-limiting illness, and their caregivers. Medical Journal of Australia 2007; 186: S77- 108 https://www.mja.com.au/journal/2007/186/12/clinical-practice-guidelines-communicating-prognosis-and-end-life-issues-adults (6) The conversation project http://theconversationproject.org (7) Respecting Patient Choices http://www.respectingpatientchoices.org.au (8) Oncotalk http://www.oncotalk.info (9) Vital talk http://vitaltalk.blogspot.com.au

Abstract
Metastatic lung cancer remains an incurable condition despite recent advances in medical therapy. New treatments can prolong survival and enhance quality of life for patients. Lung cancers can shrink and impressive responses may divert attention from the inevitable reality that the cancer will eventually acquire resistance to therapy. In the current era, patients are guided through multiple lines of therapy; first line, maintenance, second-line and so on. Such ‘lines of therapy’ extend the time on active therapeutic measures designed to alter the biological behavior of advanced lung cancer. Cancer centres that conduct clinical trials may offer potential active intervention beyond the proven ‘lines of therapy’, thereby extending time on active therapy. This may delay the transition to end-of-life care, but the likelihood of benefit with therapies beyond 2[nd] line is low. Whilst oncologists have a greater armamentarium in the fight against lung cancer, there comes a time for every patient when the medical advice will be to stop ‘active therapy’ as there is no prospect of benefit. The focus of care will be on symptom control with acceptance of palliative care. This is the time when patients accept that medical intervention will not prolong survival time, cannot alter the growth and spread of the cancer and will not influence the harmful effect of the cancer on the body. The body will fail, and in the process a functional decline will become obvious. This transition to ‘end-of-life’ care is not a clearly defined time point. Complex decision-making and emotional discussions are usually required. Several issues should be taken into account in determining when anti-cancer therapies should continue to be utilised or when such measures should be abandoned. These issues include: The likelihood of a beneficial response to a proposed therapy The safety and toxicity of the proposed intervention The patient’s ‘performance status’ – an overall measure of functional capacity Comorbidity and organ dysfunction Patient preference There are validated measures of prognosis that may assist in determining when this transition to palliative care should commence, including extent of cancer, weight loss, serum albumin, white cell count, neutrophil to lymphocyte ratio, lactate dehydrogenase and measures of critical organ function (cerebral, pulmonary, liver and renal). The synthesis of all of this information, a global assessment of the patient and knowledge of all the available treatment options is required to determine the timing of transition to ‘end-of-life’ care. The transition is usually handled through a multidisciplinary approach, with potential involvement of multiple health care professionals including the general practitioner, palliative care physician, medical oncologist, radiation oncologist and surgeon. Additional input from health care professionals in the fields of psychology, psychiatry, dietetics, social work and physiotherapy can enhance the quality of care in this transition period, depending on the degrees of distress and disability. Expert counseling is required. Involvement of the primary carers and family in discussions and decision-making is also important. Optimising care for patients during the late stages of cancer should focus on alleviating symptoms, allow planning for death, facilitate arrangements and enable personal commitments before death, dealing with grief and supporting the caregivers. These measures can improve quality of life, reduce anxiety and permit a more peaceful death. Ongoing active cancer therapies may introduce adverse events, require additional hospital visits and expose patients to invasive procedures. This can delay the introduction and implementation of optimal palliative care in the ‘end-of life’ period. The transition to end of life care should also be an opportunity to consider advanced directives. This issue needs to be discussed in a sensitive manner. Decisions around intensive medical intervention and confirmation of ‘not-for-resuscitation’ status will allow for patient wishes to be respected and permit patient autonomy at this difficult stage. An appreciation and acceptance of ‘end’-of-life’ care, when the prognosis is clearly very poor and death from an advancing lung cancer inevitable, will improve the prospect of allowing a comfortable death, a ‘good death’. This should be a goal that every cancer specialist seeks for patients with incurable cancer.

Abstract not provided

Abstract
Chemo-radiotherapy (CT-RT) is the presently the standard of care in patients presenting with a stage III-N2 NSCLC, with surgery being considered an option in selected patient subgroups [ESMO Guidelines 2013]. Improved local disease control in stage III disease translated into improved overall survival [Auperin A, 2010; Machtay M, 2012]. The best median survival reported to date in a randomized phase III trial of concurrent CT-RT is 28.7 months in RTOG 0617, when concurrent carboplatin-paclitaxel was combined with delivery of 60 Gy in once-daily fractions of 2 Gy [Bradley J, 2013]. At this dose, 25% of patients developed a local failure at 18 months follow-up. Therefore, both the optimization of radiation delivery and dose escalation are areas of active research. Evidence supporting radiation dose escalation comes from retrospective analyses and pooled studies. The randomized RTOG 0617 study compared 60 Gy (in 6 weeks) versus 74 Gy (in 7.5 weeks), in a 2x2 design where patients were also randomized to receive additional cetuximab or none [Bradley J, ASCO 2013]. Analysis of the impact of cetuximab has yet to be reported, but a difference in favor of the 60 Gy arm was seen for progression-free survival (36.6% vs. 26.3% at 18 months), median overall survival (28.7 vs. 19.5 months). A higher incidence of grade 3 esophagitis in the 74 Gy arm (21% vs. 7%, p=0.0003) was not unexpected, but the higher rates of local disease progression (34.3% vs. 25.1%, HR = 1.37, p = 0.0319) was a surprise. The latter raises the possibility of unsafe planning practices being applied when higher (and more toxic) doses have to be delivered. On multivariate analysis, factors predictive of a poorer overall survival were higher radiation doses, higher esophagitis/dysphagia grade, greater gross tumor volume, and heart volumes receiving >5 Gy. While publication of full data from the RTOG 0617 study is awaited, some issues related to toxicity and techniques should considered (below). 1. Improved radiotherapy planning using intensity modulated radiotherapy (IMRT) has permitted some parameters for lung toxicity (V20, mean lung dose) to be improved. However, not all recent or ongoing studies have applied constraints for lung volumes undergoing low-dose irradiation (V5). Furthermore, the choice of chemotherapy may increase lung toxicity, as reported by a recent meta-analysis showing that the risk of radiation pneumonitis is highest after concurrent CT-RT in patients aged >65 years, who also received carboplatin-paclitaxel [Palma D, 2012]. This meta-analysis revealed that predictors of fatal pneumonitis were use of daily dose fractions exceeding 2 Gy, the V~20~ parameter, and lower-lobe tumor locations. 2. Cardiac doses exceeding 5 Gy correlated with poorer survival in RTOG 0617 trial, and future studies must pay more attention to cardiac toxicity. Following concurrent CT-RT, cardiac doses of 45 Gy or higher correlated with myocardial hypoperfusion [Gayed 2006], and a higher risk of ischemic heart disease and cardiac dysfunction was seen patients undergoing CT-RT for left-sided lung cancer [Hardy D, 2010]. 3. After radiation doses of 74 Gy and 86 Gy delivered after chemotherapy, the incidence of bronchial stenosis was 4 and 25%, respectively, (Millar KL, 2005]. This complication may increase in incidence if higher doses of radiation are administered concurrent with chemotherapy. 4. When evaluating the results dose escalation studies, one must also consider the impact of tumor volume on survival in stage III NSCLC. For example, planning target volumes (PTV) correlated significantly with overall survival, with PTV values <350cc, 350-700cc, >700-1050cc and >1050cc having corresponding median overall survivals of 35.6 months (95% CI=0-71.3), 24.2 months (95% CI=18.3-30.2), 15.7 months (95% CI=10.5-20.9) and 10.3 months (95% CI=6.0-14.7), respectively [van Reij E, 2013]. Survival for tumors measuring 350-700cc differed significantly from the groups 700-1050 and >1050 cc (p=0.039 and p=0.002, respectively). In addition, larger tumor volumes are also associated with an increased risk of dying in the first 18 months, independently of T and N stage, but not beyond that time point [Ball D, 2012]. The latter finding suggests that a better characterization of tumor characteristics that correlate with radiocurability should be a research priority. 5. In RTOG 0617, the dose of 74 Gy arm was delivered in 7.5 weeks. A recent meta-analysis suggested that the accelerated delivery of radiation, i.e. delivering a higher higher dose in a shorter overall time, led to superior tumor control [Mauguen 2013]. Such accelerated delivery can be delivered using multiple fractions per day, or larger doses/fraction, both approaches which can increase the toxicity of concurrent CT-RT. For some subgroups of stage III-N2 disease, e.g. large volume tumors, where the benefits of concurrent CT-RT are modest [Wiersma T, 2013], sequential delivery may be a less toxic manner for investigating dose escalation. 6. Tumor volumes not infrequently regress during a course of CT-RT, and may also shift in position during the course of treatment [Lim G, 2011]. Volumetric imaging on modern treatment machines can allow for such changes to be detected, and treatment plans be adapted when required. Many recent clinical studies have not studied this aspect of treatment delivery. With dose escalation in fractions of 2 Gy falling out of favor, some groups are exploring the ‘dose painting’ hypothesis [Bentzen S, 2011]. Briefly, the former assumes that local recurrences arise from relatively radiation-resistant sub-regions; that reliable spatio-temporal mapping of such sub-regions is possible (e.g. using FDG-PET); that boosting radiation delivery to these egions improves local tumor control with acceptable side effects. Others groups are exploring the use of protons to improve local control in stage III NSCLC. A critical evaluation of the results emerging from ongoing trials is required, even if early outcomes from pioneering centers appear promising. Quality assurance data from radiotherapy trials, in which a broad range of institutions participate, reveals that a failure to comply with protocol requirements is associated lead to decreased survival, poorer local control and potentially increased toxicity [Weber DC, 2012].

Abstract
For the past decade, the standard first-line therapy for stage IV non-small-cell lung cancer (NSCLC) patients with adequate ECOG performance status (<2) has been platinum-based doublet chemotherapy with a reported median overall survival (OS) of 8–10 months, which has slightly improved to 12 months with the addition of bevacizumab. More recently, tyrosine kinase inhibitors of epidermal growth factor receptor, gefitinib and erlotinib, and anaplastic lymphoma kinase, crizotinib, have been shown to provide longer progression-free survival (PFS) and fewer side effects as first-line therapy, compared with chemotherapy in patients with certain histological subtypes and activating mutations. However, despite therapeutic developments, palliative treatment is standard for many stage IV NSCLC patients and the prognosis remains poor, with relative 5-year survival rates ≤4%, compared with an average of approximately 17% for all patients with NSCLC. Oligometastatic NSCLC is a subgroup of stage IV NSCLC with a limited number/number of sites of metastatic disease, usually 1–5 metastatic lesions. While only a small subset of patients present with such limited metastases (brain metastasis <50%, adrenal gland metastasis ~7%), they may be suitable for, and achieve long-term survival following, eradication of both the primary and metastatic tumours. For example, reported 5-year survival rates for NSCLC patients with solitary brain or adrenal metastasis who underwent surgical removal of both primary lung and metastatic disease ranged from 7–24% and 25–34%, respectively, which is higher than the average of ≤4% for all stage IV NSCLC patients. In addition to having an adequate performance status, there are additional key prognostic factors for identifying oligometastatic NSCLC patients likely to benefit from aggressive therapy; staging of the metastatic lesions, lymph node involvement and status of the primary lung tumour. A lower number of metastatic sites is predictive of good clinical outcome, with >2 metastatic sites associated with shorter PFS. The organ involved in metastatic spread may also impact clinical outcome as, for instance, patients with brain and adrenal gland metastases are more suitable for surgical intervention compared with bone or liver. Given the importance of number and site of metastatic lesions, positron emission tomography and magnetic resonance imaging are vital for accurate staging of oligometastatic disease. An association between N0 disease and increased long-term survival compared with N1/N2 disease has been observed. For example, following surgical treatment of one cohort of patients with solitary NSCLC adrenal gland metastasis, those with N2 disease had 0% median 5-year survival rate, compared with 52% for N0/N1 patients (P=0.001). Pathologic staging of lymph nodes is therefore critical. The status of the primary lung tumour also impacts clinical outcome, as the primary tumour must itself be resectable. Stage III primary disease is associated with worse survival outcome than stage I or II in patients undergoing surgical excision of brain and adrenal gland metastases, with a reported 5-year survival rate of 0% for stage III, compared with 63% and 77% for stage I and II, respectively. Similarly, patients with oligometastatic disease and a controlled primary site, or ‘oligorecurrance’, have better prognosis than those with an uncontrolled primary tumour. Histological subtype of NSCLC may also impact prognosis in oligometastatic NSCLC, with adenocarcinoma associated with the most favourable outcomes. Although an optimal disease-free interval (DFI) to define synchronous and metachronous disease has not been agreed upon, synchronous oligometastasis is generally associated with poorer survival outcomes. Patients receiving adrenalectomy for oligometastatic NSCLC with DFI ≤6 months had median OS of 12 months versus 31 months for DFI >6 months. Similar results were reported for isolated brain metastasis. Oligometastatic NSCLC is a stage IV cancer and as a guiding principle therapy should be simple and minimally invasive. Furthermore, given the diffuse nature of this disease state, management should ideally involve a multidisciplinary team as the primary and metastatic cancer must be treated, requiring a wide range of expertise. Surgery and radiosurgery (stereotactic radiosurgery [SRS] in the brain and stereotactic body radiotherapy [SBRT] in extracranial sites) are the two most common methods of tumour ablation. In general, radiosurgery is less invasive than surgery and is therefore useful for patients ineligible for surgery. Additionally, evidence suggests SBRT may be more applicable to limited extracranial metastasis to multiple organs compared with surgery. The treatment of limited brain metastases has been evaluated through several randomised clinical trials. Surgical resection plus whole brain radiotherapy (WBRT) of oligometastasis in the brain led to significantly prolonged OS compared with WBRT alone – as high as 40 weeks in one trial, compared with 15 weeks for WBRT (P<0.01). Further, a study of SRS with WBRT in patients with 1–4 brain metastases showed improved survival outcomes for patients with solitary brain metastases and improved clinical outcomes for patients with >1 site of metastatic disease, compared with WBRT alone. Overall, SRS plus WBRT is considered an acceptable choice for those patients with limited brain metastasis who are not suitable for surgery. Adrenalectomy is the standard of care for adrenal gland metastases, with OS ranging from 11–31 months. There have been no randomised trials of the use of SBRT in the adrenal gland setting, although one retrospective study reported OS of 23 months for isolated adrenal metastases treated with SBRT. Further, a recent review article of extracranial oligometastatic disease from various primary cancers suggests that survival following SBRT compares favourably to surgery. One of the key challenges for the management of oligometastatic NSCLC is continuing to improve diagnosis and prognostic factors to more accurately identify those patients with oligometastatic NSCLC who are likely to benefit from ablative treatment, as well as distinguish truly isolated metastatic disease from early-stage metastasis that later develops into widely disseminated disease. Continuing advances in imaging technology will play a role in refining diagnosis and prognosis. From a treatment perspective, challenges include cognitive problems associated with WBRT and the current lack of randomised trial data comparing surgery, radiosurgery and standard of care. To this end, there are several ongoing clinical trials, such as the randomized SABR-COMET study, which compares palliative radiation with stereotactic ablative radiation for ≤3 metastatic tumours to any single organ.

Abstract not provided

Abstract not provided

Abstract
In lung cancer field, the direction of treatment strategy is in the midst of great change since molecular-targeted-drug was introduced. One of the revolutionary drugs is EGFR-TKI for EGFR positive lung cancer. It contributes to the treatment not only of advanced lung cancer, but also resectable early lung cancer in stage II and III. Although it is common to conduct surgery as the standard of care for early-stage lung cancer treatment, Stage II and III cases show poor prognosis even though complete resection is done. According to Asamura et al, 5-year survival of stage IIA, IIB and IIIA Japanese patients are 61%, 47% and 32% respectively. In response to these results, many investigators have been trying neoadjuvant and adjuvant chemotherapy in combination with surgery. In many cases, relapse after standard resection is distant metastasis, thus it is suspected that in order to achieve better prognosis, possible micrometastasis in whole body needs to be eradicated. Currently combination therapy with platinum-doublet is commonly chosen as the strong treatment regimen for controlling the micrometastasis in advanced cancer cases. Gefitinib is an EGFR-TKI (tyrosine kinase inhibitor) approved for the treatment of unresectable or recurrent NSCLC. Since its approval in Japan in July 2002, it has been approved in about 70 countries and used as the therapeutic medication for advanced recurrent NSCLC (at the point of August 2010). When Gefitinib was initially released, 3 to 5% of drug-related ILD (Interstitial Lung Disease) and deaths were reported in Japan. After cohort study, it has been reported that those who are male, smoker, have the past illness of ILD and have low PS showed the higher ILD incidence rate than other cohorts. In 2004, it has been reported that Gefitinib shows the remarkably high response rate for EGFR positive patients, and subsequent studies demonstrated that EGFR mutation is an efficacy predictive biomarker.It is commonly known that Gefitinib induces apoptosis in EGFR positive patients. In fact, some clear anti-tumor effect have been observed in many clinical studies targeting on EGFR positive advanced NSCLC patients. The result of post-marketing Phase III clinical trial targeting on Asian patients untreated with chemotherapy (IPASS trial)was reported in September 2008. According to pre-planned subgroup analysis based on biomarker, in EGFR positive patient group Gefitinib cohort showed the significant prolongation of PFS (Progression Free Survival) compared to Carboplatin plus Paclitaxel cohort (HR 0.48, 95% CI 0.36～0.64, p<0.0001). In addition, Gefitinib cohort showed better efficacy, QOL, safety and tolerability. However, in terms of OS (Overall Survival), the difference was not observed in two cohorts (HR 1.00, 95% CI 0.76～1.33, p=0.990). On the other hand, in EGFR negative patient group, Carboplatin plus Paclitaxel cohort significantly prolonged PFS compared to Gefitinib cohort (HR 2.85, 95% CI 2.05～3.98, p<0.0001). It was reported from Japan that in the two Phase III controlled study targeting on EGFR positive advanced lung cancer patients untreated with chemotherapy, Gefitinibcohort showed better PFS (Hazard Ratio of less than 0.5) than standard chemotherapy cohort of Cisplatin plus Docetaxel (HR 0.489, 95% CI 0.336～0.710, p<0.0001) or Carboplatin plus Paclitaxel (HR 0.357, 95% CI 0.252～0.507, p<0.001) in both studies. However, the comparison of 2-year survival in the controlled study of Carboplatin plus Paclitaxel, there was no statistically significant difference (Gefitinib: 61.4%, Standard Chemotherapy of CBDCA plus PTX: 46.7%) (P=0.31). In addition, in two cohorts (Gefitinibvs CDDP plus DOC) there was no statistically significant difference in OS among them (HR 1.638, 95％CI 0.749～3.582, p=0.21), thus the superiority of Gefitinib was not demonstrated in terms of OS. In the above-mentioned three prospective controlled trial, it was demonstrated that Gefitinib dramatically prolongs PFS in EGFR positive NSCLC patients compared to standard chemotherapy, and shows high anti-tumor effect and improvement of QOL, but not in terms of OS. It is suspected this is partially because of the fact that in chemotherapy cohort many patients received Gefitinib after relapse. To sum up, Gefitinib shows higher anti-tumor effect than standard chemotherapy if the target patients are EGFR positive, and thus is an appropriate option as the adjuvant treatment regimen after complete resection. It was impliedfrom Phase I trial that Gefitinib shows anti-tumor effect for the solid tumor patients who did not benefit from standard of care or have no other treatment options and thus effective for various types of cancer. In Phase II monotherapy trial, Gefitinib showed clinically significant and continuous anti-tumor effect in locally-advanced or metastatic NSCLC patients previously treated with chemotherapy. Additionally, it improved the cancer symptoms in these patients. Orally administering 250mg or 500mg once per day showed the same level of efficacy. Although in Phase III monotherapy trial targeting on advanced NSCLC patients the efficacy of Gefitinib was demonstrated, there was no statistically significant difference in terms of OS, which is the primary endpoint. However, the response rate of Gefitinib cohort was higher than that of placebo cohort. In Phase III combination chemotherapy trial, standard chemotherapy (Gemcitabine/Cisplatin, Paclitaxel/Carboplatin) and Gefitinib 250mg per day or 500mg per day were evaluated. In this study however, statistically or clinically significant data superior to combination of standard chemotherapy with placebo was not observed. In conclusion, there are three reasons why patients with EGFR mutations should receive adjuvant EGFR-TKIs. Firstly, EGFR-TKI induces apoptosis and its exposition to EGFR positive patients potentially induces eradication of micrometastasis remaining after surgery. Secondly, for the EGFR positive patients, EGFR-TKI is superior to standard platinum-doublet in terms of safety and tolerability. Lastly, EGFR-TKI is the most potent anti-cancer drug for EGFR positive patients at the moment, which shows remarkable anti-tumor effect, prolongs PFS and improves QOL compared to platinum-doublet. Therefore, despite the issue of ILD remains, EGFR-TKI is to be considered as an appropriate treatment option in adjuvant setting for the EGFR positive patients. In order to prove this point, West Japan Oncology Group (WJOG) is currently conducting IMPACT trial, which is a prospective Phase III study.

Abstract
Adjuvant therapies are required to prevent disease recurrence and improve patient survival after surgery in early stage (stage I–IIIA) non-small cell lung cancer (NSCLC). In advanced disease, patients whose tumours have activating EGFR mutations have an approximately 60%-75% response rate to EGFR tyrosine kinase inhibitors (TKIs) treatment as first-line therapy and a progression-free survival that is significantly superior to standard chemotherapy. Although adjuvant TKI therapy is an appealing treatment strategy, can the remarkable efficacy of EGFR TKI in advanced EGFR-mutant NSCLC be extrapolated to stage I-IIIA disease? While adjuvant cisplatin-based chemotherapy prolongs survival with just 3-4 months of treatment, no trial to date has studied the value of a few months of adjuvant EGFR TKI. The phase III BR.19 trial by the National Cancer Institute of Canada designed to study 1,160 patients with fully resected stage IB, II and IIIA NSCLC randomised to receive 2 years of adjuvant gefitinib at 250 mg daily or 2 years of adjuvant placebo was greatly underpowered because enrolment was stopped prematurely when, in 2008, the SWOG 0023 investigators reported a worse overall median survival with maintenance gefitinib after definitive chemoradiation in a patient population not enriched for the presence of EGFR mutations with inoperable stage III NSCLC. EGFR mutation status was tested post-hoc in two thirds of 537 patients enrolled. There was no overall survival (OS) benefit for adjuvant gefitinib over placebo (HR 1.58; 95% CI, 0.83 - 3.0; P = 0.16) in 76 patients with EGFR mutant tumours. Despite an underpowered study, the negative finding should caution us about recommending adjuvant EGFR TKI therapy outside of a protocol setting. The phase III Randomized, Double-Blind Trial in Adjuvant NSCLC with Tarceva (RADIANT) compares 2 years of adjuvant erlotinib, at 150 mg orally per day, to 2 years of placebo following complete resection of stage I-IIIA NSCLC in 945 patients who were selected based on having either EGFR protein expression by immunohistochemistry or increased EGFR gene copy number by FISH. At the 2009 ASCO meeting, the RADIANT investigators presented the results of the biomarker analyses from the first 655 patients enrolled which showed an EGFR mutation-positive rate of 12%, which suggests that in the final result there will be approximately 38 EGFR mutation-positive patients on the placebo arm and about 76 on the erlotinib arm. This study is unlikely to provide sufficient power to a draw a firm conclusion about genotype-directed adjuvant therapy for patients with EGFR mutations. Investigators from the Memorial-Sloan Kettering Cancer Center reported a retrospective series of 167 patients at their centre with resected stages I-III NSCLC which were EGFR mutation-positive, among whom 32% received adjuvant EGFR TKI therapy. An improved 2-year disease-free survival (DFS) was observed in patients who received adjuvant TKI therapy compared to those who did not (89% vs 72%) (HR, 0.53; 95% CI, 0.28 - 1.03; P = 0.06). A trend toward more favorable 2-year OS in recipients of EGFR TKI therapy was also seen (96% vs 90%; HR 0.62, CI 0.26-1.51, P = 0.296). When the number of patients with resected EGFR mutant lung cancers was increased to 222, the same investigators in a more recent publication, showed treatment with adjuvant erlotinib or gefitinib was associated with a lower risk of recurrence or death, DFS HR 0.43 (95% CI: 0.26-0.72, P = 0.001), and a trend toward improved OS. The phase II SELECT (Surgically resected EGFR-mutant Lung cancer with adjuvant Erlotinib Cancer Treatment) trial is a single-arm, prospective study of two years of adjuvant erlotinib in 100 patients with resected, state IA to IIIA EGFR mutation positive NSCLC. Preliminary results presented at the 2012 ASCO meeting for the first 36 patients showed a DFS of 94% compared to a historical control of 70% with a median follow-up of 2.7 months. The answer provided by the SELECT trial is also unlikely to be definitive. The phase II/III Tailored Post-Surgical Therapy in Early Stage NSCLC (TASTE) trial by the French Collaborative Intergroup compares 4 cycles of standard cisplatin/pemetrexed (CP) adjuvant chemotherapy (arm A, n = 74) with customised adjuvant treatment (arm B, n = 76) based on EGFR and ERCC1 status in patients with completely resected stage IIA, IIB or IIIA (non-N2) (6[th] TNM edition) nonsquamous NSCLC. In the experimental arm, EGFR mutated patients received erlotinib 150 mg for one year. ERCC1 negative pts received four cycles of CP. EGFR mutation was identified in only 10 patients (3 in arm A, 7 in arm B). All 7 patients with EGFR mutation in arm B received erlotinib for a median duration of 276 days. The phase III was cancelled due to the unexpected unreliability of the ERCC1 immunohistochemistry read-out. Two randomised studies comparing gefitinib 250 mg daily for 2 years with adjuvant chemotherapy (4 cycles of cisplatin/vinorelbine) in patients with surgically resected stage IIA, IIB and IIIA EGFR mutation-positive adenocarcinoma are ongoing in Japan and in China (ADJUVANT CTONG 1104). In conclusion, it remains unclear whether targeted therapies improve outcomes over traditional chemotherapy in the adjuvant setting in NSCLC, and it is premature to recommend adjuvant EGFR TKI therapy outside well-designed clinical trials as there is currently no conclusive data on its role in early- and locally advanced NSCLC harbouring EGFR mutations. Given the consistent development of acquired resistance to EGFR TKI therapy in patients with metastatic NSCLC, there is good reason to question whether it is advisable to use EGFR TKI for adjuvant therapy in patients who have no evidence of disease and who can develop resistance to an otherwise effective treatment. Furthermore, the optimal duration of adjuvant therapy is yet to be defined and it is unknown whether EGFR TKI is merely by suppressing the growth of residual disease after surgery rather than eradicating minimal residual disease. Given the fact that adjuvant therapy needs to be given post-operatively for a prolonged period, the quality of life of patients while on therapy and the financial cost involved also need to be considered.

Abstract not provided

Background
KRAS proto-oncogene is one of the most frequent mutated genes in Non-Small Cell Lung Cancer (NSCLC) with greater incidence among adenocarcinomas (AD). While the clinical importance of KRAS mutation as a negative predictor for anti-EGFR therapy is not clearly understood in NSCLCs, selection of targeted therapies for KRAS mutated (MUT) patients has often focused on the inhibition of its direct downstream effectors. The aim of this study was to explore the impact of the KRAS status on the cellular signaling network of ADs of the lung harboring different KRAS mutations with a focus on ERK signaling architecture.

Methods
A total of 58 AD samples were collected from chemo-naïve patients at the H. Lee Moffitt Cancer Center & Research Institute (Tampa, FL) and at S. Maria della Misericordia Hospital (Perugia, Italy). Twenty-four tumors were KRAS wild type (WT) and 34 were KRAS MUT (G12C n=18, G12V n=9, G13D n=3 and G12D n=4, respectively). All samples were subjected to laser capture microdissection and reverse phase protein microarray to quantitatively evaluate the activation status of the MAP Kinase signaling network.

Results
Statistical analysis of signaling protein activation based on KRAS status revealed an overall increase in activation level of the MAPK signaling network in the KRAS MUT tumors compared to tumors expressing KRAS WT: ERK 1/2 (T202/Y204), Elk-1 (S383), p90RSK (S380), Smad2 (S245/250/255) and p70S6K (p<0.01; p<0.01; p<0.01, p=0.04 and p<0.01 respectively). Nevertheless, 6 KRAS WT patients (25%) showed activation of ERK greater than the median of the entire population and an overall MAPK signaling activation comparable to tumors harboring KRAS MUT. Eleven of the KRAS MUT tumors (32%) had ERK activation lower than the median of the population as a whole. Interestingly a high activation level of Estrogen Receptor alpha (ERα) (S118) was detected in the KRAS MUT tumors compared to the KRAS WT one (p=0.02). Moreover the nonparametric test performed to establish the correlation of activated ERK 1/2, Raf, B-Raf, C-Raf and Mek 1/2 with the expression/activation levels of the 152 endpoints analyzed in this study, revealed the activation of distinct pathways in the KRAS MUT tumors when compared to KRAS WT tumors. Significant correlations were detected with Akt, KRAS, their downstream substrates and with several receptor tyrosine kinases (p<0.0003).

Conclusion
Our results suggest that MAPK signaling activation was clearly observed in KRAS MUT tumors. However, the heterogeneity in the activation level of MAPK downstream substrates within KRAS MUT and WT tumors suggests that selection of patients for MAPK targeting might benefit from the evaluation not only of the mutation itself, but also from a direct analysis of the MAPK protein network architecture. In particular the role played by ERα in KRAS MUT tumors deserves further investigations as a possible novel therapeutic target in KRAS MUT adenocarcinomas of the lung.

Background
Personalized medicine frameworks centered on identification and therapeutic targeting of dominant oncogenic driver mutations are rapidly becoming a standard of care in the clinical management of patients with lung adenocarcinoma. However, little is currently known about the nature and impact of co-occurring genetic events on signaling output downstream of initiating oncogenes. This lacuna in our understanding is particularly pertinent for the subgroup of KRAS-driven tumors, where mounting data point towards considerable heterogeneity in pathway activation and clinical response to targeted therapies. Here, we report a comprehensive analysis of genetic events that co-occur with or are mutually exclusive of mutant KRAS in a cohort of 230 lung adenocarcinomas and assess the impact of individual co-mutations on signaling streams using data derived from state of the art transcriptomic and (phospho)proteomic profiling of primary tumors.

Methods
An integrated analysis of 230 lung adenocarcinomas from The Cancer Genome Atlas (TCGA) consortium was performed using mutation (whole exome sequencing), transcriptomic (RNASeq), and proteomic (reverse phase protein array) datasets. Fischer’s exact test was applied to identify secondary mutations that occurred more frequently in either KRAS-mutant (n=68) or KRAS-wild-type (n=162) tumors and (phospho)protein markers that associated with each co-mutation. Genes with a mutation rate of ≥3% in the overall cohort were included in the analysis.

Results
Mutations in 18 genes were associated with KRAS mutational status in patient tumors (p≤0.01). Mutations in EGFR (p=0.0001), NF1 (p=0.001), and TP53 (p=0.001) were negatively correlated with the KRAS mutation. On the other hand, mutations in STK11 were significantly more frequent in the KRAS-mutant cohort (p=0.004), as were mutations in ATM (p=0.023) and MTOR (p=0.045). The most significant positive association involved mutations in ARHGEF11, a gene that encodes a Rho guanine nucleotide exchange factor (p=0.0004). Mutations in STK11 (29.4%) and TP53 (29.4%), the two most highly prevalent genetic events within the KRAS-mutant cohort were mutually exclusive. Unsupervised hierarchical clustering of transcriptomic and quantitative (phospho)proteomic profiles revealed separation of STK11-mutant tumors at the first branch of the cluster dendrogram, indicating activation of distinct signaling pathways downstream of this key tumor suppressor gene. Several less frequent genetic events had prominent and consistent effects on signaling output. We focused our attention on signaling via the MAPK pathway which may impact clinical sensitivity to MEK inhibitors, one of the most promising classes of targeted agents currently in clinical development for KRAS-mutant tumors. Preliminary analysis suggests that mutations in 3 individual genes can identify a subgroup of tumors (19% of the cohort) with profoundly suppressed MAPK signaling flux.

Conclusion
Analysis of recurrent secondary genetic events may define distinct and clinically relevant subsets of KRAS-mutant lung adenocarcinoma. Efforts to refine the sub-classification further and assess the impact of co-mutations on sensitivity to molecularly targeted agents are underway and updated results will be presented at the meeting.

Background
RASSF1A gene promoter hypermethylation was previously shown to predict poor overall survival in the IFCT-0002 randomized phase 3 trial of neo-adjuvant platinum-based chemotherapy, in early stage (I & II) NSCLC. We investigated the molecular and cellular basis for such a dramatic influence.

Methods
We studied isogenic immortalized bronchial, non-tumorogenic, HBEC3 cell lines only differing by their K-Ras status (wild-type or mutant K-Ras Val12 allele), and a panel of lung cancer cell lines recapitulating the main molecular alterations encountered in lung cancer. RASSF1A protein was depleted by 80% using 2 specific siRNAs, followed by the evaluation of EMT markers and cell motility regualors using qRT-PCR, Western blot or Immunofluorescence. Migration of transfected cells was assayed by 2D wound-healing migration assays or 3D migration assays using transwell devices with or without a matrigel coating mimicking basement membrane (invasion assay), or an endothelial cell monolayer (trans-endothelial cell invasion). Phenotypic rescue was studied by using plasmids encoding full-length RASSF1A or RASSF1C isoform, and a construct encoding a SARAH-deleted RASSF1A protein, unable to interact with the Hippo/MST kinase. We also tested co-transfection of RASSF1A siRNAs together with siRNAs directed against Hippo pathway members LATS1/2, WW45, YAP. Depletion of RASSF1A was finally combined with expression of wild-type, activated or dominant negative RhoA, RhoB, Rac1 or CDC42 constructs.

Results
In each bronchial/lung cancer cell line tested, RASSF1A silencing led to EMT resulting in E-cadherin, Syndecan1, Zo-1, miR200 decrease and concurrent N-cadherin, vimentin, Twist1, miR-21 increase. RASSF1A silencing-induced EMT was associated with cytoplasmic to nucleus translocation of YAP transcription factor, the terminal effector of the Hippo signaling pathway. RASSF1A silencing reduced cell adhesion and increased 2D cell motility with collective migration features. RASSF1A knock-down increased 3D migration, invasion as trans-endothelial migration. These effects correlated with the up-regulation of RhoA, RhoC, CDC42, MMP2/14 mRNAs and down-regulation of RhoB, DIA1 and MMP9 mRNAs. We also observed an increase of adhesion/invasion signaling proteins, i.e. CD44v6, cofilin, ERM and NF2, cofilin being activated by inhibition of LIMK-induced phosphorylation. Finally we report that immortalized non-tumorogenic cell lines, unable to grow without adhesion, acquired the capacity to grow in soft agar when RASSFIA was knocked-down. Those effects were rescued by co-transfection of RASSF1A siRNAs with full-length RASSF1A cDNA, showing the specificity of the motile phenotype induced by RASSF1A silencing, but not by RASSF1C nor SARAH-deficient RASSF1A plasmids. SiRASSF1A-induced cell migration was inhibited by LATS1/LATS2/WW45 or YAP siRNAs, showing the involvement of the Lats/YAP signaling cascade. We finally show that RASSF1A knockdown-promoted migration was inhibited by using RhoB-Val14 constitutively active cDNA but not RhoBN19 dominant negative construct, and by specific RhoB GEFs or RhoB effectors (DIA1) constructs.

Conclusion
In lung cells, the RASSF1A protein acts as a migration-suppressor protein by regulating the LIMK/cofilin pathway through RhoB signaling. RASSF1A prevents YAP induced EMT by inhibiting its nuclear accumulation through LATS1/2 signaling, whereas Hippo/MST kinase seemed dispensable. We thus provide evidence, for the first time in human lung cancer cells, for a direct connection between RASSF1A signaling and the LATS/YAP pathway.

Background
Hypermethylated in Cancer-1 (Hic1) is a novel tumour suppressor gene that is frequently epigenetically silenced in adult tumours. In non-small cell lung cancer (NSCLC), loss of Hic1 expression is associated with reduced patient survival, suggesting reduced Hic1 expression is associated with malignant progression of NSCLC. However, whether Hic1 silencing is causal in lung cancer is not known. Hic1 is a transcriptional repressor that can regulate p53 function by repressing the expression of SIRT1, a class III histone deacetylase. We therefore hypothesized that loss of Hic1 function could cooperate with an oncogenic mutation in KRas (KRasG12D) to promote lung cancer initiation and/or progression in a similar fashion to genetic deletion of p53.

Methods
To address this question, we used a conditional genetic mouse model of lung adenocarcinoma in which administration of recombinant adenovirus expressing Cre can trigger recombination at loxP sites. When virus is administered to mice heterozygous for a conditional KRasG12D allele, mice develop multiple lung adenomas that progress to adenocarcinomas over 8-12 weeks. To test the function of Hic1 as a tumour suppressor, we created a conditional knockout mouse allele (Hic1[lox]), in which loxP sites flank the single coding exon.

Results
Following administration of Cre adenovirus, mice carrying the KRasG12D allele and homozygous for the Hic1[lox] allele, developed aggressive lung adenocarcinomas at a markedly accelerated rate and had a significantly shortened survival compared to KRasG12D animals. Remarkably, these tumours exhibited a highly malignant phenotype with highly proliferative micropapillary and pleomorphic features.

Conclusion
These data show that loss of Hic1 function can substitute for p53 mutation as a cooperating event in lung adenocarcinoma progression. Since the highly aggressive phenotype of KRasG12D/Hic1[lox/lox] lung tumours has not been reported in the KRasG12D/p53[lox/lox] lung cancer model, we speculate that Hic1 may function as a tumour suppressor beyond the regulation of p53 through Sirt1.

Background
Targeted cancer therapies often induce “outlier” responses in molecularly defined patient subsets.

Methods
One patient with advanced-stage lung adenocarcinoma, who was treated with oral sorafenib, demonstrated a complete clinical and radiographic remission for five years. Whole genome sequencing (WGS) and RNA sequencing (RNA-seq) on primary tumor and normal samples from this patient was performed.

Results
We identified a somatic mutation, ARAF S214C, present in the cancer genome and expressed at high levels. Additional mutations affecting this residue of ARAF and a nearby residue in the related kinase RAF1 were demonstrated across 1% of an independent cohort of lung adenocarcinoma cases. The ARAF mutants were shown to transform immortalized human airway epithelial cells and were associated with in vitro sorafenib sensitivity.

Conclusion
These results suggest that mutant ARAF may be a novel oncogenic driver in lung adenocarcinoma and an indicator of sorafenib response.

Background
The EGFR T790M mutation, commonly seen with acquired resistance to EGFR kinase inhibitors, has also been described rarely as a germline mutation in association with familial lung cancer. In a prior study (Oxnard et al, JTO, 2012), the presence of EGFR T790M at diagnosis was associated with a 50% chance of carrying an underlying germline T790M mutation. This suggests that by studying patients whose cancer was shown to carry T790M at diagnosis, it is possible to efficiently screen for a germline allele that otherwise is rare among patients with non-small cell lung cancer. We therefore initiated a prospective trial to identify patients and families carrying germline EGFR mutations in order to characterize phenotype and cancer risk.

Methods
Subjects are eligible if they (1) have a cancer harboring EGFR T790M (excluding acquired T790M), (2) are a relative of a known germline carrier, or (3) are already known to carry a germline EGFR mutation on prior testing. Subjects may present at a participating cancer center or may enroll remotely using a study website (www.dana-farber.org/T790Mstudy/). Eligible subjects receive genetic counseling in person or over the phone, and then submit a saliva and/or blood specimen for central testing in a CLIA lab. Results are disclosed to the subject if they wish but do not enter the medical record. Those subjects carrying germline EGFR mutations are given the option of inviting relatives to participate. Chest CT scans are collected from germline carriers and analyzed centrally to study nodule prevalence and characteristics. Available tumor specimens are collected for central pathology review and advanced genomic analysis.

Results
The trial was registered to clinicaltrials.gov (NCT01754025) and began accrual in December 2012. To date, 7 subjects have been enrolled and 5 are actively being screened, including 4 kindreds. More than half of the subjects have participated remotely via the study website. Of 4 probands with lung cancer and germline T790M, 3 have a family history of lung cancer, 2 of whom have children with CT scans showing multiple sub-centimeter ground-glass nodules. The fourth proband has no family history of lung cancer, suggesting variable penetrance or a de novo germline event. All cancers in germline T790M carriers have also harbored secondary EGFR kinase domain mutations.

Conclusion
Using a novel trial design, including remote accrual, genetic counseling by phone, and germline testing by mail, we have begun collecting a sizeable cohort of families affected by germline EGFR mutations. By leveraging referrals from commercial laboratories and contributing academic centers, we aim to study 100 patients over a three year period in order to better understand the natural history and risk associated with this unique familial cancer syndrome. Supported by grants from the Conquer Cancer Foundation of ASCO, the Bonnie J. Addario Lung Cancer Foundation, and the National Cancer Institute.

Abstract not provided

Background
Therapeutic regimens targeting the vascular endothelial growth factor (VEGF) pathway have been extensively tested in the treatment of malignancies including non-small cell lung cancer (NSCLC). VEGF pathway inhibitors including bevacizumab or VEGF receptor (VEGFR) tyrosine kinase inhibitors (TKIs) have been shown to prolong progression-free survival (PFS) and/or overall survival (OS). These benefits, however, have been modest, occurring only in subsets of patients. Therefore, predictive markers to identify patients likely to derive benefit are critically needed. Although expression of VEGFR-2, also known as KDR, was initially thought to localize primarily on endothelial cells, VEGFR-2 has been detected on malignant cells. We recently observed that KDR copy number gains (CNGs) were detectable by FISH in ~30% of both adenocarcinoma and squamous cell carcinoma and were associated with poor clinical outcome in early stage NSCLC patients treated with adjuvant chemotherapy. In addition to CNGs, mutations and polymorphisms within the KDR gene were also observed. The impact of these alterations is unknown. Here, we investigated KDR CNGs, polymorphisms, and mutations in NSCLC and their effects on sensitivity to VEGFR targeting agents in preclinical models and in NSCLC patients.

Methods
Cell migration was evaluated by Boyden chamber assay. NSCLC cell lines were treated with VEGF pathway inhibitors for 24 hours, and protein lysates where collected. HIF-1α levels were evaluated by ELISA assay. VEGFR, p38, and p70s6K were evaluated by Western blotting. Tumor DNA and peripheral blood DNA, were analyzed in duplicate using Affymetrix Genome-Wide SNP Array 6.0. Transformation of Ba/F3 cells was evaluated by an IL-3-independent growth assay.

Results
In tumor cells with KDR CNG, VEGF stimulation induced activation of p38 and p70S6K, and VEGFR TKIs including sorafenib and vandetanib effectively inhibited VEGF-mediated signal transduction. In tumor cell lines with KDR CNG, exogenous VEGF ligand increased cell motility and this was inhibited by VEGFR blockade with TKIs including sunitinib, sorafenib, and axitinib. Various receptor tyrosine kinases have been shown to drive HIF-1α levels, and NSCLC cells with KDR CNG express elevated levels of HIF-1α in normoxia compared to NSCLC cell lines without KDR CNG. In NSCLC cell lines with KDR CNG, VEGFR TKIs decreased protein levels of HIF-1α and HIF-1α regulated proteins. Furthermore, we report a clinical case in which a NSCLC patient with KDR CNG had a partial response to the VEGFR inhibitor, sorafenib. In addition to gene amplification, mutations and polymorphisms within the KDR gene were also observed. KDR mutation 1586A>T and polymorphism 1416A>T effectively transformed Ba/F3 cells. Finally, we report two clinical cases in which NSCLC patients with the 1416A>T polymorphism had a partial response the VEGF pathway inhibitor, bevacizumab.

Conclusion
Collectively, our data indicate that KDR amplification promotes downstream signaling events including activation of the p38, mTOR, and HIF pathways and are targetable by VEGF pathway inhibitors. KDR gene alterations may be predictive markers for VEGF pathway inhibitors.

Background
Characterization of lung cancer genomes has revealed a number of genes critical to tumorigenesis (e.g. EGFR, KRAS, EML4-ALK), resulting in significant changes to the treatment of lung cancer and an increase in survival for a subset of patients. These successes have prompted the search for additional driver alterations, leading to the discovery of a number of recurrently mutated or amplified genes and gene fusions with promising clinical utility. Distinguishing the key mechanisms and causal events driving tumorigenesis will lead not only to a better understanding of lung cancer phenotypes and biology, but also to new molecular markers and therapeutic targets. Using an integrative analysis of gene expression and copy number data to identify novel candidate oncogenes, we identified the chromosomal region at 12q13-15, and more specifically, the putative transcription factor YEATS4 (YEATS domain containing 4) as frequently amplified and overexpressed in NSCLC. Amplification of YEATS4 has been reported in dedifferentiated liposarcomas and in the earliest stages of glioma and astrocytoma.

Methods
Copy number profiles were generated for 261 NSCLC tumors (169 adenocarcinomas (AC) and 92 squamous cell carcinomas (SqCC)) and expression profiles for a subset of tumors with matched non-malignant tissue. Recurrent DNA amplifications were identified using the GISTIC algorithm. Copy number data were integrated with gene expression data to identify genes frequently amplified and overexpressed (defined as a 2-fold difference in expression between tumor and matched non-malignant tissue). The functional significance of YEATS4 was assessed by lentiviral knockdown in lung cancer cell lines with and without YEATS4 amplification and ectopic expression in human bronchial epithelial cells (HBECs). In vitro and in vivo assays measuring proliferation, anchorage independent growth, senescence, apoptosis, drug sensitivity and tumor growth were used to assess the phenotypic effect of YEATS4 gene expression manipulation.

Results
YEATS4 is gained or amplified and concomitantly overexpressed in over 20% of NSCLC tumors, with similar frequencies of amplification in both AC and SqCC. Although frequently co-amplified with MDM2, amplification of YEATS4 was observed to occur in the absence of MDM2 amplification, suggesting it is not merely a passenger event. Overexpression of YEATS4 in HBECs abrogated senescence, whereas knockdown reduced cell proliferation, impaired colony formation and induced cellular senescence in cell lines with YEATS4 amplification. Western blotting revealed increased p21, cleaved PARP and p53 in knockdown lines compared to empty vector controls, implicating YEATS4 as a negative regulator of the p21-p53 pathway. Moreover, YEAST4 expression was found to correlate with cisplatin sensitivity, as overexpression increased resistance and knockdown conferred sensitivity. Consistent with our in vitro findings, tumor size and growth were significantly reduced in mice injected with YEATS4 knockdown cells relative to control mice. Furthermore, survival analysis revealed that patients expressing high levels of YEATS display poorer outcomes.

Conclusion
Our findings reveal YEATS4 as a novel candidate oncogene frequently amplified and overexpressed in NSCLC. Gene expression manipulation resulted in distinct phenotypic changes consistent with oncogenic function, and suggesting YEATS4 amplification is a novel mechanism contributing to NSCLC tumorigenesis.

Background
Lung cancer remains the cause of the most cancer-related deaths each year, with a 5 year survival rate of less than 15%. The predominant type of lung cancer is non-small cell lung cancer, and the majority of these cases consist of the adenocarcinoma (AC) histology. Oncogenes such as EGFR and KRAS are well defined drivers of AC, but in approximately 50% of cases the driver alterations are unknown. Furthermore, not all defined drivers are drugable. Additional oncogenes are clearly involved in driving this subtype, and must be elucidated to better understand AC biology and improve treatment. ELF3 is an member of the E-Twenty Six (ETS) transcription factor family, which includes several well known oncogenes such as ETS1. Expression of ELF3 is uniquely epithelial-specific, with high expression in fetal but not adult lung tissue. ELF3 overexpression has been reported in a handful of clinical AC cases and cell lines, however a comprehensive analysis of the extent and impact of this overexpression is lacking. Therefore we conducted a multi-'omic, functional analysis of ELF3, and hypothesize ELF3 represents a novel oncogene in lung AC.

Methods
ELF3 was interrogated in a multidimensional integrative manner by assessing copy number (SNP 6.0), methylation (Illumina HM27), and expression (Illumina) data from a panel of 83 AC tumors and matched adjacent non-malignant tissues. ELF3 expression was also assessed in The Cancer Genome Atlas (TCGA) public database. Stable ELF3 mRNA knock-down models were established in AC cell lines with high ELF3 expression, and these models were used to assess the role of ELF3 in cell viability and proliferation via MTT and BrdU incorporation assay, respectively. Knock-down models were also used to assess the impact of ELF3 overexpression on tumor growth in vitro and in vivo by soft agar colony formation assay and flank injections of NOD-SCID mice. Subcellular localization of ELF3 was determined by western blot and confirmed with immunofluorescence. In addition, an ELF3 overexpression model was established in immortalized Human Bronchial Epithelial Cells (HBECs) to assess proliferation and soft agar colony formation in a non-malignant model system.

Results
ELF3 was found to be frequently overexpressed in our cohort (72%) and the TCGA cohort (80%). This upregulation correlated significantly with high frequencies of sequence gain (49%) and hypomethylation (71%), often seen within the same tumor. In fact, 82% of tumors with ELF3 overexpression had concurrent gain and/or hypomethylation of the ELF3 locus. Knock-down of ELF3 in cell models led to significantly reduced cell viability and proliferation. Western blot and IF revealed ELF3 to be predominantly located in the nucleus, indicating ELF3 likely behaves through its transcription factor activity. A similar hyperproliferative phenotype was seen in the HBEC ELF3 overexpression models.

Conclusion
The high frequency of ELF3 overexpression (>70%) observed in lung AC is accompanied by frequent DNA-level selection events. The affect of ELF3 on cell proliferation suggests that ELF3 is a novel oncogene in lung AC. Further studies are warranted to determine the mechanism by which ELF3 drives hyperproliferation and potentially other oncogenic functions to define novel drugable targets for this disease.

Background
Although platinum-based chemotherapy is the standard of care for most cases of advanced lung adenocarcinoma, its effectiveness is limited by the frequent incidence of innate chemoresistance. As a result, response rates rarely exceed 20%, even though cis-platinum and carboplatin are highly effective in other settings such as small cell lung, ovarian and testicular cancers. We hypothesized that innate chemoresistance in lung adenocarcinoma is mediated by one or more signalling pathways dependent on the expression of a single gene, and that these pathways could ultimately be targeted therapeutically.

Methods
To address this question, we developed a synthetic-lethal high throughput siRNA screen using the innately resistant A549 lung adenocarcinoma cell line. Optimisation of the screen was performed using a siRNA death control (PLK1), which induced cell death in the absence of platinum, and a sensitization control (MTOR), which enhanced cell death only in combination with a sublethal concentration of carboplatin. These independent controls revealed that the screening protocol performed within acceptable limits of variability, quality and reproducibility as determined by Z’ factor analysis. Screening was then performed using a pool of four siRNAs targeting a single gene in conjunction with vehicle treatment, or with carboplatin.

Results
After screening siRNAs targeting the 720 kinases, 256 phosphatases and 4794 “druggable” targets of the human genome, we identified 50 candidate targets based on fold change difference between platinum and vehicle treatments, and statistical significance determined by multiple t-test corrected for false discovery rate. Preliminary pathway analysis revealed a highly significant enrichment for genes in previously identified pathways as well as novel pathways.

Conclusion
These data demonstrate that a synthetic-lethal approach can be used to identify therapeutic targets that could potentially sensitize lung adenocarcinoma to platinum-based chemotherapy.

Abstract not provided

Background
Oncogenic driver alterations identify several types of lung adenocarcinoma with different prognosis and sensitivity to targeted agents. MicroRNAs (miRNAs) are a new class of non-coding RNAs involved in gene expression regulation. How miRNAs are dysregulated in lung cancer with ALK translocation, EGFR or KRAS mutation is largely unknown. In the present analysis we aimed to investigate miRNAs expression according to a specific molecular driver and to correlate miRNAs deregulation with patient outcome.

Methods
The study was conducted in a cohort of 67 lung adenocarcinoma patients (pts) including 17 ALK+ tumors, 11 ALK-/EGFR mutation+, 15 ALK-/KRAS mutation+, 24 ALK-/EGFR and KRAS wild-type and defined as triple negative cases. Matched normal lung tissues from 18 cases representative of the entire cohort were also included onto the analysis. RNA was isolated from formalin-fixed paraffin-embedded tissue (FFPE), using the Recover ALL kit (Ambion). NanoString nCounter system platform was used to generate the miRNA profile. We used Limma to test for differential expression analysis of data. Among the miRNAs evaluated, the miR-515 family expression between tissues was validated by RT-qPCRs, analyzed using the parametric t-test (unpaired, 2-tailed for validation).

Results
miRNA expression profile clusters distinctly ALK+ pts from ALK- and normal lung tissue. Within the ALK- group we found specific miRNAs subsets able to sub-stratify KRAS versus EGFR careers clustering sharply triple negative versus EGFR mutation+ and triple negative versus KRAS mutation+. miRNAs belonging to the miR-515 family seems to be the most deregulated in the ALK+ versus ALK-. Although their expression is stably high in normal tissues and ALK+ class, they are highly downregulated in KRAS mutated versus EGFR mutated and versus triple negative (p-value <0.001 for all comparisons).

Conclusion
miRNAs profile significantly differs in lung cancer pts with ALK translocation, EGFR mutations and KRAS mutations. Putative targets of deregulated miRNAs are under investigation to better define differences in driver-dependent pathway activation.

Background
Immunosuppressive regulatory T lymphocytes (Tregs) have been proved to play a critical role in immune tolerance to tumor. In this study we have analyzed several markers related to Tregs, in both tumor and stroma areas in patients with resectable NSCLC.

Methods
Tumor FFPE samples from 135 early-stage NSCLC patients were used in this retrospective study. The most representative areas of tumor cells and tumor stroma of each sample were carefully micro-dissected. RTqPCR using hydrolysis probes was performed to determine the expression of Treg markers such as: CD127, CD25, FOXP3, CTLA-4, IL-10, TGFB-1, LAG-3, GITR and TNF-a as well as CD4 and CD8. Relative gene expression was assessed using GAPDH and CDKN1B as endogenous controls and results were normalized against a human cDNA as a reference. FOXP3 protein expression was assessed by immunohistochemistry, in 80 of the 135 patients included in this study. The absolute number of FOXP3-positive lymphocytes was determined in both tumor and stroma areas by averaging the cell counts in 10 fields (400X). All statistical analyses were considered significant at p< 0.05.

Results
Gene expression analyses revealed an over-expression of CD25 (5.40X and 7.95X, respectively) and down-expression of CD127 (0.28X and 0.37X, respectively) in both, tumor and stroma. There was a tendency toward higher expression of FOXP3 (1.67X and 2.01X, respectively) and CTLA-4 (1.92X and 1.76X, respectively) as well. Paired Wilcoxon test showed significant gene expression differences between tumor and stroma in FOXP3 (p=0.006), CD25 (p<0.0001), CD4 (p<0.0001), CD8 (p=0.028), IL-10 (p<0.0001) and TGFB-1 (p<0.0001). Survival analyses revealed that patients with a “Treg profile” (↑CD25/↓CD127) had a reduced overall survival (OS), whilst those patients with higher levels of the ratio FOXP3 stroma/tumor had worse time to progression (TTP) (Table 1). Spearman test revealed a significant association between stromal FOXP3 expression levels and the number of FOXP3-positive lymphocytes (by IHC) in stroma, p=0.006. Moreover, chi-square test showed that patients with squamous cell carcinoma histology presented a higher number of FOXP3-positive lymphocytes than those patients with adenocarcinoma, p= 0.035. Table 1: OS for “Treg profile” and TTP for Ratio FOXP3 Stroma/Tumor

	OS
	Median (months)	95% CI	p
Others	74.33	65.96 - 82.69	0.003
"Treg profile"	29.90	4.91 - 6.54
	TTP
	Median (months)	95% CI	p
↓ Ratio FOXP3 Stroma/Tumor	NR	--	0.040
↑ Ratio FOXP3 Stroma/Tumor	32.50	16.25- 48.74

Conclusion
Gene expression of Treg markers in tumor microenvironment seem to play an important prognostic role in early-stage NSCLC patients. Furthermore, preliminary IHC analysis indicated a correlation between mRNA and protein levels for FOXP3 in NSCLC patients. Supported in part, by grants PS09/01149, RD06/0020/1024 and RD12/0036/0025 from Red Temática de Investigación Cooperativa en Cáncer, Instituto de Salud Carlos III (ISCIII).

Background
Estrogen receptor pathway has been reported to be interacted with epidermal growth factor receptor (EGFR) signal pathway. This study focused on the impact of intracellular ERβ localization (cytoplasmic or nuclear) on the efficacy of EGFR-TKI.

Methods
Tumor tissue specimens from 149 stage IV NSCLC patients treated with EGFR-TKI were analyzed using immunohistochemistry (IHC) for ER expression (ERαorβ) and their associations with clinicopathological variables and clinical outcomes. Significance of cyto-ERβ expression was further examined in NSCLC cell lines.

Results
The expression of ERα and ERβ was detected in 15% and 28.9% of the patients, respectively. Cyto-ERβ positive cases showed shortened progression free survival (PFS) compared with cyto- ERβ negative ones (3.1 months vs. 7.3 months, p=0.061). In the subgroup with concurrent EGFR mutation, the differences of PFS were enlarged with significant statistics (4.7 months vs. 10.9 months, p=0.042). COX’s proportional hazard model showed that female, EGFR mutation and c- ERβ negative expression were independent predictive factors for PFS. PC-9 cells present ERβ in cytoplasma as well as nucleus. Estrodial (E2) induced PC-9 cells moderately resistant to erlotinib with a 3-fold increase of IC50, and the resistance can be reversed by ER blocker (fulvestrant) or siRNA directed to ESR2. The function of E2 was accomplished by nongenomic activation (MAPK phosphorylation) caused by E2 via cyto- ERβ. Combination therapy with erlotinib and fulvestrant turned out to be far more effective than either treatment alone in PC-9 cells. Furthermore, 2 patients harboring both EGFR mutation and cyto-ERβ expression underwent PD of EGFR-TKIs, and re-obtained disease control after receiving combined EGFR-TKIs with fulvestrant.

Conclusion
Cyto-ERβ expression may predict relatively poor efficacy to EGFR-TKI compared with non- cyto-ERβ expression in NSCLC patients harboring EGFR mutation.

Background
Recent studies have demonstrated the feasibility of rebiopsy in patients (pts) with EGFR mutant NSCLC at the time of AR to the EGFR tyrosine kinase inhibitors (TKIs) erlotinib or gefitinib, and provide estimates of the prevalence of well described mechanisms of AR including the EGFR T790M mutation, MET amplification and small cell lung cancer (SCLC) transformation. HER2 amplification has also been described in cases of AR to EGFR TKIs, however, its exact frequency is still unclear. Moreover, comprehensive analysis of paired pre- and post-treatment samples to establish whether HER2 amplification is acquired during treatment with TKIs have not been performed. This prompted us to further investigate HER2 amplification in EGFR mutant NSCLC cases.

Methods
Pts with metastatic or recurrent NSCLC who developed AR while on a molecularly targeted agent were enrolled on an IRB approved repeat biopsy protocol. Tumor biopsies were obtained at the time of AR, and histopathological and molecular analyses of the tumors were performed. Known mechanisms of AR to EGFR TKIs were analyzed (T790M mutation, MET amplification and SCLC transformation) as well as amplification of HER2. The presence of T790M was assessed either by Taqman or pyro-sequencing (unless T790M status was available from an outside institution). HER2 and MET amplification were determined using fluorescence in situ hybridization (FISH).

Results
41 pts with AR to EGFR TKIs (erlotinib or gefitinib) were enrolled at YCC between Jan 2012 and May 2013. Histological analysis of all specimens revealed transformation of adenocarcinoma to SCLC in 3 cases (7%). Depending on the availability of tissue, samples were prioritized for T790M analysis followed by MET and HER2 amplification. T790M was identified in 36% of pts; MET and HER2 amplification were found in 11% and 10% of samples respectively. In the two cases with HER2 amplification, analysis of the pre-treatment specimen revealed that amplification of this receptor tyrosine kinase preceded treatment with EGFR-TKIs, however, the amplification level was lower pre-treatment in both cases. Specifically the ratio of HER2 to CEP17 probes was 2.8 pre-treatment in both cases and increased to 4.3 and 8 following TKI treatment. HER2 amplification was mutually exclusive with the other tested mechanisms of resistance.

Conclusion
T790M was the most commonly identified mechanism of AR to EGFR TKIs in the YCC cohort consistent with other studies. MET amplification, HER2 amplification and SCLC transformation were also observed. The observation that HER2 was amplified pre-treatment warrants further investigation of HER2 amplification in AR and pre-treatment specimens. Whole exome sequencing of specimens without known resistance mechanisms is ongoing.

Abstract not provided

Background
Epidermal growth factor receptor (EGFR) exon 20 insertion mutations account for ~10% of EGFR-mutated non-small-cell lung cancer (NSCLC), for the most part occur at the N-lobe of EGFR after its C-helix (after amino-acid M766) and have unsolved patterns of response to ATP-mimetic EGFR tyrosine kinase inhibitors (TKIs).

Methods
To understand the patterns of resistance or response to EGFR TKIs of EGFR exon 20 insertion mutations, we decided to study representative mutations using in vitro systems, structural models and also NSCLCs with these specific EGFR mutations.

Results
We selected three mutations located within the C-helix (A763_Y764insFQEA [identical to D761_E762insEAFQ], Y764_V765insHH and M766_A767insAI) and four mutations following the C-helix (A767_V769dupASV [identical to V769_D770insASV], D770_N771insNPG, D770_N771insSVD [identical to S768_D770dupSVD] and H773_V774insH [identical to P772_H773insH]) mutations. Our data indicates almost all EGFR exon 20 insertions are resistant to submicromolar concentrations of gefitinib or erlotinib; data that mirrors the lack of clinical response of NSCLCs with these mutations. The crystal structural and enzyme kinetic studies of a prototypical post C-helix EGFR TKI-resistant insertion, between residues D770_N771 (D770_N771insNPG), highlight that these mutations favor the active conformation (i.e., are activating), don’t alter EGFR’s ATP-binding pocket and are less sensitive than TKI-sensitive mutations. D770_N771insNPG is predicted to be 7.66 fold less sensitive than the TKI-sensitive EGFR-L858R. Unexpectedly, we identified the atypical EGFR-A763_Y764insFQEA as the only EGFR exon 20 insertion hypersensitive to EGFR TKIs using enzyme kinetic and cell line models. In patients with EGFR exon 20 mutated NSCLCs, the response rates to gefitinib or erlotinib were significantly higher for A763_Y764insFQEA (2/3; 66.6%) when compared to all other mutations within or following the C-helix (0/17, 0%; p=0.0158). The unorthodox homology model of A763_Y764insFQEA suggests a mechanism of activation (by shifting the register of the C-helix N-terminal) related to TKI-sensitive mutations (such as L858R or L861Q).

Conclusion
Our findings not only explain the intricate interplay between different EGFR mutations and their response to EGFR TKIs, but also have clinical implications for the treatment of EGFR exon 20 insertion mutated NSCLCs. Therefore, based on our data and previously published reports the aforementioned mutations affecting amino acids V765 to V774 should be classified as non-sensitizing to the reversible EGFR TKIs gefitinib and erlotinib. Our models may usher the development of EGFR TKIs specific for EGFR exon 20 insertion mutations.

Background
Genetic subtyping is increasingly being clinically relevant in NSCLC, and the search for novel targetable driver mutations is warranted. We intended to study the frequency and types of a vast number of potential druggable genetic aberrations in a large cohort of non-small cell lung cancers of all major histological subtypes. Herein we report the first findings.

Methods
Blood samples and tumor tissue was obtained from 96 operated early stage lung cancer patients admitted to Oslo University Hospital-Rikshospitalet in the period 2006-2011. Tissue was taken from the excised tumours, snap frozen in liquid nitrogen in the operation room, and stored at -80[o]C until DNA isolation. The tumor cell content in the specimens was found to be more than 70% in most samples. DNA was isolated from both tumor and corresponding blood sample according to standard procedures. High-throughput sequencing was performed using the SureSelect Human Kinome kit (Agilent Technologies), with capture probes that target 3.2 Mb of the human genome and include exons for all known kinases, select cancer-specific genes and their associated UTRs, in total 612 genes. The derived sequence reads were analyzed based on a pipeline including calling variations, somatic mutations, DNA copy number changes, indels and genomic rearrangements, as well as functional annotations.

Results
Tissues from 48 females and 48 males were analyzed; 73 adenocarcinomas, 21 squamous cell carcinomas and 2 large cell carcinomas. 55 patients were in stage I, 27 in II and 14 in stage III. 13 patients were never-smokers. 25 samples harbored a KRAS-mutation and 10 an EGFR mutation. The number of mutated genes per sample varied from 1 to 81. The median number of mutated genes was 14 in the overall cohort, 15 in the EGFR wildtype/KRAS wildtype tumors, 17 in KRAS- mutated patients, 5 in the EGFR-mutated group and 6 in the never-smoking patients (of whom 4 patients were EGFR-mutated).Figure 1

Conclusion
KRAS-mutated tumors contain the same amount of genetic aberrations as in wild-type tumors, whereas EGFR-mutated tumors show a much lower number of mutations per tumor. Never-smokers harbor a low number of mutations independent of EGFR-mutation status. Novel driver mutations are probably found in samples with low numbers of mutations.

Background
Recent advances in targeted therapy in NSCLC have achieved impressive results in advanced disease. For molecular testing,cytology samples are not commonly used since is less likely to be adequate. At ICO Badalona- Germans Trias i Pujol Hospital we have used cytology specimens when biopsy was not available. We describe the general results when using cytology specimens in NSCLC to detect EGFR mutation, KRAS mutations and ALK translocations.

Methods
From February 2007 to May 2012, 227 cytology samples from patients with NSCLC were collected at the Department of Pathology as cell block or fresh specimen over an apropiate slide. After that, tumor cells were(8-150) captured by laser microdissection. DNA sequencing for EGFR exons 18, 19, 20, 21, KRAS codons 12 and 13 was performed at Molecular Biology Laboratory( ICO-Badalona) and ALK translocation were analyzed at Pathology Department by FISH

Results
EGFR mutations were tested in 227 samples.The overall output was 86.3% (not evaluable in 15 , insufficient tissue in 8, no tumor cells in 4, not done in 4). EGFR mutation was detected in 8.81% (20/227). KRAS mutation were tested in 41 samples with results in 33, 80.5% (2 not evaluable, insufficient tumor cells 3, no tumor 1 and not done 2 samples). KRAS mutation was positive 6 (14.6%). ALk translocation were tested in 9 p with results in 6 p ( 1 not evaluable and 2 insufficient tumor cells) Both cell-block and fresh specimen over an apropiate slide were used to perform molecular testing. The output for cell-block was 83.3%(124/148) and testing was not possible in 23(11 not evaluable, 6 insufficient tumor cells, 4 not tumor and 3 not done). The output for membrane was 91.1% (72/79) and was not possible in 7(4 were not evaluable, 2 insufficient tumor cells and not done in 1). 54.7% of samples were obtained from endobronquial ultrasound guided transbronquial needle aspiration of mediastinal adenopathies, 11.3% lung mass needle aspiration and 11.7% from pleural effusion.

Conclusion
Our results support the potential use of cytology samples for molecular testing in NSCLC when biopsy specimens are not available. Both membrane preparations and cytology blocks have been used and are equally suitable for molecular testing.

Background
The majority of actionable drivers in SQCLCs occur in the PI3K (30%) and FGFR1 (20%) pathways. The biologic behaviors and natural histories of these subtypes are not well characterized. Characterization of these data may help to elucidate the biologic relevance of these putative oncogenic events.

Methods
As of October 2011, all patients with SQCLCs at MSK have undergone prospective, multiplex testing of their FFPE tumors for FGFR1 amplification (FISH, FGFR1:CEP8 ≥ 2.2), PIK3CA mutations (Sequenom and exon sequencing), PTEN loss (IHC, Cell Signaling), and PTEN mutations (exon sequencing), among others. The PI3K abberant group was defined as PIK3CA mutant, PTEN complete loss, or PTEN mutant. Patient characteristics, outcomes, and metastatic sites were identified. Survival probabilities were estimated using the Kaplan-Meier method. Group comparisons were performed with log-rank tests and Cox proportional hazards methods.

Results
77 stage IV SQCLC patients were analyzed. Genotypes were: FGFR1 amplified (23%); PTEN loss (22%), PIK3CA mutant (8%), PTEN mutant (7%). Events were non-overlapping save for 2 cases with PTEN nonsense mutations and PTEN loss. The sole significant clinical difference (KPS, age, sex, lines of tx, smoking status) was sex (women in PI3K group 52% vs. in others 23%, p=0.02). Metastatic patterns for PI3K and FGFR1 vs. all others were:

Site	PI3K	p	FGFR1	p	Other	Total
Brain	6 (22%)	0.002	0 (0%)	0.6	0 (0%)	6 (7%)
Pleura	5 (19%)	0.4	5 (28%)	0.7	9 (28%)	19 (25%)
Liver	5 (19%)	0.4	1 (6%)	1	1 (3%)	7 (9%)
Bone	8 (30%)	0.8	3 (17%)	0.7	10 (31%)	21 (27%)
Lung	12 (44%)	0.8	10 (56%)	0.2	12 (38%)	34 (44%)
Adrenal	3 (11%)	1	3 (17%)	1	4 (13%)	10 (13%)
Pericardium	1 (4%)	1	1 (6%)	0.3	0	2 (3%)

Median OS for PI3K vs. all others: 9mo (95%CI:8-NR) vs. 16mo (95%CI:11-NR), p=0.004. Median OS for FGFR1 vs. all others: 20mo (95%CI:11-NR) vs. 10mo (95%CI:9-16), p=0.06. Multivariate analysis for risk of death: PI3K HR=3.3 (95%CI:1.5-7, p=0.003); FGFR1 HR=0.5 (95%CI:0.2-1.1, p=0.06); Age ≥65, HR=1.3 (95%CI:0.6-2.8, p=0.5); KPS≤70, HR=3.2 (95%CI:1.6-.6.4, p<0.001); Lines of therapy ≥ 2, HR=2.3 (95%CI=0.8-5.7, p=0.08), male gender, HR=0.7 (95%CI:0.3-1.4, p=0.3).

Conclusion
Patients with stage IV PI3K-aberrant SQCLCs have poorer survival compared to other patients with SQCLCs while patients with FGFR1 amplified SQCLCs have a trend towards better survival. Brain metastases in SQCLC are rare, and occurred exclusively in patients with PI3K-aberrant tumors. These data suggest that PI3K pathway activation confers a distinct biology, and that targeting this in SQCLC patients with brain metastases may be an effective therapeutic strategy. Whole exome and RNA-sequencing data from 8 resected SQCLC brain metastases (4 paired with lung primaries) will be presented.

Abstract not provided

Background
Response assessment using conventional RECIST criteria after SABR of lung targets can be confounded by fibrotic response. The purpose of this study was to evaluate the utility of 4D-FDG-PET/CT and CT perfusion scans in the response assessment of single fraction SABR for inoperable pulmonary oligometastases.

Methods
This is a prospective ethics approved clinical study of patients undergoing single fraction SABR with 26Gy for pulmonary metastases. Eligible patients had 1-2 metastases with no extrathoracic disease on staging FDG-PET. Serial 3D / 4D-FDG-PET and CT perfusion studies were performed at baseline, 14 days and 70 days after therapy. Two radiologists independently reported CT perfusion scans.

Results
At a median follow-up of 16 months (range 3-27), 10 patients with 13 metastases received SABR. A further 7 patients (41%) were screened from the study due to interval progression of disease between the time of the original FDG-PET and trial 4D-FDG-PET / perfusion CT. The mean time between the original FDG-PET and trial scans was 62 days. No patient progressed locally, 7/10 patients progressed distantly of which 2/7 received subsequent SABR. At the end of study period, 5/10 patients are alive without disease. The median progression free survival was 14 months. The change in SUVmax from baseline was higher on 3D than 4D-PET by a mean of 20.6% (range 0.2%-47.2%) at 14 days and 14.8% (range 0-37.8%) at 70 days. Overall, the SUVmax increased at 14 days (mean 104.9%, p<0.01) and decreased at 70 days (mean=55.5%, p<0.01), despite persistent morphological lesions on the concurrent late timepoint CT. There was strong level of inter-observer agreement of CT perfusion interpretation with a median intraclass correlation coefficient of 89% (range 57%-98%). Perfusion parameters of Time to Peak Blood Flow and Blood Volume showed a median increase of 18.8% and 23.0% at 2 weeks post-therapy and decreased below baseline by a median 7.0% and 14.0% at 70 days (non-significant).

Conclusion
High rates of interval progression between staging scans indicates a need to expedite management of oligometastases in a timely fashion. Increased tumour perfusion and FDG-PET intensity at 2 weeks post-RT is likely due to an inflammatory response to large single dose SABR. Late PET response was associated with tumour control despite CT apparent morphological lesions. Conventional 3D PET may overestimate change in PET intensity post SABR as compared to 4D PET. These findings, in particular CT perfusion findings, require a larger patient cohort for validation.

Background
The standard of care worldwide for radiation dose scheduling in NSCLC has historically been 2Gy per fraction treating once daily over six weeks. The CHART regimen of accelerated hyperfractionated treatment first demonstrated significant survival benefit from a two week radical course, attributed to reduced repopulation and improved local control. A recent individual patient data meta-analysis confirms significant survival benefit from accelerated radiotherapy[1]. Meantime the evolving data on stereotactic radiotherapy treating early stage lung cancer over two weeks or less suggests marked improvement in local control rates compared with historical populations treated with conventional fractionation[2]. The latest challenge to 2Gy per fraction comes from the early stop to the dose escalation arm of RTOG 0617 with 74Gy actually appearing inferior to the 60Gy arm. We postulate that overall treatment time is a key factor in lung cancer radiotherapy outcomes and that standards of care need to be reviewed. This paper examines the current international guidelines on radical radiotherapy schedules, evaluates the supporting evidence and proposes new priorities for research.

Methods
Five international guidelines on the management of lung cancer were reviewed. All were published 2010-2013: ESMO Clinical Practice Guideline on early stage and locally advanced lung cancer 2010, NICE guideline (England and Wales) 2011, Australian Government Clinical Practice Guideline for treatment of lung cancer 2012, Cancer Care Ontario evidence based series lung cancer guideline 2013, National Comprehensive Cancer Network (NCCN) Lung Cancer guideline v2.2013. Recommendations on radical radiotherapy dose and fractionation for NSCLC were collated from each guideline together with the references cited in support of these recommendations to assess levels of evidence.

Results
Two guidelines specifically recommended hypofractionated SBRT for early stage inoperable disease -NICE and NCCN. The Australian guideline stated uncertainty over relative benefit SBRT versus conventional fractionation in stage I disease. England, Ontario and Australia all included CHART regimen as treatment of choice for stage II/III radical patients not receiving chemotherapy. Cancer Care Ontario undertook specific review of altered fractionation schedules identifying lack of evidence for hyperfractionation but suggesting possible benefit for hypofractionation. All five guidelines specified standard care, if given with chemotherapy, of conventional fraction size 2Gy: Ontario, ESMO and Australian guidance was a minimum of 60Gy in 30 fractions. NCCN offered a range of 60-74Gy at 2Gy/fraction. NICE alone proposed alternative standard of 55Gy in 20 fractions, a common UK schedule, or the option of 64-66Gy at 2Gy/fraction

Conclusion
International guidelines lag behind the emerging evidence for lack of benefit from dose escalation at 2Gy/fraction and apparent benefit from shorter treatment courses . We propose that accelerating treatment with hypofractionation and shorter overall treatment times should be the priority for radiotherapy development. We discuss current and pending trials examining this approach. 1. Mauguen A, Le Pechoux C, Saunders M et al: Hyperfractionated or accelerated radiotherapy in lung cancer: an individual patient data meta-analysis. J Clin Oncol 30:2788-2797, 2012 2. Timmerman R, Paulus R, Gavin J et al: Stereotactic body radiation therapy for inoperable early stage lung cancer. JAMA 303:1070-1076, 2010

Background
Reported non-small cell lung cancer (NSCLC) nodal failure rates following stereotactic body radiotherapy (SBRT) are lower than those reported in the surgical series when matched for stage. We hypothesize that this effect is due to incidental prophylactic nodal irradiation.

Methods
A prospectively collected group of medically inoperable early stage NSCLC patients (n=179) from 2004 to 2010 was used to identify a patient cohort with nodal relapses (n=19). These cases were matched, 1:2, to controls, controlling for tumour volume (i.e. same or greater) and tumour location (i.e. same lobe). Reference (normalized total) point doses at the ipsilateral hilum and carina, demographic data, and clinical outcomes were extracted from the medical record. Multivariate logistical regression analyses determined variables of interest.

Results
The case and control cohorts were well matched with respect to age, sex, method of nodal staging, SUVmax, histology subtype, dose and length of follow up.. The controls, as expected, had larger gross tumour volumes (p=0.02). The mean hilar doses were 9.6 and 22.4 Gy for cases and controls, respectively (p=0.014). Similarly, the mean carinal doses were 7.0 and 9.2 Gy, respectively (p=0.13). The mean ipsilateral hilar doses were 19.8 and 3.6 Gy for ipsilateral non-hilar and hilar nodal relapses, respectively (p=0.01). The conditional density plot appears to demonstrate an inverse dose-effect relationship between ipsilateral hilar normalized total dose and risk of ipsilateral hilar relapse (Figure 1).Figure 1

Conclusion
Incidental hilar dose greater than 20 Gy (normalized to 2Gy/fraction) appears to be correlated with lack of hilar relapses in inoperable early stage NSCLC patients treated with SBRT.

Background
Fiducial markers enable lesion tracking and localization with radiosurgery. Complications with percutaneous insertion are very common with pneumothorax rate reported as high as 67%, chest tube insertion 22% and migration of marker 19%. The purpose of this study was to assess complications associated with the bronchoscopic placement of a new commercial fiducial marker, designed for bronchoscopic insertion and to reduce migration. Twenty-one consecutive patients are reviewed in which 60 Cobra® (SuperDimension) fiducial markers were placed using electromagnetic navigational bronchoscopy. Accuracy of placement, utility of each marker, complications and migration are reported.

Methods
The use of these markers was approved by a advisory committee at our institution. Records of 21 consecutive patients (12 men, 9 women; mean age 61) referred to the Interventional Pulmonary Division for fiducial marker placement before initiation of cyber knife radiosurgery (Accuray, Sunnyvale CA) between December 15, 2012 and June 15, 2013. Indications for radiosurgery included non-surgical patients with nonsmall cell lung cancer and metatatic disease to the lungs from colorectal carcinoma and renal cell carcinoma. Our institution's radiation oncologist requested between one and three fiducials placed within or adjacent to each lesion. A total of 60 Cobra® fiducial markers were placed. In each insertion procedure, a computerized tomogram of the chest was used to preplan ideal insertion site and fiducial location in relation to tumor mass. Bronchoscopy was performed, using the SuperDimension planning and navigation, 1 to 3 Cobra® fiducial markers were placed in proximity to 22 different tumors.

Results
There were no instances of pneumothorax, with patient followup to one week. There no instances of fiducial migration greater than 3 mm from insertion site. There were 7 episodes of post procedure events, which included cough, dyspnea and hypoxemia. All events resolved prior to patient discharge from the outpatient treatment area. Imaging included chest radiograph, post procedure, on same day and CT chest within 1 week of procedure.

Conclusion
In this limited series, the Cobra® fiducial marker, using bronchoscopy and SuperDimension planning and navigation resulted in no instances of pneumothorax and no significant fiducial migration. This contrasts many reports of percutaneous fiducial placements regarding complications and migration.

Background
SBRT is an accepted safe and effective treatment modality for peripheral (P) stage I NSCLC tumors. Concern of excessive toxicity, however, limits its use for central (C) tumors. This study evaluates outcomes and toxicities after cone-beam CT (CBCT) image-guided SBRT for central vs. peripheral NSCLC.

Methods
959 lung tumors were treated with lung SBRT from 1998-2012 at five international centers participating in the Elekta Collaborative Lung Research Group; 98% underwent online CBCT IGRT. 100 cases were classified as Central (C) and 869 Peripheral (P), defined as ≤2cm vs. >2cm from the proximal bronchial tree, respectively. Staging included chest CT and routine chemistry for all; 93% had PET staging (mean time PET to SBRT 6.4 weeks); 6% had mediastinal sampling (mediastinoscopy or endobronchial ultrasound). 61% had tumor biopsy (84% C vs. 59% P, p<0.001). 89% were medically inoperable with mean baseline FEV1 of 1.6L (63% of predicted) and mean baseline DLCO of 12.1 ml/min/mmHg (56% of predicted). Mean age was 74y (42-93) with a large range in ECOG performance status (27%; 47%; 23%; 26% for 0-3, respectively). Clinical stage was T1aN0 44%, T1bN0 30%, T2aN0 23%, T2bN0 32%. Mean tumor maximum dimension was 2.5cm (range 0.5-8.5cm); C tumors were larger (mean 3.lcm vs. 2.4 cm, p<0.001). Mean SBRT prescription dose was 51.5±6.4 Gy, with mean dose per fraction of 14.5±4.0 Gy in 3.9±1.5 fractions. Mean biological equivalent dose (BED) was 126.6±26.6 Gy, higher for P vs. C tumors (129.2 vs. 104.0 Gy, p<0.001. Chemotherapy was administered more for C (9%) than P tumors (2%), p<0.001. Groups were compared with t-test & chi-square. Competing risks analyses were used, accounting for the competing risk of death.

Results
Mean follow-up for all cases was 1.8y (0.1-7.7y; mean potential follow-up 3.4y), similar for C&P. C tumors had higher Local Failure (LF) (3y-LF 16.2%C vs. 5.9%P; 5y-LF 20.4%C vs. 8.3%P, p<0.001), similar regional nodal recurrences (RR) (3y-RR 12%C vs.12%P, p=0.69) and distant metastases (DM) (3y-DM 19%C vs 20%P, p=0.75), lower cause-specific survival (CSS) (3yr-CSS 75%C vs. 88%P, p<0.001), but similar overall survival (OS) (3y-OS 50%C vs. 51%P, p=0.70). Grade > 2 pneumonitis was higher for C tumors (8%C vs. 1%P, p<0.001). Incidence of grade 3 pneumonitis, chest wall pain/myositis, rib fracture, and skin dermatitis were rare (0.8%, 0.5%, 0.4%, 0.6% respectively for all) with no differences between C&P. No grade 4 toxicities were noted, though 2 cases (1C & 1P) of fatal pneumonitis were potentially attributable to SBRT. On multivariate analysis, BED (HR:0.975, p<0.001) predicted CSS, and both BED (HR:0.978, p=0.002) and baseline SUVmax (HR:1.04, p=0.001) predicted LF. Weeks from PET-staging until SBRT (HR:1.25, p=0.004) and the percent of lungs receiving >20 Gy (HR:1.063, p=0.001) were the strongest independent predictors of OS.

Conclusion
In this large data set, pneumonitis was higher for central tumors, but both central & peripheral SBRT were safe with similar overall and cause-specific survival. LF was higher for central tumors, which were larger, had higher baseline SUVmax, and received lower dose. Results of the ongoing RTOG 0813 dose-finding study for central tumors are awaited.

Abstract not provided

Background
In the framework of a coverage with evidence development program on innovative radiotherapy techniques in Belgium, the cost of stereotactic body radiotherapy (SBRT) was calculated and compared to the cost of more standardized 3D-conformal (3D-CRT) and intensity-modulated (IMRT) radiotherapy treatments.

Methods
Activity-Based Costing methodology was used to calculate resource costs of radiotherapy treatments delivered in ten operational Belgian departments. Cost inputs were defined as personnel costs (number of full-time equivalents (FTE) devoted to the actual radiotherapy process times reference wages according to the guidelines of the Belgian Health Care Knowledge Centre (KCE)), equipment costs (including maintenance and upgrade) and specific material costs. Following KCE guidelines, overhead was accounted at 56% of global costs excluding physician wages. The activities in scope comprised all activities performed during the radiotherapy process from the first consultation, over treatment preparation, delivery and quality assurance until completion of the treatment. Products included all radiotherapy treatments delivered in each specific department and combined indication with treatment site and technical complexity. In view of the comparative analysis, products were aggregated into larger categories.

Results
The average cost of all SBRT treatments was calculated at 6,221€ (range 3,104€ - 12,649€) and compared favorably to the average cost of standard fractionated 3D-CRT (5,919€, range 4,557€ - 6,564€) and IMRT (7,379€, range 5,054€ - 8,733€). The average cost of hypofractionated 3D-CRT and IMRT was lower (3,993€ res. 4,730€). Apart from differences in investment costs, the relatively larger variability in fraction number and in time requirements for individual personnel types performing the radiotherapy activities explain the larger spread in treatment cost of SBRT compared to more standardized radiotherapy treatments. The figure demonstrates these differences for various technical SBRT solutions and for different 3D-CRT and IMRT fractionation schedules. The overall averages are shown by the bars, minimum and maximum center averages by the error bars. The number of centers is mentioned between brackets. Activity times shown combine time per personnel with number of FTE. Figure 1

Conclusion
Cost calculation of radiotherapy treatments at the multi-institutional level using Activity-Based Costing is feasible. SBRT shows larger variation in cost than more standardized radiotherapy approaches in line with the larger variability in technical solutions, time requirements and resource consumption. Its average cost however does not exceed the average cost of standard curative radiotherapy. Careful interpretation of these variables within the applicable economic context is required when using such cost data for determining financing levels.

Background
In early stage non-small cell lung cancer (NSCLC) stereotactic body radiotherapy (SBRT) has become standard of care for inoperable patients. Tumor size >3cm was reported to be a predictor of local recurrence (LR), suggesting a dose-volume effect. Recently, the dose effect relation was questioned[1]. We used a Tumor-Control-Probability (TCP) model on a large pooled multi-center cohort to test this.

Methods
850 patients were analyzed from our five institutes. Patients received a 4D CT-scan and plans were inversely optimized using advanced dose calculation algorithms. Treatment was delivered using online cone-beam CT guidance. Immobilization, margins, dose prescription and treatment planning was performed according to institute specific protocols. Median tumor diameter was 2.2 cm (range:0.7-8.0), median prescribed dose was 54 Gy (range:18-64) and median number of fractions were 3 (range:1-10). LRs were either biopsy proven or defined as a FDG-PET positive growing mass on CT-scan. The Web-Nahum TCP-model[2] was fitted to LR-data using maximum-likelihood estimation by optimizing its parameters: α representing the population-average radio-sensitivity, σ~α~ representing the population-variation in α and ρ the clonogen density. Input variables were the patient specific Gross Tumor Volume (estimated from the tumor diameter), for the dosimetric parameter PTV-D~min~, D~max~, D~mean~, D~1~, D~99~ were evaluated after conversion to Biological-Effective-Dose (BED) using the LQ-model with α/β=10Gy. We tested the optimized TCP model against a random model in which TCP was fixed independent of dose and volume. The optimal model was selected based on the Akaike-Information-Criterion (AIC).

Results
After a median follow up (FU) of 17 months (range:0-93), 43 LRs (5%) were diagnosed at 14 months FU (range:2-56), of which 25 tumors were biopsy proven and 18 recurrences diagnosed on PET-CT. The PTV-BED~mean~ based TCP model showed the best fit with parameters α=0.43Gy[-1] (CI:0.33–0.75) and σ~α~=0.17 Gy[-1] (CI:0.11–0.37). The model-fit was insensitive to ρ and set to literature values: 10[7]/cm[3]. The AIC of the optimal model was 12 units higher than the random model indicating a clear dose-volume-effect. At high PTV~mean~-BEDs, however, the volume effect is modest. Additionally, the AIC of the BED corrected model was 9.4 units higher than the BED uncorrected model. Figure 1

Conclusion
A dose-volume-effect relation in SBRT for early stage NSCLC for local control was derived in a large cohort of patients. This dose-effect relation requires validation in independent datasets and prospective trials. 1.van Baardwijk,Rad.Onc.,2012. 2.Web&Nahum,PMB,1993.

Background
Stereotactic body radiotherapy (SBRT) is an effective treatment for early-stage non-small cell lung cancer (NSCLC) and lung metastases. However, increased toxicity has been observed for SBRT to lesions near the proximal airways or mediastinal structures. Reported toxicities have primarily pertained to pulmonary complications, but little is known about the risk for esophageal toxicity. Therefore, we sought to evaluate dosimetric predictors of esophageal toxicity in this patient cohort at our institution.

Methods
We identified 125 patients who received SBRT for single lung tumors within 2 cm of the proximal bronchial tree (n=81) or whose planning target volume (PTV) intersected mediastinal structures (n=44). Ninety-one patients had primary NSCLC, 12 had recurrent NSCLC, and 22 had metastatic tumors involving the lung. Patients with prior thoracic radiotherapy were excluded. Toxicity was scored using the Common Terminology Criteria for Adverse Events v.4.0. Biological equivalent doses (BED) were calculated using the linear quadratic formula with either α/β=3 or 10 Gy. Dose-volume histogram variables for the esophagus (D~v~, minimum dose to the hottest volume v and V~d~, volume receiving doses greater than d) were then examined for all patients and correlation with toxicity was assessed using logistic regression. Log rank tests were performed using median splits for variables that were significant in logistic regression.

Results
With a median follow-up of 14.3 months, the overall rate of grade ≥2 esophageal toxicity was 12.8% (n=16), including two grade 3 events. The median prescription dose was 45Gy. The most common fractionation schemes were 45Gy in 5 fractions (n=56), 48Gy in 4 fractions (n=21), or 50Gy in 5 fractions (n=14). Highly significant logistic models were generated on the basis of D~3.5cc~, D~5cc~, and D~max ~(p<0.001). For a complication rate < 20%, D~3.5cc~ ≤ 29.4 Gy~10~, D~5cc~ ≤ 25.4 Gy~10~, and D~max~ ≤ 50.1 Gy~10~ was observed based on these models (BED~10~). Log rank tests showed that at 2 years, the probability of complication of those with a BED~10~ D~3.5cc~ > 16.6 Gy was 25% (p<0.001), D~5cc~ > 15.1 Gy was 26% (p<0.001), and a D~max~ > 29.6 Gy was 21% (p=0.032). The probability of complication for those with a D~3.5cc~, D~5cc~, and D~max~ (BED~10~) less than or equal to the above limits were 2%, 2% and 7%, respectively. The analysis was insensitive to α/β, and the same D~v~ variables were found to be significant using α/β =3.

Conclusion
This is a novel quantitative analysis providing dose guidelines for significant esophagitis in the setting of SBRT. Dose to the hottest 3.5cc, 5cc and D~max~ were the best parameters for prediction of esophageal toxicity. Converting the BED~10~ limits to physical doses, D~3.5cc ~to the esophagus should be kept less than 18.3, 19.7 and 20.8 Gy for 3, 4, and 5 fractions, respectively, to keep the esophagitis rate < 20%. However, these guidelines must be weighed against clinical considerations and potential compromise of target coverage. This information will be valuable for treatment planning and identifying patients at risk for esophageal complications from SBRT.

Background
Treatment-related radiologic change occurs commonly following stereotactic ablative radiotherapy (SABR) and often confound the interpretation of follow-up CT scans. SABR is frequently delivered using both fixed-beams and rotational-arcs, resulting in different dose distributions and it is unclear how this influences radiological change. We studied the morphology, timing and severity of radiologic change after both delivery techniques.

Methods
Twenty-nine patients with early stage non-small cell lung cancer receiving SABR by arc delivery, without clinical evidence of local recurrence, and a follow-up of more than two years, were assessed using a published scoring system [Dahele M, JTO 2011]. Here, the morphology of acute (within six months) radiologic change was characterized between ‘patchy (less than 5 cm) ground glass opacity’, ‘patchy consolidation’, ‘diffuse (more than 5 cm) ground glass opacity’, or ‘diffuse consolidation’. The late (after 6 months) morphology was characterized between ‘scar-like’, ‘mass-like’ and ‘modified conventional’. Additionally the severity of radiologic change was scored as ‘pronounced’ (more than expected), ‘expected’, ‘mild’ (less than expected) and none. These outcomes were compared to 54 patients treated with SABR by fixed-beam delivery, who we previously assessed using the same scoring system.

Results
Baseline characteristics of the arc and fixed-beam cohorts were well matched and respective median follow-ups were no different, 31.7 vs. 28.4 months (p=0.20). Patients treated by arc delivery trended towards being more likely to have any radiologic change (p=0.06). This was strongly time-dependent (p<0.001) and more pronounced early, as by two years radiologic changes were almost universally present irrespective of delivery technique. Figure 1 shows the morphology of these changes with time. Acute changes were not technique dependent (p=0.23). After six months, arc delivery resulted in a modified-conventional morphology throughout follow-up, while fixed-beam delivery resulted in an increasing probability of scar-like or mass-like morphologies. The predicted probabilities of a modified-conventional pattern following SABR by arc and fixed-beam delivery were 96.3% vs. 68.9% (p<0.001) respectively. Following arc delivery, radiologic changes were more likely to be scored as pronounced or expected (p=0.009) than mild or none, a finding that became more evident with longer follow-up (p=0.014). The predicted probability of pronounced or expected changes two years following arc or fixed-beam delivery was 83.1% and 26.2%, respectively. Figure 1

Conclusion
Patterns of radiologic change more than six months post-SABR are influenced by delivery technique. Diagnostic algorithms used to differentiate suspected local recurrence and benign change should therefore consider the delivery technique used.

Background
SABR has become a standard treatment option for medically inoperable, peripherally located early-stage NSCLC. However, using SABR for centrally located lesions remains challenging because of the potential for severe side effects. Here we sought to validate our previous experience with SABR (50 Gy in 4 fractions) for central lesions, including the dose-volume constraints, and explore a new regimen of 70 Gy in 10 fractions for cases in which dose-volume constraints cannot be met with the previous regimen.

Methods
We used 4D-based, volumetric image-guided SABR to treat 101 patients with biopsy-proven and PET/CT-staged centrally located (within 2 cm of bronchial tree, trachea, major vessels, esophagus, heart, pericardium, brachial plexus or vertebral body) T1-2N0M0 tumors (n=82) or isolated lung-parenchyma recurrent lesions (n=19). The treatment period spanned February 2005 through May 2011; follow-up visits (every 3 months for 2 years and every 6 months for the next 3 years) included chest CT or PET/CT. Endpoints were toxicity (CTCAE v3.0), survival, local control, regional control, and distant metastasis.

Results
At a median follow-up time of 30.3 months for all patients (40.5 months for those alive), median overall survival time was 56.5 months and 5-year overall survival rate was 49.0%. Three-year actuarial local, regional, and distant control rates were 96.5%, 87.2% and 77.3%. The most common toxicities were chest-wall pain (18% grade 1 and 13% grade 2) and radiation pneumonitis (10.9% grade 2 and 1.9% grade 3). No patient experienced grade 4 toxicity and one patient with tumor invading bronchial tree who received 70 Gy in 10 fractions died from hemoptysis 13 months after SABR. The distance between tumor and chest was associated with chest wall pain (≤1 cm 45% vs >1 cm 17%, p=0.002). Univariate and multivariate analyses showed that for the 82 patients receiving 50 Gy in 4 fractions, mean total lung dose (MLD) >5 Gy or ipsilateral lung V~20~ (iV~20~) >16% were independent predictors of radiation pneumonitis; 3 of 9 patients in that group with D~max~ to brachial plexus >35 Gy experienced brachial neuropathy versus none of the 73 patients with brachial D~max~ ≤ 35 Gy (p=0.001).

Conclusion
SABR for centrally located lesions produces clinical outcomes similar to those for peripheral lesions when normal tissue constraints are respected. For 50 Gy in 4 fractions, we recommend MLD ≤5 Gy, lung iV~20~ ≤16%; bronchial tree D~max~ ≤ 38 Gy, V~35~ ≤1 cm[3]; major vessel D~max~≤ 56 Gy, V~40~≤1 cm[3]; esophageal D~max~ ≤35 Gy, V~30~≤1 cm[3 ]; brachial plexus D~max~ ≤35 Gy, V~30~≤0.2 cm[3] and spinal cord D~max~ <25 Gy. Giving 70 Gy in 10 fractions is another option for challenging cases but can produce severe toxicity if significant amounts of critical structures are exposed to ≥70 Gy. Proper selection of cases (based on tumor location and normal tissue constraints) and SABR regimens and volumetric image-guided delivery are all crucial to avoid overdosing critical structures. Typically, a minimum 5-10 mm distance between critical structures and gross tumor is required to meet dose-volume constraints.

Abstract not provided

Background
NSCLC may utilize PD-L1 overexpression to escape immune surveillance. This mechanism has been suggested by recent clinical studies showing that NSCLC can respond to PD-L1/PD-1 blockade. MPDL3280A, a human monoclonal antibody containing an engineered Fc-domain designed to optimize efficacy and safety, aims to restore tumor-specific T-cell immunity by blocking PD-L1 from binding to its receptors, PD-1 and B7.1.

Methods
Patients received MPDL3280A IV q3w for up to 1 year in a Phase I dose escalation/expansion study. Objective response rate (ORR) was assessed by RECIST v1.1 and included unconfirmed/confirmed responses. EGFR and KRAS status was initially assessed locally by investigators. Archival tissue was analyzed centrally for PD-L1 expression by IHC.

Results
As of Feb 1, 2013, 52 NSCLC patients were evaluable for safety and treated at doses of 0.03-20 mg/kg. The median age of patients was 61 years (range, 24-83). 17 (33%) of patients were ECOG PS 0 and 35 (67%) of patients were ECOG PS 1. Prior treatments included surgery (89%), radiotherapy (54%) and systemic therapy (98%). 15% of patients received 1 prior regimen, 21% received 2 and 62% received ≥3. Additionally, 14%, 62% and 25% of patients were EGFR-mutation positive, EGFR WT and EGFR status unknown/undetermined, respectively, and 12%, 40% and 48% of patients were KRAS-mutation positive, KRAS WT and KRAS status unknown/undetermined, respectively. Patients received treatment with MPDL3280A for a median duration of 106 days (range 1-450). Treatment-related Gr3/4 AEs occurred in 12% of patients, including fatigue (4%) and hypoxia (4%). 1 patient experienced a Gr3/4 immune-related AE (Gr3 hyperglycemia). No Gr3-5 pneumonitis or diarrhea was reported. 41 NSCLC patients first dosed at 1-20 mg/kg prior to Aug 1, 2012, were evaluable for efficacy. An ORR of 22% (9/41) was observed in patients (squamous [n=9]/nonsquamous [n=31]) with a duration of response range of 1+ to 214+ days. Additional patients had nonconventional responses after apparent radiographic progression but were considered to have progressive disease in this analysis. All responses were ongoing or improving at data cutoff. The 24-week PFS was 46%. ORR by patient characteristics was also examined. The ORR for patients with ≤2 prior therapies was 23% (4/17) and 23% (5/22) for patients with >2 prior therapies. Additionally, the response for former/current smokers was 23% (8/35) versus 17% (1/6) for never smokers. Between EGFR-mutation positive and EGFR WT patients, the ORRs also did not differ (25% [1/4] and 19% [5/26], respectively). In contrast, PD-L1 status was associated with ORR response as patients with PD-L1–positive tumors had an ORR of 80% (4/5) and patients with PD-L1–negative tumors had an ORR of 14% (4/28). Updated data, including responses by KRAS status, will be presented.

Conclusion
Treatment with MPDL3280A was generally well tolerated, with no cases of Gr3-5 pneumonitis. Rapid and durable responses were observed, including in an EGFR-mutation positive patient. Responses to MPDL3280A did not appear influenced by the number of prior treatment regimens but did appear to be associated with PD-L1 tumor status. Additional studies have been initiated in NSCLC.

Background
Currently approved cytotoxic chemotherapies for previously treated patients with NSCLC demonstrate few objective responses, which are generally of short duration, with limited impact on progression-free survival and overall survival. Programmed death-1 (PD-1) is an inhibitory T-cell co-receptor whose activation by interaction with its ligands, PD-L1 or PD-L2, can lead to suppression of antitumor immunity. Preclinical and clinical data indicate that this pathway is important in NSCLC.MK-3475 is a humanized monoclonal IgG4 antibody against PD-1.

Methods
MK-3475 was administered at 10 mg/kg every three weeks to patients with NSCLC previously treated with two systemic regimens. At least one measurable tumor lesion, ECOG performance status of zero or one, and adequate laboratory function were required for eligibility. A new tumor biopsy no earlier than 60 days before the first dose of MK-3475 was required for study entry. Imaging assessments per investigators were performed every nine weeks until confirmed disease progression utilizing the immune-related response criteria (irRC). Independent central review of images was assessed with RECIST v1.1. PD-L1 expression on the pretreatment tumor sample was determined by immunohistochemistry. A cut-point associated with the Youden Index of the receiver-operating characteristic curve for PD-L1 staining was identified.

Results
Between April 2012 and September 2012, thirty-eight patients were enrolled. Median age was 63 years (range, 34-85 years), with 42% men and 42% with an ECOG performance status of zero. Previously treated, stable brain metastases were allowed and were present in 10%. Seven patients had an EGFR mutation, eight patients had a KRAS mutation, and one patient had an ALK gene rearrangement in their tumor. Fifty percent of patients experienced drug-related adverse events; the most common were fatigue, rash, and pruritus (16% each). The incidence of diarrhea was 13% (only grade 1 or 2 reported). One case of a drug-related grade 3-4 adverse event (grade 3 pulmonary edema: 3%) was seen. There were no drug-related fatalities. Using investigator-assessed irRC, the objective response rate (ORR; confirmed and unconfirmed) was 24%, including squamous and nonsquamous subtypes. Similar results were obtained using RECIST v1.1, yielding an ORR (confirmed and unconfirmed) of 21%. Most responses by irRC were observed by the time of first planned assessment at Week 9. The median duration of response by irRC has not been reached, with a median duration of follow-up of 9 months (minimum, 6 months). As of June 2013, seven of the nine responding patients by irRC continue on therapy. Pretreatment tumor PD-L1 expression was a statistically significant predictor of response. In patients with evaluable tumor PD-L1 expression, all confirmed responses by RECIST v1.1 (and irRC) occurred in patients with tumors strongly positive for PD-L1.

Conclusion
MK-3475 is generally well tolerated in previously treated patients with advanced NSCLC and provides durable objective responses. An additional cohort of patients whose tumors express PD-L1 is enrolling; preliminary safety and efficacy data, including PFS and OS, will be reported further at the World Conference on Lung Cancer 2013.

Background
Blockade of programmed death-1 (PD-1), a co-inhibitory receptor expressed by activated T cells, can overcome immune resistance and mediate tumor regression (Topalian S, et al. New Engl J Med. 2012;366:2443-54). We present long-term safety and efficacy outcomes from a phase 1 study of nivolumab, a fully human IgG4 PD-1 receptor blocking monoclonal antibody, in patients with advanced NSCLC.

Methods
NSCLC patients enrolled between 2008–2012 received nivolumab 1, 3, and 10 mg/kg IV Q2W on either dose escalation or subsequent expansion cohorts. Tumors were assessed (RECIST 1.0) after each 4-dose cycle. Protocol was amended (Jan. 23, 2012) to explore nivolumab’s potential to deliver prolonged overall survival (OS) for the initial and expansion cohorts and the overall population.

Results
129 pretreated NSCLC patients (non-squamous [n=74], squamous [n=54], unknown histology [n=1]) were treated as of March 2013. Responses (CR/PR) occurred in 22 patients (17%) and were durable (estimated median response duration, 74.0 weeks [6.1+, 133.9+]), and ongoing in 55% (12/22) of patients. The highest objective response rate (ORR) was at 3 mg/kg (24%) across NSCLC histologies. Responses were rapid; 50% of patients (11/22) demonstrated response at first tumor assessment (8 weeks). Among 12 responders who discontinued therapy for reasons other than disease progression, 3 responded for ≥24 weeks post therapy discontinuation, and all 3 had not progressed at the time of this analysis. An additional 6 of the 122 patients (5%) demonstrated unconventional “immune-related” responses (based on target lesions), but were not included among responders. Survival benefit was demonstrated by 1-year and 2-year landmark OS rates (42% and 14%; Table). Median OS was 9.6 months across doses and 14.9 months at 3 mg/kg across histologies. Median OS across doses was similar for squamous/non-squamous patients. Any grade drug-related select adverse events (AEs) occurred in 41% (53/129) of patients (grade 3/4 select AEs, 5% [6/129]); most common being skin (16%), gastrointestinal (12%), and pulmonary (7%). Any grade drug-related pneumonitis occurred in 6% (8/129) of patients (grade 3/4 pneumonitis, 2% [3/129]), resulting in 2 deaths early in the trial, leading to increased emphasis on management algorithms. Characteristics and management of nivolumab-related pneumonitis will be summarized.

Cohort	Dose, mg/kg	ORR[a] no. of patients/total no. of patients (%) [95% CI]	Estimated median response duration, wk (range)	Median OS,[b] mo (95% CI)	OS rate, % (95% CI); patients at risk, n
					1 y		2 y
NSCLC (n=129)[c]	All doses	22/129 (17.1) [11.0, 24.7]	74.0 (6.1+, 133.9+)	9.6 (7.8, 12.4)	42 (33, 51); 43		14 (4, 24); 5
	1	1/33 (3.0) [0.1, 15.8]	63.9 (63.9, 63.9)	9.2 (5.6, 11.1)
	3	9/37 (24.3) [11.8, 41.2]	NR (16.1+, 133.9+)	14.9 (9.5, NE)
	10	12/59 (20.3) [11.0, 32.8]	83.1 (6.1+, 117.1+)	9.2 (5.2, 12.4)
Initial cohort (n=19)	All doses			9.6 (4.5, 19.8)	42 (20, 64); 8		26 (7, 46); 5
Expansion cohort (n=110)	All doses			9.9 (7.8, 12.5)	42 (32, 51); 35
Squamous (n=54)	All doses	9/54 (16.7) [7.9, 29.3]	NR (16.1, 133.9+)	9.2 (7.3, 12.5)	39 (25, 53); 16
	1	0/15	0	8.0 (2.6, 13.3)
	3	4/18 (22.2) [6.4, 47.6]	NR (16.1, 133.9+)	9.5 (6.7, NE)
	10	5/21 (23.8) [8.2, 47.2]	83.1 (16.1, 117+)	10.5 (7.8, 12.5)
Non-squamous (n=74)	All doses	13/74 (17.6) [9.7, 28.2]	63.9 (6.1+, 74.0+)	10.1 (7.2, 13.7)	43 (31, 54); 26
	1	1/18 (5.6) [0.1, 27.3]	63.9 (63.9, 63.9)	9.9 (5.6, 22.7)
	3	5/19 (26.3) [9.1, 51.2]	74.0 (24.3, 74.0+)	18.2 (10.3, 18.2)
	10	7/37 (18.9) [8.0, 35.2]	NR (6.1+, 65.7+)	7.4 (4.6, 12.4)
[a]ORR = ([CR + PR] ÷ n) × 100.[b]OS estimates after 1 year reflect censoring and shorter follow-up for patients enrolling later in the study.[c]Non-squamous (n=74), squamous (n=54), and unknown histology (n=1) NE = not estimable; NR = not reached.

Conclusion
In advanced NSCLC patients, nivolumab produced durable responses and survival benefit (1-year OS rate, 42%), with a long-term safety profile acceptable for the outpatient setting, supporting ongoing development in phase 3 trials with survival endpoints. Additional follow-up on patient survival will be presented at the time of the meeting.

Background
MUC1, a tumor antigen, has been considered to be a promising target antigen for cancer immunotherapy because it possesses a potent immunogenicity. It is processed and presented by antigen-presenting cells in a MHC-unrestricted pattern. Dendritic cell-based vaccine immunotherapy can elicit antigen-specific cytotoxic T lymphocytes in tumor-bearing hosts, and activated cytotoxic T lymphocytes are expected to attack cancer cells. In this study, we evaluated the efficacy of MUC1-targeted dendritic cell-based vaccine immunotherapy in patients with standard treatments-refractory advanced non-small-cell lung cancer (NSCLC).

Methods
The eligibility criteria of this immunotherapy were as follows: histologic or cytologic evidence of NSCLC that express MUC1 protein abundantly; an Eastern Cooperative Oncology Group performance status of 0-2; advanced stage of diseases refractory to any standard cancer treatments. The dendritic cells were prepared from peripheral blood mononuclear cells with cytokines interleukin-4 and granulocyte macrophage colony stimulating factor, pulsed with MUC1 peptides, and subsequently administered to patients by subcutaneous injection. The vaccinations were repeated bi-weekly, and assessable patients were received at least 6 vaccinations. Tumor response was assessed according to the Response Evaluation Criteria in Solid Tumors. Adverse events were graded according to National Cancer Institute Common Toxicity Criteria.

Results
From June 2005 to December 2012, 36 patients were treated with dendritic cell-based vaccines, and 25 patients (69.4%) with median age of 61 years (range, 49-84 years) were assessable for tumor responses. The cohort consisted of 14 males and 11 females, and 22 patients had adenocarcinomas; 2 patients with squamous cell carcinomas and 1 patient with pleomorphic carcinoma. Among these patients, neither complete response nor partial response was obtained. Fourteen patients had progressive disease as the best response, and 10 patients had stable disease, yielding overall disease control rate of 40.0% (95%CI=20.8-59.2). Median survival time after the vaccines was 10.0 months, and 1-year survival rate was 32.3%. Adverse events related to the vaccines were less frequent. Immunological responses could be monitored in five patients, showing that MUC1-specific cytotoxic responses of effector immune cells were achieved in all of those patients, and the population of regulatory T lymphocytes in peripheral blood cells was decreased after the vaccines.

Conclusion
MUC1-targeted dendritic cell-based vaccine immunotherapy is feasible, and has a potential to control the diseases in patients with refractory NSCLC.

Abstract not provided

Background
Effective therapeutic strategies for mutant KRAS and other biomarkers of resistance in refractory NSCLC remain an unmet medical need, while a personalized medicine approach is increasingly adopted in NSCLC guided by tumor molecular profiling. The BATTLE-2 clinical study is using EGFR, PI3K/AKT and MEK inhibitors and is designed to identify biomarkers for optimal patient selection for these therapies (ClinicalTrials.gov NCT01248247).

Methods
This is a four-arm, open-label, multi-center, biopsy-driven, adaptive randomization, phase II clinical trial in NSCLC pts that failed at least 1 prior line of therapy. Patients are adaptively randomized to 4 arms: erlotinib, erlotinib plus the AKT inhibitor MK-2206, MK-2206 plus the MEK inhibitor selumetinib, and sorafenib. The primary objective is 8-week disease control rate (DCR). The trial is conducted in 2 stages. In Stage 1, 200 evaluable pts are adaptively randomized (AR) based on observed 8-week DCR and KRAS mutation status while predictive biomarkers are being developed by means of gene expression profiling, targeted next generation sequencing and protein expression. EGFR sensitizing mutations and EML4/ALK translocation in pts that are erlotinib and crizotinib naïve are exclusion criteria, while erlotinib resistant patients are excluded from erlotinib monotherapy. In Stage 2, the AR model is refined to include the most predictive biomarkers tested in Stage 1, with subsequent Stage 2 AR based on the new algorithm, to a total of 400 evaluable pts. Selection of Stage 2 single and/or composite markers follows a rigorous, internally and externally reviewed statistical analysis that follows a training, testing methodology with validation in stage 2 of the trial. All Stage 1 and 2 randomization biomarker assays are CLIA-certified.

Results
286 pts have been enrolled, 236 biopsies performed,172 pts randomized, and 167 pts treated. 144 pts are evaluable for the 8-week DCR endpoint. Within the randomized pts group KRAS mutation rate is 22.8%, and EGFR mutation rate 14.8%, while 36.3% patients have been previously treated with erlotinib. Treatment is well tolerated with no unanticipated toxicity.

Conclusion
Accrual updates, demographics, and further details will be presented at the meeting. (Supported by NCI R01CA155196-01A1)

Background
The potential of personalized medicine for improvement of lung cancer patient outcome has been paradigmatically shown by the treatment of advanced EGFR mutation- and ALK translocation positive NSCLC patients with the respective tyrosine kinase inhibitors. Furthermore numerous targeted drugs for molecular defined subgroups of NSCLC (e.g. ROS1- rearrangements) are in clinical development with the potential to improve outcome. Therefore one of the major challenges today is the implementation of comprehensive high-quality real time molecular diagnostics and personalized therapy for all NSCLC patients regardless of where they are treated.

Methods
To increase the availability of molecular testing and subsequently personalized treatment options for NSCLC patients in the catchment area of our cancer center, we established the Network Genomic Medicine (NGM) in January 2010. NGM is a collaborative network currently encompassing more than 40 different health care providers representing the full spectrum of lung cancer care in Germany including university hospitals, large non-university lung clinics and office based oncologists. NGM is based at the Center for Integrated Oncology (CIO), i.e. the joint comprehensive cancer center of the University Hospitals of Cologne and Bonn. At the NGM - headquarter genetic and clinical data are analysed and patients without approved targeted treatment options are screened for recruitment into NGM-linked personalized trials offered by the Lung Cancer Group Cologne (LCGC). Before the introduction of routine Next Generation Sequencing (NGS) within NGM in 06/2013 we screened lung adenocarcinomas (AD) via single gene assays for mutations in EGFR, KRAS, BRAF and PIK3CA, for amplifications in HER2 and translocations in ALK, ROS1 and RET. Squamous cell lung cancer (SCC) patients were screened for amplifications in FGFR1 and mutations in DDR2.

Results
We screened 5,145 lung cancer patients from January 2010 till April 2013. Genomic testing was feasible in 3,863 tumor samples (75%). 63% of the patients were male and 65% of samples were AD. In AD the following frequencies of genetic lesions were detected: EGFR 13.8% (288/2078); ALK 3.3% (54/1618); KRAS 33.8% (831/2457); BRAF 3.5% (76/2123); PIK3CA 3.1% (70/2190); HER2 amplified 3.6% (62/1717); RET 4.7% (4/85) and ROS1 5.1% (7/135). In SCC we found a frequency of 21% (279/1333) for FGFR1 amplification and 2.1% (11/505) for DDR2 mutations. Further we saw 18 KRAS/PIK3CA, 5 EGFR/PIK3CA, 5 BRAF/PIK3CA double mutant samples and 3 samples where a FGFR amplification was co-occurring with a DDR2 mutation. Overall 40% of NSCLC patients harboured a potentially targetable molecular alteration. In addition we could allocate more than 40 patients to early personalized clinical trials via the close collaboration of the partners within NGM and LCGC. *The frequencies of RET and ROS1 are biased, because of a preselection of pan negative patients.

Conclusion
NGM is one of the largest prospective molecular screening efforts for NSCLC worldwide, with currently more than 3000 samples analysed per year. Our experiences so far underline that central comprehensive high-quality real time molecular diagnostics is feasible in a large health care provider network and allows implementation of personalized medicine in routine clinical care of lung cancer patients.

Background
There are few new effective therapeutic options for patients with advanced, lung SCCA; overall survival for metastatic disease being less than one year. The Cancer Genome Atlas (TCGA) project and similar studies have detected a significant number of somatic gene mutations/amplifications in patients with this disease, some of which are targetable by investigational agents. However, the frequency of these changes is low (5-20%) in these patients, making recruitment and treatment very challenging in the traditional single-agent trial setting. Our approach is to use a common platform (Next Generation DNA Sequencing) to enable a single “umbrella screening protocol” to efficiently find patients with varied, uncommon molecular changes.

Methods
Figure 1 This is a prospective, multi-substudy randomized Phase II/III Master Registration Protocol in which patients with advanced stage Lung SCCA (2[nd] line therapy)are randomized to biomarker-driven targeted therapy (TT) or standard of care (SOC) as shown in the schema after undergoing genomic screening. Genomic screening of a large patient resource provided by sites participating in the NCI North American Intergroup will identify molecular targets/biomarkers with an analytically validated diagnostic assay and a new drug match, leading to appropriate drug treatment-arm assignment. Archival FFPE tumor and/or core needle biopsies will be screened by a broad analytically validated next generation sequencing (NGS) platform centrally to establish eligibility within 10-14 days. This platform will be supplemented by individual immune-histochemical (IHC) protein assays performed in a CLIA setting as necessitated by the specific experimental agent used. Patients will be screened with homogeneous eligibility criteria. The overall trial objective is to establish a mechanism to genomically screen large but homogeneous cancer populations and subsequently assign and accrue simultaneously to multiple substudies comparing new TT to SOC therapy based on the identified therapeutic biomarker-drug combination. Each sub-study will function autonomously and will open and close independently of the other sub-studies. Drug combinations in the experimental arm will be allowed in appropriate settings and where appropriate the control arm may consist of FDA approved targeted therapy such as erlotinib. Each sub-study is independently powered for OS with an interim analysis for PFS to determine the “go-no go” decision to proceed from Phase II into Phase III. Each agent, along with the paired biomarker, that is successful at the interim analysis based on PFS will advance to a Phase III randomized registration trial (on behalf of the Master Protocol Steering Committee).

Results
NA

Conclusion
NA

Abstract not provided

Background
KRAS is the most frequently mutated oncogene in NSCLC and represents an unmet need for targeted therapy. Trametinib plus pemetrexed enhances growth inhibition and apoptosis of NSCLC cell lines with and without RAS/RAF mutations in vitro when compared with either agent alone.

Methods
This 2-part, multi-arm, open-label phase 1/1B study evaluated the safety and efficacy of trametinib plus chemotherapy (NCT01192165). Part 1 determined the recommended phase 2 dose (RP2D) for trametinib (1.5 mg daily) and pemetrexed (500 mg/m[2] every 3 weeks) in patients with advanced solid tumors. In part 2, patients with NSCLC were stratified as KRAS WT or KRAS-mutant and treated at the RP2D. Primary study objectives were safety and tolerability; secondary objectives were efficacy and pharmacokinetics (PK). Next-generation sequencing was used to perform exploratory mutational profiling on available archival tissue from 21 patients (50%). Plasma from 38 patients (90%) was analyzed both for tumor-derived mutations in cell-free DNA (eg, KRAS, EGFR) using BEAMing technology as well as cytokine and angiogenic factors using a Searchlight multiplex assay.

Results
A total of 42 patients with NSCLC (19 KRAS WT [79% ≥ 2 prior therapies; 74% prior pemetrexed; 16% squamous] and 23 KRAS-mutant [57% ≥ 2 prior therapies; 43% prior pemetrexed; 4% squamous]) were enrolled and treated at the RP2D until disease progression or unacceptable toxicity. Safety and PK data were previously reported (ASCO 2013). Response rate was 17% and disease control rate was 69% for the whole population of NSCLC. Of note, we observed disease control in 75% of patients previously treated with pemetrexed (including 4 partial responses [PRs]) and in 2 patients out of 4 with squamous histology (including one PR). Progression-free survival (PFS) was 5.1 months for all patients with NSCLC. Detailed efficacy results according to mutation status are shown in Table 1. Among KRAS WT, activity was seen in cancers with EGFR mutations or ALK rearrangement. Final biomarker analyses, including assessment of their potential correlation with therapeutic response or resistance, are ongoing and will be reported upon completion. Figure 1

Conclusion
MEK inhibition with trametinib + pemetrexed demonstrated activity in both KRAS-mutant and WT NSCLC; efficacy data are encouraging and warrant further study. There was no significant difference in activity or efficacy across KRAS mutation subtypes. Interestingly, activity with this combination was broad and was seen in patients with squamous histology, patients with prior pemetrexed treatment, and those with EGFR mutation or ALK translocation.

Background
KRAS is the most frequently mutated oncogene in NSCLC and represents an unmet need for targeted therapy. Trametinib enhances docetaxel-induced growth inhibition and apoptosis of NSCLC cell lines. Cell lines with the KRAS G12C point mutation, the most common KRAS mutation subtype (≈50% of KRAS-mutant NSCLC or ≈10% of all NSCLC), are more responsive to apoptosis induced by this combination.

Methods
This 2-part, multi-arm, open-label phase 1/1B study evaluated the safety and efficacy of trametinib plus chemotherapy (NCT01192165). Part 1 determined the recommended phase 2 dose (RP2D) for trametinib (2.0 mg daily) and docetaxel (75 mg/m[2] every 3 weeks) in the presence of growth factors in patients with advanced solid tumors. In part 2, patients with NSCLC were stratified as KRAS WT or KRAS-mutant and treated at the RP2D. Primary study objectives were safety and tolerability; secondary objectives were efficacy and pharmacokinetics (PK). Next-generation sequencing was used to perform exploratory mutational profiling on available archival tissue from 17 patients (36%). Plasma from 42 patients (89%) was analyzed both for tumor-derived mutations in cell-free DNA (eg, KRAS, EGFR) using BEAMing technology as well as cytokine and angiogenic factors using a Searchlight multiplex assay.

Results
A total of 47 patients with NSCLC (22 KRAS WT [64% ≥2 prior therapies; 27% squamous] and 25 KRAS-mutant [40% ≥2 prior therapies; 0% squamous]) were enrolled and treated at the RP2D until disease progression or unacceptable toxicity. Safety and PK data were previously reported (ASCO 2013). Progression-free survival (PFS) was 4.2 months for all patients; efficacy results according to mutation status are shown in Table 1. Among KRAS-mutant patients, activity and efficacy were better in G12C compared with non-G12C subtypes. Among KRAS WT, activity was seen in cancers with EGFR mutations; clinical benefit was noted in 2 patients with ALK translocation (disease control 25 weeks and 60+ weeks). Final biomarker analyses, including assessment of their potential correlation with therapeutic response or resistance, are ongoing and will be reported upon completion. Figure 1

Conclusion
MEK inhibition with trametinib + docetaxel (+ growth factors) demonstrated activity in both KRAS-mutant and WT NSCLC; efficacy data are encouraging and warrant further study. Cancers carrying the KRAS G12C point mutation may have improved activity and efficacy compared with non-G12C subtypes, consistent with preclinical observations. Additionally, clinical benefit with this combination was broad and was seen in patients with squamous histology and those with EGFR mutation or ALK translocation.

Background
Phase II data from patients with KRAS mutation-positive NSCLC, selumetinib (AZD6244, ARRY-142886) plus docetaxel showed promising efficacy versus placebo plus docetaxel alone (Jänne et al. Lancet Oncol 2013;14:38–47). Median OS was 9.4 months (95% CI 6.8–13.6) in the selumetinib group and 5.2 months (95% CI 3.8–non-calculable) in the placebo group (HR for death 0∙80, 80% CI 0.56–1.14; one-sided p=0.21). Median PFS was 5.3 months (95% CI 4.6–6.4) and 2.1 months (95% CI 1.4–3.7), respectively (HR for progression 0∙58, 80% CI 0.42–0.79; one-sided p=0.014). 37% of patients in the selumetinib group and 0% in the placebo group had an objective response (two-sided p<0.0001). The KRAS mutation codon subtype might impact on prognosis and/or response to therapy. The BATTLE trial suggested that G12V or C KRAS mutations confer relatively poorer outcome within the KRAS mutant NSCLC sub-type (Ihle et al. J Natl Cancer Inst 2012;104:228–39). In cell lines carrying these codons, Akt phosphorylation but not ERK phosphorylation was low compared with other codons, suggesting these codons might confer greater dependence upon MEK/ERK signaling. We sought to understand if any codons or combinations of codons selected for striking treatment effects either between or within treatment groups in the Phase II study.

Methods
Post-hoc analysis explored the hypotheses that patients whose tumours carried G12C or G12V KRAS mutations would have a worse prognosis and that these patients would have a better outcome with the addition of selumetinib. Clinical benefit was measured by PFS, OS and ORR.

Results
G12V or G12C mutations were present in 57% of patients and whilst not reaching statistical significance, trends for PFS, OS and ORR support the hypothesis (see table, PFS). Patients with G12V mutations responded better to selumetinib plus docetaxel than other patients as measured by change in tumour size at week 6 (G12V=-62%, G12C=-8%, G12D=+3%, reduction across all codons=-18%; two sided p=0.007). It is therefore possible that trends supporting the primary hypothesis were driven by effects in the small number of G12V codons (n=9). Table. Summary of analysis of progression-free survival (PFS): MITT by mutation subgroup

Subgroup	Selumetinib + docetaxel, n (number of PFS events)	Docetaxel, n (number of PFS events)	Selumetinib + docetaxel vs docetaxel, PFS HR (80% CI)
G12C or G12V	24 (18)	23 (21)	0.48 (0.31–0.74)
Other	19 (17)	17 (15)	0.72 (0.44–1.16)
Overall	43 (35)	40 (36)	0.58 (0.42–0.79)

Conclusion
Any impacts of codon sub-type on the treatment effect in this trial were not sufficiently significant to be detected in this small Phase II trial of 87 patients, but the trends observed in this retrospective subgroup analysis warrant monitoring of the impact of specific codons or groups of codons in future clinical trials.

Abstract not provided

Background
Accurate staging of the mediastinum is critical in therapeutic decision making in NSCLC. PET/CT has emerged as an important modality for staging of treatment-naïve NSCLC, but like endobronchial ultrasound and conventional mediastinoscopy typically is inaccurate following neoadjuvant therapy. We sought to determine the accuracy of TEMLA in staging NSCLC after induction therapy.

Methods
A retrospective chart review looking at clinical stage assessed by PET -CT and TEMLA, pathologic stage, lymph node yield and clinical characteristics was performed. Accuracy of staging by TEMLA and PET-CT was compared.

Results
71 of 100 consecutive patients that underwent TEMLA had it for restaging after neoadjuvant therapy; 65 of these patients were also restaged by PET-CT. Clinical characteristics of these 65 patients are presented (Table 1). TEMLA was completed successfully on 63 (96.9%) patients and was associated with permanent recurrent laryngeal nerve injury in 2 (3%) patients. On average, 17 lymph nodes were obtained per TEMLA. Concomitant anatomic resections were completed in 58 (89.2%) of patients. 12 and 3 (18.5% and 4.6%) patients were classified as having N2 and N3 disease on final pathology. Compared to PET-CT, TEMLA more accurately classified these patients (95.4% vs. 80.0%; P<0.05). The sensitivity, specificity, positive predictive value and negative predictive value of PET/CT and TEMLA for detection of N2 disease are 50.0%, 86.8%, 46.1%, 88.5% and 75%, 100%, 100%, 94.6% respectively (Table 2). Of the 3 patients inaccurately classified by TEMLA, only 1 patient had N2 disease in TEMLA-accessible nodes. Figure 1

Table 2: Patient numbers according to nodal status.

	Path +	Path -	PET +	PET -
TEMLA +	9	0	4	5
TEMLA -	3	53	9	47
PET +	6	7
PET -	6	46

Conclusion
TEMLA is superior to PET/CT for restaging of the mediastinum after induction therapy. Since TEMLA showed little added morbidity despite central tumor and treatment effects, consideration should be given for its widespread adoption for mediastinal re-staging of NSCLC after neoadjuvant therapy.

Background
The current UICC/WHO nodal classification system is based on the location of metastatic lymph nodes, while some studies have revealed that the number or ratio of metastatic lymph nodes may work as more effective prognostic indicators. The Japan Lung Cancer Society proposed a new tumor site-based classification for mediastinal nodal metastases according to the tumor-bearing lobe. This study aimed to compare the prognostic power of location-based and number-based classification systems and elucidate the optimal classification.

Methods
Of 511 patients with non-small cell lung cancer (NSCLC) who underwent lung lobectomy and complete hilar and mediastinal lymph node dissection with curative intent at our institute between 1998 and 2009, 119 with confirmed lymph node metastases were retrospectively analyzed. Ten classifications were compared using a log-rank test. Four classifications were location-based: the current system, the tumor site-based classification, the classification based on presence or absence of clinical N2 disease, and the classification based on presence or absence of non-skip N2 disease. The other 6 classifications were number-based: the classifications based on the number or ratio of metastatic lymph nodes, the classifications based on that of metastatic stations, and the classifications based on that of metastatic mediastinal lymph nodes.

Results
Compared with the current system [hazard ratio (HR), 1.4; p = 0.29], the tumor site-based classification (HR, 2.8; p = 3.0E-4), the classification based on the number of metastatic lymph nodes (HR, 2.8; p = 1.7E-4), and the classification based on the number of metastatic mediastinal lymph nodes (HR, 2.3; p = 3.3E-3) were considered to be stronger predictors of overall survival. Similar results were obtained in terms of disease-free survival (current system: HR, 1.6; p = 0.047; tumor site-based classification: HR, 2.7; p = 2.3E-5; number of metastatic lymph nodes, HR, 2.3; p = 4.0E-4; number of metastatic mediastinal lymph nodes: HR, 2.4; p = 1.4E-4). A combination of the tumor site-based classification with the classification based on the number of metastatic lymph nodes (p = 9.0E-4) or the classification based on the number of metastatic mediastinal lymph nodes (p = 9.5E-4) further increased predictive efficiency.

Conclusion
The tumor site-based classification as well as the classifications based on the number of metastatic lymph nodes and the number of metastatic mediastinal lymph nodes was more predictive of surgical outcomes compared with the current nodal system. The results need to be further validated in a new set of patients. Figure 1

Background
We sought to establish an acceptable lobe-specific mediastinal lymphadenectomy protocol for non-small cell lung cancer (NSCLC) presenting as solitary pulmonary nodules (SPN) .

Methods
We retrospectively analyzed 415 patients pathologically diagnosed as NSCLC undergone lobectomy, bilobectomy or pneumonectomy with systematic lymphadenectomy from March 2004 to June 2011 in our hospital. All of the patients enrolled were considered SPN preoperatively. Information about primary tumor location, lymph node metastasis, and other baseline data were collected. Stepwise logistic regressions using N1 and lobe-specific regional mediastinal lymph nodes’ conditions as covariates were used to figure out the key lymph node station that indicated non-regional mediastinal lymph nodes metastases (NRM).

Results
As for the location of the primary tumor, 121 cases were in right upper lung (RUL), 42 in right middle lung (RML), 77 in right lower lung (RLL), 107 in left upper lung (LUL), and 68 in left lower lung (LLL). Stepwise regression showed that #2(OR (odds ratio) = 28.250, 95%CI (confidence interval): 1.756-454.422, P=0.018), N1 (OR=24.000, 95%CI: 3.346-172.121, P=0.002) and N1 (OR=21.667, 95%CI: 3.266-143.736, P=0.001) was the key lymph node station for RUL, LUL and RLL, respectively. None of the covariates show statistical significant for LLL.Patients with tumors >2 cm rarely had NRM without primary regional mediastinal involvement. Figure 1 Figure. Malignant cells’ residue when the key station shows negative metastasis

Conclusion
With rigid consideration, lobe-specific lymphadenectomy is feasible in practice. This protocol could be established when the lobe-specific key nodes show negative under intraoperative frozen section, especially for those NSCLCs presented as SPN smaller than 2 cm preoperatively. Table. Protocols for lobe-specific mediastinal lymphadenectomy for SPN

	Tumor locations
	RUL	LUL	RLL	LLL	RML
Superior mediastinum^4	◎	◎	○[b]	◎	◎
Inferior mediastinum	○[a]	○[b]	◎	◎	◎
Note: ◎: Complete regional lymphadenectomy is warranted; ○: Lymph nodes dissection could be omitted considerably. [a]: when #2 shows negative in intraoperative frozen section; [b]: when #10, 11 shows negative in intraoperative frozen section; ^4: #1-4 in RUL, #4-7 in LUL

Abstract not provided

Background
The brain is one of the most frequent sites of distant metastasis in patients with lung cancer. Surgical resection of isolated brain metastases in NSCLC patients is not widely accepted and still a matter of debate. The study was aimed to evaluate the long-term results and prognosis after surgical resection of primary tumor and solitary brain metastasis in NSCLC patients.

Methods
In this retrospective study, the data of 82 patients who underwent lung resection for primary NSCLC and brain metastasectomy for solitary metastasis between 1991 and 2011 in our clinic were analyzed. There were 68 (82,9%) males and 14 (17,1%) females, median age – 59,6 years. The most common histologic type of lung cancer was adenocarcinoma (70,7%). Synchronous brain metastasis was detected in 21 (25,6%), metachronous – in 61 (74,4%) patients. The primary lung cancer was completely resected in all cases. Surgery included pneumonectomy – in 7 (8,5%), lobectomy – in 69 (85,4%) and wedge resection – in 5 (6,1%) patients. In all cases of synchronous brain metastasis, except one, we performed brain metastasectomy first followed by lung surgery in 4-6 weeks interval. Simultaneous lung resection and brain metastasectomy was performed only in one patient. Surgery in patients with metachronous brain metastasis depended on the time of detection and varied from 4 to 38 months.

Results
Postoperative complications were registered in 10 (12,2%) patients, mortality rate was 3,7% (3 patients). Overall 1, 3 and 5-year survival after brain metastasectomy was 52,0%, 29,0% and 25,6% respectively with median survival 18,6 months. The most important prognostic factors were N-status of primary lung cancer and synchronous or metachronous diagnosis of brain metastasis. Three and 5-year survival after brain metastasectomy in patients with N0 status was significantly better than in N+ patients: 56,8% and 34,8% versus 21,4% and 6,5% respectively (p<0,01). Median survival was 19,8 months in N0 group and only 12,4 months in patients with positive lymph nodes. Five-year survival in patients with metachronous brain metastases was 19,8% versus 10,0% in synchronous group (p<0,05). Eight patients are alive free of recurrence, 10 patients – with recurrence in the brain and 64 (78,0%) patients died of disease progression in the brain or other distant sites.

Conclusion
Surgery in NSCLC patients with operable solitary brain metastasis is justified especially in N0 cases and metachronous disease. Surgical resection improves long-term results and quality of life in patients with operable brain lung cancer metastasis.

Background
It is difficult to achieve a margin-negative resection (R0) for non-small cell lung cancer (NSCLC) patients with infiltration of the pulmonary artery. We report our experience with reconstruction of the pulmonary artery with regard to long-term survival.

Methods
Clinical records of 118 patients with NSCLC who underwent partial or circumferential pulmonary artery resection during a 21-year period were reviewed retrospectively. Technical outcomes and survival were analyzed.

Results
We performed 22 pulmonary artery sleeve resections, 51 reconstructions by autologous pericardial patch, 36 tangential resections, 3 left main pulmonary artery (PA) angioplasties during pneumonectomy without cardiopulmonary bypass, and 6 by only preserving the 1[st] branch of pulmonary arterial trunk. In 41 patients, bronchial sleeve resection was associated; in 7 cases, superior vena cava reconstruction was also required. Thirty-one patients received induction therapy. Thirteen patients had stage IB disease, 41 stage II, 53 IIIA, and 11 IIIB. Ninety-three patients had squamous cell carcinoma, 22 adenocarcinoma, 2 mixed and 1 large cell carcinoma. Negative bronchial and vascular margins were achieved in all. 5 positive bronchial margins were due to limited lung function. The analysis of 118 cases yielded follow-up data in 94 cases. The mean follow-up was 70 months (range 1-156 months). There was no in hospital death, and the overall 5-year survival was 50.2%. Five-year survival for stages Iand II, versus IIIwere 63.9% versus 37.0% (p=0.0059). Multivariate analysis yielded non-squamous cell carcinoma, stage IIIand patch pulmonary arterioplasty as negative prognosis factors. PA reconstruction associated with bronchial sleeve resection was the positive prognostic factor.´

Conclusion
Pulmonary artery resection and reconstruction is feasible and safe, with favorable long-term survival. Our results support this technique as an effective alternative to selected patients with infiltration of the pulmonary artery, such as stage Iand IIand those who proved down-staged from stage III. Accurate preoperative evaluation, precise and suitable surgical techniques are crucial to achieve good results. Only preserving the anterior and apical pulmonary arteries reconstruction of the main pulmonary artery by using the artery conduit technique without cardiopulmonary bypass in association with left pneumonectomy can be performed successfully. Postoperative anticoagulation is unnecessary.

Background
In general, distant metastasis is regarded as an incurable systemic disease. Therefore, local therapies including metastasectomy or radiotherapy are rarely applied, and the treatment goals are disease control using chemotherapy or palliation. There are, however, several reports in which local therapy can contribute to long-term survival in patients with metastatic disease, especially for brain metastasis or adrenal metastasis in patients with NSCLC.

Methods
Between 1986 and 2009, among 1548 patients who underwent surgical resection for NSCLC in our institution, we identified 405 patients who experienced recurrence after R0 resection, without history of other malignancy, and detailed recurrence information available. We investigated the recurrent mode, number of metastatic focus and organ, treatment for metastasis, and prognosis.

Results
Among 405 patients, 245 patients had distant metastasis without local recurrence, 115 had local recurrence, and 45 had both local and distant metastasis. We focused on the 245 patients with distant metastasis without local recurrence, including 215 patients who had only single organ metastasis and 93 patients who had only solitary metastasis. The treatments for distant metastasis and the 5-year survival rates were shown in the Table 1. The number of organ involved and metastatic focus were significantly associated with prolonged survival. Local therapy were mainly applied for limited metastases, and associated with higher survival rates. The number of patients and the 5-year survival rates according to the metastatic organ in patients with solitary metastasis are shown in Table 2. Other metastatic organ included soft tissue in 3 patients, kidney in 3, and trachea, intestine, and abdominal lymph node in 1.Finally, 6 patients survived more than 5 years with disease-free status; these included 2 brains, 2 lungs, 1 bone, and 1 subcutaneous metastasis.

Table 1

	Multiple organ		Single organ		Multiple		Single
Treatment	Number of pts	5y OS (%)	Number of pts	5y OS (%)	Number of pts	5y OS (%)	Number of pts	5y OS (%)
BSC	8	0	48	6.4	43	7.1	5	0
Chemo Tx	3	0	32	16.3	30	17.7	2	0
Radio Tx	19	5.7	101	11.8	43	0	58	21.4
Surgery	0	-	34	38.0	6	66.7	28	33.4
Total	30	3.5	215	15.2	122	8.9	93	23.3

Table 2

Organ	Number of pts	5y OS (%)
Brain	36	19.2
Bone	24	16.7
Lung	18	32.4
Adrenal gland	6	0
Other	9	55.6

Conclusion
Prolonged survival can be achieved using local therapy in patients with limited distant metastasis irrespective of metastatic organ.

Abstract not provided

Background

Most SCLC patients initially respond to chemotherapy but then relapse and die so new therapies are urgently required. Pre-clinical data shows statins induce growth arrest and apoptosis in SCLC and several other tumour cell types and are additive with chemotherapy. This may in part be due to impaired Ras superfamily function as statins deplete mevalonate, reducing geranylgeranylation and farnesylation of these proteins. We therefore undertook this large pragmatic phase III trial in SCLC patients to determine if overall survival (OS) was affected by the addition of pravastatin to standard treatment.

Methods
Patients with limited (LD) or extensive (ED) stage SCLC were randomised to pravastatin 40mg OD or placebo for up to 2 years and given standard chemotherapy according to local practice but recommended as either cisplatin 60mg/m2 iv or carboplatin AUC 5 or 6 and etoposide 120 mg/m2iv d1 to 3 or 100 mg BD po d2 & 3; max 6 cycles plus radiotherapy as usually given. Patients were excluded if they had used statins within 12 months prior to randomisation. Stratification was: LD vs ED and ECOG 0,1 vs 2,3. Endpoints were: primary - OS; secondary - progression free survival (PFS), local PFS (local control), response rates (RR) and toxicity.

Results
Between 2007 and 2012, 846 patients were randomised, 422 (49.9.%) received pravastatin and 424 (50.1%) placebo in 93 participating sites in the UK. The median age was 64 years (range 54-69); ECOG performance status: 0: 23%; 1: 54%; 2: 17% and 3: 6%; weight 72.6 kg; LD, 357 (42.2%); ED, 479 (56.6%); 211 (24.9%) had ipsilateral effusion and 201 (23.8%) had ipsilateral SCF lymph nodes; Relative Dose intensity of cisplatin/carboplatin and etoposide was 91.6% (range 80.8 to 99.7), and 94.7% (range 85.7 to 100); 83.4% vs 86.3% completed >4 cycles of chemotherapy on the pravastatin and placebo arms respectively. Most patients completed 6 cycles of chemotherapy: 263 (62.3%) vs 265 (62.5%) in the pravastatin vs. placebo groups. Unblinded results showing the effect of pravastatin on OS, PFS, local PFS and toxicity will be presented.

Conclusion
This trial will report on whether pravastatin 40 mg OD added to standard therapy alters the outcome for SCLC patients.

Background
IP showed superior survival outcomes compared with EP in Japanese patients. However, IP failed to show the superiority in subsequent studies in Western population. We conducted a multi-center randomized controlled trial to determine whether IP regimen is superior to EP regimen in Korean patients (ClinicalTrials.gov Identifier: NCT00349492)

Methods
Patients were randomly assigned (simple randomization, stratified by ECOG performance status and center) to IP (irinotecan 65 mg/m2 IV on D1&8 plus cisplatin 70 mg/m2 IV on D1, every 3 weeks) or EP (etoposide 100 mg/m2 IV on D1-3 plus cisplatin 70 mg/m2 IV on D1, every 3 weeks) for maximum 6 cycles, until disease progression, or until unacceptable toxicity occurred. The primary objective was to compare overall survival (OS).

Results
A total of 362 patients were randomized to IP (N=173) and EP (N=189) arms. Median OS was 10.9 and 10.3 months (m) for IP and EP, respectively (hazard ratio for death in the IP group, 0.879; 95% one-sided confidence interval, 0 to 1.054; P=0.1207). Objective response rate was higher with IP than with EP (62.4%, 48.2%, P=0.0064), however, there was no significant difference in median progression free survival between IP and EP (6.5 and 5.8 m, P=0.1158). In the pre-planned subgroup analyses, IP was associated with longer OS than with EP in male (11.3 vs 10.1 m, P=0.0361), <65 years old (12.7 vs 11.3 m, P=0.0240), ECOG performance status 0/1 (12.4 vs 10.9 m, P=0.0407) patients group. Grade 3/4 anemia, nausea and diarrhea were more frequent in patients treated with IP. There was no difference in the frequency of grade 3/4 neutropenia, thrombocytopenia, neutropenic fever, infection between both arms.

Conclusion
IP failed to show superiority in overall survival compared to EP in Korean patients with ED SCLC. However, prolongation of OS was observed with IP in pre-defined subgroup of patients with male gender, young age, or good performance status.

Background
Background: NTX-010 is a naturally occurring replication-competent picornavirus with potent and selective tropism for SCLC tumor cells expressing neuroendocrine markers. A phase I study of NTX-010 showed evidence of antitumor activity in patients with SCLC.

Methods
Methods: ES-SCLC patients (pts) with SD, PR or CR after at least 4 cycles of platinum-based chemotherapy were pre-registered to confirm diagnosis of SCLC with > 1 neuroendocrine marker by a central pathology review. Eligible pts were.randomized 1:1 to placebo (B) or NTX-010 (A). NTX-010 or placebo was administered intravenously as a 1-hour infusion in 100 mL normal saline as a single dose of 1 x10[11]vp/kg. Viral studies to determine distribution, clearance of the virus and the presence of neutralizing antibodies were done. The primary goal of this trial was to compare the progression-free survival (PFS) of arm A to B based on a sample size of 45 patients per arm to detect an improvement in median PFS from 3 to 5 months (m). A pre-planned interim futility analysis was done after 40 PFS events, and reported here.

Results
Results: The trial is permanently closed to accrual. One-hundred and twenty pts were pre-registered, of whom 58 were randomized. Baseline age, gender, ECOG performance status, and histology were balanced between arms. Median age was 63 (range: 44 - 82). 31% of pts had a PS of 0 and 69% of 1. Grade 4 adverse events were seen in 3 (12.5%) patients in arm A and none in arm B. Based on the interim futility analysis, PFS was 1.7 m (95% CI: 1.3-3.1) for arm A and 1.7 m (95% CI: 1.4-4.3) for arm B. Pts with viral RNA at 7 (7 pts) and 14 (6 pts) days had worse PFS compared to those with no detectable levels within arm A (1.0 vs 1.6 m, p=0.02; 0.9 vs. 1.2 m, p=0.06). Median follow-up in pts is 6.1 m. The 3-month OS estimates are 83% (95% CI: 69%-100%) and 85% (70%-100%) for arms A and B respectively.

Conclusion
Conclusion: This phase II study showed no benefit in PFS for ES-SCLC patients receiving NTX-010. Pts with detectable virus at 7 and 14 days had worse PFS

Abstract not provided

Background
The type 1 insulin-like growth factor receptor (IGF1R) and MET, the receptor for hepatocyte growth factor (HGF)/scatter factor, appear to play key roles in SCLC. Ganitumab and rilotumumab are investigational, fully human monoclonal antibodies targeting IGF1R and HGF, respectively. A phase 1b/2 study evaluated ganitumab or rilotumumab combined with etoposide plus carboplatin (CE) or cisplatin (PE) in extensive-stage SCLC. The phase 1b results were previously reported (Lorigan et al. Ann Oncol 2010;21[supplement 8]: abstract 444P). Here, the phase 2 results are reported.

Methods
Key eligibility criteria: ≥18 years, confirmed SCLC, ECOG performance status ≤1, no prior chemotherapy. Patients were randomized 1:1:1 to receive blinded investigational product (IP) either ganitumab (18 mg/kg IV, day 1) or rilotumumab (15 mg/kg IV, day 1) or placebo, with etoposide (100 mg/m[2] IV, days 1-3) plus, at investigator’s discretion, either carboplatin (AUC=5 mg/mL*minute IV, day 1) or cisplatin (75 mg/m[2] IV, day 1) every three weeks for 4-6 cycles followed by IP monotherapy. Patients were stratified by gender and chemotherapy (CE; PE). Primary endpoint: overall survival (OS). Key secondary endpoints included progression-free survival (PFS), objective response rate (ORR), adverse events (AEs), pharmacokinetics.

Results
185 patients (ganitumab/rilotumumab/placebo: 62/62/61) were enrolled between 2 February 2010 and 12 January 2011. Male: 77%/76%/77%. Median age: 60/61/61 years. More patients received carboplatin (41/40/40) than cisplatin (21/22/21). Most common reason for discontinuation of IP was disease progression (69%/61%/74%). Among 179 patients (59/61/59) who received IP, the most frequent any grade AEs (occurring in ≥30% of patients in any arm) were neutropenia (69%/59%/71%), anemia (39%/34%/36%), nausea (41%/30%/22%), alopecia (41%/23%/27%), thrombocytopenia (22%/30%/12%), and vomiting (19%/10%/31%). Grade ≥3 AEs and serious AEs were reported in 69%/72%/80% and 39%/38%/36% of patients, respectively. There were three IP-related grade 5 AEs: cardiac arrest (rilotumumab, n=1), febrile neutropenia (rilotumumab, n=1), gastric ulcer hemorrhage (placebo, n=1). No neutralizing antibodies were observed for either ganitumab or rilotumumab. Efficacy is shown in the table. Ganitumab and rilotumumab combined with chemotherapy showed comparable exposures as those under monotherapy and did not affect the pharmacokinetics of chemotherapy.

	Ganitumab (n=62)	Rilotumumab (n=62)	Placebo (n=61)
OS
Median (95% CI) months	10.7 (8.1–14.1)	12.2 (8.8–14.6)	10.8 (9.4–11.9)
Adjusted HR[a] (95% CI)	1.01 (0.67–1.52)	0.91 (0.60–1.39)
PFS
Median (95% CI) months	5.5 (4.4–5.7)	5.4 (4.4–5.7)	5.4 (4.6–5.8)
Adjusted HR[a] (95% CI)	1.03 (0.70–1.52)	1.03 (0.69–1.52)
Objective Response
Complete response, n (%)	0 (0)	2 (3)	1 (2)
Partial response, n (%)	39 (63)	40 (65)	35 (57)
Stable disease, n (%)	13 (21)	12 (19)	16 (26)

[a]Adjusted for baseline lactate dehydrogenase levels and stratification factors. CI=confidence interval; HR=hazard ratio.

Conclusion
In this study of chemonaïve patients with extensive-stage SCLC, the combination of ganitumab or rilotumumab with CE or PE was tolerable; no unexpected toxicities were observed. There were no meaningful improvements in OS, PFS, or ORR with either combination. Survival analyses in biomarker and pharmacokinetic subgroups are ongoing.

Background
AAK, a key mitotic regulator, plays an important role in carcinogenesis. AAK is frequently overexpressed/amplified in a variety of human cancers. Inhibition of AAK has been associated with tumor inhibition in SCLC patients; thus, AAK is an attractive target in SCLC treatment. The investigational, oral, selective AAK inhibitor MLN8237 is under evaluation in patients with advanced hematologic and non-hematologic malignancies. A phase 1/2 study (NCT01045421) evaluated MLN8237 in patients with solid tumors; phase 2 data for the SCLC cohort are presented.

Methods
Patients aged ≥18 years with relapsed/refractory SCLC, ECOG PS 0–1, and ≤2 prior cytotoxic chemotherapy regimens were eligible (symptomatic and untreated brain metastases were excluded); sensitive and resistant/refractory was defined as recurrence > or ≤90 days of frontline therapy respectively. MLN8237 50 mg was administered twice-daily for 7 days (21-day cycles). Using a Simon’s optimal 2-stage design, ≥2 objective responses were required in the first 20 response-evaluable patients to proceed to the expansion stage. Primary objective: overall response rate (ORR) by RECIST v1.1. Secondary objectives: safety, duration of response (DOR), and progression-free survival (PFS). Exploratory studies included whole-exome sequencing in banked tumor specimens to identify genetic markers/mutated pathways associated with differential response.

Results
As of April 2013, 60 patients were enrolled; median age, 62 years (range 43–76); median number of cycles 2 (range 1–15). 48 patients (80%) were response-evaluable (chemo-sensitive, n=36; chemo-refractory, n=12). Prior lines of therapy and efficacy data are shown in the table. ORR was 21%, median PFS 2.11 months. Most common grade ≥3 drug-related AE included neutropenia (53%) and febrile neutropenia (8%); all AEs were reversible.14 patients discontinued due to AEs (treatment-related in 3 patients: hyponatremia, thrombocytopenia, acute generalised exanthematous pustulosis; each, n=1). 13 on-study deaths were reported; none were considered drug-related (most frequent cause was disease progression [n=6]). Whole-exome sequencing of selected tumor samples was completed. Preliminary results will be presented.

Conclusion
MLN8237 has a generally manageable toxicity profile; data from this study suggests it has single agent antitumor activity in SCLC patients with sensitive or resistant-refractory relapse, supporting further evaluation of MLN8237 in different combination strategies in this tumor type.

	All	Chemo-sensitive	Chemo-refractory
Prior lines of therapy
First line, n (%)	N=48	n=36	n=12
Platinum/etoposide	45 (94)	33 (92)	12 (100)
Other	3 (6)	3 (8)	–
Second line, n (%)	N=23	n=19	n=4
Rechallenge	13 (57)	12 (63)	1 (25)
Topotecan	3 (13)	2 (11)	1 (25)
Other	7 (30)	5 (26)	2 (50)
Efficacy data
	N=48	n=36	n=12
Median cycles (range)	2.5 (1–15)	3.5 (1–15)	2 (2–6)
Best response, n (%)
ORR (PR)	10 (21)	7 (19)	3 (25)
SD	16 (33)	13 (36)	3 (25)
PD	22 (46)	16 (44)	6 (50)
Median DOR (months)	4.1*	–**	–**
Median PFS (months)	2.11	2.58	1.74

* Numbers of events: 7; **Not enough events to give meaningful estimates.

Background
ES-SCLC patients (pts) with progressive disease (PD) following plat-based chemo have traditionally been categorized as plat-sensitive (PD >/= 90 days from last plat dose) or refractory (PD < 90 days). Plat-sensitivity status has previously been strongly associated with response and survival in the 2[nd]/3[rd] line treatment setting. However, in a recent pooled analysis of SWOG trials in 2[nd]/3[rd] line SCLC pts, plat-sensitivity status was found to no longer be a significant independent variable for survival (Lara, ASCO 2013). We subsequently developed a new SCLC prognostic model for overall survival (OS) for potential clinical trial and bedside application.

Methods
Updated data from recent SWOG trials in 2nd and/or 3rd line ES-SCLC (S0802: topotecan + aflibercept: S0435: sorafenib; and S0327: PS-341) were pooled. Accrual goals were specified for sensitive and refractory in each trial. Hazard ratios (HRs) for OS were calculated using Cox Proportional Hazard (PH) models [unadjusted and adjusted]. To investigate a predictive model for OS, recursive partitioning was performed using the likelihood tree model of LeBlanc and Crowley. The minimum node size was set at 20.

Results
Of 329 pts, 151 were classified as sensitive, 178 refractory; median age = 63 years; males = 52%; Performance Status (PS) 1 = 67%; weight loss >5% = 28%; > 2 prior chemo = 16%; and elevated LDH = 43%. HRs from unadjusted Cox models for OS for refractory vs. sensitive were 1.0 (95% CI 0.81-1.25, p=0.98) and 1.24 (95% CI 0.99, 1.57; p=0.06). Cox PH models adjusted for baseline prognostic factors showed that plat-sensitivity status was not significantly associated with OS. Elevated LDH was significantly associated with PFS while LDH, PS, weight loss, and male sex were independently associated with OS. Clinically relevant prognostic risk groups (High, Intermediate, and Low) were identified by recursive partitioning analysis, as shown below (MST= median survival time). High Risk (MST = 2 months: Elevated LDH And > 5% Weight Loss Or PS >0) Intermediate Risk (MST = 5 months: Elevated LDH but not High Risk Or Male) Low Risk (MST=8 months: Normal LDH And Female)

Conclusion
In this large database analysis, clinically relevant prognostic risk groups were identified, categorized as low, intermediate, and high risk, with differential survival outcomes observed for each group. Validation of these risk groups in an independent SCLC dataset is warranted. If validated, these risk groups will have important implications for individualized patient counseling in clinic and stratification of patients in prospective trials in the second and third line setting.

Abstract not provided