Virtual Library

Background
It is of great importance for treatment optimization to know before the initiation of treatment the different variables that may be used to help determine the result in the oncological diseases. It has been shown in preveious studies that have been performed that standard uptake values obtained in PET-CT carried out before the treatment in various cancer types enable us to make predictions regarding the tumor’s “aggressiveness”, response to treatment and survival. In this study, we researched the predictive role of pretreatment SU values of the patients with SCLC in determining the response to treatment and the survival.

Methods
Ninty patients were enrolled into this study, with whom we had follow-up in our center due to SCLC, and in whom pretreatment PET-CT was performed. The SU value of primary tumor was then determined. The relationship between these values and response to treatment is examined through ROC analysis and the cut-off value, effective and thus the determination to the response to treatment is identified. In the patient group, age, gender, stage, performance status, response to treatment, and effect of SU values upon response to treatment and survival are studied.

Results
Ninty patients were enrolled into the study. Median follow-up period is 11 (1-43) months, and the median age is 58 (39-83). Ninety percent of patients are male and the diseases in 37% of were noted to be of limited-stage. All patients received platinum+etoposide chemotherapy once every 21 days, as median amount of treatments to cure was 6 (3-7). When ROC analysis was carried out the SU value, suitable for determining the response, is found to be 10. The sensitivity of this value is found 85.7% (CI 95%; 63-96), the specificity was found to be 61.8% (CI 95%; 49-73) (AUC: 0.783; p: 0.0001). The overall response rate (complet response+partial response) is 59.1% in the other patients, while this rate is 93.3% in patients with involvement above the cut-off value(p:<0.0001). The response rates are higher in patients with age greater than or equal to 60 years old (p:0.047). No significant relationship is found between gender, stage and performance status, and the response to treatment. In the survival analyses carried out, median disease-free survival is found to be 9 months (SE:1; 95% CI: 8-10), and median general survival was determined to be 17 months (SE:3; 95% CI: 12-22). When single and multivariate analysis are carried out, the positive impact/effect of limited-stage disease and response to treatment upon the progression-free survival is determined (p<0.05). The fact that SU value is higher than cut-off value, affects the progression-free survival positively borderline (p:0.085). As a result, this is effective upon general survival, and studied using single and multivariate analysis. It is seen that only the response rates affect the general survival positively (p<0.05).

Conclusion
PET-CT is used in SCLC frequently for staging purposes. The data obtained in PET-CT may have a role in determining the prognostic characteristics of the disease as well as the response to treatment.

Background
Brain metastases are very common in patients with small-cell lung cancer (SCLC). Prophylactic cranial irradiation (PCI) has been shown to reduce the incidence of brain metastases and to improve overall survival in patients with limited-disease SCLC (LD-SCLC). However, brain metastases are often observed after PCI, and the optimal treatment for these brain metastases is still unclear. The present study investigated the recurrence of brain metastases after PCI in patients with LD-SCLC and the therapeutic efficacy of stereotactic radiosurgery for these metastases.

Methods
Between December 2000 and December 2012, 228 patients with LD-SCLC were treated and 98 of these patients with a complete response (CR) or a near CR to chemoradiotherapy underwent PCI at the National Cancer Center Hospital. We retrospectively reviewed the medical records and imaging data for these 98 patients.

Results
Twenty-four (24%) of the 98 patients developed brain metastases after PCI. The characteristics of the 24 patients were as follows: median age, 62 years (49-72 years), male/female, 21/3; performance status 0/1/2, 2/16/6. Twelve (50%) of the 24 patients had cranial recurrences only. Twelve patients had single brain metastases, and 12 patients had multiple lesions. Nine patients had neurological symptoms due to brain metastases. The median period after PCI until the appearance of the metastases was 9.9 months (1.1-34.9 months). Fifteen (63%) of the 24 patients underwent stereotactic radiosurgery (gamma knife radiosurgery [GKRS]), and one patient received whole brain radiotherapy. Six patients were treated with chemotherapy plus best supportive care (BSC), and two patients underwent BSC alone. The 15 patients who received GKRS had brain metastases with/without extracranial lesions (7 with, 8 without); three were symptomatic, and 12 were asymptomatic. The median number of brain metastases at the time of the first GKRS was one (range, 1-4). The local control rate of the lesions treated with GKRS was 86.7% (complete response in 3 patients, partial response in 7 patients, and stable disease in 3 patients). Five patients underwent further GKRS because of newly developing brain metastases (median: 4 times, range: 2-7 times). The median intracranial control time of the 15 patients was 6.8 months. The median survival time of the 15 patients was 29.3 months after the initial diagnosis, 13.7 months after the development of brain metastases, and 12.7 months after the treatment of GKRS. The median survival time of the patients without extracranial lesions was 20.2 months after the development of brain metastases and tended to be longer than that of the patients with extracranial lesions (12.6 months). Severe adverse events arising from GKRS were not observed in this series.

Conclusion
Stereotactic radiosurgery may be an effective option as a salvage therapy for brain metastases after PCI in patients with LD-SCLC.

Background
Small cell lung cancer (SCLC) is the most aggressive form of lung cancer. The prognosis for SCLC patients remains unsatisfactory, despite advances in chemotherapy. In this study, we sought to clarify the prognosis and treatment patterns of patients with SCLC.

Methods
A cohort comprising all patients diagnosed with SCLC between Jan. 2004 and Dec. 2006 was assembled from the Taiwan Cancer Database. Patients were followed up until December 31, 2009 to determine overall survival. Patient survival was estimated using the Kaplan-Meier method, and Cox’s proportional hazard model was used to determine the relationship between prognostic factors and median survival time.

Results
Among the 1684 patients diagnosed with SCLC, 1215 (72%) were diagnosed with extensive stage, and 469 (28%) with limited stage. Most of the patients were male (90%). The median survival time of patients with limited-stage (LS) and extensive-stage (ES) SCLC was 10.3 months and 5.6 months, respectively. For LS patients, surgery, chemotherapy, and combined chemotherapy and radiotherapy resulted in better survival than best supportive care (HR: 0.20, p<0.001; HR: 0.61, p<0.001; HR: 0.37, p<0.001 respectively). For ES patients, male gender was significantly associated with poor prognosis (HR:1.45, p<0.001) and chemotherapy was shown to improve overall survival more effectively than best supportive care (HR: 0.37, p<0.001).

Conclusion
For LS SCLC patients, surgery, chemotherapy, and combined chemotherapy and radiotherapy improved survival compared to best supportive care. ES SCLC patients benefited more from chemotherapy treatment than best supportive care.

Background
Belotecan is a new camptothecin analogue and a potent topoisomerase I inhibitor. The aim of this phase II study was to investigate the efficacy and toxicity of belotecan in previously untreated elderly patients with small cell lung cancer (SCLC).

Methods
A total of 26 patients, aged ≥65 years, with previously untreated, extensive-stage SCLC were enrolled in the study. Belotecan was administered by daily intravenous infusion at 0.5 mg/m[2]/day for 5 consecutive days every three weeks.

Results
The overall response rate and disease control rate of chemotherapy on an intention-to-treat basis were 35% and 54%, respectively. The median overall survival was 6.4 months, and the median time to progression was 2.8 months. The most common toxicity was hematologic. Grade 3 or 4 neutropenia occurred in 80.8% of patients, and grade 3 or 4 thrombocytopenia in 15.3%. Non-hematologic toxic effects of grade 3 or 4 were uncommon.

Conclusion
Belotecan had modest efficacy and well-tolerated toxicity in previously untreated, elderly SCLC patients. Single belotecan could be a promising treatment option, considering its lower toxicity in elderly patients who are unsuitable candidates for platinum plus etoposide chemotherapy.

Background
CONVERT is a multicentre, randomised, phase III trial open in Europe and Canada in limited-stage small cell lung cancer. Patients are randomised to twice (45 Gy in 30 fractions) or once-daily radiotherapy (66 Gy in 33 fractions) given concurrently with 4-6 cycles of chemotherapy. This study is funded by Cancer Research UK and involves centres from the UK NCRI, the ‘Groupe Francais de Pneumo-Cancerologie’, the Spanish Lung Cancer Group, the EORTC and NCIC CTG.

Methods
To identify and review the challenges in site set-up. To review time taken from site initiation to first patient randomised, number of centres opened that included 0-2 patients and number of centres that recruited the majority of all patients.

Results
In June 2013, 519/532 patients had been recruited in 9 countries; 299 from 32 UK centres, 100 from 17 French centres, 39 from 9 Canadian centres, 27 from 6 Spanish centres, 26 from 3 Belgian centres, 13 patients from 1 centre in Slovenia, 9 from 2 centres in The Netherlands and 6 patients from 1 centre in Poland. Figure 1 shows the number of centres open and patients recruited. 96 sites are currently open to recruitment (5 sites opened in 2008, 34 in 2009, 31 in 2010, 17 in 2011, 8 in 2012 & 3 in 2013, 2 sites subsequently closed early) of which 74 (77%) have randomised at least 1 patient. 24 sites (25%) recruited only 1 or 2 patients. 10 sites have recruited 49% of the total number of patients with a single site recruiting 18.5% of all patients randomised. Time taken from site initiation to 1[st] patient randomised ranged from 0–1029 days with a median of 144 days. Time taken to complete the QA exercise from initial information sent to site ranged from 14-1181 days with a median of 290.5 days. Figure 1

Conclusion
Recruitment to an academic trial in LS-SCLC is a challenge but accrual has improved considerably since 2008. This can be directly related to the increasing number of sites opened to recruitment. Duration of site set-up and completion of the QA exercise are factors explaining slower than anticipated accrual rates particularly between 2008 and 2010. We anticipate that the study will close to recruitment in July 2013. International participation has been a key factor to the success of the trial and the experience gained will be of value to the design of future radiotherapy studies to ensure target accrual.

Background
Patients with relapsing SCLC have a poor prognosis and options for treatment are limited. Paclitaxel in combination with platinum has demonstrated activity in first and second-line treatment. We conducted a phase II study to investigate the anti-tumor activity and safety profile of Docetaxel/Carboplatin (DC) in patients with refractory or relapsing SCLC.

Methods
SCLC patients who were refractory to or relapsed after first-line treatment and who did not receive platinum based chemotherapy within the 3 months before inclusion, were treated with Docetaxel (75 mg/m[2], day 1 iv) followed by Carboplatin (area under the curve 6, day 1 iv.) once every 3 weeks for 4-6 cycles. Primary endpoint was response rate (RR) and secondary endpoints were time to progression (TTP), overall survival (OS) and safety profile. Tumor response was measured according to RECIST version 1.1. Toxicity was evaluated according to Common Toxicity Criteria of the National Cancer Institute.

Results
50 patients (29 male and 21 female) were included with a median (range) age of 67 years (46-82). The majority of patients had an Eastern Cooperative Oncology Group performance status of 1; median time off first-line treatment was 12 weeks. Average number of treatment cycles was 4 (range 1-6). Evaluable for response and toxicity were 45 and 48 patients, respectively. A complete response was achieved in 1 patient, a partial response in 24 patients, stable disease in 16 patients, and progressive disease in 4 patients (RR 73.5%; 95% confidence interval, 59 to 88). Median time to progression was 133 days (range 89-177). Median overall survival was 198 days (range 77-322). One-year survival rate was 34%. Hematologic toxicity grade 3 and 4 was leukopenia 19% and 8%, thrombocytopenia 10% and 2%, and anemia 4% and 0%, respectively, in a total of 186 cycles. Nonhematologic toxicity was diarrhea grade 3-4 (15%), stomatitis grade 3 (4%) and hyponatriemia grade 3 and 4 in one patient each. Seven patients were hospitalized due to neutropenic fever; no toxic deaths occurred.

Conclusion
Second-line DC in refractory or relapsing SCLC patients yielded a high RR and toxicity was relatively mild in these poor-prognosis patients. NCT00686985

Background
Despite initial responses to chemotherapy, SCLC typically progresses within a few months. Targeting residual disease has the potential to improve outcomes. A number of tumor specific glycolipid antigens have been identified and are potential targets for immune therapies. In a series of phase I clinical trials, vaccination with each of these antigens individually was safe and induced antibody responses in the majority of patients. Preclinical data indicate that combining these antigens will expand the immunogenicity across a broader array of SCLC tumor cells. We conducted this pilot trial to determine the safety and immunogenicity of a vaccine combining five of these antigens.

Methods
Patients with limited or extensive stage SCLC who have completed initial chemotherapy +/- thoracic or cranial irradiation with a maintained response are eligible. Vaccinations must start within 3-8 weeks of the last chemotherapy, and at least 1 week after radiation. Patients receive KLH-conjugates of GD2L, GD3L, Globo H, fucosyl GM1, and N-propionylated polysialic acid (30mcg each) plus OPT-821 adjuvant (150mcg) subcutaneously on weeks 1, 2, 3, 9, 20, and 32. One cycle of etoposide/platinum chemotherapy was administered in week 6. A significant immune response is defined as an antibody titer of ≥ 1:80 by ELISA against a given antigen or a ≥ 8 fold increase over baseline for patients with a detectable baseline titer. The vaccine would be deemed worthy of further study if >5 patients had an immune response to 3 or more antigens.

Results
Ten patients were treated, including 9 with extensive stage, 4 women, 7 with prior brain radiation. The number of vaccinations administered was: 1 (1pt), 3 (2), 4 (3), 5 (2), and 6 (2). Toxicity was limited to mild skin reactions. One patient was taken off study after he developed aphasia the day after the first vaccination; MRI brain was unremarkable and symptoms resolved spontaneously. No patients met the predefined criteria for immune response. Six patients had increases in IgM titers to 1-2 antigens. The median time to progression was 4 months. The two patients with the strongest IgM responses to Globo H and fucosyl GM1, and also the only IgG responses to fucosyl GM1, had progression of disease 7 and 9 months after starting the vaccines, and both progressed initially in the brain only.

Conclusion
The polyvalent vaccine is safe, but fewer patients than expected had a significant immune response, Two patients with immune responses experienced a longer than expected time to progression. A second cohort of patients is now receiving the vaccinations over a shorter period of time and without the added cycle of chemotherapy. Five out of 10 of those patients have been enrolled, and preliminary data will be available on those patients for the meeting. Following completion of this pilot trial, a multicenter randomized trial is planned. Supported by MabVax Therapeutics’s NIH grant R41 CA128363 and a grant from FAMRI.

Background
The AJCC 7[th] TNM staging system is recommended for both non-small cell and small cell lung cancer (SCLC), although SCLC has traditionally been classified using the limited and extensive definition. We evaluated long-term survival of patients with SCLC according to the AJCC 6[th] and 7[th] TNM edition.

Methods
239 patients had been pathologically diagnosed with SCLC from January 2000 through December 2009 in our hospital. We included 187 patients who had initial CT scan of the thorax. We analyzed all available data for TNM staging using the AJCC 6[th] and 7[th] edition as well as the long-term survival rates.

Results
According to stages defined by the 6th TNM edition system, 5-year stage-specific survivals (5-YSR) were 43% for stage I and II, 36% for stage IIIA, 10% for stage IIIB, and 7% for stage IV. Applying the 7th edition system, 5-YSR were 46% for stage I and II, 24% for stage IIIA, 7% for stage IIIB, and 9% for stage IV. Using the 7[th] edition, 5-YSR of the stage IV was higher than that of stage IIIB. The number of changed stage from IIIB in the 6[th] edition to IV in the 7[th] edition was 22 and all were related to malignant pleural effusion.

Conclusion
Application of the TNM staging to SCLC stratifies survival more distinctly than limited and extensive definition by providing more substages. Slightly better survival of stage IV SCLC than that of stage IIIB in the 7[th] edition is related to malignant pleural effusion.

Background
Solitary Fibrous Tumors of the Pleura (SFTP) refer as to a heterogeneous group of mesenchymal malignancies with various anatomic and histologic features. Upfront surgical resection is the standard approach, but the outcome of patients is unpredictable. Recurrences may be aggressive and difficult to treat.The most widely accepted staging system has been proposed by De Perrot et al., and is based on the anatomy of the tumor implantation (sessile/pedunculated), and the presence of histologic signs of aggressiveness, including cellularity with crowding and overlapping of nuclei, cellular pleomorphism, high mitotic count, necrosis, or stromal/vascular invasion. Given the rarity of the tumor, limited evidence is available about the role and the modalities of perioperative and definite chemotherapy for SFTP.

Methods
Multicenter retrospective study of patients (pts) with histologically-proven SFTP with complete follow-up from surgical diagnostic to tumor recurrence and death.

Results
68 pts (28 males/40 females) were included. Median age at diagnosis was 62 year-old. Tumor stage according to the De Perrot system was 0/I for 29 pts, II for 23 pts, III for 7 pts, and IV for 4 pts. Adjuvant chemotherapy was given to 7 patients, mostly with stage III/IV SFTP, consisting of doxorubicin-based regimen. Recurrence rate and median time-to-progression (TTP) after surgery were 3%, 52%, 71%, and 75% (p<0.001), and 107, 70, 29, 11 months (p=0.006) for stage 0/I, II, III, and IV tumors, respectively. Besides tumor stage, predictors of shorter TTP were incomplete resection (p<0.001) and a higher number of histologic signs of malignancy (p=0.009). At time of tumor recurrence, 12 pts received chemotherapy. Highest disease control rates were observed with trabectedine (7/9 pts; Disease Control Rate (DCR): 78%; median TTP: 3,4 months), and gemcitabine-dacarbazine combination (2/3 pts, DCR: 66%; median TTP: 1,9 months). Median overall survival of the whole cohort was 56 months.

Conclusion
This study 1) confirms the prognostic value of the De Perrot staging system, 2) indicates a high recurrence rate in patients with stage II tumors, for which perioperative chemotherapy may be considered, and 3) suggests an interest for trabectedine in the setting of recurrent tumors. Besides clinical data, further molecular characterization, including recently identified specific gene fusions, may help to better predict the outcome of patients with SFTP.

Background
The number of patients with Malignant Mesothelioma is expected to increase over the next ten years from the number of 3.300 new cases that are reported annually nowadays. Most of the patients appear with advanced, surgically unresectable disease at the time of diagnosis and relapse is usual after the standard 1st line treatment with a platinum-pemetrexed doublet, due to intrinsic and acquired resistance to pemetrexed. This review focuses on the treatment options in Malignant Mesothelioma resistant to pemetrexed and is based on trials for possible 2nd line regimens tested the last ten years and ongoing trials of novel agents and studies exploring the basis of pemetrexed resistance.

Methods
We searched via PubMed/Medline, Clinicaltrials.gov and American Society of Clinical Oncology databases for articles and ongoing trials published until 1/2013 using the term: “mesothelioma”, in association with “2nd line chemotherapy”, “pemetrexed resistance”, “relapsed”, “pemetrexed-pretreated” and “biomarker”. Primary sources have been quoted.

Results
Data from forty eight conducted and ongoing trials, mainly phase II, single-armed, but also retrospective and phase III, are shown and discussed. Thirty different agents were tested in these trials as possible 2nd line drugs for Malignant Mesothelioma. Currently, no guidelines for 2nd line chemotherapy in Malignant Mesothelioma tumors resistant to pemetrexed are available.

Conclusion
In the absence of a “gold standard” as 2nd line treatment for Malignant Mesothelioma, results of ongoing trials are eagerly awaited and the research for novel agents remains critical. Furthermore, additional research should be done towards the understanding of the mechanisms that lead to pemetrexed resistance, the possible personalization of 1st and 2nd line treatment and the use of biomarkers that can be used as factors predicting the resistance itself. Finally, more patients should be convinced to enroll in clinical trials and more randomized trials have to be conducted.

Background
Early diagnosis of malignant pleural mesothelioma (MPM) can improve patients’ outcome but is hampered by non-specific symptoms and investigations, which delay diagnosis and result in advanced stage disease [van Meerbeeck JP, 2011]. An accurate non-invasive test allowing early stage diagnosis in asbestos-exposed persons is lacking. Breathomics aims at a non-invasive analysis of volatile organic compounds (VOCs) reflecting the cells’ metabolism. The breathogram obtained by the electronic nose does however, not allow identification of MPM-related VOCs [Chapman EA 2009, Dragonieri S 2011]. Ion mobility spectrometry (IMS) combines the advantages of online direct sampling with the possibility of VOC identification and linking to MPM pathogenesis [Baumbach JI 2009]. We investigated which VOCs could play a role in MPM pathogenesis in order to build a possible diagnostic MPM tool.

Methods
10 MPM patients and 10 healthy asbestos-exposed individuals (mean asbestos fiberyear count 14,6 (5,5) fibre.years/cc) were included after refraining from eating, drinking and smoking for at least 2 hours before sampling. They breathed tidally for 3 minutes through a mouthpiece connected to a bacteria filter. Ten ml alveolar air was sampled via a CO~2-~controlled ultrasonic sensor and analyzed using the BioScout Multicapillary Column/Ion Mobility Spectrometer (MCC/IMS, B&S Analytik, Dortmund, Germany) [Westhoff M 2009], by using N~2~ as a carrier gas. Per subject a background sample was taken. Peaks of interest were visually selected and their intensity (V) was analyzed and compared between background and breath samples via on-board VisualNow 3.2 software and SPSS v21 (IBM) using Mann-Whitney-U tests.

Results
Out of 41 peaks of interest, three show a significantly higher intensity in the exhaled breath of MPM patients than healthy controls [Table]. The high AUC~ROC~ of resp. P12 (0.877) and P24 (0.863) suggests a possible role of these associated VOCs in MPM pathogenesis and as a diagnostic marker in discriminating MPM patients from asbestos-exposed healthy controls. Figure 1

Conclusion
Several VOCs of interest were obtained in the breath of MPM patients. Two peaks were significantly discriminating between both populations. GC-MS analysis and further large cohort studies are ongoing in order to validate the accuracy of IMS as a diagnostic tool for MPM.

Background
Radiotherapy reduces local recurrence following extrapleural pneumonectomy (EPP), and forms part of a potentially curative, multimodality treatment of malignant pleural mesothelioma. Hemithoracic radiotherapy poses a significant dosimetric challenge. Conventional techniques have suffered with marked dose uncertainty, while modern IMRT techniques have been associated with increased pulmonary toxicity. We conducted a retrospective review of all patients referred to our institution for hemithoracic radiotherapy following EPP, with the aim of assessing treatment toxicity and outcomes. The present study is, to our knowledge, the largest Australian series of adjuvant radiotherapy for this disease.

Methods
53 patients were referred following EPP for malignant pleural mesothelioma, with or without neoadjuvant chemotherapy, between 2004 and 2012. 4 patients were excluded or did not commence radiotherapy due to poor performance (n = 3) or disease progression (n = 1). Radiotherapy involved a 3D conformal, mixed photon and electron technique, delivering 45-55 Gy in 25-28 fractions (2004-2009, n=31), and a 9-field IMRT technique, delivering 50.4-60 Gy in 28-30 fractions (2009-2012, n=18). We assessed toxicity, disease progression and survival in all patients who commenced radiotherapy (n = 49). Toxicity was assessed using the Common Terminology Criteria for Adverse Events version 4.0 and survival was calculated from the date of EPP using the Kaplan-Meier method.

Results
31 patients (59%) received neoadjuvant chemotherapy, with a combination of platinum agent and pemetrexed. 41 patients (84%) completed treatment as prescribed. 6 patients stopped prematurely due to toxicity, and 2 due to disease progression. Most patients discontinuing due to toxicity (n = 5) received over 90% of the prescribed dose. Low grade nausea, anorexia and fatigue were near universal, however severe (grade 3) skin toxicity, nausea and oesophagitis were 8%, 6% and 2%, respectively. One patient developed a grade 4 Pneumocystis carinii infection, however there were no cases of radiation pneumonitis. Late toxicities were rare, with the exception of a persistent elevation in the liver enzymes alkaline phosphatase (ALP) and gamma-glutamyltransferase (GGT). Of the patients treated with right-sided disease (n = 26), 9 (35%) developed grade 2-3 elevations in ALP or GGT. Grade 2-3 liver toxicity was more common in patients treated with a conventional technique (53%) than with IMRT (11%). No patients developed clinical hepatitis. With a median follow up of 19 months (range 2-102 months), median progression-free survival and overall survival were 22 and 30 months, respectively. 2-year overall survival was 53.8%. 7 patients (14%) were alive beyond 5 years.

Conclusion
Hemithoracic radiotherapy can be safely delivered in selected patients following EPP. Although associated with significant early toxicity, most patients complete treatment and late toxicity is uncommon. Our outcomes compare favourably with recently-published international series.

Background
A brief review of the quality of life of one group of survivors and carers identified that after EPP the most common symptoms affecting quality of life were fatigue, dyspnoea, insomnia and pain. Survivors struggled to return to normal levels of social and role functioning. Interviews also identified the need to share stories with others survivors. In response to this information a well living programme was integrated into an EPP survivor support group. The aim being: to provide information and develop skills to empower survivors and carers to react and improve their quality of life; to optimize the physical, emotional, and social functioning of survivors; and, to assist carers to perform their role while remaining mindful of their own care need.

Methods
Four 1 day support / well living meetings were held during 2012. Each involved invited speakers, fitness assessments (6-minute walk test), and time allocated for networking over food and beverages. Topics included: optimizing living with one lung, physical and respiratory assessments, exploring survivor and carer experiences, mindfulness healing, exploring survivor and carer resilience, creating opportunities for self-attention and nurturing, relating science to the living experience, pain management, writing and music for healing, eating for wellbeing and a goal for 2013. Physiotherapy staff contributed regularly to the meetings.

Results
The average meeting attendance was 20 participants consisting, of 50% survivors and 50% carers. Patients responded positively to the physical challenges and set their 2013 goal as participation in a community 7 Km fun walk – The Sydney Bay Walk in August. In 2013, an exercise physiologist has worked individually with some group members, either face-to-face or via telephone. This encouraged increased aerobic and resistance training. A number of survivors report improvements in overall fitness, enjoyment of daily living and satisfaction with life.

Conclusion
There appears to be consistent changes in participant confidence, and attitude toward physical fitness leading to improved enjoyment of life. The support group is an important link between long term survivors, those currently being treated and a valuable resource for new patients deciding about EPP.

Background
To examine the evidence base for radiotherapy in the treatment of pain in malignant pleural mesothelioma.

Methods
A systematic search of the Medline (1946-2013), Embase (1974-2013) and Central (The Cochrane Library Issue 9, 2012) databases was performed looking for studies which evaluated the role of radiotherapy for pain relief in people with malignant pleural mesothelioma (MPM). Studies were included if MPM was radiologically or histologically diagnosed, radiotherapy was given with the intent of improving pain and response rates to radiotherapy were reported. Documentation of the dose and fractionation of radiotherapy that was given was required. Finally, the study must have explored the relationship between radiotherapy and pain response. Systematic reviews were ineligible.

Results
Nine studies were eligible. There was marked heterogeneity among studies so quantitative synthesis of results was not possible. The most recent study reported a clinical response rate of 54% two weeks after radiotherapy given at a dose of 36 Gray (Gy) in 12 fractions. This was targeted at the area of MPM that was felt to be causing the pain. However, this assessment was performed retrospectively. A radiological response of 43% measured via CT scanning two months after irradiation was also reported in this study. Another study reported a superior response rate (50%) for those treated with a 4 Gy fraction size compared with those treated with a fraction size of less than 4 Gy (39%). However, this was not randomised and reflected a change in policy to treat sites of symptomatic disease only rather than covering the entire volume of disease. A further study suggested a benefit to hemi-thoracic irradiation at a dose of 30 Gy in 10 fractions. However, Cobalt machines were used in this study. Another study suggested no benefit to hemi-thoracic irradiation at a dose of 40 Gy in 20 fractions. However, 27 of the 47 patients in this study had no pain at study baseline. The other studies in this review report varying response rates, primarily reported in a retrospective fashion.

Conclusion
There are no high quality data to support the use of radiotherapy for pain management in MPM. Studies focusing on clear pain endpoints, improving target delineation and delivering higher equivalent doses using modern day radiotherapy are needed. Biomarkers which may predict response in a proportion of patients should be sought.

Background
ACT using genetically modified T cells has shown great promise in different malignancies (eg melanoma, chronic lymphoid leukemia.) Our group has recently applied this technology by creating T cells that express a chimeric antibody receptor (CAR) that targets the highly expressed tumor antigen, mesothelin to treat malignant mesothelioma (MM), ovarian, and pancreatic cancer and has initiated a Phase I trial in MM patients using mRNA transfected CAR T cells against mesothelin (CART-meso cells) to test its feasibility and safety.

Methods
Three patients with progressive MPM with epithelial or biphasic MPM underwent apheresis for T cell isolation/electroporation/expansion. To mitigate concerns about potential off-target toxicity due to low levels of mesothelin expression in normal tissues we developed “biodegradable” mesothelin-specific CAR engineered T cells that transiently express CAR transcripts following mRNA electroporation. T cells were electroporated with the CAR mRNA and then frozen aliquots were thawed and administered to the patients. Serum/peripheral blood/marrow mononuclear cells were acquired from patients during the trial to detect/quantify abundance of transgene and CD3e transcripts, soluble cytokine factors, human anti-mouse antibody (HAMA), CAR T cell-induced humoral responses against tumor antigens.

Results
Adoptive transfer of mRNA engineered to express “biodegradable” CAR that target mesothelin was feasible. CARTmeso cells were detected in the blood and demonstrated the predicted transience of persistence in peripheral blood. Infusions were safe from the perspective of off target toxicity due to low level mesothelin expression in normal tissues. However, in one patient who received two courses of T cells, with the second course after a 6 week delay developed an anaphylactic reaction after T cell infusion, presumably due to IgE antibodies generated against the murine single chain antibody that is part of the CAR. We observed clinical activity in one patient who showed a 50% reduction in mediastinal tumor volume (Figure 1). This patient also developed antibodies against other proteins present in mesothelioma cell lines. Figure 1 Fig 1. Greater than 50% reduction in volume of mediastinal MPM tumor in one patient after infusion.

Conclusion
ACT of mRNA engineered CART-meso cells is feasible and safe from the perspective of off target toxicity due to low level mesothelin expression in normal tissues. mRNA engineered T cells can be used to evaluate potential issues of off-target CAR-mediated toxicity and support the development of CAR-based strategies to target mesothelin.

Background
Patients with malignant pleural mesothelioma (MPM) may be treated aggressively by a multimodality approach including radical resection with a curative intent. Yet, a large proportion of them fail locally. One of the measures to reduce local relapse is heated pleural chemoperfusion (HPCP) with platinum based compounds. We report the results of a combined treatment consisting of resection and HCPC in patients with epitheloid MPM (EMPM).

Methods
A single center retrospective study of 48 patients with EMPM scheduled for the combined approach over a 16 years period. Surgery consisted of extrapleural pneumonectomy (EPP), or lesser resections (LR), followed by HPCP with cisplatinum, 100mg/m[2] at >42[o]C for 60 min. Survival analysis was done with the Kaplan-Meier method and comparisons between groups with chi square analysis.

Results
Of the 48 pts only 42 had both resection and perfusion. Operative mortality was 6.2%; all fatalities occurring after EPP. There was no toxicity. EPP was performed in 72% of advanced stage (AJCCS III+IV) compared to 54% in early stage (AJCCS I+II) disease (p=0.238). Major and minor morbidity were 8(27%) and 9(32%) Vs. 2(16%) and 3(25%), for EPP and LR respectively. The Overall median survival was 14.6 (11.1-18.2, CI-95%); 13.4 months (3.4-23.5, CI-95%) and 28.3 (7.7-48.9, CI-95%) months for EPP and LR respectively (p=0.079). Patients undergoing EPP+HPCP for advanced stage EPMP (n=21) had a median survival of 21.6 months (2.0-43.0, CI-95%). Local control was achieved in 87.5% of the EPP group and 42% of the LR group (p<0.007). The common sites of relapse or progression were ipsilateral pleura (n=9), contralateral pleura (n= 7) and peritoneum (n=5). Early stage disease (I+II) had a survival not significantly better then late stage disease (III+IV).

Conclusion
In a center with low-volume surgery for mesothelioma, resection and HPCP can be performed with acceptable mortality and morbidity and no toxicity. In EMPM, HPCP does not reduce local relapse after resections lesser than EPP. In advanced stage EMPM, EPP coupled with HCPC seems to reduce the rate of local relapse, and may contribute to a relatively long survival. The role of HCPC in conjunction with surgery should be further investigated. Figure 1

Background
The legacy of occupational and environmental asbestos exposure in Piedmont, Italy, is a continuing epidemic of pleural malignant mesothelioma (MPM). MPM still entails a poor prognosis, but progresses in its medical and surgical treatment have occurred and guidelines started to appear. We aimed to assess: (i) the appropriateness of treatment for MPM cases recently diagnosed in the Piedmont population, taking into account patients’ general conditions and disease stage at diagnosis, (ii) the end results of treatment.

Methods
We exploited the Registry of Malignant Mesotheliomas (RMM) records to identify incident cases from 2008 to 2011. Patients diagnosed/treated in hospitals including a thoracic surgery unit and in the Casale Monferrato hospital were known to represent about 70% of all MPM incident cases and were included in this study, as for them not only clinical records on diagnostic procedures, but also on treatment and follow-up were completely retrieved by RMM. Vital status at 31/12/2012 was ascertained for all MPM cases. Current analyses were limited to incident cases 2008-2009, followed at least up to 3 years. Multinomial logistic regression was used to estimate the odds ratio (OR) of receiving a specific treatment (categories: cytoreductive surgery, CRS, chemo/immunotherapy, CIT, best supportive care, BSC), conditional on individual characteristics. Survival was assessed with univariate (Kaplan-Meier) and multivariate (Cox) methods.

Results
There were 297 MPM cases. Taking CIT as reference, the OR of receiving CRS was decreased by older age and low performance status (Table 1). That of receiving BSC alone was increased by older age and low performance status as well, and by non-epithelial histotype. Median survival was 13.6 months for patients receiving CIT and 18.3 for those undergoing CRS Figure 1

Conclusion
In participating centres, MPM are currently treated in agreement with available guidelines, and treatment outcomes are consistent with expectations.

Background
Single time-point unidimensional tumor thickness measurements define measurable disease for clinical trial inclusion and also constitute a field in the IASLC prospective mesothelioma staging database. The modified RECIST guidelines for mesothelioma did not alter the 10mm minimum tumor measurement recommendation. It is unclear at what minimum thickness interobserver variability exceeds 20% of tumor thickness and how morphology, and location affect interobserver variability in a single baseline measurement.

Methods
105 thoracic CT scans were collected retrospectively from 50 mesothelioma patients. Each scan was reviewed by a medical oncologist, who identified 170 discrete sites of mesothelioma tumor across all scans that represented a range of thickness. Lesion morphology (concave rind, convex rind, convex mass, fusiform mass), and anatomic location (chest wall, mediastinum, anterior angle, or posterior angle; craniocaudal location; bone/soft tissue) were categorized. Using a custom computer interface (Abras), reference tumor thickness measurements were obtained by creating a line segment that spanned the tumor from the parietal tumor margin along the chest wall or mediastinal structures to the inner tumor margin. Measurements were selected to capture a range of tumor thicknesses and morphologies, rather than the most clinically relevant measurement sites. An observer study was conducted in which each of five other physicians was presented with the individual CT sections and the same digitally fixed location of the outer tumor margin at each of the 170 pre-defined tumor measurement sites. Each observer then independently created a line segment to capture tumor thickness at all measurement sites. Relative differences among the tumor thickness measurements of observers were estimated using a random-effects analysis of variance model (ANOVA) to identify the smallest tumor thickness at which linear measurements could be made reliably. Comparisons were made with the RECIST tumor response criterion of 20% for progression.

Results
The median mean measurement across all sites was 9.68mm, reflecting the study’s aim to investigate measurement variability as a function of tumor thickness, given that the current standard minimum is set at 10mm. The mean range of observer measurements across all sites was 15.1% of the mean per-site measurement (SD 9.1%). Measurements acquired at tumor sites with reference thickness less than 7.5 mm demonstrated inter-observer variability (as defined by the difference between the maximum and minimum measurements of the observers at each site) with a 75th percentile that included 20% of the tumor thickness, while measurements acquired at sites with mean tumor thickness >7.5mm showed inter-observer variability with a 75[th] percentile <20% of tumor thickness. Inter-observer variability tended to be lower for convex mass lesions relative to that associated with the other three tumor morphologies.

Conclusion
The results of this study have implications for the definition of minimum measurable tumor adopted by clinical trial and staging protocols.

Background
Development of targeted therapies in malignant pleural mesothelioma has been limited due to the incomplete understanding of genetic aberrations occurring in this disease. We determined mutational and copy number (CN) events in a comprehensive fashion in 12 patients with epithelial malignant pleural mesothelioma in order to identify characteristic genetic aberrations and new potential treatment targets.

Methods
Fresh frozen tumor (≥70% tumor content) and matched normal tissue from 12 patients with treatment-naïve epithelial malignant pleural mesothelioma were evaluated using targeted, massively parallel sequencing (Illumina HiSeq) for 580 cancer-relevant genes using hybrid capture target enrichment and an established bioinformatics analysis pipeline. CN analysis was performed using sequencing data (CONTRA) and validated on the NanoString nCounter. Mutations were modeled bioinformatically using the CHASM oncogenic driver prediction algorithm and reviewed for prior occurrence in COSMIC and evidence supporting usefulness as a treatment target.

Results
Loss of CDKN2A was the most common genetic event occurring in 80% of samples. INPP4B and TRAF6 were lost in 20-30% of tumors. No loss of PTEN was observed in any sample. Two BAP1 small frameshift deletions were observed. Copy number aberrations, in particular amplifications, were rare. The only gene with modest increase in copy number was MET (median CN=2.8). We identified multiple somatic missense mutations including: two APC mutations (K1245E, G2502D; both predicted to be damaging), one TSC1 (K587A) and one TSC2 (P952A, predicted driver) mutation, one KIT mutation (L799F; predicted driver), one JAK2 (P953A; predicted driver), as well as single mutations in NF1 (G849R), EPHA1 (P697S), EPHA4 (T532I), and mTOR (L2208P), all of which had driver character. Specific mutations have not been reported in COSMIC. No canonical PI3K or MAPK pathway aberrations were identified.

Conclusion
We identified several novel and known mutations and copy number events in epithelial malignant pleural mesothelioma: deletions in CDKN2A and INPP4B were the most common copy number events. Frameshift deletions in BAP1 were identified. Overall few copy number events were observed. Potentially targetable aberrations include missense mutations in NF1, JAK2, EPHA1, TSC2, mTOR and KIT, which are predicted to have driver character and additional experimental validation is indicated. Additional tumor and cell line samples are being processed and will be available at the time of presentation.

Background
Prior to the demonstration of the efficacy of cisplatin/pemetrexed chemotherapy, Malignant pleural mesothelioma (MPM) was previously considered a chemotherapy-resistant tumour. In order to assess the efficacy of chemotherapy in the pre-operative setting, we aimed to: (1) determine the radiological and pathological complete response (CR) rate; (2) evaluate any potential association between pathological CR and patient characteristics; and (3) assess if pathological CR is associated with a longer disease free survival (DFS) and overall survival (OS) in MPM patients.

Methods
A retrospective review of a prospectively collected database of MPM patients treated with induction chemotherapy followed by extrapleural pneumonectomy (EPP) at our institutions since 2003 were performed. Radiological response was determined by CT/PET scans while pathological response was assessed by examination of the EPP specimen. OS and DFS from the day of EPP was determined by the Kaplan Meier method and comparison was made by log rank test. Chi square tests were used to determine potential association between pathological CR and patient characteristics.

Results
Forty-two patients are included: median age 63 years; 81% male. Pathology was 91% epithelial and 9% biphasic subtype. A median of 3 cycles of induction chemotherapy was administered with either 67% cisplatin-based and 31% carboplatin-based treatment. The regimen in the remaining 2% of patients was unknown. All patients proceeded to EPP and 86% of patients received adjuvant radiotherapy. The median OS was 30.6 months (95% CI: 2.9 – 58.2 months), while the median DFS was 19.6 months (95% CI: 11.0 – 28.2 months). No-one achieved a radiological CR. Four patients (9.5%) had pathological CR and all were male (p=0.31) and had epithelial subtype (p=0.50). No association was found between pathological CR and the type of chemotherapy (cisplatin vs. carboplatin; p=0.76) or the number of cycles administered (≤3 vs. >3; p=0.64). Median OS was 30.6 months vs. median not reached (p=0.31), while median DFS was 19.4 vs. 36 months (p=0.34), for those without pathological CR and those with pathological CR respectively.

Conclusion
A 9.5% pathological CR rate was demonstrated after induction chemotherapy, indicative of chemo-sensitivity in a subgroup of MPM patients. There was a trend for longer OS and DFS in MPM patients achieving a pathological CR.

Background
The aim of this study was to describe the longitudinal picture of Health Related Quality of Life (HQOL) in people with Malignant Pleural Mesothelioma (MPM) post Extrapleural Pneumonectomy (EPP).

Methods
Participants receiving EPP from 2011- 2013 were assessed pre-operatively, pre and post adjuvant radiotherapy (Rt), and at 8, 12 and 24 months following surgery. Here we report Global HQOL and HQOL Domain Scores of the EORTC QLQ-C30, and Fatigue Scores from FACT-F. Least squares means were obtained from a mixed models analysis with time as a fixed effect, the pre-op assessment as a covariate and a random subject effect.

Results
Twelve men with a mean age of 65 years (range 48-78) completed pre-op and at least one post op assessment. Table 1 and Figure 1 report the mean HQOL domain scores, global HQOL and fatigue at baseline as well as the least squares mean and 95% confidence intervals at each follow up assessment. Table 1. Health related quality of life over time Figure 1 Figure 2 Figure 1: Health related quality of life over time These results suggest that people who elect to have EPP have baseline levels of HQOL comparable to the general population. As expected, HQOL declines after surgery and during adjuvant radiotherapy. Emotional functioning changes least, while physical and social functioning closely mirror each other. Role functioning is the domain most affected and remains low out to 24 months. Global HQOL is relatively stable over time, with an apparent increase at 24 months. Fatigue is worst at the conclusion of radiotherapy and gradually improves.

Conclusion
People electing to have EPP report a sudden decline in HQOL, with the nadir around the end of adjuvant radiotherapy. This gradually improves over time, returning to slightly below baseline in many domains.

Background
Current evidence-based guidelines for management of malignant pleural mesothelioma (MPM) suggest the optimal rate of chemotherapy utilisation should be around 65%. Reported rates of chemotherapy utilisation in Australia are about 54%. This reflects an under-utilisation rate of around 11%; i.e.,11% of those suitable for chemotherapy do not receive it. Here we explore patient, caregiver and health professional perceptions of chemotherapy treatment for MPM, as well as potential barriers and facilitators to accessing treatment.

Methods
Semi-structured interviews were conducted with people with MPM, caregivers and health professionals, in person or via telephone, to explore their perception of chemotherapy treatment for people with MPM. The analysis was informed by a framework approach.

Results
Overall 68 interviews were conducted: 16 people with MPM, 14 caregivers, 13 medical oncologists, 4 radiation oncologists, 7 respiratory physicians, 3 cardiothoracic surgeons and 11 lung cancer nurses. Health professionals reported varied responses to our estimate of the chemotherapy under-utilisation rate in people with MPM: some suggested it was likely much lower and others higher. Patients and caregivers demonstrated mixed views about the value of chemotherapy. Many reported choosing chemotherapy because they felt the need to “do something”; others stated that starting and/or continuing chemotherapy was their family’s preference, not their own; some believed that their medical oncologist made the decision about treatment. Those electing not to have chemotherapy were satisfied with their decision. Barriers to accessing chemotherapy included: patients’ nihilistic views regarding chemotherapy (no benefit, makes you unwell, other people have done poorly), distance from treatment centre; health professional perceptions such as nihilistic views regarding chemotherapy (no benefit, high toxicity), patient frailty, concerns about overtreatment, cases not presented at multi-disciplinary team meetings (MDTM), low volume MPM centres, lack of availability of carboplatin through Pharmaceutical Benefits Scheme. Facilitators to accessing chemotherapy included: MDTM discussion, treatment in specialist centres, policy to refer all MPM patients to a Medical Oncologist, and use of clinical practice guidelines. Other suggestions to improve access to chemotherapy for MPM included: increased access to specialist lung cancer nurses and allied health professionals, General Practitioner education regarding asbestos exposure history and investigating recurrent pleural effusions. People with MPM and caregivers highlighted the importance of information from health professionals and from others who have experienced these treatments.

Conclusion
Reasons for lower than expected rates of chemotherapy utilisation in people with MPM are multifaceted and complex to address. These qualitative results will inform the development of quantitative surveys of people with MPM, caregivers and health professionals to further investigate barriers and facilitators to chemotherapy treatment to guide the development of potential interventions.

Background
Primary mediastinal tumors (PMT) can be neoplastic or non-neoplastic lesions and originate mostly from thymus. We investigated epidemiology of these tumors and took into consideration age and gender of patients, histologic type of tumor and number of cases per year.

Methods
1615 cases clinico-radiologically defined as „mediastinal tumor” and diagnosed in National Tuberculosis and Lung Diseases Research Institute in Warsaw from the half of 1999 to the beginning of 2013 was analyzed. The diagnosis was establish by examination of totally resected tumor or biopsy specimens gained by mediastinoscopy or EBUS. WHO (2004) histological classification of tumors were applied to analysis.

Results
Material was obtained from 853 (53%) women (median age 49 years) and 762 (47%) men (median age 53 years). 375 metastatic carcinomas and 217 non-diagnostic cases were excluded from further examination as not „primary tumor”. There were 298 cases (27% of all PMT) of non-neoplastic thymic changes, 282 (26%) lymphomas, 147 (14%) thymomas, 64 (6%) neurogenic neoplasms, 44 (4%) thymic carcinomas, 44 (4%) thyroid lesions, 41 (4%) germ cell tumors, 39 (4%) cysts and 125 (12%) other tumors. Non-neoplastic thymic changes were dominated by follicular hyperplasia (249, 84%) and was observed mostly in women in 2nd - 4th decade. They constituted the most frequent PMT in women. Over half of all lymphomas (51%) were classical Hodgkin lymphoma, followed by primary mediastinal lymphoma (28%). Both types occurred mainly in young women ( the 3rd - 4th decade). The second peak of incidence in late life characteristic for Hodgkin lymphoma was not observed. Lymphomas were the most common PMT in men. The most frequent types of thymomas were AB (28%), B1 (17%) and B2 (11%). 25% of all tumors was classified as „combined”. Thymomas occurred more common in women (56%). Median age was 57 years. Neurogenic neoplasms were mostly schwannomas (58%), neurofibromas and ganglioneuromas (each 16%). The tumors were diagnosed more often in middle aged (3rd - 6th decade) female patients. Median age in thymic carcinomas were 58 years. They appeared almost equally in both genders. Group of thyroid lesions include retrosternal nodular goiter (80%), thyroid carcinomas (18%) and one thyroid lipoadenoma (2%). They occurred three times more often in women. Peak of incidence fell on the 6th decade. Germ cell tumors concerned mostly young men (M:F = 5:1) beneath 30 years old. There was no gender predylection among patients with mediastinal cysts. Median age was 52 years. The cysts were usually of thymic origin. The group of other tumors comprised sarcomas, haemangiomas, pleural lesions and enlarged mediastinal lymph nodes with granulomatous or reactive changes.

Conclusion
Both number of cases of all mediastinal tumors and percentage of PMT showed growing tendency from 2000 to 2012. The most frequent PMT in women are non-neoplastic thymic lesions (hyperplasias), followed by lymphomas, thymomas, neurogenic neoplasms and thyroid changes. In men, lymphomas are the most common, then are non-neoplastic thymic changes, thymomas, germ cell tumors, neurogenic neoplasms and cysts.

Background
WHO histological classification of thymic tumors (2004) distinguishes five main histologic types of thymomas: A, AB, B1, B2 and B3 and several special types. The classification is not easy and the diagnosis is not reproducible, because thymomas are heterogeneous and often include several different components or the components of borderline types. The evaluation is very subjective and more objective criteria should be found. D2-40 is a monoclonal antibody that recognizes podoplanin expressed by lymphatic vessel endothelium, mesothelial cells, germ cell tumors and stromal cells of lymphoid tissue. Some authors observed, that expression of podoplanin is a prognostic factor in thymomas but is less important in histological typing of the tumor.

Methods
30 thymomas were examined immunohistochemically by reaction with D2-40 antibody. Morphology of tumors was established according to WHO histological classification of thymic tumors (2004), stage was based on criteria of Masaoka staging system. Reaction was assessed only in epithelial cells as positive or negative, membranous or cytoplasmic. When the recognition of epithelial cells was problematic, AE1AE3 antibody was involved. Normal thymic tissue remnants, lymphatic vessels or lymphoid stroma with reactive lymphoid follicles were treated as positive internal control of reaction.

Results
The group of thymomas consisted of type A - 3 cases, AB - 4, B1 - 5, B2 - 9, B3 - 2, 4 combined thymomas: B2B3 - 3 cases, B1B2 - 1 and 3 cases of special types: 2 micronodular and 1 metaplastic thymoma. The tumors were staged as I in 4 cases, II - 17, III - 3 and IV - 3 cases. Podoplanin was expressed in 1 (33%) type A, 1 (25%) AB, 3 (60%) B1, 9 (100%) B2, 0 (0%) B3, 3 (100%) B2B3, 1 (100%) B1B2, 0 (0%) micronodular and 1 (100%) metaplastic type. B2 and B1 types, B2B3 and B1B2 thymomas showed membranous reaction, other types (A, AB and metaplastic) - cytoplasmic. The intensity of reaction was diverse from inconspicuous to distinct but usually weaker than observed in internal control. Median area of tumor that expressed podoplanin ranged between 0 and 40%: A - 0%, AB - 0%, B1 - 1.5%, B2 - 30%, B3 - 0%, B2B3 - 40%, B1B2 - 20%, micronodular - 0% and metaplastic - 2%. 3/4 (75%) tumors in stage I, 8/17 (47%) tumors in stage II, 2/3 (67%) tumors in stage III and 3/3 (100%) tumors in stage IV (100%) revealed positive reaction. Median positive area of tumor was 17,5 % for stage I, 1% for stage II, 2% for stage III and 17% for stage IV.

Conclusion
Our preliminary observations revealed that positive membranous reaction with D2-40 present in epithelial cells that concerns over 20% of tumor area strongly suggests B2 component. The reaction can facilitate the recognition of this high-grade thymoma. Positive reaction in lymphoid stroma or thymic remnants should not be treated as diagnostic, but can be useful as a positive internal control of reaction. Cytoplasmic reaction is not characteristic for B2 type. We did not observed any relationship between D2-40 expression and stage of tumor.

Background
In India, thymoma constitutes 0.1 to 0.2% of all malignancies whereas National Cancer Institute of USA reports 0.2 to 1.5%. The aim was to audit patients treated for thymomas and to associate various prognostic variables with survival and compare the same with western population studies to mark out any difference.

Methods
Retrospective review of 23 patients with thymoma treated from July 2008 to October 2012 was done. The various prognostic variables such as Mosaoka stage, WHO type, tumor size, extent of resection, presence of myasthenia gravis (MG), and treatment modalities were captured. Statistical analyses were done using SPSS and Kaplan-Meier method.

Results
The patients aged <50 years was 60.8% (n=14) with a male preponderance [73.9% (n=17)] and MG was present in 29.6% (n=8). Commonest stage at diagnosis was Mosoaka III-43.4% (n=12). Commonest type was WHO type AB and B2 [34.7% (n=8) in each]. Among patients who underwent surgery [91.3% (n=21)], either upfront or following neoadjuvant chemotherapy, R0 resection was in 47.8% (n=11), R1 in 30.4% (n=7), R2 in 13% (n=3) and only biopsy in 2 patients in stage IV. Postoperative radiation was given to 82.6% (n=19) when indicated (tumour more than 6 cm in stage I disease and adjuvant in stage II, III and IV). Conformal radiotherapy/IMRT was given in 60.8% (n=14). Chemotherapy with either platinum or anthracyclines or both given either as neoadjuvant, adjuvant or palliative in 56.5% (n=13) and second line chemotherapy with Docetaxel, Capecitabine, and or Everolimus was done for recurrence or progression. Three of 7 patients with MG who received radiotherapy had myasthenic crisis with treatment breaks. Three patients developed acute radiation pneumonitis, and 3 patients had associated pure red cell aplasia. With a median followup of 32 months (range: 3-66), recurrence rate was 26% (n=6) and OS 87%. The median time to recurrence or progression was 27 months (range: 1-66). In univariate analysis, the gender, presence of MG, stage, and extent of resection were significant prognostic factors for progression free survival. The presence of MG and stage were significant prognostic factors for overall survival.

Variable	Disease free survival Mean (months)	Univariate (p value)	Overall survival Mean (months)	Univariate (p value)
Age (years)		0.10		0.338
<50 years	27.5		29.1
>50 years	27.2		32.6
Gender		0.003		0.067
Male	30.17		33.7
Female	19.5		21.3
Myasthenia Gravis		0.002		0.031
Present	25.8		26.1
Absent	28.2		32.8
Stage		0.001		0.008
I AND II	20.1		21.5
III AND IV	33.0		37.4
EXTENT OF RESECTION		0.052		0.395
R0	25.9		27.1
R1	24.7		26.2
R2	40.3		43.0
Biopsy	25.5		45.0
Tumor size		0.193		0.508
<6 cm	29.7		34.6
>6 cm	26.1		28.3

Conclusion
Compared to western population studies, our patients are younger, advanced stage at diagnosis and are with male preponderance. The influence of prognostic variables such as stage and presence of myasthenia and the progression free survival was comparable with that of published series. Our study could not demonstrate the extent of resection as a significant prognostic factor.

Background
Adenoid cystic carcinoma (ACC) is a malignant neoplasm that usually arises in the salivary glands, but can also occur in the breast, skin, uterine cervix, upper digestive tract, and lung. Primary pulmonary ACC accounts for less than 0.2% of all lung cancers, and only 10% of all primary pulmonary ACCs are peripheral in origin. To the best of our knowledge, peripheral pulmonary ACC detected prior to its appearance in the tongue has not been reported previously in the literature.

Methods
We report herein a case of peripheral ACC in the right lower lobe, which was followed 27 months later by tongue ACC.

Results
A 52-year-old woman was referred to our hospital for evaluation of an incidentally found right lower lobe lung mass. She had a history of hypertension and denied recent weight loss or smoking. Her initial laboratory findings were unremarkable, and the results of a pulmonary function test were normal. A chest computed tomography (CT) scan revealed a relatively well-circumscribed and enhanced mass involving the visceral pleura in the right lower lobe, and no mediastinal lymphadenopathy. The mass was biopsied under video-assisted thoracoscopic surgery, and an intraoperative pathologic diagnosis revealed malignant lung cancer. The patient underwent right lower lobectomy and mediastinal lymph node dissection. A histopathological examination disclosed ACC with a cribriform and tubular pattern; metastasis was not found in the dissected lymph nodes. For confirmation of a primary or metastatic tumor, an otolaryngologic examination and whole-body positron-emission tomography (PET) were performed after lobectomy. The patient was diagnosed with primary peripheral pulmonary ACC because of no evidence of salivary gland tumor or other metastasis; Neither radiotherapy nor chemotherapy was not performed after lobectomy. During follow-up, the patient was diagnosed and treated for transitional cell cancer of bladder. Recurrence of the bladder cancer was suspected at 27 months after the pulmonary operation, and a whole-body PET scan showed that an enlarged and increased uptake pattern of both parotid nodules without recurrence in the lung. The bladder nodule was benign and the parotid nodules disclosed ACC of the tongue with metastasis to the parotid lymph node. The otolaryngologist performed wide excision of the tongue cancer. The pathologic diagnosis was ACC of the tongue; identical to those of the resected right lower lobe lung mass taken 27 months previously. After 5 week-adjuvant chemo-radiation therapy, she remained in good health and with no detectable recurrence at the 7-months follow-up.

Conclusion
We have presented herein a case of ACC with unusual clinical behaviors that remain to be fully defined. Although ACC is characteristically slow growing and associated with late distant metastasis, our case showed the reverse presentation pattern: early recognition of a metastasis and late presentation of the primary site. Therefore, peripheral pulmonary ACC should be carefully followed and considered in relation to head and neck ACC, even after successful management.

Background
BACKGROUND: BNETs consist of typical carcinoids (TC), atypical carcinoids (AC) and large cell neuroendocrine carcinomas (LCNEC) based on cellular differentiation, mitotic count and the presence or absence of necrosis [Travis 2004]. Since 2009, NETs are included in the TNM-classification for lung cancer [Sobin 2009]. Whilst increasing evidence points towards the prognostic value of the Ki67 index (a measure of the proliferative capacity [Rugge 2008, Grimaldi 2011]), it is unclear which classification method provides the best prognostic stratification. In this series, the prognostic value of clinicopathological characteristics was assessed, including differentiation, stage and proliferation.

Methods
METHODS: Medical records of pts with a histologically proven BNET, treated at Ghent University Hospital between January 2001 and June 2012, were retrospectively reviewed for clinical, staging, histopathological and treatment data. Overall and disease-specific survival (DSS) were estimated and correlated to clinicopathological features using uni- and multivariate analysis.

Results
RESULTS: Information on differentiation and staging was available in 55 of 57 retrieved pts (96.5%). Tumors were mostly TC (49%) or LCNEC (42%). Twelve pts (22%) presented with stage IV disease, 7 (13%) stage III, 9 (16%) stage II and 27 (49%) stage I. The Ki67 index was assessed in 14 tumors (25%), with a median of 12.5% (2.0 – 95.0%). Ki 67 levels were lowest in TCs and highest in LCNECs (p = 0.014). Forty-one (72%) patients underwent a resection, which was R0/1 in 39. Outcome data are in the table. A strong association was found between differentiation and stage (p <0.001). With a median follow-up of 25 months (0.0 – 132.0), 17 pts (29.8%) died from a tumor-related cause. Male gender (p = 0.025), peripheral tumor (p = 0.04), LCNEC (p < 0.001), higher Ki 67 index (p= 0.03) and stage IV (p < 0.001) were significant predictors of lower DSS in univariate analysis. Ki 67 index was not included in the multivariate analysis, which identified tumor stage as the independent predictor of DSS. Pts with stage IV disease had a 27-fold (95% CI 2.7 – 263.4, p = 0.005), those with stage III disease a 12-fold (1.3 – 113.1, p = 0.026) increased risk of dying from their BNET compared to pts with stage II disease. None of the 27 pts with stage I disease died from a tumor-related cause.

	Median (m)	1y SR (%)	3y SR (%)	5y SR (%)	10y SR (%)
Disease-specific survival	128	81.5	67.3	67.3	67.3
Overall survival	128	76.6	63.2	63.2	63.2

Conclusion
CONCLUSION: In this series of BNETS, tumor stage is the only independent predictor of prognosis. Further research is needed to assess the prognostic value of Ki67.

Background
Extragonadal germ-cell are rare, most of them arising in the anterior mediastinum and comprise less than 5% of all germ-cell malignancies, Although they also occur in women, the men are the most affected. This tumors are considered with a worse prognosis than their primary gonadal counterparts. Standar treatment includes chemotherapy with bleomycin, etoposide and cisplatin, sometimes followed by surgery. Five-year overall survival is about 45%

Methods
Retrospective, single institution review from January 2005 to december 2012. We included patients with mediastinal germ-cell tumors treated with chemotherapy plus surgery. We reviewed patient characteristics to identify factors associated with overall survival.

Results
We included 28 patients in the final analysis, there were 25 males (89.3%) and 3 females (10.7%), mean age 26.71 years (range 15 -39 years). Mediastinal tumor and elevated serum tumor markers were present in 22 patients; in 6 patients a pre-treatment biopsy was needed because serum tumor markers were negative. Non-seminomaotus germ-cell tumor was diagnosed in 22 patients, seminoma in 2 patients and mature teratoma in 4 patients. Except for the teratoma patients, all received preoperative chemotherapy, BEP regimen was used in 53.6% of patients. A second line of chemotherapy was used in 8 patients (28.6%) and only 1 patient received a third line of chemotherapy. Serum tumor markers prior to surgery were negative in 19 patients, meanwhile 9 patients underwent surgery with positive serum tumor markers. Surgical approach was made by median esternotomy in 15 patients, and posterior thoracotomy in 11 patients. Extensive resections (including lung resection and/or pericardial resection and/or bone resection) were performed in 16 patients. Perioperative mortality was 3.6% (one pneumonectomy patient). Median size tumor was 13.3 cm (range 4-25 cm) Complete resection was achieved in 23 patients (82.1%). Pathology in surgical specimens were viable germ-cell tumor in 12 patients (42.9%), mature teratoma in 6 patients (23.4%), inmature teratoma in 4 patients (14.3%) , necrosis in 4 patients (14.3%) and seminoma in 2 patients (7.1%). Only 7 patients (25%) received an additional line of postoperative chemotherapy. Median overall survival was 18.32 months (range 1-102 months). Seminoma and teratoma were associated with longer survival, meanwhile in non-seminoatous tumors preoperative negative serum tumor markers showed a better prognosis.

Conclusion
Surgical outcomes have improved over time with a positive impact in overall survival. Although the prognosis still worse for this subgroup of patients with extragonadal germ-cell tumors, our results indicates that a proper selection with agressive surgery entitles a better prognosis with low mortality. Factors associated with better survival were histology, serum tumor markers status prior to surgery and pathologic report of the surgical specimen.

Background
Pulmonary hemorrhage (PH) is a serious adverse event for patients treated with bevacizumab (BV). Previous studies have identified PH risk factors as tumor cavitation, location, and endobronchial invasion confirmed by computed tomography (CT). However, for endobronchial invasion, we believe confirmation should be judged by bronchoscopy. The aim of this study is to demonstrate the relevance of bronchoscopic findings for patients with PH and treated with BV.

Methods
Retrospective analysis of non-small cell lung cancer was performed on patients treated with combination therapy including BV, as a first line chemotherapy at St. Marianna University Hospital between April 2010 and June 2013. Clinical data were retrieved from medical records and criteria from previous studies were used to identify tumor locations. Bronchoscopic findings were classified as follows; epithelial, subepithelial, extraluminal, and normal.

Results
Of the twenty-nine patients analyzed in this study, 24 patents underwent bronchoscopy before BV treatment. The median age was 62 years (range 38-78), and adenocarcinoma was confirmed in all patients histologically. PH was present in 10.3% patients (3/29, all Grade1), and the location of tumors (central vs. peripheral), was not a significant risk factor for PH (p=0.63). Bronchoscopic classification of patients for epithelial, subepithelial, extraluminal, normal were; 0, 18, 0, 6, respectively. Dilatation findings of subepithelial vessels were seen in 2 cases (2/24). There was no significant difference for PH in bronchoscopic classifications (subepithelial vs. normal, p=0.38); however, patients with dilatation findings of subepithelial vessels were at significantly higher risk for PH (p=0.01).

Conclusion
It might be possible that patients were safely treated with BV in spite of central lesions confirmed by CT. However, dilatation finding of subepithelial vessels should be observed carefully under bronchoscopy, since these findings may predict PH risk factors for BV.

Background
Although the utility of virtual bronchoscopy has been reported, the software for virtual bronchoscopy has not been popular due to the high cost. OsiriX® is a reasonably priced software that is available to reconstruct virtual endoscopic images. Herein, we present the ability of OsiriX to enable virtual bronchoscopy.

Methods
Computed tomography (CT) of the chest was performed using a 16-row multidetector. Data in 2 mm slices from one lung were obtained in 10 patients with a lung nodule. Virtual bronchoscopic images were established by OsiriX version 5.5 (32-bit). To examine the ability to visualize small bronchi, we tried to visualize the distal bronchus possible. We selected B[1a] and B[10c] for the right lung and B[1+2a] and B[10c] for the left lung. In addition, to predict whether a pathological diagnosis can successfully made via transbronchial lung biopsy, we reconstructed virtual bronchoscopic images toward the lung nodule.

Results
Bronchoscopic images were successfully reconstructed for all patients. It takes several seconds to convert CT images into bronchoscopic images. The third to the seventh bronchi were visualized, except in one patient whose right B[10] was occluded by a tumor. The smallest bronchial diameter visualized was approximately 1.5 mm. For all cases, the second-order bronchus, such as B[1] and B[10], were easily visualized. To visualize the third and higher-order bronchi, it was sometimes necessary to manually adjust brightness and contrast. The pathway can be recorded and retrieved using the “Fly Through” function. The pathway of the virtual bronchoscopy was reconstructed in only 5 to 10 minutes. In all cases, the path of the virtual bronchoscopy reached the lung nodule (Fig.1). Therefore we predict that bronchoscopic biopsy would have been successful in all cases. Based on transbronchial biopsy, 4 patients were diagnosed with lung carcinoma and 1 patient was suspected of having adenocarcinoma. In all cases, the lung nodules were successfully diagnosed.Figure 1

Conclusion
OsiriX is practicable for virtual bronchoscopy at a low cost.

Background
Linear endobronchial ultrasound guided transbronchial needle aspiration (EBUS-TBNA) is widely used for tissue sampling of mediastinal and hilar lesions. Rapid On-Site Evaluation (ROSE) is a technique where TBNA samples are rapidly processed and screened for diagnostic material intra-procedure. The use of ROSE improves diagnostic yield, cost-effectiveness, and reduces procedural time. Until recently, at Austin Health, EBUS TBNA samples were prepared in the endoscopy room and evaluated off-site. Preliminary results were conveyed to the bronchoscopist via telephone leading to significant delay. We propose that the implementation of ROSE to assist EBUS-TBNA procedures will reduce the number of lymph node aspirates performed without reducing the quality or diagnostic yield.

Methods
Consecutive EBUS-TBNA cases were prospectively evaluated following institution of ROSE. The number of lymph node stations sampled and the number of aspirations per lymph node station were recorded. This was compared to a retrospective dataset of 69 consecutive cases preceeding the commencement of ROSE. Specimen Preparation Material obtained from TBNA was transferred onto numbered slides with at least one air-dried smear and one 95% alcohol fixed smear prepared per puncture. Remaining material was put into a saline pot for cell block preparation. A cytologist's assessment of specimen adequacy and presence of diagnostic material was performed on-site after a rapid H&E stain. Lymph node stations sampled and number of aspirations performed was recorded.

Results
Preliminary results

	Pre-ROSE	Post-ROSE	p value
Number of cases	69	21	-
Median number of punctures	4	4	0.23
Median number of punctures per lesion	2	2.5	0.46
> 1 lesion investigated (%)	66	48

For suspected lung cancer cases, the concordance rate between ROSE and final cytologic diagnosis was 92%. Data collection is ongoing.

Conclusion
Utilising ROSE during EBUS TBNA lead to a non-significant reduction in the number of lymph node stations sampled. However, it did not change the number of punctures performed.

Background
Virtual Bronchoscopy (VB) facilitates localization of small lesions at EBUS Guide sheath(GS). A good quality CT to create the VB can be problematic to obtain in a timely fashion. Most lung cancer patients require PET, hence we explored options of getting the VB CT at PET upon initial referral prior to EBUS GS.

Methods
We compared VB results for consecutive CTs acquired in one of 3 ways. These were: Group1 :reconstructed CT from the PET/CT ( tidal breathing), Group 2: a dedicated low dose inspiratory breath-hold CT as part of PET/CT, and Group 3 conventional diagnostic CT. CTs had volumetric acquisition and 1 mm overlapping cuts were derived giving at least 600 DICOM images per scan. DICOM images were loaded onto a Lung Point® VB device and VB created. This device gives automated results for the number of bronchial divisions visible, and nearest distance to lesion in question after creation of the VB. Data for Dose Length Product (DLP) and CT Dose Index (CTDI) were obtained for each scan. DLP for Group 1 included whole body CT whereas DLP for Group 2 is stated just for the Thorax CT.

Results

	Group 1 Tidal breathing	Group2 Breath hold	Group3 Conventional CT
n	18	17	20
Lesion size (mm)	21.1± 12.0	23.8±11.0	21.6±7.2
Bronchial divisions	3.25±0.9	4.49±1.1, p<0.01	4.55±0.5 p<0.01
proximity to target lesion (mm)	25.8±19.5	6.7±6.3, p<0.001 cf Gp1	13.3±9.2, p<0.02 cf Gp1
DLP	435 ± 31	173 ± 77	341 ±94, p<0.001 cf Gp 2
CTDI	3.9 ± 1.3	4.1 ± 1.9	8.9±2.9, p<0.001 cf Gp2
% lesions reached at EBUS GS	83	85	90
% positive pathology at EBUS GS	58	54	63

In Group 2 unexpected positive mediastinal nodes were seen on PET enabling staging and diagnostic EBUS TBNA confirming malignancy to be done instead of EBUS GS in 3 patients. Poor VB results with the reconstructed CT were due to underlying COPD with small airway closure and hence inability of the software to detect these small airways; (scans acquired in both inspiration and expiration).

Conclusion
Low dose single breath-hold CT at the time of PET gives at least equivalent results to dedicated CT with far less radiation exposure, and is technically superior to reconstructed CT. Higher visualised bronchial number allows closer proximity to lesions by the directed path. Obtaining this CT prior to EBUS GS means no additional high dose CT is required and additional nodal staging results with the PET can make for better procedural decision making.

Background
The diagnosis of peripheral lung lesions (PPLs) continues to be a commom problem in routine clinical practice. Despite bronchoscopy with guidance has evolved from fluoroscopy to endobronchial ultrasonography and electromagnetic navigation recently, conventional flexible bronchoscopy without guidance is still the most widely used because it is more readily available and inexpensive. The aim of this study was to evaluate the diagnostic yield and factors affecting diagnostic yield of flexible bronchoscopy without guidance for peripheral lung lesions.

Methods
We retrospectively reviewed the medical records of all the patients who underwent flexible bronchoscopy at St. Paul’s hospital of the Catholic University of Korea between Jan 2007 and Mar 2013. Two hundred four patients had lung lesions without endobronchial lesion. Among these, 24 patients were excluded due to follow-up loss and additional 11 patients were excluded because the lesions were deemed inactive in clinical follow up.

Results
One hundred sixty-eight patients were enrolled in this study. The overall diagnostic rate was 58.3%. Sensitivity was 43.2% in malignant disease and 78.1% in benign disease. Non-diagnostic results of bronchoscopy were compared with subsequent PCNA or surgical resection. Nodule location and the distance from the pleura did not effect on the diagnostic yield of bronchoscopy. The bronchus sign on CT imaging was present in 45.7% of the patients. The diagnostic yield was significantly higher for lesions with bronchus sign (81.2%) than for lesions not having bronchus sign (39.0%; P<0.001). The diagnostic yield of bronchoscopy increased according to different ranges of lesion diameter; 10 mm or smaller, 11-20 mm, 21-30 mm, 31-40 mm, 41-50 mm, and greater than 50mm were 16.7%, 32.0%, 50.0%, 57.1%, and 58.8% respectively (linear P<0.001). The diagnostic yield was significantly higher for lesions >3 cm (67.5%) than for lesions <3cm (47.9%; P < 0.02).

Conclusion
This study suggests that the presence of bronchus sign and larger (>30 mm) nodule are useful findings to predict outcome of conventional bronchoscopy without guidance.

Background
To determine the feasibility Endobronchial Ultrasound (EBUS) Elastography for mediastinal lymph nodes (LNs)

Methods
Mediastinal LNs in sarcoidosis (n=5) and metastatic LNs in lung cancer patients (n=7) were examined with B-mode ultrasound, Doppler and EBUS elastography. The elasticity distributions were classified into homogenous green pattern, Heterogenous blue-predominant pattern, ad Homogenous blue pattern. The strain ratio (strain of LN/ strain of fatty tissue) was measured in 3 cases of sarcoidosis and 2 cases of metastatic LNs.

Results
In the elastogram, green on the EBUS images correspond to relatively soft tissue and blue correspond to relatively hard tissue. In the elastogram, mediastinal LNs in sarcoidosis of all 5 cases were classified in homogenous green pattern. In the elastogram of metastatic LNs in 7 lung cancer patients, LNs of 5 cases were classified in heterogenous blue-predominant pattern and LNs of the residual 2 cases were classified in homogenous blue pattern. Homogenous green pattern A blue-predominant pattern, either homogenous or heterogenous, supported the diagnosis of malignant LNs. The strain ratio was 1.31, 3.16, and 5.48 in 3 cases of sarcoidosis respectively, and 17.65 and 29.0 in 2 cases of metastatic LNs respectively.

Conclusion
EBUS elastography would be a feasible method to visualize the elasticity of mediastinal LNs.

Background
Rapid on-site evaluation (ROSE) of transbronchial needle aspirates is cost-effective due to its ability to reduce biopsy number and complication rates without compromising diagnostic yield. Use of ROSE during sampling of peripheral pulmonary lesions (PPLs) has not previously been examined. We aimed to determine the ability of ROSE performed on transbronchial brushings of peripheral pulmonary lesions to accurately determine final procedural diagnosis. To determine if use of ROSE impacts on procedural time or procedural complication rates.

Methods
Prospective cohort of patients undergoing radial probe endobronchial ultrasound-guided bronchoscopy for investigation of PPLs. ROSE was performed using a Rapid Romanowsky stain. If ROSE demonstrated diagnostic malignant material the procedure was determined to be successful and no further sampling was undertaken. Non-diagnsotic ROSE assessment resulted in further sampling including transbronchial lung biopsy, and possibly sampling from different locations.

Results
Specimens obtained from 128 lesions in 118 consecutive patients in whom radial EBUS successfully localized a peripheral pulmonary lesion. Final procedural diagnoses included non-small cell lung cancer (n=76), carcinoid (3), metastatic malignancy (n=3), benign inflammatory/infective infiltrate (n=46). Positive predictive value of ROSE for a malignant bronchoscopic diagnosis was 97% (63/65). Two patients had positive diagnoses made on ROSE but final procedural diagnosis was “reactive bronchial cells” however both of these patients were subsequently confirmed to have NSCLC following alternate biopsy procedures. Procedure times were significantly shorter in those in whom ROSE specimens demonstrated malignancy than in those in whom ROSE was non-diagnostic (19+8 minutes vs. 31+11 minutes, respectively. p<0.0001) In four procedures, initial negative ROSE results prompted redirection of sampling from alternate bronchial segments resulting in positive diagnostic tissue being obtained.

Conclusion
ROSE examination of brushings specimen had high positive predictive value for bronchoscopic diagnosis of cancer. ROSE of brushings specimens has the potential to shorten bronchoscopy times, reduce complications and is likely to be cost-effective. It may also improve diagnostic performance via live feedback, allowing proceduralists to redirect subsequent sampling procedures.

Background
Peripheral pulmonary lesions (PPLs) are diagnostic challenges. Computed tomography guided transthoracic needle aspiration (CT TTNA) has high diagnostic sensitivity but also high complication rates[1]. Endobronchial ultrasound guide sheath (EBUS GS) allows confirmation of target localisation but cannot provide guidance to the target. Electromagnetic navigation bronchoscopy (ENB) allows the bronchoscopist to navigate to target without direct vision. We assessed whether ENB could diagnose PPLs that had undergone a non-diagnostic EBUS GS.

Methods
We performed 50 ENB procedures for diagnosis of PPLs between 3/2011-6/2013, 15 after non-diagnostic EBUS GS. ENB data was prospectively collected. ENB (superDimension, Minneapolis, US) was performed through a standard 5.9mm bronchoscope under general anaesthesia through a laryngeal mask airway after pathway planning using iLogic software. Once the locatable guide was close to and correctly aligned with the target, it was removed and replaced by EBUS radial probe (EBUS RP) to confirm target localisation. Samples were then taken with forceps biopsy, cytology brush, and mini bronchoalveolar lavage. If ENB was non-diagnostic patients underwent further investigation. Benign diagnoses were followed up for a minimum of 6 months to ensure a consistent clinical course. Primary outcome was diagnostic yield, procedure time, and complications. Characteristics distinguishing diagnostic from non-diagnostic ENB were assessed using the chi-squared test.

Results
15 patients (mean age 66.67, 9 females, 12 current or ex smokers, mean BMI 25.16kg/m2) with 15 PPLs who underwent non-diagnostic EBUS GS proceeded onto ENB. Lesion location and characteristics were as follows: left (7), upper lobe/lower lobe=11/4, bronchus sign positive (14), soft tissue density/ground glass=14/1. Mean maximal lesion dimension was 25.64mm+/- 12.38mm and mean closest distance from pleura was 12.04mm +/- 12.18mm. Average total procedure time was 56.83 mins +/- 13.71mins with a mean of 4.53 biopsies taken per patient. All except one procedure was performed under general anaesthesia with a laryngeal mask airway. The target was reached in 12 patients. Median closest distance to target was 12.69mm +/- 7.83mm. Target localisation was confirmed on EBUS RP without any manipulation in 10 patients; a further 2 lesions could be localised with minor manipulation. ENB provided a diagnosis in 5 of 15 patients (33.33%): adenocarcinoma (2), squamous cell carcinoma (1), fungal infection (1), organising pneumonia (1). Non-diagnostic ENB underwent the following additional procedures: CT TTNA (7), repeat EBUS GS (1), and surgical biopsy (2). The following conditions were diagnosed: mycobacterial infection (1), adenocarcinoma (4), fibrosis (1), hamartoma (1), non-small cell carcinoma (1), nodular lymphoid hyperplasia (1). There were no complications. Procedural success was independent of lesion size (p=0.378), location (p=0.714), or morphology (p=0.464), but was related to confirmation on EBUS RP without manipulation (p=0.053), and the ability to view the lesion on Maximal Intensity Projection (MIP) view in 360 degrees (p=0.053).

Conclusion
ENB can successfully diagnose PPLs that have been non-diagnostic on EBUS GS. Lesions that can be confirmed on EBUS GS after being navigated to by ENB, as well as those that can be visualised in 360 degrees on MIP view, have a higher chance of success.

Background
MicroRNAs (miRNAs) post-transcriptionally regulate hundreds of gene targets involved in tumorigenesis and miRNA expression profiling is an emerging tool for the early detection of malignancy. We assessed the ability of microRNA (miRNA) expression profiling of bronchoalveolar lavage (BAL) fluids and sputum samples to distinguish early stage non-small cell lung cancer (NSCLC) cases from cancer-free controls.

Methods
The expression levels of 3 miRNAs (miR-21, miR-210, miR-372) were quantified in BAL fluids and sputum, normalized to an endongenous control (U6) relative to a MRC-5 reference sample, using RNA reverse transcription and Quantitative real-time Polymerase Chain Reaction (RT-qPCR). All sputum samples were collected by a single spontaneous expectoration while BAL fluids were obtained just prior to surgical resection. Unsupervised hierarchical cluster analysis was performed on the experimental-normalized miRNA expression profiles using within-group linkage and cosine correlation similarity.

Results
From April 2011 to January 2012, twenty-one eligible cases and 10 controls were entered into this study. The median age of cases was 70 years of which 17 were male and 4 were females. Thirteen cases had adenocarcinoma, five had squamous cell carcinoma, and 3 had large cell carcinoma. Twelve cases had stage I and 9 had stage II NSCLC. The median short axis diameter of the primary tumour amongst cases was 1.6 cm. With exception of one case, endobronchial lesions were not detected on inspection by flexible bronchoscopic examination prior to BAL fluid collection. The vast majority of cases were smokers (20/21). The median age of the control group was 58.5 and five were healthy without active medical conditions while five had COPD. Six controls had prior or current histories of smoking while 4 were never smokers. Cluster analysis of the miRNA expression profiles of BAL samples from 21 NSCLC cases and sputum samples from 10 cancer-free controls yielded a diagnostic sensitivity of 85.7% and specificity of 100%. Cluster analysis of sputum samples from the same 21 NSCLC cases and 10 cancer-free controls yielded a diagnostic sensitivity of 67.8% and specificity of 90%. A cosine similarity analysis of matched pairs of concordant and discordant BAL and Sputum samples was conducted and indicated that sampling error accounted for 6 of the 7 false negative results. This suggests that triplicate sputum sample collection could improve the overall sensitivity of this method for use as a safe, non-invasive screening test for NSCLC or as a pre-screening test to select patients for low-dose CT screening.

Conclusion
Hierarchical cluster analysis of 3 expressed miRNAs obtained from sputum and brochoalveolar lavage samples are highly specific and relatively sensitive methods for the timely diagnosis of early stage NSCLC. Larger, population-based studies are necessary for further validation of this promising approach in both the diagnostic and screening setting.

Background
A new lung adenocarcinoma classification proposed by the International Association for the Study of Lung Cancer, American Thoracic Society and European Respiratory Society (IASLC/ATS/ERS) has recently been published in 2011.We investigated the relationship between predominant subtype, according to the (IASLC/ATS/ERS) lung adenocarcinoma classification and prognosis in stage I lung adenocarcinoma in our hospital.

Methods
Two hundred and sixty-one patients with stage I lung adenocarcinoma operated in Zhejiang Cancer Hospital were identified between 2000 and 2010. Survival curves were plotted using the Kaplan–Meier method. The Cox proportional hazard model was used for multivariate analysis.

Results
The median age of the 261 patients was 59.8 years in current cohort. Of the 261 patients, 175 patients (67.1%) were never smokers, 86 patients (32.9%) were former or current smokers. The surgical procedure employed 245 lobec­tomy, 12 bilobectomy and 4 pneumonectomy. The pathologic stage was IA in 92 patients (35.2%), IB in 169 patients (64.8%).No cases were adenocarcinoma in situ and six were minimally invasive adenocarcinoma. Two hundred and fifty-five cases were invasive adenocarcinoma, in which 80 were papillary predominant, 76 were acinar predominant, 42 were micropapillary predominant, 34 were solid predominant, 19 were lepidic predominant subtypes, and 4 were variants of invasive adenocarcinoma. The 5-year DFS and OS rates for the all patients were 66.0%, 78.2%, respectively. Micropapillary predominant subtype (p=0.037), solid predominant subtype (p=0.035) were predictive of DFS. Patients with micropapillary and solid predominant tumors had significantly worse disease-free survival compared with other subtypes predominant tumors ( p＜0.001).Multivariate analysis revealed that the new classification was an independent predictor of disease-free survival and overall survival (p=0.002 and 0.015).

Conclusion
The predominant subtype in the primary tumor was associated with prognosis in resected stage stage I lung adenocarcinoma.

Background
Seventh TNM classification for lung cancer and IASLC/ATS/ERS classification for pathological subtype of lung adenocarcinoma have been released independently. They have especially changed classification of small sized (≤ 3.0cm) and nodal negative adenocarcinomas. However, when utilizing them simultaneously, it is not clear which factors of each classification are useful to predict malignant potential of tumor.

Methods
We reviewed 154 pT1a-1bN0M0 adenocarcinoma cases resected between 1990 and 2011. Cases with adjuvant chemotherapy were excluded. 154 cases were subdivided into T1a or T1b cases according to 7th TNM classification. On the other hand, according to IASLC/ATS/ERS classification, reviewed cases were also subdivided into adenocarcinoma in situ/minimally invasive adenocarcinoma (AIS/MIA) or invasive adenocarcinoma (invasive Ad). Survival outcomes were evaluated based on these subclassfication methods.

Results
There were 84 AIS/MIA cases and 75 invasive Ad cases. T1a included 77 cases of AIS/MIA and 50 cases of invasive Ad. T1b included 7 cases of AIS/MIA and 25 cases of invasive Ad. T1a included significantly more AIS/MIA cases compared with T1b (p<0.001). The significance was confirmed about disease free survival (DFS) between T1a and T1b (96.7% vs 72.7, p<0.001). Pathological subtype also reached significance in DFS between AIS/MIA and invasive Ad (97.4% vs 85.7%, p<0.001). When combining these classifications, DFS of T1a-1b AIS/MIA (N=84), T1a invasive Ad (N=50), T1b invasive Ad (N=20) were 97.4%, 95.6% and 69.2%, respectively. The difference of DFS between T1a-1b AIS/MIA and T1b invasive Ad were more notable in comparison to single classification method. Relapse-free probability of T1a-1b AIS/MIA was 100%.

Conclusion
Combination of 7th TNM classification and IASLC/ATS/ERS classification were more useful compared with single method classification to predict T1a-1bN0M0 lung adenocarcinoma recurrence. T1b invasive Ad cases were identified as worsen subtypes than T1a-1b AIS/MIA. AIS and MIS might be handled equally regardless of tumor size.

Background
Targeted therapies aimed at specific molecular genetic alterations are revolutionizing cancer treatment, particularly in non-small cell lung cancer (NSCLC). ROS1 is an oncogene that encodes a transmembrane tyrosine kinase receptor that has high homology with the intracellular kinase domain of ALK. Driver mutations involving translocation of the ROS1 gene have recently been identified in NSCLC and show promise as a target for tyrosine kinase inhibitors. In this study we aimed to: (1) Investigate the incidence and clinicopathological features of NSCLCs harbouring ROS1 rearrangements in an Australian population. (2) Investigate the accuracy of immunohistochemistry (IHC) compared to FISH at identifying tumours with ROS1 rearrangements.

Methods
We tested for ROS1 translocations using both a FISH breakapart probe (Zytovision and Abbott Molecular) (≥15% cells with split signals or single green 3' signal considered positive for rearrangement), and immunohistochemistry (D4D6 clone, Cell Signaling Technology). Testing was undertaken on both (1) A retrospective cohort of 316 early stage lung adenocarcinomas in tissue microarrays. (2) A prospective cohort of 42 NSCLC, selected on clinical grounds for mutation testing (eg EGFR/KRAS/ALK negative samples and young age or never/light smoker).

Results
In the retrospective cohort, only 1 case was positive for ROS1 gene rearrangement by FISH (0.3% incidence). ROS1 IHC identified positive staining in 7 (2.0%) cases, including the FISH+ case. ROS1 IHC had a sensitivity of 100% and specificity of 98% for identifying ROS1 gene rearrangements. In the prospective cohort of 42 cases, 4 cases with ROS1 gene rearrangement were identified by FISH and all 4 cases showed positive ROS1 immunohistochemical staining. Of the total 5 cases with ROS1 gene rearrangement, all occurred in adenocarcinomas from female patients with an age range of 33-81 years (mean 58). Four of the five patients were non-smokers and two were of Asian ethnicity. All 5 cases were negative for ALK rearrangements and in the 4 cases where EGFR status was known, they were all wild type.

Conclusion
ROS1 gene rearrangements occur in a very small percentage of lung adenocarcinomas with distinctive clinicopathological features and appear to be mutually exclusive with other driver mutations in the small number of positive cases available for evaluation. Screening with IHC may be a suitable method of reducing the number of cases requiring FISH testing.

Background
Hetertopic bone formation in the lung, or pulmonary ossification, has been regarded so far as an extremely rare, post-mortem finding without much clinical significance. Two types of ossification are discerned in the literature: nodular ossifications, with a smooth and round edge, which are found in the alveoli themselves, and dendriform ossifications, branching through the alveolar septa, which sometimes even contain marrow formation. Nodular ossifications are associated with congestive heart failure, while dendriform ossification is related to inflammatory and interstitial lung diseases.

Methods
In five years time (January 2008 to March 2013), we have diagnosed 17 cases of pulmonary ossifications by awareness of the surgeon and the pathologist. The diagnosis was made in each case after pathological examination. Neither peroperative palpation nor CT by an experienced radiologist were able to differentiate these ossifications from solid tumors.

Results
76% of patients were male with a mean age of 65.7±4.0 years. Pulmonary ossifications occurred predominantly in the left lower lobe (35.3%) and the majority were nodular (64.7%) with a mean diameter of 3.5±0.88 mm. Pathological diagnosis in 9 cases could not reveal any pulmonary neoplasm despite strong pre-and preoperative indications. We have observed no clear association between specific pathologies and the presence of pulmonary ossifications.

Conclusion
These findings suggest that pulmonary ossifications are not as seldom as believed in the past, and may often be overlooked or mistaken for a malignant space-occupying lesion. Further research on the origin and pathogenesis of these ossifications and increased awareness may be useful. PO may be included in the differential diagnosis of a solitary pulmonary nodule and we expect the prevalence to rise in a population that continues to grow older.

Background
The INSIGHT observational study aimed at assessing the management of NSCLC patients with EGFR mutations in clinical practice in Central Europe. As part of this project, pathological findings including molecular testing methods were assessed.

Methods
Fourteen Pathology Departments from 6 Central European countries participated. Between 15 November 2011 and 31 March 2013, EGFR mutations were determined by one of the established standard methods in patients with NSCLC.

Results
Here we report data on 1009 patients who had been enrolled into the INSIGHT study. These patients consisted of 626 (62%) males and 383 (38%) females, 347 (35%) smokers, 452 (46%) former smokers and 182 (19%) never-smokers. Pathological diagnosis was based on histology (41%), cytology (19%) or both (40%) and revealed the following results: 54% non-mucinous adenocarcinomas, 4% mucinous adenocarcinomas, 21% unspecified adenocarcinomas, 9% NSCLC NOS, 7% squamous cell carcinomas, 2% adenosquamous carcinomas, and 2% others. Tumor material was obtained by bronchoscopy (44%), transthoracic needle biopsy (11%), surgery (19%), or other techniques. Specimens were either from primary tumor (88%), lymph node metastases (2.5%) or distant metastases (9.5%). EGFR mutation testing was done by PCR-RFLP (43%), Roche Cobas EGFR mutation test (26%), Sanger sequencing (18%), high resolution melting followed by sequencing (13%) or another method (11%). EGFR mutations were found in 163 (16%) of the patients. Among patients with mutations, the following mutations were found: 12 (7% of mutation-positive patients) exon 18 mutations, 82 (50%) exon 19 mutations including 63 (39%) deletions, 20 (12%) exon 20 mutations including 3 (2%) T790M, 63 (39%) exon 21 mutations including 45 (28%) L858R. Multiple mutations, both common and uncommon, were found in 12 (7%) of the patients. Mutations were found in 8% of smokers, 14% of former smokers and 43% of never-smokers. Mutations rates varied between centers which most likely reflected different patient selection criteria for EGFR mutation testing.

Conclusion
The INSIGHT project demonstrated that EGFR mutation testing by one of the standard tests in patients with NSCLC has been established in participating centers in Central Europe. EGFR mutation distribution is similar to other European and American NSCLC patient populations. This study was supported by Boehringer Ingelheim Regional Center Vienna.

Background
Terminal respiratory unit (TRU) adenocarcinoma, which is proposed as a distinct subset of lung adenocarcinomas, develops in periphery of the lung parenchyma, is similar in cell morphology to type II pneumocytes or Clara cells, is positive for the expression of thyroid transcription factor-1 (TTF-1), and harbors mutations in the gene encoding the epidermal growth factor receptor (EGFR). The clinicopathological characteristics are generally well defined. In contrast, few studies have focused on non-TRU type adenocarcinomas. We herein analyzed clinicopathologic and prognostic findings of non-TRU type adenocarcinomas.

Methods
We selected 244 consecutive patients with lung adenocarcinomas underwent pulmonary resection at Kyoto University Hospital between January 2001 and December 2007. All cases were classified according to IASLC/ATS/ERS criteria. Findings of significant prognostic factors for lung adenocarcinomas prompted analyses of lymphatic invasion, vascular invasion, pleural invasion, tumor grade. We annalyzed immunohistochemical expression of the mucins MUC5B, MUC5AC, as well as TTF-1 using TMA blocks comprising lung adenocarcinoma specimens from the 244 patients. The presence of mutations in EGFR and KRAS was also determined.

Results
TTF-1, MUC5B, and MUC5AC were detected in 219 (89.6%), 75 (30.7%) and 33 cases (13.5%), respectively. Cluster analysis of the protein expression and EGFR and KRAS mutations yielded four classes of tumors as follows: TRU-type (TTF-1(+), MUC5B(-), MUC5AC(-)); Combined-type (TTF-1(+), MUC5B(+) and/or MUC5AC(+)); Bronchiolar-type (TTF-1(-), MUC5B(+) and/or MUC5AC(+)); and Null-type (TTF-1(-), MUC5B(-), MUC5AC(-), EGFR mutations(-), KRAS mutations(-)). TRU-type tumors were significantly associated with non-mucinous adenocarcinoma in situ (AIS), whereas Bronchiolar-type tumors were associated with mucinous AIS. Combined-type cases were intermediate in morphology between TRU-type and Bronchiolar-type cases (P < 0.001). TRU-type was associated with significantly better prognosis (5-year disease free survival rate (5y-DFS-rate) = 75.2%, 5-year overall survival rate (5y-OS-rate) = 83.8%), followed by Combined-type (5y-DFS-rate = 61.7%, 5y-OS-rate = 73.2%), Bronchiolar-type (5 y-DFS-rate = 53.6%, 5y-OS-rate = 53.9%), and Null-type (5y-DFS-rate = 40.0%, 5 y-OS-rate = 40.0%). Multivariate analyses indicated that non-TRU type significantly correlated with poorer prognosis for DFS (hazard ratio (HR) = 1.946; 95% confidence interval (CI) 1.139-3.322; P = 0.015) and OS (HR = 1.933; 95% CI 1.072-3.491; P = 0.030).

Conclusion
The present study demonstrates that lung adenocarcinomas expressing MUC5B and MUC5AC could represent the non-TRU type and have a poorer prognosis than that of TRU-type lung adenocarcinomas. In addition, three distinct subtypes of non-TRU type adenocarcinomas were discovered. There are no effective treatments for patients with non-TRU type lung adenocarcinoma, while patients with TRU-type lung adenocarcinoma can be treated with EGFR inhibitors. Therefore, the characterization of non-TRU type lung adenocarcinomas described here, suggests that patients with this disease can be identified as candidates for receiving treatments that will hopefully be developed in the future.

Background
The predominant histologic subtype according to the IASCL/ATS/ERS classification has been associated with prognosis in patients undergoing curative surgical resection. It is recommended in the IASLC/ATS/ERS classification that histologic patterns present in small specimens be recorded in the final histopathology report. We investigated the relationship between histologic patterns in metastatic sites, overall survival and EGFR and KRAS mutations.

Methods
We identified patients who underwent surgical resection of non-small cell lung carcinoma in metastatic sites from 2000 to 2011. One biopsy site was selected per patient. A pathologist reviewed all slides to confirm the diagnosis of metastatic lung adenocarcinoma, recording all adenocarcinoma histologic patterns present. EGFR and KRAS mutations were scanned by high resolution melting, and confirmed by Sanger sequencing. Clinical data were extracted from the medical records.

Results
The 100 patients comprised 66 males, with a median age of 64 years. 85% had stage IV disease, and 15% had unresected stage III disease with mediastinal lymph node sampling. Most were current or former smokers (79%) of ECOG 0/1 (67%). Just over half the patients received systemic therapy (56%). The overall median survival was 10.2 months (95% CI 8.1 – 12.2 months). Metastatic sites included brain (30%), pleura (25%), bone/skeletal muscle (20%), and lymph nodes (mediastinal 18%; chest wall/neck 7%). It was possible in each biopsy to identify a major histologic pattern (Table 1). For patients receiving systemic therapy, survival was significantly shorter for those with a major component of solid pattern tumour in metastases compared to those with major acinar or micropapillary patterns in metastases (Table 1). No survival difference was noted on the basis of ECOG, stage, EGFR or KRAS mutations. For patients who did not receive systemic therapy, the major histologic pattern showed no association with overall survival (Table 1). Worse ECOG was the only significant factor in determining outcome (ECOG 0/1 vs 2+ – hazard ratio 2.18 (1.14 – 4.16, p=0.018)). EGFR mutations were significantly associated with major non-solid pattern histology in metastases (Fisher’s exact = 0.006). No association was observed by KRAS mutation status (Table 1).

The major histologic component in sites of metastatic adenocarcinoma – overall survival by the use of systemic therapy; presence of EGFR and KRAS mutations (*Comparison across 4 histological types; OS - overall survival, CI - confidence interval, HR - hazard ratio)

Solid Micropapillary Acinar Papillary Comments Major Pattern 50% 20% 29% 1% Mutations present n = 50 n = 20 n = 29 n = 1 n=100 EGFR mutation, n (column %) 2 (4%) 5 (25%) 4 (14%) 1 (100%) Fisher's exact = 0.006 * KRAS mutation, n (column %) 18 (36%) 5 (25%) 9 (31%) 0 Fisher's exact = 0.789* Systemic Therapy Given n = 29 n = 13 n = 13 n = 1 n=56 Median OS, months (95% CI) 9.4 (8.3 - 12.2) 18.9 (11.6 - 24.4) 15.9 (10.7 - 24.7) Solid vs MPA HR 0.33 (0.16 - 0.67, p = 0.002) Solid vs Acinar HR 0.32 (0.15 - 0.67, p=0.003) No Systemic Therapy n = 21 n = 7 n = 16 n = 0 n=44 Median OS, months (95% CI) 4.3 (3.3 - 7.4) 4.7 (1.5 - 11.5) 4.4 (1.9 - 16.9) No significant differences

Conclusion
Our results suggest that patients with a major component of solid pattern tumour in metastatic sites may be more resistant to systemic therapy as evidenced by shorter overall survival in comparison to those with major micropapillary and acinar pattern tumour in metastases. Furthermore, major solid pattern metastases are unlikely to harbour EGFR mutations. These findings require validation in larger patient cohorts.

Background
Lung cancer is the fifth leading cause of cancer in males in Saudi Arabia. As per current World Health Organization (WHO), lung carcinoma is subdivided into small cell and non-small cell carcinoma (NSCLC). The latter compromise 70-80% of lung carcinoma and consists of heterogenous groups that is further divided into adenocarcinoma, squamous cell carcinoma and large cell carcinoma. Due to poor prognosis of lung cancer, there is an increasing need to find molecular biomarkers which can be used for diagnosis, risk stratification, early detection, treatment selection, prognosis and monitoring for recurrence. Increasing interest in adenocarcinoma of lung has been raised lately for various reasons. One reason is the increasing incidence of adenocarcinoma, which is now the most predominant histologic subtype. Other reason is the possible use of targeted therapy in cases showing EGFR mutations or ALK rearrangements. Adenocarcinoma comprise approximately 70% of primary lung cancer in Saudi population. The aim of this study is to review the incidence of EGFR mutation in lung adenocarcinoma in Saudi patients.

Methods
A cohort of 37 primary lung adenocarcinoma diagnosed histologically and confirmed by Immunohistochemistry was collected. DNA was manually extracted from paraffin embedded tissue and was paired with histology-guided tissue macrodissection to target tumor cells. The mutation status of EGFR exons 18-21 was evaluated using Polymerase chain reaction (PCR) and bi-directional sequencing.

Results
EGFR mutation was detected in 10 cases (27%). Of the 10 cases, 80% of mutations (deletions) were located in exon 19 and 10% in exons 20 and 21 respectively. All mutations detected conferred increased sensitivity to tyrosine kinase inhibitors (TKI).

Conclusion
The incidence of EGFR mutation in lung adenocarcinoma in our patients (27%) is slightly higher than western population (15-23%). To our knowledge, this is the first molecular analysis of EGFR gene mutational analysis in lung adenocarcinoma in Saudi Arabia.

Background
Primary mucinous adenocarcinomas (MA) are relatively rare, and the clinicopathological characterisics have remained unclear. The aim of this study was to clarify the clinicopathological characteristics of MA.

Methods
We selected MA from 1450 cases of surgically resected primary lungWe selected MA from 1450 cases of surgically resected primary lung adenocarcinoma. The clinicopathological characteristics of MA (30 cases) were analyzed.

Results
MA showed a high rate (22/30, 73%) of tumor location in the lower lobe. Vascular invasion was observed in 6 cases (20%). Pulmonary metastasis was observed in 5 cases (17%). Lymphatic permeation was present in 1 case (3%). Pleural invasion was observed in no cases. Lymph node metastasis was present in 1 case (tumor size: 75mm, 3%). MA showed a significantly higher rate of cases aged 65 and over, tumor location in the lower lobe and pathological N0 stage cases, when compared with the other of adenocarcinoma. Furthemore, MA displayed a lower frequency of plural invasion, lymphatic permeation, and vascular invasion, and a high frequency of pulmonary metastasis. We compared the frequency of invasive cases in the two groups with respect to their size (tumor size; TS). In MA, the frequency of invasive cases in TS ≦ 3cm, 3cm < TS ≦ 5cm and TS > 5cm was 11% (2/18), 50% (2/4), 62% (5/8), respectively. In the other types of adenocarcinoma, the frequency of invasive cases in TS ≦ 3cm, 3cm < TS ≦ 5cm and TS > 5cm was 89% (918/1027), 98% (274/281), 100% (112/112), respectively. Therefore, even as TS became bigger, MA displayed lower invasive capacity. We compared the frequency of recurrence cases in the two groups.MA showed local recurrence in 3 of 30 cases (10%), no incidents of distant metastasis. The tumor size of all 3 cases showed more than 5 cm. Pulmonary metastasis showed 1 case (3%) in same side, 2 cases (7%) in the both side. MA showed a significantly lower rate of pulmonary metastasis and distant metastasis (P < 0.05), when compared with the other of adenocarcinoma.

Conclusion
The pathogenesis of MA might differ from that of lung adenocarcinoma without MA based on higher rate of tumor location in the lower lobe and recurrence of pulmonary metastasis. MA less than 5cm may be treated as a local disease and could omit mediastinum lymph node dissection.

Background
Currently, 5 driver genes (EGFR, K-RAS, B-RAF, ALK, RET) related to lung cancer have been widely explored and become new targets of NSCLC, however no report has been found in Chinese female lung adenocarcinoma, who were prone to EGFR mutations and therefore, are the targeted populations for TKIs therapy.

Methods
FFPE-tissues from 310 female lung adenocarcinoma adopted in Hunan province or Henan province between 2000 and 2012 were investigated. Oncogenic alterations in newly found 5 driver genes were analyzed. For EGFR, K-RAS and B-RAF mutation detection, ARMS was performed. The reverse transcription and real-time PCR were performed either to detect all ALK and RET fusions using primers specific to known rearrangement or to detect ALK and RET expression by primers specific to TK domain. Sequencing was further applied to confirm the subtypes of fusions.

Results
Among 310 samples, 149 (48.1%) EGFR mutations, 16 (5.2%) KRAS mutations, 0 (0%) BRAF mutations, 23 (7.4%) ALK fusions and 5 (1.6%) RET fusions were detected. Deletions in exon 19 and L858R in exon 21 account for the major EGFR mutations, representing 97 (65.1%) and 41 (27.5%), respectively. Only EML4-ALK fusion but no other ALK fusions were found. Further research demonstrated that cases (22, 95.6%) with high ALK expression were often accompanied by EML4-ALK fusion and poorly differentiated adenocarcinoma. Two RET fusions, namely, KIF5B-RET and CCDC6-RET were found, with 2 cases and 3 cases, respectively. The ratio of RET/ABL mRNA levels was significantly higher in CCDC6-RET samples with poorly differentiated adenocarcinoma. However, with KIF5B-RET fusion, the situation was more complex. Relatively high RET/ABL mRNA expression was detected in one KIF5B-RET sample with moderately differentiated adenocarcinoma, while no RET expression was detected in other two KIF5B-RET samples with highly differentiated adenocarcinoma.

Conclusion
This is the first research about the oncogenic alterations of 5 important driver genes in Chinese female lung adenocarcinoma, as well as the correlation between ALK fusion and ALK expression or between RET fusion and RET expression, thus laying good foundation for understanding the genetic mutation spectrum in female lung adenocarcinoma.

Background
Stage of lung cancer (TNM) is most valuable prognostic factor. Most of patient in IA and IB stage survive 5 years, but in 10% of them the relapse of the disease is noticed. Therefore identifying the negative prognostic factors is crucial. The neoplastic emboli in blood and lymphatic vessels seem to be one of them.

Methods
The retrospective analysis of lung cancer patients operated on from 2002 thru 2006 at our institution was done. In all patient included into the study group the IA or IB stage was diagnosed and the neoplastic emboli in both blood and lymphatic vessels of the tumor were found. The survival rate in this group was compared with those without such findings. Among 1648 patient operated on between 2002 and 2006 564 were classified as IA or IB (pN0) stage (after rejecting 22 of them due to R1). Among 271 pts during microscopic detection neoplastic emboli in blood or lymphatic vessels were found. The analyzed group consisted of 208 women and 356 men. The stage IA was established in 212 pts, including 115 in pT1a, 97 in pT1b. Stage IB was diagnosed in 352 pts (pT2a) among whom 149 have infiltration of parenchymal pleura detected. All of them underwent the radical resection – all surgical margins were negative. The pathological examination revealed adenocarcinoma in 230 pts, squamous cell lung cancer in 203 pts, large cell lung cancer in 30, carcinoid in 80 pts and other kind of NSCLC in remaining 21 pts.

Results
5-year survival rates recorded were as follow – in stage IA 79,9% and 74,1% in pT1A pT1b respectively, in stage IB (pT2a) 59,9%. The presence of neoplastic emboli in blood and lymphatic vessels of tumor correlated with poorer prognosis of survival (V=0 L=0 – 72%, V=1 L=0 63,3%, V=0 L=1 64,7%, V=1 L=1 58,1%).

Conclusion
Therefore we concluded that such finding should be considered a negative prognostic factor and ought to be recorded in every pathology report.

Background
Patients with lung adenocarcinoma carrying the ALK gene rearrangement show dramatic response to ALK TKIs (e.g. crizotinib). The currently approved method for detection of ALK gene rearrangements is fluorescence in situ hybridization (FISH), however gene sequencing can also be used for detecting ALK rearrangement. Aberrant ALK protein expression, as ALK is not normally expressed in the lung, can be detected with immunohistochemistry (IHC). Our experience has shown a high rate of false negative ALK FISH patients, therefore we retrospectively investigated 53 cases with IHC and Next Generation Sequence (NGS).

Methods
53 patients with NSCLC adenocarcinoma were tested for ALK rearrangement by FISH (Oncotest-TEVA Ltd, Israel). Retrospective IHC (D5F3 antibody, Cell Signaling) was done at the University of Colorado Cancer Center. Discordance cases were sequenced by "FoundationOne” (Foundation Medicine Inc).

Results
Out of the 53 cases, 4 (7.5%) were FISH positive, and 8 (15%) were IHC positive with 3 cases both FISH+ and IHC+ (Figure 1 , Table 1). One specimen was FISH+ and IHC-. NGS was performed on the 5 IHC+/FISH- mismatched samples and found 4 out of the 5 cases positive for ALK rearrangement. The true positive incidence of ALK rearrangement in our cohort increased by 100%, from 7.5% to 15%. Interestingly, two patients were found to harbor a unique ALK rearrangement at intron 19. One of them showed a dramatically improvement to crizotinib (PFS=18 months).

FISH	IHC	Patients	NGS	ALK rearrangement +
+	+	3		3
-	-	44		0*
+	-	1(1.8%)		0
-	+	5(9.4%)	4	4
Total		53	4	7

Table 1 - FISH and IHC summary Figure 1

Conclusion
The current FISH based approach to detect ALK gene rearrangement misses numerous (50%) patients who might benefit from the most efficient lung cancer therapy for their disease. Screening for ALK protein expression by IHC may identify ALK positive patients that would otherwise be missed. Borderline IHC staining should be further sequenced by NGS to cover other abnormalities such as intron19 or other uncommon rearrangements.

Background
Pulmonary neuroendocrine cells hyperplasia may be either reactive and idiopathic type. The latter is defined as diffuse idiopathic pulmonary neuroendocrine cell hyperplasia (DIPNECH) and sometimes precedes and coexists with carcinoids. According to the WHO lung tumor classification, DIPNECH is considered a form of preinvasive lung lesion. The available data regarding this rare condition are very limited. We would like to present our own experience with DIPNECH and carcinoid tumors, describe the pathologic features and immunohistochemical profile of both processes and to evaluate the frequency of DIPNECH and carcinoid tumors.

Methods
The cohort study group consisted of 229 patients diagnosed with carcinoid tumors who underwent surgery at the National Tuberculosis and Lung Diseases Research Institute between 1998-2011. We evaluated all microscopic samples from main tumor and adjacent, as well as peripheral parenchyma of the lung. When features of neuroendocrine cell proliferation were found, immunohistochemical reaction were done using markers of epithelial differentiation (cytokeratins, TTF-1, Napsin A), neuroendocrine antigens (chromogranin A, synaptophysin, NCAM/CD56). The cell proliferation index with Ki-67 and expression of CD117 were estimated.

Results
DIPNECH was confirmed in 15 patients, 14 females and 1 male with an average age 65,2 (range 32 – 79 ys). In 7 cases, carcinoid tumors were found in more than one focus, both typical (TC) and atypical (AC) carcinoid. The lesions exhibited a diameter with an average size of 1,9 (range from 0,6 to 3,0 cm). In one patient, apart from TC and AC, numerous foci of sclerosing haemangioma were found. Three patients had a history of cancer surgery (mastectomy, kidney resection, and hemicolectomy). In all cases, no metastases of primary tumors were evident. Almost all of the diagnosed carcinoid tumors, except one, were located peripherally, with mainly spindle-cell morphology, solid and/or insular pattern and foci of fibrosis. In 1 case, one of the tumorlet infiltrated the pleura but it did not extend beyond its margins. In all patients, the neureondocrine cells, tumorlets and carcinoid tumors displayed intensive positive reactions to neuroendocrine markers. Silimilarly, a positive reaction to cytokeratin was confirmed but in the most cases (73%) it was semi-intensive or weak. In all cases the expression of TTF-1 was positive but in the most cases weak. No positive reaction to napsin A and CD117 was detected. The proliferation index in DIPNECH was 1-2%, but in carcinoid depended of subtype of tumor, for TC was lower than 4%, for AC higher than 4%. Bcl-2 was evident in 9 DIPNECH and carcinoid tumor. Serotonin expression was found in 8 carcinoid tumors. 13 patients were alive at the end of follow-up, there were no information about 2 cases.

Conclusion
DIPNECH is a rare condition predominantly connected with carcinoid tumors, both TC and AC. Immunohistochemical profile is similar to peripherally located carcinoids. We suppose that there are biological differences between central and peripheral carcinoids, in that the latter stain more frequently with TTF-1, serotonin and Bcl-2. Peripheral are more often spindled in morphology and seen in cases that present with DIPNECH.

Background
Cancer is a multistep process in which tumor cells progressively harbour genetic alterations that confer growth and spread advantages. These observations are also confirmed in NSCLCs where specific genetic abnormalities are sensitive to the action of targeted drugs, even if secondary mutations could develop conferring drug resistance. Furthermore, in recent studies, histologic and immunophenotypic changes have been described in patients (pts) with a specific molecular alteration re-biopsied after receiving a targeted therapy. This finding suggest the possibility of a "clonal resistance mechanism" in which genetic-similar cell populations had or acquire selective survival features thus escaping the inhibitory drug effect. In the present study histological examination and wide genetic analyses have been performed in tumor tissue sampled both before and after treatment in an unselected NSCLC patient population in order to elucidate progressive clinic-pathological and genetic abnormalities.

Methods
NSCLC pts with adequate tissue samples before and after at least one treatment have been evaluated from July 2006 to March 2013, at Department of Oncology of San Luigi Hospital. All had ECOG Performance Status of 0, median age of 51,5 years and IIIA-IV stage at diagnosis. After histological examination, mutational analyses for EGFR, K-RAS, PIK3CA, B-RAF, and HER2 genes were performed using pyrosequencing or RealtimePCR. ALK and c-MET genomic rearrangement were tested by FISH.

Results
A total of 24 (12 males and 12 females) pts were collected. Histological diagnoses were re-confirmed in all but one (4%) case in which morphological and immunophenotypical histology of neuroendocrine small cell lung cancer (SCLC) was found. All samples were adequate for molecular analyses. At the first biopsy EGFR activating mutations were 10/24 (42%) and 3/24 (12,5%) were the exon 20 EGFR p.T790M mutation, 2/3 (66%) associated to EGFR activating mutations. At the second biopsy 6/10 (60%) EGFR activating mutations were maintained and no acquired mutations in EGFR wild type pts at first were found at second biopsy. On the other hand, 6/24 (25%) p.T790M mutation were detected at second biopsy, 5/6 (83%) de novo acquired, 1/6 (17%) maintained and only 1/5 (20%) associated to EGFR activating mutations. The patient who acquired SCLC histology, maintained the L858R EGFR activating mutation after treatment with no other acquired mutation. K-RAS mutations were found in 4/24 (16.7%) first biopsies, while 5/24 (21%) were found at the second biopsies. No mutations were found in BRAF, HER2 and PIK3CA genes. ALK rearrangement was assessed in 5/24 (20,8%) patients; otherwise MET amplification was seen in 3/24 (12.5%) cases only 1/3 (30%) in EGFR mutant cases.

Conclusion
These preliminary data showed a complex scenario of basal and acquired alterations on tumor tissue before and after treatment highlighting the need of repeated histological and genotypic assessments to guide at the best treatment decisions.

Background
Personalized treatment of lung cancers necessitates the subclassification of NSCLC into adenocarcinoma (ADC) and/or squamous cell carcinoma (SqCC). In most cases, NSCLC can be subclassified by routine histomorphological examination of tumors. However, in poorly differentiated tumors or on small biopsy specimens, the subclassification may be difficult by H&E slides alone. Therefore, a panel of immunehistochemical (IHC) markers, including TTF1, Napsin A for adenocarcinoma (ADC) and CK5/6, p63 and p40 for squamous cell carcinoma (SqCC), are usually used to aid in the subclassification, These panels need to be performed on multiple tissue sections, and may result in the exhaustion of tumor tissue for molecular tests. In order to preserve tumor tissue, we investigated the utility of a newly developed triple marker (a combination of TTF, Napsin A and p40) in the subclassification of NSCLC; and compared the sensitivity and specificity of this novel triple marker with commonly used individual markers.

Methods
Using pathology archives from Johns Hopkins Hospital, three lung cancer tissue microarrays (TMAs) were constructed using surgical resection material, including 77 cases of ADC, 75 cases of SqCC and 46 cases of metastatic lung ADC. Immunostaining patterns of the triple marker and individual markers were scored semi-quantitatively and compared.

Results
In ADC, the sensitivity and specificity of the triple marker showed 93.50% and 77.50%, respectively (Table 1). In SqCC, the sensitivity and specificity of the triple marker showed 100% and 92.50%, respectively (Table 2). In addition, the sensitivity and specificity of the triple marker in metastatic ADC showed 71.74% and 77.50%. Table 1. Immunostaining patterns of different markers in lung ADC (n=77 cases)

	IHCscores
Markers	0	1	2	3	Sensitivity	Specificity	P value
Triple marker	6.5%	10.4%	13.0%	70.1%	93.5%	77.5%
TTF	14.3%	5.2%	32.5%	48.1%	85.7%	75.0%	0.185
Napsin A	10.4%	9.1%	22.1%	58.4%	89.6%	90.0%	0.564

Table 2. Immunostaining patterns of different markers in lung SqCC (n=75 cases)

	IHC scores
Markers	0	1	2	3	Sensitivity	Specificity	P value
Triple marker	0%	29.3%	33.3%	37.3%	100%	92.5%
p40	0%	32.0%	29,3%	38.7%	100%	90.0%	1.0
p63	0%	24.0%	45.3%	30.7%	100%	80.5%	1.0
CK5/6	5.3%	22.7%	36.0%	36.0%	94.7%	80.0%	0.12

Conclusion
Our triple marker showed a similar specificity and sensitivity as individual markers and yielded optimal conservation of tissue for molecular testing.

Background
The appropriate intraoperative decision making of surgical resection for the pulmonary ground glass nodules (GGN) is often difficult. We aimed to evaluate the role of frozen section diagnosis (FSD) as for the intraoperative decision making milestone and compared its accuracy to that of preoperative CT based practice as an interim result.

Methods
We retrospectively reviewed FSD of 171 consecutive pulmonary GGN from February 2005 to June 2013 and compared the diagnostic accuracy. Initially, we used only conventional method (Group A) but recently, we adapted a embedding medium inflation method (Group B) for FSD. The qualities of FSD were compared with the final pathologic diagnoses of corresponding permanent paraffin sections. Also, we calculated the sensitivity, specificity, and predictive values of assessing the size of invasive portion in GGN between FSD using the inflation method and preoperative CT based practice.

Results
There were no differences in nodule sizes between two groups (1.45±0.6 versus 1.51±0.5, p=0.63). In group A, a correct differential diagnosis between malignancies and benign lesions were made in 138 nodules. Thirteen nodules were erroneously classified and reported as false-positive or false-negative frozen section diagnoses (Sensitivity 95.6%, Specificity 53.8%). Three nodules were under-diagnosed in FSD. One patient required a secondary operation because of false-negative frozen diagnosis at the time of initial surgery. In group B, all of 17 nodules were correctly classified by frozen section. There were no false-positive or false-negative diagnoses in terms of making a diagnosis of malignancy, resulting in 100%-sensitivity and -specificity. (Figure 1) Thirteen nodules were correctly classified as being either minimally invasive adenocarcinoma (MIA) or invasive adenocarcinoma. Three nodules were diagnosed as MIA by frozen section through measuring invasive tumor size (<5mm) concomitantly. With regards to the estimating the size of invasive components of GGN, FSD in group B was superior to measurement of solid component in GGO nodules on HRCT. (Table 1)Figure 1

Conclusion
The accuracy of FSD using the embedding medium inflation method in GGO nodules was outstanding compared to the conventional frozen method. Furthermore, this method can help surgeons plan the appropriate surgical treatment after wedge resection of a GGO nodule by providing accurate size estimation of the invasive components of the GGN.

Background
The basic aim of a biobank is to provide biomaterials of high quality for research purposes including frozen archived samples, but also fresh tissue and viable cells, combined with relevant clinical data. This study presents workflow and quality management.

Methods
Patients are enclosed after signing a broad informed consent, allowing use of the samples in future research, and including molecular analysis. Samples are collected independently of a defined research project. Double pseudonymisation ensures privacy of donors. After pathological analysis the surplus of the harvested tissue is processed in RNAlater. Tissue samples as well as pre- and postoperative sera are archived at -80°C. Corresponding FFPE blocks are stored at room temperature. Bronchial lavages, pleura fluid, bronchial brush and transbronchial biopsies are integrated. In collaboration with researchers, fresh, unpreserved tissue is processed according to individual needs, such as filling with Agarose for three-dimensional tissue culture or in media for cell isolation.

Results
Between 01/2009 and 12/2012 around 1200 patients donated tissues to the biobank. In general, samples are processed within 20 minutes. We compared processing of samples in RNAlater with direct freezing in liquid nitrogen with times of ischemia between 5 minutes and 3 hours. We could not see any differences in RNA quality or the expression level of 4 genes tested between RNAlater or liquid nitrogen processed samples. Times of ischemia of up to three hours had no influence on RNA quality or expression, possibly due to compartmentalisation of RNAses in still intact cells. Samples were appropriate for protein analytical techniques such as Western Blot and ELISA. Histological slides from RNAlater samples were well evaluable by pathologists and comparable to normal cryosections.

Conclusion
The challenge of the biobank is a proactive tissue acquisition by pulmonary physicians trained in endobronchial techniques and thoracic surgeons in close collaboration with a pathologist. As a result global scientific work is feasible in a network like the German lung science center.

Background
Recent advances in lung cancer have identified potential driver mutations that may be targeted. On the basis of routine screening for EGFR we have initiated a comprehensive large-scale sequencing analysis of genes potentially mutated in NSCLC.

Methods
Genomic DNA was extracted prospectively from untreated advanced NSCLC tumors. All materiel was obtained IRB-approved protocols and after patients’ consent (MSN trial "Melanoma – Small-cell lung cancer – Non-small cell lung cancer "). Pathology specimens were macrodissected, after DNA extraction, 106 selected exons from 38 genes were analyzed by Sanger sequencing (EGFR, KRAS, HER2,4, BRAF, PI3KCA, PIK3R1, TP53, CDK4, CDKN2A, cKIT, PDGFRA, MET, FGFR2-4, FCGR2A,3A, FLT3, CTNNB1, GNAS, HRAS, NRAS, KDR, PDPK1, TOP1,2A, ERCC1, FBXW7, TSC2, PTEN, AKT1-3, MAP2K1-2, STK11, ALK). ALK rearrangements and HER2 amplification were detected by FISH. All result therapeutic outcomes were discussed monthly in a molecular thoracic multidisciplinary staff.

Results
Thus far (between May 2009 and September 2012), 351 patients (pts) have been included. The median age was 60 years (range 22-87), 212 (60%) were male, 248 (71%) had adenocarcinoma, 286 (81%) were former/current smokers. A complete failure of the analysis was observed in 78 (22%) pts mostly due to insufficient tumor cells in the specimen (<30%) or poor quality DNA. EGFR, KRAS, HER2, BRAF, PI3CA and ALK (“standard biomarkers”), analysis were performed in 235 (67%), 233 (66%), 207(59%), 221(63%), 139 (40%) and 206 (59%) pts respectively. Depending of markers, success rate was between 77% and 86 % (failures include scarce tumor sample). Two hundred and sixty three pts had at least one result for the EGFR, KRAS or ALK, and 176 pts had all three. 107 pts had a whole genomic analysis and 244 had at least one (1-6 biomarkers) standard biomarkers analysis. Ten (3.8%) pts had concurrent oncogenic mutation. The molecular profiles were characterized by 16% EGFR, 26% KRAS, 1% HER2, 0.8% PI3KCA mutated, 7% HER2 amplification and 11% ALK rearrangement. The pts with the while genomic analysis had 12 other genes evaluated for more than 80 pts and 13 pts had mutation (STK11, PDPK1, PTEN, NRAS, MET, KDR, FGFR4, HER4). A personalized targeted therapy was proposed in most of pts with a genomic alteration. Median OS of pts with at least one mutation/translocation for EGFR, KRAS, BRAF or ALK (n=152) was 13 and 17 months (p=0.006) in wild type or mutated pts respectively. In univariate analysis for OS (median follow-up: 19 months), KRAS mutated pts had a poor prognosis (hazard ratio [HR]=1.56, p=0.037), confirmed in multivariate analysis (HR= 1.78, p=0.008), EGFR mutated pts had a good prognosis (HR=0.48, p=0.01), BRAF and ALK mutations/translocation had no prognostic value.

Conclusion
Routine mutational profiling of advanced NSCLC is feasible in the vast majority of the pts but an extensive molecular portrait can be performed only in a limited number of pts. The molecular profile may have an impact on pts treatment strategy at our cancer institute. KRAS mutation is associated with poor prognosis. Updated results will be presented.

Background
Lung cancer remains a major unmet medical need in India and worldwide. About 63,000 new cases are diagnosed each year in India with deaths estimated at approximately 52,000, accounting for about 8% of all cancer deaths.Currently available cytotoxic chemotherapy is clearly inadequate for treatment of lung cancer. Therapies targeting the EGFR and ALK tyrosine kinases have proven effective for lung adenocarcinoma patients whose tumors harbor genomic alterations of the EGFR and ALK genes, respectively. However, most lung adenocarcinomas do not harbor these alterations and are not responsive to the relevant treatments. Hence, identification of mutationally activated oncogenes in lung cancer could result in additional therapeutic targets for this deadly disease.

Methods
676 exons from 196 were capturedin a discovery set of 125 Indian lung adenocarcinomasamples (FFPE blocks)using microfluidics based RainDance RDT-1000platform. The libraries were multiplexed and sequenced on Illumina Genome Analyser IIx. to obatin 1000X coverage per samples per base.Sequencing data were analysedusing in-house customized informtaicspipeline to identify potential driver alterations. 39 significantly altered mutations (novel and those known to be altered in NSCLC)were genotyped in an additional validation set of 300lung adeomcarcinomasamples using Sequenom MassArray.

Results
This is the firstreport of most comprehensive spectrum of alterations in 300Indian lung adenocarcinoma patients. We have identifiedmutation in genes previously reported as significantly mutated in lung adenocarcinoma: TP53, EGFR, KRAS, ERBB2, PIK3CA and NRASalong with previously unknown mutation incidence in AKT1 and NF2. In addition, we identified statistically recurrent mutations in the members of the fibroblast growth factor receptor familyin a subset of NSCLC tumors.

Conclusion
We identifyFGFRfamilyas a potential therapeutic target, a new avenue of investigation for the treatment of lung adenocarcinomas of Indian origin. This study also establishes a significant technological advancement for using cutting edge technologies on FFPE clinical specimens for high definition genomics, which has been a major impediment. in clinical research.In over all, this study enhances our understanding of the genomic complexity and heterogeneity underlying NSCLC tumors.

Background
Recent development of molecular target agents encouraged us to investigate the presence of driver mutations in patients with non-small cell lung cancer. As the prevalence of individual driver mutation is different in each ethnic group, understanding of mutational profiles of a specific country is important for clinical practice as well as for decision-making process of health care. Hence, we investigated the genetic profile of lung adenocarcinoma in Korean population.

Methods
Among the patients who underwent surgical resection for lung adenocarcinoma between 2001 and 2011, we set up a total of 477 patients whose fresh frozen lung cancer tissues and paraffin blocks were available. We retrospectively searched medical records of the EGFR exons 18-21 mutation tests results. Then, we selected patients who did not harbor EGFR mutations or who had not tested for EGFR mutations. DNA was extracted from those patients’ samples, and EGFR exons 18-21 and KRAS mutation test were performed by Sanger sequencing method. Tissue microarray was made for all 477 patients, and the EML4-Alk fusion was tested by a break-apart FISH method. We also tested KIF5B-RET fusion by using a break-apart FISH method and also by inversion specific long-range PCR. We investigated any correlation between mutational status and clinical variables, such as age, gender, smoking status, stage, and long term survivals.

Results
Among 477 patients, 321 patients (67.3%) were harboring at least one of four driver mutations. The EGFR mutations were the most frequently detected (270, 56.6%), followed by KRAS mutations (37, 7.8%), and EML4-Alk fusion (19, 4.0%). We also found five patients who had KIF5B-RET fusion mutations (1.0%). There were 10 patients who had more than two driver mutations; EGFR and KRAS mutations in 4, EGFR and EML4-Alk fusion in 4, KRAS and EML4-Alk fusion in one, and EGFR and KIF5B-RET fusion in one patient. The presences of EGFR mutations were frequently observed in patients with female gender (p=0.000). Although the EGFR mutations were associated with longer overall survival in univariate analysis (log-rank test rank test p=0.007), the presence of EGFR mutation was not a prognostic factor in multivariate analysis (Cox’s regression test p=0.469). The mutational statuses were associated with neither the disease-free survival nor fthe reedom from recurrence.

Conclusion
Based on our work, we found as high as 67.3% of lung adenocarcinoma patients in Korean populations were harboring at least one driver mutation, which may get a benefit from target agents. We also found as high as 2% of patients harbored multiple driver mutations. As the target agent will eventually develop resistance, it is recommended to test each driver mutation thoroughly even if one driver mutation was detected. Furthermore, our observation suggests future molecular testing should be based on the next generation sequencing platform.

Background
Synchronic or metachronic tumors may develop in patients with a lung tumor. Determining whether these tumors originate from the same clone or are separate lesions may be challenging. Clinical, morphological and immunohistochemical criteria are often not distinctive. The aim of our study was to investigate comparative genomic hybridization (array CGH) analysis for the evaluation of clonality in patients with metachronic or synchronic tumors, having at least one intrathoracic tumor localization.

Methods
A database was constructed of consecutive patients (n=77) referred by clinicians or pathologists for assessment of clonality by array CGH from 2007 till 2012. All cases with at least one intrathoracic tumor were selected. The array CGH patterns were analyzed by a visual comparison of CGH patterns performed by two investigators and by two mathematical models on the raw data. One model uses a log likelihood ratio as described previously[1], the other a Pearson correlation between the segmented values. Clonality cut-off and p-values were set according to copy number profiles from individual patients, which are therefore by definition non-clonal. The results of the visual evaluation and the mathematical approach were correlated. A control group was formed by the specimens of one lung tumor from every patient in the database, which were all compared, using the mathematical models. 1. Ostrovnaya I, Seshan VE, Olshen AB, et al. Clonality: an R package for testing clonal relatedness of two tumors from the same patient based on their genomic profiles. Bioinformatics. 2011;27:1698-1699.

Results
Specimens of 77 patients were referred for analysis. Samples of 14 cases were not suited to array CGH due to insufficient material or bad quality of DNA. The remaining 63 cases comprised of 142 samples. In 8 patients DNA from more than 2 tumors was compared. In the mathematical model the outcome of 3 cases was missing, 23 cases were determined clonal, 22 non-clonal, 5 with clonal as well as non-clonal tumors and 10 were undetermined. In the negative control group > 96% of cases were scored as non-clonal. The visual analysis determined 40 cases clonal, 14 non-clonal, 1 with clonal as well as non-clonal tumors and 8 were undetermined. 46 cases were available for comparing the outcomes of the mathematical and visual evaluation, as for 3 cases data were missing and in 14 cases the outcome was undetermined. The concordance rate for clonal and non-clonal tumors between visual analysis and the mathematical approach was 35 out of 46 (76%). In 4 cases in which the visual judgment was clonal, the mathematical model determined clonal as well as non-clonal samples. In 7 cases discordance was noted: the visual outcome was clonal and the mathematical non-clonal.

Conclusion
Array CGH is a useful approach for evaluating clonality of synchronic or metachronic tumors.

Background
It is known that in lung adenocarcinoma, ground glass nodule (GGN) tumors have a better prognosis than solid tumors. The aim of this study is to determine whether it is more useful to evaluate the whole tumor size or only the solid component size to predict the pathologic malignancy and the prognostic outcome in lung adenocarcinoma.

Methods
Using high-resolution computed tomography (HRCT) data of 232 patients with adenocarcinoma 7 cm or less who underwent curative resection, we retrospectively measured the whole tumor and solid component sizes with lung window setting (WTLW and SCLW) and whole tumor sizes with a mediastinal window setting (WTMW).

Results
There was significant correlation between the WTLW and the measurements of pathological specimens (r=0.792, P<0.0001). The SCLW and WTMW values significantly correlated with the area of pathologically confirmed invasion (r=0.762, P<0.0001 and r=0.771, P<0.0001, respectively). The receiver operating characteristics area under the curve for WTLW, SCLW and WTMW used to identify lymph node metastasis or lymphatic or vascular invasion were 0.693, 0.817 and 0.824, respectively. Kaplan-Meier curves of DFS and OS were better divided according to SCLW and WTMW, compared with WTLW. Multivariate analysis of DFS and OS revealed that WTMW was an independent prognostic factor (HR=0.72, 95%CI=0.58-0.90, P=0.004 and HR=0.74, 95%CI=0.57-0.96, P=0.022, respectively).

Conclusion
The predictive values of the solid tumor size visualized on HRCT especially in the mediastinal window for pathologic high-grade malignancy and prognosis in lung adenocarcinoma less than 7 cm were greater than those of whole tumor size.

Background
Asbestos is known for one of the risk factors of lung cancer (LC). The subjects with asbestos exposure are considered to have increased risk of LC especially if chest X-ray shows fibrotic changes of the lung. However, there are few reports about the association between the risk of LC and fibrotic changes detected by chest CT. The aim of this study was to define the specific CT findings of risk of LC among the subjects with asbestos exposure.

Methods
The low-dose chest CT(LD-CT) was taken as LC screening with written informed consent in subjects with former asbestos exposure in a prone position with the dose less than 30mAs. Each image was independently interpreted with more than two radiologists especially focused on fibrotic changes including subpleural curvilinear shadow/subpleural dots, reticular shadow, parenchymal band, traction bronchiectasis, and honeycombing. These findings were scored according to the extent. A chi square test was used to investigate the difference of categorical variables, and Mann-Whitney U test was applied as a non-parametric test.

Results
Between 2010 and 2012, chest LD-CT was taken in 2,126 subjects at 8 institutions in Japan. During the period, 38 (1.79%) LC was detected. We compared LC group and non-LC group, and there was no difference of duration of occupational asbestos exposure, or the prevalence of fibrotic changes on chest LD-CT. The extent of the fibrosis determined by scoring of the LD-CT findings, however, was more prominent in LC group than in non-LC group, though it was not statistically significant (p=0.067). On the other hand, LC developed more frequently in subjects with some fibrotic change (2.67%) than in those without any fibrotic change (1.44%, p=0.067).

Conclusion
The fibrotic changes of the lung detected on LD-chest CT might be the risk factor of LC development in subjects with former asbestos exposure. Further examination is planned with more subjects and longer period of observation.

Background
We previously reported the correlation among the thin-section computed tomography (TS-CT) findings, the pathological factors (Noguchi’s classification) and the prognosis of the patients. The purpose of this study was to examine the tumor shadow disappearance rate (TDR) on TS-CT findings, clinical course and pathological factors of small-sized adenocarcinomas of the lung according to the 2011 IASLC/ERS Classification.

Methods
We retrospectively analyzed 111 peripheral non-mucinous adenocarcinomas of the lung ≤ 10 mm in diameter that were surgically resected at our institute between January 1997 and February 2013. CT scans were obtained by commercially available scanners (X-Vigor/Real or Aquilion M/16 CT scanner; Toshiba Medical Systems; Tokyo, Japan). TS images were obtained with a 1 mm section thickness, pitch of 1, section spacing of 1 mm, 512 × 512 pixel resolution and 1 second scanning time. TDR was defined as the ratio of the maximum diameter of the tumor opacity of the mediastinal window to that of the lung window on TS-CT. We also examined the relationship among the TDR, the patient backgrounds, pathological findings (i.e., lymph node metastasis, pleural invasion, vascular invasion and lymphatic invasion) and clinical course. The histologic subtypes were analyzed according to the 2011 IASLC/ATS/ERS International Multidisciplinary Classification of Lung Adenocarcinoma.

Results
The median age of the patients was 64 (range, 23-83) years, and 66 patients (59.5%) were female. Sixty-four patients (57.7%) were never-smokers. The average tumor size was 8.7 mm (range, 5-10 mm). Regarding the histological subtypes, 70 cases were adenocarcinoma in situ (AIS) or minimally invasive adenocarcinoma (MIA), 19 were acinar predominant (AP), 13 were papillary predominant (PP) and seven were solid predominant (SP). Two cases could not be determined. Seventy cases diagnosed AIS or MIA were all stage IA, and none of these patients relapsed. Six cases relapsed after surgery; three cases of AP, two of PP and one of a SP tumor. In a comparison of the clinical course, the pathological differentiation and the TS-CT findings, all six cases relapsed after surgery showed ≤ 40% in TDR. Four cases diagnosed with lymph node metastasis (i.e., cases diagnosed in stage IIA or higher) showed ≤ 22% in TDR. Twelve cases with pleural invasion or vascular invasion or lymphatic invasion in the pathological factors of the resected lesions showed ≤ 28% in TDR. The TDR of AIS and MIA cases were all ≥ 50%.

Conclusion
There are sometimes pathologically invasive lesions even in small-sized adenocarcinomas of the lung. We found that the TDR is related to the clinical course and pathological factors in small-sized adenocarcinomas of the lung (10 mm or less in diameter). The lesions with a TDR ≤ 40% in the TS-CT images may be a group of highly malignant with an increased risk of relapse. The TDR may contribute to the determination of the optimal therapeutic strategy. We need a more robust prospective study to validate the efficacy of TDR.

Background
Despite the widespread adoption of FDG-PET/CT in staging of lung cancer, there are no universally accepted criteria for classifying thoracic nodes as malignant. Previous studies have generally shown high negative predictive values, but there are varying reporting criteria and positive predictive values for classifying malignant involvement. Using Endobronchial ultrasound transbronchial needle aspiration (EBUS-TBNA) histology as the gold standard, we evaluated objective 18-F FDG-PET/CT criteria for interpreting mediastinal and hilar nodes and compared this to expert visual interpretation (EVI).

Methods
A retrospective review of all patients with proven/suspected primary lung cancer who had both FDG-PET/CT and EBUS-TBNA from 2008-2010 was performed. Scan interpretation was blinded to histology. Separate prediction and validation datasets were used. 104 patients from 2008/2009 formed the prediction set; 48 patients from 2010 formed the validation set. Objective FDG-PET/CT criteria were: - SUVmax lymph node (SUVmaxLN) - Ratio SUVmaxLN/SUVmax primary lung malignancy if evident (R-SUVmax primary) - Ratio SUVmaxLN/SUVaverage liver (R-SUVavg liver) - Ratio SUVmaxLN/SUVmax liver (R-SUVmax liver) - Ratio SUVmaxLN/SUVmax blood pool (R-SUVmaxBP) An experienced Nuclear Medicine Physician visually reviewed all scans and classified each thoracic nodal station as benign, malignant, or equivocal. For statistical analysis, ‘equivocal’ nodes were classified benign.

Results
87 malignant lymph nodes from 75 patients and 41 benign nodes from 21 patients were in the prediction set. All objective 18-F FDG-PET/CT criteria analysed were significantly higher in the malignant group compared to the benign group (p<0.0001 all criteria). EVI had 95.3% accuracy, with 83/87(95.4%) malignant nodes and 39/41(95.1%) benign nodes correctly classified. 34 malignant nodes from 34 patients and 19 benign nodes from 14 patients were in the validation set. The new proposed cut-off values of the objective criteria from the prediction set correctly classified 44/53(83.0%) nodes, with 28/34(82.4%) malignant nodes and 16/19(84.2%) benign nodes correctly classified. EVI had 91% accuracy, with 33/34(97.1%) malignant nodes and 15/19 (79.0%) benign nodes correctly classified. Figure 1

Conclusion
Objective analysis of 18-F FDG PET/CT can differentiate between malignant and benign nodes with high accuracy, but is not superior to EVI. For objective criteria to perform optimally, there may need to be different criteria devised for different patient populations.

Background
Biological behavior of small lung adenocarcinoma differs in each patient. High intensity area of the tumor on single slice of chest CT has been reported as a prognostic factor in several studies. However, because single slice is obviously insufficient to utilize all of the information on CT images of a tumor, we applied 3- dimensional volumetric evaluation for whole tumor volume. Our study aims to predict lymph node metastasis and tumor invasiveness by means of preoperative quantitative CT for lung cancer patients.

Methods
From January 2011 to November 2012, 236 lesions of cT1N0M0 lung adenocarcinoma were surgically resected in our institute. Among them, total 211 lesions of 193 patients were included in this study (Age: 67.2±9.5 male/female: 94/99). We analyzed preoperative CT images of 211 lesions of resected cT1N0M0 lung adenocarcinoma retrospectively. All patients were subjected to helical scanning using sections 1mm or less thick during one breath hold. We applied threshold of -800 and -300 Hounsfield units (HU) within those CT data, calculated the tumor volume, and then, integrated them with clinico-pathological information. We defined the area -300HU and over as “solid tumor volume” and between -800 to-301 HU as “GGO tumor volume”. Spearman’s rank test was utilized for statistical analyses.

Results
We divided those lesions into 3 groups by solid tumor volume; less than 0.25cm[3] (n=61), 0.25 to 1.5cm[3] (n=72), and over 1.5cm[3] (n=78). Solid tumor volume correlated with histological tumor invasiveness; less than 0.25cm[3], p1to3 (0, 0%) ly1 (0, 0%) v1 (0, 0%); 0.25 to 1.5cm[3], p1to3 (6, 77%) ly1 (1, 1%) v1 (1, 1%); over 1.5cm[3], p1to3 (14, 19%) ly1 (4, 6%) v1 (14, 19%), (p<0.01, p=0.03, p<0.01, respectively). Pathological tumor differentiation was also investigated; less than 0.25cm[3], well (34, 56%) moderate (26, 43%) poor (1, 2%); 0.25 to 1.5cm[3], well (17, 22%) moderate (57, 73%) poor (4, 5%); over 1.5cm[3], well (8, 11%) moderate (51, 71%) poor (13, 18%) (p<0.01). Lymph node metastases were found in none (0%) of solid tumor volume less than 0.25cm[3], in 2 (3%) with 0.25 to 1.5cm[3], in 6 (8%) with over 1.5cm[3] (p=0.01). Moreover we calculated the proportion of solid tumor volume / (solid tumor volume + GGO tumor volume) as “solid tumor ratio”. We divided those lesions into 2 groups by solid tumor ratio; 0.3 or less (n=123), and over 0.3 (n=88). Solid tumor ratio also correlated with histological tumor invasiveness; 0.3 or less, p1to3 (0, 0%) ly1 (0, 0%) v1 (0, 0%); over 0.3, p1to3 (20, 23%) ly1 (5, 6%) v1 (15, 17%). (p<0.01, p<0.01, p<0.01, respectively) Strikingly, lymph node metastases were found in none (0%) of solid tumor ratio 0.3 or less, but in 8 (9%) with over 0.3. (p<0.01)

Conclusion
Both tumor volume -300HU and over “solid tumor volume” and “solid tumor ratio” significantly correlated with tumor invasiveness. Preoperative quantitative CT is probably useful for predicting tumor invasiveness and lymph node metastases, and, as a result, effectively selecting operative procedure for cT1N0M0 lung cancer whether lobectomy or segmentectomy is applicable.

Background
Superior vena cava (SVC) obstruction can cause development of collateral vessels. During contrast-enhanced thoracic CT, contrast material may reflux into the collaterals such as paravertebral venous plexus. However, unusual pseudopathologic vertebral body enhancement on CT in the presence of SVC obstruction has not been studied previously. So, the aim of this study was to demonstrate clinical presentation and imaging findings of pseudopathologic vertebral body enhancement in patients with SVC obstruction.

Methods
From March 2009 to September 2012, a retrospective radiologic database review was performed to identify patients with obstruction of SVC causing contrast reflux into collateral vessels and presented with unusual vertebral body enhancement on thoracic CT. Thirteen patients (eleven men, mean age 51.4 years) with vertebral body enhancement were enrolled. The underlying diseases that caused SVC obstruction were adenocarcinoma of the lung in four, non-small cell lung cancer in two, large cell neuroendocrine carcinoma, small cell lung cancer, thymic carcinoma, sarcomatoid carcinoma, diffuse large B-cell lymphoma, Hodgkin’s lymphoma and metastatic lymphadenopathy from pancreatic cancer in one patient each. Enhancement patterns, locations and the presence of a connection between vertebral body enhancement and the paravertebral venous plexus were evaluated. Enhancement patterns of vertebral bodies were classified as nodular enhancement with round shape that occupying < 1/3 of vertebral body or polygonal enhancement that occupying ≥ 1/3 of vertebral body on axial image. The locations of enhanced areas within vertebral bodies were described using right lateral/central/left lateral, anterior/posterior, and upper/middle/lower in the x-, y-, or z-axis directions, respectively.

Results
A total of 39 vertebral body enhancements were found in the 13 patients, involving cervical (n = 12), thoracic (n = 25), or lumbar (n = 2) vertebrae. Vertebral body enhancements showed a nodular (n = 19) or a polygonal (n = 20) pattern. The central portions of vertebral bodies were more frequently involved. The connection to the paravertebral venous plexus was observed in 34 lesions (87.2%).

Conclusion
Patients with SVC obstruction with extensive collateral vessels might exhibit a pseudopathologic vertebral enhancement. They tended to involve the central portion of the vertebral body and most of them showed connection to the paravertebral venous plexus.

Background
Coronary artery calcification (CAC) is frequently detected on low-dose CT of the thorax (LDCT). Concurrent assessment of CAC and lung cancer screening using LDCT is beneficial in terms of cost and radiation dose reduction. The aim of our study was to evaluate the reliability of visual ranking of CAC on LDCT compared to Agatston score on ECG-gated calcium scoring CT.

Methods
The subjects were 576 patients who were consecutively registered for health screening and undergoing both LDCT and ECG-gated calcium scoring CT. We excluded subjects with a calcium score of zero. The final study cohort included 117 patients with CAC (97 men; mean age, 53.4±8.5). Agatston score on ECG-gated calcium scoring CT was used as the gold standard (mean score, 166.0; range, 0.4-3719.3). Two board-certified radiologists and two radiology residents participated in an observer performance study. Visual ranking of CAC was performed according to four categories (1-10, 11-100, 101-400, and 401 or higher) for coronary artery disease (CAD) risk stratification. Weighted kappa statistics was used to measure the degree of reliability on visual ranking of CAC on LDCT: kappa values between 0 and 0.40 were considered to indicate positive but poor agreement; between 0.40 and 0.75, good agreement; greater than 0.75, excellent agreement.

Results
The degree of reliability on visual ranking of CAC on LDCT compared to ECG-gated calcium scoring CT was excellent for board-certified radiologists and good for radiology residents. A high degree of association was observed with 71.6% of visual rankings in exactly the same category as the Agatston category and 98.9% varying by no more than one category.

Conclusion
Visual ranking of CAC on LDCT is reliable for prediction of categorization of Agatston score ranks. LDCT is useful for both CAD risk stratification and lung cancer screening.

Background
[18]F-fluoro-2-deoxy-glucose (FDG) uptake on Positron Emission Tomography (PET) is highly accurate in the detection of mediastinal lymph node metastasis and extrathoracic metastasis and plays an important role in the staging of lung cancer. Furthermore, recent studies suggest that PET-CT and Standardized Uptake Value (SUVmax) reflects tumor metabolism and proliferation capacity, and can be used as a prognostic factor as well as to predict response to therapy.

Methods
Our study aims to detect SUVmax differences between different lung cancer histologic sub-types. We conducted a retrospective single centre study. All lung cancers diagnosed in 2008-2009, with a PET-TC evaluation at diagnosis, and with follow-up at IPOLFG were included. Data regarding socio-demographic factors, smoking habits, histological diagnosis, staging, and performance status were obtained from clinical chart reviews. Statistical analysis was performed using STATA

Results
A total of 92 patients have been included in this study to date. 64 (69.6%) of our patients were male and 28 (30.4%) were female. Median age was 64.5 years, standard-deviation of 10.16 years, minimum of 36 and maximum of 82 years. 47 patients (51.65%) were current smokers, 31 (34.07%) were former smokers, and 11 (12.09%) were non-smokers. 54% presented ³ 40 Pack years. The most frequent histologic sub-type was Adenocarcinoma (ADC) (41.3%), followed by Squamous Cell Carcinoma (SCC) (30.4%). 42 patients had stage I or II lung cancer, 31 had stage III, and 19 with stage IV. The median size of the primary tumor lesion was 4.05cm (standard-deviation of 2.62cm), and the median SUVmax was 10.11 (standard deviation of 5.86). SUVmax by primary tumor histological sub-type revealed the following characteristics: median SUVmax in the SCC group was 12.12 (standard-deviation 6.3), ADC group 8.26 (standard-deviation 4.5), Non-small Cell Carcinoma group 12.65 (standard-deviation 7.5), Neuro-endocrine Carcinoma group 7.25 (standard-deviation 5.4) and Small Cell Carcinoma 10.67 (standard-deviation 3.6). The observed differences were statistically significant between the 5 considered groups (p=0.0233, One-way ANOVA) and specifically, we observed a highly significant difference in SUVmax between SCC and ADC groups (p=0.005, t-test). We also observed a positive correlation between tumor size and SUVmax.

Conclusion
Our preliminary results confirm the effect of tumor biology in PET-CT and SUVmax with observed differences between different histological sub-types, and suggest that PET-CT can be specifically important in the assessment of SCC and ADC subtypes. These results support further studies regarding the usefulness of PET-CT in tumor biology characterization in SCC and ADC subgroups.

Background
To describe HRCT findings of newly developed peripheral T1 lung cancer in idiopathic interstitial pneumonia (IIP) during IIP follow-up

Methods
Between November 2001 and October 2012, 66 consecutive patients (62men, 4 women; median age 64, range 40~85 years) who were diagnosed as IIP, fulfilled the American Thoracic Society diagnostic criteria and new cancer (including fourteen small cell) simultaneously, were included. Two radiologists independently reviewed 132 serial CT scans of 66 patients, determined the earliest scan showing lung cancer, and evaluated tumor size (mm), lobar location, axial location on transverse image, shape, and density of tumor. The median interval between null-IIP to new cancer-IIP was measured. Delay in diagnosis was measured from the time of the earliest scan showing lung cancer and the subsequent clinical diagnosis. Formal radiologic reports as ‘first choice’ before diagnosis of cancer were reviewed.

Results
The inter-observer agreement was good (Kappa value > 0.77). The median smallest tumor size on axial scan at presentation was 17mm (± 6.57, range, 5-30mm) with T1a/T1b (48/18). Tumor was most commonly located in right lower lobe (29/66, 43.9%), followed by left lower lobe (13, 19.7%). Thirty five tumors (53.0%) were in the interface between normal and fibrotic lung cysts such as honeycomb cysts, twenty two (33.3%) were in the midst of fibrotic lung cysts, and nine (13.6%) were in the normal lung. Fifty nine (83.3%) tumors had round or oval shape, seven (10.6%) tumors had a stellate shape, and two had a band-like shape. Most of the tumors (90.3%) presented as solid density rather than part solid, ground-glass opacity or consolidation. Lung cancers were found during the mean follow-up CT period of 513 days. The median delay in diagnosis was 440 days. Most of the lesions (70%) were interpreted as lung cancer, but nine were interpreted as pneumonia or fungal infection and seven were missed (10.6%) on HRCT.

Conclusion
About one third of the tumors were misdiagnosed including missed in ten percents. Over fifty percent of the cancers are located at the interface between normal lung and fibrotic cysts. New lung cancers usually show as tumor with a round or oval shape and solid density.

Background
Thin-section CT (TS-CT) provides us with a more precise image of small pulmonary carcinomas. Thin-slice sections with thicknesses of 0.5 mm-1mm reflect, with some accuracy, the histopathological findings; mixed ground-glass opacity (part-solid GGO) is one characteristic finding of pulmonary adenocarciomas. These findings are vary in appearance, for example; some contain mainly GGO components, and some contain mainly solid portions. CT findings of mixed GGO, pathological findings and prognoses have been reported. Presently, we do not fully understand the correlation between TS-CT findings of mixed GGO and the status of EGFR mutation.

Methods
We retrospectively reviewed the records and TS-CT scans of 115 patients with mixed GGO tumors. All patients had undergone surgical resection between 2002 and 2008. Tumor diameters measured 20mm or less in size. All TS-CT images were acquired by Aquillion CT scanner (Toshiba Medical System). TS-CT images of tumors were obtained at 135kVp at 250mAs with 0.5-1mm section thickness. All images were photographed using mediastinal (level, 40HU; width, 400HU) and lung (level, -600HU); width, 1600HU) window settings. All TS-CT images on lung window setting were classified as: (1) Predominant GGO type (pGGO; solid portion areas less than 50% of tumor), (2) Heterogeneous type (heterogenous increased density), (3) Predominat solid type (pSolid; Solid portion areas took up more than 50% of tumor). We analyzed EGFR and Kras mutations, and then studied the correlations between these TS-CT findings and the status of EGFR mutations.

Results
The tumors in all 115 cases were well-differentiated adenocarcinomas. GwS type; 24 cases, Heteogenous type; 30 cases, and SwG type 61 cases. The EGFR mutation ratio was 66.6% in pGGO type, 90% in Heterogenous type and 52.7% in pSolid type. The ratio of EGFR mutation was greater in Heterogenous types compared to pGGO and pSolid types. (pGGO/Hetero p=0.045, pSolid/Hetero p=0.00038).

Conclusion
There is a correlation between the thin-section findings of mixed GGO and the status of EGFR mutations.

Background
Pulmonary carcinoids are histologically classified into typical and atypical. It is important to identify these preoperatively for treatment planning and prognosis. Our purpose was to calculate the diagnostic sensibility of positron emission tomography/computed tomography (PET/CT) using two different tracers ([18]Fluorine-fluorodeoxyglucose [FDG], a glucose analogue and [68]Gallium-DOTANOC, a somatostatin analogue) in a series of patients with suspected bronchial carcinoids (BCs). Additionally, we have evaluated the combination of dual tracer PET/CT findings in differentiating typical and atypical pulmonary carcinoids.

Methods
Forty-four patients (16 male/28 female; mean age: 57.8 y.o.) with suspected BCs (based on radiological findings) underwent PET/CT with FDG and [68]Gallium-DOTANOC in 2 high-volume centers. Detection rates of BCs on a per patient-based analysis were calculated. Histology was used as reference standard. The Chi-square test was used to correlate histology and PET findings.

Results
After surgery, 23 typical carcinoids (TCs), 10 atypical carcinoids (ATCs) and 11 benign pulmonary lesions (amartomas) were found. Overall and histology-based sensibility rates of both PET-methods are shown in Table. In particular, no false positive results were found and the overall detection rate for BCs was 100% when combining both PET methods (one positive at least). DOTANOC-PET/CT is superior in detecting TCs (sensibility= 91.3%) while FDG-PET/CT seems to be more useful in ATCs (sensibility= 100%). A significant association between histological type and PET findings with the two tracers was found (p<0.05). Moreover, the ratio of SUVmax values (DOTANOC/FDG) was significantly higher in TC-group when compared with ATC-group (10.00 vs 0.90; p= 0.027) Figure 1

Conclusion
Overall diagnostic performance of PET-CT in detecting BCs is optimal when integrating dual tracer PET-CT findings (FDG and DOTANOC). The combination of dual tracer PET/TC findings and the ratio of SUVmax values (DOTANOC/FDG) may be useful in differentiating TCs and ATCs. Therefore, both PET/CT methods should be performed in suspected BCs or when the histological subtype of BCs is unknown.

Background
A solitary pulmonary nodule (SPN) is radiologically defined as an intraparenchymal lung lesion that is < 3 cm in diameter and is not associated with atelectasis or adenopathy. The diagnosis of SPN is challenging. A prediction model would facilitates this task. Until now, three SPN prediction models have been developed, which are Mayo model, VA model and Peking University (PU) model. We compared the accuracy of three models in Chinese patients with SPN.

Methods
From July 2003 to December 2011, 154 surgical patients with an SPN measuring 3-30mm from Guangdong Lung Cancer Institute were included in this study. Data on gender, age, cancer history, smoking, nodule size, spiculation, calcification, border and final pathological diagnosis were collected retrospectively . Each patient’s final diagnosis was compared with probability calculated by MAYO model, VA model and PU model. The accuracy of each model was assessed by area under the receiver operating characteristics (ROC) curve and calibration curve.

Results
The area under the ROC curve (AUC) of PU model(0.800; 95% CI 0.708 to 0.891) was larger than that of MAYO model(0.753; 95% CI 0.650 to 0.857) and VA model(0.728; 95% CI 0.623 to 0.833), but this difference was not statistically significant. Calibration curves showed that all the three models overestimated malignancy.Figure 1 Figure 1 Receiver operating characteristic curves of Mayo model, VA model and PU model. AUC(Mayo)= 0.753(95% CI 0.650 to 0.857). AUC(VA)= 0.728; 95% CI 0.623 to 0.833. AUC(PU)= 0.800; 95% CI 0.708 to 0.891.

Conclusion
Three prediction models are sufficiently accurate in SPN malignancy prediction in Chinese patients.

Background
A recent study revealed that maximum iodine-related attenuation (IRA~max~) of dual-energy CT (DECT) of primary tumor strongly correlates with maximum standardized uptake value (SUV~max~) of [18]FDG PET-CT in non-small cell lung cancer (NSCLC) (n = 27; r = 0.785; p = 0.001). It suggested that DECT could serve as a valuable functional imaging test in NSCLC. The aim of our study is to validate the previous results in our NSCLC cohort.

Methods
Twenty-seven patients with NSCLC who underwent both DECT and [18]FDG PET-CT were analyzed. The maximum and mean IRA as well as virtual non-contrast (VNC) images were calculated from DECT. Pearson correlation test was used to analyze the relationship between all measurements of DECT and the SUV~max~ of [18]FDG PET-CT in primary tumors and lymph nodes.

Results
Twenty-seven primary tumors and 51 thoracic lymph nodes with an SUV~max~ of >2.5 were included in analyses. In primary tumors, there was a moderate correlation between IRA~max~ and SUV~max~ (r = 0.565; p = 0.002) whereas no correlation was found between other DECT measurements and SUV~max.~ In lymph nodes, we observed no correlation between IRA~max~ and SUV~max~ (r = 0.197; p = 0.165) as well as between other DECT measurements and SUV~max.~

Conclusion
In patients with NSCLC, correlation between IRA~max~ and SUV~max ~was observed in primary tumors but not in lymph nodes. Because of relatively small population in our study and previous study, further large prospective studies are needed for validation.

Background
Genotype manifests itself as phenotype and that the one may inform the other. In terms of phenotype, imaging has the potential to assist in noninvasively characterizing the tumor, however, there are very few investigators who have pursued that potential connection between imaging features and the genetic characteristics of lung cancer. The purpose of this study was to retrospectively correlate the CT morphologic patterns of nodular lung adenocarcinomas (ADs) with pathological and molecular phenotypes in an East-Asian cohort of patients.

Methods
The institutional review board approved this retrospective study, and all patients provided informed consent. 600 primary lung ADs smaller than 3 cm in diameter that were surgically resected from 592 patients (M:F=257:335; mean age, 63) were included. CT morphologic pattern of ADs was evaluated by three board-certified thoracic radiologists and was classified into four patterns: pure GGN, GGO dominant part-solid nodule (PSN), solid dominant PSN, and pure solid nodule. EGFR mutation, ALK rearrangement, and KRAS mutation were evaluated using PCR-based direct DNA sequencing and FISH. Histologic subtype was classified according to IASLC/ATS/ERS classification of lung AD. The Fisher exact test and student t-test were used to assess statistical significance.

Results
Figure 1 In terms of CT morphologic patterns, 17.2%,15.2%, 31.8%, and 35.8% of tumors manifested as pure GGN, GGO dominant PSNs, solid dominant PSNs, and pure solid nodules, respectively. EGFR mutation was significantly more often found in ADs that manifested as subsolid nodules (69.9%, 269/385) than in ADs presented as pure solid nodules (46.7%, 100/214) (P<.0001). ALK rearrangement was more frequent in ADs that manifested as pure solid nodule (8.5%, 13/153) than in tumors presented as subsolid nodule (1.8%, 5/281) (P=.001). KRAS mutation showed no significant difference between subsolid nodules (6.6%, 8/121) and pure solid nodules (8.5%, 5/59) (P=.760). The ratio of subsolid nodule vs pure solid nodule was 72.7% vs 27.3% in ADs with EGFR mutation and was 27.8% vs 72.2% in ADs with ALK rearrangement. EGFR mutation was more frequent in minimally invasive ADs (P=.004) and lepidic predominant ADs (P=.018). ALK rearrangement was more frequent in solid predominant ADs (P=.002) and invasive mucinous ADs (P=.030). KRAS mutation was more frequent in invasive mucinous ADs (P=.001).

Conclusion
EGFR mutation was significantly more often found in ADs that manifested as subsolid nodules, and ALK rearrangement was more frequent in ADs that manifested as pure solid nodule.

Background
The distinction between SCLC and NSCLC has been recently replaced by a more detailed re-classification. As 70% of patients with LC are still not eligible for surgery, tumor characterization is often based on needle biopsy. For the management of lung masses (LM), the availability of adequate samples is critical not only for pathological diagnosis, but also for additional molecular studies. In this context, our aim was to evaluate the safety and accuracy of image-guided core biopsies (CB) in our last 4-year series.

Methods
480 consecutive patients (325 male; 33-87 y, mean 68; LM diameter 6-150 mm, m 37,4) underwent 439 CT-guided, 35 US-guided and 6 US+CT-guided lung biopsies. In 325/480 cases (68%) a CB was preferred due to the possible requirement of molecular studies. 275 CB were performed with >=18G tru-cut needle, 50 with <18G, both by a coaxial technique (inserted in a 1G larger styleted cannula). 1 to 6 sampling per patient (m 1.5) were performed. Adverse events (including major complications) were recorded and correlated with technical issues (namely, with needle size). To assess the accuracy of CB, surgical specimens, outcome of non-surgical therapy and follow-up imaging were considered as gold standards. Sensitivity, specificity, diagnostic accuracy and positive and negative predictive values of CB were calculated.

Results
81/325 (24.9%) adverse events occurred, but only 23 (7%) were major complications (MC) (22 pneumothorax and 1 hemothorax, requiring drainage and prolonged hospitalization). The incidence of MC wasn’t different between either CB and FNAB group (11/155, 7%), or larger and smaller CB needle size (20/275 vs. 3/50, p=n.s). Only the depth of the LM seemed to be significant as negative predictor for MC (p=.0061). Pathological diagnosis was of benign LM in 60 CB (18.4%), malignancy in 265 (81.5%). According to the above gold standard criteria, TP were 265, FN 13, TN 47, FP 0. Sensitivity, specificity and diagnostic accuracy were 95.3%, 100% and 96%, respectively. PPV was 100%, NPV 78.3%.

Conclusion
CB is as safe as FNAB in characterizing LM; particularly, needle size doesn’t impact on MC rate. CB is highly accurate in morphological characterization of LM, also providing additional tissue for molecular studies, when needed.

Background
Differential diagnosis of primary lung cancer and metastatic lung tumor before surgery is important. However, histological diagnosis using bronchofiberscopy is often difficult in these small peripheral lung nodules. It appears to be useful to diagnose pulmonary nodules using chest CT. As already known, primary lung cancer presents complicated appearance in chest CT. Contour of primary lung cancer is expressed using the words such as undulated, irregular, and spiculated. Contrary, metastatic lung tumor usually shows simple round shadow. These characteristics are used for the differential diagnosis of tumors. However, we often meet tumors with borderline complexity that we are not able to clearly classify. Chest CT finding is expressed by words at the diagnosis. Therefore, it is difficult to standardize or compare the diagnostic properties. Numerical evaluation of complexity of tumor outline results in the quantitative evaluation of tumor shape and may help the standardization of diagnosis of pulmonary nodules on chest CT. Malignant pulmonary tumors basically show round appearance. Therefore, complexity of tumor outline is to be expressed by the deviation from a circle. And the extent of deviation can be expressed numerically. The array data set of the deviation is to be regarded as the composition of various kinds of waves. Fast Fourier transform (FFT) analysis is suitable to evaluate these components of the wave data. In this study, we performed the quantitative analysis for the complexity of tumor outline of both primary lung cancer and metastatic lung tumor utilizing FFT analysis. And then we evaluated the usefulness and adequacy of our evaluation method.

Methods
Sequential cases of 72 histologically proven primary lung cancers (Group PL) and 54 metastatic lung tumors (Group MT) were included. The diameters of tumors in groups PL and MT were 18.9±7.4 mm and 12.2±6.1 mm, respectively. The outline of each tumor on chest CT images was described using polar coordinates, and converted to rectangular coordinates, yielding wave data of the tumor outline. The FFT was then used to analyze the wave data. The complexity index (Cxi) was defined as the sum of the amplitude of all harmonics over a fundamental frequency.

Results
The Cxi was higher (P <0.0001) for group PL (10.3±6.7 mm) than for group MT (3.2±2.4 mm), and it was correlated with tumor diameter in both groups: PL (r =0.667, P <0.0001) and MT (r = 0.809, P <0.0001). The cut-off equation “Cxi = 0.127 DT + 2.23” provided the highest diagnostic accuracy for distinguishing Group PL from Group MT such as a sensitivity of 95.8%, a specificity of 81.5%, and an accuracy of 89.7%.

Conclusion
Complexity of outline of the pulmonary nodules can be evaluated quantitatively using FFT analysis. This analytical procedure was designed from the beginning as it can be equipped on the graphic workstation, and we are now starting to develop it. This analytical method will help the diagnosis of primary lung cancer. FFT analysis appears useful for quantification of complexity of the tumor outline.

Background
Malignant pleural mesothelioma (MPM) is a rare and aggressive tumor with a complex growth pattern. Imaging plays a crucial role in diagnosis and management. The diagnosis of MPM is based on histology using essentially immunohistochemistry on pleural biopsies. Occasional long-term survival results are probably due to the biologic characteristics of the disease. 18F-fluoro-2-deoxy-d-glucose (FDG) positron emission tomography (PET)-computed tomography (CT) (FDG PET-CT) has become an invaluable tool for the diagnosis, staging, and prognosis of this severe disease as it combines both anatomic and functional information in a single imaging procedure, allowing for improved management of this disease.

Methods
From April 20011 and May 2013, eight patients with histologically proven MPM underwent integrated FDG PET and computed tomography (CT) scanning. These were analyzed: mean age was 53.6 years (range: 40-66) and histologic MPM subtypes were epithelioid (7 patients) and biphasic (1 patients). We observed a significant correlation between high SUV(max) and high-grade disease. There were 75 % male patients and 25% female patients. PET-CT images showed intense abnormal FDG uptake (SUVmax>8.0) in the pleural lesions of all 8 MPM patients at delayed phase, while it showed abnormal FDG uptake in all (100%) patients at early phase. In all patients, the values of SUVmax at delayed phase were higher than those at the early phase. PET-CT also indicated metastasis in the lymph node in 4 patients (50%) and in the peritoneal lesions in one patient (12.5%) with MPM. The results showed that 18F-FDG PET-CT at delayed phase is very useful for the diagnosis of pleural diseases and high uptake of 18F-FDG PET-CT may be a predictive factor of prognosis in MPM patients.

Results
MPM has a limited response to conventional chemotherapy and radiotherapy, thus early diagnosis of MPM is extremely critical. In these group patients, PET-CT showed all lesions with high sensitivity (%100). CT scans have limited accuracy in the differentiation between benign and malignant pleural disease. Also CT tends to underestimate early chest wall invasion and peritoneal involvement and has well-known limitations in the evaluation of lymph node metastases. Several studies have reported that 18-fluorodeoxyglucose (FDG) PET-CT plays an important role in the assessment of thoracic malignancy such as lung cancer. With a small patients experience, we suggest that PET-CT is useful as an aid for diagnosis and prognosis of MPM. PET-CT provides useful information concerning the extension of the lesions to thoracic and abdominal walls not fully evaluated by the initial conventional cross-sectional imaging. PET/CT also allows an accurate therapeutic monitoring of the disease.

Conclusion
Our results showed (even with a limited group of patients experience), PET/CT seems to be superior to convantional imaging modalities in identifying more extensive disease involvement, and detecting unexpected occult distant metastases.

Background
Non-small cell lung carcinomas (NSCLCs) are sometimes detected incidentally on imaging for unrelated causes. We studied the rate of incidental detection and its impact on long-term survival in a nation-wide cohort of patients operated for NSCLC.

Methods
This population-based study included all patients who underwent pulmonary resection for NSCLC in Iceland between 1991 and 2010. Demographics and clinicopathological features were compared in patients diagnosed incidentally and those presenting due to symptoms. The data was devided into four 5-year periods to assess trends, and multivariate analysis used to evaluate prognostic factors of cancer-specific survival (CSS), focusing on incidental detection.

Results
From a total of 512 patients, 174 (34%) were diagnosed incidentally and this proportion remained unchanged during the study period. Most tumors were detected by chest X-ray (CXR) (26%) or CT (8%), but the proportion of CT diagnoses rose to 15% during the last 5-year period. The incidentally detected tumors were smaller (3.0 vs 4.3 cm, p<0.0001) and were diagnosed at earlier TNM-stages (64 vs. 40% on TNM-stage I, p<0.0001). Furthermore, a higher fraction of the incidentally detected tumors were adenocarcinomas. Five-year CSS for patients with symptoms was 40%, those incidentally detected on CXR 57% and on CT 80% (p<0.001). By multivariate analysis patients detected incidentally on CT had significantly better CSS compared to those diagnosed incidentally by CXR or for patients with symptoms related to NSCLC (HR 0.38, 95% CI 0.16-0.88, p=0.02).

Conclusion
A third of surgically treated NSCLC-patients are detected incidentally, and this fraction has not changed for two decades. However the proportion of CT-detected NSCLCs is increasing and these patients seem to have a more favorable survival than patients detected incidentally on CXR or those who present with symptoms.

Background
PET for early prediction of tumor response to Erlotinib in patients with advanced non small cell lung cancer NSCLC has been evaluated. The aim of this prospective study was to compare the early metabolic treatment response using European Organization for Research and Treatment of Cancer (EORTC) 1999 recommendations, PET Response Criteria in Solid Tumors (PERCIST), and the standard treatment response using Response Evaluation Criteria in Solid Tumors (RECIST).

Methods
PET was performed before treatment and after 45 days of receiving Erlotinib 150 mgrs daily in advanced NSCLC pts. The hottest lesion in each patient was evaluated according to the EORTC 1999 recommendations, PERCIST and RECIST to assess metabolic and anatomic response. Response classifications were compared statistically using Wilcoxon signed-rank test. Overall survival (OS) and Progression-Free Survival (PFS) time were calculated by the Kaplan-Meier test.

Results
14 patients (10 men and 4 women, mean age: 64.1±12.0 years, range 38 to 79 years) were evaluated. Anatomic responses according to RECIST were as follows: 4 Partial Response (PR) and 10 Progressive Disease (PD). Metabolic responses according to EORTC criteria were 5 Partial Metabolic Response (PMR), 5 Stable Metabolic Disease (SMD), and 4 Progressive Metabolic Comparisons of EORTC and RECIST assessment with PERCIST were as follows:

	PERCIST
RECIST	PMR	SMD	PMD	Total
PR	4	0	0	4
SD	0	0	0	0
PD	0	6	4	10
Total	4	6	4	14
EORTC
PMR	4	1	0	5
SMD	0	5	0	5
PMD	0	0	4	4
Total	4	6	4	14

There was a significant difference in the results of response classification between metabolic classifications and RECIST (p<0.04). RECIST and PERCIST (Responders vs. Non-responders) were significant factors associated with DFS ( HR =4.070; 95% CI, 1.383-11.972; p=0.002 for RECIST and HR =0.245; 95% CI, 0.0835-0.722; p=0.001 for PERCIST) and OS ( HR =2.620; 95% CI, 0.8806-7.795; p<0.046 for RECIST and HR =0.3817; 95% CI, 0.1283-1.1356; p=0.046 for PERCIST). EORTC criteria was a significant prognostic factor for predicting DFS ( HR = 0.319; 95% CI, 0.061-1.667; p=0.028) but not for OS (p=0.65).

Conclusion
Metabolic and anatomic response were different in NSCLC pts treated with Erlotinib. PERCIST can be considered an appropriate metabolic evaluation method of therapeutic efficacy to discriminate responders from non responders. Mature results will be presented.

Background
The epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor make very important role in chemotherapy for primary lung adenocarcinoma. It is necessary to examine EGFR mutation before medication, but it is difficult to examine for all patients. Our purpose is to determine EGFR mutation presence by HRCT findings of primary lung adenocarcinoma.

Methods
This study consists of 476 primary lung adenocarcinomas which were examined EGFR mutation (exon 18, 19 and 21) by a genetic analysis. EGFR mutation-positive group (EGFR-P) were 223 cases (resected =178, non-resected =45), and EGFR wild-type group (EGFR-W) were 253 cases (resected =161, non-resected =92). At first, presence of findings, such as much pleural effusion, atelectasis, or secondary pneumonia, that come to have difficulty in evaluation of size and the character of tumor, was determined. Then, all evaluable parameters were analyzed. The high resolution-CT (HRCT) findings that were analyzed independently by 2 radiologists with special attention were 21 parameters (tumor size, well-defined margin, irregular margin, spiculation, lobulation, pure ground-glass opacity (GGO), part-solid, solid, air bronchogram, cavity, calcification, broncho-vascular convergence, pleural indentation, pleural concave, pleural thickness, lymphangitis carcinomatosis, emphysema, interstitial pneumonia, pulmonary metastasis, pleural effusion, and lymph node enlargement). The age, gender and smoking history were additionally reviewed. These parameters were evaluated with Chi square test and multivariate analysis. A p-value less than 0.01 were considered to indicate a statistically significant difference.

Results
The cases that were hard to analyze a tumor into by such as atelectasis were two (EGFR-P) and 26 (EGFR-W), respectively. A statistical significant difference was present (p=3.4E-05). Chi square test showed a statistically significant difference about "part-solid (p=5.5E-06)", "air bronchogram (p=0.0036)" and "pleural indentation (p=0.0002)" more frequently in EGFR-P than in EGFR-W. Similarly “woman (p=1.3E-09)” and “non-smoker (p=2.7E-13)” were observed more frequently in EGFR-P than in EGFR-W. On the other hand, “solid (p=2.1E-07)”, “cavity (p=0.0004)”, “emphysema (p=5.1E-14)”, “interstitial pneumonia (p=3.1E-8)” and “lymph node enlargement (p=0.0008)” were observed more frequently in EGFR-W than in EGFR-P. And Multivariable analysis showed that “cavity (p=0.003)”, “emphysema (p=0.001)” and “interstitial pneumonia (p=0.001)” were observed more frequently in EGFR-W than in EGFR-P. On the other hand, there was no parameter that became statistically significantly more frequently in EGFR-P than EGFR-W. By the multivariable analysis, there was no significant statistical difference about gender and smoking history.

Conclusion
The cases that were hard to analyze a tumor into by such as atelectasis were significant in EGFR-W. Some HRCT findings (part-solid, solid, air bronchogram, cavity, pleural indentation, emphysema and interstitial pneumonia) indicated more statistically significant usefulness for a diagnosis of the EGFR mutation. Especially multivariable analysis showed that HRCT findings (cavity, emphysema and interstitial pneumonia) were more statistically significant about EGFR mutation than gender and smoking history.

Background
Lung cancer is a leading cause of the cancer-related death in the world, and frequently presents with an advanced disease at diagnosis. Thus, early detection of lung cancer is thought as an important opportunity for decreasing mortality. It is willingly expected that novel method of early detection of lung cancer using biomarkers, such as cancer-specific up-regulation of methylation levels on genomic DNA. We hereby report that 7 CpG sites were found to be highly methylated in Non-Small-Cell-Lung Cancer (NSCLC), those which are possible promising NSCLC-specific biomarkers.

Methods
Samples for genome-wide DNA methylation analysis consisted of 10 surgical specimens including 8 NSCLC (3 adenocarcinomas, 2 squamous cell carcinomas, 2 bronchioalveolar carcinomas and 1 large cell carcinoma) and 2 adjacent normal lung tissues. Genomic DNAs were isolated from frozen tissue sections by using QIAamp DNA mini kit (QIAGEN) followed by bisulfite conversion with Epitect bisulfite kit (QIAGEN) according to manufacturer’s protocol. Bisulfite converted DNAs were hybridized to the Illumina Infinium HumanMethylation 450 BeadChip, which screens methylation levels of more than 450,000 CpG sites within a single array. Evaluation of the array screening data revealed a series of candidate CpG sites that are preferentially methylated in cancer samples as compared to normal controls. To further validate cancer-specific up-regulation of methylation levels at candidate CpG sites, the QIAGEN Pyromark Q24 system was employed to conduct bisulfite pyrosequencing analysis. Pyromark CpG assay probes were designed with the aid of PyroMark Assay Design Software 2.0 to encompass up to 300 bases region around the CpG site of interest, and used to carry out PCR amplification and pyrosequencing on bisulfite-converted genomic DNA templates. The analyzed samples included 136 surgical specimens consisting of 68 pairs of NSCLC and adjacent non-cancerous matched tissue, which are collected by macrodissection from paraffin-embedded sections. The 68 NSCLC samples are classified in terms of histological types as follows: 54 adenocarcinomas, 10 squamous cell carcinomas, 2 adenosquamous cell carcinomas, 2 large cell carcinomas, and in terms of cancer staging as follows: 37 of stage I, 15 of stage II, 12 of stage III and 4 of stage IV.

Results
Seven CpG sites were found to be highly methylated in NSCLC samples compared to adjacent controls in a statistically significant manner (two-way ANOVA, p < 0.05). The Illumina CpG cluster numbers (cg#) of the 7 CpG sites and their genomic locations in GRCh37 coordinates are as follows: L3302 (cg26917140; chr1:91184770), L3303 (cg04654530; chr2:63282702), L3304 (cg27315333; chr2:63285950), L3305 (cg17080163; chr2:123418757),　L3310 (cg15347189; chr3:187387999), L3314 (cg03148184; chr4:111562513), L3316 (cg16509851; chr4:174430614). Among them, L3303 and L3310 are located within the genomic region corresponding to orthodenticle homeobox 1 (OTX1) gene and somatostatin (SST) 1st exon, respectively.

Conclusion
These 7 CpG sites are possible promising NSCLC-specific biomarkers for screening and early detection of lung cancer, and our goal is the development of the technology to detect these novel biomarkers from a blood sample. This technology may eventually lead to early detection of lung cancer, which is thought as an important opportunity for decreasing lung cancer mortality.

Background
To present results of the first four lung cancer (LC) screening programs based on low dose chest CT (LDCT) conducted in Poland.

Methods
Four LC screening programs were conducted in different Polish cities (Gdansk, Poznan, Szczecin, Warsaw) between 2008 and 2011. Algorithms used for patient`s selection for further diagnostics were based on IELCAP protocols with local modifications. The enrolled patients were ages 50-75, being either active smokers or with history of >20 pack years. The following data were analyzed: number of LC detected and resected in both protocols, and percentage of stage I cancers detected.

Results
34810 patients were screened in accordance with protocols. Results are presented in table I.

Parameter	G	P	W	Sz	All
N[o ]of patients	8693	9357	1740	15020	34810
LC detected	91	96	17	120	324
LC stage I	58	52	12	83	205

The overall detection rate was 1 cancer per 102 CTs and varied from 1/125 to 1/90 (1.05%, 1.03%, 1.0%, 0.86%, 0.98% respectively). The majority of cancers were diagnosed in stage I (64%, 55%, 70%, 69% 64.5% respectively), while this parameter does not exceed 30% in overall population treated by thoracic surgery in Poland. The differences between centers regarding both detection rate and stage I percentage were not significant (p=0.167 and p=0.599 respectively). However, if one compares the biggest center alone (Szczecin) vs all other ones together, the detection rate in Szczecin is significantly lower (p=0.0278).

Conclusion
The detection rate of all LC as well as stage I LC in the Polish LDCT screening programs is high and comparable in different centers.

Background
The National Lung Screening Trial (NLST) compared low dose CT (LDCT) with chest radiography (CXR) in high-risk populations and found a 20% reduction in lung cancer mortality at 6 years with LDCT after an initial scan and two annual rounds of screening. This is the first randomized controlled trial (RCT) to show a mortality benefit with lung cancer screening. LDCT screening is not yet part of the standard of care and no formal process currently exists in Ontario, Canada for lung cancer screening. Injudicious use of LDCT can potentially cause more harm than benefit, including exposure of healthy persons to ionizing radiation and subsequent invasive procedures for ultimately benign lesions. When used correctly, however, LDCT screening has the potential to save lives. A practice guideline was developed to guide clinicians and healthcare policy makers with evidence-based recommendations for screening high-risk populations for lung cancer.

Methods
The guideline was developed using the methods of Cancer Care Ontario’s Program in Evidence-Based Care (PEBC). The core methodology of the PEBC’s guideline development process is a systematic review. A systematic review had recently been completed by a collaboration of the American Cancer Society, the American College of Chest Physicians, the American Society of Clinical Oncology, and the National Comprehensive Cancer Network. The evidence from that systematic review formed the basis of the current recommendations, which were reviewed, and amended where necessary, by clinical experts in the fields of medical, radiation, and thoracic oncology; diagnostic radiology; pulmonary disease; and population health. The recommendations were reviewed by the Provincial Lung Cancer Disease Site Group and underwent both internal review by an expert panel and external review by clinicians with expertise in the topic to achieve consensus.

Results
The systematic review included three RCTs comparing LDCT screening with CXR (including the NLST), 5 RCTs comparing LDCT screening with usual care (no screening), and 13 single-arm studies of LDCT in patients at risk for lung cancer. One large RCT reported a statistically significant reduction in lung cancer mortality with low-dose computed tomography at six years compared with CXR. The practice guideline recommendations generally align with the parameters of the NLST. Deviations were described and justified by the guideline working group. The recommendations support screening persons at high-risk for lung cancer with advice for defining a positive result on LDCT, appropriate follow-up, and optimal screening interval.

Conclusion
The benefits of screening high-risk populations for lung cancer with LDCT outweigh the harms if screening is implemented in a strictly controlled manner targeting the high risk population. This practice guideline forms the basis for the rationale for a screening program. An economic impact analysis will need to be done to design an appropriate cost effective lung cancer screening program prior to implementation.

Background
Lung cancer is a leading cause of cancer death. The lack of symptoms in this disease and problems associated with screening programs for early detection means that patients present late with advanced disease. Studies using low-dose computed tomography proved to be beneficial, and the efficacy of computed tomography scanning as a screening tool for lung cancer is an important and contested topic. However, it has been reported that the increased incidence of cancer after computed tomography scan exposure was mostly due to irradiation. So, computed tomography scans should be limited to situations where there is a definite clinical indication. The search for non-invasive diagnostic methods of lung cancer has led to new avenues of research, including the exploration of the exhaled breath. Previous studies have shown that lung cancer can, in principle, be detected through exhaled-breath analysis. This study evaluated the potential of exhaled-breath analysis for the detection of lung cancer.

Methods
Breath samples were taken from patients with lung cancer and from healthy volunteers. The exhaled breath was collected in 1L Analytic- Barrier[TM] bag (OMI ODOR-AIR SERVICE Co. Ltd.). Profiles of volatile organic compounds were determined by gas chromatography/mass spectrometry (Shimadzu Co. Ltd.).

Results
We collected breath samples from 111 patients with lung cancer and 29 healthy volunteers using Analytic-Barrier[TM] bags. The volatile organic compounds were extracted with solid phase micro-extraction and analyzed by gas chromatography/mass spectrometry. The number of volatile organic compounds detected in breath samples was 68. Among the volatile organic compounds 10 compounds were found in only breath of the lung cancer patients. In addition, 11 compounds were found at significantly higher concentrations in breath of the lung cancer patients compared to the controls.

Conclusion
Our data suggest that compounds in breath could possibly be taken as useful breath biomarkers for discerning potential lung cancer patients and volatile organic compounds analysis could be used as a complementary test for the diagnosis of lung cancer. The vision of exhaled breath analysis is that of a broad use in clinical routine for personalized screening, diagnosis and treatment monitoring.

Background
Background: CT Coronary Angiography provides accurate non-invasive evaluation of coronary arteries but also images lung parenchyma and other mediastinal structures. Little is known about the range and incidence of non-cardiac findings in the Australian population and while CT screening has been shown to reduce mortality in high risk individuals the significance of identified pulmonary nodules in this mixed risk population is unknown. A lack of data regarding the malignant potential of these incidentally identified nodules makes evaluation of the relative risk/benefit of both initial imaging of the lung and subsequent surveillance scanning is difficult.

Methods
Methods: A retrospective analysis was performed on reports of all cardiac CT scans done in the calendar year 2012. Descriptive data was collected including baseline patient characteristics, type of nodule and smoking history, as well as whether a full field or restricted field view was performed. Surveillance radiological data and pathology was collected on a sub-group of the population.

Results
Results: 2500 Cardiac CT scans analysed with 48% females. Reports analysed for presence of lung nodules and type of nodule with most common nodule granuloma. Total lung nodules 14% (355) with follow up recommended by specialist radiologists; significant variation from recommendation in practice was noted. 39% of population positive for smoking exposure placing them in high risk population. No episodes of malignancy within follow up CT scans with the majority of nodules being stable over the follow up period. Majority of nodules were <4mm making up 60% of the total nodules described, 4-8mm 20% nodules of described nodules, >8mm making up 12% of nodules. Pulmonary Cysts represented 4% of nodules. Subpleural nodules described separately and making up a minority of nodules.

Conclusion
Conclusion: Cardiac CTs are an increasingly common modality used to evaluate coronary artery disease. Pulmonary nodules are a common finding with a prevalence in this study of 14%; this contributes significantly to a economic burden as well as exposure of the community to ionising radiation which there is growing community and professional concern around. The findings of this audit are in line with published data of a low risk of de novo malignancy in pulmonary nodules as well as a high prevalence of pulmonary nodules across the cohort.

Background
A small percentage of at-risk individuals ultimately develop lung cancer. This warrants a search for a strategy to identify individuals who will be developing lung cancer. The bronchial epithelium represents the field of carcinogenesis and may serve as surrogate tissue for biomarker of risk. Therefore we hypothesized that proteomic alterations in the histologically normal airway epithelium from risk stratified individuals allow us to derive a signature of risk of developing lung cancer.

Methods
We have collected bronchial brushings specimens from risk stratified individuals (Bach PB et al. 2003) and categorized them as Control (group 1), low (group 2) and high (group 3) risk groups. We collected three bronchial brushings specimens from each individual in a single session for acquiring shotgun proteomics (n=30). We also collected brushings from cancer patients whose risk criteria are similar to low (group 5) and high (group 6) risk groups as well as a group of never smoker lung cancer patients (group 4). Shotgun proteomics data were acquired from 3 technical replicates of pooled specimens from 5 individuals. Candidate biomarkers were selected based on group comparison and trend analysis.

Results
We have acquired shotgun proteomics data from bronchial brushings specimen from individuals at-risk for lung cancer development. A total of 4973 proteins have been identified. Jonchere-Terpestra trend test was performed among control, low risk and high risk groups (Groups 1, 2, and 3). Expression level of 316 proteins were altered (trend p <0.05) with 238 proteins demonstrating upward and 78 proteins downward trends. DNA repair and oxidative stress and the galactose metabolic enzyme pathway were enriched in the upward trend analysis.

Conclusion
We identified a proteomic candidate signature of risk of developing lung cancer from histologically normal bronchial epithelial cells. Integration of these results with those upcoming from two other molecular platforms will allow us to narrow down the molecular aberrations most predictive of risk. This signature will be validated in an independent cohort. Such signature of risk for lung cancer may serve as a powerful tool for lung cancer risk assessment and may provide the basis of patient selection for surveillance programs and chemoprevention. This work is funded by RO1 CA102353to PPM.

Background
CXR-screening has long been considered ineffective because RPTs have failed to demonstrate significant mortality reductions in populations randomized to CXR-screening. While these studies demonstrate significant survival advantages associated with CXR-screening, these advantages are widely believed to be spurious, due to the interpretation that CXR-screening leads to substantial overdiagnosis. Indeed, the overdiagnosis hypothesis is the only alternative to the conclusion that CXR-screening saved lives in these trials. The objective of this analysis is to assess the magnitude of overdiagnosis in the context of CXR-screening, and to determine whether survival or mortality provides the best measure of efficacy in existing RPTs.

Methods
The Memorial-Sloan-Kettering and Johns-Hopkins Lung Projects were designed to assess the effectiveness of sputum cytology, since participants were randomized to annual CXR-screening and sputum cytology every four-months or to annual CXR alone. In Mayo-Lung-Project (MLP), following a normal prevalence CXR, 9,192 smokers were randomized to an experimental group (EG) undergoing CXR and cytology every 4-months for 6-years followed by 3-years of observation, or to a control group (CG) observed for 9-years. Controls were advised to obtain annual CXR.

Results
In MSKLP and JHLP, there were no differences between EG and CG regarding mortality, survival, or stage distribution. However, 5-year survival in both trials was about three-fold superior to contemporary national statistics. This suggests a beneficial effect of CXR, assuming that overdiagnosis did not confound survival in EG and CG in both studies. In MLP, lung cancer mortality was 6% higher in EG (p=0.62), although 5-year survival was significantly superior in EG (34% vs. 13%; p=0.0021). The fact that lung cancer incidence was 30% higher in EG (p=0.013) led to the hypothesis that overdiagnosis was responsible for the mortality/survival discrepancy. However, MLP data are inconsistent with overdiagnosis. Long-term survival was achieved only among those undergoing resection. Half of resected cancers were apparently cured, while no unresected patient was cured (7-year survival: 50% vs. 0%: p<0.0001). Indeed, follow-up extended to >20 years in MLP continues to show higher incidence and mortality in EG. While interpreted as supporting overdiagnosis, long-term follow-up results are primarily informative as to randomization adequacy. Mathematical modeling studies indicate that in MLP-EG higher lung cancer incidence and mortality was due to population heterogeneity, due to problems with randomization, not to overdiagnosis.

Conclusion
Abundant evidence supports that the magnitude of overdiagnosis is minimal in the context of CXR-screening. Moreover, evidence from RPTs indicates that CXR-screening is superior to no screening. The survival endpoint is not meaningless in RPTs if one can exclude or adjust for confounding due to overdiagnosis and with sufficiently long follow-up. Because overdiagnosis was not a confounder of survival in MLP, survival most accurately reflected CXR effectiveness. Indeed, long-term survival provided an unbiased surrogate for cure. While the National Lung Screening Trial demonstrates that CT-screening is superior to CXR-screening, CXR is more widely accessible and relatively inexpensive. CXR-screening has the potential to substantially reduce the global burden of lung cancer mortality.

Background
Lung cancer is the leading cause of cancer death for both males and females in New South Wales (NSW), Australia. Similar to other countries, the five-year relative survival rate for lung cancer in NSW is low, reported at 17.6% (2005-07). The poor relative survival rate suggests that in most cases, diagnosis occurs late, predominantly in stage III or IV. A multi-channel public education campaign was implemented in NSW to increase awareness of three symptoms consistent with lung cancer (persistent cough, change in cough and blood in cough) and prompt people experiencing these symptoms to see their General Practitioner (GP) immediately. If people experiencing these symptoms seek a doctor as soon as symptoms are experienced, it has the potential to reduce diagnostic delay and hence, an improvement in the five-year relative survival of lung cancer.

Methods
The public education campaign, comprised mass media television advertising implemented over 5 weeks (March/April 2013) at 100 TARPs per week in conjunction with public relations activity and a direct mail communication to GPs across NSW. Quantitative telephone surveys were conducted in July 2012 (pre-campaign) and April 2013 (post-campaign) with adults over 45 years of age residing in NSW (both; n=1000). Unprompted and prompted recall of the campaign were measured and levels of unprompted awareness of symptoms consistent with lung cancer and confidence in recognising symptoms consistent with lung cancer were compared between the two surveys.

Results
Overall, 28.7% of the targeted population recalled the campaign unprompted, increasing to 60.5% when prompted. Pre- and post-campaign results demonstrate an increase in unprompted awareness of key lung cancer symptoms emphasised by the campaign; ‘coughing up blood’ (from 39.1% to 45.4% (p = 0.004)), ‘cough that does not go away for two or three weeks’ (from 17.0% to 40.7% (p = <0.001)) but no significant different in awareness of ‘worsening or change in cough’ (6.3% to 4.6% (p = 0.096)). Post-campaign awareness of these symptoms was higher amongst non-smokers than smokers, particularly for persistent cough. Encouragingly, confidence to recognise a symptom increased from 58.8% to 67.4% (p = <0.001).

Conclusion
These findings suggest that a mass media public education campaign can be an effective approach to increase awareness of symptoms consistent with lung cancer and confidence in recognising these symptoms. Increased awareness and recognition of symptoms can lead to earlier diagnosis and treatment.

Background
The ability to capture and characterise peripheral blood circulating tumour cells (CTCs) has the potential for the development of a blood test for cancer. A number of technological platforms are available to obtain CTCs including filtration-based devices utilising advances of antibody independent capture of cells. This technique shows promising results in experimental conditions; however, its performance has not yet been well evaluated in a clinical setting. We have evaluated diagnostic performance of filtration-based technology using cytomorphologic criteria in patients undergoing surgery for lung cancer.

Methods
From 06/03/2012 to 24/01/2013 we obtained and processed blood from 74 patients undergoing surgery for known or suspected lung cancer using ScreenCell[TM] Cyto devices. Captured cells were stained using H&E and independently assessed by two pathologists (AGN, AR) for the presence of atypical cells suspicious for cancer. Results were reported as confirmed cancer, suspicious or no evidence for cancer. Diagnostic performance was evaluated against the reference of cancer identified within surgically obtained specimens reported by a principal pathologist. Sensitivity and specificity analyses were undertaken. Inter-observer agreement was established by kappa-statistics.

Results
According to histopathology assessment, 42 patients (56.7%) had primary lung cancer, 18 patients (24.3%) had metastatic cancer (predominantly of colorectal origin), and 14 patients (18.9%) had benign lung diseases. The proportion of patients in which cells suspicious for cancer were identified was 39 (52.7%) and 42 (56.7%) as reported by two pathologists. Among those cases, 6 (15.4%) and 14 (33.3%) were reported as confirmatory. The agreement between the pathologists was 77% corresponding to a kappa-statistics of 53.7% indicating moderate agreement. In metastatic cancer patients, suspicious cells were discovered in 10 (55.6%) and 9 (50%) cases by two pathologists. In non-cancer patients, suspicious cell were found in 6 (42.8%) and 5 (35.7%) cases by two pathologists, respectively. The test performance for the diagnosis of cancer using cytomorphological criteria yielded poor-to-moderate sensitivity and specificity values, high positive predictive values and low negative predictive values (Table).

The performance of the diagnosis of cancer using filter-based antibody-independent technique of CTCs trapping

Statistic	Pathologist 1	Pathologist 2
Sensitivity (95% CI), %	55.0 (41.6-67.9)	61.7 (48.2-73.9)
Specificity (95% CI), %	57.1 (28.9-82.3)	64.3 (35.1-87.2)
Positive Predictive Value (95% CI), %	84.6 (69.5-94.1)	88.1 (74.4-96.0)
Negative Predictive Value (95% CI), %	22.9 (10.4-40.1)	28.1 (13.7-46.7)

Conclusion
The results of our study highlight the potential of filter-based antibody-independent technology to develop an accurate blood test for the diagnosis of cancer in the peripheral blood. However, conventional cytomorphological criteria used for the diagnosis provide inadequate sensitivity and specificity. Improved performance with immunocytochemistry is still required prior to further clinical validation.

Background
Thailand is among the countries with a high burden of tuberculosis (TB) with 137 new cases per 100,000 persons per year reported. The radiographic findings of TB can mimic lung cancer even after a complete anti-TB treatment. Currently, no clear evidence of benefits from lung cancer screening has been established in a high-risk population residing in an endemic area of TB.

Methods
A prospective study was started in July 2012 to determine the role of low-dose computerized tomography (LDCT) as a lung cancer screening tool in Thai heavy smokers (>30 pack-years) without a history of active TB within a recent year. Abnormal LDCT findings were categorized into 3 groups: negative, indeterminate, and suspicious for malignancy according to the baseline nodular volume and volume doubling time in a follow-up LDCT.

Results
As of April 2013, 580 subjects were enrolled. At baseline, 406 cases (70%) had abnormal LDCT findings including 16 cases (2.8%) suspicious for primary lung cancer. Two hundred and forty-five cases (42%) had benign characteristic nodules, 78 cases (13.4%) had inflammation or infection patterns, and 67 cases (11.6%) were indeterminate for malignancy. Of the suspicious primary lung cancer group, 3 cases had a previous lung nodule without any evidence of nodule progression, 13 cases underwent tissue biopsy with resulting lung cancers in 7 cases, caseous granuloma in 2 cases, and negative results in 4 cases. Two cases with the inflammation/infection pattern had asymptomatic progressive lesions in a 2-month follow-up LDCT and underwent bronchoscopy. One case was proven pulmonary TB while the other was culture-negative TB; both of them responded well to anti-TB drugs. Two additional primary lung cancer cases were detected at a 3-month follow-up LDCT yielding a total of 9 positive lung cancer cases (1.5% of total subjects) which consisted of 4 - stage I cancer, 1 - stage II cancer, 1 - stage III cancer, and 3 - stage IV cancers. Early-stage cases (5/9) received potentially curative surgery. Notably, 7 and 2 lung cancer cases were detected from suspicious and indeterminate group, respectively. Three other malignancies were also detected including 1 mediastinal lymphoma, 1 small cell of unknown primary and 1 cholangiocarcinoma.

Conclusion
Our preliminary result revealed that despite a high burden of TB in Thailand, LDCT screening in heavy smokers could yield a high rate of primary lung cancer in this population at risk.

Background
The association between chronic obstructive pulmonary disease (COPD) and lung cancer has been previously reported. However, the mechanism whereby emphysema (a destructive process) promotes lung carcinogenesis (a proliferative process) has not been adequately explained. Emphysema is associated with lower lung density while lung inflammation is associated with increase in lung density. We hypothesized that lung density and emphysema are independent predictors of malignancy risk of lung nodules found on screening low dose spiral CT (LDCT).

Methods
Image analysis was performed on a subset of LDCT scans (120 kVp, 40 mAs) from the Pan Canadian Early Detection of Lung Cancer Study and the BCCA Lung Health Study using the VIDA Diagnostics CT image analysis software Pulmonary Workstation 2. The lobe with the pulmonary nodule was first segmented. The average lung density surrounding the nodule was measured. Emphysema severity was defined as percentage of the lobe with -950 Hounsfield Units (HU). Multivariate logistic regression analysis was performed to determine if lung density and degree of emphysema were independently associated with malignant lung nodules.

Results
A total of 161 subjects with lung nodules ≤20 mm were studied. The clinical and CT characteristics are shown in the Table 1. Table 1. Study variables by lung cancer status

	No Cancer	Cancer	P-value
N=	95	66
Age	64±5	63±6	0.52
Gender : Men: Women	53% : 47%	38% : 62%	0.078
Current: Former smoker	61% : 39%	45% : 55%	0.051
Family history %	21%	38%	0.022
Nodule diameter	10.5 ± 3.0	12.9 ± 4.1	<0.001
Nodule Type – solid	53%	50%
- part solid	11%	29%	0.003
- Non-solid	37%	20%
Nodule location (upper versus middle or lower)	45%	65%	0.016
Spiculation (%)	18%	39%	0.017
Emphysema (visual score) % present	67%	78%	0.152
Density of lobe with nodule	-848±32	-837±32	0.024
% emphysema in lobe with nodule	9.3%±9.7	7.1%±6.9	0.09

While the presence of emphysema of any grade in both lungs by visual score was higher in the lung cancer group (78% versus 67%), the difference was not statistically significant in the univariate or multivariate analysis. Quantitative measurement of emphysema severity (area with -950 HU) in the same lobe where the lung nodule located showed that the degree of emphysema was less in patients with lung cancer (7.1% versus 9.3 %, P = 0.041 in the final multivariable logistic regression model consisting of family history, nodule size, type and spiculation). Lung density surrounding the lung nodule was significantly higher in the lung cancer group compared to the benign nodule group in univariate analysis (P = 0.024) but not in multivariable analysis.

Conclusion
Our results suggest lung inflammation as reflected by increase in lung density may be a more important factor in lung carcinogenesis while emphysema may be more of a dosimeter for lung damage by tobacco smoke exposure. Further studies in a larger dataset are being performed to determine the incremental value of lung density in predicting the malignancy risk of lung nodules ≤ 2cm detected by screening LDCT.

Background
For patients with advanced non-small cell lung cancer (NSCLC), it is unclear whether the anatomical location of their metastases has a different impact on survival. In this project, we sought to describe these metastatic sites using the 7th edition of the TNM classification and analyse their prognostic implications. A special emphasis was placed on brain metastases, as these patients are commonly excluded from participating in clinical trials, based on perceptions of inferior outcomes.

Methods
Consecutive patients diagnosed with NSCLC, between 2005 and 2009 at Hôpital du Sacré-Coeur de Montréal were included in this retrospective analysis. They were re-classified, from the TNM 6[th] edition, in use at the time, to the TNM 7. The following metastatic locations were considered: brain, pleura, contra lateral lung, extra-thoracic lymph nodes, soft tissues, pericardium, adrenal glands and skeleton. The treatments received were classified in one of these three groups: standard, non-standard and palliative care. Cox proportional hazards analysis was used to identify predictive factors for survival among the following parameters: age, body mass index, gender, histology, smoking history, type of treatment, number and location of metastatic sites.

Results
Five hundred thirty seven patients with stage IV NSCLC were identified. Their metastatic sites were, in decreasing order of frequency: skeleton (41%), pleura (32%), contra lateral lung (28%), brain (26%), liver (18%), extra-thoracic lymph nodes (16%), soft tissues (14%), adrenal glands (14%) and pericardium (6%). Forty-five percent of patients had metastases to 1 organ, 33% had 2 and 22% had 3 to 6 organs involved. In univariable analyses, metastases to the liver, adrenal glands or skeleton were associated with worse prognosis (HR=1,37; 1,38 and 1.34, respectively; p < 0.05). The patients with liver or adrenal metastases were more likely to have at least another metastatic site (90%). In multivariable analyses, not surprisingly, better survival was found in patients who could receive standard treatment, those having adenocarcinoma histology and less metastatic sites. Having contra lateral lung metastases conferred a better prognosis (HR=0.74, p=0.0037). Remarkably, brain metastases were not a prognostic factor in uni or multivariable analyses. This group of patients was more likely to have at least one of three better-risk factors: age < 65 (68%), brain as a unique metastatic site (37%), and more frequent in patients with adenocarcinoma than squamous cell cancer (27 vs. 13%). Conversely, they were less likely to be treated with palliative care compared to patients with extra-cerebral metastases (8 vs. 35%).

Conclusion
A history of brain metastases did not have a prognostic influence in our NSCLC patient cohort. This finding could be explained by the presence of better-risk factors in that group of patients. Brain metastases should not be considered by itself an exclusion criterion from clinical trials.

Background
Lung cancer is a major cause of mortality in Russia: 51,364 deaths in 2008, 18.0% of all cancer-related deaths (GLOBOCAN). Mutations in the EGFR gene are known to predict for sensitivity to EGFR tyrosine kinase inhibitors in patients with advanced non-small-cell lung cancer (NSCLC). Clinico-pathological characteristics associated with a higher prevalence of EGFR mutations are: adenocarcinoma histology, East Asian origin, non-smoking history, and female gender. This study aimed to document the characteristics, clinical management and outcomes of Russian patients with NSCLC, and to investigate associations between EGFR mutations and clinico-pathological parameters.

Methods
A non-interventional, prospective cohort study (ClinTrials ID: NCT01069835) carried out in a representative selection of hospitals in order to assess lung cancer management in countries throughout European and Asian parts of Russia. Patients with confirmed NSCLC attending participating centres for the first time between 1 February 2010 and 31 March 2011 were enrolled and followed-up according to routine practice for a minimum of 12 months or until death.

Results
A total of 838 patients were enrolled at 33 sites across the Russian Federation. Baseline characteristics were: mean age, 58.7 (SD ±8.5) years; male, 78.4%; European, 98.0%; Russian, 79.6%; Tatar, 3.8%; Ukrainian, 3.1%; Byelorussian, 1.3%; current smokers, 49.4%; never-smokers, 26.5%; ECOG performance status (PS) 1, 65.6%; disease stage IV, 25.4%; stage III, 37.8%; stage I/II, 36.7%. Metastases were found in 38.1% of patients; the most common metastatic sites included: respiratory system, 70.2%; pleura, 17.2%; bone, 11.9%; liver, 10.3%. EGFR mutations were detected in 85/838 (10.1%) patients (84/821 [10.2%] European patients and 1/15 [6.7%] Asian patients), who were mostly women, 69.4%; <70 years old, 85.9%; never smokers, 71.8%; PS 1, 59.3%; adenocarcinoma, 58.8%. EGFR mutation rates by histological type were: squamous-cell carcinoma (SCC), 18/455 (4.0%); adenocarcinoma, 50/260 (19.2%); adenocarcinoma bronchioloalveolar, 11/54 (20.4%); large-cell carcinoma, 2/24 (8.3%); adenosquamous carcinoma, 2/19 (10.5%); other histology, 2/26 (7.7%). Logistic regression analysis of EGFR mutations; statistically significant odds ratios: male vs female, 0.086 (p<0.0001); any smoking history vs no smoking history, 0.110 (p<0.0001); non-SCC vs SCC, 5.365 (p<0.0001); increase in age (10 years), 1.391 (p=0.0227). Platinum-containing doublets were the most commonly used chemotherapy at first (83.5%) and second line (51.0%). In patients who received first- and second-line chemotherapy, objective response rate, disease progression rate and median progression-free survival (PFS) were, respectively, 16.1% and 3.6%, 63.8% and 76.8%, 35.0 weeks and 19.4 weeks. Median PFS (weeks) after first-line chemotherapy by histology and EGFR mutation status was: SCC, 36.3; non-SCC, 34.0; EGFR mutation positive, 36.9; EGFR mutation negative, 34.3.

Conclusion
In this cohort, the proportion of patients with EGFR mutation-positive NSCLC was similar to other studies of NSCLC in Caucasian populations. EGFR mutation status was significantly associated with female gender and non-smoking history, in-line with previous studies of Asian and Caucasian patients with advanced NSCLC. This study contributes to a better understanding of prognostic and predictive factors of NSCLC, which will enable optimal treatment selection in future clinical practice.

Background
The LungPath study is a national multi-centre survey of lung cancer diagnosis and staging practice in England. As part of the project, we looked for variation in access to key investigations such as EBUS and PET-CT scanning and linked this to how likely these studies were used to be in different centres and the impact these differences may have on patient care.

Methods
Twenty willing English lung cancer centres were randomly selected to participate in the LungPath study. Participating centres agreed to submit data on each new lung cancer patient seen during the study period of six months from January to June 2012. Data collected included clinical information such as age, gender and performance status, the dates of all radiological investigations performed and anonymised pathology reports from all other investigations performed. The data collected was used to map each individual patient’s diagnostic pathway. In addition, we collected information about typical waiting times for key investigations and whether these investigations were available on-site or at other institutions. We analysed the patient data to see if the availability of investigations such as EBUS and PET-CT impacted the patient pathways in each centre.

Results
There were significant differences between centres in the availability of EBUS and PET-scanning. Approximately half of the units surveyed reported waiting times for EBUS and PET-scanning of one week or less while the other half reported longer waits, typically two weeks and sometimes longer. There were large differences in the proportions of patients that underwent EBUS or PET-scanning from centre to centre with patients up to six times more likely to receive an EBUS and four times more likely to receive a PET-scan in some centres than others. There was a clear relationship between the use of the investiagtion and the waiting time. We also found that the point in the diagnostic pathways that these investigations were used varied and in many cases differed from best practice guidelines; several centres routinely performed EBUS as a separate procedure after a bronchoscopy had already been performed.

Conclusion
There are marked differences in the availability and use of EBUS and PET-scanning within different lung cancer units involved in diagnosing and staging lung cancer in England. There is a need for commissioners to ensure fairer service provision across England and opportunities for education of clinicians to make best use of the available resources.

Background
The treatment of patients (pts) with advanced non–small-cell lung cancer (NSCLC) is based on clinical trials experience and molecular characteristics. Although molecular testing has become more mainstream, knowledge gaps need to be addressed to optimize treatment decision. Delay to treatment may compromise palliative efficacy. To evaluate therapeutic delays, procedures for 71 biopsies from pts with stage IV NSCLC newly diagnosed submitted to molecular testing were retrospectively analysed.

Methods
Demographic data, site of biopsy, type of first systemic treatment, EGFR mutations in exons 18, 19 and 21 using a dual technical approach (direct sequencing of polymerase chain reaction (PCR) products followed by bidirectional sequencing by Sanger method), EML-4-ALK rearrangements by fluorescence in-situ hybridization (both done in a single central laboratory sponsored by pharmaceutical companies), year of diagnosis, time from first consultation to molecular testing results (diagnostic delay, DD), and time from first consultation to treatment initiation (therapeutic delay, TD) were obtained from 71consecutive pts with stage IV NSCLC newly diagnosed seen from 8/2010 to 5/2013 in two general oncologic institutions located on Patagonia, Argentina.

Results
Median age was 59 years (range 22-82), 38 % were women. 86% were adenocarcinoma. 62% of specimens were derived from primary tumors, 38% from metastatic sites. EGFR analyses was done in all 71 specimens, 13 have insufficient material for testing (18%) (group ND) (27% of ND in biopsies obtained during 2010-2011 versus 10% of those seen during 2012-2013, p: 0.04). 8 evaluable pts (14%) presented any EGFR mutation, whereas 50 (86%) were negative. 11 pts were evaluated for ALK translocation in 2013. Only 1 pt was positive. As a consequence, 9 pts from the entire cohort were able to be treated with molecular directed treatments as first line (13%, group POS) Median DD for EGFR analyses was 18 days (range 8-32). It was 13 days during 2013 probably due to a better multidisciplinary approach. Median DD for ALK analyses was 18 days (range 14-32). Median TD was 24 days (13-77) for the entire cohort. It was 23 days (13-36) for Non-POS pts and 39 days (21-77) for POS pts (p:0.0001) Despite positivity only 4/9 POS pts were treated with TKI as first line treatment. Main reasons for no treatment were excessive delay for drug accessibility (5) and performance status deterioration (1). DD represent 75 % of TD in non-POS pts, and 48 % in POS pts

Conclusion
1- Molecular testing permit to indicate first line TKI treatments in 13% of pts 2- DD may compromise treatment in all pts, but specially in the numerous non-POS pts who otherwise would have been starting therapy after first consultation. 3- With a multidisciplinary approach, it is possible to have better testing results and to reduce DD. 4- Drug accessibility is an important cause of TD in POS pts 5- TD for any preventable reasons should be promptly reduced as survival or quality of life may be compromised.

Background
EGFR mutations are highly associated with pulmonary adenocarcinoma in East Asian never-smoker population. The aim of this study is to evaluate the clinical characteristics of non-small cell lung cancer (NSCLC) with EGFR mutations in never smokers.

Methods
From June 2011 to Dec 2012, consecutive NSCLC patients who admitted for initial diagnosis of lung cancer without smoking habit were recruited in Seoul National University Hospital.

Results
Among 979 of lung cancer patients, 342 never smokers with NSCLC (38.6%) were analyzed. Mean age was 62.7 ± 10.6 years old and mean BMI was 23.8 ± 3.0 kg/m[2]. Adenocarcinoma (307, 89.8%) was the major pathologic diagnosis. The most common clinical stage was IV (134, 39.2%), followed by stage Ia (115, 33.6%). Among 303 patients who underwent EGFR mutation test, 180 (59.4%) patients had EGFR mutations. The never smokers with NSCLC-EGFR (+) showed better performance status (P=0.006), lower uptake of SUV in main mass (8.9 ± 5.9 vs. 11.4 ± 6.8, P=0.003), early clinical stage (≤ II, P=0.009), and higher proportion of curative lung resection (63.3% vs. 41.2%, P=0.005) compared to the patients with NSCLC-EGFR (-).

Conclusion
EGFR mutations are associated with early stage of non-small cell lung cancer in never smokers.

Background
In many cancers, the N-classification of current TNM stage is categorized by the number of positive lymph nodes. However, for non-small cell lung cancer (NSCLC), it is classified by the involvement of specific regional node groups. The aim of this study is to make a new N-classification grouping (nN stage) based upon the number of metastatic lymph nodes and to evaluate its prognostic significance in predicting outcome after resection of NSCLC.

Methods
Between January 2000 and April 2011, 2319 patients underwent surgery for NSCLC in our hospital. A total of 1982 patients was selected by excluding patients who received preoperative induction therapy or did not get curative resection and those with pN3 metastatic lymph nodes. Prospectively collected clinical information as well as pathologic variables were retrospectively analyzed. The recursive partitioning analysis was applied to define the most significant cut-off number of metastatic lymph nodes. We then analyzed overall and disease-free survival using the new nN stage grouping to test if it can provide more accurate classification compared to the conventional N stage grouping.

Results
Recursive partitioning analysis identified patients could be divided into three distinct groups according to the number of metastatic lymph nodes: nN0 (none), nN1 (1-7), nN2 (>7). Among 1982 patients, 1371 patients were nN0, 538 were nN1, and 73 were nN2. The 5-year overall survival rates were 79.8%, 62.1% and 36.1% for nN0, nN1, and nN2, respectively, and they were statistically different (p < 0.001, log-rank test). For conventional N stage, 1371 patients were pN0, 284 were pN1, and 327 were pN2. The 5-year survival rates were 79.8%, 63.8%, and 54.8% for pN0, pN1, and pN2 stages, respectively (p < 0.001, log-rank test). The chi-square value of nN stage was superior to that of pN stage (141.02 vs. 117.16). When we further analyzed those with pN2 patients, the nN1/pN2 group showed a significantly better survival rate than nN2/pN2 group (p < 0.001, log-rank test). Moreover, the overall survival of nN1/pN2 patients was not different from that of nN1/pN1 patients. (p = 0.074, log-rank test) Figure 1

Conclusion
The nN-classification seemed to predict long-term survival more accurately compared to conventional N stage grouping. Our result suggested the new N stage grouping based on the number of metastatic lymph node should be considered for the next revision of the TNM classification system for NSCLC.

Background
Mutations in the epidermal growth factor receptor (EGFR) gene have been identified in lung adenocarcinomas and are associated with a high response to EGFR tyrosine kinase inhibitors. EGFR mutations can be detected in tumour tissue, cytology specimens and blood from lung cancer patients. Thus far, EGFR mutation analysis has not been systematically demonstrated for sputum samples. The aim of the present study was to determine whether EGFR mutation analysis is feasible on sputum samples, employing different assays in a multicenter study.

Methods
Sputum samples were collected from 10 lung cancer patients with confirmed EGFR mutation in their tumour tissue, 10 lung cancer patients without evidence of an EGFR mutation, and 10 patients with chronic obstructive pulmonary disease (COPD). DNA was isolated from the sputum and used for mutation analysis by Cycleave PCR, COLD-PCR, PangaeaBiotech SL technology (PST), and High Resolution Melting, respectively. Targeted resequencing (TruSeq Amplicon Cancer Panel) and droplet digital PCR were additionally performed on the 10 samples with EGFR mutation.

Results
Dependent on the assay, EGFR mutations could be detected in 30-50% of the sputum samples of patients with EGFR mutations (Table). The different techniques revealed consistent results, with slightly higher sensitivity for PST. Neither the lung cancer patients without EGFR mutation nor the COPD controls tested positive for EGFR mutations in their sputum samples, indicating high clinical specificity of all assays.

Subject	Gender	Age (years)	Tumour stage	EGFR mutation status of tumour tissue[1]	EGFR mutation analysis on sputum specimens[2]
					Cycleave PCR	COLD-PCR	PST[3]	HRM-sequencing	Cytology[4]
A	F	72	IV	Del E746-A750	0	0	0	0	0
B	M	66	I	Del E746-A750	0	2	0	0	0
C[6]	F	78	IV	Del E746-A750	1	1	1	1	2
D	F	46	III	Del E746-A750	0	0	1	0	0
E[6]	M	54	IV	Del E746-A750	1	1	1	1	0
F	F	49	III	Del E746-A750 & c.2369C>T [p.T790M]	0	0	0	0	0
G	F	54	IV	Del E746-A750 & c.2369C>T [p.T790M]	0	0	1[5]	0	1
H	F	73	IV	c.2753T>G [p.L858R]	0	0	0	0	0
I	F	61	IV	c.2753T>G [p.L858R]	0	0	0	0	0
J[6]	M	60	IV	Del E746-A750	1	1	1	1	2

[1 ]del E746-A750= deletion exon 19 [2] mutation identified: 0=no, 1=yes, 2=dubious [3] exclusively del19 and L858R were assessed [4] tumour cells: 0=no, 1=yes, 2=in related sample of same patient [5 ]only del19 detected [6 ]TSACP and ddPCR both tested EGFR mutation (del19) positive.

Conclusion
EGFR mutations can be detected in sputum samples from patients with EGFR-mutated non-small cell lung cancer, which may replace biopsy procedure for some patients.

Background
Diagnosis of postoperative recurrence of lung cancer usually depends on radiologic examinations. However, the diagnostic yield of radiological examination is limited and it often times show false-positive result. Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) is well accepted modality for the pre-operative lymph node staging in patients with lung cancer; however, the efficacy of diagnosis for post-operative recurrence still remains unclear. In this study, usefulness of EBUS-TBNA pathologic confirmation of regional node metastasis was investigated in comparison with 18F-fluorodeoxyglucose positron emission tomography (FDG-PET).

Methods
The patients who were suspected to have lymph node recurrence by routine chest CT follow-up after radical surgery for lung cancer were retrospectively investigated, and diagnostic yields of FDG-PET and EBUS-TBNA for the recurrence were compared. The cut-off value for positive results by PET was standard uptake value (SUV max) more than 2.5. Rapid on-site cytological evaluation was performed during the procedure of EBUS-TBNA for convenience and pathological diagnosis was employed by independent pathologist. A dedicated 22-gauge needle was used for TBNA. Final decision of presence of nodal recurrence was made based on pathological findings of cancer recurrence for EBUS-TBNA sample, and that of absence of the recurrence was made based on radiologic follow-up for more than 6 month.

Results
A total of 40 patients were eligible for this study. The mean duration between thoracotomy and EBUS-TBNA was 23.5 months, and the median follow-up period after EBUS-TBNA was 21.8 months. Diagnostic sensitivity, specificity and accuracy of EBUS-TBNA was 100% for each whereas those of FDG-PET were 95.8%, 12.5%, and 62.5%, respectively. 24 patients with metastatic lymph node confirmed by EBUS-TBNA showed significantly unfavorable prognosis than 16 patients with negative result by EBUS-TBNA (p=0.024). 22 out of the 24 patients who diagnosed as recurrence received anti-cancer treatments properly. 14 patients with positive results by FDG-PET but negative by EBUS-TBNA were determined as negative (false positive) since no deterioration of the nodal status was confirmed by radiological follow-up. Pathological findings of these false-positive lymph nodes showed 12 anthracosis and 2 non-specific granuloma.

Conclusion
Minimally-invasive diagnosis by EBUS-TBNA should be indicated when regional lymph node recurrence is suspected since radiologic modalities frequently recognize benign lesions as positive. The accurate diagnosis by EBUS-TBNA reduces fertile cancer treatment and improves patient management. Figure 1

Background
Patients which are treated due to various malignant tumours, should be investigated by pneumologists because of having intrathoracic lympadenopathy and/or pulmonary infiltrates. We should consider either progression of first malignant tumour, secondary and tertiary malignancies, adverse reactions on therapy, infectious diseases, or, last but not least, systemic diseases, e.g. sarcoidosis. However in sarcoidosis patients, increased risk of hematologic and lung malignancies were reported. The occurrence of intrathoracic lymphadenopathy with or without pulmonary infiltrates is subject to frequent diagnostic difficulties in patients with cancer history. Accumulation of pulmonary infiltrates and lymphadenopathy may be mistakenly interpreted as metastases. We present 41 patients with malignancies, who contracted sarcoidosis simultaneously, subsequently or before the diagnosis of malignancy. The coincidence of both diseases hasn't been published in the Czech Republic yet. In these patients, according to a number of studies, increased prevalence of malignant diseases was found. Sarcoidosis is a non-tumor, chronic, systemic granulomatous disease of unknown etiology and characterized by T-cell dysfunction. It affects mainly the lungs and lymphatic system, but also other parenchymal organs, the skin, the eyes, the heart. The generally accepted hypothesis is some environmental factors may promote the development of sarcoidosis in genetically susceptible individuals. Characteristic lesions are noncaseating granulomas, which consist of epithelioid cells and T-lymphocytes.

Methods
During the course of 28 years we have examined around 500 patients with sarcoidosis. We selected patients with sarcoidosis and also malignant diseases from these.Using the method of retrospective study, we evaluated the frequency of the coincidence of the period since 1996.

Results
Of all the investigated we follow a group of 41 patients with malignancy and sarcoidosis. In 33 patients there was the primary tumor, in 7 patients primary sarcoidosis, 1 patient had two diagnoses simultaneously. These are 18 men and 22 women with the mean age of 62,3 years. The spectrum of malignancies covers hematological malignancy (6), lung (3), breast (13), head and neck (4), intestine (3), melanoma (4), seminoma (3) and other (5). All patients had sarcoidosis diagnosed by biopsy (23) by CT or PET / CT and the bronchioalveolar lavages (18). Mediastinal lymphadenopathy was presented by 37 patients, pulmonary infiltrates and nodules were detected in 31 patients, infiltrates only 3 patients, extrapulmonary disease in 9 patients. The interval between sarcoidosis and the primary cancer was 0-552 months with a median of 36.Treatment with corticosteroids (or in combination with immunosuppressive agents) had to be initiated in 28 patients, 12 patients were left without treatment. 31 patients showed regression or disappearance of lesions, 8 patients were stable and disease didn't progress, the development of 2 patient is still unknown.

Conclusion
Differential diagnosis of pulmonary lesions is a common problem in patients with primary malignant disease in history. Not every pulmonary lesion must be just a manifestation of the disease, but as we argue in this presentation, it may be a manifestation of sarcoidosis. Therefore, we stress the need of histological verification of each newly formed pathological finding. The exclusion of generalized malignancy should have a significant influence on the treatment and survival.

Background
A fusion gene between the anaplastic lymphoma kinase (ALK) gene and echinoderm microtubule-associated protein-like 4 (EML4) occurs in nearly 5% of cases of non-small cell lung cancer (NSCLC). This fusion gene leads to the production of the EML4-ALK tyrosine kinase, which is considered as one of the oncogene driver mutations in NSCLC. The ALK inhibitor, crizotinib, showed remarkable antitumor activity in patients with advanced ALK-positive NSCLC, and was approved in Japan in 2011. Although the break apart FISH is considered as the gold standard to identify ALK rearrangement, pre-screening by immunohistochemistry (IHC) have been proposed in Japan because diagnosable number of cases is limited. However, disagreement of IHC and FISH are indicated. Therefore, the ALK rearrangements of NSCLC in our institute were screened by the IHC, FISH and RT-PCR, and that compatibility was examined.

Methods
EML4-ALK rearrangements of 176 examples were screened by IHC(5A4), FISH, and RT-PCR among 179 patients diagnosed as NSCLC in Kansai Medical University Hirakata Hospital from June 2012 to December 2012.

Results
Ten patients (5.7%) were diagnosed as ALK positive NSCLC. Five were male, and median age was 63 (46-81). The results by each technique of ten cases are shown in Table.

patient	IHC	FISH	RT-PCR
1	positive	positive	positive
2	positive	positive	negative
3	positive	positive	negative
4	positive	positive	negative
5	positive	positive	negative
6	negative	positive	negative
7	negative	positive	negative
8	negative	positive	negative
9	negative	negative	positive
10	negative	negative	positive

Conclusion
Although IHC is useful as a large-scale screening test, many false negative do exist. According to our results, there are few cases in agreement between IHC, FISH and RT-PCR. Various laboratory tests for ALK rearrangement detection are required for effective medication.

Background
The prediction of prognosis has been essential for programing cancer treatment strategies. In NSCLC, the TNM staging system is a worldwide standard form and there have been considerate improvements in its accuracy. However, the TNM staging system still has many limitations. It cannot reflect all the detailed pronostic parameters. In the AJCC system, prognostic factors are used to make a mutually exclusive partitioning of patients and ordering of patients assumes that all the patients receive surgery only. Therefore, treatment-specific prognosis cannot be predicted in the AJCC system. The prognoses of TNM stages may be different between institutes, however the same hazard ratio represents the same prognosis. So we propose a hazard ratio scoring system as an alternative prognosis predition method which can be used for specific clinical purposes.

Methods
A nomogram was developed and validated using clinical data of 2,115 patients who received surgery for NSCLC. Patients were randomly divided into the training set (n=1,060) and validation set (n=1,055). Nomogram predictions for 3- and 5-year overall survival were calculated for each patient in training set by using Cox proportional hazard model and compared with actual survival. Validation set was used for evaluating nomogram and AJCC staging.

Results
The median overall survival was 55.6 months for the training set and 58.1 months for validation set. In the prediction of 5-year survival, the nomogram showed an area under the receiver operating characteristic curve of 0.78 (95% CI, 0.76-0.79) in the training set. The validation set showed a good discrimination with an AUC of 0.85 (95% CI, 0.82-0.89). AJCC TNM staging system showed an AUC of 0.74 (95% CI, 0.70-0.79) with a training set.

Conclusion
The nomogram represented the survival results of a single institute more accurately than AJCC TNM staging system. The hazard scoring system cannot replace the position of the AJCC TNM system, however it can be used for programing treatment strategies concerning new or site-specific treatment methods, biomarkers and future clinical parameters.

Background
Lung cancer is one of the most common cancers worldwide. It is the leading cause of cancer death in both men and women in the West countries and in Latin America too. In Argentina, 15% of cancer deaths are because of lung cancer (9000 patients per year). The histology and the analysis of EGFR and ALK must be done in all NSCLC. To obtain the tissue, the histological diagnosis and the biological studies is particularly difficult in many centers in our country because we have to send the tissue to other cities. We want to know if we meet the world’s standard in our Institutions. Objetive:The primary objective is to assess whether the material obtained by different methods is enough for correct histopathological diagnosis and molecular biology analysis. We also evaluated the prevalence of different histological subtypes, the prevalence of EGFR mutations in patients with adenocarcinoma, the methodology for obtaining tumor tissue, and the delay in evaluation of EGFR mutation status, and currently in ALK translocations.

Methods
A retrospective study analyzed 229 consecutive patients with diagnosed NSCLC in Instituto Oncológico de Cordoba. The diagnostic method used for obtaining tissue was FNA CT-guided in 120 patients (52%), bronchoscopy in 69 (30%), surgery in 29 (13%) and pleural effusion cytology in 11 (5%). EGFR mutations were performed by PCR.

Results
A subtype was determined in 218(95%) out of the 229 patients studied; 11(5%) were classified as NOS. 147 patients(64%) were adenocarcinoma, 59(26%) squamous cell carcinoma, 12(5%) large cell carcinoma and 11 carcinoma NOS (5%). In the last period, 66 patients’ tissues with adenocarcinoma were sent for EGFR mutations analysis: In 38 patients (58%), tumor sample was optimal and in 28 (42%) patients, it was suboptimal or inappropriate for analysis. Of patient with optimal sample, 22(34%) presented different mutations and 16 (24%) patients were Wild Type. The time to obtain the result was 1.86 weeks in those with adequate sample (2.1 weeks vs 1.6 weeks for those with insufficient or inadequate sample).

Conclusion
In 95% of cases, we could determine the histological subtype and only 5% were NOS. Out of the 66 patients in whom we can evaluate the EGFR, 42% of the samples could not be analyzed due to insufficient material or inadequate fixation and processing techniques. These results are higher but it is not easy to compare them with local and Latin American statistics as they are not enough. The percentage of TKIs sensitivity was 21%, similar to national statistics published (19.3%) and western countries. Latin American average is 33.2% in one study. The time to obtain the results of EGFR mutation is consistent with recent guidelines proposed by the International Association for the Study of Lung Cancer. In our region, we have to work hard in multidisciplinary issues to obtain appropriate tissues in NSCLC.

Background
The incidence of lung cancer in the Czech Republic is high. It has reached 91/100 000 in men and 31/100 000 in women. The TNM stages I and II are diagnosed in less than 20% of all patients. Patients with COPD are at higher risk of lung cancer. How exactly emphysema affects the development of malignancy and how many patients with/without emphysema will suffer from lung cancer is unknown. Likewise, we do not know how many patients with existing lung cancer are diagnosed post-mortem and we have been interested in whether this group of patients shares any common features. Therefore, we decided to carry out an autopsy study which would provide answers to our questions.

Methods
We compared clinical data with autopsy findings in patients, both adults and children, from the largest hospital in the Czech Republic. During two studied years (2011-2013) we obtained complete sets of data on 708 patients.

Results
The studied group included 398 men and 310 women, with a median age of 71 years. Autopsies found a total of 55 lung cancer cases (7.8% of the set). Out of these, 24 (44%) tumours were diagnosed only by autopsy. In 40% of autopsy cases emphysema was proved along with a pulmonary malignancy. In the group of 24 patients with newly diagnosed lung cancer, emphysema was more common than in patients with a previously known tumour (50% vs. 32%).Patients with lung cancer discovered during autopsy died due to CHF in 8 cases, pneumonia in 6 cases, exacerbation of COPD in 2 cases, stroke in 2 cases, other tumours in 3 cases, and abdominal conditions in 3 cases. Emphysema during lifetime has been described less often than post-mortem (84 vs. 197 patients). The prevalence of autopsy proven emphysema was therefore 27.8%, the prevalence determined by clinical data was 11.8%. Lung cancer was diagnosed in one of every 10 patients with emphysema (20/197) and one in every 15 patients without emphysema (35/511).

Conclusion
The number of patients undiagnosed with lung cancer during their lifetime (44%) was higher than both literature data and our expectations. Studies carried out in the past accounted for 1 newly diagnosed lung cancer case per 80 autopsies, whereas in our group the ratio was 1:30 (Ošťádal, Studia pneumologica, 1994). On the other hand 27.8% cases with emphysema confirmed by autopsy corresponded with data in literature (15-65%; Thurlbeck, W. et Wright, J. L.: Chronic Airflow Obstruction) and results of similar previous studies which proved emphysema in 53% of autopsies on cases around the age of 70 years (Dalquen, P.: Incidence of Pulmonary Emphysema, and Study of 467 randomised autopsy cases 1, Beiträge zur Pathologie, 153, 4, 330-381). Our further results confirmed our clinical experience about the more frequent occurrence of lung cancer in patients with emphysema (1 in 10) compared with those without emphysema (1 in 15). "Supported by Projects (Ministry of Health) of conceptual development of research organization 00064203 (FN Motol, Prague, Czech Republic)."

Background
Lung cancer is the leading cause of cancer related death worldwide when considering both genders. The great effort to reduce smoking and introduce of cigarette has changed lung cancer epidemiology. In developed countries the increasing incidence of adenocarcinoma and decrease of squamous cell carcinoma are well known. Other characteristic reported is the rising number of women with the disease. Better understanding of current lung cancer epidemiology is necessary for the appropriate design of public health strategies for prevention, diagnosis and treatment.

Methods
Retrospective analysis of all patients with non-small cell lung cancer (NSCLC) treated with lung resection between 1986 and 2010 in a University Hospital of South Brazil. Analysis was divided in three periods 1986-1990, 1991-2000 and 2001-2010. Histological diagnosis was performed by the same pathology group and all staging was updated according to the new IASLC 7th edition. All analyses were performed using the SAS program version 13.

Results
We studied 817 patients who underwent lung resection for NSCLC from 1986 to 2010. 70.0% were males, average age 61.4 years, 44.2% squamous cell carcinoma and 40.0% adenocarcinoma, 26.7% stage IIIA. The female proportion increased from 22.0% in the first period to 36.0% in the last decade. Mean age at surgery treatment was 52.7 years for women and 57.3 for men in the first period, and 60.1 for woman and 63.9 for men in the last period (p<0.001). The proportion of squamous cell changed from 49.1% initially to 38.7% on the last period (p=0.017). In comparison, the adenocarcinoma prevalence increased from 35.4%, 39.6% and most recently 41.21%. Of the total NSCLC patients, females with adenocarcinoma represented 9.4% in the first period, 12.5% in the second and 16.8% in last period. Patients with stage IIIA represented 27.9% in the last decade. Type of surgery was predominantly lobectomy. Pneumonectomy was the surgical procedure in 21.9%, 18.8% and 16.8% of cases in each period respectively (p<0.03).

Conclusion
In this cohort of patients in south Brazil, gender analysis shows that rates of lung cancer in females is raising over the last three decades but has not surpassed men rates. The proportion of adenocarcinoma in females has increased. The significant decrease of pneumonectomy rates probably reflects changes on surgical management techniques and indication. The mean average age for patients undergoing surgical treatment has increased for both men and women. These findings are in accordance with the data of developed countries.

Background
EGFR activating mutations in pulmonary adenocarcinoma does confer better prognosis and are also predictive of higher response rates to both chemotherapy and EGFR-tyrosine kinase inhibitors. Therefore, EGFR genotyping in these patients (pts) is a very helpful biomarker for treatment selection. Here we aimed to report the results of consecutive EGFR genotyping in our Institution in Sao Paulo - Brazil.

Methods
It is a prospective, observational study on all consecutively tested samples from pts diagnosed with pulmonary adenocarcinoma and treated at ICESP. All samples were formalin-fixed and paraffin-embedded. Tumor areas were selected and macrodissected, followed by whole DNA extraction and amplification by PCR. EGFR genotyping was performed through DNA sequencing (exons 18, 19, 20 and 21) by Sanger´s methodology.

Results
191 pts had tumor samples genotyped between Aug/2011 and Apr/2013. Median age was 64 y (17-90), 106 (56%) female. According to ethnicity, 154 pts were Caucasian (81%), 26 African-American (14%) and 11 Asian (6%). Seventy pts were classified as never-smokers (37%), 23 (12%) as light-smokers (≤ 10 p.y.) and 95 as current smokers (51%). EGFR activating mutations could be identified in 54 out of 191 samples (28%): 35 were exon 19 deletions (65%), 15 were L858R mutation in exon 21 (30%), and three were rare mutations (G719S and G719A in exon 18, and V774M in exon 20). These mutations were found to be more frequent in females than in males (56% vs. 45%, p=0.035), and in never-smokers and light-smokers than in current smokers (77% vs. 20%, p<0.0001). It is noteworthy to mention that 11 mutations were detected in current smokers. All tumors harboring EGFR activating mutations presented TTF-1 expression by immunohistochemistry, and among those seven TTF-1-negative adenocarcinomas, no mutation was detected (p=0.0969). In a mean follow-up of 12 months, 77 pts were dead. Median overall survival was not reached in those pts whose tumors harboring EGFR-activating mutations, versus 19 months in pts with wild-type EGFR tumors (HR 0.40; 95%CI 0.29-0.78, p=0.003).

Conclusion
In this group of pts, the frequency of EGFR activating mutations was 28%, being more frequent in females, and never-smokers or light smokers, as previously described. Indeed, the presence of EGFR activating mutations was a favorable prognostic factor. The data here presented does reinforce the importance of testing EGFR activating mutations in all pts with TTF-1-positive, pulmonary adenocarcinoma.

Background
Cancer and tuberculosis are two conditions that can coexist, and their association is more common in areas of high prevalence of tuberculosis like developing countries.

Methods
To report the clinical characteristics of patients with cancer and tuberculosis association treated at a chest hospital. In this restrospective study, we reviewed the records of 970 patients treated for tuberculosis during the period 1998-2010.

Results
There were 16 patients with cancer and tuberculosis association. The average age was 55.25 years, 11 males and 5 females. Cancer was diagnosed before tuberculosis in 12 patients. The median time to onset of infection was 24 months. Tuberculosis preceded the cancer diagnosis in 3 patients and was obtained simultaneously by surgical biopsy in 2 cases. The most frequent association was with pulmonary cancer and hematological malignancies, 3 patients each. Pulmonary tuberculosis was present in 11 patients, 4 were extrapulmonary and 1 patient had both pulmonary and extrapulmonary disease. Bacteriological confirmation was obtained in 49% of patients and histopathology in 31.3%. Regarding to the evolution of tuberculosis, 50% completed treatment, 25% were derived, lost to follow up in 12.5% and 6.3% died during treatment of infection .

Conclusion
The probability of association of cancer and tuberculosis in the same patient should be taken into account in countries with high prevalence of tuberculosis. The coexistence of these two diseases frequently poses diagnostic difficulties and delay in treatment of both diseases. In countries with high rates of latent TB the oncologist should consider the possibility of reactivation of TB secondary to alterations of immunity produced by tumor and chemo or radiation treatments.

Background
No clear consensus has been reached on the XPA gene A23G (rs1800975) polymorphism and lung cancer risk. We performed a meta-analysis in an effort to systematically explore the possible association.

Methods
Case-control studies in English and Chinese publications performed with human subjects were identified by searching MEDLINE, EMBASE, Wanfang and CNKI databases prior to June 2013. References of retrieved articles were also screened. According to the inclusion criteria, 10 articles (12 studies) were finally included. The fixed-effects model and the random-effects model were applied for dichotomous outcomes to combine the results of the individual studies.

Results
Overall, statistical association could be found between A23G polymorphism and lung cancer risk in recessive genetic model (AA vs. (AG+GG)) (P=0.001, OR=1.21, 95%CI [1.08–1.35], P~heterogeneity~=0.11, fixed-effects model) and in the homozygote comparison (AA vs. GG) (P=0.03, OR=1.15, 95%CI [1.01–1.31], P~heterogeneity~=0.14, fixed-effects model). In East Asians, significant association was found in allele comparison model (A vs. G) (P=0.03, OR=1.13, 95%CI [1.01–1.26], P~heterogeneity~=0.39, fixed-effects model), in recessive genetic model (AA vs. (AG+GG)) (P=0.005, OR=1.30, 95%CI [1.08–1.56], P~heterogeneity~=0.58, fixed-effects model) and in the homozygote comparison (AA vs. GG) (P=0.02, OR=1.30, 95%CI[1.04–1.63], P~heterogeneity~=0.39, fixed-effects model). No evidence suggested that A23G polymorphism might associate with lung cancer risk in the Caucasians or African-Americans. Stratification analysis was performed by histologic types and indicated that AA genotype might represent a risk factor for squamous cell carcinoma (AA vs. (AG+GG)) (P=0.01, OR=1.42, 95%CI [1.08–1.86], P~heterogeneity~=0.27, fixed-effects model); (AA vs. GG) (P=0.03, OR=1.43, 95%CI[1.04–1.96], P~heterogeneity~=0.21, fixed-effects model). No association was observed in adenocarcinoma subgroup. Stratification analysis performed by gender shown that A allele might increase the lung cancer risk in male (A vs. G) (P=0.02, OR=1.18, 95%CI [1.02–1.37], P~heterogeneity~=0.48, fixed-effects model), but did not found association in female subgroup. Figure 1 Figure – Meta-analysis for the association between XPA gene rs1800975 polymorphism and lung cancer risk in the contrast of AA vs. (AG+GG) in overall. “Events” indicates the number of the AA genotype; “Total” indicates the total number of the AG+GG genotype plus the AA genotype.

Conclusion
XPA gene A23G polymorphism might associate with lung cancer risk in Overall and East Asians. This polymorphism might also associate with lung cancer risk in male and in the squamous cell carcinoma subgroup.

Background
Lung cancer is the fourth most common invasive cancer in Australia and was the leading cause of cancer death for both males and females in 2007. Five-year relative survival for lung cancer in Australia, which compares the risk of death for that cancer with age- and sex- matched population controls and overestimates absolute survival, is 13%. Gender based disparities have been noted, with men having worse outcomes. In this study, we publish for the first time an analysis of long-term (5 year) survival for lung cancer in Queensland, a northern state of Australia, where cancer is a notifiable disease. Differential outcomes for different tumour histologies, age groups, sex and estimated socio-economic status are compared, across 25 years spanning 1982-2006.

Methods
Cancer incidence and survival data on all Queensland residents diagnosed with non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC) between 1982 and 2009 were derived from the Oncology Analysis System, Queensland Cancer Control Analysis Team. Incidence and overall all-cause survival at five years from diagnosis were aggregated over four time periods to 2006 and stratified by histological subtype, age, gender, and estimated socioeconomic status derived from postcode information. Survival at 5 years from diagnosis was calculated using Kaplan-Meier estimates.

Results
The absolute incidence of lung cancer diagnoses in Queensland has increased between 1982 and 2009, from 820 cases per year (in the time period 1982-1990) to 1,632 cases per year (time period 2006-2009). The increased incidence of non-squamous NSCLC accounts for the majority of this and has increased significantly from 322/year to 1,083/year. Approximately one out of eight cases of lung cancer are SCLC. Five year survival for those diagnosed 2002-06 was 14% for all lung cancer, with a significantly worse outcome for SCLC (6%) compared with NSCLC (15%). Survival from squamous NSCLC was marginally better than non-squamous histologies (17% vs 15%). Significant discrepancies are seen in outcome related to gender (16% for women vs 13% for men) (p<0.05) and based on age (17% for those aged less than 65, and 13% for those 65 and older) (p<0.05). The outcomes are favourable compared with the five years 1982-86, where overall 5-year survival from lung cancer was 12% (4% for SCLC and 13% for NSCLC). No significant relationship was seen with outcome based on estimated socio-economic status across the entire time period, although this was not recorded for 34% of patients. For those classed as affluent or middle class, 5-year survival was 14% compared with 12% in disadvantaged areas.

Conclusion
Five year survival for lung cancer in Queensland, Australia compares favourably with national and international norms, and has increased over the last 25 years. There is a significantly worse outcome for men seen across all tumour subtypes, which has been noted in other tumour streams. Of note, there is no significant differential in survival based on socioeconomic status, where this could be estimated.

Background
The Victorian Lung Cancer Registry (VLCR) pilot project was established in January 2011 and aims to recruit all newly diagnosed lung cancer cases across participating sites in Victoria. Case ascertainment for the registry is derived from institutional ICD-10 coding with subsequent assessment against inclusion criteria. Incomplete case ascertainment threatens data validity and several methods have been proposed for estimating this in cancer registries.

Methods
A quantitative, case finding audit was employed to evaluate the VLCR’s case ascertainment methodology at a major metropolitan hospital between 01/07/2011 and 30/06/2012. ICD-codes determined that 125 new cases were registered at the hospital. Lists of patients recorded or suspected to have a diagnosis of lung cancer were requested from the following institutional and external departments: Radiotherapy, Day Procedure Unit, Oncology Lung Multidisciplinary Team Meeting (MDM), Cardiothoracic Surgery (CTS), Victorian Cancer Registry (VCR) and Pathology. Comparisons were made between patients included in the registry and departmental lists provided. Medical records were then assessed to check eligibility of outstanding patients for inclusion in the registry.

Results
Six patient lists were compared with the VLCR. Excluding duplications and exclusions a total of 10 eligible patients had not been recruited by the registry. Investigations indicated that the underreporting of these cases was largely attributed to the use of the ICD10 R91 Code. This code is not a primary lung malignancy code and is assigned by clinical coders for abnormal findings on diagnostic imaging of the lung where lung cancer is suspected but not confirmed. Of the 10 patients eligible for inclusion in the registry, 7 were discharged with the R91 code and pending clinical confirmation were later included in the registry. The remaining 3 patients were not included in hospital data extracts as they were non admitted day patients and therefore not coded. A capture – recapture methodology is used to evaluate ascertainment completeness.

Conclusion
The completeness of cancer registry data – the extent to which all of the incident cancers occurring in the population are included in the registry database – is an extremely important attribute of a cancer registry. Only a high degree of completeness in case-finding procedures will ensure cancer incidence rates and survival proportions are close to their true value.[i] It was identified that a possible 7.5% of total lung cancer patients were not being captured by the VLCR recruitment method. The inclusion of the R Code in hospital ICD code extracts will increase the VLCR ascertainment rate. [i] Parkin, D. M. and F. Bray (2009). "Evaluation of data quality in the cancer registry: Principles and methods Part II. Completeness." European Journal of Cancer 45(5): 756-764.

Background
Previous studies demonstrate disparities in the incidence of lung cancer. Specifically, in the United States, in comparison to non-Hispanic Whites, African Americans have a higher incidence rate of lung cancer while a lower incidence rate is seen among Hispanics. Recent studies demonstrate significant progress in smoking-cessation programs for non-Hispanic Whites; however, there are still a lower number of smoking cessation programs directed towards African Americans and trends of increased smoking behavior among Hispanics. No studies have evaluated recent histologic-specific incidence trends of lung cancer that may reflect these changing smoking behavioral patterns.

Methods
Using the Surveillance, Epidemiology and End Results data from years 1992-2010, lung cancer incidence rates and average annual percentage change (APC) were computed overall and by histology for African Americans, Hispanics, and non-Hispanic Whites.

Results
Incidence rates of lung cancer steadily decreased for non-Hispanic Whites, African-Americans, and Hispanics males (APC = -2.0, -2.6, and -1.9, respectively) from 1992 to 2010. This was true for all histological subtypes. Overall incidence rates of lung cancer among females have been relatively stable, with evidence of recent declines since 2005 among African Americans and non-Hispanic Whites. These trends were seen for all histological subtypes among females, except for the notable exceptions of increases in adenocarcinoma among African Americans (APC = 2.9 from 2003-2010) and Hispanics (APC = 1.2 from 1992-2010).

Conclusion
For men, overall and histologic-specific incidence rates of lung cancer have decreased among Non-Hispanics Whites, African Americans, and Hispanics. For women, in recent years, non-Hispanic Whites demonstrate a similar pattern of decline in lung cancer incidence rates. However, the overall incidence rate of lung cancer among African American and Hispanic women while decreasing or remaining slightly stable, an increased incidence of adenocarcinoma has been observed that warrants further exploration.

Background
There is scientific evidence of increasing lung cancer incidence in women. and lung cancer pattern and risk factors in women are not the same as with men due to genetic and hormonal differences

Methods
In this study we reviewed the clinical and pathologic features in our female lung cancer patients diagnosed and followed at our clinic between 2005 and 2009.

Results
In this period, we followed 360 lung cancer patientsl. Of them, 30(8.3%) were women.Their mean age was 62.83 +/- 12.43 (range 33-83). 93.4% (n:28) were housewives, 80% (n:24) were nonsmoker, 13.3%(n:4) were exsmoker, 6.7% (n:2) were current smokers. 93.7'% were postmenapausal women. They had no history of cancer, COPD, asthma, or tuberculosis. Hypertension, coronary arterial disease, and DM were present in 20% (n:6), 10% (n:3), and 6.6% (n:2) of them respectively. Only one patient (3.3%) was aymptomatic. Most common symptoms were cough, dyspnea, chest pain and hemoptysis in order. At the time of diagnosis, 63.3%(n:19) were ECOG 0, 23.3% (n:7) were ECOG 1, 13.3% (n:4) were ECOG 2. Histopathologic diagnosis were adenocarcinoma in 53.3%, squamous cell carcinoma in 13.3%, NSCLC (not otherwise specified) in 16.7% and SCLC in 16.7%. Of NSCLC patients only 8.6% were early stage, remaining 91.4% were locally advanced or metastatic disease. 20% of SCLC patients were at limited stage, remaining 80% had extensive disease. Most common sites of metastasis were lung/pleura (40%), brain (36.7%) and bone(30%). Three patients rejected any form of therapy. Surgery +/- adjuvant therapy were applied to 2 patients, chemotherapy +/- radiotherapy were given to 22 patients.

Conclusion
In conclusion, majority of our women lung cancer patients were nonsmokers, with adenocarcinoma histology, postmenapausal and in good performance status.

Background
The incidence of NSCLC in young patients increases annually. Our objective is to characterize this specific population

Methods
Retrospective study of 61 consecutive NSCLC-AD cases in patients under 50 years old, diagnosed between 2001 and 2011. Clinical data, histology, Performance Status (PS), therapy and overall survival (OS) were evaluated.

Results
Of all the patients 70.5% were men, mean age 44.7 ± 4.12 years, most of them smokers (55.7%) with a Performance Status (PS) of 1 (83.6%), 2 (9.8%) or 3 (6.6%). Comorbidities were found in 42.6% of patients. The most frequent histology was adenocarcinoma (55.7%) followed by squamous cell (39.4%) and large cell (4.9%). Metastasis in 1, 2 or 3 organs were found in 49.2%, 27.9% and 22.9% respectively. Excluding two patients that are still alive (49.5 and 65.3 months after initial diagnosis), the global OS was 12.4 months. All patients with PS 3 were smokers with comorbidities, had more than 2 sites of metastization and underwent best supportive care only. Mean OS of this group was 0.7 months. Only 1 patient with PS 2 showed comorbidities and they were all submitted to 1[st] line chemotherapy with "platinum-based doublets". No 2[nd] line therapy was initiated and mean OS of this group was 4.7 months. In the PS 1 group (51 patients), 43.1% showed at least 1 comorbidity and 56.9% had >1 metastasized organ. The mean OS of this group was 10.8 months. All patients underwent 1[st] line chemotherapy – options were "platinum-based doublets" (50 patients) and Erlotinib (1 patient). Second line therapy was done in 60.8% (31) of these patients (Pemetrexed in 14 patients, Docetaxel in 11 patients and Erlotinib in 6 patients), and 3[rd] line in 35.5% (11) – options were Erlotinib (9 patients), Pemetrexed and Carboplantin+Vinorelbine (1 patient each). Two patients started 4[th] line therapy (3.9%) with Pemetrexed. OS for different therapeutic lines in this group (patients that underwent only one line of therapy, 2 lines or 3 lines) were respectively 5.8 months, 11.45 months and 29.2 months.

Conclusion
In our study, young patients with NSCLC present in advanced stages with important comorbidities and have an overall prognosis similar to the literature. Stronger physical reserve may allow several therapeutic lines to be completed in a significant number of cases.