Virtual Library

Start Your Search

H. Kondo

Moderator of

  • +

    MA 14 - Diagnostic Radiology, Staging and Screening for Lung Cancer I (ID 672)

    • Event: WCLC 2017
    • Type: Mini Oral
    • Track: Radiology/Staging/Screening
    • Presentations: 15
    • +

      MA 14.01 - Influence of Early Lung Cancer Screening Programme on Treated Patients' Profile and Activity of Thoracic Surgery Department (ID 8063)

      15:45 - 15:50  |  Presenting Author(s): Bartosz Kubisa  |  Author(s): M. Wojtyś, T. Grodzki, J. Pieróg, J. Wójcik, N. Wójcik, J. Alchimowicz, P. Waloszczyk

      • Abstract
      • Presentation
      • Slides

      Background:
      The objective of the study was to compare the program of early detection of lung cancer by low-dose computed tomography scan with other groups of lung cancer patients who live in the area where such a program is carried out. The study was based on the materials of one thoracic surgery clinic and the screening was carried out on the inhabitants of one district city. The objective of the study was implemented by analyzing selected factors that impact the activities of surgery ward. The study was retrospective.

      Method:
      The patients were divided into three groups. Group 1a - 52 patients operated due to primary lung cancer which were detected during the screening. Group 1b - 87 patients operated for primary lung cancer during the screening, but who did not participate in the screening program. Group 2 - 103 patients operated before the commencement of the screening. The analysis involved among others the factors described in the table below. For the statistics we utilised Statistica PL 2010 program. Non parametric Mann-Whitney U-test was used for not normally distributed data. Parametric t-Student test was used for normally distributed data and p<0.05 was considered significant.

      Result:

      Significant differences among the groups
      Group Group Group Statistics Statistics
      unit 1a 1b 2 p 1a/1b p 1a/2
      Patiens n 52 87 103 - -
      Adenocarcinoma % 58 34 38 0.01 0.03
      G2 grading % 44 28 24 0.03 0.02
      Tumour volume cm[3] 8 21 20 0.004 0.01
      T1 factor % 54 28 34 0.004 0.02
      IA stage % 50 24 32 0.003 0.02
      Right side % 73 63 59 ns 0.05
      Lobectomy % 75 53 56 0.02 0.04
      Operation time min 129 114 112 0.02 0.007


      Conclusion:
      In the screening 1a group adenocarcinoma was detected more frequently, as a smaller tumour, at an earlier stage, with the prevalent G2 factor and located mainly on the right side. In the group 1a lobectomy was performed more frequently, than the other groups. The duration of surgery of 1a group was longer than the other groups due to more often intraoperative assessment use. There were no differences according to postoperative complications and deaths among all groups. Our screening program detectes lung cancer at earlier stage and offers faster definitive surgical treatment, probably improving 5 year survival, what is being evaluated now.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

    • +

      MA 14.02 - Simulation of the Four Rounds of NELSON Lung Cancer Screening Triage Algorithm (ID 9284)

      15:50 - 15:55  |  Presenting Author(s): Takashi Hayashi  |  Author(s): H. Beaumont, A. Iannessi, D. Wormanns, N. Faye

      • Abstract
      • Presentation
      • Slides

      Background:
      Imaging screening programs are designed for specific eligibility criteria and technologies. Effectiveness of these programs is precisely known only after monitoring large populations over a long period. Ensuring generalizability of screening programs is required when targeting a population which is different from the original one tested or when modifying involved technologies. The NELSON screening program is a lung cancer triage process featuring four rounds of variable intervals (1, 2 and 2.5 years). Each screening round classifies the nodule’s malignancy according to the nodule’s volume, growth and volume doubling time, using CT. The aim of our study was to assess by simulation the influence of variable precision of measurement on the robustness of NELSON’s diagnostic algorithm.

      Method:
      We simulated 10[6] nodules using a Chi[2] distribution for nodule size [3mm; 20mm], an inverse Chi[ 2] distribution of growing. 2.1% of nodules were malignant (true positive). We tested several distributions of measurement error using a zero-mean Gaussian distribution and a standard deviation (SD) ranging [0%; 20%]. We reported positive and negative predictive values (PPV, NPV) at each round.

      Result:
      After round 4, we found that NPV decreased with increasing measurement error from 100% to 99.89%, PPV decreased from 100% to 29.6%. Figure 1 Figure 1: Detection performances of NELSON’s triage algorithm depending on measurement error. As shown in this graph, an increase of SD leads to a decrease of PPV (gray curve) and has almost no impact on NPV (yellow curve).



      Conclusion:
      Increasing measurement error of nodules significantly degrades the positive predictive value of NELSON’s diagnostic algorithm in identifying malignant pulmonary nodules, whereas the negative predictive value remained stable. We confirmed the efficacy of the successive rounds when measurement error is larger than 5%; however, the algorithm could be improved for larger measurement errors. Simulations could help us to assess better strategies lung in screening studies.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

    • +

      MA 14.03 - Lung Cancer Risk Score Analysis Using Plasma microRNA Profiles (ID 8335)

      15:55 - 16:00  |  Presenting Author(s): Jennifer Eve Gyoba  |  Author(s): W. Roa, L. Guo, S. Ghosh, E. Bedard

      • Abstract
      • Presentation
      • Slides

      Background:
      There is a need for more accurate and minimally invasive methods to screen high risk populations and diagnose lung cancer during its asymptomatic early stages. microRNAs (miRNAs) are small, non-coding strands of RNA that are shown to lead to carcinogenesis when dysregulated. miRNAs, expressed in a tissue specific manner, are stable and detectable in small quantities, thus are promising candidates for biomarkers. Through the use of previous miRNA profiling done in our group, we aim to validate this panel in a large sample size of non-small cell lung cancers (NSCLC) using miRNAs 21, 155, 210, and 223 in blood plasma to determine if this miRNA panel is able to differentiate lung cancer cases from controls.

      Method:
      A nested case-control study of 64 patients with stage I/II NSCLC and 110 healthy controls with similar age, gender, and smoking history was performed. Plasma was provided by Conservant Bio, Lung Cancer Biospecimen Resource Network, and Alberta’s Tomorrow Project. miRNA was isolated using the Qiagen miRNeasy serum/plasma kit. miR-21, 155, 210, and 223 were quantified via RT-PCR using C. elegans miR-39 as a spiked-in endogenous control. Binary logistic regression (SPSS version 15) was performed to develop a combined risk score of the patients’ risk of having lung cancer. Receiver operating curve (ROC) analysis was used to determine risk category cut-off values based on the sensitivity and specificity. Plasma samples were taken 4-7 months post resection and also analyzed and compared to pre-operative samples and controls.

      Result:
      The cases and controls showed similar age ranges (mean=61.97, SD=7.76; mean=61.38, SD=7.95) respectively. Smoking history was higher in the cases (mean=51.5, SD=29.46) than controls (mean=30.98, SD=10.84). The combined score was dichotomized at -0.4169 into high and low risk categories (sensitivity=81%, specificity=41%, AUC=72.3%), the cases pre-operative samples compared to healthy controls was significantly different (odds ratio=3, p-value=0.003, 95% C.I.=[1.440,6.249]). For the cases post-operative samples compared to healthy controls, the combined score was dichotomized at -0.3255 (sensitivity=77%, specificity=41%, AUC=67%), also showing a significant difference (odds ratio=2.3, p-value=0.023, 95% C.I.=[1.120,4.621]). There is no significant difference in the combined risk score when comparing the pre-operative and post-operative NSCLC samples.

      Conclusion:
      Through binary logistic regression miRNA profiling has the potential to assist in screening the high-risk population for lung cancer. Used in conjunction with radiologic screening, this approach could allow early detection and treatment of disease while sparing patients unnecessary investigations and biopsies.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

    • +

      MA 14.04 - Therapeutic Response Assessment of NSCLC Patients Treated with Apatinib: A Radiomics Approach Based on CT Texture Features (ID 9468)

      16:00 - 16:05  |  Presenting Author(s): Qiong Zhao  |  Author(s): L. Peng, Z. Hong, Y. Wang, X. Ye, X. Li, P. Pang, F. Chen

      • Abstract
      • Presentation
      • Slides

      Background:
      Apatinib is a novel small molecular drug targeting vascular endothelial growth factor receptor-2 (VEGFR-2), which is currently being studied in multiple tumor types. The purpose of this study was to assess the treatment response in non-small cell lung cancer (NSCLC) patients enrolled in a clinical trial of apatinib according to the response evaluation criteria in solid tumors (RECIST) using non-contrast-enhanced computed tomography (CT) texture-based radiomics approach.

      Method:
      A total of 19 NSCLC patients from our single center participated in the currently undergoing multi-center phase III ANSWER study of apatinib (NCT 02332512). Patients were categorized as responders (CR and PR) and non-responders (SD and PD) according to RECIST criteria. Radiomic texture features were extracted from target lesions in post-therapy CT of NSCLC. Lasso regression was used to establish a model to discriminate between responders and non-responders. The performance of the model was assessed with ROC in both internal and independent validation cohorts.

      Result:
      Altogether, 108 CT scans were performed. Among them, 75 scans were randomly selected as internal validation group (70%), while the remaining 33 scans (30%) were identified as an independent validation group. Three hundred and eighty-four CT texture parameters were extracted and 21 out of 384 CT texture were finally selected for the model. The area under the curve (AUC) of ROC was 0.903 in the internal validation group, and that of the independent validation group was 0.714. Figure 1



      Conclusion:
      A radiomic discriminate model was built based on post-therapy CT texture features, which demonstrated a good performance in assessing the therapeutic response in NSCLC patients treated with apatinib.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

    • +

      MA 14.05 - Discussant - MA 14.01, MA 14.02, MA 14.03, MA 14.04 (ID 10837)

      16:05 - 16:20  |  Presenting Author(s): Edward F Patz

      • Abstract
      • Presentation
      • Slides

      Abstract not provided

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

    • +

      MA 14.06 - Population Based Cohort Study to Evaluate Lung Cancer Screening Using Low Dose CT in Hitachi City (ID 8087)

      16:20 - 16:25  |  Presenting Author(s): Takeshi Nawa  |  Author(s): K. Fukui, T. Nakayama, M. Sagawa, T. Nakagawa, Hideo Ichimura, T. Mizoue

      • Abstract
      • Presentation
      • Slides

      Background:
      In 1998, low-dose CT screening for lung cancer was introduced in Hitachi City, Japan. Based on time trend analysis, a significant reduction in lung cancer mortality was observed 4–8 years after introduction of CT screening.

      Method:
      To evaluate the effectiveness of lung cancer screening, we conducted a cohort study for CT screening participants and X-ray screening participants among Hitachi residents. Citizens aged 50 to 75 who underwent CT screening from 1998-2006 were defined as the CT group, and those who underwent X-ray screening during the same period, but did not receive CT screening throughout the follow-up period were defined as the X-ray group. We investigated lung cancer mortality and all-cause mortality of both groups from the first lung cancer screening of the subject to the end of 2012 using residence registry, the regional cancer registry, and national death statistics.

      Result:
      From the CT group (17,935 cases, 9,790 men and 8,145 women), 273 cases of lung cancer (1.5%), 72 cases of lung cancer death (0.4%), and 885 cases of all-cause mortality (4.9%) were observed. On the other hand, 164 cases (1.1%) of lung cancer, 80 cases (0.5%) of lung cancer death, and 1,188 cases (7.6%) of all-cause mortality were observed in the X-ray group (15,548 cases, 6,526 men and 9,022 women). The hazard ratios of the CT group to the X-ray group adjusted for sex, age, and smoking history were 0.49 for lung cancer mortality and 0.57 for all-cause mortality.

      Conclusion:
      Low dose CT screening participants exhibited a 51% reduction in lung cancer mortality during the observation period compared with the X-ray group. Although all-cause mortality also decreased by 43% in the CT group, the decrease in proportion of lung cancer deaths was greater.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

    • +

      MA 14.07 - Randomized Lung Cancer Screening with Low-Dose CT in China: A Specific Risk-Based Screening for Chinese Population (ID 8906)

      16:25 - 16:30  |  Presenting Author(s): Baohui Han  |  Author(s): H. Wang, J. Teng, J. Ye, Q. Chen, Y. Zhang, W. Yang, F. Qian

      • Abstract
      • Presentation
      • Slides

      Background:
      The purpose of the present study was to investigate whether low-dose computed tomography (LDCT) screening is capable of enhancing the detection rate of early-stage lung cancer and reducing lung cancer mortality rate in China, thus determining the appropriate duration of screening and identifying additional risk factors for lung cancers in Chinese population.

      Method:
      A randomized lung cancer screening study was performed with participants aged 45 to 70 years old who had at least one high-risk factor as follows: 1) a history of cigarette smoking ≥20 pack-years; former smokers who had quit within the past 15 years; 2) cancer history in immediate family members; 3) personal cancer history; 4) professional exposure to carcinogens (asbestos, dust or radiation); 5) long history of passive smoking; or 6) long-term exposure to cooking oil fumes. Participants were randomly assigned to a screening group with alternating years of LDCT screening (R1, R2) or a control group with biennial questionnaire inquiries.

      Result:
      A total of 6657 eligible participants were enrolled, 3145 participants were assigned to the control group and 3512 were assigned to the baseline LDCT screening (R1) group. 1516 participants (43.2%) underwent the second round of LDCT screening (R2) in the alternate year. At R1 and R2 rounds, 19.6% and 24.0% participants showed non-calcified nodules ≥4 mm on LDCT images. Among these, lung cancer was diagnosed in 44 participants (1.3%) at R1, 12 (0.8%) at R2, and 10 (0.3%) in the control group through either biopsy or cytologic analysis. The proportions of early-stage (0 to I) lung cancer were 97.7% at R1, 91.7% at R2 and 20% in the control group, respectively. At R1, the sensitivity of LDCT for lung cancer screening was 97.7%, the specificity was 76.8%, the positive predictive value was 5.1%, and the negative predictive value (NPV) was 99.9%; at R2, both the sensitivity and the negative predictive value increased to 100%. Two cases of lung cancer-specific deaths occurred in the control group, but no death occurred in the LDCT group.

      Conclusion:
      Compared to usual care, the two biennial screenings with LDCT led to a 77.7% increase at R1 and 71.7% at R2 in detecting early-stage lung cancer and a 20% decrease in lung cancer mortality. Biennial screening may be at least as efficient as annual screening in terms of detecting rate, sensitivity and NPV. This study provides insights about the non-smoking related risk factors of lung cancer in the Chinese population.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

    • +

      MA 14.08 - Hematology/Oncology Providers’ Practices and Attitudes of Lung Cancer Screening And Tobacco Cessation at an Academic Medical Center and VA (ID 8827)

      16:30 - 16:35  |  Presenting Author(s): Jennifer Lewis  |  Author(s): H. Chen, K.E. Weaver, L. Horn, K. Sandler, Pierre P Massion, H. Tindle

      • Abstract
      • Presentation
      • Slides

      Background:
      Advances in cancer screening and therapeutics have led to an estimated 15.5 million US cancer survivors. A history of cancer is a known risk factor for lung cancer. Lung cancer screening (LCS) with low-dose CT (LDCT) and smoking cessation in high-risk populations are recommended standard-of-care practices for cancer survivors, yet knowledge and practice of these interventions is low among PCPs. Hematologists and oncologists commonly provide cancer survivorship care, and yet their practices of and attitudes toward LCS are unknown. Based on prior data, we hypothesized that very few providers (<25%) would report performing LDCT screening while most (>75%) would report providing tobacco cessation services in the last year, and that knowledge of LCS guidelines would be associated with LDCT screening.

      Method:
      We electronically surveyed all Hematology/Oncology providers (n = 104) at a large academic institution in the Mid-South and its affiliated VA from February to May 2017. The survey queried: LCS/tobacco cessation practices (LDCT screening as primary outcome), perceived cancer screening/tobacco cessation effectiveness, knowledge of USPSTF LCS guideline recommendations and CMS coverage, perceived barriers to LDCT screening, and interest in future provider/patient LCS education and reminder tools. Data were summarized using counts, proportions, means, and medians. We used logistic regression to evaluate the association of LCS guideline knowledge (primary predictor) with reported LDCT screening.

      Result:
      The overall survey response rate was 73%. Few providers (38%) reported performing LDCT screening in the past year, while almost all providers (95%) reported providing tobacco cessation services. In unadjusted analysis, providers who knew at least three LCS guideline components were more likely to perform LDCT screening (OR 5.96, CI 2.03-17.49; P = 0.001). Only 55% of providers knew at least three LCS guideline components. More providers rated Pap-smear (75%), colonoscopy (71%), smoking cessation (68%), and mammography (39%) as very effective at reducing cancer-specific mortality compared to LDCT (24%). Major perceived barriers included: lack of patient awareness (74%) and patient financial cost (51%). More VA providers (37%) rated lack of a multi-disciplinary screening program as a major screening barrier compared to academic providers (7%) (P = 0.002). Majority of providers (≥ 56%) reported interest in future provider/patient LCS education and reminders.

      Conclusion:
      LDCT screening is currently an uncommon practice among hematology/oncology providers. Future interventions aimed at the provider, patient, and health system levels are needed to ensure standard-of-care LCS practices in the cancer survivor population. Provider level interventions should incorporate education on screening/tobacco cessation effectiveness and screening guideline recommendations.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

    • +

      MA 14.09 - Impact of Lung Cancer Perceived Risk, Screening Eligibility and Worry on LDCT Screening Preference - Challenges for Engaging Patients at High Risk (ID 9669)

      16:35 - 16:40  |  Presenting Author(s): Katharine See  |  Author(s): R. Manser, E. Park, Daniel P Steinfort, F. Piccolo, D. Manners

      • Abstract
      • Presentation
      • Slides

      Background:
      Lung cancer screening is only effective at reducing lung cancer deaths when the highest risk individuals are screened and followed. An individual’s risk of lung cancer, and therefore their screening eligibility, has not been shown to correlate with their perceived risk or intention to participate in screening. While previous studies have suggested many at-risk individuals are supportive of screening, no validated risk perception questionnaire has been used to compare perceived risk and worry with screening preference between eligible and ineligible individuals.

      Method:
      Participants were current or former smokers aged 55 to 80 years old who presented for medical outpatient specialist appointments at three Australian hospitals. The survey included 1) demographics and previous cancer screening participation 2) objective lung cancer risk measured by PLCOm2012 lung cancer risk prediction model 3) perceived lung cancer risk and worry about lung cancer measured by the questionnaire developed by Park et al and validated in sub-set of National Lung Screening Trial (NLST) participants and 4) preference for screening measured by a five point Likert scale. Eligibility for screening was PLCOm2012 risk >1.5%. Ordinal logistic regression identified factors associated with screening preference.

      Result:
      760 people 55-80 years old participated, of which 306 were ever-smokers. The participation rate was 26.9%. 23 did not complete either sufficient smoking details for PLCOm2012 risk or screening preference leaving 283 responses. Mean±SD age was 66.3±6.5, 60.4% (171/283) were male, median (IQR) PLCOm2012 risk was 1.28% (0.44-3.11) and 45.6% (129/283) were eligible for screening. Overall screening preference was high; 72.1% (204/283) either agreed or strongly agreed to having screening if offered. Objective lung cancer risk (PLCOm2012) was weakly correlated with both perceived lung cancer risk (r=0.28, p<0.0001) and worry (r=0.21, p<0.001). In univariate analysis, worry (OR 1.37, 95% CI [1.18-1.60], p<0.001), perceived risk (OR 1.10, 95% CI[1.04-1.16], p=0.002) and PLCOm2012 risk (OR 1.06, 95% CI[1.01-1.12], p=0.02) were associated with higher screening preference, but not associated with higher screening eligibility (OR 1.50, 95%CI[0.97-2.30], p=0.06). Age, gender, smoking status, family history of lung cancer and previous screening practice were not associated with screening preference. Only worry remained significantly associated with screening preference (adj-OR 1.33, [95%CI 1.10-1.60], p=0.003) with multivariate analysis.

      Conclusion:
      Worry about lung cancer appears to be a more important driver for screening preference than eligibility status. This presents a unique challenge when trying to engage with eligible individuals while minimizing screening demand from the ineligible majority.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

    • +

      MA 14.10 - Discussant - MA 14.06, MA 14.07, MA 14.08, MA 14.09 (ID 10838)

      16:40 - 16:55  |  Presenting Author(s): Helmut Prosch

      • Abstract
      • Presentation
      • Slides

      Abstract not provided

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

    • +

      MA 14.11 - Malignancy Risk Prediction of Pulmonary Nodule in Lung Cancer Screening – Diameter Or Volumetric Measurement  (ID 9113)

      16:55 - 17:00  |  Presenting Author(s): Ren Yuan  |  Author(s): M. Tammemägi, A.J. Ritchie, B. Dougherty, C. Sanghera, C. Jacobs, J.R. Mayo, H.C. Schmidt, M. Gingras, S. Pasian, L. Stewart, S. Tsai, D. Manos, J.M. Seely, P. Burrowes, R. Bhatia, S. Atkar-Khattra, Renelle L Myers, Ming Sound Tsao, B. Van Ginneken, Stephen Lam

      • Abstract
      • Presentation
      • Slides

      Background:
      Nodule size is an important parameter to determine malignancy risk. Semi-automated size measurements have the potential to replace manual measurements due to their higher accuracy and reproducibility, and less inter/intra-user variation. However, controversy exists regarding the relative accuracy of 2D diameter versus 3D volumetric measurement to predict malignancy risk. The objective of this study is to compare nodule malignancy prediction models based on 2D mean diameter versus volumetric measurement, both generated by a CAD Software.

      Method:
      We analyzed baseline LDCT reconstructed using high spatial frequency algorithm from 1746 participants (47% women, 53% men, age: 62.5 ± 5.8 yrs) in the Pan-Canadian Early Detection of Lung Cancer Study (PanCan), who had ≥1 non-calcified nodules ≥3mm in diameter. CAD software (CIRRUS Lung Screening, Radboud University Medical Center, Nijmegen, the Netherlands) performed an automatic nodule segmentation, which could be optimised manually, measurement of mean diameter and volume was generated. Malignant or benign nodule status was confirmed by pathology or prolonged follow-up (median follow-up 5.5 years). Logistic regression models predicting cancer were prepared with one including mean diameter and the other including volume. The discrimination, the ability to classify cancer versus benign nodules correctly, was evaluated by the area under the receiver operator characteristic cure (AUC). The calibration - do predicted probabilities match observed probabilities, was assessed using Spiegelhalter’s z-test and graphically by plotting the observed and predicted mean probabilities of cancer by deciles of model risk.

      Result:
      There were in total 5878 nodules, including 119 cancers in 115 individuals. Both models gave similar predictive performances. AUC was 0.947 (95% CI 0.922-0.964) in the mean diameter model and 0.946 (95% CI 0.921-0.966) in the volumetric model (p=0.83). The calibrations were similar between the two models (figure). Figure 1



      Conclusion:
      The predictive performances of nodule malignancy prediction models using mean 2D nodule diameter and 3D volumetric data were indistinguishable.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

    • +

      MA 14.12 - Detecting Epidermal Growth Factor Receptor Mutation Status in Patients with Lung Adenocarcinoma Using Radiomics and Random Forest (ID 9772)

      17:00 - 17:05  |  Presenting Author(s): Tianying Jia  |  Author(s): Junfeng Xiong, X. Li, J. Ma, Y. Ren, Z. Xu, X. Cai, J. Zhang, J. Zhao, X. Fu

      • Abstract
      • Presentation
      • Slides

      Background:
      We tried a radiomics approach to build a random forest classifier for recognition of epidermal growth factor receptor (EGFR) mutation status in Chinese patients with lung adenocarcinomas using quantitative image features extracted from non-enhanced computed tomography (CT) images

      Method:
      From October 2008 to December 2015, 355 patients diagnosed with lung adenocarcinomas were included in this retrospective study. They all have complete clinical, pathological, and EGFR mutation status information, and their CT images were scanned before any invasive operation. Tumors with ground glass component or diameter smaller than 2 cm were not included. Their pathological phenotypes and EGFR mutation status were gained from surgical resections. Region of tumors on CT images were segmented semi-automatically first then manually modified by experienced clinicians. 440 quantitative image features were extracted from CT images and fall into four groups: first order statistics, shape and size based features, textural features, and wavelet features. Random forest was used to build the classification model which takes all the features into consideration and make an overall probability of mutation based on the vote of decision trees. The random forest classifier was validated using an independent set and its performance was evaluated using area under curve (AUC) values of the receiver operating characteristic

      Result:
      355 patients diagnosed with lung adenocarcinoma were enrolled in this study (170 male, 185 female; 54 smokers, 301 non-smokers). The patients all received surgery based treatment and their tumor stage varied from I to IV. EGFR mutations (mainly 19del and 21L858R) were found in 187/285(65.6%) and 48/70(68.6%) patients in training and validation sets respectively. The random forest model showed an AUC of 0.781 (95% confidence interval: 0.668-0.894, p<0.001) in the validation set. The sensitivity and specificity are 60.4% and 90.9% at best diagnostic decision point. These results were highest among published results of only using images to detect EGFR.

      Conclusion:
      The random forest classifier based on CT images showed potential ability to identify EGFR mutations in patients with lung adenocarcinomas and could be improved in future works.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

    • +

      MA 14.13 - Nodule Size Isn't Everything: Imaging Features Other Than Size Contribute to AI Based Risk Stratification of Solid Nodules (ID 8177)

      17:05 - 17:10  |  Presenting Author(s): Timor Kadir  |  Author(s): L. Pickup, J. Declerck, R. Munden, F. Gleeson, Pierre P Massion, G. Smith

      • Abstract
      • Presentation
      • Slides

      Background:
      Previously proposed risk models for the malignancy of Indeterminate Pulmonary Nodules (IPNs) detected on Computed Tomography (CT) typically incorporate a mixture of clinical factors, such as age and smoking history, and radiological factors such as nodule size and location. Of the latter, size is considered one of the most significant. Artificial Intelligence based risk stratification software has been previously proposed that uses Texture Analysis with Machine Learning to predict IPN malignancy and has been shown to achieve high classification performance. While it is assumed that such techniques can capture image texture patterns that separate benignity from malignancy, such methods also intrinsically measure nodule size. The contribution of texture to classifier performance beyond size has not been studied and we seek to quantify this. We show, for the first time, the relative contributions of texture and size on the performance of Artificial Intelligence risk stratification of solid nodules.

      Method:
      Two datasets were created from the US National Lung Screening Trial (NLST). The first (A), comprising 640 solid nodules, was built to remove size as a discriminatory factor between benign and malignant; all malignant solid nodules between 4 and 20 mm diameter were selected, and for each, a benign solid nodule was selected that most closely matched it in diameter. Any malignant nodule for which an equivalently sized benign could not be found within 0.8 mm was rejected. Sizes were measured using automated volumetric segmentation. The second dataset (B), also comprising 640 subjects, included all malignant nodules in A but benign nodules were randomly selected following the empirical size distribution of the whole NLST dataset. Therefore, nodule size cannot be a discriminative factor in A but would be in B. Two nodule stratification algorithms were developed using Texture Analysis combined with Machine Learning (Support Vector Regression) integrating 20 variables including 3D Haralick, Gabor and Shape features, from A and B respectively using five-fold cross validation and the performance compared measuring Area-Under-the-Curve (AUC).

      Result:
      The average AUC for the algorithm trained on dataset A was 0.70 whereas using size alone on the same dataset gave an AUC of 0.50. The AUC was 0.91 for the algorithm trained on B.

      Conclusion:
      On this data, Texture Analysis with Machine Learning contributes 0.20 AUC points to classfication performance. Artificial Intelligence based risk classification can identify radiological features that are predictive of solid nodule malignancy that are independent of nodule size.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

    • +

      MA 14.14 - The First 100 Days: Early findings from the Lung Cancer Screening Pilot for People at High Risk in Ontario, Canada (ID 8579)

      17:10 - 17:15  |  Presenting Author(s): Gail Elizabeth Darling  |  Author(s): H.C. Schmidt, B. Miller, M. Yurcan, E. Svara, V. Treister, S. Doig, M. Tammemägi

      • Abstract
      • Presentation
      • Slides

      Background:
      An estimated 330,000 people in the province of Ontario are at high risk of developing lung cancer and eligible for screening with low-dose computed tomography (LDCT). On June 1 2017 Cancer Care Ontario launched the Lung Cancer Screening Pilot for People at High Risk with the purpose of informing the design and implementation of a province-wide organized screening program. Organized screening is available at 3 hospitals, and provider and public recruitment strategies are being implemented to engage the target population in regional catchment areas. Key aspects of pilot design include eligibility based on the PLCO~M2012noRace~ risk prediction model, navigation support, informed participation, embedded smoking cessation services, radiology quality assurance, LDCT findings categorized in accordance with Lung-RADS™, provision of same-visit screening results and seamless transition to a Diagnostic Assessment Program (DAP) for assessment of findings suspicious for lung cancer. Data collected for 3,000 participants over a 2-year period will inform a comprehensive evaluation of the pilot.

      Method:
      Indicators were selected to assess impacts of early recruitment efforts and outcomes of key screening processes related to eligibility assessment, the LDCT scan and smoking cessation. Data were collected by the pilot sites and submitted to Cancer Care Ontario. Participant feedback on the screening experience was collected by survey. Data were collected in June and July 2017; data from August 2017 will be available for presentation.

      Result:
      The majority (87%) of the 862 people recruited into the pilot were provider-referred. Of the 472 people who completed a risk assessment, 71% were found to be eligible for screening (PLCO~M2012noRace~ 6-year risk ≥2.0%). Baseline LDCT scans were conducted for 156 participants; approximately 8% of these participants were referred to a DAP for further assessment. Uptake of smoking cessation services by current smokers was high (data to be included in presentation). Feedback surveys were received from 78 of 156 participants screened. Overall experience with the screening visit was rated as ‘excellent’ by 91% of respondents, and 70% indicated a preference to receive results during the same visit as the LDCT.

      Conclusion:
      Provider-led recruitment supports the identification of screen-eligible individuals. Implementation of navigator-guided organized screening, following a detailed screening pathway that features provision of same-visit results, has contributed to high participant satisfaction to date. To our knowledge, this pilot involves the most detailed organized screening pathway and comprehensive evaluation plan developed to date. Learnings from this pilot will be highly relevant to jurisdictions around the world that are adopting screening.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

    • +

      MA 14.15 - Discussant - MA 14.11, MA 14.12, MA 14.13, MA 14.14 (ID 10839)

      17:15 - 17:30  |  Presenting Author(s): Ugo Pastorino

      • Abstract
      • Presentation
      • Slides

      Abstract not provided

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.



Author of

  • +

    P1.02 - Biology/Pathology (ID 614)

    • Event: WCLC 2017
    • Type: Poster Session with Presenters Present
    • Track: Biology/Pathology
    • Presentations: 1
    • +

      P1.02-033 - Differentiating of Cytomorphological Characteristics in Non-Small Cell Lung Cancer Predicts Value of Radiologic Features (ID 9355)

      09:30 - 09:30  |  Author(s): H. Kondo

      • Abstract
      • Slides

      Background:
      In the 2011 IASLC/ATS/ERS classification, guidelines recommend that the major adenocarcinoma (ADC) subtypes should be classified according to the predominant histologic pattern. In published papers, the ADC classification has significant prognostic and predictive value regarding death or recurrence has been reported. The purpose of this study was to retrospectively evaluate characteristic differences between cytomorphological findings among histological subtypes in the preoperative bronchoscopic materials and radiologic features on high-resolution computed tomography (HRCT) and [(18)F]-fluorodeoxyglucose (18F-FDG) positron emission tomography (PET) /CT examinations in our database.

      Method:
      Forty-four consecutive lung cancer patients with peripheral lung nodules diagnosed by bronchoscopic biopsies underwent surgery in our hospital from April 2011 to May 2016. The cyotologic material obtained by brushing bronchoscopically was placed onto a glass slide, immediately fixed in 95% ethanol, and stained with Papanicolaou stain. The cytomorphological studies were retrospectively performed separately by two experienced cytotechnicians. Clinicopathological data with preoperative radiologic features on HRCT and 18F-FDG PET/CT in the patients was utilized for comparing to cytomorphological analyses.

      Result:
      Out of the forty-four patients with lung cancer, by excluding one typical carcinoid and four not otherwise specified (NOS), thirty-nine specimens in the patients consisted of ADC (n=32) and squamous cell carcinoma (SQCC, n=7) were analyzed. Thirty-two ADC were subclassified cytomorphologically into acinar (n=22), solid (n=8), papillary (n=1) and lepidic (n=1). The subtypes of ADC except for eight solids were categorized as nonsolid of ADC on comparisons of analyzed data. Specimens classified as solid pattern of ADC had a predominant 3D clusters (8 of 8 specimens, 100%) and conspicuous nucleoli (7 of 8 specimens, 87.5%) than nonsolid patterns of ADC. There were statistically significant differences between the nonsolid and the solid patterns about the two features (P=0.0011 and 0.0007, respectively). C/T ratio (a diameter of consolidation divided by a diameter tumor size) of the nonsolid pattern (0.9±0.1) was significantly lower than that for the solid pattern and the SQCC (1.0±0.0) (p=0.01). Maximum standardized uptake value (SUVmax) of the SQCC at 60 and 120 minutes (12.3±4.0 and 16.9±7.3) was significantly higher than that for the nonsolid pattern of ADC (7.0±3.6 and 9.0±4.6; P=0.003 and 0.004, respectively). Figure 1



      Conclusion:
      Preoperative cytomorphological subtyping might predict value of radiologic features on CT and PET examinations, therefore their features also could provide information about a degree of pathological invasiveness or biological malignancy of tumor with some decisions for treatments.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.