Virtual Library

Start Your Search

J.K. Field

Moderator of

  • +

    ORAL 09 - CT Screening - New Data and Risk Assessment (ID 95)

    • Event: WCLC 2015
    • Type: Oral Session
    • Track: Screening and Early Detection
    • Presentations: 8
    • +

      ORAL09.01 - Discerning Malignant and Benign New Nodules at Incidence Rounds of CT Lung Cancer Screening: The Role of Volume and Predicted Volume Doubling Time (ID 1358)

      J.E. Walter, M.A. Heuvelmans, G.H. De Bock, P.A. De Jong, R. Vliegenthart, M. Oudkerk

      • Abstract
      • Slides

      Background:
      Newly detected nodules after baseline screen are common findings in low-dose computed tomography (LDCT) lung cancer screening, and may complicate management. So far little research focused specifically on nodules newly detected at incidence screening rounds. These nodules develop within a known time-frame and are expected to be fast-growing and potentially malignant. Even so, the majority are benign. The aim of this study was to compare volume and predicted growth rate of benign and malignant new solid nodules detected in the incidence screening rounds of the Dutch-Belgian Randomized Lung Cancer Screening Trial (NELSON).

      Methods:
      The NELSON trial was approved by the Dutch Ministry of Health. All participants gave written informed consent. In total, 7,557 individuals underwent baseline LDCT screening. After the baseline screening, incidence screening rounds took place after 1, 3 and 5.5 years. This study included participants with solid non-calcified nodules, newly detected after baseline and also in retrospect not present on any previous screening. Nodule volume was obtained semi-automatically by Lungcare software (Siemens, Erlangen, Germany). The growth rate at first detection was estimated by calculating the slowest predicted volume-doubling time (pVDT), according to the formula pVDT=[ln(2)*Δt]/[ln(V2/V1)], using the study’s detection limit of 15mm[3] (V1), the volume of the new nodule at first detection (V2), and the time interval between current and last screen (Δt [days]). The pVDT was calculated for nodules with a predicted volume increase of at least 25% (≥ 18.75mm[3]). Lung cancer diagnosis was based on histology. Benignity was based on histology or a stable size for at least two years. Mann-Whitney U testing was used to evaluate differences in volume and pVDT between malignant and benign nodules.

      Results:
      During the incidence screening rounds, 1,484 new solid nodules in 949 participants were detected of which 77 (5.2%) turned out to be malignant. At first detection, both the median volume of malignant (373mm[3], IQR 120-974mm[3]) and benign (44mm[3], interquartile-range [IQR] 22-122mm[3]) new nodules, as well as the median pVDT of malignant (144 days, IQR 116-213 days) and benign (288 days, IQR 153-566 days) new nodules differed significantly (P<0.001 for both). The calculated median pVDT of adenocarcinomas (183 days, IQR 138-299 days) and squamous-cell carcinomas (150 days, IQR 117-223 days) was similar to previously published volume doubling-times of fast-growing baseline cancers in the NELSON trial of the same histological type (196 days, IQR 135-250 days and 142 days, IQR 91-178 days, respectively).

      Conclusion:
      At incidence LDCT lung cancer screening, volume and pVDT can be used to differentiate between malignant and benign nature of newly detected solid nodules. The pVDT is a new measure that can assist in adjusting for time differences in screening intervals.

      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.

    • +

      ORAL09.02 - Results of the Fourth Screening Round of the NELSON Lung Cancer Screening Study (ID 1354)

      U. Yousaf-Khan, C. Van Der Aalst, P. De Jong, R. Vliegenthart, E. Scholten, K. Ten Haaf, M. Oudkerk, H. De Koning

      • Abstract
      • Presentation
      • Slides

      Background:
      Although screening can reduce lung cancer (LC) mortality, the optimal screening strategy (e.g. numbers of screening rounds, screening interval) is unclear. The use of different screening intervals in the NELSON study is unique and makes it possible to investigate how the screening test performances (e.g. lung cancer detection rate, false positive rate) and characteristics of screen-detected lung cancers might change. This study describes the results of a fourth screening round that took place 2.5 years after the third round.

      Methods:
      The Dutch-Belgian randomized-controlled Lung Cancer Screening Trial (NELSON) aims to investigate whether low-dose CT screening would reduce LC mortality by at least 25% relative to no screening after ten years of follow-up. Therefore, screen group participants were screened four times: at baseline and year 1, 3, and 5.5. Screening test results were classified as negative, indeterminate, or positive based on nodule presence, volume (in case of new nodules) and volume doubling time (in case of previous existing nodules). Participants with an indeterminate test result underwent follow-up screening to classify their final screening test result as positive or negative. Participants with a positive scan result were referred to a pulmonologist for a diagnostic work-up. For this study, we included only participants who had attended all four screening rounds (n=5279). Epidemiological, radiological and clinical characteristics of lung cancers detected in the fourth round were compared with those of the lung cancers detected in the first three rounds. In addition, the risk for lung cancer detection in the fourth round (5.5 year risk) was quantified for subgroups.

      Results:
      In round four, 46 lung cancers were detected; 58.7% were diagnosed at stage I, 15.2% at stage II and 23.8% at stage III/IV. Adenocarcinomas were correlated with better cancer stage distribution, while small-cell carcinomas (SCLC) were associated with higher stage distribution (p=0.064). False positive rate after positive screening was 59.04% (62/105) and the overall false positive rate of the fourth round was 1.15% (62/5383). Relative to the results of the first three rounds, the LC detection rate was lower (0.80 vs 0.80-1.1) and LC was detected at a more advanced stage (23.8% vs 8.1%). In the fourth round more squamous-cell carcinomas (21.7% vs. 16.3%), SCLC (6.5% vs 3.8%) and bronchioloalveolar carcinomas (8.7% vs 5.3%) were detected. No large-cell carcinomas, large-cell neuroendocrine carcinomas or carcinoids were found in the fourth round. Screening results of the first three rounds led to formation of subgroups with significantly different probability of screening result in the fourth round: participants with previous exclusively negative results had a probability of 97.2% of negative screen compared to participants with ≥1 indeterminate or positive screen (94.6% and 87.1%) in the first three rounds. The risk of detecting LC in the fourth round also differed between these subgroups: exclusively negative results (<1.0%) and any time ≥1 indeterminate or positive result (1.5-1.7%).

      Conclusion:
      The LC detection rate after the third screening round was slightly lower and the stage distribution of screen-detected lung cancers in the fourth round was slightly less favorable. However, the differences seem limited.

      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.

    • +

      ORAL09.03 - The Danish Lung Cancer Screening Trial: Results 5 Years after Last CT Screening (ID 2384)

      J.H. Pedersen, M.W. Wille, A. Dirksen

      • Abstract
      • Presentation
      • Slides

      Background:
      The Danish Lung Cancer Screening Trial (DLCST) is a European randomized controlled trial comparing annual CT screening with no screening. Inclusion ran from 2004 to 2006, and participants have now been followed for 5 years since last CT screening (approximately 10 years since randomization). The American NLST showed 20% decrease in lung cancer mortality in the screening group, and DLCST is the first European trial to present comparable results regarding effect of screening on mortality, causes of death, lung cancer findings and risk stratification after sufficient follow up.

      Methods:
      4,104 participants aged 50-70 at time of inclusion and a minimum of 20 pack-years of smoking history were randomized to five annual low-dose CT scans or clinical follow up without CT scanning; thus, participants were younger and had fewer pack-years than participants from NLST. Screening was concluded in 2010. Follow up information regarding date and cause of death as well as lung cancer diagnosis, stage and histology was obtained from national registries, latest follow up date was April 7, 2015. . The effects of age, amount of smoking and COPD on lung cancer mortality in the two randomized groups were explored to evaluate possible effects of risk stratification and selection of high-risk individuals on effect of screening.

      Results:
      More cancers (100 vs. 53, p<0.001) were found in the screening group, in particular adenocarcinomas (58 vs. 18, p<0.001). Significantly more low-staged cancers (stage I+II: 54 vs. 10, p<0.001) and stage IIIa cancers (15 vs. 3, p=0.009) were found in the screening group. However, stage IV cancers were more frequent in the control group (23 vs. 32, p=0.278), and this was statistically significant for the highest-stage cancers (T4N3M1: 8 vs. 21, p=0.025). No differences in lung cancer mortality or all-cause mortality were observed between the two groups (Log Rank tests: p=0.898 and p=0.885, respectively). However, sub-group analyses including participants with higher age, presence of COPD, and more than 35 pack-years of smoking history showed significantly increased risk of death from lung cancer; the highest-risk group (with COPD and >35 pack-years) showed a 20% reduction in lung cancer mortality when screened. Though this result is not statistically significant due to small numbers, it does show compliance with the results from NLST.

      Conclusion:
      Although no statistically significant effects of 5 annual CT screening rounds on lung cancer mortality were observed in this small study, results indicate that focus on selection of high-risk individuals may be essential for the effect of CT lung cancer screening. We suggest that balancing benefits with harms—such as false positive findings and overdiagnosis— should bring focus to high-risk profiling of screening participants. Thus, the effects of age, amount of smoking, and COPD on the occurrence and mortality of lung cancer in the two randomization groups seems to indicate that limiting lung cancer screening to a higher-risk group improves the outcome of 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.

    • +

      ORAL09.04 - Discussant for ORAL09.01, ORAL09.02, ORAL09.03 (ID 3532)

      N. Peled

      • 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.

    • +

      ORAL09.05 - Lung-RADS versus the McWilliams Nodule Malignancy Score for Risk Prediction: Evaluation on the Danish Lung Cancer Screening Trial (ID 356)

      S.J. Van Riel, F. Ciompi, M.W. Wille, M. Naqibullah, C. Schaefer-Prokop, B. Van Ginneken

      • Abstract
      • Slides

      Background:
      Lung-RADS published in 2014 by the American College of Radiology is based on literature review and expert opinion and uses nodule type, size, and growth to recommend nodule management adjusted to malignancy risk. The McWilliams model (N Engl J Med 2013;369:910-9) is a multivariate logistic regression model derived from the Pan-Canadian Early Detection of Lung Cancer Study and provides a nodule malignancy probability based on nodule size, type, morphology and subject characteristics. We compare the performance of both approaches on an independent data set.

      Methods:
      We selected 60 cancers from the Danish Lung Cancer Screening Trial as presented in the first scan they were visible, and randomly added 120 benign nodules from baseline scans, all from different participants. Data had been acquired using a low-dose (16x0.75mm, 120kVp, 40mAs) protocol, and 1mm section thickness reconstruction. For each nodule, the malignancy probability was calculated using McWilliams model 2b. Parameters were available from the screening database or scored by an expert radiologist. Completely calcified nodules and perifissural nodules were assigned a malignancy probability of 0, in accordance with model guidelines. All nodules were categorized into their Lung-RADS category based on nodule type and diameter. Perifissural nodules were treated as solid nodules, in accordance with Lung-RADS guidelines. For each Lung-RADS category cut-off sensitivity and specificity were calculated. Corresponding sensitivities and specificities using the McWilliams model were determined.

      Results:
      Defining Lung-RADS category 2/3/4A/4B and higher as a positive screening result, specificities to exclude lung malignancy were 21%/65%/86%/99% and vice versa sensitivities to predict malignancy were 100%/85%/58%/32%. At the same sensitivity levels as Lung-RADS, McWilliams model yielded overall higher specificities with 2%/86%/98%/100%, respectively (red arrows in Figure 1). Similarly, at the same specificities McWilliams’s model achieved higher sensitivities with 100%/95%/85%/48%, respectively (green arrows in Figure 1). Figure 1



      Conclusion:
      For every cut-off level of Lung-RADS, the McWilliams model yields superior specificity to reduce unnecessary work-up for benign nodules, and higher sensitivity to predict malignancy. The McWilliams model seems to be a better tool than Lung-RADS to provide a malignancy risk, thus reducing unnecessary work-up and helping radiologists determine which subgroup of nodules detected in a screening setting need more invasive work-up.

      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.

    • +

      ORAL09.06 - The Cancer Risk Management Model: A Tool to Inform Canadian Policymakers Implementing Low-Dose CT Screening for Lung Cancer (ID 968)

      W.K. Evans, J. Gofffin, W. Flanagan, A. Miller, N. Fitzgerald, S. Memon, S. Fung, M. Wolfson

      • Abstract
      • Presentation
      • Slides

      Background:
      Although the National Lung Screening Trial (NLST) demonstrated that 3 annual low-dose CT (LDCT) screens reduced lung cancer specific and overall mortality at 6 years in a defined population of smokers, the decision to implement population-based screening is difficult in the absence of information on factors not evaluated in the NLST including frequency and duration of screening, characteristics of the “at risk” population, program cost and cost-effectiveness. The Canadian Partnership Against Cancer has developed a Cancer Risk Management Model for lung cancer (CRMM-LC) with a screening module informed by data from NLST that can evaluate these factors.

      Methods:
      CRMM-LC uses longitudinal microsimulation techniques that incorporate Canadian demographic characteristics, risk factors, cancer management approaches and outcomes, resource utilization and other economic factors to assess impacts on population health and costs to the Canadian healthcare system. Data sources include large national population surveys, cancer registries and census data. The diagnostic and therapeutic approaches and outcomes in CRMM-LC are based on input from Canadian lung cancer experts and survival information from medical literature. The simulated mortality reduction from LDCT screening using CRMM-LC is comparable to NLST. The model can projected incident cases, life years and quality adjusted life years over different time periods for populations defined by different age ranges and smoking histories and by screening duration and frequency (annual vs biennial). It can also inform individual provinces of the incremental resources (CT scans, invasive procedures) required for program implementation and project budget impact.

      Results:
      Based on NLST at risk criteria (55-74 yr old smokers of 30+ pack-years, the base case scenario), 1.4 million or 4% of Canadians would be candidates for LDCT screening in 2014. Annual screening over a 10 year period with a participation rate of 60% and 70% adherence would identify an additional 12,500 (4.7%) incident cases and result in 11,320 life-years saved (undiscounted). Biennial screening would identify 4,620 (1.6%) fewer cases and save 1,454 (12.8%) fewer life-years, but may be more cost-effective than annual screening. Scenarios modeling participation rates of 20, 40 and 80% (linear uptake over 10 years) yield incident cases that vary from 8,380 fewer for the lowest rate to 3,950 more for the highest with life years saved over 10 years ranging from 7,540 fewer to 3,310 more, respectively. The model projects 3,560 more cases would be detected if LDCT was introduced for younger (50 to 69 yr old), 30 pack-year smokers compared to the base case scenario and 1,760 more cases if the threshold number of pack years was decreased to 20 pack-years. The 10 year cumulative incremental cost in Canada of annual and biennial screening would be $1,107 and $709 million, respectively

      Conclusion:
      CRMM-LC, available at cancerview.ca/cancerriskmanagement, can be used by provincial analysts to estimate the impact of various scenarios on the impact of policy decisions concerning the scope of the LDCT screening program. In the current fiscally constrained healthcare environment, models that can assimilate diverse sources of information and extrapolate beyond clinical trial results can help inform decisions that healthcare administrators confront.

      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.

    • +

      ORAL09.07 - Economic Evidence for the Use of Risk-Selection and Risk-Stratification for Lung Cancer Screening Programs (ID 2928)

      S. Cressman, S. Lam, M. Tammemägi, S. Peacock

      • Abstract
      • Presentation
      • Slides

      Background:
      Screening for lung cancer according to age and smoking history alone could cost billions of dollars of public health expenditure due to the high incidence of potential participants. Risk-adapted lung cancer screening strategies such as participant selection (based on published risk prediction models such as the PLCOm2012 model) and malignancy risk based screening protocols may reduce program costs while improving outcomes among current and former smokers at risk of developing the disease. The Pan-Canadian Early Detection of Lung Cancer Study (PanCan) was designed with the objective of providing economic evidence for an affordable lung cancer-screening program in Canada.

      Methods:
      Data for 2537 screening participants in the PanCan study (median follow-up time of 4 years) and 25,914 eligible participants from the NLST-CXR arm were included in the analysis. There was adequate power and follow-up to inform the transition probabilities in model and provide the distribution to test all model parameters simultaneously in a probabilistic sensitivity analysis. The cost and health utility inputs are from patient-level trial data with defined ranges of certainty.

      Results:
      Our results show that risk selection using the PanCan risk prediction model could reduce the need to screen 21,022 (81%) of the NLST population if risk prediction were applied. If risk prediction were applied to Canadians who met the NLST criteria, 2 year program costs could be reduced by 400 million dollars and nearly half a million people could be spared the potential harms from screening that is not likely to result in a Cancer diagnosis. With the economic evidence from the PanCan and NLST trials, we report our initial cost-effectiveness results and will show, for the first time, a definitive description of the uncertainty surrounding our cost-effectiveness ratios.

      Conclusion:
      Using a model loaded with patient-level screening data has enabled us to predict the likelihood that risk-adapted screening will fall below most commonly referenced thresholds of acceptability for cancer interventions. The initial results and characterization of the parameters affecting cost-effectiveness will be presented. [*]on behalf of the PanCan study team The panCan study is sponsored by the Terry Fox Research Institute and the Canadian Partnership against Cancer, ARCC is funded by the CCSRI The authors thank the National Cancer Institute for access to NCI’s data collected by the National Lung Screening Trial. The statements contained herein are solely those of the authors and do not represent or imply concurrence or endorsement by NCI

      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.

    • +

      ORAL09.08 - Discussant for ORAL09.05, ORAL09.06, ORAL09.07 (ID 3472)

      S. Malkoski

      • 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

  • +

    MS 15 - Current Screening Trials, Current Evidence and Screening Algorithms (ID 33)

    • Event: WCLC 2015
    • Type: Mini Symposium
    • Track: Screening and Early Detection
    • Presentations: 1
    • +

      MS15.03 - UKLS Impact of Utilization of Risk Assessment in Trial Selection (ID 1914)

      J.K. Field

      • Abstract
      • Presentation

      Abstract:
      Future implementation of lung cancer screening programmes will require accurate identification of the population who will benefit the most, to ensure that the benefits outweigh the harms [1]. In the USA, the current criteria for Medicare reimbursement [2], for screening are: age 55 to 77, a smoking history of 30 pack-years or more and smoking within 14 years of entry [3]. However, an in-depth analysis of the NLST showed that there was marked variation in individual risk of lung cancer death, with some screened that had only a low chance of benefit: 20% of participants at lowest risk accounted for only 1% of prevented lung-cancer deaths). [4]Conversely, 88% of the prevented deaths were in the 60% of participants that were at highest risk. The only risk prediction model so far utilised in the recruitment of participants into a CT Lung Cancer Screening RCT, is the LLP~v2~ risk model in the pilot UK lung cancer screening trial (UKLS) [5]. The Liverpool Lung Project (LLP) risk model was based on a case-control study [6]. The LLP~v1~ model utilised conditional logistic regression to develop a model based on factors that were significantly associated with lung cancer (smoking duration, prior diagnosis of pneumonia, occupational exposure to asbestos, prior diagnosis of cancer family history of lung cancer (early onset <60 years) and exposure to asbestosis [6]. The multivariable model was combined with age-standardised incidence data to estimate the absolute risk of developing lung cancer. The discrimination of the LLP was evaluated and demonstrated its predicted benefit for stratifying patients for CT screening by using data from three independent studies from Europe and North America [7]. The LLP~v2~ was used to select subjects with ≥5% risk of developing lung cancer in the next five years for UKLS [8]. This method may improve cost-effectiveness by limiting screening to high-risk individuals. The UKLS approached 247,354 individuals in the two pilot sites, 75,958 people (30.7%) responded positively to the screening invitation. Demographic factors associated with positive response were: higher socioeconomic status, age 56-70 years, and ex-smokers. Those from lower socioeconomic groups and current smokers were less likely to respond. 8,729 (11.5%) positive responders were calculated as high risk of lung cancer. The high risk individuals were more often elderly, current smokers, of lower socioeconomic status and males (2.4x females). 4,055 were randomised into the UKLS. Forty two UKLS participants have been diagnosed with confirmed lung cancer, 34 of these were detected at baseline or three months, giving a baseline prevalence of 1.7% which is significantly higher than that reported by the NLST[9]or NELSON [10]trials. To date, 2.1% of all individuals screened have been diagnosed with lung cancer. 36/42 (85.7%) of the screen-detected cancers were identified at stage 1 or 2. Of those with a confirmed cancer, 17/42 (40.5%) were from the most deprived Index of Multiple Deprivation (IMD) quintile. Figure 1 Figure 1: Percentage of UKLS positive responders (n=75,958) with an LLP risk of >5%, by individual year of age. The positive response rate increased steadily with higher socioeconomic status: 21.7% of individuals in the lowest (most deprived) IMD quintile gave a positive response compared with 39.7% in the highest quintile (p<0.001;) (Figure 2). The proportion of individuals with a high LLP risk score decreased with higher socioeconomic status; ranging from 18.2% in the most deprived quintile to 8.3% in the least deprived quintile (p<0.001;). LLP risk were offset by, the socio-demographic spectrum of the individuals attending the clinic, which was in proportion to that of the original approached sample. People recruited into the UKLS trial therefore spanned all IMD quintiles in roughly equal numbers, including a representative proportion from more deprived postcodes. However, in the high risk sub group of individuals invited for screening, there was a trend towards individuals of higher socioeconomic status being more likely to consent to participate in the trial. Figure 2 Figure 2: Impact of socioeconomic status upon initial response rate (lower line), LLP risk (bars) and trial consent rate (upper line). The demographic and response data from the UKLS pilot trial enable specific recommendations to be made regarding the implementation of any future UK-wide lung LDCT screening programme. Such a programme would need to target those most at risk who may be least likely to take up offers of screening (i.e. the most deprived, current smokers, and the over 70s), and women.





      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.

  • +

    PRC 01 - Press Conference 1 (ID 196)

    • Event: WCLC 2015
    • Type: Press Conference
    • Track: Other
    • Presentations: 1
    • +

      PRC01.02 - Summary of the IASLC Third CT Screening Workshop - Dr. John Field, Chair, Screening Advisory Committee, IASLC (ID 3615)

      J.K. Field

      • Abstract

      Abstract not provided