Virtual Library
Start Your Search
A. McWilliams
Moderator of
-
+
O05 - Cancer Control (ID 130)
- Event: WCLC 2013
- Type: Oral Abstract Session
- Track: Prevention & Epidemiology
- Presentations: 8
- Moderators:N. Van Zandwijk, A. McWilliams
- Coordinates: 10/28/2013, 10:30 - 12:00, Bayside Auditorium A, Level 1
-
+
O05.01 - The DANTE trial, a randomized study of lung cancer screening with spiral CT: 7-year results (ID 1428)
10:30 - 12:00 | Author(s): M.V. Infante, S. Cavuto, F.R. Lutman, E. Passera, G. Chiesa, G. Brambilla, E. Angeli, M. Chiarenza, G. Aranzulla, M. Andresi, V. Errico, E. Bottoni, E. Voulaz, A. Santoro, M. Alloisio
- Abstract
- Presentation
Background
The purpose of this study was to explore the effect of screening with spiral CT on lung cancer mortality in comparison with no screening in a high-risk population. Secondary endpoints were incidence, stage and resectability.Methods
Male subjects, aged 60-75, smokers of 20+ pack-years were randomized to screening with low-dose spiral CT or control. Prospective participants were pre-assessed for eligibility and randomized during a telephone interview, while formal enrolment took place at a later date. All enrolled participants underwent a structured medical interview and physical examination, a baseline, once-only chest X-ray (CXR) and sputum cytology examination. Screening-arm subjects had a LDCT upon accrual, which was repeated every year for four additional years (5 rounds), while controls had a yearly clinical review only, with further testing only if needed.Results
Between March 2001 and February 2006, 2811 subjects were pre-assessed and randomized (CT arm: 1403, control arm:1408). 20 cases of double registration and two test records have been identified in the database, and 2540 subjects actually appeared for assessment (1229 CTR arm, 1299 CT arm), of whom 2450 (1264 CT arm and 1186 Controls) were eligible and enrolled. The two study groups are comparable for age, smoking exposure, and comorbid conditions. As per inclusion criteria, all subjects are males and 99.8 % are 60 or older. As of December 2012, median follow-up was 73.1 months in the control arm and 75.5 months in the screening arm. Altogether, 152 patients were detected during active follow-up with 161 lung cancers: 92 CT-arm subjects (7.27%) versus 60 controls (5.05%), p< 0,0237. 82% of CT-arm lung cancers were detected at scheduled CT examinations, and 55% were stage I disease at the time of diagnosis compared with 27% in the Control arm. The absolute number of lung cancer cases with stage II-IV disease was virtually the same as in the control arm. Resectability rate was similar in the two groups. After linkage with population registries, vital status information is now available for 2528 subjects (99.5%). Overall, 370 subjects have died, for 104 of whom we are currently investigating mortality causes.Conclusion
While the number of participants is relatively small, smoking exposure and average age of participants are higher than in similar trials. A comparatively high number of events has been observed. To date, 28% of such events cannot yet be attributed to a specific cause. Health registry data necessary to complete mortality comparisons is expected to become available in the next few weeks. Merging of all European trials may yield robust data about this strategy in the future.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.
-
+
O05.02 - Local, Surrounding and Global Features for Improved Computer Aided Diagnosis of Lung Cancer (ID 318)
10:30 - 12:00 | Author(s): J.C. Sieren, S. Dilger
- Abstract
- Presentation
Background
The National Lung Screening Trial reported a 20% reduction in lung cancer mortality achieved through low dose computed tomography (CT) screening of the at risk population, compared to screening with chest x-ray. Challenges with clinical implementation of CT screening for lung cancer include the high number of lesions detected that require further follow-up, approximately 97% of which are ultimately diagnosed as benign. A computer-aided diagnosis (CAD) tool can be designed to determine the probability of malignancy of a lung nodule based on objective measurements. While current CAD tools examine the pulmonary nodule’s shape, density, and border, analyzing the lung parenchyma surrounding the nodule is an area that has been minimally explored. By quantifying characteristics, or features, of the surrounding tissue, this study explores the hypothesis that textural differences in both the nodule and surrounding parenchyma exist between malignant and benign cases, which can be utilized to improve CAD performance.Methods
From CT data, several novel feature extraction techniques were developed, including a three-dimensional application of Laws’ Texture Energy Measures to quantify the textures of the parenchyma as well as the nodule. In addition, the densities of the nodule and parenchyma were summarized through metrics such as mean, variance, and entropy of the intensities. The margins of the nodule were characterized following ray casting and rubber-band straightening to analyze mean and variance of border irregularity. Basic demographics and risk factor data were also included. The large feature set was reduced by statistical testing and stepwise forward selection to a few independent features that best summarize the dataset. A neural network was used to classify the cases in a leave-one-out method.Results
To illustrate proof of concept, the CAD tool was applied to 27 lung nodule cases: 10 malignant and 17 benign. These data were diverse with regards to data acquisition protocol, reconstruction kernel and slice thickness – all of which can pose challenges to CAD. Through statistical testing, 36 features were found to be significant predictors of malignancy (p < 0.05), including many textural and parenchymal features. Two of these significant features, selected through stepwise forward selection, were utilized to classify the data: nodule variance (p = 0.0003) and parenchyma median intensity (p = 0.0028). CAD performance achieved a sensitivity of 90%, specificity of 100%, and an accuracy of 96.3%.Conclusion
Preliminary findings indicate features from both the nodule and the surrounding parenchyma have value in distinguishing benign and malignant lesions. This is particularly valuable in the analysis of early detected, small pulmonary lesions (<10mm). In these small lesions, standard CAD approaches are hindered by few CT data voxels contained within the lesion. By incorporating local, surrounding and global features, more information is included and augmented CAD performance may be achieved. Finally, many significant features were identified despite diversity in the CT data acquisition parameters which indicates the suitability of the approach to broad clinical application. We are currently working on applying the CAD tool to a larger dataset.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.
-
+
O05.03 - Screening of Lung Cancer by Low-Dose CT (LDCT), Digital Tomosynthesis (DT) and Chest Radiography (CR) in a High Risk Population: A Comparison of Detection Methods (ID 3018)
10:30 - 12:00 | Author(s): N. Triphuridet, S. Singharuksa, T. Sricharunrat
- Abstract
- Presentation
Background
LDCT has recently been recommended as a screening tool for lung cancer in a high risk population, provided a 20% reduction in lung-cancer specific mortality. Nevertheless, LDCT has some limitations with respect to its high false positive rate, accumulated radiation exposure and relatively high cost. Digital tomosynthesis (DT) is a multisection imaging technique which can improve detection ability of small lung nodules and renders much lower radiation dosage and operation costs.Methods
Thai heavy smokers (>30 pack-years) were enrolled in a prospective study starting from July 2012 to April 2013 (n=580). LDCT, DT and CR were utilized as a screening tool for lung cancer screening. All participants underwent imaging studies on the same day and the results were independently reviewed within a 1-week interval. Abnormal findings were categorized into 3 groups: negative, indeterminate (maximum diameter of pulmonary nodule >5- 9.9 mm), and suspicious for malignancy (maximum diameter of pulmonary nodule > 10 mm, consolidation, obstructive atelectasis, pleural effusion or mediastinal lymphadenopathy).Results
At baseline, LDCT and DT classified 16/580 cases as suspicious for primary lung cancer while CR detected 15/580 cases. Seven cases with positive LDCT and DT findings were tissue-proven primary lung cancer including 3 - stage I cancers, 1 - stage III cancer and 3 - stage IV cancers. CR detected only 3 proven cases of primary lung cancer and all of them were stage IV cancer. The lung cancer detection rate for pulmonary nodule > 10 mm and other suspicious findings was 1.2%, 1.2%, and 0.5% by LDCT, DT, and CR, respectively. LDCT classified 67 cases as indeterminate while DT and CR classified 21 and 11 as such cases, respectively. Two additional primary lung cancer cases were detected at a 3-month follow-up LDCT of the indeterminate group by LDCT (2 cases), DT (1 case) and CR (0 case), respectively. The lung cancer detection rate for pulmonary nodule > 5 mm and other suspicious findings was 1.6%, 1.4%, and 0.5% by LDCT, DT, and CR, respectively. The positive predictive value (PPV) for pulmonary nodule >10 mm and other suspicious findings by LDCT, DT and CR was 43.8%, 43.8%, and 20.0%, respectively, while the PPV for pulmonary nodules of> 5 mm and other suspicious for malignancy findings by LDCT, DT and CR were 10.8%, 21.6%, and 11.5%, respectively. The sensitivity and specificity was 100% and 87%, respectively, for LDCT, and 88.9% and 94.9%, respectively, for DT, and 33.3% and 96%, respectively, for CR.Conclusion
DT is a lung cancer screening modality that is comparable to LDCT, particularly for pulmonary lesions that are larger than 10 mm. and other suspicious for malignancy findings while CR was far inferior to DT and LDCT.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.
-
+
O05.04 - DISCUSSANT (ID 3991)
10:30 - 12:00 | Author(s): S. Lam
- Abstract
- Presentation
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.
-
+
O05.05 - Biennial lung cancer screening by low-dose CT scan - a simulation of cost effectiveness in Canada (ID 2313)
10:30 - 12:00 | Author(s): J.R. Goffin, W.M. Flanagan, A.B. Miller, F.F. Liu, S. Cressman, N. Fitzgerald, S. Fung, M.C. Wolfson, W. Evans
- Abstract
- Presentation
Background
Randomized data support annual screening for lung cancer among smokers using low-dose CT scans. To compare the resource implications of annual versus biennial screening, a cost-effectiveness analysis was undertaken using the Cancer Risk Management Model (CRMM version 2.0.1) in the context of the Canadian publicly funded healthcare system.Methods
The CRMM performs simulations at an individual level and incorporates demographic data, cancer risk factors, cancer registry data, diagnostic and treatment algorithms and health utilities. Outputs are aggregated and costs (in 2008 Cdn dollars) and life-years are discounted at 3% annually. Simulations were performed with a cohort 55-74 years and a ≥30 pack-year (p-y) smoking history recruited from 2012-2032. CT scan sensitivity (Sens) and specificity (Spec) and cohort outcomes were based on NLST and Canadian data. It was assumed 60% of the eligible population participates by 10 years, 70% adhere to the screening regimen, and smoking cessation rates are unchanged. Sensitivity analysis was undertaken.Results
An annual screening program incurs net costs of $2.97 billion and saves 55,000 quality-adjusted life-years (QALYs) at an incremental cost-effectiveness ratio (ICER) of $53,700 per QALY. Under default biennial screening assumptions (Table 1, scenario 3), biennial screening costs are $1.81 billion, saving 32,000 QALYs and producing an ICER of $56,200. In the least favourable stage shift scenario (1) tested, the ICER is $275,000, whereas the most favourable shift (4) results in $49,300. Using Sens/Spec 0.90/0.73 for all scans in scenario 3 produces an ICER of $61,400, whereas changing all incidence scan Sens/Spec to 0.87/0.73 gives an ICER of $60,900. Increasing age of eligibility to 55-79 cost $2.25 billion at an ICER of $58,700 per QALY while requiring a 40 p-y smoking history reduced cost to $1.3 billion at an ICER of $49,800 per QALY. Table 1.
T0, T1 refer to the NLST stage shift at specified time, where T0 equals shift at time zero screen, T1 shift at 12 month screen. T0/T1 indicates an average. CD: the unscreened Canadian stage distribution. PS: NLST post-screening stage shift. *Represents 3[rd] year and all future odd years. **Represents 4[th] year and all future even years. Hyphens indicate years without screening.Year Stage Shift Scenario Sens/ Spec 1 2 3 4 5 0 T0 T0 T0 T0 T0 0.9/0.73 1 CD PS PS PS CD - 2 T0 T0 T0/T1 T1 T1 0.89/0.84 3* CD PS PS PS CD - 4** T0 T0 T0/T1 T1 T1 0.89/0.84 ICER $275,000 $65,000 $56,200 $49,300 $104,00 Conclusion
Compared to annual lung cancer screening, biennial screening reduces net cost but may have a similar ICER. Stage shift assumptions have a significant impact on ICER values. Minor adjustments in Sens/Spec modestly change the ICER. Widening the age range increases but increasing the p-y requirement reduces system costs.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.
-
+
O05.06 - Resource utilization and costs of screening high-risk individuals for lung cancer in Canada (ID 1520)
10:30 - 12:00 | Author(s): S. Cressman, S.J. Peacock, I. Cromwell
- Abstract
- Presentation
Background
In September 2008 the Pan-Canadian early lung detection of lung cancer study recruited 2537 current or former smokers who were determined to have a high risk of developing lung cancer. An economic analysis was conducted to estimate the potential costs and benefits of screening with the aim of knowledge translation and decision aid for provincial screening programs. An analysis of prospectively collected resource utilization and cost data is presented.Methods
Screening costs have been determined, accounting for the cost of all resources utilized to confirm true negative and false positive screen tests as well as early stage treatment costs for resources applied to obtain diagnostic confirmation of true positive and false negative results for screened individuals, treat the primary disease and any subsequent lung cancer within three years. All costs have been calculated from the Canadian public payer’s perspective. The average CT-screening cost over a fixed period of 18 months for the pan-Canadian study participants who did not have cancer was determined and compared with the phase specific costs of true positive and false negative lung cancer screening participants who had a lung cancer diagnosis proven prior to Dec. 31, 2012. The costs for early-detected lung cancer were determined and presented by diagnosis, treatment and surveillance phases of care.Results
The average cost per screened individuals who did not end up having cancer in the first two years of the study was $456 (95%CI: $385-$570) per-person. The average rate of non-invasive investigations to pursue suspicious CT findings was 49% (CI: 45%-54%); depending significantly on the follow-up protocol observed in different participating sites. The rate of invasive investigations for individuals who had true negative or false positive results was low (<0.4%) as was the rate of complication (<0.004%). 85 individuals had lung cancer detected and diagnosed prior to December 31, 2012. The average cost of screening and the subsequent diagnostic workup for the most common detected lung cancer (stage IA and IB non-small cell lung cancer) was $4,233 (95%CI: $3,643-$4,822) per person. Per-person treatment and surveillance costs are presented by stage and mode of treatment for 84 lung cancers found in the early detection study.Conclusion
This information indicates that screening costs are low on average, as are the rates of complications in the screened individuals that do not receive a cancer diagnosis in the first two years of screening. These numbers arrive while Canadian and other national healthcare systems must manage the impacts of several private, opportunistic, lung-screening clinics that are already operational. This study is sponsored by the Terry Fox Research Institute and the Canadian Partnership against CancerOnly 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.
-
+
O05.07 - COPD-based eligibility increases lung cancer detection rate and lowers over-diagnosis in CT screening studies (ID 1721)
10:30 - 12:00 | Author(s): R.P. Young, F. Duan, R.J. Hopkins, X. Deng, C. Chiles, G.D. Gamble, C. Gatsonis, D. Aberle
- Abstract
- Presentation
Background
Based on a 20% reduction in lung cancer deaths in the CT screening arm of the National Lung Screening Trial (NLST), yearly CT screening for lung cancer is now widely recommended. Eligibility for the NLST was based on age and smoking history only. However, we and others propose that multivariate risk models of lung cancer that incorporate variables for chronic obstructive pulmonary disease (COPD), improve risk prediction for lung cancer. The aim of this study was to examine recently published CT screening studies for lung cancer and the effect of having COPD on outcome.Methods
We searched the literature for CT screening studies of lung cancer where spirometry had been done at baseline to assess the effects of spirometry-defined COPD on outcomes. We identified six studies where there was published data reporting spirometry results in lung cancer screening studies. Using this data we objectively measured outcomes stratified or pre-selected on spirometry-defined COPD.Results
By comparing outcomes in these single arm and randomized studies we found the following lung cancer detection rates were between 1.5 to 6 fold higher in current and former smokers eligible for screening with spirometry-defined COPD compared to those with no airflow limitation or normal lungs (Table 1). Only 15% of those screened had advanced stage COPD (GOLD 3-4) The proportion of eligible current or former smokers with COPD had less indolent lung cancers with long doubling times (Table 2), and Survival after surgical resection of early stage CT-detected lung cancers was no different between those with or without COPD at baseline screening. Figure 1 Figure 2Conclusion
We conclude that a COPD-centric approach to lung cancer screening offers a more efficient means of identifying lung cancer (higher lung cancer detection rate), with less over-diagnosis and comparable outcomes to screening those without COPD.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.
-
+
O05.08 - DISCUSSANT (ID 3992)
10:30 - 12:00 | Author(s): H.M. Marshall
- Abstract
- Presentation
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.
Author of
-
+
E08 - Early Endobronchial Tumours (ID 8)
- Event: WCLC 2013
- Type: Educational Session
- Track: Pulmonology + Endoscopy/Pulmonary
- Presentations: 1
- Moderators:N. Kurimoto, D. Fielding
- Coordinates: 10/29/2013, 14:00 - 15:30, Bayside Gallery A, Level 1
-
+
E08.1 - Screening / Early Detection of Central Malignancy (ID 408)
14:00 - 15:30 | Author(s): A. McWilliams
- Abstract
- Presentation
Abstract
Introduction: Excellent cure rates for central lung cancer can be achieved with local endobronchial therapy if lesions can be detected at a pre-invasive stage. Flexible white light bronchoscopy is the most commonly used imaging tool to evaluate the central airways but it has a limited ability to detect small pre-invasive central lung cancers. Screening with low dose thoracic CT (LDCT) has been shown to be useful in the early detection of lung cancer, but it largely detects peripheral lesions. Despite advances in CT technology, LDCT cannot detect early pre-invasive central lung cancers due to limitation of resolution. Method: Optical imaging modalities that are both established and in development will be reviewed and discussed. These include techniques such as autofluorescence bronchoscopy, narrow band imaging, optical coherence tomography, confocal microendoscopy, endocystoscopy and raman spectroscopy. Results: Autofluoresence imaging is the most well proven imaging tool to be used in conjunction with white light bronchoscopy to rapidly detect small preinvasive lesions. Narrow band imaging may also be useful but further comparative studies are needed. Optical coherence tomography and raman spectroscopy are promising techniques that can be easily applied via small probes during flexible bronchoscopy to further evaluate abnormal lesions. Further development of in-vivo microscopic evaluation of abnormal lesions is ongoing using confocal microendoscopy and endocystoscopy although tissue staining and direct contact is currently required. Conclusion: The detection of early central lung cancers requires more sophisticated tools than conventional white light bronchoscopy. The future utilisation of other imaging tools as part of a minimally invasive flexible bronchoscopic procedure appears promising. A multimodality approach will enable the rapid detection and diagnosis of early curable central lung cancers in selected high-risk populations.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.
