The Fourth International Conference on Screening for Lung Cancer
Location: Weill Medical College of Cornell University
1300 York Avenue, New York NY 10021
Friday, February 23, 2001 to Sunday, February 25, 2001
Summaries of Workshops
Workshops I-V
I. Indications for Screening: Chairs: David Burns and Robert Smith
II. Early Intervention on 'ground-glass opacities': Chairs: Nasser Altorki and David Yankelevitz
III. Behavioral questions to be answered in the I-ELCAP framwork: Chairs: Jamie Ostroff and Suzanne Miller
IV. Role of Biomarkers: Chairs: Carl Kelsey and Fred Hirsch
V. Research on overdiagnosis and curability of lung cancer: Chair: Marek Kimmel
Workshop I: Indications for Screening:
Chairs: David Burns and Robert Smith
Questions:
In the practice of screening for lung cancer, and corresponding practice-oriented research, what questions should be directed to smoking history and how should it be pursued (questionnaire, interview).
Beyond smoking history, what supplementary exposure history should enter into the indication for screening for lung cancer.
Participants:
Summary and Recommendations:
The incidence of lung cancer rises steeply among cigarette smokers after age 45, and the potential years of life saved by effective lung cancer screening programs are likely to be greatest among smokers aged 45-54 years. The relative risk of developing lung cancer declines with smoking cessation, but the absolute rate of disease incidence for former smokers appears to remain constant at a rate similar to that of current smokers of the age at which the former smoker had quit.
Recommendations for lung cancer screening of the general population should be based on the likely benefits to the individuals screened. However, research studies have the additional need to accumulate sufficient incident cases of lung cancer to establish the disease outcome benefits that result from early diagnosis of lung cancer through screening. Since the number of incident cases per thousand people scanned is directly related to the risk of disease in the population screened, research studies may be well served by targeting their programs to those with a higher incidence of lung cancer than would be appropriate to use in setting screening guidelines for the general population. At age 45, current or former smokers have lung cancer incidence rates that, in general, are too low for purposes of current research investigations, and may result in the accrual of too few cases of malignant disease to allow timely resolution of the questions of stage shifting and case fatality reduction. Therefore, targeting those who reach age fifty as current smokers (current smokers age 50 or older) is recommended for maximizing the number of lung cancer cases identified in the incidence scans for I-ELCAP participants.
Research studies also have an obligation to prove that screening programs can lead to prolonged survival with minimal morbidity from screening and diagnostic procedures. Inclusion of a large number of older subjects with shorter life expectancies, or subjects with high disease co-morbidity, may increase the number of incident cases of lung cancer required to demonstrate a statistically meaningful difference in survival resulting from screening; and it may lead to an overestimation of the complication rate for diagnostic procedures. It may be useful to target research studies to subjects age 70 and under for these reasons, even though smokers over age 70 are likely to benefit from screening and have high rates of lung cancer.
The optimal target population for the I-ELCAP combined analyses would be:
Current smokers age 50-70 years, and former smokers age 50-70 years who quit at age 50 or older. At the time of screening, candidates for participation should have had no acute change in symptoms consistent with lung cancer
Screening should begin at age 45 among smokers (or former smokers who quit at age 45 or older) who also have:
- Exposure to asbestos or other proven occupational exposures to lung carcinogens
- Abnormal FEV < 70%
- Smokers who smoke 2+ packs per day.
Subjects with a prior history of aero-digestive cancer should be treated as a separate category in the analyses.
These criteria are not intended to suggest that other populations should not be screened or that data on groups other than those specified would not be useful and important in the process of examining the benefits of lung cancer screening. The criteria are also not intended to be used as inclusion or exclusion criteria for screening programs conducted by I-ELCAP members. The criteria are intended to identify the population where the benefits of screening might be most rapidly and efficiently demonstrated by I-ELCAP, and to provide guidance for population recruitment to members seeking such guidance. More detailed and complete data will be collected at the time of the initial evaluation which would allow more complete and extensive inclusion and exclusion criteria to be applied for future studies conducted by independent I-ELCAP participants.
Development of national and international guidelines for screening of the general population will require consideration of issues beyond the question of effectiveness of the screening program in altering disease outcomes, including issues of cost and resource availability. Based on the incidence rates of lung cancer and potential years of life that could be saved with an effective lung cancer screening program, consideration of these economic and resource issues include those who reach age 45 as current cigarette smokers (current smokers age 45 and older, and former smokers who quit at age 45 or older) in estimating the cost-effectiveness of approaches that offer screening for lung cancer. Further work on the cost-effectiveness and public policy implications of screening these groups will be needed, in parallel with studies of the effectiveness of screening programs, in order to define recommendations for screening for lung cancer in the general population.
Recommendations for data collection
The individual level data that should be collected ads part of the core data for I_ELCAP for purposes of risk assessment would include information on:
Tobacco use including
- Smoking status
- Current amount smoked and possibly amount smoked one year prior to the evaluation
- Age of initiation
- Other tobacco use
- Marijuana use
- Age of quitting
- Respiratory Symptoms
- Parental or spousal smoking status
Prior diagnosis of heart or lung disease
Occupational history1
Family history of lung cancer (parents, spouses, children)
BSA BMI
Alcohol use
Urban/suburban/rural residence
Other disease co-morbidity
Reason for coming to screening
The specific questions or methods used to obtain this information should be developed by a committee with representation from those I-ELCAP members with interest in risk assessment, behavioral outcomes, measurement of biomarkers, information systems and data management. This group should also be asked to consider what information should be collected at the time of each follow-up scan.
1construction, demolition, ship construction, repair, mining, chemical industrial exposures, nuclear industry, building maintenance, with a 10-year cumulative history exposures to these occupations.
Workshop II: Early Intervention on 'ground-glass opacities':
Chairs: Nasser Altorki and David Yankelevitz
Questions:
1. What is now known about the natural history of cancer-type lesions diagnosed in radiologically identified 'ground-glass opacities'? In particular, what would be the frequency of outcome in the absence of intervention?
What is current practice in regards to intervention?
What type of research is most urgently needed?
Participants:
Hisao Asamura, David Beck, Lior Copel, Asger Dirksen, Ali Farooqi, Salvatore Giunta, Orly Goitein, Olga Gorlova, Rajiv Gupta, Ryutaro Kakinuma, Ara Klijian, Karl Klingler, Toshiaki Kobayashi, William Kostis, Xueguo Liu, Mark Logan, James Mulshine, Anna Nicolas, Harvey Pass, G. Sylvester Price, Fred Grannis, Anthony Reeves, Meyer Rosen, Dorith Shaham, Shusuke Sone, Deborah Walter, Junji Yoshida, Javier Zulueta,
Summary and Recommendations:
It was generally accepted by the group that ground-glass opacities (GGOs) can be broadly divided into two categories: simple and complex. The simple lesions were considered to be more indolent than the complex lesions. Complex lesions, in general, were thought of as similar to solid lesions. In regard to the simple lesions, it is generally understood that they are more frequently found on baseline than on repeat screening. Little data were available in regards to the frequency of their progression, however, several groups are currently following cohorts of patients with these abnormalities. A relatively high, although not quantified, percentage of these have remained stable over several years. Little is known whether waiting for these lesions to demonstrate growth has had any impact on their curability, as opposed to immediate resection upon detection. It was generally accepted that once growth was demonstrated, these lesions would continue to progress, ultimately leading to death. Overall, however, the percentage that would start demonstrating growth remains relatively unquantified.
Current practice for simple GGOs varied widely. One defining feature was size. For lesions less than 5 mm in diameter, immediate intervention was not necessary and that 6 month follow-up was reasonable. Beyond this criteria, there were greater differences in the approach, ranging from observation until growth (at any size) to percutaneous biopsy for lesions greater than 3 cm, to immediate intervention for any lesion over 5 mm. The actual type of intervention was also variable. The Japanese group used open/VATS wedge resection (mainly those within outer 1/3 of lung parenchyma) and open/VATS lobectomy for inner (central) lesions. The US group favored lobectomy with hilar/mdiastinal dissection for most lesions regardless of type and size as it is more likely to assure complete removal of the lesion. It was discussed that one important question in the natural history of these lesions is the frequency of additional lesions occurring post-resection and the consequent impact this would have on the ability to perform additional resections. There was a general belief that these lesions tend to recur and would require additional resection, thus tissue-sparing procedures would be advantageous, provided they lead to the same rate of curability. The rate of curability at present for these lesions is considered to be extremely high.
Several areas were identified that pose the most important research questions. It was recognized that there would be a great advantage to pooling of data in regards to these lesions in order to make more meaningful statements with respect to natural history. A data form would be developed and distributed to participating groups for presentation at the next meeting. These data forms would focus on learning the frequency of these lesions and their curability in the presence and absence of intervention. It was agreed that a study should proceed prospectively to further assess these issues as information about natural history will help guide all future interventions. Elements to be included in the study relate to measures of aggressiveness and curability. This includes radiologic measures, such as growth, and biomarker evaluations. It was believed that biomarker evaluation should be done at the time lesions are detected and upon choice of suitable cohorts, patients should be placed into either an early-intervention arm versus an observation arm. Growth of lesions in the observation arm would be considered an indication for immediate intervention. This would enable us to determine whether lesions had transformed from their initial state (as measured by biomarker indices) and would also allow us to determine differences in curability between these alternatives. It was also recognized that improvements in minimally-invasive tissue procurement devices and techniques would be highly desirable to obtain sufficient material for biomarker evaluation. No specific marker recommendations could be made at this point, but a mechanism for tissue banking would be highly desirable to provide for future evaluations.
It was recognized that the current staging system should be reevaluated in light of CT screening findings with particular attention to multiple small malignancies found in the same lobe or in other lobes; these are currently classified as T4 and M1, respectively, although in the screening setting they are not treated accordingly.
Workshop III: Behavioral questions to be in the I-ELCAP framwork:
Chairs: Jamie Ostroff and Suzanne Miller
Questions:
What are the potential psychological and behavioral risks and benefits of undergoing CT scanning for early detection of lung cancer?
What are the psychological and behavioral barriers for recruitment and retention of volunteers, particularly medically underserved, in screening trials?
What psychosocial and behavioral support services should included within the lung screening sites to ensure optimal quality of care and health outcomes?
Participants:
Kim Agnello, Lee Ecoles, Yolanda Faustini, Beth Heckel, Barbara Kupetzky, Scott Rivers, Paula Trief, Rose Tselentis
Background:
Low dose helical CT scanning for the early detection of lung cancer represents one of the most promising recent breakthroughs in the fight to control lung cancer. Nonetheless, prior studies of well-established cancer screening modalities (i.e., mammography) have demonstrated that psychological and behavioral issues play a critical role in translating technological advances in cancer screening into optimal public health benefits. As such, there is a critical need to examine key psychological (psychological distress and quality of life) and behavioral outcomes (smoking cessation) following low dose CT scanning. Assessment of psychological and behavioral outcomes will be essential for future efforts to reach and retain high risk smokers as well as promote positive patient-centered outcomes in lung screening programs.
Summary and Recommendations:
What are the potential psychological and behavioral risks and benefits of undergoing CT scanning for early detection of lung cancer?
At this point in the early development of CT scanning programs, psychological and behavioral outcomes of lung cancer screening have not yet been the focus of systematic investigation. Given the importance of standardized measurement of psychological and behavioral factors, we recommend the adoption of a core assessment battery for the examination of psychological and behavioral outcomes. This core assessment module should be developed and evaluated by a consensus panel of experts in cancer risk assessment and behavioral outcomes. Each site is responsible for ensuring that program participants complete the psychosocial/behavioral assessment module and forward data to the central coordinating unit for analysis and monitoring.
To date, there have been no reported findings of adverse psychological reactions secondary to lung cancer screening. Nonetheless, based on prior research with other cancer screening modalities (i.e., mammography) and clinical observation, there is a potential for adverse psychological impact, particularly elevated psychological distress among individuals with premorbid psychiatric history, and those who receive either positive or indeterminate CT findings.
Our preliminary data has demonstrated that screening program enrollees are likely to make quit attempts following enrollment in the CT scanning program and are receptive to receiving smoking cessation advice and assistance from CT scan programs. However, preliminary data suggest caution that those who receive negative CT findings may be more likely to continue smoking. We recommend that informed consent include mention of the health risks of continued smoking and advise smoking cessation.
What are the psychological and behavioral barriers for recruitment and retention of high risk volunteers, particularly medically underserved, in screening trials?
Based on the historical under-representation of select populations in medical research and their well-documented excess disease risk, each site should identify and establish strategies to overcome sociocultural barriers for ensuring equal access to the lung cancer screening program.
Due to the importance of tracking patients over time and retaining program enrollees, we recommend obtaining alternative contact information to reduce loss to follow-up. Although, to date, reported rates of nonreturn for repeat scanning have been low, the optimal outreach protocol for insuring high rates of return has not yet established. Therefore, in order to identify optimal follow-up retention procedures, we recommend that screening sites document all contact efforts made to ensure ongoing adherence to the screening protocol. This follow-up documentation is essential for explaining any observed site variation in adherence due to population or systems factors.
What psychosocial and behavioral support services should be included within the lung screening sites to ensure optimal quality of care and health outcomes?
Considering both the large number of current smokers likely to seek CT scanning as well as the availability of evidence-based consensus guidelines for smoking cessation, we strongly recommend that all screening sites provide brief, smoking cessation advice and assistance consistent with
standard care practice guidelines espoused by the Agency for Health Research & Quality (AHRQ) and the World Health Organization (WHO). In order to meet this recommendation, all sites should utilize trained personnel for these functions and these activities should be included in the comprehensive ongoing quality assurance program.
We encourage all screening sites to provide, upon request, information about psychosocial support services and/or referral to screening program participants. Along these lines, we recommend that informed consent includes mention of the potential risk of acute psychological distress inherent in undergoing any cancer surveillance as well as any provisions for support and/or referral.
Workshop IV: Role of Biomarkers
Chairs: Carl Kelsey and Fred Hirsch
Questions:
How promising is the biomarker topic in each of the following areas of potential application: risk assessment, diagnosis (determining the presence of cancer), and in prognosis (determining how fatal the cancer is).
What are the principal impediments to documenting the leading markers in screening, e.g., cost, requisite competence, etc.
Is there the option of storing specimens for future selective marker assessments?
Participants:
Members: Steven Belinsky, Christian Brambilla, Randy Hughes, Thomas Myers, Masayuki Noguchi, Branko Palcic, Mark Pasmantier, Daniel Ray, Stefan Schaller, Thomas Scherer, Joseph Schoepf, Melvyn Tockman, Michael Unger, Jim Voegele, Polly Yin, Tomoyuki Yokose.
Summary and Recommendations:
CT screening for lung cancer presents the opportunity to incorporate and assess the utility of additional biomarkers in the framework of the current protocols. At present, the field is in the process of developing, understanding and subsequently validating a wide variety of biomarkers for application in medicine and public health, including prevention, intervention, diagnosis, prognosis and therapeutic evaluation.
The workshop participants voiced agreement that there is a definite role for biomarkers beyond imaging. Biomarkers are likely to complement imaging by stage, by cell type and by location. A central theme of the workshop emerged as a consensus; The principal impediment of biomarker development has been the lack of longitudinally well characterized specimens from well designed studies that include high quality corresponding patient data. The ELCAP initiative has the potential to overcome this impediment and the workshop participants concluded that the current overriding need is for development of a Biomarker Group within the ELCAP system
Most broadly, the mission of this workshop was to develop a framework for the practical incorporation of biomarkers into the ELCAP infrastructure. Rather than address specific biomarkers or specific areas within the biomarker paradigm, the workshop aimed primarily at broadly defining the field, articulating the roles of biomarkers and then specifically mapping a strategy for incorporating biomarkers into the ELCAP program.
The participants began by operationally defining a "biomarker" as a measurement or observation that indicates risk, of affects detection or prognosis of the process of carcinogenesis. In this framework, malignant histologic or cytologic observations are not biomarkers, but measures of outcome.
The Workshop then focused upon delineation of the roles of biomarkers with emphasis upon their application to the extant screening program. The potential roles included:
1. Risk Assessment: Determination of who should be screened.
2. Triage and tool selection for detection and treatment
3. Assessment of susceptibility
4. Detection of the process of carcinogenesis
5. Monitoring intervention (use of intermediate markers)
6. Prognosis
It was recommended that:
1. The development of a formal Biomarker Working Group in ELCAP.
The Goals of the group were articulated as follows:
1. To determine the extent to which biomarkers complement imaging in the detection and early intervention in lung carcinogenesis.
2. To obtain knowledge concerning lung cancer biology in a screening population.
The workshop further suggested that the strategy for ELCAP to meet these goals should include the establishment of formal collaboration with "expert groups" for specifying markers and specimens for ELCAP-related research (including but not limited to NCI; EDRN and SPORE groups, EU Early Detection of Lung Cancer, IASLC/pathology groups).
As a final effort, the workshop drafted a charge to the working group. The I-ELCAP Biomarker Working Group should:
1. include a representative from each participating institution
2. develop a bioinformatics infrastructure with links to the existing database.
3. develop mechanisms for initiating and facilitating studies with collaborative groups
4. develop protocols for specimen collection, preservation and archiving
5. develop mechanisms for independent funding
6. work to evaluate biomarkers both in screening and subsequent diagnostic techniques (e.g.LIFE scopes and FNA).
Workshop V: Research on overdiagnosis and curability of lung cancer:
Chair: Marek Kimmel
In the studies by Flehinger et al. and Sobue et al.
a) were practice relevant parameters addressed (quantitatively)?
b) were serious biases involved in the studies?
Participants:
Nathaniel Berlin, Anat Caspi, Robert Comis, Stanley Fox, Marek Kimmel, Daniel Libby, Deborah Marshall, Colin McCulloch, Myron Melamed, Olli Miettinen, Melvin Tockman
Summary and Recommendations:
Definitions
1. Cancer (or progressive cancer): Progressive and eventually fatal, given no treatment and competing causes
2. Diagnosed case of cancer: A wider category, including progressive cancers and overdiagnosed (non-progressive, "indolent") cases.
3. Overdiagnosis: Calling lesion cancer if it is not intrinsically progressive and fatal.
4. Cure: Termination of the existence of cancer (definition relevant for progressive cancers).
5. Curability: Frequency with which intervention leads to cure.
What is needed and what is to be studied
The following concepts are relevant for medical practice:
1. Diagnostic distributions, i.e., distributions of diagnosed cases, according to indicators of how early is the diagnosis (at baseline, at repeat screens, or in interval between screens) and prognostic indicators (size, markers, etc.). These indicators define sub-categories of cases.
2. Relative frequency of overdiagnosis (specific to the means of detection), in sub-categories of cases.
3. Significance of detection in given sub-category, leading to the timing of fatal outcome, in the absence of intervention and competing cause of death, in the sub-category and curability in the category.
4. Other indices can be obtained by combining the above, depending on the screening procedure, population of subjects and other factors.
Discussion of the papers by Flehinger et al. (1993) and Sobue et al. (1992)
In these two papers the paradigm leading to the ELCAP approach was first established. Stage 1 screening-detected lung cancers were resected or not resected based on criteria independent of the prognostic category. In both papers the non-resected cases exhibited significantly worse survival.
Ensuing discussion underscored the need to precisely define the diagnostic categories and to document the causes for non-resection.
Conclusion
1. With suitably defined diagnostic/prognostic categories, assignment of cases to treated or non-treated arms is effectively random. Therefore all practice-relevant parameters are possible to study within the experimental arm of the study.
2. With suitably defined diagnostic/prognostic categories, assignment of cases to treated or non-treated arms is effectively random. Therefore, interventive issues can be validly addressed in the spirit of Flehinger and Sobue papers, i.e. using as controls patients refusing treatment or given no treatment for reasons independent of their lung cancer.
3. Were a randomized trial to be carried out, contrasting screening with non-screening, the practice-relevant parameters could b studied in the framework of experimental arm alone. However, it would provide an opportunity to apply those results to the control arm, and predicting the cause-specific mortality in it, notably as it depends on the time.
<< Back to The Fourth International Conference on Screening for Lung Cancer General Information
Stay Up-to-Date!
Sign up below to join our email list and receive notice of site updates and other I-ELCAP news.
Contact Us
Have questions? Contact us by email, telephone or mail.
Get Screened
Do you need to reach one of our member sites? Search our site locator, or skim our full list of member sites.
Thank You
I-ELCAP would like thank Idologic, Inc., for its donation of hosting and support services.