| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Thoracic Imaging |
1 From the Department of Radiology (K.G., D.A.L.) and Division of Pulmonary Medicine and Critical Care (R.L.K., Y.E.M.), University of Colorado, Denver Veterans Affairs Medical Center; Department of Biostatistics and Epidemiology, School of Medicine, University of Colorado Health Sciences Center, Denver (T.B.); and Department of Oncology, University of Colorado Comprehensive Cancer Center, Denver (K.K., A.L.K.). From the 2001 RSNA scientific assembly. Received November 20, 2001; revision requested January 11, 2002; final revision received April 22; accepted April 25. Supported in part by a grant from the University of Colorado Lung Cancer SPORE (Specialized Program of Research Excellence). Address correspondence to K.G., University of Colorado Hospital, Anchutz Outpatient Pavilion (AOP), Department of Radiology, 1635 N Ursula St, Campus Box F726, Aurora, CO 80010 (e-mail: kavita.garg@uchsc.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Subjects are being recruited into a randomized controlled trial to undergo either low-dose spiral computed tomography (CT) or observation. Subjects are from a high-risk group with known chronic obstructive pulmonary disease and sputum atypia and a moderate-risk group randomly selected from the general population of a Veterans Affairs Medical Center. All subjects must be 50–80 years of age with 30 or more pack-years of cigarette smoking and must not have undergone chest CT during the previous 3 years. Baseline screening CT is performed with 50 mA, 120 kVp, 5-mm collimation, and a pitch of 2. CT scan interpretation and management of nodules is based on Society of Thoracic Radiology guidelines. The 
2 test for categoric data was used for statistical analysis.
RESULTS: To date, 304 eligible subjects have been contacted, and 239 (79%) have agreed to participate in the trial. One hundred nineteen (88%) of the 136 subjects in the high-risk group and 120 (71%) of the 168 subjects in the moderate-risk group agreed to randomization (P < .001). To date, 190 subjects have been randomized. Of the first 92 subjects examined with CT, 22 (40%) of 55 in the high-risk group and eight (22%) of 37 in the moderate-risk group had one to six noncalcified nodules that required follow-up (P = .07). In all but three subjects, nodules were smaller than 5 mm. Two of the three larger nodules were malignancies.
CONCLUSION: Findings of this study indicate that a randomized controlled trial of CT to screen for lung cancer is feasible.
© RSNA, 2002
Index terms: Cancer screening, 60.30 • Lung neoplasms, CT, 60.12115 • Lung neoplasms, diagnosis, 60.12115
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although lung cancer alone accounts for more cancer deaths than the next three most common causes (ie, colorectal, breast, and prostate cancer) combined, lung cancer is the only one of these four cancers for which screening is not available. In the 1970s and early 1980s, screening trials conducted at Johns Hopkins, Baltimore, Md; Memorial Sloan-Kettering, New York; and the Mayo Clinic, Rochester, Minn, uniformly failed to show a reduction in mortality as a consequence of periodic chest radiography and cytologic analysis of sputum (5–8). This outcome prompted the National Cancer Institute and other health policy and research groups to conclude that large-scale radiologic or cytologic screening for early lung cancer was not efficacious.
An improved understanding of lung cancer epidemiology and biology has emerged since these trials were conducted. Incidence of lung cancer among smokers with airflow obstruction secondary to chronic obstructive pulmonary disease (COPD) is markedly increased compared with that among smokers with similar smoking histories but no airflow obstruction (9,10). In addition, cytologic atypia detected in expectorated sputum has repeatedly been associated with a higher incidence of bronchial dysplasia and the development of lung cancer (11–15). On the basis of a combination of tobacco exposure history, presence of airflow obstruction at pulmonary function testing, and presence of sputum atypia, cohorts at high risk for lung cancer can be identified.
Improvements in computed tomographic (CT) technology (low-dose spiral CT) and thoracic imaging have paralleled advances in lung cancer biology. In the early 1990s, the Japanese Anti-Lung Cancer Association (ALCA) introduced low-dose spiral CT screening for detection of lung cancer in its early stages (16,17). In the United States, the Early Lung Cancer Action Project (ELCAP) was initiated in 1992 (18). Findings of ALCA and ELCAP studies clearly showed that CT is superior to chest radiography in the detection of early lung cancer. Approximately 80%–95% of the lung cancers detected in ALCA and ELCAP studies were stage I. However, these studies were not designed to measure the effect of screening on mortality or severe morbidity from lung cancer, as they did not include a control group. Several other observational studies of an evaluation of spiral CT screening for lung cancer were initiated, but, to our knowledge, no findings of randomized controlled trials in which low-dose spiral CT was used for lung cancer screening have been reported. Because of the encouraging results from the observational studies, concern has been raised that subjects may not agree to participate in randomized controlled trials with CT and, thus, would not enable comparison with the current standard of care, which is no screening at all. This led us to our current study, the purpose of which was to assess the feasibility of conducting a randomized controlled trial for lung cancer screening among subjects with varying degrees of lung cancer risk.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Individuals who satisfied eligibility criteria signed a study-specific consent form approved by the institutional review board.
Randomization
After permission from the primary care physicians is obtained, all potential subjects are contacted by telephone. The nature of the study is explained, including the fact that there will be an equal probability that a subject would be assigned to undergo a CT screening examination or observation. Subjects are told that the CT examination could lead to further testing. Subjects are informed that this would be a 2-year study, with a repeat CT examination in 1 year. Subjects are also informed that they would be contacted 1 year after enrollment to answer a standard questionnaire. There is no cost for the examinations, and any subsequent diagnostic procedures are covered by the VA system. After the subjects sign the consent form, they are randomized to undergo either low-dose spiral CT or observation. All current smokers are offered a referral to the Smoking Cessation Clinic at the VA Medical Center, and dietary and physical activity information is provided for all subjects. Included is the American Cancer Society pamphlet Choices for Good Health: Guidelines for Diet, Nutrition, and Cancer Prevention. The basic study design is shown in Figure 1. We plan to enroll a total of 400 subjects for this pilot project (200 high-risk and 200 moderate-risk subjects).
|
Image Reading
Each study is reviewed by a board-certified chest radiologist (K.G.) with 10 years of experience in the evaluation of thoracic CT scans, and findings are recorded on a standardized score sheet. The first 20 studies and every subsequent sixth CT scan or any scan for which the first reader has questions are then independently read by a second board-certified radiologist (D.A.L.) with 15 years of experience in the evaluation of thoracic CT scans. Consensus opinion is used as the final interpretation in these examinations. Both readers are blinded to the risk category of each subject. The formal interpretation is incorporated into the patient’s medical record. The presence, size (in three dimensions), and edge characteristics (ie, well defined, poorly defined, round, lobulated, spiculated) of the nodules as well as calcification, fat content, and location (ie, peribronchovascular, subpleural, lobar) are recorded on the basis of the consensus statement of the Society of Thoracic Radiology (19). The following definitions are used: Size was defined as the average of length and two orthogonal diameters (anteroposterior and transverse) on a section. The shape was classified as round if the ratio of length and width was greater than 0.75. Nodules were classified as well-defined if the borders were clearly circumscribed. Benign calcification was diagnosed if there was complete homogeneous calcification of the nodule. If the nodule was located within 2 cm of the pleural surface and/or costal margin, it was classified as peripheral; otherwise, it was classified as central. If no noncalcified nodules were identified, the examination was classified as negative. The examination was classified as positive if one to six noncalcified nodules were identified according to the criteria used by Henschke et al (18). If low-dose CT showed benign calcification or substantial fat associated with a round well-defined nodule, the nodule was classified as benign. All other nodules were fully evaluated with diagnostic thin-section CT in accordance with the current standard of care. In four (25%) of 16 subjects with nodules smaller than 5 mm, a low-dose protocol similar to that used with baseline CT was used for follow-up CT. All subjects with negative scans will undergo repeat CT in 1 year.
Diagnostic CT Protocol
The scanning parameters for diagnostic CT were as follows: 200 mA (standard dose), 5-mm collimation, and 2:1 pitch. Scanning started from the lung apices and extended inferiorly to include the adrenals. Both lung and mediastinal windows were imaged. Thin-section scanning through the nodule was performed for characterization. Each nodule was evaluated with thin-section CT to include the entire nodule by using a collimation of 1 mm and pitch of 1:1.5. If a nodule was larger than 6 mm and indeterminate in appearance and if the subject had no known contraindication to intravenous contrast material, contrast material–enhanced CT was performed according to the protocol described by Swensen et al (20).
Management of Nodules
Nodule management was based on the consensus statement of the Society of Thoracic Radiology (19). Round nodules that showed diffuse and/or benign calcification or substantial fat content were excluded from further study. Solid smooth-edged nodules that did not show "benign calcifications," air bronchograms, or converging vessels were classified as indeterminate. They were not spiculated and were of unknown chronicity. The follow-up interval for indeterminate nodules was often dictated by the subject and the attending physician. The following strategy was commonly followed: For nodules smaller than 5 mm, follow-up was performed with thin-section spiral CT at 3 and/or 6, 12, and 24 months. If no growth is noted for 2 years, the nodule is classified as benign. Biopsy and/or removal should be considered for nodules that increase in size. For 5–10-mm nodules, thin-section spiral CT is performed at 3, 6, 12, and 24 months. Biopsy and/or removal of nodules should be considered for nodules that increase in size. Nodules larger than 10 mm were evaluated on an individual basis with dynamic contrast-enhanced CT, positron emission tomography, and/or biopsy. Nodules that enhance more than 15 HU may require a 6-week follow-up (if they are thought to be inflammatory) or percutaneous, bronchoscopic, or surgical tissue evaluation. Nodules that enhance less than 15 HU require a follow-up with thin-section CT as described previously; 99% of these nodules will be benign (20). For the purpose of this study, focal or linear opacities with nodular components that do not allow exclusion of malignancy with certainty are referred to as atypical scars. The follow-up interval for these atypical scars was also dictated by the subject and his or her attending physician.
Statistical Evaluation
Adequacy of randomization was assessed with statistical comparisons of the two randomized groups, wherein the categoric variables were assessed with the 2 test, and the continuous variables were assessed with the t test. A P value of less than .05 was used as the threshold for statistical significance. To determine the study size, the approach we used in this study was to have a size large enough to estimate the approximate response rate to the invitation to screening. We did not intend to test any deviation from a priori specified response rate. Empirically, a 50% response rate for randomization was considered good, and a response rate greater than 50% was considered high for this trial. All data were analyzed by using commercially available software (StatMost, version 2.50, Dataxiom Software, Salt Lake City, Utah).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
The baseline CT scans in 60 (65%) of these 92 subjects did not show a noncalcified nodule. Among the remaining two baseline CT studies, one individual showed more than six (approximately 15) noncalcified nodules in a pattern consistent with multiple granulomas, and the other showed multiple pulmonary masses that were subsequently proved to be metastases from a laryngeal cancer.
Twenty-two (40%) of 55 subjects in the high-risk group had noncalcified nodules, compared with only eight (22%) of 37 in the moderate-risk group (P = .07). The remaining subjects did not have nodules, but six (11%) of 55 subjects in the high-risk group and one (3%) of 37 in the moderate-risk group had atypical scars that required follow-up CT.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
It is known that COPD by itself increases the risk of lung cancer by a factor of 2–3 (13). Though the sample size of this study was small and the follow-up short, at this time it appears that most of the nodules smaller than 5 mm are benign. Sixteen subjects had nodules smaller than 5 mm; no change in nodular size or appearance was seen on their follow-up scans. The use of the same field of view and collimation helps make comparisons of two studies easier. For follow-up scans of these nodules, a regular x-ray dose of 200 mA was used in the majority of individuals. In four (25%) of 16 subjects, a low-dose protocol similar to that used with baseline CT was used for follow-up scans and was equally sensitive.
Though we have not yet systematically assessed the compliance of our individuals with follow-up studies, we are observing that it is not uncommon for them to delay their follow-up studies. Full compliance with not only the initial CT examination but also the recommended follow-up examinations will be an important aspect of this ongoing feasibility study. For nodules larger than 5 mm, thin-section CT is indicated, as has been previously suggested (18,19). Investigators (25–27) have reported that some early, more indolent, and less invasive cancers may appear as ground-glass opacities. Multisection CT protocols for lung cancer screening with thinner collimation may be more sensitive in the assessment of these ground-glass opacities and low-attenuation nodules. In the future, computer-aided detection and follow-up of nodules may become available in clinical practice.
Among the first 92 CT examinations, there was a trend toward a significantly higher prevalence of nodules in the high-risk group (40% vs 22%, P = .07). A longer follow-up period will be necessary to determine whether there will also be a higher incidence of malignancy in this group. Scars were also found more often in patients with COPD. Though most of the scars have a recognizable pattern on CT scans, in many instances nodular scars were seen that required follow-up CT at 3- or 6-month intervals, depending on the nodular soft-tissue component. A longer follow-up will also be required to assess the natural course and clinical importance of these nodular scars.
Our study has a number of limitations. The VA population pool includes few women, so the compliance rate of potential female participants cannot be assessed. The small sample size of this study does not allow conclusions that are based on the variation in agreement rates in high- versus moderate-risk subjects, which is also confounded with the high-risk group being preselected as one consisting of subjects who already have agreed to participate in another study. We do not know if a high-risk group in a future study would be more likely to agree to participate or even less likely to agree to do so. All subjects have consented to participate for 2 years. It is not clear if the agreement rates for randomization would have been different if the subjects were randomized for a longer period of time. The adherence to randomization for a longer period of time and a larger sample size will be needed to detect any effect on mortality rates.
The results of this study so far strongly suggest that a randomized controlled trial in which subjects are randomized to either CT or standard care (no imaging) is feasible. The rates of agreement to be randomized to a trial are quite high in both those at very high risk for lung cancer because they have COPD and those at only moderate risk because they smoke tobacco. A large-scale randomized controlled trial with subjects from the VA general population may well be feasible. A multiinstitutional randomized controlled trial on a much larger scale is warranted to detect any effect of CT screening on the mortality rate from lung cancer, and results of this study suggest that such a trial could be conducted.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, K.G., A.L.K., Y.E.M.; study concepts, K.G., T.B., Y.E.M.; study design, T.B.; literature research, K.G., D.A.L., T.B., Y.E.M.; clinical studies, Y.E.M., R.L.K.; experimental studies, K.G., Y.E.M., R.L.K.; data acquisition, K.G., R.L.K., Y.E.M.; data analysis/interpretation, K.G., T.B., K.K., D.A.L.; statistical analysis, T.B., R.L.K.; manuscript preparation and definition of intellectual content, K.G.; manuscript editing and revision/ review, K.G., T.B., K.K., D.A.L., Y.E.M., R.L.K.; manuscript final version approval, K.G.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. Kubo, P.-J. P. Lin, W. Stiller, M. Takahashi, H.-U. Kauczor, Y. Ohno, and H. Hatabu Radiation Dose Reduction in Chest CT: A Review Am. J. Roentgenol., February 1, 2008; 190(2): 335 - 343. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Black, R. de Verteuil, S. Walker, J. Ayres, A. Boland, A. Bagust, and N. Waugh Population screening for lung cancer using computed tomography, is there evidence of clinical effectiveness? A systematic review of the literature Thorax, February 1, 2007; 62(2): 131 - 138. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Mascalchi, G. Belli, M. Zappa, G. Picozzi, M. Falchini, R. D. Nave, G. Allescia, A. Masi, A. L. Pegna, N. Villari, et al. Risk-benefit analysis of X-ray exposure associated with lung cancer screening in the Italung-CT trial. Am. J. Roentgenol., August 1, 2006; 187(2): 421 - 429. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Bastarrika, M. J. Garcia-Velloso, M. D. Lozano, U. Montes, W. Torre, N. Spiteri, A. Campo, L. Seijo, A. B. Alcaide, J. Pueyo, et al. Early Lung Cancer Detection Using Spiral Computed Tomography and Positron Emission Tomography Am. J. Respir. Crit. Care Med., June 15, 2005; 171(12): 1378 - 1383. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Gohagan, P. Marcus, R. Fagerstrom, P. Pinsky, B. Kramer, and P. Prorok Baseline Findings of a Randomized Feasibility Trial of Lung Cancer Screening With Spiral CT Scan vs Chest Radiograph: The Lung Screening Study of the National Cancer Institute Chest, July 1, 2004; 126(1): 114 - 121. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Brenner Radiation Risks Potentially Associated with Low-Dose CT Screening of Adult Smokers for Lung Cancer Radiology, May 1, 2004; 231(2): 440 - 445. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
