DOI: 10.1148/radiol.2462070775
(Radiology 2008;246:653-654.)
© RSNA, 2008
Will the National Lung Screening Trial Be Able to Demonstrate a Mortality Reduction?
David F. Yankelevitz, MD
Department of Radiology, New York Presbyterian Hospital, Building Starr-JO30, 525 East 70th Street, New York, NY 10021
e-mail: dyankele{at}med.cornell.edu
Editor:
In the May 2007 issue of Radiology, the Executive Committee for the National Lung Screening Trial (NLST), in their editorial "Large Field Trial for Lung Cancer Screening: Putting the Wrong Cart before the Horse?" (1), criticizes the suggestion of performing a large field trial, suggesting that only a screening randomized controlled trial (RCT) can avoid the many biases associated with screening and produce evidence regarding mortality reduction.
Two conditions need to be met for a screening RCT to produce meaningful results: (a) The screening is to continue long enough for the reduction in deaths to become fully manifest, and (b) as the deaths averted as a result of the early treatment associated with screening are well into the future, it is necessary to focus the mortality comparison on an appropriate, suitably delayed time interval after the screening's initiation to document the full reduction in mortality (2). This was demonstrated in mammography screening trials, where it was found that only when screening continued for 7–8 years did the full reduction in mortality become manifest. The current design of the NLST allows for only three rounds of screening, with a maximum of 6 years of follow-up from baseline. Therefore, it is likely that the findings of this study will predictably underestimate the benefit of computed tomographic (CT) screening, possibly not showing one at all, even though a true benefit actually exists. Even continued follow-up after the trial completion will not help, as it will only further dilute the true benefit. Recognition of this design error has been credited as being the turning point in the controversy that erupted over breast cancer screening in late 2001 (3).
While no one questions the role of RCTs in treatment trials, their role in the evaluation of the usefulness of screening has led to confusion. In 2002, Congressional hearings were necessary to overcome the damage to the perceived usefulness of mammography on the basis of the results of seven completed RCTs that had been interpreted as showing no mortality benefit, even though it was generally understood that a benefit truly exists (4). Unfortunately, confusion regarding the potential benefit of mammography persists even now (5). As for lung cancer screening, the U.S. Preventive Services Task Force in making their recommendation for screening in 2004 reviewed the available evidence, including six completed RCTs and five case-controlled studies. Primarily on the basis of the conflicting results of these studies, they concluded that they "could not determine the balance between the benefits and harms of screening for lung cancer" giving it an "I" recommendation (6). Of note was that each of those studies reviewed was rated on a three-point scale as to quality (good, fair, or poor), and not a single one was given a good rating (7). Given the major design error in the NLST, it seems likely that this pattern will continue.
The author is a consultant for and shareholder in PneumRx (Mountain View, Calif) and receives royalties from Cornell University from a licensing arrangement for patented technology with General Electric.
 |
References
|
|---|
- Black WC, Aberle DR, Berg CD. Large field trial for lung cancer screening: putting the wrong cart before the horse? Radiology 2007;243(2):314–316.[Free Full Text]
- Miettinen OS, Henschke CI, Pasmantier MW, Smith JP, Libby DM, Yankelevitz DF. Mammographic screening: no reliable supporting evidence? Lancet 2002;359(9304):404–405.[CrossRef][Medline]
- Jackson VP. Screening mammography: controversies and headlines. Radiology 2002;225(2):323–326.[Free Full Text]
- Making Sense of the Mammography Controversy: What Women Need to Know. Committee on Appropriations United States Senate. 107th Congress Second Session. http://frwebgate.access.gpo.gov/cgi-bin/getdoc.cgi?dbname=107_senate_hearings&docid=f:78085.pdf. Accessed November 16, 2007.
- Gotzsche PC, Nielsen M. Screening for breast cancer with mammography. Cochrane Database Syst Rev 2006 Oct 18;(4):CD001877.
- U.S. Preventive Services Task Force. Lung cancer screening: recommendation statement. Ann Intern Med 2004;140(9):738–739.[Abstract/Free Full Text]
- Humphrey LL, Teutsch S, Johnson M; U.S. Preventive Services Task Force. Lung cancer screening with sputum cytologic examination, chest radiography, and computed tomography: an update for the U.S. Preventive Services Task Force. Ann Intern Med 2004;140(9):740–753.[Abstract/Free Full Text]
Response
William C. Black, MD *,
Denise R. Aberle, MD 
,
Christine D. Berg, MD 
, on behalf of the Executive Committee of the National Lung Screening Trial (NLST)
* Department of Radiology, Dartmouth-Hitchcock Medical Center, Lebanon, NH 03756
e-mail: william.black{at}hitchcock.org
Radiological Sciences, David Geffen School of Medicine, UCLA, Los Angeles, Calif
Division of Cancer Prevention, National Cancer Institute, Bethesda, MD
Dr Yankelevitz speculates that the NLST will fail to demonstrate a mortality reduction because of a major design flaw in RCTs of screening that he and his colleagues feel they have identified. The "delay principle" to which they lay claim (1)—that the observed effect of screening depends on the number of screening rounds and the length of follow-up—has actually been well known for decades and is described in standard textbooks of screening (2,3). Contrary to his assertion, the number of screening rounds and the length of follow-up were major considerations in the design of the NLST. In fact, the pretrial power calculations were based on a modification of a statistical model that had been developed specifically to determine the optimal number of screening rounds and the length of follow-up in screening trials (4). As designed, the NLST has a statistical power of 90% to detect a 20% mortality difference between screening with CT and chest radiography on the basis of lung cancer deaths occurring through approximately August 2008. More details about the power calculations are included in an article describing the design of the NLST, which is in its final stage of preparation.
For financial disclosure information, please see Radiology 2007;243:314–316.
|
References
|
|---|
- Miettinen OS, Henschke CI, Pasmantier MW, Smith JP, Libby DM, Yankelevitz DF. Mammographic screening: no reliable supporting evidence? Lancet 2002;359:404–405.[CrossRef][Medline]
- Morrison AS. Assessing the value of early treatment: non-experimental studies. In: Screening in chronic disease. 2nd ed. New York: Oxford University Press, 1992; 98–128.
- Prorok PC, Kramer BS, Gohagan JK. Screening theory and study design: the basics. In: Kramer BS, Gohagan JK, Prorok PC, eds. Cancer screening: theory and practice. New York, NY: Marcel Dekker, 1999; 29–53.
- Hu P, Zelen M. Planning clinical trials to evaluate early detection programmes. Biometrika 1997;84:817–829.[Abstract/Free Full Text]
TOP
References
References 
