February 15, 2007 — Recent calls for widespread computed-tomography (CT) screening to detect early lung cancer in smokers and other individuals at high risk are premature, say experts in the field: a decision on whether or not to introduce such screening should not be made before the results of ongoing randomized trials are in.
The calls for CT screening for lung cancer were sparked last October by publication of results from the International Early Lung Cancer Action Program (I-ELCAP) in the New England Journal of Medicine (NEJM) (Henschke C et al. N Engl J Med. 2006;355:1763-71). Led by Claudia Henschke, MD, PhD, from the New York-Presbyterian/Weill Cornell Medical Center, the researchers claimed that CT scanning in high-risk populations could prevent about 80% of deaths from lung cancer by detecting the disease in its earliest stages. They also claimed that rates of detection and cost-effectiveness are similar to that of mammography screening for breast cancer. The news was widely reported by all media outlets across the United States.
However, the study has come in for some criticism since its publication, both in correspondence in the February issue of the NEJM and in 2 editorials in other medical journals. Both editorials and several of the letters point out biases in the I-ECLAP study, emphasize the fact that there was no control group, and urge restraint and waiting for results from randomized studies to come in before any new move is made.
In the February issue of the Journal of the American College of Radiology, editorialist Bruce Hillman, MD, from the University of Virginia, Charlottesville, says that there were significant biases in the I-ELCAP study that could explain some or all of the positive effect observed. He points out that the ongoing National Lung Screening Trial (NLST), which he is involved with, was "specifically designed to supersede the biases in the I-ELCAP study and to definitely answer the question of whether computed tomography can improve on chest radiography in reducing the lung cancer–specific mortality rate." Results from the NLST, which is sponsored by the National Cancer Institute (NCI), are expected in 2009 but may come earlier if one of the modalities has a profound effect.
At this stage, a population-based CT scanning program is of "uncertain value and perhaps even harmful," Dr. Hillman comments. "Our best course is to wait for results of the NLST and several other randomized trials of CT lung cancer screening now under way."
The editorial in the February 10 issue of the BMJ reaches the same conclusion. "We must be careful what we promise the public," say editorialists Pamela McMahon, PhD, and David Christiani, MD, from Massachusetts General Hospital/Harvard Medical School in Boston. "Evidence-based science should inform policy," they comment, pointing out that failure to do so led to a "hasty about-face" on hormone replacement therapy when randomized trials (as opposed to observational studies) showed no overall benefit on mortality.
No Control Group
A major criticism of the I-ECLAP study is the lack of a control group. Dr. Henschke and colleagues reported that individuals who had early lung cancer detected by CT screening went on to have excellent long-term survival (88% at 10 years). But the BMJ editorial points out that the longer survival of patients with screening-detected cancers results from a combination of lead time bias (screening detects cancer earlier), length bias (periodic screening detects a large proportion of slow-growing cancers), and overdiagnosis bias (it may detect slow-growing cancers that do not need treatment).
"Without a control group, it is difficult if not impossible to distinguish between these effects or even to be sure that screening has any true effect at all," the editorialists comment.
As a case in point, they hark back to the Mayo lung project, one of several trials of chest X-ray screening performed during the 1970s. "Patients in the intervention arm had higher survival rates than those in the control arm, but screening had no effect on mortality, even after nearly 3 decades of follow-up. If this study had not included a control arm, the higher survival in patients with cancers detected by screening rather than via usual care might have prompted the inappropriate adoption of widespread chest X-ray screening," Drs. McMahon and Christiani write. "Given this hindsight, will we still adopt computed-tomography screening on a similar grade of evidence from the I-ECLAP study?"
Risk of Harm From Invasive Follow-Up Procedures
It may turn out that CT screening does reduce mortality from lung cancer, the BMJ editorial concedes. But any potential and as-yet-unproven benefit must be balanced against the risk of morbidity from invasive follow-up procedures, including biopsies and needle aspirations. The high rate of false positives with CT screening means that some of these procedures will be carried out in healthy individuals, whose pulmonary lesions turn out to be benign.
"We must have a plan for how we will effectively handle the very large number of individuals who will be identified as 'possibly' having lung cancer and who will suffer the consequences until that diagnosis is proven or not," comments Dr. Hillman in his editorial. Another issue is that of who will care for these people. As individuals who seek screening may not be under a physician's care, the radiologist will need to become the patient's physician of record, at least until that care is transferred to another doctor, he suggests. "Radiologists must be prepared to handle this responsibility to ensure that patients get appropriate follow-up care on the basis of the scan results."
Criticism of the Way the News Was Propagated
Dr. Hillman is also critical of the way in which news from the I-ELCAP study was propagated, giving the impression that there was a "cure" for deadly lung cancer. The combination of the study appearing in the NEJM, an apparently positive result for a dread disease, and "a very savvy public-relations effort on the part of the investigators and their institutions made for a media sensation," he writes.
"Media outlets appreciate a simple message," he points out, and "the I-ECLAP authors presented such a message: CT screening works, and everyone at risk ought to undergo it." It is much more difficult to explain in a short sound bite why showing longer survival attributable to a screening test, for which the lead time and length biases are undoubtedly exaggerating the effect, doesn't necessarily mean that there is any real benefit to be derived from CT scanning in reducing the number of lung cancer deaths, he comments.
Question Over Ethics of Continuing Ongoing Trial
The point about waiting for further results from randomized trials is made in several of the letters to the NEJM, including one from the principal investigators of the NLST.
However, another letter to the NEJM questions whether it remains ethical to proceed with NLST. As the results from I-ECLAP demonstrate the effectiveness of CT screening for detecting genuinely cancerous lung lesions while they are still curable, Sophie Kulaga, PhD, and Igor Karp, MD, PhD, from McGill University, in Montreal, Quebec, ask whether it is still ethical to randomly assign individuals to something other than CT imaging.
The NSLT now has the opportunity "to do the right thing" and to offer CT screening to all participants, they write. This would enhance public trust in clinical research and provide a chance of repeating the results from I-ECLAP, which is a requirement of scientific research. "We hope that the many years of struggle and controversy over the interpretation of findings from clinical trials of screening mammography may thereby be avoided," they comment.
N Eng J Med.2007;356:743-747.
J Am Coll Radiol. 2007;4:83-85.
Medscape Medical News © 2007
Cite this: Zosia Chustecka. CT Scanning for Lung Cancer: Wait for Randomized Trials - Medscape - Feb 15, 2007.