NEW YORKThe National Cancer Institute (NCI) is designing a randomized controlled trial of low-dose spiral CT lung cancer screening that would enroll 88,000 Americans.
Richard Fagerstrom, PhD, of the Biometry Research Group of NCIs Division of Cancer Prevention, outlined a proposal for a clinical trial in a presentation at the Second International Conference on Screening for Lung Cancer.
The proposal, which is still under review, calls for 44,000 people to be screened with low-dose spiral CT and a control group of the same number who would receive chest x-rays.
The age range of participants, Dr. Fagerstrom said, is likely to be 50 to 74 years. Another eligibility criterion being considered is a minimum of smoking one pack of cigarettes a day for at least 30 years (30 pack-years).
We have criteria for including former smokers, provided that they quit within the past 5 years of enrollment, Dr. Fagerstrom added.
Once launched, the study is currently expected to take 8 years, with screening done annually for 5 years. We would have about 3 years of recruitment, Dr. Fagerstrom said, and then each individual would be followed for 5 years.
The endpoint of the study, Dr. Fagerstrom explained, would be lung cancer mortality after 5 years of follow-up. Were looking for what we think is an important, and realistic, mortality reduction of 20% at 5 years, he said. Built into the proposal is an assumption that 85% of the enrollees in both arms of the clinical trial would remain in the study for the entire 5 years.
That assumption was challenged by several clinicians involved with the Early Lung Cancer Action Project (ELCAP), which is conducting an observational study of spiral CT screening without a control arm.
We would love to do a randomized trial if we could get only 15% in the control arm not to cross over and get CT somewhere else, said Robert Clark, MD, professor and chairman of radiology, University of South Florida, Tampa. I dont think that would happen. Weve estimated that at least 40% and maybe 50% of controls would seek CT screening on their own.
The reason, Dr. Clark said, is that during the informed consent process, participants would learn the relative benefits of the two methods. Those randomized to be controls, he suggested, might then opt to get a CT scan on their own. Thats a phenomenon we call contamination, he said.
Those likely to take this action, he added, would be those who could pay for it. Thats a major problem because then the people left in the control arm would likely be a socially disadvantaged group.
In an interview, Dr. Fagerstrom acknowledged that contamination is always a possibility when a procedure under study is available outside a trial.
However, he said, the informed consent process will communicate the risks of spiral CT and the unknowns that must be answered by the trial. Then people can choose whether or not to participate. Most likely, those who are strongly inclined to obtain a spiral CT screen on their own would not consent to randomization.
He also noted that consumer focus groups conducted by NCI indicated that a majority of people who would be eligible for the trial did not know about spiral CT and, when informed of the availability of the procedure, were interested in participating in a trial of it.
Most said they would stay until the trial ended, Dr. Fagerstrom said. Only a few said they might drop out of the trial to get spiral CT if they were randomized into the control arm.
Groups Advertising CT Screening
Claudia Henschke, PhD, MD, chief of chest imaging and professor of radiology at Weill Medical College, Cornell University, New York, said that private practice groups in her area are advertising CT screening and that major institutions in Boston are also offering it. I think that it would be very hard to randomize in New York City, she said, and perhaps in the whole East Coast.
Dr. Fagerstrom told ONI that physician focus groups conducted by NCI showed that both primary care physicians and specialists considered a randomized clinical trial essential to answering critical questions about the benefits, risks, and cost effectiveness of spiral CT screening. All said that they would recommend patients to enter a clinical trial rather than recommend the spiral CT procedure itself, he said.
An unidentified member of the audience asked whether it was reasonable to expect 44,000 American health care consumers to accept assignment to chest x-ray screening, an approach that the NCI has said is unacceptable for the past 20 years.
Dr. Fagerstrom acknowledged that studies have not shown a major benefit of chest x-rays for early lung cancer screening. However, he stressed, it has not been shown to be unacceptable.
In fact, he said, earlier at the conference, Pamela Marcus, PhD, of NCIs Division of Cancer Prevention, pointed out that earlier studies of chest x-ray in lung cancer screening have been flawed, and a large NCI trial, the Prostate, Lung, Colorectal, and Ovarian (PLCO) cancer Screening Trial, is investigating chest x-ray to provide a definitive answer about its effectiveness in lung cancer screening.
Moreover, Dr. Fagerstrom said, many physicians consider screening with chest x-ray to be standard care for smokers.