RSNA: NLST may trigger new screening model
Less than one month after the NLST showed a 20 percent mortality reduction among trial participants screened with low-dose helical CT compared with those screened with chest x-rays, researchers convened to provide new insights into the data and detailed the study design, challenges and future directions.
The prospective randomized trial of 50,000 asymptomatic current and former smokers age 55 to 74 years with a minimum 30 pack-year smoking history targeted lung cancer specific mortality with a 1:1 randomization into screening CT or chest x-ray, explained Aberle. Each participant underwent a baseline screen followed by two annual screens.
Researchers compared demographic characteristics of the NLST cohort to the general population using the tobacco use supplement from U.S. Census, matched the groups with respect to gender and median pack years. They reported that the NLST population is somewhat healthier than the target population. However, 25 percent of NLST participants self-reported underlying lung disease such as emphysema or chronic obstructive pulmonary disease, and a relatively high proportion reported a family history of lung cancer.
Aberle and colleagues defined a positive screen as any study with a nodule 4 mm or larger or other findings potentially related to lung cancer. Initial screening delivered a fairly high false positive rate of 27.3 percent, which dropped to 16.8 percent by the third screen. “Most positive screens did not have lung cancer,” reported Aberle. Ultimately, the CT arm produced a 2.3 increase in lung cancer detection. Plus, five-year lung cancer survival in the CT arm was 58 percent compared with 33 percent in the chest x-ray cohort.
Screening dose at a glance
Overall, when medical physicists analyzed radiation dose they found that average effective dose was 1.4 mSv, offered Fred J. Larke, MS, of University of Colorado in Denver, which means acceptable low dose can be accomplished at a fraction of the dose for a standard chest CT. However, further analysis of organ doses showed that effective dose to female participants was 30 to 50 percent higher than males due to the sensitivity of breast tissue to radiation.
False positives and followup
NLST produced a fairly high false positive rate, reported David S. Gierada, MD, from Mallinckrodt Institute of Radiology at Washington University in St. Louis, Mo. In fact, the overall positivity rate for exams with findings suspicious for lung cancer or nodule size threshold of 4 mm or more reached 24.2 percent at three screens. The rate is higher than other screening studies including mammography, flexible sigmoidoscopy and prostate specific antigen, stated Gierada.
Researchers handled positive screens with a staggered protocol. They recommended that participants with four to 10 mm CT-detected nodules undergo a follow-up low-dose CT within three to six months. Those with nodules larger than 10 mm or growing nodules were referred for more immediate, aggressive evaluation, shared Gierada. Finally, patients with positive x-ray screen were referred for follow-up CT or chest x-ray.
Gierada admitted that there may be room for improvement in some areas related to screening CT. Specifically, reader variability had a modest impact in previous chest CT screening studies. He suggested lesion specificity software or CAD might offer a remedy. Finally, “Increased awareness of the implications of a positive screen is needed.”
The large clinical trial will continue to yield new findings for years. NLST researchers gathered an array of data including quality of life, smoking behavior, healthcare utilization, cost and cost-effectiveness, shared Constantine A. Gatsonis, PhD, professional of medical science at Brown University in Providence, R.I.
Researchers collected blood, urine and sputum samples annually over three years to develop a biospecimen bank for the study of biomarkers shown to be promising in preliminary testing. The bank also includes pathology specimens of resected tissue.
The research team expects to begin manuscript submission early in 2011, and will submit the primary paper in the first quarter, followed by T0, T1 and T2 findings in the second quarter.
Although the trial produced profound results, its impact on clinical practice is less certain. “How we generalize these results to the greater population of individuals who may be at risk and from academic institutions participants is a little less certain [than the actual mortality reduction],” Aberle explained. That’s because numerous questions remain such as the exact definition of the high-risk population and appropriate follow-up protocols. “It behooves us to remember that the greatest way to prevent lung cancer mortality is through the elimination of tobacco use. Any lung screening program should be couched within a broader program that includes smoking cessation,” summed Aberle.