In early 2011 NIOSH posted a blog entry commenting on the preliminary results from the National Lung Screening Trial (NLST), funded by the National Cancer Institute, that found a 20.3% reduction in deaths from lung cancer among current or former heavy smokers who were screened with low-dose helical computed tomography (CT) versus those screened by chest X-ray (CXR). On June 29, 2011, the peer-reviewed primary results from the NLST were published online in the New England Journal of Medicine. This is an update to the previous blog.1
The National Lung Screening Trial
The National Lung Screening Trial (NLST)2 was a national randomized controlled trial launched by the National Cancer Institute (NCI) in 2002 to determine whether annual screening with low-dose helical CT would lead to earlier detection and reduced mortality from lung cancer relative to screening with CXR.
In this trial 53,454 participants at high risk for lung cancer—current and former heavy smokers of at least 30 pack-years between 55 and 74 years of age—were randomly assigned to receive low-dose helical CT or chest x-ray screenings once a year for three years, and then followed for 3.5 additional years with no further screening.
Participant compliance with screening was over 90%. During the screening phase of the trial, 39.1% of the participants in the low-dose helical CT group and 16% of the participants in the chest X-ray group had at least one positive screening result, i.e., a finding suggestive of lung cancer. Of the total number of screening tests in the three rounds, 24.2% of the low-dose CT tests and 6.9% of the X-ray tests were classified as positive. Upon follow-up, 96.4% of the “positive” low-dose CT tests and 94.5% of the “positive” X-ray tests turned out to be false positives, meaning that the positive finding did not prove to be lung cancer. Lung cancer was confirmed in 3.6% of the positive screenings in the low-dose CT group and in 5.5% of the positive screenings in the X-ray group.
The diagnostic evaluations mainly consisted of further imaging: diagnostic CT, PET-CT scans and chest radiography. Invasive procedures were performed far less frequently. The rate of at least one complication after a diagnostic evaluation for a positive screening test was less than 2% for either type of screening. Among those who did have complications, 16 participants screened with low-dose helical CT (10 of whom had lung cancer) and 10 X-ray participants (all with lung cancer) died within 60 days of a follow-up invasive diagnostic procedure. It is not known whether the complications from the diagnostic procedures caused these deaths.
Lung cancer incidence per 100,000 person-years was 645 (1,060 cancers) in the low-dose helical CT group and 572 (941 cancers) in the chest X-ray group. In the low-dose CT group, 649 cancers were diagnosed after a positive screening test, 44 after a negative screening test, and 367 among participants who either missed the screening or received the diagnosis after their trial screening phase was over. In the X-ray group, 279 cancers were diagnosed after a positive screening test, 137 after a negative screening test, and 525 among participants who either missed the screening or received the diagnosis after their trial screening phase was over. In both groups, the highest percentages of screen-detected lung cancers were in early stage, and adenocarcinoma and squamous cell carcinoma were the types of lung cancer most detected by screening. Small-cell lung cancers were not detected in early stages by either screening method.
Lung cancer mortality was 247 per 100,000 person-years in the low-dose helical CT group and 309 in the chest X-ray group, representing a reduction in lung cancer mortality in the low-dose helical CT group of 20.3%. All-cause mortality (deaths due to any cause, including lung cancer) was reduced by 6.7% for those participants who underwent low-dose helical CTs compared to those who received chest X-rays.
Key Points to Consider
Lung cancer mortality is high and better survival prognosis for early stage cases makes early detection an appealing public health strategy. For years studies have been conducted to find an effective screening method; the NLST is the first randomized trial to show a significant reduction in mortality from lung cancer with low-dose CT screening. Until this trial, no screening test has been shown to reduce the risk from dying from lung cancer. However, it is important to emphasize that the data derived from the NLST were obtained from a very specific population group—individuals at high risk for developing lung cancer due to present or past heavy smoking, aged 55 to 74, and do not necessarily apply to the general population or specific populations of workers. In addition, the participants were mostly screened at major academic medical centers, staffed by experienced, well-trained radiologists and thoracic oncologists. It is unclear if results could be reproduced at different settings.
Screening with CT scans is not risk-free. Radiation exposure from repeated CT scans is cumulative and can lead to illness, including cancer. While a “low-dose” method was used, this is relative to a full diagnostic helical CT scan (average radiation effective dose 7 mSv).3 The radiation dose for this “low-dose” method (1.5 mSv in the NLST) is about 15 times higher than a CXR (average effective dose 0.1 mSv).3 Because the harmful effect of radiation is a long-term phenomenon, any harm from exposure to radiation during the screenings could not be measured directly in NLST.
Any screening process generates suspicious findings that turn out not to be cancer in a large number of cases, producing significant anxiety, morbidity and expense. People who receive false-positive results may be subjected to unnecessary testing, including more radiation exposure, invasive diagnostic and surgical procedures; complications, and even death. In the NLST, 39.1% of those participating in low-dose CT screening had at least one positive result and 96.4% of these were false-positives and did not identify a lung cancer.
Overdiagnosis is another major concern when screening for cancer. This is because the screening process may be more likely to detect slow growing cancers or cancers that would not have become symptomatic and therefore never diagnosed. The implication for overdiagnosis is that the patients diagnosed with an indolent cancer may end up undergoing an invasive intervention that they would not otherwise need. Additional follow-up would be necessary to measure the extent and magnitude of overdiagnosis in the NLST.
Analyses on the cost-effectiveness and quality-of-life effects from the NLST are still to be released; however, the authors pointed out that 320 individuals had to be screened with CT to prevent one lung cancer death. According to the NLST researchers, cost-effectiveness analyses of low-dose CT screening including not only the screening examination itself but also the diagnostic follow-up and treatment must be rigorously analyzed before policy recommendations can be made. They advise policy makers to wait for more information before endorsing lung-cancer screening programs.
In the occupational setting, there are a number of agents associated with lung cancer. However, the excess risks for lung cancer associated with these occupational exposures vary depending on the actual exposures. Consideration of the use of any screening test in occupationally exposed groups requires a careful assessment of the risk of a given condition. The risk of lung cancer from a specific exposure will directly affect the likelihood that a positive screening test for lung cancer will actually be evidence that the cancer exists. In other words, high risk for lung cancer in the NSLT trial due to a long history of heavy smoking made it more likely that a “positive” finding on a low-dose CT scan was in fact a lung cancer. The benefit of screening for lung cancer with low-dose CT cannot be easily estimated for populations with risk profiles that are different from those of the NLST participants. For that matter, the NLST researchers are planning collaborations with the Cancer Intervention and Surveillance Modeling Network4 to investigate the potential effect of low-dose CT screening in a wide range of scenarios. Further studies using the NLST data for identification of biologic markers for lung cancer are likely to be forthcoming; such studies may help in decision-making concerning which groups may benefit from screening with low-dose helical CT, and assist in establishing recommendations in the occupational setting.
NIOSH will continue to consult with the NCI investigators, and work with stakeholders in labor, industry, and the occupational medicine clinical community to consider how these data related to cancer screening may impact workers potentially at risk for lung cancer because of occupational exposures. As noted by Dr. Harold Sox in an editorial accompanying the NLST report, its findings can be considered a landmark in the lung cancer screening research era. Now, the “focus will shift to informing the difficult patient-centered and policy decisions that are yet to come.”5
Dr. Tramma is a Senior Service Fellow/Medical Officer in the Surveillance Branch of the NIOSH Division of Respiratory Disease Studies.
Dr. Storey is Chief of the Surveillance Branch in the NIOSH Division of Respiratory Disease Studies.
Dr. Trout is Associate Director for Science in the NIOSH Division of Surveillance, Hazard Evaluations, and Field Studies.
Dr. Sweeney is Chief of the Surveillance Branch in the NIOSH Division of Surveillance, Hazard Evaluations, and Field Studies.
- NIOSH Science Blog. Helical CT Scans and Lung Cancer Screening.
- National Lung Screening Trial Research Team. Reduced Lung-Cancer Mortality with Low-Dose Computed Tomographic Screening. N Engl J Med. 2011 Jun 29. [Epub ahead of print]
- Mettler FA Jr, Huda W, Yoshizumi TT, Mahesh M. Effective doses in radiology and diagnostic nuclear medicine: A catalog. Radiology 248(1):254-263; 2008.
- Cancer Intervention and Surveillance Modeling Network (CISNET).
- Sox HC. Better evidence about screening for lung cancer. Editorial. N Engl J Med. 2011 Jun 29. [Epub ahead of print]