ABSTRACT: Lung cancer is the leading cause of cancer-related death worldwide. At the time of initial presentation, most patients are at an advanced stage of disease with an associated poor prognosis. Those diagnosed and treated at earlier stages have a significantly better outcome, with 5-year survival for stage I disease approaching 75%. Ideally, a screening strategy for lung cancer would detect disease at an earlier stage and allow for potential surgical cure in those with cancer and avoid unnecessary morbidity in those without. Until recently, there had not been a screening test that demonstrated a mortality reduction in lung cancer. The National Lung Cancer Screening Trial showed that screening high-risk persons with low-dose CT scanning could significantly reduce lung cancer mortality; however, there are difficulties in generalizing these results to the community. Several new testing techniques, including measurement of volatile organic compounds in exhaled breath, an airway epithelial gene expression biomarker, and serum sampling for antibodies to tumor-associated antigens, are currently being evaluated and may prove useful as part of a screening algorithm for lung cancer.
Introduction
Lung cancer is responsible for the largest number of cancer-related deaths worldwide. In the United States in 2011, it was expected that there would be 221,130 new cases of lung cancer, accounting for 14% of all new cancer diagnoses. It was estimated that 156,940 deaths would occur from lung cancer in 2011, accounting for 27% of all cancer deaths.[1]
At the time of presentation, only 16% of p,atients are diagnosed with a disease stage that has potential for cure by surgical resection. The overall 5-year survivorship for all patients is 15%, despite therapy. However, the 5-year survival for those with pathologic stage I non–small-cell lung cancer (NSCLC) is 58% to 73% after surgery.[2] This significant difference in survivorship makes the identification of early-stage disease desirable. Until recently, however, there had not been a screening test that demonstrated a mortality reduction in lung cancer. In this review, we discuss past attempts at lung cancer screening, the results of the National Lung Cancer Screening Trial (NLST), and innovative tests for lung cancer screening currently being evaluated.
Lung Cancer Testing: The Past
CXR/sputum screening
The first trial using chest radiographs for lung cancer screening was published in 1968. Subjects were randomly assigned to receive a chest x-ray (CXR) every 6 months for 3 years or to receive a baseline and end-of-study CXR; there was no difference in mortality between the two groups.[3] The addition of sputum cytology screening was examined in four subsequent studies published in the 1980s that randomly assigned patients to annual CXR with or without annual sputum cytology; there was again no difference in mortality.[4-7]
Recently, the findings of the lung cancer arm of the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial were reported.[8] This randomized controlled trial enrolled 154,901 participants aged 55 through 74 years at 1 of 10 screening centers across the United States between 1993 and 2001; 77,445 were assigned to annual screening CXR for 4 years and 77,456 to usual care. Participants were followed for a total of 13 years or until December 2009, whichever occurred first. Cancers diagnosed were classified according to histology and as either screen-detected or non–screen-detected (interval or never-screened). Logistic regression analysis was performed to control for smoking status. Cancer stage and tumor histology were similar between the groups. At 13 years of follow up, 1213 lung cancer deaths were observed in the intervention group, compared with 1230 in the usual-care group (mortality relative risk [RR], 0.99; 95% confidence interval [CI], 0.87-1.22). These results clearly demonstrated that a CXR screening program does not improve lung cancer mortality.
Despite repeated trials failing to show a mortality benefit with CXR screening for lung cancer, a recent survey of primary care physicians showed that 67% recommended lung cancer screening, and more than 65% of respondents said that they would use CXR over other screening modalities.[9] A follow-up survey completed by 962 primary care physicians showed that 55% had ordered CXR for asymptomatic patients even though this is not a recommendation of any expert group.[10]
Low-dose CT screening
Low-dose CT screening (LDCT), which uses low levels of radiation to generate an image, is faster and less expensive than standard helical CT scanning and detects three times as many lung nodules as chest radiographs. Several single-arm observational cohort studies provided valuable information about the ability to recruit, retain, and screen a large population over multiple years using LDCT but could not provide data on the overall effectiveness of this screening modality.[11-13]
Further investigation with prospective trials demonstrated the potential of LDCT as an effective screening tool. The Mayo Clinic LDCT prospective trial involved annual screening in a high-risk population for 5 years. Lung cancer was identified in 4% of cases, 61% of which were stage I.[14] Similarly, Bach et al combined data from three single-arm studies using LDCT for lung cancer screening and found a detection rate of 4.4 %, with 74% of cancers being early-stage.[15] Even with such promising results, these studies by design could not ascertain whether detection of early-stage cancer was truly an indication of a good screening tool that reduced mortality. This has resulted in several randomized controlled trials designed to detect a mortality benefit.
The results of the North American National Lung Screening Trial (NLST) have most recently been released, and several other trials, including the NELSON trial (Dutch Belgian randomized lung cancer screening trial) assessing screening with LDCT, are currently underway in Europe.
National Lung Screening Trial
The NLST included 33 centers across the United States. Eligible participants were between ages 55 and 74 years at the time of randomization, had a history of at least 30 pack-years of cigarette smoking, and, if former smokers, had quit within the past 15 years.[16] A total of 53,454 persons were enrolled; 26,722 were randomly assigned to screening with LDCT and 26,732 to screening with CXR. Any noncalcified nodule found on LDCT measuring at least 4 mm in any direction and CXR images with any noncalcified nodule or mass were classified as positive. The LDCT-screened group had a substantially higher rate of positive screening tests compared with the radiography group (Round 1, 27.3% vs 9.2%; Round 2, 27.9% vs 6.2%; and Round 3, 16.8% vs 5.0%). Overall, 39.1% of participants in the LDCT group and 16.0% in the radiography group had at least one positive screening result. Of those who screened positive, the false-positive rate was 96.4% in the LDCT group and 94.5% in the radiography group. The false-positive rate was evenly distributed across all three rounds.
In the LDCT 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 the completion of the screening phase. In the radiography 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 the completion of the screening phase. A total of 356 deaths from lung cancer occurred in the LDCT group and 443 in the CXR group, with a relative reduction in the rate of death from lung cancer of 20.0% with LDCT screening. Overall mortality was reduced by 6.7%. The number needed to screen with LDCT to prevent 1 death from lung cancer was 320, which is comparable to the numbers in studies on screening mammography for breast cancer in women over 50. The National Lung Screening Trial is the first large-scale randomized trial to convincingly demonstrate a mortality benefit for LDCT lung cancer screening in high-risk individuals.
One of the concerns with using LDCT for lung cancer screening is the high rate of test-positive results necessitating further work-up. Investigators from the NELSON study demonstrated that this can be overcome by utilizing semi-automated volumetry software to measure diameter and volume doubling time (VDT). Growth was defined as a change in volume between the first and the second scan of 25% or greater. Nodules meeting growth criteria were then classified into three categories based on VDT (< 400, 400-600, and > 600 days). This approach to nodule management resulted in a decrease in the rate of test-positive results at baseline from 30% to 2%.[17] The final results regarding the reduction in mortality from lung cancer in this trial are pending.
Despite the impressive results from NLST in high-risk adults, generalizing them may be problematic for a number of reasons. Participants in the NLST were enrolled in urban, tertiary care hospitals with expertise in all aspects of cancer care. LDCT studies were interpreted by dedicated chest radiologists with expertise in characterizing nodules and providing appropriate recommendations for follow-up. As a result, few patients required further invasive testing, and radiographic follow-up was sufficient for many.
In contrast, community practice gives rise to the potential for considerable variation in the management of solitary pulmonary nodules identified by screening LDCT. One study demonstrated a two-fold variation among geographic regions in use of CT-guided biopsy, ranging from 14.7 to 36.2 per 100,000 adults. This substantial variation in the management of solitary pulmonary nodules may lead to an increased number of invasive procedures with risk of harm. For instance, complications from transthoracic biopsies include a 1% rate of bleeding (with one-third of affected patients requiring transfusion) and a 15% rate of pneumothorax. More than 6% of CT-guided biopsies result in a pneumothorax requiring chest tube drainage, a clinically important complication that results in pain, serial imaging with radiation exposure, and hospitalization.[18] Older patients and those with chronic obstructive pulmonary disease (COPD) have an increased risk of biopsy-related complications that can result in longer length of hospital stay, contributing both to economic cost and higher rates of respiratory failure that can affect long-term health.[18-20] Furthermore, the psychological harm of a positive test result should not be underestimated. This effect has been demonstrated in patients with false-positive results from both breast and prostate cancer screening. A study of women recalled for diagnostic assessment following a mammogram found that those with abnormal findings reported an increase in level of concern regarding cancer that was sustained even a month after a negative result had been reported.[21] Similarly, men with false-positive screening results had increased prostate cancer–related worry and problems with sexual function.[22,23]
Another difference between the results of the NLST and community practice is that the mortality from lung cancer surgery was 1% in the NLST, whereas the national average is between 3% and 5% for a lobectomy. While the NLST allowed patients to choose where they had their evaluation and management for screen-detected nodules, it is likely that many were managed at an NLST site with high volume and dedicated thoracic surgery support, both of which have been shown to have better outcomes.[24,25] It is also important to note that while 70 is the average age at lung cancer diagnosis, only 8% of the NLST study population was over the age of 70. Thus it is probable that patients enrolled in this study were younger and healthier than persons likely to participate in broad-based screening programs.
In addition to the difficulty associated with generalizing the study results to the community, difficulties arise from the fact that smokers represent a unique group with potential barriers to lung cancer screening. One study demonstrated that current smokers are less likely than nonsmokers to believe that early cancer detection would result in a good chance of survival. Current smokers are also less likely to consider CT screening for lung cancer (71.2%) than are never-smokers (87.6%). In addition, only half of the current smokers surveyed would opt for surgical resection of a screening-diagnosed cancer.[26] Finally, it is significant that smokers make up 31% of the population below the poverty line, compared with 20% of those at or above the poverty line[27]; as a result, smokers are likely to be more difficult to reach as a target population for large-scale screening in the community.
Currently, most of the major societies either have no recommendation or recommend against screening for lung cancer in a high-risk population.[15,28,29] Representatives of major societies in the cancer community are collectively evaluating cost-efficacy and assembling guidelines for appropriate screening and management of screen-detected abnormalities. The National Comprehensive Cancer Network (NCCN) most recently published its screening recommendations for lung cancer; these recommend screening with LDCT for those determined to be high risk. The NCCN guidelines rate as “category 1” the screening recommendation for high-risk patients with the same demographics as those enrolled in the NLST (age 55 to 74 years, with at least a 30–pack-year smoking history). They also place people who are 50 years of age or older and who have at least a 20–pack-year smoking history and an additional risk factor into the high-risk category, with a 2B recommendation for screening with LDCT.[30] However, there is no evidence to support this recommendation, and we believe it unwise to extend lung cancer screening to persons outside the inclusion criteria of the NLST.
