Formidable Challenges Ahead for Lung Cancer Screening

Publication
Article
OncologyONCOLOGY Vol 26 No 2
Volume 26
Issue 2

Tanner et al provide a concise review of lung cancer screening, including discussion of past failed attempts, the success of the National Lung Screening Trial (NLST), and promising new avenues for improving on the NLST results.

Tanner et al provide a concise review of lung cancer screening, including discussion of past failed attempts, the success of the National Lung Screening Trial (NLST), and promising new avenues for improving on the NLST results. The major question currently facing the lung cancer community is how to integrate low-dose helical computerized tomography (LDCT) into health-care delivery while continuing to emphasize the over-arching goal of smoking cessation. The population of current and former smokers in the United States is in excess of 90 million. The number of individuals that fit the high-risk criteria used for entry into the NLST is approximately 7 million.[1] Tanner et al currently recommend screening this group only. However, the majority of lung cancers in the United States arise in the lower-risk cohorts, and if screening were to be limited to the NLST high-risk group, only a small percentage of the potentially achievable reduction in lung cancer mortality would be attained. The NCCN guidelines provide an initial step in expanding screening outside the NLST criteria by considering recommending screening for persons aged ≥ 50 with ≥ 20 pack-years of smoking who have additional risk elements, such as occupational exposure to arsenic or a family history of lung cancer.[2] However, the NCCN did not provide an absolute benefit to be derived from screening in the lower-risk groups in which this intervention was recommended-which might then be balanced against estimated risks.

Tanner et al mention one risk model with promise, but a limitation of that model is the need for bronchoscopy samples, with the associated acquisition and processing costs.[3] Also, how does one choose whom to test? Alternative risk models have been developed using epidemiologic data obtainable from easily administered short questionnaires. A recently developed model from Tammemagi et al is based on information obtained from the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial (PLCO) cohort.[4] A substantial improvement on this PLCO cohort model, utilizing pulmonary function test data, has also been developed by Tammemagi et al in cohorts entered into lung cancer chemoprevention trials at one institution.[5] Sputum DNA image cytometry only provided slight additional improvement on the predictive model. These types of risk models could be utilized in further deciding who should be offered LDCT.

One of the Cancer Intervention and Surveillance Modeling Network (CISNET) modeling groups, led by McMahon et al, is mentioned in the Tanner review.[6] Their approach is one of four that are now utilizing NLST data to further improve the reliability of their models and ensure that they replicate the NLST mortality reduction.[1] Extrapolation of benefits and risks to other categories can then be done with greater precision. Eventually, the information from the combined efforts of all the lung CISNET modeling groups will be utilized by the United States Preventive Services Task Force (USPSTF) as one factor in the planned review of their lung cancer screening recommendations. Frequently, the Centers for Medicare and Medicaid Services take the recommendations of the USPSTF into account when considering coverage decisions.

A factor that is critical when considering decisions about whom to screen is the cost-effectiveness of LDCT, particularly compared with the costs of other interventions, such as smoking cessation, as McMahon et al have so well delineated.[6] NLST investigators are doing a detailed cost-effectiveness analysis with data directly from the NLST. Strengths of this analysis include the detailed information collected by the American College of Radiology Imaging Network (ACRIN) portion of the NLST on healthcare utilization associated with scans that revealed substantial abnormalities suggestive of problems other than lung cancer, and the random sampling of data from individuals who had negative studies. NLST investigators also are analyzing the effect of screening on smoking behaviors.[1] Tanner et al mention one study that shows a randomized screening intervention had no impact on smoking cessation, but other approaches within randomized screening trials have shown the possibility of increasing quit rates.[7]

The application of LDCT in practice also provides formidable challenges. Tanner et al discuss one approach to dealing with high false-positive rates that is being assessed in the NELSON trial (Dutch Belgian randomized lung cancer screening trial)-that of volume averaging over serial scans.[8] Another problem is the great variability among radiologists and even in the work of a single radiologist with regard to the detection and assessment of nodules. The Lung Screening Study Reader Variability Study reported that the multi-rater Κ measure was 0.64 (95% confidence interval, 0.62–0.65) for agreement on classification as a positive or negative screening result.[9] A potential approach to improvement in this area is the development of computer-aided detection (CAD) and diagnosis algorithms. The Lung Image Database Consortium (LIDC) and eventually NLST images can serve as a resource for the research community for this work.[10] CAD is commercially viable and already utilized for second reads for mammography.

As lung cancer screening programs are started, quality assurance and quality control programs to improve performance are critical. There are many lessons to be learned from what the breast cancer and other screening communities, such as the cervical cancer community, have done. One model is the Breast Cancer Surveillance Consortium.[11] The research of the Consortium has demonstrated process improvements such as improved image quality; achievement of lower radiation doses; and enhanced monitoring of call-backs, false-negatives, and interval cancer. The Mammography Quality Standards Act administered through the Food and Drug Administration has also contributed to improvements.[12]

The promise of light-induced fluorescence endoscopy (LIFE) bronchoscopy is discussed by Tanner et al. A prospective cohort study of LDCT screening in which half of the group subsequently receives LIFE bronchoscopy is being conducted by Stephen Lam et al in Canada through a consortium.[13] This will provide some assessment of added benefit.

A new era has opened with the landmark, positive results of the NLST. The challenge for the lung cancer community is to implement these findings wisely and well, and in those most likely to benefit with the least risk. We must, however, never lose sight of the major objective of halting cigarette smoking and bringing the tragic lung cancer epidemic to an end.

Financial Disclosure: The author has no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.

References:

REFERENCES

1. Aberle DR, Adams AM, Berg CD, et al. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011;365:395-409.

2. NCCN guidelines: Lung Cancer Screening. 2011. Available from: www.nccn.org. Accessed January 20, 2012.

3. Beane J, Sebastiani P, Whitfield TH, et al. A prediction model for lung cancer diagnosis that integrates genomic and clinical features. Cancer Prev Res (Phila). 2008;1:56-64.

4. Tammemagi CM, Pinsky PF, Caporaso NE, et al. Lung cancer risk prediction: Prostate, Lung, Colorectal And Ovarian Cancer Screening Trial models and validation. J Natl Cancer Inst. 2011;103:1058-68. [Epub 2011 May 23]

5. Tammemagi MC, Lam SC, McWilliams AM, Sin DD. Incremental value of pulmonary function and sputum DNA image cytometry in lung cancer risk prediction. Cancer Prev Res (Phila). 2011;4:552-61. [Epub 2011 Mar 16]

6. McMahon PM, Kong CY, Bouzan C, et al. Cost-effectiveness of computed tomography screening for lung cancer in the United States. J Thorac Oncol. 2011;6:1841-8.

7. Taylor KL, Cox LS, Zincke N, et al. Lung cancer screening as a teachable moment for smoking cessation. Lung Cancer. 2007;56:125-34. [Epub 2006 Dec 28]

8. van Klaveren RJ, Oudkerk M, Prokop M, et al. Management of lung nodules detected by volume CT scanning. N Engl J Med. 2009;361:2221-9.

9. Gierada DS, Pilgram TK, Ford M, et al. Lung cancer: interobserver agreement on interpretation of pulmonary findings at low-dose CT screening. Radiology. 2008;246:265-72. [Epub 2007 Nov 16]

10. Armato SG 3rd, McLennan G, Bidaut L, et al. The Lung Image Database Consortium (LIDC) and Image Database Resource Initiative (IDRI): a completed reference database of lung nodules on CT scans. Med Phys. 2011;38:915-31.

11. Breast Cancer Surveillance Consortium. Available from: http://breastscreening.cancer.gov/. Accessed January 17, 2012.

12. Mammography Quality Standards Act and Program. Available from: http://www.fda.gov/Radiation-EmittingProductsMammographyQualityStandardsActandProgram/default.htm. Accessed January 17, 2012.

13. Lam S, Tsao MS, Tammemagi M, et al. The pan-Canadian early detection of lung cancer study. [Abstract]. J Thorac Oncol. 2009;4:S377–8.

Recent Videos
Video 4 - "Frontline Treatment for EGFR-Mutated Lung Cancer"
Video 3 - "NGS Testing Challenges and Considerations in NSCLC"
Related Content