NLST data clearly demonstrate that lung cancer screening is effective and safe and reduces lung cancer-specific mortality by at least 20%. There is no possible reason for CMS to further delay or restrict lung cancer screening for those at high risk.
The completion of the National Lung Screening Trial (NLST), a randomized controlled trial (RCT) of lung cancer screening (LCS), in 2010 provided powerful RCT evidence of the efficacy and safety of computed tomography–based screening; nevertheless, the study had important limitations. Failure to understand these limitations has had substantial adverse effects. Misinterpretation or misrepresentation of the results has led to underestimation of benefits and overestimation of adverse effects. When factored into predictive models, inaccurate estimates have yielded falsely low projections of potential lives saved with national implementation of LCS, exaggerated projected costs, and underestimated cost-effectiveness. When extrapolated estimates were presented to guideline groups and payer panels by screening critics, results included delay in implementation of screening, recommendations to screen only a limited high-risk subgroup, and advice to restrict LCS to otherwise undefined “centers of excellence” able to enter data into a national registry. Finally, despite the formal endorsement of LCS by a large number of prestigious guideline groups, inaccurate extrapolation of NLST data has served to convince payer panels to recommend against insurance coverage for LCS. This article reviews limitations of the NLST study design and compares its results with screening data from many other RCTs and clinical programs, with the intention of providing more accurate and comprehensive information on the benefits, risks, costs, and cost-effectiveness of LCS.
Since the initial public announcement by National Cancer Institute (NCI) director Harold Varmus on November 4, 2010, that the National Lung Screening Trial (NLST) was complete,[1] results have been cited in literally hundreds of publications, guideline and payer deliberative sessions, and media outlets, typically simultaneously omitting reference to evidence from multiple other sources.[2] National implementation of lung cancer (LC) screening (LCS) has been repeatedly delayed as critics of screening have cited low efficacy and substantial adverse effects. This delay ended in December 2013, when the US Preventive Services Task Force (USPSTF) announced a level-B recommendation for computed tomography (CT) LCS in smokers and ex-smokers 55 to 80 years old, with 30-pack-year exposure, and who had quit less than 15 years earlier.[3-5] By federal law, the USPSTF’s ruling meant that private insurers (in most cases) had to provide first-dollar LCS coverage for their clients at risk. On April 30, 2014, the Centers for Medicare and Medicaid Services (CMS) Medicare Evidence Development & Coverage Advisory Committee (MEDCAC), however, recommended against Medicare and Medicaid coverage of LCS.
Unless CMS rejects this advice-and federal law allows it to do so-on January 1, 2015, persons with private healthcare insurance will have coverage for LCS while those with Medicare and Medicaid will not. The implications of this disparity are enormous, since almost 70% of lung cancer deaths occur in people over the age of 65.[6] Furthermore, cigarette smoking and lung cancer are disproportionately greater problems among underserved CMS populations, many of whom are living on fixed incomes, which is another barrier to screening.[7]
It deserves emphasis that many of the critiques of NLST study design and personnel described herein are not retrospective. Concerns regarding rumored design features of NLST were publically discussed at meetings of the International Early Lung Cancer Action Program (I-ELCAP) investigators in 2000.[8] NCI investigators were invited to attend this meeting to discuss and debate alternatives for the design of a large LCS trial, but they declined to do so. When the final NLST design was announced by NCI in 2003, prior concerns became actualized.[9] Paradoxically, over the intervening years there has been relatively little public criticism. Many feared that it would be injudicious to openly criticize NLST. With the passage of time, as anticipated, concerns became real, yet reluctance to openly criticize NLST persisted. Why?
A major concern was that NLST was the only randomized controlled trial (RCT) with sufficient statistical power to demonstrate a reduction in lung cancer–specific mortality (LCSM)-by 20% in the NLST design. Multiple European RCTs with lower accrual numbers (NELSON,[10] Danish,[11] DANTE,[12] etc) would likely each have insufficient power to demonstrate significant LCSM reduction. Thus, the specter arose that frank public discussion of NLST flaws might, should the study prove positive, allow policy and payer groups to view NLST as noncompelling evidence and insist on further delay in order to conduct more optimal studies. Such delay would represent a public health catastrophe since, inevitably, large numbers of unnecessary LC deaths could be anticipated in the interim.[13]
The LCSM and all-cause mortality reductions demonstrated by NLST, in conjunction with results from many other studies, have now convinced many national organizations to issue formal clinical practice guideline recommendations for LCS of high-risk persons, related to current or past smoking, multiple other risk factors, or both.[14-16] Suboptimal NLST results, which will be discussed below, coupled with inaccurate underestimates of benefit and followed by inaccurate predictions of population benefit and harm from modeling studies based on NLST results, have allowed critics of screening to recommend repeatedly that guideline groups and payers delay implementation of LCS and/or restrict its application. Specifically, such recommendations have proposed limiting insurance to cover only those individuals meeting NLST entry criteria, thus excluding many millions of others at high risk. The National Comprehensive Cancer Network (NCCN) lung cancer screening guidelines, for example, includes group 2 (heavy smokers younger than 55 and older than 74, those exposed to asbestos and other carcinogens, those with prior tobacco-caused cancers, those who quit more than 15 years earlier, and many others at high risk).[17] Current recommendations also include restriction of LCS to otherwise undefined “centers of excellence” and centers capable of entering data into computerized registries.[18] Various features of the NLST are discussed below, along with evidence from other research projects that suggest far higher benefits, as well as substantially lower risks from CT screening.
NLST limitations can be considered under three broad categories: study design, interpretation of data, and investigators.
At the root of the NLST’s limitations is the unproven assumption, based upon evidence-based medicine doctrine, that only an RCT can provide sufficient evidence to assess lung cancer mortality rate and to balance benefit against potential risk of LCS.[19,20] The decision to mandate an RCT for LC arose in large part from research sponsored by the Council for Tobacco Research, which put forth the puzzling and unsubstantiated notion that there is “a large reservoir” of nonlethal, “resting,” lanthanic, or overdiagnosed (OD) LC in the US population.[21-24] “As the trial [Mayo Lung Project] failed to show a mortality benefit in the screened arm, the investigators concluded that the extra tumors must have represented overdiagnosis.”[25] Incidence estimates from experts for OD LC have ranged from 14% to 50%. Patz et al estimate NLST OD at 21%[26,27]; Reich suggests a substantially higher estimate, while Bach et al decry “a 10-fold increase in lung cancer surgeries resulting from screening.”[28,29]
A number of investigators from I-ELCAP and other screening trials questioned whether doctors would recommend participation in a control group to patients at high risk for LC, based on widely divergent rates of stage I diagnosis (16% in Surveillance, Epidemiology and End Results [SEER] vs 80% in the Early Lung Cancer Action Project [ELCAP]).[30] This concern was real. More than 900 NLST participants died of LC, in both the control and LCS arms (the latter died following the last CT screen). Assuming 60% to 80% survival if all had annual CT screening, the cost in lives lost in order to complete NLST is enormous.[31]
NCI’s decision to perform an RCT had other negative aspects associated with RCTs-enormous expense, research bureaucracy, and compromises that are almost always necessary to carry out large trials on fixed budgets. And the NLST would have to be large. Statistical power calculations indicated that in order to demonstrate a 20% reduction in LCSM, 50,000 subjects would be required in intervention and control arms. NLST’s enormous size in turn engendered another major problem: cost. Because of NLST’s $250 million price tag, other types of LCS studies would not be funded.
Another practical problem arose from the prospect of trying to accrue 25,000 subjects into an unscreened control arm. RCT control arms typically receive current standard clinical practice, which in this case meant no screening. NCI, based upon its interpretation of Mayo Clinic, Johns Hopkins, and Memorial Sloan Kettering LCS studies of the 1980s, had repeatedly advised the public that LCS with chest roentgenogram (CXR) and/or cytology was ineffective and possibly dangerous. Regardless, NLST design offered CXRs to control group participants. Why? It is not possible to answer this question definitively without review of NLST research meeting minutes, but investigators with experience in screening trials suggest that the decision was likely related to the known difficulty recruiting research participants into control arms that offer no intervention. Significant crossover to the intervention arm could be anticipated among those randomized to observation only, as in the Mayo Lung Trial and some European RCTs.[32-36] The stated justification for use of an “ineffective” CXR control arm was that results of the Prostate, Lung, Colon, Ovary (PLCO) study were-finally, after 20 years-expected to become available soon. If PLCO showed a significant benefit from CXR screening (it did not), then the necessity for another RCT would be averted.[37,38]
The next and perhaps the most important limitation in NLST design was the decision (as in the New York Health Insurance Plan Study) to provide only three rounds of screening (again, likely because of cost constraints). The medical question is not whether 3 years of cancer screening is effective, but whether annual screening saves lives. A trial of LCS lasting only 3 years is problematic. The Canadian Breast Cancer Screening Trial was widely criticized for its 5-year screening period.[39] Screening for only a brief period may miss a trial’s targeted endpoint despite effective screening. Miettinen has illustrated how few deaths are anticipated during the first years of a cancer screening study of otherwise healthy, asymptomatic individuals.[40] Although LC often causes death within 1 year of symptomatic presentation, it seldom does so following early-stage diagnosis in asymptomatic patients. In NLST, after the third annual CT scan, 2 years after the first screen, screening stopped. Any benefit in early diagnosis associated with screening should diminish progressively over time. In later years of NLST, new LC-developing in now-unscreened persons-presented symptomatically in an advanced stage, as in the unscreened US population (80% in SEER), with few survivors.
If a screening method is effective, cancer deaths are anticipated to be uncommon in both study arms during the first year or two of an RCT. Subsequently, in a high-risk population, more cancers will be diagnosed in both arms, and detection will be earlier in more patients and in smaller-size, earlier-stage tumors in the screened vs the control group. Over time, LC deaths increase in the unscreened and, to a lesser degree, in the screened population. Accordingly, after the first year or two, LCSM curves should begin to separate, diverging more widely over time. After screening ceases (2 years in NLST), following a lag, LC deaths can be expected to accelerate over time in the (now-unscreened) “screened” arm, with the slope of the LCSM curve trending upward, narrowing the difference in LCSM and ultimately approaching the slope of the control arm. Unless LCSM is measured during the window of time before this upswing, a reduction in mortality will not be appreciated; the study will have false-negative results. It is also critically important to understand that late deaths following cessation of LCS were included in calculation of LCSM in NLST, falsely lowering benefit calculations even more.
A final consequence of this design is that, because the study stopped immediately when the 20% LCSM endpoint was realized (as ethical considerations dictate), the ultimate (maximum) mortality reduction from sustained annual screening cannot be ascertained.[41,42] As succinctly stated by Yankelevitz, the NLST 20% LCSM endpoint represents “a floor, not a ceiling.”[43]
Estimation of potential harm. Critics of LCS express concern that not only would mortality reduction be insignificant in NLST, but that many of those screened would be harmed by their participation, eg, by a lessening of resolve to stop smoking, anxiety, discomfort, and expense. Physical injury and even death might occur as a consequence of diagnostic tests, unnecessary invasive procedures, and operations in patients with benign nodules or OD LC.[44] Further, Bach, for example, has stated that “essentially no one believes the current data support the hypothesis that screening is beneficial,”[45] and approximately 10 times as many thoracotomies were performed in one lung cancer screening series as would be expected in a population.[29] Also, some studies suggest that radiation exposure from CT scans might cause a large number of cancers.[46,47]
Several NLST design features had an adverse impact on potential harms. First was the decision to use a nodule size of 4 mm as the cutpoint for a “positive” CT scan. Second, the same nodule from the first (baseline) screening was to be counted as another “positive” when it was seen again on the second (first annual) screening. The most serious flaw was that the study incorporated no diagnosis/treatment algorithm. Instead, patients in whom nodules were detected would be managed by their personal physicians using standard management methods for pulmonary nodules. Numerous other study limitations lie beyond the scope of this review.
NLST: suboptimal results. NLST design features, as anticipated, resulted in suboptimal results. NLST design ignored prior data from ELCAP and I-ELCAP indicating that a nodule mean diameter of 5 mm was an effective cutpoint, limiting the number of those screened who would require further diagnostic studies, without delay in diagnosis and treatment of LC.[48] NLST’s 4-mm cutpoint led to a substantially increased percentage of “false-positive” study results, without enhancing earlier diagnosis. Unlike ELCAP statistical methods, which did not re-count the same nodule detected during the prior round of screening as another positive during the second round of screening (such nodules were already under continuing follow-up), double-counting the same nodules misleadingly redoubled the number of putative “false positives.” This method yields an inaccurate impression of the number of study subjects requiring further testing and of the expense and potential harms arising therefrom.
No guidelines for diagnosis and treatment. Another important limitation predictably arose from failure to incorporate a diagnostic/treatment algorithm to guide investigators and personal physicians in the management of lung nodules detected by LCS.[49] By 1999, ELCAP had analyzed a large volume of data on CT screening and derived and published a set of clinical practice guidelines designed to minimize risks of false-positive screens and invasive diagnostic tests and surgeries in patients with benign nodules.[50]
This guideline has been regularly updated to incorporate new knowledge from continuing research in the ELCAP, NY-ELCAP, and I-ELCAP study groups, as well as information gathered from Japanese LCS trials and clinical practice. The I-ELCAP algorithm has been incorporated into NCCN and European RCT guidelines. Instead of using this information, the NLST elected to follow CT detection of a nodule with “standard clinical practice,” eg, the Fleischner Society or similar recommendations for the workup and treatment of pulmonary nodules, as had been the practice in the PLCO trial.[51]
The consequences of the absence of a diagnostic/treatment algorithm are reflected in NLST results.[52] For example, although the diagnostic yield from standard bronchoscopic methods in the workup of small peripheral nodules is very low, large numbers of diagnostic bronchoscopic examinations were performed in NLST. Participants in I-ELCAP underwent fewer such futile examinations. Alternatively, few transthoracic needle biopsies were performed in NLST, and a substantially higher percentage of patients had benign nodules removed surgically. In ELCAP and I-ELCAP, transthoracic needle biopsies were performed frequently, and the number of surgical resections of benign nodules was substantially lower than in NLST.[53]
Another adverse factor in the rigid design of RCTs is that patients do not benefit from rapidly developing advances in diagnosis and treatment that drive improvements in clinical practice. Technology and protocols do not change dramatically during the course of an RCT because investigators must remain blinded to the relative results in the treatment groups at all times before completion of the trial.[54]
In contradistinction, I-ELCAP results were reviewed annually by the assembled principal investigators. Improvements based on emerging new data were incorporated to enhance the study protocol. This is exemplified by the improved early-stage diagnosis and survival in Chinese patients following implementation of the I-ELCAP algorithm[55] and by the anticipated lower false-positive rates when a 6-mm nodule cutpoint is used.[56]
Out-of-date equipment. Many of the CT scanners used at NLST sites were obsolete and suboptimal (single-slice scanners), which limited detection of small nodules and accurate measurement of nodule growth on serial scans.[57] This limitation became increasingly important as diameter and volume growth during short intervals in small nodules emerged as a striking feature of effective LC screening with modern multislice CT scanners.[58]
Greatly compounding problems with NLST are the misinterpretations of NLST data and the puzzling failure of NLST investigators to correct them. Of greatest import are the misrepresentations of LCSM reduction. NCI director Harold Varmus was very clear in announcing that NLST had been stopped when the study endpoint of 20% LCSM reduction was attained, and that it would be unethical not to inform participants in the control group.[59] He was also very clear in explaining that, with only three rounds of screening, it would be impossible to accurately determine maximum LCSM reduction.
Despite such admonitions, Bach and others have repeatedly asserted that only 20% of patients with CT-detected lung cancer are cured: that “4 out of 5 of the lung cancers ‘snuck through’ (in laymen’s terms) and were incurable despite routine screening.”[29,60] This egregiously low estimate is a direct consequence of terminating CT screening after three scans, of counting LC deaths in patients in whom CT screening had ceased as attributable to LCS, and of the lack of a diagnostic/treatment algorithm. As stressed by Varmus, progressive future reductions in LCSM could not be measured. Finally, NLST survival data indicate that 62% of patients (3 in 5) with LC diagnosed by screening survived at 5 years.
The 62% NLST survival statistic, however, is also an underestimate.[61] I-ELCAP investigators attribute their better survival (82% actuarial 10-year survival) to use of the I-ELCAP screening regimen-whereas, again, NLST failed to use a protocol. Although Bach has repeatedly questioned the accuracy of I-ELCAP survival data, characterizing them as an “outlier,”[62,63] nearly identical survival has been attained or exceeded in screening programs from Canada, Japan, and China.[53,55,64,65] Utilization of data from other prospective trials in modeling studies suggests much greater mortality reduction.[41] Mortality reduction estimates from Japanese screening programs are also far higher than Bach’s estimate.[66]
To date, I am unaware of any NLST investigator coming forward to clarify misapprehensions or to defend the study results. This is perhaps understandable, given that several of them have striking and clearly documented track records of opposition to cancer screening in general and to LCS in particular.[31] Two NLST investigators (Aberle and Black) have provided adverse testimony on LCS for lawyers representing Philip Morris in medical monitoring lawsuits.[67,68] Philip Goodman, research partner of NLST investigator Edward Patz, has provided such testimony twice.[69,70] Other NLST principal investigators have produced books, articles, editorials, and media comments critical of cancer screening in general and LCS in particular-specifically Miller,[71] Brawley,[72] Black,[73] Patz,[74] and Kramer.[75] While not formally associated with the NLST, Peter Bach has assumed a role as its unofficial spokesman, as part of his more general role as a long-term opponent of population LCS.
None of the predicted adverse consequences of LCS described earlier have been documented in NLST publications. There is no evidence of a “license” to continue or resume smoking.[76,77] Instead, we may have a “teachable moment.”[78] Anxiety among study subjects was minor.[79,80] Although Brawley has repeatedly asserted that 16 deaths occurred within 60 days of detection of a nodule,[81] there is no published report that even a single death was caused by an intervention in the NLST.[82]
Although NLST had a relatively higher incidence of “false positives” followed by further radiation exposure, diagnostic tests, and even operations among those with benign nodules, other research has clearly shown that, when an effective diagnostic algorithm is used, repeat CT scans, biopsies, and resections of benign nodules are minimized.[83] Furthermore, when a new cutpoint of 6 mm for a positive scan is used, false positives further decrease without delay in diagnosis.[84]
When factored into predictive models, inaccurate estimates have yielded falsely low projections of potential lives saved with national implementation of LCS, exaggerated projected costs, and underestimated cost-effectiveness.[85,86] The limitations of NLST data are multiplied when suboptimal results are plugged into modeling equations.[87,88] For example, when the 20% survival estimate is entered into modeling equations, low estimates of lives saved and cost-effectiveness are generated. Incorporation of more accurate 60% to 80% survival estimates yields estimates of lives saved and cost-effectiveness that are three and four times higher, which is an enormous difference.[89] In a recent ASCO abstract, Roth et al estimated that implementation of LCS according to the USPSTF recommendations would cost Medicare $9.3 billion over a 5-year period.[90] This estimate is substantially higher than that predicted using accurate input data. For example, cost per screening CT scan drops sharply by incorporating the CMS physician fee schedule cost ($178.04 per CT scan rather than Roth’s $500 estimate).[91] Furthermore, estimates of downstream workup costs after “false-positive” findings are markedly lower when the cutpoint for a positive screen is updated to a 6-mm nodule (I-ELCAP, NCCN, and American College of Radiology) rather than 4 mm (NLST).[92] Incorporation of cost data reflecting this more accurate input results in a much lower Medicare cost.[91]
The first consequence of failure to appreciate NLST shortcomings has been delay. Very few individuals at high risk for developing LC were screened during the 7-year study period or in the 4 years since the Varmus announcement. An opportunity to save many lives has been irretrievably lost.
More recently, presentation of inaccurate data to payer groups has directly resulted in adverse recommendations that threaten to further delay and restrict provision of high-quality LCS to those at risk. Specifically, based on emphasis on the relatively poor results of the NLST-without mention of its limitations and of far better results from other LCS RCTs and clinical programs-there have been repeated efforts to delay population implementation of LCS, to narrow the inclusion criteria for participation, and to further restrict screening by restricting the number of programs eligible for federal reimbursement.[93]
Two recent egregious examples are the decision of Blue Cross of California and the California Technology Assessment Task Force (CTAF) and the recommendation of the CMS MEDCAC panel. These two groups advocated against provision of coverage of CT LCS for high-risk individuals by Blue Cross and Medicare/Medicaid health insurance plans.
The California Technology Assessment Task Force assessment forum
In October 2012, Blue Cross convened a panel in South San Francisco to consider a positive recommendation for LCS from the CTAF. Dr. Peter Bach, who was chosen to present the evidence to the panel, limited his presentation to his interpretation of NLST results, with minimal reference to evidence from other screening studies discussed above. Despite rebuttal from a number of participants, Blue Cross and CTAF recommended against LCS coverage.[94]
In April 2014, MEDCAC convened a panel in Maryland to consider coverage of LCS by Medicare and Medicaid. Dr. Rita Redberg, another prominent screening critic, chaired the meeting, and Bach again gave a featured presentation. I have described my impressions of this meeting elsewhere.[95] Although Bach no longer authoritatively asserted that no lives are saved by LCS,[29] he again misinterpreted NLST results. Other participants again informed MEDCAC of results from multiple other sources, but once more to no avail. MEDCAC advised CMS not to cover those at high risk, despite the positive recommendation of the USPSTF. Bach and others suggested that, if coverage was to be provided, it should be limited to those meeting NLST entry criteria and restricted to otherwise-undefined “centers of excellence” that would enter screening data into a national database.
If the CMS heeds MEDCAC advice, it will lead to an anomalous and dangerous disparity in our society, in which individuals at high risk for LC who have private insurance can be screened without cost, while the underprivileged and those with fixed incomes, in whom screening would be a financial burden, would be virtually excluded. The adverse consequences to public health would be major.
Recently published information allows us to quantitate the harm that would result from failure to provide LCS coverage for Medicare patients. As an example, Ulyott summed up the CTAF/Blue Cross conclusion as saying that no more than a “few thousand” LC deaths would be prevented by private insurance coverage of LCS.[94] In contrast, Pyenson et al updated estimates and concluded that implementation of LCS according to USPSTF recommendations in the Medicare population could potentially prevent approximately 400,000 LC deaths-even if only half opted to be screened.[91] These two estimates vary by more than 100-fold! Pyenson et al estimate that this unprecedented benefit can be realized at a cost of less than $1/member/month: an increase in the total annual Medicare expenditure of approximately 0.3%. Although the MEDCAC panel expressed concern that the benefit of LCS in Medicare-age patients had not been determined, a very recent report documents no substantial difference in screening benefits among individuals above age 65 in NLST.[96] Furthermore, screening results in NCCN group 2 have been shown to be equivalent to those in group 1.[97] The modeling by McMahon et al also suggests deficiencies in screening strategy only by NLST entry criteria.[98]
With regard to the MEDCAC recommendation of participation in screening only to centers able to register data, on September 8, 2014, NCI deputy director Douglas Lowy stated unequivocally that NCI funding for such a registry was not available.[99]
NLST data clearly demonstrate that LCS is effective and safe and reduces LCSM by at least 20%. Data from a large number of other studies indicate that CT screening in the context of an effective diagnostic/treatment algorithm can prevent an enormous number of LC deaths safely and at a reasonable cost. There is no possible reason for CMS to further delay or restrict LCS for those at high risk. CMS must incorporate a recommendation for LCS in their Draft Coverage Determination due in November 2014.
Financial Disclosure: Dr. Grannis was a principal investigator in I-ELCAP 2001-7 and has received grant support (in 2001) and travel costs to attend semi-annual research meetings. He has testified against Philip Morris Corp. in three lung cancer screening medical monitoring lawsuits in the states of New York, Massachusetts, and California, between 2008 and 2011.
1. Lung cancer trial results show mortality benefit with low-dose CT: twenty percent fewer lung cancer deaths seen among those who were screened with low-dose spiral CT than with chest X-ray. Press release. NCI Office of Media Relations. Available from: http://www.cancer.gov/newscenter/newsfromnci/2010/NLSTresultsRelease. Accessed October 17, 2014.
2. National Lung Screening Trial Research Team, 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.
3. Screening for Lung Cancer: Clinical Summary of U.S. Preventive Services Task Force Recommendation. Release date: December 31, 2013. Available from: http://www.uspreventiveservicestaskforce.org/uspstf13/lungcan/lungcansumm.htm. Accessed October 17, 2014.
4. Humphrey LL, Deffebach M, Pappas M, et al. Screening for lung cancer with low-dose computed tomography: a systematic review to update the US Preventive Services Task Force recommendation. Ann Intern Med. 2013;159:411-20.
5. de Koning HJ, Meza R, Plevritis SK, et al. Benefits and harms of computed tomography lung cancer screening strategies: a comparative modeling study for the U.S. Preventive Services Task Force. Release date: December 31, 2013. Available from: http://www.uspreventiveservicestaskforce.org/uspstf13/lungcan/lungcanmodelstudy.htm. Accessed October 17, 2014.
6. American Cancer Society, surveillance research 2014. Estimated cancer deaths by sex and age (years), 2014. Available from: www.cancer.org/acs/groups/content/@research/documents/document/acspc-041775.pdf. Accessed October 17, 2014.
7. Centers for Disease Control and Prevention. Smoking & tobacco use: health disparities. CDC 24/7: saving lives: protecting people. Available from: http://www.cdc.gov/tobacco/basic_information/health_disparities/index.htm. Accessed October 17, 2014.
8. The 2nd International Conference on Screening for Lung Cancer; 2000 Feb 25-27; Weill Medical College of Cornell University; New York, NY. Available from: http://events.ielcap.org/conferences/2nd-international-conference-screening-lung-cancer. Accessed October 17, 2014.
9. National Lung Screening Trial Research Team; Aberle DR, Berg CD, Black WC. The National Lung Screening Trial: overview and study design. Radiology. 2011;258:243-53.
10. Horeweg N, van der Aalst CM, Thunnissen E, et al. Characteristics of lung cancers detected by computed tomography screening in the randomized NELSON trial. Am J Respir Crit Care Med. 2013;187:848-54.
11. Pedersen JH, Petersen RH, Hansen HJ. Lung cancer screening trials: Denmark and beyond. J Thorac Cardiovasc Surg. 2012;144:S7-8.
12. Nair A, Hansell DM. European and North American lung cancer screening experience and implications for pulmonary nodule management. Eur Radiol. 2011;21:2445-54.
13. Jemal A, Thomas A, Murray T, et al. Cancer statistics 2002. CA Cancer J Clin. 2002;52:23-47.
14. Wender R, Fontham ET, Barrera E Jr, et al. American Cancer Society lung cancer screening guidelines. CA Cancer J Clin. 2013; 63:107-17.
15. Jaklitsch MT, Jacobson FL, Austin JH, et al. The American Association for Thoracic Surgery guidelines for lung cancer screening using low-dose computed tomography scans for lung cancer survivors and other high-risk groups. J Thorac Cardiovasc Surg. 2012;144:33-8.
16. American Lung Association. Providing guidance on lung cancer screening to patients and physicians. April 23, 2012. Available from: http://www.lung.org/lung-disease/lung-cancer/lung-cancer-screening-guidelines/lung-cancer-screening.pdf. Accessed October 18, 2014.
17. Ettinger DS, Akerley W, Borghaei H, Chang AC; National Comprehensive Cancer Network. Non-small cell lung cancer, version 2.2013. J Natl Compr Canc Netw. 2013;11:645-53.
18. Bach P. Testimony on 2014 Apr 30, CMS MEDCAC lung cancer screening with low dose computed tomography. Morning session. Available from: http://www.youtube.com/watch?v=xlCaTHxleqM&index=3&list=PLaV7m2-zFKpjPEC0xkktuzL7kGm6jxwIP. Afternoon session. Available from: http://www.youtube.com/watch?v=qG3NAPVzkTk&index=2&list=PLaV7m2-zFKpjPEC0xkktuzL7kGm6jxwIP. Accessed October 17, 2014.
19. Strauss GM. Measuring effectiveness of lung cancer screening: from consensus to controversy and back. Chest. 1997;112(4 suppl):216S-28S.
20. Miettinen OS. Screening for lung cancer: Do we need randomized trials? Cancer. 2000;89(11 suppl):2449-52.
21. McFarlane MJ, Feinstein AR, Wells CK. Necropsy evidence of detection bias in the diagnosis of lung cancer. Arch Intern Med. 1986;146:1695-8.
22. McFarlane MJ, Feinstein AR, Wells CK. Clinical features of lung cancers discovered as a postmortem “surprise.” Chest. 1986;90:520-3.
23. McFarlane MJ, Feinstein AR, Wells CK, et al. The ‘epidemiologic necropsy’: unexpected detections, demographic selections, and changing rates of lung cancer. JAMA. 1987;258:331-8.
24. Patz EF Jr. Lung cancer screening, overdiagnosis bias, and reevaluation of the Mayo Lung Project. J Natl Cancer Inst. 2006;98:724-5.
25. Croswell JM, Ransohoff DF, Kramer BS. Principles of cancer screening: lessons from history and study design issues. Semin Oncol. 2010;37:202-15.
26. Patz EF Jr, Pinsky P, Gatsonis C, et al. Overdiagnosis in low-dose computed tomography screening for lung cancer. JAMA Intern Med. 2014;174:269-74. Erratum in: JAMA Intern Med. 2014;174:828.
27. Patz EF Jr, Pinsky P, Kramer BS. Estimating overdiagnosis in lung cancer screening-reply. JAMA Intern Med. 2014;174:1198-9.
28. Reich JM. A critical appraisal of over-diagnosis: estimates of its magnitude and implications for lung cancer screening. Thorax. 2008;63:377-83.
29. Bach PB, Jett JR, Pastorino U, et al. Computed tomography screening and lung cancer outcomes. JAMA. 2007;297:953-61. Erratum in: JAMA. 2007;298:518.
30. SEER stat fact sheets: lung and bronchus cancer: statistics at a glance. Available from: http://seer.cancer.gov/statfacts/html/lungb.html. Accessed October 17, 2014.
31. The National Lung Screening Trial Research Team. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med 2011; 365:395-409.
32. Fontana RS, Sanderson DR, Woolner LB, et al. Lung cancer screening: the Mayo program. J Occup Med. 1986;28:746-50.
33. Fontana RS, Sanderson DR, Woolner LB, et al. Screening for lung cancer. A critique of the Mayo Lung Project. Cancer. 199;67(4 Suppl):1155-64.
34. Flehinger BJ, Kimmel M, Polyak T, et al. Screening for lung cancer: the Mayo Lung Project revisited. Cancer. 1993;72:1573-80.
35. Saghir Z, Ashraf H, Dirksen A, et al. Contamination during 4 years of annual CT screening in the Danish Lung Cancer Screening Trial (DLCST). Lung Cancer. 2011;71:323-7.
36. Baecke E, de Koning HJ, Otto SJ, et al. Limited contamination in the Dutch-Belgian randomized lung cancer screening trial (NELSON). Lung Cancer. 2010;69:66-70.
37. Oken MM, Hocking WG, Kvale PA, et al. Screening by chest radiograph and lung cancer mortality: The prostate, lung, colorectal and ovarian (PLCO) randomized trial. JAMA. 2011;306:1865-73.
38. Hocking WG, Hu P, Oken MM, et al. Lung cancer screening in the randomized prostate, lung, colorectal, and ovarian (PLCO) cancer screening trial. J Natl Cancer Inst. 2010;102:722-31.
39. Kopans DB, Feig SA. The Canadian National Breast Screening Study: a critical review. Am J Radiol. 1993;161:755-60.
40. Miettinen O. Curability gain vs mortality reduction as measures of benefit of screening. 24th International Conference on Screening for Lung Cancer. 2011 Feb 25-26; Tempe, AZ.
41. Foy M, Yip R, Chen X, et al. Modeling the mortality reduction due to computed tomography screening for lung cancer. Cancer 2011;117:2703-8.
42. Henschke CI, Boffetta P, Gorlova O, et al Assessment of lung cancer mortality reduction from CT screening. Lung Cancer. 2011;71:328-32.
43. Barnes E. Lung cancer screening of seniors: Should CMS regulate? Aunt Minnie. June 24, 2014. Available from: http://www.auntminnie.com/index.aspx?sec=sup&sub=cto&pag=dis&ItemID=107797. Accessed October 27, 2014.
44. Marshall E. A bruising battle over lung scans. Science. 2008;20:600-3.
45. Grady D. Scientist at work: Claudia I. Henschke: When it comes to lung cancer, she doesn’t believe in waiting. The New York Times. October 31, 2006. Available from: http://www.nytimes.com/2006/10/31/health/31prof.html?pagewanted=all&_r=0. Accessed October 28, 2006.
46 de Gonzalez A, Mahesh M, Kim K-P, et al. Projected cancer risks from computed tomographic scans performed in the United States in 2007. Arch Intern Med. 2009;169:2071-7.
47. Smith-Bindman R, Kipson J, Marcus R, et al. Radiation dose associated with common computed tomography examinations and the associated lifetime attributable risk of cancer. Arch Intern Med. 2009;169:2078-86.
48. Henschke CI, Yankelevitz DF, Smith JP, et al. CT screening for lung cancer. Assessing a regimen’s diagnostic performance. Clin Imaging. 2004;28:317-21.
49. Henschke CI (principal investigator). International Early Lung Cancer Action Program: enrollment and screening protocol; 2014 Apr 1; New York, NY. Available from:
http://www.ielcap.org/sites/default/files/ielcap.pdf. Accessed October 18, 2014.
50. Henschke CI, Yankelevitz DF, Smith JP, Miettinen OS. Screening for lung cancer: the Early Lung Cancer Action approach. Lung Cancer. 2002;35:143-8.
51. MacMahon H, Austin JH, Gamsu G, et al. Guidelines for management of small pulmonary nodules detected on CT scans: a statement from the Fleischner Society. Radiology. 2005;237:395-400.
52. The National Lung Screening Trial Research Team. Results of initial low-dose computed tomographic screening for lung cancer. N Engl J Med 2013; 368:1980-91.
53. Roberts HC, Patsios D, Paul NS. Lung cancer screening with low-dose computed tomography: Canadian experience. Can Assoc Radiol J. 2007:58:225-35.
54. Oudkerk M, Heuvelmans MA. Screening for lung cancer by imaging: the Nelson study. JBR-BTR. 2013;96(3):163-6.
55. Liu X, Liang M, Wang Y, et al. The outcome differences of CT screening for lung cancer pre and post following an algorithm in Zhuhai, China. Lung Cancer. 2011;73:230-6.
56. Yip R, Henschke CI, Yankelevitz DF, Smith JP. CT screening for lung cancer: alternative definitions of positive test result based on the National Lung Screening Trial and International Early Lung Cancer Action Program databases. Radiology. 2014;132950. Epub 2014 Jun 19.
57. Cody DD, Kim H-Y, Cagnon CH, et al. Normalized CT dose index of the CT scanners used in the National Lung Screening Trial. AJR Am J Roentgenol. 2010;194:1539-46.
58. Henschke CI, Yankelevitz DF, Yip R, et al. Lung cancers diagnosed at annual CT screening: volume doubling times. Radiology. 2012;263:578-83.
59. Participant notification letter and DSMB letter. Available from: http://www.cancer.gov/clinicaltrials/noteworthy-trials/nlst. Accessed October 18, 2014.
60. Sox H, Bach P. Lung cancer screening guidelines updated. May 25, 2012. Available from: http://www.cancernetwork.com/lung-cancer/lung-cancer-screening-guidelines-updated. Accessed October 18, 2014.
61. Kates M, Swanson S, Wisnivesky JP. Survival following lobectomy and limited resection for the treatment of stage I non-small cell lung cancer ≤1 cm in size: a review of SEER data. Chest. 2011;139:491-6.
62. Knox R. Study questions early screening for lung cancer. National Public Radio. “Morning edition.” March 6, 2007 [audio file]. Available from: http://www.npr.org/player/v2/mediaPlayer.html?action=1&t=1&islist=false&id=7744559&m=7744612. Accessed October 18, 2014.
63. Press release. Memorial Sloan Kettering Cancer Center. Study shows no benefit for CT screening for lung cancer. Death rates for smokers are not improved despite early diagnosis. March 6, 2007. Available from: http://www.innovations-report.com/hrml/reports/studies/report-80336.html. Accessed October 28, 2014.
64. Sone S, Nakayama T, Honda T, et al. Long-term follow-up study of a population-based 1996-1998 mass screening programme for lung cancer using mobile low-dose spiral computed tomography. Lung Cancer. 2007;58:329-41.
65. Koike T, Yamato Y, Asamura H, et al. Japanese Joint Committee for Lung Cancer Registration. Improvements in surgical results for lung cancer from 1989 to 1999 in Japan. J Thorac Oncol. 2009;4:1364-9.
66. Grannis FW. What do cars, TVs, and lung cancer have in common? [blog] March 1, 2013. Available from: http://www.cancernetwork.com/blog/what-do-cars-tvs-and-lung-cancer-have-common. Accessed October 18, 2014.
67. Testimony of Denise R. Aberle [transcript]. Civil District Court, Parish of Orleans, State of Louisiana. Gloria Scott and Deania Jackson vs The American Tobacco Company, Inc et al. No. 96-8461; Division “I,” Section 14. June 18, 2003.
68. Declaration and Expert Report of William C. Black, MD. United States District Court Eastern District of New York. Marcia Caronia, et al vs Philip Morris USA Inc. Civil Action No. 06-0224 (CBA) (SMG). September 26, 2006.
69. Declaration and Expert Report of Philip C. Goodman, MD. United States District Court for the District of Massachusetts. Kathleen Donovan and Patricia Cawley vs Philip Morris USA Inc. Civil Action No. 06-12234 NG. December 11, 2007.
70. Deposition of Philip Goodman, MD (taken on behalf of the plaintiffs). United States District Court Eastern District of New York. Marcia Caronia, et al vs Philip Morris USA Inc. Civil Action No. 06-0224 (CBA) (SMG). Durham, North Carolina. January 30, 2007.
71. Miller AB. Screening for lung cancer with low-dose computed tomography. Oncologist. 2013;18:897-9.
72. Brawley OW. How we do harm: a doctor breaks ranks about being sick in America. New York: St. Martin’s Press; 2012.
73. Black WC, Baron JA. CT Screening for lung cancer: spiraling into confusion? JAMA. 2007; 995-7.
74. Patz EF Jr, Goodman PC. Low-dose spiral computed tomography screening for lung cancer: not ready for prime time. Am J Respir Crit Care Med. 2001;163:813-4.
75. Kramer BS, Croswell JM. Cancer screening: the clash of science and intuition. Annu Rev Med. 2009;60:125-37.
76. Townsend CO, Clark MM, Jett JR, et al. Relation between smoking cessation and receiving results from three annual spiral chest computed tomography scans for lung carcinoma screening. Cancer. 2005;103:2154-62.
77. Ashraf H, Saghir Z, Dirksen A, et al. Smoking habits in the randomised Danish Lung Cancer Screening Trial with low-dose CT: final results after a 5-year screening programme. Thorax. 2014;69:574-9.
78. van der Aalst C, van den Bergh KA, Willemsen MC, et al. Lung cancer screening and smoking abstinence: 2 year follow-up data from the Dutch-Belgian randomised controlled lung cancer screening trial. Thorax. 2010;65:600-5.
79. Gareen IF, Duan F, Greco EM, et al. Impact of lung cancer screening results on participant health-related quality of life and state anxiety in the National Lung Screening Trial. Cancer. 2014;120:3401-9.
80. van den Bergh KA, Essink-Bot ML, Borsboom GJ, et al. Short-term health-related quality of life consequences in a lung cancer CT screening trial (NELSON). Br J Cancer. 2010;102:27-34.
81. How we do harm: A webinar by SHARE with Dr. Otis Brawley. Slide 17. Available from: http://www.slideshare.net/bkling/how-we-do-harm-a-webinar-by-share-with-dr-otis-brawley. Accessed October 28, 2014.
82. Van’t Westeinde SC, Horeweg N, De Leyn P, et al. Complications following lung surgery in the Dutch-Belgian randomized lung cancer screening trial. Eur J Cardiothorac Surg. 2012;42:420-9.
83. Petersen RH, Hansen HJ, Dirksen A, Pedersen JH. Lung cancer screening and video-assisted thoracic surgery. J Thorac Oncol. 2012;7:1026-31.
84. McKee BJ, Regis SM, McKee AB, et al. Performance of ACR lung-RADS in a clinical CT lung screening program. J Am Coll Radiol. 2014;S1546-1440. Epub 2014 Aug 28.
85. Wisnivesky JP, Mushlin AI, Sicherman N, Henschke C. The cost-effectiveness of low-dose CT screening for lung cancer: preliminary results of baseline screening. Chest. 2003;124:614-21.
86. Chirikos TN, Hazelton T, Tockman M, Clark R. Screening for lung cancer with CT: a preliminary cost-effectiveness analysis. Chest. 2002;121:1507-14.
87. Ma J, Ward EM, Smith R, Jemal A. Annual number of lung cancer deaths potentially avertable by screening in the United States. Cancer. 2013;119:1381-5.
88. McMahon PM, Kong CY, Johnson BE, et al. The MGH-HMS Lung Cancer Policy Model: tobacco control versus screening. Risk Anal. 2012;32(suppl 1): S117-S24.
89. Goulart BH, Bensink ME, Mummy DG, Ramsey SD. Lung cancer screening with low-dose computed tomography: costs, national expenditures, and cost-effectiveness. J Natl Compr Canc Netw. 2012;10:267-75.
90. Roth JA, Sullivan SD, Ravelo A, et al. Low-dose computed tomography lung cancer screening in the Medicare program: projected clinical, resource, and budget impact. ASCO Annual Meeting 2014, abstract 6501. Available from: http://www.cancernetwork.com/asco-2014-lung-cancer/slide-show-lung-cancer-highlights-asco-2014. Accessed October 18, 2014.
91. Pyenson B, Henschke CI, Yip R, et al. Offering lung cancer screening to high-risk Medicare beneficiaries saves lives and is cost-effective: an actuarial analysis. Am Health Drug Benefits. 2014;7:272-82.
92. Henschke CI, Yip R, Yankelevitz DF, et al. Definition of a positive test result in computed tomography screening for lung cancer: a cohort study. Ann Intern Med. 2013;158:246-52.
93. Pinsky PF, Berg CD. Applying the National Lung Screening Trial eligibility criteria to the US population: What percent of the population and of incident lung cancers would be covered? J Med Screen. 2012;19:154-6.
94. Ulyott D. California Technology Assessment Forum: policy brief: screening for lung cancer using low dose computed tomography. Available from: http://www.ctaf.org/about-ctaf/news/2012/policy-brief-screening-lung-cancer-using-low-dose-computed-tomography. Accessed October 18, 2014.
95. Grannis FW Jr. CT lung screening meeting: a travesty of public health policy. May 8, 2014. Available from: http://www.auntminnie.com/index.aspx?sec=sup&sub=imc&pag=dis&ItemID=107339. Accessed October 18, 2014.
96. Pinsky PF, Gierada DS, Hocking W, et al. National Lung Screening Trial findings by age: Medicare-eligible versus under-65 population. Ann Intern Med. 2014 Sep 9. [Epub ahead of print]
97. McKee BJ, Hashim JA, French RJ, et al. Experience with a CT screening program for individuals at high risk for developing lung cancer. J Am Coll Radiol. Epub 2014 Aug 28.
98. McMahon PM, Meza R, Plevritis SK, et al. Comparing benefits from many possible computed tomography lung cancer screening programs: extrapolating from the National Lung Screening Trial using comparative modeling. PLoS One. 2014;9:e99978.
99. Lowy DR. NCI supported research and lung cancer. Presented at the 6th National Lung Cancer Alliance Survivor Summit; 2014 Sep 7-10; Washington, DC. Available from: http://www.lungcanceralliance.org/events/6th-national-lung-cancer-survivor-summit.html. Accessed October 18, 2014.