Steven E. Vogl, MD
Discussions of benefits and harms from screening of high-risk populations for lung cancer have missed the point. The National Lung Screening Trial (NLST) showed an early and statistically significant major benefit in all-cause mortality from computed tomography (CT) screening.1 Those referred for annual screening had 7% fewer deaths—a huge benefit for an annual test that costs $300 and takes about 30 seconds. The NLST likely underestimated the benefits of screening because it included some relatively low-risk patients and because it mandated only one prevalent and two incident annual screening CT scans. The risk of lung cancer generally increases with increasing age in heavy smokers, but screening in the NLST stopped after only 2 years. More years of screening would likely prevent more early deaths.
The NLST did not include protocols for treatment—a calcified nodule with a longest diameter ≥ 4 mm or a noncalcified nodule or mass was reported as “suspicious” to the subject’s physicians. Even with no evaluation and treatment protocol specified, the NLST showed a major benefit in all-cause mortality. In fact, low-dose CT scanning of heavy smokers is the only cancer screening test to show a decrease in all-cause mortality in a randomized, prospective clinical trial. This decrease began to appear in just 2 years—a reflection of the virulence of asymptomatic lung cancer. Before the NLST reported its results, no one had demonstrated the existence of a reduction in lung cancer deaths or an overall mortality benefit for CT screening for lung cancer. The magnitude of the latter is surprising.
“The biggest problem with lung cancer screening is that no one is getting it.”— Steven E. Vogl, MD
Tweet this quote
Guided by the principles of epidemiology research, the NLST investigators chose as their primary endpoint mortality from lung cancer, not death from any cause. This choice reflects a “real-world” decision among academic epidemiologists that studies looking at all-cause mortality would not detect effects unless the studies were extremely large (and expensive) and very long in accrual and follow-up. Those reluctant to support annual CT screening for lung cancer worry that a reduction in cause-specific mortality may not reflect a benefit to the entire screened population because of toxic effects from screening and treatment of the discovered nodules and deaths from other causes. This concern would be valid if a decrease in “lung cancer deaths” was all the NLST showed.
The major and early decrease in all-cause mortality addresses this concern completely. That is not to say that whatever was done in the NLST cannot be improved upon. The recent report of the Dutch-Belgian NELSON trial suggests that a screening protocol based on the volume of nodules detected, change in volume over time, and using central image review, reduces the percentage of “false-positive” studies and the number of unnecessary interventions. The NELSON report (presented at the 2018 World Lung Cancer Congress)2 confirms the prompt reduction in lung cancer deaths from CT screening, suggests that women may benefit more than men, and suggests that the final screening interval of 2.5 years in NELSON may have been too long.
The NELSON investigators observed a nonsignificant 3.3% reduction in all-cause mortality in a smoking population at lower risk than the NLST population (as low as 15 pack-years compared to at least 30 pack-years for the NLST). Since the risk of lung cancer and consequent death was lower in NELSON, more of the deaths would be from causes other than smoking-induced lung cancer, so the sensitivity of the trial was lower, especially since it was only about one-third the size of NLST.
Heavy current and former smokers are dying of lung cancer every year. It is immoral to delay screening them for lung cancer while we engage in a process of improving the screening, evaluation, and treatment protocols. The process of improvement is important but never-ending. There is no good reason to delay screening implementation as it proceeds.
Failure to Screen Is an Abomination
THE BIGGEST PROBLEM with lung cancer screening is that no one is getting it. This is not news anymore; using survey data, 2 American Cancer Society epidemiologists estimated that only 262,700 of 6,800,000 eligible smokers were screened in 2015.3 Using data from 1,796 accredited lung cancer screening sites compared to survey data of living heavy smokers, only 1.9% of 7,600,000 eligible U.S. residents were screened in 2016.4
“Since the higher the risk, the greater the benefit of screening, the exclusion of lower-risk patients justifies the extreme pressure I propose in favor of screening.”— Steven E. Vogl, MD
Tweet this quote
Screening is recommended by the U.S. Preventive Services Task Force as well as professional societies of oncology and pulmonary medicine, and it is covered by Medicare, Medicaid, and all private insurance companies. What is the solution to activating this massively beneficial technology so that fewer smokers die in the next few years?
I propose four steps:
1. Simplify and rationalize eligibility determination using an app or a preprogrammed tablet, with data entry by the patient, a medical assistant, or a telephone interviewer done for the physician. Eligibility could be based on a conservative view of the NLST population, pushing for screening of only those with the higher risk.
2. Offer a central radiologic facility to evaluate lung nodules in terms of prevalence (initially) and to assess later scans as a service whenever referring physicians or involved radiologists want help.
3. Medicare, insurance companies, rating agencies, and accreditation agencies should evaluate medical providers (physician groups and hospitals) by the percentage of eligible patients referred for screening, with sanctions and monetary penalties for failure to recommend screening.
4. Bring to the attention of large medical institutions and large doctor groups the risks of being sued for wrongful death of heavy smokers not offered lung cancer screening.
Simplify and Refine Who Should Be Screened
THE ENTRY CRITERIA for NLST were fine for a study—simplicity was key to recruitment. The study showed that, in a population with a particular range of risk of developing lung cancer, benefits to screening are substantial in terms of all-cause mortality. Without a doubt, the benefits will be greater among those with greater risk. The NLST has been criticized because it included some populations with a relatively low risk of cancer and excluded some with a very high risk, such as those with a prior cured lung cancer and those who had received prior mediastinal irradiation for lymphoma.
“Clearly, a recommendation by ‘august’ bodies based on overwhelming data is not enough. We have these recommendations, with only about 4% of those eligible getting screened.”— Steven E. Vogl, MD
Tweet this quote
At least nine risk prediction models can already be found in the literature. It makes sense to create 1 computer program or app, enter the appropriate patient parameters based on a questionnaire or interview, and decide to recommend screening if the patient’s risk of lung cancer in the next 5 years is above the median for the NLST population using any of the 4 best models,5 or based on clinical data such as asbestos exposure, prior mediastinal irradiation, or prior lung cancer.
The 5-year incidence of lung cancer discovered on CT in the screening group of the NLST was 900 cases in 26,732 participants, or about 3.3%. Since the higher the risk, the greater the benefit of screening, the exclusion of lower-risk patients justifies the extreme pressure I propose in favor of screening.
How Should We Get Patients to Consent?
TO INCREASE the rate of eligible patients who opt for screening, we should advertise not only about preventing the subject’s death, but about preventing grief and loss (both personal and monetary) among the subject’s spouse and children. The smoker owes it to his family and those who love him or her to be screened. An annual screening chest CT scan could be “pitched” as an intervention that mitigates the potentially devastating effect of a self-destructive addiction to tobacco.
The benefit from the screening occurs only with a positive screen—when a lung cancer that has not yet metastasized can be removed before it spreads. The advertising, therefore, will need to be long term and creative to capture the imagination of those at high risk (and those who care about them) and to repeatedly get their attention. The highest risk of a new lung cancer is among the survivors of a prior lung cancer (about 1.5%– 2.0% annually), so targeted advertising and screening should not stop after the first lung cancer is removed.
Make Screening Easier
EVERY PATIENT who walks into a doctor’s office or enters a hospital should have risk data collected and entered into a computer program or app, with assignment of a 5-year risk of developing lung cancer. For the moment, any one of the four best calibrated risk assignments above the NLST median should mandate annual low-dose chest CT unless some other problem (eg, imminent death, obvious inability to tolerate lung cancer treatment, uncontrolled infection) makes lung cancer screening senseless. The calculation of risk should be taken out of the hands of the physician and done for him.
National reference resources should be available to review serial CT scans to help in the decision to proceed with therapy (generally surgical resection). It seems that volumetric intervention criteria used by NELSON are superior to the single-diameter measurements used in the NLST, with fewer false-positives.
Pressure Institutions to Push for Lung Cancer Screening
CLEARLY, A recommendation by “august” bodies based on overwhelming data is not enough. We have these recommendations, with only about 4% of those eligible getting screened. Lung cancer risk evaluation should be prompted in electronic medical records to remind caregivers to discuss and order low-dose chest CT scans. Physician groups and medical institutions should be sanctioned professionally and monetarily by insurance companies, government agencies, and accreditation agencies for failing to assess lung cancer risk and offer screening when appropriate. Lung cancer screening should be a criterion for evaluation of quality of care , as were pneumococcal and influenza vaccinations in the recent past. Since most physicians are now employed by large groups or institutions, such mechanisms are likely to have considerable influence. If the electronic medical record has a pop-up that says the patient is eligible for screening, with places to click to order the test or print out the form by which the patient refuses the test, screening rates will skyrocket.
Lung Cancer Deaths: ‘Wrongful’ if an Opportunity to Offer Screening Was Missed
“Lung cancer screening is sufficiently important to justify an aggressive approach, even if it is coercive to physicians, patients, institutions, and businesses that manage medical care.”— Steven E. Vogl, MD
Tweet this quote
PLAINTIFF’S ATTORNEYS will soon realize how much money could be at stake if they enroll surviving family members of those who died of lung cancer. Their suits will allege wrongful death after failure to screen. It would take only a few multimillion-dollar judgments to get the lawyers of every large medical institution in the country to induce their employers to force their physicians and staff to push screening for lung cancer.
Several enlightened institutions have already started free screening programs, perhaps because they suspect it is good for business and will increase income. That is fine, but not enough. The cost of the screening CT scan is estimated at about $300, equal to 3 cartons of cigarettes in the city of New York. The cost is not the issue, especially compared to the skyrocketing cost of treating metastatic lung cancer.6
How many premature deaths can be prevented is not totally clear. In the NLST, 123 deaths over about 6.5 years were prevented with only 3 annual screening CT scans in about 27,000 subjects, some of them at relatively low risk of lung cancer. If all the NLST eligible patients in the United States get screened, this calculates to about 32,000 premature deaths averted—about 5,300 per year. Extended annual screening will likely improve the benefit. Selection of only the highest-risk patients will decrease false-positives and improve the cost-benefit ratio.
The goal of selection is not to detect every lung cancer when it is small, but to select subjects for screening whose risk of cancer is high enough so we know there is a significant benefit to annual CT scans in decreasing the risk of early death based on randomized controlled trials of CT screening.
Lung cancer screening is sufficiently important to justify an aggressive approach, even if it is coercive to physicians, patients, institutions, and businesses that manage medical care. The current voluntary system is not working. ■
At Microphone 1 is an occasional column written by Steven E. Vogl, MD, of the Bronx, New York. When he is not in his clinic, Dr. Vogl can generally be found at major oncology meetings and often at the microphone, where he stands ready with critical questions for presenters of new data.
The opinions expressed in this column are those of the author. If you would like to share your opinion on this or another topic, please write to editor@ASCOPost.com.
DISCLOSURE: Dr. Vogl reported no conflicts of interest.
1. National Lung Screening Trial Research Team, Aberle DR, Adams AM, et al: Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med 365:395-409, 2011.
2. De Koning H, Van Der Aalst C, Ten Haaf K, et al: Effects of volume CT lung cancer screening: Mortality results of the NELSON randomised-controlled population based trial. 2018 World Conference on Lung Cancer. Abstract PL02.05. Presented September 25, 2018.
3. Jemal A, Fedewa SA: Lung cancer screening with low-dose computed tomography in the United States—2010 to 2015. JAMA Oncol 3:1278-1281, 2017.
4. Pham D, Bhandari S, Oechsli M, et al: Lung cancer screening rates: Data from the lung cancer screening registry. 2018 ASCO Annual Meeting. Abstract 6504. Presented June 1, 2018.
5. Katki HA, Kovalchik SA, Petito LC, et al: Implications of nine risk prediction models for selecting ever-smokers for computed tomography lung cancer screening. Ann Intern Med 169:10-19, 2018.