To Screen or Not to Screen
By Keith Loria
Vol. 24 No. 8 P. 10
Lung cancer screening with low-dose CT (LDCT) has been considered a useful method of detecting cancer early and potentially saving lives. However, a recent meta-analysis conducted by the University of Oslo in Norway calls that assumption into question. The study concluded that cancer screenings offer limited benefits, which contrasts with the way that screenings have been promoted.
“Our study was about the statement of cancer screening as a prolongation of life, and we found that most or all do not prolong life in a meaningful way,” says Michael Bretthauer, MD, PhD, a professor of medicine at the Institute of Health and Society at the University of Oslo. “My conclusion from these data is that we stop advertising screening with the slogan related to life prolongation, as outlined in the paper.”
He cautions, however, that the question of whether lung cancer screening in general is a good idea or what can be done differently for screening tests was not within the scope of the study.
“What I am recommending is to stop saying that cancer screening saves lives when there is no good evidence that it does, which there is not, according to our results,” Bretthauer says.
The primary research in the space is the 2011 National Lung Screening Trial (NLST) study published in the New England Journal of Medicine, which found that LDCT-based lung screening reduced lung cancer mortality by 20%.
“The control set was using chest X-ray, so we know that the benefit could be even more, compared to those without any screening,” says Kaiwen Xu, a PhD candidate in medical image analysis and statistical interpretation in the lab department of computer science at Vanderbilt University in Nashville. “The conclusion was later confirmed by a study in Europe called NELSON. These large-scale studies formed a solid foundation for the current policy to promote lung screening with LDCT in high-risk individuals.”
Xu was the lead author of the study “AI Uses Lung CT Data to Predict Mortality Risk,” which mainly focused on body composition assessment using AI and how it improves the value of LDCT lung screening beyond early lung cancer detection.
“It turns out these markers not only improve the lung cancer mortality prediction but they are also useful to predict mortality of other causes like cardiovascular diseases,” he says. “This is something very new to the field. This is also a meaningful finding since the current enrollment criteria of LDCT lung screening covers older adults (>50 y) with a heavy smoking history who are exposed to higher risks, not only limited to lung cancer but also to other diseases like cardiovascular disease.”
Lack of Screening in the United States
Edwin J. Ostrin, MD, PhD, of the division of internal medicine at MD Anderson Cancer Center, says that right now, the biggest limitation in lung cancer screening is uptake. Depending on state and other factors, uptake ranges between 1% and 30%. This means that the vast majority of the 14 million people eligible for lung cancer screening in the United States are not being enrolled.
A group from the Medical University of South Carolina recently published a study that found, even in people who do enter a screening program, the number who make it back for their second annual CT is surprisingly low, with just 22% returning.
“The reasons for these low numbers are manifold, and there is a whole body of literature on the barriers to screening,” Ostrin says. “Some programs have been more successful meeting these barriers and have boosted numbers and improved compliance. It takes quite a lot of effort, and in resource strained health systems, which often serve the people who may benefit most from lung cancer screening, this amount of effort may not be feasible.”
When it comes to who is enrolling in lung cancer screening, lower numbers are expected in historically underserved populations.
The most significant variable, in Ostrin’s opinion, comes from state-level enthusiasm for lung cancer screening and implementation of programs to improve uptake. For example, Nevada and California have rates in the 1% range, while Kentucky, Vermont, and Massachusetts have rates in the 13% to 15% range. The national average is around 5%.
“These higher uptake states have state-level programs to identify and screen eligible individuals,” he says. “So, the numbers are low overall, but there are definitely things, including more centralized programs, that may help out.”
Ins and Outs of Participation
There are many barriers to screening participation, including patient-level barriers, such as low knowledge of lung cancer screening, worry about false-positive scans, and access to programs.
“While LDCTs are now paid for under Medicaid and Medicare, there is considerable overlap between those eligible for lung cancer screening and those who are underinsured, especially since eligibility has increased to those 50 years and up,” Ostrin says. “There are provider-level barriers, including worry about increased work in following positive scans, coordination with tobacco cessation programs, the need for a rather lengthy shared decision-making documentation in order to enroll in programs, and perceived lower impact of lung cancer screening in these patients who may have other tobacco-related comorbidities like COPD or heart disease.”
Mona S. Jhaveri, PhD, who has shifted from working in lab-based settings to founding a nonprofit called Music Beats Cancer, says numerous factors affect lung cancer screening.
“Unfortunately, a tiny percentage— about 6%—of at-risk people get screened,” she says. “In the US, Massachusetts has the highest screening rate at 16.3%, and California has the lowest at 1%.”
Jhaveri cites poor communication between providers and their patients as the main reason patients do not realize they are at risk for lung cancer and need to be screened.
“There is also an unawareness of screening programs and a lack of insurance to cover this cost,” Jhaveri says. “One significant barrier to screening is simply fear of getting a cancer diagnosis and its perceived stigma.”
There are system-level barriers, as well. Successful implementation of lung cancer screening requires a concerted effort to ensure appropriate follow-up per correct recommendations; knowledgeable and receptive radiologists, pulmonologists, and oncologists to evaluate positive scans; maintenance of ACR-compliant databases; and many other factors.
There are multiple guidelines and suggestions for building an effective lung cancer screening program. These include community outreach, including in some rural centers, and factors such as transportation and mobile CT units.
“Electronic decision aids and decreasing the shared decision-making requirements have helped improve provider participation,” Ostrin says. “There was an intriguing study recently comparing a centralized lung cancer screening program, which included devoted and knowledgeable staff and physicians, coupled with tobacco cessation vs decentralized approaches. The centralized approach seemed to work better.”
Jhaveri believes it is vital that the United States implement a national comprehensive screening program to raise awareness and enable access to lung cancer screening programs. As part of this, she adds, insurance companies must encourage screening by making these procedures free annually.
Challenges in Asia
The incidence and mortality of lung cancer are highest in Asia, compared with Europe and the United States, with the incidence and mortality rates being 34.4 and 28.1 per 100,000, respectively, in East Asia. Asian countries face several challenges in implementing LDCT screening, such as economic limitations, lack of early detection efforts, and lack of specific government programs.
A recent study, “Lung Cancer Screening in Asia: An Expert Consensus Report,” examined the challenges that countries in Asia face in getting people screened. Pan-Chyr Yang, MD, PhD, chair professor in the department of internal medicine at National Taiwan University Hospital, who served as lead author of the study, says awareness that screening can help detect lung cancer in an early curable stage in high-risk populations is most important to increase participation.
“Awareness of disease risk and severity, as well as the risk and benefit of LDCT for lung cancer early detection, are the keys,” he says. “Asia has the highest lung cancer disease burden in the world—60% of lung cancer incidence and 62% of lung cancer mortalities occur in Asia. If we want to improve lung cancer survival, we need to expand the screening eligibility to include nonsmokers with risk factors and not just focus on heavy smokers.”
What Is the Answer?
If, as the Oslo study suggests, LDCT screening isn’t the answer, the big question is, “What is?”
“Biomarkers, lung cancer prediction models for smokers and nonsmokers at risk, and AI radiomics may be helpful,” Bretthauer says.
Ostrin believes that LDCT needs to be a part of the answer, in conjunction with a comprehensive care setting that includes tobacco cessation, other cancer screening, and addressing comorbidities that could be discovered in screening, such as cardiovascular disease.
“I do think that there are improvements to be made with improving the personalization of care—including things like better risk calculators, blood biomarker tests that can improve clinical risk calculations, and multiple-cancer early detection tests,” he says. “But it took the most expensive trial of all time— NLST—to demonstrate the benefit of lung cancer screening, and I don’t think that we will find something that outperforms CTs in the near term. This is especially true because CTs are getting cheaper.”
He notes that at the World Congress on Lung Cancer, Chinese and Korean researchers estimated that LDCT was costing $30 to $100 out of pocket for patients in those countries, and he finds it hard to envision another test that will provide as much information at such a low cost.
An intriguing future direction, Ostrin says, will be expanding the scope of lung cancer screening to include a focus on other CT findings. For example, in England, they are billing lung cancer screening as a “chest health scan” and are including formal assessment of coronary artery calcifications.
“So perhaps that may increase the perceived value of these scans for the patient,” he says. “We would like to also consider that something like a blood biomarker, which may be able to give a patient and provider more personalized and precise information on the particular benefit for that patient of lung cancer screening, would improve participation. We may also, in the future, use more precise risk models and a blood biomarker to identify those at risk who are currently not eligible, for instance, light and never smokers.”
— Keith Loria is a freelance writer based in Oakton, Virginia. He is a frequent contributor to Radiology Today.