Read time: 6 minutes.
The 2026 World Conference on Lung Cancer (WCLC) brought the global thoracic oncology community to Seoul, South Korea, from September 12–15. This year's theme, “Science Without Boundaries: Uniting the World Against Thoracic Cancer,” was particularly important for early detection, where the central question of who gets screened and who doesn’t, is being addressed in different ways across the globe. We saw progress in how lung nodules are managed, how AI can help predict risk, and how screening can reach people that current guidelines miss.
Here are the conference highlights, organized by screening-eligible and screening-ineligible populations, followed by new findings on lung cancer risk and on lung cancer in people without a history of tobacco exposure.
Screening-Eligible Populations
Currently, lung cancer screening using low-dose CT scans is available in the U.S. for people:
- Between the ages 50-80, and
- Have a 20-pack year history of tobacco use, and
- Currently use tobacco or have quit within the last 15 years.
Do All “Hazy” Lung Nodules Need Aggressive Treatment? No
Low-dose CT screening often finds subsolid nodules, spots that look partly or fully hazy (called ground-glass opacities) rather than solid. Many go away on their own, usually because they were caused by infection or bleeding. In one screening study of about 50,000 people, roughly 30% of subsolid nodules seen on the first scan and 80% of new ones shrank or disappeared. In an Italian cohort followed for more than nine years, no lung cancer deaths came from subsolid nodules. Confirming a nodule persists, following it long-term, and choosing watchful waiting over aggressive treatment is a safe approach.
Do AI Risk Scores Reflect More Than Biology of a Lung Nodule? Possibly
Sybil is a free, open-access AI model that predicts a person's lung cancer risk from a single low-dose CT scan. A U.S.-based study of 2,638 people enrolled a diverse population in a screening program, where 62% were non-Hispanic Black, and 34% were on Medicaid. Each 0.1 rise in the Sybil score was linked to a 72% higher risk of a lung cancer diagnosis. However, scores also tracked with insurance type, education, neighborhood violent crime, and local high school graduation rates, even after accounting for other factors. AI can therefore be a useful tool to incorporate social determinants of health in lung cancer screening. However, understanding what drives these scores will be important before AI tools are widely used.
Will AI Change How Nodules Are Followed? Not Yet
A nodule's risk of developing into lung cancer changes over time. Combining serial scans with clinical data and blood-based biomarkers may sharpen predictions. However, major guidelines still treat AI as a supporting tool. Models must prove accurate across different hospitals and populations, and show that they lead to better clinical decisions and patient outcomes.
Screening-Ineligible Populations
Do Current Guidelines Miss Many People With Lung Cancer? Yes
In the U.S., screening eligibility is based on age and history of tobacco use. The Boston Lung Cancer Study found that only 46.1% of patients with lung cancer would have qualified had they been screened under the current criteria. Counting pack-years also disadvantages Black patients, given that 57.6% would have qualified versus 74% of White patients. Using 20 years of tobacco use duration instead would raise eligibility to 85.3% and 82%, respectively.
Should People Who Never Smoked Be Screened? Not Quite There Yet
In Taiwan's national program, people with a family history had a higher detection rate than heavy smokers (1.5% vs 0.9%), and 89% of their cancers were found at stage 0–I. In the Female Asian Nonsmoker Screening Study (FANSS) in New York, a first scan found lung cancer in 16 of 1,000 women, 75% at the earliest stage (IA). Every cancer carried a driver mutation, such as EGFR, that can be treated with targeted therapy. Large trials are needed to define who should be added to screening guidelines.
Can AI Find Lung Cancer on Scans Taken for Other Reasons? Yes
About 43.7 million Americans are eligible for a heart CT that measures coronary artery calcium, versus 12.8 million eligible for lung screening. Researchers applied Sybil to baseline heart scans from 5,726 people in the MESA study, half of whom had no history of tobacco use. Over 15 years, people in the top 1% of Sybil scores formed a very high-risk group among both tobacco users and never-users suggesting that coronary artery calcium scans may be a useful starting point to screen this population.
It is important to note that prospective studies are needed before this is used in the clinic. Researchers at Memorial Cancer Institute in Florida built an AI-driven incidental pulmonary nodule program. An AI tool reviewed 348,989 reports from CT scans of the chest, abdomen, spine, and heart. A nurse navigator and expert team reviewed 3,120 flagged cases, leading to 177 lung cancer diagnoses, compared with 32 through the system's screening program. Most of these patients did not qualify for screening. Because screening found cancer far more often per scan, incidental nodule programs should complement, not replace, lung cancer screening.
Defining “Risk” for Lung Cancer
Does Radon Testing Reach the Communities That Need it Most? No
Radon is a radioactive, odorless gas that seeps into homes from the ground and is the second-leading risk factor of lung cancer after tobacco. Researchers at Michigan State University analyzed about 99,300 home radon tests across 83 Michigan counties. Counties in the lowest income quartile had 37.5% higher lung cancer rates than the highest, and rural counties had higher rates than urban ones (63.8 vs 57.4 per 100,000). Despite radon being a known risk factor, county radon levels did not track with lung cancer rates because of low radon testing. Testing rates also fell as the share of Black residents in a county rose. Radon testing and mitigation need to be accessible to every community.
Does Air Pollution Affect Survival After Diagnosis? Yes
The International Lung Cancer Consortium that included 6,261 patients in Canada, the U.S., and the U.K. showed that higher exposure to PM2.5 and nitrogen dioxide at diagnosis was linked to a higher risk of death within five years, even after accounting for tobacco use. The effect of PM2.5 appeared greater in never-users. The team suggested adding environmental exposures to prognostic models and considering steps like home air filtration for high-risk patients.
How Common is Lung Cancer in People Who Never Used Tobacco? More Common Than Many Think
Using national cancer data, research from Mass General Brigham estimated about 29,983 new U.S. lung cancer cases each year among people who never smoked, or 13% of all cases. Of those who had never used tobacco, 58% of patients were diagnosed before age 40. They were also more likely to be diagnosed with advanced lung cancer, reiterating the need for early detection strategies for this population.
Looking Ahead
WCLC 2026 showed early detection is moving toward risk-based screening, less invasive nodule management, and AI tools that help us reach people who are missed by current guidelines. Expert discussions also made it clear that radon, air pollution, and other non-tobacco risk factors need to be integrated into lung cancer screening discussions.
More World Conference on Lung Cancer 2026 Coverage:
