EDITOR’S NOTE: This story was reported in collaboration with the Global Health Reporting Center with support from the Pulitzer Center.
Lung cancer kills more Americans than any other form of cancer: more than prostate cancer and colon cancer combined, and nearly three times as many as breast cancer. But while pink ribbons are a familiar sight and 4 in 5 women get their recommended mammograms, screening for lung cancer is far less common. It’s often viewed as a disease primarily of heavy smokers, when the reality is far more nuanced.
Technology offers new opportunities to catch lung cancer early, when it can be treated most easily. But official screening guidelines, which play a major role in determining whether insurance will cover the cost of screening, often fail to pinpoint who is at risk. What’s more, only 20% of people who are eligible for screening actually get checked. Some of this failure can be traced to persistent myths about lung cancer.
Juliet DuBois learned that she had cancer after she had a hip replacement two years ago, at age 46. In the weeks after the operation, a blood test showed warning signs of a possible blood clot. DuBois went to the ER, where a CT scan revealed a 1-centimeter mass in her lung.
DuBois isn’t sure whether to consider herself lucky or unlucky. “If I hadn’t gotten a hip replacement and then been extra careful, I would never have known about it,” she said.
Before the hip operation, the former dancer noticed aches and pains but no other symptoms aside from lingering fatigue and sleeping more than usual.
After four rounds of chemo, she now says she’s feeling “pretty good” and has no evidence of disease. This winter, she started an online MBA program, chasing a dream she put off for years. “It can’t be as scary as cancer,” DuBois said.
Both of DuBois’ parents were heavy smokers, which turned her off cigarettes and kept her from ever picking up the habit. Lung cancer screening was never on her radar: “I just wonder, if someone had recommended it, if I’d have done the thing.”
In fact, as many as a quarter of all lung cancer appears in “never-smokers,” defined as smoking fewer than 100 cigarettes in a lifetime. Lung cancer in nonsmokers is more common among women than men, and it’s especially common in women with East Asian or South Asian ancestry. One study found that 83% of female lung cancer patients in south Asia were never-smokers.
Albertha “Bertie” Gethers started smoking in junior high, hanging out with friends in Mattapan, Massachusetts. It was the 1960s, and for 12-year-old Bertie, Virginia Slims were the brand. “We thought it was cute,” she recalled.
Gethers never smoked more than a few cigarettes a day, even as she kept up the habit for another 56 years. Because it didn’t add up to a “20-pack-year” smoking history - the equivalent of a pack a day for 20 years - she didn’t meet official criteria for screening. That meant Medicare wouldn’t pay for it, and no doctor suggested that she seek out screening on her own.
But in 2024, a friend tipped Gethers to a research program in Boston that offered free CT scans to Black women between the ages of 50 and 80. Accompanied by an aunt, Gethers went for a scan, which to her surprise turned up three cancerous lesions on her lungs. Dr. Chi-Fu Jeffrey Yang, a thoracic surgeon at Mass General Brigham Cancer Institute, removed the growths, and Gethers made a good recovery.
The program Gethers stumbled across was the “INSPIRE” study, launched by Yang along with medical students Alex Potter and Deepti Srinivasan. A primary aim is to explore the feasibility of lung cancer screening for Black patients with any smoking history, even if they fall outside the 20-pack-year guideline.
Black patients are at particular risk; they are less likely to be screened for lung cancer despite having a higher chance of dying from the disease. One reason for the discrepancy is that Black patients are more likely to be light or irregular smokers who don’t smoke “enough” to qualify for screening. During the first phase of the INSPIRE study, several lung cancers were detected in participants – including Gethers – who don’t meet current screening eligibility criteria.
Overall, more than half of all lung cancer appears in people who don’t qualify for screening. Their cancer is found only after symptoms appear, or by accident if someone is having a CT for an unrelated reason, like a rib injury or a cardiac test.
Yang says lung cancer risk is shaped by environmental triggers like air pollution and radon gas, by genetic factors and, of course, by smoking. However, he says the “20-pack-years” standard leaves out a lot of vulnerable people. “I think it really is about how many years you smoked, rather than the intensity.”
Standard lung cancer screening uses low-dose computed tomography, known as low-dose CT or LDCT. It’s a type of X-ray that uses a low amount of radiation to create a 3D image of the lungs. Because the test typically costs between $200 and $400 and can sometimes lead to unnecessary medical procedures, major medical organizations don’t recommend routine screening for everyone, just those at highest risk.
The problem, says Dr. Jessica Donington, chief of thoracic surgery at the University of Chicago, is that their classifications are outdated. “The criteria are just way too narrow,” she said.
Until 2022, the three major organizations that issue guidance on lung cancer screening all used the same criteria. They said CT scans were warranted only for people ages 50 to 80 who have a “20-pack-year smoking history” and who still smoke or quit less than 15 years ago.
In 2022, the National Comprehensive Cancer Network changed its criteria to cover more current and former smokers. That group now recommends screening for anyone who smoked any amount for more than 20 years. The American Cancer Society has also expanded its guidelines and no longer automatically classifies someone who quit smoking more than 15 years ago as low-risk.
But the largest and most influential organization, the US Preventive Services Task Force, has not evaluated new evidence since 2021.
The USPSTF, a national panel of 16 volunteers who offer evidence-based recommendations about preventive health, is especially influential because the Affordable Care Act relies on those recommendations to determine what screenings must be paid for by insurance. Guidelines are typically reviewed every five years, which led experts to expect an update this year. However, no review has been announced. In fact, March 2025 was the last time the task force met to review evidence on any topic.
Terms for five of its 16 members expired in January, and the US Department of Health and Human Services, which appoints members, has not announced replacements.
In a notice published last week in the Federal Register, HHS asked for nominations of new members, encouraging anesthesiologists, cardiologists, oncologists, radiologists, obstetricians and other specialists to apply, although the members have typically been primary care doctors.
When low-dose CT scans were first tested in the 1990s and early 2000s, studies found that they detected far more early-stage cancers than traditional CT scans or X-rays but didn’t significantly reduce the mortality rate. Based on that, many physicians felt that the benefits did not outweigh the risks, including potential harm from surgery.
In 2011, however, the National Lung Screening Trial, which followed more than 50,000 patients over several years, found that low-dose CT screening led to a 20% decrease in mortality rates. That finding led to the first national recommendations to promote LDCT, along with widespread insurance coverage and growing use of the scans.
Yang says surgeons today typically avoid aggressive treatment. When a suspicious nodule is found, he says, the usual followup is not surgery but a repeat scan a few months later. Only in about 1% of cases, when a nodule grows or changes, is a biopsy done. Of those, 90% reveal cancerous tissue.
Growing up in State College, Pennsylvania, Yang spent countless hours playing catch and puttering in the garden with his grandfather, a chemical engineer and the family patriarch. Months after Yang left home for his freshman year of college, his grandfather was diagnosed with lung cancer. He died a year and a half later, in 2004.
“It was tough, because he was just a rock for me,” Yang said. “We’d talked about medical school, and I always have this deep regret he didn’t get to see me go.” The college freshman watched his grandfather suffer and helped keep him company after difficult chemo treatments. “It stuck with me in a very strong way and made me want to not just be a doctor but to treat lung cancer patients,” Yang said.
His grandfather had been a smoker for 30 years but died before the advent of widespread CT scans. “I’m convinced that if we’d found it early, he could have stuck around to see me be a doctor.”
Everyone told Loryn Fadus it was normal to feel exhausted. What mother of a 2-year-old isn’t? But Fadus, 34, an IT project manager in Deerfield, Massachusetts, couldn’t shake her sense that something was wrong. While trying to deal with her fatigue and then suspected pneumonia, she saw four doctors without anyone mentioning the possibility of cancer. Only after Fadus coughed so hard that she broke a rib did doctors order a low-dose CT scan, which identified cancerous growth in her lungs, ribs, liver and spine.
“I was shocked,” Fadus said. “I had never smoked. I was big into hiking and biking. Lung cancer was absolutely the last thing I ever would have suspected.”
About 1 in 10 newly diagnosed lung cancer patients is younger than 55. Although lung cancer overall is significantly more common in men, that’s not true among these younger patients. Among those under 40, women make up 52% of new cases – and an even larger proportion of cases in “never-smokers.”
For younger patients as a whole, the disease is more likely to be found at an advanced stage.
In her Chicago-based practice, says Donington, with younger patients, “the majority are diagnosed at stage IV. They only come in when they have symptoms, like having seizures or pain in a rib.”
According to the Society of Thoracic Surgeons, the 20-year survival rate is 80% when lung cancer is caught early, compared with a five-year survival rate of just 8% with a late-stage diagnosis.
Kelley Jones, an online marketing professional from Newburyport, Massachusetts, was at the nail salon when she noticed that her fingernails were curving downward, a condition known as nail clubbing that’s often a sign of a heart or lung condition. Jones, 36, had had months of severe bone pain and swelling in her legs, and this was the final straw that pushed her to see a pulmonologist. She was reluctant, but her sister, a nurse, pressed the issue.
An initial battery of tests – including an echocardiogram, a lung function test and an X-ray – picked up nothing unusual. But a CT scan told Jones and her doctors that she had lung cancer and that they’d caught it at stage I.
“That’s extremely rare, and I was especially fortunate to have good health insurance through my job,” Jones said.
She reflects on how easily things could have been different. “I remember the doctor finally said ‘we’ll do a CT scan,’ ” she recalled. “He said, ‘you’re not a smoker, so it’s probably nothing, but let’s do it.’ And ironically, that’s the test that saved my life.”
It’s true that a cough that doesn’t go away is a common symptom of lung cancer, as is shortness of breath, chest pain, unexplained weight loss, constant fatigue or coughing up blood. But severe inflammation or neurologic symptoms may also be early signs, especially in younger adults.
Predictions are hard, but Yang says state-of-the art computer models may do a better job than official guidelines in identifying who is likely to develop cancer. An experimental AI program called Sybil, developed by MIT computer scientists and Yang’s colleague Dr. Lecia Sequist, can “look” at a single CT scan and generate a “risk score” corresponding to the likelihood of the person developing cancer over any period up to six years.
In 2023, its developers reported that Sybil was between 86% and 94% accurate in identifying which patients were at higher risk and which were at lower risk of developing lung cancer within a year.
“The whole idea with early detection is to try and find cancer when it’s small,” said Sequist, director of the Cancer Early Detection and Diagnostics Program at Mass General Brigham Cancer Institute. “With radiology screening, you’re stuck in a narrow window where you can see it on the scan. It’s just luck.”
This approach is distinct from FDA-approved AI models that analyze CT scans to assist a human radiologist in detecting cancerous nodules and determining a course of treatment. Sybil is also distinct as being open-source software: No company owns or promotes the technology, but it can be freely shared through academic channels.
Sybil is being used for research projects in about a dozen US hospitals, including Mass General Brigham, and in more than 30 countries around the world.
Yang and Sequist are preparing a trial that would follow people whose smoking history doesn’t qualify them for regular screening but who received a CT scan for some other medical reason. Sybil will assign a risk score to these patients, and they’ll be followed prospectively, with the highest-risk patients receiving more frequent screenings. “I expect we’ll find quite a few lung cancers that way,” Yang said.
Ultimately, Sequist hopes Sybil could be used to personalize screening regimens. “Ideally, everyone would get an initial CT scan, and based on the results, they’d be stratified into different categories,” she explained. Higher-risk patients might be told to come back in six months or a year, while others could rest easy and wait five or six years to be screened again.
“We could be smarter about this if the system isn’t so closely tied to smoking,” Sequist said.