There is no cancer more deadly than lung cancer. It will kill about 130,000 Americans this year, more than the next three most common causes of cancer death — colorectal, breast, and prostate cancer — combined. Lung cancer is the most consequential cancer throughout the world.
Nevertheless, some lung cancers are of no consequence — unless they are found. Last month in the journal JAMA Internal Medicine, colleagues and I reported on the disturbing trend of lung cancer overdiagnosis among women in Taiwan. From 2004 to 2018, there was a sixfold increase in the diagnosis of early-stage lung cancer in women — but no change in their frequency of late-stage lung cancer. This combination strongly suggests that the new cancers being found were generally not those destined to progress.
The culprit? Widespread, indiscriminate screening for lung cancer.
“Overdiagnosis” refers to the detection of abnormalities that meet the criteria for disease yet are not destined to cause either symptoms or death. Although it may be hard to imagine you can overdiagnose the world’s deadliest cancer, lung cancer experts have recognized the possibility for decades.
Get The Gavel
A weekly SCOTUS explainer newsletter by columnist Kimberly Atkins Stohr.
In the 1970s, the Mayo Lung Project put 9,200 heavy cigarette smokers in a randomized trial for six years. Half received a chest X-ray every four months; the others got medical care as usual. At the end of the trial there were more cases of lung cancer in the screening group than in the usual-care group. That’s to be expected. Screening can bring cases forward in time.
But 20 years later, long after screening had stopped, there were still significantly more cases of lung cancer in the group that got screened. That means the chest X-rays had detected some cancers that would never have otherwise become apparent. That’s unexpected. That’s overdiagnosis.
If lung cancer overdiagnosis could occur with low-resolution chest X-rays in the 1970s and ’80s, imagine the potential problem with high-resolution CT scans today. In fact, investigators in Japan reported in 2001 that CT screening detected virtually the same amount of lung cancer in people who never smoked as in smokers.
That’s what happened in Taiwan. Hospitals bought CT scanners, priced exams low (some even offered free scans), and widely promoted screening — with little attention paid to distinguishing high-risk smokers from low-risk never-smokers. Promotions often featured images of young women preparing to enter the doughnut-shaped CT scanner, despite the fact that fewer than 5 percent of Taiwanese women smoke cigarettes.
Over the last 15 years, roughly 10,000 Taiwanese women have been overdiagnosed with lung cancer. This is not the problem of false alarms, which is what happens when the results of a test are worrisome but are ultimately shown to be in error. These are women who were told they had lung cancer and had surgery to treat them for it.
By the way, these women generally do quite well. They survive their surgery and their “cancer.” Some become strong advocates for screening. Collectively, they contribute a salutary effect on survival statistics: The five-year survival rate has more than doubled. Now Taiwanese women have the highest lung cancer survival rate in the world, at 40 percent — despite no change in their lung cancer death rate.
Pop quiz: What is the fastest way to increase cancer survival rates?
Answer: Diagnose everybody with cancer.
There is good news in lung cancer. Every year, 30,000 fewer Americans die from the disease than did in 1998. This decline has been by far and away the biggest contributor to the decline in US cancer deaths overall. But it’s not because of screening. Nor is it because of treatment. It’s because of the reduced exposure to the most important risk factor: cigarette smoking.
But as cigarette smoking declines, smoking-related lung cancer declines. Even if there is no change in the amount of lung cancer in never-smokers, a higher proportion of lung cancers will occur in never-smokers. In turn, this will increase the pressure to screen for lung cancer in never-smokers. Last March, the US Preventive Services Task Force expanded the criteria for lung cancer screening to include those with more moderate smoking exposure. Expect calls for the task force to expand the criteria further.
Lung cancer is a bad disease. But screening low-risk people is harmful. It comes with overdiagnosis and overtreatment — not to mention false alarms, more diagnostic procedures, and even a few radiation-induced cancers. Don’t assume it is always good to find more lung cancer. It’s not.
H. Gilbert Welch, a senior investigator in the Center for Surgery and Public Health at Brigham and Women’s Hospital, is the author of several books, including “Less Medicine, More Health: 7 Assumptions That Drive Too Much Medical Care.”