Can cancer scans lead to overdiagnosis?
The decision last week by the U.S. Preventive Services Task Force recommending annual lung cancer screenings for heavy smokers using low-dose computed tomography scans has reignited the controversy among physicians about the trade-off between benefits and risks for a number of cancer-screening procedures.
Conducting broad screenings for various cancers, such as for the breast and lung, has generally been credited for improving survival rates by catching the disease in its early stages.
But the practice has come under increasing fire from experts who contend screening has been overused. It has led to overdiagnosis of pre-cancerous lesions that may never evolve to full-blown cancers, and subjected people to unnecessary biopsies and treatments, the critics say.
“Overdiagnosis is a huge issue in cancer and is under-recognized among physicians,” said Dr. Otis Brawley, chief medical officer for the American Cancer Society. “If it’s under-recognized among physicians, you can’t expect it to be something patients appreciate a great deal.”
Brawley said he applauded the recommendations a group of doctors recently published on the Journal of the American Medical Association’s website, which suggested redefining how physicians diagnose cancer by excluding the term when only premalignant conditions were found.
In such cases, what’s usually found—known as incidentalomas—are so slow-growing it is unlikely they would ever become a health concern within a patient’s lifetime. This has become a major concern for elderly men diagnosed with prostate cancer.
Some experts, however, contend the benefits of screening some cancers outweigh the risks. The task force estimated that for every 320 people screened, one life was saved in the case of lung cancer, while one life was saved for every 900 to 1,900 mammograms taken in breast cancer cases.
Screening proponents say the issue of overdiagnosis is not real because no doctor can definitively tell a patient whether a cancer will progress to become life-threatening. “I think I would be very cautious about changing terminology that may convey the sense to the patient that they will not develop an invasive cancer when we can’t be sure that they won’t,” said Dr. Larry Norton, medical director for the Evelyn H. Lauder Breast Center at Memorial Sloan-Kettering Cancer Center in New York.
But problems arise when screenings produce false positives, which occurred in 96% of CT lung cancer tests that initially tested positive, according to the panel’s draft report. Fully 24% of the screens for 53,000 heavy smokers were positive for cancer. But the 96% false-positive rate means 24 out of 25 of those cases would lead to further tests, exposing the patients to more radiation and could lead to more invasive and costly procedures that carried higher risks of complications.
Despite those concerns, the task force still recommended the use of CT scans, which could reduce the number of lung cancer deaths among high-risk patients by 16%. The task force tentatively gave it a B rating, with a final ruling due in several months. The Patient Protection and Affordable Care Act requires insurers to provide firstdollar coverage for preventive services given an A or B rating by the government advisory body.
While Brawley acknowledged the benefits of broad screenings for some cancers, he said for others, such as prostate cancer, the results have been less encouraging. He estimated 60% or more of positive screenings are actually examples of overdiagnosis that do not require additional treatment.
According to the JAMA study, which was funded by the National Cancer Institute, even though screening has increased the overall number of early cancer detections during the past 30 years, it has not produced a proportional decline in the number of cancer-related deaths.
The problem has been the assumption that all lesions that are detected during a screening have the potential to be life-threatening, said Dr. Laura Esserman, lead author of the JAMA report and director of the University of California, San Francisco’s Carol Franc Buck Breast Care Center. She said such assumptions have led to overdiagnosis in 20% to 50% of cancer cases.
With the added health risks that can result from an overdiagnosis comes the added cost of treating a patient.
Brawley and Esserman said the costs related to overdiagnosis possibly amount to billions of dollars.
As more is learned about the disease, Esserman said she felt confident more physicians would begin to develop a more nuanced approach in the way they approached a cancer diagnosis.
“We have to work hard to educate the public to let them know that we’re not playing Russian roulette with their life—that it’s common to have conditions that are not serious,” Esserman said. “Just like when you’re trying to test a new treatment to see if you can improve the chance of survival, we have to start testing how we can safely do less.”