Are We Hunting Too Hard?
"Get screened!" "Find it early!" When it comes to cancer, these dictums are considered self-evident. But what if getting screened for cancer and finding cancer ever-earlier does not save lives? What if too much probing does more harm than good? Several DMS physicians have been asking provocative questions like these for more than a decade.
The human body is full of imperfections—most of which we simply didn't know about until recently. Now, thanks to sophisticated scanning technology, like computed tomography (CT) and magnetic resonance imaging (MRI), we're able to see ourselves at a level of detail that has never before been possible. "Because we're now able to see every millimeter of the body, we of course find a lot more abnormalities in the body than we ever knew existed," says Dartmouth radiologist William Black, M.D. "What the imaging does is it makes us think, 'Oh, there is this ton of tumors out there and other diseases, so disease must really be increasing in frequency.'" But is it?
All cancers are not created equal. Some grow rapidly and invade other tissue, others grow slowly and remain noninvasive, and some don't grow at all or may even recede. Many of the cancers that doctors are finding and treating today, says Black, are what's called "pseudodisease"—tumors that will never cause harm, let alone death. The trouble is that pseudodisease is nearly impossible to identify for sure in an individual who is still living, because the medical community doesn't know enough about some cancers to predict how they will behave over time. So it's safer, they reason, to label a questionable abnormality as "cancer" and to treat it, than it is to risk its growing out of control. Only after an untreated person dies from other causes can a cancer be declared pseudodisease. Only then is it clear that treatment of the cancer would have provided no benefit, only potential harm. In other words, you can't tell an "overdiagnosed," or overtreated, person from a person who has been cured. "One of the biggest downsides to cancer screening is overdiagnosis, but you don't know which people have been overdiagnosed," says Black. "And so a person who has been overdiagnosed will think they've been cured."
These ideas may seem revolutionary, even radical, but they're hardly new. Black and fellow DMS physician-researcher H. Gilbert Welch, M.D., M.P.H., have spent more than a decade exploring the potential harms of looking harder and harder for disease. As early as 1993, in a paper published in the New England Journal of Medicine, Black and Welch drew attention to the way that advances in diagnostic imaging can distort physicians' perceptions of disease prevalence and inflate the effectiveness of treatments.
During his 14-year career at Dartmouth, Black has become nationally known and respected as one of the most vocal radiologists when it comes to questioning screening tests. He has coauthored numerous papers on the topic, often with Welch, and served on several National Cancer Institute (NCI) committees, including the 1993 International Workshop on Screening for Breast Cancer, which concluded that there is no proven benefit from mammography for women in their forties.
Although four medical groups—including the American College of Physicians (ACP) and the U.S. Preventive Services Task Force—supported the NCI workshop's conclusion, many other medical and cancer advocacy organizations—including the American Medical Association and the American Cancer Society—strongly opposed the NCI position. Even Congress got involved and, after several hearings in 1994, condemned the workshop's recommendation in a report titled "Misused Science." Eventually, after pressure from Congress, the public, and medical and scientific organizations, the NCI reversed its decision, as did the U.S. Preventive Services Task Force. (The ACP has no current recommendation regarding mammography screening but usually supports the Task Force's guidelines on other matters.)
Today, Black is a cochair of the NCI's multicenter National Lung Screening Trial (NLST), as well as the principal investigator for the DHMC-based arm of the trial. The NLST is comparing two ways of detecting lung
cancer—spiral computed tomography (CT) and standard chest x-rays—to determine if either can reduce deaths from the disease. "You can't figure out" whether certain screening tests really work, Black maintains, "short of doing a huge experiment" like the NLST, which will randomly assign 50,000 patients to one treatment or the other.
While Welch credits Black with initially sparking his own interest in cancer screening, Welch has made significant contributions to the debate, too, in both the scientific literature and the popular press. Most recently, Welch authored a book titled Should I Be Tested for Cancer? Maybe Not and Here's Why. Early in the book, he takes on the concepts of overdiagnosis and pseudodisease, using prostate cancer as an example.
"The most compelling evidence that pseudodisease is a real problem comes from our national experience with prostate cancer," Welch writes. Prostate cancer is the second-leading cause of cancer-related death in American men, and over the last 30 years, more and more of it has been found. In 1975, about 100,000 new cases were diagnosed; in 2003, about 220,000. At first glance, one might conclude that prostate cancer is on the rise. However, if a cancer is "really increasing," says Welch, "you'd expect death rates to rise."
And that hasn't happened with prostate cancer. The death rate has remained more or less constant, hovering around 30,000 deaths per year in the U.S., with a slight decline in recent years. (See the graph here for a visual representation of this data.) Some argue that the decline "attests to the usefulness of testing for prostate cancer," writes Welch, "but it could just as easily re- flect better treatment." Regardless of the small changes in the death rate, Welch believes that most of the new cases represent "nothing more than pseudodisease: disease that would never progress far enough to cause symptoms—or flat-out would never progress at all.
"But what, you might ask, is the harm in finding all this pseudodisease?" Welch writes. "Simply put: unnecessary treatment. Most of the million men whose prostate cancer is found because of superior screening have to undergo some sort of treatment, whether radical surgery or radiation. . . . [And] many experience significant complications: 17% need
Jennifer Durgin is Dartmouth Medicine's senior writer.