Cancer Screening: Overdiagnosed, Overtreated, and Blind to the Risks?

Posted 01/12/2014

At first blush, cancer screening seems like a no-brainer, sort of like getting your teeth cleaned. You don’t relish the project, but you know it’s good for you. And if a screening detects a God-forbid-bite-your-tongue-don’t-say-the-word diagnosis, at least you’ve likely caught it early.

Oh, were life so straightforward. Ponder this a moment. When your doctor wants to write a prescription, you’ll ask about side effects, in your mind weighing the benefits versus risks, and frequently you’ll decline. With surgery, too, you may get a second or even third opinion, keeping the surgeon at arm’s length while you decide for yourself.

But when it comes to cancer screening, for most people it’s show me where to sign up.

Of course, what doctors don’t tell you, because most of them don’t know themselves, are the risks you’ll subject yourself to with cancer screening. And it turns out that once people are told the risks, a lot of them start having second thoughts.

The two risks of cancer screening are overdiagnosis and overtreatment

First let’s get our definitions straight:

Overdiagnosis is finding an abnormality that may meet certain criteria for cancer (“irregular mass on chest x ray”), but will never progress to disease because there was no cancer in the first place. That alleged mass they visualized three months ago is probably an old scar, not cancer. Of course, during those three months they wanted you to wait to see if it changed in any way on a repeat x ray or CT scan you’ve not had a restful night’s sleep, but at least you probably stopped smoking.

Overtreatment is exactly what it sounds like. The vigorous, aggressive treatment of something found during diagnosis that doesn’t require treatment. Like everything in medicine, overtreatment can definitely be fraught with harm, including surgery, chemo, and radiation. “Looks like we’ve caught that lung mass early. We’ll start with radiation and then remove it from your lung.” The radiation fries your chest and then the surgeon cracks it open and removes the mass. Soon a pathology reports comes back reading “Old scar tissue; no cancer found.”  Now you’re lying in bed, weak as a limp rag, glowing in the dark from radiation, your chest wound raw and infected, and for what? Removing some scar tissue? Of the millions of men who underwent PSA (prostate-specific antigen) screenings for prostate cancer, tens of thousands had their prostates cut out, not a few dying as a result.

Let’s say that before you had a cancer screening–a mammogram, for example–you were to learn these facts:

• That on average, out of 2000 mammograms performed just one woman fewer will die of breast cancer. Other women may be discovered to have breast cancer, but not by mammograms, and most will survive, but naturally some will succumb. The point is that of 2000 mammograms performed, the life of only one woman will be saved. Now you might say, as many doctors themselves do, “Screenings are worthwhile if only for this one woman.” That would be okay except what about the women who suffer (and die) from overdiagnosis and overtreatment?
• How many? Of these 2000 mammograms, ten women will be overdiagnosed and undergo unnecessary surgery/chemotherapy/radiation therapy.

Or you’re a man and your doctor suggests you have a PSA test to screen for prostate cancer. Would you bother if you knew that the same number of men die of prostate cancer who’ve never had their PSAs measured as those who have a high PSA and had their prostate removed? Read that carefully: you’ll have the same outcome with prostate cancer if you never ever have your PSA measured or you have it measured, find it elevated, and have full treatment with surgery, radiation, and chemo.

How would these facts soften your enthusiasm for screening mammograms, PSAs, and other cancer screening procedures?

Is there an acceptable limit to overdiagnosis?

That question, as an online survey, was sent out to 317 men and women between the ages of 50 and 69 (the main years for undergoing cancer screening) and reported on in JAMA Internal Medicine, accompanied by a thoughtful commentary. None had a previous cancer diagnosis. Indeed, a high percentage of the 317 had been having regular cancer screenings (CSs), as follows:

• 19% had one CS (usually a regular mammogram or Pap smear).
• 36% had two CSs.
• 27% three or more CSs.
• 17% had had no cancer screenings whatsoever.

Of these 317 garden-variety patients, fewer than 10% had been told by their doctors of the possible risks involved with any particular screening test. And a full 80% of those surveyed would have liked to have heard those potential problems described. Of course there will always be patients who want a certain cancer screening despite the risks of overdiagnosis and overtreatment. Again, more survey answers: 51% of responders said they’d stop the specific cancer screening on themselves if they knew that for every life saved there would be one person unnecessarily overtreated (with the possibility that they could be the one person).

Predictably, as the number of patients who’d be overdiagnosed and overtreated increased, enthusiasm for the test dropped considerably.  Nearly 70% reported they’d have serious second thoughts about screening if they knew the risks were high. However, a full 58.9% would continue their screening even if they knew that for every life saved there might be ten overdiagnosed and overtreated—the current situation with mammograms.

Obviously what’s needed to help a person make an informed decision about participating in a cancer screening is someone who can clearly describe the risks and benefits. And here’s the real crux of the problem: doctors themselves don’t know.  Surveys among physicians reveal that almost two thirds are clueless about the potential dangers of cancer screenings.

The most common answer from physicians was a variation of the previously mentioned “Saving just one life makes screening worthwhile,” and as emotionally charged as that may sound, the statistics simply don’t justify it. And, by the way, try explaining that to the families of patients who died during unnecessary surgical procedures, chemo, and so forth.

The hot new screening test that has me a little worried because of overenthusiasm for it in medical journals is having high-risk smokers (pack a day for 30 years) between ages 55 and 79 undergo lung CT scans. Although the phrase “could save about 20,000 lives annually” sounds impressive, it’s more complex than that. CTs of chronic smokers can be quite difficult to interpret due to years of smoke damage and/or previous lung infections. An abnormal mass seen on a scan needs to be confirmed as cancer via biopsy (not exactly a gentle procedure). If the biopsy results are positive, the patient, likely with significantly compromised breathing–remember, he’s a smoker–will need major lung surgery. Despite the fact that CT scans expose the patient to a lot of radiation, the amount received during 24 years (55 to 79) of annual CT scans, which increase cancer risks, was felt “acceptable.” And the risks operating on a 79-year-old who has smoked not 30, but more like 60, years were also found acceptable.

At this early stage, we really don’t know how many of these abnormal CT scans will lead to unnecessary surgery (and, yes, some lives might be saved), but the cost to those who have an abnormal yet cancer-free CT scan could be grim indeed.

Like I said, it’s complicated.

Be well,
David Edelberg, MD