To Screen-or Not?
Questions to put to your physician before having a test
Corrected on 8/12/2008: An earlier version of this article incorrectly reported that neuroblastoma is a brain cancer. It's a neuroendocrine tumor.
Once cancer cells break away from their original home and take up lodging elsewhere in the body, the disease can often be managed but is rarely curable-so finding cancers before they spread and symptoms appear makes sense. Hence the emphasis on timely screening to sniff out the very first signs of disease.
But any screening test is a balancing act. On one side is the obvious benefit of detecting cancer in time to save a life. On the other are less obvious risks. The medical community is in the midst of hotly debating which tests are right for which people-and whether ferreting out every last sign of cancer is always good. "The idea that a good screening test is the one that finds the most cancers is totally wrong," says H. Gilbert Welch, a physician researcher at the VA Outcomes Group in White River Junction, Vt., and author of Should I Be Tested for Cancer? Maybe Not and Here's Why. A good screening test, he says, is the one that saves the most lives and harms the fewest healthy people in the process.
Some tests, such as the Pap smear and those for colorectal cancer screening, have cleared that barrier with room to spare. Others, such as the PSA test for prostate cancer and CT screening for lung cancer, are murkier. Your comfort level depends on a talk with your doctor about whether a given test is right for you. Here are the key questions to ask when a test is proposed or when you want to discuss one you've heard about.
1. Why should I have this particular test?
With no symptoms as guidance, who should have a test comes down to the risk of getting the disease. Cancer is most often a disease of aging, so logically, screening for common cancers such as breast and prostate can wait until midlife. Other risks may depend on family history, specific genetic patterns (mutations in the BRCA genes dramatically increase the chance of getting breast cancer) or lifestyle factors such as smoking or exposure to carcinogens. If a doctor recommends a screening test, it's presumably because she thinks you are at some risk, however small.
Understanding the nature and degree of that risk is vital, says Dartmouth Medical School radiologist William Black, an expert in cancer screening. To bring it home, you can say you want to know your chances of getting the disease or dying from the cancer for which you would be screened within a year. How about 10 years? A lifetime? Out of 1,000 people with a risk profile like yours, how many will get the disease? And how does the physician know that? "If the doctor isn't well versed in that kind of information, a request should be made to be referred to one who is," says gastroenterologist Moshe Shike, who heads the screening recommendation committee at Memorial Sloan-Kettering Cancer Center in New York.
2. Will the test lower my chance of dying?
You'd think the answer would always be yes, but it's not that straightforward. The most reliable and definitive way to know is to take a number of healthy people in the population who might be candidates for screening and assign them randomly to a screened or an unscreened group, then follow them to see which group has fewer deaths. But such randomized controlled trials are expensive, and because most cancers develop slowly, conclusions can't be drawn for years. Meanwhile, people who might benefit are not getting the test. So other kinds of studies are often performed instead that use different effectiveness measures, such as whether use of the test raises the average survival rate over a certain period.
What you really want to know, though, is whether a test saves lives by catching and permitting the treatment of cancers that would be deadly when detected months or years later. Take a tumor of the neuroendocrine system called neuroblastoma. It is the most common malignant tumor in children, so any and all efforts to find it early would seem to be worthwhile. A screening test was developed that examined infants' urine for certain chemicals produced by this cancer. The test was used in Japan for 20 years starting in the mid-1980s, but it was abandoned after researchers surveyed the accumulated data. Many more infants had been diagnosed with early-stage disease-but those same cases almost always had the biological hallmarks of the kind of neuroblastoma that has a high survival rate after it's detected because of symptoms. Many of the infants had had tumors that never would have caused a problem later. That means they were unnecessarily treated with surgery, radiation, or chemotherapy, all of which carry the risk of harm or even death. There was no drop in the incidence of the most dangerous cases. And using the test had no effect on mortality.
For some tests, no one knows whether having them will reduce deaths from the disease. The U.S. Preventive Services Task Force, an independent advisory panel that issues recommendations on preventive care, puts screening for lung, prostate, oral, and skin cancer in this category (box). For example, the task force says the PSA test for prostate cancer can definitely find early cancers, but whether that actually saves lives is unclear.
"Some people will say, gee, if you don't know the answers, maybe I shouldn't get [the test], while others will think they still may want it," says Russell Harris, a member of the task force and professor of medicine at the University of North Carolina. Laurie Fenton, president of the Lung Cancer Alliance, an advocacy group, is in the latter camp. The results of randomized trials of CT screening for people at high risk for lung cancer aren't due out for a few years, but she says people have the right to get the test if they understand the possible harm as well as the potential benefits. A possible factor: Your insurance may not cover the cost of tests that aren't recommended by government groups or the American Cancer Society.
3. What's the likelihood the test will miss the cancer or say I have it if I don't?
No screening test is perfect. Even mammography and colonoscopy can occasionally miss a cancer. Ask about the chance of such a false negative. Tests also exhibit false positives-identifying a nonexistent cancer. Usually a false positive is merely an annoyance, but not always. It ups the ante, often leading to follow-up tests to determine whether that suspicious spot on the film or screen is cancerous or benign or nothing at all. No man wants to go through an unnecessary prostate biopsy because of a worrisome PSA test. Certain follow-up tests, such as open lung biopsy, carry a small possibility of complications. And some tests have a high rate of false positives-about 11 percent of mammograms, for example.
How would you handle a false alarm? Would it keep you up at night for months or roll off your back? If this is an annual test, are you comfortable with the possibility of many false alarms over your lifetime? Again: Ask your doctor to put the risk in meaningful terms, such as the number of false positives per 1,000 people taking the test compared with the number of real cancers that will be found.
4. What other risks does the test pose?
The biggest is overdiagnosis-finding real cancers that don't need treating because they'll never amount to anything. That is especially common for melanoma as well as cancers of the prostate, breast, and thyroid, says Welch. "Even within a given cancer, they're not all the same," says Harris. "Some cancers grow really fast and they're bad, and others just sort of sit there and don't do very much." If screening reveals any cancer, there is considerable pressure to treat it. That is hardly pleasant and has risks of its own; chemotherapy, for example, can damage the heart. And picking up possibly harmless, slow-growing tumors is no guarantee of finding dangerous ones that grow quickly and might slip by between screenings.
For any test, the scope of overdiagnosis isn't always known. Some cases of ductal carcinoma in situ-cancerous cells confined to the milk ducts-progress to invasive cancer; others don't. It's the same for prostate cancer. Worse, medicine can't yet reliably predict whether a particular tumor will be dangerous. "We can't question a polyp and ask its intentions," says Robert Smith, director of cancer screening with the American Cancer Society. Given that uncertainty, it may be hard to calculate whether treating a potentially deadly cancer early will outweigh treating people who didn't need it. But knowing whether research suggests overdiagnosis may play a factor in your decision to have a test.
5. What will happen if the results are positive?
It could mean anything from a repeat test in a few months to a biopsy to more invasive surgery. "If you don't want to take the next step, why do the test?" asks Welch. Don't assume you'll figure out what to do when the time comes.
6. How and where should I have it done?
If you elect to go ahead with a test, take steps to see that it's done right, says Barnett Kramer, associate director for disease prevention at the National Institutes of Health. Wherever you're sent for a test, "it's useful to know how experienced the center is and what techniques are being used," he says. Then ask the doctor how those techniques stack up and what their track records are for false negatives and positives. A recent cautionary tale is computer-aided detection technology in mammography, which has become a hot tech upgrade at imaging centers because software highlights suspicious areas on the mammogram. A government-sponsored study released this month showed that CAD flagged more women for unnecessary biopsies and didn't catch more cancers. For every woman diagnosed with breast cancer using CAD, the researchers figured, 156 women had tests they didn't need and 14 had unnecessary biopsies. Newer is not necessarily better.
7. What are my options besides this test?
Ask your doctor what you can do to protect yourself besides getting a test now, advises Douglas Owens, an internist at the VA Palo Alto Health Care System and Stanford University. It could be to wait to do the test because the odds of developing breast cancer in your 40s don't outweigh your concerns about the downside of a mammogram, so you'd rather hold off until you're 50.
Or, if you've decided against a specific test for a particular reason-maybe you can't get over your squeamishness about a screening colonoscopy even if Katie Couric says it's not that bad-ask about other options. A fecal occult blood test, a CT or "virtual" colonoscopy, or a sigmoidoscopy (which doesn't reach very far into the colon) may be more personally acceptable. Any test for colon cancer is better than none. But since that maxim doesn't hold for every screening test, your decision to have one should be thoughtful rather than automatic.
Thumbs Up and Down
For which cancers is screening indicated for healthy people with typical risks? Which ones aren't? The following advice is taken from guidelines set by the United States Preventive Services Task Force and the American Cancer Society.
Colon: Several methods. The ACS has schedules for each.
Cervical: Pap smear advised for women who are sexually active; not recommended for women over 65.
Breast: Mammography benefits are strongest for women in their 50s, says the USPTF.
INSUFFICIENT EVIDENCE OR POSSIBLE HARM
Prostate: Not enough evidence for a routine PSA test, says the USPTF. The ACS says it should be offered but doesn't endorse it.
Lung, oral, skin: Neither group takes a position for or against testing.
Bladder, ovarian, pancreatic, testicular.
Note: Details are at www.ahrq.gov/clinic/cps3dix.htm#cancer and www.cancer.org/docroot/PED/content/PED_2_3X_ACS_Cancer_Detection_Guidelines_36.asp?sitearea=PED
This story appears in the April 23, 2007 print edition of U.S. News & World Report.