Cancer screening as we age

Does it make sense to get a mammogram if you're 80? A colonoscopy if you're 85? Experts are still sorting it out.

Experts have battled over whether women should start getting screening mammograms for breast cancer at age 40 or 50. Hit the half-century mark these days, and chances are that your doctor has a present waiting for you: a referral for a colonoscopy. It's a given that women will start getting Pap smears, the screening test for cervical cancer, when they turn 21 or even sooner, depending on when they become sexually active.

But now attention — and with it, controversy — is shifting to questions about how old people should be when they stop getting these screening tests.

Americans are now accustomed to getting screening tests regularly, so as they get older it's only natural that they'd wonder about how long they should continue getting these tests. And pressure seems to be building to push the age limits for screening tests into people's late 70s and 80s — even their 90s. There are a number of reasons for this. Americans are living longer, so there are more years of life to save if a treatable cancer is found. Cancer treatment is more tolerable than it used to be — especially when the cancer is caught early — so treatment is a more realistic option for older people.

Opening a Pandora's box

But expanding screening to include older people may also have some serious drawbacks. Many of the cancers we screen for become more common with age, so screening elderly people is bound to find more cancers. Cancer in older people often tends to grow and spread slowly, if at all. Oncologists refer to these cancers as "indolent tumors." So screening programs that include the elderly could mean diagnosis and treatment of a lot of cancers that aren't causing any real harm and weren't likely to in the future, both because of the nature of the cancer and limited life expectancy. When you factor in the extra testing and the false positives, we could be opening up a Pandora's box of medical intervention.

There are treatment issues, too. Notwithstanding the recent advances, if a serious cancer is found, it may be more difficult to treat in many older people because of other conditions they have. Just the practical issues of getting people to their appointments and having them adhere to complicated regimens will present some challenges.

Right now, there are more questions than answers about cancer screening in older people. Open up the research cupboard and the shelves are pretty bare. Most screening studies have enrolled few, if any, people over age 75. The good news is that this is a lively enough issue that it's now beginning to attract some research interest and dollars. The danger is that clinical practice will get ahead of the research, which has happened many times before with cancer screening.

Here is a brief rundown on age-related issues for four screening tests for important cancers:

Breast cancer

The American Cancer Society sets no age limits on mammograms, recommending that women get the exams every year as long they are in good health. The U.S. Preventive Services Task Force, whose recommendations are the closest thing we have to official government guidelines, says pretty much the same thing: women should get mammograms annually or every other year as long as their life expectancy is not limited by other diseases — a more precise way of saying as long as they're in good health. The American Geriatrics Society recommends that women over 75 get mammograms every two to three years as long as their life expectancy is more than four years.

In the absence of research that addresses the issue of mammography in older women head-on, researchers have conducted several studies that have examined the question in a roundabout way by categorizing older women who have been diagnosed with breast cancer by how often they got mammograms before their diagnosis. For the most part, the results have been "pro-mammogram." Researchers have found that women who have gotten mammograms regularly are more likely to have their cancer diagnosed at an earlier stage than women who have gotten mammograms infrequently or not at all.

In 2008, researchers at the M.D. Anderson Cancer Center in Texas reported results of an analysis of data on over 12,000 breast cancer patients ages 80 and over. Like other researchers, they found an association between regular mammograms and the cancer being diagnosed at an earlier stage. Quite reasonably, the association is presented as evidence that even women in their 80s stand to benefit from regular screening because cancers are being found at a more treatable stage. It falls well short, though, of the gold standard for judging screening tests, which is evidence that the people who are screened live longer than those who aren't.

The M.D. Anderson study did hint at a survival advantage, although such studies are plagued by so-called healthy user bias: women who go to the trouble of doing something like getting mammograms regularly may be fundamentally healthier than women who don't. Researchers use statistical techniques to compensate for this bias, but it's difficult to weed out entirely.

Despite these qualms, it's hard not to be impressed by the growing number of studies that suggest older women should get mammograms regularly. Dutch investigators presented findings in 2008 that showed women between the ages of 75 and 79 were less likely to die from breast cancer if they had continued to get mammograms regularly through age 75.

And yet part of the math of mammography — indeed, any screening test — is that it will generate extra diagnostic testing. The National Cancer Institute estimates that 85 out of every 1,000 women over age 65 who have a mammogram will be referred for diagnostic tests, and out of those 85, only nine will have cancer. The cost, inconvenience, and anxiety caused by the diagnostic tests must be weighed against finding and treating early-stage breast cancer.

Cervical cancer

Of all the screening tests, the Pap smear for cervical cancer is the one closest to having an agreed-upon age cutoff. The American Cancer Society says women who are 70 or older who have had three or more normal Pap tests in a row and no abnormal test results in the past 10 years can choose to stop having the test. The government's Preventive Services Task Force recommends that women older than 65 stop getting screened for cervical cancer if their recent Pap smears have been normal. The guidelines don't spell out what is meant by recent.

It's safe for older women with a history of normal Pap smears to stop getting screened for a couple of reasons. Cervical cancer originates with infection with the human papillomavirus (HPV), a virus that's spread through sexual contact. But it takes years, even decades, for cervical cancer to develop from the cellular abnormalities caused by the virus. A recent history of normal Pap smears is good evidence that those cellular changes haven't taken place, and there's little chance that they will in old age. Moreover, older women rarely contract new HPV infections. Even if they were to do so, a new infection isn't likely to develop into cervical cancer in their lifetime.

Colon cancer

Colon cancer screening in older people is largely a blank slate. The American Cancer Society and other groups are silent on the subject and don't set any kind of age cutoff. The American Geriatrics Society has a general position paper on screening tests with a warning that older people may be less willing to cope with the discomfort of screening tests like colonoscopy, but it makes no specific recommendations.

Although there are other ways to screen for colon cancer, colonoscopy has taken off since Medicare started covering it in 2001. With the heavy-duty laxative preparation needed to clean out the colon before the exam and the sedation used during it, there's much more involved in getting a colonoscopy than the other cancer screens. Studies have found colonoscopies done on older people (usually defined as ages 75 and up) take longer, are less likely to yield a complete view of the colon, and, most troubling of all, are more likely to result in perforations of the colon.

Several years ago, Columbia University researchers reported results showing that perforations are rare, with just 77 occurring among the 39,286 colonoscopies included in the study. That works out to 0.2%. But they also found that the risk of a perforation occurring was nearly four times higher for people ages 75 and older compared with the risk for those in the 65 to 69 age bracket.

Research findings have also sown some doubts about the benefit side of the risk-benefit equation. A study published in 2006 in The Journal of the American Medical Association found that colonoscopies are more likely to discover cancerous growths in people ages 80 and older than in younger patients. That is just as one would expect, since cancer becomes more common with age. But when the researchers did some calculations on what detecting those cancers meant for life expectancy, they found the gain, on average, was much lower in the older group than the younger groups. Why? In part because most of the growths found by colonoscopies have a long lag time before they develop into cancer, so many elderly patients are likely to die from other causes long before the growths become serious enough to be life- or health-threatening.

The American Journal of Gastroenterology published an interesting debate in 2006 on whether a hypothetical 88-year-old healthy woman should get a colonoscopy. The doctor who argued the affirmative, Dr. Robert E. Schoen of the University of Pittsburgh, noted that a very healthy 88-year-old woman has a life expectancy of eight more years, so catching and treating a cancer early might prolong her life, whereas it might not do so for a less healthy individual that age. It's an argument that underscores just how important life expectancy is to any calculation of a screening test's benefits, and that chronological age is just a single factor in that life-expectancy equation, albeit an important one.

Dr. Schoen also suggested that colonoscopies in older people could be made safer if doctors focused on removing polyps that were cancerous, or very nearly so, and left the noncancerous ones alone, on the theory that there is little chance that they'll develop into cancers within the person's lifetime.

Prostate cancer

Screening men for prostate cancer at any age is controversial. The government's Preventive Services Task Force sent out mixed signals in its 2002 recommendation, saying that the prostate-specific antigen (PSA) test can detect early-stage cancer but also leads to frequent false positives and unnecessary anxiety and biopsies. It concluded that the evidence is "insufficient" to decide whether the benefits outweigh the harms. The American Cancer Society has come down in favor of PSA testing and digital rectal examination for men starting at age 50, but it also hedged its bets. The recommendation is that the tests be offered to men, and only to those who have at least a 10-year life expectancy.

There's good reason to be even more skeptical about the value of PSA screening in older men. Many older men have prostate cancer that causes no illness. Autopsy studies have shown that up to 80% of men in their 70s have microscopic amounts of cancerous prostate cells. Furthermore, the forms of prostate cancer that the PSA test helps find in elderly men are often slow-growing. Many of the tumors wouldn't cause any symptoms for years — if ever. If men are screened, these cancers will be discovered and are likely to be treated.

PSA screening of older men is common. Doctors from the U.S. Department of Veterans Affairs reported results of a prostate cancer screening study that included over half a million veterans ages 70 and older. Over 36% of the veterans ages 85 and older had been screened for prostate cancer, yet only a fraction of them had the requisite 10-year life expectancy. Other studies have found that over half of American men ages 80 and older have had a PSA test in the past year.

A big part of the problem with the PSA test is the number of false positives. By one estimate, 75% of the men who have had a prostate biopsy because of a high PSA level (4.0 to 10.0 nanograms per milliliter) don't have cancer. The false positive rate among older men is especially high. More sophisticated ways of interpreting PSA results are being developed, but they all present problems.

Some experts have suggested that one way to improve PSA testing would be to start it earlier, between the ages of 40 and 45, and then set an age limit at 75, or even 65, if a man has a history of low PSA levels (0.5 to 1.0 ng/ml).