close

click image to close
Home / Cancer / Early Detection

Early Cancer Detection Isn't Always an Advantage

Font Size: (-)(+)
 
August 26, 2005; Page B1

When Richard Bloch, co-founder with his brother of H&R Block, died of heart failure in 2004 at age 78, he was a medical success story. In 1978, he was diagnosed with terminal lung cancer. A decade later, he had colon cancer. He beat both, and went on to found a cancer hotline, a survivors group and other services. He was counted as someone whose cancer was detected early enough to save his life.

Should he have been? Nothing has greater intuitive appeal than the claim that cancer screening leads to early detection, which leads to longer survival. Whether it is the PSA test for prostate cancer, mammograms, endoscopy for colon cancer or -- in the wake of Peter Jennings' untimely death -- X-ray screening for lung cancer, intuition screams that the earlier a cancer is caught, the better the odds that you'll be alive in five years. Like Mr. Bloch.

Cancer researchers are now augmenting that intuition with data, and the result isn't pretty. The impact of cancer screening "on reducing cancer mortality," says Elaine Jaffe of the National Cancer Institute, "still isn't proven for a number of cancers."

How can that be? Part of the answer is that many tumors are so slow to progress -- indolent, scientists call them -- that they'll hang out in an organ for decades with no ill effects. "Early on, the idea of an indolent tumor was just a theoretical construct," says Barnett Kramer of the NCI. But indolence was found to characterize many neuroblastomas (a cancer of the nervous system), "and then it was found that tumors in the prostate, lung and now breast can also be indolent."

That doesn't mean cancer screening is useless. Without question, some of the tumors it finds would, if left untreated, have killed patients before their time, and some of the improvement in survival rates after breast cancer likely reflect earlier detection. But you can be misled into attributing the decades of life you enjoy after "beating" cancer to early detection and treatment rather than to the properties of the tumor itself.

Left to its own devices, the tumor might well have left you alone until you died of something else entirely. "Overdiagnosis of cancer as a result of screening is the rule rather than the exception," says Dr. Kramer. This overdiagnosis isn't the false positives that tests such as mammograms can spit out. In that case, what is detected might look like cancer, but on further examination is not. False positives cause great anxiety and cost, as patients undergo more tests. But diagnosing an indolent cancer is arguably worse, as patients undergo treatments that often have debilitating, even dangerous, side effects.

Overdiagnosis has another effect: on perceptions of progress in the war on cancer. More-sensitive screening means tumors are detected at ever-earlier stages. Let's say that, as a result of such a screening, a patient begins treatment on Aug. 26, 2005. She does well, and celebrates her five-year survival on Aug. 26, 2010.

If she succumbs to a recurrence or a spread of her initial cancer in, say, 2015, she still counts as a five-year survivor. But if she had a slow-growing cancer she might have made it to 2015 anyway, without early diagnosis and treatment. She is scored as a victory for cancer warriors, but in fact they didn't buy her a single extra day of life. All she got was more years knowing she had a dreaded disease.

"The improvement in long-term mortality may be due to the higher proportion of small or slow-growing tumors being detected, which means you start counting earlier," says Dr. Jaffe. That's why longer survival, measured from the time of diagnosis, is a misleading measure of progress against cancer, and no substitute for reductions in mortality.

The more scientists study cancers, the more indolent ones they discover. Researchers in Japan, for instance, find that CT scans detect almost as many lung lesions in nonsmokers as in smokers. But since nonsmokers have a mortality rate from lung cancer less than 10% that of smokers, the vast majority of what CT scans picked up would never have progressed to anything life-threatening. And a Mayo Clinic study found that although X-rays detect lung cancers at earlier stages, and lead to more five-year survivors, early detection does not lower death rates.

For colon cancer, the fecal occult blood test "does decrease your risk of dying of this cancer," says Dr. Kramer. "But for colonoscopy and sigmoidoscopy, which appeal to our intuition [about early detection], the evidence is not great." They pick up polyps earlier, but not all polyps become cancers, "and we don't know what proportion would lead to death."

The Pap test for cervical cancer has saved lives, but many of the abnormal cells it finds wouldn't go on to become cancer. Most women with low-grade or even high-grade lesions would have been fine anyway. Similarly, the PSA test for prostate cancer picks up tumors that are biologically nonaggressive.

The discovery that many tumors are innocuous casts doubt on the value of new screening tests. "You may fool yourself into thinking a test is twice as sensitive," says Dr. Kramer, "but the only extra cancers it picks up are those that wouldn't have harmed the patient.