boston.com your connection to The Boston Globe

Routine lung cancer CT scans don't save lives, study finds

More research is underway

LOS ANGELES -- Only five months after a major study recommended routine use of CT scans to detect lung cancer, a second study recommends the opposite, concluding that the scans do not save lives.

Both studies found an estimated 10-year survival as high as 90 percent among patients whose cancers were detected early. But the new study, reported today in the Journal of the American Medical Association, found no difference in the number of lung cancer deaths between the screened group and a control group.

"CT scanning is an experimental procedure, and there is no evidence that it is reducing the death rate," said epidemiologist Colin B. Begg of the Memorial Sloan-Kettering Cancer Center in New York, senior author of the paper. "It really should not be used outside of clinical trials."

The primary drawback to CT scanning is that it can detect very small tumors that may not present a risk of turning into advanced lung cancer. Nonetheless, patients may have to undergo surgery, which carries its own risks.

Dr. Claudia Henschke of the Weill Cornell Medical Center in New York, who led the first study, argued that the results from the two studies were very similar.

She maintained that Begg and his coauthors were simply interpreting the results "in the most negative way possible."

The conflict of opinion is reminiscent of the early days of mammography, when competing studies showed both benefit and lack of benefit, said Dr. Richard Wasley, director of imaging at Orange Coast Memorial Medical Center in Fountain Valley, Calif., who was not involved in either study.

"It took 30 years to show that mammography was useful," he said.

More than 200,000 Americans develop lung cancer every year and an estimated 163,000 will die from it in 2007, according to the American Cancer Society.

The goal of CT scanning is to find lung cancers while they are still at an early, curable stage.

Chest X-rays have previously been shown to be useless for lung screening because they do not detect small tumors. CT scans, which produce a sophisticated three-dimensional image of the chest using X-rays, can see much smaller objects.

Begg, Bach, and their colleagues at the two medical centers in the United States and one in Italy, studied 3,246 asymptomatic men and women who were at high risk because they had smoked or were still smoking.

In the three yearly screenings, they found about three times as many tumors as would be expected in the population, and performed 10 times as many surgical biopsies as would be expected, the team reported.

As a direct result of the tests, an additional 99 people were diagnosed with lung cancer and an additional 98 had surgery. But that intervention had no effect on the death rate, they reported.

There were 38 deaths due to lung cancer, and 39 would have occurred without screening, according to the report.

"Early detection and intervention did not save lives, but did subject patients to invasive and possible unnecessary treatments," Bach said.

Dr. David Johnson of Vanderbilt University, a former president of the American Society of Clinical Oncology who was not involved in either study, added that the new study "interjects a note of caution" into the debate over lung cancer screening.

But Henschke argued that Begg and Bach excluded some deaths in the control group that occurred in the first year of the study and unfairly included some deaths among the CT scanned patients. Those patients were already showing overt signs of lung cancer, such as coughing up blood, and thus should not have been enrolled as asymptomatic subjects.

Taking both of those into account, she said, the reduction in deaths was "possibly as much as 45 percent."

Begg countered: "At the end of the day, the studies are not contradictory. What is contradictory is the interpretation of the data. . . . We obviously believe our opinion is right."

Two large, randomized trials of CT screening are now underway to help resolve the issue. One is a US study sponsored by the National Cancer Institute; the other is taking place in the Netherlands. Results from those are expected as soon as 2009.

Clinicians hope those results will be "definitive," Johnson said.

SEARCH THE ARCHIVES