Is that enough of a difference to save lives? By June, government-funded scientists will begin enrolling 50,000 current and former smokers in a massive study to find out — research that proponents hope will point toward a screening test for the nation's top cancer killer.
"It's an exciting time," says Dr. James Jett of the Mayo Clinic, one of the test sites.
"It's the most important screening trial" for any cancer in the works today, adds Dr. Harmon Eyre of the American Cancer Society.
More than 150,000 Americans die of lung cancer every year. It's one of the most dismal cancers: Only 15 percent of sufferers live five years, mainly because they're diagnosed very late, after symptoms begin.
But catch a lung tumor when it's still small enough to cut out, and the survival rate jumps beyond 50 percent. The question is how to find them.
Neither the government nor cancer specialty groups recommend an annual chest X-ray, because studies during the 1970s concluded that X-ray screening didn't save lives.
The screenings must not have caught the deadliest tumors early enough to make a difference, doctors thought — or maybe over-diagnosed less dangerous ones, putting patients through risky treatments for little reason.
Now three new developments are putting the long-taboo issue of lung screening back on the table:
A reanalysis of one of those 1970s X-ray studies, published in last week's Journal of Clinical Oncology, argues that regular X-rays could help after all.
The original Mayo Lung Project compared X-rays every four months with no regular screening, and found no difference in lung cancer deaths. But Dr. Gary Strauss, a Rhode Island oncologist who has long criticized the study's interpretation, contends a statistical flaw that somehow lumped 30 percent more cancer patients into the screened group skewed the death comparison.
Strauss says the key instead is long-term survival — something 29 percent of the screened group achieved, and just 13 percent of the others.
"Chest X-ray is far better than nothing," he says. But he has yet to persuade many other cancer experts.
Still, X-rays are being considered anew for another reason. Smokers today get a different form of cancer, called adenocarcinoma, in a different part of the lungs than they did in the 1970s, before filtered cigarettes were popular. Could modern X-rays spot these tumors better?
A National Cancer Institute project now is studying 140,000 adults, comparing annual X-rays to no screening, to see. Preliminary results are expected in 2005.
But the big excitement surrounds the new spiral CT scans.
Patients lie in a rotating scanner that uses X-rays to image the chest from all angles. Hold a breath; it's over in 20 seconds, with slightly more radiation than an old-fashioned X-ray. A computer then assembles the images into a three-dimensional model of the lungs.
Early studies suggest tumors found by spiral CT average half the size of those spotted on a regular X-ray. Cornell University's Dr. Claudia Henschke, a prominent proponent, has estimated the new method could find about 80 percent of tumors when they're in the earliest stage.
Don't race to get one just yet, Eyre said. Although some hospitals already aggressively advertise spiral CT, it's vital to prove the method really saves lives before testing becomes widespread, he said.
Why? The test can mistake scar tissue from an old infection or some other benign lump for cancer. No one knows how high this false-positive rate is, but in one Mayo study half of baseline scans detected "incidental nodules."
That's important because even a biopsy can be riskier for lung cancer than other cancers — the needle can collapse a lung. Doctors could watch whether a suspicious spot grows instead of immediately biopsying, but no one yet knows how long to wait.
Plus, unless spiral CT proves itself, insurance won't begin paying the $400 cost. And even these smaller tumors have lurked in the body a long time, notes Jett, who predicts scientists eventually will develop blood or sputum tests that could find cancerous cells far earlier.
Spiral CT is "a new tool, it allows you to detect it early, and the question is at what cost," he said. "Would I recommend this for the masses right now? No. ... I would be an advocate of people going in the trial."