Lung cancer kills more Americans than breast, prostate and colon cancer combined. But unlike those other cancers, there's no agreed-upon screening test to catch lung cancer when it's early and can be curable. A study in this week's New England Journal of Medicine may help change that. But experts disagree about when the evidence will be strong enough to recommend that people at risk of lung cancer get the screening test, called spiral CT.
Spiral CT is a computerized X-ray scan that takes hundreds of detailed pictures of the chest in the space of three seconds. It can spot lung tumors smaller than a pea.
Seven years ago, a group at Cornell University published one of the first big studies suggesting that spiral CT might be used to screen the lungs of smokers and former smokers for signs of lung cancer, long before symptoms appear. Now the same group is publishing results on more than 31,000 patients from around the world.
Dr. Claudia Henschke, who heads the Cornell group, says spiral CT turned up nearly 500 lung cancers.
"The overall survival was 80 percent," says Henschke, "which is quite dramatic compared to what the rates are in the U.S. ... Only about five percent of the people who are diagnosed with lung cancer eventually are cured of it. So that's a tremendous improvement."
The results look even better among the great majority of patients in the study diagnosed with early-stage lung cancer. Based on results so far, an impressive 92 percent of them are projected to be alive ten years after diagnosis.
"I think our study gives compelling evidence that screening is useful and can be done responsibly now," Henschke says.
Henschke is the most prominent enthusiast for lung cancer screening. Others in the medical community are just as enthusiastic. Dr. James Mulshine of Rush University in Chicago, a veteran cancer-prevention expert who sits on Henschke's advisory board, calls the new results a landmark.
"In one fell swoop, you go from 15 percent, five-year survival to 92 percent, 10-year survival," Mulshine says. "That's amazing. Claudia's result is spectacular. The question is, is it generalizable? And should it be the basis of public health policy changes?"
Mulshine says the new results should force a public debate on that question.
"This is a very strong suggestion that lung cancer screening can conform to the kind of experience we've had with colon cancer and breast cancer and cervical cancer," Mulshine says.
Other experts urge caution.
"This study basically demonstrates that those individuals who have a lesion detected by CT, who are then diagnosed with early-stage lung cancer, do fairly well, which we've known for over 40 years," says Dr. Ned Patz, a radiologist at Duke University.
He says the study hasn't resolved whether individuals who are screened for lung cancer will fare better in the long run than individuals who aren't.
That's exactly the question that Dr. Denise Aberle of UCLA Medical Center is trying to answer. She heads a giant, federally financed study that will cost hundreds of millions of dollars. It has enrolled 53,000 smokers and former smokers. Half are getting annual CT scans, and half are getting annual chest X-rays.
Aberle says the study published this week can't determine whether spiral CT actually reduces lung cancer deaths in the long run because there was no comparison group that didn't get spiral CT.
"We know it prolongs survival," Aberle says. "We absolutely need to know whether or not the screening intervention reduces lung cancer mortality."
She says prolonging individual patients' survival isn't the same thing as making a big dent in the overall lung cancer death rate, and only studies like hers and others under way in Europe can nail down that question.
But results from those studies won't be available for at least three years. In that time, 480,000 Americans will die from lung cancer, and some experts think spiral CT might save some of those lives.
Still, there's not likely to be an official recommendation until further big studies are concluded.
For the 80 million Americans at risk of lung cancer who choose to seek a spiral CT test in the meantime, the American Cancer Society warns that there are risks from spiral CT. A positive result can lead to lung biopsies and major surgery for findings that may turn out not to be lung cancer.
The Cornell group found ways to reduce those risks. But everyone agrees that American doctors in general are a long way from being ready to deal with the complicated issues that arise from screening for lung cancer.