New research suggests that the use of computed tomography (CT) in lung cancer screening may not reduce deaths from the disease and may expose some individuals to invasive and unnecessary treatments.

CT technology has generated considerable interest as a screening tool because it can detect very small growths in the lungs of current and former smokers. Two large, ongoing randomized studies - the NCI-sponsored National Lung Screening Trial (NLST) and the NELSON trial in the Netherlands - are evaluating whether CT scans can save lives by detecting cancers before they become incurable.

A study in today's Journal of the American Medical Association (JAMA) addresses this question using data from CT screening studies at the Mayo Clinic, the H. Lee Moffitt Cancer Center, and the Instituto Tumori in Italy. The analysis included more than 3,200 asymptomatic individuals who had smoked for an average of 39 years.

Because the studies lack control groups, Dr. Peter Bach of Memorial Sloan-Kettering Cancer Center and his colleagues used statistical modeling to create artificial control groups. They then compared the results from screening with what might have been expected in the absence of screening.

Screening led to a 3-fold increase in the number of lung cancers diagnosed and a 10-fold increase in lung cancer surgeries compared with what was expected without screening. Screening did not save lives and led to additional testing and treatments for growths that may never have caused harm, the researchers found.

"CT screening is an experimental procedure with numerous potential downsides," says Dr. Bach. These include exposure to radiation, anxiety from false-positive results, and harms caused by detecting and treating an indolent disease.

Participants in the study received an initial CT scan and at least three subsequent exams; they were followed for 5 years.

"The study found no decrease in the number of advanced cases of lung cancers or lung cancer deaths, and that's what is really alarming about the findings," notes Dr. William Black of Dartmouth-Hitchcock Medical Center, who co-authored an editorial in JAMA.

The results will be compared with findings last October showing that CT screening resulted in a 10-year survival rate of 88 percent for patients with stage I disease. Reporting their findings in the New England Journal of Medicine, the investigators suggested that CT screening in high-risk individuals could prevent 80 percent of lung cancer deaths.

The huge contrast between the two studies reinforces the idea that randomized clinical trials are needed to understand what is going on, says Dr. Black. The most likely reason for the discrepancy was that the first study used survival as the primary measure of outcome while the second used mortality, he says.

"The new study is well done and critically important," says Dr. Christine Berg of NCI's Division of Cancer Prevention, who is co-leader of NLST. "The findings underscore the importance of a prospective trial to address this question with mortality as an endpoint."

The editorial suggests that while rigorous cancer-screening trials are expensive and time-consuming, they are cost-effective compared with the broad adoption of expensive screening interventions that cause more harm than good.

Dr. Bach is awaiting the NLST results, but he went into the study hoping and expecting that CT was going to work. "We were all very disappointed because, like everyone else, we want to have a solution to this terrible disease," he says.

By Edward R. Winstead