One of the biggest clinical trials ever run by the National Cancer Institute (NCI) came to an end today with hopeful but budget-threatening news. NCI Director Harold Varmus announced at a press conference that the National Lung Screening Trial (NLST) has been halted by a monitoring panel because it has strong evidence that lives can be saved by using high-tech x-ray imaging to screen for lung cancer. Preliminary data from the study, which enrolled 53,000 heavy smokers aged 55 to 74, show that deaths among those screened with low-dose computed tomography (CT) were 20% lower than among people who got a standard chest x-ray. The study, which began in 2002, did not investigate why the CT-screened people did better. Varmus said they presumably got early and effective treatment.
This is good news because no other method of screening for lung cancer has ever reduced deaths from the disease. Lung cancer is by far the biggest cause of cancer deaths in the United States, and is predicted to kill about 157,000 people this year. Worldwide, lung cancer deaths are about 1.3 million a year. If the NLST study is right, it points to a way to achieve "the greatest single reduction of cancer mortality in the history of the war on cancer," says James Mulshine, a cancer researcher associated with the pro-screening advocacy group, the Lung Cancer Alliance, and vice president for research at Rush University Medical Center in Chicago, Illinois.
But the trial results are worrisome, too, Varmus said. They're likely to place significant new demands on health care. The main concern, said NCI deputy director of basic science Douglas Lowy, is the huge cost. It's not just the price of each scan, which is estimated to be about $300. What concerns Lowy is the high rate of false positives likely to result from mass screening. About 25% of the scans in the NLST trial identified anomalies, but the vast majority of those were not dangerous cancers. Lowy and other experts worry that thousands of people will receive false alarms and thus undergo needless x-rays and biopsies. Investigating each false alarm will be expensive and potentially quite risky.
U.S. health agencies won't make any recommendations for individuals based on the NLST data right away, Varmus said: The first step will be to publish a final, peer-reviewed report on the study. Varmus expects it to be out within 2 or 3 months. Then review panels will decide who should consider getting a CT scan, when, and how often.