Shots - Health Blog
Wed October 26, 2011
Looking For Lung Cancer With A Yearly X-Ray Doesn't Reduce Deaths
The drive to test healthy people for common cancers rests on the idea that finding malignancies early can trigger life-saving treatment. But the evidence that some of the tests will actually reduce mortality is sometimes lacking or is less than clear cut for the people who'll get tested.
Now, a big federally funded study finds that yearly chest X-rays to look for lung cancer aren't worth doing because they don't save lives. The findings were just published online by JAMA, the Journal of the American Medical Association.
Some previous studies of regular chest X-rays for screening suggested they weren't very helpful, though the evidence wasn't conclusive. Now there's an answer: Don't bother.
In the latest study, more than 154,000 people, ages 55 to 74, got either four annual chest X-ray or regular medical care without a screening X-ray. Patients were checked for as long as 13 years after getting the X-rays.
Among the more than 77,000 people who got X-rays, 1,213 died from lung cancer. For those in an equal-size group who didn't get the screening test, the death toll was 1,230. Statistically, there was no difference.
The types of lung cancer were pretty much the same in both groups.
An accompanying editorial says the study "provides convincing evidence that lung cancer screening with chest radiography is not effective. The study is important for putting this question to rest...."
Now is there another way to find lung cancer that would make a difference? Recent findings from another big study suggested that low-dose CT scans could help detect cancers in older people with a history of smoking versus a chest X-ray.
The approach remains controversial, especially as some hospitals and imaging centers tout inexpensive scans to all comers.
Update 4:19 p.m.: I asked the American Cancer Society for comment on the findings. In an emailed statement, epidemiologist Robert A. Smith, director of cancer screening for the group, had this to say:
Back in the late 1980s there was general agreement that the lung cancer screening trials that were done in the 1970s had serious methodological limitations....
All in all, while the results were null, there were a long list of concerns that led to a lack of confidence in the findings. [National Cancer Institute] investigators thought there was a real chance that there was a true benefit of 10-15 percent..., so they included a lung cancer screening arm....
Keep in mind that the NCI funded most of the trials in the 1970s. So, the goal here was to try and do a trial that would produce results that would stand up to scrutiny. It is true that many people thought the question had been resolved by the old trials, but they really were a bad lot, which is why the NCI was prepared to address the question one more time.
I think what we observed in the run up to use of spiral CT was an attempt to cast doubt on the value of lung cancer screening with any technology, which meant dressing up the old trials with new credibility.