By EVELYN MALDONADO
Dartmouth Medical School professor Harold Sox has received national attention for his analysis of the two studies funded by the National Cancer Institute that evaluated the effectiveness of X-rays and CT scans in screening cancer patients. Sox, a former Annals of Internal Medicine editor, was chosen to provide an expert opinion on the research data due to his experience in evaluating evidence and his current lack of involvement in cancer research, he said in an interview with The Dartmouth.
The first study, entitled "Prostate, Lung, Colorectal and Ovarian Cancer Study," tracked a range of high-risk smokers and low-risk non-smokers to examine if annual chest X-rays over a three-year period influenced the subjects' survival rate, according to Sox.
The National Lung Screening Trial also conducted its own study of former heavy smokers aged 55 to 74 to determine the effectiveness of annual CT scans compared to the effectiveness of annual chest X-rays over a three-year period.
Sox assessed the studies' results and concluded that both findings indicated that X-rays were not effective in reducing the death rates of cancer patients. The second study, however, did observe a 20-percent lower death rate for patients who received low-dose CT scanning rather than with chest X-rays, according to Sox.
The purpose of the two studies was not to try to prevent lung cancer, but to improve methods of locating cancer while it is still treatable, according to Sox. Previous trials lacked a control group of individuals who did not receive X-ray screening, and therefore yielded inconclusive results when attempting to determine chest X-rays' effectiveness, Sox said.
"Despite the previous evidence, there was reason to think still the chest X-rays might be effective," Sox said. "But they found out they weren't."
In the "Prostate, Lung, Colorectal and Ovarian Cancer Study," researchers ensured that only a small proportion of the control group 10 percent of individuals studied received chest X-rays, creating a substantial discrepancy in X-ray exposure between the test subjects and the control group, according to Sox.
Both studies found that chest X-ray screenings led to little difference in death rates for high-risk patients, according to Sox.
"Screening didn't have an effect on the whole population or the high-risk members of the population," Sox said.
Sox stressed that even the two recent trials have shortcomings. In both studies, most of the deaths from cancer were deaths that occurred after the screening was completed, he said.
"Chest screening for three years doesn't have much of an effect, and if screening would work at all it would be for every year in the patient's life," Sox said. "A negative test doesn't seem to lower your chance of developing cancer over the next year. Lung cancers seem to be a really nasty cancer that can go from being undetectable to being detectable and even fatal in as little as one year."
Overdiagnosis was another problem common to both studies, Sox said. CT scans are more effective than chest X-rays in picking up smaller tumors, as CT scans have the ability to find tumors that are as small as a quarter-inch in diameter, according to Sox. In a small fraction of cancer patients, the detected tumor either stops growing or goes away completely, Sox said.
"As a result, some patients are getting treatment, often big surgery, for cancers that were never going to get any bigger than when they were first discovered," Sox said. "If we were really smart, we'd be able to tell which cancers will be able to keep growing and which ones will be able to stop."
The results of the study are not expansive or conclusive enough to affect the cancer treatments at the present time, according to Sox. Further analysis of the data, including the costs and benefits of the screening, is still needed, Sox said.
"Usually practice doesn't change until more evidence is in and it's been examined closely by groups like the United States Preventative Services Task Force and the American Cancer Society," Sox said. "I don't expect to see recommendations from these two organizations for at least a couple years."
