Low-dose helical CT scans reduced death from lung cancer by 20% among high-risk patients compared with chest x-ray -- the first time a large, randomized controlled trial has found a benefit for lung cancer screening, the National Cancer Institute announced today.
The agency reported early results from the long-awaited National Lung Screening Trial (NLST) during a press conference Thursday, after its independent data and safety monitoring board decided that the benefits were significant enough to halt the trial.
Based on the trial results, physicians would have to screen 300 high-risk patients with CT to prevent one lung cancer death, according to Harold Varmus, MD, director of the NCI.
"We have the potential to save many lives of those at greatest risk for lung cancer," Varmus said during the briefing. But he warned that the message shouldn't be interpreted as a free pass to start or continue smoking.
"Screening does not prevent lung cancer, and it does not protect the large majority of subjects from death from lung cancer," he said. "Not smoking and quitting smoking remain … the best defenses against lung cancer."
Lung cancer is the leading cause of cancer-related mortality in the U.S. and worldwide. Only 15% of patients can expect to be alive five years after their diagnosis.
Screening has been controversial because some studies have found no mortality benefit, while others that saw an improvement in detection with low-dose helical CT (also called spiral CT) had methodological limitations.
So in August 2002, the NLST was launched to clarify the issue. The researchers enrolled 53,454 current and former heavy smokers ages 55 to 74. These high-risk patients were assigned to receive three annual screens with either low-dose helical CT or standard chest x-ray.
All participants had a history of at least 30 pack-years, and were either current or former smokers without signs, symptoms, or a history of lung cancer.
As of Oct. 20, 2010, there were a total of 354 deaths from lung cancer in the CT group, compared with 442 in the chest x-ray group.
That amounts to a 20.3% reduction in lung cancer mortality -- a finding that the study's independent data and safety monitoring board decided was statistically significant enough to halt the trial and declare a benefit.
Varmus said that a validated approach that can reduce lung cancer mortality "by even 20% has the potential to spare very significant numbers of people from the ravages of this disease."
The early results also showed that all-cause mortality was 7% lower in the CT screening group.
Historically, the results of lung cancer screening trials have been less decisive. In 2006, the I-ELCAP trial, led by Claudia Henschke, MD, PhD, of Weill Cornell Medical Center in New York and published in the New England Journal of Medicine, found that screening could prevent some 80% of deaths from lung cancer in an at-risk population.
But researchers were quick to point out that study's lack of a control group and its inability to find a reduction in mortality.
Henschke's study also received criticism because it had been funded with undisclosed tobacco company money.
A 2007 study, led by Peter Bach, MD, of Memorial Sloan-Kettering Cancer Center in New York and published in the Journal of the American Medical Association, found that spiral CT diagnosed three times as many lung cancers as predicted and resulted in 10 times as many operations.
However, it didn't decrease mortality or advanced stage cancers. It also did not include a control arm.
Bach told 鶹ý and ABC News that the NLST findings of reduced mortality "constitute a step forward, but are not a panacea for this dreadful disease."
"We don't yet know what problems screening caused for individuals who ended up not being found to have lung cancer, or why 80% of the people who died of lung cancer were not helped by screening."
The trial did, in fact, appear to have a high false-positive rate, although one that is comparable to other screening procedures.
Of all screens, spiral CT detected 24.2% positive scans, compared with only 6.9% positive scans for x-ray. But of those, only 3.6% of the CT positives actually turned out to be cancer, the corresponding percentage for x-ray was 5.5%.
The subsequent issue is an increased number of unnecessary tests that involve additional doses of radiation or risky surgical procedures, particularly for this high-risk, unhealthy population.
Christine Berg, MD, of the National Cancer Institute, said radiation is not a large concern for initial screens because the doses used in the trial are lower than that used in standard mammography (1.7 mSv versus 2.5 to 3.5 mSv).
Varmus pointed out another caveat -- the fact that some patients who do have very small or early cancerous lesions removed may not have died of those cancers anyway.
Lee Green, MD, MPH, of the University of Michigan, took issue with the fact that the trial was stopped when it reached the numbers that it had been statistically powered to detect.
"Stopping a study when the results reach significance, rather than at a specified preplanned time, tends to overestimate benefit," Green said in an e-mail to 鶹ý and ABC News. "Statistics naturally fluctuate up and down, and if you declare victory when you reach a certain height you're often just catching the peak of a fluctuation. The real average is lower."
But Varmus explained that scans had only been done during the first three years of the trial, so adding additional time to the study, when no scanning was being done, would inaccurately dilute the results.
Weill Cornell's Henschke told 鶹ý that the mortality benefits would likely continue if screening were continued.
"If screening had continued for 10 years rather than three," she said in an e-mail, "the mortality reduction would eventually be what is estimated by the curability rate we published in the NEJM."
The researchers are also working on a cost-efficacy analysis. Douglas Lowy, MD, deputy director of the NCI, said Medicare and most insurance carriers do not reimburse for screening CT for lung cancer. Some reimburse for diagnostic CT, but at that point, it is usually too late in disease stage.
The findings are still far from policy implication, as several groups -- including the American Cancer Society, the American College of Radiology, and the U.S. Preventive Services Task Force (USPSTF) -- will eventually take the findings into account when they draft guidelines.
On behalf of the American Society for Clinical Oncology, Bruce Johnson, MD, of the Dana-Farber/Harvard Cancer Center in Boston, gave a barometer reading for how cancer groups may interpret the findings when writing their guidelines.
He said the findings show lung cancer screening in high-risk patients can "save nearly as many lives as the number of people who die from breast cancer per year. We as a medical community now need to figure out how to do this in a way that the cost is acceptable to the public."
This article was developed in collaboration with ABC News.