Historically, the classical teaching to physicians in training was that deaths from lung cancer could not be prevented by screening chest x-rays. As opposed to breast and colon cancers, where mammography and colonoscopy can reduce mortality, lung cancer prognosis could not be improved by screening because by the time a cancer was big enough to see on a chest x-ray it was probably incurable.
All that changed in 2006 with publication of the Early Lung Cancer Action Program (ELCAP) trial. Researchers at Memorial Sloan-Kettering Cancer Center in New York claimed an 80% reduction in cancer mortality in patients who underwent CT scans of the chest before symptoms of lung cancer developed. The survival of patients whose cancer was discovered at Stage I was 88%, much higher than previously seen. Many patients in this trial had abnormal CT scans resulting in lung surgery that turned up no cancer. Nonetheless, there was immediate excitement about these results and the medical community began to be swayed about the value of lung cancer screening.
Subsequent analysis, however, called these results into question. First, the study was uncontrolled (there was no group studied at the same time that did not undergo CT) and secondly it was revealed that the study was underwritten by the tobacco industry, a relationship not initially divulged. As a result of these perceived flaws the medical community declined to embrace CT scanning as useful in asymptomatic smokers.
The next chapter in the story was written only recently. The National Lung Screening Trial (NLST), started in 2002, before the ELCAP results were published, was designed to examine whether chest CT was better than plain chest x-ray in reducing lung cancer mortality. There was no group in the study that received no screening. More than 53,000 current and former smokers were enrolled.
Last November, the National Cancer Institute abruptly ended the trial ahead of schedule when their preliminary analysis showed that study subjects who underwent CT screening has a 20% lower chance of dying of lung cancer than the group who received only chest x-ray. It is relatively unusual for a large federally-funded clinical trial to be terminated prematurely but this was done for ethical reasons because the CT group was doing so much better. To put this in perspective, in the CT group there were 354 deaths from lung cancer and in the chest x-ray group there were 442, for a 20% reduction in mortality.
This may not seem like much with so many people in the study but the statisticians were impressed. Furthermore all of these people had some sort of screening versus no screening, which until now had been the community standard.
Whether the NLST trial will establish a new standard of care is uncertain. For anyone entertaining the idea of being screened proactively for lung cancer be aware that if you have smoked there is a substantial possibility (up to 25%) that your CT scan will show something and that this finding could lead to lung surgery that will turn up nothing. Surely the final chapter in this story has not yet been written.