The much-anticipated manuscript from the NCI-sponsored National Lung Screening Trial (NLST) was just published on line in the New England Journal of Medicine, with an editorial by Dr. Harold Sox. As Dr. West outlined in a previous post, we've known since last year that this trial demonstrated an improvement in lung-cancer specific survival with low-dose spiral CT screening of high-risk individuals. However, we have not had access to the details of this study until now. One of the fundamental problems we face is that lung cancer is most often diagnosed in an advanced stage. This has generated intense interest in screening and early detection. However until this trial, no screening test has been shown to reduce the risk from dying from lung cancer, which is the benchmark we use to judge the effectiveness of any screening modality. As the first trial that shows lung cancer screening can save lives, the NLST will no doubt have a significant impact on how we practice in this country and should be viewed as a very hopeful result for lung cancer advocates. However, many questions remain and as is the case with any medical intervention, individuals considering low dose CT screening will need to weigh not only the potential benefit (as defined for the first time by this trial), but also the potential risks associated with this approach. The NLST will go a long way towards informing this decision.

The trial was started in 2002 and randomized 53,454 high-risk individuals to three annual screenings with either a low-dose CT scan of the chest or chest x-ray. The individuals who were enrolled were between 55 and 74 years of age, current or former smokers (who quit within the previous 15 years), and had a cigarette smoking history of at least 30 pack-years (average number of packs of cigarettes smoked per day x the number of years smoking). The screening took place in 33 centers across the country. Most centers were university-based. All scans were interpreted at the screening centers by NLST-approved Radiologists. Although specific practice guidelines were not mandated, some NLST centers developed guidelines for subsequent evaluation of a positive screening test and both the results of the test and recommendations for further work-up were provided to the patient and their health care provider.

The trial overall was very well done. The adherence to the screening protocol was high for both groups. During the screening phase, 39% of participants in the low-dose CT group and 16% in the chest x-ray group had a positive result. The most common diagnostic evaluation was further imaging rather than invasive diagnostic procedures. The rate of complications after a diagnostic evaluation for a positive screening test was low (1.4% and 1.6% in the CT- and chest x-ray groups, respectively). Overall, 96.4% of the positive CT scans and 94.5% of the positive chest x-rays were false positive results (no malignancy identified). Over the 3 years of screening and 3 additional years of follow up, there were 1060 lung cancers diagnosed in the CT group and 941 in the chest-xray group. Most cancers were diagnosed during the screening period, although in both groups some cancers were diagnosed despite a negative screening test or after the screening phase was over. Most screen-detected cancers were early stage, particularly in the CT-group and predominantly adenocarcinoma histology.

Most importantly, there were fewer deaths in the CT group (356) compared to the chest x-ray group (443), which translated to a relative reduction in the rate of death from lung cancer with low-dose CT screening of 20%. The number needed to screen with a low-dose CT scan to prevent one death from lung cancer was 320.

So, should all individuals who have the same characteristics as the participants in the NLST trial undergo low-dose CT screening? Although the risks associated with CT screening on this trial was low, there are a number of important considerations for an individual before deciding to undergo this test (I'll get to the policy implications below). The first has to do with the high rate of false-positive results. Overall almost 40% of patients on the CT arm were told they had a positive screening test, the overwhelming majority of whom did not have cancer. In the Midwest, where I practice, that number can be even higher. The current use of newer, more sensitive CT scans then were evaluated in this trial may also increase the risk for false-positive results (although may also detect more cancers). Also, there remains a concern about the potential for "over-diagnosis" with low-dose CT screening. This has to do with the possibility that low-dose CT screening may be more likely to detect cancers that would not have become symptomatic or impact on the life of the individual patient (i.e., identifying slow growing cancers but missing faster growing and more biologically aggressive cancers). The implication is that some patients diagnosed with an indolent cancer may end up undergoing an invasive intervention that would not otherwise be necessary. An earlier screening trial conducted by the Mayo Clinic suggests that over-diagnosis is a problem in lung cancer screening. Unfortunately, it will take many more years of follow up of the NLST data to know how much over-diagnosis was present in this study.

There are a number of other issues that limit the ability to generalize these results. Although some of the centers participating in the NLST were community-based, most were academic university centers with established thoracic oncology programs. Also, while a specific guideline to evaluate positive screening tests was not mandated, several centers did develop guidelines for the diagnostic follow up that was communicated to the patient and their treating physician. It remains uncertain, therefore, whether these results can be re-created in centers without the necessary infrastructure and guidelines or standard operating procedures in place to safely conduct low-dose CT screening.

There are also a number of issues that policy makers will need to work through before current guidelines are updated. The authors themselves argue against basing changes in current guidelines on the NLST data alone. The question remains which individuals should be screened? According to the authors, there are currently 7 million individuals in the US who meet the entry criteria for the NLST. However, there are currently 94 million current or former smokers in the US. How should individuals with a lighter smoking history be approached? How does family history impact on this decision? What about never-smokers? Are three annual scans enough? The fact that the number of cancers diagnosed every year during the screening phase in the low-dose CT group was the same, suggests that a negative result does not necessarily predict lower risk of lung cancer in the future. Cost will also need to be addressed, which includes not only of the cost of doing an annual low dose CT scan but also the "downstream" costs of false-positive test results. Cost effectiveness analysis of this data is planned and will be helpful in addressing this question. Lastly, there may also be long-term implications for prolonged radiation exposure from low-dose CT scans of an entire population, which has not yet been addressed in this setting.

So, what is the bottom line? The NLST is the first trial that shows that lung cancer screening can save lives. This is a very positive and hopeful result for all lung cancer advocates. The potential risks of screening have been described previously and are supported by the NLST. However, both individuals who are concerned about their risk of lung cancer and their doctors have not had evidence of the potential benefit of low-dose CT screening to weigh against these risks. The results of the NLST will now help inform that discussion. It will be a while before policy makers will make formal recommendations and it remains to be seen whether low-dose CT scans will be covered by insurance. In the meantime, patients and their physicians should get to know the results of the NLST so that they can have a balanced discussion of risks and benefits before deciding how to proceed (ideally in centers who have the necessary infrastructure in place). Additional trials are ongoing that will hopefully address many of the questions listed above. Moving forward, research will continue to focus on the identification of additional factors that may help guide who is most likely to benefit from lung cancer screening. However, one can argue that the results of the NLST have changed the discussion from asking if lung cancer screening is effective to who should be screened and what is the best approach.