Defining the Right Population for Lung Cancer Screening: Why

[NikoleV]

Defining the Right Population for Lung Cancer Screening: Why Shouldn’t it be Everyone?

http://blog.lungevity.org/2012/10/20/de ... -everyone/

October 20th, 2012 - by Dr. Jack West

A commentary piece on About.com suggests that CT screening being restricted to a population with a significantly increased risk of lung cancer, such as people age 55-75 with a significant smoking history that were the subject of the the National Lung Cancer Screening Trial (NLST) published in the New England Journal of Medicine last year, only detects approximately a quarter of the lung cancer out there. This post implies that broadening CT screening efforts to a wider range of ages and smoking history would be appropriate in order to detect more lung cancer and improve outcomes. I need to clarify that while I consider myself a clear proponent of lung cancer screening for a defined population, it is a profound mistake to presume that screening more people is definitely better. It may well be worse. Why might that be?

The NLST documents a 20% improvement in overall survival in the screened population, who had a significant risk for lung cancer as defined by being old enough to have a greater probability of developing cancer but not too old to be especially likely to have competing serious health risks that limit the ability to tolerate treatments for lung cancer, and they also had enough of a smoking history (30 “pack-years”) that nodules found by screening would have a meaningful probability of representing cancer vs. a benign finding like inflammation, infection, prior scarring, etc., which is always a real possibility. We know that CT screening for lung cancer detects many lung nodules that are not cancer; this is associated with additional scans, a significant risk of pursuing invasive tests to chase down findings, and significant anxiety in those affected. There is also some potential risk, still not really well defined over the long term, from the radiation administered with scanning.

Screening for any cancer makes sense if you can focus on the people with a high enough probability of having what you’re trying to find that the benefit exceeds the risks. But there are risks, which aren’t just limited to the very real financial cost of doing CT scans tens of thousands of additional people every year. If screening is applied to a broader group with a lower probability of actually having cancer (such as looking at younger patients and those with a less significant smoking history, or even never-smokers), or not being fit enough to pursue treatment for a cancer detected early (such as a 79 year-old with significant heart disease), the more likely that the screening effort will confer negative consequences without the benefits. In other words, screening will be a lower yield proposition if applied to a broader population, and there is a far greater probability that any suspicious findings will entail the repeat scans, interventions, and anxiety without sufficient probability of a real cancer to counterbalance those factors.

While it would be possible to identify more lung cancers by doing more and more CT scans, the survival benefit would be diluted and potentially even lost in the face of a small but real risk from complications from invasive procedures or radiation over years and years.

In the real world today, lung cancer screening has yet to be widely accepted and pursued, at least where I practice, in spite of the real evidence to support it. This is likely because of cost issues related to who should pay for these scans, as well as the belief expressed by many primary care physicians, and perhaps also insurers, that the benefit isn’t truly “worth it” for the cost of the intervention. That’s a shame, but this situation will only be made worse by diluting the benefit by screening into an overly broad population in whom the vast majority of questionable nodules will lead to a lot of anxiety, additional scans, and potentially invasive procedures that will likely be more likely to ultimately document a benign process.

The evidence supports screening a population with a high enough probability of actually having lung cancer…but it’s a mistake to presume that screening more people will lead to better outcomes. It may lead primarily to more scans, more cost, more radiation, more procedures, and a lot more people subjected to the anxiety of being told they have nodules that need to be evaluated further.