The findings reported yesterday in the Annals of Internal Medicine that lung CT scans used to screen moderate to heavy smokers for early lung cancer resulted in up to 33% of the studies incorrectly suggesting that the participants may have had lung cancer is no surprise to me, and probably shouldn’t be a surprise to you.
After all, as I have written about previously in this blog, even reasonably healthy non-smokers like me have had to cope and contend with a false positive scan result from a chest CT scan. And, as the use of CT scans becomes even more widespread, we are going to find ourselves as doctors and patients dealing more often with the question of what to do when something unexpected and unanticipated shows up on one of these tests.
Let’s briefly review how this study was done.
Currently, there is a nationwide study underway to find out whether or not lung CT scans in otherwise asymptomatic long term smokers can find lung cancer early and reduce deaths from that disease. Given the fact that lung cancer is the leading cause of cancer deaths in this country for both men and women, that is a pretty important question. Hopefully, we will have meaningful results from that research in the next couple of years.
In preparation for that larger study, a small preliminary research study was done where moderate to heavy smokers agreed to enter a similar trial that was conducted on a much smaller scale. It is that trial that provided the results on which the current Annals article is based.
In that trial, participants were randomly chosen to get either CT scans or routine chest x-rays at the time they entered the study and one year later. In analyzing the data, the researchers found that for those who had the CT scans, about 33% of the participants had what we call a false positive result by the second scan. That means the patient was told there was a finding on their CT scan that could be lung cancer, but eventually was proven not to be the case. For those participants who had chest x-rays on a similar schedule, the false-positive rate was 9%, or about 1 in 10.
Some of those patients underwent more extensive studies; including an “invasive” test which means they had their doctors do something more aggressive to find out whether or not the lesion seen on the CT scan or the x-ray was in fact cancer. 7% of the participants who had a false positive CT scan had such a test, while 4% of those with a false positive chest x-ray underwent an invasive procedure. 2% of the participants had major surgery for what eventually turned out to be benign disease.
The authors of the study concluded that “risks for false-positive results on lung cancer screening tests are substantial after only 2 annual examinations, particularly for low-dose CT.”
Unlike mammography, where a false positive mammogram may result in an extra mammogram or ultrasound, when you start doing invasive tests on patients with suspected lung cancer the risks can be much greater. That is because these folks are not generally healthy to begin with. Many have underlying serious heart and/or lung disease, for example. That’s why this issue is so important.
I can actually speak “first hand” to the question of false positive lung CT scans.
As discussed in a prior blog, I underwent a CT scan of my heart to see if I had underlying coronary artery disease. The good news was that my arteries were clear. The not-so-good news was that the radiologist saw a small nodule in my lung.
Follow-up CT scan 6 months later showed the nodule was stable and a couple of other similar nodules were present in my right lung. I made the decision—based on the size of the nodules (which were very small, and within the range where experts said no further evaluation was warranted given my extremely low risk of lung cancer)—not to undergo further testing. Over two years later, and I am doing fine in that regard.
The impact for me was real until I looked into the science of the situation and consulted with colleagues who are very knowledgeable about these things. But most people don’t have that option. At the same time, I was applying for an increase in my life insurance, and the insurance company—although not outright denying my request—wanted further serial scans before committing to a larger increase than was already possible under my policy. If I didn’t have a reasonable amount of insurance already, that could have been very practical “fallout” from this episode.
As I talk with colleagues around the country, this issue of false positive CT scans has become more prominent and concerning in ways that you probably can’t even imagine.
For example, Medicare recently decided not to cover CT scans to look for early colon cancer or colon polyps. One of the reasons was that they were concerned about the rate of lesions being found elsewhere in the body that may be “false positives” (that is, abnormalities seen on the CT scans which turn out not to be serious problems but require extensive and possibly expensive further testing to find out).
And then there is the issue of CT scans finding things that may appear to be serious, and have to be taken seriously, but in fact may never have caused anyone any problem. That in no small part—when it comes to cancer, for example—is due to the fact that we don’t have reliable tests to determine consistently for many cancers which ones are really bad actors and which ones would never have caused difficulty.
As we improve our technology and our testing capabilities, I suspect we are going to find more and more of these situations where the “good” of our testing (such as finding cancers early) will run into the consequences of moving closer and closer to finding cancer before it can even be seen on available tests. Think in terms of blood tests routinely spotting changes in a blood sample that suggests a cancer is present, or a genetic test which suggests a cancer might be present, and then not being able to find where the cancer is coming from.
That is where the future will be taking us. For now, we have to deal with more “mundane” but no less real situations where we can diagnose things that may never be a problem, such as certain breast cancers, prostate cancers and probably lung cancers.
As we look for these cancers, inevitably we will find those things that look bad but turn out not to be so. Eventually someone is going to be hurt by that process. Maybe the “hurt” will be psychological, or maybe the “hurt” will be an extra test or two or maybe the “hurt” will be being denied life insurance to protect one’s family. But it is also very possible that the “hurt” could be serious injury or even death, which is what has so many of us concerned about lung cancer screening.
That’s why it is so important that we get this lung cancer screening thing right. We need to know whether or not the benefit of screening smokers to find early lung cancer really saves lives. And, in the process of saving lives, it is important to find out how many lives were truly harmed by the effort.
This report is just a first step in understanding those harms. I for one won’t be particularly surprised if we eventually find out in the larger study currently underway that those harms in this particular group of people turn out to be pretty large in number and substantial in impact.
So if you are contemplating getting screened for lung cancer—and from what I hear there are a fair number of smokers thinking of doing just that—be certain you understand what we don’t know about the benefits of getting such a test (which means we can’t say that it really saves lives) and what we do know about the downsides, as reported in this study.
Remember, just because we doctors can do things doesn’t always means it’s always the best thing to do. So be cautious, and be informed.
Feedback
