Are lung cancer breath tests more than hot air?

By Ted Gansler, MD, MBA, MPH

Can breath tests (like those used to check whether drivers have been drinking alcohol) be used for lung cancer screening? Or, is this (pardon the pun) just a lot of “hot air?” Although breath tests for lung cancer are “not ready for prime time,” there has been some encouraging research.

There are 3 main ways to fight cancer – prevention, screening, and treatment. Although lung cancer remains the leading cause of death from cancer worldwide and in the United States, researchers are making progress against this disease on all 3 fronts. 

Over nearly a half century, researchers tried several tests for lung cancer screening, none of which were accurate enough for widespread use. Because of research results released in 2010, the American Cancer Society and several other organizations now recommend that people at high risk for lung cancer (certain groups of current and former smokers) ask their doctor about CT scans for lung cancer screening.

On average, people in these high risk groups who have this test every year according to the ACS guidelines can reduce their risk of dying from lung cancer by about 20%. This can save a lot of lives and prevent a lot of suffering, so if you are a current or former smoker, you should read more about our lung cancer screening recommendations.

Research into easier lung cancer screening

One challenge with CT scans is that they find some lung nodules that are neither clearly cancer nor clearly benign (not cancer). This question is usually figured out by follow-up scans, but sometimes biopsies are needed. These biopsies can pose significant risks, which is one reason screening isn’t recommended for people whose risk of lung cancer isn’t as high as that of heavy smokers. So researchers are looking for ways to make screening easier and more accurate, faster and more affordable. [more]

One interesting area of current research is blood, sputum (phlegm), and breath tests that might help resolve uncertain scan results, or perhaps even pre-screen people so that some might not even need to have CT scans.

Today, I’m going to write about exhaled breath research because it’s interesting and surprising. Exhaled breath, just blowing out, is probably the easiest way to get a sample for medical testing – it’s also one of the least used methods. So, why aren’t clinical laboratories routinely testing exhaled breath for cancer-related chemicals (such as volatile organic compounds, or VOCs)?

The first reason, in my opinion, is that these chemicals are hard to find because they are present at such low levels. Compared to gases like nitrogen, oxygen, carbon dioxide, and argon (how many of you knew that air is nearly 1% argon?), levels of the volatile organic compounds (VOCs) that may be linked to cancer are about a billion times lower. Consequently, measuring these VOCs requires very expensive instruments and very highly skilled chemists.

These measurements are still complex, but thanks to some new technologies, they are getting simpler and more affordable. One of the most fascinating of these technologies is a category of instruments known as “electronic noses” that copy some aspects of the noses of animals, whose sense of smell is much more sensitive than that of humans.

The second reason, also in my opinion, is that few cancer researchers have been interested in VOCs in exhaled breath. Breath and air consist mostly of gases, which tend to be rather small and simple molecules. Much of the basic cancer research of the past few decades has focused on identifying the DNA, RNA, and protein changes that distinguish cancer cells from normal cells; in contrast to gases, these molecules are very large and complex. And one thing that researchers have learned over that past half century is that changes in DNA are the “main event” in cancer. Changes in DNA are necessary for cells to become cancer. These DNA changes result in RNA changes, which then influence proteins. So, there are valid reasons why scientists have been focusing on these large molecules.

This research, called genomics and proteomics, has paid off in terms of numerous tests that can be done on samples or small pieces of tumors, which help predict a patient’s prognosis (outlook) and help oncologists pick the drugs most likely to be effective. Moreover, changes in DNA, RNA, and proteins can eventually cause changes in small and simple molecules, including the “lowly” VOCs. The bad news is that no individual VOC has been shown to accurately predict the presence of lung cancer. The good news is that tests based on patterns of VOC levels are pretty good.

Breath tests showing promise

Using “traditional” instruments of analytic chemistry or electronic noses to recognize which VOCs are rarer in people with lung cancer and which are abundant, researchers have developed some tests that, based on preliminary studies, seem to have about the same accuracy as current lung cancer screening tests (like CT scans). In other words, they have similar rates of false-negatives (the tests says there isn’t cancer, when there really is cancer) and false-positives (the test says there is cancer, when there really isn’t cancer). Other researchers have used the “biological noses” of trained dogs to recognize VOC patterns, with similarly impressive results.

For example, preliminary results presented at the American Association for Thoracic Surgery 2014 Annual Meeting on April 29, 2014, showed that a breath test of VOC patterns could help recognize early lung cancer, with a false-negative rate slightly higher than that of PET scanning (18% and 10%, respectively), but a false-positive rate that was much lower (25% and 61%, respectively). Likewise, according to a preliminary report from the May 2014 meeting of the American Society for Clinical Oncology, a test of exhaled breath was 85% accurate in distinguishing patients with early lung cancer from those with benign lung problems. 

I have never done any VOC research, but long ago I dabbled in research with DNA, RNA, and proteins, so I was a little surprised by the promising results of these studies (and quite a few others) of VOCs. I was even more surprised (and a little embarrassed) by a recent study that fits into the category of “why didn’t anyone else think of this before?” research.

By simply measuring the temperature of exhaled breath in 82 people with abnormal x-ray tests suggesting that they might have lung cancer, the researchers were able to correctly predict in nearly all instances whether lung cancer was actually present. Forty patients turned out to truly have lung cancer, and the temperature correctly predicted this diagnosis in all of them. Among the 42 patients who did not have lung cancer, the temperature test correctly classified 39 and incorrectly predicted that 3 had cancer (a false-positive rate of 9%).

The point of this blog is not that I’m predicting breath tests for temperature or VOC patterns will become the standard for lung cancer screening. At this point, there’s no way to tell whether these tests will be as accurate as other experimental blood tests for protein or DNA or RNA patterns, or some new kind of diagnostic imaging scan that will eventually replace CT scans. The research I’ve described here is still in its early stages; many more studies will need to confirm that this approach actually works before it can be considered for widespread use.

The point is that there’s a lot of exciting research underway, and one of these methods, or perhaps a combination of more than one, may eventually save even more lives than screening with CT scans.

But until then the most important message (in addition to helping to support this research financially) is to stay away from tobacco products and other exposures that could increase lung cancer risk. Ask your health care professional for help in quitting if you currently smoke, and ask your doctor about screening with CT scans if you’re a current or former heavy smoker.

 

Dr. Gansler is director of medical content for the American Cancer Society. 

Leave a Reply

Your email address will not be published. Required fields are marked *