I held my breath when I saw a Wall Street Journal story, “What Your Breath Reveals,” lead off with claims such as:
“ultimate noninvasive medical test: A growing number of health problems can be diagnosed by analyzing a patient’s breath alone…
Breath tests are also painless, faster to return results and potentially less expensive than blood tests—and easy to repeat as often as needed, even while patients are sleeping or exercising.”
Sure they are – but how strong is the evidence about accuracy? Yes, it’s intriguing research, but the story included some far-flung and unfounded projections.
It’s at times like this that I often to turn to ace story reviewer and guest blogger Harold DeMonaco, MS, to comment on what he saw. Here is his guest post:
Using breath odor to diagnose disease is certainly not a new concept. The Greek physician Hippocrates recommended that physicians “sniff” their patients as a regular part of the diagnostic workup.
A recent online story in the Wall Street Journal provided a good overview of the development process. It also noted the difficulties in the development of technologies that can identify biomarkers of disease using the patient’s breath rather than their blood stream. But the opening sentence caught my eye, “It’s the ultimate noninvasive medical test: A growing number of health problems can be diagnosed by analyzing a patient’s breath alone.”
The concept of a non-invasive method to diagnose disease or follow its progress using the presence of volatile organic compounds in breath is appealing. Attempts to develop what is now commonly referred to as an E-nose have been ongoing since the 1970’s with some success. But we are a long way from the “medical scanner” used so successfully by Dr. McCoy in the Star Trek series.
There are literally hundreds of volatile organic compounds exhaled by all of us on a daily basis. Some are the product of healthy metabolism. Some are likely markers of disease. Some are the result of food we have eaten and some are the result of exposure from the environment. So, identifying a biomarker that has a high degree of sensitivity (the proportion who have the disease and are correctly identified by the test) and specificity (correctly identifying the proportion of healthy people without the disease) is a complicated and demanding task.
I was particularly struck by the comments related to a recent study published in the Journal of Thoracic Oncology. Here’s the quote:
“In a study in the Journal of Thoracic Oncology this month, researchers from Israel and Colorado reported that breath analysis could distinguish between benign and malignant pulmonary nodules in a group of 72 patients with 88% accuracy; the test could also assess the specific type and stage of the lung cancers.”
The researchers used a combination of gas chromatography and mass spectroscopy and a newly developed set of chemical nanoarrays in their “explorative study” to compare benign and malignant pulmonary nodules. The test is clearly not ready for prime time.
The development process is long, time consuming and expensive for most diagnostic tests. While there have been great strides in the past decade, the technology is still in its infancy and many obstacles are likely to be encountered. It is a story worth watching but it is likely to read more like a novel than a short story.