Biomarkers--In Your Breathe?

I read an article on today about using "breath prints" as a way of diagnosing a disease or monitoring patient health. The link to the article is here if you're interested.

For you that work in science or medicine, a biomarker is a familiar concept; but for the sake of those who many not be so familiar, let me briefly describe it:

A biomarker is a molecule that has been tested and validated to indicate the presence or absence of a health condition.

For instance, biomarkers are heavily studied in the diagnosis of cancer. If research confirms that "molecule A" increases when a large cohort of people develop lung cancer, then "molecule A" may be considered a biomarker.

In the article I linked to on, folks in the pulmonary vascular program at the Cleveland Clinic are narrowing in on the idea of using exhaled air as a sample source to look for biomarkers. Rather than looking for a particular molecule though, they propose testing a large array of molecules to give a "breath print".

This concept is not new, but the use of exhaled air is somewhat novel. Technology to test people's breath is mostly associated with confirming blood alcohol levels--the test which has affectionately become known as the "breathalyzer."

But how will this technology be different from other biomarker pursuits?

The struggles surrounding biomarker validation are staggering. 

Patient populations vary in genetic background (ethnicity), diet, weight, age, gender, lifestyle, and many other factors. After stratifying patients into relevant categories, the patient populations become small, making statistical significance difficult to attain.

Going back to the analogy of lung cancer and "molecule A", consider that only 35% of lung cancer patients have elevated "molecule A", and 35% have moderate levels, and 30% have low levels.

Is the relationship between molecule A and lung cancer causative, correlative, or non-existent?

That is the billion dollar question.

Then there is the million dollar question: Are these scientists certain that "molecule A" diffuses through the alveolar membranes of the lung and dissolves into air in the lungs? That question has partially been answered, but it will need to be addressed for every molecule that is analyzed for every disease to be diagnosed.

Other questions for these scientists include:

Are there other conditions or scenarios that could cause "molecule A" to increase? 

Stress? Pregnancy? Eating lots of Chinese food?

What threshold level of "molecule A" will be considered "elevated", "moderate", and "low"?

Perhaps some people have inherently higher or lower levels of "molecule A", meaning a baseline level must be established for each patient prior to diagnosis. 

Another problem is establishing a baseline is difficult because people are always changing. When can you consider a patient "normal"? And even if you knew when they were "normal", could you get them to come to the doctor to obtain a sample? Probably not. People don't go to the doctor when they're "normal".

I'm not being negative. I'm trying to be rational.

These questions hang over the head of these biomarker-based technologies. Conceptional challenges must be addressed before people start celebrating a breathalyzer that can diagnose early stage cancer--or any other disease.

In my opinion, the most useful application for a breath-based analysis technology is to market it as a convenience to medical professionals, their patients, and law enforcement. 

Develop ways to test the exact things that are already being tested for (glucose, narcotics, hormones, etc.), but allow them to give their patients a breath test instead of a blood test.

Nobody likes needles.

No comments:

Post a Comment