06 February 2012

Blood Test For Depression May Soon Be Ready For Clinical Use

A private company has developed a proprietary blood test for depression that costs $745. It showed a positive result in 91% of cases where someone has been diagnosed with clinical depression by conventional means, and shows a negative result in 81% of cases where there is not a diagnosis of depression by conventional means, when applied to a sample of 70 depressed people and 43 controls.

This is probably not ready for prime time yet. The paper describing the test reveals that one could conduct a much simpler and cheaper test that isn't proprietary that is almost as accurate. And, at $745 per test, it still isn't cost competitive with having somebody with a graduate degree that trained them to diagnose clinical depression visit with and evaluate a patient. It doesn't yet serve the purpose that one would really add value to this kind of diagnostic test.

[A]ll these data demonstrate is that the test can distinguish between people with MDD and entirely healthy people. But how often are doctors going to need to do that?

More likely, they'll want to distinguish depression from other things that are often confused with it, such as: bipolar disorder, anxiety disorders, chronic fatigue syndrome, bereavement, "stress", and all manner of physical illnesses e.g. thyroid problems. Daniel Carlat said last year that:

"If the test cannot distinguish different psychiatric problems, then the MDDScore is simply a non-specific 'biomarker' for emotional difficulties of all stripes, and would be essentially useless. How disorder-specific is the MDDScore? This paper doesn't tell us."

But, despite the appropriate grounds for skepticism of this particular test as of 2012, the promise is clearly there. Mass production of medical tests can dramatically reduce their costs, larger data sets can determine what characteristics distinguish the false positives and false negatives, and the extent to which these biomarkers (or new ones similarly motivated) can discriminate between different causes of emotional difficulties can be determined.

It isn't hard to imagine a world ten or twenty years from now in which someone can buy a depression test at a pharmacy much the way that one can by pregnancy tests, drug tests, and paternity tests now. Given the social stigma of mental illness, the difficulty that someone who thinks that they might be experiencing clinical depression who is isolated has in self-diagnosing subjective symptoms, and the high cost in time and inconvenience involved in locating a credible mental health professional and scheduling, attending and paying for a diagnostic appointment, an investment of say, $15-30 in a drug store depression test might be worthwhile.

Blood tests for mental health conditions also help to validate the idea that mental health conditions have a biological basis, that they have an objective reality, and that they can be addressed by medical means. The impact of the mere existence of a blood test for depression on how the public views depression could be as important for how our culture conceptualizes depression as the increasing scientific evidence that sexual orientation is largely congenital is for gay rights.

And, there is synergy here too. To the extent that we can determine if someone is depressed, and can determine how depressed they are from a blood test, we can improve the rigor of clinical trials of drugs designed to treat depression which are forever haunted by questions regarding the accuracy and consistency of subjective diagnostic criteria, particularly in drug company run trials where there is strong incentives to subtly push the boundaries of these criteria in a manner that will show a new drug to be effective.

Biomarker tests are really better suited than genome wide association studies for determining the mechanism that is implicated in mental health conditions and for distinguishing causally different subtypes of conditions that have similar symptoms but might respond differently to treatment because of their different causes. And, it is possible to make much more credible determinations of the role of gene and environmental interactions in a condition when the candidate genes can be linked to a biological process that produces a measurable biomarker that is associated with the mental health condition in question. Without knowing the biological mechanism of a mental condition, it is far too easy to generate genetic assocations that are nothing more than statistical flukes or products of people with a condition having some sort of ancestral relationship associated with the people in whom the mutations that gave rise to the genetic components of the condition arose.

One can even imagine forensic applications. It is already possible to determine the identity of someone at an incident scene from a DNA test of a blood sample. In a few decades, it might be possible to also determine if, for example, that person who currently in the throes of clinical depression, a clue that might, for example, help a coroner help judge whether someone was a suicide victim or a homicide victim, to evaluate the veracity of a claim that clinical depression was an important factor in a child abuse or neglect incident, or to help an investigator more accurately profile how a suspect whose blood was left on the scene was likely to be behaving to better devise a search for that suspect. In these kinds of applications, a $745 price tag would be quite tolerable.

No comments: