WHAT do we mean when we say that an elevated level of a biomarker "relates" to symptoms, as Pattinson et al1 do in their article? This "precision" target approach tends to beg the question of what we are really studying.

In the field of traumatic brain injury (TBI), a search of PubMed, the National Library of Medicine database/search engine, yields 45 679 citations. Of that number, 1875 have something to do with the study of biomarkers. There are also 2065 articles on TBI-associated proteins. A search on TBI and behavior produced 5881 citations of published articles. I think we can all agree that, taken together, this represents a substantial collection of facts! But now what?

Despite all this information from preclinical studies at both the molecular and behavioral levels of analysis, 100% of large, multicenter clinical trials for treating TBI have failed, and our collective frustration has led to the idea that if only we could identify the biomarker or even a set of molecular biomarkers to predict behavioral outcomes, all would be well. This is a relatively rigid definition of the field that can result in an overemphasis on mechanistic markers of injury at the cost of attention to other factors like objective, quantifiable, behavioral measures to evaluate patient functional improvements over time. It has been my observation that, because of all the problems with clinical trials, increasing emphasis is being placed on identifying molecular markers that will facilitate "precision medicine" approaches that can ignore more holistic issues, like how patients respond to treatments at the functional level involving the activities of their everyday life.

While I have no intention of discounting the signi-ficance of brain biomarkers in many contexts, maybe it is time to reconsider the assumption that important and very complex disease conditions like TBI, posttraumatic stress disorder, depression, stroke, and other central nervous system (CNS) disorders can be understood (and treated) by identifying and modifying molecular markers extracted from biofluid (blood or serum, saliva, and cerebrospinal fluid). What if these "brain" diseases are, in fact, systemic and affect most, if not all, organs of the body? What are the biomarkers marking?

The growing interest in highly targeted biomarkers is facilitated by pressures to get quick answers to assure diagnoses, prognoses, and more rapid treatments. This trend is driven by an ever-increasing dependence on high-technology imaging and quantitative measurements of molecular and cellular processes that often target a single gene, or a few genes, proteins, signaling pathways, or inflammatory factors that, if suitably modified, promise efficient and effective treatment.

So, which approach is most promising-the highly targeted or the more holistic? It is worth noting here that behavior, as an end-point and descriptive measure, is also a critical biomarker that can be used to determine and likely predict functional outcomes over the short and long course of a brain trauma. Do we need to be reminded that behavior is also a product of multiple organismic events and brain processes that are altered by, and can alter, outcomes of CNS injuries? Overemphasis on seeking tissue-type biomarkers can also overlook the fact that specific behaviors like substance and dietary abuse can alter the expression of biomarkers and need to be considered in any prognoses.

Even if we could all agree that some selection of specific biomarkers are valid and reliable measures of an injury (eg, 3D imaging of the extent of a brain lesion), we still have to address the question of whether patient behaviors, as often measured in clinical trials for TBI, are also adequate to mirror and reflect the severity of the injury and the process of repair in both the acute and chronic stages of brain damage (see Stein2 for a fuller discussion of this issue).

In other words, both molecular and behavioral biomarker studies are necessary to determine and predict severity and recovery, and both need more serious testing in clinical situations. Despite the intrinsic elegance and allure of its associated technology, overemphasis on molecular markers as the primary focus for understanding functional recovery and treating disorders draws resources away from support of research seeking holistic, but precise and valid, outcome measures relevant to researchers, patients, and their families. Maybe we need a consensus selection of subsets of both, and then head-to-head comparisons, first at the preclinical level and then in human trials, to determine which markers work best. TBI research needs more than 1 methodology-maybe even more than 2.

REFERENCES