High Serum S100B Levels for Trauma Patients without Head Injuries

To the Editor: During the past few years, there has been an increase in research articles dealing with protein S100B as a surrogate marker of brain injury. Because of its commercial availability and easy application in a clinical setting, S100B has been used as an indicator of the severity of damage to glial brain tissue. The majority of studies, however, simply replicate previous findings without reflecting on the drawbacks posed by the biochemical and physiological properties of the protein. Therefore, the criticism of this methodology in the article by Anderson et al. (1) is most welcome but unfortunately is of limited value, because it is made on the basis of very poor data. Anderson et al. studied early S100B serum concentrations in a small “consecutive” series of patients with “significant trauma without head trauma” (1, p 1256). The lack of brain injury was ruled out by “physical examinations and patient interviews” (1, p 1256). The authors did not present any significant data to reduce the probability of brain injury in their patients, however. Because the lack of head trauma associated with damage to brain tissue was the main focus of their investigation, one would expect the authors to have made a major effort to rule out brain injury by the use of well-established methods. The authors did not report any data based on conventional or advanced neuroimaging methods. Furthermore, they did not perform basic neuropsychological testing that could have provided a hint of brain dysfunction even in the absence of signs of brain injury in conventional brain imaging. These types of basic information are all the more important because both burns and respiratory distress syndromes, as reported in this study, could lead to cellular and subcellular brain dysfunction. Therefore, the discussion of the possible limitations of the use of S100B as a brain-specific surrogate marker of brain damage, although heads in the right direction, is based on a patient sample that was not evaluated appropriately to rule out brain injury. Protein S100B (as well as most of the other proteins that are supposed to be brain-specific) also is expressed in a variety of cells in addition to glial cells. This fact, although known for quite a long time and intensively discussed in the literature, does not necessarily reduce the value of S100B as a reliable marker of the severity and the type of brain injury in a patient. It alerts one to question the specificity of individual serum values of S100B. As the authors pointed out, more information on the temporal profile of S100B release is needed, especially with regard to multitraumatized patients and in patients with surgery-related traumata. Studies in “pure” neurological conditions (e.g., stroke) have shown that the predictive value of S100B is highly dependent on its release pattern during the first 3 to 4 days. The major aim of Anderson et al.’s study was to draw attention to the contamination of brain-originated S100B by extracerebral sources in multitrauma patients. The data presented by this group from Karolinska Hospital do not substantiate the lack of brain injury, however, and therefore do not suit their purpose. Manfred Herrmann