Sepsis affects the brain, and the impairment of brain function resulting from sepsis is often associated with severe infectious disease. The effects of sepsis on the brain are detectable in previously healthy brains but are amplified in cases with concomitant brain injury, as after traumatic brain injury or subarachnoid hemorrhage. Previous injuries, in fact, increase brain vulnerability to the complex cascade of events summarized in the term “septic encephalopathy.”
In fatal cases, various anatomical lesions have been demonstrated, such as proliferation of astrocytes and microglia in the cortex, cerebral infarcts, brain purpura, multiple small white matter hemorrhages, central pontine myelinolysis, or disseminated microabscesses (1). In survivors, on the contrary, reversible changes, such as reduction of cerebral blood flow, capillary leakage, and dysfunction of the blood-brain barrier (BBB), are involved (2,3).
The interplay between direct effects of sepsis resulting from toxic mediators and indirect effects, such as arterial hypotension, hyperthermia, and increased intracranial pressure, contributes to the confused picture of the brain during sepsis. A marker of cerebral damage during sepsis could have considerable value, both for clarifying the mechanisms involved and for quantifying the degree of brain injury.
In the study by Larsson et al. (4) in this issue of the journal, a possible mechanism is investigated and a marker proposed in a porcine model of endotoxemic shock. After endotoxin infusion, the plasma concentration of S-100B, a protein thought to be a marker of glial damage, increases. This significant increase could derive from glial or Schwann cell damage, accompanied by an opening of the BBB. The study detects an S-100B increase concurrent with the endotoxin infusion that is not replicated in the control group. A clear association between the infusion, its temporal profile, and the S-100B plasma levels is therefore demonstrated.
It would be tempting to conclude that this S-100B increase indicates brain damage, suggesting glial destruction and BBB opening. If that is the case, a pathogenetic mechanism for brain damage resulting from endotoxin and the role of S-100B as a marker of brain dysfunction could be advocated. Unfortunately, the specific setting of this experiment does not provide all the information required for such a promising conclusion.
S-100B has been acknowledged as a marker of brain damage; for instance, after traumatic brain injury (5). However, there are extra-cerebral potential sources of S-100B. In a recent study considering cerebral and extra-cerebral infections, one fourth of patients without cerebral infection showed an increase of serum S-100B (6). There is the possibility that this increase was a result of brain damage concurrent with extra-cerebral infections, but this was not confirmed.
In trauma, S-100B is detectable in the absence of brain injury (7,8). The largest study (387 patients and healthy volunteers) concluded that serum S-100B is a marker of brain damage that correlates with injury severity, but, as major extracranial injuries also increase S-100B levels, specificity is still a problem (9). Unfortunately, data proving or excluding potential extra-cerebral sources of S-100B during sepsis are not available.
In the experiment by Larsson et al., a large variability of the baseline concentration of S-100B, before endotoxin infusion, confirms that in some animals a large concentration of the marker is present with an intact brain, but all the reasoning is based on the assumption that S-100B increase reflects brain damage. That is likely, but not proven, as we do not have data measured in the cerebrospinal fluid, in the cerebral tissue, or in the cerebral veins.
Another intriguing question that cannot be answered in this experiment is the relative role of BBB disruption. Assuming that S-100B tissue concentration in the brain increases as a result of endotoxin infusion, the concentration detected in the plasma depends both on this increase, which creates a gradient between tissue and intravascular concentration, and on the reflection coefficient of the interface between tissue and blood. If the BBB permeability is slightly increased, a much larger tissue concentration should be suspected for achieving the plasma concentrations detected in this study compared with a situation in which BBB freely allows the transit of proteins.
The measurement of an independent marker of BBB permeability would have been extremely useful to clarify the results of this experiment. First, the degree of BBB disruption is important per se, as BBB opening, leading to vasogenic brain edema, is thought to be an important mechanism of septic encephalopathy. Moreover, if the degree of BBB opening could be measured, some estimate of the tissue concentration of S-100B could be attempted.
The multifaceted properties of S-100B, which acts both as a marker of glial damage and of BBB function, require a separation of the two components. Direct measurements of brain variables, both in vivo (such as microdialysis, intracranial pressure, and sagittal sinus sampling) and postmortem, would greatly increase the amount of information obtainable from this relatively simple experimental model.
“Brain and sepsis” remains a difficult and relatively unexplored topic, and contributions such as the paper by Larsson et al. (4) are welcome. The authors clearly admit the weaknesses of their work but also highlight the potential of this field of research.
1. Jackson AC, Gilbert JJ, Young GB, Bolton CF. The encephalopathy of sepsis. Can J Neurol Sci 1985;12:303–7.
2. Bowton DL, Bertels NH, Prough DS, Stump DA. Cerebral blood flow is reduced in patients with sepsis syndrome. Crit Care Med 1989;17:399–403.
3. Voigt K, Kontush A, Stuerenburg HJ, et al. Decreased plasma and cerebrospinal fluid ascorbate levels in patients with septic encephalopathy. Free Radic Res 2002;36:735–9.
4. Larsson A, Lipcsey M, Sjolin J, et al. Slight increase of serum S-100B during porcine endotoxemic shock may indicate blood-brain barrier damage. Anesth Analg 2005;101:1465–9.
5. Pelinka LE, Kroepfl A, Leixnering M, et al. GFAP versus S100B in serum after traumatic brain injury: relationship to brain damage and outcome. J Neurotrauma 2004;21:1553–61.
6. Unden J, Christensson B, Bellner J, et al. Serum S100B levels in patients with cerebral and extracerebral infectious disease. Scand J Infect Dis 2004;36:10–3.
7. Anderson RE, Hansson LO, Nilsson O, et al. High serum S100B levels for trauma patients without head injuries. Neurosurgery 2001;48:1255–8.
8. Unden J, Bellner J, Eneroth M, et al. Raised serum S100B levels after acute bone fractures without cerebral injury. J Trauma 2005;58:59–61.
9. Savola O, Pyhtinen J, Leino TK, et al. Effects of head and extracranial injuries on serum protein S100B levels in trauma patients. J Trauma 2004;56:1229–34.