Volume 17, Issue 1 , Pages 1-3, February 2003
S100β and NSE: Stroke surrogate signals or natal neural noise?
Article Outline
- Surgery and S100β and NSE
- The search for a surrogate—Inherent limitations
- The current findings and their importance
- Conclusions and the true surrogate
- References
- Copyright
The S-100 Proteins constitute one of the largest subfamilies of the calcium-binding proteins and have been extensively characterized. Their genomic organization now is known, and many of their target proteins have been identified.1* Among their biologic functions are differentiation, signal transduction and transcription, and cell motility and cycle regulation. Because of these latter properties, S-100 proteins are expressed in a broad range of chronic and acute diseases, including cardiomyopathies, cancer, inflammation, and neurologic diseases. However, and importantly, their role in these diseases still is undetermined. Of the 3 isoforms of S-100, the ββ subunit (S100β) is present in high concentrations in central and peripheral glial and Schwann cells, Langerhans and anterior pituitary cells, and fat, muscle, and bone marrow tissues. Consequently and not unexpectedly, S100β has been reported in blood and cerebrospinal fluid after acute injury, such as with ischemic stroke,2 encephalopathy (diabetic, hepatic, cardiac arrest), traumatic brain injury (infants, adults), cancer, and mania, as well as with chronic diseases, such as Alzheimer's disease, Down syndrome, acquired immunodeficiency syndrome–related dementia, multiple sclerosis, schizophrenia, and major depression.
The second set of markers under this consideration are the enolases—a large class of enzymes that catalyze intermediaries of the glycolytic pathway. Among these is neuron-specific enolase (NSE). Containing both α and γ subunits, (NSE) appears to be specific for neurons being detected in acute ischemic stroke,3 perinatal brain damage, neuroblastoma, Alzeheimer's disease, seizures, encephalitis, lyme disease, and retinal detachment. However, NSE also has been found in patients with cancer of the lung, thyroid, pancreas, prostate, and skin. Importantly, NSE is “carried” in red blood cells and platelets (including megakaryocytes), the latter suggesting that dissociation and recombination of NSE subunits occur in platelets during aging. Furthermore, their association with blood cell carriers affects their assay vis à vis the dynamics imposed by blood loss and replacement, as well as intravascular thrombosis.
Surgery and S100β and NSE
A substantial literature exists for both these potential markers, with the majority of published studies suggesting association with both chronic and acute perioperative cerebral ischemic dysfunction.4, 5, 6, 7, 8, 9, 10, 11, 12 However, a number of cautionary publications addressing limitations (described later) also have been published, which suggest extracranial sources of the markers,13, 14, 15 as well as the existence of potentially confounding effects of concurrent thrombus, emboli formation,2, 16 and inflammation.17 Albeit, these “negative studies” have not been given appropriate recognition, and when considered, the logical conclusion is that controversy exists regarding both the specificity and sensitivity of S100β and NSE for cerebral injury.
The search for a surrogate—Inherent limitations
Clearly, discovery of a surrogate for brain injury—especially if easily measured—would be valuable to clinicians, clinical researchers, drug developers, and, importantly, the public. That task, although easy to articulate, is daunting in its challenge because the impediments are far greater than those encountered in like quests for surrogates of myocardial infarction or renal failure. Cerebral ischemic injury is quite unique and especially peculiar. For example, unlike that for other organs, the volume of cerebral injury per se is less important than its location, making any neural or glial breakdown by product—such as enzyme spillage—less reliable either as an associate of, or a surrogate marker for, clinically relevant injury. Seemingly insignificant amounts of brain tissue destruction, when located in critical regions, may portend substantial functional disability, even more disability than that associated with much larger volumes of ischemia when located in less critical regions of the brain. As a result, accurate interpretation of central ischemic “spillage” is inherently limited—the magnitude of such spillage although indicative of the volume of injury does not necessarily relate to the degree of functional impairment. Only in the circumstance of profound global injury will neural spillage be a certain harbinger of clinical disability.
As well, accurate interpretation of brain spillage is beset by a second limitation—also seemingly peculiar to the brain, that is the critical importance of the integrity (or lack thereof) of a barrier—the blood-brain barrier. When intact, this unique barrier actually prevents accurate peripheral (arterial, venous) estimation of brain tissue ischemic spillage, especially when spillage manifests in the form of a large molecule. Disruption of the barrier, instigated by emboli or inflammation, for example, will allow more ready (and likely more accurate) peripheral assessment of central injury, assuming that the site of barrier injury is proximate to the region of injured brain in question.
As if the previously mentioned spatial limitations were not enough, a third temporal-limitation likely exists, for both central spillage and barrier permeability are not unwavering but temporally sporadic, making accurate peripheral sampling difficult. For example, some suggest that barrier disruption may be induced early, at the time of aortic manipulation and cross-clamping and secondary emboli release. Permeability rises geometrically with aperture size, allowing greater spillage into the vascular space, which may allow more accurate peripheral assessment of central spillage. However, central spillage is known to be protracted, accelerating over hours to days—likely peaking 48 hours after initial insult—that is, when barrier integrity may be well into recovery. Thus, the accuracy of a peripheral sampling of central spillage critically is a function of the accumulated spillage at the time of sampling, as well as the barrier permeability proximate to the injury, and lastly the peripheral sampling frequency.
Finally—and perhaps most importantly here—is the impact of secondary (noncerebral) sources of spillage, as well as secondary (nonplasma/cellular) carriers of spillage, both of which may create substantial noise that can obfuscate the central injury signal. In the first case, noise emanates from secondary sources of spillage (such as fat for S100β), which have release dynamics that are characteristically unpredictable. Differentiating signal (central brain spillage) from noise (peripheral fat spillage) becomes difficult, if not impossible, if such secondary sources are contributory. In the second case, noise is generated because the carrier (such as a red blood cell or platelet for NSE) releases spillage erratically as the integrity of its membrane is altered (such as by cardiotomy suction) or when intravascular carrier mass is altered (such as by red cell or platelet dilution or loss).
Clearly again, then, accurate peripheral assay of central neural damage is itself a very difficult endeavor, and no less is its validation as a bona fide surrogate of irreversible injury. Given that more than 2 decades were required to establish appropriate guidelines for myocardial creatine-kinase myocardial band release as a surrogate of infarction, it should not be unexpected that at least that amount of time would be needed to discern accurate cerebral markers.
The current findings and their importance
The article by Ishida and colleagues18 then is important, and its cautions must be taken seriously. Far too many studies assessing S100β and its association with clinical injury endpoints have emerged prematurely. Few studies have taken a less glamorous, but more scientific, approach to understanding such markers—namely, the validation of measurement technique and the painstaking detail-driven analysis of sampling-time implications, the effect of confounding factors, and the specific technique limitations. The Ishida team has shown, importantly, the potential of a noncentral noise source for S100β—previously speculated to be fat, thymus, or other nonneural tissue—documenting the existence of such by straightforward measurement in patients undergoing coronary artery bypass graft surgery. As well, these investigators have re-educated clinicians regarding NSE measurement and its relationship to the dynamic of the red blood cell (a carrier), especially in the environment of cardiac surgery when red cell flux, loss, damage, and replacement assume an important dynamism.
Recognizing the controversy in the literature regarding the association of these markers with neurologic/neuropsychologic “outcomes” then, on balance, it can be concluded from this evidence—and that now provided by the present study—that assay of these potentially important markers is not straightforward, and before even association is claimed, the accuracy of the methodologies first must be substantiated for the specific clinical setting under consideration.
Conclusions and the true surrogate
Among the greatest clinical research challenges is the identification of the true surrogate. Unfortunately, surrogacy is readily claimed and prematurely advertised but often not painstakingly assessed, making the supposed surrogate vulnerable to exaggerated claim. Key is painstaking validation of methodology—a task that is not glamorous, is necessarily protracted, and then mastered only by a select few. However, such efforts are the essence of science, and, moreover, such efforts are absolutely essential for any putative factor to rise to the rank of a “true surrogate.”
References
- . Glial protein S100β modulates long-term neuronal synaptic plasticity. Proc Natl Acad Sci. 2002;19:4037–4042
- . S-100 protein and neuron-specific enolase concentrations in blood as indicators of infarction volume and prognosis in acute ischemic stroke. Stroke. 1997;28:1956–1960
- Brain-specific proteins in the cerebrospinal fluid of severely asphyxiated newborn infants. Acta Paediatr. 2001;90:1171–1175
- Cerebral emboli and serum S100β during cardiac operations. Ann Thorac Surg. 1998;65:1645–1650
- . Release of S100β during coronary artery bypass graftings reduced by off-pump surgery. Ann Thorac Surg. 1999;67:1721–1725
- Neuromonitoring and neurocognitive outcome in off-pump versus conventional coronary bypass operation. Ann Thorac Surg. 2000;69:1162–1166
- Neural tissue-related proteins (NSE, GO alpha, 28-kDa calbindin-D, S100β and CK-BB) in serum and cerebrospinal fluid after cardiac arrest. J Neurol Sci. 1994;123:134–139
- Serum S100 protein: A potential marker for cerebral events during cardiopulmonary bypass. Ann Thorac Surg. 1996;61:88–92
- Postoperative course of S-100β protein and neuron-specific enolase in patients after implantation of continuous and pulsatile flow LVADs. J Heart Lung Transplant. 2001;20:1116–1310
- Significance of serum S100 release after coronary artery bypass grafting. Ann Thorac Surg. 1998;65:1639–1644
- Neuropsychological changes after cardiopulmonary bypass for coronary artery bypass grafting. Thorac Cardiovasc Surg. 1998;46:207–212
- . Neuronal and astroglial injuries in patients undergoing coronary artery bypass grafting and aortic arch replacement during hypothermic cardiopulmonary bypass. Anesth Analg. 2000;91:1066–1072
- Is there an association between release of protein S100β during cardiopulmonary bypass and memory disturbances?. Scand Cardiovasc J. 2002;36:117–122
- . Biochemical markers of neurologic injury in cardiac surgery: the rise and fall of S100β. J Thorac Cardiovasc Surg. 2001;122:1019–1020
- . Extracranial sources of S100β. Br J Anaesth. 2001;86:601
- Intraoperative embolus formation during cardiopulmonary bypass affects the release of S100β. J Thorac Cardiovasc Surg. 1999;47:166–169
- Leukocyte filtration improves brain protection after a prolonged period of hypothermic circulatory arrest: A study in a chronic porcine model. J Thorac Cardiovasc Surg. 2000;120:1131–1141
- Are serum S100β protein and neuron specific enolase predictors of cerebral damage in cardiovascular surgery?. J Cardiothorac Vasc Anesth. 2003;17:4–9
PII: S1053-0770(02)47701-4
doi:10.1053/jcan.2003.1
© 2003 Published by Elsevier Inc.
Volume 17, Issue 1 , Pages 1-3, February 2003
