S100B protein is present in high concentrations in glial

Transcription

1 Serum S100B and Hypothermic Circulatory Arrest in Adults Kausik Bhattacharya, FRCS, Stephen Westaby, FRCS, Ravi Pillai, FRCS, Susan J. Standing, MRCPath, Per Johnsson, MD, PhD, and David P. Taggart, MD(Hons) Oxford Heart Centre and Department of Clinical Biochemistry, John Radcliffe Hospital, Oxford, England Background. Cerebral injury is the most important complication of cardiac operations with cardiopulmonary bypass. Prolonged total circulatory arrest (TCA) can expose patients to an even greater risk of cerebral injury. We sought to detect the degree of cerebral injury in adults who had thoracic aortic operations with TCA by measuring S100B protein, which is released into the circulation after cerebral injury. Methods. Serial measurements of S100B protein, a highly specific serum marker of astroglial damage, were performed in 26 patients who had complex aortic operations, of whom 13 required cardiopulmonary bypass alone (for aortic root replacement), and in 13 patients who required an additional period of TCA (for type A aortic dissections and arch aneurysms). Blood samples were taken preoperatively, at skin closure, and 5 and 24 hours postoperatively. Results. There were significant increases in serum S100B concentrations in all patients, and peak levels occurred at skin closure. The magnitude of the increase in S100B was significantly greater at all postoperative time points and persisted longer in the TCA group. There was a significant correlation between the duration of the TCA and S100B concentration at 5 hours (r 0.66, p 0.01) and 24 hours (r 0.63, p 0.02) postoperatively. Conclusions. S100B levels were higher in all patients who had complex aortic operations and were significantly greater in patients requiring a period of TCA. The duration of the TCA period correlated with S100B levels 5 hours and at 24 hours postoperatively. Circumstantial evidence, in accordance with other studies, suggests that S100B protein is a marker for cerebral injury during cardiac operations. (Ann Thorac Surg 1999;68:1225 9) 1999 by The Society of Thoracic Surgeons S100B protein is present in high concentrations in glial and Schwann cells [1]. Measurement of S100B protein in serum has been used as a diagnostic and prognostic tool for patients with stroke, central nervous system tumors, and head injury [2 4]. Recently, elevated serum levels have been detected after adult cardiac operations complicated by stroke or subtle cerebral injury [5]. We have previously characterized the pattern of S100B release for patients who have coronary artery operations with cardiopulmonary bypass (CPB) [6]. Comparison of S100B levels during coronary artery bypass grafting (CABG) and valve operations showed significantly greater release in the latter group [7]. We also reported that an arterial line filter reduces S100B release in patients who have CABG [8]. We investigated the pattern of release of S100B protein in adults who had profound hypothermic circulatory arrest for major thoracic aortic repair. This procedure produces three different risk factors for brain injury: cerebral ischemia, particulate or air embolism, and hypothermia. Although hypothermic circulatory arrest at 18 C for 30 to 45 minutes is generally regarded as safe, previous studies suggest a greater risk of cerebral injury Accepted for publication March 22, Address reprint requests to Dr Bhattacharya, Department of Cardiothoracic Surgery, Royal Hospital For Sick Children, Yorkhill, Glasgow G3 8SJ, Scotland. for both adult and pediatric patients than after CPB alone [9 11]. To explore the significance of circulatory arrest on S100B levels over and above that of thoracic aortic repair, we studied a group of patients who had a major operation on the aortic root but with continuous cardiopulmonary bypass. Material and Methods Twenty-six patients who had complex aortic operations were studied prospectively, using blood samples taken on induction of anesthesia, at skin closure, and at 5 and 24 hours postoperatively. Thirteen patients who had aortic root repair or replacement were supported with continuous CPB at a mean temperature of 28.5 C. Thirteen other patients had operations involving the aortic arch with periods of hypothermic circulatory arrest at 20 C (Table 1). Anesthetic and Surgical Protocol Premedication was achieved with morphine (10 to 15 mg) and scopolamine (0.3 to 0.4 mg). Anesthesia was induced with fentanyl (1 mg), pancuronium (8 mg), and etomidate (4 to 10 mg). Before CPB, anesthesia was maintained with a combination of oxygen, nitrous oxide, and halothane. During CPB propofol 6 mg/kg per hour was used. No anesthetic agent was administered during hypothermic 1999 by The Society of Thoracic Surgeons /99/$20.00 Published by Elsevier Science Inc PII S (99)

2 1226 BHATTACHARYA ET AL Ann Thorac Surg S100B IN CIRCULATORY ARREST 1999;68: Table 1. Patient Characteristics Variable No TCA TCA p (Mean SEM) (Mean SEM) Value a Number Age (y) Emergency cases Cardiopulmonary bypass time (min) TCA time (min) Temperature ( C) a Student s t test. SEM standard error of the mean; TCA total circulatory arrest. circulatory arrest. Cardiopulmonary bypass was with a Cobe CML membrane oxygenator (Cobe Cardiovascular Inc, Arvada, CO) and a roller pump producing nonpulsatile flow. The circuit was primed with 1.5 L of Hartman s solution. Alpha stat CO 2 tension management was used, and the mean arterial pressure was maintained pharmacologically between 50 and 80 mm Hg. At normothermia the pump flow was 2.4 L/m 2 per minute and was reduced to 1.6 L/m 2 per minute at temperatures below 28 C. Circulatory arrest was undertaken at between 18 and 20 C. Myocardial protection was with antegrade crystalloid cardioplegia (1 L of St. Thomas s cardioplegic solution) at 4 C repeated every 40 minutes as necessary. For patients who had circulatory arrest, venous blood was drained into the oxygenator and the patient was tipped head down. Aortic arch repair was done with the brachiocephalic arteries open to air. Three of the 13 patients who had circulatory arrest had retrograde cerebral perfusion through the superior vena cava (200 ml of blood per minute at a pressure less than 25 mm Hg). S100B Assay All blood samples, obtained from arterial cannulas, were centrifuged for 5 minutes at 3,000 g. The resultant serum was frozen at 20 C and saved for batch analysis. S100B levels were measured using a monoclonal immunoradiometric assay (Sangtec 100, AB Sangtec Medical, Bromma, Sweden), which uses three monoclonal antibodies (SMST 12, SMSK 25, and SMSK 28) to detect the chains in the and dimers of S100B. This methodology has been described in detail elsewhere [8]. Briefly, the samples were incubated in a bovine serum albumen buffer and a plastic bead coated with S100B antibody. After washing the beads, S100B antibody labelled with iodine 125 was added. The antibody was seized by the bead-bound S100B and after incubation and further washing, the radioactivity of the beads were read using a gamma counter. The mean duplicate gamma count was compared with known standards, and any value over 0.2 ng/ml were recorded, as this was the limit of sensitivity of the assay. Samples were analyzed in duplicate to reject those with more than 10% variation, but this did not apply to any patients in this series. Table 2. Differences in S100B Release Patterns Between the Groups Variable Neurologic Assessment All patients had preoperative and daily postoperative clinical neurologic examination by one observer (KB) for signs of overt cerebral injury. Statistics SPSS for Windows (version 6.1.1; SPSS Inc, Chicago, IL) was used for data analysis. Demographic data are expressed using means and standard error of the mean, and serum S100B as medians and interquartile ranges. Group comparisons were made using either Mann- Whitney or Student s t tests, depending on data distribution. Spearman s correlation coefficient was used to identify correlations between S100B and total circulatory arrest (TCA) time, CPB time, and lowest core temperature. Results No TCA median (IQR) TCA median (IQR) p Value a Number Preoperative S100B Undetectable Undetectable b... (ng/ml) Skin closure S100B 0.26 ( ) 0.54 ( ) (ng/ml) 5-hour S100B (ng/ 0.00 ( ) 0.13 ( ) ml) 24-hour S100B (ng/ ml) 0.00 ( ) 0.13 ( ) a Mann-Whitney test. b Two patients had detectable preoperative levels. IQR interquartile range; TCA total circulatory arrest. The patient details, including mean age, elective or emergency operation, overall duration of CPB, and TCA time, together with the lowest intraoperative temperature are shown in Table 1. All patients had normal results of preoperative neurologic examination. In 2 patients who had emergency operation for acute type A dissection, there were minor increases in preoperative S100B levels (0.09 ng/ml in both patients). There were highly significant postoperative increases in S100B protein in both groups, with peak values at skin closure, shown in Table 2 and Figure 1. Patients who had a period of TCA had significantly higher serum S100B protein at skin closure ( p 0.06), 5 hours ( p 0.001), and 24 hours ( p 0.01) postoperatively. Neither the duration of CPB nor the duration of TCA correlated with peak S100B concentrations at skin closure. However, there was a significant correlation between the duration TCA and S100B concentration at 5 hours (r 0.66, p 0.01) (Fig 2A) and at 24 hours (r 0.63, p 0.02) (Fig 2B) postoperatively. There was a negative correlation between temperature and S100B at 5 hours (r 0.45, p 0.05).

3 Ann Thorac Surg BHATTACHARYA ET AL 1999;68: S100B IN CIRCULATORY ARREST 1227 Of the 2 patients with elevated preoperative levels of S100B, 1 had a considerably higher postoperative S100B concentration than the other patients before it returned to within normal limits by 24 hours. His S100B profiles were 0.09, 3.28, 1.14, and 0.18 ng/ml. One patient with a particularly prolonged circulatory arrest period (82 minutes) maintained low serum S100B levels throughout the postoperative period. This patient was treated with continuous retrograde cerebral perfusion during the arrest period. Of the remaining 2 patients who received retrograde cerebral perfusion, 1 had significantly elevated S100B levels and the other, who had greatly elevated S100B levels at five and 24 hours, died of an extensive stroke (Fig 3). Fig 1. A boxplot graph showing the differences in S100B profiles between the no total circulatory arrest (TCA) group and the TCA group. The whiskers represent the range of the data, the boxes express the upper and lower quartiles and the median is shown by the bold line. Patients who are outliers (more than 1.5 times the interquartile range away from the whiskers are not shown). A preoperative value; B at skin closure; C 5 hours postoperatively; D 24 hours postoperatively. The median is shown by the bold line. Outliers are represented as circles and extreme outliers are represented by asterisks. Comment This study confirms that CPB causes significant postoperative increases in S100B in most TCA and non-tca patients. In both TCA and non-tca groups the temporal response in S100B release was similar to that in patients who had coronary bypass [6, 8] or valve operations [7], where peak elevations occurred at the time of skin closure. Despite a similar overall duration in CPB (not including TCA time) in the non-tca and TCA groups, serum S100B levels were significantly higher and remained elevated for much longer in the TCA group. This result concurs with those of previous studies which suggested a greater risk of cerebral injury in both adult and pediatric patients for TCA than for CPB alone [9 11]. In the present study there was no correlation between the duration of CPB and S100B in either group at any time point. This finding is consistent with our previous finding that S100B levels do not correlate with duration of CPB in patients who have valve operations [7] and contrasts with studies that documented a correlation between the duration of CPB and S100B levels in CABG patients [6]. For patients who have open cardiac operations where some degree of air or particulate embolism is inevitable, the relationship between duration of CPB and S100B release might be obscured by microembolic events. In contrast to the duration of CPB, the duration of TCA did correlate with peak S100B levels at 5 and 24 hours postoperatively (but not at skin closure). A possible explanation for the absence of correlation at skin closure is that the cooled brain is not uniformly or completely rewarmed at that time: Consequently the cellular processes that release S100B might not have been fully activated. The inverse relationship between temperature and S100B might attenuate the S100B level at skin closure, whereas in the fully rewarmed patient (5 and 24 hours postoperatively) the correlation becomes clearer. Fig 2. Scattergraphs showing the relation between serum S100B concentrations 5 hours (A) and 24 hours (B) postoperatively and duration of total circulatory arrest.

4 1228 BHATTACHARYA ET AL Ann Thorac Surg S100B IN CIRCULATORY ARREST 1999;68: Fig 3. Trend in serum S100B levels in a patient who sustained a perioperative stroke compared with the rest of the TCA group who had no overt clinical neurologic injury. (Pre-op preoperative.) Although temperature itself could alter the release of S100B protein, at 5 hours postoperatively there was only a weak negative correlation between hypothermia and S100B (r 0.48, p 0.05). Recently, Amory and colleagues [12] showed no significant difference in S100B release for patients who had coronary artery bypass at 28 C versus 36 C. This suggests no effect of moderate hypothermia on S100B release and that the weak negative correlation described above is a function of circulatory arrest and not temperature. Similar findings were noted by others [13]. The finding of greatly elevated S100B with a progressive increase between 5 and 24 hours postoperatively is characteristic of stroke and suggests irreversibility of the cerebral event. This finding has been well documented in other studies and clearly suggests a role for S100B protein as a prognostic marker. To this end we are pursuing the development of a rapid bedside assay for S100B. As in traumatic head injury and subarachnoid hemorrhage, the ability to diagnose an acute cerebral event earlier could increase the scope for therapeutic intervention. The pathophysiology of cerebral injury after TCA can be attributed in part to a prolonged disturbance in cerebral blood flow and cerebral vascular resistance. A period of hypoperfusion triggers the excitotoxic cascade, releasing oxygen free radicals and platelet-activating factor into the interstitium, which promotes further vasoconstriction and extension of the ischemic zone. Intracellular calcium concentrations then exceed the critical level and cause brain cell injury and death (Mangano CM. Cerebral morbidity during cardiac surgery: mechanisms of injury and the effect of temperature management during cardiopulmonary bypass. Presented at the 16th Pathophysiology and Techniques of Cardiopulmonary Bypass Conference, San Diego, CA, March 1996). Microcirculatory embolism sustained during cardiopulmonary bypass can cause focal areas of hypoxia and critical hypoperfusion, resulting in cell death from energy failure. The total extent of cerebral injury will depend on the resistance of the tissues adjacent to the ischemic zone (the penumbra zone) to triggering of the excitotoxic cascade. Hypothermia could delay this process and partly explain the correlation between TCA time and S100B in the intermediate postoperative period. Two patients with acute aortic dissection had slightly high preoperative S100B levels and normal preoperative neurologic examination. Although that increase could indicate a degree of cerebral malperfusion from the dissection process, it is more likely that it represents S100B release from other sources. In our operations the key difference between patients in the continuous CPB or TCA groups was the need for an open anastomosis for aortic arch repair. Otherwise the patients were similar with respect to duration of CPB and preoperative S100B levels. The TCA patients had aneurysms extending into the aortic arch so that the brachiocephalic vessels were open to air. The continuous CPB patients were also at risk from residual air in the aorta or left ventricle after completion of the root replacement and from the inevitability of incomplete removal of air. Nevertheless, the likelihood of cerebral embolism must be greater in the open arch cases. Although the respective contributions of the variables circulatory arrest, temperature, and cerebral air embolism are unknown, the fact that those at greater risk from cerebral injury manifest higher levels of S100B protein reinforces the evidence for this agent as an important marker. The study was limited by the current sensitivity and specificity of the S100B assay. The true significance of S100B levels can be determined only by comparison with results from a standardized battery of neuropsychometric tests or by correlation with known volumetric cerebral damage. We are pursuing this line of investigation in patients who have elective operations, but there are difficulties in comparing detailed neuropsychometric analysis with S100B levels in this type of patient. Jugular bulb catheter sampling might record higher S100B levels than central venous sampling and disclose other potential causes of cerebral injury. The recent identification of glutamate receptors, thought to have a mediatory role in the excitotoxic cascade, has prompted interest in specific receptor antagonists [14]. S100B might have a role in evaluating the efficacy of potential therapies aimed at glutamate and other central nervous system receptors. Furthermore, the differential sensitivities of various parts of the brain might show a disparity between the S100B concentrations measured and other clinical parameters observed [15]. Although our results suggest that patients who have major thoracic aortic operations with total circulatory arrest are at risk of cerebral injury, the role of total circulatory arrest itself might be obscured if other factors promote S100B release. Nevertheless, we suggest that S100B can be used to monitor astroglial injury during thoracic aortic operations and that much earlier availability of results could allow evaluation of treatment strategies. We thank Mario Cortina-Borja, PhD, Department of Statistics, University of Oxford for statistical guidance; David DeKeyzer (Cambridge Life Sciences) and Madeleine Höckenström (AB

5 Ann Thorac Surg BHATTACHARYA ET AL 1999;68: S100B IN CIRCULATORY ARREST 1229 Sangtec Medical) for their support; Susmita Mukherjee, BSc, for help in preparing this manuscript; and the cardiac anesthetists (Drs Ainsworth, Evans, Fisher, Fordham, Grebenick, Nagle, and Sinclair) for collecting intraoperative blood samples. References 1. Johnsson P. Markers for cerebral injury after cardiac surgery. J Cardiothorac Vasc Anesth 1996;10: Ishiguro Y, Kato K, Ito T, Nagaya M. Determination of three different enolase isoenzymes and S100B protein in various tumours in children. Cancer Res 1983;43: Persson L, Hardemark HG, Gustafsson J, et al. S-100 protein and neuron-specific enolase in cerebrospinal fluid and serum; markers for cell damage in the human central nervous system. Stroke 1987;18: Ingebrigstein T, Romner B, Kongstad P, Langbakk B. Increased serum concentrations of protein S-100 after minor head injury: a biochemical serum marker with prognostic value? J Neurol Neurosurg Psychiatry 1995;59: Johnsson P, Lundqvist C, Lindgren A, Ferencz I, Alling C, Stahl E. Cerebral complications after cardiac surgery assessed by S-100 and NSE levels in blood. J Cardiothorac Vasc Anaesth 1995;9: Westaby S, Johnsson P, Parry A, et al. Serum S-100 protein: a potential marker for cerebral events during cardiopulmonary bypass. Ann Thorac Surg 1996;61: Taggart DP, Mazel J, Bhattacharya K, et al. A comparison of serum S100 levels during open and closed heart surgery. Ann Thorac Surg 1997;63: Taggart DP, Bhattacharya K, Meston N, et al. Serum S100 protein after cardiac surgery: a randomized trial of arterial line filtration. Eur J Cardiothorac Surg 1997;11: Mezrow CK, Gandas A, Sadeghi AM, et al. Metabolic correlates of neurologic and behavioural injury after prolonged circulatory arrest. J Thorac Cardiovasc Surg 1995; 109: Alamanni F, Agrifoglio M, Pompilio G, et al. Aortic arch surgery: pros and cons of selective cerebral perfusion. A multivariable analysis for cerebral injury during hypothermic circulatory arrest. J Cardiovasc Surg (Torino) 1995;36: Newburger JW, Jonas RA, Wernovsky G, et al. A comparison of the neurologic effects of hypothermic circulatory arrest versus low-flow cardiopulmonary bypass in infant cardiac surgery. N Engl J Med 1993;329: Amory D, Croughwell N, Lowry E, et al. Hypothermic vs normothermic cardiopulmonary bypass influence on serum S-100 levels. Circulation 1996;94 (Suppl I):I Kumar P, Dhital K, Hossein-Nia M, Patel S, Holt D, Treasure T. S-100 protein release in a range of cardiothoracic surgical procedures. J Thorac Cardiovasc Surg 1997;113: Redmond JM, Gillinov AM, Zehr KJ, et al. Glutamate excitotoxicity: a mechanism of neurologic injury after hypothermic circulatory arrest. J Thorac Cardiovasc Surg 1994;107: Treasure T, Naftel DC, Conger KA, Garcia JH, Kirklin JW, Blackstone EH. The effect of hypothermic circulatory arrest time on cerebral function, morphology, and biochemistry. J Thorac Cardiovasc Surg 1983;86: