| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2001;71:1913-1919
© 2001 The Society of Thoracic Surgeons
a Michael E. DeBakey Department of Surgery, Division of Cardiothoracic Surgery, Baylor College of Medicine, The Methodist Hospital, Houston, Texas, USA
Address reprint requests to Dr Coselli, 6560 Fannin, #1100, Houston, TX 77030
e-mail: jcoselli{at}bcm.tmc.edu
Presented at the Thirty-sixth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 31–Feb 2, 2000.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and methodsResultsCommentDiscussionReferences |
|---|
Methods. Thirty-nine consecutive patients underwent thoracic aortic repairs during HCA; RCP was used in 25 patients. Serum S100ß was measured preoperatively, after cardiopulmonary bypass, and 24 hours postoperatively.
Results. Neurologic complications occurred in 3 patients (8%). These patients had higher postbypass S100ß levels (7.17 ± 1.01 µg/L) than those without neurologic complications (3.63 ± 2.31 µg/L, p = 0.013). Patients with S100ß levels of 6.0 µg/L or more had a higher incidence of neurologic complications (3 of 7, 43%) compared with those who had levels less than 6.0 µg/L (0 of 30, p = 0.005). Retrograde cerebral perfusion did not affect S100ß release.
Conclusions. Serum S100ß levels of 6.0 µg/L or higher after HCA correlates with postoperative neurologic complications. Using serum S100ß as a marker for brain injury, RCP does not provide improved cerebral protection over HCA alone.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentDiscussionReferences |
|---|
Astrocyte protein S100ß has recently emerged as a potential serum marker for ischemic cerebral injury. Elevated serum S100ß has been reported after routine cardiopulmonary bypass (CPB) in both adult and pediatric cardiac procedures [7, 8]. S100ß levels have also been shown to correlate with adverse neurologic outcomes after CPB [9, 10]. Studies of the role of S100ß as a marker for cerebral damage in patients undergoing aortic reconstruction requiring HCA are limited. Although total circulatory arrest time has been shown to correlate with elevated serum S100ß levels, there have been no studies correlating elevated serum S100ß with adverse neurologic events in this group of patients [11, 12]. Additionally, as a potential marker for cerebral injury, S100ß may be useful in assessing the efficacy of adjuncts for cerebral protection. Therefore, the goals of this study were [1] to determine whether elevated serum S100ß correlates with neurologic complications in patients requiring HCA during thoracic aortic repair, and [2] to determine the impact of RCP on S100ß release in this setting.
|
Material and methods |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentDiscussionReferences |
|---|
Anesthetic and surgical protocol
Standardized anesthesia for all patients included intravenous fentanyl (50 µg/kg), pancuronium, and etomidate for induction and isoflurane or enflurane inhalation for maintenance. All patients received mannitol (0.5 mg/kg) intravenously for cerebral protection. Pentothal was not used during the study period.
All procedures were performed by a single surgeon (J.S.C.) using techniques for HCA, RCP, and aortic replacement as described previously [13, 14]. For proximal aortic operations, standard CPB was established through a full median sternotomy. The femoral artery or ascending aorta was used for arterial cannulation. Bicaval cannulation was used whenever possible so that RCP could be administered. In cases with extremely large ascending aortic aneurysms or severe adhesions resulting from previous operations, bicaval cannulation was not feasible, precluding the use of RCP; in these situations, a single dual-stage atriocaval cannula was placed. A left ventricular sump was placed through the right superior pulmonary vein. Cardiotomy suction effluent was recirculated through the CPB circuit.
During the cooling phase, electroencephalography (EEG) was monitored continuously. Ice packs were placed around the head to enhance cerebral hypothermia. Circulatory arrest was initiated after electrocerebral silence was obtained. Alpha-stat pH management was used.
When RCP was used during HCA, perfusate was delivered continuously through the superior vena cava cannula at flows ranging between 150 and 500 mL/min to maintain a central venous pressure below 25 mm Hg. After completion of the distal anastomosis, the aortic graft was cannulated for arterial flow, CPB was resumed, and RCP was discontinued. During rewarming and completion of the proximal anastomosis, myocardial protection was provided with combined antegrade and retrograde blood cardioplegia. Rewarming was carried out to a rectal temperature of 37°C. Blood from the cell-saving device was administered to the patient as needed.
One patient required HCA for repair of an aneurysm involving the distal transverse arch and descending thoracic aorta. The aortic repair was approached through a left posterolateral thoracotomy. Cardiopulmonary bypass and HCA were achieved using femoral arterial and venous cannulation and EEG monitoring.
Serum S100ß levels
Blood samples were taken from a central venous catheter at induction of anesthesia, 30 minutes after the completion of CPB, and 24 hours postoperatively. After centrifugation, the samples were frozen to -20°C for batch analysis. Serum S100ß levels were measured using a standard immunoradiometric assay (Santec 100; Santec Medical AB, Broma, Sweden).
Outcome definitions
Operative mortality includes all deaths occurring in hospital or within 30 days, as well as any subsequent deaths clearly related to the procedure. The patients were evaluated daily for clinical signs of neurologic dysfunction by a physician on the surgical service. Neurologic complications were defined as stroke (transient or permanent), encephalopathy, or seizures. Strokes were defined as any new clinically evident brain injury present after operation, including focal and global deficits, and transient and permanent deficits. Respiratory failure was defined as reintubation or tracheostomy for prolonged ventilator support. Cardiac complications included myocardial infarction, dysrhythmias, persistent low cardiac output requiring inotropic support or intraaortic balloon pump placement, cardiac tamponade requiring drainage, and congestive heart failure.
Statistical analysis
The statistical analysis was performed using the SAS system for Windows (release 6.12; SAS Institute, Cary, NC). Continuous variables are reported as mean ± standard deviation. Fisher exact and Student’s t test were used for comparing categorical and continuous variables, respectively; p values less than 0.05 were considered statistically significant.
In addition to analyses based on absolute S100ß levels, the patients were divided into groups based on peak S100ß levels (ie, 6.0 µg/L or higher versus less than 6.0 µg/L) and whether they received RCP during HCA.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentDiscussionReferences |
|---|
Demographic details for the 37 patients are given in Table 1. Most patients (65%) underwent graft repair of the ascending aorta and transverse hemiarch. One patient underwent graft repair of transverse hemiarch alone after a previous composite valve graft aortic root replacement. One patient underwent graft repair of the distal transverse arch and descending thoracic aorta through the left chest. Retrograde cerebral perfusion was used in 25 patients (68%) with a mean duration of 20.8 ± 8.8 minutes.
|
|
|
Comparing patients in the RCP and non-RCP groups respectively, the mean serum S100ß levels were 0.09 and 0.09 µg/L (p = 0.83) preoperatively, 3.8 and 4.2 µg/L (p = 0.60) 30 minutes after CPB, and 0.82 and 0.53 µg/L (p = 0.41) 24 hours postoperatively. The patient characteristics for these subgroups are compared in Table 3.
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentDiscussionReferences |
|---|
Elevated S100ß levels have been associated with neurologic complications after standard CPB [9, 10, 18–20]. Although the degree of S100ß release varies with the type of operation performed, in all cases the levels peak at 30 minutes after the operation [21]. In 4 patients with postoperative cerebral complications after coronary artery bypass or valve replacement, Blomquist and colleagues [9] reported a mean S100ß level of 6.16 ± 1.13 µg/L. Three of the 4 patients (75%) in their study had levels exceeding 6.0 µg/L.
Operations performed during HCA are associated with the highest increase in S100ß levels [17]. A report by Johnsson and coworkers [18] included 2 patients who had aortic arch repairs requiring HCA; 1 patient who suffered a fatal stroke had an S100ß level of 9.5 µg/L on the second postoperative day. Astudillo and coworkers [11] reported a mean post-CPB S100ß level of 2.37 µg/L for 10 patients who underwent HCA. In the 1 patient who had a postoperative stroke, the post-CPB level was 10.89 µg/L. Although these anecdotal reports are suggestive, a clear association between S100ß elevation and neurologic complications after HCA has not been established. Our data support the association between elevated S100ß levels and postoperative neurologic complications.
To be a useful marker in the clinical setting, S100ß must either mirror subtle subclinical neurologic injury or provide early predictive information of overt neurologic damage (ie, during operation or before awakening from anesthesia). Ideally, S100ß levels would predict the degree of neurologic damage in a quantitative fashion. Practically, however, a threshold level would suffice to guide the initiation of treatment or the use of a particular adjunct. A threshold level of 6.0 µg/L—chosen based on the report by Blomquist and coworkers [9]—appears to be particularly sensitive for neurologic injury. Patients with S100ß levels higher than 6.0 µg/L had a higher incidence of postoperative neurologic complications than patients with lower levels.
A limitation of this study is that only major neurologic complications were analyzed. We did not attempt to evaluate postoperative neurocognitive dysfunction or other more subtle forms of neurologic injury. A correlation between S100ß and neurocognitive dysfunction has been difficult to establish [22]. This difficulty may reflect a fundamental difference in the location of injury resulting in S100ß release, which is primarily the white matter tracks, compared with those resulting in neurocognitive defects, which reflect primarily gray matter. Additional research in this area is warranted.
Recent studies have shown a correlation between S100ß release and total CPB time depending on the type of cardiac procedure performed [7, 10, 19, 23, 24]. Taggart and coworkers [24] found that S100ß release correlated with CPB time during coronary artery bypass but not during primary intracardiac operations (ie, valve replacement). This difference was attributed to microembolic events in the intracardiac group that may have obscured the relationship between CPB and S100ß release. Although studies in patients undergoing HCA for aortic arch repairs have shown no correlation with S100ß release and CPB time, S100ß release did correlate with total circulatory arrest time [11, 12]. In our patients, however, there was no correlation between S100ß level and either total CPB or HCA times. This lack of correlation may be due to the small number of patients studied and the variation in extent of aortic reconstruction performed in these patients combined with the inconsistent use of RCP.
Factors affecting S100ß release and potential contamination sources may confound correlations with postoperative neurologic outcomes. During the first 24 hours after CPB, S100ß levels can be elevated to levels similar to those found after major cerebral strokes [25]. The levels rapidly decline in patients without neurologic injury, but remain elevated in those who have a severe neurologic insult. Thus, it is questionable whether S100ß levels measured within the first 24 hours after operation will accurately predict adverse neurologic outcomes. Although this may limit the predictive role of S100ß in patients undergoing routine CPB, it is unclear whether similar limitations will affect patients undergoing HCA, in whom peak S100ß levels are much higher [22].
The possibility of S100ß contamination from cardiotomy suction effluent has been raised recently. In patients undergoing CPB for coronary operations, Jönsson and coworkers [20] found significantly higher S100ß levels in shed mediastinal blood from the cardiotomy suction reservoirs, compared with levels in the washed cell-saver blood. The introduction of cardiotomy suction effluent back into the CPB circuit was found to increase the serum S100ß concentration in a linear fashion depending on the volume of reinfused blood. In contrast, washed cell-saver blood had only a small fraction of the S100ß found in cardiotomy suction effluent and had minimal effects on serum S100ß when reinfused. This study has raised the question as to whether reinfusion of cardiotomy effluent into the bloodstream causes a transient rise in S100ß that may be confused with a cerebral source of S100ß. Although the addition of mediastinal-derived S100ß to the bloodstream would be expected to influence measured serum values, this effect should be similar in all patients undergoing a given operation through a standard sternotomy. An unusual rise in serum levels above that expected from routine CPB and reinfusion of mediastinal-derived blood would likely represent a cerebral source.
The potential for using S100ß to assess both the mechanism and the efficacy of protective adjuncts is illustrated by recent data regarding RCP. The theoretical benefits of RCP include [1] retrograde flushing of embolic and gaseous debris, [2] cerebral metabolic support and removal of metabolite waste, and [3] more homogenous cooling of the brain [26, 27]. Wong and Bonser [28] measured the S100ß gradient between afferent blood (superior vena cava) and efferent blood (carotid artery) during RCP and demonstrated true reversal of brain blood flow in humans. Our data corroborate previous studies that have compared S100ß levels in patients undergoing HCA with and without RCP; the use of this adjunct has not reduced post-HCA S100ß levels [29–31]. With the recent resurgence of interest in using antegrade cerebral perfusion during aortic arch repair, future clinical trials comparing these adjuncts are inevitable; S100ß release will be an important variable in such studies.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentDiscussionReferences |
|---|
I would like to ask why do you think the S100 levels you found were so elevated, and why were they increased so early after CPB? Previous literature on S100 associated with CPB describes high levels later, at the end of the operation, and up to 6 hours postoperatively. As second question, I would like to know whether you considered age-adjusting the S100 levels in your patients, and whether you had a chance to screen them for Down’s or Alzheimer’s.
Thank you.
DR COSELLI: Let me take the last question first, and I appreciate your comments very much.
First of all, none of these patients had Down’s syndrome or Alzheimer’s disease. I cannot explain why the S100 levels that we measured were maybe higher than some other series, but most other series combined patients with HCA and patients without HCA. But in all series in which patients are included with HCA, the levels are higher than those of patients without HCA. But I really cannot explain why our particular measurements went higher.
The other thing that I might add is that we standardized, so as to not confuse the issue, the use of cardiotomy suction, return of cell-saver blood, and so forth for all of these patients. They were all managed equally. And although these things have been found to affect the secretion of S100, both in CSF and in the serum postoperatively, at least in our study it was standardized across the board and balanced out for it I think.
DR JOHN W. HAMMON, JR (Winston-Salem, NC): Doctor Coselli, I want to congratulate you for your courage in measuring S100 levels in your patients. I think it is important for our understanding of these phenomena.
As you know, in patients with acute stroke, S100 levels rise to similar values and then persist for several days after the stroke. This phenomenon is similar to your patients who had the neurologic deficits following your operations. Do you think they had a formal cerebral infarction? And if so, what caused it? Do you think it was particulate matter that was introduced into the cerebral circulation either during the perfusion, during cooling, before the operation, or that particular matter was not washed out with your retrograde cerebral perfusion?
And, as a last question, would you care to maybe randomize some patients between RCP and no RCP to see if it made a difference?
DR COSELLI: Thank you for your comments. Again, I will take the last question first.
At the outset, we are looking at carrying out a similar evaluation with retrograde perfusion, nonretrograde perfusion, and antegrade perfusion, which I think will provide us with some more information. Again, this sort of study at the outset is exceedingly preliminary.
DR RAUL GARCIA RINALDI (San Juan, Puerto Rico): Doctor Coselli, you used RCP in only 60-some percent of the cases. What determined whether you used it or not?
And the second question is, would these strokes occur in the people who did not have cerebral perfusion?
DR COSELLI: All of the strokes were in patients who had RCP. We use RCP in the more complex cases. When the procedure was analyzed as a variable, it did not fall out as significant.
And to add further to that and come back to one of Dr Hammon’s questions earlier that I passed over, all of the patients with neurological deficits had postoperative CT and/or MRI scans. The 1 patient that had a definitive stroke and died from it had a positive finding on those studies.
The patient with encephalopathy and the patient with seizures on MRI and CT scanning postoperatively were normal.
|
References |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Ruel, C. Bianchi, T. A. Khan, S. Xu, J. R. Liddicoat, P. Voisine, E. Araujo, H. Lyon, I. S. Kohane, T. A. Libermann, et al. Gene expression profile after cardiopulmonary bypass and cardioplegic arrest J. Thorac. Cardiovasc. Surg., November 1, 2003; 126(5): 1521 - 1530. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Vaage and R. Anderson Biochemical markers of neurologic injury in cardiac surgery: The rise and fall of S100{beta} J. Thorac. Cardiovasc. Surg., March 1, 2003; 125(90030): S31 - 33. [Full Text] [PDF]
|
|
|
|
 |
|
 D. Harrington, C. H. Wong, and R. S. Bonser Neurological Complications of Aortic Surgery Seminars in Cardiothoracic and Vascular Anesthesia, March 1, 2002; 6(1): 7 - 16. [Abstract] [PDF]
|
|
|
|
|
|
J. Vaage and R. Anderson Biochemical markers of neurologic injury in cardiac surgery: The rise and fall of S100{beta} J. Thorac. Cardiovasc. Surg., November 1, 2001; 122(5): 853 - 855. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
).