| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the Departments of *Anesthesia and 
Pediatrics, ||Division of Pediatric Cardiology; #Division of Biostatistics, University of California at San Francisco, San Francisco, California; Department of Anesthesiology, Division of Pediatric Cardiovascular Anesthesiology, Baylor College of Medicine, Houston, Texas; Department of Pediatrics, Division of Pediatric Cardiology, Stanford University School of Medicine, Stanford, California; and ¶Department of Anesthesiology, University of Utah, Salt Lake City, Utah.
Address correspondence and reprint requests to Kathryn Rouine-Rapp, MD, Department of Anesthesia, University of California-San Francisco, 521 Parnassus Ave., San Francisco, CA 94143-0648. Address e-mail to rouinerk{at}anesthesia.ucsf.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
In infants with transposition of the great arteries (TGA) who have undergone an arterial switch operation, postoperative ischemia is a cause of morbidity and mortality (5–11). In these patients, SWMA have been used to define myocardial ischemia, detect coronary artery perfusion impairment, and assess cardiac function in the intensive care unit in the immediate postoperative period (6,7,12,13). Despite the use of TEE to detect SWMA in these neonates, it is not known if SWMA predict myocardial injury.
In pediatric patients, after cardiac surgery, cTnI levels have been used as biochemical markers of myocardial injury, to define lesion-specific patterns of cTnI levels, and to predict postoperative clinical outcome (14–16). However, in neonates after an arterial switch operation, the association between cTnI levels and clinical outcome has been inconsistent (15,16). In addition, SWMA and cTnI levels have not been correlated. Therefore, the present study was conducted in neonates undergoing an arterial switch operation to determine if new SWMA after the termination of CPB correlate with elevated cTnI levels as a measure of myocardial injury or predict early clinical outcome.
|
METHODS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Known risk factors for increased postoperative morbidity and mortality after an arterial switch operation were recorded with preoperative data and included postnatal diagnosis, smaller birth weight, presence or absence of chromosomal abnormalities, right ventricular (RV) hypoplasia, and aortic arch repair at a separate procedure before an arterial switch operation (17,18).
Anesthetic Summary
For induction and maintenance of anesthesia before CPB, neonates received up to 3% sevoflurane or 0.5% isoflurane in combination with 10–20 mcg/kg of fentanyl. Pancuronium 0.2 mg/kg was given to provide muscle relaxation. During CPB, neonates received up to 200 mcg/kg of fentanyl, 0.4 mcg/kg of midazolam, 0.2 mg/kg of pancuronium, and up to 0.5% isoflurane. After separation from CPB, the neonates received up to 50 mcg/kg of fentanyl, 0.4 mg/kg of midazolam, and pancuronium 0.2 mg/kg.
TEE Examination
After placement of the endotracheal tube and an intraarterial catheter, a pediatric biplane TEE probe was inserted. The probe inserted was either a dual-frequency pediatric biplane HP® TEE probe (Hewlett-Packard, Palo Alto, CA) or a dual-frequency pediatric biplane Acuson® TEE probe (Acuson, Mountain View, CA). For study purposes, two cross sections of the left ventricle (LV) were obtained to provide myocardial segments supplied by the right, circumflex and left anterior descending coronary arteries and to provide cross sections at the apical, mid, and basal levels. LV transgastric mid short axis and mid esophageal four-chamber cross-sectional views were obtained and recorded for subsequent analysis three times; before sternal incision, after CPB immediately after administration of protamine sulfate, and at the time of chest closure.
TEE Analysis
During analysis, we determined the tomographic correctness of each cross section acquired. A cross section was eliminated from further analysis, if any essential anatomical structure was missing. A myocardial segment was judged adequate for interpretation using published criteria (19). The LV was divided into eight segments, using a modification of the model recommended by the American Society of Echocardiography Council for Intraoperative Echocardiography and the Society of Cardiovascular Anesthesiologists Task Force for Certification in Perioperative Transesophageal Echocardiography (20); that is, three subdivisions with respect to the long axis: basal, mid, and apical, and four with respect to the short axis: anterior, lateral, inferior, and septal, resulting in the following segments for analysis in the neonates: lateral basal, septal basal, anterior midpapillary, lateral midpapillary, inferior midpapillary, septal midpapillary, lateral apical, and septal apical. Segments seen in more than one cross section were counted only once when abnormal segments were tallied.
Assessment of Segmental Function
Each segment was inspected visually for myocardial thickening and the percent shortening of an imaginary radius from the endocardial border to the center of the LV. A grade for each segment was assigned on the basis of the area within each segment with the most abnormal motion and thickening: 0 = inadequate segmental resolution; 1 = normal; 2 = mild hypokinesis; 3 = severe hypokinesis; 4 = akinesis; and 5 = dyskinesis. Grades 1 and 2 were classified as normal function and grades of 3, 4, and 5 as abnormal function. Segments from the intraoperative period were graded independently by two investigators (KRR and MKC) who were blinded to each other’s analysis, patient identity, and clinical information. Agreement between these investigators was defined as independently assigned grades of 0 or grades within the normal (Grades 1 or 2) or abnormal (Grades 3–5) range. The classification assigned to segments when the investigators independently agreed was considered final. When one investigator classified function as normal and the other abnormal, or when the two investigators differed in their assessment of segmental resolution (Grade 0 vs any other grade), the investigators met and assigned a class of function by consensus. The segment was eliminated from further analysis, if the investigators could not agree on a consensus classification.
ECG and 24-h Holter Evaluation for Ischemia
Each patient’s cardiac rhythm status was evaluated for ST segment changes suggestive of ischemia and for the presence of atrial or ventricular tachyarrhymthias. All patients underwent a standard pediatric 15-lead ECG preoperatively and within 24 h of admission to the cardiac intensive care unit. The 15-lead ECG consists of limb leads, precordial leads (V1–V6), two right precordial leads (V3R, V4R), and an additional left precordial lead (V7). In the immediate postoperative period, if the patient had an open sternum or had large sternal bandages, then the ECG was completed with as many leads as possible.
In addition to the ECG, the initial 10 neonates who entered into the study also had a 24-h continuous dual-channel ECG recording (Holter monitoring). In these patients, five additional ECG leads were placed intraoperatively for Holter monitoring. The recordings began immediately after removal of the aortic cross-clamp (AoXC) and resumption of cardiac electrical activity. One neonate was deleted from study because of the failure of the tape to record the ECG. ECG data were stored for subsequent analysis. Holter ECG tapes were scanned first by the Zymed® Holter Scanner (2010 DOS) 1990 E (Zymed Laboratories, South San Francisco, CA), then by an experienced technician. After automated and technician scanning, they were reviewed by an expert pediatric electrophysiologist (KC), who was blinded to patient identity and clinical information and who identified findings suggestive of ischemia and atrial and ventricular arrhythmias.
For study purposes, findings suggestive of ischemia included ST segment changes of more than 1 mm ST depression or elevation from baseline measured at 20 ms after the J point in the limb leads or ST segment changes of more than 2 mm ST depression or elevation from baseline in the precordial leads. Supraventricular tachycardia included sustained (>6 beats) and nonsustained narrow complex tachycardia. Ventricular tachycardia included sustained (>6 beats) and nonsustained wide complex tachycardia. Any therapy for the arrhythmia, including administration of antiarrhythmic medications, pace termination, or cardioversion also was recorded.
Cardiac Troponin I
A baseline cTnI level was drawn before sternal incision. In the first six neonates, blood samples were obtained 6, 12, 24, 48, and 72 h after removal of the AoXC. Adult values peak 6 h after removal of the AoXC, but reported peak values are inconsistent in pediatric patients (21). To accurately determine the time of cTnI peak levels, in subsequent neonates an additional blood sample was obtained 3 h after removal of the AoXC. One of two commercial assays was used to determine cTnI levels, the Dade Stratus II® (Dade Behring, Deerfield, IL) or the Abbott AxSYM® (Abbott Laboratories, Abbott Park, IL). Both assays use a two-site fluorescence immunoassay to determine cTnI levels. Although the assays changed during the study, all assays are considered useful for detection of cardiac injury, despite differences in absolute results (22). Thus, we assigned cTnI values a score using a scale of 1–5 on the basis of each assay. Five was assigned to the highest cTnI value in each assay; then, quartiles were established to define scales 1–4. For the Dade Stratus II® assay, Score 5 = absolute value of 50 µg/L, Score 4 = 37.6–49.9 µg/L, Score 3 = 25.1–37.5 µg/L, Score 2 = 12.6–25 µg/L, and Score 1 = 1–12.5 µg/L. For the AxSYM® assay, Score 5 = absolute value of 400 µg/L, Score 4 = 300–399 µg/L, Score 3 = 200–299 µg/L, Score 2 = 100–199 µg/L, and Score 1 = 0–99 µg/L. These scores were used for subsequent analysis. Postoperatively, clinicians responsible for primary care of the neonates were blinded to the cTnI levels.
Surgical Summary
After systemic heparinization to provide an activated clotting time of >475 s and the start of CPB, each patient was cooled to 15°C over a period of 20 min. During cooling, pH stat was used. After cooling and placement of the AoXC, the heart was arrested using a single dose of 20 mL/kg of cold cardioplegic solution (Plegisol®, Abbott Laboratories, North Chicago, IL). During CPB, flow rates ranged from 140–180 mL · kg–1 · min–1 and were decreased to 60 mL · kg–1 · min–1, while the AoXC was in place. Coronary artery origins and epicardial course were determined by the attending surgeon and classified as usual [the left coronary ostium originated in the left-facing aortic sinus and gave rise to the left anterior descending and circumflex (Cx) coronary arteries, and right coronary ostium originated in the right-facing sinus and gave rise to the right coronary artery (RCA)], or a variant. To provide appropriate positioning of a coronary artery, the attending surgeon decided to use a trap-door technique for coronary implantation in some patients. After completion of the surgical repair, RV pacing wires and intracardiac lines were placed. At the discretion of the attending surgeon, all neonates were given dopamine 5–10 mcg · kg–1 · min–1, and if the TEE suggested that LV function was not optimal, then they were given epinephrine 0.03–0.05 mcg · kg–1 · min–1. Other recorded data included return to CPB because of the poor LV function, total time of CPB and AoXC, and method of sternal closure.
Postoperative Transthoracic Echocardiogram
Postoperatively, all neonates underwent a TTE before hospital discharge. Myocardial segments were analyzed and assessed by a single investigator (KRR) who was blinded to patient identity and clinical information. All segments graded 3, 4, or 5 and a random selection of segments graded 1 or 2 were analyzed and assessed by a second investigator (IR) who also was blinded to patient identity and clinical information. Segments were reported as abnormal only when there was agreement between investigators.
Statistical Analysis
Primary outcome were SWMA, cTnI levels, life-threatening cardiac arrhythmias, cardiac failure, myocardial ischemia, or MI and death. Secondary outcomes were days of inotrope use and mechanical ventilation, days to sternal closure, and lengths of stay in the pediatric intensive care unit (PICU) and hospital. Except where noted, results are expressed as mean ± sd.
For ordered variables, the Mann–Whitney and Kruskal– Wallis tests were used for comparisons among groups. Comparisons for categorical variables were performed using the Fisher’s exact test. To compare two different cTnI assays, cTnI levels were categorized within each assay. Alternative scoring systems such as z-scores and percentiles were explored and shown to correlate >0.95 with our scoring system. Missing cTnI scores were estimated using linear interpolation or last observation carried forward for 72-h data points.
Area under the curve (AUC) for cTnI scores was calculated on the basis of quartile classification within each assay. Confidence limits for the AUC of patients with normal wall motion were defined, then compared with patients with SWMA. Sensitivity and specificity of the SWMA to discriminate quartile-based AUC troponin scores were explored for several AUC cut-off points.
Associations with postoperative SWMA were analyzed using Spearman nonparametric correlation and Mann–Whitney tests. Statistical analysis was performed by a biostatistician (AB) at UCSF, using the computing software SAS®, Version 8.2 (SAS Institute Cary, NC). A P value of <0.05 was considered significant.
|
RESULTS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Intraoperative SWMA
Prior to sternal incision, a total of 232 myocardial segments were stored for analysis. In all of these myocardial segments, resolution was adequate and wall motion was normal. After CPB, a total of 464 myocardial segments were stored for analysis. Sixteen segments received a grade of 0 (inadequate segmental resolution), and 356 received a grade of 1 or 2 and thus were classified as normal wall motion, and 92 received a grade of 3, 4, or 5 and thus were classified as abnormal wall motion. Nine patients maintained normal segmental wall motion during each of our three TEE examinations. In 20 patients, at least one myocardial segment with abnormal wall motion was observed following surgical repair. SWMA were transient in five of the 20 patients and in these five, the median number of abnormal segments was one. In 15 of the 20 patients, SWMA were present at the time of chest closure and in these 15, the median number of abnormal segments was four. Patients in whom SWMA were present at the time of chest closure had more abnormal segments than patients in whom the SWMA were transient (P < 0.001). SWMA did not occur more often in patients who underwent patch closure of a ventricular septal defect (VSD).
Surgical Data
The preoperative diagnosis of Taussig-Bing malformation was confirmed at the time of surgery in one neonate. In all others, the diagnosis of TGA was confirmed. In neonates with normal wall motion, coronary artery anatomy was usual (n = 6), single RCA (n = 2), and Cx from the RCA (n = 1). All five neonates with transient SWMA and 10 patients with SWMA at chest closure had usual coronary artery anatomy. In the remaining five patients with SWMA at chest closure, coronary artery anatomy was single RCA (n = 2), inverted R and Cx (n = 1), Cx from the RCA (n = 1), and one neonate had coronary artery anatomy previously associated with ischemia, a single left coronary artery with an intramural course (5). A trap-door technique was used for coronary reimplantation in three patients in the normal SWMA group and four in the group with SWMA at chest closure. In neonates with normal wall motion, four had an intact ventricular septum (IVS) and five had a VSD. Three of the patients with transient SWMA had an IVS, two had a VSD, and of the 15 patients with SWMA at chest closure, 11 had an IVS and four had a VSD. Coronary artery anatomy, the decision to use a trap-door technique for coronary reimplantation, and presence or absence of a VSD were not different among the three patient groups.
Cardiopulmonary Bypass
Mean total CPB and AoXC times for patients with normal wall motion, transient SWMA, and SWMA at chest closure are listed in Table 1. Five patients in the group with SWMA at chest closure returned to CPB immediately after initial separation because of decreased LV systolic function. CPB times in these five patients were added and reported as total CPB. CPB time trended higher in patients with transient SWMA and SWMA at chest closure than those with normal wall motion, but the difference was not significant. Mean AoXC times were not different among the groups, but were longer in patients with a VSD than those without, 110 ± 24 vs 86 ± 18 min (P = 0.01).
|
Prior to weaning from CPB and at the discretion of the attending surgeon, dopamine 5–10 mcg · kg–1 · min–1 was given for all neonates. Additional inotropes given for 23 neonates were epinephrine 0.05 mcg · kg–1 · min–1 or milrinone 0.5 mcg · kg–1 · min–1. Three neonates in the group with persistent SWMA also received an IV infusion of nitroglycerine 1–3 mcg · kg–1 · min–1.
Primary and Secondary Outcome
Presternotomy cTnI levels were normal in all neonates. Mean AUC cTnI scores were different among the three groups of patients (P = 0.047). Between-group pairwise comparisons showed a difference between neonates with normal wall motion and SWMA at chest closure (P = 0.02). For neonates with normal wall motion, transient SWMA and SWMA at chest closure, the AUC were 16.8 ± 17.8, 38.7 ± 26.7, and 62 ± 55.5, respectively. Mean quartile cTnI scores for each group are shown in Figure 1. SWMA optimally discriminated AUC divided at 16.5. The sensitivity and specificity of SWMA to discriminate among quartile-based AUC is shown in Table 2. Maximum cTnI score occurred at the 3 or 6-h time point and was not different in patients with or without SWMA.
|
|
Twenty-four Holter ECG recordings were available in nine neonates. In two patients, wall motion was normal intraoperatively, and no ST segment changes or arrhythmias were detected by Holter ECG. ST segment changes suggestive of myocardial ischemia were detected by Holter ECG in all of the remaining seven patients in whom SWMA were transient (n = 1) or present at chest closure (n = 6). In this subset of patients, the probability of the absence of ST segment changes suggestive of myocardial ischemia in all patients with normal wall motion and presence of ST segment changes suggestive of myocardial ischemia in all patients with SWMA by Fisher’s exact test is P = 0.026.
Postoperatively, 15-lead ECGs were available in 22 of 29 patients. ST changes suggestive of myocardial ischemia were detected in two patients with normal wall motion and three with SWMA at chest closure. Of note, myocardial ischemia was not detected by 15-lead ECG in any of the nine patients who underwent Holter ECG recordings. Overall, of the patients in whom ECG data were available, ST segment changes suggestive of myocardial ischemia were present by Holter or 15-lead ECG in two of eight (25%) patients with normal wall motion, one of five (20%) patients with transient SWMA and nine of 13 (69%) patients with SWMA at chest closure. Among the three groups, the rate of ischemia detection was not different (P = 0.09).
ST segment changes suggestive of myocardial ischemia were detected by Holter or 15-lead ECG in all five patients who returned to CPB because of depressed LV systolic function.
Life-threatening cardiac arrhythmias (sustained ventricular tachycardia and incessant atrial tachycardia) were detected in three neonates who had SWMA at chest closure and Holter or 15-lead ECG evidence of ischemia; one neonate with incessant atrial tachycardia underwent cardioversion and drug therapy with amiodarone. One neonate with SWMA at chest closure and postoperative 15-lead ECG, evidence of ischemia was noted to have an apical area of MI at the time of sternal closure in the PICU. There were no deaths. Days of inotrope use and mechanical ventilation, days to sternal closure, and lengths of stay in the PICU and hospital were not different among patient groups and are given in Table 3.
|
Postoperative TTE
TTE done within 60 days of surgery were available in 24/29 patients. A total of 192 myocardial segments were analyzed. Twenty-six segments received a grade of 0 (inadequate segmental resolution), 125 received a grade of 1 or 2 and thus were classified as normal wall motion, and 41 received a grade of 3, 4, or 5 and thus were classified as abnormal wall motion. Two patients with normal wall motion intraoperatively and eight patients with SWMA at chest closure had SWMA at the time of the postoperative TTE. Intraoperative and postoperative abnormal myocardial segments, cTnI scores, and ECG data are summarized in Table 4.
|
Patients with life-threatening arrhythmias versus patients without had more abnormal myocardial segments during the postoperative TTE, 6 ± 1.7 vs 1 ± 2, respectively (P = 0.007). Postoperative abnormal segments correlated with the number of intraoperative abnormal myocardial segments and cTnI scores at 12, 24, and 48 h, and number of intraoperative abnormal myocardial segments. Confidence intervals, Spearman rank correlation, and P values for these data are given in Table 5.
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
In adults, SWMA have been established as sensitive markers of myocardial ischemia (1,3,23) and the association between ECG ischemic changes and SWMA increases when at least one-third of the myocardial segments are abnormal (24). At least one-third of myocardial segments were abnormal in 10 of 15 patients with SWMA at the time of chest closure. In addition, patients who undergo an arterial switch operation are at risk for ischemia in the immediate postoperative period (11,12). We are unaware of previous studies that have shown an association between SWMA at the termination of an arterial switch operation and ischemia in the immediate postoperative period.
In adults, the severity of SWMA may be an important predictor of myocardial injury (23). To ensure the reliability of our qualitative assessment of SWMA, we used stringent criteria to assign a wall motion grade for each myocardial segment. Only segments with severe hypokinesia, akinesia, or dyskinesia were classified as abnormal. Thus, our data represent severe SWMA which were present in multiple myocardial segments.
Not all SWMA may be indicative of ischemia. The etiology of SWMA immediately after CPB could include reversible physiologic changes that produce stunned or hibernating myocardium. Some of the SWMA in our patients could be caused by such physiologic changes, but our cTnI data suggest that the likely cause was myocardial injury. Further evidence for the association between SWMA and ischemia is the correlation between cTnI levels and SWMA present at the time of the postoperative TTE. Hernandez-Pampaloni et al. (25) examined three infants with severe SWMA up to two months after an arterial switch operation and found a decrease in myocardial segmental perfusion by high resolution positron emission tomography and coronary artery stenosis by angiography.
The pattern of our cTnI levels is similar to those reported by others in neonates who have undergone an arterial switch operation (14,16). However, inconsistencies in cTnI assays, sampling times, and differences in absolute values preclude direct comparisons between studies. Assays to measure cTnI levels are considered useful for detection of myocardial injury despite differences in absolute results (22), and we developed a scoring system to compare cTnI levels from the two assays used in this study. Our scoring system was validated by the use of alternative scoring systems such as z-scores and percentiles that demonstrated excellent correlation (>0.95).
Others have established a cut-off level of cTnI to predict early clinical outcome after pediatric cardiac surgery, but similar predictions used for patients after an arterial switch operation have shown weak correlations (16,26). In our study, transient SWMA or SWMA at chest closure showed reasonable sensitivity and specificity in discriminating quartile-based troponin AUC levels. Thus, we believe that these data demonstrated a relationship between SWMA and elevated cTnI levels.
Our study is limited in that we included a small number of neonates and did not extend our observations beyond the perioperative period. Thus, we cannot determine the long-term significance of the perioperative ischemia in these patients. However, we have determined that persistent severe SWMA found in neonates after an arterial switch operation defines a group of patients at increased risk for myocardial ischemia in the immediate postoperative period.
|
ACKNOWLEDGMENTS |
|---|
|
Footnotes |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. P. Graham Jr The Year in Congenital Heart Disease J. Am. Coll. Cardiol., July 24, 2007; 50(4): 368 - 377. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
