Cerebral oximetry during infant
cardiac surgery: evaluation and
relationship to early postoperative
outcome
Cerebral oxygen saturation-time
threshold for hypoxic-ischemic
injury in piglets

Article Reviews

1. Kussman BD, Wypij D, DiNardo JA, Newburger JW, Mayer JE Jr, del Nido PJ, Bacha EA, Pigula F, McGrath E, Laussen PC. Cerebral oximetry during infant cardiac surgery: evaluation and relationship to early postoperative outcome. Anesth Analg. 2009 Apr;108(4):1122-31.

Reviewer:
Erin Gottlieb, M.D., Texas Children's Hospital

Introduction:
The aim of this observational study was to monitor cerebral oximetry during infant heart surgery and evaluate the relationship to anatomic diagnosis and early postoperative clinical outcome. The relationship between cerebral oxygen saturation and early neurologic outcome would also be evaluated.

Methods:
In this study, 104 patients who were less than 9 months of age with either d-transposition of the great arteries (D-TGA), tetralogy of Fallot (TOF) with or without pulmonary atresia or truncus arteriosus, and ventricular septal defect were anesthetized for definitive surgical repair.

Anesthetic technique, cooling protocols, the use of pH-stat strategy during cooling and the administration of steroids, phentolamine, lasix and mannitol were fairly standardized, although the anesthetic technique was not controlled.

Regional cerebral oxygen saturation data was collected at specified time points during the case and postoperatively. Postoperative data included serum lactate, cardiac output by thermodilution, hematocrit, arterial blood gas, mixed venous oxygen saturation. Postoperative events and length of intubation, ICU and hospital stay were also recorded.

The minimum regional cerebral oxygen saturation during the critical phases of CPB were recorded. Based on earlier research, an rSO2 of 45% was chosen as the threshold. The total duration and longest duration of rSO2 less than or equal to 45% and the intergrated rSO2 ≤45% were used for the analysis. The same analyses were done with 50%, 55% and 60% as the possible threshold value. Associations were assessed between NIRS values and outcome measurements.

Results:
Analysis of the variables showed no difference in perioperative outcome between patients who had NIRS monitoring versus those that did not. All patients survived to discharge and there were no adverse early neurologic events.

Regional cerebral oxygen saturation and cerebral oxygen extraction differed among diagnostic groups with the TOF group having the highest rSO2 and lowest cerebral oxygen extraction prebypass. The D-TGA group had the highest rSO2 during bypass. The D-TGA group had the highest average rSO2 and the VSD group had the lowest rSO2. There were no differences in mean minimum rSO2 among diagnosis groups. The rSO2 was ≤ 45% in 22 infants, but there was no difference among diagnosis groups as far as incidence, total and longest duration ≤ 45%.

There were no adverse neurologic events in the study patients, so the relationship between NIRS monitoring an early neurologic outcome could not be determined. There was no statistically or clinically significant association between the cerebral saturation and postoperative clinical outcomes. There was no outcome change for patients with NIRS ≤ 45% or AUC for rSO2 ≤ 45%.

The D-TGA group underwent DHCA, and some observations were made. First, there was no correlation between the rSO2 at the onset of DHCA and the rate of decline in rSO2. Second, there was no correlation between the hematocrit at the onset of low flow before DHCA and the rSO2 before CA or the rate of decline during arrest. Also, no patient reached a nadir in rSO2 by 25 minutes. The nadir could be estimated to occur at 38 minutes of DHCA using a nonlinear exponential decay model.

Discussion:
Intraoperative cerebral oxygen saturation varies according to diagnosis. Some of this reflects the baseline differences in oxygen saturation, but different lesions require different management techniques which can also be reflected in the cerebral oxygenation.

Intraoperative cerebral oxygen saturation was not associated with early hemodynamic and clinical outcome. The relationship between intraoperative cerebral oxygen saturation and early neurologic outcome could not be determined because there were no patients with seizures, stroke, or choreathetosis postoperatively.

Twenty-three of the patients had rSO2 ≤ 45%, the threshold that is associated with a bad neurologic outcome in pigs and is associated with new/worsened ischemia on MRI in neonates after the Norwood procedure. The duration of rSO2 ≤ 45% was quite short in the patients in the study and was not related to adverse early neurologic outcome. The short time of DHCA and the small sample size made it difficult to determine the nadir rSO2 during circulatory arrest and its application to early outcome.

There is still an undefined low rSO2 threshold in children. It is difficult to define for a number of reasons including the low incidence of adverse early neurologic outcomes in infants having heart surgery. In addition, since baseline cerebral saturation differs with diagnosis, it is difficult to generalize about changes.

Comments:
This study looked at intraoperative cerebral oxygen saturation for infants undergoing biventricular repair with three basic diagnoses and examined the relationship between NIRS values and early outcomes. The study confirmed that cerebral oxygen saturation differed among the diagnosis groups. The authors point out that the difference among groups is related to differing management strategies. I agree that the management strategies among the groups would differ, but it would seem that the management of patients within each group might differ as well. For example, in the TOF group, the intraoperative oxygenation and ventilation strategy for a “regular” TOF would be quite different from that of a truncus arteriosus which would certainly impact the NIRS values. In other words, there were intragroup differences in rSO2 as well as intergroup differences.

The study also examined the relationship between intraoperative cerebral oxygen saturation and early hemodynamic and clinical outcome and found that there was no relationship. The authors comment that although the study population, a group of patients with two ventricles and no aortic arch obstruction, may not have a correlation between intraoperative cerebral oxygen saturation and early outcome, there may be a correlation in patients with more complex anatomy and physiology like hypoplastic left heart syndrome or interrupted aortic arch. It may be that the study groups were “too healthy” to show a relationship. In addition, it is unclear if the anesthesiologist had access to the NIRS monitor during the case and was intervening in response to the data. Intervention by the anesthesiologist in response to a trend in NIRS data could improve the early hemodynamic and clinical outcome.

Interestingly, as the authors point out, low postoperative cerebral oxygen saturation is associated with new/worsened ischemia on brain MRI after the Norwood procedure [1]. A recently published study by Phelps, et al, concludes that low cerebral oxygen saturation after the Norwood procedure is strongly associated with adverse outcome [2].

The incidence of early neurologic complications like seizure, stroke and choreoathetosis in pediatric patients undergoing heart surgery is low. It is commendable that there were no adverse neurologic events in the study group. Therefore, there could be no relationship drawn between intraoperative NIRS and early postoperative neurologic outcome. There are uncommon but potentially neurologically devastating cases in which a major problem with cardiopulmonary bypass like aortic cannula malposition with no flow to the head during bypass. It is these cases in which intraoperative cerebral oxygen saturation would be related to early neurologic outcome if no intervention was taken to troubleshoot low rSO2. If the sample size were larger, perhaps the incidence of adverse neurologic outcome would not be zero.

In a piglet model, the time at the nadir of cerebral oxygenation during DHCA correlates with injury to the brain. In this study, the periods of circulatory arrest were short and center specific bypass management including the use of pH-stat, higher hematocrit, and longer duration of cooling. No patient in this study reached a nadir after 25 minutes of circulatory arrest. Estimation using nonlinear exponential decay suggests that a nadir may be reached after 38 minutes. This illustrates definite progress in management of bypass and DHCA.

This study examined the relationship between cerebral oxygen saturation and early neurologic, hemodynamic and clinical outcome. However, the relationship between intraoperative cerebral oxygen saturation and neurodevelopmental outcome is an area of important ongoing research.

References:

Dent CL, Spaeth JP, Jones BV, Schwartz SM, Glauser TA, Hallinan B, Pearl JM, Khoury PR, Kurth CD. Brain magnetic resonance imaging abnormalities after the Norwood procedure using regional cerebral perfusion. J Thorac Cardiovasc Surg 2005;130:1523-30
Phelps HM, Mahle WT, Kim D, Simsic JM, Kirshbom PM, Kanter KR, Maher KO. Postoperative cerebral oxygenation in hypoplastic left heart syndrome after the Norwood procedure. Ann Thorac Surg 2009;87:1490-4

2. Kurth CD, McCann JC, Wu J, Miles L, Loepke AW. Cerebral oxygen saturation-time threshold for hypoxic-ischemic injury in piglets. Anesth Analg. 2009 Apr;108(4):1268-77.

Reviewer:
Pablo Motta, MD, Texas Children's Hospital

Background:
Hypoxic-ischemic (H-I) brain injury is common in critically ill infants. Currently there is no gold standard bedside monitoring technique for the early diagnosis of hypoxic-ischemic brain injury in the immature brain. The diagnosis of H-I is made by clinical and radiological grounds which occurs after the brain is damaged irreversibly.
Near-infrared spectroscopy is noninvasive bedside cerebral oxygen saturation (ScO2) monitor that measures the hemoglobin saturation of the venous, capillary and arterial components. In piglets the normal ScO2 is 60%-75% and in a H-I piglet model reduction of the ScO2 30%-45% produces neurophysiologic impairment evidenced by increased brain tissue lactate, decreased adenosine triphosphate concentrations and slowing of electrocortical activity. This suggests that there is a viability threshold for the NIRS monitor.
The objective of this study was to define the viability-time threshold to predict neurologic injury in neonatal H-I model. The author’s hypothesis was that there is a distinct time threshold for the development of neurologic injury.

Methods:
The animal model consisted in 5 – 10 day old piglets which had the carotid arteries exposed surgically under anesthesia. The induction was performed with intramuscular ketamine 33 mg/kg and acepromazine (neuroleptic of veterinarian use) 3.3 mg/kg. After intravenous access was secured anesthesia was maintained with fentanyl 10 μg.kg-1.h1 and midazolam 0.1 μg.kg-1.h1. No muscle relaxation was used.
The neuromonitors used were the NIRS device (Near-infrared Monitoring, Philadelphia, PA, optical probe and a main unit) and cerebral function monitor (CFM 6000 Olympic Medical, Seattle, WA, single channel amplitude integrated EEG) . CFM monitors moment to moment variation in the electrocortical activity (ECA) and correlates well with EEG. NIRS and CFM data were collected every 15 minutes through the study.
The groups were randomly assigned to increasing times of H-I which it was define as a NIRS reading of ScO2 35%. There was one control group (n = 7) and 6 H-I groups which experienced ScO2 35% for 1 h (n = 4), 2 h (n = 6), 3 h (n = 6), 4 h (n = 7), 6 h (n = 8), or 8 h (n = 8). After 10 minutes stabilization a ScO2 35% was achieved by bilateral carotid occlusion and adjusting the FiO2 if needed. .
The variables measured included arterial blood pressure, blood gases, and pH. These were recorded at baseline, during H-I and at 15 min during reperfusion after H-I. Once the 75% of the planned H-I period the anesthetic infusion was discontinued to allow prompt arousal of the animal and extubation.
Post procedure animals were inspected hourly for the initial 4 h, and then every 6–8 h. Neurobehavioral examinations were performed 8, 24, and 48 h after extubation by a blind observer. These parameters evaluated were level of consciousness, sensory function, gait and motor tone.
After a two day survival periods the piglets were euthanized, sooner if they were severely disabled. A blinded neuropathologist evaluated the slides for neuronal cell death, inflammation, hemorrhage, and infarction. A semi-quantitative score was used grade neuronal damage. Animals were considered to have an abnormal neurologic outcome if the had either mild to severe disability or neuropathologic damage.

Results:
As expected H-I caused significant decrease in the ScO2 and ECA not related to PaO2 changes. During the reperfusion period the ScO2 was significantly greater than control of H-I values. On the other hand reperfusion ECA peak and trough values were significantly higher that H-I and significantly lower than the control ones.
In the H-I groups longer than four hours (4, 6 and 8 hours groups) during the reperfusion period the ScO2 was significantly increased while the ECA was significantly decreased. This increased ScO2 and decreased ECA can be explained by permanent neurological damage which is unable to utilize O2.
The neurological outcome was normal in the control group and the H-I 1 and 2 hours. The H-I groups of more than 2 hours have an increasing incidence of neurologic injury. This incidence increased approximately 15% per hour of H-I. The incidence of neurologic injury was 100% in the H-I group of 8 hours.

Conclusion:
The authors demonstrated in this piglet H-I model a viability-time threshold for H-I injury was ScO2 of 35% for 2–3 hours, indicated by abnormalities in NIRS and CFM during reperfusion. They also suggest that NIRS and CFM might be used together to predict neurological outcome, and illustrate that there is a several hour window of opportunity during H-I to prevent neurological injury.

Comments:
The ultimate goal of animal research is to determine the optimal strategy for human disease diagnosis and treatment. Specifically the neonatal piglet model resembles the developing human brain and is validated H-I model.
The viability-time threshold for H-I injury of ScO2 of 35% for 2 hours demonstrated in this piglet model could be used therapeutic threshold to optimize brain protection during the repair of complex congenital heart diseases. Two hours with a ScO2 of 35% seems to be a long period of time to be considered safe even though in the animal model no neurologic sequela was found on the neurobehavioral examination or the pathologic study. The neurobehavioral examination in this study was only gross looking only for mayor neurological damage such as con level of consciousness, sensation and/or tone.
Learning disabilities and neurocognitive dysfunction which are common after cardiac surgery are subtle and difficult to detect in an animal model. Outcome studies following patients long term after different severity of H-I evidence by low NIRS or ECA values could bring light to the safety of procedures that are performed daily like deep hypothermia circulatory arrest or low flow perfusion states.
Factors like brain temperature, anesthesia, preexisting disease and neuroprotective strategies like anterograde cerebral perfusion can affect the viability-time threshold and the risk of structural brain damage. Nevertheless, the concept of a viability-time threshold along with technologies to measure cerebral blood flow forms a strategy in clinical care to prevent neurologic injury during the repair of complex congenital heart diseases.

References:
1. Newman M, Kirchner JL, Phillips-Bute BS et al: Longitudinal assessment of neurocognitive function after coronary-artery bypass surgery. N Engl J Med. 2001 Feb 8;344 (6):395-402
2. Toet MC, Lemmers PMA, van Schelven LJ et al: Cerebral oxygenation and electrical activity after birth asphyxia: their relation to Outcome Pediatrics 2006;117;333-339
Hoffman GM: Neurologic monitoring on cardiopulmonary bypass: what are we obligated to do? Ann Thorac Surg 2006;81:S2373-80

Article Reviews

© 2009 CONGENITAL CARDIAC ANESTHESIA SOCIETY