| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2004;77:1698-1706
© 2004 The Society of Thoracic Surgeons
a Department of Neurology, San Francisco, CA, USA
b Department of Pediatrics, San Francisco, CA, USA
c Department of Radiology, San Francisco, CA, USA
d Department of Epidemiology, San Francisco, CA, USA
e Department of Cardiothoracic Surgery, University of California, San Francisco, California, USA
Accepted for publication October 28, 2003.
* Address reprint requests to Dr Miller, 521 Parnassus Ave, Room C-215, San Francisco, CA 94143-0663, USA
e-mail: smille{at}itsa.ucsf.edu
Presented at the Fortieth Annual Meeting of The Society of Thoracic Surgeons, San Antonio, TX, Jan 26–28, 2004.
 |
Abstract |
|---|
 TopFootnotes Abstract IntroductionMaterial and methodsResultsCommentAcknowledgmentsDiscussionReferences |
|---|
METHODS: Ten term newborns with TGA undergoing an arterial switch operation were studied with MRI, MRSI, and neurologic examination preoperatively and postoperatively at a median of 5 days (2 to 9 days) and 19 days (14 to 26 days) of age, respectively. Five term historical controls were studied at a median of 4 days (3 to 9 days). Lactate/choline (marker of cerebral oxidative metabolism) and N-acetylaspartate (NAA)/choline (marker of cerebral metabolism and density) were measured bilaterally from the basal ganglia, thalamus, and corticospinal tracts.
RESULTS: Four TGA newborns had brain injury on the preoperative MRI. The only new lesion detected on the postoperative study was a focal white matter lesion in one newborn with a normal preoperative MRI. The MRSI of age-adjusted lactate/choline was quantitatively higher in newborns with TGA compared with those without heart disease (p < 0.0001), even in newborns without MRI evidence of preoperative brain injury. Lactate/choline decreased after surgery but remained elevated compared with controls. In newborns with TGA, those with preoperative brain injury on MRI had lower NAA/choline globally (p = 0.04) than those with normal preoperative MRI. Five newborns had a decline in NAA/choline from the preoperative to postoperative studies.
CONCLUSIONS: Abnormal brain metabolism and injury was observed preoperatively in newborns with TGA. Brain injury is not solely related to the operative course.
|
Introduction |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsDiscussionReferences |
|---|
Interventional studies to prevent brain injury have focused on intraoperative factors as the primary mechanism of injury, with less consideration of preoperative factors [1, 3, 6, 7]. Many studies have focused on infants with transposition of the great arteries (TGA) yielding the best description of neurodevelopmental outcome in these patients. These studies clearly demonstrate that the use of circulatory arrest rather than low-flow cardiopulmonary bypass was associated with worse early neurodevelopmental outcome [1, 6]. Yet at later ages, neurodevelopmental outcome in both groups was below population norms, indicating that other factors must also affect neurodevelopmental outcome [8]. Other major risks for brain injury in newborns include postoperative complications, such as low cardiac output or cardiac arrest. Only recently has it been recognized that more than half of newborns with CHD have evidence of neurobehavioral and neurologic abnormalities before surgery, and that these abnormalities are a significant risk factor for later neurodevelopmental impairment [2, 9].
Magnetic resonance imaging (MRI) is a powerful tool for detecting subacute brain injury in the newborn [10]. Proton magnetic resonance spectroscopy (MRS) provides an in vivo quantitative measure of brain biochemistry that can be performed sequentially [11, 12]. Of the compounds measured by MRS, N-acetylaspartate (NAA) and lactate are the most useful in assessing brain injury [11]. In normal infants, NAA increases with increasing cerebral maturity, whereas decreases indicate impaired cerebral integrity or function [11, 13]. Lactate elevation is an excellent marker of cellular anaerobic metabolism, such as that secondary to hypoxic-ischemic injury. The severity of NAA and lactate metabolic abnormalities have been associated with neurodevelopmental outcome following hypoxic ischemic brain injury in newborns with and without CHD, suggesting their usefulness in detecting early brain injury [14–16]. However, the prior studies utilizing MRS to detect brain injury related to CHD were not limited to newborns [14, 16]. These studies also used a technique that is limited to a single region of interest in the brain and thus may have missed significant brain injury in other locations. To overcome this limitation we applied a newer MRS technique, lactate-edited three-dimensional magnetic resonance spectroscopic imaging (MRSI), which can detect and localize the distributions of these compounds throughout the newborn brain at 1 cm3 resolution [17].
Preoperative MRI, postoperative MRI, and three-dimensional MRSI were used to study a cohort of newborns with TGA and intact ventricular septum, a group that is relatively homogeneous with respect to cardiac and extracardiac anatomy. Repair is performed in the immediate neonatal period and results in a near normal cardiovascular anatomy and physiology, allowing a focus on perioperative events influencing neurodevelopmental outcome.
|
Material and methods |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsDiscussionReferences |
|---|
The 10 newborns had two magnetic resonance (MR) studies each, with a neurologic assessment by a pediatric neurologist within 24 hours of the MRI studies. The first MR study was performed preoperatively as soon as the baby was stable enough to be transported safely to and from the MR scanner. The second study was performed postoperatively, as soon as possible after epicardial pacemaker wires were removed (usually 7 to 14 days postoperatively). Five term newborn controls without CHD were studied as part of an ongoing study of MR predictors of outcome in neonatal encephalopathy [15]. This study utilized broad inclusion criteria to include newborns with minimal perinatal depression and normal outcomes; the five controls had normal MRI and normal neurodevelopmental outcome at 12 months of age.
Our Institutional Committee on Human Research approved the protocol. Infants were studied after voluntary, informed parental consent was obtained.
Neuromotor evaluation
A pediatric neurologist performed a neurologic exam before each MR study (thereby blinded to the imaging results), noting abnormalities of the cranial nerves, motor tone, bulk, and power.
MRI studies
All studies were performed on a 1.5-Tesla Signa EchoSpeed system (GE Medical Systems, Waukesha, WI) using an isolette specifically designed for neonatal brain imaging research studies. MRI of the brain in all newborns included: T1 weighted sagittal and axial spin echo images (4-mm thickness) using repetition time (TR) equal to 500 ms, echo time (TE) equal to 11 ms, one excitation, and 192x256 acquisition matrix; and dual-echo T2 weighted spin echo (4-mm thickness) with TR equal to 3000 ms, TE equal to 60 ms and 120 ms, and 192x256 acquisition matrix; coronal volumetric three-dimensional gradient echo images with radiofrequency spoiling images (SPGR; 1.5-mm thickness) with TR equal to 36 ms, TR equal to 9 ms, flip angle of 35 degrees, and the number of excitations equal to one.
A pediatric neuroradiologist reviewed all MRI studies blinded to the newborn's clinical course. Newborns with TGA were classified based on the presence of preoperative brain injury, defined as the presence of hemorrhagic or nonhemorrhagic gray matter injury in a vascular distribution (arterial or venous), white matter injury (manifested as T1 shortening or T2 prolongation), or intraventricular hemorrhage on the preoperative study.
Three-dimensional proton MRSI
A specialized lactate-edited point-resolved spectroscopic imaging sequence (PRESS-MRSI) was used to obtain 8x8x8 arrays of 1.0 cm3 spectral voxels from selected regions of typically 150 to 250 cm3 in the brain [17, 18]. The lactate-editing MRSI scheme provided the accurate detection of lactate as well as choline, creatine, NAA, and lipid resonances (Fig 1)
[18]. The raw MRSI data were analyzed offline using software developed at our institution for three-dimensional MRSI processing. The analysis software was used to position 1 cm3 spectral voxels in specific anatomic locations, as defined from the MRI, with a custom-designed region-of-interest tool and metabolite intensities were calculated. To assess the MR spectra in various anatomic locations in the brain, spectral voxels were retrospectively centered bilaterally in the thalamus, basal ganglia, and corticospinal tracts (Fig 1).These regions were chosen because of their previously demonstrated sensitivity to hypoxic-ischemic injury in newborns [10, 15]. The regions of interest were defined according to anatomic landmarks and chosen to avoid signal from the cerebrospinal fluid in the ventricular system. The areas under the choline, NAA, and lactate resonances were calculated for each voxel; NAA/choline and lactate/choline were calculated for each region bilaterally (Fig 1).
|
|
-stat pH management. The CPB prime included normosol, methylprednisolone, cefazolin, and albumin. Fresh whole blood or packed red blood cells and fresh frozen plasma were added to the prime to maintain hematocrit at 24% to 27%. Generally, flow was maintained at 150 mL · kg–1 · min–1 throughout the procedure with the lowest flow no less than 75 mL · kg–1 · min–1 and no low-flow (50 mL · kg–1 · min–1) or circulatory arrest. After the repair and separation from CPB, all patients underwent modified ultrafiltration to remove the crystalloid priming volume.
Data collection
Perinatal, preoperative, intraoperative, and postoperative data were collected from medical chart review and departmental databases. A composite of physiologic and laboratory data were used to calculate score of neonatal acute physiology–perinatal extension (SNAP-PE) [19], a validated measure of overall neonatal illness severity.
Data analysis
Statistical analyses were performed using Stata (Stata Corporation, College Station, TX). The clinical and surgical characteristic of the patient with and without preoperative brain injury were compared using the Mann-Whitney U test for continuous or ordinal data and the Fischer's exact test for categoric data. In order to determine the percent difference in the cerebral metabolite ratios among those without CHD and those with TGA (with and without preoperative injury), we modeled the change in log transformed NAA/choline and lactate/choline, adjusting for gestational age at the MR study in each analysis [20]. Because patients were studied twice, we used linear regression for repeated measures (generalized estimating equation) [21].
|
Results |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsDiscussionReferences |
|---|
|
|
Neurologic examination
None of the newborns had focal deficits on the neurologic examination to suggest a focal lesion of the brain. However, abnormalities of tone or reflexes were common in newborns with and without preoperative injury on MRI (Table 3).
None of the newborns had deficits of motor power, such as hemiparesis. Furthermore, none of the newborns had clinically suspected seizures.
|
|
Although NAA is expected to increase with age, half of the TGA newborns had a decline in NAA/choline from the preoperative to postoperative studies: 3 newborns without preoperative brain injury and 2 newborns with preoperative injury on MRI (Fig 4).
|
|
Comment |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsDiscussionReferences |
|---|
Preoperative brain injury in this cohort, the largest MRI study of newborns with TGA, is consistent with a recent study of 24 newborns with a variety of types of CHD, where 24% had preoperative ischemic lesions on MRI [14]. However, in that study 54% of newborns had single ventricle physiology, including hypoplastic left heart syndrome, a group known to have a high incidence of congenital brain abnormalities and poor neurodevelopmental outcome. The preoperative injury noted in our study is all the more remarkable because newborns with TGA represent a homogenous group with favorable neurodevelopmental outcome compared with infants with hypoplastic left heart syndrome.
Preoperative mechanisms of brain injury in newborns with TGA: focal and diffuse
In addition to issues of timing, multiple mechanisms have been identified that may contribute to the high frequency of abnormal neurodevelopmental outcomes observed in infants with CHD. These can be broadly characterized as the following: global hypoxia; global or focal ischemia/reperfusion (including hemorrhage); inflammatory; and developmental (including genetic). In this relatively homogenous cohort of infants with TGA, we found diffuse disturbances in cerebral oxidative metabolism in newborns with and without preoperative focal ischemic events. Three of four preoperative lesions noted on MRI were consistent with focal embolic strokes. All 4 infants with preoperative abnormalities on MRI had atrial septostomy, raising the possibility that some debris embolized during this procedure. Although we could not establish a relationship of these lesions with the atrial septostomy given our sample size, this possibility requires further investigation.
The significantly higher lactate/choline, a marker of impaired cerebral oxidative metabolism, in newborns with TGA compared with those without heart disease is consistent with the qualitative elevation of lactate seen in 53% of newborns with a wide variety of congenital heart lesions [14]. The elevation in lactate/choline occurred in newborns with TGA that were hemodynamically stable preoperatively without evidence of other organ hypoperfusion and was not regionally specific, being evident in the basal ganglia, thalamus, and corticospinal tract regions. The similar elevation in lactate levels in those newborns with and without MRI detectable preoperative brain injury suggests that the impairment in lactate/choline may relate to impaired global cerebral oxygen delivery due to cyanosis or oxygen utilization. Alternatively, the elevation in lactate/choline may be secondary to relative immaturity of cerebral development as lactate/choline is seen in some premature newborns without MRI detectable brain injury [13]. A relative cerebral immaturity is also suggested by the presence of intraventricular hemorrhage and periventricular white matter injury in term newborns with TGA because these lesions are more common in premature newborns.
Cardiopulmonary bypass related brain injury
The elevation in lactate/choline observed in newborns with TGA improved from the preoperative to postoperative studies. This suggests that correction of the defect improves cerebral oxidative metabolism. Additionally, only one new lesion was identified postoperatively, an ischemic focus in the paratrigonal white matter. This is significantly different than the high frequency of new lesions identified in the early postoperative period by Mahle and colleagues [14] in a cohort of patients including single ventricle physiology. Another explanation for the low incidence of new lesions may relate to institutional differences in the CPB strategy used. This cohort of patients was repaired with full-flow CPB (150 mL · kg–1 · min–1) compared with a high incidence (88%) of deep hypothermic circulatory arrest of significant duration (median 50 min) in the cohort of Mahle and coworkers [14]. Low-flow CPB compared with deep hypothermic circulatory arrest is associated with better perioperative neurodevelopmental outcome, and full-flow CPB may compare even more favorably comparing full scale intelligence quotient testing [22].
In half of the newborns the NAA/choline decreased, indicating a detrimental effect of the intraoperative or postoperative events on cerebral cellular metabolism in these patients. In this small cohort of newborns with TGA, perioperative indices of cardiac function and CPB were not significantly associated with three-dimensional MRSI metabolite ratios. However, the decrease in NAA/choline in individual newborns from the preoperative to postoperative studies suggests that some newborns indeed develop impaired cerebral integrity and function over this time period. The lower NAA/choline in newborns with TGA and preoperative brain injury suggests that the role of preoperative injury in predisposing to impaired cerebral energy metabolism during bypass needs to be explored further.
Role of neuroimaging in newborns with CHD
Reduction in NAA ratios and the presence of lactate have previously been associated with neurodevelopmental outcome in infants and children with CHD [16]. Determining the relationship between preoperative brain injury, reduced NAA, and elevated lactate with neurodevelopmental outcome will improve our ability to identify newborns at high risk of neurodevelopmental impairment, and will suggest when to intervene to prevent brain injury in this high-risk population. A consideration of the preoperative cerebral condition and the exact timing and nature of acquired injury will be required to optimally design and test new strategies of neuroprotection and will also impact considerations of the timing of cardiac surgery.
|
Acknowledgments |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsDiscussionReferences |
|---|
|
Footnotes |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsDiscussionReferences |
|---|
|
Discussion |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsDiscussionReferences |
|---|
The authors' finding that four out of five macroscopic brain lesions detected by imaging were present preoperatively is important new information. In a prospective randomized trial of 171 neonates who also underwent repair of transposition, we found that 23% of our patients had possible or definite abnormalities on magnetic resonance imaging (MRI) scan at 1 year of age, but we had no knowledge whether these lesions were present preoperatively because we only performed scans postoperatively. Like the authors, we found no relationship between intraoperative factors and the presence of MRI lesions, though we did find a relationship between preoperative acidosis and presence of an MRI lesion.
Like the authors, we also found that many of the MRI lesions appeared to be clinically silent and were not detected by neurological examination. In fact, overall we found that the neurological examination was an insensitive endpoint in contrast to developmental testing, which suggested a high incidence of subtle developmental delay related to intraoperative factors. Developmental testing as well as our laboratory studies have helped us to understand that the technique of bypass that we used in the late 1980s with an alkaline pH and severe hemodilution almost certainly caused a global hypoxic ischemic injury to many patients. The authors' own previous reports of developmental testing of their patients has also demonstrated developmental abnormalities in many of their patients. The technique of bypass used by the authors in this study, as in their previous reports, employed an alkaline pH strategy, moderate hemodilution, but higher flows and temperatures than we have used.
My question relates to the authors' finding that half of their patients had a decline in their N-acetylaspartate (NAA)/choline ratio, suggesting a global injury during repair. Why do you think half of your patients showed a decline in the NAA/choline ratio? Is it possible that this was caused by your technique of high flow cardiopulmonary bypass, which might have increased the inflammatory response to bypass, particularly when conducted at the relatively warm temperature of 28°? Is this a more likely explanation of your previously published developmental studies showing speech and motor abnormalities rather than local embolic or hemorrhagic lesions, which would surely cause focal deficits?
Secondly, do you have plans to expand this study and to include long-term outcome data in which you will attempt to define the significance of the deterioration in the NAA/choline ratio?
Once again, congratulations on a very difficult study, an excellent presentation, and thank you for the opportunity to comment.
DR ANDREW L. CARNEY (Chicago, IL): I would like to comment on two things: one, the use of "stroke" and then "brain injury." For many surgeons, stroke means a middle cerebral artery injury with hemiparesis or hemiplegia. Brain injury is more common and does not involve that territory. The second thing is that Barkovich has written a great deal about the value of diffusion scans to identify brain injury in newborns and pointing out the limitation of physical examination. That same limitation is also true in adults who have brain injury. To say an adult having cardiopulmonary bypass has no stroke doesn't mean he does not have brain injury. But I think you beautifully point out the value of diffusion scans in identifying acute brain injury following and preceding surgery. In my opinion, it would be highly unlikely that bilateral or symmetrical lesions are due to embolism. Thank you.
DR MILLER: First, to begin with Dr Jonas' question, we have expanded our cohort to newborns beyond transposition of the great arteries as we are trying to get at the issue of are different types of cardiopulmonary bypass and is cardiopulmonary bypass the critical issue? So we will be looking at other types of congenital heart disease. And very importantly, as I failed to point out in my talk, we are following this cohort longitudinally through childhood now to determine what are the best MR predictors of developmental outcome in this population.
Our experience to date in other term newborns with hypoxic ischemic brain injury is that NAA/choline ratios in the basal ganglia, the area most selectively vulnerable to injury in the term newborn, have been most predictive of neurodevelopmental outcome, but the burden is clearly on us now to show that this is the case in newborns with congenital heart disease. We are also hoping that this expansion of the cohort will allow us to try and address specifically the cardiopulmonary bypass-related measures that you have raised, inflammatory mediators, for example, and their role in the genesis of brain injury.
And to address the second discussant's comments, we have been finding that stroke and brain injury as a whole are very intertwined. Stroke is a form of brain injury, as you know, in the newborn but in this cohort seemed to have been the tip of the iceberg as we did see lower overall levels of NAA/choline ratios in the newborns with stroke. A hypothesis that we are pursuing actively is whether those preoperative strokes are the critical risk factors for them being at risk for intraoperative or cardiopulmonary bypass-related brain injury. Thank you.
|
References |
|---|
|
TopFootnotesAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. P. Miller, P. S. McQuillen, S. Hamrick, D. Xu, D. V. Glidden, N. Charlton, T. Karl, A. Azakie, D. M. Ferriero, A. J. Barkovich, et al. Abnormal Brain Development in Newborns with Congenital Heart Disease N. Engl. J. Med., November 8, 2007; 357(19): 1928 - 1938. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. P Miller and P. S McQuillen Neurology of congenital heart disease: insight from brain imaging Arch. Dis. Child. Fetal Neonatal Ed., November 1, 2007; 92(6): F435 - F437. [Full Text] [PDF]
|
|
|
|
 |
|
 M. Arrica and B. Bissonnette Therapeutic hypothermia. Seminars in Cardiothoracic and Vascular Anesthesia, March 1, 2007; 11(1): 6 - 15. [Abstract] [PDF]
|
|
|
|
|
|
S. D. Markowitz, R. N. Ichord, G. Wernovsky, J. W. Gaynor, and S. C. Nicolson Surrogate markers for neurological outcome in children after deep hypothermic circulatory arrest. Seminars in Cardiothoracic and Vascular Anesthesia, March 1, 2007; 11(1): 59 - 65. [Abstract] [PDF]
|
|
|
|
 |
|
 P. S. McQuillen, A. J. Barkovich, S. E.G. Hamrick, M. Perez, P. Ward, D. V. Glidden, A. Azakie, T. Karl, and S. P. Miller Temporal and Anatomic Risk Profile of Brain Injury With Neonatal Repair of Congenital Heart Defects Stroke, February 1, 2007; 38(2): 736 - 741. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. H. Freed, C. M.T. Robertson, R. S. Sauve, A. R. Joffe, I. M. Rebeyka, D. B. Ross, J. D. Dyck, and the Western Canadian Complex Pediatric Therapies P Intermediate-term outcomes of the arterial switch operation for transposition of great arteries in neonates: Alive but well? J. Thorac. Cardiovasc. Surg., October 1, 2006; 132(4): 845 - 852. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D.H. Kim, A.J. Barkovich, and D.B. Vigneron Short Echo Time MR Spectroscopic Imaging for Neonatal Pediatric Imaging AJNR Am. J. Neuroradiol., June 1, 2006; 27(6): 1370 - 1372. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. S. McQuillen, S. E.G. Hamrick, M. J. Perez, A. J. Barkovich, D. V. Glidden, T. R. Karl, D. Teitel, and S. P. Miller Balloon Atrial Septostomy Is Associated With Preoperative Stroke in Neonates With Transposition of the Great Arteries Circulation, January 17, 2006; 113(2): 280 - 285. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Azakie, J. Muse, M. Gardner, K. L. Skidmore, S. P. Miller, T. R. Karl, and P. S. McQuillen Cerebral oxygen balance is impaired during repair of aortic coarctation in infants and children J. Thorac. Cardiovasc. Surg., September 1, 2005; 130(3): 830 - 836. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. M. Ferriero Neonatal Brain Injury N. Engl. J. Med., November 4, 2004; 351(19): 1985 - 1995. [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 |
