|
|
||||||||
J Thorac Cardiovasc Surg 2006;131:659-665
© 2006 The American Association for Thoracic Surgery
Cardiopulmonary Support and Physiology |
a Department of Neonatology, Stanford University School of Medicine, Stanford, Calif
b Department of Cardiothoracic Surgery, Stanford University School of Medicine, Stanford, Calif
c Department of Comparative Medicine, Stanford University School of Medicine, Stanford, Calif
d Department of Anesthesia, Stanford University School of Medicine, Stanford, Calif
Received for publication May 23, 2005; revisions received September 2, 2005; accepted for publication September 13, 2005. * Address for reprints: Valerie Chock, MD, Division of Neonatology, Stanford University Medical Center, 750 Welch Rd, Suite 315, Stanford, CA 94305. (Email: vchock{at}stanford.edu).
| Abstract |
|---|
|
|
|---|
METHODS: Neonatal piglets undergoing cardiopulmonary bypass were randomized to deep hypothermic circulatory arrest or antegrade cerebral perfusion for 45 minutes. Animals were killed after 6 hours of recovery, and brain tissue was stained for evidence of cellular injury and for the apoptotic markers activated caspase 3 and cytochrome c translocation from mitochondria to cytosol.
RESULTS: Piglets from the antegrade cerebral perfusion group exhibited less apoptotic or necrotic injury (4 ± 3 vs 29 ± 12 cells per field, P = .03). The piglets undergoing antegrade cerebral perfusion also had less evidence of apoptosis, with fewer cells staining for activated caspase 3 (57 ± 8 vs 93 ± 9 cells per field, P = .001) or showing cytochrome c translocation (6 ± 2 vs 15 ± 4 cells per field, P = .02).
CONCLUSIONS: The use of antegrade cerebral perfusion in place of deep hypothermic circulatory arrest reduces evidence of apoptosis and histologic injury in neonatal piglets. Neonates with congenital heart disease might benefit from antegrade cerebral perfusion during complex cardiac surgery to improve their overall neurologic outcome.
| Introduction |
|---|
|
|
|---|
We used a piglet model of CPB to compare the degree of neurologic injury after ACP or DHCA. Injury was assessed histologically, and specific markers of apoptotic injury were compared. A piglet model was chosen because of its similarity in weight and brain development at birth to the human neonate.
8
In addition, this piglet model of CPB with DHCA has already been shown to induce apoptotic brain injury
9,10
by morphology, terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) staining, and caspase 3 and caspase 8 activation in neurons of the neocortex and hippocampus.
We studied specific markers of apoptosis because apoptosis is known to play a prominent role in the evolution of hypoxic-ischemic injury in the neonatal brain.
11
Caspase 3 was assessed because it is a major effector protease in the execution of cellular apoptosis. We also studied cytochrome c translocation because it is released from mitochondria and participates in forming the apoptosome, which activates caspase 3 through the intrinsic pathway. We hypothesized that the piglets undergoing ACP would have reduced neurologic injury on brain histology, specifically with less caspase 3 activation and cytochrome c translocation. These initial studies assessed the induction of apoptosis at an early survival time point.
| Methods |
|---|
|
|
|---|
Cardiopulmonary Bypass and Perfusion
The CPB circuit consisted of a roller pump, membrane oxygenator (Minimax; Medtronic Bio-Medicus, Minneapolis, Minn), and sterile
-inch tubing. The circuit was primed with blood previously harvested from a donor pig mixed with crystalloid prime solution (Normasol; Abbot Labs, Chicago, Ill) to maintain the hematocrit value at greater than 30%. In addition, dexamethasone (30 mg/kg), heparin (2500 U), mannitol (0.5 g/kg), and sodium bicarbonate (20 mEq) were added to the priming solution. CPB was then initiated, and the animal was cooled to 18°C over 40 minutes at a pump flow of 200 mL · kg1
· min1 by using pH-stat arterial blood gas management. The piglets were randomized to DHCA (n = 7) or ACP (n = 7) for 45 minutes. For the arrest period, the aorta was clamped, and cold cardioplegic solution was administered to the aortic root (Plegisol, Abbot labs). For those animals in the ACP group, the proximal innominate artery was clamped, and ACP flow rate was maintained at 40 mL · kg1
· min1 with continued pH-stat management.
This ACP flow rate correlates with cerebral blood flow of 24 ± 5 mL · kg1
· min1, as measured with a carotid flow probe.
13
In this study we found that an ACP flow rate of 40 mL · kg1
· min1 corresponded with cerebral flow rates of 19 to 24 mL · kg1
· min1 and cerebral capillary oxygen saturations of 50% to 70%. Lower flow rates resulted in cerebral capillary oxygen saturations of less than 40%, reflecting unacceptable tissue ischemia. Near-infrared spectroscopy measurements by others demonstrate the need for regional perfusion of 20 mL · kg1
· min1 to maintain baseline cerebral saturations in neonates undergoing regional low-flow perfusion.
6
The piglets were then rewarmed to 37°C by using a combined pH-stat and alpha-stat strategy for arterial blood gas management. They were weaned from bypass, and hemoconcentration was performed to achieve a hematocrit value of 45%. The piglets were extubated and allowed to recover for 6 hours before death.
Histology
Piglets were killed with sodium pentobarbital and phenytoin (Euthanol; MTC Pharmaceuticals, Cambridge, Ontario, Canada) and perfused with 1 L of chilled saline and 1 L of 4% paraformaldehyde. Brains were removed and immersed overnight in 4% paraformaldehyde for fixation. Five-millimeter coronal blocks were cut and routinely processed into paraffin. Sections were cut from each block at a thickness of 5 µm and either stained with hematoxylin and eosin (H&E) or reserved for immunohistochemistry.
Immunohistochemistry
Paraffin sections were rehydrated, treated with 0.5% H202 for 30 minutes to inactivate endogenous peroxidases, and blocked with 2% horse serum for 30 minutes. Primary antibody to caspase 3 (rabbit monoclonal, 1:2000; PharMingen, San Diego, Calif) or cytochrome c (mouse monoclonal, 1:200; Santa Cruz Biotechnology, Santa Cruz, Calif) was added and incubated for 2 hours. After incubation with the species-appropriate secondary IgG antibody (Vector Laboratories, 1:200) for 45 minutes, the sections were washed and incubated with an avidin-biotin complex (ABC 1:100, Vector Laboratories), and immunoreactivity was visualized with 3,3' diaminobenzedine tetrahydrochloride (Sigma, St Louis, Mo). For negative controls, the primary antibody was omitted.
Counting Protocol
Ten randomly generated 400x fields of H&E-stained caudate nucleus, putamen, hippocampus, cortex, and cerebellum were examined for each animal by using bright-field microscopy. All slides were blinded before examination by a pathologist. Cell injury was defined as necrotic or apoptotic by morphology, and counts were averaged for the ten 400x fields. Necrotic cells were identified on the basis of pyknotic or absent nuclei and shrunken hypereosinophilic cytoplasm and angular cell margins. Apoptotic cells had fragmented or condensed nuclei with shrunken cytoplasm and rounding of cytoplasmic margins. Positively stained cells for activated caspase 3 and cytochrome c translocation were counted in a similar manner in caudate nucleus, putamen, neocortex, hippocampus, and cerebellum by a blinded observer. Cytochrome c release from mitochondria was counted when staining was distributed diffusely through the cytoplasm.
Statistical Analysis
Data are presented as means ± the standard error of the mean. One-way analysis of variance with the post-hoc Newman-Keuls test was used to compare results between groups. Statistical analyses demonstrated no significant differences in the histology of brains of the control piglets that underwent cannulation with anesthesia and those of control piglets that did not. Therefore these animals (n = 5) were pooled into a single control group for histology but kept as 2 distinct groups for immunohistochemistry.
| Results |
|---|
|
|
|---|
|
Figures 2, A through C, and 3, A through C, show differences in the intensity of immunostaining in the striatum for activated caspase 3 and cytochrome c in the DHCA, ACP, and control groups. Both neurons and glial cells showed positive staining for activated caspase 3, whereas cytochrome cpositive cells were neurons. Control piglets showed minimal positive staining cells for either antibody. Immunostaining was also done in the hippocampus, cortex, and cerebellum, but because of the milder injury in these brain regions, differences did not reach statistical significance (data not shown).
|
|
| Discussion |
|---|
|
|
|---|
Other porcine models of CPB have also found less severe injury in a regional low-flow perfusion group.
5,14
In an adult pig model of hypothermic ACP, Hagl and colleagues
5
found earlier electroencephalographic recovery and lower intracranial pressure increase compared with that seen in their pigs undergoing 90 minutes of DHCA. During a neonatal piglet study with 90 minutes of DHCA, a cerebral flow rate of 10 mL · kg1
· min1, and a 7-day recovery period, Myung and associates
14
found a trend toward less severe injury in their ACP group by means of H&E staining and decreased apoptosis by means of TUNEL staining. We chose a more clinically relevant time for DHCA of 45 minutes, an ACP flow rate of 40 mL · kg1
· min1, and an earlier 6-hour time point to analyze histologic injury. We also used more specific measures of apoptosis: activated caspase 3 analysis and cytochrome c analysis.
Length of circulatory arrest, duration of recovery, and degree of cerebral perfusion are key variables differentiating our model from others. In an adult pig model of DHCA, it takes at least 75 to 90 minutes of hypothermic circulatory arrest to consistently produce cerebral damage by histopathology.
5
Perhaps longer arrest times would have resulted in increased injury in additional brain regions and even more pronounced protection in our ACP compared with DHCA groups. Similarly, after flow is restored after bypass, brain oxygen levels still remain decreased, even at 2 hours of recovery.
15,16
As injury continues to evolve beyond a 6-hour recovery, delayed effects, such as inflammation, will require analysis with longer survival intervals. Inflammatory mediators contributing to brain injury might not peak until 72 hours to 1 week after injury.
17
In this global model of brain ischemia, we demonstrated a selective vulnerability by brain region, with striatum being the most sensitive to injury. Other investigators using this piglet model saw more cortical and hippocampal injury and less striatal injury
9,10
; however, they used a 90-minute DHCA time. Different mechanisms of neuronal injury might occur in different selectively vulnerable brain regions at various times. Striatum is involved in control of motor behavior and is especially vulnerable to hypoxic-ischemic injury in the neonate.
18
Indeed, after severe perinatal asphyxia in term infants, characteristic injury occurs in the basal ganglia region, as determined by means of magnetic resonance imaging.
19
Consistent with our findings, significant striatal damage occurs after cardiac arrest in newborn piglets.
20
Several mechanisms might account for striatal vulnerability in our model. Increased levels of striatal dopamine, a mediator of neuronal injury, and increased striatal orthotyrosine, a measure of free radical generation, were reported in a piglet model of DHCA in comparison with a low-flow bypass condition.
21
Others have further shown the dopaminergic system of piglet striatum to be very sensitive to local oxygen pressure, with extracellular dopamine increasing as tissue oxygenation decreases.
22,23
Increased striatal dopamine after hypoxia and ischemia in piglets was also associated with decreased cAMP response element binding protein (CREB) phosphorylation
24
and suppression of Na+, K+-adenosine triphosphatase activity,
25
mechanisms potentially contributing to neuronal death in this region. After hypoxia-ischemia, altered N-methyl-D-aspartate (NMDA) receptors and decreased intracellular calcium clearance are also found in piglet striatum, implicating excitotoxic injury as an additional factor in striatal vulnerability.
26,27
In addition to excessive dopaminergic excitatory activity, after a hypoxic insult, neonatal piglet striatal neurons express almost no heat shock protein 72, suggesting that protective mechanisms in this brain region at this developmental stage are limited.
26,28
Although apoptosis can be identified on the basis of typical changes in cellular morphology and DNA fragmentation, these characteristics are not entirely specific for apoptosis. In the setting of ischemia, mixed and intermediate morphologies are often observed. Therefore we investigated several key apoptotic proteins. Cytochrome c is translocated from dysfunctional mitochondria into cell cytoplasm in response to ischemia activating the intrinsic pathway of apoptosis.
29
It forms a complex with Apaf-1, which activates caspase 9 to activate caspase 3. Caspase 3 can then initiate cleavage of DNA. We found significantly increased cells with activated caspase 3 and translocated cytochrome c in the DHCA group of piglets compared with that in the ACP group at 6 hours of reperfusion, suggesting that the protective effects of ACP might be due to a decrease in apoptotic activity. In addition, there was no difference in apoptotic markers between the ACP and control groups, despite increased histologic injury in the ACP group. This discrepancy might reflect differences in the timing of cellular morphologic changes. Ditsworth and colleagues
9
found a similar increase in caspase 3 activity from 4 to 72 hours after DHCA in piglets, whereas cytosolic cytochrome c levels, as determined by means of Western blotting, peaked at 1 hour after DHCA and remained mildly increased for up to 72 hours. Their study further found an increase in caspase 8 levels at 24 hours of reperfusion after DHCA.
9
This finding might reflect an additional contribution to apoptosis through the extrinsic pathway and might be related to later inflammatory effects of DHCA, which we were unable to assess with our early 6-hour reperfusion time point.
Both apoptosis and necrosis occur in striatum after transient global ischemia, although it is impossible to distinguish between these 2 histologic injuries on the basis of clinical outcome alone.
11
Neurologic injury is typically a spectrum between apoptosis and necrosis. However, apoptosis tends to occur in regions with less severe ischemia,
30
whereas necrotic injury is the form of neuronal death in piglet striatum after severe asphyxic cardiac arrest.
31
Although not quantified, the observation that piglets undergoing DHCA had more necrotic cells and piglets undergoing ACP had more apoptotic cells further supports the attenuation of brain injury caused by ACP.
As an animal model, our study is subject to several experimental constraints. As previously discussed, the benefit of using ACP in our model is restricted by a shorter circulatory arrest time and early survival time point. Although clinically relevant, our relatively short circulatory arrest time of 45 minutes might preclude assessment of the protective effects of ACP when typical time limits have been exceeded during surgical repair. Histologic evaluation at 6 hours of survival prevents us from analyzing the evolution of apoptosis in multiple brain regions. Furthermore, it is unclear to what extent increased apoptotic brain injury correlates with clinical deficits of motor integration and neurologic function in our model. In a similar piglet model of DHCA, despite ongoing histologic cell death, neurologic function improved over a 7-day survival period.
10
In addition, species differences might limit application of our findings to human neonates with congenital heart disease, although piglet size and brain maturation is similar to that of the term human newborn.
8
Our study supports the use of ACP during hypothermic CPB in this piglet model by a reduction in histologic injury and specific apoptotic markers. Moreover, striatum was the most vulnerable brain region to injury, a finding consistent with known patterns of hypoxic-ischemic injury in the neonate. Application of our model with longer survival periods and investigation of other markers of injury at later time points might yield additional insight into the protective role of ACP and the mechanism of cerebral injury during CPB.
| Acknowledgments |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. Allibhai, R. DiGeronimo, J. Whitin, J. Salazar, T. T.-S. Yu, X. B. Ling, H. Cohen, P. Dixon, and A. Madan Effects of moderate versus deep hypothermic circulatory arrest and selective cerebral perfusion on cerebrospinal fluid proteomic profiles in a piglet model of cardiopulmonary bypass J. Thorac. Cardiovasc. Surg., December 1, 2009; 138(6): 1290 - 1296. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Salazar, R. Coleman, S. Griffith, J. McNeil, H. Young, J. Calhoon, F. Serrano, and R. DiGeronimo Brain preservation with selective cerebral perfusion for operations requiring circulatory arrest: protection at 25 {degrees}C is similar to 18 {degrees}C with shorter operating times Eur. J. Cardiothorac. Surg., September 1, 2009; 36(3): 524 - 531. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. D. Salazar, R. D. Coleman, S. Griffith, J. D. McNeil, M. Steigelman, H. Young, B. Hensler, P. Dixon, J. Calhoon, F. Serrano, et al. Selective cerebral perfusion: real-time evidence of brain oxygen and energy metabolism preservation. Ann. Thorac. Surg., July 1, 2009; 88(1): 162 - 169. [Abstract] [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 |