| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Division of Cardiac Surgery, The Johns Hopkins Medical Institutions, Baltimore, Maryland
b Department of Neurology, The Johns Hopkins Medical Institutions, Baltimore, Maryland
c Division of Neuropathology, The Johns Hopkins Medical Institutions, Baltimore, Maryland
d Kennedy-Krieger Institute, Baltimore, Maryland
e Departments of Neuroscience and Psychiatry, Center for Traumatic Brain Injury Studies, University of Florida, Gainesville, and Banyan Biomarkers Inc, Alachua, Florida
Accepted for publication April 3, 2009.
* Address correspondence to Dr Baumgartner, Division of Cardiac Surgery, Department of Surgery, The Johns Hopkins University School of Medicine, Blalock 618, The Johns Hopkins Hospital, 600 N Wolfe St, Baltimore, MD 21287 (Email: wbaumgar{at}jhmi.edu).
Presented at the Basic Science Forum of the Fifty-fifth Annual Meeting of the Southern Thoracic Surgical Association, Austin, TX, Nov 5–8, 2008.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
II-SBDPs) in a canine model of hypothermic circulatory arrest (HCA) and cardiopulmonary bypass.
Methods: Canine subjects were exposed to either 1 hour of HCA (n = 8; mean lowest tympanic temperature 18.0 ± 1.2°C) or standard cardiopulmonary bypass (n = 7). Cerebrospinal fluid samples were collected before treatment and 8 and 24 hours after treatment. Using polyacrylamide gel electrophoresis and immunoblotting, SBDPs were isolated and compared between groups using computer-assisted densitometric scanning. Necrotic versus apoptotic cell death was indexed by measuring calpain and caspase-3 cleaved II-SBDPs (SBDP 145+150 and SBDP 120, respectively).
Results: Animals undergoing HCA demonstrated mild patterns of histologic cellular injury and clinically detectable neurologic dysfunction. Calpain-produced II-SBDPs (150 kDa+145 kDa bands–necrosis) 8 hours after HCA were significantly increased (p = 0.02) as compared with levels before HCA, and remained elevated at 24 hours after HCA. In contrast, caspase-3 II-SBDP (120 kDa band–apoptosis) was not significantly increased. Animals receiving cardiopulmonary bypass did not demonstrate clinical or histologic evidence of injury, with no increases in necrotic or apoptotic cellular markers.
Conclusions: We report the use of II-SBDPs as markers of neurologic injury after cardiac surgery. Our analysis demonstrates that calpain- and caspase-produced II-SBDPs may be an important and novel marker of neurologic injury after HCA.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
A technique of importance for understanding brain injury after cardiac surgery is hypothermic circulatory arrest (HCA). Hypothermic circulatory arrest remains a widely used form of neuroprotection for complex cardiac cases involving aortic arch abnormalities and congenital malformations. Despite its utility, patients who receive HCA are at particularly high risk for neurologic dysfunction such as stroke, seizures, developmental delays, and neurocognitive decline [3, 4].
In clinical practice, a reliable method of detecting brain injury after cardiac surgery does not exist. Brain injury is, unfortunately, often diagnosed several hours or even days after surgery when it becomes clear that a patient is not progressing as expected. Given the substantial mortality, morbidity, and cost resulting from brain injury, developing readily obtainable biomarkers for both diagnosing the injury early and predicting the magnitude of injury is of paramount importance for patients, physicians, and families.
To this end, several potential serum biomarkers have been devised for acute brain injury with modest success. Specifically, S-100B [5], neuron-specific enolase [6], glial fibrillary-associated protein, and myelin basic protein [7] are among those widely studied. No biomarker for neurologic injury has yet proven ideal. For example, S100B and neuron-specific enolase are nonspecific for brain injury and do not predict outcomes with high sensitivity [5]. Furthermore, no biomarker exists specific for brain injury resulting from cardiopulmonary bypass (CPB) or HCA. The identification of a reliable specific serum biomarker for neurologic injury after cardiac surgery would be very beneficial.
Recent work has focused on the use of disease-specific proteolysis of II-spectrin as a biochemical marker of brain injury [8, 9]. Intact II-spectrin is a major structural component of the cortical membrane cytoskeleton abundant in axons and presynaptic terminals [10, 11]. Importantly, II-spectrin is a major substrate for cysteine proteases involved in necrotic (calpain) and apoptotic (caspase-3) cell death [12]. There is considerable evidence that II-spectrin is processed to cleavage products also known as spectrin breakdown products (SBDP), of molecular weight 150 kDa (SBDP150) and 145 kDa (SBDP145) by calpain. In addition, it is cleaved to a major cleavage product of molecular weight 120 kDa (SBDP120) by caspase-3 (Fig 1). There is evidence for the detectable presence of SBDPs in in vitro neuronal cell culture models of injury [13], in vivo studies of mice [14], and in humans studies of traumatic brain injury [15, 16].
|
II-SBDPs as biomarkers for neuronal injury, we sought to apply this concept to neurologic injury resulting from cardiac surgery. We therefore tested for the presence of II-SBDPs in an established canine model of HCA and CPB alone. We hypothesize that for animals subjected to HCA for a clinically relevant duration, II-SBDPs will be present and identifiable in the cerebral spinal fluid (CSF) and serve as indicators of neuronal injury after HCA. We further hypothesize that II-SBDPs will be present, but to a lesser extent, in animals subjected to CPB alone. Identification of these proteins is the CSF is a first step toward the identification of a biomarker. It is our hope that in future experiments we will identify II-SBDPs in serum and that they may therefore serve as a clinically applicable biomarkers for diagnosis of post–cardiac surgery neurologic injury, even when that injury is subtle on clinical examination. |
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Experimental Design
Canine subjects were exposed to either 1 hour of HCA (1H HCA [n = 8]), or standard CPB (n = 7) and survived to either 8 hours (n = 4 for 1H HCA and n = 3 for CPB) or 24 hours (n = 4 for 1H HCA and n = 4 for CPB) after treatment. Cerebrospinal fluid samples were collected immediately before treatment and 8 hours and 24 hours (when available) after treatment. For animals that survived to 24 hours after treatment, clinical neurologic scoring (performed independently by two study team members) using the Pittsburgh Veterinary Scoring System (higher scores indicate worse neurologic function) [21] was performed before giving any sedation or narcotic pain medication. The score utilized includes 22 specific clinical questions relating to level of consciousness, respiration, cranial nerve function, reflexes, behavior, and motor and sensory function. All animals utilized had no levels of neurologic impairment before experimentation, and no additional sedation is given within 12 hours of neurologic assessment.
At the conclusion of the experiment, all canine subjects were sacrificed by exsanguination, during which selective perfusion of the head with cold saline (4°C) was performed and brains were harvested for histologic analysis. The II-SBDPs, which were isolated from CSF samples, were compared between groups using computer-assisted densitometric scanning. Necrotic versus apoptotic cell death was indexed by measuring calpain and caspase-3 cleaved II-SBDPs (SBDP 145+150 and SBDP 120, respectively).
Historical Controls
For purposes of comparison of neurologic outcomes, two sets of historical controls were used (data not published elsewhere). The first set (n = 5) was subjected to 2 hours of HCA and sacrificed at least 24 hours after the treatment (2H HCA-24s). A second set of controls (n = 8) was subjected to 2 hours of HCA and sacrificed at 8 hours after HCA (2H HCA-8s). The historical control experiments were performed between February 20, 2006, and October 23, 2006. With the exception of HCA duration and lack of CSF samples, no differences in experimental protocol, setup, or equipment existed for these animals relative to the study subjects. These additional groups of animals are included for comparison of neurologic outcomes with our experimental groups of 1H HCA and CPB alone.
Detailed Procedures
Surgical HCA procedure
Our experimental model has been described previously [17, 19, 20]. In brief, canine subjects are induced with sodium thiopental (25 mg/kg intravenously) endotracheally intubated and maintained on isoflurane inhalational anesthesia (0.5% to 2%), 100% inspired oxygen, and intravenous fentanyl at 150 to 200 µg/dose during invasive procedures. Tympanic membrane probes (which correlate closely with cerebral temperature), nasopharyngeal probes, and rectal probes are placed to monitor temperatures throughout the experiment. A left femoral artery cannula is placed through an open surgical cut-down technique before the initiation of CPB, for monitoring of blood pressures and sampling of arterial blood gases.
The CPB circuit consists of a Cobe membrane oxygenator with 40-µm arterial filter (Cobe Laboratories, Lakewood, CO), and Sarns roller pump system (Sarns, Ann Arbor, MI). The circuit is primed with lactated Ringer's solution with potassium chloride (20 mEq). After heparinization (300 U/kg intravenously), the right femoral artery is cannulated (12F to 14F), and the cannula is advanced into the descending thoracic aorta. Venous cannulae (18F to 20F) are advanced to the right atrium from the right femoral and right external jugular veins. All arteries and veins are accessed through an open cutdown technique. Closed-chest CPB is initiated, and the animals are cooled until tympanic membrane temperature reaches 18°C (approximately 30 minutes). Pump flows of 60 to 100 mL · kg–1 · min–1 are required to maintain a mean arterial pressure of 50 to 60 mm Hg. Once tympanic temperatures reach 18°C, the pump is stopped and blood is drained by gravity into the reservoir.
The animals undergo 1 hour of HCA with standard hemodilution and alpha-stat regulation of arterial blood gases. Once HCA is finished, CPB is restarted and the animals are rewarmed to a core temperature of 37°C over the course of 2 hours. If sinus rhythm does not return spontaneously, the heart is defibrillated at 32°C. At 37°C, the animal is weaned from CPB, the cannulas are removed, vessels are ligated, wounds are appropriately closed, and heparin is reversed with protamine (3 mg/kg intravenously).
The animals then recover from anesthesia while intubated, with frequent monitoring of vital signs, arterial blood gases, and urine output. Once hemodynamically and clinically stable, they are extubated and transferred to their crate for recovery.
Cardiopulmonary bypass
After induction and cannulation, animals undergo 2.5 hours of CPB with no HCA. Animals are cooled to 32°C, similar to routine clinical cardiopulmonary bypass. The heart continues to beat during this operation. The animals recover from anesthesia as described above.
CSF and serum collection
For both baseline and subsequent CSF collection, under sedation and in a routine sterile fashion, the spinal canal is entered with a 22G needle through the cisterna magnum (at the base of the skull posteriorly). Samples are immediately frozen in a –80°C freezer. Blood samples are obtained through previously placed peripheral intravenous catheters, cold centrifuged to collect serum, and frozen at –80°C.
Sacrifice and histologic evaluation
Animals are sacrificed by exsanguination. They are sedated, intubated, and maintained on inhalational anesthesia. The dogs then undergo median sternotomy and cannulation of the ascending aorta using a 22F cannula. CPB is initiated after clamping the descending aorta to ensure the brains are perfused with 12 L cold saline (4°C) at 60 mm Hg. The right atrial appendage is transected, and the venous return is allowed to drain into a reservoir. The brains are harvested with the right hemibrain fixed in 10% formalin and embedded in paraffin. Left hemibrains are stored in saline for future biochemical work (–80°C). Paraffin embedded tissues are sectioned into 8-µm slices and are stained with hematoxylin and eosin stains for histologic evaluation in a blinded manner by a single neuropathologist (J.C.T.).
Eleven distinct regions of the canine brain are evaluated for the presence of apoptosis and necrosis. These regions include midfrontal cortex, superior parietal cortex, basal ganglia, hippocampus (dentate gyrus and pyramidal gyrus), entorhinal cortex, amygdala, cerebellum (molecular layer, Purkinje layer, and granule layer), and brainstem. A semiquantitative scale is used to assess the degree of necrosis and apoptosis in each area. The scale ranges from a minimum value of 0 (no damage) to a maximum of 99 (severe neuronal damage).
CSF analyses for proteolytic fragments of AII-spectrin
All CSF analyses for the presence of SBDPs were performed by colleagues at Baynan Biomarkers (Alachua, FL). Protein concentrations of CSF were determined by DC-Protein assays (BioRad, San Diego, CA) with albumin standards. Protein-balanced samples (7 µL CSF samples) were prepared for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) in twofold loading buffer containing 0.25 M Tris (pH 6.8), 0.2 M DTT, 8% SDS, 0.02% bromophenol blue, and 20% glycerol in distilled water. Protein, 20 µg per lane, was resolved by SDS-PAGE on 6.5% Tris/glycine gels for 2 hours at 200 V. After electrophoresis, separated proteins were laterally transferred to polyvinylidene fluoride membranes in a transfer buffer containing 0.500 M glycine and 0.025 M Tris-HCl (pH 8.3) 10% methanol at a constant voltage of 20 V for 2 hours at 4°C in a semidry transfer unit (Bio-Rad).
Immunoblot analyses
After electrotransfer, membranes were blocked for 1 hour at ambient temperature in 5% nonfat milk in TBS and 0.05% Tween-2 (TBST), then incubated in primary monoclonal II-spectrin, antibody (FG6090; Affinity Research Products, Nottingham, UK) in TBST with 5% milk at 4°C overnight, followed by four washes with TBST and a 2-hour incubation at ambient temperature with either a secondary antibody linked to horseradish peroxidase (enhanced chemiluminescence method, for CSF samples), or biotinylated secondary antibody followed by a 30-minute incubation with strepavidin-conjugated alkaline phosphatase (colorimetric method; for tissue lysate). Enhanced chemiluminescence reagents were used to visualize the immunolabeling on X-ray film. Molecular weights of intact proteins and their potential breakdown products were assessed by running alongside rainbow-colored molecular weight standards. Semiquantitative evaluation of protein and the breakdown product levels were performed through computer-assisted densitometric scanning (Epson XL3500 high-resolution flatbed scanner) and image analysis with Image J software (National Institutes of Health, Bethesda, MD). The results were expressed in arbitrary densitometric units.
Statistical Analysis
All data are presented as means ± SD. Comparisons of both neurologic and histologic scores among groups were performed by one-way analysis of variance (ANOVA). For all subjects, the paired comparisons t test was examined differences in levels of SBDP from baseline to 8 hours after treatment. For animals that survived to 24 hours after treatment, repeated-measures ANOVA was used to account for the repeated CSF samples from within the same canine subject over time. For both one-way and repeated-measures ANOVA, post hoc pairwise comparisons between specific treatments time points were conducted using the Tukey-Kramer method. Statistical analysis was performed with the aid of STATA software, version 9.2 SE (StataCorp LP, College Station, TX).
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Neurologic Scores
The canine scoring system ranges from 0 to 480, with higher scores indicating worse neurologic function [21]. For animals undergoing 1H HCA, mean neurologic scores at 24 hours after HCA ranged from 10 to 105.5 points, with a mean score of 57.6 (median 57.5, SD 53.5). That was in sharp contrast to scores from 2H HCA historical control animals, whose mean neurologic scores were 171 ± 17.8 (range, 149.5 to 186.6; p < 0.05; Fig 2). In contrast to 2 dogs in the 1H HCA group that were effectively normal, with neurologic scores less than 20, no dog receiving 2H HCA had a neurologic score less than 140—all were severely impaired neurologically. Animals undergoing CPB alone had only mild decreases in cognitive function, with mean scores at 24 hours of 6.25 ± 4.8 (range, 0 to 10). No dog in the CPB group appeared to be cognitively impaired at 24 hours, indicating at most mild neurologic injury.
|
|
|
II-spectrin (280 kDa) as well as 150+145 and 120 kDa II-SBDP were visible on immunoblots. Note that these fragments were not present at baseline (Fig 4). For animals that underwent 1H HCA, the levels of the necrotic marker (SBDP 145+150) rose significantly from baseline at 8 hours after HCA (55 ± 48 to 203 ± 128 densiometric units, p = 0.02). The apoptotic marker (SBDP 120) also rose from 37 ± 16 to 121 ± 42, but this change was not statistically significant (p = 0.11). In the series of animals followed and sacrificed at 24 hours after HCA, both SBDP 145+150 and SBDP 120 continued to remain above baseline (242 ± 296 for SBDP 145+150, and 82 ± 96 for SBDP 120; Fig 5). Animals that underwent CPB alone had similar baseline levels to HCA dogs and did not show significant increases in SBDPs. There was a trend toward increased SBDP 120 at 8 hours after treatment in the CPB group (p = 0.06).
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
II-SBDPs in the CSF. Dogs undergoing shortened HCA had a significant increase in the level of the necrosis marker SBDP 145+150 when compared with baseline. In addition, the apoptotic marker SBPD 120 also increased from baseline, although this result was not statistically significant. Importantly, the presence of biomarkers were associated with subtle changes on both histologic and neurologic examination (when compared with dogs undergoing prolonged HCA for 2 hours). This result indicates that SBDPs may be a useful indicator when neurologic injury is mild or difficult to detect. Importantly, for animals undergoing CPB alone, there was no significant increase in either SBDP 145+150 or SBDP 120, indicating that the breakdown products appeared to be consistent with neurologic dysfunction and not from the common insult of having been on CPB.
The fact that the necrosis marker increased in the 1-hour HCA group correlated well with the high levels of necrosis seen on histologic examination at 8 hours in the 1-hour HCA dogs (5.5%). Although a surprising finding, given the high functional level of the animals, SBDPs did correlate with histologic evidence of injury. When SBDP levels at 24 hours after treatment were examined, the necrotic markers continued to rise while apoptotic markers plateaued. The significance of this trend is unclear, but may indicate a difference in time course of generation of the 145+150 fragment and 120 fragment of II-spectrin.
We chose to use 1 hour of HCA in this model, because that is the approximate duration of HCA used clinically. Our goal was to create subtle neurologic injury that would lend itself to the use of biomarkers for diagnosis. Historical 2-hour HCA controls were used to verify that this subtle injury pattern had been achieved. We are confident that 1 hour of HCA produces a more clinically relevant neurologic injury than the traditional model of severe injury produced with 2 hours of HCA. The group of dogs undergoing CPB alone were included both to serve as controls and to determine if SBDPs would increase in the presence of CPB alone. Although definitive injury was not observed with CPB alone through either histology or clinical examination, it is noteworthy that there was a trend toward increasing apoptotic markers in dogs receiving CPB alone. This observation raises the possibility that CPB can lead to subtle neurologic injury in these animals. Additional work is needed, however, to see whether these changes in biomarkers translate into documented neurologic injury or changes in neurocognition.
Use of II-Spectrin Breakdown Products as Brain Injury Biomarkers
The use of II-SBDPs as biomarkers offers numerous advantages over existing biomarkers. Although II-spectrin is not specific to the CNS, it is present in high levels in neurons. Furthermore, its presence in glial cells is minimal, making it highly specific for neuronal damage. Alpha II-spectrin is not found in erythrocytes. This important property makes blood contamination unlikely to affect results of sample collection [9]. Importantly, studies of traumatic brain injury have documented increases in calpain and caspase-3 proteases after injury, translating into increases in II-SBDPs released in the CSF [8]. These properties make II-SBDPs attractive candidates for biomarkers despite the relative lack of brain specificity. That there are two fragments, specific for necrosis (calpain-mediated 145/150 kDa fragment) and apoptosis (caspase-mediated 120 kDa fragment), further reinforces its potential utility. Necrosis and apoptosis represent the two main cell-death pathways in the central nervous system. Different injuries confer distinct patterns of necrosis and apoptosis, and as a result, differential levels of the fragments may yield important insights into injury mechanism. Studies in humans have further shown temporal changes in calpain- and caspase-related fragments over time (16). These unique properties underscore the potential importance of II-SBDPs for monitoring of brain injury from a variety of etiologies, including those that occur from CPB and HCA.
Prior Work
A significant amount of basic research has now confirmed the importance of II-SBDPs as potential biomarkers in brain injury [8, 9, 22]. Specifically, there have been in vitro neuronal cell culture models of mechanical stretch injury [23] as well as excitotoxicity [24] and glucose oxygen deprivation [25] that confirm the potential importance of these biomarker in human injury after cardiac surgery. In vivo work in mice has suggested that the presence of SBDPs correlates with permanent neurodegeneration in a model of hippocampal traumatic injury [14].
Cardali and colleagues [15] have recently demonstrated the presence of SBDPs in the CSF of human trauma patients. In a case-control study, patients with Glasgow coma scores of less than 8 were examined for the presence of both caspase- and calpain-derived II-SBDPs. Both were substantially higher in cases than controls, and patients with improved neurologic scores 6 months after injury had lower levels of SBDPs. These findings demonstrate that levels of II-SBDPs may correlate with long-term prognosis after brain injury. A similar study was subsequently performed with Pineda and colleagues [16] with similar results.
Further important work relating to cardiovascular surgery has been performed by Simon and colleagues [26, 27] at the University of Pennsylvania School of Medicine. Using an approach whereby they identified proteins released from cultured neurons, the group identified several important proteins for neurologic degeneration. Those specifically identified include several fragments from II-spectrin cleaved by calpain, plus a deubiquitinating enzyme UCH-L1. The group recently published an important study demonstrating the upregulation of several of these SBDPs and UCH-L1 in 19 surgical cases of aortic arch repair, including 7 performed with HCA [28].
Taken together, these results clearly point to a potential role for examination of II-SBDPs in the setting of neurologic injury from HCA or CPB. Further work with these biomarkers will help to establish the time course and patterns of protein expression in cases of neurologic injury after cardiac surgery.
Study Limitations
Our study is limited by small sample size and limited information on the temporal pattern of biomarker expression. In this initial study, our goal was to confirm the presence of SBDP in our established canine model, evaluate the effect of CPB, and focus on short-term changes. We acknowledge that the current study provides little information on long-term changes in this biomarker. The study is further limited by the use of CSF for detection. Clearly, cannulation of the spinal canal is not possible in the post–cardiac surgery setting when patients are coagulopathic. In cases where preoperative spinal drains are place, CSF samples may be more readily obtainable. We are in the process of developing the ability to reliably measure SBDPs in the serum. As of yet, we do not have such a method. The use of CSF, however, provides initial proof of concept that II-SBDPs are increased with HCA in our canine model. Our experimental design is further limited by a lack of animals receiving anesthesia alone. In this experiment, all animals received inhalational anesthetic agent, and we do not know the effects of that treatment on spectrin breakdown product levels. Finally, no control group received CBP alone (without HCA) at very low temperatures (18°C), and we therefore do not know the incremental effect of arrest beyond hypothermia on brain injury in these animals.
In conclusion, brain injury resulting from various etiologies is a significant international health concern. Unlike other organ-based diseases in which biomarkers exist for diagnosis and guidance of treatment, no such definitive tests exist for brain injury. Our analysis demonstrates that calpain- and caspase-produced II-SBDPs represent an important and novel marker of neurologic injury after HCA that may be useful for identifying subtle injuries and guiding prognosis.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. J. Arnaoutakis, T. J. George, K. K. Wang, M. A. Wilson, J. G. Allen, C. W. Robinson, K. A. Haggerty, E. S. Weiss, M. E. Blue, C. C. Talbot Jr., et al. Serum levels of neuron-specific ubiquitin carboxyl-terminal esterase-L1 predict brain injury in a canine model of hypothermic circulatory arrest. J. Thorac. Cardiovasc. Surg., October 1, 2011; 142(4): 902 - 910.e1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. G. Allen, E. S. Weiss, M. A. Wilson, G. J. Arnaoutakis, M. E. Blue, C. C. Talbot Jr, C. Jie, M. S. Lange, J. C. Troncoso, M. V. Johnston, et al. Transcriptional Profile of Brain Injury in Hypothermic Circulatory Arrest and Cardiopulmonary Bypass Ann. Thorac. Surg., June 1, 2010; 89(6): 1965 - 1971. [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 |
Value corresponds to results from repeated-measures analysis of variance.