| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2007;84:1166-1173
© 2007 The Society of Thoracic Surgeons
a Division of Cardiothoracic Surgery, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts
b Genomics Center, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts
c Division of Cardiac Surgery, University of Western Ontario, London, Ontario, Canada
Accepted for publication April 30, 2007.
* Address correspondence to Dr Sellke, Division of Cardiothoracic Surgery, BIDMC, LMOB 2A, 110 Francis St, Boston, MA 02215 (Email: fsellke{at}caregroup.harvard.edu).
Presented at the Forty-third Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 29–31, 2007.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Methods: Patients undergoing coronary artery bypass grafting or valve procedures who exhibited new-onset PAF (n = 11) and those who remained in sinus rhythm (n = 13) were prospectively matched based on preoperative, intraoperative, and postoperative characteristics. Postoperative atrial fibrillation was assessed by electrocardiogram and must have required initiation of antiarrhythmic therapy or anticoagulation. Right atrial and skeletal muscle samples were harvested before and after cardiopulmonary bypass for oxidative protein immunostaining (Oxyblot assay). Serum samples were collected preoperatively and postoperatively at 6 hours and day 4 for microarray assessment of gene expression and to quantify total peroxide levels.
Results: Patients with PAF had significantly more elevation in total peroxide levels in serum compared with patients in sinus rhythm at 6 hours (5.83 ± 1.9 versus 2.02 ± 0.2 fold, respectively; p = 0.039) but not at day 4 (3.81 ± 1.2 versus 2.17 ± 0.5 fold, respectively; p = 0.188). Patients with PAF also had significantly more myocardial oxidation compared with patients in sinus rhythm at 6 hours (4.19 ± 1.4 versus 0.94 ± 0.3 fold, respectively; p = 0.021). Increased serum peroxide levels in patients who exhibited PAF correlated with elevated myocardial protein oxidation but not peripheral muscle oxidation. Gene expression analysis revealed a differential genomic response in patients with new-onset PAF (more oxidation) compared with patients in sinus rhythm (more reduction).
Conclusions: Patients who exhibit PAF after cardiac surgery have significantly increased acute oxidative stress, which translates into increased myocardial oxidation. Also, patients with PAF have a differential oxidative genomic response after cardiopulmonary bypass that may predispose them to oxidative stress.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
A recent study found that a more pronounced increase in postoperative white blood cell count independently predicts development of PAF [5]. Findings also point to the oxidative and inflammatory response after cardiopulmonary bypass (CPB) as possible etiologic factors in the development of PAF. The most important reactive oxygen species include nitric oxide, superoxide, hydrogen peroxide, and peroxynitrite. Other sources of reactive oxygen species include xanthine oxidase, cyclooxygenase, and cytochrome P-450. Cardiopulmonary bypass elicits the production of reactive oxygen species formation, which likely contributes to post-CPB myocardial dysfunction. Reactive oxygen species cause cardiomyocyte injury by altering and inactivating enzymes, leading to membrane permeability and impairment of ion transport [6, 7], and may be an important cause of arrhythmias [8]. In 2001, Carnes and colleagues [9] showed a direct effect of oxidative stress on electrical remodeling in an animal model of AF.
It has been proposed that electrical remodeling plays a role in the high incidence of early PAF after the ischemia–reperfusion myocardial injury caused by cardioplegic arrest [10]. Certain studies have also demonstrated that administration of agents with antioxidant and antiinflammatory properties, such as statins, have proven effective in preventing atrial effective refractory period shortening and susceptibility to AF in a dog model of rapid atrial pacing [11]. Although the mechanisms of oxidative injury within the fibrillating atrial myocardium are not entirely clear, there is increasing evidence that formation of the oxygen-derived free-radical superoxide by nicotinamide adenine dinucleotide phosphate oxidases contribute to myocardial disease and the myocardial oxidative injury during AF [12, 13].
In other studies, it was demonstrated that the reactive oxygen species scavenger N-acetylcysteine reduced the neutrophil oxidative burst response after CPB or cardioplegic arrest [14], and Tossios and associates [15] reported that it attenuates myocardial oxidative stress after CPB or cardioplegic arrest. However, none of these studies directly investigated, in the context of PAF, the effects of oxidative stress response on myocardial tissue compared with peripheral oxidative responses. The present study examined myocardial, peripheral (skeletal muscle), and systemic (serum) oxidative responses after CPB or cardioplegic arrest in a cohort of patients who exhibited PAF compared with matched patients in sinus rhythm (SR). We also examined the differences in gene expression profiles of patients who exhibit PAF after cardiac surgery compared with those who do not exhibit this complication in an effort to improve our understanding of PAF pathophysiology and treatment as well as prevention through gene-based risk-stratification strategies.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Patients with preoperative documented chronic or paroxysmal AF or who were taking antiarrhythmic medications (eg, amiodarone, sotalol) were excluded. Similarly, subjects with significant comorbidities such as advanced hepatic disease (cirrhosis) and chronic renal failure (serum creatinine > 2.0 mg/dL) were excluded.
New-onset PAF was defined as a sustained episode of AF, as assessed by electrocardiogram, before postoperative day 5 that required electrical or chemical cardioversion or anticoagulation.
Subjects with PAF were matched prospectively with subjects who remained in sinus rhythm (SR) on the basis of preoperative baseline characteristics (age, sex, hypertension, hypercholesterolemia, diabetes mellitus, leukocyte count, and ß-blocker use), intraoperative characteristics (procedure type, CPB time, aortic cross-clamp time, antifibrinolytic use, cardiotomy suction use), and postoperative characteristics (time to extubation and ß-blocker use). Because of the small, diverse sample, the matching process was appropriate to allow comparisons of characteristics that might influence the onset of PAF. All molecular and serologic studies were carried out after the matching process between SR and PAF groups in a blinded fashion.
Anesthetic and Surgical Technique
Conventional operative approach at our institution was followed, including induction of general anesthesia, invasive monitoring, midline sternotomy, and systemic heparinization. Cardiopulmonary bypass was initiated through cannulation of the right atrium and ascending aorta with a nonpulsatile system, membrane oxygenator, and a 40-µm arterial filter. Crystalloid pump priming solution was used. For all patients, mild hypothermic CPB (minimum temperature, 32° to 34°C) with intermittent cold-blood hyperkalemic (25 mmol/L) cardioplegia was used. Serum glucose levels were monitored and attempted to be maintained at less than 130 mg/dL by intermittent intravenous insulin injections or insulin infusion. During CPB, pump blood flow was maintained at 2 to 2.4 L · min–1
· m–2 body surface area. Arterial partial oxygen pressure was kept between 150 and 250 mm Hg, and alpha-stat pH monitoring was used. Mean blood pressure was maintained at between 50 and 90 mm Hg using conventional vasoactive medications.
Tissue Harvesting and Oxyblot Immunostaining
Samples of right atrial appendage and skeletal muscle were harvested from enrolled patients undergoing coronary artery bypass grafting surgery before and after exposure of the heart to blood cardioplegia and short-term reperfusion under conditions of CPB as previously described [16]. Samples were harvested with cold sharp dissection and handled in a nontraumatic fashion. Double 3-0 polypropylene pursestring sutures (Ethicon, Somerville, NJ) were placed in the atrial appendage. During placement of the venous cannula, the first sample of atrial appendage was harvested (pre-CPB). The superior suture was tightened to secure the venous cannula. The inferior suture remained loose to allow this portion of the atrium to be perfused with blood, exposed to CPB and cold blood cardioplegia, and reperfused after removal of the aortic cross-clamp. The cardioplegia consisted of a 4:1 mixture of oxygenated blood with a hyperkalemic crystalloid solution. The second sample of atrial appendage (post-CPB) was harvested after CPB during removal of the venous cannula. Skeletal muscle samples were taken from the internal mammary artery harvest site. Myocardial and skeletal muscle tissue (10 x 10 x 2 mm) was immediately frozen in liquid nitrogen for Oxyblot immunostaining.
The Oxyblot Oxidized Protein Detection Kit (Chemicon International, Temecula, CA) was used for detection of overall carbonyl groups introduced into proteins by oxidative reactions with ozone or oxides of nitrogen. 1,3-Dinitrophenylhydrazine (DNPH) derivatization was carried out for 15 minutes following the manufacturer’s instruction on 6 µg of protein obtained from the tissue lysate. One-dimensional electrophoresis was carried out on 12.5% sodium dodecyl polyacrylamide gel after derivatization. Proteins were transferred to nitrocellulose membranes and then stained by Red Ponceau. After incubation with anti-dinitrophenylhydrazine antibody, blots were developed using a chemiluminescence detection system. Densitometry was performed on scanned gels using an NIH image system. Fold change was calculated (post-CPB divided by pre-CPB) for right atrial and skeletal muscle samples of matched patients.
Serum Peroxide Levels and Microarray Processing
For each of the 40 patients enrolled, blood samples were collected from the central venous catheter preoperatively after induction of anesthesia and before skin incision as well as 6 hours postoperatively in the cardiovascular intensive care unit and on postoperative day 4. Samples were collected into sterile vacuum tubes and immediately centrifuged at 15,000g for 15 minutes, and serum or plasma samples were frozen at –80°C until the time of assay at the conclusion of the study. Total peroxide levels in serum were quantified in duplicate as a measure of oxidative stress using the Oxystat assay (OxyStat, ALPCO Diagnostics, Salem, NH).
For microarray gene expression analysis, whole blood samples were collected preoperatively and at 6 hours postoperatively into PAXgene tubes (QIAGEN Inc, Valencia, CA) for immediate mRNA stabilization and extraction as per manufacturer’s recommendation. Transcriptional profiles of samples were probed using Affymetrix HG U133 Plus 2.0 chips for a total of 84 chips. Total RNA extraction and purification, cDNA synthesis, in vitro transcription reaction for production of biotin-labeled cRNA, hybridization of cRNA with Affymetrix GeneChips, and scanning of arrays were done according to previously described protocols [17]. All arrays went through stringent quality control assessment with regard to 3'-5' ratios (using glyceraldehyde-3-phosphate dehydrogenase and ß-actin probes), percentage of present calls, and array outlier call percentage within 2 standard deviations of the mean. All scanned array images passed the quality controls and were analyzed by dChip [9], which has been shown to be more robust than Affymetrix software Microarray Analysis Suite (MAS) 5.0 in signal calculation for about 60% of genes [18]. In the dChip analysis a smoothing spline normalization method was applied before obtaining model-based gene expression indices, also called signal values. There were no outliers identified by dChip, so all samples were carried on for subsequent analysis.
Unsupervised and Differential Expression Analysis
A hierarchical clustering technique was used to construct an unweighted pair group method with arithmetic mean (UPGMA) tree using Pearson’s correlation as the metric of similarity. When comparing two groups of samples to identify genes enriched in a given group, we used the lower confidence bound (LCB) of the fold change (FC) between the two groups (6 hours versus preoperative) as the cutoff criteria. If 90% LCB of FC between the two groups was above 1.2, the corresponding gene was considered to be differentially expressed. The LCB is a stringent estimate of the FC and has been shown to be the better ranking statistic [9]. Recently, dChip’s LCB method for assessing differentially expressed genes have been shown to be superior to other commonly used approaches, such as MAS 5.0 and Robust Multiarray Average based methods [19, 20]. Using custom arrays and quantitative reverse transcriptase real-time polymerase chain reaction, it has been suggested that Affymetrix chips may underestimate differences in gene expression [21]. On the basis of this work and others [22], a criterion of selecting genes that have an LCB above 1.2 most likely corresponds to genes with an "actual" fold change of at least three in gene expression.
Real-Time Polymerase Chain Reaction
After total RNA extraction, concentrations were determined by spectrophotometry; each sample yielded a minimum of 10 µg of total RNA with an A260/A280 ratio ranging between 1.7 and 1.9. Integrity of total RNA was confirmed by agarose gel electrophoresis, and only samples with a ratio of 2:1 28S/18S were used for analysis.
The quantitative real-time SYBR Green reaction was performed in duplicate according to recommended protocols provided by the TaqMan ABI 7700 (Applied Biosystems, Foster City, CA). The RT2-PCR primers for the human genes analyzed were obtained commercially (Superarray Bioscience Corporation, Frederick, MD). Primers for the 3' ends of genes used for microarray validation included downregulated genes HLA-DQA1, FBP2, and PYGB and upregulated genes CYCL2, IFIH1, and TFAP2B.
Clinical Data Analysis
Clinical data were expressed as mean ± standard deviation. The 
2 test or Fisher’s exact test were used to compare proportions, and the Student’s t test was used to compare continuous variables. Software packages used are GraphPad Prism 4 (GraphPad; Prism 4, San Diego, CA) and SPSS (SPSS 11.5 for Windows; SPSS Inc, Chicago, IL).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
Myocardial and Skeletal Muscle Oxidation
Myocardial oxidative stress was significantly increased in the PAF group compared with the SR group after CPB (4.19 ± 1.4 versus 0.94 ± 0.2 fold change, respectively; p = 0.021) as summarized in Figure 1. On the other hand, skeletal muscle samples did not demonstrate a significant difference in the level of oxidation between the PAF group compared with the SR groups (0.90 ± 0.1 versus 0.81 ± 0.1 fold change, respectively; p = 0.575).
|
|
Unsupervised Analysis of Gene Expression
Using the described microarray GeneChip, we generated a nearly comprehensive database of gene expression in patients in the PAF and SR groups before and after CPB. A complete list of gene expression, upregulated and downregulated gene lists, and functional and correlation analysis are provided in the online supplementary data Web site (http://www.bidmcgenomics.org/AFIB/index.html).
Unsupervised cluster analysis of complete array expression values produced expression similarities across the samples (Figure 3). Hierarchical clustering of all samples demonstrated a clear distinction based on the sample collection time: preoperative versus 6 hours. In the PAF group, 723 genes were upregulated and 5,779 genes were downregulated at 6 hours compared with preoperative. In the SR group, 1,365 genes were upregulated and 6,690 genes were downregulated in 6 hours compared with preoperative. When these gene lists were compared, we found that 322 genes were commonly upregulated at 6 hours versus preoperative regardless of the rhythm status, 723 genes were uniquely upregulated in PAF patients, and 1,043 genes were uniquely upregulated in SR patients. Among the genes that were downregulated in 6 hours versus preoperative, 4,877 genes were commonly downregulated regardless of the rhythm status, 902 genes were uniquely downregulated in patients with PAF, and 1,813 genes were uniquely downregulated in SR patients (Fig. 4).
|
|
Differential Expression of Oxidative Stress Genes in PAF Group
Genes that were uniquely upregulated or downregulated at 6 hours compared with preoperative time points in either group (PAF or SR) were subjected to further detailed analysis for identification of genes involved in oxidative metabolism.
Table 2 summarizes differentially expressed genes involved in oxidative metabolism and stress in the PAF or SR groups. Genes included those involved in mitochondrial redox metabolism, pyruvate dehydrogenase kinase, cytochrome c oxidase, nicotinamide adenine dinucleotide dehydrogenase, biliverdin reductase A, and amine oxidase. Genes uniquely upregulated in the SR group were predominantly genes whose products were responsible for cellular reduction reactions (gaining electrons or reducing oxidative stress). Genes uniquely upregulated in the PAF group were mostly genes responsible for mediating oxidation reactions within the cell (removing electrons) and thus contribute to cellular oxidative stress.
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Results from this study demonstrate that in similarly matched cardiac patients undergoing cardiac surgery, patients who exhibit PAF were associated with a significantly larger amplitude increase of oxidant stress systemically (as measured by total peroxide levels) as well as increased oxidative stress at the myocardial level (as measured in the right atrium). These findings are significant for two reasons. First, this study confirms the strong association between the development of PAF after CPB and oxidant stress. Second, it points out that in patients with PAF, this differential response is present at the systemic level, not just within the myocardium. Such findings indicate that patients who exhibit PAF have a differential response at the systemic level. This response then is associated with and translates to increased oxidation in the heart as a result of ischemia–reperfusion, but not in other areas that are continuously perfused such as skeletal muscle. In fact, myocardial oxidation was directly associated with increasing serum peroxide levels.
Moreover, in an interesting finding that supports the previous two results, we found that PAF patients have a differential genomic response after CPB compared with SR patients. There were genes in this group that were uniquely differentially expressed depending on whether they were in the PAF or SR groups. When these genes were further analyzed, it was found that most redox genes uniquely upregulated in the PAF group were genes that supported reactions promoting oxidation reactions, contributing to the perpetuation of oxidative stress. Such genes included pyruvate dehydrogenase kinase, isozyme 1 (PDK1), nicotinamide adenine dinucleotide dehydrogenase (ubiquinone) 1 (NDUFAB1), and aldo-keto reductase family 1-dihydrodiol dehydrogenase 1 (AKR1C1). On the other hand, genes that were uniquely upregulated in the SR group mostly supported reduction reactions that diminished the overall oxidative stress load. Examples of such genes are superoxide dismutase (SOD2), carbonyl reductase 4 (CBR1), and methionine sulfoxide reductase A (MSRA). It is possible that in the PAF and SR groups, the overall redox balance favors oxidation and reduction, respectively.
Other studies have examined the effect of reactive oxygen species scavengers and antioxidant administration during cardiac surgery. As early as 1994, Yau and colleagues [24] examined the effects of vitamin E on patients undergoing coronary artery bypass grafting operations and found that there was a small, but significant, effect and recommended its use in high-risk patients. More recently, although animal models have been mostly positive, human studies have reported conflicting evidence regarding the benefits of N-acetylcysteine and L-arginine [15, 25, 26]. In an interesting study, Knott and colleagues [27] reported that pyruvate-fortified cardioplegia suppresses oxidative stress and enhances phosphorylation potential of arrested myocardium.
Findings from our study shed new light on the pathophysiology of PAF as they point out that patients who exhibit PAF may be different at the genomic response level and are predisposed to a higher magnitude of oxidative response after the stress of CPB. Further studies may allow us to tailor systemic therapy and cardioplegic arrest methods to patients who may be genetically predisposed to such a response. Also, it may be that such data would form the basis for a database of genetic information that may provide beneficial gene-based risk-stratification strategies. We hope that further study of differentially expressed genetic pathways, in addition to established knowledge of clinical risk factors and perioperative factors, will translate into better understanding of PAF pathophysiology after CPB.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Let me please ask you whether you’ve looked at the actual cholesterol levels in the two groups, ie, in patients who developed AF versus those who did not. As you know, there is accumulating evidence that statins may be involved in preventing postoperative AF. There is no class 1 evidence yet, but certainly there is some suggestion of it.
Therefore, I was wondering whether you have compared the proportion of treated patients and doses of statins in patients from each group, and also whether there was a difference in preoperative cholesterol levels between the groups, which also could impact oxidative stress?
DR RAMLAWI: Absolutely. This is a very valid point. Oxidative stress has been shown to be decreased with statin use, but in this small study we did not use it for the matching.
Post hoc analysis, which we looked at, revealed that there is no difference in statin use between the two groups; but we do expect that further studies would show a difference.
DR RUEL: And were you able to actually compare the actual cholesterol levels as opposed to the proportion of patients who had high cholesterol?
DR RAMLAWI: No, we did not do that.
DR MICHAEL E. JESSEN (Dallas, TX): I enjoyed this paper.
It looked like about a third of your patients had a half-dose of aprotinin and there was balance in both groups. But if you went back and looked at the ones that got aprotinin, did that affect oxidative stress levels in the heart at all?
DR RAMLAWI: We did not do that analysis, no. But a lot of the studies involving aprotinin have involved full-dose aprotinin. And in those studies, there is a significant reduction, as we know, in the inflammatory response. So we do, by correlation, expect a certain amount of decrease in oxidative stress, but we did not do that for this particular group for the half-dose aprotinin.
DR JESSEN: And do you plan to look at off-pump patients to see if they exhibit less oxidative stress with this type of analysis?
DR RAMLAWI: That would certainly be interesting. We did not look at that.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. H. Rennison and D. R. Van Wagoner Impact of Dietary Fatty Acids on Cardiac Arrhythmogenesis Circ Arrhythm Electrophysiol, August 1, 2009; 2(4): 460 - 469. [Full Text] [PDF]
|
|
|
|
 |
|
 A. Anselmi, G. Possati, and M. Gaudino Postoperative inflammatory reaction and atrial fibrillation: simple correlation or causation? Ann. Thorac. Surg., July 1, 2009; 88(1): 326 - 333. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. L. Baker, M. W. Anglade, E. L. Baker, C. M. White, J. Kluger, and C. I. Coleman Use of N-acetylcysteine to reduce post-cardiothoracic surgery complications: a meta-analysis Eur. J. Cardiothorac. Surg., March 1, 2009; 35(3): 521 - 527. [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 |
