HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Data Supplement Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Feezor, R. J. Articles by Seeger, J. M. Search for Related Content PubMed PubMed Citation Articles by Feezor, R. J. Articles by Seeger, J. M.

Genomic and Proteomic Determinants of Outcome in Patients Undergoing Thoracoabdominal Aortic Aneurysm Repair¹

Robert J. Feezor^*, Henry V. Baker, Wenzhong Xiao, W. Anthony Lee^*, Thomas S. Huber^*, Michael Mindrinos, Raymond A. Kim, Laurence Ruiz-Taylor, Lyle L. Moldawer^2,*, Ronald W. Davis and James M. Seeger^*

Departments of ^* Surgery and Molecular Genetics and Microbiology, University of Florida College of Medicine, Gainesville, FL 32610; Stanford Genome Technology Center, Palo Alto, CA 94305; and Zyomyx, Hayward, CA 94545

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Thoracoabdominal aortic aneurysm repair, with its requisite intraoperative mesenteric ischemia-reperfusion, often results in the development of systemic inflammatory response syndrome, multiorgan dysfunction syndrome (MODS), and death. In the present study, an adverse clinical outcome following thoracoabdominal aortic aneurysm repair was identified by blood leukocyte genomic and plasma proteomic responses. Time-dependent changes in the expression of 146 genes from blood leukocytes were observed (p < 0.001). Expression of 138 genes (p < 0.001) and the concentration of seven plasma proteins discriminated between patients who developed MODS and those who did not, and many of these differences were evident even before surgery. These findings suggest that changes in blood leukocyte gene expression and plasma protein concentrations can illuminate pathophysiological processes that are subsequently associated with the clinical sequelae of systemic inflammatory response syndrome and MODS. These changes in gene expression and plasma protein concentrations are often observed before surgery, consistent with either a genetic predisposition or pre-existing inflammatory state.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Patients who undergo thoracoabdominal aortic aneurysm (TAAA)³ repair or mesenteric revascularization for chronic ischemia develop organ failure and die with alarmingly high frequency. Rectenwald et al. (1) reported a mortality of 10% following elective TAAA repair, but strikingly, noted that 77% of the patients developed significant postoperative complications. A recent national review found similar mortality and complication rates of 20.3 and 62.2%, respectively, for TAAA repair (2). Reperfusion following intraoperative visceral ischemia associated with TAAA repair is presumed responsible for the release of proinflammatory cytokines and other inflammatory mediators. This activation of innate immunity may lead to the development of a systemic inflammatory response syndrome (SIRS), multisystem organ dysfunction syndrome (MODS), and even death (3, 4).

Svensson et al. (5) first reported that the duration of intraoperative intestinal ischemia was one of several independent predictors of postoperative morbidity following TAAA repair, others including intraoperative blood loss and shock. We subsequently observed that MODS occurred more commonly in patients undergoing TAAA repair whose intraoperative mesenteric ischemia exceeded 40 min (6). Increased circulating levels of TNF-, IL-6, IL-8, IL-10, and shed TNF receptors after TAAA repair were also associated with longer intraoperative mesenteric ischemic times, and the frequency of postoperative organ dysfunction correlated with elevated postoperative plasma concentrations of IL-6 and/or TNF- (7).

Unfortunately, it is often difficult to identify or predict which patients will have an adverse outcome after acute injuries such as TAAA repair. The advent of microarray technology and multiplex protein analyses has permitted a large-scale survey of the genomic and proteomic responses to surgical and visceral ischemia-reperfusion injury. In the present study, gene expression patterns from peripheral blood leukocytes and circulating plasma protein levels were determined in patients undergoing TAAA repair to explore whether an assessment of genome-wide expression and inflammatory protein production could be used to discriminate between a dichotomous clinical outcome. Such approaches have been used to characterize the malignant phenotype of tumors and their responsiveness to antineoplastic therapies (8, 9, 10, 11). The present findings demonstrate that there are differences in leukocyte gene expression and plasma protein concentrations that may identify outcome following visceral ischemia-reperfusion injury. More importantly, these patterns of genome-wide expression and plasma protein concentration are often evident before the surgical trauma and the induction of visceral ischemia-reperfusion injury.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Patient selection

Ten patients undergoing repair of thoracoabdominal aortic aneurysms were enrolled. All procedures were performed electively by one of three surgeons at a single institution over 17 mo. The study was approved by the Institutional Review Board at the University of Florida, and each patient provided signed informed consent.

Blood sample collection and RNA isolation

Whole blood was collected into heparinized phlebotomy tubes from each patient after induction of anesthesia, but before the start of surgery (presurgery), and at 2 and 24 h after establishment of mesenteric reperfusion. Each sample was immediately centrifuged at 4°C, 400 x g for 10 min. The plasma was separated and saved for proteomic analysis (see below), and the buffy coat layer containing the leukocyte populations was recovered for genomic analysis. After contaminating erythrocytes were removed by lysis (buffer EL) (7) (Qiagen, Valencia, CA), total cellular RNA was isolated from the buffy coat using a commercial kit (RNeasy) (7) (Qiagen). Purity was confirmed by spectrophotometry (A₂₆₀/A₂₈₀ ratio), an ethidium bromide-stained RNA agarose gel, and RT-PCR amplification of a housekeeping gene (Cu/Zn superoxide dismutase) (12). Total and differential white blood cell counts were also obtained from an additional blood sample through the diagnostic laboratories at Shands Hospital (Gainesville, FL).

cRNA synthesis and chip hybridization

cRNA was synthesized from 10 µg of total cellular RNA based on the protocol outlined by Affymetrix (Santa Clara, CA), with few modifications as outlined below. cRNA was transcribed in vitro incorporating biotinylated nucleotides using an ENZO BioArray HighYield RNA Transcript Labeling Kit (T7) (Enzo Life Sciences, Farmingdale, NY) and hybridized onto Hu95aVer2 oligonucleotide arrays (Affymetrix). The microarrays were hybridized for 16 h at 45°C, stained, and washed according to an Affymetrix protocol (EukGE-WS2v4) using an Affymetrix fluidics station, and scanned with an Affymetrix scanner. Fluorescence intensity measurements were calculated and normalized by Microarray Suite, version 5.0 (MAS 5.0; Affymetrix).

Microarray data analysis

After excluding Affymetrix control probe sets, each microarray detected the hybridization intensity of 12,558 probe sets. A total of 5,163 probe sets (41%) was never detected above background intensity levels on any array, and was removed from the dataset. Clustering analysis was performed on a normalized variance of each remaining probe set wherein the mean expression across all microarrays was set to 0 and the SD to 1. Supervised analysis on the cube root of the MAS 5.0 signal expression values was accomplished using significance analysis of microarrays (SAM) (13), which uses modified t and F tests, and data permutation to identify probe sets with statistically significant differences in hybridization signal intensity, while allowing for tuning to control the false discovery rate (FDR). The variation in hybridization intensity used to infer expression of the genes represented by the probe sets identified in this manner was clustered using Cluster (Eisen Laboratory, Stanford University) and displayed with TreeView (Eisen Laboratory) (14).

For each time point studied, further analysis was performed on the mean fold change in expression level for each probe set between patients who developed MODS and those who did not. MicroArray Pathway Profiler Finder (MAPPFinder; Gladstone Institutes, University of California, San Francisco, CA) (15) identified classes of genes according to the Gene Ontology Consortium (16) that were statistically overrepresented among genes 2-fold up- or down-regulated (MODS vs no MODS) at each time point. MAPPFinder uses a statistical probability to rank the likelihood that the specific gene ontology is represented in the gene list of interest more often than would occur by chance alone.

Proteomic analyses

Plasma was obtained by whole blood centrifugation at 4°C, 400 x g before surgery, and after induction of anesthesia, immediately after the reperfusion, and at 1, 2, 4, 6, 8, 24, 48, and 72 h after reperfusion. Samples were drawn into EDTA-containing tubes, except when the samples corresponded to those designated for genomic analysis (presurgery, and 2 and 24 h after reperfusion). After initial centrifugation of the whole blood, plasma was removed and frozen at 80°C until further analysis. Patient plasma samples were analyzed using the Zyomyx Protein Profiling Biochip System and the Zyomyx Human Cytokine Biochip (Hayward, CA). The Human Cytokine Biochip measures 30 different cytokines, chemokines, and growth factors in a multiplexed assay format. The samples were assayed without any sample treatment. Duplicate data were collected, with each assay performed on separate biochips. Quantification of cytokine concentrations was obtained through six-point multianalyte calibration run concurrently with the sample biochips. Median feature intensity was background subtracted after outlier removal, and statistical significance of the signal intensity above background was determined for a 95% confidence interval with a modified Z-factor parameter calculated with 2 SDs (17).

To analyze the differential plasma concentrations of the 30 proteins between patients who developed MODS and those who did not, statistical significance was estimated by a modified t test with permutations of data as described in the SAM software program (13). Because of the limited number of analytes, probability values were reported, instead of FDRs. A conservative significance level of p < 0.003 was used, corresponding to probability <0.1 after Bonferroni correction for multiple testing (p-B).

Clinical organ failure assessment

Clinical postoperative organ failure was based on the Modified Brussels Organ Failure Score (18). This system uses physiologic parameters in six different organ systems to assess multiorgan dysfunction, and each patient received an organ failure score daily for 14 postoperative days. Statistical analyses on clinical data were performed by Student’s t test or Mann-Whitney rank sum test, where appropriate, and are reported as mean ± SD.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
The mean age of the patients was 71.0 ± 6.0 years (range 64–85), and five of the patients were male (50%). Comorbidities included tobacco use (100%), hypertension (70%), chronic obstructive pulmonary disease (40%), coronary artery disease (40%), peripheral vascular occlusive disease (40%), diabetes mellitus (30%), and cerebrovascular occlusive disease (20%). The mean aortic cross clamp time among patients was 49.8 ± 18.5 min, and the estimated intraoperative blood loss was 4510 ± 5180 ml. The mean hospital length of stay was 28.3 ± 17.3 days.

The mean postoperative organ failure score among all 10 patients was 3.20 ± 1.93. The 5 patients who developed multiorgan dysfunction by definition had higher MODS scores than those who did not (5.0 vs 1.4, respectively). Patients who developed MODS also had a longer duration of intraoperative visceral ischemia (62.4 ± 13.3 vs 37.2 ± 15.5 min, p = 0.025), higher intraoperative blood loss (6700 ± 6900 ml vs 2300 ± 490 ml, p = 0.032), and a longer postoperative hospitalization (42.6 ± 8.9 days vs 14.0 ± 9.1 days, p < 0.001). The organ system most often rendered dysfunctional was pulmonary, followed, in order, by the coagulation system, renal system, CNS, cardiac, and hepatic system. Three of the patients with MODS died, giving an overall operative mortality of 30%.

Time-dependent changes in gene expression were present in peripheral blood leukocytes. As shown in Fig. 1, the expression of 152 probe sets (with an FDR of 8 or 5%) representing 146 genes discriminated among presurgery, 2 h, and 24 h after reperfusion (Fig. 1 and Supplemental Table A).⁴ Although the separation in gene expression among the time points was not perfect, examination of the dendrogram revealed that the 24-h expression patterns clustered separately (with the exception of one sample) from both the preischemia and 2-h patterns. This temporal expression pattern could be broken down into several categories, depending upon the kinetics of gene expression. The expression of 119 genes increased after surgery, whereas the expression of 27 genes decreased. The largest category identified 78 genes whose expression increased 24 h after reperfusion injury, followed by 41 genes whose expression increased at 2 h. Several of the genes whose expression increased early after reperfusion injury were expected early activation genes, and genes involved in the innate immune response (Supplemental Table A). They included leukocyte secretory products involved in innate immunity (perforin, granzyme B, TNF--induced protein 3), cytokines and their receptors (TNFRII, CXCR3), and signal transduction intermediates (STAT4, cytokine-inducible kinase, mitogen-activated protein kinase, mitogen-activated protein kinase, kinase, C/EBP, C/EBP). In contrast, genes whose expression decreased after the surgery included a number of MHC class II peptides, and other cytokine receptors (CXCR 1, IL-2R chain).

View larger version (71K): [in this window] [in a new window]   FIGURE 1. A supervised (SAM) analysis identifies 152 probe sets (146 genes) that successfully discriminate gene expression among peripheral leukocytes from samples obtained before surgery (brown blocks), and at 2 (blue blocks) and 24 h (yellow blocks) after reperfusion injury. The colors associated with the cluster analysis represent variance normalized expression for individual genes in which the intensity of the color red indicates relative expression greater than the mean for that individual gene, green indicates expression less than the mean, and black indicates mean expression. The resulting dendrogram provides a relative relationship among the 30 samples, as determined by the Pearson correlation coefficient. The vertical distance to common nodes of intersection on the dendrogram represents the degree of similarity between the samples. Identified on the vertical axis are three bins that identify the relative changes in gene expression that occur with time. The preaortic cross clamp samples and 2-h postreperfusion samples are more similar to each other than they are to the 24-h postreperfusion samples. There appear to be three predominant patterns of expression following the surgical and reperfusion injury: a transient 2-h increase in expression, a 24-h increase in expression, and a decrease in gene expression at 2 and 24 h.

View larger version (69K): [in this window] [in a new window]   FIGURE 2. A supervised analysis revealed that the expression pattern of 146 probe sets (138 genes) could discriminate between patients who developed MODS (fuchsia blocks) and those who did not (green blocks). A detailed explanation for the dendrogram and cluster analysis is provided in the legend for Fig. 1. Presurgical patterns of gene expression did not cluster together, but rather were equally distributed between patients with and without the subsequent development of MODS. These findings suggest that the divergent patterns of gene expression were already apparent presurgery in this patient population.

We subsequently asked whether the changes in gene expression after surgery or in patients who developed MODS were secondary to changes in the differential leukocyte pattern. Gene expression profiles were based on a constant amount of leukocyte RNA, and variations in the differential count could explain the changes in gene expression. As shown in Table I, however, the total and differential leukocyte count did not change dramatically over the first 24 h, with the exception that the percentage of lymphocytes declined with time. More importantly, the differential leukocyte counts did not significantly differ in the first 24 h when the genomic analyses were performed between patients going on to develop MODS, and those who did not. Therefore, it is unlikely that the differences in expression can be simply explained by differences in the leukocyte pattern.

View larger version (79K): [in this window] [in a new window]   FIGURE 3. The Zyomyx Human Cytokine Biochip measured the circulating levels of 30 known cytokines. For each cytokine, the concentrations have been normalized, and the colors of the squares indicate the normalized concentrations (red (+3 SD) to blue (–3 SD)). Individual samples from the five patients who developed MODS are represented on the left side of the figure, and individual samples are listed chronologically to include: presurgery (preaortic occlusion; PAC), immediately after reperfusion (ACR), and then at 1, 2, 4, 6, 8, 24, 48, and 72 h after reperfusion. Clearly, while not all cytokines measured are informative at discriminating between the divergent clinical outcomes of patients, MCP-1, IL-10, MIG, IL-6, IL-8, IP-10, and sCD23 were globally predictive.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
SIRS is the clinical manifestation of a nonspecific activation of the immune system commonly observed following multiple trauma, thermal injury, pancreatitis, or visceral ischemia-reperfusion injury. The incidence of MODS and subsequent death with SIRS is alarmingly high, and despite relative success in animal models of systemic inflammation, most current clinical approaches are limited to supportive therapy. Current injury severity scoring systems provide a static assessment of a patient’s = s physiological condition, but, at present, there are no reliable methods to predict which patients will develop SIRS or MODS. Patients with similar baseline characteristics and nearly identical injuries often have very differing clinical outcomes.

Although the mechanisms by which SIRS progresses to MODS are complex and not fully resolved, inappropriate activation of the innate immune response has been implicated from preclinical and animal studies (3). Unfortunately, clinical trials aimed at inhibiting the inflammatory mediators responsible for the propagation of SIRS have generally failed to improve outcome (19, 20).

Patients undergoing TAAA repair afford unique opportunities to study the SIRS response because the patient population is relatively homogeneous, the injury is reproducible, the timing of the injury is precisely known, and the incidence of MODS is high. In the present study, we investigated whether genome-wide leukocyte gene expression patterns or plasma protein levels, as indicators of the innate immune response, could identify patients at risk of developing MODS.

For genome-wide analyses, we used the Affymetrix U95a oligonucleotide microarray, which simultaneously interrogates 12,558 probe sets. These expression patterns represent a significant fraction of the entire human transcriptome, estimated to be the product of roughly 30,000 genes (21). In addition, we measured simultaneously the plasma concentration of 30 analytes. A supervised analysis identified 146 genes and 7 plasma proteins whose patterns were associated with an adverse clinical outcome. Several of these genes are involved in the innate immune response, and reflect activation of leukocyte subpopulations (neutrophil collagenase (matrix metalloproteinase-8), RANTES, pentaxin-related gene, monocyte to macrophage differentiation-associated factor, platelet factor-4, and vascular endothelial growth factor). Others represent a generalized stress response (increased expression of heat shock proteins, and proteins of the ubiquitin-proteosome system). Simultaneously, there was a down-regulation in the expression of several genes involved in the MHC class I response.

Probably, the most compelling observation was that many of these changes in leukocyte gene expression and plasma protein concentrations were evident before the induction of visceral ischemia-reperfusion injury. Furthermore, the expression of only 1 of these 146 genes was common to the gene expression response to the surgery and ischemia-reperfusion injury. Thus, the genes that appeared to discriminate between a favorable and adverse outcome were not the same genes involved in the response to the surgical and reperfusion injury. Similarly, although the duration of visceral ischemia, blood loss, and length of hospitalization significantly differed between patients who did and did not develop MODS, several of the presurgery protein concentrations and gene expression levels were also significantly different. These findings suggest that patients who go on to develop MODS most likely differed in either their genetic predisposition or their pre-existing inflammatory state, and these differences were reflected in both plasma protein concentrations and blood leukocyte gene expression patterns before the surgical injury.

There is some precedent for this observation. Foulds et al. (22) reported that preoperative neutrophil expression of CD11b identified patients at risk of developing postsurgical complications following TAAA repair. Similarly, the activation state of NF-B in leukocytes from these patients preoperatively also predicted the incidence of MODS following TAAA repair (23). Taken together, these findings suggest that some subset of patients undergoing TAAA repair is at increased risk for developing MODS, and these patients may be identified by differences in their patterns of leukocyte activation, gene expression, and plasma protein concentrations.

By surveying 30 known cytokines simultaneously, we observed patterns of circulating concentrations of several of these proteins to be associated with an adverse clinical outcome. In the present study, circulating levels of MCP-1, IL-10, MIG, IL-6, IL-8, IP-10, and sCD23 were all significantly elevated in patients who progressed to MODS. Several of these observations are novel, and have not been reported previously in patients with SIRS and MODS. Although there have been a number of studies examining the prognostic ability of increased serum IL-6, IL-8, and IL-10 concentrations (7, 24, 25, 26), there has been only a limited exploration for many of these other analytes in acute inflammatory conditions. For example, although increased MCP-1 concentrations have been reported in patients with sepsis (27), this is the first documentation that serum MCP-1, IP-10, sCD23, and MIG concentrations are increased in patients at risk of developing MODS and death after surgically induced, ischemia-reperfusion injury. The sCD23 data are particularly interesting because the concentrations of this marker of B cell differentiation have only been reported in patients with cancer or chronic autoimmune or inflammatory diseases, such as type I diabetes (28). Similarly, increases in MIG and IP-10 concentration have not been previously reported in SIRS patients progressing to MODS. It should be noted, however, that these preliminary results from a small number of patients lack the statistical power to predict clinical behavior based on a single analyte or an individual time point. Furthermore, we have not attempted to correlate the genomic and proteomic data in any multivariate analysis because of the small sample sizes. It is interesting, however, that there was very little correlation between the changes in gene expression profiles from the whole blood leukocytes, and the corresponding protein concentrations in the plasma. Although we have focused on the genomic response by circulating leukocytes, our earlier studies have suggested that these cell populations are generally not responsible for the appearance of the immunoinflammatory cytokines in the blood (29). Rather, inflammatory cells in the splanchnic bed, and not peripheral blood leukocytes, appear to be the primary source of blood-derived TNF- and IL-6.

Based on these results, however, it is evident that there are broad patterns of leukocyte gene expression and protein production that change in patients undergoing surgical aortic reconstruction and associated visceral and/or lower torso ischemia-reperfusion. Furthermore, there are also distinct patterns of gene expression and protein production before and following the ischemic injury that are associated with the subsequent development of MODS. In general, a global view of the leukocyte gene expression and plasma protein responses is consistent with the earlier observation that patients who develop MODS following this surgical injury are generally manifesting an exaggerated innate immune and inflammatory response. This conclusion is confirmed by the increased plasma concentration of a number of cytokines involved in the innate immune and inflammatory response, as well as by the overrepresentation of gene ontologies associated with the inflammatory response. However, equally importantly, the genome-wide expression analyses suggest that the changes in leukocyte gene expression associated with the development of MODS in these patients are much more complex. An adverse outcome is associated with an integrated up-regulation of genes involved in the heat shock response, ubiquitin-proteosome-dependent protein degradation, and the increased expression of extracellular proteins (peptidases), as well as a sustained decrease in the expression of genes involved in the MHC class I response. These studies are of course preliminary, and future investigations will be required to confirm these studies in a prospective manner, and to determine whether the changes in gene expression are merely associative or causative in nature.

	Footnotes

 
¹ This work was supported in part by Grant R37 GM-40586 (to L.L.M.), awarded by the National Institute of General Medical Sciences. R.J.F. was supported in part by a training grant (T32, GM-08721) from the National Institute of General Medical Sciences in burn and trauma research.

² Address correspondence and reprint requests to Dr. Lyle L. Moldawer, Department of Surgery, University of Florida College of Medicine, P.O. Box 100286, Gainesville, FL 32610-0286.

³ Abbreviations used in this paper: TAAA, thoracoabdominal aortic aneurysm; FDR, false discovery rate; IP-10, IFN--inducible protein-10; MCP, monocyte chemoattractant protein; MIG, monokine induced by IFN-; MODS, multisystem organ dysfunction syndrome; p-B, Bonferroni correction for multiple testing; SAMS, significance analysis of microarrays; sCD23, soluble CD23; SIRS, systemic inflammatory response syndrome.

⁴ The on-line version of this article contains supplemental material.

Received for publication December 3, 2003. Accepted for publication March 12, 2004.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Rectenwald, J. E., T. S. Huber, T. D. Martin, C. K. Ozaki, M. Devidas, M. B. Welborn, J. M. Seeger. 2002. Functional outcome after thoracoabdominal aortic aneurysm repair. J. Vasc. Surg. 35:640.[Medline]
2. Derrow, A. E., J. M. Seeger, D. A. Dame, R. L. Carter, C. K. Ozaki, T. C. Flynn, T. S. Huber. 2001. The outcome in the United States after thoracoabdominal aortic aneurysm repair, renal artery bypass, and mesenteric revascularization. J. Vasc. Surg. 34:54.[Medline]
3. Oberholzer, A., C. Oberholzer, L. L. Moldawer. 2001. Sepsis syndromes: understanding the role of innate and acquired immunity. Shock 16:83.[Medline]
4. Moore, F. A.. 1999. The role of the gastrointestinal tract in postinjury multiple organ failure. Am. J. Surg. 178:449.[Medline]
5. Svensson, L. G., E. S. Crawford, K. R. Hess, J. S. Coselli, H. J. Safi. 1993. Variables predictive of outcome in 832 patients undergoing repairs of the descending thoracic aorta. Chest 104:1248.[Free Full Text]
6. Harward, T. R., M. B. Welborn III, T. D. Martin, T. C. Flynn, T. S. Huber, L. L. Moldawer, J. M. Seeger. 1996. Visceral ischemia and organ dysfunction after thoracoabdominal aortic aneurysm repair: a clinical and cost analysis. Ann. Surg. 223:729.[Medline]
7. Welborn, M. B., H. S. Oldenburg, P. J. Hess, T. S. Huber, T. D. Martin, J. A. Rauwerda, R. I. Wesdorp, N. J. Espat, E. M. Copeland III, L. L. Moldawer, et al 2000. The relationship between visceral ischemia, proinflammatory cytokines, and organ injury in patients undergoing thoracoabdominal aortic aneurysm repair. Crit. Care Med. 28:3191.[Medline]
8. Ramaswamy, S., K. N. Ross, E. S. Lander, T. R. A. Golub. 2003. Molecular signature of metastasis in primary solid tumors. Nat. Genet. 33:49.[Medline]
9. Pomeroy, S. L., P. Tamayo, M. Gaasenbeek, L. M. Sturla, M. Angelo, M. E. McLaughlin, J. Y. Kim, L. C. Goumnerova, P. M. Black, C. Lau, et al 2002. Prediction of central nervous system embryonal tumor outcome based on gene expression. Nature 415:436.[Medline]
10. Staunton, J. E., D. K. Slonim, H. A. Coller, P. Tamayo, M. J. Angelo, J. Park, U. Scherf, J. K. Lee, W. O. Reinhold, J. N. Weinstein, et al 2001. Chemosensitivity prediction by transcriptional profiling. Proc. Natl. Acad. Sci. USA 98:10787.[Abstract/Free Full Text]
11. Garber, M. E., O. G. Troyanskaya, K. Schluens, S. Petersen, Z. Thaesler, M. Pacyna-Gengelbach, M. van de Rijn, G. D. Rosen, C. M. Perou, R. I. Whyte, et al 2001. Diversity of gene expression in adenocarcinoma of the lung. Proc. Natl. Acad. Sci. USA 98:13784.[Abstract/Free Full Text]
12. Tannahill, C. L., K. Fukuzuka, T. Marum, Z. Abouhamze, S. L. MacKay, E. M. Copeland III, L. L. Moldawer. 1999. Discordant tumor necrosis factor- superfamily gene expression in bacterial peritonitis and endotoxemic shock. Surgery 126:349.[Medline]
13. Tusher, V. G., R. Tibshirani, G. Chu. 2001. Significance analysis of microarrays applied to the ionizing radiation response. Proc. Natl. Acad. Sci. USA 98:5116.[Abstract/Free Full Text]
14. Eisen, M. B., P. T. Spellman, P. O. Brown, D. Botstein. 1998. Cluster analysis and display of genome-wide expression patterns. Proc. Natl. Acad. Sci. USA 95:14863.[Abstract/Free Full Text]
15. Doniger, S. W., N. Salomonis, K. D. Dahlquist, K. Vranizan, S. C. Lawlor, B. R. Conklin. 2003. MAPPFinder: using Gene Ontology and GenMAPP to create a global gene-expression profile from microarray data. Genome Biol. 4:R7.[Medline]
16. Ashburner, M., C. A. Ball, J. A. Blake, D. Botstein, H. Butler, J. M. Cherry, A. P. Davis, K. Dolinski, S. S. Dwight, J. T. Eppig, et al 2000. Gene ontology: tool for the unification of biology: The Gene Ontology Consortium. Nat. Genet. 25:25.[Medline]
17. Zhang, J. H., T. D. Chung, K. R. A. Oldenburg. 1999. Simple statistical parameter for use in evaluation and validation of high throughput screening assays. J. Biomol. Screen. 4:67.[Abstract]
18. Bernard, G., G. Doig, L. Hudson. 1995. Quantification of organ failure for clinical trials and clinical practice. Am. J. Respir. Crit. Care Med. 151: (Abstr.).
19. Abraham, E.. 1999. Why immunomodulatory therapies have not worked in sepsis. Intensive Care Med. 25:556.[Medline]
20. Vincent, J. L.. 1998. Search for effective immunomodulating strategies against sepsis. Lancet 351:922.[Medline]
21. Subramanian, G., M. D. Adams, J. C. Venter, S. Broder. 2001. Implications of the human genome for understanding human biology and medicine. J. Am. Med. Assoc. 286:2296.[Abstract/Free Full Text]
22. Foulds, S., M. Mireskandari, P. Kalu, W. Jackson, N. J. Cheshire, A. O. Mansfield, M. Schachter. 1998. Visceral ischemia and neutrophil activation in sepsis and organ dysfunction. J. Surg. Res. 75:170.[Medline]
23. Foulds, S., C. Galustian, A. O. Mansfield, M. Schachter. 2001. Transcription factor NF B expression and postsurgical organ dysfunction. Ann. Surg. 233:70.[Medline]
24. Friedland, J. S., J. C. Porter, S. Daryanani, J. M. Bland, N. J. Screaton, M. J. Vesely, G. E. Griffin, E. D. Bennett, D. G. Remick. 1996. Plasma proinflammatory cytokine concentrations, Acute Physiology and Chronic Health Evaluation (APACHE) III scores and survival in patients in an intensive care unit. Crit. Care Med. 24:1775.[Medline]
25. Pruitt, J. H., M. B. Welborn, P. D. Edwards, T. R. Harward, J. W. Seeger, T. D. Martin, C. Smith, J. A. Kenney, R. I. Wesdorp, S. Meijer, et al 1996. Increased soluble interleukin-1 type II receptor concentrations in postoperative patients and in patients with sepsis syndrome. Blood 87:3282.[Abstract/Free Full Text]
26. Rogy, M. A., S. M. Coyle, H. S. Oldenburg, C. S. Rock, P. S. Barie, K. J. Van Zee, C. G. Smith, L. L. Moldawer, S. F. Lowry. 1994. Persistently elevated soluble tumor necrosis factor receptor and interleukin-1 receptor antagonist levels in critically ill patients. J. Am. Coll. Surg. 178:132.[Medline]
27. Bossink, A. W., L. Paemen, P. M. Jansen, C. E. Hack, L. G. Thijs, J. Van Damme. 1995. Plasma levels of the chemokines monocyte chemotactic proteins-1 and -2 are elevated in human sepsis. Blood 86:3841.[Abstract/Free Full Text]
28. Kretowski, A., M. Szelachowska, M. Pietruczuk, I. Kinalska. 1999. CD23 antigen expression on B lymphocytes and soluble CD23 levels in peripheral blood of high-risk type 1 diabetes subjects. Scand. J. Immunol. 49:78.[Medline]
29. Fong, Y., M. A. Marano, L. L. Moldawer, H. Wei, S. E. Calvano, J. S. Kenney, A. C. Allison, A. Cerami, G. T. Shires, S. F. Lowry. 1990. The acute splanchnic and peripheral tissue metabolic response to endotoxin in man. J. Clin. Invest. 85:1896.

This article has been cited by other articles:

				A. Pachot, M.-A. Cazalis, F. Venet, F. Turrel, C. Faudot, N. Voirin, J. Diasparra, N. Bourgoin, F. Poitevin, B. Mougin, et al. Decreased Expression of the Fractalkine Receptor CX3CR1 on Circulating Monocytes as New Feature of Sepsis-Induced Immunosuppression J. Immunol., May 1, 2008; 180(9): 6421 - 6429. [Abstract] [Full Text] [PDF]

				H. Kuivaniemi, C. D. Platsoucas, and M. D. Tilson III Aortic Aneurysms: An Immune Disease With a Strong Genetic Component Circulation, January 15, 2008; 117(2): 242 - 252. [Full Text] [PDF]

				L. L. Moldawer Opening the Window on Genome-wide Expression Analyses in Sepsis Am. J. Respir. Crit. Care Med., October 1, 2007; 176(7): 631 - 632. [Full Text] [PDF]

				A. S. Shanbhag, A. M. Kaufman, K. Hayata, and H. E. Rubash Assessing Osteolysis with Use of High-Throughput Protein Chips J. Bone Joint Surg. Am., May 1, 2007; 89(5): 1081 - 1089. [Abstract] [Full Text] [PDF]

				P. Matt, T. Carrel, M. White, I. Lefkovits, and J. Van Eyk Proteomics in cardiovascular surgery J. Thorac. Cardiovasc. Surg., January 1, 2007; 133(1): 210 - 214. [Abstract] [Full Text] [PDF]

				T. Liu, W.-J. Qian, M. A. Gritsenko, W. Xiao, L. L. Moldawer, A. Kaushal, M. E. Monroe, S. M. Varnum, R. J. Moore, S. O. Purvine, et al. High Dynamic Range Characterization of the Trauma Patient Plasma Proteome Mol. Cell. Proteomics, October 1, 2006; 5(10): 1899 - 1913. [Abstract] [Full Text] [PDF]

				R. Tabibiazar, R. A. Wagner, A. Deng, P. S. Tsao, and T. Quertermous Proteomic profiles of serum inflammatory markers accurately predict atherosclerosis in mice Physiol Genomics, April 13, 2006; 25(2): 194 - 202. [Abstract] [Full Text] [PDF]

				M. V. Podgoreanu and D. A. Schwinn New Paradigms in Cardiovascular Medicine: Emerging Technologies and Practices: Perioperative Genomics J. Am. Coll. Cardiol., December 6, 2005; 46(11): 1965 - 1977. [Abstract] [Full Text] [PDF]

				V. Tomic, S. Russwurm, E. Moller, R.A. Claus, M. Blaess, F. Brunkhorst, M. Bruegel, K. Bode, F. Bloos, J. Wippermann, et al. Transcriptomic and Proteomic Patterns of Systemic Inflammation in On-Pump and Off-Pump Coronary Artery Bypass Grafting Circulation, November 8, 2005; 112(19): 2912 - 2920. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Data Supplement Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Feezor, R. J. Articles by Seeger, J. M. Search for Related Content PubMed PubMed Citation Articles by Feezor, R. J. Articles by Seeger, J. M.