Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Circulation. Author manuscript; available in PMC 2012 May 24.
Published in final edited form as:
PMCID: PMC3115694

Identification and Classification of Acute Cardiac Rejection by Intragraft Transcriptional Profiling

Cecile TJ Holweg, PhD,1 Luciano Potena, MD, PhD,1 Helen Luikart, RN,1 Tianwei Yu, PhD,5 Gerald J Berry, MD, PhD,3 John P Cooke, MD, PhD,1 Hannah A Valantine, MD, MRCP,1 and Edward S Mocarski, PhD4



Treatment of acute rejection (AR) in heart transplantation relies on histopathological grading of endomyocardial biopsies (EMB) according to International Society for Heart and Lung Transplantation (ISHLT) guidelines. Intragraft gene expression profiling may be a way to complement histological evaluation.

Methods and results

Transcriptional profiling was performed on 26 EMB and expression patterns were compared with the 1990 ISHLT AR grades. Importantly, transcriptional profiles from settings with an equivalent AR grade appeared the same. In addition, grade 0 profiles could not be distinguished from 1A and grade 3A profiles could not be distinguished from 3B. Comparing the AR groupings (0+1A, 1B, and 3A+3B), 0+1A showed more striking differences from 1B than from 3A+3B. When these findings were extrapolated to the 2005 revised guidelines, the combination of 1A and 1B into a single category (1R) appears to have brought together EMBs with different underlying processes that is not evident from histological evaluation. Grade 1B was associated with upregulated immune response genes, as one categorical distinction from grade 1A. Although, grade 1B was distinct from the clinically relevant AR grades 3A and 3B, all of these grades shared a small number of overlapping pathways consistent with common physiological underpinnings.


The gene expression similarities and differences identified here in different AR settings have the potential to revise the clinical perspective on acute graft rejection, pending the results of larger studies.

Keywords: heart, biopsy, transplantation, gene expression


Despite the success of heart transplantation as a therapy for end-stage heart failure, acute rejection (AR) continues to reduce long-term survival in transplant recipients. AR develops as a consequence of the recognition of non-self histocompatibility antigens and elaboration of an immune response against allogeneic heart muscle that leads to compromised cardiac function and graft loss. The histological evaluation of right ventricular endomyocardial biopsies (EMB) is part of the current standard of care is based on the International Society for Heart and Lung Transplantation (ISHLT) guidelines1, as revised in 20052. The greatest incidence of AR occurs in the first year post transplant3. Patients who experience AR grades 3A, 3B and 4 (revised grade 2R or 3R) within this time frame exhibit poor five-year survival as well as an increased risk of cardiac allograft vasculopathy4, 5 despite intervention. Several studies have drawn attention to discordance between histological grade and graft function, particularly in settings where declining cardiac function and increasing mortality proceed independently of histological grade6, 7. These observations point to the need for greater understanding of underlying processes that may unveil alternative or complementary approaches for AR surveillance with the goal of accurate and early diagnosis. An accumulating body of evidence suggests that AR surveillance based on gene expression profiling in EMB or peripheral blood cells may help improve clinical management.8-11

Intragraft gene expression studies have provided insight in the process of AR12-19, although recent global intragraft gene expression profiling studies have not revealed a correlation with 2005 ISHLT AR grades 11. Direct assessment of transcriptional profiles have proven valuable as an adjunct to tumor grading20 as well as in kidney21-23, lung24 and liver25 transplant settings. In cardiac transplantation, such microarray analysis has suggested biomarkers for diagnosis of rejection and to distinguish AR from infection26.

We specifically investigated the relationship between tissue wide gene expression patterns and 1990 ISHLT AR grades to address key issues including whether (i) intragraft gene expression patterns, grouped by 1990 ISHLT grade, are consistent from patient to patient; (ii) particular EMB-derived transcriptional profiles typify particular grades considering either the 1990 or 2005 guidelines; and (iii) expression patterns provide insights into pathways or mechanisms contributing to AR.


Patients and biopsies

Samples were selected from a cohort of 83 patients, enrolled from 109 consecutive cardiac transplant recipients at Stanford between January 2002 and May 2004, provided written informed consent approved by the institutional review board (IRB). The baseline immunosuppressive regimen consisted of induction with an interleukin-2 receptor antagonist (daclizumab), and a maintenance regimen of Prednisone; Cyclosporine A or Tacrolimus; and Mycophenolate Mofetil (MMF) or Sirolimus, as previously described27. All patients underwent routine AR surveillance for which 4 to 6 endomyocardial biopsies were obtained from the right side of the interventricular septum. One, collected in RNAlater™ or OCT (snap frozen in liquid nitrogen and stored at -140°C), was used for microarray experiments, and the remainder were formalin-fixed, paraffin-embedded, sectioned and histologically graded for AR by a cardiac pathologist according to the 1990 ISHLT criteria1. The most severe grading observed across multiple sections determined the AR grade. Because ISHLT guidelines were revised during the study course2, both the old and revised grades are shown in Table 1. Patients receiving a score of ≥3A were treated with high dose corticosteroids and/or anti-thymocyte-globulin.

Table 1
Original scoring of EMB for diagnosis of AR.

In an effort to avoid confounding factors, we employed the following stringent criteria to select EMB for inclusion in this study: 1) absence of HLA antibodies in patients prior to transplantation, 2) the EMB collected immediately prior to the EMB subjected to assay was free of signs of rejection (i.e. grade 0), 3) absence of treatment for infections at time of and in the weeks before the EMB collection, 4) absence of hemodynamic compromise at the time of EMB collection, 5) RNA of sufficient quality and quantity for use in microarray analysis. This resulted in a cohort of 23 patients providing 26 EMB (grade 0, n=9; 1A, n=5; 1B, n=5; 3A, n=4; 3B, n=3). Two patients provided more than one biopsy, each from a different grade and separated by a period that was free of rejection. As a result, only one biopsy per patient was included in comparative analyses (supplemental Table 1-3).

RNA isolation, amplification and hybridization

For total RNA extraction, EMB were homogenized in 500 μl TRIzol®. Following chloroform extraction, an RNA RNeasy micro kit (Qiagen, Valencia, CA, USA) was used on the aqueous phase. mRNA was amplified and labeled with Cy5 using the Low RNA Input Fluorescent Linear Amplification Kit (Agilent Technologies, Palo Alto, USA). The samples were mixed with equal amounts of amplified Cy-3 labeled human reference RNA (Stratagene, nr 740000, La Jolla, CA, USA), and hybridized onto 22K oligomicroarrays (H1A-v2, Agilent Technologies) using the In Situ Hybridization Kit Plus (Agilent Technologies). All kits are used according to the manufacturer's instructions.

Data and gene filtering and statistical analysis

Data were collected using an Agilent Microarray Scanner, (model G2565AA) and processed, including Lowess normalization, using Agilent Feature Extraction Software (Agilent G2566AA), version A.6.1.1. Spot and gene filtering was performed using the standard settings for Agilent microarrays using the Stanford Microarray Database (, averaging values for genes represented by multiple spots on the arrays. Data from non-uniform features; spots with intensities below background (defined by the feature extraction software); and outliers were excluded from the analysis. Arrays with 80% remaining data were used for analysis. Genes were median centered and Significance Analysis of Microarrays (SAM)28 multi-class and two-class unpaired analyses were performed. Missing log2 based values were calculated using the K-nearest neighbor algorithm set to 10. A false discovery rate (FDR) <1% was used for both the multi-class and two-class analysis. Principal component analysis using the genes of the multi-class comparison was performed to display data set relatedness. Two-way hierarchical clustering of statistical significant genes was employed using the average linkage clustering algorithm in Cluster29 and visualized in Java-Treeview30. The gene expression data is available through, accession number GSE9377 and through

Patient characteristics were compared using the Fisher exact for discrete and Mann-Whitney test for continuous variables.

Ingenuity Pathway Analysis (IPA, version 8.8) Software (Mountain View, CA, USA) was used to compute a molecular network based on published direct and indirect physical and functional interactions. Differentially expressed genes were entered into the Ingenuity Pathways Knowledge Base and those recognized were further evaluated using their proprietary software package.


Gene expression patterns in relation to AR grading

We evaluated 26 EMB from 23 heart transplant recipients collected when signs of rejection or infection had been absent during the weeks preceding collection. After filtering the primary data, a set of 16,575 genes remained to explore relationships between expression patterns and histological AR grade. Multi-class SAM analysis followed by principal component analysis was performed to estimate relatedness of transcriptional profiles and AR grade. The first two principal components generally revealed three groupings (Figure 1). ISHLT grade 1B separated as a single group, with the other grades distributing into two additional groupings, 0+1A and 3A+3B. To further evaluate relationships and consistency of gene expression patterns across AR grades, we performed pairwise statistical comparisons. We found no significant differences between grade 0 and 1A biopsies, however, we identified differentially expressed genes when comparing either grade 0 or grade 1A to grade 1B EMB (Figure 1A-C). Only a few genes were found to be differentially expressed comparing grade 1B to either 3A or 3B (Figure 2D and E). Grade 3A and 3B samples from different patients were indistinguishable (Figure 2F). In addition, when a grade 3B sample collected at a different time from patient S (table 1) was included in the analysis, these AR groups remained indistinguishable (data not shown). These observations lead us to focus on comparing three groupings: 0+1A, 1B, and 3A+3B. There were no significant differences between groups (0+1A; 1B, 3A+3B) with respect to donor and recipient age and gender, HLA mismatch, ischemic time, and duration post-transplant were found (Table 2).

Figure 1
Principal component analysis scatterplot of SAM multi-class significant genes. The first and second principal components (PC) were used to depict relationships between samples. For patients with multiple samples of different ISHLT grades, only one sample ...
Figure 2
SAM plots of the comparisons between grades of AR and of the three AR groups 0+1A, 1B and 3A+3B based on similarity in gene expression profiles. A: grade 0 vs. 1A, B: grade 0 vs. 1B, C: grade 1A vs. 1B, D: grade 1B vs.3A, E: grade 1B vs. 3B, F: grade ...
Table 2
Patient characteristics

Most pronounced differences in transcriptional profiles were observed when comparing grouping 0+1A versus grade 1B, with 458 statistically different genes, of which 343 were upregulated and 115 downregulated in 1B. Comparing groupings 0+1A and 3A+3B yielded 205 different genes of which 136 were upregulated and 69 were downregulated in 3A+3B. Comparing grade 1B with 3A+3B revealed only 11 differences, all of which indicated downregulation in 3A+3B relative to 1B. (Figure 2G-I). Supplemental Tables 4 through 8 provide lists of significant different genes for each comparison. Overall, the intragraft gene expression patterns agreed with grading by histological assessment using the 1990 ISHLT system.

To visualize relationships between expression profiles and AR grades, we performed two-way hierarchical cluster analysis with each set of differentially expressed genes. This distinguished and revealed striking consistency in three distinct AR groupings (Figure 3). Twenty three out of 26 (88%) EMB profiles clustered according to the histological AR grade. Every grade 1B (5/5) and 13 out of 14 grade 0+1A EMB clustered together when the profiles of grade 0+1A versus 1B were subjected to this analysis. The outlier 0+1A EMB (#6) clustered with grade 1B (Figure 3A). Although the number of genes from the grade 1B versus 3A+3B comparison was small, a cluster of the grade 1B versus 3A+3B revealed that grade 1B expression patterns were highly related and independent of 3A+3B. All but one grade 3A+3B EMB profile clustered together (Figure 3B). Using the genes from the comparison 0+1A vs. 3A+3B, all grade 0+1A EMB clustered, and five out of seven 3A+3B EMB clustered together. The two grade 3A+3B samples that did not cluster with their corresponding grade, seemed to form a separate cluster (Figure 3C), but further sample collection and analysis will be needed to determine any relevance of this small cluster.

Figure 3
Heatmap of two-way hierarchical clustering with differentially expressed genes. A: grade 0+1A vs. 1B, B: grade 1B vs. 3A+3B, C: grade 0+1A vs. 3A+3B. Columns illustrate EMB's and rows represent genes. Green cells indicate transcript levels higher and ...

The histological grades of the non-clustering EMB were re-reviewed by the pathologist (GB) without any revision of the initial histological grade. Intriguingly, the apparent relationship between transcriptional profiles and histological AR grade was not influenced by variable section-to-section histology (Table 1), which raises a tantalizing possibility that transcriptional profiling may provide a more uniform, less subjective read-out than histological evaluation.

Functional pathway analysis according to AR grading

To gain insights in the biological basis for differences in gene expression, we used Ingenuity Pathway Analysis (IPA) Software. To identify functions associated with each of the AR grade groupings, functional analysis was performed using 434 out of 458 genes of the 0+1A versus 1B comparison for which biological data was available. The top five significant biofunction categories of this analysis were inflammatory response; DNA replication, recombination and repair; nucleic acid metabolism; small molecule biochemistry; hematological disease. A similar analysis on 197 out of 205 significant genes from the 0+1A versus 3A+3B comparison for which biological information was available yielded top biofunction categories of inflammatory response; cellular development; hematological system development and function; hematopoiesis; cellular function and development. The 1B versus 3A+3B comparison contained too few genes to conduct a functional analysis.

To identify pathways that characterize the most dramatic differences, we used IPA's canonical pathways tool on the genes from the 0+1A versus 1B comparison. The top pathway identified was the antigen presentation pathway. Seven genes (B2M, CIITA, HLA-A, HLA-B, HLA-C, HLA-E, TAP1) involved in classical and non-classical MHC class I antigen presentation were upregulated in grade 1B. The second most significant pathway was related to antigen presentation, with allograft rejection pathway genes CD40 and granzyme B upregulated in addition to major histocompatability complex (MHC) genes involved in antigen presentation.

Antigen presentation also predominated in the 0+1A and 3A+3B comparison, with upregulation of six genes (CD74, HLA-C, HLA-DPA1, HLA-DPB1, HLA-DRB5, TAP1), representing both MHC class I and MHC class II. This differed from 1B rejections where only MHC class I seems to be activated. Two other significant pathways in the 3A+3B grouping were Natural Killer Cell Signaling and Dendritic Cell Maturation, although neither one of these was significant in the grade 1B group. Despite the clinical relevance of the histological score, the allograft rejection pathway did not reach statistical significant in 3A+3B profiles. Besides the MHC genes, no other genes in this pathway seem to have changed in 3A+3B grouping according to the IPA knowledge base. Thus, although some similar processes and pathways may be present, a predominance of distinct processes characterize grade 1B as compared to clinically relevant grades 3A or 3B AR. A summary of the IPA output for the comparisons 0+1A vs. 1B and 0+1A vs. 3A+3B is provided as supplemental data.


Current methods of AR surveillance rely on histological evaluation of EMB, an area that would benefit from approaches requiring fewer EMB and providing more consistent sample-to-sample information. To explore a possible alternative approach for diagnosing AR using only a single EMB, this small study collected tissue-wide intragraft transcriptional profiles that generally correlated with histological AR grade based on the 1990 ISHLT guidelines. We observed remarkable sample-to-sample consistency as highlighted by the statistical evaluation that grouped patients with similar grades of AR. The most surprising transcriptional differences were observed in grade 1B, compared to less severe 0+1A. Despite the obvious need to apply these methods to larger sample sets, the genes associated with the different AR grades in this small study suggest that common tissue-wide pathways predominate in similar histological 1990 ISHLT AR grades and may require fewer EMB to assign grades.

Our observations raise important questions as to whether the 19901 or revised 2005 ISHLT guidelines2 will ultimately bear a closer relationship to tissue-wide transcription profiles and provide the most clinically useful information. Whereas the 1990 grading system accommodated the groupings that resulted from our small set of gene expression profiles, the 2005 guidelines collapsed grades that showed distinct patterns and separated grades that showed similar profiles. Using an Affymetrix microarray approach, Mengel did not detect any correlation between gene expression profiles and histological classification when applying the 2005 ISHLT grading system. Although, they did not attempt to correlate their expression profiles with the 1990 AR grades, their results might have been influenced by the merger of rejection grades 1A and 1B in the revised guidelines, thereby averaging out the most striking gene expression differences observed here.11

Widespread application of the 1990 guidelines led to (i) inconsistencies in grading among transplant centers and (ii) resulted in the recognition that grades 1, 2 and some forms of 3A rarely (< 25%) progress; whereas, grade 3B progresses and 4 are associated with allograft failure. This provided an impetus to revise the grading system such that grades 1A, 1B and 2 were combined into a single grade, 1R; 3A became 2R; and 3B together with 4 became 3R. Our data reveals strikingly different expression patterns between grade 1B compared to grades 0 and 1A, suggesting that grade 1B represents a distinct tissue state. The clinically innocence of grade 1B remains suspect among physicians, leading some transplant centers to modulate immunosuppression when diagnosing 1B31, 32. Combining 1A and 1B into 1R certainly undervalues any differences. Should the clinical relevance of grade 1B increase and larger transcriptional profiling studies confirm our results, the criteria employed for revised AR scoring might need to be revisited.

Grades 3A and 3B share a characteristic myocyte damage profile but differ histologically, a multifocal inflammatory infiltration pattern characterizing grade 3A and a diffuse pattern characterizing 3B (Table 1). Grade 3B exhibits more dramatic edema, hemorrhage or vasculitis than 3A. However, gene expression patterns for grades 3A and 3B were indistinguishable. Although profiling reveals tissue-wide gene expression patterns, rigorous statistical evaluation tends to eliminate variable inflammatory markers from the analysis1, 2. Our failure to distinguish grade 3A from 3B by profiling could have resulted from insufficient sampling, individual variability or dilution of relevant mRNA when assaying total tissue samples. The infiltrates in the grade 3B EMB collected for evaluation in this study were diffuse, not widespread (Table 1). Severe AR may have become less common due to improved immunosuppressive therapy but will continue to benefit from histological evaluation. The fact that grades 3A (2R) and 3B (3R) were not distinguished by profiling a single EMB suggests that this these more clinically relevant AR grades benefit from the direct histological review of multiple EMBs.

Overall, it was striking that we observed such a high level of consistency in tissue-wide transcriptional patterns based on statistical analysis and clustering of EMB into three groups, despite variability in histological grading of EMB (Table 1). Microarray profiling may smooth variability by sampling a greater percentage of tissue or by detecting qualities distinct from those observed under the microscope. While inter-reader, subjective variability in histological grading of rejections is recognized33, the small number of significant genes between grades 0+1A and 3A+3B groupings may have resulted from the use of a single EMB for microarray. The most severe reading drives diagnosis even though this may be observed in fewer than half of evaluated EMB. Further evaluation employing equivalent numbers of biopsies for both approaches, possibly by splitting each biopsy in two would be necessary to address this issue.

Although, pathway analysis and interpretation is difficult when based solely on mRNA expression, as transcript levels are not always proportional to protein levels, we believe that changes in the expression of multiple genes in a pathway may implicate the pathway in the studied disease and may provide new insights into underlying disease. An interesting observation is the fact that in 1B AR, inflammatory processes are ongoing, similar to 3A and 3B AR. However, nucleic acid metabolism together with DNA repair and replication seem to be more prevalent in 1B rejections, suggesting that both proliferation and repair processes are ongoing. Although MMF may cause these changes due to effects on purine and pyrimidine synthesis34, 35, the standard dose of MMF used here in all patients reduces the possibility that gene expression differences result from MMF treatment differences.

The most prominently changed pathway in both 1B and 3A+3B AR was antigen presentation. Interestingly, all identified HLA molecules in grade 1B belonged to MHC class I, while in grade 3A and grade 3B rejections, genes implicated in MHC class II were also upregulated. The absence of MHC class II in the 1B AR might result from the use of HMG-CoA reductase inhibitors that can have an effect on MHC class II upregulation. Alternatively, grade 1B AR patterns may reflect antigen presentation to cytotoxic T-cells, while 3A and 3B rejections may broader immune activity involving CD4+ and CD8+ T-lymphocytes.

In addition to the recognized limitations of this study, particularly the small sample size and use of a single EMB for profiling, histological review was by a single, albeit experienced pathologist. Although the samples were re-evaluated without any changes, we were unable to address potential reader-to-reader variability. This is a particular concern with grades 2, 3A and 3B, where grade 2 had the lowest concordance rate among pathologists. For this reason we excluded grade 2, which has been ascribed to extend from quilty lesions and recategorizes often to 1A/1B or 3A/3B33, from this study. Overall, we cannot estimate the extent to which quilty lesions contribute to the scores in our study.

Our reliance on a single EMB sample for microarray analysis means that sampling error and variability cannot be addressed directly. The stringent EMB selection criteria employed to avoid confounding factors means that the number of biopsies we analyzed was small. However, one of the most important of our findings was that histologically similar AR gradings grouped together, suggesting that the expression pattern in the single sample was influenced by the factors that drive histological grading. Thus, intragraft gene expression may turn out to be uniform and valuable as a way to determine physiological status of the transplanted heart.

We did not collect peripheral blood for comparison to EMB by expression profiling. Comparing intragraft and peripheral blood gene expression patterns could have provided potentially complementary information. The predominant reason for not initially planning to profile this compartment was the fact that changes in leukocyte subpopulations are a dominant contributor to gene expression differences that are observed, thereby limiting utility. Transplant patients exhibit swings in CBC that would certainly complicate evaluation of whole blood expression profiles. Nevertheless, peripheral blood profiling for AR has been used to discriminate between rejection grades 0 and ≥3A and more importantly, these studies have also suggested that grade 1B should be viewed as a distinct subset8, 36.

Despite the limitations of our study, our data reveals gene expression patterns that are consistent with the histologically most severe grade in multiple sections of biopsies. This suggests that transcriptional profiling could complement histopathology in guiding treatment strategies. Additional larger studies with replicate and complementary tissues are needed to fully elucidate the gene expression signatures for each grade of rejection. Our observations provide a basis for further study of correlation with clinical outcome and novel directions for more detailed investigation of mechanisms underlying AR.

Clinical perspective

Surveillance for acute cellular rejection (ACR) after heart transplantation relies on histological evaluation of serial endomyocardial biopsies (EMB), an approach limited by sampling error and inter-observer variations in the interpretation, even when performed by the most experienced pathologist. This clinical area would benefit from technologies requiring fewer biopsies and more consistent sample-to-sample information. We have used intragraft transcriptional profiling to explore a possible alternative approach for diagnosing acute rejection. Using only one EMB, our results show a consistent expression pattern between similar grades of acute rejection and distinctly different expression patterns between different rejection grades that was not influenced by variable section-to-section histology. Furthermore, the expression pattern distinguished Grade 1B from all others, and was thus more in accordance with the original ISHLT grading scale (1990) than with the revised rejection grading system (2005 ISHLT guidelines) that combines several rejection grades. These observations raise important clinical implications for ACR surveillance: 1) transcriptional profiling may provide a more uniform read-out than histological evaluation, with less subjective section-to-section variability. 2) If confirmed in a larger sample set and a conclusive transcriptome identified for each rejection grade, transcriptional profiling can be an additional and complementary tool for diagnosing acute rejection, which can ultimately change the clinical practice to collecting fewer biopsies to diagnose rejection. 3) The consistent expression pattern for Grade 1B suggests that the ISHLT grading system should be re-evaluated to limit the combining of tissues with different underlying processes, taking into account new insights from advanced technologies.

Supplementary Material


Funding Sources: Primary funding source was program grant PO1 HL079355 from the NIH, with funding from RO1 AI020211 (to ESM) used to support data analysis.


Disclosures: The authors have no conflicts of interest to disclose.


1. Billingham ME, Cary NR, Hammond ME, Kemnitz J, Marboe C, McCallister HA, Snovar DC, Winters GL, Zerbe A. A working formulation for the standardization of nomenclature in the diagnosis of heart and lung rejection: Heart Rejection Study Group. The International Society for Heart Transplantation. J Heart Transplant. 1990;9:587–593. [PubMed]
2. Stewart S, Winters GL, Fishbein MC, Tazelaar HD, Kobashigawa J, Abrams J, Andersen CB, Angelini A, Berry GJ, Burke MM, Demetris AJ, Hammond E, Itescu S, Marboe CC, McManus B, Reed EF, Reinsmoen NL, Rodriguez ER, Rose AG, Rose M, Suciu-Focia N, Zeevi A, Billingham ME. Revision of the 1990 working formulation for the standardization of nomenclature in the diagnosis of heart rejection. J Heart Lung Transplant. 2005;24:1710–1720. [PubMed]
3. Klingenberg R, Gleissner C, Koch A, Schnabel PA, Sack FU, Zimmermann R, Katus HA, Dengler TJ. Impact of pre-operative diabetes mellitus upon early and late survival after heart transplantation: a possible era effect. J Heart Lung Transplant. 2005;24:1239–1246. [PubMed]
4. Taylor DO, Edwards LB, Boucek MM, Trulock EP, Waltz DA, Keck BM, Hertz MI. Registry of the International Society for Heart and Lung Transplantation: twenty-third official adult heart transplantation report-2006. J Heart Lung Transplant. 2006;25:869–879. [PubMed]
5. Valantine H. Cardiac allograft vasculopathy after heart transplantation: risk factors and management. J Heart Lung Transplant. 2004;23:S187–193. [PubMed]
6. Mills RM, Naftel DC, Kirklin JK, Van Bakel AB, Jaski BE, Massin EK, Eisen HJ, Lee FA, Fishbein DP, Bourge RC. Heart transplant rejection with hemodynamic compromise: a multiinstitutional study of the role of endomyocardial cellular infiltrate. Cardiac Transplant Research Database. J Heart Lung Transplant. 1997;16:813–821. [PubMed]
7. Bhalodolia R, Cortese C, Graham M, Hauptman PJ. Fulminant acute cellular rejection with negative findings on endomyocardial biopsy. J Heart Lung Transplant. 2006;25:989–992. [PubMed]
8. Deng MC, Eisen HJ, Mehra MR, Billingham M, Marboe CC, Berry G, Kobashigawa J, Johnson FL, Starling RC, Murali S, Pauly DF, Baron H, Wohlgemuth JG, Woodward RN, Klingler TM, Walther D, Lal PG, Rosenberg S, Hunt S. Noninvasive discrimination of rejection in cardiac allograft recipients using gene expression profiling. Am J Transplant. 2006;6:150–160. [PubMed]
9. Pham MX, Deng MC, Kfoury AG, Teuteberg JJ, Starling RC, Valantine H. Molecular testing for long-term rejection surveillance in heart transplant recipients: design of the Invasive Monitoring Attenuation Through Gene Expression (IMAGE) trial. J Heart Lung Transplant. 2007;26:808–814. [PubMed]
10. Pham MX, Teuteberg JJ, Kfoury AG, Starling RC, Deng MC, Cappola TP, Kao A, Anderson AS, Cotts WG, Ewald GA, Baran DA, Bogaev RC, Elashoff B, Baron H, Yee J, Valantine HA. Gene-expression profiling for rejection surveillance after cardiac transplantation. The New England journal of medicine. 2010;362:1890–1900. [PubMed]
11. Mengel M, Sis B, Kim D, Chang J, Famulski KS, Hidalgo LG, Einecke G, de Freitas DG, Tymchak W, Burton J, Halloran PF. The molecular phenotype of heart transplant biopsies: relationship to histopathological and clinical variables. Am J Transplant. 2010;10:2105–2115. [PubMed]
12. Alvarez CM, Fernandez D, Builes M, Zabaleta J, Restrepo LM, Villegas A, Garcia LF. Intragraft cytokine expression in heart transplants with mild or no histological rejection. Clin Transplant. 2001;15:228–235. [PubMed]
13. Baan CC, Holweg CH, van Gelder T, Knoop CJ, Niesters HG, Zondervan P, Mochtar B, Balk AH, Weimar W. Redundancy of the cytokine network in the development of rejection after clinical heart transplantation. Transpl Int. 1998;11(Suppl 1):S512–514. [PubMed]
14. Baan CC, Niesters HG, Balk AH, Mochtar B, Zondervan PE, Weimar W. The intragraft cytokine mRNA pattern reflects the efficacy of steroid antirejection therapy. J Heart Lung Transplant. 1996;15:1184–1193. [PubMed]
15. de Groot-Kruseman HA, Mol WM, Niesters HG, Maat AP, van Gelder T, Balk AH, Weimar W, Baan CC. Differential intragraft cytokine messenger RNA profiles during rejection and repair of clinical heart transplants. A longitudinal study. Transpl Int. 2003;16:9–14. [PubMed]
16. Fahmy NM, Yamani MH, Starling RC, Ratliff NB, Young JB, McCarthy PM, Feng J, Novick AC, Fairchild RL. Chemokine and chemokine receptor gene expression indicates acute rejection of human cardiac transplants. Transplantation. 2003;75:72–78. [PubMed]
17. Perez EC, Shulzhenko N, Morgun A, Diniz RV, Almeida DR, Musatti CC, Gerbase-DeLima M. Expression of Fas, FasL, and soluble Fas mRNA in endomyocardial biopsies of human cardiac allografts. Hum Immunol. 2006;67:22–26. [PubMed]
18. Souza AI, Felkin LE, McCormack AM, Holder A, Barton PJ, Banner NR, Rose ML. Sequential expression of three known protective genes in cardiac biopsies after transplantation. Transplantation. 2005;79:584–590. [PubMed]
19. Van Hoffen E, Van Wichen D, Stuij I, De Jonge N, Klopping C, Lahpor J, Van Den Tweel J, Gmelig-Meyling F, De Weger R. In situ expression of cytokines in human heart allografts. Am J Pathol. 1996;149:1991–2003. [PubMed]
20. Cole KA, Krizman DB, Emmert-Buck MR. The genetics of cancer--a 3D model. Nat Genet. 1999;21:38–41. [PubMed]
21. Sarwal M, Chua MS, Kambham N, Hsieh SC, Satterwhite T, Masek M, Salvatierra O., Jr Molecular heterogeneity in acute renal allograft rejection identified by DNA microarray profiling. The New England journal of medicine. 2003;349:125–138. [PubMed]
22. Scherer A, Krause A, Walker JR, Korn A, Niese D, Raulf F. Early prognosis of the development of renal chronic allograft rejection by gene expression profiling of human protocol biopsies. Transplantation. 2003;75:1323–1330. [PubMed]
23. Hauser P, Schwarz C, Mitterbauer C, Regele HM, Muhlbacher F, Mayer G, Perco P, Mayer B, Meyer TW, Oberbauer R. Genome-wide gene-expression patterns of donor kidney biopsies distinguish primary allograft function. Lab Invest. 2004;84:353–361. [PubMed]
24. Gimino VJ, Lande JD, Berryman TR, King RA, Hertz MI. Gene expression profiling of bronchoalveolar lavage cells in acute lung rejection. Am J Respir Crit Care Med. 2003;168:1237–1242. [PubMed]
25. Borozan I, Chen L, Sun J, Tannis LL, Guindi M, Rotstein OD, Heathcote J, Edwards AM, Grant D, McGilvray ID. Gene expression profiling of acute liver stress during living donor liver transplantation. Am J Transplant. 2006;6:806–824. [PubMed]
26. Morgun A, Shulzhenko N, Perez-Diez A, Diniz RV, Sanson GF, Almeida DR, Matzinger P, Gerbase-DeLima M. Molecular profiling improves diagnoses of rejection and infection in transplanted organs. Circ Res. 2006;98:e74–83. [PubMed]
27. Potena L, Holweg CT, Chin C, Luikart H, Weisshaar D, Narasimhan B, Fearon WF, Lewis DB, Cooke JP, Mocarski ES, Valantine HA. Acute rejection and cardiac allograft vascular disease is reduced by suppression of subclinical cytomegalovirus infection. Transplantation. 2006;82:398–405. [PubMed]
28. Tusher VG, Tibshirani R, Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci U S A. 2001;98:5116–5121. [PubMed]
29. Eisen MB, Spellman PT, Brown PO, Botstein D. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A. 1998;95:14863–14868. [PubMed]
30. Saldanha AJ. Java Treeview--extensible visualization of microarray data. Bioinformatics. 2004;20:3246–3248. [PubMed]
31. Hornick P, Rose M. Chronic rejection in the heart. Methods Mol Biol. 2006;333:131–144. [PubMed]
32. Frigerio M, Bonacina E, Gronda E, Andreuzzi B, Anjos MC, De Vita C, Mangiavacchi M, Masciocco G, Oliva F, Pellegrini A. A semiquantitative approach to the evaluation of acute cardiac allograft rejection at endomyocardial biopsy. J Heart Lung Transplant. 1997;16:1087–1098. [PubMed]
33. Marboe CC, Billingham M, Eisen H, Deng MC, Baron H, Mehra M, Hunt S, Wohlgemuth J, Mahmood I, Prentice J, Berry G. Nodular endocardial infiltrates (Quilty lesions) cause significant variability in diagnosis of ISHLT Grade 2 and 3A rejection in cardiac allograft recipients. J Heart Lung Transplant. 2005;24:S219–226. [PubMed]
34. Devyatko E, Zuckermann A, Bohdjalian A, Roedler S, Dunkler D, Grimm M, Weigel G. Activation of the purine salvage pathway in mononuclear cells of cardiac recipients treated with mycophenolate mofetil. Transplantation. 2006;82:113–118. [PubMed]
35. Qiu Y, Fairbanks LD, Ruckermann K, Hawrlowicz CM, Richards DF, Kirschbaum B, Simmonds HA. Mycophenolic acid-induced GTP depletion also affects ATP and pyrimidine synthesis in mitogen-stimulated primary human T-lymphocytes. Transplantation. 2000;69:890–897. [PubMed]
36. Bernstein D, Williams GE, Eisen H, Mital S, Wohlgemuth JG, Klingler TM, Fang KC, Deng MC, Kobashigawa J. Gene expression profiling distinguishes a molecular signature for grade 1B mild acute cellular rejection in cardiac allograft recipients. J Heart Lung Transplant. 2007;26:1270–1280. [PubMed]