ABSTRACT: Acute lung rejection is a risk factor for chronic rejection, jeopardizing the long-term survival of lung transplant recipients. At present, acute rejection is diagnosed by transbronchial lung biopsies, which are invasive, expensive, and subject to significant sampling error. In this study, we sought to identify groups of genes whose collective expression in BAL cells best classifies acute rejection versus no-rejection. BAL samples were analyzed from 32 unique subjects whose concurrent histology showed acute rejection (n=14) or no rejection (n=18). Global BAL cell gene expression was measured using Affymetrix U133A microarrays. The nearest shrunken centroid method with 10-fold cross validation was used to define the classification model. 250 runs of the algorithm were performed to determine the range of misclassification error and the most influential genes in determining classifiers. The estimated overall misclassification rate was below 20%. Seven transcripts were present in every classifier and 52 transcripts were present in at least 70% of classifiers; these transcripts were notable for involvement with T-cell function, cytotoxic CD8 activity, and granulocyte degranulation. The proportions of both lymphocytes and neutrophils in BAL samples increased with increasing probability of acute rejection; this trend was more pronounced with neutrophils. We conclude that there is a prominent acute rejection-associated signature in BAL cells characterized by increased T-cell, CD8+ cytotoxic cell, and neutrophil gene expression; this is consistent with established mechanistic concepts of the acute rejection response. Experiment Overall Design: Lung transplant recipients surviving at least 30 days after transplantation at the University of Minnesota were eligible for enrollment in the study. The study was approved by the Universityâs Institutional Review Board, and all patients provided written informed consent. Subjects underwent scheduled surveillance (n = 22) and clinically indicated (decreased pulmonary function tests or new radiographic abnormalities) bronchoscopies (n = 10), in which 100-200 cc of sterile saline was instilled into a single anatomic location (right middle lobe or lingula), and recovered using gentle suction. 4-6 transbronchial biopsies (TBB) were obtained from the same subsegmental location as the BAL, and 4-6 additional TBB were obtained from the lower lobe of the same lung. After sending samples for clinical tests, approximately 50 ml of the recovered BAL effluent was immediately placed on ice. BAL cell counts and differentials were determined with a hemocytometer. Specimens from subjects with active bronchopulmonary infections (as determined by history, exam, chest radiograph, and the results of routine laboratory tests and cultures) were excluded from analysis. Experiment Overall Design: All biopsy specimens were graded according to standard International Society for Heart and Lung Transplantation criteria8. Vascular and airway cellular infiltrate were scored separately on A (vascular) and B (airway) scales, each score ranging from 0 to 48;23;24. For the purposes of this analysis, acute rejection was defined as a combined A+B score greater than or equal to 2; and no-rejection was defined as a combined A+B score less than 2. Experiment Overall Design: The BAL samples used in this study were selected according to strict criteria that were designed to maximize potential differences in gene expression between acute rejection and non-acute rejection BAL samples; and to minimize bias from confounding factors. Each subject (n=32) was represented by one BAL sample in this study. Control subjects and samples (n=14) were selected according to the following criteria: 1) A+B score was 0 or 1 for all biopsies collected from each subject during his/her post-transplant course; 2) at least three biopsies had been graded from each subject; 3) the subject had survived at least one year post-transplant; 4) the selected BAL sample was culture negative for pathogenic bacteria, fungi, and viruses; and 5) the selected sample was from a patient that had not developed BOS or was collected at least 6 months prior to the development of BOS. Rejection subjects and samples (n=18) were selected according to the following criteria: 1) the BAL sample was culture negative for pathogenic bacteria, fungi, and viruses; 2) for subjects with more than one acute rejection sample, we used the first acute rejection sample (A+B score > 1); 3) BOS grade was 0 at the time of BAL sampling (although some subjects subsequently developed BOS).
