393 The Lung Microbiome and Development of Bronchiolitis Obliterans Syndrome after Pediatric Lung Transplantation M. Sagar, 1 S.A. Crabtree, 1 E.B. Hollister, 2 J.K. Runge, 2 Y. Shang, 2 J.S. Heinle, 3 E.D. McKenzie, 3 G.B. Mallory, 1 R.A. Luna, 2 M.G. Schecter. 1 1 Pulmonology, Texas Children’s Hospital, Houston, TX; 2 Pathology, Immunology, Microbiome Center, Texas Children’s Hospital, Houston, TX; 3 Surgery, Cardiovascular, Texas Children’s Hospital, Houston, TX. Purpose: Bronchiolitis obliterans syndrome (BOS) is a major compli- cation following lung transplant. Infection is a known risk factor for the development of BOS. Microbiome analysis has an advantage over traditional culture as it detects uncultivable organisms. In addition, reduced bacterial diversity has been associated with increases in morbidity and mortality. We hypothesize that significant differences exist in the microbiome of post lung transplant patients with BOS vs. non-BOS. Methods and Materials: Ninety-four bronchoalveolar lavage specimens from 21 patients were collected during surveillance bronchoscopies throughout the first year post-transplant. Bacterial DNA was extracted, and the V3V5 variable regions of the 16SrRNA gene were amplified using universal barcoded primers. Subsequent 454 sequen- cing produced an average of 6,000 reads per sample with an average length of 500 bases. Resulting sequences were analyzed by the Genboree Microbiome Workbench. Results: In lung transplant recipients (LTRs) who developed BOS, Firmicutes decreased and Proteobacteria increased over time. Con- versely, the non-BOS group revealed a longitudinal increase in Firmicutes and a decreased in Proteobacteria. 33 genera were noted in the BOS compared to 49 genera in the non-BOS group, indicating an overall decreased bacterial diversity in the BOS group. Bacterial diversity was similar in both the groups at 2 weeks post transplant but steadily declined in the BOS group while it increased in the non-BOS group during the first year post transplant. Conclusions: Longitudinal genus-level variation clearly differed in the BOS vs. non-BOS groups. Moreover, BOS was associated with reduced bacterial diversity, which has been associated with an increase in morbidity and mortality in previous studies of the lung microbiome. Further studies are required to investigate if a change in the lung microbiome of post-transplant individuals is an indication of the future development of BOS. 394 Pediatric Thoracic Multiorgan Transplantation in the United States M.S. Khan, 1 C.D. Fraser III 1 D.L.S. Morales, 2 E.D. McKenzie, 1 I. Adachi, 1 M.G. Schecter, 3 G.B. Mallory, 3 J.S. Heinle. 1 1 Congenital Heart Surgery, Texas Children’s Hospital, Baylor College of Medicine, Houston, TX; 2 Pediatric Cardiothoracic Surgery, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH; 3 Pediatric Pulmonology, Texas Children’s Hospital, Baylor College of Medicine, Houston, TX. Purpose: Thoracic multiorgan transplantations (TMOTx) are complex procedures that when performed in adults have shown similar survival to single organ transplantation (TX). The overall number of TMOTx has been gradually increasing, however, the absolute number remains low compared to single organ TX. In the pediatric population, TMOTx constitute only a small fraction of the total pediatric TXs performed. A national database was used to review the overall experience with pediatric TMOTx in the United States. Methods and Materials: United Network of Organ Sharing database was queried through 2011 for pediatric (r18y) TMOTx. Of the 2102 TMOTx, 238 (12%) were pediatric TMOTx: 181 (76%) were heart-lung TX, while others included heart-kidney (32,13%), lung-liver (11,5%), heart-liver (10,4%), heart-lung-liver (3,1%) and lung-kidney (1,1%). TMOTx accounted for 3% (238/7945) of the total pediatric thoracic TX performed. Results: Analysis of outcomes shows that graft survival in combined heart-lung TX was similar to survival in lung-only TX (p¼0.127) but was worse than in heart-only TX (po0.0001). Graft survival for all TMOTx involving heart, except heart-lung, was similar to pediatric heart-only TX survival (Figure 1). Similarly, graft survival for TMOTx involving lung, except lung-liver (p¼0.05), was not different than lung- only survival (Figure 2). Conclusions: Thoracic multiorgan transplantation offers a promising option for pediatric patients with multiorgan failure, while combining abdominal organ transplant with heart or lung transplant does not portend a worse outcome. Although rare in the pediatric population, these procedures can be performed successfully with an expectation of similar graft survival to single organ transplantation. 395 Metabolites in Human Ex Vivo Lung Perfusates Are Differentially Affected in Brain Death (BDD) vs. Cardiac Death Donors (DCD) and Modified by Duration of Normothermic Perfusion M.K. Hsin, M. Cypel, R. Zamel, T. Machuca, M. Chen, M. Liu, S. Keshavjee. Latner Thoracic Surgical Research Laboratoy, University of Toronto, Toronto, ON, Canada. Purpose: Normothermic Ex Vivo Lung Perfusion (EVLP) has increased our clinical lung transplantation (LTx) activity by 20% by salvaging marginal lungs, with satisfactory clinical outcomes. Re- warming these lungs on EVLP from 41C to 371C normalizes metabolic activity, and studying the dynamic metabolomic profile may help us understand lung injury and predict the ultimate function of the transplanted lung. Metabolic profiling of EVLP perfusate may elucidate the metabolic changes of donor lungs on EVLP. Methods and Materials: Donor lungs from 28 BDD and 22 DCD donors were assessed on EVLP for 4 hours prior to LTx. For each lung, EVLP perfusate was sampled at 1 hr and 4 hr perfusion, and global metabolomic profiling was performed with Mass Spectrometry. We examined the results over time and with respect to BDD vs. DCD status. Results: Global metabolomics detected 312 metabolites. Statistical analysis with Random Forest (RF) classification showed a dramatic difference in metabolite profiles between EVLP 1hr and 4hr with a 96% accuracy in predicting whether a sample was taken at the start or end of EVLP perfusion. Interestingly, the metabolic profiles grouped by donor status were sufficiently different to be able to predict with RF whether a sample belongs to BDD or DCD donor, with an accuracy of 80% at 1hr and 78% at 4 hr EVLP. Conclusions: As we search for opportunities to personalize the diagnosis and treatment of the donor lung, it is evident that EVLP perfusate is Abstracts S151