S402 The Journal of Heart and Lung Transplantation, Vol 37, No 4S, April 2018 improve musculoskeletal conditioning and exercise efficiency during the pre-transplant period, which may impact post-transplant outcomes. ( 1035) Description of Longitudinal Measurement of Donor Fraction of Cell- Free DNA Following Rejection Treatment and Correlation to Clinical Outcomes W. Ragalie , 1 S. Kindell, 2 S. Zangwill, 3 M.E. Mitchell. 4 1 Surgery, Medical College of Wisconsin, Milwaukee, WI; 2 Pediatrics, Medical College of Wisconsin, Milwaukee, WI; 3 Pediatrics, Pheonix Children's Hospital, Pheonix, AZ; 4 Pediatric Cardiothoracic Surgery, Children's Hospital of Wisconsin, Medical College of Wisconsin, Milwaukee, WI. Purpose: Donor Fraction (DF) of cell-free DNA in transplant recipients has been correlated with rejection and allograft injury. Treatment of rejection results in a decrease in DF levels. Little is known about the clinical signifi- cance of rebound of, or increase in, DF following initial decrease associated with rejection treatment. Methods: A cohort of 88 heart transplant recipients were prospectively followed. DF was quantified using a targeted assay, myTAIheart test (TAI Diagnostics, Milwaukee Wisconsin). 7 subjects were treated for rejection and had longitudinal samples available for analysis with serial DF levels before and treatment. Clinical end points were death, need for mechanical circula- tory support (MCS), and recurrent or progressive rejection. Results: Two patients did not demonstrate rebound in DF following treatment and did not experience near-term adverse events. Mean pre-treatment DF was 2.67% and post-treatment was 0.15%. Of the five patients who demonstrated rebound in DF, two required MCS within 19 days following DF rebound and subsequently died. One patient with DF rebound developed progression of previously present cardiac allograft vasculopathy (CAV) within 42 days fol- lowing rebound. The two remaining subjects who demonstrated DF rebound did not experience clinical adverse events. Conclusion: We found that initial treatment of rejection lowers DF in general. Rebound of DF following treatment of rejection appears to be correlated with near-term adverse clinical events. Larger studies are needed to define the precise prognostic significance of this observed treatment effect. ( 1036) Pediatric Donor-Recipient Size Matching - Seeing is Believing, Or is it? J. Plasencia , 1 J.R. Ryan, 2 D. Velez, 2 K. Lagerstrom, 2 J.J. Nigro, 3 T. Karamlou, 2 Y. Kamarianakis, 1 D.H. Frakes, 1 S.G. Pophal, 2 S.D. Zangwill. 2 1 Arizona State University, Tempe, AZ; 2 Phoenix Children's Hospital, Phoenix, AZ; 3 Rady Children's Hospital, San Diego, CA. Purpose: Size matching in pediatric heart transplant (PHTx) remains subjective and weight based, starting at listing, when a maximum acceptable donor weight is entered into UNET. We aimed to develop a process using a library of 3D “donors” to derive a recipient specific data driven max acceptable donor weight. Methods: 45 PHTx pts from our institution had data and CT/MR images reviewed retrospectively. Outcome-blinded surgeons virtually transplanted healthy heart reconstructions, i.e., strategically fused geometries into CT/MR images, sequentially until they identified the max volume they were willing to accept. Surgeons assessed for graft overlap onto surrounding anatomy; looking for compressive effects. Corresponding virtual max body weights were compared to the actual donors for these pts, and to contemporaneous cohorts from PHTS and UNOS. Max donor weight to recipient weight ratio at listing (MaxDR) and donor to recipient weight ratio (DRWR) at HTx were compared. PHTS and UNOS listing data from 2011-2015 were obtained. Results: 45 local PHTx cases were compared with 3007 UNOS listings and 1903 HTx cases from PHTS. Our center lists pts with a higher MaxDR than other UNOS centers (mean 2.9 vs 2, p<.0001) and has a higher actual DRWR than PHTS counterparts (mean 1.7 vs 1.4, p=.003). Mean surgeon derived MaxDR was 2.1 and corresponding mean max DR ratio by total cardiac volume (TCV) =1. In 3 cases, actual DRWR exceeded MaxDR by > 30% - including the 1 case with fit related complications. Actual donor TCV (derived by a previously described algorithm) exceeded surgeon derived max TCV in 9 cases. In 2, there was > 20% pt growth between scan and HTx. Of the remaining 7, one had serious fit related morbidity. transplant (HTx) recipients are possible. We tested the hypothesis that con- genital heart disease has increased post-HTx resource utilization but poorer outcomes compared to other HTx indications. Methods: We retrospectively analyzed a merged UNOS-PHIS cohort of pedi- atric HTx recipients at participating centers between 2004 and 2015. Patients were categorized as congenital (CHD), myocarditis (MYO), or cardiomyopathy (CM) based on UNOS-defined primary indication for first HTx. Standardized costs were estimated using cost-to-charge ratios for each hospital. Chi-square test and Wilcoxon tests were used to compare across indications. Results: Of 2264 pediatric HTx (1106 CHD, 106 MYO, 1039 CM), CHD recipients were younger (median years 2 [IQR 0-10] vs. 6 [IQR 0-12] vs. 7 [IQR 1-14], respectively; p<0.001) and less likely to have ventricular assist device at time of HTx (CHD 3%, MYO 27%, CM 13%; p<0.001); Mortality during transplant admission (10% vs. 5% vs. 1%, p<0.001), and at 1-year (14% vs 8% vs 3%; p<0.001) was higher in CHD compared to MYO and CM. Post-HTx length of stay (LOS) was longer in CHD than MYO or CM (median days 25 [IQR 15-45] vs. 21 [IQR 12-35] vs. 16 [12-25]; p<0.001). CHD had longer ICU LOS (median days 9 [IQR 5-22] vs. 6 [IQR 4-25] vs 5 [IQR 3-9]; p<0.001), longer mechanical ventilation (median days 7 [IQR 3-19] vs. 4 [IQR 2-12] vs. 3 [IQR 2-6]; p<0.001) and longer vasoactive sup- port (median days 7 [IQR 5-14] vs. 6 [IQR 4-10] vs. 5 [IQR 3-8]; p<0.001) compared to MYO and CM. Need for ECMO within 1 year after HTx was higher in CHD (15%) and MYO (15%) compared to CM (6%) (p<0.001), while re-HTx within 1 year was similar (CHD 0.4%, MYO 0.9%, CM 0.4%; p=0.83). Number of readmissions were similar (median 1 [IQR 0-2] for all; p=0.23). Post-HTx costs were higher for CHD compared to MYO and CM (median $316k [IQR $201-503k] vs. $259k [IQR $179-390k] vs. $222k [IQR $145-319k]; p<0.001), primarily driven by longer LOS. Conclusion: CHD has poorer outcomes and increased resource utilization, including costs, during the first year after HTx. Further work is ongoing to understand more fully these differences. ( 1034) Exercise Training in Pediatric Ventricular Assist Device Recipients D.S. Burstein , M. McBride, J.W. Rossano, M.J. O'Connor, K.Y. Lin, S.M. Paridon. Cardiology, Children's Hospital of Philadelphia, Philadelphia, PA. Purpose: Children with end-stage heart failure may require ventricular assist device (VAD) while awaiting heart transplantation. Currently, no data exist on the safety of exercise training in pediatric patients on VAD support. The purpose of this study was to determine the safety and feasibility of an exercise training program for pediatric heart failure patients on VAD support while awaiting heart transplantation. Methods: After VAD placement and ambulatory recovery, patients were enrolled into an exercise training program while awaiting heart transplanta- tion between 1998 and 2017. Both inpatient and outpatient exercise sessions were scheduled three times a week lasting from 30-60 minutes and consisted of aerobic and musculoskeletal conditioning. Results: 25 patients (56% male, mean age 14 ± 3.2 years) were included with a median VAD duration of 120 ± 109 days. Patient diagnoses consisted of dilated cardiomyopathy (n = 17), restrictive cardiomyopathy (n = 2), myocar- ditis (n = 2), graft failure (n = 1), and biventricular congenital heart disease (n = 4). Pediatric VAD support included both pulsatile (Thoratec, n = 15; Berlin EXCOR, n = 2 including 1 biventricular Berlin EXCOR; Syncardia Total Artificial Heart, n = 1) and continuous flow devices (Heartware, n = 8; Heartmate I, n = 1). 1017 (85%) of a possible 1192 exercise training sessions were conducted. Reasons for non-compliance with exercise training included patient cooperation, staffing and unavailability of subject due to conflicting testing or procedures. No adverse episodes of hypotension or significant complex arrhythmia occurred and there was no termination from the program. No complications due to device malfunction or damage to the VAD device, including driveline, pump or cables, occurred. Serial exercise stress tests after VAD placement performed in four patients showed no change in peak VO2 (mean 0 +/- 0 mL/kg/min, p = 0.36) but trend towards increased work rate (mean 10 +/- watts, p = 0.07), suggestive of improved musculoskeletal conditioning and efficiency. Conclusion: Pediatric patients on VAD support for advanced heart failure can safely participate in exercise training programs with relatively high compliance. Exercise training programs in pediatric VAD recipients may