- Email: [email protected]
AMR in the Current Era: Clinical, Pathologic, and Immunologic Features in Pediatric Heart Transplant Recipieints
Abstracts S273 pancreatitis. The ages of the pancreatitis and non-pancreatitis groups were compared using the Student’s T test. Results: 266 heart transplants were performed over the time period. 10 cases of pancreatitis were identified (incidence 3.7 %). The mean age of the pancreatic group was 11.9 years, versus 8.3 years in the non-pancreatitis group (p = 0.02). All 10 patients had abnormal biochemistry, either elevated amylase (1 patient), lipase (6 patients), or both amylase and lipase (3 patients). All pancreatitis patients also had imaging of their pancreas attempted using either abdominal computed tomography (CT) or ultrasound, but in 3 cases it was not possible to visualise the pancreas (all ultrasound scans). 2 cases had a normal pancreatic appearance (1 CT and 1 ultrasound). 5 cases had radiological evidence of pancreatitis (3 CT and 2 ultrasound). 2 patients required surgical intervention for pancreatits (1 laparotomy, 1 percutaneous drainage of pancreatic pseudocyst and gastrocystic stenting). Of the 10 patients with pancreatitis, 2 died before discharge from hospital. 1 patient required long term treatment with insulin for impaired glucose tolerance. Conclusion: This is the largest study yet presented on pancreatitis following paediatric heart transplant. Our incidence of pancreatitis is similar to previously published studies in both children and adults. The patients who developed pancreatitis were significantly older than those who did not. Pancreatitis is associated with a substantial pre-discharge mortality, and can result in long-term morbidity for some patients. 7( 19) Improving Outcomes of Cardiac Transplantation in Patients with Cavopulmonary Connection L.E. De León ,1 J.S. Heinle,1 E. McKenzie,1 C.M. Mery,1 L.C. Kane,1 W. Dreyer,2 S.W. Denfield,2 J.F. Price,2 A.G. Cabrera,2 C.D. Fraser Jr.,1 I. Adachi.1 1Division of Congenital Heart Surgery, Texas Children's Hospital; Michael E. DeBakey Department of Surgery, Baylor College of Medicine, Houston, TX; 2Lillie Frank Abercrombie Section of Pediatric Cardiology, Texas Children's Hospital, Houston, TX. Purpose: Previous studies have reported unfavorable early mortality rates (up to 35%) following transplantation in patients with cavopulmonary connection(s) either with bidirectional Glenn (BDG) or Fontan procedures. We sought to evaluate outcomes of cardiac transplantation in this group of patients at a single pediatric institution. Methods: All patients who had undergone cardiac transplantation after a BDG or Fontan procedure between 1995 and 2016 were included. Actuarial survival was analyzed using the Kaplan-Meier method. Results: The cohort included 53 patients [median age 7 years (7 months - 20 years)], with 25 (47%) at BDG and 28 (53%) at Fontan stage. Six patients (11%) were mechanically bridged. Median time from BDG or Fontan to transplantation was 4 years (2 months - 19 years). Protein losing enteropathy (PLE) and/or plastic bronchitis (PB) were seen in 11 (39%) of Fontan patients. There were 2 (3.7%) operative deaths; one due to graft failure at 43 days postoperatively in a BDG patient who had a positive crossmatch, and the other one due to multi-organ failure at 306 days in a Fontan patient. There were 16 (30%) late mortalities, eight in each group of patients. Actuarial survival at 1- and 5-years after transplant was 89% and 59%, respectively. When compared to all transplant recipients with congenital heart disease (CHD) but without cavopulmonary connections, there was no statistical difference in operative [2/53 (3.7%) vs 1/45 (2.2%), p= 0.55] or late mortality [16/53 (30%) vs 16/45 (36%), p= 0.57]. The Fontan non-survivor had severe PLE/ascites, and these conditions persisted even after restoration of normal cardiac physiology with transplantation. This was in stark contrast to the rest of the Fontan patients, most of whom [9/10 (90%)] experienced resolution of PLE/PB. Conclusion: With appropriate patient selection, cardiac transplantation after BDG and Fontan procedures can be performed with low operative mortality. Resolution of PLE and PB may be expected unless transplantation is offered at a too advanced stage. Although the presence of cavopulmonary connections is not a negative predictor among the CHD population, further improvement is necessary in posttransplant outcomes since late outcomes in this cohort remain suboptimal relative to the overall pediatric transplant outcomes. 7 ( 20) Primary Graft Dysfunction After Pediatric Heart Transplant: A Single Center Experience
K. Mowers , K. Simpson, P. Eghtesady, O. Petrucci, C. Canter, C. Castleberry. Washington University School of Medicine, St. Louis, MO. Purpose: Primary graft dysfunction (PGD) has been identified as a complication after transplantation (OHT) and a significant cause of mortality, particularly in infant recipients. Lack of standardized definition of PGD in the pediatric population makes prevalence and impact of PGD unclear. Methods: We applied adult ISHLT PGD consensus guidelines, which include number and doses of inotropes and need for mechanical circulatory support (MCS), to 208 pediatric OHT recipients from Jan 2005-May 2016. PGD was defined as: moderate PGD - inotrope score > 10 on post-operative day 1 (24-48 hours), and severe PGD - MCS within 24 hours (in the absence of rejection). Results: PGD occurred in 36 patients (17%); 18 of which had severe PGD (8.5%). All with severe PGD were on ECMO. Multivariate risk factors for PGD included a positive cross match (PCM) (OR 7.23, 95% 1.9,29.8, p< 0.01) and cardiopulmonary bypass time (CPBT) (OR 10.14/10 min, 95% 10.05,10.2, p< 0.01), but not age, weight, ischemic time, or primary diagnosis. Risk factors for severe PGD also included PCM (OR 7.87, 95% 1.26,55.4, p= 0.03) and CPBT (OR 10.3/10 min, 95% 10.05,10.2, p= 0.03). Patients with PGD vs those without PGD, had longer hospital courses (28 vs 10 days (med), p< 0.01); yet when comparing patients with moderate PGD to without PGD, they also had longer courses (26 vs 10 days) but this was not statistically significant (p= 0.08). Patients with PGD had decreased overall survival (69% vs 85%, p= 0.03), which remained significant on multivariate analysis (HR 2.82, 95% CI 1.01, 7.67, p< 0.01 ). Severe PGD was associated with the worst survival; yet survival was comparable between moderate PGD and those without PGD (89% vs 85%, p= 0.65) [Figure 1]. Conclusion: In the current era, infants do not have increased risk of PGD; however, it appears to be related to CPBT and PCM. Need for MCS, but not high dose inotropes, is associated with increased mortality. The association of PCM and severe PGD in the absence of detectable rejection requires further investigation.
7( 21) AMR in the Current Era: Clinical, Pathologic, and Immunologic Features in Pediatric Heart Transplant Recipieints D. Drogalis-Kim ,1 M.J. Hickey,2 M.C. Fishbein,3 J.C. Alejos,1 E.F. Reed,2 N.J. Halnon.1 1Pediatrics, UCLA, Los Angeles, CA; 2Immunogenetics Center, Department of Pathology and Laboratory Medicine, UCLA, Los Angeles, CA; 3Pathology and Laboratory Medicine, UCLA, Los Angeles, CA. Purpose: Right ventricular (RV) endomyocardial biopsy (EMB), an invasive procedure, is currently the mainstay for monitoring for allograft rejection. Over time, criteria for the pathologic diagnosis of antibody mediated rejection (pAMR) have evolved based on new diagnostic methods. However,
S274
The Journal of Heart and Lung Transplantation, Vol 36, No 4S, April 2017
clinical correlates, prognosis, and optimal treatment for pAMR in pediatric heart transplant recipients remain unclear. Our aim is to describe clinical, pathologic, and immunologic features present at the time of pAMR diagnosis and at next follow-up EMB. Methods: Charts of pediatric patients transplanted at UCLA from 2010-2015 were reviewed retrospectively to define a cohort of patients with pAMR diagnosed by RV EMB. Clinical data including symptoms, heart rate, left ventricular ejection fraction by echocardiogram, pathology findings by RV EMB, and presence and strength of donor specific HLA antibody (DSA) as well as immunomodulatory therapy given at the time of initial pAMR+ EMB and at next follow up EMB were recorded. Results: At our institution, 15/82 pediatric patients transplanted had pAMR a median 416 days (range: 6-1446d) post-OHT. (5 pAMR2, 7 pAMR1(H+), 3 pAMR1(I+), 1pAMR1(H+/I+). Median age at time of transplant was 13.0 years (range: 342 days-20 years). At the time of EMB, 6/15 were symptomatic and/or had clinical abnormality but with little correlation with pAMR grade. HLA DSA were present in 8/14 tested. AMR treatment generally consisted of combinations including ATG, IVIG, Rituxan, plasmapheresis or steroids. Complete resolution of pAMR (median time 75d to biopsy) occurred in 12/15 patients, including 2/4 cases of pAMR2, and not resolved in 3 cases. In all those without DSA, pAMR resolved without treatment. At the time of follow up EMB, DSA initially weak/moderate in strength (average: 5133 MFI; range: 1203-12087 MFI) were significantly reduced. Therapy was generally not effective at reducing strong DSA (average: 9665 MFI; range: 1128-16257 MFI). Conclusion: Clinical features in our pediatric patients with pAMR varied and were seldom helpful in predicting grades of rejection though absence of DSA was associated with resolution of pAMR without treatment. Further study is indicated to determine optimal treatment and role of non-invasive monitoring for AMR in children. 7( 22) Preoperative Recipient Factors Predict Outcome in Pediatric Heart Transplant Recipients Using a Novel Risk Factor Score M.J. O'Connor ,1 A.C. Glatz,1 J.W. Rossano,1 R.E. Shaddy,1 R.M. Ryan,1 C. Ravishankar,1 S. Fuller,2 T.L. Spray,2 C.E. Mascio,2 J.W. Gaynor,2 K.Y. Lin.1 1Division of Cardiology, The Children's Hospital of Philadelphia, Philadelphia, PA; 2Division of Cardiothoracic Surgery, The Children's Hospital of Philadelphia, Philadelphia, PA. Purpose: Appropriate candidate selection in pediatric heart transplantation (HT) is critical to optimizing outcome, but remains challenging in part due to inadequate risk prediction models. Methods: As part of a quality improvement initiative, outcomes on all HTs at our center between 1/2010 and 8/2016 were reviewed. Multiple organ recipients were excluded. Recipient clinical parameters were investigated; those identified by univariate Cox regression to be significant risk factors for mortality were assigned a point value, based on magnitude of the hazard ratio, to derive a risk score. The discriminant ability of this score to predict mortality was assessed using area under the receiver-operating characteristic (ROC) curve; survival estimates were generated stratified by number of risk factors. Results: Seventy-four HTs were performed, with unadjusted 13.5% (10/74) operative mortality (prior to discharge). Overall median follow-up was 2.7 y (range 2 d - 6.1 y). Risk factors (hazard ratio, 95% CI, P value) for overall mortality in univariate analysis included single ventricle CHD (3.2, 1.0-9.5, P = 0.04), pre-HT BiVAD (4.8, 1.1-22.2, P = 0.04), ≥ 4 sternotomies prior to HT (3.9, 1.2-12.8, P = 0.02), pre-HT PRA > 10% (4.4, 1.4-14.2, P = 0.01), prior surgery at other institution (3.2, 1.1-9.6, P = 0.04), and pre-HT pulmonary venous obstruction (4.7, 1.0-21.4, P = 0.05). The presence of ≥ 4 risk factors predicted mortality with 57% sensitivity and 90% specificity (area under ROC curve 0.76). There were significant differences in survival when stratified by number of risk factors (P = 0.0001, Figure). Conclusion: Pediatric HT candidates at higher risk for post-HT mortality may be identified preoperatively through recipient-specific factors not captured by present risk prediction models. In this study, ≥ 4 risk factors predicted mortality with high specificity. These risk factors should be assessed in larger populations and considered for inclusion in models assessing center performance.
7( 23) CMV-Specific T-Cell Immunity and CMV NAAT in Assignment of CMV Status in Pediatric SOT Candidates C. Burton ,1 J. Preiksaitis,1 M. Sester,2 Y. Tong,1 F. Wong,1 S. Urschel.1 1University of Alberta, Edmonton, AB, Canada; 2Saarland University, Homburg, Germany. Purpose: Presence of passive transfusion-acquired or maternal antibodies, in infants, can make serologic determination of CMV status unreliable. We evaluated detection of CMV-specific CD4 T-cells and CMV nucleic acid amplification testing (NAAT) in assignment of CMV status in children awaiting solid organ transplant. Methods: We enrolled 24 children awaiting SOT (5 heart, 2 lung, 11 liver, 6 renal) and 24 pediatric immunocompetent controls, and collected a throat swab, urine, saliva and blood sample on each. Using flow cytometry CMVspecific CD4+ cells were identified by CD69 expression and intracellular IFN-γ quantification after stimulation with CMV-antigen lysate. CMV NAAT was performed on all specimens using our in- house real time PCR assay. CMV serology was determined by enzyme immunoassay. Subjects were considered to have potential passive antibody (PA) if < 18 mos. and/or received blood products within 2 mos. of collection. Results: Median ages (yrs.) of subjects and controls were 1.45 (0.07-17.4) and 2.4 (0.2-17.0). 9/24 subjects and 7/24 controls were CMV seropositive. 12/24 subjects were inpatients and 6 were in ICU. CMV NAAT was positive in 6/11 seropositive children with PA; CMV-specific T cells were only detected in 2/11 (Figure 1). Of the 4 with positive CMV NAAT without detectable CMV-specific T cells 3 were < 18 mos, 2 controls and 1 outpatient pre-liver candidate, and 1 was a 15 yr. old on ECMO awaiting lung transplant. CMV-specific T cells were detected in 5/5 CMV seropositive children without PA; 2/5 (both 4 yrs.) had positive CMV NAAT. No CMV seronegative child had a positive CMV NAAT, 1 had detectable CMV-specific T cells. Conclusion: CMV NAAT is useful when positive, to identify true CMV infection in infants with PA. Detection of CMV-specific CD4 T-cells was not reliable in assignment of CMV status in these children, likely due to recently acquired infection and/or the critically ill condition of the subjects. An algorithm including CMV serology, NAAT and CMV-specific T cells may be promising.
K. Mowers , K. Simpson, P. Eghtesady, O. Petrucci, C. Canter, C. Castleberry. Washington University School of Medicine, St. Louis, MO. Purpose: Primary graft dysfunction (PGD) has been identified as a complication after transplantation (OHT) and a significant cause of mortality, particularly in infant recipients. Lack of standardized definition of PGD in the pediatric population makes prevalence and impact of PGD unclear. Methods: We applied adult ISHLT PGD consensus guidelines, which include number and doses of inotropes and need for mechanical circulatory support (MCS), to 208 pediatric OHT recipients from Jan 2005-May 2016. PGD was defined as: moderate PGD - inotrope score > 10 on post-operative day 1 (24-48 hours), and severe PGD - MCS within 24 hours (in the absence of rejection). Results: PGD occurred in 36 patients (17%); 18 of which had severe PGD (8.5%). All with severe PGD were on ECMO. Multivariate risk factors for PGD included a positive cross match (PCM) (OR 7.23, 95% 1.9,29.8, p< 0.01) and cardiopulmonary bypass time (CPBT) (OR 10.14/10 min, 95% 10.05,10.2, p< 0.01), but not age, weight, ischemic time, or primary diagnosis. Risk factors for severe PGD also included PCM (OR 7.87, 95% 1.26,55.4, p= 0.03) and CPBT (OR 10.3/10 min, 95% 10.05,10.2, p= 0.03). Patients with PGD vs those without PGD, had longer hospital courses (28 vs 10 days (med), p< 0.01); yet when comparing patients with moderate PGD to without PGD, they also had longer courses (26 vs 10 days) but this was not statistically significant (p= 0.08). Patients with PGD had decreased overall survival (69% vs 85%, p= 0.03), which remained significant on multivariate analysis (HR 2.82, 95% CI 1.01, 7.67, p< 0.01 ). Severe PGD was associated with the worst survival; yet survival was comparable between moderate PGD and those without PGD (89% vs 85%, p= 0.65) [Figure 1]. Conclusion: In the current era, infants do not have increased risk of PGD; however, it appears to be related to CPBT and PCM. Need for MCS, but not high dose inotropes, is associated with increased mortality. The association of PCM and severe PGD in the absence of detectable rejection requires further investigation.
7( 21) AMR in the Current Era: Clinical, Pathologic, and Immunologic Features in Pediatric Heart Transplant Recipieints D. Drogalis-Kim ,1 M.J. Hickey,2 M.C. Fishbein,3 J.C. Alejos,1 E.F. Reed,2 N.J. Halnon.1 1Pediatrics, UCLA, Los Angeles, CA; 2Immunogenetics Center, Department of Pathology and Laboratory Medicine, UCLA, Los Angeles, CA; 3Pathology and Laboratory Medicine, UCLA, Los Angeles, CA. Purpose: Right ventricular (RV) endomyocardial biopsy (EMB), an invasive procedure, is currently the mainstay for monitoring for allograft rejection. Over time, criteria for the pathologic diagnosis of antibody mediated rejection (pAMR) have evolved based on new diagnostic methods. However,
S274
The Journal of Heart and Lung Transplantation, Vol 36, No 4S, April 2017
clinical correlates, prognosis, and optimal treatment for pAMR in pediatric heart transplant recipients remain unclear. Our aim is to describe clinical, pathologic, and immunologic features present at the time of pAMR diagnosis and at next follow-up EMB. Methods: Charts of pediatric patients transplanted at UCLA from 2010-2015 were reviewed retrospectively to define a cohort of patients with pAMR diagnosed by RV EMB. Clinical data including symptoms, heart rate, left ventricular ejection fraction by echocardiogram, pathology findings by RV EMB, and presence and strength of donor specific HLA antibody (DSA) as well as immunomodulatory therapy given at the time of initial pAMR+ EMB and at next follow up EMB were recorded. Results: At our institution, 15/82 pediatric patients transplanted had pAMR a median 416 days (range: 6-1446d) post-OHT. (5 pAMR2, 7 pAMR1(H+), 3 pAMR1(I+), 1pAMR1(H+/I+). Median age at time of transplant was 13.0 years (range: 342 days-20 years). At the time of EMB, 6/15 were symptomatic and/or had clinical abnormality but with little correlation with pAMR grade. HLA DSA were present in 8/14 tested. AMR treatment generally consisted of combinations including ATG, IVIG, Rituxan, plasmapheresis or steroids. Complete resolution of pAMR (median time 75d to biopsy) occurred in 12/15 patients, including 2/4 cases of pAMR2, and not resolved in 3 cases. In all those without DSA, pAMR resolved without treatment. At the time of follow up EMB, DSA initially weak/moderate in strength (average: 5133 MFI; range: 1203-12087 MFI) were significantly reduced. Therapy was generally not effective at reducing strong DSA (average: 9665 MFI; range: 1128-16257 MFI). Conclusion: Clinical features in our pediatric patients with pAMR varied and were seldom helpful in predicting grades of rejection though absence of DSA was associated with resolution of pAMR without treatment. Further study is indicated to determine optimal treatment and role of non-invasive monitoring for AMR in children. 7( 22) Preoperative Recipient Factors Predict Outcome in Pediatric Heart Transplant Recipients Using a Novel Risk Factor Score M.J. O'Connor ,1 A.C. Glatz,1 J.W. Rossano,1 R.E. Shaddy,1 R.M. Ryan,1 C. Ravishankar,1 S. Fuller,2 T.L. Spray,2 C.E. Mascio,2 J.W. Gaynor,2 K.Y. Lin.1 1Division of Cardiology, The Children's Hospital of Philadelphia, Philadelphia, PA; 2Division of Cardiothoracic Surgery, The Children's Hospital of Philadelphia, Philadelphia, PA. Purpose: Appropriate candidate selection in pediatric heart transplantation (HT) is critical to optimizing outcome, but remains challenging in part due to inadequate risk prediction models. Methods: As part of a quality improvement initiative, outcomes on all HTs at our center between 1/2010 and 8/2016 were reviewed. Multiple organ recipients were excluded. Recipient clinical parameters were investigated; those identified by univariate Cox regression to be significant risk factors for mortality were assigned a point value, based on magnitude of the hazard ratio, to derive a risk score. The discriminant ability of this score to predict mortality was assessed using area under the receiver-operating characteristic (ROC) curve; survival estimates were generated stratified by number of risk factors. Results: Seventy-four HTs were performed, with unadjusted 13.5% (10/74) operative mortality (prior to discharge). Overall median follow-up was 2.7 y (range 2 d - 6.1 y). Risk factors (hazard ratio, 95% CI, P value) for overall mortality in univariate analysis included single ventricle CHD (3.2, 1.0-9.5, P = 0.04), pre-HT BiVAD (4.8, 1.1-22.2, P = 0.04), ≥ 4 sternotomies prior to HT (3.9, 1.2-12.8, P = 0.02), pre-HT PRA > 10% (4.4, 1.4-14.2, P = 0.01), prior surgery at other institution (3.2, 1.1-9.6, P = 0.04), and pre-HT pulmonary venous obstruction (4.7, 1.0-21.4, P = 0.05). The presence of ≥ 4 risk factors predicted mortality with 57% sensitivity and 90% specificity (area under ROC curve 0.76). There were significant differences in survival when stratified by number of risk factors (P = 0.0001, Figure). Conclusion: Pediatric HT candidates at higher risk for post-HT mortality may be identified preoperatively through recipient-specific factors not captured by present risk prediction models. In this study, ≥ 4 risk factors predicted mortality with high specificity. These risk factors should be assessed in larger populations and considered for inclusion in models assessing center performance.
7( 23) CMV-Specific T-Cell Immunity and CMV NAAT in Assignment of CMV Status in Pediatric SOT Candidates C. Burton ,1 J. Preiksaitis,1 M. Sester,2 Y. Tong,1 F. Wong,1 S. Urschel.1 1University of Alberta, Edmonton, AB, Canada; 2Saarland University, Homburg, Germany. Purpose: Presence of passive transfusion-acquired or maternal antibodies, in infants, can make serologic determination of CMV status unreliable. We evaluated detection of CMV-specific CD4 T-cells and CMV nucleic acid amplification testing (NAAT) in assignment of CMV status in children awaiting solid organ transplant. Methods: We enrolled 24 children awaiting SOT (5 heart, 2 lung, 11 liver, 6 renal) and 24 pediatric immunocompetent controls, and collected a throat swab, urine, saliva and blood sample on each. Using flow cytometry CMVspecific CD4+ cells were identified by CD69 expression and intracellular IFN-γ quantification after stimulation with CMV-antigen lysate. CMV NAAT was performed on all specimens using our in- house real time PCR assay. CMV serology was determined by enzyme immunoassay. Subjects were considered to have potential passive antibody (PA) if < 18 mos. and/or received blood products within 2 mos. of collection. Results: Median ages (yrs.) of subjects and controls were 1.45 (0.07-17.4) and 2.4 (0.2-17.0). 9/24 subjects and 7/24 controls were CMV seropositive. 12/24 subjects were inpatients and 6 were in ICU. CMV NAAT was positive in 6/11 seropositive children with PA; CMV-specific T cells were only detected in 2/11 (Figure 1). Of the 4 with positive CMV NAAT without detectable CMV-specific T cells 3 were < 18 mos, 2 controls and 1 outpatient pre-liver candidate, and 1 was a 15 yr. old on ECMO awaiting lung transplant. CMV-specific T cells were detected in 5/5 CMV seropositive children without PA; 2/5 (both 4 yrs.) had positive CMV NAAT. No CMV seronegative child had a positive CMV NAAT, 1 had detectable CMV-specific T cells. Conclusion: CMV NAAT is useful when positive, to identify true CMV infection in infants with PA. Detection of CMV-specific CD4 T-cells was not reliable in assignment of CMV status in these children, likely due to recently acquired infection and/or the critically ill condition of the subjects. An algorithm including CMV serology, NAAT and CMV-specific T cells may be promising.