ASAIO Journal 2014
Pediatric Circulatory Support
Human Leukocyte Antigen Sensitization in Pediatric Patients Exposed to Mechanical Circulatory Support Borah J. Hong,* Meghan Delaney,†‡ Anthony Guynes,‡ Paul Warner,† David M. McMullan,§ Mariska S. Kemna,¶ Robert J. Boucek,¶ and Yuk M. Law¶
Human leukocyte antigen (HLA) sensitization of pediatric heart recipients increases their risk of rejection and graft loss. As more children are placed on mechanical circulatory support (MCS) as a bridge to transplant, the risk factors for development of sensitization warrant further study. A single-center retrospective review of 36 children who received MCS identified 22 patients supported with either extracorporeal membrane oxygenation (ECMO) (n = 15) or ECMO-ventricular assist device (VAD) (n = 7) with paired (pre-MCS/post-MCS) panel reactive antibodies (PRA) or only negative post-MCS PRAs. Four patients (18%) became sensitized post-MCS (one ECMO-only patient, three ECMO-VAD patients). No difference was found between sensitized and nonsensitized patients in terms of congenital heart disease versus primary cardiomyopathy (p = 0.096), duration of MCS (38 days vs. 14 days, p = 0.233), or volume of blood product transfusions (358.6 ml/kg vs. 612.7 ml/kg, p = not significant). By multivariable analysis, the association of sensitization with older age at MCS (p = 0.076) and history of homograft (p = 0.064) approached significance. Pediatric patients supported with MCS are at low risk of developing HLA sensitization. Diagnosis, MCS duration, and volume of transfused blood products do not appear to be associated with HLA sensitization, but there is a suggestion of an association with older age at MCS and history of a homograft. ASAIO Journal 2014; 60:317–321.
orthotopic heart transplantation (OHT), human leukocyte antigen (HLA) sensitization becomes a greater concern. Patients who are sensitized have to wait longer for OHT, have higher rates of primary graft dysfunction compared with nonsensitized patients, and are more likely to develop rejection.1,2 As evidenced in adult studies, HLA sensitization rates following ventricular assist device (VAD) implantation are as high as 66%.3 The increased sensitization rates in adults are attributed to the blood product transfusions, particularly platelet transfusions, and the inclusion of women with prior pregnancies.4,5 However, sensitization of pediatric patients is likely to be different from adults in terms of their inflammatory and immunological response to human tissue exposure, blood product requirements, and mode of MCS. In children, extracorporeal membrane oxygenation (ECMO) is an important and widely used mode of MCS either as a transitional mode of support to a durable VAD or used alone such as in postcardiotomy, with biventricular failure, or when a short period of support is expected.6–8 There have been only two single-center studies in children reporting HLA sensitization frequency in a cohort (n = 20) of predominantly VAD patients9 and a combination of ECMO (n = 13) and VAD patients (n = 9).10 As the cumulative experience is small, we sought to examine the frequency of HLA sensitization and the risk factors in a modern cohort of pediatric transplant candidates who required MCS.
Key Words: pediatric mechanical circulatory support, HLA sensitization
Methods This was a single-center retrospective cohort study of all patients placed on MCS with either ECMO or VAD between 2005 and 2011. Institutional review board approval was obtained. Inclusion criteria included having all prior surgeries performed at our institution, having complete records of blood product transfusions before and after MCS initiation, and having availability of panel reactive antibodies (PRA) at least post-MCS. Patient demographics and medical data including cardiac diagnosis, prior cardiac surgeries, prior homograft placement, MCS device type, and duration of support were obtained. Blood product transfusion volumes for packed red blood cells, platelets, plasma, and fresh whole blood for each patient before MCS and while on MCS were noted. All blood products were leukocyte reduced and irradiated. Human leukocyte antigen sensitization was defined as PRA greater than 10% for either HLA class I or class II. Panel reactive antibodies were measured by flow cytometry from 2005 to 2008 and subsequently by Luminex platform-based bead assay (FlowPRA Screening; One Lambda Inc., Canoga Park, CA). We collected
As more pediatric patients with end-stage heart failure are
placed on mechanical circulatory support (MCS) as a bridge to
From the *Division of Pediatric Cardiology, Texas Children’s Hospital, Baylor College of Medicine, Houston, Texas; †Puget Sound Blood Center, Seattle, Washington; ‡University of Washington School of Medicine, Seattle, Washington; §Division of Pediatric Cardiothoracic Surgery, Seattle Children’s Hospital, University of Washington, Seattle, Washington; and ¶Division of Pediatric Cardiology, Seattle Children’s Hospital, University of Washington, Seattle, Washington. Submitted for consideration September 2013; accepted for publication in revised form January 2014. Disclosure: The authors have no conflicts of interest to report. Reprint Requests: Borah J. Hong, Division of Pediatric Cardiology, Texas Children’s Hospital, 6621 Fannin Street, MC-19345-C, Houston, TX 77030-2399. Email: [email protected] Copyright © 2014 by the American Society for Artificial Internal Organs DOI: 10.1097/MAT.0000000000000053
317
318 HONG et al. all PRA results, pre-MCS, and post-MCS that were obtained and used the latest PRA before MCS and the latest PRA after MCS initiation to determine sensitization. We included those patients (10 of the 22 patients) who only had post-MCS PRAs if their PRA class I and class II were less than 10% as they were assumed to not have been sensitized before MCS if they were not sensitized following MCS. None of the patients underwent desensitization therapy before, during, or after MCS. Statistical Analysis Patient demographic data are expressed as median with interquartile range (IQR). Comparison of categorical variables between sensitized and nonsensitized patients was performed using the Fisher’s exact test where appropriate. Comparison of continuous variables between sensitized and nonsensitized patients was performed using the Mann–Whitney U test where appropriate. Multivariable analysis of predictor variables was performed by binary logistic regression. A p value of less than 0.05 was considered statistically significant. All statistical analysis was performed using IBM SPSS Statistics Version 18.0 (SPSS Inc., Chicago, IL). Results Patient Population There were 36 patients who were placed on MCS from 2005 to 2011. Three patients were excluded because they were placed on MCS following OHT, and 11 patients had insufficient data such as no post-MCS PRAs as they were not considered candidates for transplantation. Complete medical data were available for a cohort of 22 patients. Patient characteristics are shown in Table 1. Mechanical Circulatory Support ECMO was used in all 22 patients in the cohort. Seven of the 22 patients (32%) were additionally supported with VAD. One patient was initially placed on a Berlin Heart EXCOR (Berlin Heart, Inc., The Woodlands, TX) and subsequently transitioned to ECMO. Of the 21 patients who were initially placed on ECMO, five were subsequently transitioned to a left ventricular assist device (four onto Berlin Heart EXCOR, one onto MicroMed Debakey VAD [MicroMed Technology, Inc., Table 1. Patient Characteristics Number (%) or Median (Interquartile Range) Total patients Male Age at MCS (months) Weight at MCS (kg) Cardiomyopathy Congenital heart disease Homograft ECMO as initial support ECMO to LVAD ECMO to biventricular assist device LVAD to ECMO
22 (100) 10 (45) 7.5 (0.4–107.6) 7.0 (3.3–30.7) 10 (45%) 12 (55%) 3 (14%) 21 (95%) 5 (23%) 1 (5%) 1 (5%)
MCS, mechanical circulatory support; ECMO, extracorporeal membrane oxygenator; LVAD, left ventricular assist device.
Houston, TX]). One patient was transitioned from ECMO to biventricular support with the Berlin Heart EXCOR. Table 2 lists all the patients in the cohort and the type of MCS device received. Human Leukocyte Antigen Sensitization Of the 22 patients in this cohort, the median number of postMCS PRAs that were obtained was 4 (IQR, 2–6 PRAs). The median time from MCS initiation to latest PRA obtained was 56 days (IQR, 35.8–118.5 days). Of the cohort, four (18%) developed new anti-HLA antibodies post-MCS that were above a PRA of 10% and therefore considered sensitized (Table 2). Three of these patients had anti-HLA antibodies p re-MCS, although two were at low levels (PRA < 10%). All three had new or higher anti-HLA-II antibodies post-MCS, which allowed us to classify these patients as sensitized after MCS. The one patient (patient 1) who had a negative PRA before MCS but developed sensitization post-MCS had a history of congenital heart disease, but no homograft material was used for surgical repair. Upon further examination of the four patients who became sensitized, only one patient was exclusively supported on ECMO, whereas the other three patients were supported with ECMO-VAD. This estimates a sensitization rate of 7% (1/15) in the ECMO-only group and 43% (3/7) in the VAD group (p = 0.077). The median days to latest PRA in the ECMO group was 56 days (IQR, 37–150 days) and in the VAD group was 41 days (IQR, 28–101 days), which was not significantly different (p = not significant). Comparison of Sensitized Versus Nonsensitized Patients Comparison of the nonsensitized and sensitized patients following MCS demonstrated no significant difference in any of the covariates examined (Table 3) including the time from MCS to the latest PRA collected (p = 0.469). However, several suggestive risk factors are apparent. All four sensitized patients had a history of congenital heart disease (p = 0.096) and two of them had homografts. Of the three patients with a homograft, two were in the sensitized group (p = 0.073). Patients 2, 10, and 14 all had prior homografts, but patient 2 was not sensitized pre-MCS nor at the latest PRA post-MCS. However, when looking at sequential PRAs for patient 2, the patient did have an elevated PRA post-MCS with class I 0%/class II 19% on post-MCS day 87 but then had negative PRAs of 0%/0% by post-MCS day 217. Of the two other patients with homografts who became sensitized after MCS, patient 10 was highly sensitized pre-MCS (class I 96%/class II 0%), and patient 14 had some HLA antibodies pre-MCS (class I 8%/class II 0%). Transfused Blood Products Sensitized patients received a median of 358.6 ml/kg (IQR, 80.6–965.9 ml/kg) of total blood products before the latest PRA, whereas nonsensitized patients received a median of 612.7 ml/ kg (IQR, 196.7–990.9), which was not statistically different (p = 0.496). When looking specifically at different blood product types transfused, there was no difference between the sensitized and nonsensitized groups in terms of packed red blood cells, platelets, fresh-frozen plasma, cryoprecipitate, and fresh whole blood (Table 3). When comparing the patients based on
319
HUMAN LEUKOCYTE ANTIGEN SENSITIZATION Table 2. Pre-MCS and Post-MCS Panel Reactive Antibodies
Patient 1 2* 3 4 5 6 7 8 9 10* 11 12 13 14* 15 16 17 18 19 20 21 22
MCS Device
Pre-MCS HLA Class I/II
Post-MCS HLA Class I/II
Sensitized Pre-MCS
Sensitized Post-MCS
0/0 N/A N/A 0/0 0/4 N/A 0/0 21/22 N/A
15/0 0/0 0/0 0/0 0/16 0/0 0/0 0/0 0/0
No No No No No No No Yes No
Yes No No No Yes No No No No
96/0 0/0 N/A N/A 8/0 N/A N/A 0/0 0/0 0/0 N/A N/A 5/0
100/90 0/0 0/0 0/0 0/92 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0
Yes No No No No No No No No No No No No
Yes No No No Yes No No No No No No No No
ECMO → Berlin Heart LVAD ECMO ECMO ECMO ECMO → MicroMed Debakey LVAD ECMO ECMO ECMO ECMO → Berlin Heart biventricular assist device Berlin Heart LVAD → ECMO ECMO → Berlin Heart LVAD ECMO ECMO ECMO ECMO ECMO ECMO → Berlin Heart LVAD ECMO ECMO ECMO ECMO ECMO → Berlin Heart LVAD
*History of homograft. MCS, mechanical circulatory support; HLA, human leukocyte antigen; ECMO, extracorporeal membrane oxygenator; LVAD, left ventricular assist device; N/A, not applicable.
the type of MCS support of ECMO only versus ECMO-VAD, there was no significant difference (p = 0.503) in total blood products transfused with ECMO-only patients receiving a median of 592.9 ml/kg (IQR, 200.3–985.1 ml/kg), whereas the ECMO-VAD patients received a median of 546.6 ml/kg (IQR, 65.3–1,008.3 ml/kg). Multivariable regression analysis demonstrated that the duration of MCS and the volumes of individual blood products transfused including platelets, packed red blood cells, and plasma were not associated with development of sensitization. However, there was a suggestion of HLA sensitization with older age at MCS (p = 0.076) and a history of a surgical repair with a homograft (p = 0.064).
Discussion In this study, we observed 18% of the MCS cohort to be sensitized. There was a trend toward significance with older age, VAD, and homograft use, which is suggestive of these variables as risk factors. However, there was no association between sensitization and any specific blood products or transfusion volume. Unlike the adult population, there are only two published studies on children, and they report sensitization frequencies of 27% and 35%, which puts our present data in line with these similarly sized studies.9,10 In both previously published pediatric studies, there was no association between blood product transfusions and development of sensitization.
Table 3. Comparison of Sensitized Versus Nonsensitized Patients
Median age at MCS (months) (IQR) Duration of MCS (days) (IQR) Weight at MCS (kg) (IQR) Duration from MCS to final panel reactive antibodies (days) (IQR) Congenital heart disease Cardiomyopathy Homograft Yes No Median packed red blood cells transfused in ml/kg (IQR) Median platelets transfused in ml/kg (IQR) Median fresh-frozen plasma and cryoprecipitate transfused in ml/kg (IQR) Median fresh whole blood transfused in ml/kg (IQR) Median total blood products transfused in ml/kg (IQR) MCS, mechanical circulatory support; IQR, interquartile range.
Sensitized Patients, n = 4 (18%)
Nonsensitized Patients, n = 18 (82%)
p
138.9 (27.1–174.2) 38 (16.3–90.5) 36.5 (8.8–57.3) 56.5 (18.3–133.8)
4.7 (0.4–16.2) 14 (4.8–40.5) 5.6 (3.2–8.8) 56 (37–135.3)
0.233 0.233 0.148 0.469
8 (44%) 10 (56%)
0.096
4 (100%) 0 (0%) 2 (50%) 2 (50%) 155.4 (46.9–514.8) 97.1 (11.7–323.4) 34.0 (9.6–98.4)
1 (6%) 17 (94%) 239.5 (104.7–451.5) 148.6 (77.4–394.6) 69.4 (29.3–172.5)
0.073
0.0 (0.0–113.6) 358.6 (80.6–965.9)
0.0 (0.0–57.3) 612.7 (196.7–990.9)
0.957 0.496
0.670 0.444 0.307
320 HONG et al. Our study is unique in that the majority of the patients were only supported with ECMO. However, when we subdivided this small group of patients by whether ECMO only or ECMO-VAD was used, there is a suggestion toward sensitization in the ECMO-VAD group compared with the ECMO-only group. Of the four patients who became sensitized post-MCS, only one of them was exclusively on ECMO. As there were 15 patients in the ECMO group, this is a 7% absolute risk of developing sensitization when supported with ECMO alone. In comparison, of the seven ECMO-VAD patients, three (43%) became sensitized post-MCS. The sensitization detection was also not related to the duration between MCS initiation and latest PRA as the time from MCS to latest PRA was actually longer for the ECMO group. These findings are similar to the study by Yang et al.10 which demonstrated that none of their 13 ECMO patients became sensitized, whereas two-thirds (6/9) of their VAD patients became sensitized. The study by O’Connor et al.9 looked exclusively at VAD patients and reported a sensitization rate of 35%. Based on these three pediatric studies, there may be a difference between the risk of sensitization between ECMO and VAD. Such a difference may reside in a difference in the blood–circuitry interface, management, or duration and degree of the proinflammatory state between ECMO and VAD even though ECMO patients in general are more critically ill and require more medical intervention. When evaluating risk factors for development of sensitization, there did not appear to be any increased risk in terms of duration of MCS, volume of transfused blood products, or type of transfused blood products. Interestingly, the nonsensitized patients received more transfusions. This is in contrast to previously published adult VAD studies, which reported an association between sensitization and platelet transfusions.4,11,12 The lack of association found in this study and the other two pediatric studies could be attributed to the small sample size or that younger children do not mount an anti-HLA response that is proportional to the amount of antigen exposure compared with adults. Finally, in the era of leukocyte reduction, the lack of association could also be explained by better filtration of leukocytes. Despite not finding an association between MCS duration and transfused blood products with sensitization, it is interesting that older age at MCS and a history of a homograft with prior surgeries were suggestive as risk factors for HLA sensitization based on the multivariable analysis. The older patients could have a more developed immune system compared with younger patients and mount a more robust immune response to HLA-antigen exposure. Prior studies have demonstrated that patients with a preserved allograft had a significantly increased risk of sensitization.13–16 Meyer et al.17 compared neonates with hypoplastic left heart syndrome who underwent the Norwood procedure with a cryopreserved allograft patch with neonates with transposition of the great arteries who underwent the arterial switch operation with no allograft material. Their study found the allograft group to be more likely to be sensitized. Therefore, it is possible that even without MCS, the patients in our study with a homograft would have been sensitized over time but that MCS accelerated the development of HLA antibodies. A more recent study looking at predictors of higher PRA at transplant found that having a VAD at transplant, congenital heart disease, and previous Norwood operation were the
three highest predictors.18 If indeed it is the homograft alone that predisposes our cohort to sensitization, then the likelihood of sensitization associated with MCS alone becomes even lower, 2/19 or 11% if we excluded all patients with a homograft. More specifically, if we only looked at the ECMO subgroup and excluded homograft patients, it would be 0% (0/13 patients). There are several limitations to this study. This is a small retrospective, single-center study in which we could not include all patients during this time period who were supported by MCS as there were ECMO patients who did not have baseline or post-MCS PRAs. Given the small sample size, we are limited in our statistical analysis and conclusions that can be drawn from it. We included those patients who did not have baseline pre-MCS PRAs if their post-MCS PRAs were negative and designated those patients as nonsensitized with the assumption that their pre-MCS PRAs were likely negative. However, we excluded patients who only had a positive PRA post-MCS, which could be underestimating the frequency of sensitization in this cohort. Although we were able to collect and analyze the effect of the total volumes of blood product transfusions on sensitization, we were unable to obtain donor exposures in our patient cohort. The exposure data may have been more reflective of the total spectrum of HLA-antigen exposure. Finally, the type of VAD in our patient cohort was predominantly the pneumatically driven pulsatile Berlin Heart EXCOR. Therefore, we cannot extend our conclusions onto those patients who developed sensitization while supported with the newer generation continuous flow devices. Despite these limitations, we were able to make some conclusions regarding the risk of HLA sensitization in children supported by MCS. Sensitization frequency in this largely ECMO cohort is low compared with adult or prior pediatric studies. This study is suggestive of a lack of association between blood product transfusions including platelet transfusions and sensitization, which contrasts with the adult data. Nevertheless, serial surveillance of PRAs in this patient population is important as PRAs can fluctuate during the MCS period. Finally, risk factors other than MCS itself, such as prior surgical repair with homograft material, could increase the risk of sensitization. References 1. Arnaoutakis GJ, George TJ, Kilic A, et al: Effect of sensitization in US heart transplant recipients bridged with a ventricular assist device: Update in a modern cohort. J Thorac Cardiovasc Surg 142: 1236–1245, 1245.e1, 2011. 2. Smedira NG, Hoercher KJ, Yoon DY, et al: Bridge to transplant experience: Factors influencing survival to and after cardiac transplant. J Thorac Cardiovasc Surg 139: 1295–1305, 1305. e1, 2010. 3. John R, Lietz K, Schuster M, et al: Immunologic sensitization in recipients of left ventricular assist devices. J Thorac Cardiovasc Surg 125: 578–591, 2003. 4. Massad MG, Cook DJ, Schmitt SK, et al: Factors influencing HLA sensitization in implantable LVAD recipients. Ann Thorac Surg 64: 1120–1125, 1997. 5. Leukocyte reduction and ultraviolet B irradiation of platelets to prevent alloimmunization and refractoriness to platelet transfusions. The Trial to Reduce Alloimmunization to Platelets Study Group. N Engl J Med 337: 1861–1869, 1997. 6. Chen JM, Richmond ME, Charette K, et al: A decade of pediatric mechanical circulatory support before and after cardiac transplantation. J Thorac Cardiovasc Surg 143: 344–351, 2012.
HUMAN LEUKOCYTE ANTIGEN SENSITIZATION
7. Barth E, Durand M, Heylbroeck C, et al: Extracorporeal life support as a bridge to high-urgency heart transplantation. Clin Transplant 26: 484–488, 2012. 8. Almond CS, Singh TP, Gauvreau K, et al: Extracorporeal membrane oxygenation for bridge to heart transplantation among children in the United States: Analysis of data from the Organ Procurement and Transplant Network and Extracorporeal Life Support Organization Registry. Circulation 123: 2975–2984, 2011. 9. O’Connor MJ, Menteer J, Chrisant MR, et al: Ventricular assist device-associated anti-human leukocyte antigen antibody sensitization in pediatric patients bridged to heart transplantation. J Heart Lung Transplant 29: 109–116, 2010. 10. Yang J, Schall C, Smith D, et al: HLA sensitization in pediatric pre-transplant cardiac patients supported by mechanical assist devices: The utility of Luminex. J Heart Lung Transplant 28: 123–129, 2009. 11. Drakos SG, Stringham JC, Long JW, et al: Prevalence and risks of allosensitization in HeartMate left ventricular assist device recipients: The impact of leukofiltered cellular blood product transfusions. J Thorac Cardiovasc Surg 133: 1612–1619, 2007. 12. Moazami N, Itescu S, Williams MR, Argenziano M, Weinberg A, Oz MC: Platelet transfusions are associated with the
321
development of anti-major histocompatibility complex class I antibodies in patients with left ventricular assist support. J Heart Lung Transplant 17: 876–880, 1998. 13. Shaddy RE, Hunter DD, Osborn KA, et al: Prospective analysis of HLA immunogenicity of cryopreserved valved allografts used in pediatric heart surgery. Circulation 94: 1063–1067, 1996. 14. Hooper DK, Hawkins JA, Fuller TC, Profaizer T, Shaddy RE: Panel-reactive antibodies late after allograft implantation in children. Ann Thorac Surg 79: 641–644, discussion 645, 2005. 15. Hoekstra F, Witvliet M, Knoop C, et al: Donor-specific anti-human leukocyte antigen class I antibodies after implantation of cardiac valve allografts. J Heart Lung Transplant 16: 570–572, 1997. 16. Hawkins JA, Breinholt JP, Lambert LM, et al: Class I and class II anti-HLA antibodies after implantation of cryopreserved allograft material in pediatric patients. J Thorac Cardiovasc Surg 119: 324–330, 2000. 17. Meyer SR, Campbell PM, Rutledge JM, et al: Use of an allograft patch in repair of hypoplastic left heart syndrome may complicate future transplantation. Eur J Cardiothorac Surg 27: 554–560, 2005. 18. Mahle WT, Tresler MA, Edens RE, et al; Pediatric Heart Transplant Study Group: Allosensitization and outcomes in pediatric heart transplantation. J Heart Lung Transplant 30: 1221–1227, 2011.
Pediatric Circulatory Support
Human Leukocyte Antigen Sensitization in Pediatric Patients Exposed to Mechanical Circulatory Support Borah J. Hong,* Meghan Delaney,†‡ Anthony Guynes,‡ Paul Warner,† David M. McMullan,§ Mariska S. Kemna,¶ Robert J. Boucek,¶ and Yuk M. Law¶
Human leukocyte antigen (HLA) sensitization of pediatric heart recipients increases their risk of rejection and graft loss. As more children are placed on mechanical circulatory support (MCS) as a bridge to transplant, the risk factors for development of sensitization warrant further study. A single-center retrospective review of 36 children who received MCS identified 22 patients supported with either extracorporeal membrane oxygenation (ECMO) (n = 15) or ECMO-ventricular assist device (VAD) (n = 7) with paired (pre-MCS/post-MCS) panel reactive antibodies (PRA) or only negative post-MCS PRAs. Four patients (18%) became sensitized post-MCS (one ECMO-only patient, three ECMO-VAD patients). No difference was found between sensitized and nonsensitized patients in terms of congenital heart disease versus primary cardiomyopathy (p = 0.096), duration of MCS (38 days vs. 14 days, p = 0.233), or volume of blood product transfusions (358.6 ml/kg vs. 612.7 ml/kg, p = not significant). By multivariable analysis, the association of sensitization with older age at MCS (p = 0.076) and history of homograft (p = 0.064) approached significance. Pediatric patients supported with MCS are at low risk of developing HLA sensitization. Diagnosis, MCS duration, and volume of transfused blood products do not appear to be associated with HLA sensitization, but there is a suggestion of an association with older age at MCS and history of a homograft. ASAIO Journal 2014; 60:317–321.
orthotopic heart transplantation (OHT), human leukocyte antigen (HLA) sensitization becomes a greater concern. Patients who are sensitized have to wait longer for OHT, have higher rates of primary graft dysfunction compared with nonsensitized patients, and are more likely to develop rejection.1,2 As evidenced in adult studies, HLA sensitization rates following ventricular assist device (VAD) implantation are as high as 66%.3 The increased sensitization rates in adults are attributed to the blood product transfusions, particularly platelet transfusions, and the inclusion of women with prior pregnancies.4,5 However, sensitization of pediatric patients is likely to be different from adults in terms of their inflammatory and immunological response to human tissue exposure, blood product requirements, and mode of MCS. In children, extracorporeal membrane oxygenation (ECMO) is an important and widely used mode of MCS either as a transitional mode of support to a durable VAD or used alone such as in postcardiotomy, with biventricular failure, or when a short period of support is expected.6–8 There have been only two single-center studies in children reporting HLA sensitization frequency in a cohort (n = 20) of predominantly VAD patients9 and a combination of ECMO (n = 13) and VAD patients (n = 9).10 As the cumulative experience is small, we sought to examine the frequency of HLA sensitization and the risk factors in a modern cohort of pediatric transplant candidates who required MCS.
Key Words: pediatric mechanical circulatory support, HLA sensitization
Methods This was a single-center retrospective cohort study of all patients placed on MCS with either ECMO or VAD between 2005 and 2011. Institutional review board approval was obtained. Inclusion criteria included having all prior surgeries performed at our institution, having complete records of blood product transfusions before and after MCS initiation, and having availability of panel reactive antibodies (PRA) at least post-MCS. Patient demographics and medical data including cardiac diagnosis, prior cardiac surgeries, prior homograft placement, MCS device type, and duration of support were obtained. Blood product transfusion volumes for packed red blood cells, platelets, plasma, and fresh whole blood for each patient before MCS and while on MCS were noted. All blood products were leukocyte reduced and irradiated. Human leukocyte antigen sensitization was defined as PRA greater than 10% for either HLA class I or class II. Panel reactive antibodies were measured by flow cytometry from 2005 to 2008 and subsequently by Luminex platform-based bead assay (FlowPRA Screening; One Lambda Inc., Canoga Park, CA). We collected
As more pediatric patients with end-stage heart failure are
placed on mechanical circulatory support (MCS) as a bridge to
From the *Division of Pediatric Cardiology, Texas Children’s Hospital, Baylor College of Medicine, Houston, Texas; †Puget Sound Blood Center, Seattle, Washington; ‡University of Washington School of Medicine, Seattle, Washington; §Division of Pediatric Cardiothoracic Surgery, Seattle Children’s Hospital, University of Washington, Seattle, Washington; and ¶Division of Pediatric Cardiology, Seattle Children’s Hospital, University of Washington, Seattle, Washington. Submitted for consideration September 2013; accepted for publication in revised form January 2014. Disclosure: The authors have no conflicts of interest to report. Reprint Requests: Borah J. Hong, Division of Pediatric Cardiology, Texas Children’s Hospital, 6621 Fannin Street, MC-19345-C, Houston, TX 77030-2399. Email: [email protected] Copyright © 2014 by the American Society for Artificial Internal Organs DOI: 10.1097/MAT.0000000000000053
317
318 HONG et al. all PRA results, pre-MCS, and post-MCS that were obtained and used the latest PRA before MCS and the latest PRA after MCS initiation to determine sensitization. We included those patients (10 of the 22 patients) who only had post-MCS PRAs if their PRA class I and class II were less than 10% as they were assumed to not have been sensitized before MCS if they were not sensitized following MCS. None of the patients underwent desensitization therapy before, during, or after MCS. Statistical Analysis Patient demographic data are expressed as median with interquartile range (IQR). Comparison of categorical variables between sensitized and nonsensitized patients was performed using the Fisher’s exact test where appropriate. Comparison of continuous variables between sensitized and nonsensitized patients was performed using the Mann–Whitney U test where appropriate. Multivariable analysis of predictor variables was performed by binary logistic regression. A p value of less than 0.05 was considered statistically significant. All statistical analysis was performed using IBM SPSS Statistics Version 18.0 (SPSS Inc., Chicago, IL). Results Patient Population There were 36 patients who were placed on MCS from 2005 to 2011. Three patients were excluded because they were placed on MCS following OHT, and 11 patients had insufficient data such as no post-MCS PRAs as they were not considered candidates for transplantation. Complete medical data were available for a cohort of 22 patients. Patient characteristics are shown in Table 1. Mechanical Circulatory Support ECMO was used in all 22 patients in the cohort. Seven of the 22 patients (32%) were additionally supported with VAD. One patient was initially placed on a Berlin Heart EXCOR (Berlin Heart, Inc., The Woodlands, TX) and subsequently transitioned to ECMO. Of the 21 patients who were initially placed on ECMO, five were subsequently transitioned to a left ventricular assist device (four onto Berlin Heart EXCOR, one onto MicroMed Debakey VAD [MicroMed Technology, Inc., Table 1. Patient Characteristics Number (%) or Median (Interquartile Range) Total patients Male Age at MCS (months) Weight at MCS (kg) Cardiomyopathy Congenital heart disease Homograft ECMO as initial support ECMO to LVAD ECMO to biventricular assist device LVAD to ECMO
22 (100) 10 (45) 7.5 (0.4–107.6) 7.0 (3.3–30.7) 10 (45%) 12 (55%) 3 (14%) 21 (95%) 5 (23%) 1 (5%) 1 (5%)
MCS, mechanical circulatory support; ECMO, extracorporeal membrane oxygenator; LVAD, left ventricular assist device.
Houston, TX]). One patient was transitioned from ECMO to biventricular support with the Berlin Heart EXCOR. Table 2 lists all the patients in the cohort and the type of MCS device received. Human Leukocyte Antigen Sensitization Of the 22 patients in this cohort, the median number of postMCS PRAs that were obtained was 4 (IQR, 2–6 PRAs). The median time from MCS initiation to latest PRA obtained was 56 days (IQR, 35.8–118.5 days). Of the cohort, four (18%) developed new anti-HLA antibodies post-MCS that were above a PRA of 10% and therefore considered sensitized (Table 2). Three of these patients had anti-HLA antibodies p re-MCS, although two were at low levels (PRA < 10%). All three had new or higher anti-HLA-II antibodies post-MCS, which allowed us to classify these patients as sensitized after MCS. The one patient (patient 1) who had a negative PRA before MCS but developed sensitization post-MCS had a history of congenital heart disease, but no homograft material was used for surgical repair. Upon further examination of the four patients who became sensitized, only one patient was exclusively supported on ECMO, whereas the other three patients were supported with ECMO-VAD. This estimates a sensitization rate of 7% (1/15) in the ECMO-only group and 43% (3/7) in the VAD group (p = 0.077). The median days to latest PRA in the ECMO group was 56 days (IQR, 37–150 days) and in the VAD group was 41 days (IQR, 28–101 days), which was not significantly different (p = not significant). Comparison of Sensitized Versus Nonsensitized Patients Comparison of the nonsensitized and sensitized patients following MCS demonstrated no significant difference in any of the covariates examined (Table 3) including the time from MCS to the latest PRA collected (p = 0.469). However, several suggestive risk factors are apparent. All four sensitized patients had a history of congenital heart disease (p = 0.096) and two of them had homografts. Of the three patients with a homograft, two were in the sensitized group (p = 0.073). Patients 2, 10, and 14 all had prior homografts, but patient 2 was not sensitized pre-MCS nor at the latest PRA post-MCS. However, when looking at sequential PRAs for patient 2, the patient did have an elevated PRA post-MCS with class I 0%/class II 19% on post-MCS day 87 but then had negative PRAs of 0%/0% by post-MCS day 217. Of the two other patients with homografts who became sensitized after MCS, patient 10 was highly sensitized pre-MCS (class I 96%/class II 0%), and patient 14 had some HLA antibodies pre-MCS (class I 8%/class II 0%). Transfused Blood Products Sensitized patients received a median of 358.6 ml/kg (IQR, 80.6–965.9 ml/kg) of total blood products before the latest PRA, whereas nonsensitized patients received a median of 612.7 ml/ kg (IQR, 196.7–990.9), which was not statistically different (p = 0.496). When looking specifically at different blood product types transfused, there was no difference between the sensitized and nonsensitized groups in terms of packed red blood cells, platelets, fresh-frozen plasma, cryoprecipitate, and fresh whole blood (Table 3). When comparing the patients based on
319
HUMAN LEUKOCYTE ANTIGEN SENSITIZATION Table 2. Pre-MCS and Post-MCS Panel Reactive Antibodies
Patient 1 2* 3 4 5 6 7 8 9 10* 11 12 13 14* 15 16 17 18 19 20 21 22
MCS Device
Pre-MCS HLA Class I/II
Post-MCS HLA Class I/II
Sensitized Pre-MCS
Sensitized Post-MCS
0/0 N/A N/A 0/0 0/4 N/A 0/0 21/22 N/A
15/0 0/0 0/0 0/0 0/16 0/0 0/0 0/0 0/0
No No No No No No No Yes No
Yes No No No Yes No No No No
96/0 0/0 N/A N/A 8/0 N/A N/A 0/0 0/0 0/0 N/A N/A 5/0
100/90 0/0 0/0 0/0 0/92 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0
Yes No No No No No No No No No No No No
Yes No No No Yes No No No No No No No No
ECMO → Berlin Heart LVAD ECMO ECMO ECMO ECMO → MicroMed Debakey LVAD ECMO ECMO ECMO ECMO → Berlin Heart biventricular assist device Berlin Heart LVAD → ECMO ECMO → Berlin Heart LVAD ECMO ECMO ECMO ECMO ECMO ECMO → Berlin Heart LVAD ECMO ECMO ECMO ECMO ECMO → Berlin Heart LVAD
*History of homograft. MCS, mechanical circulatory support; HLA, human leukocyte antigen; ECMO, extracorporeal membrane oxygenator; LVAD, left ventricular assist device; N/A, not applicable.
the type of MCS support of ECMO only versus ECMO-VAD, there was no significant difference (p = 0.503) in total blood products transfused with ECMO-only patients receiving a median of 592.9 ml/kg (IQR, 200.3–985.1 ml/kg), whereas the ECMO-VAD patients received a median of 546.6 ml/kg (IQR, 65.3–1,008.3 ml/kg). Multivariable regression analysis demonstrated that the duration of MCS and the volumes of individual blood products transfused including platelets, packed red blood cells, and plasma were not associated with development of sensitization. However, there was a suggestion of HLA sensitization with older age at MCS (p = 0.076) and a history of a surgical repair with a homograft (p = 0.064).
Discussion In this study, we observed 18% of the MCS cohort to be sensitized. There was a trend toward significance with older age, VAD, and homograft use, which is suggestive of these variables as risk factors. However, there was no association between sensitization and any specific blood products or transfusion volume. Unlike the adult population, there are only two published studies on children, and they report sensitization frequencies of 27% and 35%, which puts our present data in line with these similarly sized studies.9,10 In both previously published pediatric studies, there was no association between blood product transfusions and development of sensitization.
Table 3. Comparison of Sensitized Versus Nonsensitized Patients
Median age at MCS (months) (IQR) Duration of MCS (days) (IQR) Weight at MCS (kg) (IQR) Duration from MCS to final panel reactive antibodies (days) (IQR) Congenital heart disease Cardiomyopathy Homograft Yes No Median packed red blood cells transfused in ml/kg (IQR) Median platelets transfused in ml/kg (IQR) Median fresh-frozen plasma and cryoprecipitate transfused in ml/kg (IQR) Median fresh whole blood transfused in ml/kg (IQR) Median total blood products transfused in ml/kg (IQR) MCS, mechanical circulatory support; IQR, interquartile range.
Sensitized Patients, n = 4 (18%)
Nonsensitized Patients, n = 18 (82%)
p
138.9 (27.1–174.2) 38 (16.3–90.5) 36.5 (8.8–57.3) 56.5 (18.3–133.8)
4.7 (0.4–16.2) 14 (4.8–40.5) 5.6 (3.2–8.8) 56 (37–135.3)
0.233 0.233 0.148 0.469
8 (44%) 10 (56%)
0.096
4 (100%) 0 (0%) 2 (50%) 2 (50%) 155.4 (46.9–514.8) 97.1 (11.7–323.4) 34.0 (9.6–98.4)
1 (6%) 17 (94%) 239.5 (104.7–451.5) 148.6 (77.4–394.6) 69.4 (29.3–172.5)
0.073
0.0 (0.0–113.6) 358.6 (80.6–965.9)
0.0 (0.0–57.3) 612.7 (196.7–990.9)
0.957 0.496
0.670 0.444 0.307
320 HONG et al. Our study is unique in that the majority of the patients were only supported with ECMO. However, when we subdivided this small group of patients by whether ECMO only or ECMO-VAD was used, there is a suggestion toward sensitization in the ECMO-VAD group compared with the ECMO-only group. Of the four patients who became sensitized post-MCS, only one of them was exclusively on ECMO. As there were 15 patients in the ECMO group, this is a 7% absolute risk of developing sensitization when supported with ECMO alone. In comparison, of the seven ECMO-VAD patients, three (43%) became sensitized post-MCS. The sensitization detection was also not related to the duration between MCS initiation and latest PRA as the time from MCS to latest PRA was actually longer for the ECMO group. These findings are similar to the study by Yang et al.10 which demonstrated that none of their 13 ECMO patients became sensitized, whereas two-thirds (6/9) of their VAD patients became sensitized. The study by O’Connor et al.9 looked exclusively at VAD patients and reported a sensitization rate of 35%. Based on these three pediatric studies, there may be a difference between the risk of sensitization between ECMO and VAD. Such a difference may reside in a difference in the blood–circuitry interface, management, or duration and degree of the proinflammatory state between ECMO and VAD even though ECMO patients in general are more critically ill and require more medical intervention. When evaluating risk factors for development of sensitization, there did not appear to be any increased risk in terms of duration of MCS, volume of transfused blood products, or type of transfused blood products. Interestingly, the nonsensitized patients received more transfusions. This is in contrast to previously published adult VAD studies, which reported an association between sensitization and platelet transfusions.4,11,12 The lack of association found in this study and the other two pediatric studies could be attributed to the small sample size or that younger children do not mount an anti-HLA response that is proportional to the amount of antigen exposure compared with adults. Finally, in the era of leukocyte reduction, the lack of association could also be explained by better filtration of leukocytes. Despite not finding an association between MCS duration and transfused blood products with sensitization, it is interesting that older age at MCS and a history of a homograft with prior surgeries were suggestive as risk factors for HLA sensitization based on the multivariable analysis. The older patients could have a more developed immune system compared with younger patients and mount a more robust immune response to HLA-antigen exposure. Prior studies have demonstrated that patients with a preserved allograft had a significantly increased risk of sensitization.13–16 Meyer et al.17 compared neonates with hypoplastic left heart syndrome who underwent the Norwood procedure with a cryopreserved allograft patch with neonates with transposition of the great arteries who underwent the arterial switch operation with no allograft material. Their study found the allograft group to be more likely to be sensitized. Therefore, it is possible that even without MCS, the patients in our study with a homograft would have been sensitized over time but that MCS accelerated the development of HLA antibodies. A more recent study looking at predictors of higher PRA at transplant found that having a VAD at transplant, congenital heart disease, and previous Norwood operation were the
three highest predictors.18 If indeed it is the homograft alone that predisposes our cohort to sensitization, then the likelihood of sensitization associated with MCS alone becomes even lower, 2/19 or 11% if we excluded all patients with a homograft. More specifically, if we only looked at the ECMO subgroup and excluded homograft patients, it would be 0% (0/13 patients). There are several limitations to this study. This is a small retrospective, single-center study in which we could not include all patients during this time period who were supported by MCS as there were ECMO patients who did not have baseline or post-MCS PRAs. Given the small sample size, we are limited in our statistical analysis and conclusions that can be drawn from it. We included those patients who did not have baseline pre-MCS PRAs if their post-MCS PRAs were negative and designated those patients as nonsensitized with the assumption that their pre-MCS PRAs were likely negative. However, we excluded patients who only had a positive PRA post-MCS, which could be underestimating the frequency of sensitization in this cohort. Although we were able to collect and analyze the effect of the total volumes of blood product transfusions on sensitization, we were unable to obtain donor exposures in our patient cohort. The exposure data may have been more reflective of the total spectrum of HLA-antigen exposure. Finally, the type of VAD in our patient cohort was predominantly the pneumatically driven pulsatile Berlin Heart EXCOR. Therefore, we cannot extend our conclusions onto those patients who developed sensitization while supported with the newer generation continuous flow devices. Despite these limitations, we were able to make some conclusions regarding the risk of HLA sensitization in children supported by MCS. Sensitization frequency in this largely ECMO cohort is low compared with adult or prior pediatric studies. This study is suggestive of a lack of association between blood product transfusions including platelet transfusions and sensitization, which contrasts with the adult data. Nevertheless, serial surveillance of PRAs in this patient population is important as PRAs can fluctuate during the MCS period. Finally, risk factors other than MCS itself, such as prior surgical repair with homograft material, could increase the risk of sensitization. References 1. Arnaoutakis GJ, George TJ, Kilic A, et al: Effect of sensitization in US heart transplant recipients bridged with a ventricular assist device: Update in a modern cohort. J Thorac Cardiovasc Surg 142: 1236–1245, 1245.e1, 2011. 2. Smedira NG, Hoercher KJ, Yoon DY, et al: Bridge to transplant experience: Factors influencing survival to and after cardiac transplant. J Thorac Cardiovasc Surg 139: 1295–1305, 1305. e1, 2010. 3. John R, Lietz K, Schuster M, et al: Immunologic sensitization in recipients of left ventricular assist devices. J Thorac Cardiovasc Surg 125: 578–591, 2003. 4. Massad MG, Cook DJ, Schmitt SK, et al: Factors influencing HLA sensitization in implantable LVAD recipients. Ann Thorac Surg 64: 1120–1125, 1997. 5. Leukocyte reduction and ultraviolet B irradiation of platelets to prevent alloimmunization and refractoriness to platelet transfusions. The Trial to Reduce Alloimmunization to Platelets Study Group. N Engl J Med 337: 1861–1869, 1997. 6. Chen JM, Richmond ME, Charette K, et al: A decade of pediatric mechanical circulatory support before and after cardiac transplantation. J Thorac Cardiovasc Surg 143: 344–351, 2012.
HUMAN LEUKOCYTE ANTIGEN SENSITIZATION
7. Barth E, Durand M, Heylbroeck C, et al: Extracorporeal life support as a bridge to high-urgency heart transplantation. Clin Transplant 26: 484–488, 2012. 8. Almond CS, Singh TP, Gauvreau K, et al: Extracorporeal membrane oxygenation for bridge to heart transplantation among children in the United States: Analysis of data from the Organ Procurement and Transplant Network and Extracorporeal Life Support Organization Registry. Circulation 123: 2975–2984, 2011. 9. O’Connor MJ, Menteer J, Chrisant MR, et al: Ventricular assist device-associated anti-human leukocyte antigen antibody sensitization in pediatric patients bridged to heart transplantation. J Heart Lung Transplant 29: 109–116, 2010. 10. Yang J, Schall C, Smith D, et al: HLA sensitization in pediatric pre-transplant cardiac patients supported by mechanical assist devices: The utility of Luminex. J Heart Lung Transplant 28: 123–129, 2009. 11. Drakos SG, Stringham JC, Long JW, et al: Prevalence and risks of allosensitization in HeartMate left ventricular assist device recipients: The impact of leukofiltered cellular blood product transfusions. J Thorac Cardiovasc Surg 133: 1612–1619, 2007. 12. Moazami N, Itescu S, Williams MR, Argenziano M, Weinberg A, Oz MC: Platelet transfusions are associated with the
321
development of anti-major histocompatibility complex class I antibodies in patients with left ventricular assist support. J Heart Lung Transplant 17: 876–880, 1998. 13. Shaddy RE, Hunter DD, Osborn KA, et al: Prospective analysis of HLA immunogenicity of cryopreserved valved allografts used in pediatric heart surgery. Circulation 94: 1063–1067, 1996. 14. Hooper DK, Hawkins JA, Fuller TC, Profaizer T, Shaddy RE: Panel-reactive antibodies late after allograft implantation in children. Ann Thorac Surg 79: 641–644, discussion 645, 2005. 15. Hoekstra F, Witvliet M, Knoop C, et al: Donor-specific anti-human leukocyte antigen class I antibodies after implantation of cardiac valve allografts. J Heart Lung Transplant 16: 570–572, 1997. 16. Hawkins JA, Breinholt JP, Lambert LM, et al: Class I and class II anti-HLA antibodies after implantation of cryopreserved allograft material in pediatric patients. J Thorac Cardiovasc Surg 119: 324–330, 2000. 17. Meyer SR, Campbell PM, Rutledge JM, et al: Use of an allograft patch in repair of hypoplastic left heart syndrome may complicate future transplantation. Eur J Cardiothorac Surg 27: 554–560, 2005. 18. Mahle WT, Tresler MA, Edens RE, et al; Pediatric Heart Transplant Study Group: Allosensitization and outcomes in pediatric heart transplantation. J Heart Lung Transplant 30: 1221–1227, 2011.
