Study Rationale, Design and Pre-Transplant Alloantibody Status: A First Report of Clinical Trials in Organ Transplantation in Children-04 (CTOTC-04) in Pediatric Heart Transplantation

Abstract

Anti-HLA donor-specific antibodies are associated with worse outcomes following organ transplantation. Among sensitized pediatric heart candidates, requirement for negative donor-specific cytotoxicity crossmatch increases wait times and mortality. However, transplantation with positive crossmatch may increase post-transplant morbidity and mortality. We address this clinical challenge in a prospective, multi-center, observational cohort study of children listed for heart transplantation (CTOTC-04). Outcomes were compared among sensitized recipients who underwent transplant with positive crossmatch, non-sensitized recipients, and sensitized recipients without positive crossmatch. Positive crossmatch recipients received antibody removal and augmented immunosuppression, while other recipients received standard immunosuppression with corticosteroid avoidance. This first CTOTC-04 report summarizes study rationale and design, and reports on pre-transplant sensitization status using solid phase technology. Risk factors for sensitization were explored. Of 317 screened patients, 290 were enrolled and 240 underwent transplantation. Core laboratory evaluation demonstrated that over half of patients were anti-HLA sensitized. Over 80% of sensitized patients had class I (with or without class II) HLA antibodies, and one third of sensitized patients had at least one HLA antibody with median fluorescence intensity (MFI) ≥8000. Logistic regression models demonstrated male sex, weight, congenital heart disease history, prior allograft and ventricular assist device are independent risk factors for sensitization.

Introduction

The presence of ‘preformed’ antibodies specific against HLA prior to pediatric heart transplantation has been associated with high wait-list mortality, partially reflecting an historical requirement for negative donor-specific complement-dependent cytotoxicity crossmatch (CDC-XM) (1–4). By contrast, transplant across a positive CDC-XM may be associated with worse outcomes due to increased rejection, graft coronary vasculopathy, graft dysfunction and failure (3, 5–8). In recent years, select centers have offered transplantation across a positive CDC-XM to sensitized candidates with high risk of pre-transplant mortality (1, 9–12). Early results have been encouraging, even when retrospective CDC-XM was positive. However, optimal strategies for transplant and management of sensitized pediatric heart candidates remain unknown, in part due to challenges of drawing conclusions from small numbers of subjects in single center studies. Therefore, we developed a prospective, multi-center study to assess the impact of pre-transplant sensitization on pre- and post-transplant outcomes in pediatric heart candidates, focusing on the safety and efficacy of transplant across a positive CDC-XM and the impact of donor specific antibody (DSA) on post-transplant outcomes. The study was developed within the infrastructure of the National Institutes of Health (NIH)–sponsored Clinical Trials in Organ Transplantation in Children (CTOTC) program (www.ctotc.org). The aims of this first CTOTC-04 report are to: 1. Describe study rationale and design; 2. Report frequency and characterize pre-transplant alloantibodies using contemporary solid phase assays; and 3. Define risk factors for sensitization in the study population.

Methods

Study Design Overview

This is a prospective, observational, multi-center cohort study of pediatric heart transplant candidates. The primary objective is to compare clinical outcomes of sensitized recipients with positive CDC-XM at transplant (managed with a specialized treatment plan; see below), to non-sensitized recipients, or sensitized recipients without positive CDC-XM (managed with standard immunosuppression). The primary hypothesis is that highly sensitized candidates with positive CDC-XM can achieve first year outcomes similar to non-sensitized candidates when managed with perioperative antibody removal. Secondary objectives focused on outcomes based on assessment of DSA, independent of CDC-XM results.

In addition, a series of mechanistic studies was designed to evaluate graft accommodation in the setting of circulating DSA, and for development of a biomarker for antibody-mediated rejection (AMR) based on evaluation of levels of cell-bound complement activation products in peripheral blood. Full description of mechanistic studies is outside the scope of this report, and will be detailed in future publications (see also www.clinicaltrials.gov for summary of all study objectives and endpoints).

Study Organization and Participating Sites

The study was performed as part of a cooperative research program, CTOTC, sponsored by the National Institute of Allergy and Infectious Diseases (NIAID). CTOTC is an investigative consortium for conducting multi-institutional clinical and associated mechanistic studies leading to improved short- and long-term graft and patient survival in children who have undergone solid organ transplantation (www.ctotc.org). All study activities were approved by Institutional Review Boards at each of the participating centers. Statistical and clinical coordinating was performed by Rho (Chapel Hill, NC). The study was monitored by an external Data Safety and Monitoring Board appointed by NIAID. The study is registered at clinicaltrials.gov (NCT01005316).

Eight North American pediatric heart transplantation centers participated in this study: Boston Children’s Hospital; Children’s Healthcare of Atlanta; Children’s Hospital at Montefiore, Bronx, NY; Children’s Hospital of New York; Children’s Hospital of Philadelphia; Children’s Hospital of Pittsburgh of UPMC; Hospital for Sick Children, Toronto, ON, CA; St. Louis Children’s Hospital. Core laboratories for mechanistic studies were: Alloantibody (PI, A. Zeevi, University of Pittsburgh), Pathology (PI, A. Demetris, University of Pittsburgh), Endothelial Cell Culture (PI, T. Mohanakumar, Norton Thoracic Institute, Phoenix), Angiogenesis Factors (PI, D. Briscoe, Boston Children’s Hospital), Cell Bound Complement Activation Products (PI, J. Ahearn, Allegheny Health Network, Pittsburgh).

Study Duration

The planned study duration was four years, including three years of accrual and minimum of one year of follow-up. Enrollment commenced in February, 2011 and the last transplant occurred in December, 2013. Follow-up and data collection was completed on December 31, 2014.

Diagnosis of Sensitization and Cohort Assignment

Pre-transplant sensitization was first defined as positive Luminex LABScreen^® Mixed (LSM12) (One Lambda; Canoga Park, CA) for class I and/or class II with confirmation of the presence of HLA antibodies with median fluorescence intensity (MFI) ≥1,000 using Luminex LABScreen^® single antigen (LSA) beads (One Lambda; Canoga Park, CA). To minimize false negative results, the core laboratory subsequently performed LSA bead testing on all enrolled subjects, including those with negative Luminex LABScreen^® Mixed assay (see below). Local site results were used to assign non-sensitized patients to Cohort A and patients sensitized to class I and/or class II HLA to Cohort B. Cohort assignment status was used for evaluation of wait-list outcomes. Results from the Core Laboratory (with pre-transplant LSA testing of all subjects) were used to define subjects as ‘sensitized’ or ‘non-sensitized’, and are used for analysis of all post-transplant outcomes in future reports.

Overall study design is summarized in the flow diagram (Figure 1).

Figure 1.

CTOC-04 study design.

Study Procedures

Schedules for study visits, collection of blood and endomyocardial biopsy (EMB) samples are in online supplemental Tables S1a and S1b. Pre-transplant study visits occurred at enrollment (visit ‘PT1’) and six-month intervals until transplant (visit 0). Post-transplant study visits occurred at post-transplant days 1 and 7, months 1, 3, 6, 12, 24, and 36, and at initial post-transplant hospital discharge.

Participant Selection and Enrollment

Study inclusion and exclusion criteria are presented in Table 1. All patients evaluated for transplant at the eight participating centers were screened for enrollment. Limited screening data were collected on all patients <21 years-old who were listed for transplant. For patients not enrolled in the study, the reason was also collected. After listing, informed consent was obtained by the local investigator. Assent was obtained based on patient age and ability. Patients who met eligibility criteria and provided consent/assent were enrolled and assigned a unique study identification number. Data collected at enrollment included diagnosis, United Network for Organ Sharing (UNOS) status, donor and recipient demographics, medical/surgical history, blood type, use of artificial ventilation and/or inotropic agents, local alloantibody results, and history of sensitizing events. Blood specimens were obtained for baseline assessment (Table S1a).

Table 1.

Study inclusion/exclusion criteria.

INCLUSION CRITERIA Participants less than 21 years of age listed for heart transplantation at participating centers.

EXCLUSION CRITERIA Listed for multiple organ transplantation. Inability or unwillingness of the participant or parent/guardian to give written informed consent or comply with the study protocol. Condition or characteristics, which in the opinion of the investigator, makes the participant unlikely to complete at least one year of follow-up. Current participation in other research studies that would, or might, interfere with the scientific integrity or safety of the current study.

Pre-Transplant Follow-Up

Scheduled assessments prior to transplant included chart review for new sensitizing events, change in UNOS listing status, and recent laboratory results. ‘Unscheduled visits’ occurred with change in listing status or new sensitizing event. De-listed patients were followed for death or re-listing.

Day of Transplant

Prior to transplant, patients were assigned to cohort A or B (as described above) and blood specimens were obtained prior to any perioperative antibody removal (Table S1b). Donor selection was determined by the local clinical site as part of standard clinical care and there was no protocol requirement for negative virtual crossmatch (VXM) or CDC physical crossmatch. For the purposes of this study, transplantation across a positive VXM was defined as a sensitized subject with evidence of DSA (MFI≥1000) as determined by the core laboratory. Data on organs turned down for transplantation based on VXM results, or other criteria, was not collected for this study. T and B cell CDC-XM was performed at the time of transplantation according to standard of clinical care at each sites’ HLA laboratory, and were not modified for the purposes of this observational study. Sites used anti-human globulin augmented CDC-XM with and without dithiothreitol. Modified Amos method was used at three sites. Strength of CDC-XM positivity was reported and we used the final diagnostic determination from the local Histocompatibility Laboratory Directors’ report to define a positive CDC-XM. In subsequent analyses, CDC-XM was deemed to be a false positive if there was absence of DSA by the core laboratory. ABO incompatible transplant was permitted according to local center and UNOS criteria.

Post-Transplant Follow-up

Clinical data, blood for alloantibody analysis, EMB samples and mechanistic study samples were obtained at regularly scheduled post-transplant visits (Table S1b). Additional ‘unscheduled’ post-transplant visits occurred at the time of key clinical events, including: all EMB and coronary angiography procedures occurring at times other than scheduled visits, and all major post-transplant clinical events including hospitalization for any cause, rejection, infection, new-onset diabetes mellitus, post-transplant lymphoproliferative disease, re-transplantation, and death. Additional blood collection for alloantibody analysis was performed at all EMB and when rejection was clinically suspected in the absence of EMB.

Study Endpoints

The primary endpoint (Table 2) is a composite of the incidence of death, re-transplantation, and rejection with hemodynamic compromise at 12 months post-transplant.

Table 2.

Study endpoints.

PRIMARY ENDPOINT Composite incidence of death, re-transplantation or rejection with hemodynamic compromise at 12 months post-transplant.

SECONDARY ENDPOINTS Pre-Transplant Secondary Endpoints: Wait-list mortality. Time from listing to transplantation, death, or re-listing. Presence and quantification of anti-HLA IgG antibodies. Presence of anti-MICA antibodies.

Post-Transplant Secondary Endpoints: Individual components of the primary endpoint. Incidence of acute rejection (classified by type). Incidence of de novo allo-antibody production post-transplant and assessment of the impact on graft and participant outcomes. Specificity and time course of production of post-transplant de novo allo-antibodies and risk factors for their development. Incidence of hospitalization. Incidence of severe infections. Time to diagnosis of chronic rejection (graft coronary artery disease). Time to post-transplantation lymphoproliferative disorders. Time to new-onset diabetes mellitus.

Secondary endpoints are listed in Table 2, and are assessed at 12 months post-transplant as well as at subsequent follow-up up to three years post-transplant.

Alloantibody Core Laboratory

Samples of whole blood were sent at room temperature by overnight courier from clinical sites to the Alloantibody Core Laboratory, where serum was separated, aliquoted and stored at −80°C. Specimens were analyzed for the presence of HLA antibodies using pools of beads coated with purified class I or class II HLA or MICA antigens. LSM12 beads were used for the detection of HLA antibodies as well as antibodies to MICA, and LSA1 and LSA2 beads were used for class I and class II single-antigen analysis, respectively. Manufacturer’s positive cutoff for the ratio of LABScreen:Normalized Background (1.4) was used for screening. The reactions of a given HLA specificity were expressed as MFI, with MFI≥1000 considered positive. Single antigen testing and typing for MICA were not performed. All Core Laboratory antibody determinations were performed in batched fashion by a single senior technician (CB).

All results refer to those from the Alloantibody Core Laboratory unless explicitly stated as local laboratory results. Donor and recipient molecular typing was by low resolution molecular technique (including for HLA-C and –DQ), and was performed at local sites and entered into the CTOTC database. All donor/recipients were typed for DQB1* and when possible for DQA1*, and for DSA assignment we considered DQB1*/DQA1* pair based on DQA1* typing or DR/DQB1* association. There was variable availability of –DP molecular typing and HLA-DP antibodies were only assessed for DSA status when donor and recipient –DP typing were available. All DSA determinations were made by a single senior investigator (Core Director, A. Zeevi, University of Pittsburgh).

Immunosuppression Management and Clinical Care Guidelines

All participating centers followed the same standardized clinical care guidelines for immunosuppression, reflecting uniform contemporary clinical care (online supplement ‘Clinical Care Guidelines’). Patients received thymoglobulin induction therapy beginning within 24 hours of transplant and continuing for five total doses (total cumulative dose 7.5 mg/kg) and maintenance immunosuppression with tacrolimus and mycophenolate mofetil. Corticosteroids were used prior to each dose of thymoglobulin, and routine maintenance corticosteroids were not given. Target tacrolimus levels were 8–12 ng/ml for the first three months, 6–10 ng/ml from 3–12 months, and 5–8 ng/ml thereafter.

Sensitized patients at transplant underwent intra-operative plasma exchange (or immediate pre-operative plasmapheresis), with the goal of 1–3 times volume exchange. Centers could elect to omit this procedure in the setting of a negative VXM. Sensitized subjects with negative CDC-XM at transplant received the same initial post-transplant immunosuppression as non-sensitized subjects. Subsequent adjustments were made based on rejection status (online supplemental material). Positive CDC-XM results without evidence of DSA were considered false positive, and in these instances immunosuppression was adjusted at the discretion of the clinical site to that for CDC-XM negative subjects. Sensitized patients with DSA and positive CDC-XM underwent a 5-day course of post-transplant plasma exchange or plasmapheresis and received six total doses of intravenous immunoglobulin (IVIG) at monthly intervals (2g/kg/dose). Maintenance immunosuppression in patients with positive CDC-XM included routine maintenance corticosteroids for a minimum of six months in addition to tacrolimus and mycophenolate mofetil.

Participating centers followed local site clinical care guidelines for infectious disease prophylactic medications and surveillance monitoring for cytomegalovirus and Epstein-Barr virus. Concomitant use of other medications was defined by each center’s standard of care, with no prohibited medications.

Rejection Surveillance and Treatment

All patients underwent serial, invasive EMB for screening of rejection according to standard clinical care guidelines (online supplemental material). Rejection episodes were diagnosed based on local site EMB interpretation. Decisions to treat for rejection were made locally, and followed standard clinical care guidelines (online supplement material). Rejection on EMB was classified according to the standard (revised) criteria of the International Society for Heart and Lung Transplantation (ISHLT) (13, 14). Acute cellular rejection was defined as ISHLT grade 2R (two or more foci of mononuclear infiltrate with associated myocyte damage) or greater. Acute AMR was also defined according to the criteria of the ISHLT. Acute mixed rejection was defined according to the criteria of the ISHLT as evidence of both the presence of acute cellular rejection (ISHLT ≥ grade 2R) and the histopathologic and/or immunopathologic characteristics of AMR concurrently. Acute clinical rejection was defined as augmentation of immunosuppression based on clinical findings in the absence of histologic confirmation. Acute rejection with hemodynamic compromise was defined as echocardiographic fractional shortening of <26% with ≥5% decrease from last echocardiogram and/or new-onset heart failure, and was deemed severe when there was concomitant use of intravenous inotropic agents.

Statistical Considerations

We planned to enroll 370 subjects with an estimated 330 subjects undergoing transplantation. Based on prior literature, we estimated approximately 40 Cohort B subjects would have a positive CDC-XM. Initial power calculations were based on comparison of incidence rates of the composite primary endpoint between Cohort A (n=264) and Cohort B CDC-XM positive patients (n=40), and would have at least 82% power to detect an increase from 0.15 or less in cohort A to 0.35 or more in cohort B using a one-tailed, 0.05 level Fisher’s exact test.

For the current report, data are summarized using descriptive statistics for categorical (counts and percentages) and continuous (medians and IQRs) variables. Univariate comparisons were performed using Fisher’s exact test or chi-squared test for categorical variables and Wilcoxon rank-sum test for continuous variables. Logistic regression models were developed to estimate probabilities of sensitization with MFI≥1000 and MFI≥8000, separately. Risk factors were identified using backwards elimination variable selection strategy with α=0.10 threshold for model inclusion. All statistical analyses were performed using SAS Version 9.4 (SAS Institute, Inc., Cary, NC).

Results

Patient Characteristics

The consort diagram (Figure 2) follows the course of patients within the study. There were 317 screened patients at the eight participating centers, and 290 were enrolled. Compared to enrolled patients, screen failures were similar in age, sex, and history of previous heart surgery, but were more likely non-white/non-black race (33.3% vs. 6.6%, p=0.0005) and to have a diagnosis other than congenital heart disease (CHD) or cardiomyopathy (25.9% vs. 2.1%, p<0.0001). For enrolled patients (Table 3), median age at listing was 5 years (IQR 0–13) and median weight was 16.7 kg (IQR 7.9–46.3); 55.2% were male; 58.3% were white and 20.3% black; 49.7% had diagnosis of CHD and 48.3% cardiomyopathy. At time of listing, the majority of patients were hospitalized (69.3%), listed status 1A (64.8%) and had at least one prior sensitizing event (79.0%). Of 290 enrolled patients, 240 underwent transplant during the study period. Among those not transplanted, 20 died awaiting transplant. Transplanted patients were more likely to be older (6 vs. 2 years, p=0.0115), heavier (18.4 vs. 12.6 kg, p=0.0089), have diagnosis of cardiomyopathy (53.3 vs. 24.0%, p=0.0003), be listed status 1A at listing (67.9 vs. 50.0%, p=0.0197), not have had a prior sensitizing event (76.3 vs. 92.0%, p=0.0129), and were less likely to have been on extracorporeal membrane oxygenation support at listing (1.7 vs. 8.0%, p=0.0323) (Table 3).

Figure 2.

Consort diagram by core laboratory sensitization status (numbers based on actual enrollment).

Table 3.

Baseline characteristics by transplant status for enrolled subjects.

Characteristics, n (%)	Not Transplanted (n = 50)	Transplanted (n = 240)	p-Value	Total (n = 290)
Age at Listing in years, median (IQR)	2 (0–7)	6 (0–13)	0.0115	5 (0–13)
Weight at Listing in kg, median (IQR)	12.6 (6.0–22.5)	18.4 (8.5–49.0)	0.0089	16.7 (7.9–46.3)
Diagnosis			0.0003
Cardiomyopathy	12 (24.0)	128 (53.3)		140 (48.3)
Congenital Heart Disease	36 (72.0)	108 (45.0)		144 (49.7)
Other	2 (4.0)	4 (1.7)		6 (2.1)
Race			0.0689
White	36 (72.0)	133 (55.4)		169 (58.3)
Black or African American	5 (10.0)	54 (22.5)		59 (20.3)
Non-White/Non-Black	1 (2.0)	18 (7.5)		19 (6.6)
Unknown or Not Reported	8 (16.0)	35 (14.6)		43 (14.8)
Ethnicity			0.2901
Hispanic or Latino	5 (10.0)	30 (12.5)		35 (12.1)
Not Hispanic or Latino	39 (78.0)	161 (67.1)		200 (69.0)
Unknown or Not Reported	6 (12.0)	49 (20.4)		55 (19.0)
Male	32 (64.0)	128 (53.3)	0.1677	160 (55.2)
Blood Type			0.0931
A	12 (24.0)	77 (32.1)		89 (30.7)
AB	1 (2.0)	8 (3.3)		9 (3.1)
B	4 (8.0)	42 (17.5)		46 (15.9)
O	33 (66.0)	113 (47.1)		146 (50.3)
UNOS Status at Listing			0.0197
1A	25 (50.0)	163 (67.9)		188 (64.8)
1B	8 (16.0)	38 (15.8)		46 (15.9)
2	17 (34.0)	38 (15.8)		55 (19.0)
7	0	1 (0.4)		1 (0.3)
Prior Sensitizing Event	46 (92.0)	183 (76.3)	0.0129	229 (79.0)
Surgery	35 (70.0)	130 (54.2)	0.0397	165 (56.9)
Blood Transfusion	38 (76.0)	127 (52.9)	0.0027	165 (56.9)
VAD	10 (20.0)	55 (22.9)	0.6528	65 (22.4)
ECMO	8 (16.0)	28 (11.7)	0.3979	36 (12.4)
Any MCS	14 (28.0)	71 (29.6)	0.8229	85 (29.3)
Aortic or Pulmonary Allograft	10 (20.0)	37 (15.4)	0.4237	47 (16.2)
Prior Transplant	2 (4.0)	14 (5.8)	1.0000	16 (5.5)
Pregnancy	0	1 (0.4)	1.0000	1 (0.3)
Hospitalized at Listing	32 (64.0)	169 (70.4)	0.3708	201 (69.3)
ICU at Listing	23 (46.0)	110 (45.8)	0.9828	133 (45.9)
Ventilator at Listing	8 (16.0)	36 (15.0)	0.8577	44 (15.2)
ECMO at Listing	4 (8.0)	4 (1.7)	0.0323	8 (2.8)
VAD at Listing	2 (4.0)	28 (11.7)	0.1054	30 (10.3)
MCS at Listing	5 (10.0)	32 (13.3)	0.5204	37 (12.8)

IQR, interquartile range; UNOS, United Network for Organ Sharing; VAD, ventricular assist device; ECMO, extracorporeal membrane oxygenation; MCS, mechanical circulatory support; ICU, intensive care unit

Sensitization Status

Samples were available at the Alloantibody Core Laboratory for 272 of 290 patients at enrollment and 237 of 240 patients at transplant. Sensitization at enrollment occurred in 156 of 272 patients (57.4%), and at transplant in 129 of 237 patients (54.4%). In 62 transplanted patients, there was only one core laboratory sample received at enrollment and prior to transplant, and 36 (58.1%) of these patients were sensitized. This reflects the short time between listing and transplant for many patients. Of the remaining 175 transplanted patients, 97 (55.4%) were sensitized at enrollment and 93 (53.1%) were sensitized at transplant. However, 21 patients crossed over from sensitized to non-sensitized and 17 crossed over from non-sensitized to sensitized between enrollment and transplant. There were 76 patients who were sensitized and 61 who were non-sensitized on both samples (i.e. at enrollment and at transplant).

Distribution of HLA antibody class and strength (based on MFI) in transplanted patients is shown in Table 4. There was little difference between enrollment and transplant samples. In both cases, over 80% of patients had class I HLA antibodies (± class II). Almost half of patients had maximum MFI in the 1000–3999 range, but approximately one-third had at least one HLA antibody with peak MFI≥8000. The majority (30 of 38, 78.9%) of the patients that had sensitization status ‘crossover’ between enrollment and transplant had peak MFI in the lowest range (1000–3999).

Table 4.

Core laboratory antibody data for samples obtained at the time of transplantation.

Characteristics, n (%)	Sensitized (n = 129)
Antibody Class
Class I Only	56 (43.4)
Class II Only	24 (18.6)
Class I and II	49 (38.0)
Class I (with or without Class II)	105 (81.4)
Class II (with or without Class I)	73 (56.6)
MICA Antibody
Negative	110 (85.3)
Positive	19 (14.7)
Presence of at Least 1 Antibody in MFI Threshold
1000–3999	122 (94.6)
4000–7999	54 (41.9)
>8000	42 (32.6)
Maximum MFI Threshold
1000–3999	64 (49.6)
4000–7999	23 (17.8)
>8000	42 (32.6)
Maximum MFI Threshold for Class I Antibodies
1000–3999	56 (43.4)
4000–7999	20 (15.5)
>8000	29 (22.5)
No Class I Antibodies	24 (18.6)
Maximum MFI Threshold for Class II Antibodies
1000–3999	42 (32.6)
4000–7999	8 (6.2)
>8000	23 (17.8)
No Class II Antibodies	56 (43.4)
Calculated PRA per UNOS Calculator, mean ± standard deviation
>1000	41.3 ± 35.76
>4000	23.8 ± 35.02
>8000	15.4 ± 30.38

MICA, major histocompatibility complex class I-related chain A; MFI, median fluorescence intensity; PRA, panel reactive antibody; UNOS, United Network for Organ Sharing

The breadth of anti-HLA was assessed by calculated panel reactive antibody (cPRA), utilizing the UNOS cPRA calculator (https://www.unos.org/transplantation/allocation-calculators/) at the standard study definition of sensitization (MFI≥1000), and also with MFI cutoffs of ≥4000 and ≥8000 (Table 4).

Antibody against MICA antigens was identified in 25 patients at enrollment (9.2%) and transplant (10.5%), respectively. This included 11.5% of the sensitized and 6.0% of the non-sensitized patients at enrollment and 14.7% of the sensitized and 5.6% of the non-sensitized patients at transplant.

Overall, the proportion of sensitized subjects was similar between the local and core laboratories (59.6% vs. 54.4%, respectively). Of 97 non-sensitized patients by local laboratory testing, 70 (72.2%) were non-sensitized and 27 (27.8%) were sensitized by core laboratory testing (no available core laboratory sample in 3 patients). Of 143 sensitized patients by local laboratory testing, 102 (71.3%) were sensitized and 38 (26.6%) non-sensitized by core laboratory testing (no available core laboratory sample in 3 patients). Overall, there was discordance between local and core laboratories in 27.4% patients. Of the discordant samples, sensitization was associated with low peak MFI (1000–3999) in the majority (86.8%) of cases.

Risk Factors for Sensitization

Patient characteristics for enrollment and transplant samples stratified by sensitization status are shown in Table 5. Univariable analyses demonstrated sensitization at enrollment associated with heavier weight (20.5 vs. 15.6 kg, p=0.0287), black race (25.0 vs. 15.5%, p=0.0087), male sex (61.5 vs. 46.6%, p=0.0140), being listed as an out-patient (35.3 vs. 24.1%, p=0.0489), prior heart surgery (66.0 vs. 45.7%, p=0.0008), blood transfusion (64.1 vs. 45.7%, p=0.0025), aortic or pulmonary allograft placement (21.8 vs. 9.5%, p=0.0069), and heart transplant (9.6 vs. 0.9%, p=0.0024). When examining sensitization status at transplant, the significant associations remained the same except that race was no longer significant and CHD diagnosis (52.7 vs. 34.3%, p=0.0047) was associated with sensitization. Prior sensitizing events were more common in patients with CHD compared to cardiomyopathy (99.2% vs. 59.3%). Strength of sensitization (peak MFI) differed between cardiomyopathy and CHD patients. The majority of sensitized cardiomyopathy patients had peak MFI 1000–3999 (62.9%), which was true of only 36.6% of sensitized CHD patients.

Table 5.

Baseline patient characteristics by core laboratory sensitization status for samples obtained at study enrollment (a) and at the time of transplantation (b).

Characteristics, n (%)	a: Enrollment Sample				b: Transplanted Sample
	Non-Sensitized (n=116)	Sensitized (n = 156)	p-Value	Total (n = 272)	Non-Sensitized (n=108)	Sensitized (n = 129)	p-Value	Total (n = 237)
Listing Characteristics
Age at Listing in years, median (IQR)	4 (1–13)	7 (1–14)	0.0849	6 (1–14)	5 (1–13)	8 (1–14)	0.0675	6 (1–14)
Weight at Listing in kg, median (IQR)	15.6 (6.9–40.2)	20.5 (9.8–50.4)	0.0287	18.2 (8.5–47.8)	15.5 (6.8–40.1)	21.5 (10.2–52.9)	0.0057	18.4 (8.5–49.0)
Diagnosis			0.1854				0.0047
Cardiomyopathy	64 (55.2)	71 (45.5)		135 (49.6)	70 (64.8)	58 (45.0)		128 (54.0)
Congenital Heart Disease	51 (44.0)	80 (51.3)		131 (48.2)	37 (34.3)	68 (52.7)		105 (44.3)
Other	1 (0.9)	5 (3.2)		6 (2.2)	1 (0.9)	3 (2.3)		4 (1.7)
Race			0.0087				0.1230
White	71 (61.2)	86 (55.1)		157 (57.7)	62 (57.4)	70 (54.3)		132 (55.7)
Black or African American	18 (15.5)	39 (25.0)		57 (21.0)	20 (18.5)	34 (26.4)		54 (22.8)
Non-White/Non-Black	13 (11.2)	5 (3.2)		18 (6.6)	11 (10.2)	6 (4.7)		17 (7.2)
Unknown or Not Reported	14 (12.1)	26 (16.7)		40 (14.7)	15 (13.9)	19 (14.7)		34 (14.3)
Ethnicity			0.0886				0.5093
Hispanic or Latino	11 (9.5)	24 (15.4)		35 (12.9)	12 (11.1)	18 (14.0)		30 (12.7)
Not Hispanic or Latino	87 (75.0)	98 (62.8)		185 (68.0)	74 (68.5)	85 (65.9)		159 (67.1)
Unknown or Not Reported	18 (15.5)	34 (21.8)		52 (19.1)	22 (20.4)	26 (20.2)		48 (20.3)
Male	54 (46.6)	96 (61.5)	0.0140	150 (55.1)	44 (40.7)	82 (63.6)	0.0005	126 (53.2)
Blood Type			0.7305				0.8506
A	33 (28.4)	47 (30.1)		80 (29.4)	37 (34.3)	38 (29.5)		75 (31.6)
AB	5 (4.3)	4 (2.6)		9 (3.3)	3 (2.8)	5 (3.9)		8 (3.4)
B	21 (18.1)	23 (14.7)		44 (16.2)	19 (17.6)	22 (17.1)		41 (17.3)
O	57 (49.1)	82 (52.6)		139 (51.1)	49 (45.4)	64 (49.6)		113 (47.7)
UNOS Status at Listing			0.1700				0.1161
1A	79 (68.1)	99 (63.5)		178 (65.4)	79 (73.1)	83 (64.3)		162 (68.4)
1B	21 (18.1)	21 (13.5)		42 (15.4)	18 (16.7)	19 (14.7)		37 (15.6)
2	16 (13.8)	35 (22.4)		51 (18.8)	11 (10.2)	26 (20.2)		37 (15.6)
7	0	1 (0.6)		1 (0.4)	0	1 (0.8)		1 (0.4)
Prior Sensitizing Event	86 (74.1)	130 (83.3)	0.0636	216 (79.4)	77 (71.3)	105 (81.4)	0.0666	182 (76.8)
Surgery	53 (45.7)	103 (66.0)	0.0008	156 (57.4)	45 (41.7)	84 (65.1)	0.0003	129 (54.4)
Blood Transfusion	53 (45.7)	100 (64.1)	0.0025	153 (56.3)	48 (44.4)	78 (60.5)	0.0138	126 (53.2)
VAD	27 (23.3)	36 (23.1)	0.9693	63 (23.2)	24 (22.2)	31 (24.0)	0.7425	55 (23.2)
ECMO	10 (8.6)	24 (15.4)	0.0953	34 (12.5)	8 (7.4)	20 (15.5)	0.0545	28 (11.8)
Any MCS	34 (29.3)	48 (30.8)	0.7954	82 (30.1)	30 (27.8)	41 (31.8)	0.5026	71 (30.0)
Aortic or Pulmonary Allograft	11 (9.5)	34 (21.8)	0.0069	45 (16.5)	7 (6.5)	30 (23.3)	0.0004	37 (15.6)
Prior Transplant	1 (0.9)	15 (9.6)	0.0024	16 (5.9)	2 (1.9)	12 (9.3)	0.0154	14 (5.9)
Pregnancy	1 (0.9)	0	0.4265	1 (0.4)	1 (0.9)	0	0.4557	1 (0.4)
Hospitalized at Listing	88 (75.9)	101 (64.7)	0.0489	189 (69.5)	84 (77.8)	84 (65.1)	0.0326	168 (70.9)
ICU at Listing	59 (50.9)	64 (41.0)	0.1070	123 (45.2)	55 (50.9)	54 (41.9)	0.1631	109 (46.0)
Ventilator at Listing	20 (17.2)	20 (12.8)	0.3086	40 (14.7)	19 (17.6)	17 (13.2)	0.3457	36 (15.2)
ECMO at Listing	2 (1.7)	4 (2.6)	1.0000	6 (2.2)	1 (0.9)	3 (2.3)	0.6278	4 (1.7)
VAD at Listing	13 (11.2)	17 (10.9)	0.9358	30 (11.0)	11 (10.2)	17 (13.2)	0.4771	28 (11.8)
MCS at Listing	15 (12.9)	20 (12.8)	0.9785	35 (12.9)	12 (11.1)	20 (15.5)	0.3244	32 (13.5)
Transplant Characteristics1
Age at Transplant in years, median (IQR)					5 (0–13)	8 (1–14)	0.0761	6 (1–14)
Weight at Transplant in kg, median (IQR)					15.6 (8.0–40.1)	25.2 (10.7–51.7)	0.0129	20.6 (9.2–50.0)
UNOS Status at Transplant							0.5744
1A					92 (85.2)	112 (86.8)		204 (86.1)
1B					12 (11.1)	10 (7.8)		22 (9.3)
2					4 (3.7)	7 (5.4)		11 (4.6)
Hospitalized at Transplant					84 (77.8)	89 (69.0)	0.1292	173 (73.0)
ICU at Transplant					48 (44.4)	53 (41.1)	0.6025	101 (42.6)
Ventilator at Transplant					17 (15.7)	16 (12.4)	0.4598	33 (13.9)
ECMO at Transplant					6 (5.6)	3 (2.3)	0.3069	9 (3.8)
VAD at Transplant					18 (16.7)	26 (20.2)	0.4915	44 (18.6)
MCS at Transplant					24 (22.2)	29 (22.5)	0.9621	53 (22.4)

IQR, interquartile range; UNOS, United Network for Organ Sharing; VAD, ventricular assist device; ECMO, extracorporeal membrane oxygenation; MCS, mechanical circulatory support; ICU, intensive care unit

¹Data only available for transplanted recipients.

Using pre-transplant samples, multivariable logistic regression models demonstrated sensitization to be associated with prior transplant (OR 11.009, p=0.0283), prior allograft placement (OR 3.438, p=0.0179), male sex (OR 2.540, p=0.0016), diagnosis of CHD (OR 2.179, p=0.0216), and weight at transplant (OR 1.013, p=0.0345). In a second multivariable model of sensitized patients with HLA antibodies of high strength (peak MFI≥8000), identified risk factors were prior transplant (OR 58.410, p<0.0001), diagnosis of CHD (OR 6.288, p=0.0008), history of ventricular assist device (VAD) (OR 4.282, p=0.0047), and prior allograft (OR 3.004, p=0.0229).

Discussion

Sensitization among pediatric patients considered for heart transplantation presents a significant challenge, and controversy exists regarding optimal strategy for pre- and post-transplant management. Furthermore, there is no consensus on criteria for donor selection, even when candidate antibody status is known and donor HLA typing is available. Almost all prior studies in this area are limited to single center reports or registry data (3, 4, 8, 15–22). There have been no prior prospective, multi-institutional studies that address these issues in the pediatric heart transplant population. Furthermore, the methods of detection of anti-HLA have evolved from the initial cell based CDC-PRA assays to newer solid-phase, flow-bead techniques such as the Luminex^® assay that mix patient serum with latex beads bound with single HLA. This has enabled the identification of individual anti-HLA specificities with much greater sensitivity, and determination of relative strength, enhancing accuracy of VXM prior to organ acceptance (23, 24). This approach has proven to be a strong predictor of a positive CDC-XM at transplant (23, 25), and is used practically by many institutions to guide donor organ acceptance (26). However, this may come at the cost of increased wait times and wait list mortality for sensitized candidates (2, 9).

Several centers in the CTOTC-04 consortium had previously noted through retrospective evaluation that transplant in the setting of a positive donor-specific CDC-XM can be achieved with similar early post-transplant survival (9–12). This led us to hypothesize that sensitized pediatric heart candidates with positive CDC-XM can achieve outcomes similar to non-sensitized candidates when managed with perioperative antibody removal and augmented immunosuppression. We further hypothesized that use of donor organs irrespective of VXM and CDC-XM result would reduce wait-list mortality. In this first report from the CTOTC-04 study, we aimed to outline in detail our study rationale and design for future reference, report the frequency and characterization of pre-transplant alloantibodies using contemporary solid phase assay, and define risk factors for sensitization in the study population. An accompanying report describes the primary study clinical outcomes (27).

Frequency of Pre-Transplant Sensitization

Previous reports in pediatric patients found pre-transplant sensitization rates between 14 and 23% using a definition of T and/or B cell CDC-PRA > 10% (1, 3, 26). In our population, defining sensitization as positive Luminex^® screen and detection of at least one anti-HLA with MFI≥1000 identified by single antigen bead testing, we identified much higher frequency of sensitization. Over half of candidates were sensitized, and among sensitized candidates approximately one-third had one or more anti-HLA with peak MFI≥8000, a value considered to represent concern for early poor graft outcomes in other transplant populations (28–30).

The reasons for the unexpectedly high number of sensitized patients are likely several. We anticipated higher prevalence of sensitized patients compared to historical cohorts evaluated by CDC-PRA since the solid phase platforms are known to be more sensitive. Furthermore, with no identified singular significant cutoff for sensitization defined in the literature (31), we chose to err on the side of being more inclusive and define sensitization at an MFI cutoff of 1000 as one of the future directions of this consortium is to determine the importance of low strength pre-formed alloantibodies on outcomes after transplant. In particular, it will be important to define the appropriate criteria for excluding donors during the performance of a VXM, especially when identified DSA are present at low strength. At low MFI, it may be difficult to distinguish ‘false positive’ findings from ones that are real but potentially clinically unimportant. Also contributing to high rates of sensitization is the large proportion of patients in this cohort with CHD (48.2%), with the vast majority of these patients having had a prior sensitizing event, which was not surprising given that the majority of these patients underwent previous surgical palliation(s). However, the number of cardiomyopathy patients who had a prior sensitizing event was surprisingly high (59.3%), reflecting, in part, the increasing use of VAD as a bridge to transplant in pediatrics, which accounted for over half of these events.

Risk Factors for Sensitization

While our results demonstrate a greater frequency of pre-transplant sensitization than prior studies, risk factors for sensitization mostly remain the same. These include CHD with prior cardiac surgery (especially allograft placement), male sex, weight at transplant, and VAD usage. Prior heart transplant was the strongest predictor of sensitization. There were only minor differences in the models depending on the MFI threshold used to define sensitization, with male sex and weight identified as risk factors for sensitization at ≥1000 MFI but not at higher strength (≥8000 MFI), whereas VAD use was only a risk factor for sensitization at higher strength anti-HLA.

Study Limitations

We enrolled fewer patients than originally planned, but this study still represents the largest cohort of pediatric heart transplant candidates enrolled in a prospective study. The final enrollment will influence the power of the primary outcome (27), but still provides a rich dataset for analysis of the frequency, antibody characteristics, and risk factors for pre-transplant sensitization in the contemporary era. We recognize that antibody characterization is limited to class and strength in this initial analysis. Predictive value for clinical outcomes may be enhanced by more detailed characterization of pre-transplant antibody such as analysis of immunoglobulin class and isotype as well as subtype, dilutional analysis of sera, and complement fixing ability. Such studies are currently underway. This study only addresses pre-transplant sensitization status. Future studies will also focus on the predictive value of detection of de novo DSA post-transplant in the large cohort of patients who achieved transplantation within CTOTC-04.

Conclusion

Using solid-phase assay, sensitization against HLA is more common in children requiring heart transplantation than previously reported. Traditional risk factors associated with sensitization are identified in the current study, though we have further modeled risk factors based on HLA antibody strength. Longer-term follow-up of this large group of well-phenotyped patients offers a unique opportunity to understand the role played by these alloantibodies on late outcomes. Future analyses will provide invaluable information as to risk stratification, including how to optimally perform donor organ selection and the role played by VXM.

Supplementary Material

Supp TableS1. Table S1.

(A) Pre-Transplant Schedule of Events (for all enrolled subjects). (B) Transplant and Follow-Up Schedule of Events (for all cohort A and B subjects).

Abbreviations

AMR

antibody-mediated rejection

CDC-XM

complement-dependent cytotoxicity crossmatch

CHD

congenital heart disease

cPRA

calculated panel reactive antibody

CTOTC

Clinical Trials in Organ Transplantation in Children

EMB

endomyocardial biopsy

HLA

Human Leukocyte Antigen

IRB

Institutional Review Board

ISHLT

International Society for Heart and Lung Transplantation

IVIG

intravenous immunoglobulin

LSA

Luminex LABScreen^® Single Antigen

LSM12

Luminex LABScreen^® Mixed

MFI

median fluorescence intensity

MICA

major histocompatibility complex class I-related chain A

NIAID

National Institute of Allergy and Infectious Diseases of the National Institutes of Health

UNOS

United Network for Organ Sharing

VAD

ventricular assist device

VXM

virtual crossmatch