Compatibility of ABO blood groups is usually regarded as essential for solid organ transplants (1). There are occasional reports of ABO-incompatible renal and liver transplants (1–6), usually after preparation with techniques such as plasmapheresis or splenectomy (5, 6). There is one collected series of ABO-mismatched cardiac transplants, all of which were performed inadvertently. Despite aggressive treatment including retransplantation, only two of the eight patients that were described survived (7).
The availability of suitable donors is a major factor limiting heart transplantation in infants. At all ages, blood group O recipients have significantly longer waiting times and a higher mortality on the waiting list because it has generally been accepted that they can only receive hearts from group O donors, whereas group O donor hearts may be used for any otherwise suitable recipient (Table 1). Moreover, figures in the United Kingdom suggest that a significant proportion of donor hearts from infants and children go unused. In many cases, these are from a donor of blood group A, B, or AB for whom no recipient is identified during the period when the heart is available.
A group based at the Hospital for Sick Children in Toronto has recently described a series of successful ABO-incompatible transplants in infants (8). The program was based primarily (1) on the fact that neonates do not produce anti-A and anti-B antibodies (isohemagglutinins) until they have been exposed to the carbohydrate antigens A and B that are also present on bacteria colonizing the gut after birth (9) and (2) on the immaturity of the infant immune system, which has, for instance, lower levels of complement activity (10).
The use of ABO-incompatible donors considerably increases the opportunity for transplant infant recipients, as has been demonstrated by the Toronto group. After their experience, we embarked on a similar program of ABO-incompatible heart transplantation.
PATIENTS AND METHODS
Since January 2000, we have adopted a policy of undertaking ABO-mismatched heart transplantation in infants, or in children with low levels of isohemagglutinins, for whom a suitable ABO-compatible organ was not available. The protocol was drawn up on the advice of the Hospital for Sick Children in Toronto, and all parents gave written informed consent. To date, we have considered five infants and performed three ABO-incompatible cardiac transplants in infants aged 2, 3, and 21 months (Table 2).
Blood Group and Antibody Monitoring
The blood type of donors was provided by the referring donor hospital. Serum from all infants was tested for the presence of anti-A and anti-B antibodies by means of standard agglutination tests. Then, 0.2 mL of the patient’s serum diluted with saline in a ratio ranging from 1:1 to 1:512 was mixed with 0.1 mL of a suspension of 3% to 5% erythrocytes of blood group A or B and incubated for 40 min at room temperature (21–23°C). Agglutination was assessed macroscopically without prior centrifugation by the “tip-and-roll” technique and negatives confirmed microscopically. Positives were graded from 4+ to 1+, the endpoint being the presence of easily discernible agglutination on microscopic examination as judged by experienced operators. All tests were performed by experienced staff. Appropriate controls were used to exclude anti-HI and other cold reacting antibodies. The test will detect both immunoglobulin (Ig) M and IgG antibodies but is not as sensitive as an indirect antiglobulin test for detection of IgG (11).
Titers of serum isohemagglutinins were monitored twice weekly while recipients were awaiting heart transplantation. After a decision was made to transplant an ABO-incompatible heart, titers were measured during the operation after each plasma exchange, before the release of the aortic cross-clamp, and every 6 hr for 4 days after transplantation. Testing was subsequently repeated daily for 2 weeks, weekly for 2 months, and thereafter monthly for 12 months.
When the possibility of ABO-incompatible heart transplantation was under consideration, blood products transfused into the patient were chosen to minimize transfusion of blood donor isohemagglutinins.
Blood products were chosen with the intent of minimizing transfusion of relevant isohemagglutinins in donor plasma and avoiding transfusion of platelets that would carry donor A or B antigens. The bypass circuit was primed with donor group or group AB plasma. Transfused red blood cells were of recipient blood group, from donors known to be negative for high-titer A or B antibodies, and washed and resuspended in optimal additive solution consisting of saline, adenine, glucose, and mannitol (SAG-M; Baxter Healthcare Ltd., Newbury, United Kingdom). When platelet transfusion was required, platelets of recipient blood group were washed and resuspended in platelet optimal additive solution (T-sol; Baxter Health-care) (Table 3).
Technique of plasma exchange was different in each case. In case 1, blood equal in volume to the estimated circulating blood volume was removed from the infant during the initiation of bypass through the venous return line and replaced with an equal volume of washed red cells from the blood bank and group AB plasma to maintain normovolemia. The recovered infant’s blood was separated into plasma and red blood cell fractions by the local transfusion center. The plasma was discarded and the washed red blood cell fraction was returned through the bypass circuit. This procedure was repeated three times until the concentrations of circulating antibodies to blood group antigens were reduced to undetectable levels. Absence of measurable isohemagglutinins was confirmed before release of the aortic cross-clamp.
In cases 2 and 3, at initiation of hypothermic bypass, the infant was effectively exsanguinated through the venous line of the cardio-pulmonary bypass circuit and the blood was discarded. Normovolemia was then reinstated by transfusion of washed red blood cells and appropriate plasma in the pump prime through the aortic return line. This process was completed over a 5-min period, and full cardiopulmonary bypass was immediately initiated. Serum isohemagglutinins were confirmed to have been reduced to undetectable levels before release of the aortic cross-clamp.
The immunosuppressive regimen consisted of preoperative cyclosporine 4 mg/kg and azathioprine 4 mg/kg administered orally. At the time of release of cross-clamp, 30 mg/kg of methylprednisolone was administered intravenously. Immediately after the operation, they were administered 10 mg/kg of anti-human thymocyte equine immunoglobulin (Lymphoglobuline; IMTIX-SangStat, Lyon, France) intravenously. This was then administered daily for a total of 5 days at varying doses to maintain peripheral blood CD3+ lymphocytes measured by flow cytometry (FACScan; Becton Dickinson, Franklin Lakes, NJ) at less than 50×109/L. Primary triple immunosuppression consisted of cyclosporine (Neoral; Novartis, Basel, Switzerland) (initially administered intravenously [2–4 mg/kg/day in two divided doses] and then orally [8 –10 mg/kg/day in two divided doses]; with a target 12-hr trough plasma concentration of 300 –350 ng/mL, according to enzyme-linked immunoassay [EMIT assay; Dade-Behring Ltd., Milton Keynes, United Kingdom]), azathioprine (2–4 mg/kg/ day administered orally), and methylprednisolone (3 mg/kg administered every 8 hr for three doses only).
The dose of cyclosporine was decreased over time (target 12-hr trough plasma concentration 1 year after transplantation of 150 –200 ng/mL), to maintain immunosuppression at the lowest level required to prevent rejection. None of our infants that underwent ABO-incompatible heart transplantation had acute cellular rejection, but any such episode would have been treated with pulse methylprednisolone (administered in intravenous doses of 10 mg/kg/day for 3 days, followed by a tapering oral dose starting at 1 mg/kg/day) and further plasmapheresis if relevant isohemagglutinins were detectable above a titer of 1.
It is not our unit policy to routinely subject infants after heart transplantation to endomyocardial biopsies. They undergo endomyocardial biopsy if possible rejection was indicated by changes in clinical status (e.g., fever, hypotension, lethargy, irritability). Because of the unusual circumstance, all three infants who received hearts from ABO-incompatible donors underwent endomyocardial biopsy at 1 week postoperatively. In view of her greater age at time of transplantation, patient 5 also underwent an additional two biopsies at weekly intervals. The biopsy specimens were assessed according to the criteria of the International Society for Heart and Lung Transplantation (12). Interstitial edema or hemorrhage and endothelial cell swelling, if observed, are generally regarded as histologic features of humoral rejection.
Since January 2000, five infants were considered for ABO-incompatible cardiac transplantation at the Freeman Hospital, Newcastle upon Tyne (Table 2). Three underwent successful transplantation with hearts from ABO-incompatible donors. Two were considered unsuitable for ABO-incompatible heart transplantation because of titers of anti-A or anti-B antibodies consistent with established antibody production (Table 4). Subsequently, both underwent successful transplantation with hearts from ABO-compatible donors. During this period, a further 22 patients in the pediatric age group (<16 years old) underwent transplantation with hearts from ABO-compatible donors.
Patients Considered for ABO-Incompatible Heart Transplantation
Patient details and serologic findings are summarized in Tables 2 and 4.
A male infant of blood group O was diagnosed with neonatal dilated cardiomyopathy in the first week of life and had numerous hospital admissions with worsening heart failure. At 3 months, when referred for transplantation, he was ventilated with increasing inotropic requirements. He underwent transplantation with a heart from a blood group A donor (group A subtyping of the donor was not performed).
A 31-month-old girl of blood group B and with postviral dilated cardiomyopathy, ventilated and inotrope dependent, was referred for consideration of cardiac transplantation. When a suitable size-matched heart from a blood group AB donor became available, an attempt to reduce the antibody titers by plasmapheresis was made. Five plasma exchanges failed to reduce the antibody titers to less than 1 in 16, and instead she was successfully bridged to transplantation from a blood group B donor with a biventricular assist device.
A 2-month-old infant of blood group O with neonatal dilated cardiomyopathy and complete heart block was urgently referred for cardiac transplant assessment because of progressive worsening congestive cardiac failure and increasing requirement for inotropic support. He developed aggressive necrotizing enterocolitis and underwent right hemicolectomy and ileostomy. He recovered well and received a heart transplant from a blood group B donor.
An 18-month-old girl of blood group O with postviral dilated cardiomyopathy was placed on extracorporeal membrane oxygenation (ECMO) for worsening cardiac and respiratory failure. She was considered for ABO-incompatible heart transplantation, but her antibody titers were considered too high and indicative of established antibody production. She was successfully weaned off ECMO and underwent a successful heart transplant from an ABO-compatible donor
A 21-month-old girl of blood group A with a 4-week history of acute myocarditis, who was ventilated and on high-dose inotropes, was referred for possible mechanical support or transplantation. Anti-B titers were considered low for her age, at 1 in 2. A blood group B donor heart with a good size match became available and she underwent an ABO-incompatible transplant.
Serum Antibody Titers before Transplantation
Positive titers of anti-A and anti-B, though at low levels, were detected before transplantation in each recipient of an ABO-incompatible heart. In patients 1 and 3, these are likely to have been of maternal origin. Anti-A and anti-B is often present as an IgG antibody with activity against both A and B in blood group O individuals. IgG will cross the placenta and will give positive titers in neonates by saline agglutination tests (9). The levels of anti-B antibodies in patient 5 were lower than normal for age (9, 11). Patients 2 and 4 had serum antibody titers consistent with normal isohemagglutinin production for their age (Table 4).
Mortality and Morbidity
There was no mortality in this group of patients. All three recipients of hearts from ABO-incompatible donors are doing well 40 months, 30 months, and 12 months posttransplantation. Recipients of hearts from ABO-compatible donors are also doing well 37 and 26 months posttransplantation.
Patients 1, 3, and 5 (the ABO-incompatible group) underwent endomyocardial biopsies at 7 days posttransplantation. These biopsies did not show any evidence of cellular or humoral rejection. Patient 5 underwent two additional biopsies at weekly intervals, which showed grade 1A minimal and grade 1A mild rejection, respectively. No modulation of immunosuppression was required. No further biopsies were performed, as they were not indicated on clinical or echocardiographic grounds. All three patients continue on calcineurin-inhibitor therapy (cyclosporine) and weaning from adjunct immunosuppressive therapy (azathioprine) continues.
Patient 3 initially had erratic cyclosporine absorption because of fluctuating ileostomy losses, but that soon stabilized. The ileostomy has since been reversed. Specifically, no cases of posttransplantation lymphoproliferative disease, severe opportunistic infections, or diabetes mellitus have developed in recipients with ABO-incompatible donors, although the follow-up is relatively short.
All three recipients with ABO-incompatible donors had isohemagglutinin titers of less than 1 in 4 before transplantation. Patient 1, who received a group A heart, has developed low titers (1 in 1) of anti-A hemagglutinins 30 months after transplantation, and patient 3, who received a group B heart, has no detectable anti-B hemagglutinins 20 months posttransplantation. Serum anti-B hemagglutinin titers in patient 5, who received a group B heart, have remained low (1 in 2) 12 months posttransplantation (Table 4). Despite the low level of antibody production, there have been no adverse effects on clinical status or echocardiographic studies in patients 1 and 5.
In contrast, antibody production to the A or B blood group antigens not expressed in donor or recipient has developed as expected in patients 1 and 3 (Table 4). This suggests that reduced production of antibody to the antigens in the ABO-incompatible graft is not solely attributable to immunosuppression.
Cardiac transplantation has now been performed for 15 years as therapy for neonates and infants with lethal myopathic or structural heart disease. Since Bailey and colleagues reported the first series of three infants with hypo-plastic left heart syndrome who underwent orthotopic heart transplantation (13), infants have accounted for approximately 40 percent of all heart transplantations in children. The results in this group are particularly good, with 5-year survival rates approaching 80% (14). However, the number of such procedures in children, including infants, has remained stable since 1995 (15), in part because of mortality among infants and children who were waiting for a suitable donor (16). Efficient allocation of the rarely available organs is hampered by the presumed need for a recipient who has a blood type that is compatible with the donor.
Humans have two major antigen systems relevant to transplantation, the ABO system and the human leukocyte antigen system. In solid organ transplantation, the ABO system is by far the most important. The reason is twofold: (1) ABO antigens are expressed on almost all cells in the body; and (2) according to the law formulated by Landsteiner (17), humans have antibodies against those ABO antigens absent in the individual’s own tissues, which usually induce a hyperacute rejection of grafts expressing foreign A or B antigens. ABO-incompatible transplantations have been undertaken for kidney (1–4, 6) and liver (5, 7) to overcome the donor organ shortage. Several techniques such as plasma-pheresis (3, 18), immunoadsorption (2), administration of competitive soluble carbohydrate antigen (19), or splenectomy (4) have been used with some success.
Thorpe et al. described expression of A and B blood group antigens in human heart tissue (20). A and B antigens, as studied by immunofluorescence, were found on mesothelial cells on the surface of the epicardium and on the cardiovascular endothelium. No blood group antigens could be found on the cardiac muscle cells. Nonetheless, considerations of the particularly devastating effects of antibody-mediated rejection, both hyperacute and delayed, have prevented the intentional use of organs from ABO-incompatible donors. A landmark article by Cooper found only two survivors of eight cases of inadvertent ABO-incompatible heart transplantation (7).
The immune system in young infants is immature. Stimulation by T-cell–independent polysaccharide antigens, such as the capsular components of bacteria, does not lead to serum antibody response in infancy (21). The antibody production to the carbohydrate blood-group antigens begins at the age of 3 to 6 months as an immune response after the colonization of the gut with polysaccharide-bearing bacteria (9). Antibody titers increase to a maximum at 5 to 10 years of age, with a slightly more rapid increase in individuals of blood group O compared with those of blood group A or B (11). Individuals of blood group O tend to produce higher final levels of anti-B than those of blood group A, and may produce more IgG antibodies than individuals of blood group A or B (9, 11). Approximately 20% of group A donors will be of the subtype A2, with a lower density of antigens on the red blood cells and endothelial cells, and probably with a lower likelihood of causing organ rejection. However, subtyping of blood group A by standard procedures is unreliable in small infants because of the lower density of A antigen on neonatal cells (11). In addition, potential donors will have often received red blood cell transfusions, further confusing the picture. Therefore, we did not attempt subtyping of A in potential donors.
The importance of these observations to heart transplantation in infants, however, remains uncertain. Clinical data to support the concept of neonatal tolerance of foreign antigens are now just beginning to appear (e.g., lower rates of rejection of heart and lung transplants (22), better late survival, and perhaps a lower incidence of chronic graft atherosclerosis in infants than in adults (23)). To this body of evidence can be added the results of ABO-incompatible heart transplantation in infants as reported by West and colleagues (8). On the basis of their experience, we started our program of assessing infants for ABO-incompatible heart transplantation and performed three such transplants.
We corroborate the results of West and colleagues that heart transplantation can be safely performed during early infancy in spite of ABO incompatibility. Other than plasma exchange during cardiopulmonary bypass, no other intervention is required.
Among our patients, two have developed low levels of antibodies to the donor blood-group antigens. In the follow-up so far, these patients have shown no deviation from our ABO-compatible infant heart transplants. It is possible that even with accumulation of antibodies to donor antigen there is a degree of protection accorded to the graft because of accommodation (2, 3, 19, 24, 25).
Patients 2 and 4 represent a different phase of immunologic maturation. Isohemagglutinin production was established in these older infants, with relatively high serum antibody titers. A high risk of hyperacute rejection prevented us from performing ABO-incompatible transplantation in these infants. The good results in our ABO-incompatible heart transplant recipients, who have 40, 30, and 12 months of follow-up, suggests that excessive immunosuppression is not necessary in these cases.
The excellent results seen after ABO-incompatible heart transplantation, both in this series and as reported by West et al., suggest it should be offered to suitable infants awaiting transplantation. To our knowledge, patient 5 is the oldest deliberate ABO-incompatible heart transplant recipient reported in the literature. Absence of hyperacute rejection in the most vulnerable period in this case suggests that even children outside of infancy, who require urgent transplantation, can be considered for ABO-incompatible heart transplantation. Attempts to bridge these critically ill patients to transplantation using mechanical assist devices or ECMO are not without their own morbidity and mortality. The maximal age at which this protocol may be used safely remains to be determined, but detailed assessment of immune status is necessary. Indeed, the relative absence of antibodies against the ABO type of the donor posttransplantation may be an advantage. If a blood group O recipient who received a heart from a blood group A donor requires retransplantation in the future, the range of possible donors (i.e., blood groups O and A) is considerably expanded.
1. Paul LC, Baldwin WM. Humoral rejection mechanisms and ABO incompatibility in renal transplantation. Transplant Proc 1987; 19: 4463–4467.
2. Bannett AD, Bensinger WI, Raja R, et al. Immunoadsorption and renal transplant in two patients with a major ABO incompatibility. Transplantation 1987; 43: 909 –911.
3. Alexandre GPJ, Squifflet JP, De Bruyere M, et al. Present experiences in a series of 26 ABO-incompatible living donor renal allografts. Transplant Proc 1987; 19: 4538 –4542.
4. Sutherland DE, Fryd DS, So SK, et al. Long-term effect of splenectomy versus no splenectomy in renal transplant patients: Reanalysis of a randomised prospective study. Transplantation 1984; 38: 619 –624.
5. Gugenheim J, Samuel D, Reynes M, et al. Liver transplantation across ABO blood group barriers. Lancet 1990; 336: 519 –523.
6. Osorio AV, Sullivan EK, Alexander SR, et al. ABO-mismatched renal transplantation in children: A report of the North American Paediatric Renal Transplant Cooperative Study (NAPRTCS) and the Midwest Organ Bank (MOB). Pediatr Transplant 1998; 2: 26 –29.
7. Cooper DK. Clinical survey of heart transplantation between ABO blood group-incompatible recipients and donors. J Heart Transplant 1990; 9: 376–381.
8. West LJ, Pollock-Barziv SM, Dipchand AI, et al. ABO-incompatible heart transplantation in infants. N Engl J Med 2001; 344: 793–800.
9. Fong SW, Qaqundah BY, Taylor WF. Developmental patterns of ABO isoagglutinins in normal children correlated with the effects of age, sex, and maternal isoagglutinins. Transfusion 1974; 14: 551–559.
10. Ferriani VP, Barbosa JE, de Carvalho IF. Serum hemolytic classical and alternate pathways of complement in infancy: Age-related changes. Acta Paediatr Scand 1990; 79: 322–327.
11. Mollison PL, Englefriet CP, Contreras M. Blood transfusion in Clinical medicine [ed 9]. Oxford, Blackwell Scientific Publications 1993.
12. Billingham ME, Cary NR, Hammond ME, et al. A working formulation for the standardization of nomenclature in the diagnosis of heart and lung rejection: Heart Rejection Study Group. The International Society for Heart Transplantation. J Heart Transplant 1990; 9: 587–593.
13. Bailey LL, Nehlsen-Cannarella SL, Doroshaw RW, et al. Cardiac allo-transplantation in newborns as therapy for hypoplastic left heart syndrome. N Engl J Med 1986; 315: 949–951.
14. Dapper F, Bauer J, Kroll J, et al. Clinical experience with heart transplantation in infants. Eur J Cardiothorac Surg 1998; 14: 1–5.
15. Boucek MM, Faro A, Novick RJ, et al. The Registry of the International Society for Heart and Lung Transplantation: Fourth Official Pediatric Report—2000. J Heart Lung Transplant 2001; 20: 39–52.
16. Morrow WR, Naftel D, Chinnock R, et al. Outcome of listing for heart transplantation in infants younger than six months: Predictors of death and interval to transplantation. J Heart Lung Transplant 1997; 16: 1255–1266.
17. Landsteiner K. The specificity of serological reactions. Cambridge, Harvard University Press 1945.
18. Eid A, Zamir G, Yaron I, et al. Liver transplantation across the ABO barrier: The role of plasmapheresis. Transplant Proc 1998; 30: 701–703.
19. Cooper DYC, Ye Y, Miekrasz M, et al. Specific intravenous carbohydrate therapy: A new concept in inhibiting antibody-mediated rejection—Experience with ABO-incompatible cardiac allografting in the baboon. Transplantation 1993; 56: 769–777.
20. Thorpe SJ, Hunt B, Yacoub M. Expression of ABH blood group antigens in human heart tissue and its relevance to cardiac transplantation. Transplantation 1991; 51: 1290–1295.
21. Cadoz M. Potential and limitations of polysaccharide vaccines in infancy. Vaccine 1998; 16: 1391–1395.
22. Ibrahim JE, Sweet SC, Flippin M, et al. Rejection is reduced in thoracic organ recipients when transplanted in the first year of life [abstract]. Transplantation 2000; 69(suppl): S274.
23. Chinnock RE, Cutler D, Baum M. Clinical outcome 10 years after infant heart transplantation. Prog Pediatr Cardiol 2000; 11; 165–169.
24. Platt J. A perspective on xenograft rejection and accommodation. Immunol Rev 1994; 141: 127–149.
25. Bach FH, Ferran C, Hechenleitner P, et al. Accommodation of vascularized xenografts: Expression of “protective genes” by donor endothelial cells in a host Th2 cytokine environment. Nat Med 1997; 3: 196–204.