← Back to October 2014

The Science and Fiction of ABO Incompatible Transplantation

links image

Simon Urschel, MD
University of Alberta
Edmonton, Alberta, CANADA

The ABO blood group system, which was first described by Nobel Laureate Karl Landsteiner in 1901, is based on polysaccharide antigens expressed on red blood cells. These antigens are also expressed on a variety of other cells including vascular endothelium. Consequently following some disastrous early experiences it was recognized that transplantation of solid organs across ABO blood group incompatibility mostly led to hyperacute rejection and loss of the organ, and, in the case of heart transplantation, associated loss of the patient. Adults and older children are regularly found to have preformed antibodies against their non-self blood group, predominantly but not exclusively of IgM type. The lack of expressed A or B antigens makes blood group O persons universal donors while the presence of both antigens in AB humans and consequent absence of antibodies to either blood group make them universal recipients.

Observational and experimental studies in the 1960s revealed that newborns completely lack agglutinating antibodies against the blood groups and typically only develop them between 6 and 18 months of age, sometimes even later. Contact to human blood or tissue is not required, but molecular mimicry and cross reactivity with the surfaces of intestinal bacteria induce the immune response that also agglutinates and non-self erythrocytes causes their lysis.

Driven by a desperate wait list mortality with only half of the newborns and infants surviving to transplantation, a team at the Hospital for Sick Children in Toronto under the lead of Dr. Lori West and Dr. Ivan Rebeyka planned to cross this assumed insurmountable barrier in carefully selected children who had not yet developed relevant quantities of blood group antibodies. In 2001 the successful intentional ABO incompatible (ABOi) transplantation of the first 10 infants of whom 8 survived at least for 11 months post-transplant was reported [1], breaking with a dogma and initiating a new view on transplantation in young children.

In the initial cohort, none of the patients had shown antibody mediated rejection including 2 of 8 surviving children who eventually developed antibodies towards the blood group antigens of their donor. Further follow up revealed an unprecedented finding of antigen specific tolerance in these patients: the vast majority did not develop antibodies or detectable B-cellular immune response towards their donor blood group [2]. Recent assessment shows that to present none of the patients has developed titers higher than 1:32 towards their donor [3] compared to average titers of 1:256 in healthy children older than 2 years. It is well illustrated that this difference is not the consequence of global immune suppression in blood group O individuals who have received incompatible hearts: recipients of blood group A hearts develop normal titers towards blood group B but absent or very low titers to blood group A. The reverse pattern was documented for O recipients of B hearts and persistence of this tolerance was found up to 15 years post-transplant [4]. In addition, the isoagglutinins towards the donor blood group measured in erythrocyte agglutination tests of stepwise diluted plasma, which in principle are performed with the same laboratory technique as 100 years ago, is questionable. Recent investigations suggest that the endothelium of the heart expresses only some of the antigen subtypes and that antibodies towards these subtypes remain absent.

So which are the unique features of the immature immune system that allow indeterminate tolerance of the non-self blood group antigen, and until which age do they persist? The absence of the AB-blood group system and low potency of antibody mediated rejection in small rodents makes the generation of animal models difficult. Crucial hints on the mechanisms were drawn from infectious disease and vaccine research. Many pathogenic bacteria are protected by polysaccharide capsules of similar structure to the blood group antigens. Infections with these bacteria are more frequent and more severe in the same age range that permits AB-tolerance. Pure polysaccharide vaccines towards these infections were found not to induce antibody response up to 2 years of age. Polysaccharides cannot be presented in the type II major histocompatibility complex and therefore not induce the classical T cell mediated B cell response. Alternate activation of B cells occurs though linkage of the B cell co-receptor component CD21 to complement factor C3d bound to the antigen. While CD21 was found to be lacking in the spleen of children in the first 6 months of life, our group recently showed significantly reduced C3d levels specifically in children after ABOi transplantation [5]. Lipopolysaccharides can also be presented to NKT-cells via CD1d, a surface structure shown on regulatory B cells in mice and humans. While we found infants to have between 10 and 100 times the proportion of this B cell phenotype (CD5+CD1dhi) a correlation to ABOi transplant was not identified. In contrast, the development of B cell memory was shifted to less non-switched IgM+ memory B cells after ABOi compared to compatible transplantation. Further studies have to assess whether these immune alterations lead to consequences in immune response towards other antigens. To our surprise we found that infants after ABOi transplantation were significantly less likely to develop class II HLA antibodies compared to ABO compatible recipients [6], so ABOi transplantation might provide a tolerizing effect towards other donor antigens as well, although the power of this finding is low due to small numbers of patients and warrants verification in future studies.

From desperate experiment to standard procedure:
The 2001 paper induced huge attention and fascination; however, as every breach of long term dogmas, it also induced skepticism and limited adaption into clinical practice. Canada continued to pioneer the approach after 2001 from East to West. The two major centers in the United Kingdom and one German center could reproduce the Toronto success and publication of their safety data along with the encouraging long term results from Canada eventually convinced the US policy makers to offer this approach. The initial approach in the US was very careful: ABOi allocation only to patients with titers below 1:4 in the first year of life and only if the organ could not be compatibly allocated. Based on several studies showing that rejection rates and clinical outcome were similar despite selection of a sicker and higher risk cohort for ABOi hearts [7], UNOS revised their policy in 2013 to a more liberal ABOi allocation process, closer to the standard practice in the UK and Canada. Other jurisdictions remain hesitant despite the increasing evidence of the safety of ABOi transplant in infants and the associated shortening of wait times and, in some studies, reduction of wait list mortality.

While from a Canadian point of view it appears almost unethical not to offer ABOi transplantation to infants, the applicability for older children and adults remains hazy. The UK cohort published in 2008 included a 4.5 year old patient with primary and a 7 year old child with ABOi re-transplantation. Meanwhile, our center has successfully performed several cases in children up to 4 years old, and oral presentations at the last ISHLT meeting mentioned toddlers ABOi transplanted in the UK. A Swedish group has published the successful ABOi transplantation of two carefully selected adult patients in 2012.

Looking beyond the heart:
Doctors in Japan have developed extensive experience with ABOi kidney transplantation due to the lack of a deceased donor program until recently. Using intensified immune suppressive regimes, outcomes were found to be vastly comparable to ABO compatible transplants. Early approaches including splenectomy were replaced by routine application of B-cell depleting agents (rituximab); however, recent data suggests that even this may not be required. From an immunological perspective there are no overt reasons why infants receiving ABOi liver or lung transplants should not follow the same course as observed in the heart patients since the key to success seems to lie in the maturity of the immune system rather than in the type of transplanted organ.

Where do we go from here?
ABOi transplantation provides a unique insight into mechanisms of the human immune system and its immaturities. It may reveal some of the reasons why children show the best long term graft-survival of all age groups after heart transplantation. Whether targeted intervention such as interruption of B-cell maturation can induce similar benefits and ABO-tolerance in the more mature immune system remains undetermined but should be the subject of future research. ■

Disclosure Statement: the author has no conflicts of interest to report.


  1. West LJ, Pollock-Barziv SM, Dipchand AI, et al. ABO-incompatible heart transplantation in infants. New England Journal of Medicine. 2001 Mar 15; 344(11):793-800.
  2. Fan XH, Ang A, BarZiv SMP, et al. Donor-specific B-cell tolerance after ABO-incompatible infant heart transplantation. Nature Medicine. 2004 Nov;10(11):1227-33.
  3. Conway J, Manlhiot C, Allain-Rooney T, et al. Development of donor- specific isohemagglutinins following pediatric ABO-incompatible heart transplantation. Am J Transplant 2012;12:888-95.
  4. Urschel S, Larsen IM, Kirk R, et al. ABO-incompatible heart transplantation in early childhood: an international multicenter study of clinical experiences and limits. J Heart Lung Transplant 2013;32: 285-92.
  5. Urschel S, Ryan LA, Larsen I, et al. C3d plasma levels and CD21 expressing B-cells in children after ABO-incompatible heart transplantation: Alterations associated with blood group tolerance. J Heart Lung Transplant 2014, epub ahead of print May 9. pii: S1053-2498(14)01121-8.
  6. Urschel S, Campbell PM, Meyer SR, et al. Absence of donor-specific anti-HLA antibodies after ABO-incompatible heart transplantation in infancy: altered immunity or age? Am J Transplant. 2010 Jan;10(1):149-56.
  7. Everitt MD, Donaldson AE, Casper TC, et al. Effect of ABO-incompatible listing on infant heart transplant waitlist outcomes: analysis of the United Network for Organ Sharing (UNOS) database. J Heart Lung Transplant. 2009 Dec;28(12):1254-60.

Share via:

links image    links image    links image    links image