Finding a compatible donor for kidney transplant candidates requires overcoming immunological barriers such as human leukocyte antigens (HLA) compatibility and ABO compatibility, which may prolong time to transplant. Emerging data suggest a role for red blood cell antigens (RCA) in renal transplant outcomes [1–3]. As of July 2023, there are 45 recognized blood group systems, many of which are expressed on tissue other than the surface of red blood cells (RBC) [4]. The functions of these antigens are not all fully understood but some function as chemokine receptors and others as membrane transporters [5]. Of the 45 recognized blood groups the kidney is known to express Duffy, Lewis, Kidd and MNS blood groups [1, 5]. Antibodies to RCA (other than ABO) are usually not naturally occurring and require exposure for development. Transfusions are the main source of exposure for most RCA with each transfusion event increasing the likelihood of alloimmunization by 0.2%. The probability of developing RCA allosensitization is dependent on multiple factors including underlying disease, immunogenicity of the antigen, dose and frequency of transfusion, age, and gender among other factors [6]. The incidence of RCA alloimmunization in multiply transfused individuals, such as end stage renal disease patients requiring frequent transfusions, can be as high as 60% [2, 6].

Whether antibodies to RCA can mediate renal graft rejections remains debatable. Various studies produced mixed results when looking at retrospective data and trying to correlate presence of antibodies with kidney graft survival [1, 3]. There are some data to suggest that donor-recipient Duffy blood group (Fy) mismatches may increase the risk for chronic allograft damage [3]. However, while it is routine to screen renal transplant candidates for ABO antigens, detailed RCA phenotyping of the donor kidney is not routinely tested. In this paper, we review the current data on the role of Fy in renal transplantation and discuss the potential mechanisms of its biological function.

The Duffy blood group system was first described in 1950 in a patient named Duffy who presented with hemolytic transfusion reactions after having received multiple transfusions [7, 8]. Fy is a seven transmembrane domain glycoprotein that has multiple epitopes for which antibodies can be formed. It is encoded on chromosome 1 by two codominant alleles (FY*A and FY*B). These two alleles differ by a single nucleotide polymorphism at position 125 (G/A) resulting in the presence either of glycine or aspartic acid in position 42 of the polypeptide chain that gives rise to the two antigens, Fy^a and Fy^b (Figure 1). Depending on the alleles present, four main phenotypes may be found in the population: Fy(a+ b−), Fy(a− b+), Fy(a+ b+) and Fy(a− b−), although the phenotypic expression in tissue versus blood can vary depending on the specific genotype (Table 1) [7, 9]. The most common antigens are Fy^a and Fy^b, but Fy3, Fy5, and Fy6 have also been described [5]. Antibodies for Duffy are almost never naturally occurring and are a result of exposure to the antigen. Antibodies to Fy^a are more frequently seen than antibodies to Fy^b by about a 20-fold increase. These antibodies are predominately IgG1 with a small percentage of presentations consisting of IgM (25%) and IgG2 (18%) [5, 7]. Antibodies against both Fy^a and Fy^b cause immediate and delayed hemolytic transfusion reactions and have been associated with hemolytic disease of the fetus and newborn [5].

The functional role of DARC in recruitment of inflammatory cells is still not fully understood, and several mechanisms have been suggested: 1) DARC helps presenting chemokines on the endothelial cells to leukocytes expressing corresponding receptors; 2) The binding of chemokines to DARC helps to create a gradient flow to keep an influx of chemokines coming to the site; 3) Neutralization of bound chemokines [19, 20]. Although the mechanism of DARC is not clear, there is evidence of upregulation of DARC on peritubular capillaries during both humoral and acute cellular rejection episodes [19]. The question remains whether this upregulation is an attempt to bind and neutralize chemokines and control the inflammation or is this an attempt to recruit more inflammatory cells? It is important to note that most of the studies investigating Duffy’s role in renal transplantation rely on RBC Fy phenotyping, which may be different from the Fy phenotype in the renal tissue. It is possible that DARC is able to pursue different functions according to the spatial and the temporal context (Figure 2).

The idea that anti-Fy antibodies may participate in renal allograft rejection is supported by a case report by Watorek et al. The authors reported a 41 years-old Caucasian woman, Fy(a− b+) phenotype, who had anti-Fy^a antibodies detected 2 years prior to her transplant, although at time of transplant her anti-Fy^a antibodies were undetectable. Her post-transplant hospital course was complicated by DGF and a biopsy at 26 days post-transplant revealed acute rejection, both of which are associated with upregulation of DARC expression in the kidney [14, 19]. Two months post-transplant her serum tested positive for anti-Fy^a antibodies and a repeat biopsy at 3 months showed acute and C4d+ antibody-mediated rejection in the absence of HLA donor-specific antibodies (DSA). The authors concluded that the unfavorable outcome of her transplant is a result of the presence of antibodies to Fy^a [21]. At the time of writing this paper, this is the only case report found linking the cause of kidney rejection to anti-Fy antibodies. However, case reports have suggested the involvement of antibodies to other RBC groups in renal allograft rejection [22–25].

At our center, we recently evaluated a 30 years-old African female with a history of multiple RBC transfusions due to anemia for a kidney transplant. Her HLA panel reactive antibody (PRA) was 0%. An ABO type and screen identified antibodies to Fy3, JkB, E, C, and K. The finding of the anti-Fy3 suggested that this patient carries the very rare true Fy null phenotype. Despite her ability (albeit limited) to receive transfusions due to the absence of Fy on RBC in most AA, her ability to find a compatible kidney donor is virtually impossible due to the presence of Fy on nearly every donor kidney. Even with a perfectly HLA and ABO matched donor, there is a potential for rejection due to anti-Fy3 reactivity and therefore this patient was deemed not suitable for transplantation at our center. Currently, there are no guidelines for donor-recipient RCA matching, mostly since more data is needed and since this will narrow down the donor pool for many transplant candidates. However, a more detailed screening process for RBC antibodies and RCA phenotyping of the donor may be warranted in patients experiencing antibody-mediated rejection (ABMR) with no HLA-DSA, especially if a recent post-transfusion reaction was observed prior to the diagnosis of ABMR. RCA antibody characterization might also be helpful prior to re-transplant following an ABMR involving RCA antibodies, and in cases where the recipient has a rare null phenotype. Differentiating between the true null and the erythrocyte silent mutations can be achieved by utilizing DNA-based typing instead of serologic phenotyping.

In conclusion, most of the emphasis in overcoming immunological barriers in solid organ transplantation is rightfully put on matching donors and recipients for HLA and ABO, however, the Duffy blood group system can present a unique and rare barrier to transplantation and potentially impact transplant outcomes.