In most countries, a paired donation program to circumvent the immunological challenge of ABO incompatibility is precluded by law. Therefore, a kidney transplant (KTx) candidate with an ABO-incompatible (ABOi) living donor has a valuable option to wait for a deceased ABO-compatible donor with long-term dialysis therapy. However, recent meta-analysis has shown lower patient and graft survival in ABOi KTx than ABO-compatible KTx.^1,2 In ABOi KTx, over immunosuppression, leading to life-threatening infections, may cause lower patient survival.^1,2 In addition, acute antibody-mediated rejection (AABMR), due to anti-A or -B antibodies (Abs), contributes to lower graft survival.^1,2 Ab titers against donor blood group antigen are an AABMR predictor following ABOi KTx. Isohemagglutinin assays employing red blood cells (RBCs) are the most common assay used to measure Ab titer in ABOi KTx. However, ABO blood group antigens expressed on RBCs are not identical to those of the kidney due to different proteins linked to ABO carbohydrate antigens.³ In some cases, Ab titers do not correlate with clinical outcome; AABMR does not occur in some patients with high Ab titers, and vice versa.^4-6 Pecam1 (CD31) is the most abundant protein linked to ABO blood group antigens on kidney endothelial cells (KECs), which is different from Band3 mainly expressed on RBCs.³ A method to determine Ab titer specifically against ABO blood group antigens expressed on KECs is necessary to prevent over immunosuppression or precisely predict AABMR following ABOi KTx.

Figure 1. A new method to measure anti-A and -B antibody titer specific to ABO antigen on renal endothelial cells.⁷ a, The development of recombinant CD31 proteins containing ABO carbohydrate antigen. b, The microarray specifically detects anti-A and -B antibodies. FUT1, α1,2 fucosyltransferase. GT-A, α1,3 N-acetylgalactosaminyltransferase. GT-B, α1,3 galactosyltransferase. HEK, human embryonic kidney.

We developed a novel method to examine antibody titer against ABO antigens expressed on kidney.⁷ Recombinant CD31 containing ABO carbohydrate antigens were produced in glycogene-modified human embryonic kidney (HEK293) cells (Fig. 1, a) and spotted onto epoxysilane-coated glass slides using a noncontact microarray printing robot (Fig. 1, b). Mass spectrometry analysis revealed that recombinant CD31 of microarray has ABO antigens which are similar to those of human KECs. We compared anti-ABO Abs between this new method (CD31-ABO microarray) and conventional isohemagglutinin assays. We found that Ab levels in the CD31-ABO microarray varied even in samples with the same isohemagglutinin titers, suggesting Ab levels against renal ABO antigen are significantly different in each patient (Fig. 2).

Figure 2. Comparison of anti-A and -B antibodies between the isohemagglutinin and CD-31 ABO microarray methods in volunteers and hemodialysis patients.⁷ Antibody levels in the CD-31 ABO microarray varied even in samples with the same isohemagglutinin titers, suggesting antibody levels against renal ABO antigens are significantly different in each patient.

To investigate whether Ab levels in the CD31-ABO microarray would more accurately predict AABMR after ABOi KTx than the isohemagglutinin method, initial anti-A and -B Abs of the samples obtained before desensitization therapy were compared. In transplant recipients, any initial IgG or IgM antibody intensity >30,000 against the donor blood type in the CD31-ABO microarray showed higher sensitivity, specificity, positive predictive value and negative predictive value to predict AABMR following ABOi KTx, compared to isohemagglutinin assays. Representative data are shown in Figure 3 (blood group A-incompatible KTx). Ten out of 12 patients with AABMR (83.3%) had anti-A IgG Ab levels >30,000 in the CD31-ABO microarray. In contrast, only 1 out of 17 patients without AABMR (5.9%) had anti-A IgG Ab levels >30,000. No one had anti-A IgM Ab levels >30,000 in the AABMR (-) group.

Figure 3. Comparison of anti-A IgG (a) and IgM (b) antibodies between the isohemagglutinin and CD-31 ABO microarray methods in blood group A-incompatible kidney transplant patients. All samples were collected before desensitization therapy for ABO-incompatible kidney transplant.⁷ Ten out of 12 patients with antibody-mediated rejection (83.3%) had initial anti-A IgG or IgM antibody levels >30,000 (red or yellow circle) in A-incompatible kidney transplant, as shown by the CD-31 ABO microarray.

According to the results of anti-A and -B Abs against renal ABO blood group antigens by CD31-ABO microarray before KTx, we’ll be able to strengthen or reduce immunosuppression therapy, resulting in decreased numbers of AABMR and infectious events. We believe that this new method will contribute opportunities for safe ABOi KTx to end-stage renal disease patients.