Repeat ABO-B-incompatible living-donor kidney transplantation

To the Editor, We present a 34-year-old woman admitted in July 2023 for a repeat ABO-B-incompatible (donor: ABO-B, recipient: ABO-O) living-donor kidney transplantation (LDKT). Her first allograft (ABO-compatible) failed at 5 years because of medication non-adherence leading to (anti-HLA-A2 positive) antibody-mediated rejection (ABMR). Her second graft [first ABO-incompatible (ABOi) LDKT] likewise developed ABMR despite negative donor-specific anti-HLA antibodies and low anti-B titers by gel-card testing. In 2023 the patient developed end-stage kidney failure and was prepared for repeat ABO-B-incompatible LDKT Fig. 1. Timeline of the patient’s clinical course, ABO antibody titers and graft function. Top row: LDKT, including the first LDKT (ABOc) in 2010; the first ABO-B-incompatible transplant in 2015 preceded by pre-transplant conditioning (plasmapheresis, IAS and rituximab); biopsy-proven cABMR in 2021; a 3-month course of daratumumab in 2022; the discrepancy in anti-B-II levels between hemagglutination assays and an Luminex-based multiplexed assay in 2023 during preparation for the repeated ABO-B-incompatible LDKT; and the repeat ABO-B-incompatible transplant in 2023, again preceded by pre-transplant conditioning (IAS and rituximab). Lower panel: ABO antibody titers (anti-B IgG, solid red line; anti-A IgG, solid blue line) and serum creatinine (dashed orange line) plotted against the left and right y-axes, respectively. Shaded bars indicate phases of HLA-DSA positivity (pink) and negativity (green). Uneven timeline intervals are denoted by double slashes. Created in BioRender. Daoudi, I. (2025) https://BioRender.com/f3ruyzf. The patient underwent her first ABO-compatible (ABOc) LDKT in 2010 for end-stage kidney failure secondary to bilateral renal atrophy of unknown etiology. In 2015, end-stage graft failure occurred due to documented medication non-adherence, resulting in biopsy-proven kidney allograft rejection (Banff II) and the formation of anti-HLA-A2 donor-specific antibodies (DSAs) with a mean fluorescence intensity (MFI) of 4300. HLA antibody testing was performed using a Luminex Single Antigen Bead assay (One Lambda, Thermo Fisher). In 2015, the patient underwent her first ABOi kidney transplantation (recipient: ABO-O; donor: ABO-B), receiving an HLA-haploidentical organ from her father. The previously formed anti-HLA antibodies (anti-HLA-A2) were not donor-specific. Luminex screening revealed a virtual panel-reactive antibody of 9% for HLA class I and no detectable HLA class II antibodies. Thirteen days before transplantation, she received 620 mg of rituximab. Plasmapheresis sessions were conducted on Days –11 and –10 to reduce the initial anti-B immunoglobulin G (IgG) titer of 1:256 (anti-A IgG titer 1:64). Subsequent immunoadsorption sessions (IAS) using Glycosorb® B column (Glycorex, Inc.) on Days –8, –6, –4, –1 and 0 reduced the ABO antibody titers, with anti-B levels declining to a nadir of 1:1 after the fifth session. The immunosuppressive regimen included basiliximab for induction, with tacrolimus, mycophenolate mofetil and prednisolone for maintenance. Postoperatively, the serum creatinine declined to 1.4 mg/dL by Day 2, with anti-B IgG titers remaining low (max. 1:4) without further IAS. In subsequent years, kidney function declined progressively, and multiple biopsies confirmed chronic active ABMR (cABMR) as the underlying cause. The index biopsy in August 2021 demonstrated the following activity and chronicity parameters per the contemporaneous Banff classification (2019 update): activity—g2, i0, ti1, t1, v0, ptc1; chronicity—ah2, cg2, ci2, ct1, cv1, mm0, i-IFTA1. This alloimmune response occurred despite documented adherence to the immunosuppressive regimen, the absence of detectable HLA-DSAs and consistently low anti-B IgG titers (max. 1:4), as measured by sequential hemagglutination assays. In 2022, a 3-month course (May–August) of daratumumab was attempted as a rescue therapy for the cABMR; however, it failed to improve the patient’s condition. A subsequent follow-up biopsy, performed in September 2022, demonstrated a glomerulopathy most consistent with chronic, largely inactive antibody-mediated injury, with minimal activity and pronounced chronicity (Banff: g0, i0, t0, v0, ptc0, ti2; ah3, cg2, ci3, ct3, cv2, mm0; i-IFTA2, t-IFTA1). By May 2023, hemodialysis was reinitiated. In 2023, our institution’s ABO antibody titer assessment—performed using hemagglutination assays—revealed a low anti-B IgG titer (1:2), in contrast to a markedly higher anti-A IgG titer (1:64). These assays are performed using a gel-card system, which measures agglutination both at room temperature (NaCl phase) and at 37°C (anti-IgG Coombs phase). However, subsequent testing performed in Edmonton using an in-house Luminex-based multiplexed assay revealed the presence of anti-B-II IgM and IgG, demonstrating moderate reactivity at room temperature and low but persistent reactivity at 37°C. The corresponding MFIs for IgM anti-B-II were ∼15 000 prior to immunoadsorption, 15 000 following immunoadsorption and 12 000 post-transplantation. In contrast, anti-B-II IgG measured at the same time points showed substantially lower levels, with MFIs of 1900, 300 and 1000, respectively Fig 2. The Luminex assay is a bead-based immunoassay primarily used to detect anti-HLA antibodies. Creation of a multiplexed assay for measurement of ABO antibodies and its use for ABOi transplants has recently been reported by the University of Alberta group [1]. Longitudinal anti-B-II MFI values by Luminex Single-Antigen Bead assay before and after the repeat ABO-B-incompatible LDKT in 2023. IgM anti-B-II (red solid line) exhibited moderately high and persistent levels, with MFIs of ∼15 000 pre-immunoadsorption, 15 000 post-immunoadsorption and 12 000 post-transplantation. In contrast, IgG anti-B-II (blue solid line) showed substantially lower levels, with MFIs of 1900, 300 and 1000 at the same time points. Tx, transplantation. The discrepancy between the two testing methods significantly influenced our transplant preparation, leading us to initiate a standardized desensitization protocol—including immunoadsorption and rituximab. Ten days before surgery, 1 g of rituximab was administered. Over the 3 days preceding the transplant, the patient underwent daily anti-B immunoadsorption sessions using Glycosorb-B. The immunosuppressive regimen again consisted of basiliximab for induction and standard triple drug therapy for maintenance. Postoperatively, the patient’s urine output increased, and serum creatinine decreased to a nadir of 2.8 mg/dL by Day 12, where it plateaued, probably due to small infarctions at the mediocaudal pole of the transplanted kidney detected on a postoperative computed tomography scan. These infarctions were caused by the elderly donor kidney’s double arterial blood supply, where only one artery could be successfully anastomosed due to the small size of the upper pole artery. A kidney graft biopsy performed 14 days post-transplant revealed no pathologies but expected C4d positive immunohistochemistry was observed. The patient was discharged with a serum creatinine of 2.5 mg/dL. Because early post-transplant isoagglutinin rebound, particularly of complement-fixing IgG, can precipitate acute ABMR and necessitate rescue apheresis, we closely monitored isoagglutinins after transplantation. Neither gel-card hemagglutination nor bead-based (Luminex) testing showed a clinically meaningful increase: the anti-B IgG titer remained 1:2 throughout, and anti-B-II IgG MFI rose only modestly from ∼300 to ∼1000 (as detailed above). This intensive surveillance was pursued a priori given the anticipated risk of rebound in the setting of a repeated identical ABO mismatch; however, this concern was not borne out, and, accordingly, no additional IAS were required [2]. During the first 6 months, serum creatinine was stable at ∼2.0 mg/dL, the urinary albumin-to-creatinine ratio (uACR) remained <150 mg/g, and cytomegalovirus and BK viremia were undetectable. Follow-up during this period was conducted at our center; thereafter, care was transferred to a facility near the patient’s residence. At the latest follow-up (May 2025), the anti-B IgG titer was 1:2, serum creatinine remained ∼2.0 mg/dL and anti-HLA donor-specific antibodies (DSA) were negative at all time points during follow-up (HLA donor–recipient mismatch for kidney transplantations 1–3; see Supplementary data, Fig. S1). The immunological landscape of ABOi kidney transplantation differs markedly from that of HLA-mismatched transplantation. Unlike the adaptive responses typically seen with HLA antigens, natural anti-A and anti-B antibodies develop early in infancy presumably through exposure to gut microbiota. In the setting of ABOi transplantation, failure to adequately remove these ABO antibodies pre-transplant may result in acute rejection, as the antibodies bind to their corresponding antigens expressed in the renal microcirculation [2, 3]. Notably, antigen expression in the ABO-A kidney varies according to the ABO-A subgroup. In ABO-A1 kidneys, multiple glycans (A-II, A-III and A-IV) are expressed; however, in the vascular endothelium, only A-II glycans are expressed. In contrast, non-ABO-A1 kidneys (e.g. ABO-A2, A3) display only A-II glycans and at a lower density than in A1 kidneys. ABO-B kidneys express solely B-II glycans [4]. As a result, kidneys from ABO-A1 donors present significant antigenic challenges for ABO-O and ABO-B recipients [1, 5]. Our case demonstrates that repeated ABO-B-incompatible transplantation is feasible, while the practicability of repeating an ABO-A1 mismatch remains uncertain due to its greater immunologic challenge [5]. However, it is reasonable to expect that this framework extends to other ABO histocompatibility scenarios as well—with A2 grafts functioning as “ABOi lite” options for many recipients (and repeat ABOi A2 procedures already reported as successful) as well as A1 to B pairings, which may be immunologically feasible given lower average anti-A-II IgG MFI outside group O (median anti-A2 IgG MFI: ABO-O ≈ 28 000 vs ABO-B ≈ 4000) [1, 6, 7]. In this context, it is important to address the pathogenesis of chronic active ABMR observed after the patient’s second LDKT (i.e. her first ABOi LDKT). Notably, no anti-HLA DSAs were detected throughout the post-transplant course, raising the question of whether anti-B-II IgM and IgG, identified by the Luminex assay performed during preparation for the repeat ABOi LDKT, could have contributed to the alloimmune process. Defining their precise pathogenic role remains challenging: bead-based testing was performed only once, 8 years after transplantation, and there is no clear evidence to date that ABO antibodies alone drive a cABMR phenotype. Glycan-specific, bead-based ABO antibody testing may refine risk assessment and mechanistic understanding in this context, as discussed below [1, 8]. Another key finding in our case is the discrepancy in antibody detection at the time of the third LDKT. At our institution, hemagglutination assays revealed a high anti-A IgG titer (1:64), whereas anti-B IgG titers remained low (1:2). Several immunologic mechanisms such as the selective adsorption of anti-B antibodies by graft endothelial B-antigen epitopes or, more compellingly, donor-specific downregulation of anti-B–producing B cells have been reported in the literature, which could, at least in part, explain the observed antibody dynamics [9, 10]. However, assay-specific limitations of gel-card hemagglutination must be acknowledged, particularly considering findings from Luminex-based assays demonstrating the presence of anti-B-II antibodies of both the IgM and IgG isotypes. This novel Luminex assay was assembled as follows: subtype-defined tetrasaccharide neoglycoconjugates (A-I to A-VI, and B-I to B-VI) were synthesized and subsequently covalently coupled to fluorescent polystyrene microspheres (Luminex) using adapted bead-coupling protocols [1]. Compared with hemagglutination, glycan-targeted Luminex testing offers several advantages: high reproducibility, good inter-laboratory concordance, reliance on a single standardized platform rather than multiple gel/tube variants, reliable discrimination between IgM and IgG, and the ability to interrogate antibodies to specific ABO glycans, a clinically relevant feature because type II A/B structures are those expressed on vascular endothelium. At the same time, important caveats remain: limited collective experience, uncertain MFI cut-offs for clinical decision-making and incomplete understanding of the complement-activating capacity of the detected antibodies [1, 8, 11, 12]. In view of these findings, we deemed a repeat ABO-B-incompatible LDKT, following desensitization and targeted reduction of anti-B-II IgG, a reasonable and safe course, given that both the pathogenic role of anti-B-II IgM and clinically meaningful Luminex MFI cut-offs remain uncertain, whereas current evidence implicates IgG as the principal driver of acute ABMR [1, 2]. In summary, we report a 34-year-old woman who successfully underwent a repeat ABO-B-incompatible LDKT, demonstrating that sequential ABOi transplantation is feasible. More broadly, repeat ABOi LDKT can help overcome inequities driven by population blood-group distributions (e.g. the relative scarcity of compatible donors for blood group O) and patient preference to proceed with their intended donor, i.e. parents. In relation to kidney-paired donation (KPD), ABOi does not replace KPD but complements it, offering a viable route when timely compatible matches are unlikely, chains are operationally complex, or pairs prioritize direct donation. Finally, continued refinement of bead-based, glycan-specific assays should advance ABO histocompatibility assessment beyond transfusion-surrogate testing: because red blood cells and renal endothelium display different ABO glycan repertoires and densities, not all anti-A/B specificities detected by RBC-based methods are clinically relevant for the graft. Aligning assays with endothelial (type II) antigen expression may prevent the unnecessary exclusion of candidates from solid organ transplantation. The author gratefully acknowledges Prof. Oberbauer for his invaluable guidance and support throughout the development of this research letter. Special thanks are due to Prof. West and Prof. Halpin for their critical review of the manuscript and insightful comments on the antibody measurement techniques. The authors also thank the staff of the HLA laboratory at AKH Vienna for providing detailed testing data. This research project was supported by a grant from the Vienna Science and Technology Fund (#WWTF #LS20-081 to R.O.). The authors of this manuscript report that there is no affiliation or involvement in an organization or entity with a financial or non-financial interest in the subject matter or materials discussed in this manuscript. I.D. drafted the manuscript and prepared overview figure; R.O. and A.H. assisted with the acquisition and interpretation of clinical and laboratory data and A.H. also provided supervision. L.W. and A.H. contributed methodological expertise on antibody measurement techniques and critically reviewed, edited the manuscript and provided figures illustrating the Luminex assay results. All authors have read and approved the final version for publication. The Ethics Committee of the Medical University of Vienna has been consulted and confirmed that this fully anonymized research letter complies with institutional requirements for publication. Patient confidentiality has been maintained throughout. All data generated or analyzed during the preparation of this research letter are included in this published article and referenced in the manuscript. None declared.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.