The majority of graft losses in the late post-transplant phase are thought to result from chronic antibody-mediated rejection owing to de novo development of donor-specific HLA alloantibodies (DSA).1 The anti-CD20 antibody rituximab has, therefore, gained widespread attention as a therapeutic agent against B-cell immunity and is used in the majority of ABO-incompatible or crossmatch-positive kidney transplant protocols.2

Rituximab binds to the CD20 molecule, which is not expressed on plasma blasts or plasma cells and, therefore, has no effect on the production of pre-existing antibodies; however, it can target memory B cells. In addition, rituximab is thought to inhibit antigen presentation to T cells through B cells and, thus, prevent T-cell mediated damage to the kidney. Such damage can stimulate de novo DSA production, especially in presensitized patients with HLA antibodies in their serum that were acquired after blood transfusions, pregnancy or a previously failed organ transplant. Many transplantation centres have, therefore, started to use rituximab in the pretransplant and post-transplant phase to prevent alloantibody production.2,3

New results by Ashimine et al. challenge this practice.4 The investigators analysed the impact of pretransplant rituximab on the development of de novo HLA antibody production in a series of 30 ABO-incompatible kidney transplants. 51 splenectomized patients who received an ABO-incompatible transplant and 228 untreated recipients of ABO-compatible kidney transplants served as control groups. Patients with pretransplant DSA were excluded. Recipients of ABO-incompatible transplants who were treated with rituximab had a de novo HLA antibody (DSA or non-DSA) development rate of 14.3% 34.3 months after transplantation. A similar de novo HLA development rate (13.9%) was seen in recipients of ABO-compatible transplants without rituximab and a 13.2% rate in splenectomized recipients of ABO-incompatible transplants. These results clearly show that neither splenectomy nor rituximab treatment in the medium-term follow-up period inhibits de novo HLA antibody production, and suggest that rituximab might not prevent short-term antibody-mediated allograft injury, at least in immunologically low-risk patients without DSA in their pretransplantation serum.

However, the study is a retrospective single-centre study with a small number of patients, lacks histopathological data from biopsy specimens and has a limited study period, drawbacks the authors recognize. All three points are critical: small numbers of patients mean that even small changes in the rate of events can greatly influence the results. In addition, a lack of (standardized) histopathological data prevents firm conclusions on the rate of chronic antibody-mediated changes between treatment groups. Finally, the limited study period of less than 3 years in the rituximab arm might fail to detect the majority of de novo HLA antibodies. Indeed, in a similar study that also excluded patients with pretransplant DSA, regular antibody screenings revealed the occurrence of de novo DSA at a mean of 4.6 years after transplantation, although the T cell-mediated damage in the organ occurred earlier (28% of patients with de novo DSA had clinical rejection within the first 6 months after transplantation).5

Superficially, the findings of Ashimine et al. are in clear contradiction to the results from a similar Japanese cohort.6 Despite a lower rituximab dose (200 mg versus 400 mg per patient), Kohei et al. observed a significantly lower de novo DSA development rate in ABO-incompatible kidney transplant recipients under rituximab induction therapy (1.7%) than the 18.1% rate in a matched group of ABO-compatible patients without such therapy (P = 0.029; Table 1). As in the study of Ashimine et al., the follow-up of 2 years for de novo DSA development was short. However, the number of rituximab-treated patients was higher (57 versus 30 patients), and histopathological data from biopsies were available. Chronic antibody-mediated rejection rate after 2 years was strikingly lower with rituximab induction (3.5%) than without rituximab induction (28.9%; P = 0.0001).

Table 1 Comparison of rituximab induction therapy on the de novo DSA development
Full size table

Importantly, when DSA alone (instead of total HLA antibodies) was analysed in the study by Ashimine et al., a trend towards a lower rate of de novo occurrence of DSA was seen in the group receiving rituximab induction therapy compared with the ABO-compatible control group (3.6% versus 8.8%). Another striking difference between the two studies is the inclusion of presensitized patients with DSA (31.6% and 28.9% of patients with or without rituximab had pretransplant DSA, respectively) by Kohei et al.; these patients were excluded by Ashimine et al. Highly alloreactive presensitized patients are expected to be more prone to the development of de novo DSA, and this might explain the higher rate of de novo DSA occurrence reported by Kohei et al. compared with the findings of Ashimine et al. (18.1% versus 8.8% for ABO-compatible transplants). Differences in the efficacy of rituximab induction therapy in different cohorts might depend on the alloimmune state of the recipient. Although non-sensitized patients, such as those recruited to the present study, might not benefit from B-cell depleting agents, (highly) sensitized recipients or patients with anti-blood group antibodies against the donor organ might benefit from the B-cell depleting properties of this therapy. Therefore, this strategy was implemented in many algorithms for the transplantation of patients who are at a high risk of antibody-mediated allograft injury.2,7,8,9

Although the study by Ashimine et al. questions the use of rituximab to eliminate B cells and thereby prevent de novo HLA antibody production in non-sensitized kidney graft recipients,4 a further randomized controlled study of rituximab induction in an ABO-compatible cohort also found a clear trend towards a lower occurrence of de novo DSA.10 In patients who were moderately sensitized (panel-reactive antibody ≤50%), only 3.0% who received a single dose of 375 mg/m2 developed de novo DSA after 3 years compared with 15.8% of patients in the non-rituximab treated control group (P = 0.1). Similarly, Kohei et al. clearly demonstrated that rituximab might prevent de novo DSA synthesis in a group that contained presensitized recipients.6 In summary, rituximab might be useful in patients with pre-existing HLA (or ABO) antibody, but its use in immunological low-risk kidney graft recipients remains questionable.