Correspondence: Dr F Martin, Department of Immunology, Genentech Inc., 1 DNA Way, South San Francisco, CA 94080, USA. E-mail: Flavius@gene.com

¹The author is employed by Genentech Inc.

Although multidrug immunosuppressive protocols have significantly increased short-term graft survival rates, chronic rejection still remains the major cause of long-term transplantation failure. A recent study in Nature Medicine by Liu et al.¹ provides important clues to achieving long-term survival of pancreas islet transplants.

In renal transplant, a shortage of appropriately matched donors and sensitization to donor antigens due to previous transplants, blood transfusions or pregnancy has resulted in renewed efforts to improve transplantation across immunological barriers. In pancreas islet transplantation, insulin independence is frequently lost within 5 years even with optimized transplant and immunosuppression protocols. While novel protocols improve graft survival, patients on lifelong immunosuppression develop severe infections and secondary malignancies. Therefore, it is important to improve regimes for inducing and maintaining tolerance to donor alloantigens. To this end, the study by Liu et al.¹ provides important clues to achieving this goal.

The authors show that adding B-cell depletion (anti-CD20—rituximab) to a standard-of-care T-cell immunotherapy (antithymocyte globulin—ATG and maintenance with rapamycin) results in a benefit in a cynomolgous monkey model of islet transplantation. While transplantation regimes incorporating B-cell depletion with rituximab have been reported in case studies and small human patient series,^{2, 3, 4, 5} the value and novelty of this study lies in the abundant longitudinal immune functional characterization in a controlled study. It also provides an example of successful cyclosporine replacement from standard-of-care immunosuppressive regimes. Surprisingly, a similar study using CTLA4-Ig (instead of ATG) in combination with rituximab without maintenance therapy did not find an elongation in graft survival.⁶ While the effects of ATG on cell types other than T cells have been documented,⁷ this discrepancy suggests that specifics of standard-of-care T-cell inhibition during induction and maintenance therapy are also critical to clinical success.

The major mechanisms of graft rejection are thought to be driven by T cells and by high-affinity alloreactive antibodies produced by B cells. Many immunosuppressive compounds (cyclosporine, anti-CD3, CTLA4-Ig, anti-CD25, antithymocyte globulin (ATG)) primarily, but not exclusively, target T cells. Since T cells are important for B-cell activation, the production of alloreactive antibodies is also inhibited. Despite the dominant T-cell pathology during graft rejection, agents that affect T-cell pathways show suboptimal results in chronic regimes, suggesting that other non-T-cell effector mechanisms are involved in long-term graft survival. Within these mechanisms, the best supported by preclinical and clinical data is the role of alloantibodies and B cells in early and late graft rejection.^{8, 9, 10} For example, B-cell gene signatures have been associated with poor prognosis, even in patients without obvious graft antibody and complement depositions. These results suggest that additional B-cell-dependent and potentially antibody-independent mechanisms are important.¹¹

So what are the B-cell-dependent mechanisms relevant to transplantation therapy and how do they contribute to graft rejection? Data from mouse models⁹ elegantly show that while B-cell elimination on its own did not benefit graft survival, combined B and T-cell inhibition conferred significant advantage over T-cell suppression, and this effect was due to both antibody-dependent and -independent mechanisms. While these mouse studies showed that it may 'take two to tango,' they also generated challenging questions. First, what is the antibody-independent B-cell-driven pathogenic mechanism? Second, is this observation translatable to the clinic and of value for the treatment of human transplant patients? The results from Liu et al. increase our mechanistic understanding on how answers to these questions suggest that it might also 'take two to treat.'

From the B-cell pathogenic mechanisms described in transplant rejection (Figure 1), antigraft antibodies cause a serious barrier for transplant eligibility and successful long-term graft survival in a large group of patients. Anti-CD20 therapy has been shown to efficiently eliminate CD20-positive B cells but not most CD20-negative plasma cells. Thus, depending on whether antibodies are preformed or not, anti-CD20 therapy has different mechanisms of action. Antibody responses to novel antigens are efficiently decreased or abrogated through depletion of CD20-positive B cells because CD20-negative plasma cells are never produced. In contrast, preexisting antibody titer decrease depends primarily on the intrinsic half-life of the plasma cells and their decrease after B-cell depletion. Owing to the indirect nature of this mechanism, in clinical practice, intravenous immunoglobulin (IVIG) and plasmapheresis have been combined with B-cell depletion to counteract antibody-mediated pathogenicity. Precedents for this mechanism have been described in rituximab-treated patients with several autoimmune diseases (rheumatoid arthritis—rheumatoid factor, systemic lupus erythematosus—anti-dsDNA, immune thrombocytopenic purpura—antithrombocyte antibodies).¹²

Figure 1.

Potential mechanisms of B-cell pathogenesis in transplantation: differentiation to antigraft antibody producing plasma cells, antigen presentation, cytokine production, lymphoneogenesis and malignant transformation.

Full figure and legend (144K)

Besides antibody-driven pathogenesis, several B-cell-driven mechanisms have been implicated in human autoimmunity and transplant rejection (Figure 1). Of these, antigen presentation, cytokine production and novel lymphoid structure formation can all be contributed by B cells that infiltrate grafts and contribute to poor prognosis. Efficient graft and systemic B-cell depletion may reduce or prevent these functions. Finally, B-cell depletion has been used successfully to treat patients with posttransplant lymphoproliferative syndrome dominated by highly malignant B-cell clones.

Out of these clinical alternatives and their corresponding B-cell mechanisms of pathogenicity (Figure 1), the work by Liu et al. models de novo antigraft antibody generation and all other antibody-independent mechanisms. It also makes the observation that long-term responders to anti-CD20 therapy have efficient depletion and late reconstitution with naive B cells. This is important, since comparable results have been obtained with B-cell depletion in human autoimmune disease patients, suggesting that similar mechanisms of B-cell deregulation apply across several human pathologies.

Recently, human transplant case reports and patient series treated with rituximab have shown various degree of success. Most importantly, these studies open the door for controlled, blinded trials to prove the clinical utility of B-cell depletion combined with standard-of-care T-cell immunotherapy. Additional clinical research to identify diagnostic markers of response and failure may contribute to the ultimate success of adding B-cell immunotherapy to transplantation standard-of-care regimes.