The past, present, and future of costimulation blockade in organ transplantation

Abstract

Purpose of review

Manipulating costimulatory signals has been shown to alter T cell responses and prolong graft survival in solid organ transplantation. Our understanding of and ability to target various costimulation pathways continues to evolve.

Recent findings

Since the approval of belatacept in kidney transplantation, many additional biologics have been developed targeting clinically relevant costimulation signaling axes including CD40-CD40L, inducible costimulator-inducible costimulator ligand (ICOS-ICOSL), and OX40-OX40L. Currently, the effects of costimulation blockade on posttransplant humoral responses, tolerance induction, and xenotransplantation are under active investigation. Here, we will discuss these pathways as well as preclinical and clinical outcomes of biologics targeting these pathways in organ transplantation.

Summary

Targeting costimultion is a promising approach for not only controlling T cell but also B cell responses. Consequently, costimulation blockade shows considerable potential for improving outcomes in antibody-mediated rejection and xenotransplantation.

INTRODUCTION

The alloimmune response in transplantation that leads to allograft rejection is mediated by the adaptive immune response through the action of T and B lymphocytes. The adapative immune response in general requires three signals: through the antigen-specific receptors [T cell receptor (TCR) and B cell receptor (BCR)], costimulation through various pathways (CD28 and CD40), and cytokines promoting both autocrine and paracrine signaling [interleukin (IL)2, IL4, B cell activating factor, and so on]. The necessity of costimulatory pathways in the process of T and B cell activation in the context of alloimmunity was first postulated in the late 1970s by Lafferty and Woolnough [1]. Subsequent work demonstrated that activation of signal 1 in the absence of costimulation (signal 2) leads to anergy, or antigen-specific hyporesponsiveness, a key mechanism of peripheral tolerance [2]. These findings sparked further investigation into the receptors and pathways involved in costimulation. In broad terms, there are two families of costimulatory pathways that are most important for adapative immune responses: the Ig-related superfamily [CD28-CD80/86, inducible costimulator-inducible costimulator ligand (ICOS-ICOSL)] and the TNF-related superfamily (CD40-CD40L, OX40-OX40L, 41BB-41BBL). Although these pathways serve to help activate the adapative immune response, there are a number of coinhibitory pathways related to each family [cytotoxic T lymphocyte antigen (CTLA)4-CD80/86 and Programmed death 1 (PD1)-PDL1/2] that help to regulate and fine tune the adaptive immune response (see Fig. 1).

FIGURE 1.

Costimulatory molecules and the adaptive immune response. The adaptive immune response requires three signals for full lymphocyte activation. Signal 1 is through the TCR-Ag-MHC complex, signal 2 is through costimulatory molecules, and signal 3 is through cytokines and their receptors. There are two main families of costimulatory molecules. The Ig-like costimulatory family (right, blue) and the TNF-like costimulatory family (left, red), which provide both costimulatory (+) and coinhibitory (−) signals to fine-tune the adaptive immune response. Illustrated by Megan Llewellyn, MSMI, CMI; Duke Surgery.

THERAPEUTIC MANIPULATION OF COSTIMULATORY PATHWAYS IN TRANSPLANTATION

Manipulating the second signal of T cell activation with checkpoint inhibitors has revolutionized cancer therapy [3-8]. Modulating these same pathways in the opposite way made them ideal targets for therapeutic manipulation in transplantation. Many of the molecular targets were initially identified using monoclonal antibodies that had profound effects on lymphocyte activation and proliferation. The discovery of these important costimulatory pathways coincided with the development of recombinant engineering techniques for generating Ig-fusion proteins, which created an arsenal of putative therapeutic options for manipulation of adaptive immune responses that could potentially suppress allograft rejection with less off-target effects than other immunosuppressive strategies. Table 1 shows a number of the completed and ongoing clinical trials of costimulatory blockade agents in transplantation, the results of which will be discussed in the following sections.

Table 1.

Clinical trials involving costimulation blockade agents in organ transplantation

Study ID	Phase	Indication	Drug regimens	Results
NCT00035555	2	KTx	Belatacept MI vs. belatacept LI vs. CsA	No differences in acute rejection rates between belatacept and CsA groups
NCT00114777	3	ECD KTx	Belatacept MI vs. belatacept LI vs. CsA	No difference in patient/graft survival or acute rejection rates between belatacept and CsA groups
NCT00256750	3	KTx	Belatacept MI vs. belatacept LI vs. CsA	Significantly improved patient and graft survival in belatacept groups compared to CsA
NCT00402168	2	KTx	Switch to Belatacept vs. remain on CNI	Similar patient and graft survival but increased rate of acute rejection in belatacept group
NCT00455013	2	KTx	Belatacept + MMF vs. belatacept + sirolimus vs. Tac + MMF	Similar acute rejection rates but improved graft function in Belatacept groups compared to Tac group
NCT00565773	2	KTx	Alemtuzumab induction + belatacept + sirolimus + BM infusions	Well tolerated and effective use of belatacept-based regimen avoiding use of steroids and CNI
NCT00578448	2	KTx	Belatacept 10 mg/kg vs. 5 mg/kg	4/12 acute rejection; 1 death; 1 graft loss
NCT01780844	2	KTx	Tac + MMF vs. bleselumab + Tac vs. bleselumab + MMF	No difference in acute rejection rates in bleselumab compared to Tac groups
NCT01791491	1	KTx	Single dose of belatacept 7.5 mg/kg	Similar pharmacokinetics in adolescent patients receiving 7.5mg/kg belatacept as adults receiving 5mg/kg belatacept
NCT02152345	4	KTx	Belatacept vs. Tac	Completed, results unpublished
NCT02217410	2	KTx	CFZ533 + Tac vs. CFZ533 + anti-IL2 induction vs. Tac + anti-IL2 induction	CFZ533 noninferior to CNI in preventing graft rejection
NCT00001857	2	KTx	BG9588 +/− MMF and steroids	Terminated because of high rate of thromboembolic complications
NCT00346151	2	KTx	ATG induction + belatacept + sirolimus + steroids	Terminated because of acute rejection threshold met
NCT00555321	2	LiverTx	Basiliximab induction + belatacept MI + MMF vs. belatacept MI + MMF vs. belatacept LI + MMF vs. Tac + MMF vs. Tac	Terminated because of higher rates of acute rejection, death, and graft loss in Belatacept groups at 6 months
NCT01436305	2	KTx	Alemtuzumab induction + Tac vs. alemtuzumab induction + belatacept vs. basiliximab induction + Tac + belatacept	Terminated because of safety concerns and change in alemtuzumab availability
NCT01790594	2	SKPTx	Tac + belatacept vs. Tac	Terminated because of projected accrual goal not achieved
NCT01856257	2	KTx	Tac vs. belatacept	Terminated because of safety concerns
NCT02130817	4	KTx, AMR	PLEX/IVIG + ATG induction + belatacept + Tac withdrawal + MMF + steroids	Withdrawn because of low recruitment
NCT03504241	1	KTx	MSC infusions + belatacept-based immunosuppression	Recruiting
NCT00719225	3	KTx	Belatacept compassionate use	Approved for marketing
NCT01729494	4	KTx	Alemtuzumab induction + Belatacept vs. ATG induction + Belatacept vs. ATG induction + Tac	Active, not recruiting
NCT01820572	3	KTx	Belatacept vs. CNI	Active, not recruiting
NCT01837043	2	KTx	Switch to belatacept vs. remain on CNI	Recruiting
NCT01921218	3	Failing KTx	Belatacept vs. CNI	Active, not recruiting
NCT02137239	2	KTx	Belatacept + everolimus vs. Tac + MMF	Active, not recruiting
NCT02213068	4	KTx	Belatacept + MMF vs. belatacept + Tac vs. Tac + MMF	Recruiting
NCT02314403	1	BM + KTx	Conditioning regimen that includes belatacept	Recruiting
NCT02560558	4	KTx	Belatacept 8-weekly vs. belatacept 4-weekly	Active, not recruiting
NCT02738918	2	KTx	Belatacept	Active, not recruiting
NCT03388008	2	LungTx, AMR	Switch to belatacept vs. remain on Tac	Not yet recruiting
NCT03504241	1	KTx	MSC infusions + alemtuzumab induction + belatacept + sirolimus + MMF + steroids	Recruiting
NCT03663335	2	KTx rejection	CFZ533 dose A + steroids vs. CFZ533 dose B + steroids vs. Tac + steroids vs. CFZ533 dose C +/− steroids vs. Tac +/− steroids	Active
NCT03781414	2	LiverTx rejection	CFZ533 dose A vs. CFZ533 dose B vs. Tac	Not yet recruiting
NCT03805178	2	LungTx desensitization and AMR	Belatacept + carfilzomib	Not yet recruiting

AMR, antibody mediated rejection; ATG, antithymocyte globulin; BM, bone marrow; CNI, calcineurin inhibition; CsA, cyclosporine; ECD, extended criteria donor; IVIG, intravenous immunoglobulin; KTx, kidney transplant; LI, less intensive; LiverTx, liver transplant; LungTx, lung transplant; MI, more intensive; MMF, mycophenolate; MSC, donor mesenchymal stem cells; PLEX, plasmapheresis; SKPTx, simultaneous kidney pancreas transplant; Tac, tacrolimus.

COSTIMULATION BLOCKADE WITH BELATACEPT: SUCCESSES AND CHALLENGES

One of the first Ig-fusion proteins generated by recombinant engineering was a CD28Ig fusion protein, but because of its relatively low affinity for CD80/86 its efficacy in dampening adaptive immune responses in experimental systems was modest [9]. Thus, when its higher affinity co-inhibitory receptor CTLA4 was identified and its fusion protein CTLA4Ig produced, there was much excitement when it showed greater efficacy in modulating adaptive immune responses [10,11]. In-vivo testing of CTLA4Ig demonstrated reduced T cell-dependent immune responses and long-term xeno and allograft survival [12-14]. These results prompted clinical trials in autoimmune disease ultimately leading to Food and Drug Administration (FDA) approval of abatacept in 2005 as the first selective costimulatory modulating agent in clinical practice for treatment of rheumatoid arthritis. [15,16]

These promising results with abatacept in autoimmune disease were not recapitulated when abatacept was tested in nonhuman primate (NHP) allotransplant studies, which demonstrated limited efficacy in preventing rejection [17,18]. This prompted a redesign of the CTLA4Ig fusion protein to a higher affinity version, LEA29Y, with enhanced affinity for CD80/86. The higher affinity CTLA4Ig fusion protein, belatacept, significantly prolonged allograft survival in NHPs paving the road for the first clinical trials of belatacept in transplantation [17]. In 2005, the phase II clinical trial results of belatacept as maintenance immunosuppression in renal transplantation were reported showing similar efficacy to cyclosporine in preventing rejection and improved allograft function at 1 year [19]. The phase III Belatacept Evaluation of Nephroprotection and Efficacy as First-line Immunosuppression Trial (BENEFIT) with longer term follow-up also showed improved renal function, reductions in cardiac and metabolic toxicities, and a survival advantage at 7 years posttransplant compared to cyclosporine [20]. These results led to the FDA approval of belatacept for immunosuppression in kidney transplantation in 2011.

Despite the success of these trials some limitations to belatacept-based immunosuppression in transplantation still exist. For instance, belatacept is only available as an intravenous infusion, limiting its use to patients with monthly access to an infusion center. Furthermore, belatacept-based immunosuppression carries an increased risk of posttransplant lymphoproliferative disease in Epstein-Barr virus (EBV)-seronegative recipients [21]. Perhaps the most notable limitation is the observation that transplant recipients receiving belatacept-based immunosuppression have higher rates of early acute cell-mediated rejection [20]. These episodes were usually treatable, but suggest that not all T cell subsets are equally susceptible to costimulatory blockade with belatacept. More recent analyses of immune phenotypes in recipients of belatacept who experienced rejection have demonstrated subsets of T cells associated with belatacept resistance including CD45RA-CCR6+ memory Th17 cells and CD57+PD1-CD4+ T cells [22,23^■]. These results have prompted further investigation into novel costimulation blockade based immunosuppressive regimens in transplantation.

COSTIMULATION BLOCKADE COMPATIBLE AGENTS IN TRANSPLANTATION

Standard maintenance immunosuppression for kidney transplant recipients is based on calcineurin inhibition and corticosteroids. Although these regimens have proven to reduce acute rejection episodes, they failed to improve long-term allograft survival [24-26]. Additionally, the principal side-effects include diabetes, hypertension, hyperlipidemia, neurotoxic, and nephrotoxic effects [27,28]. Therefore, the focus has shifted toward calcineurin inhibitor-free and steroid avoidance regimens. Costimulation blockade has emerged as an essential part of these regimens. Most notably, belatacept spares many of the above listed side-effects of conventional immunosuppression [17,29]. Concerns about increased costs and incidence of post-transplant lymphoprolierative disease in EBV seronegative recipients have prevented widespread acceptance of belatacept as primary therapy [21].

Therefore, optimization was studied in the form of combined belatacept and T cell depletion and added mycophenolate (MMF) or sirolimus, with the goal to wean patients to belatacept monotherapy long-term. Although calcineurin inhibitors (CNIs) successfully prevent early acute rejection, they also antagonize the mechanism by which costimulation blockade facilitates long-term allograft acceptance. In contrast, mammalian target of rapamycin inhibitors promote the effects of costimulation blockade [30,31]. Despite this evidence, a recent retrospective analysis of 745 kidney transplant recipients who received a combination maintenance immunosuppression of belatacept and transient tacrolimus for at least 9 months, has shown comparable rates of acute rejection episodes compared to standard CNI-based regimens with superior long-term graft function [32^■]. In the same study, a significant reduction of de-novo donor-specific antibody (DSA) formation in the belatacept group was observed after 1 year (8.82% vs. 4.14%, P = 0.04) which likely contributes to the improved long-term graft function [32^■]. The ability of belatacept to prevent de-novo DSA formation and antibody-mediated rejection (AMR) has also been shown in the BENEFIT trial and in NHP renal allotransplantation [33,34^■].

In a single-institution randomized trial with 89 patients, immunosuppression with a combination of beletacept and MMF, belatacept and sirolimus, or tacrolimus and steroids had comparable graft and patient survival with acceptable rates of acute rejection (12% vs. 4% vs. 3%, respectively) and improved renal function [35]. However, belatacept in combination with MMF seemed to have higher rates of acute rejection which because of the cohort size did not reach statistical significance. Additionally, the most recent clinical trial CTOT-16 ( NCT01856257) evaluating the efficacy of belatacept and MMF in combination with thymoglobulin induction therapy had to be terminated because of high rejection rates. Belatacept and sirolimus in combination with alemtuzumab have been successfully applied with low rejection rates and excellent graft function with 7 out of 10 patients transitioned to belatacept monotherapy [36]. These two pilot studies have demonstrated the value of belatacept in combination with mammalian target of rapamycin inhibition as potent CNI-free and steroid avoiding immunosuppression in kidney transplantation with excellent long-term graft function and a favorable side-effect profile. A clinical trial comparing maintenance immunosuppression with belatacept and everolimus to Tacrolimus and MMF after thymoglobulin induction is currently ongoing ( NCT02137239).

TARGETING ALTERNATIVE COSTIMULATORY PATHWAYS

Targeting CD40L: thromboembolic complications drive development of modified CD40L-binding agents

Blockade of the CD40-CD40L axis has remained an attractive therapeutic target in transplantation as work in 1996 showed that monotherapy with anti-CD40L monoclonal antibody (mAb) MR1 prolonged cardiac allograft survival in both naive and sensitized mice [37]. Early preclinical work on CD40L blockade was closely tied to co-administration of CTLA4Ig, as the combination of these agents greatly increased allograft survival in mice and in NHPs [38,39]. However, subsequent studies showed that an extended course of anti-CD40L monotherapy prolonged kidney allograft survival in NHPs [40]. Anti-CD40L as monotherapy or used in combination with other drugs showed extraordinary promise in NHP models of renal, islet, skin, and heart allotransplantation (reviewed in [41]).

Based on the preclinical success of anti-CD40L therapy, a clinical trial was initiated in the late 1990s using a humanized anti-CD40L mAb engineered by Biogen Inc. (Cambridge, Massachusetts, USA) (ruplizumab/BG9588) in kidney transplantation [42]. Ruplizumab and a competing anti-CD40L mAb (toralizumab/IDEC-131) made by Idec Pharmaceuticals Corp. (San Diego, California, USA) were used in clinical trials for multiple autoimmune conditions, including multiple sclerosis, immune thrombocytopenic purpura, and systemic lupus erythematosus. The Biogen trials were halted by the FDA after 10 of 100 total trial patients taking ruplizumab/BG9588 experienced thromboembolic complications. After Biogen and Idec merged, development of both ruplizumab/BG9588 and toralizumab/IDEC-131 was discontinued because of concerns about thromboembolic complications. Similar clotting complications associated with anti-CD40L monoclonal antibodies have been reported in NHPs [43]. Further study has shown that anti-CD40L monoclonal antibodies are able to directly activate platelets through CD40L expressed on the platelet surface [44,45]. Anti-CD40L monoclonal antibodies can also form immune complexes with soluble CD40L, and these immune complexes are capable of powerfully activating platelets through the Fc-γ receptor IIA (FCGR2A) [46,47, 48^■, 49].

Concerns about FCGR2A-mediated platelet activation have driven development of three novel CD40L-binding molecules, all of which have undergone early phase clinical trials for autoimmune indications, and one of which has shown efficacy in NHP kidney transplantation (reviewed in [50]). All three molecules have been modified to avoid the FCGR2A-mediated thrombosis associated with the original anti-CD40L monoclonal antibodies.

Letolizumab/BMS-986004 is an anti-CD40L domain antibody (anti-CD40L dAb) engineered to express a mutated IgG1 that lacks Fc-binding and complement-fixing effector functions. At high doses, this agent successfully prolonged kidney allograft survival in rhesus macaques without evidence of thromboembolic complications [48^■]. Letolizumab has undergone safety and efficacy evaluation in an open label phase I/II clinical trial for immune thrombocytopenic purpura, with results pending ( NCT02273960). Dapirolizumab pegol/CDP7657 is an anti-CD40L Fab’ antibody fragment that lacks a functional Fc domain and is conjugated to polyethylene glycol (PEG). [51] This molecule was shown to inhibit the immune response to tetanus toxoid with no evidence of thromboembolic complications in NHPs [51], and a phase IIb clinical trial for systemic lupus erythematosus has been completed, with data evaluation ongoing ( NCT02804763). VIB4920 (formerly MEDI4920) is a CD40L-binding protein made from two Tn3 proteins fused to human serum albumin [52]. This molecule does not have an Fc domain. A phase Ib clinical trial for adult-onset rheumatoid arthritis has been completed ( NCT02780388), and a phase II trial is planned.

Targeting CD40

Owing to the early thromboembolic complications associated with anti-CD40L monoclonal antibodies, CD40 became an increasingly attractive therapeutic target, with initial studies using anti-CD40 agents showing prolonged pancreatic islet allograft survival in NHPs and leading to the clinical development of several anti-CD40 drugs [53-55]. CFZ533 is an anti-CD40 mAb with a modified Fc domain that has led to prolonged allograft survival in NHP kidney allotransplantation [56]. In a phase I/II clinical trial, CFZ533 showed efficacy similar to standard triple drug immunosuppression but was associated with fewer infectious complications and serious adverse events [57^■]. Bleselumab/ASKP1240 is a fully human IgG4 anti-CD40 mAb that has been used as maintenance monotherapy to promote both liver and kidney allograft survival in NHP models [58,59]. A phase IIa clinical trial of bleselumab/ASKP1240 for plaque psoriasis has been completed with no evidence of clinical benefit [60], and a phase II trial is ongoing for prevention of focal segmental glomerulosclerosis in kidney transplant patients ( NCT02921789). BI-655064 is a humanized anti-CD40 mAb that has not been tested in transplantation but is involved in an ongoing phase II clinical trial for lupus nephritis ( NCT03385564).

Targeting CD40L/CD40: preclinical molecules

A suite of novel molecules targeting CD40L and CD40 are under preclinical investigation, including monoclonal antibodies, peptides, and small interfering RNA, and are reviewed in reference [50].

The ICOS-ICOSL and OX40-OX40L pathways: targeting secondary costimulatory pathways to overcome costimulation blockade-resistant rejection

Blockade of the ICOS-ICOSL pathway has been proposed as a therapeutic strategy with the potential to synergize with blockade of the CD28-CD80/86 pathway. Because ICOS is not expressed by naïve T cells but is rapidly upregulated after T cell activation, it was hypothesized that ICOS-ICOSL blockade could control costimulation blockade-resistant T cell populations and could ultimately be used as adjuvant therapy to reduce the rate of costimulation blockade-resistant rejection seen with belatacept therapy. Although combined ICOS-ICOSL and CD28-CD80/86 blockade prevented allograft rejection in mouse studies, ICOS-ICOSL blockade using a novel ICOS-Ig human Fc-fusion protein failed to prolong kidney allograft survival in a NHP model when used alone or in combination with belatacept [61]. Given the lack of efficacy in NHPs, ICOS-ICOSL blocking agents have not advanced to human clinical trials in transplantation.

On the other hand, blockade of the OX40-OX40L pathway has shown promising results. Like ICOS, OX40 is a so-called ‘secondary’ costimulatory molecule that is preferentially expressed on activated T cells and is involved in generating T cell memory [61]. It was shown recently that a humanized anti-OX40L antibody (huMAb OX40L) given in combination with belatacept significantly prolonged kidney allograft survival in NHPs compared to either drug given alone [62^■]. This promising NHP data combined with safety data from a phase II clinical trial of huMAb OX40L for asthma [63] make OX40-OX40L blockade an appealing strategy in transplantation, especially for addressing the problem of costimulation blockade-resistant rejection.

Alternative CD28-blocking agents

Novel agents that directly bind to CD28 have been developed in an effort to selectively block costimulatory signaling without interfering with normal coinhibitory signaling through CTLA-4, a theoretical advantage over CTLA-4Ig (reviewed in [64]). Briefly, this strategy suffered a setback when a phase I clinical trial using a CD28-specific mAb known as theralizumab (TGN1412) reported life-threatening cyotokine release syndrome and nonspecific T cell activation requiring admission to the intensive care unit in all six healthy trial participants [65]. More recently, this strategy has shown encouraging results in NHP models of kidney and heart allotransplantation [66], leading to the development of FR104, a humanized pegylated anti-CD28 Fab’ antibody fragment [67]. FR104 itself has demonstrated encouraging effects in NHP kidney allotransplantation, including prolonged allograft survival, inhibition of donor-specific antibodies, and superiority to belatacept in preventing steroid-resistant acute rejection [67,68^■]. Lulizumab, a humanized pegylated anti-CD28 domain antibody, has also been effective in preclinical models of allotransplantation [69,70]. To date, FR104 and lulizumab are devoid of the cytokine release syndrome associated with TGN1412 and clinical trials in autoimmunity are underway using both agents [29].

EMERGING FIELDS FOR COSTIMULATION BLOCKADE

Despite two phase III trials, BENEFIT and BENEFIT-EXT, showing noninferior outcomes as well as improved renal function and cardiovascular outcomes at one year compared to cyclosporine [21,71], belatacept implementation clinically has been slow. Over 90% of US kidney transplant patients are under tacrolimus-based regimens [72]. Nevertheless, costimulation blockade has shown some advantages with respect to targeting the humoral response and in xenotransplantation in experimental models.

Tolerance induction via promoting chimerism

Clinical tolerance induction in kidney transplantation has been actively attempted in a few institutions using bone marrow transplantation coupled with myeloablative conditioning to facilitate allogeneic bone marrow engraftment and chimerism (reviewed in [73]). To reduce toxicity, costimulation blockade had been evaluated to avoid myeloablative conditioning [74]. Several NHP studies were done to establish a conditioning regimen using belatacept, anti-CD40 mAb, and/or anti-CD40L mAb [75-77,78^■]. Anti-CD40L mAb promoted transient mixed chimerism (11/13) and long-term tolerance (6/13) but later these study participants showed late onset of AMR (3/6) with EBV-related lymphoma or thrombogenic complications (3/6). Interestingly, combined CTLA4Ig and anti-CD40 mAb (dual costimulation blockade) for tolerance induction showed less robust outcomes such that all animals rejected their grafts early, mostly because of AMR leading the authors to conjecture that anti-CD40mAb limits Treg expansion and promotes B cell responses [76].

Costimulation blockade and allo-humoral response

AMR is considered to be the major contributor to immunologic graft loss in kidney transplantation. Conventional immununosuppressive agents do not sufficiently suppress the B cell response. The effect of costimulation blockade on alloimmune humoral responses is well known (reviewed in [79]). Clearly, tacrolimus and rapamycin also have shown substantial suppression of humoral responses either directly or indirectly [80,81]. It is likely that there are some T and B cell repertoires resistant to these agents. There are many potential reasons for the superiority of costimulation blockade in controlling the B cell response. For instance, Tfh cells are specialized T cells that provide B cell help in the germinal center (GC) (reviewed in [82]). It has been postulated that the lack of durable control over Tfh cells by conventional immunosuppression could unleash the GC-driven antibody response. Conceptually, costimulation blockade should be excellent agents to target the GC response as this interaction relies on many costimulatory signalings including CD28-CD80/86, CD40-CD40L, ICOS-ICOSL, PD1-PDL1/2, and so on (reviewed in [82]). Clincal data reveal that patients on belatacept-based immunosuppression have lower incidences of donor-specific antibody production and AMR [36,83^■]. We reported that preemptive treatment with CTLA4Ig or anti-CD40 mAb abrogated the development of posttransplant GC responses as well as AMR injury in a NHP model [33]. Blocking the CD28-CD80/86 costimulation signal before immune activation, more specifically T–B conjugation, down-modulated the allo-B cell response. Limited efficacy of costimulation blockade was expected after initiation of an immune response as activated or memory T and B cells do not have the same requirements for costimulation as naïve cells. However, we observed a great reduction of GC responses as well as memory T cells in sensitized NHPs with belatacept and anti-CD40mAb [84^■]. We and others suggested targeting GC activation using costimulation blockade together with proteasome inhibitors to target plasma cells [85,86] and reduce an ongoing allo-humoral response and its rebound (Fig. 2).

FIGURE 2.

Costimulatory blockade as an adjunct for desensitization in transplantation. (a) Desensitization strategies using a single agent targeting plasma cells such as proteasome inhibitors (middle panel) leads to rapid compensation by the germinal center and rebound humoral immunity (right panel). (b) Combining the strategy in (a) with costimulatory blockade in a dual targeting approach (middle panel) quiets the germinal center in the follicles of the lymph node and prevents rapid humoral rebound (right panel). Illustrated by Megan Llewellyn, MSMI, CMI; Duke Surgery.

Xenotransplantation

Initially, the development of genetically modified pigs such as galactose-a1,3-galactose (aGal) knock out [87,88] and later pigs transgenic for human complement regulatory proteins (hCD55), human membrane cofactor protein (hCD46), and/or human thrombomodulin (hTBM) had overcome the profound barrier of hyperacute rejection and prolonged xenograft survival (reviewed in [89]). However, overall graft survival of life-supporting organs remained only a few weeks. In 2016, mean xenograft survival of 433 days was reported for heterotopic heart transplantation [90^■]. Later in 2018, Längin et al. reported up to 195 days graft survival with in a pig-to-baboon orthotopic heart transplantation model [91^■]. Most recently, Adams et al. [92^■] reported more than 400 days of kidney xenograft survival in a pig-to-rhesus model. Costimulation blockade, specifically anti-CD40L mAb, was used in these studies. Targeting costimulation controls the xeno-immune response effectively compared to conventional CNI-based immunosuppression [93,94]. Notably, targeting the CD40-CD40L signaling axis seems more crucial to promoting long-term graft survival in xenotransplantation than the CD28-CD80/86 axis [95-97]. Costimulation blockade targeting the CD40/CD40L axis may be very useful at least in initial approaches to clinical xenotransplantation.

CONCLUSION

Transplant specialists increasingly recognize the importance of downregulating the B cell alloimmune response to prevent or suppress an alloantibody response and the associated graft injury. The value of costimulatory blockade will likely grow in importance and clinical use given the efficacy and safety of such agents in transplantation. In addition to belatacept, anti-CD40 biologics are in development as are anti-CD40L and others including biologics targeting ICOS and OX40. As experimental results of xenotransplantation appear to be superior with anti-CD40L, it may be that costimulation blockade will allow clinical introduction of xenotransplants. Future clinical trials will be needed to compare the different costimulatory blockers to understand their unique properties and benefits. We are fortunate to have a plethora of agents to evaluate for their potential applications to solid organ transplantation and await their rational introduction.

KEY POINTS.

• Belatacept became the first costimulation blockade agent approved by the FDA for immunosuppression in kidney transplant recipients in 2011.

• Limitations such as the intravenous formulation of belatacept, higher rates of acute rejection (belatacept resistance), and higher rates of post-transplant lymphoprolierative disease in EBV seronegative recipients have prompted new clinical trials to find costimulation blockade compatible agents for optimal maintenance immunosuppression.

• Novel agents targeting other costimulatory molecules such as CD28, CD40, CD40L, OX40L, and ICOS are in preclinical or early phase clinical trials with some promising results in preventing allograft rejection.

• Preclinical experience with costimulation blockade agents in tolerance induction regimens, xenotransplantation, and treating antibody-mediated rejection are shaping the design of future clinical applications for costimulation blockade in organ transplantation.