(Universal Activity Number H01- P) Please note: If you participated in the activity with UAN L01- P, you are

Transcription

1 CME Information Release Date: January 2, 2018 Expiration Date: January 2, 2019 UAN Live Release Date: June 12, 2016 Estimated Time to Complete this Activity: 1.25 hours Overview This article reviews the pathophysiology of antibody- mediated rejection (AMR) with emphasis on the role of complement in allograft injury. It reviews the required features for a definitive diagnosis of AMR according to the latest Banff criteria, which incorporate recent findings delineating the complex spectrum of AMR and the diverse phenotypes seen. Clinical data on current and investigational treatments are summarized, and strategies to improve outcomes are presented. Lastly, a discussion of relevant predictive and surrogate endpoints provides a framework for future studies. Target Audience This activity has been designed to meet the educational needs of physicians, surgeons, scientists, nurses, organ procurement personnel, and pharmacists who care for patients with AMR. Educational Objectives After participating in this educational initiative, the participant should be better able to: Explain the significance of activation of the complement system in the development of AMR Select appropriate therapeutic agents and protocols to prevent and treat AMR based on pathophysiology Evaluate the use and clinical implications of surrogate endpoints for clinical trials of emerging treatments for AMR Provide accurate and appropriate counsel as part of the treatment team Faculty Robert A. Montgomery, MD, DPhil, FACS Professor of Surgery Director, NYU Langone Transplant Institute NYU Langone Medical Center New York, NY, USA Alexandre Loupy, MD, PhD Associate Professor Paris Translational Research Center for Organ Transplantation and Department of Nephrology and Kidney Transplantation, Hôpital Necker, INSERM U 970, Paris Descartes University Paris, France Dorry Lidor Segev, MD, PhD Professor of Surgery and Epidemiology Associate Vice Chair of Research, Department of Surgery Johns Hopkins University Baltimore, MD, USA Statement of Support This activity is jointly provided by RMEI Medical Education, LLC and Postgraduate Institute for Medicine. This activity is supported by an independent medical educational grant from Shire. For questions, please contact PIM at Joint Accreditation Statement In support of improving patient care, this activity has been planned and implemented by the Postgraduate Institute for Medicine and RMEI Medical Education, LLC. Postgraduate Institute for Medicine is jointly accredited by the Accreditation Council for Continuing Medical Education (ACCME), the Accreditation Council for Pharmacy Education (ACPE), and the American Nurses Credentialing Center (ANCC), to provide continuing education for the healthcare team. Physician Continuing Medical Education The Postgraduate Institute for Medicine designates this enduring material for a maximum of 1.25 AMA PRA Category 1 Credit(s) TM. Physicians should claim only the credit commensurate with the extent of their participation in the activity. Continuing Pharmacy Education Postgraduate Institute for Medicine is accredited by the Accreditation Council for Pharmacy Education as a provider of continuing pharmacy education. Postgraduate Institute for Medicine designates this continuing education activity for 1.25 contact hour(s) (0.125 CEUs) of the Accreditation Council for Pharmacy Education. [Correction added on 22 March, 2018, after first online publication: The accreditation statement and CEUs were updated.] Am J Transplant. 2018;18(Suppl. 3):3 17. amjtransplant.com 2017 The American Society of Transplantation 3 and the American Society of Transplant Surgeons

2 (Universal Activity Number H01- P) Please note: If you participated in the activity with UAN L01- P, you are not able to claim credit for this activity. For Pharmacists: Upon successfully completing the post- test with a score of 75% or better and the activity evaluation form, transcript information will be sent to the NABP CPE Monitor Service. Activity Type: Application Method of Participation and Request for Credit There are no fees for participating and receiving CME credit for this activity. During the period, January 2, 2018 through January 2, 2019, participants must read the learning objectives and faculty disclosures and study the educational activity. PIM supports Green CME by offering your Request for Credit online. If you wish to receive acknowledgment for completing this activity, please complete the post- test and evaluation on On the navigation menu, click on Find Post- test/evaluation by Course and search by course ID Upon registering and successfully completing the post- test with a score of 75% or better and the activity evaluation, your certificate will be made available immediately. Processing credit requests online will reduce the amount of paper used by nearly 100,000 sheets per year. Fee Information There is no fee for this educational activity. Media Print Journal Supplement Software Requirements for Online Version This activity requires a modern web browser (Internet Explorer 7.0 or later, Mozilla Firefox, Apple Safari, Google Chrome). Disclosure of Conflicts of Interest Postgraduate Institute for Medicine (PIM) requires instructors, planners, managers, and other individuals who are in a position to control the content of this activity to disclose any real or apparent conflict of interest (COI) they may have as related to the content of this activity. All identified COI are thoroughly vetted and resolved according to PIM policy. PIM is committed to providing its learners with high- quality CME activities and related materials that promote improvements or quality in healthcare and not a specific proprietary business interest of a commercial interest. The faculty reported the following financial relationships or relationships to products or devices they or their spouse/life partner have with commercial interests related to the content of this CME activity: Robert A. Montgomery, MD, DPhil, FACS, has affiliations with Alexion, Astellas, Genentech/Roche, Hansa Medical, Immucor, Novartis, Shire, and True North/iPierian (Travel and Consulting Fees); and Alexion, CSL Behring, Hansa Medical, Immune Tolerance Network, and ViroPharma (Contracted Research/Grants). Alexandre Loupy, MD, PhD, has no affiliations with commercial interests to disclose. Dorry Lidor Segev, MD, PhD, has no affiliations with commercial interests to disclose. The RMEI Medical Education, LLC planners reported the following financial relationships or relationships to products or devices they or their spouse/ life partner have with commercial interests related to the content of this CME activity: Jacqui Brooks, MBBCh, MRCPsych, reported that she does not have any affiliations with commercial interests to disclose. Sherri Kramer, MD, reported that she does not have any affiliations with commercial interests to disclose. Kyra Sheahan reported that she does not have any affiliations with commercial interests to disclose. Amy Reeve reported that she does not have any affiliations with commercial interests to disclose. Marie Sabo Recine, MS, MT (ASCP), reported that she does not have any affiliations with commercial interests to disclose. The following PIM planners and managers, Judi Smelker- Mitchek, RN, BSN; Trace Hutchison, PharmD; Samantha Mattiucci, PharmD, CHCP; and Jan Schultz, RN, MSN, CHCP, hereby state that they or their spouse/life partner do not have any financial relationships or relationships to products or devices with any commercial interest related to the content of this activity of any amount during the past 12 months. Disclosure of Unlabeled Use This educational activity may contain discussion of published and/or investigational uses of agents that are not indicated by the US Food and Drug Administration (FDA). The planners of this activity do not recommend the use of any agent outside of the labeled indications. The opinions expressed in the educational activity are those of the faculty and do not necessarily represent the views of the planners. Please refer to the official prescribing information for each product for discussion of approved indications, contraindications, and warnings. Disclaimer Participants have an implied responsibility to use the newly acquired information to enhance patient outcomes and their own professional development. The information presented in this activity is not meant to serve as a guideline for patient management. Any procedures, medications, or other courses of diagnosis or treatment discussed or suggested in this activity should not be used by clinicians without evaluation of their patients conditions and possible contraindications and/or dangers in use, review of any applicable manufacturer s product information, and comparison with recommendations of other authorities. Acknowledgments The authors wish to thank Marie Sabo Recine, MS, MT (ASCP); Jacqui Brooks, MBBCh, MRCPsych; Sherri Kramer, MD; and Kyra Sheahan for their assistance in writing and editing this article. The individuals mentioned report no potential COI and have given their written permission to be named.

3 Received: 17 August 2016 Revised: 26 October 2017 Accepted: 4 November 2017 DOI: /ajt SPECIAL ARTICLE Antibody- mediated rejection: New approaches in prevention and management R. A. Montgomery 1 A. Loupy 2 D. L. Segev 3 1 Department of Surgery and NYU Langone Transplant Institute, NYU Langone Medical Center, New York, NY, USA 2 Paris Translational Research Center for Organ Transplantation and Department of Nephrology and Kidney Transplantation, Hôpital Necker, INSERM U 970, Paris Descartes University, Paris, France 3 Department of Surgery, Johns Hopkins University, Baltimore, MD, USA Correspondence Robert A. Montgomery Robert.Montgomery@nyumc.org Funding information Funding support provided in the form of an independent medical educational grant from Shire. Despite the success of desensitization protocols, antibody- mediated rejection (AMR) remains a significant contributor to renal allograft failure in patients with donorspecific antibodies. Plasmapheresis and high- dose intravenous immunoglobulin have proved to be effective treatments to prevent and treat AMR, but irreversible injury in the form of transplant glomerulopathy can commonly manifest months to years later. There is an unmet need to improve the outcomes for patients at risk for AMR. Updated Banff criteria now take into account the increasing understanding of the complex and heterogeneous nature of AMR phenotypes, including the timing of rejection, subclinical and chronic AMR, C4d- negative AMR, and antibody- mediated vascular rejection. Treatment for AMR is not standardized, and there is little in the way of evidencebased treatment guidelines. Refining more precisely the mechanisms of injury responsible for different AMR phenotypes and establishing relevant surrogate endpoints to facilitate more informative studies will likely allow for more accurate determination of prognosis and efficacious intervention using new therapeutic approaches. In addition to plasma exchange and intravenous immunoglobulin, a number of other add- on therapies have been tried in small studies without consistent benefit, including anti- CD20, proteasome inhibitors, complement inhibitors, anti- interleukin- 6 receptor blockers, and immunoglobulin G- degrading enzyme of Streptococcus pyogenes (called IdeS). KEYWORDS antibody-mediated (ABMR), autoantibody, basic (laboratory) research/science, clinical research/ practice, kidney transplantation/nephrology, liver transplantation/hepatology, organ procurement and allocation, rejection: 1 INTRODUCTION Antibody- mediated rejection (AMR) is a significant complication following kidney transplantation that contributes toward both shortand long- term injury in approximately 1% to 10% of kidney transplant recipients. 1 The presence of antibodies that recognize donor human leukocyte antigens (HLAs), as well as incompatible ABO blood group antigens and other endothelial or xenogeneic antigens, has been associated with AMR and subsequent graft loss for some time, and the presence of donor- specific antibodies (DSAs) was once considered a contraindication to transplantation. 2 Despite use of desensitization protocols, up to one-third of highly sensitized recipients may develop acute AMR following transplantation. 1 DSAs present in the serum of sensitized patients are produced from long- lived plasma cells (LLPCs) Abbreviations: ahus, atypical hemolytic uremic syndrome; AMR, antibody-mediated rejection; AMVR, antibody-mediated vascular rejection; C1-INH, C1 esterase inhibitor; cg [score], chronic glomerulopathy; CI, confidence interval; DSA, donor-specific antibody; egfr, estimated glomerular filtration rate; ENDAT, endothelial-associated transcript; FDA, Food and Drug Administration; g [score], glomerulitis; HLA, human leukocyte antigen; HR, hazard ratio; IdeS, immunoglobulin G-degrading enzyme of streptococcus pyogenes; IF/TA, interstitial fibrosis and tubular atrophy; IgG, immunoglobulin G; IL-6R, interleukin-6 receptor; IVIg, intravenous immunoglobulin; LLPC, long-lived plasma cell; MFI, mean fluorescence intensity; ptc, peritubular capillary; SOC, standard of care; TCMR, T cell mediated rejection.

4 6 protected and nurtured in bone marrow niches, a scenario equivalent to a vaccine response from a previous exposure to allogeneic HLA molecules. 2 These LLPCs do not have a significant capacity to increase basal antibody production in response to donor HLA. Thus, AMR following transplantation in a sensitized patient likely results from a primed memory B cell response in regional lymphoid tissue. B cells clonally expand and differentiate into plasmablasts and plasma cells in response to HLA activation and T cell derived growth signals. These cells may traffic to the spleen as short- lived blasts and plasma cells or to the bone marrow as LLPCs. AMR is also of significant concern in nonsensitized individuals, as de novo DSAs can develop early or late after transplantation. 2 Early appearance of de novo DSAs weeks to months after transplantation may cause acute AMR and can be severe. AMR is seen in up to 5% of first transplants and in some cases may represent an anamnestic response from cryptic sensitization in which HLA- specific memory B cells are present but soluble antibody is not being produced or is not detectable. Late or chronic AMR may result from de novo DSA formation, incomplete elimination of DSA following an acute AMR episode, or persistence of preformed DSA after desensitization. AMR is often diagnosed late and may only manifest as an insidious rise in serum creatinine. De novo antibody formation and chronic AMR are thought to be a significant cause of premature graft failure. The deleterious impact on graft half- life of all forms of AMR can be profound. 3 Orandi and colleagues found a fold higher risk of graft loss in patients with AMR (95% confidence interval [CI], ; P =.006). 4 In the case of deceased donor transplants in which lowlevel DSA is present, they found a somewhat attenuated AMR effect on graft loss (hazard ratio [HR] = 2.39; 95% CI, ; P =.028), while in HLA- incompatible live donor transplants, the impact of AMR appeared even stronger (HR = 6.29; 95% CI, ; P <.001), likely due to higher pre- desensitization antibody strength. The best treatment for AMR is prevention. While AMR will always be a feature of the practice of allogeneic transplantation, the incidence can be greatly reduced by measures to reduce risk by avoiding strong DSA. The new kidney allocation system, which gives priority to the allocation of non- normative donor HLA genotypes to highly sensitized patients, is an example of how policy can influence transplant outcomes and equity. Kidney paired donation, combining paired donation with desensitization, and acceptable mismatch strategies all are ways of reducing the risk of AMR and its consequences PATHOPHYSIOLOGY OF COMPLEMENT- MEDIATED AMR There has been an increasing awareness of the important role of complement in the pathophysiology of acute AMR. The link between complement and AMR was confirmed when the complement split product C4d was demonstrated in transplant biopsies. 9 Corroboration by additional investigations led to development of consensus diagnostic criteria for acute and chronic AMR in renal transplantation that included detection of DSA in the serum and C4d deposition in the allograft. 10,11 The major mechanism involved in antibody- mediated kidney injury is activation of the classical complement pathway by the binding of DSA to HLA and subsequent binding of the C1 complex, which ultimately leads to formation of the membrane attack complex (C5b- C9) and results in direct injury to the allograft (Figure 1). 12 The extent of complement activation by immune complexes is influenced by a variety of local and systemic factors, including antibody strength, 13,14 antibody isotype, epitope density, and the local concentration of complement regulatory proteins. 15 Immunoglobulin G (IgG)3 and IgG1 DSAs have strong capability to bind C1q and activate the classical pathway, whereas IgG2 and IgG4 DSAs have weaker C1- fixing properties. 15,16 FIGURE 1 The role of the classical complement pathway in acute AMR in sensitized renal transplant recipients. Following binding of DSA to the allograft vascular endothelium, the C1 complex activates the serine esterases C1s and C1r, resulting in the cleavage of C4, deposition of C4d, and the assembly of the classical pathway C3 convertase. C3 convertase cleaves C3 into C3a, a potent pro- inflammatory mediator, and C3b, which propagates the complement cascade and leads to the formation of the pro- inflammatory mediator C5a and the lytic membrane attack complex (C5b- C9). Investigational therapies are shown in red. Adapted from Stegall et al. 12 AMR, antibody- mediated rejection; C1- INH, C1 esterase inhibitor; DSA, donor- specific antibody; HLA, human leukocyte antigen

5 7 The target antigens in AMR are primarily located on the vascular endothelium, resulting in histological findings of acute microvascular injury (glomerulitis and peritubular capillaritis) and chronic vascular injury (transplant glomerulopathy). Endothelial damage also leads to platelet activation and the formation of microthrombi. 17 C4d, a degradation product of C4, binds at the site of complement activation and remains covalently attached and detectable by immunohistochemistry for several weeks. 12 The byproducts of complement activation (C3a and C5a) act as anaphylatoxins, leading to chemotaxis and amplification of inflammation through innate immune responses. Longstanding inflammation results in cell proliferation, basement membrane duplication, and mesangial interposition that is apparent as glomerular basement membrane splitting and peritubular capillary (ptc) basement membrane multilayering, respectively, on light and electron microscopy AMR DIAGNOSIS, CLASSIFICATION, AND RISK STRATIFICATION Previous Banff schema (2001/2007) for diagnosis of acute and chronic AMR in renal allografts recognized the role of DSAs in graft injury. 10,11 Required features for a definitive diagnosis of AMR at the time included morphologic evidence (primarily microvascular inflammation), immunohistologic evidence (ptc C4d staining), and serologic evidence (presence of circulating DSAs). The complex spectrum of AMR in renal allografts is only recently becoming better defined. AMR has multiple phenotypes related to the mechanism of DSA production and injury (preformed DSA vs de novo DSA), the strength of antibody reactivity, and the timing of rejection. 2,18 Updated 2013 Banff criteria for AMR now take into account recent data supporting the recognition of subclinical AMR as part of the AMR spectrum, the existence of C4d- negative AMR, and antibodymediated vascular rejection (AMVR) characterized by intimal arteritis (Table 1). 19 Although AMR phenotypes are now organized as acute AMR, subclinical AMR, and chronic AMR, it is recognized that AMR is a continuous process with varying degrees of activity and damage. Further refinement of the Banff criteria will help determine whether this continuum corresponds to distinct pathophysiology, molecular signatures, involvement of complement- dependent and independent processes, and responses to precision therapies, which were topics of discussion during the 13th meeting of the Banff Conference on Allograft Pathology. 20 At present, there is still significant ambiguity in the Banff criteria in terms of the classification and behavior of the different forms of AMR. The Banff classifications are primarily focused on the presence or absence of detectable complement activation (type 1 vs type 2). The timing of the AMR relative to the transplant and the presence of presensitization are important factors in determining phenotype and response to therapy. Patients with preformed DSA, many of whom undergo desensitization protocols, are at greatest risk for early acute AMR, primarily due to anamnestic responses. However, these patients can develop acute AMR at any point in time after transplantation. TABLE 1 Summary of revised (Banff 2013) classification of AMR in renal allografts. From Haas 19 Acute/active AMR: all 3 features must be present a,b 1. Histologic evidence of acute tissue injury, including 1 or more of the following: Microvascular inflammation (g > 0 and/or ptc > 0) Intimal or transmural arteritis (v > 0) Acute thrombotic microangiopathy Acute tubular injury 2. Evidence of current/recent antibody interaction with vascular endothelium, including at least 1 of the following: Linear C4d staining in peritubular capillaries At least moderate microvascular inflammation ([g + ptc] 2) Increased expression of validated gene transcripts in the biopsy tissue indicative of endothelial injury 3. Serologic evidence of donor-specific antibodies (DSAs) (HLA or other antigens) Chronic, active AMR: all 3 features must be present a 1. Morphologic evidence of chronic tissue injury, including 1 or more of the following: Transplant glomerulopathy (cg > 0), if no evidence of chronic thrombotic microangiopathy (TMA) Severe peritubular capillary basement membrane multilayering Arterial intimal fibrosis of new onset 2. Evidence of current/recent antibody interaction with vascular endothelium, including at least 1 of the following: Linear C4d staining in peritubular capillaries At least moderate microvascular inflammation ([g + ptc] 2) Increased expression of validated gene transcripts in the biopsy tissue indicative of endothelial injury 3. Serologic evidence of DSAs (HLA or other antigens) C4d staining without evidence of rejection: all 3 features must be present 1. Linear C4d staining in peritubular capillaries 2. g = 0, ptc = 0, cg = 0, v = 0; no TMA, no peritubular capillary basement membrane multilayering, no acute tubular injury 3. No acute cell-mediated rejection or borderline changes AMR, antibody- mediated rejection; cg, Banff chronic glomerulopathy score; DSA, donor- specific antibody; g, Banff glomerulitis score; HLA, human leukocyte antigen; ptc, peritubular capillary; v, Banff arteritis score. a For all AMR diagnoses, it should be specified whether the lesion is C4dpositive or without evident C4d deposition. b These lesions may be clinically acute, smoldering, or subclinical. Early acute AMR can be severe and result in graft loss, but it is also potentially responsive to treatment and, when it is, often results in the disappearance of DSA. 21 Late acute AMR (more than 3 months posttransplant) can be a mixed cellular and antibody rejection, and it is often recalcitrant to standard of care (SOC) therapy, as is chronic AMR and, in some cases, subclinical AMR, which are commonly associated with transplant glomerulopathy and reduced graft half- life. This review will attempt to make these distinctions when describing histologic phenotypes and efficacy of available treatments. 3.1 Complement activation and AMR Findings from a number of investigations have demonstrated that DSAs may cause graft injury independent of C4d deposition. 22,23 For example, Sis et al found that in renal transplant patients with DSAs,

6 8 high endothelial- associated transcript (ENDAT) expression was an indicator of active antibody- mediated allograft damage and was associated with a significantly increased rate of graft loss, even in the absence of C4d. 24 The findings of Loupy et al support the existence of C4d- negative AMR that, although less severe than C4d- positive AMR, is still associated with the development of chronic allograft changes and reduced allograft half- life. 25 These investigators evaluated the presence of subclinical AMR in patients receiving deceased donor grafts and correlated histology with immunological and functional parameters. Patients with DSA (n = 45) could be classified into 1 of 3 categories based on pathologic findings at 3 months posttransplantation: (1) no evidence of AMR, (2) subclinical AMR characterized by histologic features of AMR (glomerulitis and peritubular capillaritis) and ptc C4d staining, and (3) histologic features of AMR but no C4d staining. At 1- year posttransplantation, patients with evidence of subclinical AMR at 3 months had reduced graft function, all had interstitial fibrosis and tubular atrophy (IF/TA), and 43% also had transplant glomerulopathy. Patients with AMR histology who were C4d negative at 3 months had an intermediate course between the 2 other cohorts with regard to creatinine clearance and frequencies of IF/TA (77%) and transplant glomerulopathy (18%) at 1 year. This work was further developed when the authors studied 80 patients who had preformed DSA with protocol biopsies at 3 and 12 months. Ninety biopsies were C4d- positive, and 67 were C4d- negative. Almost all the C4d- positive biopsies were associated with microvascular inflammation, and more than half of the C4d- negative biopsies showed microvascular inflammation, again suggesting that complement activation is not a requirement for inflammation, injury, and reduced graft half- life. 26 According to the Banff 2013 classification, lesions in acute and active AMR may be clinically acute, smoldering, or subclinical. 19 The requirement for C4d positivity for diagnosis of AMR has been replaced by a category of current and recent evidence of antibody interaction with the endothelium that includes at least 1 of the following: C4d staining, at least moderate microvascular inflammation, or molecular evidence based on a thoroughly validated assay. In a recent comparison of the risk of allograft loss among patients with consistently C4d- negative AMR vs patients with C4d- positive AMR at a single center, both phenotypes were associated with increased graft loss compared with AMR- free matched controls. 27 Although graft outcomes overall were better for C4d- negative AMR compared with C4d- positive AMR, the differences in 1-, 2-, and 3- year post- AMR- defining biopsy graft survival between patients with C4dnegative (93.4%, 90.2%, and 90.2%, respectively) and C4d- positive AMR (86.8%, 82.6%, and 77.7%, respectively) were not statistically significant (P =.4). No clinical characteristics that reliably differentiated C4d- negative and C4d- positive AMR were identified. Various techniques for detecting complement activation by allo- HLA have been described, including immunofluorescence, immunoperoxidase, and C1q binding. 28,29 All are adept at demonstrating or predicting C4d- positive AMR. The classic diagnostic pattern of C4d deposition is diffuse staining in the ptc. 30 In a large clinical series, Kikić et al found less commonly recognized C4d deposition patterns, including focal linear C4d in ptc, endothelial C4d in glomeruli, or arteriolar C4d, which correlated with glomerulitis, presensitization, and poor graft survival. 31 These diverse techniques and recognition of novel staining patterns have further refined the diagnosis of C4d- positive AMR and have helped to identify new phenotypes and prognostic criteria. Additionally, the strength of the DSA can be prognostic for complement activation and the more deleterious C4d- positive phenotype. In a study by Eskandary et al, clinically silent (subclinical) AMR was found to be present in about half of their cohort of patients who have DSA more than 6 months after transplantation, and the strength of the antibody (mean fluorescence intensity [MFI]) added predictive value in determining which patients will have AMR on surveillance biopsies Antibody- mediated vascular rejection AMVR is a previously unrecognized type of allograft rejection characterized by the association of circulating DSA with intimal arteritis. Using a population- based approach, Lefaucheur et al were able to distinguish AMVR as a distinct clinical phenotype that differed from T cell mediated vascular rejection, T cell mediated rejection (TCMR) without vasculitis, and AMR without vasculitis. 33 Classification of rejection was based on 7 variables: glomerulitis, peritubular capillaritis, DSA, C4d deposition, interstitial inflammation, tubulitis, and endarteritis. Of the 4 rejection phenotypes, AMVR carried the highest risk of graft loss. Based on these findings, intimal arteritis is now included among those lesions meeting the updated Banff criteria as histologic evidence of acute AMR TREATMENT OF AMR The current strategy for treating AMR is use of a combination of modalities to address multiple pathophysiologic pathways. Treatment regimens of plasmapheresis 34 or immunoadsorption 35 followed by low- dose intravenous immunoglobulin (IVIg; 100 mg/kg) or high- dose IVIg (2 g/kg) with or without steroids have become the SOC based on a preponderance of supporting data rather than randomized controlled trials. 36,37 High- dose IVIg is often used in combination with anti- CD20 (rituximab) despite mixed evidence of its added benefit (see the section that follows on anti- CD20). Kidney Disease: Improving Global Outcomes clinical practice guidelines recommend the use of these 2 treatment regimens, and most insurers will reimburse for therapy involving this combination of modalities. 36,38 A recent clinical practice survey found that AMR treatments vary widely, with most centers using a combination of plasmapheresis and IVIg, with a number also incorporating rituximab. 39 Plasmapheresis and immunoadsorption directly remove IgG from the serum in a fairly predictable fashion but are inefficient inasmuch as IgG is removed only from the vascular space and then must reequilibrate with the interstitium before further antibody removal is optimal, a process that takes about 48 hours. There is evidence that low- dose IVIg dampens endogenous antibody rebound after depletion with plasmapheresis. IVIg also serves to restore antimicrobial immunoglobulins removed during plasmapheresis. Few randomized controlled

7 9 trials are available to assess the effectiveness of plasmapheresis and immunoadsorption in treating AMR. A small but well- designed study by Böhmig et al using immunoadsorption vs conventional immunosuppression for severe C4d- positive AMR was terminated early due to the high rate of graft loss in the control arm. 35 Another early randomized controlled trial comparing plasmapheresis with standard immunosuppression for early vascular rejection (not biopsy- proven AMR) showed no difference in graft survival. 40 Other randomized controlled studies (summarized in Roberts et al) of low or uncertain quality were inconsistent in showing a benefit of plasmapheresis for the treatment of AMR. 36 Investigations in the early 1990s suggested the therapeutic potential of IVIg was due primarily to anti- idiotypic interactions with HLA antibodies. 36,41 At high doses, IVIg exhibits immunomodulatory effects on T and B cells, including induction of B cell apoptosis and modulation of B cell signaling. 42 In a multicenter, placebo- controlled, randomized trial, IVIg alone was shown to increase transplant rates for highly sensitized patients and resulted in a modest reduction in panel reactive antibodies. 43 Other studies have suggested that highdose IVIg alone is only effective against low- level antibodies. 44,45 In an uncontrolled study, the group at Cedars- Sinai Medical Center (in Los Angeles, California, USA) found improved graft survival rates when anti- CD20 was combined with high- dose IVIg. 46 High DSA levels may exceed the therapeutic capacity of both plasmapheresis and IVIg, and tissue destruction due to complement fixation will occur. Anti- CD20, proteasome inhibitors, C5 and C1 esterase inhibitors (C1- INH), antiinterleukin- 6 receptor (IL- 6R) blockers, and a new drug called IdeS [IgG- degrading enzyme of Streptococcus pyogenes] are all being tested as add- on therapy to the SOC plasmapheresis and IVIg. A small phase III, multicenter, double- blind study evaluating rituximab as an addition to plasma exchange, IVIg, and corticosteroids for treatment of acute AMR (n = 38) found no additional benefit of rituximab on a composite primary endpoint of graft loss and renal function at day 12, compared with placebo (52.6% vs 57.9%, respectively; P =.744). 53 At 1 year, no deaths occurred, but 1 graft loss was seen in each group. The authors conclude that the study was underpowered and that important differences between groups may have been missed, but there did not appear to be a benefit to using rituximab for established AMR. There is some evidence that rituximab could be effective in desensitization protocols by preventing anamnestic responses. In a study by Zachary et al, tetramers were used to determine the frequencies of B cells with HLA specificities that are not producing soluble antibody. Patients with elevated B cell frequencies for donor- specific allo- HLA who were treated with rituximab did not produce DSA posttransplant, whereas those who did not undergo B cell ablation had a high rate of developing new DSA posttransplant. 54 In 1 retrospective study comparing 25 desensitized patients using plasmapheresis with rituximab vs 25 patients using plasmapheresis alone, the investigators found significantly less HLA antibody rebound in the rituximab- treated patients, and the magnitude of the increase was significantly larger among patients who did not receive rituximab. 55 A group at Cedars- Sinai Medical Center showed a graft survival advantage to desensitization with IVIg (high dose) combined with rituximab vs IVIg alone. 46,56 Other groups have demonstrated the limited efficacy of high- dose IVIg when used without rituximab. 44 Rituximab is widely used off- label for the prevention and treatment of AMR, without clear evidence of efficacy. 4.1 Anti- CD20 Rituximab, a chimeric monoclonal IgG antibody directed against the CD20 antigen expressed on the surface of pre- B and mature B cells, causes B cell lysis via both complement- dependent cytotoxicity and antibody- dependent cell- mediated cytotoxicity. 47 It is approved for use in several B cell lymphomas and leukemias as well as in rheumatoid arthritis. Rituximab appears to be ineffective for the treatment of AMR. Antibody- secreting plasma cells do not express CD20 and would not be expected to have any response to anti- CD One systematic review and meta- analysis of 10 small retrospective or nonrandomized trials evaluating use of rituximab in AMR suggested that use of this agent is associated with improved graft outcome; however, the favorable effect was driven by 2 larger studies included in the analysis. 49 A systematic review of 21 studies (2 randomized controlled trials and 19 retrospective cohort studies) did not support the use of rituximab in desensitization in highly sensitized recipients. 50 The first controlled trial of rituximab, which was not well designed, showed that adding rituximab to plasmapheresis and IVIg provided improved graft outcome over that seen with IVIg alone. It is likely that the benefit could be attributed to plasmapheresis and not to the rituximab. 51 A study by Oblak et al failed to show any benefit of adding rituximab to SOC treatment for AMR Proteasome inhibitors Bortezomib, which is approved for use in multiple myeloma and mantle cell lymphoma, is a proteasome inhibitor that disrupts the normal intracellular protein degradation process. 57 Plasma cells are sensitive to proteasome inhibition due to their continuous production of immunoglobulins, and in 1 study bortezomib has been shown to induce plasma cell apoptosis and block anti- HLA antibody production. 58 However, bortezomib- induced plasma cell depletion may trigger increased antibody production through germinal center B cell and follicular helper T cell expansion; this may be why only modest effects on HLA antibody strength have been achieved with monotherapy. 59,60 A number of case reports and a small series of patients have noted beneficial effects of bortezomib when used in combination with plasmapheresis and IVIg or rituximab in the treatment of acute AMR in kidney transplant recipients. 48,61 A small study comparing use of 1 cycle of bortezomib with historical controls treated with a single dose of rituximab, both in conjunction with plasmapheresis and IVIg, showed that renal function at 9 months was better in the bortezomib group compared with the historical rituximab group; in addition, graft survival at 18 months was higher (60% vs 11%, P =.0171). 62 Other studies have shown no improvement in glomerular filtration rate after bortezomib when used as add- on therapy with

8 10 plasmapheresis and IVIg for chronic AMR. 63 The finding that bortezomib has little effect on HLA class II antibody production may explain its lack of efficacy for late or chronic AMR IL- 6R blockade A humanized monoclonal antibody against the IL- 6R (tocilizumab) has been used in phase I/II studies for the treatment of chronic active AMR unresponsive to SOC therapy and with high- dose IVIg for patients who are difficult to desensitize. 65 In 1 study of 75 patients with chronic active AMR with or without transplant glomerulopathy, 37 patients who had failed high- dose IVIg, rituximab, and plasmapheresis received monthly doses of tocilizumab for 6 to 18 months and were found to have good outcomes when compared with published results. Anti- IL- 6R therapy has also been used in combination with high- dose IVIg and anti- CD20 for patients who failed SOC desensitization, and it appeared well tolerated and safe. 66 There are no published randomized controlled trials showing the benefit of tocilizumab at this time. 5 COMPLEMENT INHIBITOR STRATEGIES IN AMR Since the activation of complement in the setting of high levels of DSA plays a major role in the pathogenesis of acute AMR, it makes sense that complement blockade would be an important strategy for prevention and treatment of AMR. Agents targeting C5 and C1 esterase have been evaluated in clinical trials. 5.1 C5 inhibition Eculizumab is a humanized monoclonal IgG antibody that binds to complement protein C5, inhibiting its cleavage to C5a and C5b and blocking the generation of the terminal complement complex C5b- C9 (Figure 1). 67 Approved by the FDA for use in the treatment of paroxysmal nocturnal hemoglobinuria and primary atypical hemolytic uremic syndrome (ahus), eculizumab has been used off- label in the kidney transplant setting for the prevention and treatment of ahus, antiphospholipid syndrome, and catastrophic antiphospholipid syndrome The first case report on the use of eculizumab to treat severe AMR demonstrated that administration of the drug in combination with plasmapheresis and low- dose IVIg led to a marked decrease in C5b- C9 complex deposition in the kidney and a reversal of the AMR episode. 72 In a single- arm phase I/II trial conducted at the Mayo Clinic that included 26 highly sensitized recipients of living- donor renal transplants (NCT006707), eculizumab in combination with plasmapheresis resulted in a statistically significant reduction in AMR in the first 3 months compared with 51 historic controls receiving plasmapheresis alone (7.7% vs 41.2%, respectively; P =.0031). 73 In this study, eculizumab was dosed at 1200 mg immediately prior to transplantation, 600 mg on postoperative Day 1, and then 900 mg weekly thereafter for 4 weeks; patients with persistently high DSA continued treatment (1200 mg at week 5 and every 2 weeks afterward) until the B flow crossmatch channel shift fell below 200. In this study, eculizumab does not appear to affect DSA levels, as the percentage of patients who developed high levels of DSA posttransplantation was similar among eculizumab- treated patients and the control group. Among patients with high DSA levels and C4d+ staining, significantly fewer patients receiving eculizumab developed AMR compared with historical controls (15% vs 100%, P <.0001). However, despite the lower incidence of AMR at 3 months, the incidence of chronic AMR at 1- to 2- years posttransplant was not significantly reduced. 74 In addition, no significant differences were noted at 1 and 2 years between patients receiving eculizumab and historical controls with regard to transplant glomerulopathy (Banff chronic glomerulopathy [cg] score > 0; 26.7% vs 31.9% for 1 year and 45.4% vs 63.6% for 2 years, respectively; P =.62), peritubular capillaritis (cg score 2; 60.0% vs 60.0%; P = 1.00), or C4d positivity (33.8% vs 13.5%; P =.08). However, eculizumab appeared to be somewhat protective against transplant glomerulopathy if DSA remained low (B flow crossmatch channel shift < 200). In a phase II randomized, open- label, multicenter trial of eculizumab vs placebo as an add- on to SOC desensitization protocols in living kidney donor recipients (NCT ), the eculizumab group failed to achieve a statistically significant reduction in the primary composite endpoint (occurrence of biopsy- proven AMR, graft loss, patient death, or loss to follow- up at the end of the 9- week treatment period posttransplant) when compared with placebo (9.8% vs 15.7%, P =.554). 75 Although the rate of AMR in the eculizumab group was similar to results from the Mayo Clinic phase I/II study, the rate of AMR in the placebo group in the multicenter trial was substantially less than historic controls used in the Mayo study, and that difference resulted in the failure to show efficacy for this indication. Additionally, there was significant intersubject variability in antibody strength, and the sponsors lowered the DSA- level thresholds during the study to try to increase enrollment. Any or all of these design problems could have impacted the results and led to a negative study. In a randomized, open- label, multicenter phase II study in sensitized deceased donor kidney transplant recipients (NCT ), 1- year data show that eculizumab (1200 mg on postoperative Day 0 prior to reperfusion; 900 mg on Days 1, 7, 14, and 28; and 1200 mg at Weeks 5, 7, and 9) was effective in reducing the incidence of acute AMR (7.5% vs 30% seen with historical controls). 76 In a Johns Hopkins study of 24 patients who developed severe, oliguric AMR after desensitization and transplantation with a live donor kidney, Orandi et al found that a combination of splenectomy with eculizumab as an add- on therapy to plasmapheresis/low- dose IVIg and rituximab resulted in the highest graft salvage rate and protection from transplant glomerulopathy compared with splenectomy alone or eculizumab alone as an add- on therapy. 77 A potential mechanistic explanation for the success of this combined therapy is that splenectomy removes antibody- producing plasmablasts and plasma cells that traffic to the spleen from regional lymphoid tissue, 78 and eculizumab protects the endothelium from injury while plasmapheresis clears residual circulating antibody. In other words, splenectomy increases rescue

9 11 rates by reducing DSA production, and eculizumab prevents transplant glomerulopathy by protecting the microvascular circulation. However, the Mayo Clinic s study suggests that eculizumab s protective effect may not be durable if strong DSA is allowed to persist for long periods of time. In the Mayo Clinic s study, DSA was not reduced or controlled after desensitization and transplantation. In a small pilot, randomized, controlled trial by Kulkarni et al, eculizumab was compared to observation for chronic active AMR. 79 The treatment group showed a modest improvement in estimated glomerular filtration rate (egfr) trajectory and no difference in ENDAT. Inclusion in this study was based on the development of de novo DSA and deteriorating renal function and not on biopsy- proven chronic AMR. The primary endpoint was change in egfr, which is not a good outcome to establish efficacy as changes in egfr occur late after ongoing injury and are not a specific finding for chronic AMR. Eculizumab failures have been reported in cases of C4d- negative AMR. 80 The presence of glomerular and capillary inflammation after eculizumab treatment suggests that terminal complement inhibition, which does not affect deposition of C4d or generation of the pro- inflammatory factor C3a, has limitations, especially when DSA is present and persists at high strength. AMR phenotypes, in relation to complement binding capacity 81 or IgG subclass, 16 change over time, and thus the design of eculizumab studies in the future needs to be more refined. The complexity of AMR requires a more personalized therapy with a more precise assessment of phenotypes that are responsive to C5 inhibition. Data from additional ongoing studies will provide important information regarding use of this agent in these patient populations. 5.2 C1 esterase inhibition C1- INH is a serine protease inhibitor that inactivates both C1r and C1s and has multiple effects. C1- INH regulates proteases in the classical and lectin complement pathways and has major effects on regulation of the coagulation cascade as well as vascular permeability and inflammation by kinins (Figure 2). 82 Following antibody/immune complex activation of C1qrs, C1- INH dissociates C1r and C1s from the activated C1 macromolecule. This prevents proteolytic activation of C4 and C2 that form C3 convertase, which is important in the context of C4d deposition in AMR. 83 Two C1- INH products that are approved for use by the FDA in the treatment of hereditary angioedema have been evaluated in small pilot studies for AMR: Berinert (CSL Behring, Kankakee, IL, USA) 84 and Cinryze (Shire ViroPharma Inc., Lexington, MA, USA). 85 A recent double- blind, placebo- controlled, randomized phase I/II trial (NCT ) evaluated C1- INH (Berinert) in sensitized renal transplant patients for the prevention of acute AMR. 83 In this trial, 20 patients were desensitized with high- dose IVIg and rituximab with FIGURE 2 Role of C1- INH in complement, coagulation, and contact (kallikrein- kinin) pathways and AMR. Primary effects indicated by solid lines; secondary effects indicated by dashed lines. Adapted from Levy et al. 82 AMR, antibody- mediated rejection; C1- INH, C1 esterase inhibitor; FDP, fibrin degradation product; HMWK, high molecular weight kininogen; IL, interleukin; KK, kallikrein; MASP, MBP- associated serine protease; MBP, mannose- binding protein; TNF, tumor necrosis factor; tpa, tissue plasminogen activator

10 12 or without plasma exchange and received C1- INH (20 IU/kg) or placebo intraoperatively, followed by 7 doses administered twice weekly. Twenty percent of C1- INH patients developed AMR outside of the study period, whereas 30% of placebo patients developed AMR, 10% during the study period. Delayed graft function developed in 1 patient receiving C1- INH and 4 patients receiving placebo. C1- INH treatment also reduced C1q- positive HLA antibodies in all 5 patients with these antibodies at time of transplantation. In a single- arm pilot study, the investigators found that CI- INH (Berinert) in combination with high- dose IVIg improved allograft function in kidney recipients with acute AMR that was nonresponsive to conventional therapy. 86 Among 26 patients with active AMR, 6 were nonresponsive to plasmapheresis, rituximab, and IVIg (denoted by negative change in the egfr at 3 months after diagnosis, DSA >3000 MFI, and persistent active AMR on the allograft biopsy (Banff glomerulitis [g] score + ptc 2) and received C1- INH (20 U/kg on Days 1, 2, and 3 and then twice weekly) and high- dose IVIg (2 g/kg every month) for 6 months. All treated patients showed an improvement in egfr between study inclusion and 6 months (from 38.7 ± 17.9 to 45.2 ± 21.3 ml/min/1.73 m 2, P =.028). The improvement in egfr in patients receiving C1- INH was significantly higher than a historical control group (N = 21) treated with high- dose IVIg alone (+16.6 ± 9.9% vs 13.0 ± 33.1%; P =.012). There was no change in the histological features in patients in the C1- INH group, except for a significant decrease from inclusion biopsy in the C4d deposition rate at 6 months (from 83% to 17%, P =.046). In addition, DSA C1q positivity was significantly reduced in the C1- INH group at 6 months (from 100% to 17%, P =.025). A randomized, double- blind, placebo- controlled, multicenter phase IIb study (NCT ) evaluated the safety and effect of C1- INH (Cinryze) for the treatment of acute AMR in 18 recipients of donorsensitized kidney transplants. 87 Patients received plasmapheresis, lowdose IVIg, and rituximab every other day, and they received an initial infusion of 5000 U C1- INH or placebo on Day 1 after plasmapheresis followed by 6 more doses of C1- INH (2500 IU) or placebo every other day after plasmapheresis over a 2- week period. The primary endpoint was the change from baseline in histopathology endpoints at Day 20 (C4d, margination, glomerulitis, vasculitis, glomerulosclerosis, chronic glomerulopathy, interstitial fibrosis, and chronic vasculitis scores). Histopathological resolution of AMR was seen in 7 out of 7 C1- INH patients analyzed and 4 out of 6 placebo patients analyzed. No statistically significant differences in histology scores between the C1- INH and placebo groups were noted at this early time point. However, surveillance 6- month biopsies performed in 14 patients showed no transplant glomerulopathy (ptc + cg 1b) in the 7 patients in the C1- INH group, whereas 3 out of 7 patients in the placebo group showed transplant glomerulopathy. In addition, transplant glomerulopathy tended to develop in those patients in the placebo group who had very low levels of C1-INH. 88 Interestingly, plasmapheresis itself was found to decrease endogenous serum C1- INH antigen concentrations by 17.6% (P <.05) and C1- INH functional activity by 43.3% in the placebo group, potentially worsening complement- mediated injury (Figure 3). 87 However, C1- INH- treated patients maintained serum C1- INH concentrations at 1.73 times the physiologic levels during plasmapheresis, which may be important in providing protection for the allograft from injury when high levels of DSAs are present. This new finding suggests that C1- INH replacement may be necessary to maintain physiologic levels during plasmapheresis. An ongoing, randomized, double- blind, placebocontrolled phase III study evaluating C1- INH (Cinryze) as an adjunct to DSA reduction therapy (plasmapheresis, plasma exchange, and/or immune adsorption treatments and IVIg) for the treatment of acute AMR is currently enrolling patients (NCT ). 5.3 IdeS Phase I/II testing (NCT ) has begun for a new compound called IgG- degrading enzyme of Streptococcus pyogenes (IdeS). 89 This enzyme cleaves at a very specific amino acid sequence in the hinge region of human IgG and essentially neutralizes all of the IgG in the body within 4 hours of administration. There is a period of about 7 days during which both soluble IgG and the B cell receptor are not detectable, after which it begins to rebound and can reconstitute fully by Day The enzyme is immunostimulatory, and the anti- IdeS antibody usually appears after 1 or 2 doses. Attempts to modify the molecule s immunogenicity might allow repeated dosing in the setting of acute or chronic AMR. Jordan et al recently published the results of a diverse population of 25 sensitized patients with DSA at 2 institutions (Cedars- Sinai Medical Center in Los Angeles, CA, USA, and Uppsala University in Uppsala, Sweden) who received a single treatment of IdeS (at varying doses) prior to transplantation. 91 The 14 patients at Cedars- Sinai received high- dose IVIg and rituximab in an attempt to desensitize them prior to enrolling in the trial and receiving an HLA- incompatible deceased donor kidney. They also received additional doses of IVIg and rituximab after the transplant. The Uppsala patients were not previously desensitized and received HLA- incompatible live or deceased donor kidneys after IdeS. There was 1 hyperacute rejection, and 10 out of the remaining 24 patients had an episode of AMR that was successfully treated with SOC therapy. Thirty- eight serious adverse effects occurred in 15 patients; 5 were possibly related to the study drug (NCT , NCT , and NCT ). Enrollment is open for a phase II study to evaluate the efficacy of IdeS to desensitize transplant patients with a positive crossmatch (HighIdeS; NCT ). Patients who qualify and are highly unlikely to receive a kidney transplant using SOC desensitization and who have a positive crossmatch (flow cytometric or complement- dependent cytotoxic) are treated with IdeS. The primary endpoint is conversion of a positive crossmatch to a negative crossmatch. Patients who achieve the primary endpoint receive either a living or deceased donor transplant. During the period prior to the DSA rebound, they receive single doses of IVIg and rituximab. A total of 20 patients at 3 centers will be enrolled. 6 HOW CAN AMR OUTCOMES BE IMPROVED? Successful management of HLA sensitization and AMR continues to represent an important unmet need in solid organ transplantation.