Humoral rejection in heart transplantation in the current era

Abstract

Immune response to the human leukocyte antigens (HLA) of a donor heart, resulting in antibody-mediated rejection (AMR), remains one of the greatest challenges in current transplant medicine. Besides the risk of acute functional deterioration of the graft during rejection, there is clear evidence of induction and acceleration of the development of coronary allograft vasculopathy with recurrent or severe AMR and development of donor–specific HLA antibodies. Management guidelines and consensus statements currently have to rely on limited evidence.

In this review, we illustrate some of the immune mechanisms underlying AMR, summarize the clinical impact, and discuss the therapy concepts of the recently emerging armamentarium of medical interventions. We also highlight some pitfalls for AMR-therapy protocol design and suggest a staged approach for decision-making at the occurrence of de novo donor-specific antibodies post transplant.

Background

Orthotopic heart transplant is a lifesaving intervention for many pediatric heart conditions, and diligent work across eras has seen continued improvement in long-term graft survival.¹ Unfortunately, antibody-mediated rejection (AMR) in pediatric heart transplantation is the most common early source of graft failure, and morbidity and mortality² remain a major concern for transplant cardiologists due to limited evidence for successful therapies. AMR is the result of antibody presence or development toward human leukocyte antigens (HLA) or blood group ABO polysaccharide antigens (in case of ABO incompatible transplants) expressed on the transplanted donor organ that initiates harmful immune cascades. AMR represents an acute risk to graft function and survival,³ and has been associated with the development and acceleration of cardiac allograft vasculopathy (CAV),4, 5, 6 and reduced long-term graft survival, especially when occurring in the first year after transplant.⁷ Although the true incidence of AMR is not known, it is estimated to be between 11% and 20% in pediatric heart transplant recipients³ within the first 3 years post transplant. Some children face the additional challenge of pretransplant sensitization, especially if they are transplanted following surgery for congenital heart disease⁸ with the highest risk arising from the use of HLA–expressing human vascular graft material for vessel reconstruction.⁹ Increasing use of long–term mechanical circulatory support also contributes to the risk of sensitization before transplant.9, 10, 11 To that end, this review aims to inform clinicians on the diagnosis of AMR, the mechanisms by which AMR develops, and provide guidance on the treatment of AMR with the latest clinical tools.

Definition of AMR

AMR remains challenging to define clearly and succinctly; however, based on the recently updated International Society for Heart and Lung Transplantation (ISHLT) consensus and for the purposes of this article, we define AMR as a clinical entity combining specific clinical, immuno-histopathologic features, with serologic characteristics, resulting in short- and long-term consequences post heart transplant. This definition is distinct from the diagnostic criteria,5, 6, 7 a strictly pathologic diagnosis endorsed by all major international transplant societies.8, 12, 13. AMR may occur throughout the life of a donor organ. Hyperacute rejection, occurring in the first minutes to 7 days following transplant, has become exceedingly rare in the modern era due to improved detection and quantification of HLA and AB-antibodies and respective organ selection or patient peritransplant treatment that allows for avoidance of clinically significant early AMR. Otherwise, AMR is categorized as early (within a year of transplant) or late (beyond a year after transplant).

Pathogenesis of AMR

Antibodies are the result of a sophisticated response to a specific antigen and forming a key part of the adaptive immune response. While they are the product of B cells having evolved into antibody–producing plasma cells, their development is closely interlinked with other parts of the adaptive (antigen-presenting cells [APC], T cells, etc.) and the innate (complement cascade, macrophages, etc.) immune system.

The predominant pathway of B cell activation is the type 2 T helper–mediated B cell immune response.¹⁴ Here, antigens are presented to CD4 T cells by macrophages, dendritic cells, B cells, or other APC via the major histocompatibility class II complex (in humans named HLA Class II). The CD4 T cells are then activating B cells via direct communication (CD40 to CD40Ligand, inducible costimulator (ICOS), and others) and soluble cytokines. In the B cell, this induces hypersomatic mutation of the antigen-binding areas of the immunoglobulin genes, selection of the highest affinity clones,¹⁵ transformation into plasma cells, antibody class switch from IgM to IgG, and formation of memory B cells for later quick reactivation.¹⁴ Alternatively, B cells can be activated via T-independent (TI) B cell response.¹⁶ TI type 1 occurs via cross-activation of multiple B cells all bound to the same long-chained antigen, such as bacterial lipopolysaccharide resulting in polyclonal antibody response of mostly IgM type. TI type 2 uses the interaction of the B cells with the innate immune system: if the antigen binding to the B cell receptor is accompanied by signaling through the B cell coreceptor triggered by linkage of CD21 to its counterpart complement split product C3d, the threshold for B cell activation without a T cell signal is reduced by 100-fold or more.¹⁷ As polysaccharide antigens, including those defining the blood groups, are not presented via the major histocompatibility II complex, TI plays a key role in transplantation in the response to A/B antigens and ABO incompatible transplantation.¹⁸

The graft is affected by antibodies in various ways. Aside from complement activation and subsequent direct damage to the cells to which the antibodies are bound, chemotaxis effects attract other immune cells to the targeted structure and opsonization facilitates invasion of macrophages and other APC perpetuating the immune attack. In the case of the graft cardiovascular wall, this results in inflammation, swelling, fibrosis, and endothelial proliferation with resulting narrowing of the coronary lumen reflecting CAV.⁵ B cells, as well as endothelial cells, also act as APC further enhancing the immunologic attack.

Diagnosis of AMR

In 2004, a task force assembled by the ISHLT provided the first standardized definition of AMR, which was updated in 2013 (Table 1).¹⁹ This definition of AMR forms the pathologic gold standard by which AMR is diagnosed in heart transplantation independent of age and includes both immunohistochemical and histologic features. It does not take clinical or other diagnostic features into account and does not mandate the presence of causative antibodies in the blood. Endomyocardial biopsy via catheterization allows for minimally invasive tissue sampling and is generally low risk.²⁰ However, in the pediatric population, patient size, more adverse effects of biopsies, clinical stability, threshold of suspicion for rejection, and the need for sedation or anesthesia are additional risk factors that suggest a more restrictive application than in adults.

Table 1.

ISHLT Pathologic Definition of Antibody-Mediated Rejection

AMR grade	Definition	Histopathologic and immunologic findings
pAMR0	No AMR	Histology or immunologic staining negative
pAMR1(H+)	Histopathologic AMR	Histology positive Immunopathology negative
pAMR1(I+)	Immunopathologic AMR	Histology negative Immunopathology positive
pAMR2	Pathologic AMR	Both histology and immunopathology positive
pAMR3	Severe AMR	Interstitial hemorrhage, capillary fragments, mixed inflammatory infiltrates, edema, and immunopathologic findings

Abbreviations: AMR, antibody-mediated rejection; ISHLT, International Society for Heart and Lung Transplantation; pAMR, pathology defined antibody mediated rejection.

Although biopsy–proven pathologic diagnosis is the gold standard for AMR,¹² noninvasive techniques are essential to screen for rejection. Subclinical symptoms predominate early, and thorough evaluation of function by echocardiogram, electrocardiogram (ECG), and Holter should be used. Echocardiogram may show impaired ejection fraction, reduced tricuspid annulus plane systolic excursion, and abnormal tissue Doppler indices, although these parameters are more sensitive than specific.²¹ Myocardial thickening, suggestive of edema, may also be evident by echo, although it is more apparent on cardiac magnetic resonance imaging (MRI) which can also be a useful screening tool. In both human and animal models, T2-weighted MRI has been shown to predict AMR with 89% sensitivity and may differentiate subtle rejection not seen on biopsy.22, 23

Inappropriate sinus tachycardia, an increased burden of atrial or ventricular ectopy, or new arrhythmias on ECG or Holter monitor are all signs of rejection that should prompt more invasive testing. Integration of the clinical history is essential for interpretation of newly detected HLA antibodies or lower degree AMR, including recent episodes of low immune suppression levels, nonadherence, or viral infections.

Recent efforts have highlighted new noninvasive methodologies to screen for AMR. Cell-free DNA (cf-DNA)²⁴ and molecular microscope technology (MMDx)²⁵ are 2 methods available in pediatric heart transplantation. Cf-DNA relies on recognizing an increase in circulating donor-specific DNA in the presence of tissue damage, while MMDx relies on machine learning derived patterns of microarray–based gene expression in the biopsy. Cf-DNA accuracy can be further increased by combining fractional donor cf-DNA and absolute quantitation.²⁶ This is an actively evolving area in pediatrics, with thresholds and interpretation validity against histopathologic biopsy still in development.25, 27, 28 To date, Cf-DNA has been shown to have good negative predictive value, which is especially promising in pediatrics given the need for anesthesia with invasive biopsy. MMDx shows moderate agreement with biopsy.²⁸ The patterns of gene expression associated with AMR in MMDx may be recognizing early rejection in a subclinical phase, but this remains to be determined and concerns about reliability have been raised, especially early post-transplant phase and in the presence of Quilty lesion.²⁹ In patients requiring close surveillance or with subclinical rejection, these technologies may offer further clarity and help guide decisions to treat. Other methods to better classify the nature of HLA antibodies, such as eplet mismatch and Predicted Indirectly Recognizable HLA Epitopes score, may be helpful,³⁰ but their role in AMR in pediatric heart transplantation is not yet clear.

Any child post-transplant who presents to a clinician must be considered for the possibility of rejection, as AMR exists along a wide clinical spectrum ranging from subtle new dysfunction or tachycardia to severe symptomatic dysfunction with hemodynamic compromise. For those without obvious clinical signs, a high index of suspicion is required to diagnose AMR in pediatric transplant patients and the appropriate clinical history regarding immunosuppressive medication adherence or new cardiac symptoms.

Targets and principles of therapeutic intervention

Figure 1 illustrates key components of donor-specific antibodies (DSA) development and graft affection as well as the targets of the various therapeutic modalities. When naïve B cells get activated by the mechanisms outlined above, they evolve into plasma cells, clonally producing DSA, losing CD20 expression during maturation, but also memory B cells for later rapid reactivation upon repeat exposure to the same antigen. In addition, long–lived plasma cells home into the bone marrow and remain there with low antibody-producing activity. Class I HLA is expressed on all nucleated cells, while Class II is limited to APCs, which includes the vascular endothelium, explaining a more global pattern of injury, including myocardial dysfunction with class I DSA and a stronger association to CAV for class II DSA. Some therapeutic options do not target the DSA themselves but downstream activating and injuring effects.

• A.Direct antibody removal or inactivation steps are fast and effective and therefore especially useful in situations of acute antibody–mediated graft dysfunction and rejection. Plasmapheresis or plasma exchange removes all protein components of the plasma, including immunoglobulins of all types and proinflammatory cytokines; however, also coagulation and other homeostatic proteins and protein-bound therapeutics. Immunoadsorption can specifically remove IgG³¹ or blood group–directed antibodies³² with less disturbance of the remaining homeostasis. Imlifidase is a streptococcal enzyme that cleaves the crystallizing (Fc) part from the antigen-binding (Fab) section of IgG and disables the DSA from activating their downstream effects,³³ but can only be used once due to sensitization against the medication itself.
• B.Intravenous immunoglobulins (IVIG) are the pooled antibodies from thousands of donors which provide the recipient with a broad range of protective and immune-modulating antibodies but may also contain smaller concentrations of HLA and A/B antibodies. For the latter, specific products with reduced blood group antibodies are available for use in patients after ABO incompatible transplant. High dose (2 g/kg) IVIG reduce antibody production by negative feedback, reduce downstream activation by blockade of the Fc-receptors of respective immune cells, modulate the cytokine milieu and autoreactive responses, reduce chemotaxis, and provide protection from many infectious diseases.³⁴ Isolated IVIG therapy at higher or lower dose has been found insufficient to treat AMR or achieve sufficient desensitization³⁵; however, it is an excellent addition to any AMR protocol reducing the risk for infectious complications of the other components aside from the above-mentioned rejection-directed effects.
• C.B cell depletion removes the “supply lines” of future plasma cells and omits the antigen-presenting effect of B cells. Depleting CD20 antibodies, such as rituximab, has been used for many years to effectively deplete B cells for typically 3 to 12 months in various autoimmune disorders and for AMR and desensitization. In contrast to others, Obinutuzumab depletes B cells via complement-independent mechanisms. As plasma cells do not express CD20, they are not depleted by these medications resulting in a delayed effect on the antibody production and plasma-concentration with measurable decline expected only after about 2 to 3 months. While CD19 antibodies deplete cells and plasma cells in experimental settings,³⁶ no effective clinically applied formulations are available.
• D.Plasma cell depletion has been applied in a variety of AMR and desensitization protocols in heart and other organ transplants initially using proteasome inhibitors³⁷ and more recently monoclonal antibodies targeting CD38³⁸ (daratumumab, felzartamab). While proteasome inhibitors mostly deplete highly active plasma cells and may miss sufficient effect on long–lived plasma cells, CD38 is expressed on all plasma cells but also a variety of B cells and other immune cells, including natural killer cells. This results in more effective antibody depletion and clinical effect on rejection³⁹ but also enhances the infectious risks with especially cytomegalovirus (CMV) reactivation and intestinal affection being reported on CD38 antibody therapy. Recent preclinical studies aim to enhance the effectiveness by combining the CD38 antibodies with a T cell–activating anti-CD3 component.⁴⁰ As plasma cell depletion directly reduces the antibody-producing cells, the effect is immediate, especially when combined with antibody-removing strategies, making this approach the prime choice for hemodynamically compromising AMR.
• E.Downstream activation inhibition includes therapies that do not directly affect the antibody concentration but reduce their ability to cause harm on the graft. Eculizumab and ravulizumab block the complement cascade activation at the level of C5, thereby preventing one of the main harm-inducing effects of antibodies on the graft.⁴¹ However, other antibody-mediated effects are not impacted and the complement blockade also hampers the effect of all infectious agent-targeting antibodies, including those transferred with IVIG, hence going along with a significant risk of infections, especially urinary tract infections and bacteremia.⁴²
• F.Augmentation of overall immunosuppression: As AMR rarely occurs in isolation treatment for acute cellular rejection is commonly applied in parallel. Besides steroids with a broad and potent effect on nearly all branches of the immune system, Antithymocyte Globulin has effects on B cells and plasma cells and the downstream activation via predominant T cell depletion. Mammalian target of rapamycin inhibitors, sirolimus and everolimus, have an effect on endothelial cell proliferation and can reduce the evolution and progression of CAV post AMR. Recent studies investigate specific effects of coreceptor blockade and Interleucin 6 interference (tocilizumab).⁴³

Figure 1.

Overview on B cell activation and targets of AMR, including expression of specific surface molecules as well as target points of the various therapeutic principles (details see text). Red arrows indicate natural evolution, blue bar-lines indicate inhibiting or modulating effects. AMR, antibody-mediated rejection; APC, antigen-presenting cells; IDeS, imlifidase; IVIG, intravenous immunoglobulins; mTORi, mammalian target of rapamycin inhibitors.

Updated and modified from Amdani et al.⁴⁴

Practical approach to AMR treatment

To date, centers rely on expert consensus and institutional experience to decide how and when treatment of AMR is attempted. Recent consensus guidelines have described the many options that exist for AMR but primarily focus on the principles of treatment rather than specific suggested approaches.13., 32, 45 Figure 2 is an example decision tree to help direct treatment of AMR based on both these consensus documents as well as local institutional experience.

Figure 2.

Treatment algorithm for AMR by presentation. Patients with suspicion for rejection based on clinical presentation may follow this suggested algorithm as a guide for potential rejection treatment, coupled with suggested treatment plans (Table 1). Patients with suspicion of rejection but no clinical signs may require escalating treatment for new DSA if adjunct monitoring, such as cell-free DNA or the MMDx phenotype, suggests graft damage. Patients presenting with hemodynamic instability require antibody removal and treatment for rejection, but suggested therapy may require adjustment based on clinical stability. AMR, antibody-mediated rejection; DSA, donor-specific antibodies; MMDx, molecular microscope technology.

Broadly, we can separate patients presenting with AMR into the 3 uppermost categories: (1) asymptomatic with new DSA, (2) rejection based on clinical or echocardiographic signs, and (3) those with hemodynamic instability due to rejection. We have paired this with Table 2, in which we offer suggested protocols for treatment in each circumstance. As discussed above with the AMR diagnosis section, the clinical context must be applied to each presentation. Factors such as recent illness, changes in immunosuppression approach or levels, or recent complications (CMV infection, post transplant lymphoproliferative disorder, etc.) should be explored even in the context of rejection causing hemodynamic instability as they offer insight into the anticipated course with treatment. While not clearly documented in the literature, our and other centers’ experience has also been that poor initial response and recurrence of AMR are associated with worse long-term outcomes and may warrant more aggressive therapy or faster escalation.

Table 2.

Suggested Treatment Plans

Treatment plan 1—low intensity		Treatment plan 2—moderate intensity		Treatment plan 3—high intensity/unstable
Day 1	Rituximab (375 mg/m2/dose)	Day 1	Rituximab (375 mg/m2/dose) Methylprednisolone 2mg/kg IV	Days 1-5	Methylprednisolone 10 mg/kg IV + Plasmapheresis followed by 0.1 mg/kg IVIG
Day 3	If B cells 0% IVIG 1 g/kg	Day 3	Plasmapheresis	Day 5	Rituximab (375 mg/m2/dose) and Daratumumab (16 mg/kg) Post plasmapheresis
Day 4	IVIG 1 g/kg	Day 5	If B cells 0% IVIG 1 g/kg	Days 1, 3, 5 (7)	HLA testing (prepheresis)
Every 28 days	IVIG 2 g/kg	Day 7	IVIG 1 g/kg	Days 7-12	If DSA are still present: continue plasmapheresis for 48 hours
		Every 28 days	IVIG 2 g/kg	Every 28 days	IVIG 2 g/kg

Abbreviations: DSA, donor-specific antibodies; HLA, human leukocyte antigens; IVIG, intravenous immunoglobulins.

Adjunct therapy for treatment-resistant or recurrent AMR: Bortezomib—1.3 mg/m² in 4 doses divided over 10 days; Daratumumab—16 mg/kg post plasmapheresis weekly for 3 weeks.

Due to the protein load, the first dose of 2 g/kg IVIG is split over 2 doses, if tolerated, application in a single dose is pursued thereafter.

A severe presentation with hemodynamic instability is often conceptually more straightforward to approach, and there is little argument for the need to treat. intensive care unit management and stabilization of the patient on appropriate inotropes is paramount, including extracorporeal life support if necessary. Following initial stabilization, prompt initiation of treatment (Table 2: protocol 3) is required. Pulse steroids are included as a broad-spectrum treatment where acute cellular rejection may be present as well, and which cannot be ruled out in very acute presentations pre biopsy. In these severe presentations, rapid antibody removal with daily plasmapheresis or immunoadsorption is our preferred step, as donor–specific circulating antibodies are already present in likely high amounts. Once clinical stability permits, endomyocardial biopsy to confirm the diagnosis before initiation of further treatments is preferred before moving on to the next treatment steps. Treatment with rituximab on day 5 should be done post biopsy if possible. Rituximab has been shown in case series to be effective in cardiac and other organ recipients, and its safety and efficacy are derived from the oncologic literature.46, 47 Our practice has been to avoid treatment with IVIG or plasmapheresis for 48 hours post Rituximab to maximize treatment effect.

If there are clear signs of rejection by echo, ECG, or clinical history without hemodynamic instability, then our preferred approach is to confirm the diagnosis with endomyocardial biopsy before rapid initiation of rituximab. Treatment plan 2 (Figure 2) initiates treatment by B cell depletion and steroid treatment, followed by plasmapheresis to remove circulating antibodies. Confirmation of B cell depletion is followed by IVIG for the antibody neutralization, as well as the infectious protection for 6 months.

In cases presenting with subclinical features (Figure 2: asymptomatic with new DSA), the aforementioned clinical context, as well as new diagnostic approaches, are key in determining the individual threshold to treat. Asymptomatic patients may have a clinical reason (i.e. recent nonadherence to immunosuppression, history of previous AMR) that would prompt treatment. If so, proceeding to biopsy and treatment (Figure 2: treatment plan 1) can be warranted as these DSA may lead to poorer outcomes.⁴⁸ A recent large registry analysis showed negative impact on graft survival with no significant dependency on the degree of AMR or presence of DSA, highlighting the importance of careful assessment and monitoring of also patients perceived as “mild cases.”⁷ In those patients without clinical reasons to treat, basic and advanced noninvasive parameters may help decision-making, including elevation of C-reactive protein and N-terminal pro-B-type natriuretic peptide and more specific Cf-DNA or MMDx. There are ongoing efforts to validate these markers together into clinical prediction tools for use.⁴⁹ If absolute and fractional cf-DNA are negative,²⁶ clinical monitoring alone may be sufficient. Circulating DSA in the absence of evidence of cellular damage may reflect early-stage AMR, so repeat cf-DNA at appropriate clinical intervals may prompt subsequent treatment. If noninvasive techniques are suggestive of graft injury, biopsy and treatment are indicated in our opinion to remove DSA. For cases in which initial treatment does not show a decrease in antibodies, we suggest subsequent options for escalation of therapy. Bortezomib and Daratumumab are appropriate next steps for persistent or hemodynamically compromising AMR as these therapies directly target plasma cells and result in immediate effects. Current evidence for Daratumumab is limited to case reports and series in the pediatric population.50, 51 Despite this, Daratumumab represents a promising new agent for refractory DSA. Optimal treatment routine is yet to be determined; however, weekly treatment for 3 weeks has proven reasonable, although longer courses have been attempted.

Conclusion

As this review attempts to highlight, there are many current areas in which our understanding of AMR is evolving. Earlier diagnosis and better treatment options have improved the clinical outlook for patients who develop AMR, but our understanding remains incomplete. The suggested approach herein is a guideline for potential treatment, supported by core immunologic concepts and the existing literature. Future research into the treatment of AMR is needed to improve the ability to compare institutional approaches to this challenging clinical entity.

Funding

No funding was received for the generation or development of this manuscript.

Author contributions

This manuscript was entirely written and developed without help of artificial intelligence. Dr Sheridan and Dr Urschel have both conceptualized, written, and edited the manuscript and developed the figures and tables with the exception of the definition of AMR taken from the ISHLT consensus statement.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.