Update on diagnostic and therapeutic strategies for antibody-mediated rejection in kidney transplantation

Abstract

Antibody-mediated rejection (AMR) remains a leading cause of kidney allograft failure, posing significant clinical and economic challenges. Donor-specific antibodies against human leukocyte antigens or non-human leukocyte antigens are critical risk factors for AMR and graft loss. The diagnostic criteria and classification of AMR have evolved considerably over the past three decades, driven largely by the Banff classification. The latest Banff 2022 classification introduced two additional subcategories of “microvascular inflammation, donor-specific antibody-negative, C4d-negative” and “probable AMR”. Traditionally, graft monitoring has relied on non-specific markers such as serum creatinine and proteinuria, and the invasive biopsies. Noninvasive tools using blood and urine biomarkers, including cellular assays and molecular profiling, are increasingly being investigated. Technologies such as the Molecular Microscope Diagnostic System show promise, with studies reporting 80% sensitivity and 90% specificity in detecting AMR. Treatment of AMR remains inconsistent. Recent advances, including CD38 antibodies, have demonstrated up to 60% efficacy in reversing AMR, while complement inhibition shows potential in severe early cases. Ongoing clinical trials evaluating high-dose intravenous immunoglobulin, efgartigimod, fostamatinib, and other novel therapies aim to expand treatment options. These developments highlight the need for well-designed clinical trials to validate biomarkers and therapies and to improve long-term outcomes for kidney transplant recipients.

Key Words: Antibodies; Kidney; Transplantation; Rejection; Diagnosis

Core Tip: Antibody-mediated rejection remains a leading cause of kidney graft dysfunction and failure. While donor-specific antibodies are a key risk factor, not all are pathogenic. Current diagnostic markers lack sensitivity and often miss subclinical changes, with definitive diagnosis still relying on invasive biopsy. The clinical adoption of noninvasive biomarkers continues to face significant hurdles. Despite advances in understanding antibody-mediated rejection biology, no standardized or approved treatment currently exists. Present guidelines are largely based on earlier consensus, although emerging therapies such as CD38 antibodies and complement inhibitors show promise, particularly in early or severe cases.

INTRODUCTION

Antibody-mediated rejection (AMR), also referred to as humoral rejection, is a major contributor to allograft dysfunction and remains one of the foremost challenges in achieving sustained graft survival in kidney transplantation. It accounts for approximately 76% of death-censored graft failures occurring beyond the first year post-transplantation[1].

AMR exists along a pathological continuum and often overlaps with T-cell-mediated rejection (TCMR)[2]. At one end of the spectrum, acute AMR is characterized by microvascular inflammation (MVI), endothelial injury, and the presence of donor-specific antibodies (DSAs) in circulation, and is often amenable to therapeutic intervention[3]. At the other extreme, chronic AMR culminates in transplant glomerulopathy (TG), an advanced state of glomerular injury and architectural remodeling, for which current treatments are largely ineffective in halting progression[3]. A systematic review of 28 studies reported AMR incidences ranging from 3% to 12%, with chronic AMR affecting 7.5%-20.1% of patients over 10 years[4].

DSAs, pivotal to the development of AMR, are recognized as independent risk factors for graft failure. However, not all DSAs exhibit pathogenic behavior. Advances in genomic research and the development of sophisticated assays for DSA detection and characterization have significantly enhanced our understanding of AMR’s underlying mechanisms[5]. Emerging biomarkers and predictive models hold promise for evaluating graft failure risk and tailoring therapeutic strategies.

The therapeutic landscape for AMR is diverse and continues to evolve, partly due to the absence of standardized diagnostic frameworks and variability in clinical practice[6]. Management of acute AMR typically involves a combination of plasmapheresis, intravenous immunoglobulin (IVIG), corticosteroids, and adjunctive therapies tailored to the patient’s needs. In contrast, the clinical management of chronic AMR remains particularly challenging[6]. Current therapeutic strategies primarily aim to slow the progression of graft injury rather than reverse established damage, in part because of the limited availability of agents that effectively target antibody-producing plasma cells.

Nevertheless, emerging therapies such as interleukin-6 receptor antagonists like tocilizumab[7,8], clazakizumab[9], and CD38 monoclonal antibody like felzartamab[10] offer promising avenues for intervention and may help reshape the treatment landscape for chronic AMR.

Despite significant progress over the past two decades, many diagnostic and therapeutic challenges remain unresolved. This comprehensive review highlights recent studies that seek to advance our understanding of AMR, improvements in diagnostic methodologies, and current therapeutic strategies based on the latest evidence.

LITERATURE SEARCH STRATEGY

A comprehensive literature search was conducted using electronic databases, including PubMed, MEDLINE, Scopus, and Web of Science. The search covered publications from January 1965 to February 2025 to incorporate both foundational and recent studies. A wide range of keywords and Medical Subject Headings terms were used to identify relevant literature, including “antibody-mediated rejection”, “donor-specific antibodies”, “de novo DSA”, “preexisting DSA”, “T cell-mediated rejection”, “The Transplantation Society guidelines”, “Kidney Disease: Improving Global Outcomes clinical practice guidelines”, “pathophysiology”, “advancements in diagnostic methodologies”, “updated management”, “noninvasive biomarkers”, “invasive biomarkers”, “new treatment strategies”, and “kidney transplantation”. Boolean operators (AND, OR) were applied strategically to refine and optimize the search results.

Inclusion criteria were: (1) Peer-reviewed original research articles, systematic reviews, or meta-analyses; (2) Studies involving human subjects with kidney transplants; (3) Publications addressing the pathophysiology, diagnosis, biomarkers, or treatment of AMR and/or TCMR; and (4) Articles published in English. Exclusion criteria were: (1) Conference abstracts without full text, case reports with insufficient detail, editorials, and commentaries; and (2) Studies focusing solely on non-renal organ transplants without relevant kidney transplant data.

Two reviewers independently screened titles and abstracts to identify potentially relevant studies, followed by a full-text review (Figure 1). Discrepancies were resolved through discussion or consultation with a third reviewer. The reference lists of all included studies were also manually searched to identify additional eligible articles. Data extraction was performed to synthesize evidence for this descriptive review, including information on study design, population characteristics, diagnostic criteria, interventions, outcomes, and key findings. The methodological quality of included studies was not systematically assessed due to the narrative nature of this article.

Figure 1  Flow diagram showing study methodology for selecting the articles.

PATHOPHYSIOLOGY OF AMR

AMR in kidney transplantation is a complex immune process initiated by the recognition of donor antigens, primarily human leukocyte antigens (HLA), by the recipient’s immune system. Sensitized B lymphocytes and plasma cells produce DSAs, most often of the IgG class, which circulate and bind to HLA molecules expressed on the vascular endothelium of the transplanted kidney[11-13]. Once bound, these antibodies activate the classical complement cascade, beginning with C1q binding and progressing to the formation of the membrane attack complex (C5b-9), which disrupts endothelial integrity (Figure 2). A hallmark of the process is deposition of C4d along peritubular capillaries, which is widely used as a diagnostic marker of AMR[14]. In addition to direct complement-mediated damage, the complement split products C3a and C5a act as potent chemoattractants, recruiting inflammatory cells such as neutrophils, macrophages, and natural killer (NK) cells. These cells exacerbate vascular injury through the release of cytokines, reactive oxygen species, and proteolytic enzymes, and NK cells can also mediate antibody-dependent cellular cytotoxicity. The net effect is MVI, endothelial swelling, capillary congestion, and fibrin deposition within glomeruli and peritubular capillaries[1,2].

Figure 2 Mechanism of antibody-mediated rejection. (Source: Biorender.com). NK cells: Natural killer cells; IFN: Interferon; TNF: Tumour necrosis factor; HLA: Human leukocyte antigen.

When this process is persistent or inadequately controlled, it transitions into a chronic phase characterized by ongoing endothelial injury and maladaptive remodeling. Chronic AMR is typified by TG, seen histologically as duplication of glomerular basement membranes, multilamination of peritubular capillary basement membranes, and progressive loss of capillary networks. This remodeling leads to impaired microcirculation, ischemia, and progressive nephron loss. Interstitial fibrosis (IF) and tubular atrophy also develop as downstream consequences of chronic inflammation and hypoxia. Together, these structural changes result in declining renal function, proteinuria, hypertension, and, ultimately, graft failure. Thus, AMR represents a continuum from acute antibody and complement-driven vascular injury to chronic scarring and fibrosis, and remains one of the most significant barriers to long-term kidney allograft survival[3-5].

PATHOLOGY OF AMR

The understanding of antibody-mediated mechanisms in transplant pathology has advanced considerably over the past three decades. In its earliest form, the Banff classification acknowledged the existence of alloantibodies but did not fully integrate their role in graft injury due to limited supporting evidence[15]. At that time, antibodies were believed to be relevant only in the immediate post-transplant period. However, accumulating evidence demonstrated that antibodies contribute to both acute and chronic forms of rejection, leading to major revisions in the Banff criteria[16-18]. This paradigm shift was driven by several key developments, including the identification of C4d as a biomarker of AMR, improved techniques for detecting anti-donor antibodies, an increasing number of re-transplants, and the higher frequency of transplants across immunologic barriers in the context of organ shortages[19-22].

From a pathological perspective, AMR is characterized by injury to vascular endothelial cells, with the nature and severity of damage determined by the antibodies involved. As knowledge expanded, the Banff classification progressively refined both its terminology and diagnostic criteria. Initially, the process was described as hyperacute rejection. In 1999, the terminology was changed to AMR, with subdivisions based on clinical presentation[15]. By 2003, the role of antibodies beyond the immediate post-transplant period was recognized, prompting the incorporation of C4d staining and serological testing into diagnostic protocols. For the first time, pathological criteria for diagnosing and classifying AMR were formally established at the Banff meeting of that year[16].

Further refinements followed. In 2007, diagnostic criteria for chronic AMR were introduced, including the definition of specific pathological lesions[18]. In 2008, the category was renamed antibody-mediated changes, a term intended to reflect the broader spectrum of antibody effects, and a new subcategory was created for isolated C4d deposition without evidence of active rejection[19]. In 2016, focal C4d staining was accepted as a diagnostic criterion for diagnosis, and the presence of any degree of vascular inflammation (v > 0) was recognized as sufficient for AMR[20]. At the same time, molecular diagnostics began to be integrated into the classification[21-25]. By 2019, the framework had expanded further with the introduction of chronic inactive AMR as a distinct subtype[26].

Over time, the Banff consensus has shifted away from rigid diagnostic requirements. The absolute need for C4d and DSAs to confirm AMR has been removed. The contemporary approach relies on an integrated assessment that combines histological features, immunohistochemical findings such as C4d, serological testing for DSAs, and molecular diagnostic tools[26-29].

Artificial intelligence and machine learning in kidney allograft pathology reporting

Artificial intelligence (AI) and machine learning (ML) are increasingly being applied in kidney allograft pathology reporting to enhance diagnostic accuracy, predict outcomes, and streamline the evaluation process. These technologies enable the automated analysis of histopathological images, offering the potential for more efficient and consistent interpretation. The Banff Digital Pathology (DP) Working Group was established during the joint American Society for Histocompatibility and Immunogenetics/Banff Meeting, September 23-27, 2019, held in Pittsburgh, Pennsylvania, to address unmet needs and challenges in the widespread adoption of digitization in kidney allograft pathology interpretation[3]. The main roles these emerging technologies perform are outlined below.

Image analysis: AI algorithms, particularly deep learning models, are used to analyze digital slides and histopathological images. These models can identify patterns and abnormalities in tissue samples, such as cellular rejection or fibrosis, which may be overlooked by human pathologists. For example, convolutional neural networks have been applied to classify various types of rejection and graft injury with high accuracy, thereby reducing diagnostic variability.

Predictive models: ML algorithms trained on large datasets of clinical and pathological features can predict the likelihood of graft rejection or failure. These models incorporate variables such as DSAs, serum creatinine levels, and biopsy findings to provide personalized risk assessments and guide treatment decisions.

Automated reporting: AI-driven systems can generate preliminary pathology reports based on image analysis and clinical data, reducing the time required for manual documentation and allowing pathologists to focus on complex cases. These tools can suggest potential diagnoses, thereby supporting decision-making.

Biomarker discovery: ML algorithms are also used to analyze molecular profiles (e.g., gene expression or protein levels) to identify novel biomarkers for AMR and other kidney transplant pathologies. This facilitates the development of noninvasive diagnostic tests for early detection of rejection.

The Banff 2022 meeting highlighted the significant progress in applying ML to DP, particularly in kidney transplant pathology[3]. The Banff Banff DP Working Group has since produced two publications on the subject. Increasingly, kidney transplant pathologists are incorporating DP into their daily diagnostic practices, research, and clinical trials, benefiting from the ability to share digitized whole slide images (WSI) without compromising quality. Additionally, many pathologists are also curating annotated and unannotated WSI datasets to train models for diagnostic tasks.

Deep learning-based decision support systems are being developed for kidney transplant pathology. These include applications for segmenting microanatomical structures, quantifying morphometric data, detecting objects like inflammatory cells, and classifying slides as rejection vs non-rejection, all of which can assist nephropathologists. DP tools are also improving the evaluation of complex lesions such as TG, offering prognostic insights. Recent studies demonstrate that digital features correlate more closely with Banff Lesion Scores, with improved sensitivity to subtle pathological changes and better predictive accuracy for graft loss, suggesting potential applications beyond the Banff grading system.

A key challenge in deep learning model development is the limited availability of large annotated WSI datasets that include both pathology and clinical outcomes. Methods such as the Human AI Loop are helping to reduce the need for extensive human annotation. To ensure fair and transparent algorithm development, practices such as algorithmic auditing, sensitivity analysis, and adherence to AI-specific reporting guidelines (e.g., the Developmental and Exploratory Clinical Investigations of DEcision support systems driven by AI) are being emphasized.

In summary, AI, ML, and related algorithms are transforming kidney allograft pathology by improving diagnostic precision, facilitating early detection of rejection, and supporting personalized treatment approaches, ultimately enhancing patient outcomes. The pathology of AMR is expected to evolve further with greater emphasis on molecular profiling, recognition of non-HLA antibodies, and the incorporation of novel biomarkers. These advances reflect the growing understanding of the immunological mechanisms underlying rejection and the development of more sophisticated diagnostic strategies, ultimately aimed at improving the accuracy of AMR diagnosis and long-term kidney transplant outcomes.

In summary, the pathology of AMR is expected to evolve further with increasing emphasis on molecular profiling, the recognition of non-HLA antibodies, and the incorporation of new biomarkers. These changes reflect the growing understanding of the immunological mechanisms underlying rejection and the development of more sophisticated diagnostic strategies, ultimately aimed at improving the accuracy of AMR diagnosis and enhancing outcomes in kidney transplantation.

DEFINITION AND CLASSIFICATION OF AMR

Diagnosing and classifying AMR in kidney transplant patients remains a complex challenge due to overlapping clinical, immunological, and pathological factors. A major difficulty lies in the heterogeneity of presentation: Some patients may have subclinical AMR detectable only through protocol biopsies, while others present with rapid allograft dysfunction. Diagnosis typically requires integrating histological evidence of microvascular injury, detection of DSAs, and C4d staining. However, each of these markers has limitations. For example, some patients with AMR may have negative C4d staining, while others may have circulating DSAs without overt histological injury, complicating interpretation.

Another challenge is the evolving classification criteria. The Banff consensus has undergone repeated revisions to incorporate new knowledge and biomarkers. Despite these updates, diagnostic “gray zones” persist, particularly in distinguishing active from chronic AMR or in cases where molecular diagnostics suggest AMR but traditional criteria remain inconclusive. Limited access to advanced molecular tools in many centers further contributes to variability in diagnosis. These challenges not only hinder timely therapeutic decisions but also complicate prognostication and research, as consistent case definitions are essential for comparing outcomes across studies and centers.

Since its initial inclusion in the international Banff classification, the diagnosis of AMR has been based on three interdisciplinary criteria: (1) Histological signs of AMR activity or chronicity; (2) Evidence of antibody interaction with donor endothelial cells; and (3) Detection of DSAs in the recipient’s serum. Over the past two decades, these diagnostic criteria have undergone several revisions (Table 1), each contributing to an increase in the number of confirmed AMR cases.

Table 1 Chronology of antibody mediated rejection advancements and defining developments.

Year	Key developments
1964[11]	Recognition of immediate kidney rejection caused by pre-existing antibodies
1970[12]	Emergence of obliterative vascular lesions in recipients with de novo DSA, leading to poor transplant outcomes
1990[13]	Characterization of AMR through a three-factor framework: (1) Impaired graft function; (2) Presence of neutrophils in peritubular capillaries; and (3) Detection of class I HLA antibodies, independent of conventional T-cell-mediated rejection
1993[14]	Confirmation of the role of DSA in rejection with C4d deposition in PTC
1997[15]	Banff consensus establishes histological guidelines for kidney transplant biopsies: AMR is determined as rejection partially or fully attributed to anti-donor antibodies, categorized as hyperacute (immediate) or accelerated acute (delayed reaction)
2001[16-17]	Classification of AMR. Chronic rejection is defined as progressive deterioration of renal function, marked by hypertension and proteinuria beyond three months post-transplant: C4d staining in PTC distinguished AMR from non-specific CAN. Cases lacking detectable antibodies are classified as “suspicious for AMR.”
2005[18]	Banff classification revised to replace CAN with chronic active AMR. Discovery of AT1R-activating antibodies. Isolated tubulitis under the borderline category
2007[19]	Banff developed grading for PTC inflammation, Focal C4d, C4d scoring, and interpretation of C4d deposition absent overt histological rejection. ti score. Grading of zero-time and protocol biopsies
2009	Subclinical AMR. Creation of Banff Working Groups
2013[20]	Major revision of AMR definition: C4d staining is no longer an absolute necessity; microvascular inflammation or validated GEP is used as an additional diagnostic marker. Cg1a scoring. HLA DSA testing by SA
2015	Specific reports for each organ. i-IFTA. Role of non-HLA-DSA
2017[21]	Banff incorporates C4d staining and molecular classifiers as surrogate indicators of DSA presence. AMR without anti-HLA DSA
2019[22]	Introduction of machine learning-based FFPE-based molecular diagnostics to refine AMR classification. FFPE-based molecular diagnostics
2021	Banff incorporated multi-omics approaches (genomics, transcriptomics, proteomics) to improve AMR detection
2023	Artificial intelligence-assisted digital histopathology was introduced for automated grading of AMR severity. Routine molecular diagnostics as a “companion criterion standard.” Open source analytic pipeline. Machine learning based biopsy contextualization
2025[23-25]	Banff further refines AMR classification with four advancements: (1) C4d and molecular classifiers remain key markers; (2) Single-cell RNA sequencing aids in identifying early AMR signatures; (3) Non-invasive biomarkers (circulating donor-derived cell-free DNA) validated for AMR monitoring; and (4) Personalized immunosuppressive strategies based on molecular profiling introduced

DSA: Donor-specific antibody; AMR: Antibody mediated rejection; HLA: Human leucocyte antigen; PTC: Peritubular capillary; CAN: Chronic allograft nephropathy; AT1R: Angiotensin receptor; GEP: Gene expression profiling; IFTA: Interstitial fibrosis and tubular atrophy; Cg: Chronic glomerulopathy; FFPE: Formalin-fixed paraffin- embedded; SA: Single antigen.

According to the Banff 2019 classification, AMR is categorized into three distinct diagnostic entities (Table 2)[26]: (1) Active AMR defined by the presence of all three of the following: Histologic features of MVI, such as glomerulitis and peritubular capillaritis, evidence of ongoing or recent antibody interaction with the endothelium, typically indicated by C4d-positive staining, and Serologic confirmation of DSAs; however, C4d positivity or validated endothelial gene transcript analysis may substitute DSAs, when they are undetectable; (2) Chronic active AMR shares the core diagnostic criteria with active AMR but is distinguished by histologic evidence of chronic tissue injury, including findings like TG; and (3) Chronic (inactive) AMR characterized by chronic tissue injury, such as TG and basement membrane duplication, in the absence of both active MVI and C4d deposition in peritubular capillaries.

Table 2 Antibody-mediated rejection spectrum according to Banff 2019 classification report.

Criterion	Active AMR	Chronic active AMR	Chronic (inactive) AMR
Criterion 1: Histopathology	Acute tissue injury, including one or more of the following: (1) MIV (glomerulitis and/or ptc) in the absence of recurrent or de-novo GN, intimal or transmural arteritis; (2) Acute TMA, in the absence of any other apparent cause; and (3) ATI, in the absence of any other apparent cause	Chronic tissue injury, including one or more of the following: (1) TG (GBM duplication in the absence of subendothelial immune complex deposits) if no evidence of chronic TMA in the absence of recurrent or de-novo glomerulonephritis; (2) Severe peritubular capillary basement membrane multilayering (requires EM); and (3) Transplant arteriopathy (arterial intimal fibrosis of new onset) and mild to moderate acute tissue injury (MIV)	Chronic tissue injury: (1) TG, and/or severe ptc basement membrane multilayering (requires EM) without MVI; and (2) Significant loss of ptc (capillaries simply no longer exist to show capillaritis)
Criterion 2: Evidence of antibody interaction with the endothelium	These include at least one of the following: (1) C4d deposition in ptc, OR; (2) At least moderate MIV, OR; and (3) Increased expression of gene transcripts/classifiers in the biopsy tissue	C4d deposition in ptc, or at least moderate MIV, or increased expression of gene transcripts/classifiers in the biopsy tissue	C4d negative
			There may be prior evidence of antibody interaction with the endothelium
Criterion 3: Serologic evidence of DSAs (HLA or other antigens)	Detectable serum anti-HLA DSA	Detectable serum anti-HLA DSA	Anti-HLA DSA may be undetectable
	If anti-HLA DSA is undetectable, non-HLA antibody testing	If anti-HLA DSA is undetectable, non-HLA antibody testing	However, there should be prior evidence of anti-HLA or non-HLA DSA
	C4d staining or expression of validated transcripts/classifiers may substitute for DSA	C4d staining or expression of validated transcripts/classifiers may substitute for DSA

AMR: Antibody-mediated rejection; ptc: Peritubular capillary; GN: Glomerulonephritis; TMA: Thrombotic microangiopathy; ATI: Acute tubular injury; TG: Transplant glomeruopathy; GBM: Glomerular basement membrane; EM: Electron microscope; HLA: Human leukocyte antigen; DSA: Donor-specific antibody; MVI: Microvascular inflammation.

The Banff classification continues to evolve to reflect advances in the understanding of rejection pathobiology and its phenotypic spectrum. Notably, the Banff 2019 and 2022 Kidney Meeting Reports introduced important updates related to AMR and MVI[3,26,27]. The Banff 2022 update added two new subcategories within the broader AMR/MVI classification: (1) “MVI, DSA-negative, C4d-negative”; and (2) “Probable AMR”, characterized by subtle MVI in patients with circulating DSAs.

Among these, the MVI, DSA-negative, C4d-negative phenotype involves MVI above the histologic threshold in the absence of both DSA and C4d staining (Figure 3). While this pattern may appear histologically similar to AMR, its immunopathologic underpinnings remain uncertain, rendering it a descriptive rather than diagnostic entity. Importantly, these cases have been observed across the spectrum of renal function, from stable grafts to those with dysfunction. Before designating a case as DSA-negative, a comprehensive evaluation is required, including testing across all HLA loci and close collaboration with the HLA laboratory. The STAR guidelines highlight both the technical limitations of DSA testing and the potential for missed or low-level antibodies undetectable by conventional assays[28]. This caution helps avoid premature classification of patients into ambiguous categories without excluding occult alloimmunity.

Figure 3 An update and comparison of the Banff 2019 and Banff 2022 on antibody-mediated rejection and microvascular inflammation. AMR: Antibody mediated rejection; MVI: Microvascular inflammation; g: Glomerulitis; ptc: Peritubular capillaritis; v: Arteritis; TMA: Thrombotic microangiopathy; cg: Chronic glomerulopathy; ptcml: Peritubular capillary basement membrane multilayering; DSA: Donors-specific antibodies.

The differential mechanisms for DSA-negative and C4d-negative MVI are broad and extend beyond conventional AMR. Potential explanations include missed or low-affinity HLA-DSA, alloreactive T-cell-mediated endothelial injury, autoreactive or non-HLA antibodies, primary NK cell activation through mechanisms such as “missing self”, and innate immune activation. Additionally, non-immunologic factors such as ischemia-reperfusion injury and viral infections may also mimic or contribute to these findings. This complexity underscores that not all microvascular injury is synonymous with alloantibody-driven rejection, and that a multifactorial lens is necessary when interpreting biopsy findings. Ultimately, these observations highlight the evolving understanding of AMR and MVI and reinforce the importance of ongoing research to determine prevalence, biological mechanisms, clinical implications, and therapeutic strategies for these challenging phenotypes[29,30].

Another distinct phenotype of AMR within this classification is “probable AMR”, defined by the presence of circulating DSA in combination with subtle but subthreshold MVI lesions (glomerulitis + peritubular capillaritis < 2) and the absence of C4d deposition in the peritubular capillaries. This phenotype raises important clinical questions, as it suggests antibody activity despite limited histological correlates. It can be observed in patients with either normal or impaired graft function, highlighting that serological evidence of alloimmunity may exist even when tissue injury is not overt. In such cases, antibody-targeted therapies may be considered, particularly if the clinical context indicates risk for progression; however, the lack of robust outcome data to date makes decision-making challenging. Current literature emphasizes the need for further research to establish the prevalence, clinical significance, and most effective treatment strategies for this phenotype, which may ultimately refine patient stratification and avoid both under- and over-treatment[31].

Clinically, patients with active AMR are at a heightened risk for future rejection episodes, progression to chronic AMR, and eventual graft failure. Likewise, those diagnosed with chronic AMR face increased risks of graft loss and mortality.

DIAGNOSIS

For many years, monitoring kidney allograft function primarily depended on nonspecific indicators such as serum creatinine, glomerular filtration rate, and proteinuria[32]. However, these conventional markers have notable limitations, especially in detecting early or subclinical graft injury. This underscores the urgent need for innovative, reliable, and preferably noninvasive diagnostic tools to accurately assess both acute and chronic allograft damage. While several novel biomarkers have demonstrated promising diagnostic and prognostic value, their integration into routine clinical practice remains limited due to a lack of compelling evidence that they significantly improve graft outcomes[33].

Noninvasive biomarkers

A range of noninvasive monitoring strategies has emerged, leveraging accessible biological fluids such as blood and urine to enable frequent and longitudinal assessment of the recipient’s immune activity. The advancement of “omics” technologies, including genomics, transcriptomics, proteomics, and metabolomics, has driven the identification of several promising biomarker candidates in transplantation[25,34].

Functional cell-based immune monitoring: One such approach is the interferon-γ (IFN-γ) enzyme-linked immunospot assay, which quantifies IFN-γ release by recipient T cells upon exposure to donor antigens. This assay has been employed to assess anti-donor T cell alloreactivity[35]. Multiple studies indicate that heightened pre-transplant T cell reactivity, as detected by enzyme-linked immunospot, correlates with an increased risk of post-transplant acute rejection and reduced graft function[36]. Despite its potential utility in monitoring T cell activity and memory B cells, its widespread use is limited due to technical complexity, time intensity, and limited scalability in routine clinical settings.

Molecular blood biomarkers: These biomarkers are typically identified through microarray platforms that detect specific messenger ribonucleic acid (mRNA) or DNA signatures in peripheral blood to predict or diagnose acute rejection.

Donor-derived cell-free DNA: This biomarker consists of fragmented DNA continuously released into the bloodstream from an injured graft, with a short half-life of about 30 minutes. Several donor-derived cell-free DNA (dd-cfDNA) assays have gained regulatory approval (e.g., Medicare) for transplant recipients[37]. Elevated or rising plasma levels, particularly above 1% have been linked to acute rejection episodes[38]. In one study, a threshold of > 1% dd-cfDNA demonstrated a positive predictive value (PPV) of 61% and a negative predictive value (NPV) of 84% for biopsy-confirmed AMR[39].

Kidney solid organ response test: This polymerase chain reaction based assay quantifies expression levels of 17 genes implicated in acute rejection or leukocyte trafficking, with reported sensitivity of 92% and specificity of 93%. However, it does not effectively differentiate between TCMR and AMR, and it is not yet commercially available[40].

TruGraf v1 assay: This microarray-based blood test was developed to reduce the reliance on surveillance biopsies in recipients with stable graft function by ruling out subclinical rejection[24]. A study demonstrated that combining TruGraf with dd-cfDNA enhanced the detection accuracy of subclinical rejection[41].

Peripheral blood gene expression profiling: This technique stratifies patients into “Transplant Excellence” (TX), indicating immune quiescence, or “not-TX,” suggesting a higher risk of immunologic activity and graft rejection. In a study involving 208 recipients and 428 surveillance biopsies, the non-TX profile had a PPV of 47% and NPV of 82% for detecting subclinical rejection[41].

Additional peripheral blood assays: A 17-gene expression signature identified patients with subclinical or borderline TCMR three months post-transplant[42]. An eight-gene panel, developed through the BIOMArkers of Renal Graft Injuries consortium, demonstrated utility in AMR diagnosis[43]. MicroRNA (miRNAs) profiling (e.g., miR-15B, miR-16, miR-103A, miR-106A, and miR-107) has shown improved diagnostic performance for rejection when used as a combined signature[44].

Urine biomarkers: Urine-based biomarkers provide a noninvasive means to detect and monitor acute rejection, encompassing a variety of molecular and protein targets, including mRNAs, miRNAs, proteins, and peptides: (1) Urinary mRNAs: Transcriptomic analyses of urinary cell RNA have revealed gene expression profiles and pathways enriched in both TCMR and AMR, offering diagnostic insights beyond traditional histology[45]; (2) Urinary miRNAs: Combined profiling of biopsy and urinary miRNAs has demonstrated potential in tracking graft function and forecasting progression toward chronic allograft dysfunction[46]; (3) Urinary proteins: Immune-related proteins such as chemokine C-X-C Motif Chemokine Ligand (CXCL) 10 have been identified as elevated in individuals experiencing acute rejection, particularly those with AMR[47]; and (4) Urinary proteomics/peptidomics: Proteomic and peptidomic analyses have yielded high-sensitivity and specificity signatures for noninvasive detection of acute rejection in several studies[48].

Limitations of non-invasive biomarkers: Non-invasive biomarkers for the diagnosis of AMR show promise but remain limited by important shortcomings. The main shortcomings of these biomarkers, along with approximate cost estimates, are shown in Table 3.

Table 3 Limitations of biomarkers and their validation status.

Biomarker/test	PPV	NPV	Estimated cost	Validation status
dd-cfDNA (> 1%)	61%	84%	About 2800 dollars per test	FDA-approved; medicare reimbursed
kSORT (17-gene PCR panel)	92% (sensitivity)	93% (specificity)	Not commercially available	Research-stage; limited clinical use
TruGraf v1	Combined PPV/NPV with dd-cfDNA	Improved detection	About 2000-2500 dollars per test	Validated in stable graft recipients
Peripheral blood GEP	47%	82%	About 1500-2000 dollars	Clinically available; moderate uptake
ELISPOT (IFN-γ)	Variable	Variable	High (labor-intensive)	Limited scalability; research use
Urinary CXCL10	Moderate	Moderate	Low (about 100-300 dollars)	Research-stage; promising results
Urinary miRNAs	Variable	Variable	Moderate (about 500-1000 dollars)	Experimental; under validation
MMDx	64%	91%	About 3000-4000 dollars per biopsy	Validated; used in specialized centers
Banff transcript panel	Not yet published	Not yet published	TBD	Under evaluation by Banff committee

PPV: Positive predictive value; NPV: Negative predictive value; cfDNA: Cell free DNA; FDA: Food and Drug Administration; kSORT: Kidney Solid Organ Response Test; PCR: Polymerase chain reaction; GEP: Gene expression profiling; ELISPOT: Enzyme linked immuno spot; IFN: Interferon; CXCL10: C-X-C motif chemokine 10; miRNAs: MicroRNAs; MMDx: Molecular Microscope Diagnostic System; TBD: To be determined.

Overall, these biomarkers face common challenges, including high costs, limited accessibility, variable accuracy (particularly low PPV), and incomplete validation across broad patient populations. Consequently, while they offer valuable adjunctive information, they cannot yet replace biopsy as the gold standard for AMR diagnosis.

Comparison of clinical utility of blood and urinary biomarkers: Blood and urine biomarkers offer distinct yet complementary clinical utilities in transplant monitoring. Blood-based markers such as dd-cfDNA reflect systemic allograft injury by detecting fragments of donor DNA released into the recipient’s circulation. They are highly specific for graft damage, particularly AMR, and can signal injury even before a rise in creatinine. However, dd-cfDNA testing requires specialized assays, is costly, and may be elevated in non-immune injuries or systemic conditions, limiting its specificity for rejection alone. In contrast, urine biomarkers such as CXCL10 capture local immune activation within the kidney, as this IFN-γ-induced chemokine rises during TCMR and subclinical inflammation. Urinary CXCL10 is advantageous for its non-invasive sampling, lower cost, and particular utility in pediatric patients, though it lacks specificity since infections and proteinuria can also raise its levels. When combined, these markers provide complementary insights: Elevated dd-cfDNA with high CXCL10 suggests active rejection with injury, high CXCL10 with normal dd-cfDNA indicates subclinical immune activation, and elevated dd-cfDNA with normal CXCL10 points to non-immune mediated damage such as ischemia or drug toxicity. Thus, dd-cfDNA offers robust early detection and is clinically validated, but comes at a higher cost. Urinary CXCL10 is a promising, low-cost alternative for monitoring immune activity, especially in AMR, though it requires further validation for routine use (Table 4).

Table 4 Comparison of clinical utility of blood vs urine biomarkers in kidney transplant monitoring.

Feature	dd-cfDNA	Urinary CXCL10
Sample type	Peripheral blood	Urine
Target	Fragmented donor DNA from injured graft	Chemokine linked to immune activation
Clinical role	Detects active graft injury and acute rejection	Predicts and monitors acute rejection, especially AMR
Sensitivity/specificity	About 59% sensitivity, about 85% specificity at 1% threshold	High NPV for ruling out rejection
PPV/NPV	PPV: 61%, NPV: 84%	NPV: High; PPV varies
Timing of detection	Early detection due to short half-life (about 30 minutes)	Reflects ongoing inflammation; may lag behind dd-cfDNA
Regulatory status	FDA-approved; reimbursed by Medicare	Research-stage; not yet standardized
Cost	About 2800 dollars per test	About 100-300 dollars per test
Advantages	High specificity; validated in multicenter trials	Noninvasive, inexpensive, easy to collect repeatedly
Limitations	Costly; may be affected by other injuries	Limited standardization; influenced by urine concentration
Use in subclinical rejection	Useful when combined with other assays (e.g., TruGraf)	Promising but less validated for subclinical cases

dd-cfDNA: Donor-derived cell-free DNA; CXCL10: C-X-C motif chemokine 10; AMR: Antibody mediated rejection; NPV: Negative predictive value; PPV: Positive predictive value; FDA: Food and Drug Administration.

DSA monitoring: With advances in detection technologies, the sensitivity and specificity for identifying low-level DSAs have markedly improved, enhancing risk stratification for AMR. In a cohort of 402 deceased-donor kidney transplant recipients, a remote positive complement-dependent cytotoxicity crossmatch demonstrated a sensitivity, specificity, and PPV of 41%, 97%, and 54%, respectively, for predicting AMR[49]. Comparatively, performing DSA detection using a single-antigen (Luminex) bead assay achieved sensitivity, specificity, and PPV values of 91%, 85%, and 35%, respectively.

Several DSA characteristics correlate with poorer outcomes in AMR: (1) Subclass: IgG4-dominant DSAs have been linked to late-onset glomerulopathy and IF, while IgG3-dominant DSAs are associated with early rejection, C4d deposition, and graft failure[50]; (2) Strength: Elevated mean fluorescence intensity of anti-HLA DSAs (> 6000) confers over a 100-fold increased risk of AMR compared to lower mean fluorescence intensity levels (< 465)[49]; (3) Type: De novo DSAs, often arising from immunosuppression lapses, are more deleterious than preexisting DSAs and are linked to worse post-transplant outcomes[51,52]; (4) Complement binding: Complement-activating DSAs, assessed via C1q assays, have been associated with heightened allograft loss risk[53]; and (5) Therapeutic response: Declining DSA levels post-treatment may signal a favorable graft prognosis, although a validated definition of DSA response remains elusive[54].

Endothelial and DSA-selective transcripts: Molecular assays detecting endothelial cell–associated transcripts and DSA-targeted gene signatures capture microvascular injury not evident on conventional histopathology. These tools have shown promise in identifying active antibody-mediated injury and may predict adverse graft outcomes[55]. The Banff committee is currently evaluating a multiorgan transplant gene expression panel, the Banff Human Organ Transplant Panel, for broader diagnostic utility[27].

Imaging techniques: Noninvasive imaging modalities are being explored as complementary approaches for evaluating graft health. 18 fluorodeoxyglucose positron emission tomography combined with CT scanning has demonstrated utility in distinguishing acute rejection from non-rejection states[56]. 31-phosphorus magnetic resonance spectroscopy assesses renal allograft energy metabolism by measuring high-energy phosphate compounds, offering metabolic insights into graft function[57].

Invasive biomarkers

When noninvasive assessments suggest a high likelihood of active disease, invasive diagnostic approaches are often essential to confirm the diagnosis. They also remain indispensable in detecting subclinical graft injury when sensitive noninvasive tools are unavailable.

MMDx: MMDx is a microarray-based diagnostic platform that evaluates mRNA expression profiles in transplant biopsy specimens to detect acute rejection. It utilizes predefined computer algorithms referred to as classifiers to generate standardized reports (typically within 29 hours), comparing a patient’s biopsy transcriptome with a reference database of histologically verified samples. The report yields molecular scores indicating the probability of various conditions, including TCMR[58], AMR[59], any form of rejection[60], acute kidney injury[61], and IF/tubular atrophy[62]. In a study involving 403 renal transplant biopsies, including 56 with histologic/DSA-confirmed AMR, a classifier was developed to assign AMR molecular scores across samples. A positive AMR score (≥ 0.2) demonstrated a PPV of 64% and an NPV of 91% for histologic AMR confirmation. These molecular scores were also significantly correlated with DSA presence and served as independent predictors of graft failure[63]. Discrepancies between histological and molecular findings may result from interpretive variability, algorithmic limitations, or biopsy sampling inconsistencies (i.e., different cores analyzed by MMDx vs standard pathology)[64]. Beyond acute rejection, microarray analysis has been leveraged to assess the risk of fibrosis and long-term allograft loss[65]. Additional transcriptomic approaches such as RNA sequencing[66] and NanoString technology[27] are being explored for more nuanced profiling of kidney allograft biopsies, expanding diagnostic capabilities and research potential.

Kidney allograft biopsy: Despite advances in noninvasive diagnostics, kidney allograft biopsy remains the definitive method for assessing graft dysfunction and is widely regarded as the gold standard[67,68]. Due to this status, it is inherently challenging to evaluate its sensitivity and specificity, even though biopsies are known to have limitations. Surveillance or protocol biopsies are commonly employed to detect subclinical rejection, histopathological evidence of acute rejection that occurs without a concomitant rise in serum creatinine. However, only about 17% of transplant centers routinely perform these biopsies at scheduled intervals post-transplantation[69]. The utility of such surveillance in the contemporary era, where noninvasive biomarkers like dd-cfDNA and gene expression panels are available, remains uncertain. Biopsies are invasive, costly, and logistically burdensome. They carry risks of complications, are subject to sampling variability, and often yield inconsistent interpretations, even among experienced pathologists. Nonetheless, protocol biopsies may still serve a valuable purpose in rigorously structured clinical trials aimed at investigating subclinical graft injury or evaluating emerging therapies.

PREVENTION AND MANAGEMENT

Despite being recognized for over 25 years, AMR continues to pose a formidable challenge in kidney transplantation. Although our understanding of its underlying biology has expanded considerably, no regulatory-approved or evidence-backed standard therapy currently exists, and AMR remains closely associated with poor graft survival[70]. Multiple systematic trials targeting key pathophysiological pathways have failed to demonstrate definitive clinical benefit[70]. Growing evidence suggests that effectively preventing and managing AMR requires a multifaceted approach aimed at disrupting B-cell development, maturation, and function. Current therapeutic recommendations are primarily based on expert consensus and are drawn from case series, observational studies, and a limited number of controlled trials.

Preventive measures

In the absence of approved therapies, prevention remains paramount in mitigating the risk of AMR. The preventive approach largely hinges on the timing of DSA detection, either before transplantation (preformed DSA) or emerging post-transplant (de novo DSA).

Patients with preexisting (preformed) DSAs: These individuals are at significantly increased risk of AMR and subsequent graft failure compared to nonsensitized recipients[71,72]. The degree of risk correlates with DSA strength. For instance, patients with a positive complement-dependent cytotoxicity crossmatch exhibit a higher propensity for AMR and graft loss than those with a positive flow cytometric crossmatch. In turn, flow crossmatch positive patients are at higher risk than those with only a virtual crossmatch, where antibodies are detected using single-antigen bead assays[72,73]. The approach to prevent AMR in patients with preexisting DSAs is multifaceted and starts before transplantation, continues during the perioperative period, and extends into post-transplant immunosuppression and monitoring Table 5[74-77].

Table 5 Approach to prevent the development of antibody-mediated rejection in patients with preexisting donor-specific antibodies.

Type of donor	Type of crossmatch	Pre-transplant treatment	Induction and maintenance immunosuppression therapies	Monitoring after transplant
Patients with a potential living donor	Patients with a positive CDC crossmatch or a strongly positive flow crossmatch	Prefer to use KPD programs[74], rather than desensitization[75]	Appropriate for patients at high risk for the development of acute rejection (plasmapharesis/IVIG and glucocorticoids[77]	Routinely monitor DSA levels at months 1, 3, 6, and 12 post-transplant and then annually
			Patients with positive C4d staining, plasmapheresis (two to three sessions), IVIG, and a single dose of rituximab 375 mg/m would be added	Perform kidney allograft biopsies in all patients who develop a de novo DSA
	Patients with a positive virtual crossmatch1 or a mild to moderate flow crossmatch2	Employ HLA desensitization strategies3[76]	Maintenance on a triple therapy4	-
Patients without a potential living donor	-	Employ HLA desensitization strategies strategies3	-	-

¹Antibodies detected by single antigen bead technology.

²Median channel shift of < 200.

³Include treatment with plasmapheresis, recombinant antithymocyte globulin-thymoglobulin, and rituximab.

⁴Regimen include tacrolimus, mycophenolate, and prednisone.

CDC: Complement-dependent cytotoxicity; KPD: Kidney paired donation; IVIG: Intravenous immune globulin; DSA: Donor-specific antibody; HLA: Human leukocyte antigen.

Patients with de novo (post-transplant) DSA: Recipients who develop DSAs following transplantation are predisposed to late-onset AMR and generally experience poorer clinical outcomes than those with preformed DSAs. The primary drivers of de novo DSA formation include medication nonadherence and suboptimal immunosuppression. The latter can result from intentional minimization strategies, episodes of acute TCMR, malignancies, or opportunistic infections such as BK polyomavirus or cytomegalovirus that necessitate reducing immunosuppressive therapy[78,79].

A proactive immunosuppressive regimen consisting of tacrolimus, mycophenolate mofetil, and prednisone is recommended. Tacrolimus trough levels should be monitored monthly during the first three years post-transplant, and quarterly thereafter. For patients intolerant to tacrolimus, belatacept is preferred over alternatives like sirolimus or everolimus[80,81]. Routine annual DSA surveillance is advised, with kidney allograft biopsies indicated for any patient in whom de novo DSAs are detected.

Treatment of active AMR

General considerations: Initiate treatment in all patients with biopsy-confirmed active AMR. While surveillance or protocol biopsies are not universally adopted across transplant centers, patients diagnosed with subclinical AMR through such biopsies should still receive appropriate treatment. Patients with C4d-negative AMR should be managed similarly to those with C4d-positive findings, as the therapeutic approach remains consistent.

Therapeutic goals: The primary objectives of AMR treatment are to lower the titers of pathogenic DSAs, eliminate the responsible B cell or plasma cell clones, inhibit complement activation, minimize endothelial injury, and ultimately preserve graft function[6].

Initial therapeutic strategy: Due to limited high-quality evidence, current treatment guidelines align with the 2023 Kidney Disease: Improving Global Outcomes clinical practice guidelines and the 2019 consensus recommendations from The Transplantation Society[6,82]. These frameworks emphasize that histologic findings alone are insufficient for understanding the pathophysiology of AMR. Recent findings reveal that microvascular injuries may occur independently of alloantibodies, highlighting the need for a nuanced clinical evaluation[83]. Key contextual factors such as the presence of preformed vs de novo DSAs and timing of AMR onset (early vs late) should inform management decisions (Table 6).

Table 6 The Transplantation Society 2019 consensus-based treatment recommendations in alignment with the Kidney Disease: Improving Global Outcomes 2023 guidelines for the treatment of antibody-mediated rejection.

2019 consensus	DSA	Banff 2017 phenotypes	Standard of care	Consider adjunctive therapies	KDIGO 2023 updates
Early acute (< 30 days)	Preexisting DSA (or nonimmunologically naive)	Active AMR	PP/IVIG1 glucocorticoids	Complement	Eculizumab use is restricted to select cases with complement activation; not routine. KDIGO recommends early biopsy confirmation, dd-cfDNA and GEP monitoring, and tailored immunosuppression
				Inhibitor (e.g., eculizumab)
				Rituximab3
				Splenectomy
	Preexisting DSA	Active AMR	PP/IVIG1 glucocorticoids	Rituximab3	KDIGO advises against routine complement blockade unless complement-mediated injury is proven
Late (> 30 days)	De novo DSA	Chronic AMR	Optimize baseline immunosuppression2	-	KDIGO supports noninvasive monitoring (e.g., dd-cfDNA, GEP), and individualized therapy based on graft function
		Active AMR	Optimize baseline immunosuppression2	PP, IVIG	KDIGO encourages molecular diagnostics (e.g., MMDx) and consideration of adherence issues
				Rituximab
		Chronic AMR	Evaluate and manage nonadherence	IVIG	KDIGO highlights importance of patient education, adherence tracking, and long-term graft surveillance

¹Daily or alternate day plasmapheresis sessions × 6 based on donor-specific antibody titer; intravenous immunoglobulin 100 mg/kg after each plasmapheresis treatment or intravenous immunoglobulin 2/kg at the end of plasmapheresis treatments.

²Includes the use of tacrolimus with the goal through levels of > 5 ng/mL and use of maintenance steroid equivalent to prednisone 5 mg/day.

³If age < 70 years, estimated glomerular filtration rate ≥ 20 mL/minute/1.73 m³, lower chronicity scores on biopsy (interstitial fibrosis + tubular atrophy + fibrous intimal thickening + allograft glomerulopathy < 8) and evidence of severe disease [higher donor-specific antibody, diffuse C4d staining, or more extensive microvascular inflammation (i.e., glomerulitis score + peritubular capillary score ≥ 4) on biopsy].

AMR: Antibody-mediated rejection; DSA: Donor-specific antibody; IVIG: Intravenous immunoglobulin; PP: Plasmapheresis; KDIGO: Kidney Disease: Improving Global Outcomes; dd-cfDNA: Donor-derived cell-free DNA; GEP: Gene expression profiling; MMDx: Molecular Microscope Diagnostic System.

Prophylactic measures: All patients undergoing treatment for active AMR should receive antimicrobial and antiviral prophylaxis consistent with early post-transplant protocols. Additionally, antifungal agents and a histamine-2 blocker should be considered to prevent opportunistic infections and stress ulceration, though prophylactic strategies may vary between institutions.

Monitoring therapeutic response: Serum creatinine, electrolytes, and complete blood counts should be monitored before each plasmapheresis session or weekly over four weeks if plasmapheresis is not employed. The utility of serial DSA monitoring post-treatment remains controversial, and practices differ across centers. In cases of persistent allograft dysfunction, a repeat kidney biopsy should be considered[84].

Following treatment, the one-year graft survival rates are approximately 80% for clinical AMR and 95% for subclinical AMR[85]. Successful reversal of AMR within three months is defined by meeting all of the following criteria: (1) Serum creatinine reduction to within 20%-30% of baseline; (2) Normalization of proteinuria; (3) ≥ 50% decline in immunodominant DSA levels; and (4) Histologic resolution of AMR findings on follow-up biopsy.

Adjustments to maintenance immunosuppression: Increase the daily dose of oral prednisone; Maximize the dose of the antiproliferative agent; Raise tacrolimus dosage to achieve trough levels 20%-25% above baseline.

For patients who do not respond to first-line therapy, a repeat allograft biopsy is essential. If histologic findings do not support ongoing rejection, therapy should be discontinued. Conversely, if active AMR persists, escalation to second-line (rescue) treatments is warranted.

Second-line therapies for refractory AMR

In cases where AMR proves unresponsive to first-line treatment, escalation to second-line therapy may be warranted. However, the intensity of further interventions, particularly plasmapheresis, should be carefully balanced against the patient’s comorbid conditions and the associated risks of infection and malignancy.

Bortezomib: A proteasome inhibitor primarily used in the treatment of multiple myeloma, bortezomib has demonstrated improved DSA reduction and histologic response in early-onset rejection (within six months post-transplant). However, its efficacy in late AMR remains limited[86,87].

Complement inhibitors: Given the central role of complement activation in AMR pathogenesis, targeting this pathway offers a mechanistic rationale for therapeutic intervention.

Eculizumab: This fully humanized monoclonal antibody inhibits the complement cascade by binding to the C5 component, thereby preventing formation of the membrane attack complex. While eculizumab has been employed prophylactically in desensitized, high-risk recipients[88], its effectiveness remains unproven in randomized trials. Nonetheless, case reports suggest potential utility as a rescue agent in refractory AMR[89]. Recommended dosing includes an initial 1200 mg intravenous infusion followed by 900 mg weekly for 3-4 weeks. Due to the risk of severe or fatal meningococcal infections, patients must receive meningococcal vaccination at least two weeks before initiation. Notably, the drug has shown limited efficacy in C4d-negative or chronic AMR, indicating a more restricted benefit in acute, complement-mediated cases[90,91].

C1 inhibitors: Since DSA-mediated binding to C1q is linked to adverse graft outcomes and severe AMR phenotypes[92], C1 inhibition has emerged as a potential strategy. However, clinical adoption remains limited, pending more robust evidence on safety and therapeutic effectiveness.

Immunoadsorption: Protein A-based immunoadsorption has been utilized as an alternative to plasmapheresis for antibody removal in AMR management[93]. Selective immunoadsorption offers an appealing option by obviating the need for IVIG, potentially reducing treatment costs and minimizing immunomodulatory burden, assuming comparable clinical efficacy.

Splenectomy: Although not a standard approach, splenectomy has been considered in select cases of refractory AMR unresponsive to plasmapheresis and/or IVIG. Given the paucity of supporting evidence, its use remains highly selective and largely institution-specific[94].

Special populations in acute rejection

Mixed acute rejection (AMR + TCMR): For patients within 1 year post-transplant, the recommended regimen includes: (1) Glucocorticoids; (2) Plasmapheresis; (3) IVIG (every other day × 4 doses); and (4) rATG-Thymoglobulin (1.5-3 mg/kg every other day × 3 doses).

Subclinical rejection: Diagnosed via protocol biopsy without elevated serum creatinine. Treatment mirrors clinical AMR, even in the absence of symptoms[95].

C4d-negative AMR: Despite lacking C4d staining, these cases are now recognized as genuine AMR. Treat similarly to C4d-positive AMR, as outcomes can still be poor if untreated[96].

Patients with non-HLA DSA: Non-HLA DSAs include antibodies such as anti–angiotensin II type 1 receptor[97], anti-endothelial antibodies[98], perlecan fragment LG3[99], anti-Ro/Sjögren’s Syndrome-A, and anti-centromere protein[100].

Treatment aligns with that for AMR involving anti-HLA DSA: For patients with anti- angiotensin II type 1 receptor antibodies, an angiotensin II receptor blocker should also be administered to mitigate antibody-mediated effects[97,98].

Treatment of chronic AMR

General considerations: Chronic AMR is the leading cause of graft failure and poses a greater therapeutic challenge than active AMR due to irreversible tissue injury and significantly reduced graft survival[101]. Currently, there is a lack of high-quality evidence to guide optimal treatment, with most recommendations based on observational studies[101-103].

Therapeutic goals: The primary aim is to suppress B-cell development, maturation, and activity. While combination therapies are often used in active AMR, their safety and efficacy in chronic AMR remain uncertain.

Initial treatment approach: (1) Management mirrors that of active AMR occurring beyond the first post-transplant year. Although intensifying immunosuppression may seem logical, the limited supporting data and heightened risk of infections warrant a cautious approach; (2) First-line therapy: Glucocorticoids and IVIG; and (3) Optional addition: Rituximab may be considered in patients under 70 years of age, with relatively preserved graft function and severe disease.

Management of non-responders: For patients who do not respond to initial therapy, the following agents may be considered: (1) Tocilizumab: An anti-interleukin (IL)-6 receptor monoclonal antibody (8 mg/kg IV, max 800 mg, monthly). Studies have shown it can stabilize renal function, reduce DSAs, and improve histologic findings in chronic AMR[7]; (2) Clazakizumab: Another IL-6/IL-6 receptor–targeting antibody[9]. Although the IMAGINE trial faced enrollment challenges, a phase 3 trial of tocilizumab is ongoing in this population[8]; and (3) Felzartamab: A CD38 monoclonal antibody targeting plasma and NK cells. It has received orphan drug and breakthrough therapy designations from the Food and Drug Administration and European Commission[10,104]. Following encouraging phase 2 results, a phase 3 trial is underway, incorporating personalized treatment guided by dd-cfDNA monitoring[104]. An evidence-based comparison of therapies in acute and chronic AMR is given in Table 7.

Table 7 Evidence-based comparison of acute vs chronic antibody-mediated rejection therapies.

Therapy	Use in acute AMR	Use in chronic AMR	Evidence summary	KDIGO 2023 position
PP	First-line for antibody removal; often combined with IVIG	Used selectively; less effective in chronic injury	Supported by multiple case series and consensus guidelines	Recommended in acute AMR; limited role in chronic AMR
IVIG	Adjunct to PP; modulates immune response	Used in chronic AMR with DSA presence	Moderate evidence; variable dosing strategies	Supported in both acute and chronic AMR
Rituximab	Targets CD20+ B cells; used in combination with PP/IVIG	Sometimes used in chronic AMR with active inflammation	Mixed results; better efficacy in early AMR	Considered in both settings; not universally effective
Eculizumab	Complement inhibition in severe or refractory acute AMR	Not recommended for routine chronic AMR	Effective in complement-mediated AMR; high cost, limited trials	Restricted use: Only in complement-driven AMR
Bortezomib	Used in refractory acute AMR; targets plasma cells	Limited efficacy in chronic AMR; poor outcomes in late-stage fibrosis	Early studies showed promise; later trials failed to show consistent benefit[83]	Not recommended for chronic AMR; use in acute AMR remains investigational
Tocilizumab (IL-6 blocker)	Investigational; targets inflammatory cytokine pathways	Emerging therapy in chronic AMR	Phase 2/3 trials ongoing; promising results in chronic inflammation	Under evaluation; not yet standard of care
CD38 antibodies (e.g., felzartamab)	Targets long-lived plasma cells and NK cells	Promising in chronic AMR with persistent DSA	Recent trials show reversal of AMR activity	KDIGO supports further research; not yet routine
Molecular diagnostics (MMDx, dd-cfDNA, GEP)	Used to confirm and monitor acute AMR	Valuable for detecting subclinical chronic AMR	High NPV; improves diagnostic precision	Strongly recommended for both acute and chronic AMR monitoring

PP: Plasmapheresis; IVIG: Intravenous immunoglobulin; IL: Interleukin; MMDx: Molecular Microscope Diagnostic System; dd-cfDNA: Donor-derived cell-free DNA; GEP: Gene expression profiling; AMR: Antibody-mediated rejection; DSA: Donor-specific antibody; NK: Natural killer; NPV: Negative predictive value; KDIGO: Kidney Disease: Improving Global Outcomes.

Emerging therapies under investigation for AMR

Fostamatinib: Preclinical studies in sensitized rat models suggest that fostamatinib, a spleen tyrosine kinase inhibitor, may suppress the production of DSAs[105]. A phase 2 clinical trial is currently evaluating its potential in treating chronic active AMR.

High-dose IVIG: The efficacy of prolonged high-dose IVIG therapy, administered over six months alongside steroid pulse therapy, is being explored in a randomized open-label trial conducted in Australia. Results from this study are pending publication.

BIVV020: This next-generation anti-C1s monoclonal antibody targets the classical complement pathway. A phase 2 trial is underway to assess its safety and efficacy in both the prevention and treatment of AMR when used in conjunction with standard-of-care therapy[106].

Efgartigimod: An FcRn antagonist approved for myasthenia gravis, efgartigimod works by inhibiting neonatal Fc receptor-mediated IgG recycling, thereby reducing circulating IgG levels[107]. A phase 2 trial is currently investigating its role in managing late-stage AMR.

Ethical dilemmas in experimental therapies

Experimental therapies such as felzartamab raise complex ethical dilemmas that sit at the crossroads of hope, innovation, and inequity. One major concern is cost–access disparity: While high-income patients or those in trial hubs may gain access, resource-limited populations are often excluded, creating inequities where survival depends on geography and socioeconomic status rather than medical need. Clinical trials themselves tend to recruit from well-resourced centers, leaving out diverse patient groups and raising questions about justice, exploitation, and the fair distribution of risks and benefits. At the same time, while such drugs offer hope to patients with few options, their long-term efficacy and safety remain uncertain, creating conflicts between respecting patient autonomy and the duty to avoid harm. Another dilemma lies in resource allocation; funding a handful of patients on expensive experimental agents may divert resources from basic care that benefits many. Post-trial access is also problematic: Withdrawing a drug from patients who may be benefiting once a study ends undermines continuity of care. Globally, pharmaceutical development often prioritizes profitable markets over the greatest disease burden, deepening disparities between high- and low-income settings. Finally, restrictive compassionate use policies force regulators and companies to weigh early access against the need for robust scientific evidence and public trust. Taken together, these dilemmas highlight enduring challenges of justice, equity, and responsibility in the advancement of experimental therapies.

FUTURE DIRECTIONS

The future roadmap for diagnostic and therapeutic updates of AMR in kidney transplantation will focus on improving early detection, refining diagnostic tools, and developing more effective treatments. Advancements in biomarker discovery, including the identification of specific antibodies and immune mediators, will enable more precise diagnosis and monitoring of AMR. The use of next-generation sequencing and high-resolution molecular assays will improve the sensitivity of detecting DSAs and the presence of pathogenic immune responses at earlier stages. In terms of therapeutics, the development of targeted therapies to inhibit complement activation, B-cell depletion, and the modulation of the humoral immune response will offer promising strategies to prevent or mitigate AMR. Agents like eculizumab, a complement inhibitor, are being explored for their potential to reduce graft injury in AMR, while new immunosuppressive agents will be tailored to specifically target humoral immunity. Additionally, personalized treatment regimens that account for individual patient risk profiles and genetic predispositions will enhance the management of AMR. Ongoing research and clinical trials will play a pivotal role in improving patient outcomes by advancing both diagnostic and therapeutic approaches to this complex immune-mediated condition.

CONCLUSION

AMR remains a major cause of kidney graft dysfunction and failure. While DSAs are a key risk factor, not all DSAs are pathogenic. Current diagnostic tools often lack sensitivity, missing subclinical changes, and diagnosis still depends on invasive biopsy. The adoption of non-invasive biomarkers is limited by technical and validation challenges. Despite advances in AMR biology, no standardized or approved treatment exists. Current guidelines are based on earlier consensus, with apheresis or complement inhibitors typically reserved for early AMR and steroid pulses used for concurrent TCMR. Many therapeutic trials have yielded disappointing outcomes. However, emerging therapies such as CD38 antibodies and complement inhibitors show promise, especially in early or severe cases. Long-term safety data are still lacking, and results from ongoing phase 3 trials are awaited. Novel strategies under investigation offer hope for improved management.