Objectives: Late and chronic active antibody-mediated rejection represents distinct phenotypes after kidney transplant, yet their comparative outcomes remain poorly defined. This study compared clinical outcomes, treatment patterns, and long-term graft survival between late and chronic active antibody-mediated rejection.
Materials and Methods: This single-center retrospective cohort study included 111 kidney transplant recipients with biopsy-proven antibody-mediated rejection (50 late, 61 chronic active) between 2009 and 2024. Groups were matched on baseline demographics, donor type, comorbidities, and maintenance immunosuppression. Outcomes included treatment intensity, complications, renal function, and 5-year graft survival.
Results: Both groups had comparable baseline characteristics. Patients with late antibody-mediated rejection received significantly more intensive therapy: higher rituximab use (52.0% vs 34.4%; P = .⁰⁰¹), more plasma exchange sessions (5.04 vs 2.11; P < .⁰⁰¹), and higher rituximab doses (513.4 vs 252.6 mg; P = .⁰⁰⁴). Patients with late antibody-mediated rejection showed trends toward higher BK viremia (6.0% vs 1.6%), whereas patients with chronic active antibody-mediated rejection had more new-onset diabetes after transplant (21.3% vs 8.0%). At presentation, patients with chronic active antibody-mediated rejection had worse graft function (creatinine 176 vs 139 µmol/L; P = .⁰¹³). Despite divergent initial management, both groups converged at 5 years with no significant difference in graft function or survival; overall graft loss exceeded 70% in both groups (P > .⁰⁵).
Conclusions: Late and chronic active antibody-mediated rejection received different treatment intensities based on perceived acuity, yet both groups had similarly poor 5-year outcomes with >70% graft failure. These findings challenge current therapeutic paradigms and emphasize the urgent need for earlier diagnosis and more effective agents beyond standard regimens of steroids, plasma exchange, intravenous immunoglobulin, and rituximab.

Key words : Acute versus chronic, anti-HLA antibody-mediated rejection, Kidney graft rejection

Introduction
Kidney transplantation remains the optimal treatment for end-stage kidney disease, offering superior quality of life and survival compared with long-term dialysis. However, the long-term success of this life-saving intervention is persistently threatened by the specter of allograft rejection. Among rejection phenotypes, antibody-mediated rejection (ABMR) has emerged as the principal cause of late allograft attrition, a formidable barrier to achieving the goal of lifelong graft function.¹ Driven by donor-specific antibodies (DSA) that target HLA or other endothelial antigens, ABMR provokes a cascade of endothelial injury, microvascular inflammation, and progressive tissue scarring, culminating in transplant glomerulopathy and graft failure.² The Banff classification, the international histopathological schema for diagnoses of transplant pathology, has evolved to recognize distinct yet overlapping clinical-pathological entities. Late ABMR, typically diagnosed more than 1 year posttransplant, often presents with an acute or subacute decline in graft function and is characterized by significant microvascular inflammation (g+ ptc+) with or without early chronic changes.³ In contrast, chronic active ABMR is defined by the coexistence of both active microvascular inflammation and established chronic architectural damage, most notably transplant glomerulopathy (cg+), signifying an ongoing injurious process superimposed on irreversible fibrosis.⁴ This histological distinction implies different underlying disease kinetics, with a potentially more fulminant immune activation in late ABMR versus a smoldering, progressive injury in chronic active ABMR. This nosological separation carries profound clinical implications, particularly for therapeutic decision-making. Although consensus guidelines outline general principles for ABMR management, often involving a combination of pulse corticosteroids, plasma exchange (PE), intravenous immunoglobulin (IVIG), and B-cell-depleting agents like rituximab, the application of these therapies is highly variable in practice.⁵ Clinicians often intuit that the more “active” phenotype of late ABMR warrants a more aggressive, multi-modal intervention akin to acute rejection protocols, aiming to abort the alloimmune assault. Conversely, the diagnosis of chronic active ABMR, with its burden of fibrosis, presents a therapeutic dilemma, balancing the potentially marginal benefit of intensified immunosuppression against the significant risks of infection, malignancy, and metabolic toxicity in setting of likely irreversible damage.⁶ Literature has often conflated ABMR subtypes or have focused on general risk factors, leaving a gap in our understanding of the comparative natural history and treatment responsiveness of late versus chronic active ABMR. In this study, we performed a detailed comparative analysis of outcomes of late ABMR versus chronic active ABMR in kidney transplant recipients.

Materials and Methods
We performed a single-center retrospective cohort study of consecutive kidney transplant recipients who developed biopsy-proven late ABMR or chronic ABMR between 2009 and 2024. Patients were diagnosed and classified for ABMR with the contemporaneous Banff criteria, distinguishing active/late ABMR from chronic active and chronic (inactive) ABMR based on microvascular inflammation, chronic glomerulopathy, peritubular capillary basement membrane multilayering, C4d staining, and DSA status. Eligible patients were adults with a functioning kidney allograft at the time of event biopsy, histologic features consistent with ABMR, and available clinical and follow-up data for at least 5 years after rejection diagnosis. We excluded patients with combined organ transplants, primary nonfunction, and biopsies not meeting Banff ABMR criteria. Late ABMR was defined as biopsy-proven active ABMR occurring >6 months posttransplant without predominant chronic lesions, whereas chronic ABMR was defined by transplant glomerulopathy and/or chronic microvascular injury with or without ongoing activity, according to Banff 2019 refinements. We extracted demographic and baseline transplant variables from electronic records, including age, sex, original kidney disease, type of donor (living or deceased), pretransplant dialysis modality and duration, pretransplant comorbidities, and immunologic risk profile. We also recorded induction and maintenance immunosuppression (calcineurin inhibitor, antimetabolite, corticosteroids, and any mechanistic target of rapamycin inhibitor), as well as time from transplant to ABMR diagnosis and presence, class, and strength of DSAs when available. We matched the 2 groups (late ABMR and chronic ABMR) a priori on key baseline variables (age, sex, donor type, dialysis modality, comorbidities, original kidney disease, and maintenance immunosuppression) to minimize confounding in comparisons of treatment and outcomes. We reviewed pretransplant renal replacement modality (hemodialysis, peritoneal dialysis, pre-emptive transplantation) and immediate graft function versus delayed graft function; immediate graft function was defined as the absence of dialysis requirement in the first posttransplant week and an early decline in serum creatinine. Management of ABMR followed our center’s standard protocols, which are broadly aligned with international practice: all patients received optimization of baseline maintenance immunosuppression and at least 1 of high-dose intravenous methylprednisolone pulses (500 mg daily for 5 days), IVIG (total dose of 2 g/kg), and PE, with or without rituximab at the treating physician’s discretion. We collected dosing and number of sessions of PE and cumulative IVIG and rituximab doses from infusion records, recognizing that rituximab is widely used as adjunctive therapy in ABMR but with heterogeneous regimens across centers. We obtained posttransplant complications prospectively, focusing on infectious events (bacterial, viral, and fungal), particularly BK viremia and BK virus-associated nephropathy, and metabolic complications such as new-onset diabetes after transplant (NODAT), given their known association with intensified immunosuppression. Standard definitions were applied: BK viremia was defined by detectable plasma BK viral load of >1000 copies/mL, biopsy-proven BK nephropathy required characteristic histology with SV40 positivity, and NODAT was defined according to American Diabetes Association criteria beyond the immediate posttransplant period. Renal function was assessed using serum creatinine and estimated glomerular filtration rate at 3 predefined time points: at the start of ABMR treatment (index), 3 months after completion of the initial rejection management, and at last follow-up approximately 5 years after ABMR diagnosis. Graft outcomes included death-censored graft failure, defined as return to chronic dialysis or preemptive retransplant.

Statistical analyses
We presented continuous variables as mean ± SD or median (interquartile range) according to distribution and made comparisons using the t test or the Mann-Whitney U test, as appropriate. We presented categorical variables as counts and percentages, which we compared with the χ2 or the Fisher exact test. P < .⁰⁵ was considered statistically significant. We used SPSS version 26 for our analyses.

Results
Most patients in the 2 groups (late ABMR vs chronic ABMR) were men (62.0% vs 70.5%; P = .³⁴) with median age of 32.8 ± 15.0 versus 35.9 ± 14.2 years (P = .²¹), respectively. Table 1 presents demographics of the 2 groups, which were similarly matched (P > .⁰⁵). Most patients had immediate graft function without significant difference between the 2 groups (P > .⁰⁵). We observed that most of the study patients (32 in late ABMR [64.0%] vs 42 in chronic ABMR [68.9%]) underwent hemodialysis pretransplant without significant difference between groups (P > .⁰⁵). The management of rejection was comparable between the 2 groups (augmented maintenance immunosuppression, pulse steroid, IVIG or PE; P > .⁰⁵), but the number of patients who received rituximab was significantly higher in the late ABMR group (P = .⁰⁰¹). The 2 groups were comparable regarding the mean dose of pulse steroid or IVIG (P > .⁰⁵). The mean number of PE sessions was significantly higher in the late ABMR versus the chronic ABMR group (5.04 ± 4.15 vs 2.11 ± 3.43 sessions, respectively; P < .⁰⁰¹). Moreover, the mean dose of rituximab was significantly higher in late ABMR group (513.4 ± 57 vs 252.6 ± 35 mg, respectively; P = .⁰⁰⁴). Regarding posttransplant complications (viral and non-viral infections) (Table 2), the late ABMR group showed higher BK viremia and nephropathy, whereas the chronic ABMR group showed higher prevalence of NODAT (P > .⁰⁵). Moreover, graft function (as represented by serum creatinine) was significantly higher in the chronic ABMR group at the start of management of rejection and at 3 months after (Table 3; P < .⁰⁵). At last follow-up (5 years after ABMR diagnosis), no significant difference in graft function was shown between the 2 groups (P > .⁰⁵). This was reflected for both patient and poor graft outcomes (Table 4, Table 5) (>70% graft loss), which was similar in the 2 groups (P > .⁰⁵).

Discussion
This comparative study of late and chronic active ABMR provided a compelling and sobering narrative on the management and outcomes of this primary barrier to long-term kidney allograft survival. The key findings, which showed equivalent baseline characteristics, divergent initial therapeutic strategies, and a convergent, dismal long-term outcome, prompt nuanced discussion on the integration of a contemporary understanding of ABMR pathogenesis, treatment paradigms, and prognostic determinants. The lack of significant differences in donor type, comorbidities, original disease, and immunosuppressive regimens (induction and maintenance) between the late and chronic active ABMR groups is a critical strength. Lack of differences between groups effectively isolated the phenotype and timing of the ABMR diagnosis as the primary variable of interest, minimizing the confounding influence of a preexisting risk profile. The male predominance (62%-70.5%) aligned with global transplant registries, which often reported a higher proportion of male recipients, potentially reflecting underlying disease epidemiology or access to transplant.⁷ The comparable rates of pretransplant hemodialysis and immediate graft function further suggested that early perioperative factors and baseline allograft quality were unlikely drivers of the subsequent divergent ABMR presentations. The most pronounced difference between the groups was in the intensity and modality of antirejection therapy. The late ABMR group had intensified B-cell depletion and antibody removal. The significantly higher use of rituximab (P = .⁰⁰¹) and plasmapheresis (PE) (mean 5.04 vs 2.11 sessions; P < .⁰⁰¹) in the late ABMR group reflected a more aggressive intervention for this phenotype. Late ABMR, often presenting with a rapid decline in graft function and de novo DSAs, was perceived as an active, potentially fulminant alloimmune assault. This triggered a response modeled on acute rejection protocols, combining antibody clearance (PE) with targeted B-cell depletion. This approach is consistent with consensus recommendations that advocate for multi-modal therapy in active ABMR.⁵ In contrast, treatment in the chronic active ABMR was less intensive, likely reflected the challenging clinical dilemma, with biopsy showing both active microvascular inflammation (g >0) and chronic, irreversible changes like transplant glomerulopathy (cg). Physicians may be hesitant to escalate immunosuppression aggressively in the face of established fibrosis, weighing the marginal potential benefit against risks of infection and metabolic complications. This practice aligns with observations that chronic active ABMR is often less responsive to therapy, leading to a more conservative or palliative approach focused on slowing progression rather than achieving reversal.⁸ This therapeutic dichotomy is the study’s pivotal observation. Regarding infection and metabolic trends, no significant differences were shown with regard to higher BK viremia/nephropathy in the late ABMR group and increased incidence of NODAT in the chronic ABMR group. BK virus reactivation is a recognized consequence of intensive immunosuppression, particularly with agents affecting T-cell and B-cell function. The combination of rituximab, PE, and pulse steroids in the late ABMR group created a perfect storm for viral replication.⁹ Conversely, NODAT is strongly associated with prolonged exposure to calcineurin inhibitors and corticosteroids. The chronic ABMR group, likely living with higher baseline immunosuppression for a longer duration to manage smoldering rejection, may have been predisposed to this metabolic complication.¹⁰ Concerning graft function at presentation, the chronic active ABMR group had significantly higher serum creatinine levels (reflecting worse function) at diagnosis and 3 months posttreatment, which was expected and almost tautological. The Banff criteria for chronic active ABMR require the presence of chronic tissue injury (eg, cg). This structural damage directly correlates with impaired function.¹¹ The Late ABMR group, potentially diagnosed on a for-cause biopsy after a functional decline but before extensive fibrosis sets in, presented with a relatively better-preserved parenchymal reserve. The most profound and clinically significant finding is the complete convergence of graft function and survival at 5 years, with >70% graft loss in both groups. This result demands careful interpretation and has major implications. Our data suggested that the more intensive, multi-modal therapy used in the late ABMR group did not confer a long-term survival advantage. This challenges the assumption that aggressively treating a seemingly “acute” late presentation can reset the alloimmune course. Potential explanations include irreversible injury at diagnosis, where even “late active” ABMR may have substantial subclinical chronicity. In addition, the diagnostic biopsy may sample a less scarred area, but the allograft may already be committed to failure. Another explanation is the persistence of the alloimmune response despite the current therapies (rituximab, PE, IVIG), which may modulate but rarely eradicate the memory B-cell and long-lived plasma cell clones responsible for DSA production. Therefore, recrudescence of disease is common.⁶ The poor prognosis of established ABMR is phenotype independent. The 70% graft loss rate underscores that any diagnosis of clinical ABMR (whether labeled late/active or chronic/active) is a watershed event portending a high risk of graft failure. This aligns with large registry analyses that have shown that ABMR is the leading cause of late graft loss, with 5-year graft survival rates postdiagnosis often below 50%.¹² Our study confirmed that once the pathological process is advanced enough to cause clinical concern and warrant a biopsy diagnosis, the long-term prognosis is grim regardless of the histological subtleties. The armamentarium of steroids, PE, IVIG, and rituximab is insufficient to change the natural history of established ABMR. This calls for a paradigm shift. The goal should be to prevent ABMR through improved HLA matching, adherence monitoring, and protocols for early DSA surveillance with protocol biopsies to detect subclinical injury before irreversible chronicity sets in.¹³ Future therapeutic trials must stratify by ABMR phenotype (eg, using Banff scores for activity [g, ptc] and chronicity [cg, ct]) and DSA characteristics. The uniform poor outcome does not mean all ABMR is the same; rather, it means our current one-size-fits-all approaches are ineffective. Hope lies in novel agents targeting the alloimmune pathway, such as complement inhibitors (eculizumab, anti-C5) and newer C1s inhibitors (eg, sutimlimab) that aim to block the final common pathway of antibody-mediated injury, showing promise in early trials for active ABMR.¹⁴ Agents like daratumumab (anti-CD38) or bortezomib offer the potential to deplete the source of pathogenic antibodies, moving beyond B-cell depletion to target antibody-producing plasma cells directly.⁶ Drugs like clazakizumab (anti-interleukin-6) aim to modulate the inflammatory microenvironment and plasma cell survival, addressing a key cytokine in ABMR pathogenesis.¹⁵

Conclusions
Our study demonstrated that, while late and chronic active ABMR elicit different initial therapeutic responses based on perceived acuity, they represent points on a continuum of a destructive process that, once clinically manifest, leads to graft failure in most patients within 5 years. The convergence of outcomes despite divergent management underscores the limited efficacy of current standard therapies and the urgent need for earlier diagnosis through vigilant monitoring and the integration of novel, more effective agents into treatment algorithms. The battle for long-term graft survival will be won not in treating advanced ABMR but in preventing it and intercepting it at its earliest, most vulnerable stage.