Introduction
Kidney transplantation (KT) is considered the optimal treatment for patients with end-stage kidney disease [1]. However, the lifespan of a transplanted kidney is not permanent, and the kidney eventually loses its function. There are various causes of graft dysfunction in transplanted kidneys; however, chronic active antibody-mediated rejection (cABMR) is considered to be the most important cause of late allograft failure [2–4]. However, no treatment has been proven to be effective for cABMR [5]. Rituximab (anti-CD20) combined with intravenous immunoglobulin (IVIG) has failed to show efficacy in delaying the decline in estimated glomerular filtration rate (eGFR) [6], and bortezomib (a proteasome inhibitor) has not demonstrated efficacy in reducing donor-specific anti-human leukocyte antigen antibody (DSA) production [7]. Eculizumab (a C5 complement inhibitor) also showed no clear effectiveness, with no significant improvement in eGFR trajectory or reduction in endothelial cell-associated transcripts [8].
Meanwhile, interleukin-6 (IL-6) is a key cytokine involved in the adaptive immune response and is closely associated with kidney allograft injury [9,10]. IL-6 promotes the maturation of naïve T cells toward T follicular helper cells (Tfh), the formation of germinal centers, and the development of antibody-producing plasma cells from naïve B cells [10]. Increased expression of IL-6 mRNA transcripts in allograft tissue [11] or elevated serum IL-6 levels [12] have been observed in KT recipients with allograft rejection, indicating a possible association between IL-6 and kidney allograft rejection. Tocilizumab (TCZ), an anti-IL-6 receptor (IL-6R) monoclonal antibody, has demonstrated efficacy in animal study [13] and clinical trial for desensitization [14]. In the first trial by Choi et al. [15] on the treatment of cABMR, TCZ treatment demonstrated improvements in allograft survival, reduction in the mean fluorescence intensity (MFI) levels of DSA, and stabilization of eGFR. However, subsequent studies reported inconsistent results. While some studies have reported reductions in DSA MFI levels, stabilization of allograft function, and decreased inflammation and microvascular lesions on follow-up biopsies [16], others have not demonstrated significant treatment effectiveness [17–19].
Based on the background above, this study aimed to investigate the effectiveness and safety of TCZ in KT recipients with cABMR. We analyzed the clinical data of 18 patients diagnosed with cABMR by allograft biopsy who underwent TCZ treatment to assess its potential as a therapeutic option for cABMR in KT recipients.
Methods
Study population
This single-center, retrospective, observational study included 18 KT recipients who were diagnosed with cABMR and received TCZ between November 2019 and October 2023. All patients were histologically confirmed to have cABMR by a single pathologist with extensive experience in transplant nephrology based on the 2019 Banff classification [20]. C4d was assessed by indirect immunofluorescence using monoclonal antibodies (dilution 1:50; Biogenesis). In one case, because of poor immunofluorescence staining quality, paraffin sections were re-stained by immunohistochemistry. C4d positivity was defined as diffuse (>50%) and linear staining of peritubular capillaries. All patients consented to off-label TCZ treatment despite its high cost and lack of insurance coverage.
Tocilizumab treatment protocol
Our treatment protocol is summarized in Fig. 1. Before starting TCZ treatment, the potential reactivation of viral infections, including hepatitis B virus, hepatitis C virus, cytomegalovirus (CMV), and BK virus (BKV), was screened, along with an interferon gamma release assay for latent tuberculosis. Nystatin, trimethoprim/sulfamethoxazole, and valacyclovir were prophylactically administered during treatment. TCZ was administered monthly at a dose of 8 mg/kg (up to a maximum of 800 mg) for 6 months. Patients were routinely premedicated with antihistamines, steroids, and acetaminophen. Serum immunoglobulin G (IgG) levels were measured before initiating TCZ treatment and before each subsequent administration. Initially, patients with serum IgG levels of <600 mg/dL received an initial dose of 1.0 g/kg IVIG. For subsequent TCZ administration, if the serum IgG levels remained below 600 mg/dL, a 0.5 g/kg IVIG was administered. IVIG was administered the day after TCZ administration. Following treatment, patients were closely monitored every 2 weeks for serum creatinine levels, eGFR, proteinuria, and adverse events (AEs). CMV quantitative polymerase chain reaction (PCR) and BKV real-time PCR assays were performed every 3 months. TCZ was discontinued in cases of allograft failure or serious AEs. Allograft failure was defined as a return to dialysis dependence or subsequent retransplantation.
Immunosuppression and donor-specific anti-human leukocyte antigen antibody monitoring
Induction immunosuppressive drugs, either antithymocyte globulin (ATG) or basiliximab (anti-IL-2R monoclonal antibody), were selected based on immunological risk, medical condition, risk of infection, and morbidity. Maintenance immunosuppressive drugs consist of tacrolimus as the main drug for all patients, combined with mycophenolic acid and prednisolone as a triple regimen unless adverse reactions occur. During TCZ treatment, the maintenance immunosuppressive regimen remained unchanged. All patients underwent DSA monitoring immediately before starting treatment and at 3 and 6 months posttreatment. DSA was identified using LABScreen single-antigen human leukocyte antigen (HLA) class I-combi and class II-group 1 kits (One Lambda, Inc.) [21]. The MFI values of the detected DSA were quantified using a positive criterion set at an MFI level >1,000. Anti-HLA antibodies other than DSA were also assessed, and MFI levels were measured to evaluate non-donor-specific anti-HLA responses. Non-HLA antibodies were not evaluated as DSA in this study.
Clinical outcomes
The primary outcome was allograft survival. Secondary outcomes included changes in eGFR, proteinuria levels, MFI of immunodominant DSA and non-DSA anti-HLA antibodies, and the occurrence of AEs. Allograft failure was death-censored as no patient death occurred during the study period. Immunodominant DSA or non-DSA anti-HLA antibodies were defined as those with the highest MFI detected in the patient serum samples. Serum creatinine levels were measured 3 and 6 months before treatment, at treatment initiation, and 3 and 6 months posttreatment. The eGFR was calculated using the MDRD (Modification of Diet in Renal Disease) study equation [22]. Patients who experienced allograft failure were excluded from the eGFR analysis. The delta eGFR (ΔeGFR) was calculated to represent the rate of decline in eGFR based on the differences observed at 6 months before treatment, at treatment initiation, and at 6 months posttreatment. The amount of proteinuria was assessed using the spot urine protein/creatinine ratio (UPCR) in grams per gram (g/g). UPCR was measured at treatment initiation and at 3 and 6 months posttreatment. Infection-free survival was estimated based on the duration from treatment initiation to the occurrence of an infectious event that necessitated hospitalization. AEs were defined as any unfavorable medical occurrence in a patient, regardless of their relationship to TCZ. AEs included immediate side effects such as fever, chills, or rash, as well as complications such as infections, leukopenia, and hypogammaglobulinemia. Leukopenia was defined as a white blood cell (WBC) count of less than 4,000/μL. Serious AEs included events resulting in death, life-threatening conditions, hospitalization, and significant disability.
Statistical analysis
Continuous variables are presented as mean ± standard deviation or median with interquartile range (IQR), depending on their distribution. Categorical variables are presented as frequencies and percentages. The Student t test or Mann-Whitney U test was used to compare continuous variables, and the chi-square test or Fisher exact test was used to compare categorical variables contingent on data distribution. Survival curves were generated using the Kaplan-Meier method, and group differences were assessed using the log-rank test. Cox proportional hazards model analysis was performed to investigate predictors of graft failure. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the ability of eGFR at treatment initiation to predict allograft failure. The area under the curve (AUC) was calculated, and the optimal cutoff value was determined using the Youden index [23] to maximize sensitivity and specificity. Wilcoxon signed-rank test was used to compare the median MFI values of immunodominant DSA and anti-HLA antibodies before and after TCZ treatment. The paired t test was used to compare ΔeGFR before and after treatment. A linear mixed model was used to compare changes in repeated measurements of eGFR and proteinuria over time between the groups. All statistical analyses were performed using R software (version 4.4.2; R Foundation for Statistical Computing). Data visualization was performed using GraphPad Prism (version 10.0.0; GraphPad Software, Inc.). Statistical significance was defined as a two-sided p-value of <0.05.
Ethics statement
This study was conducted in adherence to the Declarations of Helsinki and Istanbul and received approval from the Institutional Review Board of The Catholic University of Korea, Seoul St. Mary’s Hospital (No. KC24RASI0459). The requirement for informed consent was waived because of the retrospective study design and the use of noninvasive procedures.
Results
Baseline characteristics
Baseline demographic, clinical, and histological characteristics of the patients are presented in Table 1. The mean patient age was 54.8 ± 9.9 years, and 10 (55.6%) were male. Three patients (16.7%) had previously undergone treatment with rituximab, plasmapheresis, and IVIG for cABMR before receiving TCZ, whereas 15 patients (83.3%) received TCZ as first-line treatment for cABMR. None of the patients experienced a delayed graft function. Most patients (17/18, 94.4%) underwent living donor KT, of whom two were ABO-incompatible. The mean total ischemic time was 63.6 ± 27.1 minutes. Patients were scheduled to receive six TCZ doses; however, treatment was discontinued in four patients (three for allograft failure and one for leukopenia). The median number of doses administered was 6.0 (IQR, 5.3–6.0).
The mean number of HLA mismatches was 3.7 ± 1.4. Anti-HLA antibodies were detected in most patients (16/18, 88.9%); however, six patients (33.3%) were confirmed to have DSA at the time of cABMR diagnosis, all of which were de novo DSA. Additionally, in five patients, anti-HLA antibodies were suspected to be de novo DSA. However, these could not be confirmed because the corresponding donor HLA type was not identified (three anti-HLA-DQ antibodies, one anti-HLA-B antibody, and one anti-HLA-Cw antibody). These antibodies were considered possible DSA. Among the 12 patients without DSA at biopsy, nine had historical and/or possible DSA, and three were C4d-positive, supporting the diagnosis of cABMR (details in Supplementary Table 1, available online). Ten patients (55.6%) received basiliximab as induction therapy, four (22.2%) received ATG, and three (16.7%) did not undergo any induction regimen. One patient underwent KT at an external center, and the induction regimen was unavailable. All patients were on tacrolimus as the main drug for maintenance immunosuppression at the initiation of TCZ treatment, with a mean tacrolimus trough level of 7.2 ± 3.0 ng/mL. The TCZ dose administered was a median of 6.0 (IQR, 5.3–6.0). At the initiation of TCZ treatment, the mean eGFR was 29.3 ± 14.6 mL/min/1.73 m2, and the UPCR was 1.9 ± 2.2. The mean time from KT to the start of TCZ treatment was 13.2 ± 6.6 years, and the mean time from cABMR diagnosis to treatment was 15.4 ± 27.4 weeks. The mean follow-up period was 18.2 ± 10.7 months.
The allograft biopsy specimens exhibited severe microvascular inflammation with high glomerulitis (g) plus peritubular capillaritis (ptc) scores (median, 5.0; IQR, 4.0–5.0). In addition, biopsy results indicated advanced tissue chronicity with a high chronic glomerulopathy (cg) score (median, 2.0; IQR, 1.0–2.8) and interstitial fibrosis and tubular atrophy (IFTA) score (median, 2.0; IQR, 2.0–2.0). The C4d score was positive in eight patients and negative in 10 patients. None of the patients was accompanied by concurrent T cell-mediated rejection.
Allograft survival
Allograft failure occurred in eight patients (44.4%) during the follow-up period (Fig. 2A), with a median period of 7.9 months (IQR, 5.4–14.1 months). The cause of allograft failure was cABMR in all patients. In the univariate Cox proportional hazards model (Table 2), low eGFR at treatment initiation was the only significant predictor of allograft failure (hazard ratio, 0.854; 95% confidence interval [CI], 0.761–0.959). In the ROC curve analysis using eGFR at treatment initiation (Fig. 2B), the AUC was 0.900 (95% CI, 0.740–1.000; p < 0.05). The optimal cutoff value for predicting allograft failure was determined to be an eGFR of 25.5 mL/min/1.73 m2, with a sensitivity of 87.5% and a specificity of 90.0%.
Changes in mean fluorescence intensity levels of anti-human leukocyte antigen antibodies, estimated glomerular filtration rate, and proteinuria
The MFI of immunodominant DSA and non-DSA anti-HLA antibodies before and after TCZ treatment is presented in Fig. 3A. All six patients with identified DSA showed a significant reduction in MFI, with the median value decreasing from 9,203 to 6,228 (p = 0.03). Among the 10 patients positive for non-DSA anti-HLA antibodies, nine showed a reduction in MFI, including two with negative conversion. Overall, the median MFI significantly decreased, from 4,804 to 2,808 (p = 0.006). Fig. 3B illustrates the changes in eGFR before and after TCZ treatment. The mean eGFR decreased from 40.8 ± 16.2 mL/min/1.73 m2 at 6 months pretreatment to 29.3 ± 14.6 mL/min/1.73 m2 at treatment initiation and further to 26.5 ± 12.0 mL/min/1.73 m2 at 6 months posttreatment. Fig. 3C shows the changes in ΔeGFR, which improved significantly, from –11.5 ± 10.7 mL/min/1.73 m2 during the 6 months pretreatment to –2.9 ± 6.6 mL/min/1.73 m2 during the 6 months posttreatment (p = 0.03). Changes in the amount of proteinuria are shown in Fig. 3D. UPCR decreased from 1.9 ± 2.2 g/g at treatment initiation to 1.3 ± 1.4 g/g at 3 months posttreatment and remained stable at 6 months posttreatment (1.3 ± 1.4 g/g), although the differences were not statistically significant (p = 0.50).
Adverse events
No immediate side effects were observed. Five patients experienced serious infectious AEs requiring hospitalization (two cases of bacterial pneumonia, two of coronavirus disease 2019 [COVID-19], and one of colitis). All patients recovered from the infections with appropriate treatment, and TCZ administration was not discontinued. Leukopenia occurred in two patients, one of whom did not recover from severe leukopenia (WBC count, 1,540/μL), resulting in the discontinuation of TCZ after three doses. These patients were receiving enteric-coated mycophenolate sodium (1,440 mg/day and 720 mg/day, respectively) with the patient requiring TCZ discontinuation being on 1,440 mg/day. Hypogammaglobulinemia (IgG <600 mg/dL) developed in six patients, all of whom continued TCZ treatment while administering IVIG according to our protocol.
Subgroup analysis based on baseline estimated glomerular filtration rate
Subgroups were defined using 25.5 mL/min/1.73 m2, the optimal cutoff for predicting allograft failure identified by ROC analysis (Fig. 2B). Baseline characteristics are presented in Table 3. The eGFR at TCZ initiation was significantly lower in the reduced eGFR group compared to the corresponding results in the preserved eGFR group (17.2 ± 5.6 mL/min/1.73 m2 vs. 38.9 ± 12.2 mL/min/1.73 m2, p < 0.001). The reduced eGFR group had fewer HLA mismatches, although the prevalence of anti-HLA antibody did not differ between the groups. Four patients in the reduced eGFR group discontinued treatment due to allograft failure or leukopenia before completing the six scheduled doses; in contrast, all patients in the preserved eGFR group completed the protocol. No other baseline characteristics differed between groups.
Subgroup outcomes are shown in Supplementary Fig. 1 (available online). Allograft survival was significantly lower in the reduced eGFR group than in the preserved eGFR group (p < 0.001) (Supplementary Fig. 1A, available online). Allograft failure occurred in seven of eight patients (87.5%) in the reduced eGFR group (mean, 8.3 ± 6.1 months after treatment). In contrast, only one of 10 (10.0%) in the preserved eGFR group experienced allograft failure occurred at 37.8 months. In both the reduced and preserved eGFR groups, DSA MFI tended to decline (from 24,903 to 13,882 and from 7,941 to 6,147, respectively), although the changes were not statistically significant (Supplementary Fig. 1B, available online). Similarly, non-DSA anti-HLA antibody MFI tended to decline (from 2,997 to 2,800 in the reduced eGFR group and from 6,611 to 5,220 in the preserved eGFR group), without statistical significance (Supplementary Fig. 1B, available online). ΔeGFR improved in both groups during the 6 months posttreatment compared with 6 months pretreatment, with a greater tendency in the preserved eGFR group than in the reduced eGFR group, although the difference was not statistically significant (p = 0.07 and p = 0.31, respectively) (Supplementary Fig. 1C, available online). In the preserved eGFR group, UPCR decreased from 1.6 ± 2.1 g/g at treatment initiation to 1.3 ± 1.6 g/g at 3 months and 1.0 ± 1.0 g/g at 6 months. In the reduced eGFR group, UPCR decreased from 2.4 g/g at treatment initiation to 1.3 ± 1.2 g/g at 3 months but subsequently rebounded to 2.0 ± 1.9 g/g at 6 months. Overall, UPCR changes differed significantly between groups (p = 0.007) (Supplementary Fig. 1D, available online). Infectious AEs occurred in two of eight patients (25.0%) in the reduced eGFR group and three of 10 (30.0%) in the preserved eGFR group (p > 0.99). Leukopenia was observed only in the reduced eGFR group (n = 2).
Discussion
IL-6 is a pivotal cytokine that mediates inflammatory and immunomodulatory pathways and serves as a regulator of T-cell/B-cell development and function [9,10]. As classical approaches, such as plasmapheresis, IVIG, and corticosteroids, as well as novel immunotherapies, such as rituximab, bortezomib, and eculizumab, have failed to demonstrate effectiveness in treating cABMR in KT recipients, targeting IL-6/IL-6R signaling has emerged as a promising therapeutic strategy [24]. In our previous report [25], we demonstrated elevated circulating IL-17 and IL-6 levels, along with a higher proportion of T-helper 17 (Th17) cells, in KT recipients with chronic allograft dysfunction than in those with stable allograft function. These results suggest that the IL-6/Th17 pathway may be a potential therapeutic target. Therefore, we evaluated the effectiveness and safety of TCZ treatment in KT recipients diagnosed with cABMR.
First, we analyzed allograft survival following TCZ treatment. The allograft survival rate at 18 months posttreatment was 55.6%, which was significantly lower than that reported in previous studies of TCZ [16,26,27] and comparable to that reported for conventional therapies, such as steroids, IVIG, and rituximab [28]. In a large single-center study, 76% of patients with cABMR lost their grafts; the median survival was 1.9 years, with graft survival at 18 months being approximately 55% [28]. This poor outcome could be attributed to several factors: 1) a long time duration between transplantation and the start of treatment (mean, 13.2 years); 2) patients in our study exhibited advanced transplant glomerulopathy (median cg score of 2.0); 3) a low eGFR at the time of treatment initiation (mean, 29.3 mL/min/1.73 m2). Collectively, these factors suggest delayed treatment in our cohort, potentially allowing persistent alloimmune activity to lead to irreversible injury.
In this study, a low baseline eGFR was identified as the only significant predictor of allograft failure in the Cox proportional hazards model (hazard ratio, 0.854; 95% CI, 0.791–0.959). In the ROC analysis, the baseline eGFR demonstrated a high AUC value of 0.900 for predicting allograft failure. When a cutoff value of 25.5 mL/min/1.73 m2 was applied, it showed high sensitivity (87.5%) and specificity (90.0%). This is not surprising because a low eGFR indicates that the allograft kidney has already suffered a significant loss of functioning nephrons and fibrotic changes due to cABMR, resulting in a diminished reserve capacity. This aligns with previous studies that identified a low eGFR (or high serum creatinine level) as a risk factor for allograft failure in KT recipients with cABMR [28–30].
Second, we analyzed the impact of TCZ on DSA and non-DSA anti-HLA antibodies. Previous studies have shown that TCZ reduces alloantibody levels not only during desensitization in highly sensitized patients [14], but also when treating cABMR [16,27]. Our study demonstrated consistent findings, showing an overall reduction in the strength of immunodominant DSA and non-DSA anti-HLA antibodies. This reflects the mechanism by which TCZ blocks the IL-6R, inhibits B-cell differentiation into plasma cells, and reduces alloantibody production.
Third, we examined changes in allograft function and proteinuria following TCZ treatment. Although the mean eGFR continued to decline after treatment, the rate of decline significantly slowed, as reflected by an improvement in ΔeGFR from the 6 months pretreatment to the 6 months posttreatment (p = 0.03). Proteinuria showed a trend toward reduction, with UPCR decreasing at 3 months and remaining stable at 6 months, although the changes did not reach statistical significance. These findings suggest that TCZ may contribute to stabilizing allograft function in patients with cABMR.
Fourth, we evaluated the occurrence of AEs. The frequency of AEs in our study was lower than previous reports [16,26]. We thoroughly screened for infectious diseases with the potential for reactivation before TCZ treatment and implemented chemoprophylaxis for fungal, Pneumocystis Jirovecii, and CMV infections. We monitored the patients every 2 weeks after initiating treatment and assessed infectious and noninfectious AEs. IgG levels were measured before every TCZ administration, and patients were administered IVIG when IgG levels decreased to <600 mg/dL. This rigorous monitoring, combined with early diagnosis and prompt management of AEs, allowed us to continue TCZ treatment without interruption, except in only one case of severe leukopenia. There were no cases of allograft failure or death related to AEs.
Subgroup analyses based on baseline eGFR provided further insights into treatment response. Patients with preserved eGFR showed markedly better allograft survival than did those with reduced eGFR, among whom most grafts failed within 1 year despite TCZ treatment. In the reduced eGFR group, early allograft failure resulted in treatment discontinuation in three patients before completion of the six scheduled doses. ΔeGFR trajectories revealed a more pronounced tendency toward stabilization in the preserved eGFR group compared with the reduced eGFR group, although not statistically significant. Changes in proteinuria significantly differed between groups (p = 0.007), with sustained improvement observed only in the preserved eGFR group. In terms of safety, the frequency of infectious AEs did not differ between groups, and leukopenia occurred only in the reduced eGFR group. This finding suggests a potential vulnerability to hematologic side effects in patients with advanced allograft dysfunction. Although caution is warranted in interpreting these results, due to the small sample size, we can conclude that a baseline eGFR of 25.5 mL/min/1.73 m2 may serve as a practical reference when considering the initiation of TCZ treatment in patients with cABMR.
Failing allograft causes significant distress to both patients and transplant physicians, with cABMR being the leading cause of late allograft failure [3,4]. However, determining whether and how to treat cABMR remains challenging because of insufficient evidence to support the effectiveness of available treatments [5]. TCZ is a high-cost drug, and without clear guidelines on its dosage and duration, its cost-effectiveness requires careful consideration, as observed in rheumatoid arthritis and COVID-19 [31,32]. Administration of TCZ in KT recipients with reduced eGFR not only limits therapeutic effectiveness [16,28–30] but may also increase unnecessary financial burdens and the risk of AEs [17]. Thus, TCZ treatment should focus on patients with preserved allograft function, and a specific eGFR threshold should be established. Delayed cABMR diagnosis leads to a decline in allograft function and progression of histological changes, which can reduce the feasibility of TCZ treatment. Therefore, proactive DSA monitoring and early biopsy are essential for early cABMR diagnosis and allograft function preservation.
This study has several limitations. First, it was a retrospective single-center study with a small sample size, which may have limited the generalizability of the results, and the absence of a control group precluded direct comparison with other therapeutic approaches. Second, only eight allograft failure events occurred, precluding multivariable Cox regression because the number of events per variable was less than 10, the widely recommended minimum for reliable modeling [33]. Therefore, we presented only univariate analyses, which may reduce statistical power and limit adjustment for potential confounders. Third, the lower number of TCZ doses compared with previous studies may have insufficiently suppressed the ongoing inflammatory process in cABMR. In addition, heterogeneity in treatment initiation timing and background immunosuppressive regimens may have influenced the outcomes and should be considered potential confounders. Fourth, the absence of posttreatment allograft biopsies limits the assessment of histological changes. Fifth, as patients were included based on their consent for off-label use of TCZ, selection bias cannot be excluded. Therefore, more extensive, multicenter prospective studies are warranted to determine the effectiveness and safety of TCZ in KT recipients with cABMR.
In conclusion, TCZ treatment reduces alloantibody levels in most KT recipients with cABMR. However, baseline eGFR levels significantly influenced clinical outcomes such as allograft survival, attenuation of eGFR decline, and reduction in proteinuria. Given the high cost of TCZ and its potential AEs, including infections, it should be selectively used in patients with preserved allograft function.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement table 1
Print
