A retrospective analysis was performed on data of kidney transplant recipients who were treated in the Erasmus MC, University Medical Center Rotterdam, with alemtuzumab (Campath®, Sanofi Genzyme, United States) because of AR between January 2012 and January 2018. The study was approved by the medical ethical review board of the Erasmus MC (number 2018-1430). The patients were identified by the electronic medication prescription system of our hospital pharmacy. Patients with blood group AB0-incompatible kidney transplantations were excluded from the analysis, because they receive alemtuzumab as induction therapy (13).

The outcomes were compared to those of a cohort of patients treated with rATG (Thymoglobulin®, Sanofi Genzyme, United States) for AR between January 2002 and January 2012. The characteristics and outcomes of this cohort were described in detail previously (10).

All AR episodes (including recurrent AR) were biopsy-proven and biopsies were re-evaluated according to the Banff 2015 (for rATG-treated patients) and Banff 2017 classification (for alemtuzumab-treated patients) by one dedicated renal-pathologist (M.C.C-v.G.) (14–16). The presence of donor-specific anti-HLA antibodies (DSA) and non-donor-specific HLA antibodies against HLA-A, HLA-B, HLA-DR, and HLA-DQ were examined in alemtuzumab-treated patients using the single-antigen bead Luminex assay on serum samples collected at the time of AR. DSA directed against Cw and DP HLA molecules were not tested. The presence of DSA was not routinely tested in the period 2002–2012 when rATG still was the therapy of choice (10). Therefore, the biopsies of the rATG-treated patients could not be reclassified according to the Banff 2017 criteria (14).

Of patients treated with alemtuzumab, patient survival, allograft function [estimated glomerular filtration rate (eGFR); Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) (17)], allograft survival (censored for death), variables that could influence allograft survival (patient and donor characteristics, type of immunosuppressive therapy, and type and grade of rejection), and adverse events were assessed. Baseline eGFR was defined as the highest eGFR in the 3 months prior to AR. Delayed graft function (DGF) was defined as the need for dialysis in the first week after transplantation. Allograft loss was defined as the need for dialysis or retransplantation. The follow-up period for allograft loss and infection was from the day of T cell-depleting therapy until death, retransplantation, or loss to follow-up. Malignancies and mortality were evaluated until the last follow-up visit, which could be after subsequent kidney transplantation. The Dutch national pathology archive PALGA (Pathologisch-Anatomisch Landelijk Geautomatiseerd Archief, https://www.palga.nl/) was used for collecting of data relating to the occurrence of malignancy. Infections were considered serious if the infection necessitated hospitalization or occurred during hospital admission for another reason.

The allograft- and patient survival data of patients who had received a kidney transplant in the same time periods in our center and were not treated with T cell-depleting therapy was also compared to the patients treated with T cell-depleting therapy for AR.

The standard immunosuppressive regimen included induction therapy with basiliximab (Simulect®, Novartis Pharma, Basel, Switzerland) 20 mg intravenously on days 0 and 4 after transplantation, followed by maintenance therapy with tacrolimus (Prograf®, Astellas Pharma, Leiden, the Netherlands), mycophenolate mofetil (MMF; Cellcept®, Roche Pharmaceuticals, Basel, Switzerland) and glucocorticoids.

Basiliximab became part of our standard immunosuppressive regimen in 2009. Dosing of tacrolimus and MMF was based on pre-dose concentrations (C₀). Target C₀ for tacrolimus was, respectively, 10–15 μg/L (weeks 1–2), 8–12 μg/L (weeks 3–4), 5–10 μg/L (weeks 5–12), and 4–8 μg/L from month 4 onwards. MMF was started at 1,000 mg twice daily and subsequent dosing was based on C₀ (target C₀ was 1.5–3.0 mg/L). Glucocorticoids were given as an intravenous dose of 100 mg on days 0–3 and followed by a dose of 20 mg/day (days 4–14). Thereafter, glucocorticoids were tapered off and completely withdrawn around month 4.

The first-line treatment of aTCMR was methylprednisolone 1,000 mg (Solu-Medrol®, Pfizer, New York, the United States) intravenously daily for 3 consecutive days, followed by a second-line treatment with alemtuzumab or rATG in case of a glucocorticoid-resistant, recurrent or severe aTCMR (Banff grade IIA or higher). rATG was administered as a single bolus [4 mg/kg (actual bodyweight, no maximum dose limit)] intravenously (10). Alemtuzumab was administered subcutaneously (18). The first 14 patients were treated with alemtuzumab (30 mg) daily for 2 consecutive days. Since T cell-depletion already occurred after one dose of alemtuzumab, the next patients received a single dose (30 mg). To prevent infusion-related side effects patients were premedicated with glucocorticoids (50 mg intravenously), acetaminophen (4 times daily 1,000 mg), and clemastine (4 mg intravenously). The alemtuzumab-treated patients were discharged the same day if no severe side-effects were noted. T- and B cell counts were measured with BD FACSCanto™ software every 3 months until the T cell count was >200 × 10⁶/L. In the patients treated with rATG, a CD3⁺ T cell count <200 × 10⁶/L was aimed for a duration of 2 weeks during which patients were hospitalized (10). If CD3⁺ T cell counts increased during this period, a repeat dose of rATG was administered. All patients received prophylaxis for Pneumocystis jirovecii (sulfamethoxazole/trimethoprim) and cytomegalovirus (CMV; valganciclovir, except for CMV seronegative recipients with CMV seronegative donors) until the T cell count was >200 × 10⁶/L. Patients with aABMR or mixed type AR could additionally be treated with intravenous immunoglobulins (IVIg), plasma-exchange, or both according to the Kidney Disease: Improving Global Outcomes (KDIGO) guideline (9).

Categorical variables are presented as number (percentage). Continuous variables are presented as mean with standard deviation for normally distributed variables or median with interquartile range (IQR) for non-normally distributed variables. For differences between unpaired non-normally distributed continuous data or unpaired categorical data, the Kruskal-Wallis and Mann-Whitney U-tests, and the Chi-squared and Fisher's exact tests were used. Kaplan-Meier survival analysis was used to examine subgroups (e.g., age categories and rejection types) within the alemtuzumab group and to compare allograft- and patient survival between alemtuzumab-treated patients and patients transplanted in the same period and who were not treated with alemtuzumab.

The influence of predictor variables on allograft survival in alemtuzumab-treated patients was analyzed with multivariable Cox proportional hazard regression analysis. Due to the number of events (41 allograft losses), the number of variables that could be included per analysis was limited. The influence of the most significant variable was tested in the presence of all the other variables one by one, and the non-significant variables were eliminated from the model by backward elimination.

Propensity scores were used to control for baseline differences between the patients treated with rATG and alemtuzumab (19). They were acquired by performing a logistic regression with therapy type as the outcome variable. Covariates included in the logistic model were: age of the patient at time of AR, gender, primary kidney disease, donor type (living/deceased), induction therapy (43% of patients treated with rATG received induction therapy, vs. 97.3% of alemtuzumab treated patients), maintenance therapy, time to AR, and type of AR. The resulting propensity score was used as a covariate in Cox proportional hazards regression models (for calculation of the patient-, allograft-, and infection-free survival), in linear regression models (for continuous outcomes), and in logistic regression models (for categorical outcomes).

A two-sided p < 0.05 was considered statistically significant. GraphPad Prism, version 5 (San Diego, CA, USA), SPSS version 21 (SPSS Inc., Chicago, IL, USA), and R (R Foundation for Statistical Computing, Vienna, Austria, version 3.5.1) were used for the statistical analysis.

Between January 2012 and January 2018, 1,214 patients received a kidney transplant at our center. Of these, 113 patients (9.3%) were treated with alemtuzumab for AR. Three patients were treated with alemtuzumab twice because of two separate rejection episodes of the same kidney transplant. Between January 2002 and January 2012, 1,107 patients were transplanted with a kidney and 108 patients of these (9.8%) were treated with rATG for AR (10). The median cumulative dose of rATG per patient was 7.4 mg/kg. Baseline characteristics of patients treated with either alemtuzumab or rATG for AR are presented in Table 1.

Induction therapy with basiliximab was given to 303 (27.4%) of all patients transplanted between 2002 and 2012 (the rATG period) and to 1,065 (87.8%) of all patients between 2012 and 2018 (the alemtuzumab period). As a result, significantly more patients treated with alemtuzumab (93.8%) had previously received basiliximab induction therapy compared to rATG-treated patients [29.2%; p < 0.0001 (Table 2)]. A tacrolimus- and MMF-based maintenance therapy was given to 81% of alemtuzumab-treated patients and to 72.2% of rATG-treated patients (p = 0.08; Table 2). First line therapy for AR was methylprednisolone in 94.8% of alemtuzumab-treated patients, and to 86.1% or rATG-treated patients (Table 2).

Sixty-four alemtuzumab-treated patients (55.2%) and 64 (59.3%) rATG-treated patients had an early AR (within 3 months after transplantation; Table 3). The distribution of the Banff grade of AR was not different between the patients treated with alemtuzumab or rATG (p = 0.89; Table 3). In 18 patients (15.5%), a second kidney allograft biopsy was performed after the initial treatment with methylprednisolone and immediately before alemtuzumab treatment to confirm ongoing AR (Table S1).

Patient survival of patients treated with either alemtuzumab or rATG for AR is depicted in Figure 1A. Compared with the historical rATG cohort, the patient survival of the alemtuzumab group was not different [hazard ratio (HR) 1.14, 95%-confidence interval (CI) 0.48–2.69, p = 0.77; Figure 1A), also when only those patients who were treated with basiliximab induction therapy were included (HR 1.74, 95%-CI 0.68–4.46, p = 0.25; Figure 2A).

The patient survival of alemtuzumab-treated patients was significantly lower compared with the patients transplanted in the same period and who were not treated with alemtuzumab (HR 2.38, 95%-CI 1.25–4.54, p = 0.0036, Figure S1A). In the total follow-up period [median 2.8 years (IQR 1.3–3.8 years)], 18 patients died after a median of 1.45 years (IQR 0.92–2.93; Table S2). A univariable Cox proportional hazard regression analysis was performed to investigate which variables influenced the risk of death in patients treated with alemtuzumab. The only variable that was associated with the risk of death was age of the recipient (HR per year 1.09, 95%-CI 1.04–1.14, p < 0.0001; Table S3). This increased risk of death was seen in alemtuzumab-treated patients older than 50 years at the time of transplantation (Figure S2).

A comparison between the patient survival of rATG-treated and patients transplanted in the same period and who were not treated with rATG is shown in Figure S1C and was described previously (10).

Death-censored kidney allograft survival of alemtuzumab-treated patients was not different compared to that of patients who received rATG for AR (HR 0.82, 95%-CI 0.45–1.50, p = 0.52; Figure 1B). A similar survival was also observed when only those patients who were treated with basiliximab induction therapy were included (HR 1.10, 95%-CI 0.57–2.10, p = 0.78; Figure 2B). Additional information about the kidney allograft function after alemtuzumab or rATG therapy for AR is provided in Figure S3.

The allograft survival of alemtuzumab-treated patients was significantly worse compared to the allograft survival of patients that were transplanted in the same period and who were not treated with alemtuzumab (HR 258.0, 95%-CI 112.0–591.3, p < 0.0001; Figure S1B). During the follow-up (median 2.2 years, IQR 1–3.5), 41 (35.3%) patients lost their kidney allograft after alemtuzumab therapy for AR, of which six never had a functioning graft [primary non-function (PNF)]. To investigate which variables influenced allograft survival in alemtuzumab-treated patients, a Cox proportional hazard regression analysis was performed. In the univariable model, age of the recipient, number of HLA mismatches, glucocorticoid maintenance treatment, timing of AR, and the Δ eGFR (percentage change between baseline eGFR and eGFR at the moment of AR) significantly influenced the risk for death-censored allograft loss (p < 0.05, Table S4) in alemtuzumab-treated patients. The variables glucocorticoid use and timing of rejection were related because all patients with an early acute rejection used glucocorticoids as maintenance therapy, while only 56.6% of patients with a late acute rejection used glucocorticoids (p < 0.0001). The Banff grade of rejection did not influence allograft survival (p = 0.19). Allograft survival of alemtuzumab-treated patients suffering from either aTCMR or aABMR is shown in Figure S4.

The final multivariable model showed that patients with actual panel reactive antibodies (PRA) >6%, and patients with a Δ eGFR of more than 50% had an inferior allograft survival after alemtuzumab therapy (Figure 3). Patients using glucocorticoids at time of AR, and patients with more HLA mismatches, showed a superior allograft survival after alemtuzumab therapy (Figure 3). Several variables were compared between patients with an HLA mismatch of 0–3 and an HLA mismatch of 4–6 (Table S5). One variable was significantly different: 42 (72%) recipients with 4–6 HLA mismatches received a living unrelated donor kidney (Table S5), while 13 (23%) recipients with 0–3 HLA received a living unrelated donor kidney (p < 0.001; Table S5).

The allograft survival of patients treated with survival of rATG was worse compared to that of patients who were not treated with rATG and transplanted in the same time period (HR 15.9, 95%-CI 9.2–27.4, p < 0.0001; Figure S1D). A multivariable Cox proportional hazard regression analysis was performed for patients treated with rATG for AR and reported previously (10). This analysis demonstrated that allograft survival was superior in patients with an early AR compared with a late AR (10).

Infusion-related side effects also occurred less frequently in alemtuzumab-treated patients compared with the patients treated with rATG (Table 4). No alemtuzumab-treated patients experienced serum sickness vs. five patients in the rATG group (p = 0.02). No patients experienced cytokine release syndrome or pulmonary edema after alemtuzumab. The median duration of hospitalization was 3 days (IQR 1–6) in patients treated with alemtuzumab and 15 days (IQR 13–19) in rATG-treated patients.

The infection-free survival (excluding CMV, EBV, and BK virus infections) in the first year after alemtuzumab treatment was significantly better compared with the infection-free survival of the rATG-treated patients (HR 0.41, 95%-CI 0.25–0.68, p < 0.002; Figure 4). CMV reactivation occurred in 25 patients (21.6%) treated with alemtuzumab (Table S6), compared to 27 patients (25%) in the rATG group (p = 0.10). In both the alemtuzumab- and rATG-treated groups, two patients experienced a primary CMV infection (p = 0.50). Additional information on the occurrence of infections is presented in Table S6.

Secondary autoimmune events have been described after administration of alemtuzumab (6). In the current study, two patients developed inflammatory polyneuropathy (one case of Guillain-Barre syndrome and one case of chronic inflammatory demyelinating polyneuropathy) after alemtuzumab treatment (20). No patients were diagnosed with autoimmune thyroid disorders, idiopathic thrombocytopenic purpura or autoimmune nephropathy.

Repopulation of T cells >200 × 10⁶/L occurred in 55.7% of alemtuzumab-treated patients in the first year after administration (Figures S5A,B). In 40.2% of the patients, repopulation of B cells >100 × 10⁶/L (Figures S5C,D) was seen at 1 year.

Solid tumors were diagnosed in seven alemtuzumab-treated patients during the total follow-up [median 2.8 years (IQR 1.3–3.8); Table S7]. The incidence of solid tumors was 2.3 per 100 person-years with a median time after alemtuzumab therapy of 28 months (IQR 9–38), and the age at the time of diagnosis was 65 years (IQR 60–76). Seven patients were diagnosed with skin cancer: 21 basal cell carcinomas and 12 squamous cell carcinomas. In the rATG treated patients, 14 malignancies were diagnosed after a mean time of 63 months (standard deviation 45 months) during the follow-up of 6.8 years (IQR 4.9–9.1; Table S8) (10).

DH has received grant support, lecture, and consulting fees from Astellas Pharma and Chiesi Pharmaceuticals, as well as a lecture fee from Hikma Pharma and grand support from Bristol Myers-Squibb. MC-V has received grant support from Astellas Pharma. MH has received grant support from Novartis and Shire and lecture fees from Astellas Pharma, Chiesi Pharmaceuticals, MSD, Sanofi/Genzyme, Shire, and Vifor Pharma. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.