Chronic Antibody-Mediated Rejection Is Reduced by Targeting B-Cell Immunity During an Introductory Period


Kazunari Tanabe,


Transplantation across blood group antigen and human leukocyte antigen (HLA) barriers are immunologically high risk. Both splenectomy and rituximab injection were developed to overcome those immunological barriers. The idea behind these treatments is to control B-cell immunity before and after renal transplantation and antibody production. Between January 2001 and December 2004, recipients underwent pretransplant double-filtration plasmapheresis (DFPP) and splenectomy at the time of transplantation in the ABO-incompatible group (ABO-I-SPX; n= 45). From January 2005 to June 2009, a low dose of rituximab was given as an alternative to splenectomy (ABO-I-RIT; n = 57). As a control group, we selected 83 cases of ABO-C living-donor kidney transplantation between January 2001 and December 2007 (ABO-C). We compared the graft survival rate and chronic antibody-mediated rejection (C-AMR) rate between ABO-C and ABO-I kidney transplantation with induction treatment. C-AMR rates 2 years after the operation were 8.8, 3.5 and 28.9%, and de novo donor-specific anti-HLA antibody (DSHA) positive rates were 2.2, 1.7 and 18.1% in the ABO-I-SPX, ABO-I-RIT and ABO-C groups, respectively. The ABO-C group showed the highest rate of C-AMR and de novo DSHA. B-cell depletion protocols, such as splenectomy or rituximab administration, reduced C-AMR after kidney transplantation.

ABO-C, ABO-compatible; ABO-I, ABO-incompatible; A-AMR, acute antibody-mediated rejection; C-AMR, chronic antibody-mediated rejection; CDC, complement-dependent cytotoxicity; DFPP, double filtration plasmapheresis; DSHA, donor-specific anti-HLA antibody; FCXM, flow cytometry cross-match analysis; MP, methylprednisolone; HLA, human leukocyte antigen; NDSHA, nondonor-specific anti-HLA antibody; RIT, rituximab; RTx, renal transplantation; SPX, splenectomy.



Although it is commonly believed that chronic renal allograft failure results from the cumulative effects of multiple injurious factors, anti-HLA antibodies are thought to have the most adverse effect on graft survival and may lead to C-AMR (1). Evidence suggests that the humoral response to alloantigens plays an important role in the development of C-AMR. The prognosis of C-AMR is unfavorable because there is no established treatment protocol (2).

Antiblood group antibodies act as another barrier against transplantation. Historically, ABO-I was considered an absolute contraindication to kidney transplantation. Such transplants, when performed initially, usually resulted in rapid and nearly uniform loss of the allograft due to A-AMR (3). We previously reported that the long-term outcome of ABO-I with splenectomy was almost equivalent to that of standard ABO-C kidney transplantation and a lack of correlation between anti-A/B antibody titers (4).

In 2005, our institution introduced a new preconditioning regimen, consisting of DFPP and low-dose rituximab injection with basic potent immunosuppressive agents (5).

Under these conditions, we observed a significantly lower incidence of C-AMR in ABO-I renal transplantation. These findings led us to retrospectively analyze long-term outcomes of ABO-I, compared with ABO-C, in terms of the incidence of C-AMR, graft survival and de novo DSHA.

Study design and patients

Patient characteristics are shown in Table 1. All study procedures were approved by the Ethics Committee of Tokyo Women's Medical University. There were an unequal number of samples at various time points because the analysis was retrospective. Clinical and laboratory data were extracted from electronic databases and patient medical records.

Table 1.  Patient characteristics
 ABO-I -SPX (n = 45)ABO-I -RIT (n = 57)ABO-C (n = 83)p-Value*p-Value**
  1. *ABO-I-SPX versus ABO-C, **ABO-I-RIT versus ABO-C.

  2. 1Data are expressed as means ± standard deviation.

  3. 2 FGS, focal glomerular sclerosis.

  4. 3MPGN, membranoproliferative glomerulonephritis.

  5. 4Only one recipient died of cardiovascular disease 7 years after transplantation.

Recipient age (mean ±SD)140.4 ±12.344.0 ±14.840.4 ±12.6  
Recipient gender (M / F)25 / 2042 / 1559 / 24  
Donor age (mean ±SD)57.2 ±10.757.3 ± 9.0 56.8 ±10.4  
Donor source   N.S
 Parents (father/mother)8/176/2314 / 38  
 Siblings 6 410  
Duration of hemodialysis (months)32 (13–69)26 (19–57)44 (16–86)  
DSHA N(%) 15 (33.3%)18 (31.6%)24 (28.9%)  
Underlying disease     
 Chronic glomerulonephritis151324  
 Reflux nephropathy 6 3 8  
 Rapidly progressive glomerulonephritis 0 0 2  
 Cystic kidney 1 5 2N.S
 Renal hypoplasia 0 0 3  
 FGS2 1 3 5  
 MPGN3 0 1 2  
 Rupus nephritis 0 1 0  
 IgA nephropathy111317  
 Diabetic nephropathy 4 5 3  
 Toxemia of pregnancy 0 0 2  
 Others 71313  
Cause of graft loss     
 Graft failure4 (8.9%)1 (1.7%)9 (10.8%)0.7270.040
 Death with functioning graft1 (1.2%)40.4600.406
 Acute rejection2 (4.4%)0.053
 Chronic rejection2 (4.4%)1 (1.7%)8 (10.2%)0.2960.062
Most common adverse events     
 Herpes zoster01 (1.8%)1 (1.2%)0.4600.788
 Cytomegalovirus infection 6 (13.3%)2 (3.5%) 9 (10.8%)0.6760.113
 Hepatopathy1 (2.2%)1 (1.8%)4 (4.8%)0.4690.337
 Anemia4 (8.9%) 7 (12.3%)3 (3.6%)0.2100.050
 Leucopenia1 (2.2%)14 (24.6%)3 (3.6%)0.6660.0002

From January 2001 to December 2004, ABO-I recipients underwent pretransplant DFPP and splenectomy at the time of operation (ABO-I-SPX). Between January 2005 and June 2009, splenectomy was not performed in ABO-I recipients, and a protocol that included a pretransplantation injection of a low dose of rituximab (200 mg/person) was used (ABO-I-RIT).

As a control group, we performed ABO-C living-donor kidney transplantation by the standard protocol between January 2001 and December 2007.

All recipients had negative cross-matches as assessed by the T-cell CDC assay as well as T-cell FCXM assay performed before the transplantation. As shown in Table 1, there were no significant differences in the prevalence of DSHA positivity among the three groups prior to transplantation.

The patients selected for this study underwent the first renal transplant protocol biopsy within 6 months of transplantation and the second biopsy during the second year after the operation. Patients were hospitalized for 2 days and 1 night for the renal graft biopsy and serum samples were collected during this hospitalization. The frozen serum samples used in this study were kept frozen at −80° until thawing for the analysis. Patients who refused to provide consent for collection of the serum samples or had not undergone a protocol biopsy were excluded from this analysis, and finally, 45 cases enrolled in the ABO-I-SPX group, 57 in the ABO-I-RIT group and 83 in the ABO-C group. However, there were no significant differences in laboratory data, pathological findings and clinical course between selected patients and excluded patients in this study.

Desensitization protocol and immunosuppression

At 1 week prior to transplantation, all patients received a triple-drug immunosuppressive regimen that consisted of FK, MMF and MP as described previously (5). The anti-CD25 monoclonal antibody (basiliximab) was also given at the dose of 20 mg on days 0 and 4 after transplantation as a desensitization protocol. Patients in the ABO-I-SPX and ABO-I-RIT groups underwent three or four sessions of DFPP before surgery (Figure 1). The patients’ antiblood group antibody titers (i.e. anti-A immunoglobulin G (IgG)/IgM titers and/or anti-B IgG/IgM titers) were reduced to 1:32 or lower.

Figure 1.

Immunosuppressive protocols in ABO-I-SPX or ABO-I-RIT. MP = methylprednisolone; FK = tacrolimus; MMF = mycophenolate mofetil; DFPP = double filtration plasmapheresis. DFPP was performed three or four times.

Pathological findings

All rejection episodes were confirmed by the graft biopsy. Renal biopsies were performed as reported previously (6). Written informed consent was obtained from all patients for the biopsy. Protocol biopsies were performed twice: the first biopsy was at 6 months of the transplant and the second biopsy was during the second year after the operation. Episode biopsies were performed for clinical indications, such as when serum creatinine levels increased by 0.3 mg/dL above baseline or the patient had symptoms (e.g. oliguria or fever).

We rediagnosed the pathology specimen based on the newest Banff classification (7) and the data of antiblood group antibodies and anti-HLA antibodies for this study.

FACS analysis of the T-cell and B-cell populations in the peripheral blood

We used CD3 as the T-cell marker and CD19 as the B-cell marker in the peripheral blood. Because rituximab was used, CD19 was examined as a B-cell surface marker. A fluorescein isothiocyanate monoclonal antibody against CD3 (Becton Dickinson, San Jose, CA, USA), or a phycoerythrin-conjugated monoclonal antibody against CD19 (Becton Dickinson) were added to whole blood, followed by incubation at 4°C for 30 min. FACS data were analyzed by the Simultest software (Becton Dickinson).

Detection of antiblood group antibodies and anti-HLA antibodies

Antiblood type antibodies were measured as described previously (4,5). All patients were evaluated for DSHA, determined by CDC assay, T-cell FCXM assay and a solid-phase technique (LABScreen 100 Luminex, System; One Lambda Inc., Canoga, Park, CA, USA). The assay was performed according to the manufacturer's protocol (6). The serum was reacted with 10 mM of dithiothreitol at 37°C for 10 min to eliminate IgM.

Adverse events

CMV infection and disease were defined according to previously published criteria (8). Anemia was defined as hemoglobin <13.5 g/dL in adult males, <12.0 g/dL in adult females and <5th percentile in children. Leukopenia was defined as grade 2 (2000 < N < 3000/mm2) or more severe (N < 2000/mm2) according to KDIGO clinical practice guidelines for the care of kidney transplant recipients (9).

Statistical analysis

All data were analyzed using the SAS software (ver. 9.1; SAS Institute, Cary, NC, USA). Data were expressed as mean and standard deviations (SD). Means of normally distributed data were compared using Student's t-tests, whereas a χ2-test was used to compare proportions. Cumulative probabilities of graft and patient survival curves were estimated using the Kaplan–Meier method; differences among the curves were examined using a log-rank test.

In all analyses, p < 0.05 (two-tailed) was taken to indicate statistical significance.


Patient and graft survival

Overall patient survival was 100% in all groups until the 6th year postsurgery. Only one recipient in the ABO-C group died, of cardiovascular disease 7 years after transplantation. The 5-year graft survival rates were 91.1, 98.1 and 90.3% in the ABO-I-SPX, ABO-I-RIT and ABO-C groups, respectively (Figure 2).

Figure 2.

Graft survival rate (Kaplan–Meier analysis). Years after transplantation.

Allograft function

Median serum creatinine levels at 6 months postsurgery were 1.31, 1.26 and 1.42, mg/dL, while those at 2 years postsurgery were 1.13, 1.11 and 1.30 mg/dL in the ABO-I-SPX, ABO-I-RIT and ABO-C groups, respectively. There was no significant difference in graft function among the three groups.

Rejection episodes

The number and type of rejection episodes are summarized in Table 2. The incidence rates of AMR within 6 months after transplantation were 13.3, 3.5 and 10.8% in the ABO-I-SPX, ABO-I-RIT and ABO-C groups, respectively. The ABO-I-RIT group showed the lowest AMR rate within 6 months; however, the differences among the groups were not statistically significant.

Table 2.  Rejection episodes
 ABO-I -SPX (n = 45)ABO-I -RIT (n = 57)ABO-C (n = 83)p-Value*p-Value**
  1. *ABO-I-SPX versus ABO-C, **ABO-I-RIT versus ABO-C.

  2. IF/TA = interstitial fibrosis/tubular atrophy; AMR = antibody mediated rejection; BC = borderline change; ACR = acute cellular rejection; C-AMR = chronic AMR; AVR = acute vascular rejection; IgA = IgA nephropathy.

6 months after RTx
 AMR 6 (13.3)2 (3.5) 9 (10.8)0.6750.113
 C-AMR01 (1.8)00.226
 AVR 6 (13.3)1 (1.8) 9 (10.8)0.6750.040
 ACR2 (4.4)05 (6.0)0.7070.059
 C-TMR003 (3.6)0.1970.147
 BC2 (4.4)2 (3.5)14 (16.9)0.0420.015
 IF/TA001 (1.2)0.4600.406
 Others01 (1.8)1 (1.2)0.4600.788
 No rejection30 (66.7)42 (73.7)40 (48.2)0.0450.003
 Unknown 8 (14.0)1 (1.2)0.4600.002
The second year after RTx
 AMR002 (2.4)0.2940.298
 C-AMR4 (8.8)2 (3.5)24 (28.9)0.009 0.0001
 AVR1 (2.2)000.173
 ACR001 (1.2)0.4600.406
 C-TMR01 (1.7)2 (2.4)0.2940.793
 BC 5 (11.1)4 (7.0)1 (1.2)0.0110.069
 IF/TA3 (6.7)2 (3.5)7 (8.4)0.7220.243
 Others3 (6.7)1 (1.7)1 (1.2)0.0900.788
 No rejection26 (57.8)34 (59.6)36 (43.4)0.1200.059
 Unknown3 (6.7)13 (22.8)11 (13.2)0.2540.141

The incidence rates of C-AMR 2 years after transplantation were 8.8, 3.5 and 22.9% in the ABO-I-SPX, ABO-I-RIT and ABO-C groups, respectively. The ABO-I-SPX and ABO-I-RIT groups showed significantly lower incidences of C-AMR than the ABO-C group (Table 4).

Table 4.  Relevance to C-AMR of antiblood group antibody and DSHA
 ABO-I -SPX (n = 45)ABO-I -RIT (n = 57)ABO-C (n = 83)p-Value*p-Value**
  1. *ABO-I-SPX versus ABO-C, **ABO-I-RIT versus ABO-C.

  2. Anti-A/B antibody titers (IgM and IgG) were evaluated postoperatively after 6 months.

Preoperative antiblood group antibody titers decreased to ≤1:3245 (100%)57 (100%)
Postoperative antiblood group antibody titers increased to >1:6401 (2.2%)
C-AMR with postoperative antiblood group antibody titers increased to >1:6400
Preoperative DSHA positive15 (33.3%)18 (31.6%)24 (28.9%)0.6040.736
Class I/class II/both9/3/39/7/212/9/3  
Postoperative DSHA positive including de novo DSHA (the second year after RTx)2 (4.4%)2 (3.5%)29 (34.9%)0.0001<0.0001
Class I/class II/both0/2/00/2/014/12/3  
De novo DSHA (the second year after RTx)1 (2.2%)1 (1.7%)15 (18.1%)0.0100.029
Class I/class II/both0/1/00/1/06/8/1  
C-AMR with postoperative DSHA positive including de novo DSHA2 (4.4%)2 (3.5%)19 (22.9%)0.0070.002
C-AMR with de novo DSHA008 (9.6%)0.0320.016

Peripheral blood lymphocyte fraction

We compared the lymphocyte subpopulations in the datasets obtained at the three time-points; preoperative, 6 months postoperative and 2 years postoperative (Table 3). The rate of CD19 among all lymphocyte cells was significantly lower in the ABO-I RIT group than in the ABO-I SPX group and in the ABO-C group. The imbalance of the T-cell and B-cell fractions was maintained for at least 2 years postoperatively.

Table 3.  T-cell and B-cell population in the peripheral blood
  ABO-I -SPXABO-I -RITABO-Cp-Value*p-Value**
  1. *ABO-I-SPX versus ABO-C, **ABO-I-RIT versus ABO-C.

  2. Median: %, (range) CD3 is a T-cell marker and CD19 is a B-cell marker.

CD3Before RTx72.9 (55–91)74.9 (49–95)73.1 (52–91)0.8410.709
 6 months after RTx77.6 (57–87)87.9 (64–97)81.2 (65–93)0.5500.187
 The second year after RTx68.8 (40–83)79.8 (58–90)78.1 (48–91)0.1670.582
CD19Before RTx11.6 (4–21)10.8 (3–28)12.3 (4–25)0.9220.301
 6 months after RTx19.3 (7–31)1.26 (0–7)15.2 (4–28)0.0470.001
 The second year after RTx14.5 (6–24)2.35 (0–12)11.6 (4–18)0.0930.002

Antiblood group antibodies and DSHA

We evaluated anti-A/B antibody titers (IgM and IgG). One recipient in the ABO-I-RIT group showed a postoperative increase in antiblood group IgG titer to ≥1:64; however, this patient had excellent graft function without developing A-AMR or C-AMR (Table 4). Preoperative DSHA-positive rates were 33.3, 31.6 and 28.9% in the ABO-I-SPX, ABO-I-RIT, and ABO-C groups (Table 1), respectively. There was no significant difference among the three groups; however, the de novo DSHA-positive rates were 2.2, 1.7 and 18.1% in the ABO-I-SPX, ABO-I-RIT and ABO-C groups, respectively. Moreover, we did not observe any significant differences in titers between anti-HLA class I and class II antibodies, either preoperatively or postoperatively (de novo).


Most common adverse events are shown in Table 1. No patient in any group died because of infection. The incidence rates of leucopenia were 2.2, 24.6 and 3.6% in the ABO-I-SPX, ABO-I-RIT and ABO-C groups, respectively (p < 0.001). Administration of G-CSF facilitated recovery in all patients, and no patients developed serious complications.


ABO-C renal transplantation was thought to be associated with a low risk of rejection and its immunosuppression protocol was considered to be standard. However, the low rate of deceased donor organ transplantation is insufficient to meet the increased demand for organs. Thus, ABO-I is being performed. Until about 2000, short-term graft survival was significantly poorer in ABO-I than in ABO-C. In most patients, early graft loss was caused by A-AMR; thus, removal of antiblood group antibodies was recognized as an important factor.

Alexandre et al. (10) reported that splenectomy was a prerequisite for successful ABO-I renal transplantation. However, excellent short-term outcomes following ABO-I without splenectomy were reported in 2003 (11). Instead of splenectomy, rituximab was administered to suppress B-cell function. Recently, in the case of ABO-I, many studies have shown significant decreases in early graft loss due to AMR, and good graft survival with rituximab and standard immunosuppressive agents. Thus, the long-term outcome of ABO-I is almost equivalent to that of standard ABO-C kidney transplantation (5).

However, anti-HLA antibodies act as another difficult immunological barrier against transplantation. Due to the development of more sensitive assays, it is now clear that DSHA levels vary widely among patients and DSHA serum levels may be a major determinant in allograft injury (12). In this study, the preoperative DSHA positive rate was about 30% in all three groups. Lefaucheur et al. reported an HLA antibodies positive rate in patients waiting for a kidney transplant (13). In this report, 502 patients were registered on the waiting list for deceased-donor kidney transplantation at their institution. A total of 174 transplant candidates (34.7%) had antibodies against class I or class II HLA on at least one pretransplantation serum, which is comparable to our result of HLA antibodies positive rate in this study.

However, the ABO-C group showed the highest detection rate of de novo DSHA and the highest incidence of C-AMR. Furthermore, there was no influence of antiblood group antibodies after more than 6 months postsurgery. The presence of anti-HLA antibodies is regarded as the main cause of AMR in ABO-C kidney transplantation. Terasaki and Cai (1) suggested a strong causal link between anti-HLA antibodies and the development of C-AMR. Anti-HLA antibodies were found in the serum of 2278 (20.9%) kidney recipients evaluated 6 months after renal transplantation. By the 1-year follow-up, 6.6% of the recipients in whom anti-HLA antibodies were detected had lost their grafts, compared with a graft loss rate of 3.3% for those in whom no anti-HLA antibodies were detected (p = 0.0007) (14). Additionally, there are many reports showing that nonsensitized recipients begin to produce de novo DSHA after transplantation, leading to C-AMR and graft loss (15). As reported by us previously (16), and also in this study, the appearance of de novo DSHA was detected at 1226 ± 789 postoperative days on average, with 5115 ± 4980 intensity measured by Luminex assay (data; not shown). There were no significant differences in its appearance days and its strength between ABO-I SPX, ABO-I RIT and ABO-C groups. In addition, we need to discuss the significance of NDSHA in kidney transplantation, both from the qualitative and quantitative aspects. However we did not focus on the appearance of NDSHA in this study.

It is difficult to treat C-AMR because it is often diagnosed after irreversible damage to the graft occurs, and its prognosis is usually poor (2,3). Even if inflammation is transiently suppressed using a high-dose immunosuppressant, the risks of infectious disease, sclerosing vasculitis increase and extensive interstitial fibrosis remain in the graft. Fehr et al. reported four patients with C-AMR, who were treated with a combination of rituximab and intravenous immunoglobulin. In their report, in one patient an acute rejection episode occurred 12 months after the treatment, and another patient had severe, possibly rituximab-associated lung toxicity (17).

The ABO-I renal transplantation protocol includes perioperative treatment with splenectomy or rituximab, in addition to the standard ABO-C renal transplantation protocol; however, the regimen of maintenance immunosuppressive agents used was the same in all three groups. The onset of chronic antibody-mediated rejection was suppressed; thus, the effect of the perioperative treatment was prolonged.

The effects of changes in the absolute numbers and distribution of lymphocyte subsets on the immune response after splenectomy are unclear. Our data showed that the rate of T-cell fraction of splenectomized patients was the lowest in the three groups. The spleen is the largest single secondary lymphoid organ and is the most important component of the reticuloendothelial system. B cells, T cells and plasma cells are thought to mature in the white pulp of the spleen. Healthy individuals who have undergone post-traumatic splenectomy show long-term impairment of humoral and cellular immunity (18). Although there are significant numbers of B cells in other parts of the body, including the lymph nodes and bone marrow, most plasma cells and memory cells exist in the spleen. The major purpose of splenectomy is to remove antibody-producing plasma cells and/or memory cells that may cause AMR during the posttransplant period (10).

Rituximab probably eliminates memory B cells, which, once they are stimulated by antigens in the transplanted graft, produce large numbers of antibody-producing plasma cells in various sites, including the lymph nodes and spleen. B cells are not just plasma cell precursors. They also represent an important population of antigen-presenting cells that are particularly efficient in sensitized transplant recipients because they carry surface immunoglobulins as antigen-specific receptors, which are involved in the uptake and presentation of donor antigens to T cells. It is now recognized that B cells play a much wider role in transplant immunology than previously thought. This suggests that B-cell depletion during the induction period may play an important role in preventing humoral rejection.

In this analysis, in the lymphocyte fraction 2 years after surgery, the ratio of the CD19 positive lymphocytes was the lowest in a case using RIT. Genberg et al. (19) reported that a single dose of rituximab at 375 mg/m2 along with FK, MMF/azathioprine and MP in renal transplant recipients completely eliminated B cells from peripheral blood, and the B-cell population remained suppressed for several years. The most significant reason for the exceptionally long-term depletion of B cells after transplantation is probably the combination of rituximab with continuous oral administration of standard immunosuppressive agents.

Tydén et al. published a report of a randomized, double-blind, placebo-controlled, multicenter study in which the effect of administering a single dose of rituximab (375 mg/m2 within 24 h before revascularization) was compared with the use of placebo as induction therapy (20). They found a tendency toward fewer (11.6% vs. 17.6%) and milder rejections during the first 6 months in the rituximab group. But their report was of a short follow-up period and they used more high doses of rituximab than our protocol. Given the current tendency to reduce the intensity of immunosuppression, we propose that low-dose rituximab therapy is a valid option for desensitization.

Clatworthy et al. described their use of rituximab compared with daclizumab for induction therapy (21). In their report, five of six patients (83%) who received rituximab had an episode of biopsy-confirmed acute rejection in the first 3 months after transplantation, compared with one of seven patients (14%) in the daclizumab group. They explained that proinflammatory cytokine release associated with B-cell depletion would be the cause. However, they used a steroid free protocol. Although it is possible to reduce the dose of an immunosuppressant by using various immunosuppressants, it is difficult to reduce the number of kinds of an immunosuppressant.

In this study, there was no increase in the risk of infectious complications after ABO-I-RIT transplantation despite rituximab induction and repeated apheresis. These findings indicate that our pretransplant conditioning and post-transplant immunosuppressive regimen is not too strong and did not cause troublesome opportunistic infections. The incidence of leukopenia in the ABO-I-RIT group was significantly higher than in the other two groups; however, there were no serious complications and all patients who became leukopenic recovered promptly by treatment with G-CSF.

In conclusion, compared with ABO-C transplantation without desensitization, ABO-I in conjunction with desensitization, including rituximab or splenectomy, showed a significantly reduced incidence of C-AMR. Our protocol for ABO-I prevented the production of not only antiblood group antibodies, but also anti-HLA antibodies. The onset of C-AMR and infectious disease are suppressed by using immunosuppressants appropriately from an introductory period. Although, this is retrospective study, a prospective, randomized study of anti-CD20 for the prevention of de novo antibody formation and chronic AMR would be a valuable contribution.


The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.