The effect of rituximab dose on infectious complications in ABO-incompatible kidney transplantation

Abstract

Background

Rituximab (RIT) improves the outcomes of ABO-incompatible (ABOi) kidney transplantation (KT), but it has been associated with infectious complications. The aim of this study was to investigate infectious complications according to the dose of RIT in ABOi KT.

Methods

We analyzed 213 recipients [118 ABO-compatible (ABOc) KT and 95 ABOi KT] who underwent living donor KT between 2010 and 2014. ABOi KT patients were categorized by RIT dose: standard RIT (375 mg/m², n = 76) versus reduced RIT (200 mg, n = 19). All patients received basiliximab and maintained on triple immunosuppression consisting of tacrolimus, prednisone and mycophenolate mofetil. Infectious complications and post-transplant outcomes were analyzed for 1 year following KT.

Results

The rates of overall infectious complications among the three groups were comparable (22.9% in ABOc KT, 38.2% in standard RIT and 26.3% in reduced RIT, P = 0.069). In the standard RIT group, hepatitis B virus reactivation occurred in three recipients (3.9%) with hepatitis B surface antigen[−]/anti-hepatitis B core antibody[+]. Three cases (3.9%) of Pneumocystis jirovecii pneumonia occurred in the standard RIT group. Serious infections developed in 13 of the ABOc KT (11.0%), 20 from the standard RIT group (26.3%) and 2 from the reduced RIT group (10.5%, P = 0.015). Standard-dose RIT was found to be an independent risk factor for serious infections [hazard ratio: 2.59 (95% confidence interval: 1.33–5.07), P = 0.005]. There were no significant differences in rejection, renal function, graft survival and patient survival between standard and reduced RIT groups.

Conclusions

Standard RIT increased the risk of serious infection when compared with reduced-dose RIT. Reduced-dose RIT might be sufficient for ABOi KT without increasing the risk of serious infection.

INTRODUCTION

Recent advancements in desensitization protocols and potent immunosuppression have resulted in excellent long-term outcomes of ABO-incompatible (ABOi) kidney transplantation (KT) [1]. Current desensitization strategies are composed of antibody removal, B-cell depletion and/or intravenous immunoglobulin (IVIG) [2].

Rituximab (RIT), a chimeric monoclonal antibody specific for CD20 expressed on B cells, has become an important component of desensitization strategies in the current era. B-cell depletion by RIT has facilitated the expansion of ABOi KT by obviating the need for splenectomy [3, 4]. We previously reported excellent short-term outcomes of ABOi KT with the standard dose of RIT [5]. However, the optimal dose of RIT needed for desensitization in ABOi KT has not been established. Furthermore, long-term complications of RIT in immunocompromised patients remain unknown, and as a result, there is a trend toward reducing RIT dosage to avoid adverse events [6]. Recently, a Japanese group reported favorable long-term outcomes of ABOi KT with a reduced dose of RIT (a fixed dose of 200 mg) [7]. Considering the similar body habitus of Asian populations, a fixed dose regimen might not be significantly different from body surface area-based regimens. Hence, we adopted a fixed dose of RIT (200 mg) in ABOi KT recipients with a baseline anti-A/B immunoglobulin G (IgG) titer of ≤1 : 128.

Controversial results have been published regarding the impact of RIT on infectious complications [8–14]; however, few studies have determined the effect of RIT dose on infectious complications [7]. Accordingly, we attempted to identify the effect of RIT dose on the development of infectious complications after ABOi KT.

MATERIALS AND METHODS

Patients

This retrospective analysis was conducted on all consecutive living donor KT between June 2010 and July 2014 at Severance Hospital. We included in this study KT recipients who were negative for both complement-dependent cytotoxicity crossmatch with donor T and B cells and flow cytometry crossmatch with donor T and B cells. ABOi KT patients were categorized to either the standard RIT group (375 mg/m²) or the reduced RIT group (200 mg). ABO-compatible (ABOc) KT patients not administered RIT were assigned to the control group (Figure 1). Four hundred and fifty-four living donor KTs were performed during the study period, and 213 recipients that met the inclusion/exclusion criteria were included in the analysis. Ninety-five ABOi KT patients underwent desensitization with RIT and plasmapheresis, and of these 95 patients, 76 received standard-dose RIT and 19 received reduced-dose RIT.

Isoagglutinin titer and donor-specific antibody measurement

Isoagglutinin (anti-A and/or anti-B) titers were determined by testing 2-fold serial dilutions of the serum with commercially available A/B indicator red cells using a previously described method [5]. The presence of donor-specific antibodies was determined using single antigen beads assay (Lifecodes LSA class I and class II; Gen-Probe Transplant Diagnostics, Inc., Stamford, CT, USA). Donor-specific antibodies were analyzed against donor human leukocyte antigen (HLA)-A, -B, -DR and -DQ antigens. Results are expressed as mean fluorescence intensities (MFI). A normalized value of >1000 MFI was considered positive for donor-specific antibodies.

Desensitization protocol and immunosuppression

All ABOi KTs were performed according to the desensitization protocol shown in Figure 2. RIT was administered at a single dose (375 mg/m² or 200 mg) within 7 days before transplantation in ABOi KT patients. RIT dosage was reduced to 200 mg when the baseline antidonor isoagglutinin (anti-A and/or anti-B) IgG titer was ≤1 : 128 from August 2013. Plasmapheresis was performed before KT to remove isoagglutinin antibodies, until IgG titer decreased to a level of 1 : 16. IVIG (100 mg/kg) was administered after each session of plasmapheresis in seven consecutive cases. Afterward, we did not use IVIG in cases of baseline IgG titer of ≤1 : 128. All patients received basiliximab 20 mg on Days 0 and 4. The immunosuppressive protocol consisted of tacrolimus, methylprednisolone and mycophenolate mofetil (MMF). Initial tacrolimus was administered orally at 0.1 mg/kg twice daily. Subsequent doses were adjusted to maintain a target trough concentration between 5 and 8 ng/mL. The initial dose of methylprednisolone (500–1000 mg) was tapered to oral prednisolone (5–10 mg/day). The initial dose of MMF was 1.5 g/day, and this dose was adjusted to minimize adverse events, such as gastrointestinal trouble or leukopenia.

Diagnosis of infections

Infections were determined based on identification of infectious organisms. Patients with any of the following were excluded: fever of undetermined origin, potential infection (where antibiotics were given, but no infectious etiology was identified), asymptomatic candiduria and upper respiratory infection (presumed viral). Urinary tract infection (UTI) was defined as a quantitative bacterial count of ≥10⁵ colony-forming units/mL for an appropriately collected urine specimen in the presence of symptoms or signs of urinary infection. BK virus (BKV) viremia was defined as a blood BKV DNA level of >10 000 copies/mL. Cytomegalovirus (CMV) viremia was defined as one blood sample with a CMV DNA level of >1000 copies/mL. CMV disease was classified as CMV syndrome and tissue-invasive CMV disease according to recent guidelines [15]. Tissue-invasive CMV disease was confirmed by immunohistochemical staining.

Hepatitis B virus (HBV) carrier was defined as seropositivity for hepatitis B surface antigen (HBsAg) at the time of transplantation. In HBsAg[+] patients, HBV reactivation was defined as a >10-fold increase of HBV DNA, compared with previous nadir. In HBsAg[−] but hepatitis B core antibody (anti-HBc)[+] patients, HBV reactivation was defined as HBsAg reverse seroconversion after transplantation.

We defined serious infections as any (bacterial, viral or fungal) infection requiring hospitalization, or prolongation of hospitalization. Tissue-invasive CMV disease was viewed as a serious infection.

Prophylaxis protocol and infection monitoring

All patients received trimethoprim–sulfamethoxazole as Pneumocystis jirovecii pneumonia (PCP) prophylaxis (minimum 6 months). Fungal prophylaxis consisted of 4 mL of oral nystatin four times daily for 12 months. No prophylaxis against CMV infection was administered to any patient. Instead, a preemptive approach was used in high-risk patients (donor-seropositive and recipient-seronegative). HBsAg[+] patients were treated with anti-HBV nucleot(s)ide analogs regardless of HBV DNA.

BKV surveillance was undertaken at 3, 6, 9 and 12 months post-transplant by blood quantitative polymerase chain reaction (PCR) according to recent practice guidelines [16]. Blood CMV DNA was monitored by quantitative PCR in cases where infection was suspected. Post-transplant serological follow-ups, which included HBsAg and hepatitis B surface antibody (anti-HBs), were performed at 12 months in all patients.

Diagnosis and management of acute rejection

Renal biopsies were performed in cases of acute allograft dysfunction (creatinine increase or proteinuria). All acute rejections were biopsy proven and classified using the Banff Criteria 2011. All biopsy specimens were stained for C4d. Patients with acute cellular rejection (ACR) were treated by methylprednisolone pulse therapy (500 mg/day, three to four times). Steroid-resistant ACR patients received anti-thymocyte globulin (ATG). Antibody-mediated rejection (AMR) was treated with combination of plasmapheresis and IVIG.

Renal function and lymphocyte subpopulation

Renal function was assessed using serum creatinine and glomerular filtration rates (using the Modification of Diet in Renal Disease formula). Peripheral CD19+ cell counts were measured by flow cytometry to evaluate B-cell depletion before RIT infusion and during the postoperative period.

Primary endpoints

The primary study endpoints were the incidences of bacterial, viral, fungal and serious infections up to 12 months post-KT.

Secondary endpoints

The secondary study endpoints were the incidences of biopsy-proven acute rejection (BPAR), graft function, lymphocyte subpopulation, and patient and graft survival during the first year following transplantation.

Statistical analyses

Demographic information is summarized using frequency counts (in percentage), or as means ± standard deviations, depending on data type. The χ² test with Fisher's exact test was used to compare categorical variables, and one-way analysis of variance was used to compare continuous variables. Univariate analysis and multivariate logistic regression analyses were performed to determine risk factors of the development of serious infections. Freedom from events and survival were estimated by Kaplan–Meier product limit method and were compared using the log-rank test. The analysis was performed using SPSS software (version 20.0; SPSS Inc., Chicago, IL, USA), and P-values <0.05 were considered significant.

Ethics statement

The study was approved by the institutional review board of Severance Hospital, Yonsei University Health System.

RESULTS

Baseline characteristics

Baseline characteristics are summarized in Table 1. There were no significant differences in age, gender and dialysis. Viral hepatitis and CMV serostatus were comparable among the groups. Only 2.3% of patients were CMV donor-seropositive and recipient-seronegative. There was a significant difference in mean number of HLA mismatches, which was significantly higher in the reduced RIT group.

Infectious complications

There were 27 patients (22.9%) in the control group, 29 patients (38.2%) in the standard RIT group and 5 patients (26.3%) in the reduced RIT group diagnosed with an infection during the 12-month follow-up period (Table 2). Rates of bacterial, viral and fungal infections were comparable in the three groups.

Viral infections were the most commonly observed infectious complications, occurring in 23.9% of patients. CMV was the most common viral infection in both standard and reduced RIT groups. Although not statistically significant, there was a tendency toward higher incidence of CMV disease in ABOi KT than in ABOc KT. Two patients developed tissue-invasive CMV disease in the standard RIT group and one patient in the control group. Two cases of CMV disease occurred in five high-risk patients (donor-seropositive and recipient-seronegative). BKV was the most common viral infection in the control group. However, only one patient in the control group and one in the reduced RIT group developed BKV nephropathy. All patients with pretransplant HBsAg were treated with anti-HBV nucleot(s)ide analogs (nine entecavir, two lamivudine and two telbivudine). No virologic deterioration was observed in HBsAg[+] patients. On the other hand, HBV reactivation occurred in three patients (HBsAg[−]/anti-HBc[+]) from the standard RIT group during the 12 months. Entecavir was initiated for the patients upon HBV reactivation. However, one patient died from hepatic failure (Day 203) despite active antiviral treatment.

Bacterial infection rates were similar in the standard RIT, reduced RIT and control groups (P = 0.388). UTI was the most common, accounting for 65.4% of all bacterial infections.

Six patients in the standard RIT group developed a fungal infection: three PCP, two fungal pneumonia [diagnosed by antigen assay (Aspergillus Galactomannan and/or Candida) and computed tomography] and one fungal UTI. All episodes of PCP occurred after discontinuation of trimethoprim–sulfamethoxazole prophylaxis. One death at 32 days post-KT was related to fungal pneumonia.

A significant intergroup difference was observed in the rate of serious infections, and the standard RIT group had a significantly higher rate of serious infection than the reduced RIT or control group. Incidence of serious infections with regard to antirejection treatments is shown in Figure 3. A tendency toward a higher incidence of serious infection was observed in patients with ATG, but this tendency did not reach statistical significance by multivariate analysis (Table 3).

Risk factor analysis for serious infection

The induction agent and maintenance immunosuppressive regimens were similar in the standard RIT, reduced RIT and control groups. Mean serum tacrolimus concentrations during the entire study period were similar in the three groups (P > 0.05, Figure 4). Independent factors associated with serious infection were assessed by multivariate analysis (Table 3). Standard-dose RIT (375 mg/m²) was found to have a 2.6-fold higher risk of serious infection within a year of transplantation (P = 0.005). Preoperative plasmapheresis (five or more sessions) did not increase the risk of serious infections.

BPAR and graft function

BPAR occurred in 18 of 118 ABOc KT patients (15.3%), compared with 23 of 95 ABOi KT patients (24.2%). In ABOi KT, BPAR rates in the standard and reduced RIT groups were similar (23.7% versus 26.3%, respectively; P = 0.773). Of the patients with ACR (28 ACR only and 9 ACR + AMR), 23 (62.2%) were treated with steroid pulse (methylprednisolone 500 mg/day, three to four times) alone and 14 (37.8%) were treated with ATG following steroid pulse. No significant intergroup difference was found for any type of BPAR or for serum creatinine and glomerular filtration rates (Table 4).

Changes of lymphocyte subpopulation

Changes of lymphocyte subpopulation according to RIT dose are shown in Figure 5. Reduced RIT depleted CD19+ cells successfully, and the impact of RIT on lymphocyte subpopulation lasted for 12 months regardless of RIT dose. The CD19+ cell counts were not significantly different in the standard and reduced RIT groups at any time point.

Graft and patient survival of ABOi KT according to RIT dose

Overall graft and patient survival were comparable in the standard RIT and reduced RIT groups (Figure 6). Graft survival rates were 94.7% in the standard RIT group and 100% in the reduced RIT group, respectively, at 12 months (P = 0.312). Patient survival rates in the standard RIT and reduced RIT groups were 97.4 and 100%, respectively (P = 0.477). Two patients in the standard RIT group died during the 12 months post-KT: one due to fungal pneumonia and the other due to hepatic failure by hepatitis B reactivation.

DISCUSSION

This study showed that standard-dose RIT was associated with serious infectious complications during the first year after transplantation. From the multivariate analysis, we found that standard-dose RIT had a 2.6-fold higher risk of serious infections.

Reported prevalences of overall infectious complications among KT recipients treated with RIT vary widely from 45.5 to 63.8% [8–13]. Differences in reported prevalences are probably due to a lack of uniform diagnostic criteria, use of different prophylaxis regimens and different follow-up durations. In the present study, infection rates were lower than those previously reported (38.2% in the standard RIT group and 26.3% in the reduced RIT group). We believe that this difference was possibly due to the relatively low blood trough levels of tacrolimus [17, 18], basiliximab induction [19] and short study period.

CMV infection remains one of the most common infections after KT. Studies concerning the impact of RIT have reported a higher incidence of CMV disease [10, 11]. In addition, a recent study described a correlation of ABOi desensitization and CMV-specific cellular immunity [14]. In line with previous studies, there was a tendency toward higher incidence of CMV disease in ABOi KT than in ABOc KT, although not statistically significant. However, there was no significant difference in CMV infection according to RIT dose in the present study, which concurs with a Japanese study in which no significant difference in CMV infection rates were found between a standard RIT group and a reduced RIT group [7]. Unfortunately, the majority of studies conducted on the topic, including the present study, are limited by heterogeneities in terms of donor/recipient CMV serostatus, immunosuppression and prophylaxis. Both prophylactic and preemptive approaches were effective in CMV high-risk patients. However, CMV prophylaxis is associated with high cost, late CMV infection and hematologic toxicity. In addition, even in the high-risk group, some patients received unnecessary treatment. In contrast, preemptive strategies have shown similar efficacies with lower incidences of late CMV infection and hematologic toxicity [20, 21]. Therefore, we decided to use a preemptive approach in high-risk patients. During the study period, two of five high-risk patients experienced CMV disease.

We adopted blood PCR for routine BKV screening due to its higher positive predictive value than BKV viruria [16]. RIT did not increase the incidence of BKV infection in the present study, which is consistent with previous reports [9, 12]. In contrast, recent studies suggested a higher risk for BKV nephropathy in ABOi KT [22, 23], and Barbosa et al. [24] reported a higher incidence of BK viremia in patients desensitized with RIT and IVIG. Nonetheless, the results of prior studies should be interpreted with caution as they included heterogeneous desensitization protocols. Furthermore, the risk of BKV infection seems to be largely related to the total burden of immunosuppression, rather than a specific agent [25]. In particular, high blood trough concentration of tacrolimus (>8 ng/mL) and ATG induction are known risk factors of BKV infection [17]. Therefore, relatively low tacrolimus blood trough concentration and no ATG induction might have contributed to the low incidence of BKV infection encountered in this study.

In September 2013, the US Food and Drug Administration issued a new boxed warning for RIT in terms of hepatitis B reactivation [26]. However, there is a lack of information regarding HBV reactivation in organ transplantation patients receiving RIT [8, 27]. In the present study, pretransplant HBsAg[+] patients received anti-HBV nucleot(s)ide, whereas HBsAg[−]/anti-HBc[+] patients did not receive prophylactic antiviral therapy. Given the high seroprevalence of anti-HBc in our cohort (41.7% in ABOc group and 40% in the ABOi group), HBV reactivation was an important consideration. HBV reactivation occurred in 10% of HBsAg[−]/anti-HBc[+] patients from the standard RIT group (3/30), and this might have been due to RIT-related HBV reactivation after KT [28]. Accordingly, prophylactic therapy might be considered for HBsAg[−]/anti-HBc[+] as well as HBsAg[+] KT patients scheduled to receive RIT for desensitization.

With respect to fungal infections, the standard RIT group showed higher incidence of fungal infections (7.9%) than the other groups, although this was not statistically significant. In particular, all three cases (3.9%) of PCP occurred in the standard RIT group. Host defenses against PCP have largely been attributed to CD4+ lymphocytes [29]. However, animal studies have indicated that B cells also play a vital role in host defense against PCP [30]. Recent studies have also suggested that RIT may be associated with the development of PCP [31–33]. Consequently, prolonged PCP prophylaxis should be considered for KT recipients receiving RIT [9].

The present study confirmed that standard-dose RIT (375 mg/m²) is an independent risk factor for serious infections. Two recent studies of ABOi KT demonstrated similar results to ours [11, 14]. RIT might contribute to serious infections in solid organ transplant recipients who simultaneously receive concomitant T-cell immunosuppressive agents [34, 35]. Despite the significant advancement made in preventive strategies, serious infections after KT remain a substantial cause of mortality and contribute to decreased graft survival [36, 37]. Interestingly, the reduced RIT group showed comparable results to the control group in terms of serious infection. Therefore, more caution is needed in standard RIT use for ABOi KT.

The risk of infection after transplantation is known to be significantly affected by the net state of immunosuppression [38]. Determining the specific effect of RIT on the risk of infection is because of coexisting confounding factors, such as combination and strength of immunosuppression. Actually, discordant results have been published regarding the impact of RIT [8–14], and these discordances might be associated with different exposures to immunosuppressants and different RIT doses [10, 12]. To minimize these confounding factors as possible, we included patients with highly uniform immunosuppressive regimens, except for the dose of RIT. In fact, mean serum tacrolimus concentrations at all time points were similar irrespective of RIT dose. In addition, we evaluated the exposure to antirejection agents and found no difference.

Another important concern regarding reduced-dose RIT is efficacy. A single dose of RIT at 375 mg/m² is generally accepted as being adequate in a renal transplant setting; little evidence exists to indicate whether this is the optimal dosage for renal transplant recipients [3, 4, 39, 40]. Toki et al. [6] reported that reduced doses of RIT (10–375 mg/m²) potently depleted B cells in spleen and peripheral blood. In addition, a Japanese group reported excellent results with reduced-dose RIT in ABOi KT [7, 41]. In the present study, target antibody titers (≤1 : 16 at the time of transplantation) were achieved successfully in all patients in the reduced RIT group. In addition, no differences were observed in terms of BPAR incidence, graft function, lymphocyte subpopulation, and graft and patient survival between the two RIT doses.

The present study has some limitations that warrant consideration. First, it was performed at a single institution and was retrospective in design. However, this made it possible to maintain homogeneity in the desensitization protocol and immunosuppressive regimen. Second, as the desensitization protocol was modified recently, the number of patients in the reduced RIT group was relatively small. Third, we could not ascertain CMV viremia prevalence in low- to intermediate-risk patients without any symptoms. We monitored CMV PCR in case of suspicion of viral infection, such as fever or neutropenia.

In conclusion, our results suggest that the risk of serious infection is substantially related to the dose of RIT administered. Furthermore, the available evidence and our findings indicate that RIT at a reduced dosage might be sufficient for ABO blood type incompatible desensitization without increasing the risk of serious infection. Appropriate prophylaxis against hepatitis B reactivation and PCP might be considered in KT patients receiving RIT.

ACKNOWLEDGEMENT

This work was supported by a faculty research grant of Yonsei University College of Medicine for 2013 (6-2013-0042).

CONFLICT OF INTEREST STATEMENT

None declared.

REFERENCES