Outcome after Desensitization in HLA or ABO-Incompatible Kidney Transplant Recipients: A Single Center Experience

Background The shortage of deceased donors led to an increase of living donor kidney (LDK) transplantations performed in the presence of donor-specific antibodies (DSA) or ABO incompatibility (ABOi) using various desensitization protocols. Methods We herein analyzed 26 ABOi and 8 Luminex positive DSA patients who were successfully desensitized by anti-CD20, antigen-specific immunoadsorption and/or plasmapheresis to receive an LDK transplant. Twenty LDK recipients with non-donor-specific HLA-antibodies (low risk) and 32 without anti-HLA antibodies (no risk) served as control groups. Results 1-year graft survival rate and renal function was similar in all 4 groups (creatinine: 1.63 ± 0.5 vs 1.78 ± 0.6 vs 1.64 ± 0.5 vs 1.6 ± 0.3 mg/dl in ABOi, DSA, low risk and no risk group). The incidence of acute T-cell mediated rejections did not differ between the 4 groups (15% vs 12, 5% vs 15% vs 22% in ABOi, DSA, low risk and no risk), while antibody-mediated rejections were only found in the DSA (25%) and ABOi (7.5%) groups. Incidence of BK nephropathy (BKVN) was significantly more frequent after desensitization as compared to controls (5/34 vs 0/52, p = 0.03). Conclusion We demonstrate favorable short-term allograft outcome in LDK transplant recipients after desensitization. However, the desensitization was associated with an increased risk of BKVN.


Methods
We herein analyzed 26 ABOi and 8 Luminex positive DSA patients who were successfully desensitized by anti-CD20, antigen-specific immunoadsorption and/or plasmapheresis to receive an LDK transplant. Twenty LDK recipients with non-donor-specific HLA-antibodies (low risk) and 32 without anti-HLA antibodies (no risk) served as control groups.

Conclusion
We demonstrate favorable short-term allograft outcome in LDK transplant recipients after desensitization. However, the desensitization was associated with an increased risk of BKVN.

Introduction
The considerable shortage of organ donors and increasing number of patients with end-stage renal disease has led to an extended waiting time for potential renal allograft recipients. In addition, approximately 30% of patients on the kidney waiting list are sensitized to HLA antigens of potential donors facing a significant increased waiting time [1]. The introduction of solid phase assays, such as the Luminex technology, which allows a more sensitive and specific identification of donor-specific anti-HLA antibodies (DSA), may impact the organ allocation process even further, in that even more patients are considered to be sensitized and rejected as recipients [2;3]. However, the impact of DSA detected only by solid phase assays on transplant outcome remains under debate [4]. While the identification of DSA in the presence of a negative complement depending cytotoxicity (CDC) in pre-transplant sera were linked with increased immunogenic graft loss in retrospective analysis [5;6], smaller prospective series questioned their relevance [7;8]. In particular, the degree of mean fluorescent intensity (MFI) that is of clinical relevance remains unclear, with large retrospective series showing no apparent difference in graft loss in recipients with DSA at an MFI < 3000 [9]. Therefore, several centers developed desensitization protocols to allow immunized patient to receive a kidney from a donor across the HLA barrier. Using different therapeutic strategies with new immunosuppressive medications, modern apheresis techniques and/or antigen-specific immunoadsorption, promising short-term outcomes have been reported [10][11][12]. Furthermore, ABO-incompatible (ABOi) living kidney donor (LDK) transplantation has become a popular alternative to expand the donor pool [13].
However data from those studies have been difficult to interpret or compare because of heterogenecity among immunologic testing techniques, DSA levels, desensitization strategies, and demographic and clinical characteristics of donor and recipient populations. We therefore aimed to reevaluate our own practice by assessing 1-year graft outcome after desensitization in renal transplant recipients who received a kidney from a living donor across blood group and/ or HLA barriers in comparison to two matched control groups with low immunological risk. We furthermore aimed to evaluate the impact of the intensified immunosuppressive regiment on the rate of BK virus infections.

Study design and patients
We conduct a retrospective cohort study with 91 adult patients (> 18 years) who received a LDK between January 2007 and June 2012 at our center. The study was approved by the ethics board of the Ludwig-Maximilians-University Munich. The patients gave their written consent. This consent procedure was also approved by the ethics committee of the Ludwig-Maximilians-University Munich (Approval number 513-12). The patients selected for the study had either Luminex-detected donor-specific antibodies with a MFI > 3000 or DSA with a MFI < 3000 but a positive CDC B-cell and/or Luminex crossmatch prior to transplantation (DSA group, n = 8) or had received a kidney from an ABO blood group incompatible donor (ABOi group; n = 26). The patients selected for the control groups were recipients, who received a living-donor kidney during the same time period and were maintained on the same long-term tripple immunosuppressive regiment and had Luminex-detected non-donor-specific antibodies (nDSA) (low risk group, n = 20) or no anti HLA-antibodies (no risk group, n = 32). All patients had a negative T-cell CDC crossmatch. However, within the DSA group 4/8 patients had a positive B-cell CDC crossmatch and 6/8 patients had a positive Luminex crossmatch before desensitization.

HLA-antibody screening and C1q assay
Recipients were screened for the presence of HLA-antibodies before transplantation and during the routine follow-up by means of Luminex (Life Screen Deluxe, Gen-Probe, USA). In positively screened patients antibody specificity in relation to the donor was confirmed by Single Antigen Beads (SAB) (LSA, Gen-Probe, USA). An MFI of > 3000 was used as cut-off. Recipients in the DSA group were screened for C1q binding HLA-antibodies by C1q-SAB assay using Luminex (LabScreen, One lambda, USA).

Anti-infective prophylaxis
All patients with a high-risk CMV constellation (D +/R -) received a prophylaxis with valganciclovir for 3 months post-transplantion. Furthermore, oral PcP prophylaxis (trimethoprim/sulfamethoxazole 160 mg/800 mg, 3 times a week) was administered 6 months post-transplant to every patient.

BKV
Systematic BKV screening using a polymerase-chain-reaction (PCR) assay was routinely performed at baseline (transplantation) and every three months thereafter. The cut off used to considerer the PCR as a positive viremia for BK virus was 1,000 copies/ml. In patients with diagnosed BK viremia and/or BK nephropathy (BKVN) immunosuppression was reduced, in that MMF was slowly tapered to half the dose and blood through levels of tacrolimus were reduced (goal 4-5 ng/ml). In the event of BKVN with concomitant acute rejection MMF was replaced by leflunomide. Leflunomide was administered with a loading dose of 100 mg for 3 days, followed by maintenance dose of 20 mg/d, adjusted for blood levels of 50-100 μg/ml.

Biopsies/Rejections
Allograft biopsies were taken when renal graft function was impaired or BK viremia developed. Rejection was determined according to the diagnostic criteria proposed at the 2007 Banff Conference [14]. C4d staining was routinely performed in paraffin sections of all biopsies. C4d positive staining in peritubular capillaries (PTC) was evaluated semi-quantitatively as follows: minimal (<10% of PTC), focal (11-50% of PTC) and diffuse (>50% of PTC) [15]. The diagnosis of BK nephropathy was based on histologic features in combination with positive staining for SV40.

Demographic and clinical information
The following parameters were evaluated:

Patient characteristics
The patient characteristics are shown in Table 2. No significant differences were noted in the age and sex of donors and recipients, relationship between them, kidney diseases and the number of HLA mismatches. As expected the number of prior transplants was significantly higher in the DSA group (75%) and low risk group (40%) resulting in significantly higher PRA (panel reactive antibodies) IgG levels as in the no risk and ABOi group. Interestingly, time on dialysis was also higher in the DSA group.

Desensitization within the ABOi and DSA group
The median anti-donor IgG isoagglutinine titer in the ABOi group was 1: 512 (range 1:2 up to 1:1024) prior of treatment. The 26 patients underwent a mean of 4 ± 3 IA sessions before surgery. Five of 26 patients (19.2%) patients required an IA session post-transplantion because of a rise in isoagglutinine titers. All 5 patients had a higher titer before transplantation. Within the DSA group 1 patient was desensitized due to DSA against HLA-class I only, 1 patient due to DSA against HLA-class I and II, while the remaining 6 patients were positive for DSA against HLA-class II only. The mean DSA against HLA-class I MFI was 1000 ± 1927 and the mean DSA against HLA-class II MFI was 9100 ± 7088 before desensitization. With 6 routine PE treatments the mean DSA against HLA-class I MFI could be reduced to 200 ± 385 while the mean DSA against HLA-class II MFI remained significantly elevated by 6800 ± 7484. Four patients were transplanted with DSA against HLA-class II MFI above 3000.

Graft function
Renal function, as assessed by serum creatinine levels and estimated GFR (MDRD formula) at 10 days, 1, 4, and 6 months (data not shown) was similar in all 4 groups, while it was slightly reduced in the DSA group at 12 months after transplantation (creatinine: 1.78 ± 0.6 mg/dl; GFR: 42.6 ± 10 ml/min) as compared to the ABOi, low risk and no risk group (creatinine: 1.63 ± 0.5 vs 1.64 ± 0.5 vs 1.6 ± 0.3 mg/dl; GFR: 48.05 ± 18 vs 48.5 ± 13 vs 45.4 ± 10 ml/min, respectively). (Fig 1) However, the difference was not statistically significant. While no patient in the no risk group and ABOi group developed significant proteinuria defined as > 500 mg protein/ g creatinine ratio, 1 DSA (2950 mg protein/g creatinine) and 2 low risk patients (530 mg protein/g creatinine and 1010 mg protein/g creatinine, respectively) showed significant proteinuria (p = 0.935).

HLA antibody screening post-transplant
Persistence of pre-existing DSA. Pre-existing DSA against HLA-class II levels with a MFI value above 3000 were still detectable in 6 (75%) patients including those 3 who experienced an acute rejection episode. In 50% of patients the antibodies were able to fix complement C1q. However, C1q-DSA did not seem to predict an acute rejection. (Table 3) Persistence of pre-existing nDSA. In the low risk group we found evidence of persistent nDSA in 5/17 patients. All of these patients were biopsied due to impaired renal function. However, none of them experienced an acute rejection.

Discussion
In this report, we present our experience in patients that underwent desensitization due to ABO or HLA incompatible living donor kidney transplantation in comparison to two control collectives. Using desensitization with anti-CD20, antigen-specific IA and/or PE we demonstrated favorable short-term allograft survival and function after ABOi renal transplantation. However, in patients with Luminex-detected DSA pre-transplant we noted an increase in rejection episodes and a significant decline in renal function 12 months after transplantation. Furthermore, we noted an increase in the rate of BK virus infections after the intensified desensitization protocol in ABOi and HLA recipients. Our results are in line with the extended experience of other groups showing that ABO incompatible kidney transplantation is a reasonable way to extend the donor pool. Takahashi K et al recently reported excellent patient (91%) and graft (83%) survival 9 years after kidney transplantation within the largest cohort of 1427 ABO incompatible recipients with the longest follow-up of more than 20 years [16].
In the context of LDK transplant recipients with DSA Klein et al. showed that desensitization results in good outcomes with low rate of side effects [17]. In contrast, Bentall et al., also showed an impaired 5 year graft survival in kidney allograft recipients with a positive pretransplant flow cytometric crossmatch who were desensitized by PE and IVIG in comparison to a crossmatch-negative control group (70.7% vs. 88.0%) [18]. This is in line with our own results showing a reduced renal function with a loss of GFR of 3.1 ml/min within the first year after transplantation across HLA barriers as compared to the controls. Furthermore, we experienced more rejections in the DSA group with the persistence of DSA beeing the greatest risk to develop a rejection in our cohort. Comparably, Bagnaso et al experienced an incidence of clinical and subclinical rejections of 66% in serial graft biopsies in HLA-incompatible kidney allograft recipients with pre-transplant DSA in the first year [19]. Furthermore, patients with DSA post-transplantation had a significant higher probability to develop transplant glomerulopathy. However, we were unable to show an association of rejection episodes with high MFI DSA titers, C1q-binding DSA and/or a positive Luminex and/or B-cell crossmatch, although this might be an effect of our small sample size. In addition, C1q-binding capacity and the MFI DSA titer did not correlate with CDC crossmatch, as patient number 6 had a very low anti-DQ titer, which was C1q negative, but a positive CDC B-cell XM. He was desensitized due to the positive CDC-B-cell XM. The positive CDC XM might be caused by non-HLA antibodies. Thus while expanding the donor pool by transplanting over HLA barriers one should consider the higher immunologic risk. On the other hand, it has been demonstrated that the longer waiting time on dialysis in immunized versus non immunized recipients is a major risk factor for a poor patient outcome after kidney transplantation [20]. Furthermore, LDK transplantation after desensitization appears beneficial for the patient, in that survival rates in desensitized LDK transplant recipients were strikingly higher than in dialysis patients waiting for a HLA compatible deceased donor organ despite a diminished graft function (80.6% vs 49.1%) [21].
However, the impact of potential side effects of desensitization such as infections or malignancies is still unknown. The BKV is a member of the polyoma virus family with a seroprevalence between 46% -94% in the general population [22]. With the renal tubular system being the main target for virus infiltration reactivation of the latent BKV in kidney allograft recipients receiving immunosuppression can cause BK specific interstitial nephritis. It mostly develops within the first year and can result in a 50% rate of graft loss [23][24][25][26]. A recent increase in prevalence BKVN (1% -10%) has been attributed, in part, to the use of more potent immunosuppressive regimens and an increased use of antirejection therapy [27;28]. However, it seems the overall intensity of immunosuppression, rather than one specific drug, leading to an increased risk [29]. In line with this theory, we noted a significant increase in the incidence of BKVN in patients after desensitization in comparison to the control groups. The intensified immunosuppressive therapy due to the desensitization and/or the more frequent antirejection therapy during the first year may have rendered those groups more susceptible for BKV reactivation.
However, another reason might be a lower maintenance therapy within the low risk groups. These data are in agreement with recent reports including our own showing an increased incidence of BKVN (20-25%) after desensitization in ABOi or HLA incompatible transplant recipients [30][31][32].
In conclusion we demonstrate impaired short-term allograft outcome in LDK transplant recipients undergoing desensitization for preformed DSA (HLA-incompatible). However, regarding the increased rejection rate we did not succeed in identifying any predictive serologic parameter. Prospective and multicenter studies are needed to standardized desensitization therapies and to proof the predictive values of MFI cut-off or C1q-binding assays.