Objectives: We investigated outcomes in living-donor kidney transplant recipients with preformed donor-specific antibodies (detected with flow cytometry and specified with the LABScreen single antigen test) under desensitization pretransplant and immunosup-pression posttransplant.
Materials and Methods: Of 15 recipients included, 8 had ABO-incompatible kidney transplant. Six patients had sensitization caused by pregnancy, 8 by blood trans-fusion, 5 by previous transplants, and 1 by unknown cause. Desensitization was initiated using calcineurin inhibitors, methylprednisolone, and mycophenolate mofetil 30 days pretransplant, with rituximab administered 1 and 10 days pretransplant. Patients underwent plasmapheresis 1, 3, and 5 days pre-transplant. Antithymocyte globulin was admi-nistered for 5 days posttransplant as induction therapy. At 3 and 12 months posttransplant, all recipients underwent protocol renal allograft biopsies, with donor-specific antibodies simultaneously measured with the single antigen test.
Results: T-cell complement-dependent cytotoxicity crossmatch was negative in all 15 recipients, but T-cell and B-cell flow cytometry was positive in 8 and 14 recipients, respectively. Anti-HLA class I antibodies became negative, except in 1 recipient 3 months posttransplant. Class II antibodies remained positive in 8 recipients 3 months posttransplant. No clinical or subclinical T-cell-mediated rejection occurred, but 1 recipient experienced clinical acute antibody-mediated rejection. At 3 and 12 months posttransplant, 8 and 5 recipients had subclinical acute antibody-mediated rejection. Cytomegalovirus test showed positivity in 14 recipients, but none developed cytomegalovirus disease. BK viremia was detected in 2 recipients, with 1 developing BK virus nephropathy, which was reversed by reducing immunosuppression.
Conclusions: Transplant patients with preformed donor-specific antibodies showed good outcomes in terms of desensitization and immunosuppression. However, most anti-HLA class II donor-specific anti-bodies remained, and microvascular inflammation score could indicate long-term risk of renal allograft dysfunction.
Key words : Antibody-mediated rejection, Renal allograft dysfunction, T-cell-mediated rejection
The complement-dependent cytotoxicity (CDC) crossmatch test is unable to detect low titers of donor-specific antibodies (DSAs); therefore, a more sensitive flow cytometry (FCM) crossmatch test has been adopted in deceased-donor kidney transplants since 2013 in the Japan Organ Transplant Network. Since then, the number of nonprimary functions in deceased-donor kidney transplants has decreased,1 which may be because of fewer incidences of acute antibody-mediated rejection (AMR). In most major kidney transplant centers, the FCM crossmatch test is also popular in living-donor kidney transplants (LDKT). In addition, DSAs are specified using the LABScreen single antigen test. Patients with high titers of DSAs (mean fluorescence intensity [MFI] > 1000) should be desensitized pretransplant to avoid acute AMR posttransplant. In Japan, the primary relationship between a recipient and donor in LDKT is a spousal relationship.2 The number of spousal LDKTs has exceeded the number of LDKTs between parent and child. Matching a wife with DSAs grafted from a husband can induce acute AMR post-transplant. Therefore, with this matching, DSAs should be studied in detail pretransplant. Other causes of generation of DSAs are blood transfusion and previous history of transplants. In Japan, the wait time for deceased-donor kidney transplant is > 14 years,3 and paired donations of LDKT are not systematized because these types of transplants are prohibited by the ethical guidelines of the Japan Society for Transplantation.4 Therefore, we often need to perform LDKT for a recipient with preformed DSAs from a donor with HLA-specific antigens.
Desensitization involves the administration of rituximab, mycofenolate mofetil (MMF), steroids, calcineur in inhibitors (CNIs), intravenous immuno-globulin (IVIG) and the procedure of double -filtration plasmapheresis pretransplant (Figure 1). This desen-sitiz-ation protocol should result in negative T-cell CDC crossmatch. We evaluated whether desensitization pretransplant and antithymocyte globulin induction triple therapy (MMF, steroid, and CNI) post-transplant could prevent clinical acute, hyperacute, and sub-clinical AMR. In addition, we assessed whether such heavy immunosuppression could induce serious complications, including life-threatening infections.
Materials and Methods
Recipients underwent CDC and FCM crossmatch tests; when either the T- or B-cell FCM crossmatch test was positive, DSAs were specified with the LABScreen single antigen test. Fifteen recipients with DSAs were enrolled in the present study (Table 1), with 8 recipients having ABO-incompatible kidney transplants. Six patients had sensitization caused by pregnancy, 8 by blood transfusion, 5 because of previous transplants, and 1 because of unknown reasons (Table 1).
For pretransplant desensitization, CNIs (6 pa-tients received cyclosporine and 9 patients received tacrolimus), methylprednisolone, and MMF were initiated 30 days before transplant, and rituximab (100 mg/day) was administered 1 and 10 days before transplant (Figure 1). The recipients underwent plasmapheresis 1, 3, and 5 days before transplant. Nine study patients received IVIG (100 mg to 1 g/kg) pretransplant. After transplant, antithymocyte globulin (1.5 mg/kg/day) was administered for 5 days (days 1-4) as induction therapy in 13 of 15 recipients. An MFI cutoff of < 1000 was defined as negative.
All rejection events were pathologically proven with biopsies. All 15 recipients underwent protocol renal allograft biopsies 3 and 12 months posttransplant. Acute AMR was diagnosed according to microvascular inflammation (MVI) score, C4d ≥ 2 (immunofluo-rescence), glomerulitis plus peritubulitis score (g + ptc) ≥ 2, and positive DSA based on Banff classification 2013.5 Incidences of rejection, complications (including infections and late-onset neutropenia), serum creatinine levels, and patient and graft sur-vival rates were evaluated. Neutropenia was classified as mild (1000-1500 cells/μL), moderate (500-1000 cells/μL; grade 3), or severe (< 500 cells/μL; grade 4) based on absolute neutrophil count.6
The paired t test, Welch t test, and chi-square test with Yates continuity correction were used for statistical analyses.
This study was conducted in accordance with the principles of the Declarations of Helsinki and Istanbul. All study procedures were approved by the Ethical Committee of the Toho University, Omori Medical Center (approval number: 27-203).
T-cell CDC crossmatch was negative in all 15 reci-pients, but T-cell FCM crossmatch was positive in 8 recipients on day of transplant (Table 2). B-cell FCM crossmatch was positive in 14 recipients (Table 2). Ten of 11 recipients (91%) with class I-positive DSAs developed negative DSAs within 12 months post-transplant (Figure 2A). Conversely, 4 of 15 recipients with class II DSAs maintained high MFI values (> 10 000), and 6 of 15 recipients (40%) developed nega-tive class II DSAs (MFI < 1000) (Figure 2B).
The negative rate of DSAs posttransplant was significantly higher in class I than in class II DSAs (P < .026; chi-square test with Yates continuity correction). The MFI values of class II DSAs versus class I DSAs were significantly higher at 3 months (4221 ± 5855 vs 720 ± 1662; P < .022) and 12 months (5498 ± 8345 vs 520 ± 1001; P < .019) posttransplant despite no significant differences in MFI values before transplant (7969 ± 6346 vs 5244 ± 5116).
No clinical or subclinical T-cell-mediated rejection occurred, but clinical acute AMR occurred in 1 ABO-compatible kidney transplant recipient. The biopsy of a recipient with a clinical acute AMR episode showed positive MVI (g + ptc = 3) and intimal or transmural arteritis (v = 1) although C4d was negative and the recipient had DSAs. In contrast, 8 and 5 of 15 recipients had subclinical acute AMR 3 months and 1 year posttransplant, respectively (Tables 3 and 4). Scores for g + ptc in protocol biopsies increased in 3 of 15 recipients at 12 versus 3 months posttransplant (Tables 3 and 4). C4d scores increased in 4 of 15 recipients at 12 versus 3 months posttransplant in protocol biopsies (Tables 3 and 4).
Severity of MVI on light microscopy did not appear to develop in all recipients. Scores of tubulitis (t), interstitial infiltration (i), and vasculitis (v) based on Banff classification in biopsies at 3 versus 12 months posttransplant were 0.40 ± 0.63 versus 0.47 ± 0.64, 0.13 ± 0.35 versus 0.13 ± 0.35, and 0.00 ± 0.00 versus 0.00 ± 0.00 (all not significant). Protocol biopsies in 8 and 7 recipients 3 months posttransplant and 10 and 5 recipients 12 months posttransplant showed ct1 and ct0, respectively. Protocol biopsies in 1 and 14 of 15 recipients at 3 months posttransplant and 7 and 8 of 15 recipients 12 months posttransplant showed ci1 and ci0, respectively. Interstitial fibrosis, but not tubular atrophy, developed (P < .039; chi-square test with Yates continuity correction) 12 months posttransplant.
Cytomegalovirus (CMV) antigenemia test became positive in 14 of 15 recipients. One of 2 CMV-seronegative patients who received a graft from a seropositive donor did not show positivity for antigenemia, and no recipients developed CMV disease. BK virus in urine was detected in 3 recipients, with 2 showing BK viremia and 1 developing BK virus nephropathy, which was treated by reducing immunosuppressant doses. One recipient had a urinary tract infection. Another had pneumonia. Anemia, late-onset neutropenia, new-onset diabetes, and hypertension occurred in 9, 6, 5, and 11 recipients, respectively. Absolute neutrophil counts in the 6 recipients with late-onset neutropenia were 986 cells/μL at 4 months, 528 cells/μL at 13 months, 1091 cells/μL at 3 months, 224 cells/μL at 5 months, 782 cells/μL at 5 months, and 1321 cells/μL at 7 months, respectively.
Serum creatinine levels in the 15 recipients did not increase until 1 year posttransplant (1.09 ± 0.32, 1.08 ± 0.32, 1.07 ± 0.32 mg/dL at 3, 6, and 12 months posttransplant) (Figure 3). However, 12 recipients showed increased levels 2 years posttransplant, with mean level (1.20 ± 0.23 mg/dL) significantly higher than that at 1-year posttransplant (P < .012). Proteinuria (0.6 g/g creatinine) developed in 1 recipient at 1-year posttransplant. Patient and graft survival rates were 100% at 2 years; however, 1 patient died of lung cancer 2 years and 10 months posttransplant. So far, the other 14 recipients have survived with functioning grafts.
In the past, the CDC crossmatch test to detect DSAs was the standard; however, sensitivity is so low that low titers of DSAs cannot be detected. The FCM crossmatch test is more sensitive and has been recently used for both deceased-donor kidney transplant and LDKT in Japan. In 2012, 2 recipients from a brain-dead donor and 9 recipients from a donor after cardiac arrest had primary graft nonfunction.1 However, 4 and 1 recipients in 2013, 1 and 0 in 2014, and 3 and 2 in 2015 had primary graft nonfunction. The introduction of FCM crossmatch test in 2013 may have reduced the incidence of acute AMR, resulting in fewer primary nonfunctioning grafts after deceased-donor kidney transplant.
In Japan, the spouse is the most popular donor in LDKT.2 In 2016, 506 of 1331 LDKT procedures (38.0%) were from spousal donors. A kidney donated from a husband to a wife may induce acute AMR due to DSAs after a wife is pregnant. Sagasaki and associates reported a case of acute AMR due to anti-HLA-DQ antibody after pregnancy and delivery in a female kidney transplant recipient7 who underwent LDKT with her mother as a donor. After delivery, she developed anti-HLA-DQ5 de novo antibody against her husband. Coincidentally, her mother had HLA-DQ5; thus anti-HLA-DQ5 de novo antibody was a DSA against her renal allograft. Previous organ transplants and history of blood transfusion have also been reported as risk factors for generating DSAs and inducing acute AMR.
Acute AMR due to preformed DSAs is an immunologic reaction of the remaining and accelerated producing DSAs against vascular endothelium cells. The use of double-filtration plasmapheresis and plasma exchange, IVIG, and desensitization involving rituximab can reduce DSA titers. The immune response against a renal allograft in the remaining antibodies must be minimized, and hyperacute rejection must not occur. However, accelerated production of DSAs is a second-onset antibody response in terms of memory B cells and is not easily controlled. Memory B cells rapidly produce massive DSAs when the same antigens, such as with a renal allograft, are exposed. The unique cytoplasmic domain of immunoglobulin G (IgG) causes the prompt activation of antigen-experienced IgG memory B cells.8 Antigen-experienced IgG1 memory B cells rapidly differentiate into plasma cells, whereas nonex-perienced IgG1 B cells do not, suggesting the impor-tance of stimulation history. Repression of the Bach2 transcription factor, which results from the antigen experience, contributes to the predisposition of IgG1 memory B cells to differentiate into plasma cells.8
In the United States,9 Europe,10 and South Korea,11 paired donations have been popular in LDKT with preformed DSAs. However, the ethical guidelines of the Japan Society for Transplantation prohibit living organ donations from a third-party donor.4 Therefore, paired donations (donor exchange), donations from an emotional-related donor, and altruistic donations in LDKT are prohibited in Japan. In 2016, the numbers of deceased-donor kidney transplants in Japan were considerably less than numbers in South Korea and Taiwan as well as in Western countries.12 The mean wait time for deceased-donor procedures exceeds 14 years in Japan.3 Therefore, we have been challenged to perform LDKT with preformed DSAs.
It is interesting to note that most class I DSAs became negative within 12 months posttransplant versus the class II DSAs. This may be because of the antigenic structure of the kidney, with HLA class I antigens having a wide distribution and class II antigens having restricted expression on renal vasculature.13 A renal allograft contains a larger amount of class I than class II antigens. When a renal graft is transplanted, a part of class I DSAs may be adsorbed in the renal tissues with a large amount of class I antigens. However, class II DSAs are not adsorbed into the renal tissue with a trace amount of class II antigens. Therefore, class II DSAs may remain at higher levels than class I DSAs. The HLA-DQ antigen was not shown to be expressed in normal renal tissues,14 but the HLA-DR antigen has been shown to be faintly expressed in proximal tubular cells.15 However, all 3 HLA class II antigens (HLA-DR, HLA-DP, and HLA-DQ) have been detected on renal endothelial cells with a markedly increased expression of HLA class II antigens in renal allografts undergoing rejection.16 The increase in HLA-DQ expression allows endothelial cells to behave as potential inducers of T-cell activation and as targets of DSAs directed against HLA-DQ. Patients who produce HLA class II DSAs in kidney allografts have a higher risk of developing transplant glomerulopathy associated with reduced long-term graft survival.17 In our study, although t, i, and ct scores were not different between at 3 and 12 months, scores of interstitial fibrosis (ci) significantly increased at 12 months posttransplant, suggesting that interstitial fibrosis gradually developed and could be related to the remaining HLA class II DSAs.
In the present study, acute AMR was diagnosed according to the MVI score and g + ptc score based on 2013 Banff classification.5 Nine of our recipients had C4d of ≥ 2, although 6 of 9 had ABO-incom-patible LDKT. Haas and associates reported that peritubular capillary C4d is often present on protocol biopsies, even when the histologic features of acute AMR are lacking in ABO-incompatible grafts,18 suggesting that diffuse peritubular capillary C4d deposition without rejection is associated with a lower risk for scarring in ABO-incompatible renal allografts. Therefore, it was difficult to evaluate the significance of positive C4d in the 6 ABO-incompatible LDKT recipients.
Scores of g + ptc ≥ 2 indicate moderate MVI based on the 2013 Banff classification.5 In our study, 10 and 7 of 15 recipients had scores ≥ 2 at 3 and 12 months posttransplant, and subclinical acute AMR was suspected due to MVI severity. Loupy and associates reported that patients with subclinical acute AMR had worse graft survival at 8 years posttransplant (56%) than those with subclinical T-cell-mediated rejection (88%) and those without rejection (90%).19 At 1 year, subclinical acute AMR was shown to be independently associated with a 3.5-fold increase in graft loss and more rapid progression to transplant glomerulopathy.19 Early lesions involving the glo-merular and peritubular capillaries on subclinical acute AMR included glomerular endothelial swelling and vacuolization, glomerular endothelial edema with subendothelial electron-lucent widening, early duplication/multilayering in glomerular basement membrane with subendothelial electron-lucent widening, and early multilayering in PTC basement membrane.20 Multilayering in the PTC basement membrane could be an early specific sign of chronic active. AMR. Haas and associates reported that rituximab, IVIG, and plasmapheresis for acute AMR can reduce the subsequent development of overt transplant glomerulopathy.20 Therefore, protocol biopsies should include electron microscopy findings to detect early signs of chronic active AMR in LDKT with preformed DSAs, and treatments for subclinical acute AMR should be considered.
Fourteen of our 15 recipients had positive CMV antigenemia; however, no patient developed CMV disease. Strict monitoring of CMV antigenemia and prophylactic use of valganciclovir appeared to keep CMV infection from developing into a serious infection. However, 1 of 2 recipients with BK viremia developed BK virus nephropathy, although the patient recovered after doses of MMF and tacrolimus were reduced. There were no incidences of life-threatening infection even under heavy immunosuppression. Late-onset neutropenia associated with rituximab occurred in 6 recipients. In the present study, mild, moderate, and severe late-onset neutropenia occurred in 2, 3, and 1 recipient, respectively, at about 3 to 13 months posttransplant. All 6 recipients recovered without serious infection due to repeated administration of granulocyte colony-stimulating factor. The increased serum creatinine levels in 12 recipients at 2 years versus 1 year posttransplant (1.20 ± 0.23 mg/dL; P < .012) and proteinuria (0.6 g/g creatinine) in 1 of 15 recipients at 1 year posttransplant could have been due to chronic AMR.
In LDKT with preformed DSAs, desensitization and immunosuppression could largely prevent clinical acute AMR but not always subclinical acute AMR. In our patients, HLA class I DSAs disappeared, but some class II DSAs remained at high levels posttransplant. We found that higher MVI and g + ptc scores indicated acute AMR severity and may result in chronic AMR. However, short-term outcomes of LDKT with preformed DSAs were satisfactory despite the remaining risk of chronic AMR.
Volume : 17
Issue : 1
Pages : 43 - 49
DOI : 10.6002/ect.MESOT2018.L42
From the Departments of 1Nephrology and 2Pediatric Nephrology, Toho University,
Faculty of Medicine, Tokyo, Japan
Acknowledgements: There is no funding for this study, and the authors have no conflicts of interest to declare. Yoshihiro Itabashi and Atsushi Aikawa collected and analyzed all the data. Hideyo Oguchi and Taichi Arai studied the pathological findings of the protocol biopsies. The other authors were clinically involved in the kidney transplants.
Corresponding author: Atsushi Aikawa, Department of Nephrology, Toho University, Faculty of Medicine, 6-11-1, Omorinishi, Otaku, Tokyo 143-8541, Japan
Phone: +81 3 3762 4151
Table 1. Characteristics of Living-Donor Kidney Transplant Recipients With Preformed Donor-Specific Antibodies
Table 2. Flow Cytometry Crossmatch on Day of Transplant
Table 3. Protocol Biopsies 3 Months Posttransplant According to the 2013 Banff Classification
Table 4. Protocol Biopsies 12 Months Posttransplant According to 2013 Banff Classification
Figure 1. Desensitization and Immunosuppression in Living-Donor Kidney Transplant Recipients With Preformed Donor-Specific Antibodies
Figure 2. Mean Fluorescence Intensity Values in HLA Class I (A) and Class II (B) Donor-Specific Antibodies
Figure 3. Serum Creatinine Levels Posttransplant