Human leukocyte antigen (HLA) class I- and II-specific antibodies present before kidney transplantation or induced after transplantation have been associated with hyperacute and acute vascular rejection episodes, graft loss, and poor long-term graft survival (1). Historically, donor-specific antibodies have been detected in complement-dependent cytotoxicity assays after incubation of patient serum with donor T- and/or B-lymphocytes in the presence of rabbit complement (2). Flow cytometric crossmatch (FCXM) techniques added sensitivity to the detection of donor-reactive antibodies (3). More recently, ELISA (4) and flow cytometry protocols (5) have been developed using soluble HLA antigens coated onto micro-titer wells or microbeads, allowing detection and specificity determination of donor HLA-specific antibodies. The link between the presence of HLA-specific antibodies and kidney graft dysfunction was recently strengthened by Terasaki et al. (6), who showed in prospective trials that the presence of HLA-specific antibodies in patients with functioning grafts was strongly associated with graft failure during a 4-year follow-up.
Even though the success rates of kidney grafts from HLA identical sibling donors are significantly better than those from HLA-mismatched donors (7), rejections of HLA identical grafts do occur, suggesting that antigenic systems other than HLA are important (8). A recent publication by Opelz (9) strongly supports this notion. He showed that, among recipients of HLA-identical sibling transplants, patients with no panel reactive antibodies (PRA) had significantly higher 10-year graft survival than patients with PRA; again suggesting a significant role of non-HLA immunity for graft survival. Furthermore, despite negative conventional donor lymphocyte XMs several cases of hyperacute and acute vascular kidney graft rejections have been reported, implicating an important clinical role for target antigens not expressed on lymphocytes (10).
Endothelial cells of donor organs are the most immediate targets for the host’s immune system during allograft rejection (11), and it has been claimed that anti-endothelial cell antibodies (AECAs) are clinically important in kidney (12–14), heart (15), and liver (16) transplantation. Methods currently used to detect AECAs are laborious, time-consuming, and impractical for use as clinical routine diagnostics. Recently, one of us developed a method by which donor-specific AECA could be detected in a FCXM using donor precursor endothelial cells as target cells (17). A novel crossmatch kit, XM-ONE, has been developed based on this method. Here, we report the results from a multicenter study in which the pretransplant presence or absence of AECAs as detected by XM-ONE, was correlated with kidney transplantation outcome.
MATERIALS AND METHODS
Subjects were recruited at six (two Swedish and four US) transplantation centers between June 2005 and October 2006. Patients were accepted for transplantation and included into the study, based on the standard T- and B-lymphocyte complement-dependent cytotoxicity and FCXM currently in use in respective centers. PRA values were recorded in 146 of 147 evaluated cases using FlowPRA, Luminex, or ELISA (four cases) and were positive in 33 cases. The HLA classes I and II antibody specificities were determined in 22 of 33 cases using flow cytometric solid-phase assays. Four of the six centers only recruited live donors, whereas at two centers also deceased donors (n=25; see Table 1) were included. Even though four centers for practical reasons did not manage to recruit deceased donors, inclusion of patients was consecutive and no selection based on diagnosis, degree of sensitization, or patient characteristics was performed. Clinical care and immunosuppression were according to the routine protocols at each participating center as described below.
In 131 of 147 cases, the immunosuppression protocol used was reported. The regimens were as follows: tacrolimus, mycophenolate mofetil (MMF), and steroids (n=35); cyclosporine, steroids, rapamycin, anti-lymphocyte (ALG), or anti-thymocyte globulin (ATG) (n=26); tacrolimus, MMF, steroids, and ALG/ATG (n=22); tacrolimus, MMF, steroids, and other (n=18); and other combinations (n=30). There was no statistically significant difference in rejection frequency between patients receiving different immunosuppressive protocols.
Of the endothelial cell crossmatch (ECXM) positive patients, 16 experienced a total of 22 rejection episodes (5 patients had >1 rejection episode). In the ECXM negative group, 13 patients had a total of 16 rejection episodes (1 patient had >1 rejection episode). The antirejection therapy was as follows: steroids only (2 ECXM+ and 9 ECXM−), ATG only (3 ECXM+ and 2 ECXM−), steroids and ATG (5 ECXM+ and 2 ECXM−), plasmapheresis, or immunoadsorption in combination with other therapy (5 ECXM+ and 0 ECXM−), other (2 ECXM+ and 3 ECXM−), and not reported/not treated (5 ECXM+ and 0 ECXM−).
The study protocol for the participating Swedish centers was reviewed and approved (Docket no. 2005/222-31/1; 2005) by the Stockholm Regional Human Ethics Committee. Each participating US center had approval from its respective institutional review board. If a patient was diagnosed with a rejection episode, the responsible physician was allowed to access the ECXM results when making decisions on rejection therapy. Further, an interim analysis was required when 100 transplanted patients had reached 3 months follow-up with a requirement to abort the study if the results showed that the ECXM predict rejection episodes with significant certainty. Because this was the finding of the interim analysis, further recruitment of patients into the study was stopped at that point. The total number of patients recruited was 195.
The EC FCXM measuring IgG and IgM antibodies reacting with donor precursor endothelial cells was performed simultaneously with the standard pretransplant XM procedure. The result from the ECXM was not available to the responsible clinician(s) for deciding whether or not to proceed with the transplantation.
Of the 195 patients (two Swedish centers included 76 patients and four US centers included 119 patients) registered into the web-based report form, 48 patients were excluded for the following reasons: (1) 19 patients were not transplanted or had no ECXM performed; (2) 4 patients were pretreated by plasmapheresis; (3) 8 patients had no follow-up data; (4) the negative control was outside the specified range in 14 cases; and (5) 3 grafts were lost due to surgical complications in the immediate postoperative period, resulting in 147 patients evaluated in the data analysis. Four of the 147 patients received ABO incompatible grafts and were pretreated by GlycoSorb-ABO (Glycorex Transplantation AB, Lund, Sweden) immunoadsorptions. The surgical complications included one graft lost at day 12 because of a rupture of the iliac artery during angioplastic dilatation of a proximal arterial stenosis. This graft had never functioned and prior biopsies had showed acute tubular necrosis and ischemia, but no signs of rejection. The second lost graft had two arteries, anastomosed to the external iliac artery using a Y reconstruction. A thrombosis of the artery branch supplying the lower part of the graft was diagnosed on the following day and verified at surgical exploration. The recipient was regrafted on day 9, and the explanted graft showed no circulation of the lower pole and necrosis of the ureter. Histological examination revealed necrosis and no sign of rejection. In the third case, the recipient died on day 2 posttransplantation because of septicemia. The 147 patients fulfilled the following inclusion criteria: the ECXM yielded a channel value for the negative control serum between 200 and 300 on a 1024 scale and follow-up data from 3 to 6 months or biopsy-verified rejection episodes were provided.
Endothelial precursor cells (EPC) expressing the angiopoietin receptor, Tie-2, were isolated from donor peripheral blood mononuclear cells using the XM-ONE kit according to the instructions of the manufacturer (AbSorber AB, Stockholm, Sweden). Briefly, 32 mL of peripheral blood were collected in Vacutainer CPT tubes with heparin (BD, Franklin Lakes, NJ) and centrifuged for 15 min at 1500 to 1800 g in a swing-out rotor. Donor peripheral blood mononuclear cells were recovered from above the gel layer and EPCs were isolated by immunomagnetic separation on nanobeads coated with anti-Tie-2 antibody. For crossmatch testing, washed Tie-2 positive cells were incubated with patient serum. EPC-reactive IgG and IgM were detected by flow cytometry using fluorochrome-conjugated secondary antibodies. The channel value for the negative control serum was set to between 200 and 300 on a 1024 scale. Positive and negative ECXM test values were established for each immunoglobulin isotype using standard procedures used to validate the lymphocyte FCXM assay at the Karolinska University Hospital. A shift of more than or equal to 80 channels above the negative control serum was determined as a positive test value for IgM and a shift of more than or equal to 50 channels for IgG. During the course of the study, the positive test value for the IgG isotype was modified to include a negative channel shift value of more than or equal to 80 channels below the negative control serum. At one center the positive test values were set at 300 channels for IgG and 400 channels for IgM because a flow cytometer of a different make was used at this center.
Phenotyping of Tie-2+ Precursor Endothelial Cells Isolated Using XM-ONE
Single color fluorescence was used to characterize the Tie-2+ cells isolated using the XM-ONE kit phenotypically. Phycoerythrin-conjugated antibodies specific for major histocompatibility complex (MHC) class I (HLA-A, -B and -C; clone G46–2.6; IgG1κ; BD Pharmingen, San José, CA), MHC class II (HLA-DR; clone TÜ36; IgG2bκ; BD Pharmingen), VEGF-R2 (clone 89106; IgG1; R&D systems, Minneapolis, MN) and CD32 (clone 190723; IgG2a; R&D systems) were analyzed and compared with a PE-conjugated isotype control (clone 11711.11; mouse IgG1; R&D systems). The phenotype of Tie-2+ cells has been previously characterized by Vermehren et al. (17).
Oversight of data collection, management, and statistical analysis was performed by Pharma Consulting Group (Uppsala, Sweden). Before the study, it was estimated that approximately 8% of all transplanted patients would have antibodies specific for endothelial cells. Thus, to meet 80% power with a two-sided test and a 5% level of significance, 250 fully evaluated subjects were needed in the study. To meet this power, we estimated that 350 patients would need to be recruited.
The primary efficacy variable analyzed was the difference between a positive and negative ECXM with respect to the proportion of patients with rejection episodes. Except where noted, differences in proportions between groups were tested for significance with Fishers’s two-sided exact test. A P value less than 0.05 was considered to be significant. All data are presented with standard descriptive statistics by group.
Phenotypic Characterization of Tie-2 Positive Cells From Peripheral Blood
Analysis by flow cytometry revealed two major populations of Tie-2+ cells in the side versus forward scatter dot plot following isolation using XM-ONE (Fig. 1A). The more abundant one (R1) was CD3− and contained larger and more granulated cells, whereas the CD3+ population contained smaller, less granulated lymphocytes (Fig. 1A). In addition to previously described markers (17), Tie-2+ CD3− cells expressed VEGF-R2 and CD32 (Fig. 1B). The reduced expression of these latter markers can be used to distinguish between the Tie-2 positive and negative cells of the CD14+ population (not shown). The FACS analysis indicated that Tie-2+ cells isolated by XM-ONE expressed HLA class I and II antigens in addition to a few endothelial cell markers (Fig. 1B and not shown).
On average, approximately 1.7±1.1×106 Tie-2+ cells/32 mL of blood (n=11) can be obtained from single donors.
Demographic data of the 147 patients studied is given in Table 1. The patients had a median follow-up time of 350 days, ranging from 1 to 447 days. There were 59% males in the whole study population and no statistical difference in the gender distribution between patients with positive and negative ECXM tests. There were significantly more patients of Afro/Caribbean origin in the ECXM+ group of patients. There was no difference in the fraction of live donors between ECXM positive and negative patients. The other parameters listed in Table 1 will be commented on below.
Patients With a Positive Pretransplant ECXM Have a Higher Risk for Rejections
AECA were identified in 24 of 147 patients using a positive ECXM test value of more than or equal to 80 channels above the negative control serum for IgM and a shift of more than or equal to 50 channels for IgG. Interestingly, two centers reported an increased incidence of rejection among patients with IgG channel shifts of more than or equal to 80 channels below the negative control serum (Table 2). Further testing in two such cases showed that the ECXM became positive, as defined by a channel shift value of more than or equal to 50 for IgG, following dilution of the serum in negative control serum. Because of the uncertainty as to whether the patients with more than or equal to 80 channel shifts below the negative control for IgG are really positive or negative, they were excluded from the initial analysis. The incidence of rejection was significantly higher among patients (9 of 24, 37.5%) with AECA compared with those without (13 of 112, 12%; P<0.005; Table 2). Furthermore, a significantly higher fraction of patients 8 of 24 (33%) with AECA had rejections within 2 weeks after transplantation when compared with those without 6 of 112 (5%; P<0.0005; Table 2). Because dilution of the negative channel shift sera resulted in a positive ECXM (a positive shift of ≥50 channels) and because there was a high frequency of rejections among these patients, we have in the following presentation of the data chosen to include these patients in the collective group of ECXM+ patients. Therefore, the total number of patients possessing AECA was increased from 24 to 35 patients. The incidence of rejection remained significantly higher among patients with AECA (16 of 35; 46%) compared with patients without (13 of 112, 12%; P<0.00005; Table 2).
Of the ECXM+ patients that also had HLA antibodies (13 of 35), two had donor-specific HLA antibodies as determined by solid-phase assays, and in three cases the HLA-antibody specificity was not determined. One of the patients with donor-specific HLA antibodies had one antibody-mediated, C4d positive rejection episode within 2 weeks, whereas the other had no rejection. After excluding both patients with verified donor specific HLA antibodies and those with antibodies detected in the solid phase PRA screen but where the HLA specificities were not determined, there were still significantly more rejections recorded in the ECXM+ group (13 of 30 vs. 13 of 107; P<0.001).
Autologous ECXM tests were performed in seven cases, but were not part of the protocol. Three patients with a positive ECXM, one with IgG and two with IgM, also had a positive autologous ECXM. None of these patients had any rejection episodes. Furthermore, there was no difference between ECXM positive and negative patients or between patients with and without rejections with respect to their immunosuppressive treatment (not shown).
Donor-Reactive AECA of Both IgG and IgM Class Were Associated With Graft Rejections
The majority of patients had AECA of IgG or IgM class, and only one patient had both. Patients with IgG or IgM reactivity against donor endothelial cells experienced significantly more rejections than patients without (Table 2). Four of nine (44%) patients with a positive channel shift of more than or equal to 50 for IgG, 8 of 12 (67%) patients with a negative channel shift of more than or equal to 80 for IgG, and 5 of 15 (33%) patients with IgM in the ECXM had rejections during the follow-up when compared with 13 of 112 patients without (P<0.005, P<0.000005, and P<0.05, respectively). The corresponding numbers for rejections within 2 weeks were 4 of 9, 6 of 12, and 4 of 15 compared with 6 of 112 patients without AECA (P<0.005, P<0.0005, and P<0.05, respectively; Table 2).
Correlation Between Sensitization Status and a Positive ECXM
In the patient population tested for the presence of HLA-specific antibodies by solid-phase assays, 113 of 146 (77.5%) were nonsensitized (PRA of <10%), 25 of 146 (17%) were sensitized (PRA of 10%–80%), and 8 of 146 (5.5%) were highly sensitized (PRA>80%). Of the patients with AECA, 25 of 35 (71%) were nonsensitized, 8 of 35 (23%) were sensitized, and 2 of 35 (6%) were highly sensitized (Table 1). There seemed to be more female patients with a history of pregnancy (11 of 17 vs. 17 of 43; P=0.09) and patients with previous blood transfusion (14 of 35 vs. 27 of 112; P=0.084) in the ECXM+ group compared with the ECXM− one (Table 1). However, these differences were not statistically significant.
Thirteen of the 35 (37%) patients with a positive ECXM had antibodies to HLA class I, HLA class II or both, and 10 of these patients at a level of more than or equal to 10% PRA. Two of the 13 patients were shown to have donor-specific HLA antibodies, and in three patients the specificities of the HLA antibodies were not determined, suggesting that in fact at least 30 of 35 ECXM+ patients had donor-reactive antibody that was not specific for HLA. Six of 13 ECXM+ patients with HLA antibodies had rejections and one patient had a primary-nonfunctional graft; 5 of 6 rejections occurred within 2 weeks and 2 of 7 were classified as antibody-mediated.
Of the ECXM+ patients, 30 were not previously transplanted, four had one previous transplant, and one patient had two previous transplants. In the ECXM− group the corresponding numbers were 96, 14, and 2 patients, respectively (Table 1).
Correlation of a Positive ECXM With Rejection Type and Timing
Only biopsy-verified rejections were recorded and classified according to the Banff ‘97 criteria (18, 19). Table 3 shows the distribution of the different rejection types among patients with positive or negative ECXM results and with or without HLA-specific antibodies. It is clear that verified antibody-mediated rejections (18) and the only type III rejection—a rejection type strongly suggestive of an antibody-mediated component (19)—were only found in ECXM+ patients. The antibody-mediated rejections in the ECXM positive patient group were equally distributed between those with and without HLA-specific antibodies. Only one of the patients with rejection had verified donor-specific HLA antibodies. The conclusion that patients with a positive ECXM are at risk for antibody-mediated rejections was further supported by C4d staining; six of seven C4d positive biopsies were found among ECXM+ (3 with IgM, 2 with IgG, and 1 with IgG negative channel shifts) patients. Rejections among ECXM+ patients occurred early after transplantation, with 81% of the rejections occurring within 2 weeks.
Donor-Reactive AECA Are Significantly Correlated With Decreased Kidney Function at 3 and 6 Months
To assess kidney function among patients with and without AECA, serum creatinine values were measured during follow-up at 3 and 6 months. Serum creatinine levels were significantly higher in patients with donor-reactive AECA than without at 3 months (152 [n=34] vs. 122 [n=108] μmol/L, P<0.05) and 6 months (156 [n=33] vs. 126 [n=94] μmol/L, P<0.05) posttransplant (Fig. 2). The highest creatinine values at 3 and 6 months were found among kidney transplant recipients having HLA-specific antibodies and a positive ECXM test before transplantation.
The clinical importance of HLA-specific antibodies to graft rejection is well established. However, several investigators have shown that non-HLA-specific antibodies also negatively impact graft survival. In the most dramatic cases, the graft may actually be lost hyperacutely or acutely because of such antibodies. In many instances, non-HLA-specific antibodies react with donor endothelium, and the transplant community has for at least the past two decades tried to establish a test by which the presence of AECA can be detected before transplantation. The methods used include the crossmatch tests against monocytes (12), donor skin (20) or vessel (13), and flow cytometric assays using cultured, primary endothelial cells or the hybrid cell line EA.hy 926 as targets (11, 21, 22). These assays are time consuming, difficult to perform, and of poor reproducibility, and therefore not suitable as routine diagnostic tests. Here, we have evaluated a novel ECXM kit, XM-ONE, which allows detection of antibodies reacting with donor precursor endothelial cells immuno-magnetically isolated on Tie-2-specific antibody-bearing nanobeads. The assay is as easy to perform as a conventional FCXM using lymphocytes as targets.
We found that, 24% (35 of 147) of the patients fulfilling the study inclusion criteria were AECA+ in the XM-ONE assay and that a positive ECXM was strongly associated with rejection within 2 weeks and during the follow-up of a least 3 months (P<0.00005). In a previous study, Cerilli et al. (13) reported that 15 of 55 (∼27%) consecutive recipients of deceased donor renal allografts had a positive donor-specific vascular endothelial cell (VEC) XM pretransplant despite negative T- and B-lymphocyte crossmatches. There was a strong correlation between the VEC XM results and the clinical course such that 6 of 7 patients with irreversible rejections and 4 of 5 with early, reversible rejections, but only 5 of 43 with a benign clinical course were VEC XM+ (13). Paul et al. (23) reported that 7 of 61 (∼11%) renal allograft recipients had AECA, defined by a positive donor-specific VEC XM despite negative lymphocyte crossmatches, and of these, six patients lost their grafts to vascular rejections. Thus, AECAs may be found in as many as 25% of kidney transplant recipients with negative conventional T- and B-lymphocyte XM and are strongly correlated with rejection episodes. The association of a positive ECXM detected pretransplant with rejection was seen for both IgG and IgM in this study.
Importantly, 25 of 35 (∼71%) ECXM+ patients were nonsensitized as defined by a PRA reactivity of <10%, and in 22 of 35 (63%) ECXM+ patients no HLA-specific antibodies were detected. Even though autologous ECXMs were performed in seven patients and were positive in three patients with donor-reactive AECAs, it was not in the protocol to perform autologous ECXMs. Thus, it is unclear at this time whether the AECAs of nonsensitized patients are autoreactive, part of the natural antibody repertoire, or antibodies crossreactive with microbial determinants. Nevertheless, even autoreactive antibodies such as antibodies to vimentin have been associated with graft rejection (24–26). Future studies should reveal the frequency of autoreactive AECA in patients with a positive ECXM.
An interesting observation was that there were significantly more rejections in a group of patients with IgG channel shift values of more than 80 channels below the negative control serum, compared with patients with a negative ECXM (8 of 12 vs. 13 of 112). In two cases, diluting such a serum resulted in an ECXM with a channel shift value of more than or equal to 50 (J.A.M., unpublished data, 2007). The explanation for this observation still needs to be established, but may include blocking (27) or autoquenching (28) effects.
The fact that the majority of ECXM positive cases (54% during the follow-up) do not experience rejection, can be explained by the fact that some antigens present on the EC precursors may not be expressed within the renal vascular endothelium, may not be relevant to transplant rejection, or may contribute to chronic rejection which is not being evaluated in this study. Concomitant HLA sensitization, antibody class, subclass, and titer are all important factors that may determine the pathogenicity of AECA as detected by the ECXM test. In addition, the humoral responses toward a transplanted allograft are dynamic, why predicting transplant rejection with a high degree of certainty using a single pretransplant time point is difficult (29).
In light of the poorer long-term survival of kidney grafts in African Americans compared with non-African Americans (30), an interesting observation was the significantly raised number of Afro-Caribbean recipients in the ECXM positive group. Whether a higher prevalence of AECAs can partially explain the higher immunologic risk that African Americans face after kidney transplantation, and whether the ECXM test is particularly suited for identifying high-risk patients of African American origin remains to be shown. Future studies are justified in which healthy individuals and end-stage kidney disease patients of different ethnicity are examined with regard to the prevalence of AECAs.
The fact that the study was performed at six different centers with their own routines in terms of assays performed and equipment used is likely to have caused larger variations in the ECXM results than if the investigation would have been performed at a single center or at centers with homogenous testing routines. However, it may also be considered a strength of the study that significant results were obtained despite the diverse practices in use at the different centers.
Currently, the molecular targets for AECAs detected by the ECXM are unknown. As Tie-2+ cells carry both endothelial cell and monocyte markers, antigens recognized by antibodies detected in the XM-ONE crossmatch test can be of both endothelial and monocyte origin. Candidates include the MHC class I chain-related antigens A (MICA) and B (MICB) and the angiotensin II type 1 receptor (31). The former two are expressed on endothelial cells, and anti-MICA/B antibodies have been shown to be associated with kidney graft rejection (6). Mizutani et al. (32) investigated the frequency of MIC-specific antibodies in 139 kidney transplant recipients, with and without the presence of HLA-specific antibodies, who rejected their grafts. Of 46 patients who lost their grafts and had HLA-specific antibodies, 26% had MIC-specific antibodies, whereas 37% of patients who rejected their grafts but lacked HLA-specific antibodies, had MIC antibodies (32). Staining for MICA on the Tie-2+ precursor endothelial cell has been inconsistent (data not shown). In a recent study of 33 kidney transplant recipients with refractory vascular rejection, 20 lacked HLA antibodies. Of these, 16 had malignant hypertension and IgG against the angiotensin II type 1 receptor (31). The frequency of antibodies specific for the angiotensin II type 1 receptor in ECXM+ sera should be investigated. Identification of target antigens in addition to those mentioned above that might be responsible for the ECXM reactivity will be important.
In conclusion, XM-ONE detects an antibody population that is not detected by lymphocyte crossmatches and that is not HLA-specific, but which is strongly associated with rejection episodes and decreased kidney function at 3 and 6 months. On the basis of the results of the multicenter study, we conclude that XM-ONE testing identifies patients at risk for rejection who therefore may benefit from a modified immunosuppressive protocol. Further studies are needed to identify the significance of, and reason for, ECXM tests with negative channel shifts as well as the importance of the autologous ECXM test in kidney transplantation.
1. Terasaki PI. Humoral theory of transplantation. Am J Transplant
2003; 3: 665.
2. Terasaki PI, McClelland JD. Microdroplet assay of human serum cytotoxins. Nature
1964; 204: 998.
3. Scornik JC. Detection of alloantibodies
by flow cytometry: relevance to clinical transplantation. Cytometry
1995; 22: 259.
4. Zachary AA, Delaney NL, Lucas DP, et al. Characterization of HLA class I specific antibodies by ELISA using solubilized antigen targets. I. Evaluation of the GTI QuikID assay and analysis of antibody patterns. Hum Immunol
2001; 62: 228.
5. Pei R, Lee JH, Shih NJ, et al. Single human leukocyte antigen flow cytometry beads for accurate identification of human leukocyte antigen antibody specificities. Transplantation
2003; 75: 43.
6. Terasaki PI, Ozawa M, Castro R. Four-year follow-up of a prospective trial of HLA and MICA antibodies on kidney graft survival. Am J Transplant
2007; 7: 408.
7. Opelz G, Wujciak T, Dohler B, et al. HLA compatibility and organ transplant survival. Collaborative Transplant Study. Rev Immunogenet
1999; 1: 334.
8. Kalil J, Guilherme L, Neumann J, et al. Humoral rejection in two HLA identical living related donor kidney transplants. Transplant Proc
1989; 21: 711.
9. Opelz G. Non-HLA transplantation immunity revealed by lymphocytotoxic antibodies. Lancet
2005; 365: 1570.
10. Sumitran-Holgersson S, Holgersson J. Clinical importance of non-HLA antibodies in solid organ transplantation. Curr Opin Organ Transplant
2006; 11: 425.
11. Glotz D, Lucchiari N, Pegaz-Fiornet B, et al. Endothelial cells as targets of allograft rejection. Transplantation
2006; 82: S19.
12. Cerilli J, Bay W, Brasile L. The significance of the monocyte crossmatch
in recipients of living-related HLA identical kidney grafts. Hum Immunol
1983; 7: 45.
13. Cerilli J, Clarke J, Doolin T, et al. The significance of a donor-specific vessel crossmatch
in renal transplantation. Transplantation
1988; 46: 359.
14. Paul LC, Claas FH, van Es LA, et al. Accelerated rejection of a renal allograft associated with pretransplantation antibodies directed against donor antigens on endothelium
and monocytes. N Engl J Med
1979; 300: 1258.
15. Fredrich R, Toyoda M, Czer LS, et al. The clinical significance of antibodies to human vascular endothelial cells after cardiac transplantation. Transplantation
1999; 67: 385.
16. Sumitran-Holgersson S, Ge X, Karrar A, et al. A novel mechanism of liver allograft rejection facilitated by antibodies to liver sinusoidal endothelial cells. Hepatology
2004; 40: 1211.
17. Vermehren D, Sumitran-Holgersson S. Isolation of precursor endothelial cells from peripheral blood for donor-specific crossmatching before organ transplantation. Transplantation
2002; 74: 1479.
18. Racusen LC, Colvin RB, Solez K, et al. Antibody-mediated rejection criteria - an addition to the Banff 97 classification of renal allograft rejection. Am J Transplant
2003; 3: 708.
19. Racusen LC, Solez K, Colvin RB, et al. The Banff 97 working classification of renal allograft pathology. Kidney Int
1999; 55: 713.
20. Moraes JR, Luo Y, Moraes ME, et al. Clinical relevance of antibodies to non-HLA antigens in organ transplantation. Clin Lab Med
1991; 11: 621.
21. Le Bas-Bernardet S, Hourmant M, Coupel S, et al. Non-HLA-type endothelial cell reactive alloantibodies
in pre-transplant sera of kidney recipients trigger apoptosis. Am J Transplant
2003; 3: 167.
22. Sumitran-Karuppan S, Tyden G, Reinholt F, et al. Hyperacute rejections of two consecutive renal allografts and early loss of the third transplant caused by non-HLA antibodies specific for endothelial cells. Transpl Immunol
1997; 5: 321.
23. Paul LC, van Es LA, van Rood JJ, et al. Antibodies directed against antigens on the endothelium
of peritubular capillaries in patients with rejecting renal allografts. Transplantation
1979; 27: 175.
24. Azimzadeh AM, Pfeiffer S, Wu GS, et al. Humoral immunity to vimentin is associated with cardiac allograft injury in nonhuman primates. Am J Transplant
2005; 5: 2349.
25. Jurcevic S, Ainsworth ME, Pomerance A, et al. Antivimentin antibodies are an independent predictor of transplant-associated coronary artery disease after cardiac transplantation. Transplantation
2001; 71: 886.
26. Mahesh B, Leong HS, McCormack A, et al. Autoantibodies to vimentin cause accelerated rejection of cardiac allografts. Am J Pathol
2007; 170: 1415.
27. Landsteiner K. The specificity of serological reactions. Cambridge, MA: Harvard University Press 1946.
28. Stryer L. Fluorescence energy transfer as a spectroscopic ruler. Annu Rev Biochem
1978; 47: 819.
29. Christiaans MH, Overhof-de Ros R, Nieman F, et al. Donor-specific antibodies after transplantation by flow cytometry: relative change in fluorescence ratio most sensitive risk factor for graft survival. Transplantation
1998; 65: 427.
30. Eckhoff DE, Young CJ, Gaston RS, et al. Racial disparities in renal allograft survival: a public health issue? J Am Coll Surg
2007; 204: 894.
31. Dragun D, Muller DN, Brasen JH, et al. Angiotensin II type 1-receptor activating antibodies in renal-allograft rejection. N Engl J Med
2005; 352: 558.
32. Mizutani K, Terasaki PI, Shih RN, et al. Frequency of MIC antibody in rejected renal transplant patients without HLA antibody. Hum Immunol
2006; 67: 223.