A number of studies have demonstrated a significant association between the presence of preexisting or newly formed circulating anti-human leukocyte antigen (HLA) antibodies (Abs) and acute rejection (AR), (1–3) chronic rejection, and graft loss (4–7) in kidney transplant recipients. It has been suggested that screening patients for development of anti-HLA Abs after transplantation may be useful to identify patients at risk for rejection (1, 5). Recently, it has been shown that dynamic early monitoring of anti-HLA Ab may help predict acute humoral rejection (AHR) in high-risk patients with a positive flow cross-match at the time of transplantation (8). Testing for de novo anti-HLA Abs after transplantation is expensive and is currently only recommended to establish the diagnosis of AHR (9). This study was designed to determine whether screening for development of de novo anti-HLA Abs after transplantation could predict AR in nonsensitized kidney transplant recipients before there was clinical evidence of allograft dysfunction.
This prospective observational study was conducted in the two adult kidney transplant programs in British Columbia, Canada between February 2003 and January 2005 and was approved by our local research ethics board. Adult only (older than 18 years) kidney transplant recipients were eligible for the study. All patients were anti-human globulin enhanced cytoxicity (AHG-CDC) cross-match negative before transplantation. Recipients of HLA-identical transplants and patients who required chronic anticoagulation precluding kidney allograft biopsy were not eligible for the study. Patients had their immediate pretransplant serum screened for anti-HLA Abs by flow cytometry using latex beads conjugated with HLA antigens. Patients with any anti-HLA Abs present in pretransplant serum were excluded from further study. There were no other restrictions on eligibility.
An overview of the study protocol is provided in Figure 1.
Anti-HLA Antibody Detection After Transplantation
Simultaneous screening for HLA class I and class II Abs was performed using FlowPRA screening (#FL12-60, One Lambda Inc, Canoga Park, CA) immediately before transplantation (day 0) and 10, 20, 30, 60, 90, 180, and 365 days after kidney transplantation, as previously described (10). In brief, after incubation of serum with FlowPRA beads (a pool of 30 class I and 30 class II beads with each bead conjugated to a variety of purified HLA class I and class antigens), followed by staining with fluorescein isothiocyanate-labeled anti-human IgG Ab, the anti-HLA IgG-positive serum showed a fluorescent channel shift compared with negative serum. Percentage of panel reactive Ab (PRA) was determined by the percentage of beads that reacted positively with the serum.
Positive anti-HLA reactivity on initial FlowPRA screening was further evaluated using FlowPRA specific class I (#FL1SP) or class II (#FL2SP) tests (One Lambda Inc). In brief, each test consisted of a panel of 32 FlowPRA beads consisting of four HLA groups with eight antigen beads in each group. Each group of beads was incubated separately with the test serum followed by staining with fluorescein isothiocyanate-conjugated anti-human IgG Ab. A positive reaction was determined by a channel shift (A) compared with a negative reaction. Each of the eight beads showed a different channel shift (B). By analyzing A versus B dot plots, the HLA specificities were determined (11).
FlowPRA single antigen assays (One Lambda Inc) (for class I or class II single antigens) were used to uncover hidden specificities for patients with high PRA sera already identified using the methodology described earlier. They consisted of 32 single antigen beads in four groups (12). In addition, FlowPRA high definition class I single antigen assays (#FL1HD05 through FL1HD10) were six groups of class I single antigen beads that provided supplemental specificities for the FlowPRA class i single antigen panel.
Serum samples were tested immediately before transplantation (day 0) and on days 10, 20, 30, 60, 90, 180, and 365 after kidney transplantation (Fig. 1). In addition, anti-HLA Ab determination was performed at the time of any clinically suspected AR episode during the first posttransplant year. Among patients with biopsy-proven AR without initial evidence of circulating anti-HLA Abs, additional serum samples were obtained for Ab testing weekly for a period of 1 month. The duration of circulating anti-HLA Ab was determined in all patients who developed anti-HLA Abs by testing serum samples weekly after the time of first Ab detection for a period of 1 month and then monthly until Ab was no longer detectable.
All assays were performed in the provincial immunology laboratory by a single technologist (V.W.). Ab results were not incorporated in clinical decision making, unless there was clinical and pathological evidence of AHR in which case the presence and specificity of anti-HLA Abs were revealed to the treating physician.
Protocol Kidney Graft Biopsies
Protocol kidney graft biopsies were performed at the time of engraftment, 3 and 12 months posttransplantation, at the time of new anti-HLA Ab development, or when clinically indicated (our standard practice is to confirm all clinically suspected AR episodes with renal biopsy). All biopsies were uniformly processed and stained for C4d (C4d staining was performed on frozen sections applying a mouse monoclonal anti-C4d Ab (Quidel, San Diego, CA) as previously described (13) and were assessed for the presence of acute and chronic rejection according to the Banff 1997 schema (14) by an experienced renal pathologist (A.M.) who was blinded to the anti-HLA Ab status of the patient.
During the study period, our standard of care for low-immunological risk kidney transplant patients (patients with pretransplant PRA<20%) consisted of induction with basiliximab, a calcineurin inhibitor (either cyclosporine or tacrolimus), and mycophenolate mofetil, with or without maintenance prednisone. In addition, seven patients enrolled in this observational study were also participants in a randomized clinical trial comparing FTY720 5 mg daily plus reduced dose cyclosporine, FTY720 2.5 mg/day plus reduced dose cyclosporine, with mycophenolate mofetil plus full-dose cyclosporine (15).
Treatment of Rejection
Biopsy-confirmed AR episodes (Banff grade 1) were treated with pulse intravenous methylprednisolone (500 mg intravenous daily for 3 days), followed by oral prednisone taper. Depleting Abs (OKT3 or thymoglobulin) were used to treat steroid-resistant AR or rejection with evidence of vascular involvement (Banff grade II or greater). AHR was treated with plasmapheresis and cytomegalovirus hyperimmune globulin with or without rituximab. The duration of AHR treatment was determined by the treating physician. Protocol kidney biopsies are not routinely performed in our center and no change in treatment was instituted in patients with subclinical rejection.
Patients were followed up for the outcomes of patient and graft survival, AR, and estimated glomerular filtration rate (GFR) (using the abbreviated equation derived from the Modification of Diet in Renal Disease Study) (16) until August 2007.
The characteristics of patients who developed de novo anti-HLA Abs during the first posttransplant year were compared with those who did not using the chi-square test or analysis of variance as appropriate. The time to anti-HLA Ab formation (donor-specific Abs [DSA] or non-DSA) after transplantation was determined with the Kaplan-Meier method and group differences were compared with the log-rank test. Similarly, the Kaplan-Meier method was used to compare the time to the first AR episode between patients with and without de novo anti-HLA Abs and between patients who developed DSA and non-DSA anti-HLA Abs.
Patient and graft survival were determined using the Kaplan-Meier method and differences between patients with and without de novo anti-HLA Abs were compared with the log-rank test. The GFR at last follow-up was estimated using the four-variable equation derived from the Modification of Diet in Renal Disease Study, and group differences were determined with the t test. All analyses were performed using SAS version 9.1 (Cary, NC) and S-Plus 8.0 (Insightful Corporation, Seattle, WA).
Among the 84 kidney transplant recipients who consented to participate in this prospective study, 14 had anti-HLA Abs detected in serum samples obtained before kidney transplantation and were excluded from further study. The demographic and clinical characteristics of the 70 study patients and a comparison with and without de novo anti-HLA Abs during the first posttransplant year is shown in Table 1. A majority of participants were white, there were no African Americans, and 31% were of Asian descent. A majority of participants were living donor recipients. There were a number of numeric but few statistically significant differences among the 11 (16%) of patients who developed de novo anti-HLA Abs during the first posttransplant year and the 59 patients who did not. Maintenance immunosuppressant medication use differed significantly between patients with and without de novo anti-HLA Abs. Of the seven patients also enrolled in the FTY720 clinical trial, four developed anti-HLA Abs during the first posttransplant year.
Timing of De Novo Anti-HLA Antibody Development
The median time to de novo Ab development after transplantation was of 21 days (q1-q3=15–201 days). Figure 2 shows that de novo Abs developed early posttransplantation (7 of 11 patients formed de novo anti-HLA Abs by 1 month posttransplantation). Three patients developed only de novo non-DSA Abs and eight patients developed de novo DSA Abs alone or in combination with non-DSA Abs. Patients with only non-DSA Abs developed Abs more rapidly than patients who developed DSA Abs (median time to development of non-DSA Abs was 0.5 months (q1, q3=0.3, 0.7)) vs. 4.7 months (q1, q3=0.7, 9.6) for DSA Abs (P=0.03; Fig. 3).
During the first posttransplant year, 19 of 70 (27%) patients had AR including three patients with subclinical rejection identified on protocol biopsies. Among the 11 patients who developed de novo anti-HLA Abs, six (55%) had AR, whereas 13 of 59 (22%) of patients who did not develop de novo anti-HLA Abs had a rejection (including three subclinical rejections found on protocol biopsy at 3 months posttransplantation), P=0.01 (Fig. 4A). AR occurred in one of the three patients who developed non-DSA Abs only, whereas five of the eight patients who developed DSA Abs (alone or in combination with non-DSA Abs) had AR (P=0.03; Fig. 4B).
Anti-HLA Antibody and Acute Rejection During First Posttransplant Year
Figure 5 shows the time course of Ab formation and AR in each of the 11 patients who developed anti-HLA Abs. The first five patients who developed de novo anti-HLA Abs did not develop AR. Two of these five patients (patients numbered 1 and 4 in Fig. 5) died with a functioning graft during the first posttransplant year. In the six patients who developed both anti-HLA Abs and AR, the anti-HLA Abs developed at the same time of a first AR episode or after the occurrence of a first AR episode. Importantly, in the patients numbered 6, 7, and 10 in Figure 5, Abs were detected on serum samples that were obtained at the time of renal allograft biopsy for clinically suspected AR, and in none of these cases did a positive screening Ab test lead to a diagnosis of AR that was not already clinically evident.
There were three humoral rejection episodes in two patients, and all of these rejections occurred at least 6 months after transplantation. Both patients had previous rejection episodes that had no evidence of humoral involvement (i.e., C4d-negative and no circulating anti-HLA Abs). One of the two patients who developed humoral rejection (patient 9, Fig. 5) had de novo anti-HLA Abs detected before developing late humoral rejection, raising the possibility that monitoring for anti-HLA Abs after a first cellular rejection may be useful in predicting subsequent events. However, patients numbered 8 and 11 in Figure 5 also developed de novo anti-HLA Abs after an acute cellular rejection but did not develop further rejection episodes.
Subclinical Rejection in Protocol Biopsies
Patients underwent protocol biopsies at 3 and 12 months posttransplantation. Subclinical rejection was identified in three patients on the 3 months posttransplant protocol biopsy. The subclinical rejections had borderline changes, according to the Banff Classification, were C4d negative and concurrent serum samples tested negative for anti-HLA Abs. Sequential samples obtained weekly for a period of 1 month after the subclinical rejection also tested negative for anti-HLA Abs. The three patients who developed subclinical rejection were not treated for rejection.
Acute Rejection After First Year
We continued to follow-up patients after the first year for clinical AR. The five patients who developed de novo anti-HLA Abs but not rejection during the first posttransplant year did not have any clinically evident AR episodes during follow-up. Among the six patients who developed de novo anti-HLA Abs and had an AR during the first posttransplant year, one patient had a humoral rejection after the first year (patient 9, Fig. 5). None of the 13 patients who had an AR in the first posttransplant year without de novo anti-HLA Abs had further rejections during follow-up.
During the median follow-up of 3.7 years (Q1, Q3=1.0, 4.4), the overall patient and graft survival (including graft loss from any cause) were 94.3% and 92.9%, respectively. There was no difference in patient deaths (P=0.11), graft loss from any cause (P=0.17), or death-censored graft loss (P=0.11) between patients with and without de novo anti-HLA Abs during the first posttransplant year. Among patients who developed de novo anti-HLA Abs, there were two deaths with graft function (one patient died from sepsis 4 months after transplantation and the second patient died from a cardiac arrest at 6 months after transplantation). There were also two deaths with graft function among patients who did not develop de novo Abs (one patient died at 7 months from esophageal carcinoma and the second patient died at 24 months of pneumonia). One patient who did not have de novo Abs had graft failure at 20 months after transplantation.
Among patients with graft function at last follow-up, there was no difference in estimated GFR between patients with and without de novo anti-HLA Abs (median GFR 54.8 [Q1, Q3=47.2, 58.2] and 56.0 [Q1, Q3=45.0, 70.0] mL/min/1.73 m2, respectively; P=0.79). Unfortunately, proteinuria was not collected in study patients.
Currently, the diagnosis of renal transplant rejection relies on elevation of the serum creatinine followed by identification of pathological signs of immune-mediated injury on allograft biopsy. Identification of biological markers to permit early diagnosis of rejection would clearly be helpful in preventing AR or minimizing its clinical impact. A number of previous studies have suggested that posttransplant monitoring of anti-HLA Abs may be useful in predicting patients at risk for acute or chronic rejection of heart, lung, and renal allografts (3, 7). Recently, early monitoring of anti-HLA Abs was demonstrated to be helpful in predicting acute humoral rejection in high-risk kidney transplant recipients (positive pretransplant flow cross-match) (8). This prospective observational study was specifically designed to determine whether monitoring for development of de novo anti-HLA Abs at fixed time points after transplantation in patients without evidence of pretransplant anti-HLA Abs could predict development of AR. Consistent with other studies (1, 17), we found a clear association between anti-HLA Abs and AR. However, not all patients who developed de novo anti-HLA Abs experienced AR. Further, the identification of Abs did not predate the development of a clinically evident rejection in any of the patients who developed both de novo anti-HLA Abs and rejection. Thus, detection of anti-HLA Abs did not identify patients at risk for rejection who had not yet already manifest clinical signs of rejection. Importantly, although de novo Abs were only found in 11 of 70 patients, the other 59 patients were informative, by virtue of the fact that they did not develop any de novo Abs while they were tested during the first posttransplant year. These findings do not support routine testing for de novo anti-HLA Abs in nonsensitized patients to identify of patients at risk for rejection.
Our findings should not necessarily be interpreted as evidence against the humoral hypothesis of rejection. The fact that de novo anti-HLA Abs were present in serum samples obtained at the time of renal biopsy for clinically suspected rejection in three of the six patients who developed both Abs and rejection suggests the importance of humoral mechanisms in the early pathogenesis of rejection. Notably, none of the early AR episodes that predated the development of de novo anti-HLA Abs or that triggered testing and detection of de novo anti-HLA Abs were C4d positive or had other pathological signs suggestive of humoral rejection. These findings are consistent with those of Zhang et al. (1) who reported an association between de novo DSA and acute cellular rejection in a prospective study of patients at risk for rejection and suggest that anti-HLA Abs may be involved in the pathogenesis of rejection episodes that demonstrate only pathological signs of acute cellular rejection. Indeed, it remains possible that Abs were present in the circulation before development of rejection but without sufficient lead time to permit detection before rejection became clinically evident or that Abs predated rejection but were present at levels below the threshold for detection of our assay possibly because of the presence of soluble HLA antigens that may have complexed with circulating anti-HLA Abs in the serum or because of binding of Abs to allograft (18–20). Whatever the mechanisms, our findings show that despite frequent testing using sensitive methods, serial testing for anti-HLA Abs did not identify patients at risk for rejection before development of clinical signs of rejection.
Although other prospective studies have shown that the presence of anti-HLA Abs after transplantation is associated with an increased risk of graft loss, few published studies have specifically addressed the question of screening for de novo anti-HLA Abs to predict AR. In a single-center Italian study, 120 deceased donor kidney transplant recipients between 1992 and 1997 who were CDC cross-match and flow cross-match negative before transplantation were prospectively screened for anti-HLA Abs on days 7, 14, 30, 60, 90, 120, 150, 180, 365 after transplantation by flow cytometry cross-match (17). All patients who were IgG positive by flow cross-match after transplantation went on to be tested by FlowPRA and the association of anti-HLA Abs with clinical AR episodes during the first two posttransplant years was determined. Overall, 24% developed DS anti-HLA Abs. Patients with anti-HLA Abs had a higher incidence of clinical AR episodes (62% vs. 13%), more allograft failure (34% vs. 1%), and higher creatinine levels at 2 years (2.5±1.3 vs. 1.7±0.5 mg/dL) compared with patients without anti-HLA Abs. In this study, 9 of 18 patients who became flow cross-match positive after transplantation and had rejection, DSA preceded any clinical manifestation of AR by 1 to 10 days. Because patients in this study were included on the basis of a flow cross-match against donor cells, it is possible that pretransplant anti-HLA Abs would have been identified in some of these patients if they had been screened using a solid-phase assay as in our study. Zhang et al. (1) studied a selected population of 49 adult renal allograft recipients transplanted between 2001 and 2003 at the University of California, Los Angeles who were at increased risk for rejection because of delayed graft function, previous transplantation, or poor HLA matching. Patients were prospectively tested at predefined time points for anti-HLA Abs using enzyme-linked immunosorbent assay and Luminex Flow Beads. All patients were CDC and flow T- and B-cell cross-match negative before transplantation except one patient who had a weakly positive T-cell flow cross-match. AHR and acute cellular rejection were diagnosed in 16.3% and 22.4% of patients. Of the eight patients diagnosed with AHR, three developed DSA before the AHR, whereas four developed DSA at the time of the diagnosis of AHR. DSA were not associated with development of acute cellular rejection. These patients were at higher immunological risk than our patients with 20 of 49 patients having anti-HLA Abs present before transplantation.
The strengths of this study include: 1) the exclusion of patients with preformed anti-HLA Abs before transplantation who are known to be at increased risk for rejection and in whom the interpretation of de novo anti-HLA Abs may be problematic because of concomitant changes in the levels of preformed anti-HLA Abs that are difficult to quantify with solid-phase assays, 2) testing for de novo anti-HLA Abs at predefined time points in all patients by a single technologist using a consistent, sensitive, and reproducible methodology, 3) protocol-driven allograft biopsies performed at implantation, 3 and 12 months after transplantation and at the time of first anti-HLA Ab detection reduced the possibility that subclinical allograft rejection was not detected. In addition, all clinically suspected ARs were confirmed by renal biopsy, and all patients had antibody determination at the time of any allograft biopsy and ongoing testing after a biopsy-confirmed rejection minimizing the risk that low levels of circulating anti-HLA Abs were not detected. Readers should consider whether the findings of our study would be applicable to their practice. The study population was representative of our patient population and included mainly living donor recipients, a relatively high proportion of Asian patients, but no African American patients. Similarly, the immunosuppressive protocols used included the familiar triple immunosuppressive regimen, but patients were also treated with a rapid steroid elimination, and a small number of patients were enrolled in a clinical trial of FTY720. The fact that FTY720 will likely not be used in the future does not dilute our findings. Indeed, because de novo Abs were more frequent in FTY720-treated patients, the exclusion of such patients from our analysis would have resulted in four fewer patients with de novo anti-HLA Abs, and even stronger evidence that Ab monitoring in low-risk patients is likely unnecessary. According to our HLA laboratory protocol at the time of study testing, Abs were defined as present or absent only, and quantification was not performed. However, because Abs were only detected at and not before the time of rejection, the quantification of Ab would not have added any additional predictive value in this particular group of subjects. Of note, the overall incidence and timing of anti-HLA Abs in our study were consistent with that reported in other published studies using similar methods of antibody detection (7, 17). Patients were not tested for DP and DQ Abs or non-HLA Abs (21, 22), which may be associated with AR. Finally, the relatively small sample size may be considered a limitation of our study. However, to our knowledge, no other study using the same rigorous methodology has been conducted to address this issue. In summary, we found an association between development of de novo anti-HLA Abs and rejection. However, the anti-HLA Abs were identified after or at the same time as a clinically evident rejection episode, and in no instance did de novo anti-HLA Abs identify a patient who had not demonstrated clinical evidence of rejection. Therefore, our findings do not support routine monitoring of anti-HLA Abs after transplantation for the purposes of predicting AR in patients without evidence of anti-HLA Abs before transplantation.
The authors like to acknowledge the contributions of Teri Dahlgren, Caren Rose, and Gary Nussbaumer to this study.
1.Zhang Q, Liang LW, Gjertson DW, et al. Development of posttransplant antidonor HLA antibodies is associated with acute humoral rejection and early graft dysfunction. Transplantation
2005; 79: 591.
2.Crespo M, Pascual M, Tolkoff-Rubin N, et al. Acute humoral rejection in renal allograft recipients: I. Incidence, serology and clinical characteristics. Transplantation
2001; 71: 652.
3.Hourmant M, Cesbron-Gautier A, Terasaki PI, et al. Frequency and clinical implications of development of donor-specific and non-donor-specific HLA antibodies after kidney transplantation. J Am Soc Nephrol
2005; 16: 2804.
4.Mauiyyedi S, Pelle PD, Saidman S, et al. Chronic humoral rejection: Identification of antibody-mediated chronic renal allograft rejection by C4d deposits in peritubular capillaries. J Am Soc Nephrol
2001; 12: 574.
5.Terasaki PI. Humoral theory of transplantation. Am J Transplant
2003; 3: 665.
6.Mao Q, Terasaki PI, Cai J, et al. Extremely high association between appearance of HLA antibodies and failure of kidney grafts in a five-year longitudinal study. Am J Transplant
2007; 7: 864.
7.Terasaki PI, Ozawa M. Predicting kidney graft failure by HLA antibodies: A prospective trial. Am J Transplant
2004; 4: 438.
8.Burns JM, Cornell LD, Perry DK, et al. Alloantibody levels and acute humoral rejection early after positive crossmatch kidney transplantation. Am J Transplant
2008; 8: 2684.
9.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.
10.Pei R, Wang G, Tarsitani C, et al. Simultaneous HLA class I and class II antibodies screening
with flow cytometry. Hum Immunol
1998; 59: 313.
11.One Lambda. FlowPRA specific antibody detection test. Available at: http://126.96.36.199/pdffiles/productsheets/FlowSpec_ps_r200711.pdf
. 2009. Accessed May 1, 2009.
12.One Lambda. FlowPRA single antigen. Available at: http://188.8.131.52/pdffiles/productinserts/flspec_new_pi_EN.pdf
. 2009. Accessed May 1, 2009.
13.Herzenberg AM, Gill JS, Djurdjev O, et al. C4d deposition in acute rejection
: An independent long-term prognostic factor. J Am Soc Nephrol
2002; 13: 234.
14.Racusen LC, Solez K, Colvin RB, et al. The Banff 97 working classification of renal allograft pathology. Kidney Int
1999; 55: 713.
15.Tedesco-Silva H, Pescovitz MD, Cibrik D, et al. Randomized controlled trial of FTY720 versus MMF in de novo renal transplantation. Transplantation
2006; 82: 1689.
16.Levey A, Greene T, Kusek J, et al. A simplified equation to predict glomerular filtration rate from serum creatinine. J Am Soc Nephrol
2000; 11: 155A.
17.Piazza A, Poggi E, Borrelli L, et al. Impact of donor-specific antibodies on chronic rejection occurrence and graft loss in renal transplantation: Posttransplant analysis using flow cytometric techniques. Transplantation
2001; 71: 1106.
18.Akalin E, Pascual M. Sensitization after kidney transplantation. Clin J Am Soc Nephrol
2006; 1: 433.
19.Martin L, Guignier F, Bocrie O, et al. Detection of anti-HLA antibodies
with flow cytometry in needle core biopsies of renal transplants recipients with chronic allograft nephropathy. Transplantation
2005; 79: 1459.
20.Adeyi OA, Girnita AL, Howe J, et al. Serum analysis after transplant nephrectomy reveals restricted antibody specificity patterns against structurally defined HLA class I mismatches. Transpl Immunol
2005; 14: 53.
21.Sumitran-Holgersson S, Wilczek HE, Holgersson J, et al. Identification of the nonclassical HLA molecules, mica, as targets for humoral immunity associated with irreversible rejection of kidney allografts. Transplantation
2002; 74: 268.
22.Alvarez-Marquez A, Aguilera I, Gentil MA, et al. Donor-specific antibodies against HLA, MICA, and GSTT1 in patients with allograft rejection and C4d deposition in renal biopsies. Transplantation
2009; 87: 94.