By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Please be advised that we experienced an unexpected issue that occurred on Saturday and Sunday January 20th and 21st that caused the site to be down for an extended period of time and affected the ability of users to access content on Wiley Online Library. This issue has now been fully resolved. We apologize for any inconvenience this may have caused and are working to ensure that we can alert you immediately of any unplanned periods of downtime or disruption in the future.
The purpose of this study was to determine the relationships between acute rejection, anti-major histocompatibility complex (MHC) class I and/or class II-reactive alloantibody production, and chronic rejection of renal allografts following kidney or simultaneous kidney-pancreas transplantation. Sera from 277 recipients were obtained pretransplant and between 1 month and 9.5 years post-transplant (mean 2.6 years). The presence of anti-MHC class I and class II alloantibodies was determined by flow cytometry using beads coated with purified MHC molecules. Eighteen percent of recipients had MHC-reactive alloantibodies detected only after transplantation by this method. The majority of these patients produced alloantibodies directed at MHC class II only (68%). The incidence of anti-MHC class II, but not anti-MHC class I, alloantibodies detected post-transplant increased as the number of previous acute rejection episodes increased (p = 0.03). Multivariate analysis demonstrated that detection of MHC class II-reactive, but not MHC class I-reactive, alloantibodies post-transplant was a significant risk factor for chronic allograft rejection, independent of acute allograft rejection. We conclude that post-transplant detectable MHC class II-reactive alloantibodies and previous acute rejection episodes are independent risk factors for chronic allograft rejection. Implementing new therapeutic strategies to curtail post-transplant alloantibody production, and avoidance of acute rejection episodes, may improve long-term graft survival by reducing the incidence of chronic allograft rejection.
For kidney transplant recipients, it is generally accepted that acute allograft rejection is the greatest risk factor for subsequent development of chronic allograft rejection (1–7). Many transplant centers have reported a significant correlation between acute allograft rejection and post-transplant alloantibody production, generally detected within the first 6 months after engraftment (8–11). Further, several studies have noted an association between late post-transplant alloantibody production (> 1 years) and chronic graft loss (12–15). However, the complex relationships between acute allograft rejection, alloantibody production, and chronic allograft rejection have not been directly addressed.
McKenna et al. (16) recently reviewed the current studies dealing with the relationship between the post-transplant production of major histocompatibility complex (MHC)-reactive alloantibodies and acute allograft rejection or graft loss. In most of these studies patient sera were obtained within the first 3–6 months after transplantation and tested for the presence of MHC class I-reactive alloantibodies. All of these studies found a significant association between the presence of MHC class I-reactive alloantibodies and the development of acute rejection (8–11, 17, 18) or subsequent graft loss (9, 12, 13, 17). In contradistinction, fewer studies tested for the presence of MHC class II-reactive alloantibodies. When tested, these alloantibodies were detected in conjunction with MHC class I-reactive alloantibodies, and were also associated with the development of acute rejection (11, 18). There are no studies that concomitantly examine the role of alloantibodies and acute rejection episodes on the subsequent development of chronic allograft rejection.
Determination of the relationships between post-transplant MHC-reactive alloantibodies and acute or chronic allograft rejection has been complicated by the use of differing methods for alloantibody detection. The conventional methods of complement-dependent lymphocytotoxicity (CDC) and anti-human globulin modified complement-dependent lymphocytotoxicity (AHG-CDC) only detect complement fixing MHC class I-reactive antibody binding to T cells or MHC class II-reactive antibody binding to B cells. However, a positive result due to antibodies that bind non-MHC molecules on the T- or B-cell surface cannot always be excluded. The flow cytometric method of antibody detection is more sensitive than the CDC or AHG-CDC methods, and detects anti-MHC class I or class II-reactive immunoglobulins (IgG) without regard to their ability to fix complement. But, like the CDC and AHG-CDC methods, it cannot exclude a positive result due to detection of antibodies binding to non-MHC molecules on the T- or B-cell surface. The recent development of enzyme-linked immunosorbent assay (ELISA) methods, and most recently microparticle beads coated with purified MHC molecules for use by flow cytometry, have made it possible to test sera specifically for the presence of any anti-MHC class I or class II-reactive IgG, whether or not they fix complement.
This study used the newer flow bead analysis to determine if an association exists between the presence of post-transplant MHC-reactive alloantibodies and clinical outcome. We were specifically interested in the relationships between acute rejection, anti-MHC class I or class II alloantibodies, and chronic rejection of the renal allograft following kidney or simultaneous kidney-pancreas transplantation.
Materials and Methods
Two hundred and seventy-seven primary kidney or simultaneous kidney-pancreas transplant recipients were included in this study. By design, all patients in this study were primary transplant recipients. Re-transplant patients were excluded to simplify the alloantibody analysis. As per our routine clinical protocol for first transplant recipients, these patients were transplanted following a negative flow cytometric (n = 213) or CDC (n = 64) T-cell cross-match, but pretransplant B-cell cross-matches were not performed. The mean post-transplant follow-up time for the entire cohort of patients was 4.1 ± 2.6 years (range 1–16.7 years). One pretransplant (final cross-match) and one post-transplant serum sample were analyzed from each recipient. Of the post-transplant sera, approximately 10% were obtained from patients during hospital admission for post-transplant complications, 15% were obtained from a group of consecutively transplanted patients involved in a clinical study over a 2-year period, and 75% were obtained from patients during routine post-transplant evaluation. Post-transplant sera were obtained between 1 month and 9.6 years (mean 2.6 years, median 2.1 years) after transplantation. The incidence of detectable pre- or post-transplant alloantibodies was similar for cadaveric donor kidney, living donor kidney, and simultaneous kidney-pancreas recipients. The intent of this study was to determine the impact of graft-induced alloantibodies on post-transplant outcome. Thus, the 48 patients with detectable alloantibodies pretransplant were excluded from all post-transplant outcome analyses. The donor and recipient demographics for the remaining 229 patients are included in Table 1.
Table 1. : Donor/recipient demographics and clinical data for 229 adult kidney or simultaneous kidney-pancreas transplant recipients negative for MHC-reactive alloantibodies pretransplant
SD = standard deviation;
b patient number (percent) unless otherwise indicated;
LRD = living donor kidney transplant;
Cad = cadaveric donor kidney transplant;
SKP = simultaneous kidney-pancreas transplant;
pancreas graft loss = 7/53 or 13% of SKP recipients.
Of these 229 patients, 198 patients (86%) received induction therapy and 31(14%) received no induction. Induction therapy consisted of either anti-CD3 mAb (n = 112), polyclonal antilymphocyte antibodies (n = 35), or anti-IL2R mAb (n = 51). Prednisone was initiated at 2 mg/kg and was tapered to 0.15 mg/kg by 1 month post-transplant. For induction patients, oral cyclosporine was begun at 8 mg/kg/day when the serum creatinine fell below 2.5–3.0 mg/dL, whereas it was begun immediately when no induction agent was administered. The cyclosporine dose was adjusted to obtain appropriate levels or to avoid toxicity. Eleven patients received no additional immunosuppressant therapy. The remainder received either mycophenolate mofetil (n = 151), azathioprine (n = 59), or rapamycin (n = 8), which was adjusted as necessary for toxicity. The type of induction therapy did not influence the incidence of acute rejection, or detectable anti-MHC alloantibodies post-transplant. However, post-transplant acute rejection episodes and detectable alloantibodies were both decreased in patients who received mycophenolate mofetil together with prednisone and cyclosporine (p = 0.04 and p = 0.008, respectively, chi-square analysis).
All acute rejection episodes were biopsy proven prior to treatment. Of the first acute rejection episodes (n = 65), 33 were treated with OKT3 plus steroids, 12 were treated with anti-thymocyte globulin (ATG) plus steroids, and 20 were treated with steroids alone. Of the second acute rejection episodes (n = 32), 14 were treated with OKT3 plus steroids, 3 were treated with ATG plus steroids, and 15 were treated with steroids alone. The type of treatment for the first or second acute rejection episodes did not correlate significantly with the incidence of alloantibodies detected de novo post-transplant. The mean time from treatment of the first acute rejection episode to collection of patient sera for anti-MHC alloantibody analysis was 2.6 years (median of 2.0 years).
Chronic rejection was diagnosed in 21 patients who experienced a progressive loss in renal function unrelated to any other known causes. The mean serum creatinine at the time of testing for alloantibodies for these patients was 4.2 ± 2.0 mg/dL (range 1.8–7.9). Ten patients subsequently lost their grafts. The mean serum creatinine at last follow-up for the remaining 11 patients with functioning grafts was 5.4 ± 2.6 (range 2.7–10.7). All 21 patients were hypertensive, being treated with a mean of 3.5 anti-hypertensive agents (range 1–7). Significant proteinuria (> 500 mg/24 h) was present in 17 of 21 patients (mean 2690 ± 2384 mg/24 h). The majority of these patients (14/21) had the diagnosis of chronic rejection confirmed by biopsy. Of these 14 biopsies, 9 were determined to be grade III, 4 were grade II, and 1 was grade I chronic/sclerosing allograft nephropathy using the Banff 97 working classification. The patient with grade I allograft nephropathy had significant hypertension (treated with 3 antihypertensive agents) and proteinuria (1920 mg/24 h) and suffered progressive loss in renal function resulting in graft loss and return to dialysis.
Flow cytometric alloantibody analysis
A commercially available pool of microparticle beads coated with various purified MHC antigens of known specificity, was used according to manufacturer's instructions (FlowPRA, OneLambda, Canoga Park, CA, USA). Briefly, 20 µL of recipient sera was incubated with 5 µL of MHC class I plus 5 µL of MHC class II microparticle beads for 30 min at RT while shielded from visible light and gently agitated. The beads were washed twice with buffer, spun in a centrifuge at 10 000 r.p.m. for 2 min and the supernatant was discarded. Beads were re-suspended in 100 µL solution containing FITC-conjugated goat anti-human IgG and incubated for 30 min at RT, shielded from visible light and gently agitated. The wash step was repeated and the beads were re-suspended in 500 µL of wash buffer. Negative control serum using pooled sera from nontransfused males were similarly prepared. Samples were read with the aid of a Beckman Coulter XL2 flow cytometer. The fluorescence profile obtained with negative control sera was used as the baseline fluorescence. MHC class I and class II beads were readily distinguishable since they are fluorescent (excited at 488 nm and maximum emission at 580 nm) and have unique emission spectra. The positive/negative cut-off was empirically determined for each assay by setting a histogram region that excluded 98% of the peak obtained with the negative control serum. The median channel associated with this cut point was recorded for each assay. The percentage of positive reactivity was defined as the percentage of beads shifted to the right of the cut-off point. A shift of 6% or greater was considered significant. Representative results from two patients are shown in Figure 1, one patient with detectable MHC-class I reactivity only, and one patient with detectable MHC-class II reactivity only.
Due to concerns regarding the possibility of obtaining false positive flow bead results due to antibody binding to the bead material itself, rather than the MHC antigens bound to it, additional testing of patient sera against non-MHC coated beads was performed. We found that 95% of patients tested (35/37, 1 serum sample per patient) were nonreactive with the naked beads yet reactive to MHC class I and/or class II coated beads (≥ 6% bead fluorescence).
Student's t-test and Pearson chi-square test were used for statistical comparison of means (± SEM) and proportions between groups, respectively. Forward logistic regression was used for multivariate analysis of independent risk factors for chronic rejection. Kaplan–Meier product-limit estimate was used for the univariate analysis of graft survival time with group comparisons performed via the log-rank test.
Pretransplant detection of alloantibodies
Forty-eight of 277 patients (17.3%) had MHC class I and class II-reactive alloantibodies detectable by the flow bead method prior to transplantation (Table 2). Of these, 22 patients had alloantibodies that were MHC class I-reactive only, 12 that were MHC class II-reactive only, and 14 that were MHC class I and class II reactive. These antibodies were not detected during the pretransplant T-cell cross-match (either CDC (n = 213) or flow cytometric (n = 64) methods). Thus the anti-MHC class I alloantibodies were presumed to have been reactive to antigens not displayed by the donor graft. However, it is not known how many of the anti-MHC class II-reactive alloantibodies were graft-reactive, since pretransplant B cell cross-matches were not performed. A disproportionate percentage of antibody-positive patients were female (p = < 0.001, Chi-square analysis). There were no significant differences in age, race, or degree of MHC mismatch between antibody-positive and -negative recipients. In this group of patients, there was no correlation between pretransplant alloantibody status and the occurrence of acute or chronic rejection post-transplant (data not shown).
Table 2. : Pretransplant vs. post-transplant alloantibody status for 277 primary kidney or simultaneous kidney-pancreas recipients
tx = transplant;
n = number of patients.
Post-transplant detection of alloantibodies
As shown in Table 2, of the entire group of 277 patients in this study, 193 (70%) patients had no detectable MHC-reactive alloantibodies and 84 had detectable MHC-reactive antibodies at the time of post-transplant testing, which ranged from 1 month to 9.6 years. Of the 193 patients who had no detectable post-transplant MHC-reactive antibodies, only 14 had detectable alloantibodies pretransplant. Of the 84 patients who had detectable MHC-reactive antibodies after transplantation, 34 had detectable MHC-reactive alloantibodies pretransplant and 50 patients had no detectable MHC-reactive alloantibodies before transplantation. Thus, 18% of all patients (50/277) had MHC-reactive alloantibodies detected after, but not before, transplantation by the flow bead method. For the purposes of this manuscript, these 50 patients will be referred to as the de novo post-transplant detectable alloantibody patients. It should be noted that the time of serum collection post-transplant was similar for all 4 groups of patients in Table 2 (mean of collection time for the groups ranges from 3.6 to 4.2 years, p = 0.85 by analysis of variance). Further, there were no differences in the 4 groups of patients in Table 2 in regard to the type of donor organ(s) received, the type of induction therapy, or the type of treatment for the first or second acute rejection episodes. Finally, within each group the clinical outcome (kidney loss or chronic rejection) did not correlate with the type of donor organ(s) received, the type of induction therapy, or the type of treatment for the first or second acute rejection episodes. The 8-year renal allograft survival for the 50 de novo post-transplant detectable alloantibody patients was significantly worse than for the other patients (58% vs. 97–100%) (Figure 2). Interestingly, the majority of patients in this subgroup who lost their grafts did so between 2 and 8 years post-transplant, suggesting that chronic rejection, rather than acute rejection, may be the most significant cause of kidney loss in these patients. The poor graft survival for these patients prompted us to examine them in greater detail.
To simplify our studies relating de novo post-transplant alloantibody detection to clinical outcome, the 48 patients with detectable MHC-reactive alloantibodies pretransplant were excluded from all subsequent analyses (34 of these patients made detectable MHC-reactive alloantibodies post-transplant). Characterization of the MHC class I or class II reactivities for the 50 patients with de novo post-transplant alloantibodies revealed the following. Five patients (of 213 mismatched to their donor for MHC class I) made detectable alloantibodies directed only at MHC class I. Thirty-four patients (of 204 mismatched to their donor for MHC class II) made detectable alloantibodies directed only at MHC class II. Eleven patients (of 198 mismatched to their donor for both MHC class I and MHC class II) made detectable alloantibodies directed at both MHC class I and class II molecules (Table 3). These patients with MHC-reactive alloantibodies detected de novo post-transplant were significantly more likely to have a history of acute allograft rejection as compared to patients with no detectable alloantibodies (44% vs. 24%, p = 0.006 by Chi-square analysis). This association also exists if the 48 patients with detectable pretransplant MHC-reactive alloantibodies were included in the analysis (40% vs. 24%, p = 0.004). They also tended to be younger (mean of 42 vs. 45-year-old) and African-American (24% vs. 14.5%), but these findings were not statistically significant.
Table 3. : Post-transplant anti-MHC alloantibody reactivities for 229 patients with no detectable anti-MHC alloantibodies pretransplant
No. pts. = number of patients (% of all 229 patients).
Anti-MHC class I only
Anti-MHC class II only
Anti-MHC class I and class II
To determine the stability of detectable alloantibodies in our patients' sera over time, we evaluated paired serum samples for the presence of MHC-reactive alloantibodies that were obtained from the 229 patients with no flow bead detectable alloantibodies before transplantation (median of 370 days apart, range 43–3317 days). Multiple samples for comparison were available for 145 of these patients. The concordance for MHC class I-reactive alloantibodies between the 2 sera samples was 92% (2% gained and 6% lost reactivity), and for MHC class II-reactive alloantibodies it was 83% (10% gained and 7% lost reactivity) (data not shown). Thus, the presence or absence of detectable alloantibodies in these patients' sera was relatively stable over time.
Alloantibodies detected de novo post-transplant vs. acute rejection
The strong association noted above between acute rejection and alloantibodies detected de novo post-transplant was explored by comparing the percentages of patients with detectable alloantibodies stratified by their number of acute rejection episodes (0, 1, or > 1) (Table 4). There was a significant increase in the percentage of patients with de novo detectable alloantibodies post-transplant as the number of acute rejection episodes increased (17%, 24%, and 44% for 0, 1, and > 1, respectively). Similar results were obtained when the 48 patients with detectable pretransplant MHC-reactive alloantibodies were included in this analysis (26%, 31%, and 52% for 0, 1, and > 1, respectively). A subanalysis was performed comparing the percentage of patients who had de novo detectable alloantibodies post-transplant that were anti-MHC class I reactive only, anti-MHC class II reactive only, or anti-MHC class I plus class II reactive for patients with 0, 1, or > 1 acute rejection episodes. The percentage of patients who had MHC class II-reactive, or MHC class I- plus class II-reactive alloantibodies significantly increased as the incidence of treated acute rejection episodes increased (Table 4). This was not true for patients with only MHC class I-reactive alloantibodies. Even when the 48 patients with detectable pretransplant MHC-reactive alloantibodies were included in this subanalysis (34 of whom had detectable alloantibodies post-transplant), the results were similar. This suggests that even for these patients there is an association between acute rejection and the presence of detectable circulating alloantibodies.
Table 4. : Post-transplant MHC-reactive alloantibody status vs. the incidence of acute rejection for 229 patients with no detectable anti-MHC alloantibodies pretransplant
Alloantibodies detected de novo post-transplant vs. chronic rejection
In our transplant program, acute rejection is a risk factor for chronic rejection (1). As noted in this study, a history of acute rejection(s) increased the likelihood of de novo detectable alloantibodies post-transplant. The next studies were undertaken to determine the relative contributions of acute rejection and alloantibodies detected de novo post-transplant towards the development of chronic rejection. Initially, patients were stratified according to their acute rejection and MHC-reactive alloantibody status. This resulted in 4 groups of patients (acute rejection negative/antibody negative, acute rejection positive/antibody negative, acute rejection negative/antibody positive, and acute rejection positive/antibody positive) (Table 5). The incidence of chronic rejection was determined for each group. Chronic rejection was rare (4%) for patients with no acute rejection history and no detectable de novo MHC-reactive alloantibodies. A history of acute rejection and no detectable MHC-reactive alloantibodies and vice versa resulted in a similar incidence of chronic rejection (approximately 15%). The highest incidence of chronic rejection was noted for those patients with a history of acute rejection who also had detectable de novo MHC-reactive alloantibodies post-transplant (32%). By univariate analysis, the presence of detectable de novo anti-MHC class II alloantibodies and > 1 episode of acute rejection correlated best with the development of chronic rejection (Table 6). However, we had already noted a strong association between acute rejection(s) and detectable de novo post-transplant anti-MHC class II alloantibodies. Therefore, multivariate analysis was preformed to determine whether detectable post-transplant anti-MHC class I or class II alloantibodies were a risk factor for chronic rejection independent of acute rejection episodes (Table 6). A history of 2 or more treated acute rejection episodes was the best predictor of chronic rejection (p = 0.002). However, de novo detected post-transplant anti-MHC class II alloantibodies did not lose significance as a risk factor for chronic rejection in this model (p = 0.008), indicating it has predictive value independent from acute rejection. The recipient race (African-American) was also included in this analysis because that too was found to be a significant risk factor for chronic rejection. The above results comparing chronic rejection with a history of acute allograft rejection and detectable post-transplant alloantibodies (and their MHC reactivity) were similar when the 48 patients with detectable pretransplant MHC-reactive alloantibodies were included (data not shown).
Table 5. : Patient acute rejection and MHC-reactive alloantibody status vs. chronic rejection in 229 kidney or simultaneous kidney-pancreas transplant recipients with no detectable anti-MHC alloantibodies pretransplant
Table 6. : Variables predictive of chronic rejection by univariate and multivariate analysis. Variables not found to be significant risk factors for chronic rejection include donor age, race (African-American vs. all others), gender, type of transplant received, recipient age, or the degree of MHC class I or class II mismatch. Model significant to p < 0.001
The recent availability of a flow bead alloantibody detection method has enabled us to accurately determine the presence of anti-MHC class I and class II IgG antibodies in our transplanted patients. Using the flow bead method, 34 patients had MHC-reactive alloantibodies that were detectable pre- as well as post-transplant. By design, this method detects MHC-reactive antibodies, but does not determine donor-reactivity. Thus, for these patients the likelihood that some of their post-transplant alloantibodies were donor reactive remains uncertain. In contradistinction, for the 50 patients who had detectable MHC-reactive alloantibodies after but not before transplant, previously published data suggest that in > 90% of cases the alloantibodies are donor reactive (15). Within our own program, we analyzed post-transplant sera from a separate group of patients (n = 77) who had no detectable anti-MHC antibodies by the flow bead method prior to transplantation. When we compared the donor-specific ELISA results with the flow bead results, we found an 87% agreement for anti-MHC class I antibodies and an 86% agreement for anti-MHC class II antibodies between the results obtained by the two methods. Further, donor-specific ELISA testing of pretransplant sera for 67 patients who had no flow bead detectable pretransplant anti-MHC antibodies demonstrated that 100% lacked antidonor MHC class I and 96% lacked antidonor MHC class II antibodies (submitted for publication). Thus it appears that most alloantibodies are donor reactive when detected after, but not before, transplantation. Sensitizing events such as blood transfusions (undetermined in this study), pregnancies (2/50 patients in this study), and even infections can account for the post-transplant production of nondonor MHC-reactive alloantibodies in some of our patients. Because of the above observations, we selected patients who had flow bead detectable alloantibodies after, but not before, transplantation to investigate the relationship between alloantibodies and clinical outcome, assuming that in the majority of patients these alloantibodies were donor reactive. Analysis of a subset of these patients who had 2 sera samples analyzed a median of 370 days apart suggested that the presence or absence of detectable alloantibodies remained relatively stable over time.
A recent study using the same flow bead method to detect anti-MHC class I and class II alloantibodies after transplantation has recently been reported. The results were analyzed in regard to acute rejection and graft survival at 3 years. No relationship was noted between acute rejection and alloantibody production, but a correlation between anti-MHC class II alloantibody production and early (< 1 year) graft loss was reported (19). However, recipient sera were analyzed during the first month post-transplant and patient numbers in this study were limited. Our data found associations between post-transplant alloantibodies detected by flow bead analysis and the development of acute and chronic rejection. Importantly, we evaluated recipient sera that were obtained much later than 1 month after transplantation.
A high incidence of circulating MHC class I-reactive alloantibodies has been reported for patients experiencing acute rejection after transplantation (8–11, (c)17, (c)18). Acute rejection in alloantibody-positive patients has been noted to be more severe and associated with a poor graft survival (9). It is well accepted that T-cell help is necessary for B cells to produce antibodies against a new antigenic stimulus. Thus, the correlation between alloantibodies and a history of acute rejection, a clinical marker of T-cell sensitization, seen in this study was not surprising. However, the preponderance of alloantibodies reactive with MHC class II molecules, rather than MHC class I, was unexpected. There are at least two possible explanations for these results. First, there may be a high frequency of early graft loss in recipients who develop post-transplant anti-MHC class I antibodies in association with acute rejection in our transplant program, similar to that reported by Halloran et al. (9). Perhaps anti-MHC class I alloantibodies, unlike MHC class II-reactive alloantibodies, result in severe graft damage and early graft loss such that persistent MHC class I-reactive alloantibody production is incompatible with the longer-term graft survival needed for chronic rejection to develop. If true, this would create a selection bias towards anti-MHC class II antibody production when testing recipients who have functioning allografts later (> 1 year) post-transplant. Of the 22 patients we studied who lost their renal allograft, 7 were lost < 1 year post-transplant, and 4/7 had detectable anti-MHC class I alloantibodies. Thus, this explanation seems plausible. Second, alloantibody production may spread from MHC class I reactivity to MHC class II reactivity with time, especially if graft MHC class II expression is increased during episodes of graft inflammation such as occurs during episodes of acute rejection. Currently, we are serially testing our kidney and kidney-pancreas recipients after transplantation to determine if this indeed occurs. However, at the present time, the likelihood of this explanation cannot be ascertained.
We, as well as others, have noted that previous acute rejection episodes are the best predictors of the subsequent development of chronic allograft rejection (1–7). Acute rejection episodes correlated well with detectable de novo alloantibody production in the current study (Table 4). Thus, the significant correlation by univariate analysis between detectable anti-MHC class II alloantibodies and chronic rejection could simply be a reflection of their relationship to acute rejection. However, when including both acute rejection episodes and detectable de novo anti-MHC class II alloantibodies in a multivariate analysis, the presence of anti-MHC class II antibodies remained a statistically significant predictor of chronic rejection, independent of acute rejection. We hypothesize that in our transplant program, acute rejection and circulating anti-MHC class II alloantibodies reflect cellular and humoral allosensitization, both of which provide related, but independent accelerating factors for the development and progression of the chronic rejection process (20). Other studies have noted a strong correlation between the presence of anti-MHC antibodies (13) or antidonor alloantibodies (14) in the post-transplant sera of renal transplant recipients and increased rates of long-term graft loss, presumably due to chronic rejection. However, none have concomitantly examined the relative contributions of cellular (acute rejection) and humoral (alloantibody) sensitization.
Given the results obtained in this study, it would seem advisable to eliminate anti-MHC alloantibody production following kidney or simultaneous kidney-pancreas transplantation. In our program, this is especially true for anti-MHC class II alloantibodies. As one approach, we could match all recipients to their donors, especially for MHC class II. However, this would be impractical and place additional constraints on organ allocation. A more practical approach would be to implement immunosuppressive strategies that avoid acute rejection episodes in the greatest percentage of patients, as these are the patients at greatest risk for alloantibody production. Additionally, early identification and treatment of recipients who develop donor-reactive anti-MHC antibodies in an effort to terminate alloantibody production would be advantageous. One possible strategy might be treatment using a monoclonal anti-B-cell preparation (e.g. anti-CD20) plus rapamycin to block ongoing and future alloantibody production, respectively. Strategies such as this, designed to block alloantibody production, may have a significant role in pre-emptive or active therapy for chronic rejection in the future.
We would like to thank Donguan Xia and Maria Bellizzi for their technical assistance, Barbara Heckendorn for her clerical assistance, and Marsha Stalker and Sandra Dickens for their assistance in preparing this manuscript.
Grant Support: This study was supported by National Institutes of Health grants R29 AI40909 (amv), PO1-AI/HL40150, RO1-AI43578, RO1-HL61966, RO1-AI48623 (cgo), and in part by grant P30-CA16058 (cgo), National Cancer Institute, Bethesda MD.