Approximately 20% of patients on the kidney transplantation waiting list have a history of kidney transplant failure, and the number of patients with allograft failure returning to dialysis has increased in recent years (1). There is a growing concern about the outcomes and appropriate management of patients after transplant failure because mortality rates are greater than those observed in active transplantation patients and in patients listed for primary transplantation (2). High rates of hospitalization and sepsis after transplant failure have been noted (3). Moreover, inflammation related to the retained allograft may add to morbidity and mortality after transplant failure. Symptoms of late rejection, including fever, pelvic pain, and hematuria, have long been recognized in patients returning to dialysis therapy (4). Such symptoms tend to resolve after transplant nephrectomy, but transplant nephrectomy has been associated with alloantibody sensitization (5). A survival benefit is gained with retransplantation (6), but morbid events and sensitization diminish the likelihood of retransplantation.
We observed several patients who experienced overt rejection symptoms after returning to dialysis therapy. Many patients required hospitalization and allograft nephrectomy for such symptoms and were found subsequently to be highly sensitized. We hypothesized that nephrectomy and alloantibody sensitization were direct consequences of weaning immunosuppression after allograft failure. We sought to determine whether patients would develop sensitization after weaning even in the absence of allograft nephrectomy for cause. An analysis of our findings with such patients demonstrated that weaning of immunosuppression was an independent predictor of high sensitization after controlling for allograft nephrectomy, human leukocyte antigen (HLA) matching, and other factors.
From a cohort of 300 patients with a history of kidney transplant failure between January 1998 and January 2010, 119 were included in the analysis of late panel reactive antibody (PRA). Demographic data describing the study cohort are shown (see Table, SDC, http://links.lww.com/TP/A692). Patients in the study had a mean±SD age of 42±13 years and a median allograft survival time of just more than 7 years. More than one third of patients were African American, and approximately three fourths received kidneys from deceased donors. Most patients were relisted for transplantation, which prompted late PRA testing. Median PRA rose from 0% before transplantation to 57% and 63% by late PRA testing for classes I and II antibodies, respectively (P<0.001). Ninety-five patients (80%) weaned immunosuppression, and 39 patients (33%) underwent allograft nephrectomy before late PRA testing (see Table, SDC, http://links.lww.com/TP/A692).
Risk factors for sensitization after transplant failure were analyzed. For the time frame studied (6–24 months after transplant failure), there was no correlation with time from failure and late PRA by class I or II (r=0.074). Figure 1 shows the distribution of late PRA data by classes I and II antibodies for all 119 patients. Of the patients, 56% were highly sensitized (PRA ≥ 80%) against class I or II antigens, whereas 27% were highly sensitized against both classes I and II. Alternatively, just 16% of the patients maintained a 0% PRA against both classes I and II antigens.
We found a correlation with HLA matching from the original transplant and late PRA (r=0.202, P=0.002). This correlation was strengthened when the analysis was limited to the 95 patients who weaned immunosuppression before late PRA testing (r=0.352, P<0.001). Of the patients who weaned immunosuppression, those with an HLA mismatch of two or fewer with the original transplant showed a median late PRA level of 12% versus a median late PRA level of 90% in patients with three or more HLA mismatches (P<0.001). Patients who did not wean immunosuppression were less likely to become sensitized regardless of HLA matching. In the 24 patients who maintained immunosuppression, the mean number of HLA mismatches from the original transplant was 3.8±1.7; however, only 2 (8%) of the 24 patients became highly sensitized by late PRA testing. Overall, there was no correlation with HLA mismatching and late PRA testing in this cohort (r=−0.150). In contrast, 65 (68%) of the 95 patients who weaned immunosuppression became highly sensitized by class I or II antibodies (P<0.001).
Consistent with previous studies, history of nephrectomy for cause also correlated with sensitization after transplantation. Seventy-nine percent of patients who underwent nephrectomy before late PRA were highly sensitized compared with 38% of patients without history of nephrectomy (P<0.001). However, weaning of immunosuppression predicted the need for nephrectomy. None of the 24 patients maintained on immunosuppression with calcineurin inhibitor therapy underwent nephrectomy, whereas 39 (41%) of the 95 patients who weaned immunosuppression underwent nephrectomy for cause before late PRA testing (P<0.001). These latter patients had discontinued calcineurin inhibitor and antimetabolite therapies, although approximately half (20/39) remained on low-dose prednisone monotherapy at the time of nephrectomy, which occurred at 3.8±2.1 months after failure. We examined the rates of sensitization in patients who weaned immunosuppression stratified by a history of nephrectomy (Table 1). In these 95 patients, there was a higher percentage of those with class I PRA greater than 80% among patients who required nephrectomy (72% vs. 40%, P=0.002). For class II PRA, there was no difference between patients with and without a history of nephrectomy (46% vs. 55%, P=not significant). Overall, there was a trend for a higher percentage of highly sensitized patients for class I or II in the subset who underwent nephrectomy. However, high sensitization was common (62%) in patients who weaned immunosuppression in the absence of nephrectomy (Table 1).
Of the 24 patients maintained on immunosuppression, 16 presented with a functioning pancreas transplant requiring the continuation of immunotherapy. Because of patient or physician choice, the remaining eight patients maintained immunosuppression. There were no differences in age, gender, graft survival, or donor source between groups, but patients who maintained immunosuppression were less likely to be African American (8% vs. 43% of patients weaned, P<0.001). Because most patients who maintained immunosuppression after failure received kidney-pancreas (KP) transplants, we analyzed the risk of late sensitization specifically in KP recipients (n=21). None of the 16 KP patients who maintained immunosuppression became highly sensitized. Alternatively, four of five KP patients who weaned immunosuppression after loss of both kidney and pancreatic allograft function became highly sensitized against class I or II antigens (P<0.001 for the comparison; Fig. 2). Sensitization developed despite the absence of nephrectomy in all of these patients.
Variables associated with the risk of becoming highly sensitized (PRA I or II >80%) were examined in a logistic regression model (Table 2). By univariate analysis, longer allograft survival and KP transplantation were associated with lower risk. Alternatively, African American ethnicity, greater HLA mismatch status, weaning of immunosuppression, and history of allograft nephrectomy were all associated with a greater risk of sensitization. In the multivariate model (Table 2), African American ethnicity and HLA mismatch remained significant correlates with high sensitization. The weaning of immunosuppression also maintained a positive association (odds ratio, 14.342; 95% confidence interval: 2.334–88.144; P=0.004), whereas history of nephrectomy did not.
It was unclear as to what percentage of patients became highly sensitized before allograft failure as opposed to after allograft failure. To better address this question, flow PRA data were collected from the time of failure (±30 days) in patients still on immunosuppression with calcineurin inhibitor therapy at the time of collection. From the cohort of patients who later weaned immunosuppression after failure, 28 had PRA data at the time of failure. Median PRA at the time of failure in these 28 patients was 7% and 0% for classes I and II, respectively. Median PRA increased to 52% and 82% for classes I and II, respectively, by late PRA in this cohort (P<0.001 for both comparisons). Of 28 patients, 6 (21%) were highly sensitized at the time of failure. After weaning immunosuppression, an additional of 13 patients (68% overall) became highly sensitized (P<0.001 vs. the time of transplantation). Of the 13 patients who became highly sensitized, only 4 had undergone a transplant nephrectomy before late PRA testing. The distribution of PRA at failure and at 6 to 24 months after failure is shown in Figure 3.
Retransplantation and Survival
We analyzed rates of retransplantation and survival comparing patients who maintained or weaned immunosuppression after failure. There was a trend for a higher rate of retransplantation (46% vs. 29%) and a shorter time from relisting to retransplantation (median, 17 [range, 7–55] vs. 36 [range, 3–72] months) in patients who maintained immunosuppression. Mortality rates were similar between groups, with a 33% mortality rate at a median of 33 months after failure in patients who maintained immunosuppression versus a 25% rate at a median of 39 months in patients who weaned. Almost all deaths occurred in patients on dialysis, and patients who underwent retransplantation showed a low mortality rate overall (5%). There were two malignancy-related deaths in the maintenance group, which were potentially related to immunosuppressive therapy (melanoma and central nervous system lymphoma). Alternatively, most deaths in the withdrawal group were related to cardiac arrest or sudden death in sensitized patients on the waiting list.
In this analysis, we sought to determine the rate and risk of alloantibody sensitization after kidney allograft failure. We found a high rate of antibody sensitization at 6 to 24 months after failure, with 56% of patients highly sensitized against class I or II antigens at this interval. Of the patients studied, 80% were relisted for transplantation after failure, and the high rate of sensitization represented a significant barrier to retransplantation. We found that a decreased HLA matching at the time of transplantation was an independent predictor of greater sensitization, as previously shown (7). However, this correlation only pertained to patients who weaned immunosuppression, and weaning of immunosuppression was highly predictive of sensitization after controlling for allograft nephrectomy and other factors.
Allograft nephrectomy was previously shown to correlate with sensitization after transplant failure (5, 8, 9), and we similarly found an independent association with nephrectomy and class I PRA after failure. However, even in the absence of nephrectomy, most patients who weaned immunosuppression became highly sensitized. Furthermore, allograft nephrectomy was required exclusively in patients who weaned immunosuppression. We therefore conclude that weaning of immunosuppression triggered immune activation and broad HLA antibody formation. After weaning, a subset of patients developed rejection severe enough to cause symptoms such as fever, allograft pain, and hematuria, prompting medical attention and transplant nephrectomy. It is not surprising that nephrectomy correlated with PRA because most patients who required nephrectomy presented with overt rejection before surgery.
Many of the patients maintained on immunosuppression had a functioning pancreas transplant, necessitating maintenance immunotherapy despite kidney failure. Alternatively, KP recipients with failure of both organs showed a high rate of sensitization after weaning immunosuppression. Although this subanalysis was limited by a small number of patients, it seemed that KP recipients were at similar risk as the remaining cohort for sensitization after weaning. Our results are similar to recently published data showing the absence of sensitization after failure in 11 patients who maintained immunosuppression early after allograft failure (10). We expand on that data by demonstrating low and stable HLA antibody status in 24 patients at 6 to 24 months after failure.
One potential strategy to avoid de novo alloantibody sensitization would thus be to maintain immunosuppression after transplant failure, particularly in patients seeking another transplantation. Patients likely to undergo retransplantation quickly by means of living donation or early relisting may be reasonable candidates for whom to continue immunotherapy. Furthermore, maintaining immunosuppression after failure has been advocated as a method to maintain residual renal function in patients on dialysis therapy (11). Immunosuppression would, however, increase infectious risk, and there is already a high rate of death related to sepsis after transplant failure (2, 3). Smak Gregoor et al. (12) showed a greater risk of infection and death in patients who were maintained on immunosuppression early after transplant failure compared with patients who had weaned further out from failure. Infectious mortality after failure was shown to be highest in the first few months during the transition to dialysis therapy (13) and may relate to maintenance immunosuppression along with other risk factors such as hemodialysis catheters (3). The risk of malignancy may also increase with prolonged immunosuppressive therapy after failure, and we observed two malignancy-related deaths in the subgroup of patients who maintained immunosuppression. Whereas the rate of retransplantation was higher in this subgroup, mortality rates were similar between groups.
An alternative approach may be to perform an early allograft nephrectomy while still on immunosuppression, and this strategy has been advocated in cases where kidney allograft survival is less than 1 year (14). While nephrectomy for cause after transplant failure is associated with sensitization, it remains to be shown whether elective nephrectomy in patients on immunotherapy may conversely prevent sensitization by removing the stimulus for late rejection after weaning. Multiple studies have analyzed the role of nephrectomy in alloantibody sensitization, and some have theorized, without substantive data, that the allograft may serve as an adsorptive sink for antibodies (9). Khakhar et al. (5) found a significant increase in PRA within 6 months of nephrectomy compared with baseline levels, but a statistical difference was only seen in patients with early nephrectomy performed within 6 months of transplantation, suggesting that patients further out from transplant failure were already sensitized before nephrectomy.
Early nephrectomy may also be beneficial in patient well-being on dialysis therapy. The retained kidney allograft has been identified as a source of inflammation that may contribute to morbidity and mortality (15). One study of dialysis patients with transplant failure identified increased markers of inflammation and anemia that resolved after allograft nephrectomy (16). Hypoalbuminemia and markers of inflammation have long been associated with worse survival on dialysis (17), and a recent registry analysis described a more adjusted survival rate in patients who had undergone allograft nephrectomy (18). Although prospective trials are lacking, such reports are leading the transplantation community to question whether the retention of a failed allograft is more detrimental than beneficial.
There are limitations to this study. First, there is an absence of information about blood transfusions after allograft failure because many patients were treated in outside facilities. Scornik and Kriesche (10) recently reported on an association between blood transfusions and sensitization after posttransplant failure. Transfusions likely contribute to the risk of late antibody sensitization because anemia is common after transplant failure (16). This study is also limited by the smaller number of patients relative to recent registry analyses, although it provides details on weaning of immunosuppression unavailable in the registry data. Finally, this analysis is retrospective and has limitations inherent with a retrospective study. Prospective randomized trials are needed to examine the risks and benefits of weaning versus maintaining immunosuppression and the risks and benefits of elective nephrectomy after allograft failure.
In conclusion, we found that high levels of antibody sensitization were common at 6 to 24 months after kidney transplant failure, and sensitization rates increased markedly from the time of failure in patients who weaned immunosuppression. Weaning of immunosuppression was an independent risk factor for both classes I and II PRA elevation after controlling for allograft nephrectomy, HLA matching, and other factors. Many patients with failed transplants remain on the waiting list with a high PRA that developed only after returning to dialysis therapy. The transplantation community needs to develop new strategies to avoid late rejection and sensitization after kidney allograft failure.
MATERIALS AND METHODS
Patients and Data
After approval from the institutional review board, we retrospectively analyzed the outcomes of 623 patients at University Hospitals Case Medical Center with kidney transplant failure occurring between January 1998 and December 2009. Patients who died with a functioning graft or within 30 days of failure (n=315) were excluded from analysis. Patients with allograft thrombosis and allograft nephrectomy within 72 hours of transplantation (n=8) were also excluded. The remaining 300 consecutive patients with allograft failure were analyzed and included in the study group if they underwent PRA testing at 6 to 24 months after failure (late PRA). Patients who underwent retransplantation within 6 months of failure (n=25) were excluded from the analysis, and late PRA data were available in 131 remaining patients. Of these patients, 12 showed a cytotoxic PRA level for class I or II of 20% or more at the time of the original kidney transplant. To study a patient population with low or no sensitization at the time of transplantation, these 12 patients were excluded, leaving 119 patients in the analysis. This final cohort included 21 patients who had received a combined KP transplantation and who experienced failure of the kidney alone (n=16) or of both kidney and pancreas (n=5).
Demographic data were collected, including age at the time of allograft failure, gender, African American ethnicity, previous donor status (living vs. deceased), KP transplantation, allograft survival (months), relisting status, retransplantation status, weaning history, and history of transplant nephrectomy. Weaning of immunosuppression before late PRA testing was defined as the elimination of all immunosuppressive therapy. All patients who weaned immunosuppression had done so by 120 days after failure. All patients who maintained immunosuppression at the time of late PRA testing were on calcineurin inhibitor therapy in addition to antimetabolite or corticosteroid therapy. Finally, a history of allograft nephrectomy before late PRA testing was recorded. All nephrectomies were performed for cause and were related to symptomatic rejection or infection after failure. Because patients on dialysis had potentially received blood transfusions in facilities outside of dialysis units, blood transfusion data were incomplete and not included in the analysis.
All PRA data were obtained from the Human Leukocyte Antigen Laboratory at University Hospitals Case Medical Center dating back to 1980. PRA measurements before kidney transplantation were measured using a cytotoxic assay, whereas PRA at the time of failure and late PRA data were derived exclusively from flow cytometric analyses using FlowPRA beads (One Lambda Inc., Canoga Park, CA). PRA values were recorded if available in patients at the time of allograft failure ±30 days while still on immunosuppressive therapy. All late PRA data were collected in patients who were on renal replacement therapy. If patients underwent repeat PRA testing at 6 to 24 months after failure, the first PRA test measured during that time interval was used in the analysis. Class I or II flow PRA of 80% or more was used to define patients as highly sensitized after allograft failure.
Baseline demographic data between the study group and the remaining cohort of transplant failure patients were analyzed using Student’s t test for continuous variables and Pearson chi-square test for dichotomous variables. A Spearman rank correlation test was used to analyze the relationship between late PRA I and II and time from failure and PRA and HLA matching. A Mann-Whitney U test was used to compare late PRA values by HLA mismatch at transplantation. Univariate and multivariate logistic regression models were used to analyze variables associated with high sensitization by late PRA. The multivariate model included variables associated with sensitization based on P≤0.10. A Wilcoxon rank sum test was used to compare PRA values at the time of transplant failure and late PRA in a subset of patients. A two-sided P value less than 0.05 was considered to indicate statistical significance. All analyses were performed using SPSS version 18 (SPSS, Chicago, IL).
1. U.S. Renal Data System. USRDS 2008 Annual Data Report
. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases 2008.
2. Kaplan B, Meier-Kriesche HU. Death after graft loss: an important late study endpoint in kidney transplantation
. Am J Transplant
2002; 2: 970.
3. Johnston O, Zalunardo N, Rose C, et al.. Prevention of sepsis during the transition to dialysis may improve the survival of transplant failure patients. J Am Soc Nephrol
2007; 18: 1331.
4. Silberman H, Fitzgibbons TJ, Butler J, et al.. Renal allografts retained in situ after failure. Arch Surg
1980; 115: 42.
5. Khakhar AK, Shahinian VB, House AA, et al.. The impact of allograft nephrectomy on percent panel reactive antibody and clinical outcome. Transplant Proc
2003; 35: 862.
6. Ojo A, Wolfe RA, Agodoa LY, et al.. Prognosis after primary renal transplant failure and the beneficial effects of repeat transplantation: multivariate analyses from the United States Renal Data System. Transplantation
1998; 66: 1651.
7. Meier-Kriesche HU, Scornik JC, Susskind B, et al.. A lifetime versus a graft life approach redefines the importance of HLA matching in kidney transplant patients. Transplantation
2009; 88: 23.
8. Sumrani N, Delaney V, Hong JH, et al.. The influence of nephrectomy of the primary allograft on retransplant graft outcome in the cyclosporine era. Transplantation
1992; 53: 52.
9. 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.
10. Scornik JC, Kriesche HU. Human leukocyte antigen sensitization after transplant loss: timing of antibody detection and implications for prevention. Hum Immunol
2011; 72: 398.
11. Jassal SV, Lok CE, Walele A, et al.. Continued transplant immunosuppression
may prolong survival after return to peritoneal dialysis: results of a decision analysis. Am J Kidney Dis
2002; 40: 178.
12. Smak Gregoor PJ, Zietse R, van Saase JL, et al.. Immunosuppression
should be stopped in patients with renal allograft failure. Clin Transplant
2001; 15: 397.
13. Knoll G, Muirhead N, Trpeski L, et al.. Patient survival following renal transplant failure in Canada. Am J Transplant
2005; 5: 1719.
14. Bennett WM. The failed renal transplant: in or out? Semin Dial
2005; 18: 188.
15. Ayus JC, Achinger SG. At the peril of dialysis patients: ignoring the failed transplant. Semin Dial
2005; 18: 180.
16. López-Gómez JM, Pérez-Flores I, Jofré R, et al.. Presence of a failed kidney transplant in patients who are on hemodialysis is associated with chronic inflammatory state and erythropoietin resistance. J Am Soc Nephrol
2004; 15: 2494.
17. Kaysen GA. The microinflammatory state in uremia: causes and potential consequences. J Am Soc Nephrol
2001; 12: 1549.
18. Ayus JC, Achinger SG, Lee S, et al.. Transplant nephrectomy improves survival following a failed renal allograft. J Am Soc Nephrol
2010; 21: 374.