Objectives: The impact of allograft nephrectomy on the outcome of a subsequent renal transplant is unclear. This study was conducted to assess the effects of the first allograft nephrectomy on outcomes of a second transplant.
Materials and Methods: This study included 118 patients who received a second transplant between 1994 and 2015. Before the second transplant, 59 patients did not undergo a first allograft nephrectomy (group A). Group B comprised 59 patients who had undergone a first allograft nephrectomy. We compared sensitization, acute rejection, and survival of the second graft between groups. The risk factors of a second graft loss were assessed.
Results: The first graft survival was significantly longer in group A than in group B (100.6 vs 3.7 months; P < .001). Prevalence of preformed donor-specific antibodies before the second allograft was similar between both groups (28.8% vs 39.0% for group A vs group B; P = .243). Numerically higher acute rejection rates occurred in group B than in group A (23.7% vs 15.3%; P = .245). In group A, graft survival rates at 1, 3, and 5 years were 93.0%, 87.0%, and 82.3% and were significantly higher than for group B (76.7%, 69.1%, and 62.5%; P < .05). On multivariate analysis, survival of the second graft was affected by acute rejection (hazard ratio = 2.24; 95% confidence interval, 1.10-4.45; P = .027) and the interval from first graft loss to second transplant (hazard ratio = 1.11; 95% confidence interval, 1.02-1.19; P = .008).
Conclusions: A first allograft nephrectomy was associated with inferior second graft survival. We recommend that recipients of second transplants should be considered as high risk if they had undergone prior allograft nephrectomy.
Key words : Acute rejection, Graft failure, Kidney retransplant
An increasing number of patients are waiting for a second kidney transplant after failure of their first graft.1 Graft survival for a second renal transplant has been shown to be significantly lower than that of first renal transplant irrespective of donor type.2 Renal retransplant (RRT) presents additional challenges for health care professionals such as whether an allograft nephrectomy should be performed in an asymptomatic failed allograft patient. Currently, there is no universal consensus about allograft nephrectomy prior to subsequent transplant.
Common indications for allograft nephrectomy are vascular complications, hemorrhage, rejection, and infection.3 However, allograft nephrectomy may increase the risk of unmasking anti-human leukocyte antigen (HLA) antibodies, with a consequently prolonged wait time and higher immunologic risk resulting in inferior graft survival.4-7 Moreover, RRT after allograft nephrectomy is associated with delayed graft function (DGF) and higher incidence of acute rejection.8-12 Other disadvantages of allograft nephrectomy include the risk of surgical complications, including blood transfusion and loss of residual renal function.3,13 On the other hand, an in situ failed graft has the potential risk of ongoing subclinical low-grade rejection, resulting in a chronic inflammatory state.14,15 Maintaining immunosuppression therapy may prevent an immunologic response against an in situ graft but will subject patients to a higher risk of infection.16
The aim of this study was to compare the graft outcomes of a second renal transplant following allograft nephrectomy with those who had an in situ failed graft.
Materials and Methods
We analyzed 118 patients who received a second transplant at the Royal London Hospital between November 1994 and January 2015. First transplants were performed between April 1972 and January 2012. Seventy-eight patients had the first transplant at our institution, and the remaining 40 had it performed in different hospitals. The follow-up period was defined as the time from the second transplant to patient death or the conclusion of this study. The median follow-up was 42.1 months (interquartile range [IQR], 9.8-74.7 mo). Patients were divided into 2 groups: patients with an in situ first graft at the time of retransplant (group A; n = 59) and patients who had undergone allograft nephrectomy before the second transplant (group B; n = 59).
Both groups were compared for patient survival and second transplant outcomes, which included graft survival, estimated glomerular filtration rate (eGFR), DGF, prevalence of donor-specific antibody (DSA), duration from first graft loss to second transplant, time from listing on transplant wait list for the second transplant, incidence of acute rejection, incidence of infection, and causes of graft loss.
Data collection was performed retrospectively. Clinical records from patients who underwent second transplants were collected. The demographic and medical data were obtained from a validated database, Filmmaker Pro 5.5v1 (FileMaker Inc, Santa Clara, CA, USA), which was in use in our center. These data included recipient sex, age at time of the second transplant, recipient ethnicity, and first transplant history (donor type, graft survival time, time from first graft loss to listing on transplant wait list, time from listing to transplant wait list to second transplant, and time from first graft loss to second transplant).
Data about the second transplant were collected and included donor type, donor sex, donor age, preemptive transplant, HLA mismatch, type of immunosuppression (including induction therapy), DSA, DGF, incidence of acute rejection and infection, causes of graft loss, eGFR, graft survival, and patient survival. Acute rejection was diagnosed and categorized according to the Banff classification.17 Estimated glomerular filtration rate was calculated using the Modification of Diet in Renal Disease. Delayed graft function was defined as a requirement for dialysis in the first week after transplant. Graft failure was defined as a return to permanent dialysis.
All study patients received induction therapy with antithymocyte globulin or anti-CD25 monoclonal antibody on day 0 and on day 4 posttransplant.
The maintenance immunosuppression consisted of calcineurin inhibitor with either cyclosporine or tacrolimus, an antiproliferative agent (mycophenolate mofetil or azathioprine), and a tapering dose of steroid.
HLA and donor-specific antibody detection and immunologic analyses
Screening of HLA and panel reactive antibody (PRA) had been performed by complement-dependent cytotoxicity crossmatch from 1994 until 2000. Since 2000, enzyme-linked immunosorbent assay was the main method, and Luminex assays (Luminex, Austin, TX, USA) have been available since 2007. Mean fluorescence intensity of over 500 was considered as positive. The calculated panel reactive antibody (cPRA) was calculated using a cPRA calculator through the website of Organ Procurement and Transplantation Network (http://optn. transplant.hrsa.gov).
All values are shown as means and standard deviation, median (IQR), or percentage. Continuous variables were compared using the t test, and quantitative variables were compared using the Mann-Whitney nonparametric test. Categorical variables were analyzed using the chi-square or Fisher exact test. Univariate and multivariate associations with second graft loss were analyzed using Cox proportional hazard regression, estimated hazard ratio (HR), and 95% confidence intervals (CI). Results with P < .05 from the univariate analysis were included in the multivariate analysis. Graft and patient survival rates were calculated by the Kaplan-Meier method and compared using the log-rank test. The differences were considered statistically significant at P < .05. JMP version 12 software (SAS Institute Inc, Cary, NC, USA) was used for all statistical analyses.
Patient characteristics and demographics
Patient characteristics for both groups are summarized in Table 1. There were no significant differences in recipient sex, age at time of second transplant, and ethnicity. Group A had a significantly longer first graft survival rate than group B (100.6 months [IQR 63.7-134.7 mo] vs 3.7 months [IQR 0.1-47.8 mo]; P < .001). Time from first graft loss to listing on the transplant wait list was not significantly different between group A and group B (5.3 months [IQR 1.2-16.5 mo] vs 8.7 months [IQR 5.0-21.5 mo]; P = .809). Time from listing on the transplant wait list for a second transplant was numerically longer in group B than in group A but not statistically different. Time from first graft loss to receiving a second transplant was similar between group A and group B (36.7 months [IQR 15.0-62.2 mo] vs 53.8 months [IQR 28.0-75.6 mo]; P = .245). For the second transplant, transplant type and sex and age of the second donor did not differ between the groups. HLA-A/B/DR mismatches were comparable in both groups. There were no significant differences in the type of calcineurin inhibitor and induction therapy (P = .468 and .647). Overall preformed DSAs (anti-HLA class I or II antibodies) for the second graft were detectable in 28.8% of group A compared with 39.0% of group B patients (P = .243). Anti-HLA class I antibodies were more commonly present. Anti-HLA class I antibodies in group B were higher but not different from those in group A (35.6% vs 25.4%; P = .230). There was also no difference in anti-HLA class II antibodies between both groups. Before the second transplant, 45.8% in group A had < 30% of cPRA, whereas 59.3% in group B had ≥ 30% of cPRA. However, the sensitization status, defined as nonsensitized (cPRA < 30%), sensitized (cPRA 30-80%), and highly sensitized (cPRA > 80%), was similar in both groups.
Outcome of second transplant
The incidences of DGF, de novo DSA, overall rate of acute rejection, and infections were not statistically different between the 2 groups (Table 2). The overall rate of acute rejection after a second transplant was 19.5%. Group B had a higher rate of acute rejection than group A (23.7% vs 15.3%; P = .245). T-cell-mediated rejection was the most common type of acute rejection in the entire study cohort and was present in 11.9%, whereas combined T-cell-mediated rejection with antibody-mediated rejection (AMR) was observed in 5.9% of cases. The incidence of posttransplant infection was similar in both groups (23.7% vs 33.9% in group A vs B; P = .223), and predictably the most common infection type was urinary tract infection in 11.9% of patients. At 1 week posttransplant, eGFR results were 31.1 ± 25.2 mL/min in group A and 30.8 ± 25.1 mL/min in group B, which improved to 48.9 ± 19.6 mL/min and 47.8 ± 18.5 mL/min, respectively, at 2 months posttransplant and became stable thereafter (Figure 1). In both groups, eGFR results were similar and did not reach statistical significance until 3 years of follow-up.
Forty-two patients (35.6%) lost their graft by the end of this study (Table 2). In group B, 44.1% had graft failure compared with group A (27.1%; P = .054).
In group A, the causes of graft loss were chronic allograft nephropathy in 7 patients (11.9%), vascular thrombosis in 1 patient (1.7%), and infection in 1 patient (1.7%). Seven other patients (11.9%) died with functioning graft. In group B, the causes of graft loss were chronic allograft nephropathy in 7 patients (11.9%), vascular thrombosis in 5 patients (8.5%), primary nonfunction in 4 patients (6.8%), infection in 1 patient (1.7%), recurrent glomerulonephritis in 2 patients (3.4%), technical failure in 1 patient (1.7%), acute rejection in 1 patient (1.7%), and death with functioning graft in 5 patients (8.5%).
Sixteen grafts failed within the first 3 months after a second transplant. Group B patients were more likely to experience graft loss within 3 months posttransplant than group A (22.0% vs 5.1%, P = .007). A second allograft nephrectomy was performed in 19 patients, with 15 (25.4%) in group B, which was higher than in group A (6.8%; P = .006). Seven patients (11.9%) died in group A versus 10 patients (16.9%) in group B. However, the mortality rate was similar in both groups (P = .619). In both groups, the overall second graft survival rate was 87.9% at 6 months, 85.0% at 1 year, 78.3% at 3 years, and 73.0% at 5 years posttransplant. In group A, graft survival rate was 94.9% at 6 months, 93.0% at 1 year, 87.0% at 3 years, and 82.3% at 5 years, which were superior results than those in group B (80.7%, 76.7%, 69.1%, and 62.5%, respectively; P < .05) (Figure 2). The overall patient survival rates in our patients were comparable to published data at 6 months (96.6%), 1 year (94.6%), 3 years (92.5%), and 5 years (91.2%). Moreover, patient survival rate was similar between the 2 groups (Figure 3). The 1-, 3-, and 5-year patient survival rates in group A were 96.4%, 94.3%, and 92.0% versus 92.9%, 90.6%, and 90.6% in group B (P = .396).
Risk factors for second graft failure
Our univariate analyses showed that first allograft nephrectomy, interval from first graft loss to second transplant, preformed DSAs (anti-HLA class I antibody) for second transplant, DGF, and acute rejection were associated with survival after second graft (Figure 4). However, our multivariate analyses showed that interval from first graft loss to second transplant and acute rejection emerged as the independent determinants of outcomes after a second graft. Acute rejection had the highest HR for second graft loss (HR = 2.24, 95% CI, 1.10-4.45; P = .027). The interval from first graft loss to second transplant showed significant HR (1.11, 95% CI, 1.02-1.19; P = .008).
Currently, there is uncertainty whether allograft nephrectomy should be performed prior to RRT. We therefore conducted this study using comprehensive and nearly complete data from a large series of patients in a single center to compare the 2 approaches.
The principal finding of this study was that graft survival was inferior in patients who underwent allograft nephrectomy prior to RRT compared with survival in those who did not undergo allograft nephrectomy. This is in line with the findings of other small studies.6,8 However, some studies have demonstrated a neutral effect of allograft nephrectomy for regraft survival.10,12,18-20 In a recent meta-analysis that included 8 studies and over 1000 patients, no overall harm or benefit was seen in graft survival rates for patients with allograft nephrectomy.21 As is usual with meta-analysis, the causes and the timing of graft loss were lacking. In our series, we demonstrated that early graft loss (≤ 3 mo posttransplant) was more likely to occur in patients who received a previous allograft nephrectomy. In these patients, the early graft loss translated to inferior long-term overall graft survival.
Moreover, similar to other studies, we were unable to show either a beneficial or deleterious effect of allograft nephrectomy on overall patient survival.6,10,21 Furthermore, the incidence of DGF was not affected by allograft nephrectomy, although the latter 2 studies also demonstrated inferior graft function at 12 months posttransplant in patients with allograft nephrectomy.19,20 In our study, however, graft function was not affected by prior allograft nephrectomy when considered longitudinally over a 3-year period. Similarly, in a recent meta-analysis, graft function as assessed by serum creatinine was similar between an allograft nephrectomy group and a group without allograft nephrectomy at 1 year posttransplant.21 Our study, however, reported renal function as assessed by eGFR at 3 years. Our study, in addition to others, does demonstrate that allograft nephrectomy does not confer benefit in a second graft recipient with regard to DGF and graft function.
Allograft nephrectomy allegedly increases immunologic risk of acute rejection due to higher levels of PRA.6,8-10 Sumrani and associates demonstrated that 57% of patients with allograft nephrectomy had PRA > 30% before retransplant compared with 33% in those without allograft nephrectomy.9 Johnston and associates compared the rate of PRA < 30% before a first transplant and a retransplant.22 In patients who had PRA < 30% before a first transplant, the level of PRA before retransplant with allograft nephrectomy became significantly higher than those with a graft in situ (P < .001). A high PRA-positive rate after allograft nephrectomy was confirmed in a recently published meta-analysis.21 Tsapepas and associates demonstrated in general that recipients with preformed DSAs had significantly higher rates of acute rejection (54.8% vs 34.8%; P = .01), including AMR (32.3% vs 7.1%; P < .001), and shorter duration to first acute rejection (P = .014) than those without preformed DSAs.23 In our study, overall, 17.5% of patients with preformed DSAs had AMR compared with only 2.7% of recipients without DSAs (P = .005). Moreover, a high level of cPRA (> 80%) also showed association with AMR (P = .008). In our study, the prevalence of preformed DSA was significantly higher with an elevated risk of acute rejection. We presented that allograft nephrectomy had a numerically high rate of DSAs but that allograft nephrectomy was clearly not related to the production of DSAs. It is still uncertain why anti-HLA antibodies increase after allograft nephrectomy. Withdrawal of immunosuppression after allograft nephrectomy may lead to rebound immune reactivity against the residual donor tissue left in situ, such as renal capsule, renal artery and vein, and possibly ureter. The other reason is that a failed allograft may act as an immunoadsorbent for alloantibodies and keep their level low before allograft nephrectomy, thus masking their detectability. High DSA positivity and sensitization after allograft nephrectomy may lead to overt and more importantly subclinical rejection, resulting in inferior graft outcomes.
The management of immunosuppression is an important issue after graft loss. Immunosuppression can be stopped immediately after allograft nephrectomy. However, there is a risk of the appearance of anti-HLA antibodies after allograft nephrectomy. This may lead to acute rejection, inferior second graft survival, and a prolonged wait time for a second transplant, resulting in higher dialysis vintage. Extended vintage of dialysis before transplant is reported to be a risk factor for poor graft prognosis compared with short-duration dialysis.24 Weaning off of immunosuppressive agents is associated with sterile fever and an increased risk for allograft nephrectomy. Therefore, an in situ failed graft requires immunosuppression to avoid rejection. Prolonged immunosuppressive maintenance results in less sensitization compared with early withdrawal.25 However, continuing immunosuppression is associated with a high incidence of infection and mortality.26 We did not analyze the relationship between allograft nephrectomy and immunosuppression before second transplant due to a lack of information regarding the management of immunosuppressive agents used after first graft loss. However, our study demonstrated that allograft nephrectomy before a second transplant did not influence the second graft survival.
Various risk factors such as recipient and donor sex, donor age, renal function of donor, PRA level at the time of retransplant, a long interval from previous graft failure to retransplant, DGF, and previous allograft nephrectomy have been associated with suboptimal second graft outcomes.6,8,10,12,20 The associations between complications following a first transplant and their impact on subsequent transplant outcomes were studied by Heaphy and associates.27 In their experience, occurrence of DGF, acute rejection, and hospitalization within 1 year after first transplant were more likely to recur after a second transplant. We found that acute rejection after a second transplant and interval from first graft loss to second transplant were associated with poor second graft survival. Unlike Heaphy and associates, we did not find any association between the complications following the first graft and their recurrence in the second graft. However, patients with first allograft nephrectomy were more likely to undergo repeat allograft nephrectomy after a second graft failure. The underlying reason for this seems to be related to the clinical practice of allograft nephrectomy after early graft failure.
This study has several limitations inherent to its observational design. The number of cases was small in this single-center study. However, it benefited from having complete clinical and immunologic data. In our center, failed grafts are managed by a standard protocol, which includes first-line allograft nephrectomy for graft failure within 3 months posttransplant, graft infections, and inflamed graft diagnosed clinically (resistance to erythropoietin-stimulating agents) or serologically (persistently elevated C-reactive protein). In remaining cases, allograft nephrectomy was usually avoided except in certain cases where the protocol was not adhered to and the graft was removed, as influenced by literature reports.
Our study provides additional information from a large cohort of ethnically diverse patients and suggests that allograft nephrectomy before a second transplant is associated with early graft loss and overall inferior second graft survival, possibly due to increased overt and subclinical rejections. We recommend that allograft nephrectomy should be avoided in the absence of any clinical indication. Moreover, patients with allograft nephrectomy should be considered as high risk for inferior second transplant survival and monitored closely with a high index of suspicion for both overt and subclinical rejections.
Volume : 16
Issue : 3
Pages : 259 - 265
DOI : 10.6002/ect.2018.0046
From the 1Renal Medicine and Transplantation Department, The Royal London
Hospital, London, United Kingdom; the 2Nephrology Department, Toho University
Faculty of Medicine, Tokyo, Japan; and the 3Cellular Pathology Department and
the 4Clinical Transplant Laboratory, The Royal London Hospital, London, United
Acknowledgements: The authors have no sources of funding for this study and have no conflicts of interest to declare. *Magdi Yaqoob and Carmelo Puliatti contributed equally and are joint senior authors.
Corresponding author: Masaki Muramatsu, Department of Nephrology, Toho University Faculty of Medicine, 6-11-1, Omori-nishi, Ota-ku, Tokyo 143-8541, Japan
Phone: +81 3 3762 4151
Figure 1. Estimated Glomerular Filtration Rate After Second Transplant
Figure 2. Second Graft Survival in Patients With or Without First Allograft Nephrectomy
Figure 3. Survival in Patients With or Without First Allograft Nephrectomy
Figure 4. Univariate and Multivariate Analyses of Risk Factors for Second Transplant Survival
Table 1. Patient Characteristics
Table 2. Outcomes of Second Transplant