1. Introduction
The number of patients returning to dialysis after kidney graft failure is increasing, because of limited survival of the kidney graft, increased number of kidney transplantations and decreased access to a new transplantation due to immunisation [1].
According to several recent studies, kidney graft failure is the fourth leading cause to dialysis initiation [2] [3].
This transition period from renal transplantation to dialysis is associated with high high risk of complications. Indeed, patients on dialysis after kidney graft failure are reported to have a higher mortality rate than those who retain a functional graft [1].
This is explained by an inflammatory state due to the allograft, leading to anemia via erythropoietin resistance and hypoalbuminemia.
In addition, maintaining immunosuppressive therapy after return to dialysis in order to preserve residual kidney graft function and prevent graft necrosis increases the risk of infectionandtumors [4].
Psychological difficulties have also been reported in these patients following the change of care teams, but especially due to the end of autonomy related to renal transplantation [5].
This finding should draw attention to the importance of monitoring these patients during this period.
Much has been written regarding the risks incurred by this population. However, no clearly defined management strategy is available: the timing of dialysis initiation is not well defined, the management of immunosuppressive drugs is not standardised and indication for detransplantation remains controversial.
In the light of the data, we undertook a study of patients returning to dialysis after kidney graft failure in our center.
We describe the clinical and paraclinical characteristics, therapeutic management and evolutionary profile of those patients.
2. Material and Methods
This is a retrospective, descriptive, monocentric study conducted in the Nephrology-Dialysis-Transplantation Department of IbnSina University Hospital in Rabat.
The study period is spread over 23 years, from January 1998 to December 2021.
We included all kidney transplant patients followed in our unit who presented a kidney graft dysfunction and returned to dialysis.
We excluded from this study kidney transplant patients with a follow-up after return to dialysis less than 1 year.
Variables and definitions:
We analyzed the demographic, clinical, paraclinical, therapeutic and evolutionary data of these patients.
Data were collected from medical files in the renal transplant unit, and the analysis of the data concerned 3 periods:
For which the analysis was essentially descriptive.
This period was analyzed in terms of the patients’ age, their comorbidities, their clinical examination and the results of biological tests. We also recorded the modality and the circumstances of the return to dialysis as well as the need to place a central venous catheter.
We noted the dosage of the immunosuppressive treatment, i.e. anti-metabolite, anticalcineurin and corticosteroids at Day 0, day of the return to dialysis, at 1 month, 3 months, 6 months and 1 year.
We defined early cessation of immunosuppressive therapy as cessation of anti-metabolite and anticalcineurin before 120 days after return to dialysis.
We also reported infectious episodes, the hospitalisation requirement, the occurrence of rejection and the need for detransplantation.
We judged patient outcomes based on the occurrence of infection, graft intolerance syndrome, need for detransplantation, detection of HLA antibodies, and patient survival at 1 year.
The statistical analysis was carried out using JAMOVI software version 2.3.21.0.
Quantitative variables are expressed as means and standard deviation, median and interquartile range, and qualitative variables as number and percentage.
The data are compared using the Chi-2 test or Fisher’s exact test.
3. Results
Among 166 kidney transplant patients, we identified 20 patients who returned to dialysis during the inclusion period, corresponding to a prevalence of 12%.
The mean age of our patients at the time of return to dialysis was 38.7 ± 11.9 years, with a sex ratio M/F of 2.3.
1) Period of dialysis before renal transplantation
Initial nephropathy was glomerular in 3 patients: 1 case of membranous nephropathy and 2 cases with IgA nephropathy. Nephroangiosclerosis, Alport syndrome and polycystic kidney disease were noted in 1 case each.
Chronic tubulointerstitial nephritis was involved in 3 patients. While the initial nephropathy remained undetermined in 11 patients.
The most frequent dialysis modality before renal transplantation was haemodialysis for 19 patients with a mean dialysis duration of 34.5 ± 29.5 months.
2) Period of renal transplantation:
Three patients received a kidney from a brain-dead donor (BDE) and 17 patients from a living donor, with a median kidney graft survival time of 85.5 months [42 - 186].
The causes of kidney graft dysfunction are listed in Table 1.
3) Period of return to dialysis after renal transplantation:
The clinical characteristics of patients after return to dialysis are summarised in Table 1 and the biological data in Table 2.
At the time of return to dialysis, 14 patients (70%) remained hypertensive with more than 3 antihypertensive treatments, 2 patients had heart disease (10%) and only 1 patient developed a new-onset diabetes after transplant (NODAT) (5%). Chronic smoking was reported in only one patient (5%).
The average BMI of our patients was 22 kg/m2 ± 3, with extremes ranging from 18.3 to 26 kg/m2. In addition, 4 patients presented psychological disorders requiring psychiatric follow-up (20%).
Biologically, an inflammatory syndrome was found in the majority of our patients with a median C-reactive protein (CRP) of 15.5 [3.8 - 27].
The initiation of dialysis was urgent in 10 patients (50%), with a low average GFR of 5.1 ml/min/1.73m2. This dialysis was carried out using a central venous catheter in 7 patients (35%), one of whom was tunnelled, while 8 patients had a previously prepared AVF. Once the emergency was over and after education sessions, 15 patients opted for HD and 5 for PD.
Regarding the management of immunosuppressive therapy, we kept oral corticosteroid therapy at low dose (5 mg/d) for the first year. The antiproliferative agent is reduced by 50% per month until discontinuation and the anticalcineurin is reduced by 1mg every two weeks for tacrolimus and 25 mg every two weeks for ciclosporin until discontinuation.
Thus, weaning from mycophenolatemofetil (MMF) and anticalcineurins is achieved progressively over a period of 3 months and 5 months respectively. Three patients were allowed to stop immunosuppressive treatment early without clinical consequences on the kidney graft (Figure 1).
During the first year of follow up, 13 patients (65%) required one or two hospitalisations, with an average duration of hospitalisation of 15 days. Eight patients were hospitalised for infections: 4 cases of tuberculosis, 3 case of dialysis catheter related infection and 1 case of peritonitis in PD. The evolution was favorable for all patients after appropriate antibiotic therapy.
Figure 1. Management of immunosuppressive therapy.
Table 1. Characteristics of patients after return to dialysis.
Variables |
Value |
Age of patients at return to dialysis (in years)a |
38.7 ± 11.9 |
Genderb |
|
Male |
14 (70%) |
Female |
6 (30%) |
IMC (kg/m2)a |
22 ± 3 |
Comorbiditiesb |
|
NODAT |
1 (5%) |
Hypertension |
14 (70%) |
Heart disease |
2 (10%) |
Smoking |
1 (5%) |
Depression |
4 (20%) |
Causes of kidney graft dysfunctionb |
|
Rejection |
15 (75%) |
Recurrence of initial nephropathy |
3 (15%) |
Toxicity of anticalcineurins |
1 (5%) |
Graft renal vein thrombosis |
1 (5%) |
Lifetime of kidney graft (in months)c |
85.5 [42 - 186] |
Urgent initiation of dialysis after renal transplantationb |
10 (50%) |
Post-transplantation dialysis modalityb |
|
HD |
15 (75%) |
PD |
5 (25%) |
Hemodialysis vascular accessb |
|
Catheter |
7 (35%) |
AVF |
8 (40%) |
Hospitalisation (patients)b |
13 (65%) |
Duration of hospital stay (in days)a |
15 ± 7.19 |
Infectionb |
|
Tuberculosis |
4 (8%) |
Catheter related infection |
3 (13%) |
Peritonitis in PD |
1 (5%) |
Kidney graft intolerance syndromeb |
|
Total |
8 (40%) |
During the 1st year |
5 (25%) |
Detransplantationb |
6 (30%) |
Deathb |
1 (5%) |
aExpressed as mean ± standard deviation, bExpressed as number (percentage), cExpressed as median [interquartile range].
Table 2. Biological characteristics of patients returning to dialysis.
Variables |
Value |
Uremia (g/l)a |
2.35 ± 0.8 |
Creatinine serum level (mg/l)a |
116 ± 48.3 |
GFR (ml/min)a |
5.1 ± 2.2 |
Kalemia (mEq/l)a |
4.75 ± 0.7 |
Alkaline reserve (mEq/l)a |
15.7 ± 6.7 |
Natremia (mEq/l)a |
133± 5.1 |
Calcemia (mg/l)a |
81.4 ± 8.7 |
Phosphorus (mg/l)a |
58.5 ± 20 |
Uricemia (mg/l)a |
80.3 ± 24.1 |
Albumin serum level (g/l)a |
32.9 ± 6 |
Haemoglobin serum level (g/dl)a |
8.6 ± 1.9 |
C-reactive protein (mg/l)b |
15.5 [3.8 - 27] |
aExpressed as mean ± standard deviation, bExpressed as median [interquartiles].
After a mean follow-up of 22 months, 8 patients presented a kidney graft intolerance syndrome, 5 of them during the first year.
The clinical symptomatology was suggestive of kidney graft sensitivity in 8 cases, associated with haematuria in 6 cases and resistant hypertension in 2 cases. All patients had an elevated CRP, averaging 92.9 ± 76.6 mg/l with extremes between 25 and 207 mg/l. Doppler ultrasound of the kidney graft performed in all patients showed infiltration of the graft associated with foci of pyelonephritis in 2 patients.
Increasing the dose of steroids from 5 to 20 mg improved the clinical signs in 2 patients, while 6 patients required detransplantation. The histological examination of the graft showed foci of necrosis in 5 patients and pyonephrosis in one patient.
The evolution was favorable for all patients after detransplantation with a regression of clinical symptoms, normalization of blood pressure and resolution of the biological inflammatory syndrome.
Despite the deteriorated condition of the patients at the time of initiation of dialysis, only one death occurred after 16 months following a covid-19 infection.
In statistical analysis, the comparison between the early weaned group of immunosuppresive therapy (<120 days) and the late weaned group (>120 days) showed a significant association between the prolongation of immunosuppressive therapy beyond 120 days and the occurrence of infection (p < 0.05) (Table 3).
In terms of psychological care, our patients experienced psychological difficulties during the period of transition. The beginning of dialysis marks a period of adaptation, as the patient’s rhythm of life is affected by the frequency of dialysis sessions.
Table 3. Risk factors associated with duration of immunosuppressive therapy.
Variables |
Weaning IS > 120 jours |
p-Value Chi-2 |
p-Value Fisher |
No n = 17 |
Yes n = 3 |
Death |
1 (6%) |
0 (0%) |
0.6 |
1 |
Infection |
5 (29%) |
3 (100%) |
0.02 |
0.049 |
Detransplantation |
6 (35%) |
0 (0%) |
0.2 |
0.52 |
Graft intolerance syndrome |
8 (47%) |
0 (0%) |
0.1 |
0.24 |
Detection of HLA antibodies |
12 (70%) |
3 (100%) |
0.2 |
0.53 |
IS: Immunosuppressive therapy; HLA: Human leukocyte antigen.
The majority of patients felt nervous, stressed, anxious, had no desire for anything, and experienced mood swings. Some of them were restless and had sleeping difficulties.
Four patients were depressed, refused dialysis and required psychiatric follow-up.
These psychological difficulties can be attributed to changes in medical teams, from the transplant doctor to the dialysis doctor, but mostly due to the loss of autonomy provided by the kidney transplant.
4. Discussion
4.1. Initiation of Dialysis: Circumstances and Modalities
The optimal time to start dialysis after kidney graft failure is not clearly defined in the literature. Some authors start dialysis early to avoid uremic complications, others delay dialysis until the onset of clinical signs [4] [5].
A cohort study by the United States Renal Data System (USRDS) registry showed that high GFR at dialysis initiation is associated with an increased risk of mortality of 4% for each ml/min/1.73m2 (HR: 1.04, 95% - CI: 1.02 - 1.06) [6]. However, this study is biased since patients with comorbidities are those who started dialysis early.
Molnar et al. used a propensity score to evaluate outcomes of 747 patients with kidney graft failure. According to the adjusted model, GFR did not emerge as a risk factor for mortality. (HR: 1.02, 95% IC: 0.96 - 1.07). However, higher mortality was noted in young and healthy patients. The conclusion of this study is that there is no benefit to initiate dialysis early [7]. These results are in agreement with those of the IDEAL study [8].
In our study, 10 patients initiated dialysis urgently with a mean GFR of 5 ml/min/1.73m2. We expect a lower percentage of urgent dialysis since all our patients are followed regularly and closely. This underlines the difficulty of nephrologists to manage these patients who are for the majority reticent to return to dialysis.
Thus, the initiation of dialysis for patients with kidney graft dysfunction must take into account not only the GFR but also the clinical status of patients, their comorbidities and the existence of symptoms related to graft dysfunction.
Dialysis modality offered to patients with kidney graft dysfunction is debated in the literature. Many studies have compared HD with PD and have shown equal survival with both techniques [9]-[12].
4.2. Management of Immunosuppressive Therapy
The management of immunosuppression after return to dialysis is a real challenge for nephrologists. In the absence of recommendations, practices vary from one centre to another.
Some propose to withdraw the immunosuppressive therapy in order to avoid its side effects, 76 while others carry out a gradual weaning. In fact, both approaches have their benefits and risks.
Indeed, maintaining immunosuppression after kidney graft dysfunction preserves residual graft function, prevents graft intolerance syndrome and avoids HLA immunisation. On the other hand, this immunosuppression increases the risk of infections and complications [13]-[15].
Several studies have evaluated the impact of immunosuppression on patients undergoing dialysis after kidney graft dysfunction (Table 4).
In our study, 8 patients developed graft intolerance syndrome after a mean follow-up time of 22 months, 6 of them required detransplantation, without any statistical correlation with early withdrawal of immunosuppressive therapy.
Table 4. Studies analysing the management of immunosuppression.
Authors/Year |
Cohort size |
Results |
Augustine et al. 2012 [16] |
119 |
Withdrawal of immunosuppressive therapy was
predictive of HLA immunisation (P = 0.004), graft
intolerance syndrome and detransplantation
(P < 0.001). |
Kosmoliaptsis
et al. 2014 [17] |
131 |
Maintaining two immunosuppressive drugs
reduces significantly the risk of developing HLA
antibodies. |
Martin et al. 2021 [18] |
134 |
Continued immunosuppressive therapy did not
improve HLA immunisation or outcomes of the
second kidney graft, but reduced significantly the
risk of detransplantation (P = 0.01). |
Knoll et al. 2022 [19] |
269 |
Continued immunosuppression is not associated
with a higher risk of death or hospitalisation for
infection, but remains insufficient to prevent HLA immunisation and graft intolerance syndrome. |
However, infection was statistically higher in patients with prolonged immunosuppression over 120 days (P = 0.02). Almost half of these infections were due to dialysis catheter related infection (3/8). Their occurrence could be prevented by early creation of a vascular access for patients with graft dysfunction as soon as a GFR of 20 ml/min/1.73m2.
4.3. Detransplantation
The detransplant rate after kidney graft failure varies widely from one centre to another. According to the USRDS, the probability of detransplantation at 1 week, 3 months, 6 months and 1 year after graft dysfunction is respectively 5.3%, 17.6%, 25%, and 30.9% [20]. Thus, 89.3% of detransplants are performed within the first year [21].
There is no systematic recommendation for detransplantation after return to dialysis [22]-[24]. The majority of centres follow indications that are related either to the kidney graft or to immunosuppression (Table 5).
A less invasive alternative to detransplantation described in 1993 by Lorenzo et al., is kidney graft embolisation [25].
In 2018, a meta-analysis and systematic review compared detransplantation to graft embolisation in 2421 patients in terms of morbidity and mortality. They found that mortality and morbidity in the detransplantation group were higher compared to the embolisation group. However, 20% of patients required subsequent detransplantation [26]. This result may be of interest for patients with a high surgical risk.
4.4. Patients Surviving on Dialysis after Kidney Graft Failure
Survival of patients with graft failure is poor, estimated at 40% at 10 years compared to 75% for patients with a functional graft [27]. The most common causes of death reported in the literature are infections and cardiovascular disease [28]-[30].
Table 5. Indications for detransplantation afterkidney graft failure [15] [18] [19] [24] [31].
Absolute indications |
Relative indications |
Primary non-function of renal allograft |
Early graft loss (<6 - 12 months) |
Acute rejection resistant to immunosuppression |
Resistant BK virus nephropathy |
Severe graft pyelonephritis/urosepsis |
|
Graft lymphoproliferative syndrome |
|
Refractory graft intolerance syndrome |
|
Graft vascular complications: Hemorrhage/Thrombosis |
|
Creating space for a second graft |
|
In fact, patients with a dysfunctional graft tend to have hypoalbuminemia, anemia and resistance to erythropoietin in the context of chronic inflammation, which may increase the risk of death [32]-[34].
In addition, these patients have a long history of uraemia and immunosuppressive therapy which increases the risk of cardiovascular disease, diabetes, infections and cancer [35] [36].
Management of these modifiable risk factors may improve survival in such patients.
The use of central venous catheters in two-thirds of patients returning to dialysis after kidney graft dysfunction contributes to a significant rate of catheter related infections [6] [27] [37]-[39]. Thus, the creation of a permanent vascular access could reduce the morbidity and mortality linked to infections [30].
A significant increase in cardiovascular events has been reported in patients maintained on immunosuppression, which increases the risk of mortality by 4.9 times (95% CI: 1.8 - 13.5), mainly by myocardial infarction [40]. Assessment and treatment of cardiovascular risk factors is essential to improve patient survival.
In our cohort, we report one death due to infection after 16 months follow up: this was an acute respiratory distress syndrome related to the COVID-19 pandemic.
4.5. Limits of the Study
Our study has several limits: the retrospective nature of the study exposes us to missing data, the small sample size, and the essentially descriptive analysis of data which limits the scientific contribution of this work.
However, these data could improve the management of these patients during this delicate transition period and consequently improve their vital prognosis.
5. Conclusion
The transition from kidney graft to dialysis is associated with high morbidity and mortality [28]-[30] [38]. However, there are no practical recommendations for the management of these patients. We believe that adequate preparation for dialysis initiation, early management of comorbidities and creation of a vascular access could improve outcomes and reduce risks in this population.