Effects of pretransplant peritoneal vs hemodialysis modality on outcome of first kidney transplantation from donors after cardiac death

Background The effect of pretransplant peritoneal dialysis (PD) or hemodialysis (HD) modality on outcomes of kidney transplantation (KT) for end-stage renal disease (ESRD) is debatable. We evaluated the outcomes these modalities in KT from donor after cardiac death (DCD). Methods A cohort of 251 patients on HD, PD or pre-emptive who underwent first KT from DCD between January 2014 and December 2016 were prospectively analyzed to compare for outcomes on recovery of renal function, complications as well as patient and graft survival. The patients were followed till August 2017. Data on 104 HD and 98 PD were available for final comparative outcome analysis, 5 pre-emptive were analyzed as the control group. Results Both HD and PD group patients were well matched for demographic and baseline characteristics. The follow-up period was 12.5 (3.0, 22.0) months in HD and 12.0 (6.0, 20.0) months in PD patients. Post-transplant renal functions between the two groups showed no differences. Among PD patients, 16 (16.3%) suffered delayed graft function, versus 19 (18.3%) in HD, with no statistical differences (p = 0.715). Complications of acute rejection, infections were comparable between the groups. The patient survival, graft survival and death-censored graft survival were similar for HD and PD after adjusting for other multiple risk factors. Conclusions Our results indicate that outcome of first KT from DCD is not affected by pretransplant dialysis modality of PD or HD in aspects of recovery of renal function, complications as well as patient and graft survival.

More studies are needed to clarify the identical or even better function of PD compared with HD. Therefore, we conducted this prospective cohort study to compare the effects of pretransplant HD vs PD on outcomes of renal function, post-transplant complications, graft as well as patient survival of first KT from Donor after cardiac death (DCD).

Study population
This was a prospective cohort study of ESRD (defined as eGFR< 15 ml/min/1.73 m 2 ) patients who received their first kidney transplantation from DCD between January 2014 and December 2016 in renji hospital affiliated to School of Medicine of Shanghai Jiaotong University, a hospital at Pudong New District, Shanghai, China. During the transplantation, the technical issues that may affect the outcome of transplant like organ transplantation, preservation as well as surgical operation were all performed by same transplant team in our hospital. DCD was defined as awaiting cardiac arrest after withdrawal of life-supporting treatment in the intensive care unit.
Patients above 18 years of age who had been on the same dialysis modality (hemodialysis or peritoneal dialysis) for at least 3 months without a switch or underwent transplantation before the initiation of dialysis (pre-emptive kidney transplantation, PKT) were included. Patients who were living donor transplant, second KT or multiple organ transplants and lost to follow-up were excluded. Follow-up was terminated on August 2017. Finally data on 104 HD and 98 PD group met the inclusion criteria and were included for comparative outcome analysis. 5 PKT patients were analyzed as the control group (Fig. 1).
The protocol of this study was reviewed and approved by the Ethics Committee of Renji hospital, and patients were included only after signing informed consent.

Data source
Donor variables included age, gender, BMI, blood group, percentage of hypertension, mean time of intensive care unit(ICU) stay, HLA mismatching, estimated glomerular filtration rate (eGFR) and the causes of death. The baseline variables of recipients in two groups of patients on hemodialysis or peritoneal dialysis included age, gender, body mass index (BMI), blood group, duration time on pretransplant dialysis, post-transplant hospital stay and follow-up time, preoperative medical condition, percentage of anti-hypertensive drugs required, percentage  Fig. 1 Flowchart of patients who received KT from DCD. ERSD end stage renal disease, HD hemodialysis, PD peritoneal dialysis, KT kidney transplantation, DCD donor after cardiac death of hepatitis B virus (HBV) infection, native kidney diseases, pretransplant urinary volume, percentage of anuric patients and immunosuppresion therapy. The laboratory parameters of white blood cell count, creatinine, cholesterol, triglyceride of both donors and recipients were collected. The percentage of neutrophilic granulocyte and lymphocyte, haemoglobin, serum urea nitrogen, serum uric acid, serum albumin, alanine aminotransferase (ALT), blood glucose, serum potassium, serum sodium, parathyroid hormone, serum calcium, serum phosphate, low-density lipoprotein (LDL), high-density lipoprotein (HDL), and lymphocyte subtypes were also obtained from recipients. Among which the eGFR was estimated using the chronic kidney disease epidemiology collaboration (CKD-EPI) equation which was calculated according to the gender, the serum creatinine, the age and the race.
Post-transplant variables included renal function (serum creatinine, 24 h urine volume and the eGFR), haemoglobin, serum albumin, cholesterol, triglyceride, serum calcium and serum phosphate. The postoperative complications during hospitalization of delayed graft function (DGF), acute rejection (AR) and surgical complications including the urinaryfistula, hydronephrosis and hematoma after tranplantation were recorded. The infective complications including viral infection(cytomegalovirus, JC virus, BK virus, varicella zoster virus), fungal infection as well as bacterial infection(tuberculosis, urinary tract infection, acute bacterial pneumonia, gastrointestinal infection) were recorded during hospitalization and the whole follow-up time.   The DGF was defined as the requirement for dialysis in the first week after transplantation, or serum creatinine level increased, remained unchanged, or decreased by less than 10% per day immediately after surgery [14,15]. All patients with biopsy-proven acute rejection and those with features of antibody-mediated rejection, with borderline changes and allograft dysfunction who received treatment for acute rejection were considered to have rejection [14,15]. The patient, graft and death-censored graft survival were compared between HD and PD groups. The causes of patient mortality and graft failure were recorded. Death-censored graft failure was defined as suffering graft failure without death.

Statistical methods
The statistical analysis was performed by using the SPSS 22 version software. All numeric variables were tested for normality of their distribution. Independent samples t-test and Mann-Whitney U test were respectively used for analyzing data whose distribution are normal and abnormal. Results are described as mean ± standard deviations (SD) for normally distributed data, median and interquartile range (IQR, pp. 25-75) for abnormally distributed data. The Chi square or Fisher's exact tests was utilized to compare the categorical variables between the two groups. The results were expressed in numbers and relative frequencies [n(%)]. The patient and graft survival were calculated from the date of transplantation to the endpoints of the study. The univariate and multivariate analysis were conducted for risk factors for graft failure in HD and PD groups. The univariate analysis was conducted to study the risk factors of patient mortality and graft failure. Variables whose p < 0.05 in the univariate analysis or clinical meaningful were enrolled into multivariate analysis. The cox proportional regression models were used to assess the relative risks.
Variables of p values < 0.05 were considered to be statistically significant. All statistical tests were two-tailed.

Baseline clinical characteristics
The baseline information of recipients and donors in HD, PD and PKT group was analyzed. Among them, PKT group was analyzed as the control group. Patients in both PD and HD groups were comparable for demographic and clinical characteristics ( Table 1). The pretransplant laboratory parameters in PD group were lower compared with HD group with regard to haemoglobin, serum albumin, serum potassium and serum calcium (p < 0.05). On the contrary, the serum creatinine, serum urea nitrogen and serum uric acid were higher in PD group (p < 0.05) ( Table 2). Donors characteristics was comparable for recipient in both PD and HD groups (Table 3).

Post-transplant renal function outcomes
The post-transplant renal function outcomes of PKT were labeled as control. In view of the HD and PD group, the post-transplant serum creatinine showed no differences throughout the follow-up between the HD and PD groups (p > 0.05) (Fig. 2). Pretransplant serum creatinine was higher in PD patients (p < 0.05). There was no differences (p = 0.210) in serum creatinine reduction to half of baseline in two groups, with 1.0 (1.0, 2.0) d in HD  3) d in PD group. The serum creatinine levels at different post-transplant time points reduced statistically compared to pretransplant, which were coherent between the HD and PD groups (p < 0.001). There were significant differences in serum creatinine at different time points throughout the follow-up within two groups (p < 0.001). The 24-h urinary volume remained similar between the two groups during the follow-up period (p > 0.05) (Fig. 3) (Table 4).

Post-transplant complications
There were only 2 infective complications in PKT group. For HD and PD group, the hyperacute rejection didn't appear in both groups. Surgical and infective complications throughout the hospitalization and follow-up period were did not differ between two groups (Table 5).

Patient mortality and graft failure
There were no death and graft failure in PKT group. There were total 7 deaths, 4 in the HD and 3 in the PD group. The patient survival rate between two groups showed no significant differences (p = 1.000). There were 13 graft failure, 7 in HD and 6 in PD group, and causes of graft failure were statistically different between two groups (p < 0.001). The graft survival rates were similar  between the two groups (p = 0.860). The death-censored graft failure i.e. graft failure without death (3 in each of HD and PD group) was not different between the two groups (p = 1.000) ( Table 6).
The cox proportional hazards model showed pretransplant dialysis modality (HD and PD) had no correlation with patient survival or graft failure or death-censored graft survival After adjusting for other related multiple risk factors, the PD patients had similar rates of graft failure compared with HD in multivariate cox proportional hazards analysis (Table 7). When separately analyzed for HD and PD groups, the surgical complications in HD patients were independent stimulating factors of graft failure and DGF was an independent factor inversely correlated with graft survival in PD patients (Table 8).

Discussion
Up to 30-40% of patients are can be effectively treated by PD, far away from current 11% [16] and many suitable PD candidates are treated with HD [17]. The use of PD is lower than HD owing to the aging of dialysis population, comorbidity and social conditions that make home PD difficult. More studies are needed to   Causes of graft failure, % (n) < 0.001* Acute rejection 0(0/0) 0(0/7) 28.6(2/7) Our results indicated both the immediate and long-term renal function, the serum creatinine and urine output, were similar between the HD and PD, consistent with other studies [6,11]. In our study, the baseline serum creatinine was higher in PD than HD patients. The HD just before the transplantation could have lowered the serum creatinine. In contrast, PD lowers creatinine in moderate ways and high baseline serum creatinine in PD patients doesn't mean it is inferior to HD in creatinine reduction. This could be the reason why both groups had similar renal function after transplantation.
We had no significant differences in the incidence of AR, which is in line with recent studies [3,6,11,12]. This may be due to the availability of and rational novel immunosuppressive protocols nowadays. Our study shows similar incidence of DGF in both PD and HD patients, as reported by others [6,10]. We also found that DGF was inversely associated with the graft survival and death-censored graft survival regardless of dialysis modality. The DGF is associated with greater risk of patient death in addition to graft and death-censored graft failure [18]. Ischemic-reperfusion of donated kidney caused by postischemic acute tubular necrosis and interstitial inflammation results in DGF [19]. The PD patients has lower Table 7 The Univariate and Multivariate Analysis for Effects of HD vs PD and Other Factors on Outcomes of Non-preemptive Kidney Transplantation HD hemodialysis, PD peritoneal dialysis, P p value, RR relative risk, CI confidence interval, BMI body mass index, DGF delayed graft failure incidence of DGF in comparison with the HD [9,11,12]. This could be due to better hydration status and preservation of residual renal function (RRF) in PD patients [13]. Additionally, the PD patients has less oxidative stress which can exacerbate ischemic-reperfusion injury in kidney compared with HD patients [20].
Our results shows similar patient, graft and death-censored graft survival rate in PD and HD groups, consistent with most other studies [5,6,12]. Earlier, in a large cohort study of 22,776 patients concluded a higher rate of early graft failure (during the first 3 months after KT) in PD, possibly due to higher incidence of early graft thrombosis [18]. While the long-term graft failure and patient mortality remained similar. Some studies report PD had better patient survival, better quality of life, better nutritional status and fewer blood transfusions [9][10][11]. The differences might be associated with the different sample size and the follow-up time.
Some of the limitations of our study could be a singlecenter and inclusion of first transplantation from DCD only may not be applicable to all renal transplantations. In addition, the study variables of donors were incomplete, with some statistically analysis based on the less data compared to recipients. Besides, the pre-emptive kidney transplantation group in the cohort had only 5 patients, with people too less to be statistically comparative analyzed, finally simply summarized and displayed as the control group. And the follow-up period was not long enough, with further study and investigations to go on.

Conclusions
The choice of dialysis modality, HD or PD, prior to kidney transplantation had no influences on the patient, graft and death-censored graft survival. The immediate and longterm complications after transplantation, and renal function between the two groups were similar. Thus we can conclude that PD is equally good with potential for wider applicability as pretransplant modality of dialysis.  HD hemodialysis, PD peritoneal dialysis, P p value, RR relative risk, CI confidence interval, BMI body mass index, DG delayed graft failure