Epidural analgesia for living donor nephrectomy is associated with better early graft function in recipients after transplantation

Background: The bene�cial effects of epidural analgesia (EDA) in terms of pain control and postoperative convalescence are widely known and led to a frequent use for patients who underwent living donor kidney nephrectomy. The objective of this study was to determine whether general anesthesia (GA) plus EDA compared to GA only, administered for living donor nephrectomy has effects on postoperative graft function in recipients. Methods: In this monocentric, retrospective cohort analysis we analyzed the closed �les of all consecutive donor-recipient pairs who underwent living donor kidney transplantations from 2008 to 2017. The outcome variable was delayed graft function (DGF), de�ned as at least one hemodialysis within seven days postoperatively, once hyperacute rejection, vascular or urinary tract complications were ruled out. Statistical analyses of continuous variables were calculated using the two-tail Student’s t test and Fisher exact test for categorical variables with a signi�cance level of p<0.05, respectively. Results: The study enclosed 291 consecutive living donor kidney transplantations. 99 kidney donors received epidural analgesia whereas 192 denied epidural analgesia. The groups showed balanced pretransplantational characteristics and comparable donors´ and recipients’ risk factors. 9 out of all 291 recipients needed renal replacement therapy (RRT) during the �rst 7 days due to delayed graft function; all of these donors received no EDA. The observed rate of DGF in recipients whose kidney donors received epidural analgesia was signi�cantly lower (0% vs. 4.6%; p<0.031). Conclusions: In our cohort we observed a signi�cantly lower rate of DGF when epidural analgesia for donor nephrectomy was administered. Du to restrictions of the study design this observation needs further con�rmation by prospective studies.


Background
Living kidney transplantation showed superior results compared with deceased donor kidney transplantation in terms of graft survival, accessibility, waiting time and cost containment for public health services [1][2][3] .For patients undergoing surgical procedures for another one´s bene t, it is important to minimize perioperative risk and inconvenience.Furthermore it is the healthcare providers' duty to maximize the bene cial impact of the donation for the recipient.
In numerous studies major outcome bene ts like mortality of EDA could neither be con rmed nor denied 4,5 .However, the bene cial effects of EDA in terms of intra-and postoperative pain control, intestinal motility, early mobilization and duration of ICU-hospitalization are widely known and nd broad acceptance [6][7][8][9] .Therefore it is not surprising, that continuous EDA is a mandatory part of many surgical fast track programs [10][11][12] .In order to provide these advantages also for kidney donors and to increase their convalescence and speed up their reintegration in daily life, we offered EDA to patients for donor nephrectomy if contraindications were ruled out and patients gave their informed consent.The primary intent of providing perioperative EDA for donor nephrectomy are the bene cial effects for the donor.
These EDA effects are mostly mediated by perioperative sympathicolysis which probably may have effects on the kidney intended for transplantation.Potential effects on graft function of kidneys explanted from donors with EDA have not been reported yet.Therefore the aim of this hypothesis generating study was to determine whether GA plus EDA compared to GA only, administered for living donor nephrectomy is associated with bene cial effects on postoperative graft function after transplantation.

Methods
This retrospective cohort study was approved by the local Institutional Review Board, University of Freiburg, Germany (approval number EK 555/17).The study was conducted at the Department of Anesthesiology and Critical Care and the Department of General and Visceral Surgery, Medical Center -University of Freiburg, Faculty of Medicine -University of Freiburg Germany.The study was planned and designed in accordance with the initiative for Strengthening the Reporting of Observational Studies in Epidemiology (STROBE), using the suggested checklist for epidemiological cohort studies 13 .The study was initiated and designed in March 2018; the retrospective data collection was conducted in June 2018.
The onset of data collection is analogous to the existence of an electronic patient data management system on ICU which enabled data acquisition.As we enclosed only closed les and the data collection started in June 2018, cases after December 31 th 2017 were not enclosed.The study cohort consists of all consecutive living donor kidney transplantations between October 2008 and December 2017 which determines the sample size.A priori sample size calculation is not applicable in this fully retrospective and observational study design.Figure 1 shows the protocol of data collection and statistical processing.
Recipient and donor evaluation were based on a check-up examination which lead to con rmation of donor suitability.Ahead of transplantation all donor-recipient pairs were evaluated by an ethics committee of the District Medical Association Suedbaden.A positive vote of this ethics committee was mandatory for transplantation.Decision on epidural analgesia was based on the patients´ preference.All patients that received epidural analgesia gave their informed consent on that procedure.Epidural analgesia was performed directly preoperatively according a speci c departmental standard operating procedure: Epidural catheter was placed between the 7 th and the 11 th thoracic intervertebral space, followed by an application of 25 mg sufentanil and 10 ml ropivacaine 0.2 %.After the initial dose a continuous epidural application of 45 ml ropivacaine 0.2 % mixed with 25 mg sufentanil (= ropivacaine 0.18 % and sufentanil 0.5 mg/ml) with a infusion rate of 8 ml/h during the surgery was established.Anesthesia procedures for donor nephrectomy with and without epidural analgesia followed corresponding standard operating procedures (SOP).
Postoperatively all donors were transferred to a transplantation ICU.Patients, who received epidural analgesia, were visited daily by the acute pain service of our department.Epidural catheters were removed between the second and the fth postoperative day by the acute pain service.
Anesthesia for Transplantation was performed without epidural analgesia for the recipient and followed a departmental SOP which was established and revised where necessary in close collaboration between the responsible surgeons and anesthetists.The SOP addresses the need for 250 mg Prednisolone and 10 g Mannitol ahead of reperfusion.With the onset of reperfusion of the transplant kidney 125 mg Furosemide were administered.Intraoperative uid and catecholamine management was performed by the attending anesthetist referring to the SOP.
Renal replacement therapy (RRT) was initiated when patients were threatened by volume overload or increased serum potassium levels.Delayed graft function was de ned as any renal replacement therapy in the rst postoperative week, once hyperacute rejection, vascular or urinary tract complications were ruled out [14][15][16] .

Statistics:
The data was collected in a MS Excel™ (Microsoft, Redmond, USA) datasheet.Further statistical processing was performed using SPSS™ (IBM, Armonk, USA).Statistical analyses of continuous variables were calculated using the two-tail Student's t test and Fisher Exact test for categorical variables with a signi cance level of p<0.05, respectively.

Results
The study enclosed 291 consecutive living donor kidney transplantations between October 2008 and December 2017.99 kidney donors received epidural analgesia whereas 192 denied epidural analgesia.All recipients underwent kidney transplantation due to end stage renal disease.No mortality was reported in either group.In the EDA group no epidural catheter associated complications were found.
The perioperative characteristics are shown in table 1.The two study groups showed no signi cant differences in several donors´ risk factors except a signi cantly shorter nephrectomy time (135 vs. 144 min, p<0.003).The intraoperative uid consumption (1813 vs. 2191 ml; p=0.053) and maximum dose of vasopressor (0.03 vs. 0.06 mg/kg/min; p=0.300) showed no statistically signi cant difference.None of the recipients´ pre-and intratransplantation data showed a signi cant difference (table 1).After transplantation, 14 out of all 291 recipients needed renal replacement therapy during the rst 7 days after transplantation, but only 9 cases because of delayed graft function.The other 5 recipients suffered from humoral rejection, thrombosis of the iliac vessel or bleeding complications with the need of a surgical revision (table 2).All kidney donors to these 9 recipients received GA without epidural analgesia.The incidence of DGF was signi cantly higher in recipients whose donors did not receive epidural analgesia (4.6% vs. 0%; p<0.031) ( gure 2).In line with this nding the serum creatinine level as well as the maximum serum potassium level within 7 days were signi cantly lower in the recipients whose donors received EDA (2.17 vs. 2.04 mg/dl, p<0.036; 5.15 vs. 5.11 mmol/l, p<0.001).

Discussion
The bene ts of EDA regarding pain control, ICU stay, intestinal motility and early mobilization are frequently reported [17][18][19][20][21][22] .The bene cial effects of EDA in terms of pain control and return to normal daily activities speci cally for kidney donors have also been reported in the past 23,24 .This retrospective cohort study of 291 living donor kidney transplantations compared 99 cases whose donors received EDA with 192 cases that received GA only, with regard to DGF in the recipients.The main result is that DGF is signi cantly more frequent in patients whose kidney donors did not receive EDA.The incidence of DGF in cohorts of living kidney transplantations varies from 4 to 10% and increases morbidity, healthcare costs, hospitalization times and complicates post-transplantation care [25][26][27][28] .DGF predisposes for chronic rejection, chronic allograft nephropathy and seems to be causal for increased rates of graft failure and mortality 29,30 .
In line with the signi cantly lower rate of DGF in EDA group, we found that serum creatinine level, as well as the maximum serum potassium level within 7 days, were signi cantly lower in the EDA group.Although these ndings are statistically signi cant, their measured levels and differences in numbers are clinically questionable.Even when looking at the decline of the serum creatinine levels over the rst two days postoperatively there can be found no signi cant or clinically relevant difference.The recipients of the no-EDA group start at a slightly higher level of serum creatinine which should be taken in account.
Further baseline characteristics of donors and recipients showed no statistically signi cant difference or clinically relevant imbalance between the donors and recipients of both groups.An increased intraoperative uid and vasopressor consumption in the EDA group could be associated with the EDA mediated inhibition of the sympatho-adrenal response with consecutive vasodilatation.However, neither intraoperative uid nor vasopressor consumption showed a statistically signi cant difference in our study.

The standard surgical technique for donor nephrectomy in our institution is an open anterior extraperitoneal minimal incision laparotomy. Open surgical technique for donor nephrectomy is
associated with inferior cosmetic result, longer hospitalization and more intra-and postoperative pain with consecutively increased need for pain medication 31,32 .However, the open surgical approach showed superior results in terms of warm ischemia period, surgical costs, length of operation, intraperitoneal complications, recovery of graft function, recipient anastomosis di culties and incidence of acute tubulus necrosis [33][34][35] .It is reported that up to 25% of the living kidney donors after open surgical technique nephrectomy suffer from chroni cation of postoperative pain 36 .A reduction of somatic pain within the rst six postoperative weeks is associated with improved mental health of kidney donors 32 .
These ndings underline the need for EDA from the donors' perspective.The described clinical bene ts of EDA for the donor might lose their relevance and have to be reconsidered if the surgical approach in our institution changes to laparoscopic technique.
The reasons why kidney grafts fail to function immediately after transplantation when acute rejection, urological or vascular reasons are ruled out are associated with the transplanted kidney.DGF is modulated and caused by complex mechanisms of hypoxic and ischemic injuries and insu cient repair mechanisms 37 .These cascades seem to be induced by the operative trauma and the corresponding physiological stress response during donor nephrectomy.It is known that surgical procedures and the physiological stress response are associated with intra-and postoperative hypercoagulability which results in postoperative thromboembolic and vaso-occlusive events 38,39 .Increased levels of tissue factor, tissue plasminogen activator, plasminogen activator inhibitor-1, and von Willebrand factor which all contribute to hypercoagulability are reported to be found proximately after surgical stimulus 40 .Due to inhibition of nociceptive and non-nociceptive pathways of sympathetic innervation of the adrenal glands, EDA with local anesthetics leads to a perioperative sympathicolysis 41 .It is also reported that EDA modulates postoperative hypercoagulability by normalizing antithrombin III-activity and decreasing of platelet aggregation [42][43][44] .This could be one of the reasons that we found a signi cantly increased rate of DGF in the group of patients who refused perioperative EDA.
The present study has several severe limitations.
First, the retrospective and non-randomized design implies that a study protocol which addresses randomization on who receives EDA is missing.It is speculative why patients opted for or against EDA, possibly the way whether EDA was offered by the visiting anesthesiologist or rather recommended plays an important role.Perhaps patients who opted for EDA were more trusting of their physicians and therefore had less anxiety or stress levels which may have in uenced DGF of their donated kidney.RRT was initiated by visiting nephrology specialists and the request of the attending ICU physician when patients were threatened by volume overload or increased serum potassium levels.We were fully aware that living donor kidney transplantation is a highly complex procedure.The outcome quality is affected by various confounding variables for which we have not adjusted in our study due to the limited number of cases with DGF.
There are also several slightly different de nitions on delayed graft function in literature.In our study DGF was de ned as any renal replacement therapy in the rst postoperative week, when hyperacute rejection, vascular and urinary tract complications were ruled out.More than 22 different de nitions of DGF are described, the most common de nition refers to any RRT within the rst posttransplantational week 45,46 .Due to the manageable size of our cohort we could screen every case of RRT for the underlying reasons.Knowing these reasons leading to RRT for every patient, we decided to choose a more speci c de nition of delayed graft function.Beyond the discussion about the de nition, we have to state, that in the EDA group none of the patients who received RRT showed graft associated reasons leading to RRT.In no case of the EDA group, graft perfusion de cits or insu cient otherwise unexplainable graft dysfunction led to RRT.Finally, we can report of an association between EDA for donor nephrectomy and a lower rate of DGF in our study.However, we were fully aware that correlation does not proof causality.Especially in such a multifactorial context like living kidney transplantations, larger numbers of prospectively randomized assigned patients are needed to provide stronger evidence.

Conclusions
In this retrospective cohort study, we found an association between epidural analgesia for living kidney donors and signi cantly less delayed graft function in the corresponding kidney recipients.These results favor not only the bene cial analgesic effect of epidural analgesia for donors but also shows signi cant bene cial effects for kidney recipients.As our study has relevant shortcoming in terms of study design, number of index cases and adjustment of confounding variables, our ndings have to be con rmed by prospective randomized trials.The datasets generated and analyzed during the current study are not publicly available due institutional restrictions but are available from the corresponding author on reasonable request.
Authors contributions: initiated the study, performed statistically processing, contributed to the writing of the manuscript, UG developed the study design, contributed to data collection and proofread the manuscript, HB advised the study design, is responsible for the anesthesia SOPs and proofread the manuscript, BJ operated most of the patients is responsible for the postoperative care of the patients and contributed to data collection, KK contributed to data collection and writing of the manuscript, WB contributed to data collection and wrote the manuscript.

Tables
Surgical procedure was standardized to a maximum as only two different surgeons contributed to the transplantations in this cohort.The donor nephrectomy was performed in supine position over an open anterior extraperitoneal minimal incision laparotomy.Transplantations were performed in the established technique, to the right iliac fossa of the recipient.
Abbreviations -graft function

Figures Figure 1
Figures

Table 1 :
Main results of the study.Continuous variables are given as mean ± standard deviation, categorical variables are given as absolute number and percentage.

Table 2 :
Underlying reasons for renal replacement therapy (RRT) and met definition of delayed graft function (DGF).