Outcomes after laparoscopic living donor nephrectomy: comparison of two laparoscopic surgeons with different levels of expertise


Correspondence: Frank Friedersdorff, Charitéplatz 1, 10117 Berlin, Germany.

e-mail: frank.friedersdorff@charite.de


What's known on the subject? and What does the study add?

  • Laparoscopic living donor nephrectomy has evolved as the procurement method of choice for living kidney donation. Given that this is a technically challenging procedure with potential risks for the healthy donor, skills transfer from an experienced laparoscopist to a novice is critical.
  • The present study investigates donor and recipient outcomes during a novice's early experience with this procedure. Previous training in laparoscopic renal surgery and mentoring by the expert helps the novice to generate acceptable outcomes. However, longer warm ischaemia times during the learning phase may affect short-term graft function.


  • To test the effect of surgeon experience on donor and recipient outcomes after laparoscopic living donor nephrectomy (LLDN). Results of a LLDN expert were compared with those of an LLDN novice.

Patients and Methods

  • Between October 2008 and October 2010 the last 20 cases of a series of 130 consecutive LLDNs, performed by an expert (EXP) were compared with the first 20 cases of an LLDN novice (NOV).
  • Donor and recipient outcomes were evaluated.
  • The novice was mentored by the expert during his initial four LLDN cases.


  • Donor and recipient demographics were not different between the two surgeon groups.
  • Total operating time and warm ischaemia time during LLDN was significantly longer in the NOV group compared with the EXP group (273 min vs 147 min and 213 s vs 162 s, respectively).
  • The incidence of donor complications was low in both groups. Length of hospital stay among donors did not differ between groups.
  • Although delayed graft function, rejection rates and postoperative serum creatinine levels indicated slightly poorer recipient outcomes in the NOV group, differences did not reach statistical significance.


  • Mentoring by an experienced urological laparoscopist may help an LLDN novice to generate acceptable donor and recipient outcomes.
  • Whether or not prolonged operating times and warm ischaemia times during the early phase of an LLDN experience are risk factors for impaired graft function needs further evaluation.

laparoscopic living donor nephrectomy


delayed graft function


Laparoscopic procurement of living donor kidneys is considered the gold standard for living donor renal transplantation in numerous renal transplant centres worldwide. Our LLDN experience is among the largest in Europe. We started LLDNs in 1999 and to date we have carried out more than 300 pure laparoscopic non-hand-assisted procedures. We perform ≈130 renal transplants per year with the proportion of living donor renal transplants ≈30%.

Safety concerns and reports on increased donor and recipient morbidity after LLDN were published regularly during the early days of LLDN [1, 2]. Attempts to minimize donor complications during the early phase of an LLDN experience include the use of hand-assisted laparoscopic techniques and guidance of the LLDN novice by an experienced LLDN surgeon throughout the initial cases [3-5]. Laparoscopic surgical procedures are known to have a relatively slow learning curve. In a large multicentre study, Vickers et al. [6] characterized the learning curves of open vs laparoscopic radical prostatectomy. They found that, in laparoscopic surgery, improvements in patient outcome accumulate more slowly than in open surgery [6]. Furthermore, increasing surgical expertise is associated with substantial reductions in cancer recurrence after laparoscopic radical prostatectomy.

The aim of our study was to show that an LLDN novice has acceptable donor and recipient outcomes during his early experience. We therefore compared an LLDN novice's initial 20 cases with the last 20 cases of an LLDN expert. The expert had previously performed 130 consecutive LLDNs and was largely experienced in other urological laparoscopic procedures. During his initial cases the novice received mentorship by the LLDN expert.

Patients and Methods


Our institution's electronic database was used to identify all LLDNs performed between August 2009 and October 2010. Both left and right side LLDNs were included. Patient records were retrospectively reviewed for donor and recipient complications and for graft outcomes. Outcomes of the initial 20 LLDNs performed by a novice for this procedure were compared with the last 20 cases of the 130 consecutive LLDNs performed by an LLDN expert.

Preoperative Evaluation of the Living Donor

All donors underwent extensive medical evaluation to rule out malignancy or chronic disease. Renal vascular imaging was performed with multi-slice CT angiography and magnetic resonance angiography. Routine assessment of donor split renal function was carried out using 99mTC-MAG3 renography. In the case of a relevant difference between the left and the right donor kidney (i.e. >5% difference in GFR), we procured the kidney with inferior function, regardless of the renal vascular anatomy.

Surgical Techniques

Laparoscopic donor nephrectomy

LLDN was performed by two surgeons (S.D. and T.F.F.) as a pure laparoscopic, non-hand-assisted procedure using a four-port transperitoneal access. A pneumoperitoneum of 15 mmHg was routinely used. Arteries were secured with 12 mm hemolock and titanium clips. For right-side LLDN, the renal vein was secured with an Endo-TA multifire stapler, while the left renal vein was typically secured and transsected with a vascular Endo-GIA stapler. Kidneys were extracted via a flank incision ≈6 cm in length before being flushed with ice-cold histidine tryptophan α-ketoglutarate solution.

Renal transplantation

Renal transplantation was performed by three experienced urological transplant surgeons (C.K., A.M. and S.H.), using a standard extraperitoneal technique through a Gibson incision. End-to-side anastomoses between the renal vessels and the external iliac vessels were performed. Ureteroneocystostomy was accomplished using the extravesical Lich–Gregoir technique. A ureteral JJ stent was routinely used. The stent was typically removed 6 weeks after transplant.

Previous Laparoscopic Training of the LLDN Novice

The inexperienced surgeon had previously assisted in 60 LLDNs. He performed more than 30 unsupervised transperitoneal laparoscopic tumour nephrectomies before embarking on LLDN. During the initial four LLDN cases, the novice was aided by the LLDN expert throughout the entire procedure. During the following six cases, which were performed under full responsibility of the novice, the LLDN expert served as backup and was quickly available in the case of an intraoperative complication. The LLDN novice performed his first 20 cases in a consecutive manner with a frequency of two cases per month.

Assessment of Graft Function

Graft function was monitored by Doppler ultrasound, serum creatinine and urine output measurements. Delayed graft function was defined as the need for at least one haemodialysis session during the first 7 days after transplantation. Delayed graft function and elevated (>0.8) resistance indices upon repeat measurements as well as a 20% rise in serum creatinine were indications for renal biopsy around postoperative day 7 to rule out rejection.


In all recipients, immunosuppression was started 1 day before transplantation. Immunosuppression consisted of prednisone, mycophenolate mofetil and a calcineurin inhibitor. Oral prednisone was tapered to a dose of 10 mg during the first 6 months. Cyclosporine and tacrolimus target serum levels were between 150 and 250 ng/mL and 10 and 12 ng/mL, respectively. Induction therapy with anti-lymphocyte agents or CD25 antibodies was applied in patients with a high immunological risk. Rejection episodes were treated with high-dose steroids.

Statistical Analysis

Statistical analysis was performed using SPSS Statistics 19.0. As our null hypothesis assumed worse patient outcomes in the novice group, all statistical tests were one-tailed. Means ± standard deviation of continuous variables were compared between the two study groups using Student's t test. Categorical variables were compared using Fisher's exact test. P < 0.05 was considered statistically significant.


Donor and Recipient Demographics

Donor and recipient demographic data are outlined in Tables 1 and 2, respectively. In both donors and recipients, patient age, gender distribution and body mass index (BMI) were not different between study groups. The distribution of left vs right side donor nephrectomies was similar between the two groups (Table 1). The mean human leucocyte antigen (HLA) mismatches were 3.1 for the expert group vs 2.6 in the novice group (Table 2).

Table 1. Donor demographics and postoperative outcomes.
 LLDN expert, n = 20LLDN novice, n = 20Mean difference (95% CI)P
Age (mean ± sd)50.7 (±10.1)45 (±10.1)5.6 (0.22–11.1)0.043
No. of left/right donor kidneys11/913/7  
No. of females/males12/89/11  
BMI, kg/m2 (mean ± sd)25.7 (±4.1)25.3 (±4.0)0.4 (–3.2–2.4)0.387
Warm ischaemia time, s (mean ± sd)162.2 (±32.9)213.2 (±67.4)51 (22.6–79.4)0.004
Total operating time, min (mean ± sd)147 (±27.5)270.9 (±47.8)123.9 (103–144.9)<0.001
Serum creatinine at discharge, mg/dL (mean ± sd)1.22 (±0.14)1.24 (±0.2)0.12 (–0.16–0.13)0.433
Length of hospital stay, days (mean ± sd)5.7 (±1.5)6.4 (±2.7)0.7 (–0.702–2.102)0.151
Complications32 0.500
Small haematoma10  
Wound infection10  
Clavien–Dindo classification    
Grade I20  
Grade II12  
Grade III00  
Grade IV00  
Table 2. Recipient demographics and outcomes.
 LLDN expert, n = 20LLDN novice, n = 20Mean difference (95% CI)P
Age (mean ± sd)39 (±13.9)35 (±13.8)3.55 (−3.8–10.1)0.423
No. of females/males10/108/12  
BMI (mean ± sd)24.1 (±4.41)24.2 (±6.38)0.1 (−3.6–3.8)0.962
HLA mismatches (mean)3.12.6  
Recipient diagnoses    
Polycystic kidney disease58  
Haemolytic-uraemic syndrome21  
IgA nephropathy23  
Vascular anastomosis time, min (mean ± sd)44.6 (±11.8)47.9 (±11.7)3.35 (−3.02–9.72)0.191
Acute rejection rate (%)2040 0.150
Delayed graft function (%)1020 0.331
1-year graft survival (%)100100  
Serum creatinine after transplantation    
1 week (mean ± sd)2.15 (±1.56)3.17 (±2.22)1.02 (−0.01–2.04)0.051
4 weeks (mean ± sd)1.55 (±0.69)1.64 (±0.53)0.09 (−0.24–0.42)0.324
26 weeks (mean ± sd)1.3 (±0.42)1.57 (±0.68)0.27 (−0.04–0.58)0.078
52 weeks (mean ± sd)1.27 (±0.38)1.38 (±0.44)0.11 (−0.14–0.35)0.229

Donor Outcomes

Postoperative donor outcomes are shown in Table 1. All 40 donor nephrectomies were completed laparoscopically without conversion to open nephrectomy. Mean warm ischaemia time during LLDN was significantly longer in the novice group (213.2 s vs 162.2 s; P = 0.004). Mean total operating time was significantly longer in the LLDN novice group compared with the expert group (253 min vs 147 min; P < 0.001). The incidence and severity of donor complications was low and did not differ between groups. All complications were characterized according to the Clavien–Dindo classification and resolved without sequelae after conservative treatment (Table 1). Average hospital stay was slightly longer in the novice group, without reaching statistical significance (6.4 days novice vs 5.7 days expert). Mean serum creatinine 1 week after LLDN was 1.2 mg/dL in both groups.

Recipient Outcomes

Recipient outcomes are shown in Table 2. Demographic data were similar between groups. One-year graft survival was 100% in both the LLDN expert and the LLDN novice group. The incidence of acute rejection was 40% in the novice group and 20% in the expert group (P = 0.15). The incidence of delayed graft function (DGF) in the novice and expert group was 20% and 10%, respectively (P = 0.33). Recipient complications in the LLDN expert group were as follows: three cases of an asymptomatic lymphocele and three cases of a postoperative haematoma. All complications resolved without surgical intervention. In the LLDN novice group two recipients required reoperation due to stenosis of the renal artery. Both patients recovered renal function early after surgical revision. In two recipients laparoscopic lymphocelectomy was necessary due to symptomatic lymphocele. Postoperative haemorrhage occurred in one recipient and was treated conservatively. Wound infection was observed in two recipients. Multiple renal arteries were observed in two cases in the experienced LLDN surgeon group and three cases in the inexperienced LLDN surgeon group. The evolution of graft function as measured by serial serum creatinine levels is shown in Fig. 1. One-week post transplant grafts in the LLDN expert group showed quicker recovery of graft function than grafts in the LLDN novice group (2.15 mg/dL vs 3.17 mg/dL; P = 0.051). Serum creatinine levels converged after 1 and 6 months (1.55 mg/dL vs 1.64 mg/dL and 1.3 mg/dL vs 1.57 mg/dL, respectively; P = 0.324 and P = 0.07). One-year post transplant serum creatinine levels were 1.27 mg/dL in the LLDN expert group vs 1.38 mg/dL (P = 0.23) in the LLDN novice group (Fig. 1).

Figure 1.

Comparison of recipient mean serum creatinine levels after transplantation.

Evolution of LLDN Total Operating Time in the Novice Group

Figure 2 shows total operating times of all LLDNs, including right and left side, performed by the novice. Left-sided LLDN had longer operating times than right-sided LLDN. The trend-line shows a gradual decrease in total operating time, most noticeable after case number 10.

Figure 2.

Evolution of the novice's LLDN total operating time.


Over the past decades, LLDN has become the gold standard especially in high volume renal transplant centres and in centres with a large expertise in laparoscopic urological surgery [7, 8]. Large series comparing open donor nephrectomy with LLDN have underscored that LLDN, although technically demanding, significantly reduces donor morbidity without compromising graft outcomes [9]. As with most complex surgical procedures showing a slow rate of learning over time, LLDN raised donor safety concerns among transplant professionals in its early days [10-12]. In an attempt to increase donor safety during the early phase of an LLDN experience, some authors have evaluated the benefits of a team approach for this procedure. Teams included either two proficient laparoscopists or a proficient laparoscopist and an urologist experienced in open donor nephrectomy [5, 8]. The major benefit of this approach includes a low donor morbidity rate coupled with favourable graft outcomes [5, 8].

At our institution, LLDN was introduced 12 years ago and included both the left and the right kidney. During the early phase of our experience, LLDN was performed by two proficient laparoscopists. After an initial team approach, both laparoscopists took turns in performing LLDN, in order to achieve an equal distribution of cases per surgeon within a reasonable period of time. Before embarking on LLDNs, the novice, whose early experience is herein presented, had significantly less exposure to laparoscopic urological surgery than his two predecessors. He had previously performed 30 unsupervised laparoscopic radical nephrectomies, while the other two surgeons were highly experienced in all types of laparoscopic urological procedures. Laparoscopic training of the LLDN novice has been detailed in the Methods section. Given that the evolution of the novice's total operating time could be used to measure learning, 20 LLDNs may be a reasonable number to show significant progress, especially after the first 10 cases (Fig. 2). Despite the fact that our data are purely descriptive, we feel that the most important determinant of learning speed during a novice's early LLDN experience is his level of previous laparoscopic expertise.

Others have chosen cut-offs of 37 and 44 cases, respectively, to define a surgeon's early LLDN experience [5, 13]. In these publications, authors have described the benefits of a team approach during the initial 40 cases to ensure maximum donor safety. In our opinion, an LLDN novice should be assigned full responsibility for this operation as early as possible, with an expert providing backup as needed. However, expert guidance through the initial four to five cases and previous mentored training in laparoscopic radical nephrectomy may help the novice to produce acceptable donor and recipient outcomes during his early LLDN experience.

Not surprisingly, we found that the novice surgeon had longer total operating times and warm ischaemia times than the expert surgeon. In one of the largest single institution retrospective studies of recent years, Chin et al. [14] found that, despite longer operating times and warm ischaemia times, donor and graft outcomes are excellent during a surgeon's initial LLDN experience. In our cohort, the average operating time and average warm ischaemia time are in line with those previously reported by Chin et al. In most large LLDN series warm ischaemia times vary between 2 and 5 min [13, 15-18].

Prolonged warm ischaemia time has been considered the major disadvantage of pure laparoscopic LLDN over hand-assisted laparoscopic or open donor nephrectomy [18]. Simforoosh et al. [19] reported that warm ischaemia times up to 10 min during LLDN have no adverse effects on graft function and survival. Soulsby et al. [17] examined the effect of warm ischaemia time during LLDN on graft function and rejection episodes. Recipients were divided into two groups. In one recipient group warm ischaemia times during organ retrieval were shorter than 3 min. The authors failed to show a detrimental effect of warm ischaemia times >3 min on 1-year graft function. However, they noticed a trend towards a higher acute rejection rate in the group with prolonged warm ischaemia times. A large retrospective study on 946 recipients of laparoscopically procured living donor kidneys demonstrated that poor early graft function after LLDN has a deleterious effect on long-term graft function, acute rejection-free survival and overall graft survival [20]. Among others, a high recipient BMI and prolonged warm ischaemia time were identified as risk factors for poor early graft function in this population. We found a higher incidence of acute rejection and DGF rates as well a slower recovery of recipient serum creatinine in the novice group (Fig. 1). Our findings appear to corroborate those of Nogueira et al. [20] in that prolonged warm ischaemia times during LLDN may translate into higher DGF and acute rejection rates. As our study design allotted a relatively small patient number to each study group, we were unable to prove a statistical correlation between prolonged warm ischaemia times produced by the LLDN novice and poorer graft outcomes. Comparing our two study populations, we noticed that recipient serum creatinine values converged at 12 months after transplant (Fig. 1), suggesting full recovery of impaired initial graft function in the LLDN novice group.


We found that an LLDN novice generates acceptable graft outcomes with minimal donor morbidity during his 20 initial cases. We recommend that before embarking on LLDN, surgeons should be trained in laparoscopic radical tumour nephrectomy, preferably under the guidance of an experienced urological laparoscopist. Whether or not prolonged warm ischaemia times, which are unavoidable during the early phase of a surgeon's LLDN experience, increase acute rejection and DGF rates should be subject to future investigation.

Conflict of Interest

None declared.