Laparoscopic Procurement of Kidneys with Multiple Renal Arteries is Associated with Increased Ureteral Complications in the Recipient

Abstract

This study investigates the effect of renal artery multiplicity on donor and recipient outcomes after laparoscopic donor nephrectomy. Three-hundred and sixty-one sequential procedures were performed over a 4-year period. Forty-nine involved accessory renal arteries; of these, 36 required revascularization and 13 were small polar vessels and ligated. The 312 remaining kidneys with single arteries served as controls. Study variables included operative times, blood loss, hospital stay, graft function and donor and recipient complications.

Kidneys with multiple revascularized arteries had a longer mean warm ischemia time (35.3 vs. 29.2 min, p = 0.0003), and more ureteral complications (6/36 vs. 10/312, p = 0.0013) than single-artery controls. In contrast, ligation of a small superior accessory artery had no significant effect on donor operative time, blood loss, or complication rate while providing similar recipient graft function compared to single-artery controls.

Renal artery number is important in selecting the appropriate kidney for laparoscopic procurement. Given the current excellent results with right-sided donor nephrectomy, kidneys with single arteries should be preferentially procured, irrespective of side.

Introduction

Since the introduction of laparoscopic donor nephrectomy in 1995, the rate of living donor kidney transplantation in the United States has grown such that the number of annual living kidney donors now exceeds the number of cadaveric donors (1–3). The main benefits of the laparoscopic approach to the donor are reduced pain, shorter hospitalization, earlier return-to-work, improved cosmesis and greater overall satisfaction when compared with the traditional open approach to donor nephrectomy (4–8). In the recipient, early reports showed a higher incidence of ureteral complications associated with the laparoscopic approach (9). With subsequent technical refinement, however, the rate of ureteral complications has been reduced to that of the open approach (10). In fact, most centers currently report that overall recipient and graft outcomes with the open and laparoscopic approaches are equivalent (11,12).

Before the laparoscopic era, the most common indication for right-sided donor nephrectomy was the presence of multiple renal arteries in the left kidney (13–15). The rationale was to avoid the higher rate of vascular and ureteral complications associated with multiple renal arteries (16–24). In contrast, many centers now practicing laparoscopic donor nephrectomy prefer left kidneys for procurement, even those with multiple renal arteries (9,12,25). Laparoscopic procurement of right kidneys is commonly avoided because initial reports showed high rates of venous complications and graft loss, even though more recent studies demonstrate essentially equivalent outcomes as compared with left laparoscopic nephrectomies (26,27).

Experience with laparoscopic procurement of kidneys with multiple arteries is limited. Most studies to date have shown similar donor and graft outcomes when compared with single-artery kidneys, but these results may have been influenced by small cohort size and failure to differentiate small, accessory arteries, which can be safely ligated, from larger caliber and inferior pole accessory arteries, which require revascularization (28–30). In the present study, we describe our experience with laparoscopic donor nephrectomies focusing on the effect of multiple arteries on donor and recipient outcomes.

Materials and Methods

Patients and data collection

From November 1, 1999 to November 30, 2003, 361 consecutive laparoscopic donor nephrectomies were performed at the University of California, San Francisco by three surgeons (C.F., S-M.K., A.P.). During this period, 312 (86%) procured kidneys had a single artery and 49 (14%) had multiple arteries. Pre-operative donor evaluations included a history and physical examination by a transplant nephrologist, renal function assessment with serum electrolytes and creatinine and a urinalysis with spot urinary protein. Renal anatomy was evaluated with renal angiograms and intravenous pyelograms in all but two of the more recently evaluated donors, in whom high-resolution CT angiograms with three-dimensional reconstructions were used as the sole imaging study.

Demographic data were gathered from patient charts and clinic visits and were analyzed retrospectively after protocol approval by the Committee on Human Research, which is the Institutional Review Board of the University of California, San Francisco. The following intra-operative donor variables were collected prospectively during each case: (i) operative time, defined as the time from initial incision to closure of the final skin incision, (ii) extraction time, defined as the time from the first renal artery clipping to initiation of the back table flush, and (iii) estimated blood loss. Donor variables also included age, gender, height, weight, body mass index, side of nephrectomy, length of hospital stay, need for blood transfusions and other complications. Recipient variables included age, gender, height, weight, body mass index, warm ischemia time (defined as the time from removal of the kidney out of slush to reperfusion), requirement for dialysis in the first week after transplant, renal function (assessed by decrease in creatinine over the first 24 h after transplant, creatinine level 1 month after transplant, and estimated creatinine clearance by the Cockcroft-Gault formula 1 month and 4–6 months after transplant) and graft-related complications.

Operative technique

The operative procedure employed in this study has been described previously for both left and right sides, (27,31,32). Briefly, the patient is positioned in a flexed, modified lateral decubitus position and pneumoperitoneum is achieved using a Veress needle inserted into the left or right subcostal space. Three 11-mm trochars and one 12-mm trochar are then introduced into the right or left upper abdomen. The 12-mm port is used for introduction of the clip applier and vascular stapler. The entire dissection is completed without the use of hand ports or other manual assist devices and care is taken to preserve generous amounts of peri-ureteral tissue to minimize disruption of the ureteral blood supply. On the left side, the ureter is dissected together with the gonadal vein and separated only on the back table. On the right side, the ureter is dissected independent of the gonadal vein taking care to preserve all peri-ureteral tissue.

When the kidney has been fully mobilized, the pneumoperitoneum is relaxed and a 7–8 cm transverse suprapubic (Pfannenstiel) incision is made. The patient receives intravenous lasix, mannitol and heparin, after which the distal end of the ureter is double-clipped and transected. The renal artery is double-clipped using a combination of a locking plastic clip (Hem-o-lok) and a metal clip and divided distal to both clips with endoshears, taking care to preserve an adequate cuff of 2–3 mm on the clipped stump. If there are multiple arteries, each vessel arising from the aorta is occluded with both a locking vascular and metal clip. Superior pole accessory arteries are ligated in the donor procedure only when the diameter is 2 mm or less, or when the renal cortical area supplied is 10% or less, as judged intra-operatively. The heparin is reversed with protamine and the renal vein is stapled using a 2.5-mm vascular TA stapler (US Surgical, Norwalk, CT, USA). The vein is divided distal to the staple line with laparoscopic scissors and the kidney is removed through the suprapubic incision. Post-operatively, patients receive patient-controlled analgesia machines for the first night and are encouraged to begin walking as soon as possible. Oral intake is begun with clear liquids on the day of surgery and advanced as tolerated.

Graft preparation

Organs were perfused with ice-cold preservation solution instilled through all non-ligated renal arteries immediately after removal from the donor. Back-table dissection of the renal hilum was kept to a minimum and performed only in instances where additional mobilization of renal arteries was required to perform an ex vivo reconstruction into a common arterial channel. All vascular reconstructions were performed using standard microvascular techniques.

Statistical analysis

Differences among the three study groups were compared using the Kruskal-Walis test for continuous variables and χ² test for categorical variables. The odds ratios (ORs), 95% confidence intervals (CIs) and p-values for the development of ureteral complications associated with donor, recipient and transplant predictors were determined by univariate logistic regression. In order to determine whether the association of ureteral complications with multiple revascularized arteries could be explained by any other predictor(s), a bivariate regression model was constructed. The bivariate model adjusted the OR for ureteral complications from multiple revascularized arteries for every other predictor in the model, in pairwise fashion. Predictors in the univariate analysis found to be significant (p < 0.05), or with a trend toward association (p < 0.2), were included in the bivariate model. Multivariate models with three or more simultaneous predictors were not examined because of the limited number of ureteral complications and because results of two predictor models suggested that this was unnecessary. Statistical significance was defined at p < 0.05. All statistical analyses were performed by the UCSF Department of Biostatistics using SAS, version 9.0 software (SAS Institute, Cary, NC).

Results

We compared three groups with respect to donor and recipient outcomes. The first group consisted of 13 kidneys in which the accessory artery was ligated during the donor procedure. In all cases, these were small superior pole arteries with diameters less than 2 mm. The second group consisted of 36 kidneys in which multiple arteries were revascularized. The third group consisted of 312 donor kidneys with single renal arteries and served as the control group for this study.

In patients with single renal arteries, veins and ureters bilaterally, the left kidney was preferentially procured. In patients with multiple renal arteries on one side, the kidney with a single artery (right or left) was preferentially procured. Indications for procurement of kidneys with multiple vessels included multiple arteries bilaterally (62%), multiple left renal arteries with early bifurcation of the right renal artery (20%), accessory artery not seen on pre-operative imaging (9%), early fibromuscular dysplasia (6%) and ureteropelvic junction narrowing (3%).

Donor and recipient demographics are listed in Table 1. No significant differences in donor age, gender, height, weight, body mass index or side of donated kidney were observed. Similarly, no significant differences in recipient age, gender, height, weight or body mass index were observed between cohorts.

Table 1.  Donor and recipient demographics

	Multiple arteries		Single arteries Controls (n = 312)	p-values*
	Revascularized (n = 36)	Ligated (n = 13)
*Kruskal-Walis for continuous variables; χ^2 for categorical variables. ^†Mean ± standard deviation.
Donor age (years)^†	42.6 ± 11.6	43.1 ± 6.0	39.7 ± 11.2	0.1287
Donor gender
Male (%)	21 (58)	4 (31)	127 (41)	0.0896
Female (%)	15 (42)	9 (69)	185 (59)
Donor height (cm)^†	171 ± 9	166 ± 12	168 ± 10	0.1834
Donor weight (kg)^†	75.1 ± 11.6	72.4 ± 22.9	73.9 ± 15.3	0.3692
Donor BMI^†	25.6 ± 2.7	26.8 ± 7.1	25.8 ± 4.2	0.9608
Side of nephrectomy
Left (%)	30 (83)	12 (92)	270 (87)	0.7117
Right (%)	6 (17)	1 (8)	42 (13)
Recipient Age (years)^†	39.2 ± 15.9	36.1 ± 12.3	42.8 ± 15.4	0.1308
Recipient gender
Male (%)	18 (50)	6 (46)	190 (61)	0.2794
Female (%)	18 (50)	7 (54)	122 (39)
Recipient height (cm)^†	165 ± 15	164 ± 10	169 ± 14	0.0625
Recipient weight (kg)^†	67.9 ± 23.5	63.7 ± 16.3	71.6 ± 18.9	0.1006
Recipient BMI^†	24.4 ± 5.9	23.2 ± 3.6	24.8 ± 4.7	0.4903

Donor outcomes are listed in Table 2. No significant differences in operative time, blood loss, extraction time or length of stay were observed among the three groups. Additionally, no significant differences in the overall rate of donor complications were observed. Specifically, in kidneys with multiple revascularized arteries, one donor required a transfusion of three units of packed red blood cells for port site bleeding exacerbated by heparin over anti-coagulation that was not recognized immediately. Two patients sustained small bowel injury that was repaired intra-operatively without sequelae; the first requiring conversion to an open procedure. A third patient developed a wound infection that required readmission and surgical incision and drainage. In the donors with ligated accessory arteries, no complications were observed. In the control group, there was one open conversion for a stapler malfunction which required transfusion, and three other donors who also required transfusions. The remainder of the observed complications are listed in Table 2.

Table 2.  Donor outcomes

	Multiple arteries		Single arteries Controls (n = 312)	p-values*
	Revascularized (n = 36)	Ligated (n = 13)
*Kruskal-Wallis for continuous variables; χ^2 for categorical variables. ^†Mean ± standard deviation.
Operative time (min)^†	203 ± 64	219 ± 36	202 ± 46	0.0641
Blood loss (mL)^†	76 ± 33	71 ± 23	76 ± 64	0.6398
Extraction time (s)^†	290 ± 121	272 ± 87	256 ± 87	0.0945
Length of stay (days)^†	3.4 ± 1.1	3.2 ± 0.6	3.2 ± 1.1	0.5070
Donor complications
Conversion to open	1	0	1
Transfusion	1	0	4
Bowel injury	2	0	1
Wound infection	1	0	4
Urinary infection	0	0	1
Venous thrombosis	0	0	1
Pulmonary embolism	0	0	1
Rhabdomyolysis	0	0	1
Pneumonia	0	0	1
Port site hernia	0	0	1
Death	0	0	0
Total complications (%)	5 (14)	0 (0)	16 (5)	0.0690

Several techniques were used to revascularize the accessory vessels in the 36 kidneys with reimplantable multiple arteries (Table 3). In 20 cases (56%), the arteries were implanted separately into the external iliac artery. In 11 cases (30%), arteries were joined ex vivo to form a common channel and then implanted into the external iliac artery. In three cases (8%), the smaller accessory artery was joined end to side into the main renal artery. In one case (3%), the kidney's two major arteries were joined ex vivo to form a common channel and a third artery was anastomosed separately into the external iliac artery. In the last case (3%), the main renal artery was implanted into the external iliac artery and a 3-mm inferior pole accessory artery was anastomosed to the inferior epigastric artery end to end.

Table 3.  Ureteral complications in kidneys with multiple revascularized arteries stratified by technique of vascular reconstruction

Technique of reconstruction (n)	Ureteral complications		p-value
	Yes	No
χ^2= 1.058; p = 1.000.
Separate implantation to external iliac artery (20)	3 (15%)	17 (85%)	1.000*
Joined ex vivo to form common channel (11)	2 (18%)	9 (82%)
Accessory artery joined end to side to main renal artery (3)	1 (33%)	2 (67%)
Other (2)	0	2 (100%)
Total (36)	6 (17%)	30 (83%)

Recipient outcomes are listed in Table 4. Mean warm ischemia time for kidneys with multiple revascularized arteries was longer (35.3 ± 9.5 min) than that observed in the ligated artery (28.2 ± 5.9 min) and control (29.2 ± 7.2 min, p = 0.0003) groups. No significant differences were observed among the three groups in the initial drop of serum creatinine over the first 24 h, serum creatinine 1 month after transplant and estimated creatinine clearance (by the Cockcroft-Gault formula) at 1 month, but there was a trend toward slightly lower creatinine clearance in kidneys with ligated accessory superior pole arteries. This trend did not progress; however, the mean creatinine clearance in this group 4–6 months after transplantation was higher than at 1 month (65 ± 15.6 vs. 54 ± 13, respectively) and was comparable to the 6-month mean creatinine clearance of the multiple revascularized artery group (61.1 ± 16.9). Dialysis requirement in the first week after transplant, rate of lymphocele formation requiring drainage and graft loss were similar among the three groups. The one graft loss of a kidney with multiple revascularized arteries occurred 10 months after transplant and resulted from a combination of polyoma virus and calcineurin inhibitor nephrotoxicity.

Table 4.  Recipient outcomes

	Multiple arteries		Single arteries Controls (n = 312)	p-values*
	Revascularized (n = 36)	Ligated (n = 13)
*Kruskal-Walis for continuous variables; χ^2 for categorical variables. ^†Mean ± standard deviation. ^‡Mean ± standard deviation, as calculated from Cockcroft-Gault formula = (140 − age) × weight/72 × creatinine, ×0.85 if female.
Warm ischemia time (min)^†	35.3 ± 9.5	28.2 ± 5.9	29.2 ± 7.2	0.0003
Serum creatinine change in first 24 h^†	−2.7 ± 1.5	−3.4 ± 2.2	−3.1 ± 2.3	0.8472
Serum creatinine at 1 month^†	1.4 ± 0.6	1.6 ± 0.5	1.4 ± 0.6	0.1636
Estimated creatinine clearance at 1 month^‡	68 ± 24	54 ± 13	71 ± 35	0.1226
Dialysis in first week (%)	3 (8)	0 (0)	14 (4)	0.4210
Ureter stricture/leak (%)	6 (17)	1 (8)	10 (3)	0.0013
Vascular thrombosis (%)	1 (3)	0 (0)	1 (0.3)	0.1965
Lymphocele (%)	2 (6)	0 (0)	2 (0.6)	0.0547
Graft loss (%)	1(3)	0 (0)	4 (1)	0.4227

Two vascular complications were observed in our series. One recipient received a kidney with an accessory inferior pole artery implanted into the sidewall of the main renal artery. This inferior pole vessel thrombosed within 24 h of transplant, which did not result in graft loss but did require ureteropyelostomy as a result of distal ureteral necrosis. The other vascular complication consisted of graft loss from thrombosis following myocardial infarction in a recipient who received a left kidney with a single artery and vein.

Ureteral complications occurred in 6 of 36 (17%) recipients of kidneys with reimplanted accessory arteries compared with 10 of 312 (3%) control recipients and 1 of 13 (8%) recipients of kidneys with ligated accessory arteries (p = 0.0013). Four of the patients receiving kidneys with revascularized accessory arteries developed distal ureteral necrosis that required operative revision. The fifth patient developed a ureteral stricture 10 days following transplant and the sixth developed ureteropelvic junction obstruction 3 months after transplant. Although the sample sizes were small, no statistically significant correlation between the type of vascular reconstruction and ureteral complications was observed (Table 3). Only one ureteral complication was observed in the 13 recipients of kidneys with ligated accessory arteries; this consisted of a late ureteral stricture requiring operative repair 4 months after transplant.

To further assess predictors of ureteral complications after laparoscopic donor nephrectomy, univariate logistic regression using donor, recipient and transplant variables was performed (Table 5). Univariate analysis showed that three variables: multiple revascularized arteries (OR 5.71; 95% CI = 1.97–16.53; p = 0.0013), donor age > 50 years (OR 4.04; 95% CI = 1.50–10.89; p = 0.0057) and recipient male gender (OR 0.36; 95% CI = 0.13–0.99; p = 0.0471) were significantly associated with the development of recipient ureteral complications. No significant associations were found with donor gender, weight, body mass index, operative time, blood loss, side of kidney procured, recipient age, height, weight, body mass index, warm ischemia time, institutional experience, individual surgeon or individual surgeon's experience.

Table 5.  Univariate analysis of risk factors for developing ureteral complications following laparoscopic kidney procurement

Variables	Odds ratio	95% confidence interval	p-values
Multiple revascularized arteries	5.71	1.97–16.53	0.0013
Donor age > 50 years	4.04	1.50–10.89	0.0057
Donor male gender	0.41	0.13–1.28	0.1230
Donor weight (per kg)	0.98	0.95–1.02	0.3224
Donor weight > 80 kg	0.45	0.13–1.59	0.2150
Donor BMI (per 1.0 increment)	0.98	0.86–1.11	0.7482
Donor BMI > 25 kg/m^2	1.75	0.63–4.84	0.2808
Right-sided kidney	1.39	0.38–5.02	0.6169
Recipient age > 50 years	0.82	0.28–2.37	0.7078
Recipient height (per 10 cm)	0.81	0.61–1.06	0.1221
Recipient weight (per kg)	0.99	0.96–1.01	0.3138
Recipient weight > 80 kg	0.73	0.23–2.30	0.5929
Recipient BMI (per 1.0 increment)	1.01	0.91–1.11	0.9189
Recipient BMI > 25 kg/m^2	0.73	0.26–2.03	0.5492
Recipient male gender	0.36	0.13–0.99	0.0471
Warm ischemia (per minute)	1.05	0.99–1.10	0.0826
Warm ischemia > 30 minutes	1.70	0.62–4.68	0.3051
Operative time (per minute)	1.01	1.00–1.01	0.2780
Operative time > 240 min	2.45	0.83–7.26	0.1055
EBL > 75 cc	1.00	0.34–2.91	0.9976
Institution's first 100 cases	0.90	0.31–2.63	0.8508
Surgeon
No. 1 vs. others	1.31	0.42–4.12	0.6409
No. 2 vs. others	1.10	0.35–3.48	0.8665
No. 3 vs. others	0.01	0.01–3.48	0.3967
Surgeon's first 50 cases	0.81	0.28–2.35	0.6960

To ensure that the association between multiple revascularized arteries and ureteral complications could not be explained by any other predictor(s), a bivariate regression model was constructed, which estimated the OR for developing ureteral complications in kidneys with multiple revascularized arteries, adjusting for each predictor in the univariate model having a p-value <0.2 (Table 6). When adjusted for each predictor separately, the risk of developing ureteral complications in kidneys with multiple revascularized arteries remained very high (OR = 5.20–7.26) and retained a high degree of statistical significance (p = 0.0005–0.0036) in all cases.

Table 6.  Risk of developing ureteral complications in kidneys with multiple revascularized arteries after adjustment for other predictors

	Risk of ureteral complications
	Odds ratio for multiple revascularized arteries	95% confidence interval	p-values
Univariate model (from Table 4)	5.71	1.97–16.53	0.0013
Bivariate model, also adjusted for:
Donor age > 50 years	5.20	1.75–15.46	0.0030
Donor male gender	7.26	2.39–22.04	0.0005
Recipient height (per 10 cm)	6.02	2.04–17.73	0.0011
Recipient male gender	5.33	1.82–15.62	0.0023
Warm ischemia (per min)	5.48	1.74–17.26	0.0036
Operative time > 240 min	5.55	1.90–16.20	0.0017

Discussion

Although laparoscopy has become the preferred approach for procuring kidneys from living donors, this technique has generally been limited to the procurement of left kidneys (9,12,25). Laparoscopic procurement of right kidneys has been avoided because early reports described higher rates of graft loss and longer duration of delayed graft function, both of which were attributed to the shorter length of the right renal vein and resultant difficulty in implanting the organ. Mandal et al. reported that of eight right kidneys procured laparoscopically, three developed vascular thromboses resulting in graft loss (26). These early findings prompted some centers performing laparoscopic nephrectomy to preferentially procure left kidneys, even if this entailed using organs with more complex vascular anatomy.

The outcomes associated with multiple renal arteries in the laparoscopic era are not well characterized. Published studies have found no statistically significant differences in donor outcome, recipient ureteral complications or recipient vascular complications associated with multiple arteries in kidneys procured laparoscopically versus single-artery kidneys (28–30). In the largest series published to date, Hsu et al. reported on 353 laparoscopic donor nephrectomies and found no significant differences in donor operative time, warm ischemia time, blood loss, length of stay, surgical complications, graft survival or serial creatinine concentrations in single versus multiple artery kidneys (29). Urological complications occurred in 16 of 277 (6%) recipients of single-artery kidneys versus 4 of 71 (6%) recipients of kidneys with multiple-artery kidneys (29).

These results, however, conflict with data from the pre-laparoscopic era regarding the effect of multiple arteries on recipient outcome. Roza et al. reported on 42 living donor open nephrectomies with multiple renal arteries, in which 8 (19%) urological complications (including 5 leaks and 3 obstructions) and 3 (7%) vascular complications were observed (22). Similarly, Belzer et al. found a higher incidence of urinary fistulae in recipients of kidneys with multiple arteries (4.8%), particularly when one of the arteries was injured (10.4%), than in single-artery transplants (1.3%) (21). Several other authors have reported increased ureteral complications associated with multiple arteries, especially in living donor open nephrectomies (23,24). These complications may be related to the absence of a Carrell patch, which facilitates reimplantation of vessels in cadaveric organs, and the smaller diameter of accessory arteries, making them more susceptible to thrombosis and technical errors.

In our series, we noted several important differences in donor and recipient outcomes among kidneys with single versus multiple arteries. First, we found that small accessory superior pole arteries could be safely ligated without a significant effect on serum creatinine concentration, graft function or other recipient outcomes. These accessory arteries were all less than 2 mm in diameter, supplied the superior pole only and resulted renal cortical ischemia of less than 10% based upon intra-operative findings. Although there was a trend at 1 month toward increased creatinine (1.6 vs. 1.4 mg/dL) and reduced estimated creatinine clearance by the Cockcroft-Gault formula (54 vs. 71 cc/min), these differences did not achieve statistical significance in this small cohort of 13 patients and did not persist, as evidenced by the increased 6-month mean creatinine clearance in the ligated superior pole artery group. Future investigation with a larger patient population will be required to further characterize the effect of ligating small superior pole vessels in the donor procedure. Second, warm ischemia times for kidneys with multiple renal arteries requiring revascularization are on average 6.1 min longer than those for kidneys with single arteries. Third, kidneys with multiple arteries that require revascularization are more likely to develop ureteral complications (Tables 4–6). This finding conflicts with the current literature on laparascopically procured kidneys with multiple vessels, and recapitulates findings from the open nephrectomy era. One possible reason for this difference is that prior studies of laparoscopic donor nephrectomy did not differentiate between revascularized and ligated accessory renal arteries.

We hypothesize that the increase in ureteral complications associated with multiple revascularized renal arteries is most likely a consequence of insufficient inferior pole perfusion producing relative ischemia in the ureter. Such relative ischemia may result in leak (from necrosis of the distal ureter) or stricture. Although we did not prospectively image the renal artery flow in this study, we speculate malperfusion of accessory arteries may result from several factors: small anastomoses producing a flow-limiting stenosis, greater susceptibility of small arteries to traction or cautery injury, greater turbulence and shear forces resulting from increased flow velocity, greater risk of technical errors or a combination thereof. Interestingly, the development of ureteral injuries in the present study did not appear to correlate with the type of arterial reconstruction that was used, although the number of patients in each group was quite small.

Although unintended disruption of the peri-ureteral blood supply during laparoscopic dissection could theoretically explain the increased complications observed, we believe this is unlikely since the overall rate of ureteral complications in our single-artery group is 3%. This low rate of ureteral complications compares favorably with recent literature: Ratner et al. reported a 3% ureteral complication rate in 100 consecutive laparoscopic donor nephrectomies after technical improvement in the donor ureteral dissection, compared with a 6.3% incidence in 48 consecutive open cases (12). Furthermore, in our logistic regression, ureteral complications were not associated with institutional experience, individual surgeon or individual surgeon's experience, eliminating a ‘learning curve’ effect as a possible explanation for the observed association. Similarly, it seems unlikely that extensive hilar dissection of kidneys in an effort to gain vessel length is responsible for the increased rate of ureteral complications in this series, since we do not routinely perform such ex vivo hilar dissections at our institution.

Given these results, preferential selection of left kidneys for laparoscopic procurement in the setting of multiple left renal arteries is called into question. We have performed 49 laparoscopic right-donor nephrectomies at our center with equivalent donor and recipient outcomes to laparoscopically procured left kidneys (33). As a result, we have now modified selection of the appropriate kidney for laparoscopic procurement at our center in a way that is similar to the criteria used in the past when the open procedure was the standard of care.

In conclusion, this study describes our 4-year experience with laparoscopic donor nephrectomy of kidneys with multiple arteries. Laparoscopically procured kidneys with small, superior pole accessory arteries less than 2 mm diameter and supplying <10% of the renal cortex may be managed by simple ligation with acceptable outcome. On the other hand, kidneys with multiple vessels requiring revascularization are associated with longer warm ischemia time and increased ureteral complications compared to their single-artery counterparts. In such circumstances, laparoscopic procurement of a kidney with a single artery, left or right, is preferable.