LPDI

O

M

R

C

Ptcs(tccimtaTfwv

FC

R9

4

Adult Urology

aparoscopic Partial Nephrectomy in theresence of Multiple Renal Arteries

inesh Singh, Antonio Finelli, Mauricio Rubinstein, Mihir M. Desai, Jihad Kaouk, andnderbir S. Gill

BJECTIVES As experience with laparoscopic partial nephrectomy (LPN) expands, inevitably tumor-bearingkidneys with anomalous renal vasculature will be subjected to LPN. We evaluated LPN inkidneys with multiple arteries and compared those outcomes with the LPN outcomes in patientswith conventional renal arterial anatomy.

ETHODS Since September 1999, we have performed LPN for tumors in 333 patients. From this prospec-tively maintained database, we identified 60 patients with multiple renal arteries and 273patients with a single renal artery to the operated kidney. All patients underwent three-dimensional computed tomography preoperatively for accurate delineation of the tumor andrenal vascular anatomy. The clinical and operative data were reviewed to assess critical out-comes.

ESULTS The baseline parameters, including tumor size (P � 0.87), were similar in the two groups.Intraoperatively, the method of vascular control, tumor parenchymal extension depth (P �0.40), number requiring pelvicaliceal repair (P � 0.62), and specimen weight (P � 0.49) weresimilar between the two groups. Similarly, the warm ischemia time (P � 0.60), operative time(P � 0.15), blood loss (P � 0.37), and intraoperative (P � 0.52), postoperative (P � 0.48), andlate complication (P � 0.64) rates were similar between the two groups.

ONCLUSIONS LPN can be efficaciously performed in the presence of multiple renal vessels. Preoperative evaluationwith three-dimensional computed tomography is recommended to have preoperative knowledge ofthe renal vasculature and thereby minimize iatrogenic injury. UROLOGY 69: 444–447, 2007. © 2007

Elsevier Inc.

teaaack

M

Fppiaohtsaavtps

artial nephrectomy has become the standard ofcare as a nephron-sparing procedure in properlyselected patients with a Stage T1 tumor. Long-

erm 10-year follow-up data have shown oncologic out-omes similar to radical nephrectomy for comparablytaged patients.1,2 Laparoscopic partial nephrectomyLPN) is increasingly being used as a minimally invasiveechnique in the management of Stage T1 tumors. Be-ause the detection rate of small renal tumors has in-reased with the increased use of intraabdominal imag-ng, and because approximately 20% of kidneys haveultiple renal arteries, it is increasingly likely that pa-

ients with kidneys containing both a Stage T1 tumornd multiple renal arteries will be considered for LPN.he relevance of the issue of multiple renal arteries stems

rom the standpoint of hilar control, in that, comparedith open surgery, laparoscopic control of multiple renalessels may be technically more difficult, with a poten-

rom the Section of Laparoscopic and Robotic Surgery, Glickman Urological Institute,leveland Clinic Foundation, Cleveland, OhioReprint requests: Inderbir S. Gill, M.D., M.Ch., Section of Laparoscopic and

obotic Surgery, Glickman Urological Institute, A100, Cleveland Clinic Foundation,

l500 Euclid Avenue, Cleveland, OH 44195. E-mail: gilli@ccf.orgSubmitted: April 25, 2006, accepted (with revisions): October 5, 2006

44 © 2007 Elsevier Inc.All Rights Reserved

ially greater chance of complications. LPN in the pres-nce of multiple renal arteries has not previously beenddressed in published reports. We present our techniquend experience with LPN in kidneys with multiple renalrteries and retrospectively compared outcomes with aontemporaneous group of patients undergoing LPN inidneys with a conventional, single renal artery.

ATERIAL AND METHODS

rom September 1999 to June 2004, 333 LPNs have beenerformed at our institution by one surgeon. The data from thisrospectively collected computer database were reviewed todentify 60 patients (18%) (group 1) who had at least two renalrteries to the operated kidney. For the kidney to be classified asne with multiple renal arteries, at least two separate arteriesad to be independently originating from the aorta supplyinghe kidney. The remaining 273 patients (82%) (group 2) had aingle artery to the operated kidney. The preoperative variablesnalyzed included patient age, sex, laterality of surgery, surgicalpproach, tumor size, tumor depth, and tumor location (centralersus peripheral). The operative variables analyzed includedhe depth of intrarenal tumor extension and the distance to theelvicaliceal system, both measured by intraoperative ultra-onography, the method of hilar clamping, repair of the col-

ecting system, the method of renal parenchymal repair, and the

0090-4295/07/$32.00doi:10.1016/j.urology.2006.10.047

uiemt

esaapncmioetgmrsccpcf

fSs

RIr

sotAt

wns(sgt

owsa

(cd

ogists

U

se of any adjunct hemostatic agents. The outcomes assessedncluded the estimated blood loss, operative time, warm isch-mia time, specimen weight, final pathologic findings, surgicalargin status, length of stay, and the incidence of intraopera-

ive, postoperative, and late complications.Our technique of LPN has been previously described.3 How-

ver, in regard to LPN in the presence of multiple renal vessels,everal technical caveats apply. We prefer the transperitonealpproach when feasible, because this allows more working roomnd, most importantly, better suturing angles. Typically, in theresence of a preoperatively identified single renal artery, we doot advocate skeletonizing the renal vessels during LPN, be-ause it (a) is unnecessary for obtaining adequate clamping; (b)ay cause renal artery vasospasm; (c) risks iatrogenic vascular

njury; and (d) takes approximately 30 minutes of valuableperating time, which detracts from the primary mission. How-ver, in patients with multiple vessels, it is critically importanthat each vessel be securely clamped. Using the informationleaned from the preoperative three-dimensional computed to-ography (CT) scan, the precise locations of the accessory

enal vessels were noted. The polar vessels were individuallykeletonized. Generally, even the multiple vessels can be safelyontrolled en bloc with a single Satinsky clamp. However, inases in which a distant polar artery could not be safely encom-assed en bloc with the main renal hilum using the Satinskylamp, we used a dedicated, additional bulldog clamp (Micro-rance/Medtronic, Minneapolis, Minn) for the accessory artery.

Statistical analyses were performed using Statistical Packageor Social Sciences software (SPSS, Chicago, Ill) using thetudent t tests for continuous variables and either the chi-quare test or Fisher’s exact test for categorical variables.

ESULTSn our cohort, 60 (18%) of 333 patients had at least two

Table 1. Preoperative characteristics

CharacteristicGroup 1 (Multi

Group;

Mean age (yr) 62Sex (% male) 74Mean BMI 28Mean ASA 2Laterality (% right side) 62Preoperative serum creatinine (mg/dL) 1Mean tumor size by CT scan (cm) 2Location of tumor (% central) 36

BMI � body mass index; ASA � American Society of Anesthesiol

Table 2. Intraoperative variables

VariableGroup 1

Arte

Mean intrarenal extension of tumor (cm)Mean distance to PCS (mm)Tumors abutting PCS (%)Patients requiring PCS repair (%)Transperitoneal laparoscopic approach (%)Retroperitoneal laparoscopic approach (%)

PCS � pelvicaliceal system.

enal arteries. The preoperative characteristics of age, o

ROLOGY 69 (3), 2007

ex, body mass index, American Society of Anesthesiol-gists classification, tumor laterality, and CT-determinedumor size were similar between the two groups (Table 1).

trend was noted for more men (P � 0.054) and centralumors (P � 0.059) in group 1.

The measured intraoperative tumor characteristicsere similar between the two groups in terms of intrare-al tumor extension, tumors abutting the pelvicalicealystem, and frequency of pelvicaliceal system repairTable 2). The type of hilar clamp used differed, withignificantly more patients in the multiple renal arteryroup requiring the adjunctive use of a bulldog clamphan in the conventional group (P � 0.005; Table 3).

Both groups were similar in terms of the perioperativeutcomes, including estimated blood loss, operative time,arm ischemia time, weight of specimen resected, length of

tay, final pathologic findings, margin status, and intraoper-tive, postoperative, and late complications (Table 4).

Intraoperative complications occurred in 5 patients8.3%) in group 1 and 16 patients (5.9%) in group 2. Theomplications in the control group have been previouslyescribed.4 In group 1, the complications included hem-

enal Artery60)

Group 2 (Single Renal ArteryGroup; n � 273) P Value

59.4 0.15457.4 0.05429.8 0.3172.51 0.879

53.1 0.2691.00 0.1342.87 0.867

32.2 0.059

; CT � computed tomography.

ltiple Renaln � 60)

Group 2 (Single RenalArtery; n � 273) P Value

50 1.64 0.4029 3.81 0.5306 52.5 0.065 79.6 0.520 67.7 0.920 32.3 0.92

Table 3. Vascular clamp used

GroupSatinsky

(n)Bulldog

(n)

Satinskyand

Bulldog (n)

NoClamp

(n)

Multiple renalarteries

41 12 3 0

Single renalartery

201 63 3 1

ple Rn �

.0

.1

.9

.52

.1

.09

.84

.0

(Muries;

1.3.

39.81.70.30.

rrhage in 4 patients and ureteral transaction requiring

445

lrampacacfgmclwbrCpbc

gcpraiip

CTttpcethctm

tmtvptapat

etspdwebdti

epAbWeiisSc

ttoraim

4

aparoscopic ileal ureter in 1 patient. All four hemor-hagic complications resulted from a missed accessoryrtery. Of the four hemorrhagic complications, one wasanaged with open radical nephrectomy. In the other 3

atients, the hemorrhage was controlled laparoscopically,llowing successful completion of LPN. Postoperativeomplications occurred in 8 patients (13.3%) in group 1nd 24 patients (8.8%) in group 2. The group 1 compli-ations included hemorrhage managed with blood trans-usion in 2 patients, bilateral arm rhabdomyolysis andluteal hematoma in 1 patient, bilateral pleural effusionsanaged conservatively in 1 patient, ileus requiring

olonoscopy for decompression in 1 patient, acute tubu-ar necrosis requiring transient hemodialysis in 1 patientith compromised baseline renal function; devitalizedowel in an elderly patient with generalized atheroscle-osis requiring left colectomy on postoperative day 6,T-guided needle aspiration of a perirenal abscess in 1atient on postoperative day 14, and gastrointestinalleeding on postoperative day 26 managed endoscopi-ally in 1 patient.

Late complications occurred in 6 patients (10%) inroup 1 and 25 patients (9.2%) in group 2. Group 1omplications included readmission for an ileus in 1atient, delayed hemorrhage in 2 patients, one of whomequired angiographic embolization, perirenal hematomand sepsis treated with blood transfusion and antibioticsn 1 patient, gross hematuria in 1 patient, and myocardialnfarction requiring coronary stenting on placement onostoperative day 12 in 1 patient.

OMMENThe outcome of patients undergoing partial nephrec-

omy, using either the open or laparoscopic approach, inhe presence of multiple renal arteries has not beenreviously reported. Although minimal technical impli-ations may accrue during open surgery, the potentialxists for multiple renal vessels to considerably increasehe technical challenge during LPN. As such, this issueas relevance. The widespread use of CT scans has in-reased the detection of early-stage kidney tumors. Ashe worldwide experience with LPN continues to grow,

Table 4. Outcomes assessed

OutcomeMultiple R

(n

Blood loss (mL) 3Operative time (min) 2Warm ischemia time (min)Weight of specimen (g)Length of stay (hr)Final pathologic finding (% malignant)Intraoperative complications (n) 5 (Postoperative complications (n) 8 (Late complications (n) 6 (

Data in parentheses are percentages.

ore patients with multiple renal arteries and Stage T1 i

46

umors will require counseling regarding open versusinimally invasive nephron-sparing options. Demonstra-

ion of equivalent LPN outcomes in patients with singleersus multiple renal arteries will allow LPN to be ap-ropriately applied with confidence to the 20% of pa-ients with Stage T1 tumors who also have multiple renalrteries ipsilaterally. Our data, with 60 (18%) of 333atients undergoing LPN having multiple renal arteries,re consistent with the reported 20% incidence of mul-iple renal arteries.

Because before this analysis, it had been our subjectivexperience that the outcomes between patients with mul-iple renal arteries and those with a single renal artery areimilar, this vascular anomaly had no bearing on ouratient selection for LPN (reflected by the 20% inci-ence of multiple arteries in our series). No differencesere found in the preoperative clinical variables or op-rative variables between the two groups. The outcomesetween groups 1 and 2 were similar. Importantly, noifferences were found in the complications between thewo groups, particularly in terms of estimated blood loss,schemia time, transfusion rate, and oncologic outcomes.

Although similar outcomes to standard LPN can bexpected when performing LPN on patients with multi-le renal arteries, several caveats should be kept in mind.s is true for any partial nephrectomy technique, a

loodless field is a prerequisite for precise tumor excision.e prefer using a laparoscopic Satinsky clamp, which

nables en bloc hilar clamping without dissection ofndividual vessels, thus minimizing the risk of iatrogenicnjury to the renal vessels. In some instances, an acces-ory renal artery could not be incorporated into theatinsky clamp and, therefore, an additional bulldoglamp was used.

The number of renal arteries per kidney reported inhis study was determined from the CT findings and nothe intraoperative findings. We have previously describedur technique, which avoids the skeletonization of theenal vessels and the individual identification of arteriesnd veins because such maneuvers can lead to hilar vesselnjury. Therefore, many of our patients will have hadultiple renal arteries detected preoperatively, but not

Arteries)

Single Renal Artery(n � 273) P Value

241 0.37203 0.15

31.4 0.6049.7 0.4975.6 0.7371.1 0.28

16 (5.9) 0.54) 24 (8.8) 0.48

25 (9.2) 0.64

enal� 60

401832.259.178.866.18.3)13.310)

ndividually identified intraoperatively. A logical ques-

UROLOGY 69 (3), 2007

tfuopcprraer

oabpt

CLmcroovm

R1

2

3

4

5

EToL

Lcrtdceca

pcinrpmlMcgcelpcm

anwdoOmeopstpa

R1

JP

d©

U

ion would be whether preoperative imaging is adequateor detecting patients with multiple renal arteries. Thetility of this imaging modality has been demonstrated inur laparoscopic donor nephrectomy experience. In arospective study of 60 patients, three-dimensional CTorrectly identified the number of renal arteries in 59atients (98%) compared with conventional, invasiveenal arteriography.5 A three-dimensional CT scan isoutinely obtained as the only radiographic imaging forlmost all patients undergoing LPN. Having this preop-rative knowledge of the renal vasculature minimizes theisk of iatrogenic injury to vessels.

Finally, we prefer the transperitoneal approach becausef its larger working space and superior suturing angles. Itlso allows for space for placing a Satinsky clamp and aulldog clamp in cases in which this is necessary. Onlyosterior, medial tumors are approached using retroperi-oneoscopy, as previously described.

ONCLUSIONSPN can effectively be performed in patients who haveultiple renal arteries to the operative kidney. The out-

omes are similar to patients with a conventional, singleenal artery undergoing the same type of operation. Pre-perative evaluation with three-dimensional CT is rec-mmended to have preoperative knowledge of the renalasculature and thereby minimize iatrogenic injury. Careust be taken to obtain safe, adequate hilar clamping.

eferences. Fergany AF, Hafex KS, and Novick AC: Long-term results of

nephron sparing surgery for localized renal cell carcinoma: 10-yearfollowup. J Urol 163: 442–445, 2000.

. Herr HW: Partial nephrectomy for unilateral renal carcinoma and anormal contralateral kidney: 10-year followup. J Urol 161: 33–35,1999.

. Singh D, Rubenstein M, and Gill IS: Contemporary technique oflaparoscopic partial nephrectomy. J Endourol 19: 451–455, 2005.

. Ramani AP, Desai MM, Steinberg AP, et al: Complications oflaparoscopic partial nephrectomy in 200 cases. J Urol 173: 42–47,2005.

. El Fettouh HA, Herts BR, Nimeh T, et al: Prospective comparisonof 3-dimensional volume rendered computerized tomography andconventional renal arteriography for surgical planning in patientsundergoing laparoscopic donor nephrectomy. J Urol 170: 57–60,2003.

DITORIAL COMMENThe authors should again be congratulated on exploring an-ther interesting component of their robust experience with

PN. The association of complex renal vascular anatomy and A

ROLOGY 69 (3), 2007

PN will become an increasingly salient issue in urology be-ause of the continued increase in the number of small corticalenal neoplasms that are identified and the growing dissemina-ion and application of LPN. Additionally, their report ofuplication of the renal artery in 18% of patients is veryonsistent with the results described in a recent cadavericxtrarenal vascular anatomic study. In that study, Weld andolleagues1 demonstrated a 12% to 25% incidence of renalrtery duplication (depending on the definition used).

Several important technical considerations may help to im-rove the outcome and minimize morbidity with these complexases. As per the suggestion of the authors, high-quality axialmaging is important. Close attention to the vascular compo-ents of any high-quality contrast-enhanced CT or magneticesonance imaging study will, in most cases, help the surgeon torospectively identify challenging renal vascular anatomy andight help prevent vascular injury. Additionally, when using

aparoscopic bulldog clamps from any manufacturer (Aesculap,ircofrance, or Klein), the compressive force of a single bulldog

lamp is inadequate to stop arterial blood flow in any arteryreater than 4 mm in diameter. The application of two bulldoglamps is therefore useful to achieve a bloodless field. Of great-st significance, however, is the use of the duplex feature ofaparoscopic ultrasonography. After vascular clamping, the ap-lication of the duplex feature of laparoscopic ultrasonographyan definitively confirm complete vascular interruption andinimize the risk of intraoperative bleeding.As with most challenges in life, complex renal vascular

natomy during LPN also offers the urologic surgeon opportu-ity. In the Columbia University Medical Center experience,e often exploit complex vascular anatomy, such as renal arteryuplication, to minimize renal ischemia through the applicationf selective arterial ischemia by selective arterial clamping.ften with renal artery duplication, the tumor and surroundingargin targeted for extirpation can be selectively made isch-

mic by clamping part of the renal arterial vascular supply (eg,ne of two duplicated arteries, a segmental arterial branch, or aresegmental arterial branch). The duplex feature of laparo-copic ultrasonography can then be used to confirm that theargeted area (tumor and surrounding margin of normal renalarenchyma) is avascular for safe excision and reconstruction inbloodless field.

eferences. Weld KJ, Bhayani SB, Belani J, et al: Extrarenal vascular anatomy of

the kidney: an assessment of the variations and their relevance topartial nephrectomy. Urology 66: 985–989, 2005.

aime Landman, M.D.; Department of Urology, Columbia-resbyterian Medical Center, New York, New York

oi:10.1016/j.urology.2007.01.0952007 Elsevier Inc.

ll Rights Reserved

447