S

RRPRA

O

M

R

C

Rtti1trm8riascucto

FC

I

1

urgical Challenges in Radical Prostatectomy

obotic Repair of Rectovesical Fistulaesulting From Open Radicalrostatectomy

ene Sotelo, Robert de Andrade, Oswaldo Carmona, Juan Astigueta,lejandro Velasquez, Gustavo Trujillo, and David Canes

BJECTIVES Rectovesical fistula (RVF) is a rare complication of radical prostatectomy. A 57-year-old manunderwent open radical prostatectomy with recognized rectal injury, primary closure of the rectalwall, and loop colostomy. The patient developed urine leakage per rectum after colostomyclosure. We diverted the fecal stream with end colostomy and placed a suprapubic tube. An opentranssacral (Kraske) repair failed 1 month later. We have previously described the laparoscopicapproach, and report the technique and results of our first robotic assisted operation.

ETHODS The operative steps were as follows: (1) cystoscopy, (2) RVF catheterization (3) five-porttransperitoneal laparoscopic initial dissection (4) mobilization of omental pedicle flap, (4)cystotomy extending toward the fistulous tract, (5) robot docking (6) dissection of the recto-vesical plane, (7) interrupted rectal closure, (8) omental interposition, (9) bladder closure, and(10) drain placement.

ESULTS Operative time was 180 minutes. Hospital stay was 1 day. The suprapubic tube was removed at2 months after normal cystography. Bowel continuity was restored at 4 months, with no fistularecurrence at 1-month follow-up.

ONCLUSIONS We await longer follow-up and experience in larger series. For now, robotic repair of rectovesicalfistula appears feasible and represents an attractive alternative to open and laparoscopic

approaches. UROLOGY 72: 1344–1346, 2008. © 2008 Elsevier Inc.

fe

ssdbasoti

iabtefi

P

Ahd

ectourinary fistulae (RUF) are uncommon. Al-though they may develop in patients with inflam-matory bowel disease and perirectal abscesses,

hey most frequently appear as an iatrogenic complica-ion of extirpative or ablative prostate procedures. Thencidence of rectal injuries after radical prostatectomy is%-11%.1,2 A review of complications after radical pros-atectomy in the Medicare population (n � 25 561)evealed a 1% incidence of RUF.3 With ablative treat-ents to the prostate, RUF incidence is as follows: 0.4%-

.8% after brachytherapy,1 0%-6% after external beamadiotherapy,4 and 0.4% after cryotherapy.5 Clinical find-ngs suggestive of RUF include pneumaturia, fecaluria,nd urine leakage per rectum.6 Imaging and endoscopictudies aid in delineating the fistulous tract. Cystoscopy,ystography, colonoscopy, and barium enema have beensed. Cystoscopy has a sensitivity of 80%-100%,7 butombined endoscopic and radiographic evaluation affordshe most precise anatomic depiction of the fistula. Distalbstruction (urinary or fecal) is a risk factor for repair

rom the Glickman Urological and Kidney Institute, Cleveland Clinic Foundation,leveland, OhioReprint requests: Rene Sotelo, MD, Centro de Cirugio Robotica y Invasión Minima,

7nstituto Medico La Floresta, Caracas, Venezuela. E-mail: renesotelo@mac.com

Submitted: April 21, 2008, accepted (with revisions): June 9, 2008

344 © 2008 Elsevier Inc.All Rights Reserved

ailure and must be determined to plan surgical repairffectively.

Regardless of the surgical approach, the best chance ofuccess is the first surgical attempt. The following clearurgical tenets must be satisfied: (1) fecal and urinaryiversion, (2) elimination of distal obstruction, (3) de-ridement to healthy tissue, and (4) tension-free tissuepproximation. When possible, and particularly in theetting of poor tissue healing potential (such as radiationr nutritional deficiency), interposition of vascularizedissue (omentum, peritoneum, muscle flaps, and so forth)s advisable.

Several open surgical approaches have been described,ncluding perineal, transanal, abdominal, and combinedbdominoperineal. Encouraged by our experience in ro-otic surgery, and having demonstrated feasibility withhe laparoscopic approach,8 we report our initial experi-nce with a robotic assisted approach to rectovesicalstula repair.

ATIENTS AND METHODS

57-year-old male with a clinically localized prostate cancerad undergone an open radical prostatectomy for high-riskisease (prostate-specific antigen of 20, Gleason score of 3�4 �

, pT2N0M0). A rectal injury occurred and was repaired pri-

0090-4295/08/$34.00doi:10.1016/j.urology.2008.06.017

mnposeseadmt(ttvtlr

PTdrflg

PTds

CWttf

PUpuptpabftwi

sfiwswarua1

barsdblppdef

Tcolc

CRtpsbo5oftsptf

mrisdhsb

dWtladrb

vsoic

U

arily in two layers. A loop colostomy was performed simulta-eously. Two months later, the colostomy was reversed. Onostoperative day 7, the patient presented with acute intestinalbstruction requiring laparotomy. At exploration, several adhe-ions were found. On the 8th postoperative day, the patientxperienced urine leakage per rectum. An end colostomy anduprapubic cystostomy were performed. A urethral Foley’s cath-ter was also placed. The fistula precluded radiation therapy forpositive surgical margin; therefore the patient received an-

rogen suppression with bicalutamide. Leakage persisted despiteaximal drainage, and 7 months later, the patient underwent a

ranssacral, sphincter-preserving, rectovesical fistula repairKraske). He remained dry for approximately 1 month, andhen again reported urinary leakage per rectum. The patient washen referred to us for further management. Cystoscopy re-ealed a fistulous tract localized at the trigone, in the region ofhe urethrovesical anastomosis. Based on prior experience withaparoscopic repair of rectourethral fistula repair, we planned aobotic approach.

atient Preparationhe patient received a full mechanical bowel preparation theay before surgery. Enemas were used to clear the 25-cm distalectosigmoid stump of mucous. Antibiotics covering entericora were given perioperatively. The patient was induced witheneral anesthetic.

atient Positioninghe patient was placed in a low lithotomy and steep Tren-elenburg position in stirrups. We applied sequential compres-ion stocking to the lower extremities.

ystoscopye performed cystourethroscopy and cannulated the fistula

ract with an open-ended ureteral catheter, pulled it throughhe fistula into the rectum, and retrieved it through the anus toacilitate intraoperative identification.

ort Placement and Techniquesing the 4-arm standard daVinci Surgical System, we em-

loyed a six-port transperitoneal approach with a port config-ration identical to that commonly used for da Vinci radicalrostatectomy. The configuration was appropriately shifted tohe right to avoid injuring the left colostomy. After establishingneumoperitoneum and placing trocars, we carefully performeddhesiolysis from prior surgeries. We created an omental flapased off the right gastroepiploic artery. This portion was per-ormed with standard laparoscopic instruments before dockinghe robot, because robotic access to the left hypochondriumith ports positioned toward the pelvis would have been lim-

ted.We created a vertical midline cystotomy with ultrasonic

hears and carried it distally to the posterior aspect of thestulous tract. Nonviable necrotic tissue was excised sharplyith scissors. The robot was docked into position and robotic

cissors were then used to dissect the fistula tract; the rectumas released from the bladder for a distance sufficient to achievetension-free rectal wall closure subsequently. We closed the

ectum with a 2-0 poliglecaprone suture on a UR-6 needle,sing an interrupted one-layer technique. The assistant grasped

previously laparoscopically created omental flap, using a

0-mm 30° EndoEYE laparoscope (Olympus Surgical, Orange- e

ROLOGY 72 (6), 2008

urg, NY) and a laparoscopic grasper, each placed throughssistant ports. This facilitated retrieval without undocking theobot, because the omentum recedes into the upper abdomen inteep Trendelenberg. We used the needle from the initial, mostistal suture placed for rectal closure to anchor the omental flapetween the bladder and rectum. The bladder was closed in oneayer in a running fashion using a 2-0 poliglecaprone suture. Welaced a new suprapubic tube. No urethral Foley’s catheter waslaced because of the trigonal location of the repair, to avoidirect pressure on the anastomosis. We performed an abdominalxit (assessing hemostasis, confirming drain location, and trocarascial closure) in standard fashion.

We removed the Blake drain on the third postoperative day.he suprapubic tube was removed at 2 months after normalystography. The patient was voiding normally and continentf urine. At 4 months, bowel continuity was restored withaparoscopic assistance. At 1 month since restoration of bowelontinuity, there has been no fistula recurrence.

OMMENTectourinary fistula is a rare but devastating complication

hat can develop after prostatic ablative or extirpativerocedures. Treatment of this entity is challenging. Con-ervative management consisting of urinary diversion,road-spectrum antibiotics, and parenteral nutrition isften initially attempted. Success rates as high as 25%-0% have been reported,9,10 but conservative measuresften fail. If the fistula remains open in 3-6 months,urther healing is unlikely.11-20 More than 40 surgicalechniques for the management of RUF have been de-cribed and there are no data clearly favoring one ap-roach.7 Transanal, transanorectal, transsphincteric,ransabdominal, perineal, and combined approaches arerequently used.

Basic surgical principles include excision and debride-ent of the fistula tract to healthy vascular tissue, sepa-

ation of the rectal and bladder suture lines with tissuenterposition, and effective urinary and/or fecal diver-ion.1 Some authors believe that fecal diversion is man-atory; others feel that it is necessary only when thereave been previous failed repairs, complex fistulae, pelvicepsis, or a history of radiotherapy, or in the absence ofowel preparation.2

The laparoscopic approach to RVF repair is modeledirectly after open principles, with notable nuances.hen RVF fistulae develop after radical prostatectomy,

hey usually occur along the vesicourethral anastomoticine. In that case, the technique involves an extravesicalpproach, opening of the bladder toward the fistula tract,issection between the rectum and bladder, closure of theectum, interposition of omentum, and closure of theladder.Our technique offers the advantages of minimally in-

asive surgery, with excellent visualization, adequatepace for instrument maneuvering, and the availability ofmentum, peritoneum, or epiploic appendages for tissuenterposition. If needed, a cystostomy, colostomy, or bothan be performed simultaneously. The bladder neck is

asily accessible with this approach, even with a “deep”

1345

pr

ieytahiialdfit

twsmmfibipt

CRsrwt

R

1

1

1

1

1

1

1

1

1

1

2

1

elvis. However, this technique requires advanced lapa-oscopic experience.

We have had a 100% success rate after minimallynvasive fistula repair, and no complications, but ourxperience is limited to 5 laparoscopic cases with only 1ear of follow-up. Follow-up for this robotic case is short-erm, at 1 month after restoration of bowel continuity,nd the patient is without recurrence. At this writing, weave performed one additional case of robotic rectoves-

cal fistula repair. This iatrogenic fistula recurred follow-ng open trans-perineal closure, and has remained closedt 1 month following robotic repair. More experience andonger follow-up is required before definitive recommen-ations can be made for the role of minimally invasivestula repair in the surgical armamentarium for this en-ity.

Certainly, our subjective impression is that features ofhe robot have simplified the robotic approach comparedith standard laparoscopy. These include three-dimen-

ional vision, six degrees of freedom with wrested move-ents, and downscaling of movements (for instance,odulation of the amplitude of surgical motions by up to

ve times). Mounting experience with deep pelvic ro-otic surgery, in addition, lay the groundwork for famil-arity with maneuvering in tight spaces. Robotics mightermit the popularization and careful diffusion of thisechnique.

ONCLUSIONSobotic rectourinary fistula repair is feasible, and repre-

ents an attractive alternative for the management ofectourinary fistulae. Longer follow-up and experienceith more patients are required before its ultimate role in

reating these difficult cases can be defined.

eferences1. Zinman L. Managing complex rectourethral fistulas. Contemp Urol.

2005;17:30-38.

346

2. Harpster LE, Rommel FM, Sieber PR, et al. The incidence andmanagement of rectal injury associated with radical prostatectomyin a community based urology practice. J Urol. 1995;154:1435-1438.

3. Benoit RM, Naslund MJ, Cohen JK. Complications after radicalretropubic prostatectomy in the Medicare population. Urology.2000;56:116-120.

4. Chrouser K, Leibovich BC, Sweat SD, et al. Urinary fistulas fol-lowing external radiation or permanent brachytherapy for thetreatment of prostate cancer. J Urol. 2005;173:1953-1957.

5. Badalament RA, Bahn DK, Kim H, et al. Patient-reported compli-cations after cryoablation therapy for prostate cancer. Urology.1999;54:295-300.

6. Trippitelli A, Barbagli G, Lenzi R, et al. Surgical treatment ofrectourethral fistulae. Eur Urol. 1985;11:388-391.

7. Shin PR, Foley E, Steers WD. Surgical management of rectourinaryfistulae. J Am Coll Surg. 2000;191:547-553.

8. Sotelo R, Garcia A, Yaime H, et al. Laparoscopic rectovesicalfistula repair. J Endourol. 2005;19:603-607.

9. Garofalo TE, Delaney CP, Jones SM, et al. Rectal advancementflap repair of rectourethral fistula: a 20-year experience. Dis ColonRectum. 2003;46:762-769.

0. Noldus J, Fernandez S, Huland H. Rectourinary fistula repair usingthe Latzko technique. J Urol. 1999;161:1518-1520.

1. Bukowski TP, Chakrabarty A, Powell IJ, et al. Acquired rectoure-thral fistula: methods of repair. J Urol. 1995;153:730-733.

2. Parks AG, Motson RW. Peranal repair of rectoprostatic fistula. Br JSurg. 1983;70:725-726.

3. Wilbert DM, Buess G, Bichler KH. Combined endoscopic closureof rectourethral fistula. J Urol. 1996;155:256-258.

4. Gecelter L. Transanorectal approach to the posterior urethra andbladder neck. J Urol. 1973;109:1011-1016.

5. Stephenson RA, Middleton RG. Repair of rectourinary fistulasusing a posterior sagittal transanal transrectal (modified York-Ma-son) approach: an update. J Urol. 1996;155:1989-1991.

6. Renschler TD, Middleton RG. 30 years of experience with York-Mason repair of recto-urinary fistulas. J Urol. 2003;170:1222-1225.

7. Ryan JA Jr, Beebe HG, Gibbons RP. Gracilis muscle flap for closureof rectourethral fistula. J Urol. 1979;122:124-125.

8. Zmora O, Potenti FM, Wexner SD, et al. Gracilis muscle transpo-sition for iatrogenic rectourethral fistula. Ann Surg. 2003;237:483-487.

9. Dafnis G, Wang YH, Borck L. Transsphincteric repair of rectoure-thral fistulas following laparoscopic radical prostatectomy. IntJ Urol. 2004;11:1047-1049.

0. Pieretti RV, Pieretti-Vanmarcke RV. Combined abdominal andposterior sagittal transrectal approach for the repair of rectourinary

fistula resulting from a shotgun wound. Urology. 1995;46:254-256.

UROLOGY 72 (6), 2008