Robotic Intracorporeal Orthotopic Neobladder during Radical Cystectomy in 132 Patients Mihir M. Desai,*,† Inderbir S. Gill,‡ Andre Luis de Castro Abreu, Abolfazl Hosseini, Tommy Nyberg, Christofer Adding, Oscar Laurin, Justin Collins, Gus Miranda, Alvin C. Goh, Monish Aron§ and Peter Wiklund From the USC Institute of Urology, University of Southern California, Los Angeles, California, and Karolinska Institute, Stockholm, Sweden
Abbreviations and Acronyms EBL ¼ estimated blood loss PLND ¼ pelvic lymph node dissection USC ¼ University of Southern California Accepted for publication July 1, 2014. Study received institutional review board approval. * Correspondence and requests for reprints: USC Institute of Urology, University of Southern California, 1441 Eastlake Ave., Suite 7416, Los Angeles, California 90089 (telephone: 323-8653749; FAX: 323-865-0120; e-mail: mihir.desai@ usc.edu). † Financial interest and/or other relationship with Hassen Medical and Baxter. ‡ Financial interest and/or other relationship with Hassen Medical and EDAP. § Financial interest and/or other relationship with Intuitive Surgical.
Purpose: We present a 2-institution experience with completely intracorporeal robotic orthotopic ileal neobladder after radical cystectomy in 132 patients. Materials and Methods: Established open surgical techniques were duplicated robotically with all neobladders suture constructed intracorporeally in a globular configuration. Nerve sparing was performed in 56% of males. Lymphadenectomy was extended (up to aortic bifurcation in 51, 44%) and superextended (up to the inferior mesenteric artery in 20, 17%). Ureteroileal anastomoses were Wallacetype (86, 65%) or Bricker-type (46, 35%). The learning curve at each institution was assessed using chronological subgroups and by trends across the entire cohort. Data were prospectively collected and retrospectively queried. Results: Mean operating time was 7.6 hours (range 4.4 to 13), blood loss was 430 cc (range 50 to 2,200) and hospital stay was 11 days (median 8, range 3 to 78). Clavien grade I, II, III, IV and V complications within 30 days were 7%, 25%, 13%, 2% and 0%, respectively, and between 30 and 90 days were 5%, 9%, 11%, 1% and 2%, respectively. Mean nodal yield was 29 (range 7 to 164) and the node positivity rate was 17%. Operative time, blood loss, hospital stay and prevalence of late complications improved with experience. During a mean followup of 2.1 years (range 0.1 to 9.8) cancer recurred in 20 patients (15%). Five-year overall, cancer specific and recurrence-free survival was 72%, 72% and 71%, respectively. Conclusions: We developed a refined technique of robotic intracorporeal orthotopic neobladder diversion, duplicating open principles. Operative efficiency and outcomes improved with experience. Going forward, we propose a prospective randomized comparison between open and robotic intracorporeal neobladder surgery. Key Words: robotics, cystectomy, urinary diversion, urinary bladder neoplasms
DURING the last decade there has been a steady increase in the use of robotic radical cystectomy for the treatment of invasive bladder cancer.1 Despite growing confidence in the technical and oncologic efficacy of the extirpative (cystectomy and lymphadenectomy) part of the robotic procedure,
most centers continue to perform the urinary diversion extracorporeally.2 Specifically, there has been a reluctance to perform orthotopic neobladders intracorporeally, with concerns centering around perceived technical difficulty, longer operative times and a higher complication rate.
0022-5347/14/1926-1734/0 THE JOURNAL OF UROLOGY® © 2014 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC.
http://dx.doi.org/10.1016/j.juro.2014.06.087 Vol. 192, 1734-1740, December 2014 Printed in U.S.A.
ROBOTIC INTRACORPOREAL ORTHOTOPIC NEOBLADDER DURING RADICAL CYSTECTOMY
Laparoscopic intracorporeal orthotopic neobladder was first described by Gill et al in 2002.3 Robotic intracorporeal orthotopic neobladder was first presented as a reproducible technique with operative efficiency and acceptable perioperative outcomes.4,5 We present our updated 2-center experience with robotic intracorporeal ileal neobladder in 132 patients with specific reference to perioperative data and complications.
PATIENTS AND METHODS At the Karolinska University Hospital and the University of Southern California a total of 136 intracorporeal neobladders were attempted between 2003 and 2013. Of these cases 4 were converted to extracorporeal diversion, resulting in 132 cases being successfully completed intracorporeally. Data were prospectively entered into each institution’s database after institutional review board approval and retrospectively analyzed. Mean patient age was 60.0 10.0 years (range 35 to 82), body mass index was 26.8 5.1 kg/m2 (range 18.0 to 60.0) and 32.6% were ASAÒ (American Society of Anesthesiologists) class 3 or 4 (table 1). Clinical stage was T3 or T4 in 14 patients (10.9%) and 26.1% had received neoadjuvant chemotherapy. The techniques of robotic radical cystectomy and intracorporeal neobladder have been described previously from both institutions.4,6 Both techniques have certain key steps in common. The most dependent and mobile point of the terminal ileum, at least 15 to 20 cm from the ileocecal junction, that seems to best reach the pelvic floor is identified as the future site for urethro-ileal anastomosis. All measurements to isolate the bowel segment for Table 1. Demographic and perioperative data Mean SD pt age No. male (%) Mean SD kg/m2 body mass index No. ASA (%): 1-2 3-4 Missing No. histology (%): Transitional cell Ca Others No. clinical stage (%): cT2 or less cT3-4 Missing No. neoadjuvant chemotherapy (%): Missing No. lymph node dissection (template)(%): PLND Limited/standard Extended (aorta bifurcation) Superextended (inferior mesenteric artery) Missing Mean SD ml estimated blood loss No. blood transfusion (%) Mean SD operative hrs Mean SD days hospital stay No. nerve sparing (male)(%): Missing
60.0 10.0 114 (86.4) 26.8 5.1 62 (67.4) 30 (32.6) 40 130 (98.5) 2 (1.5) 115 (89.1) 14 (10.9) 3 30 (26.1) 17 1 (0.9) 44 (37.9) 51 (44.0) 20 (17.2) 16 430 374 6 (4.5) 7.6 1.7 10.6 9.1 64 (74.4) 28
creating the neobladder are made from that reference point. An approximately 59 cm length of ileum is isolated, 44 cm for the neobladder pouch and 15 cm for the afferent limb. The entire neobladder pouch is constructed with robotic intracorporeal suturing without the use of any staplers. Finally, a side-to-side stapled ileoileal anastomosis is created to restore bowel continuity. There are also certain differences in techniques between the institutions. The Karolinska technique involves performing the urethro-ileal anastomosis at the outset, once its site has been identified, and all other steps including isolation of the segment, bowel anastomosis and creation of the pouch are performed subsequently. In the USC technique the future site of urethro-ileal anastomosis is marked at the outset. The actual urethro-ileal anastomosis is performed only after the posterior plate has been suture constructed and rotated 90 degrees. Additionally, the Karolinska technique uses a Wallacetype ureteral anastomosis, whereas with the USC technique the ureters are spatulated and individually implanted into the afferent limb. A urethral catheter is left without use of a suprapubic tube. A tube drain and Double-JÒ stents are also positioned. The drain is typically removed at 1 to 2 weeks once the drainage is less than 500 cc per day. The urethral catheter is removed at 3 weeks and ureteral stents are removed at 4 to 6 weeks. Demographic, perioperative, pathological and functional data were analyzed. Continuous variables were reported as mean SD (range). Survival and recurrence were assessed with the Kaplan-Meier estimator, using deaths from other causes as censoring time points when evaluating cancer specific and recurrence-free survival. To analyze learning curve trends, patients at each institution were chronologically ordered according to dates of operation and the Jonckheere-Terpstra test was used to assess trends in perioperative outcomes with increasing institutional experience. For visualization these data are shown in scatterplots together with the corresponding linear regression curves and LOESS smoothed curves. We also present the outcomes of patients from each institutional cohort divided into chronological groups of 15. All tests were performed 2-sided at the 5% significance level.
RESULTS Perioperative data are detailed in table 1. All 4 conversions occurred in the early learning curve, and were due to failure to bring the ileum to the urethra (3) and failure to progress (1). Mean operative time was 7.6 1.7 hours, EBL was 430 374 cc and hospital stay was 10.6 9.1 days. Mean and median hospital stay at institute 1 was 10.3 and 8 days, and at institute 2 was 11 and 7 days, respectively. Nerve sparing was performed in 64 males (74%). PLND was done in 1, limited or standard in 44 (38%), extended in 51 (44%) and superextended in 20 (17%) cases. Complications are listed in table 2. Overall, 30-day complications were observed in 62 (47%) patients and 30 to 90-day complications in 36 (27%).
ROBOTIC INTRACORPOREAL ORTHOTOPIC NEOBLADDER DURING RADICAL CYSTECTOMY
Table 2. Summary of early (30-day) and late (30 to 90-day) complications and categories 30-Day No. gastrointestinal (7.6%): Bowel obstruction Bowel leak Constipation Ileus Neobladder-bowel fistula Totals No. infectious (28.8%): Sepsis Urinary tract infection Intra-abdominal abscess Epididymitis Totals No. hematologic (7.6%): Anemia Anemia requiring transfusion Hematoma/bleeding Totals No. genitourinary (21.2%): Acute urinary retention Urine leak Ureteroenteric stricture Hydronephrosis Reservoir stones Bladder neck contracture Neobladder-vaginal fistula Lymphocele Totals No. other (9%): Acute renal failure Dehydration Fever Compartment syndrome Nerve palsy Arrhythmia Deep vein thrombosis Totals No. highest Clavien grade of complication (%):* 0 I II IIIa IIIb IVa IVb V
1 1 1 4 1 8 13 6 1
30 to 90-Day
e e e
1 2 1 4 2 10
19 17 1 1 38
e e e e
4 3 3 10
1 1 2 6 11 e
4 3 3 10 1 6
1 e e e e e
5 3 3 1 1 2 16 e e e e e
1 2 1 1 1 2 2 10 70 9 33 12 5 3 0 0
1 1 2 (53) (6.8) (25) (9.1) (3.8) (2.3)
96 7 12 3 11 1 0 2
2 6 5 3 3 1 1 8 29 1 2 1 1 1 3 3 12
(72.7) (5.3) (9.1) (2.3) (8.3) (0.8) (1.5)
* Clavien grade was assigned to each case based on the highest grade recorded for the period.
Rates of Clavien grade I, II, IIIa, IIIb, IVa, IVb and V complications within 30 days were 6.8%, 25%, 9.1%, 3.8%, 2.3%, 0% and 0%, and between 30 and 90 days were 5.3%, 9.1%, 2.3%, 8.3%, 0.8%, 0% and 1.5%, respectively. The highest grade complications included infective in 38 (28.8%), gastrointestinal in 10 (7.6%), genitourinary in 28 (21.2%), bleeding and hematologic in 10 (7.6%), and miscellaneous in 12 (9%). Specifically there were 2 bowel anastomotic leaks (2.3%), 5 benign ureteroenteric anastomotic strictures (3.8%) and 8 lymphoceles (9.2%). Ureteroenteric strictures were noticed in 3 of 86 patients undergoing Wallace reimplantation vs 2 of 46 patients undergoing Bricker reimplantation.
Pathological data are presented in table 3. Of the cases 28.7% were pT0, 6.2% pTis, 5.4% pTa, 9.3% pT1, 34.9% pT2, 12.4% pT3 and 3.1% pT4. There was 1 (0.8%) positive margin in the entire cohort. Mean nodal yield was 29 (7 to 164) and the node positivity rate was confirmed in 22 (17%). During a mean followup of 2.1 years (0.1 to 9.8) cancer recurred in 20 patients (15%). Five-year overall, cancer specific and recurrence-free survival was 72%, 72% and 71%, respectively (table 4, fig. 1). Continence data were available for 73 patients who had completed at least 6 months followup. Continence was achieved in 62 patients (0 to 1 pad per day), whereas 11 required 2 or more pads for persistent incontinence. The impact of learning curve on perioperative outcomes at each institution is represented in figure 2 and table 4. Trends based on chronology at institute 1 showed a decrease in median operative time (p