Fax: (11) 43 (512) 504 8411
Urethra-sparing cystectomy and orthotopic urinary diversion in women with malignant pelvic tumors
Article first published online: 9 OCT 2001
Copyright © 2001 American Cancer Society
Volume 92, Issue 7, pages 1864–1871, 1 October 2001
How to Cite
Stenzl, A., Jarolim, L., Coloby, P., Golia, S., Bartsch, G., Babjuk, M., Kakizoe, T. and Robertson, C. (2001), Urethra-sparing cystectomy and orthotopic urinary diversion in women with malignant pelvic tumors. Cancer, 92: 1864–1871. doi: 10.1002/1097-0142(20011001)92:7<1864::AID-CNCR1703>3.0.CO;2-L
- Issue published online: 9 OCT 2001
- Article first published online: 9 OCT 2001
- Manuscript Accepted: 9 JUL 2001
- Manuscript Revised: 12 MAR 2001
- Manuscript Received: 14 SEP 2000
- bladder neoplasm;
- pelvic tumor;
- urinary diversion;
- transitional cell cancer;
- pelvic recurrence
To the authors' knowledge, few data exist regarding the functional and oncologic outcome of pelvic tumors in women with urethra-sparing cystectomy and orthotopic urinary diversion to the urethra.
PATIENTS AND METHODS
The combined data of 102 women age 28–79 (mean, 59 yrs) years who underwent a urethra-sparing cystectomy and orthotopic urinary diversion for either primary bladder cancer (96 patients), carcinoma of the uterine cervix (2 patients), carcinoma of the vagina (1 patient), primary fallopian tube carcinoma (1 patient), uterine sarcoma (1 patient), or rectal carcinoma (1 patient) were reviewed. The histology of the 96 primary bladder tumors was 81 transitional cell carcinomas (TCC), 8 adenocarcinomas, 5 squamous cell carcinomas, 1 small cell carcinoma, and 1 unclassified. Follow-up ranged from 1.5–100 months (mean, 26 mos; median, 24 mos). In all patients, the bladder neck and up to 1 cm in length of the adjacent urethra were removed with the bladder. An ileal orthotopic neobladder procedure was performed if staging biopsies of the bladder neck and intraoperative frozen section of the urethral margin revealed no tumor.
There was no perioperative mortality, and an early and late complication rate requiring secondary intervention in 5 (5%) and 12 (12%) patients. With 88 of 102 patients alive and 83 of 102 patients disease free, a disease-specific survival of 74% and a disease-free survival of 63% was estimated at 5 years. No pelvic recurrence was seen in 81 patients with TCC. Three pelvic recurrences occurred, two tumors of the inner genitalia and one adenocarcinoma of the bladder, none of them in the area of the urethra or its supplying autonomic nerves. Daytime continence was 82%; nocturnal continence was 72%. Twelve (12%) patients were unable to empty their bladders completely and needed some form of catheterization.
The functional and oncologic outcome of female patients with an orthotopic urinary diversion to a remnant urethra was found to be comparable to that found in large studies on males. An orthotopic neobladder proved to be an oncologically safe option for women with pelvic tumors and was found to provide quality of life when there was adherence to previously defined selection criteria. Cancer 2001;92:1864–71. © 2001 American Cancer Society.
Orthotopic reconstruction of the lower urinary tract in the male is an almost standard operation now with a high acceptance rate both by patients and doctors and an acceptably low local recurrence rate.1–3 Until recently, however, only small series of urethra-sparing cystectomies with subsequent orthotopic bladder reconstruction were performed in women, and they had a relatively short follow-up.4–11 Whereas, in the majority of cases surgical removal of the bladder is performed for transitional cell cancer (TCC), a larger percentage of operations performed in most of these series is for indications other than TCC, indicating that preservation of the urethra is considered dangerous with TCC of the bladder. An undefined risk of the extent of TCC spreading into the female urethra was probably the reason for almost unequivocally performing total urethrectomy in combination with anterior pelvic exenteration until then. Lack of data about secondary urethral tumors and insufficient knowledge about the functional anatomy of the isolated female urethra and its sphincter were the biggest obstacles in the development and common use of an orthotopic reconstruction of the lower urinary tract after cystectomy for bladder cancer. Additionally, the use of intestinal pouches with ileal segments as urethral substitutes with or without an artificial urinary sphincter had not gained wide acceptance.12
Several authors, therefore, looked specifically at the risk of urethral tumors in the remnant female urethra after cystectomy for TCC and step-sectioned urethrocystectomy specimens of female bladder cancer patients.13, 14 They found urethral tumor involvement in 6.4–10.7% of the specimens. A strong correlation was seen with bladder cancer at the bladder neck and/or the trigone, and subtotal urethrectomy was, therefore, thought to be feasible. Another report addressed secondary urethral tumors in all patients treated for bladder cancer at a single institution over several decades and found an overall incidence of urethral tumor involvement in 2% and in 1% of patients with localized invasive TCC, who nowadays would be candidates for radical cystectomy.15 The only consistent risk factor for secondary urethral tumors in this study, again, was simultaneous primary tumor involvement of the bladder neck.
Bearing in mind that the risk of secondary tumors in females was even lower than in studies looking at male bladder cancer patients,16 it could be concluded that from an oncologic standpoint urethra-sparing cystectomy with subsequent orthotopic urinary diversion is also an option in selected women with bladder cancer.
PATIENTS AND METHODS
From November 1989 until March 1999, 102 with malignant tumor of the pelvis underwent radical cystectomy with subsequent orthotopic urinary diversion to a remnant portion of the urethra at 3 European centers. The underlying disease was bladder cancer in 96 patients, recurrent cervical carcinoma in 2, squamous cell carcinoma of the vagina in 1, carcinoma of the uterine tube in 1, recurrent uterine sarcoma in 1, and infiltrating adenocarcinoma of the rectum in 1. Four of the five patients with gynecologic tumors had preoperative definitive radiotherapy. The histology of 96 primary bladder tumors was 81 TCCs, 8 adenocarcinomas, 5 squamous cell carcinomas, 1 small cell carcinoma, and 1 not classified tumor. Tumor staging of the primary bladder tumors followed the terminology of the TNM 1992 classification of the American Joint Committee on Cancer/International Union Against Cancer (AJCC/UICC) and is listed in Table 1. Patient age at the time of cystectomy ranged from 28–79 years (mean, 59 yrs; median, 63 yrs).
|Nodes||Nodes||Nodes||Tumor grade||Tumor grade||Tumor grade||Tumor grade|
|pT stage||No. of cases||N-0||N-1||N-2||CIS||G1||G2||G3|
|Total no. of cases||96||86||6||4||2||2||13||78|
Candidates for orthotopic neobladder to the urethra were selected according to tumor extension, urethral competence, general performance status, and motivation. We unconditionally excluded patients with tumor at the bladder neck and/or in the urethra, any positive surgical margin on intraoperative frozen section, patients with macroscopically enlarged positive lymph nodes and patients with distant metastases. Apart from the usual preoperative staging with bimanual and transurethral tumor evaluation and biopsy, abdominal and thoracic imaging, biopsies of the bladder neck were performed routinely. Any macroscopically enlarged suspicious lymph nodes were checked either intraoperatively by frozen section or preoperatively by biopsy with the patient under local anes- thesia.
Urethral competence was assessed by patient history, endoscopy, radiography, and, in the majority of patients, by intraluminal pressure profile (UPP). Patients with either a history of Grade II stress incontinence or higher due to an incompetent sphincter, marked urethral hypermobility, or a maximal resting pressure in the UPP of less than 30 cm H2O generally were excluded.
A preoperative general performance status of a minimum of 80% on the Karnofsky scale was thought essential because these patients need their strength for, among other things, continence training. Motivation included the understanding and handling of specific problems of orthotopic neobladders in females, such as possible urinary retention or postvoid residuals, initial nocturnal and diurnal incontinence, timed voiding, and nycturia.
Preoperative radiation therapy to the pelvis was not a contraindication for urethra-sparing surgery and orthotopic bladder substitution, unless urethral competence as outlined above became insufficient and oncologic outcome was compromised.
Anterior exenteration was performed according to the technique previously published.17, 18 Pelvic lymphadenectomy, removal of the inner genitalia, excision of the anterior vaginal wall, and removal of the bladder neck and of a small urethral segment of up to 0.5 cm in length were thus performed in all patients. A clam vaginoplasty was sufficient for vaginal reconstruction in all but one patient, where a pedicled ileal patch was used to augment the vagina. All three centers where these surgeries were performed attempted to preserve autonomic nerves to the remnant urethra. In case of a risk of tumor infiltration or extensive scarring due to previous surgery, either unilateral or no nerve preservation was done. A section of the urethra was sent for intraoperative evaluation by pathology.
An orthotopic diversion was offered as one of the choices of urinary diversion, provided informed consent had been obtained preoperatively, and if there were no contraindications found intraoperatively as outlined in previous publications.5, 13, 19 The pouch consisted of a low-pressure ileal reservoir whose most dependent portion was anastomosed to the remnant urethra with 6 absorbable sutures over an 18 or 20 mm balloon catheter. The ureteral catheters were brought out through individual openings in the pouch and either separately through the lower abdominal wall or through the wound itself. At the end of the procedure, a J-omentum flap was brought down and led around the bottom part of the pouch.20 Alternatively, portions of the ileal pouch adjacent to this anastomosis were sutured to the anterior and lateral pelvic walls as well as to the remnant vaginal sac both to avoid the formation of obstructive folds over the urethral anastomosis and to reduce a “pouchocele”21 because of descent of the reservoir postoperatively.
During the first 2 years postoperatively, the patients usually were followed at 6 month intervals with clinical examination, routine chemistry studies including renal function parameters and serum bicarbonate levels, computerized tomography of the pelvis and abdomen, plain chest X-ray, and cytology. All patients were questioned regarding quality of life, continence, and voiding pattern according to a standard protocol. Patients were considered continent if they used maximally 1 safety pad for the occasional loss of a few drops of urine. “Normal” micturition anticipated no straining and a residual urine of less than 150 cc. Obstructive voiding specified a condition of intermittent straining during micturition and/or a residual volume of more than 150 cc and/or any signs of upper urinary tract deterioration. Sixty-five patients had a pre- and postoperative urodynamic evaluation. Endoscopy of the pouch and urethra was performed only when there were suspicious findings during any of the aforementioned routine examinations.
Progression-free survival and overall-survival rates were obtained by applying the Kaplan–Meier method. The correlation between couples of variables of interest was evaluated statistically using the chi-squared test and the G-statistic. Exact tests and confidence intervals were calculated when the subgroups were small.22 The analysis was performed using S-PLUS 2000 software for Windows NT (Microsoft, Redmond, WA).
There was no perioperative mortality among these patients. Early complications requiring secondary surgery in 5 (5%) patients included 2 ileus, 1 pouch wall necrosis, 1 wound dehiscence, and 1 self-knotted ureteral catheter. A pouch wall necrosis involving approximately two-thirds of the ileum occurred in one patient due to unknown reasons. Upon the patient's wish, the remnant pouch segment was changed to an ileum conduit-type diversion. The knotted ureteral catheter was removed percutaneously without consequence for the ureterointestinal anastomosis.
An intervention due to 14 pouch-related late complications was necessary in 12 patients and included transurethral incision and resection of an ileal fold overlying the neobladder outlet in 5 patients, transvaginal closure of a pouch-to-vagina fistula in 2, lithotripsy for pouch lithiasis in 1, and nephrolithiasis in 1 patient, dilatation of a unilateral stenosis of the ureteroileal junction in 1, and submucosal bladder neck injection in 3, and bladder neck suspension in 1 patient for the treatment of postoperative stress incontinence. The diagnosis of obstructing ileal valves and their treatment has been reported previously.21 Both fistulas, which occurred in patients after a preceding transurethral incision of ileal valves that were obstructing the neobladder outlet and a subsequent early catheter removal, were closed transvaginally and healed without problems. The bladder neck suspension was performed in an obese woman weighing 110 kg who had an anterior vaginal prolapse. The prolapse was corrected, but the patient went into urinary retention after the bladder neck suspension. She refused further interventions and opted for permanent catheterization.
Survival and Recurrence
Follow-up in these women ranged from 1.5–100 months (mean, 26 mos). At completion of the study, 88 of 102 patients were alive, and 83 of 102 patients were alive and disease free. Disease-specific 5 year-survival was estimated as 74% (Fig. 1). Disease-free survival at 5 years was 63% (Fig. 2). The 95 % confidence interval for the disease-specific 5 year-survival (0.745) was estimated as 0.613–0.904. The 95% confidence interval for the disease-free survival at 5 years (0.627) was estimated as 0.461–0.852.
All patients were free of tumor at the bladder neck, whereas the tumor distribution was relatively uniform for all other bladder regions. None of the 81 patients with TCC of the bladder developed either urethral or pelvic recurrence. Pelvic tumor recurrence was seen in one patient with an adenocarcinoma of the bladder, in one patient with uterine sarcoma, and in one patient with recurrent cervical cancer. In the latter patient, the neobladder was converted to an open abdominal diversion because the tumor infiltrated both rectum and neobladder.
Autonomic Nerve Preservation
In 66 patients, the bilateral preservation of autonomic nerves to the urethra was attempted. Because of the location of the tumor or scarring from previous surgery, the majority of autonomic nerves had to be dissected on both sides in 7 patients and on one side in 28 patients. Nerve preservation was correlated with the voiding pattern and the need for periodic catheterization (Table 2).
|Nerve preservation||Not performed||%||Unilateral||%||Bilateral||%|
For the 7 women without nerve preservation, the exact 95% confidence interval for the percentage with clean intermittent catherization (CIC) was 29.0–96.3%, for the 28 women with unilateral preservation it was 0.0–12.3%, and for the 66 women with preservation on both sides it was (3.4–18.7)%. Although the confidence interval for the 7 women was very wide, the percentage of women with CIC was clearly different from those who had some form of nerve preservation because the confidence intervals do not overlap.
Terminal ileum was used for all pouches. The spheroid reservoir was created either according to the Hemi-Kock2 (12 patients), Hautmann23 (54 patoemts), T-Pouch24 (35 patients), or VIP (1 patient) technique depending on the center where surgery was performed. Pouch capacity was measured in 93 patients at their last follow-up and ranged from 150–800 cc with a mean volume of 408 cc.
Eighteen patients with less than 6 months follow-up and 1 patient on permanent catheterization were excluded. Of 83 patients where continence status was thus evaluable, 68 (82%) patients reported daytime continence with maximally 1 safety pad. In the same group, nighttime continence with 1 safety pad only was reported by 60 (72%) patients.
Daytime micturition intervals ranged from 1.5–5 hours (mean and median, 3 hrs). Nocturnal micturition intervals were 2–6 hours (mean and median, 3 hrs). Ninety patients voided spontaneously without any regular catheterization, whereas 11 (11%) patients needed some form of periodic catheterization. One patient with a failed bladder neck suspension and subsequent permanent catheterization was excluded. Of the 90 patients who voided spontaneously, 76 reported voiding “normally” without straining. Residual urine in these patients was less than 150 cc. Fourteen patients reported intermittent difficulties with voiding that could be overcome by straining. Repeated measurements of residual urine ranged from 0–200 cc. Those patients satisfied with their conditions have been counseled and followed carefully. None of these patients has reflux to the upper urinary tract or any sign of renal deterioration.
An increasing number of radical cystectomies in the last decades have led to a growing desire for appliance-free urinary diversions to improve quality of life and body image. There are reports demonstrating that as many as 96% and 79% of patients with orthotopic neobladders resume their daily activities and occupational status and that 85% are totally continent day and night, whereas the remainder seem to manage partial incontinence well with pads.25, 26 Among appliance-free diversions performed by these authors, continence was better for neobladder patients compared with those who had cutaneous continent diversions (85% vs. 61%). In another study, both male and female patients with orthotopic neobladder were compared with those who had a wet stoma.27 Patients with orthotopic neobladder had significantly shorter postoperative rehabilitation, were socially reintegrated to a higher extent with less restriction of their leisure activities, and experienced an overall higher rate of self-confidence. There also appears to be evidence that the orthotopic reservoir is slightly advantageous to a heterotopic diversion when assessing quality of life with the Sickness Impact Profile.26
Whereas an orthotopic urinary diversion has been available to male patients for several decades, its habitual use in female patients started just a few years ago.28 One of the concerns regarding leaving the urethra in female patients with bladder cancer was the probability of leaving either tumor or a locus with increased risk for tumor recurrence behind. This problem has been approached by several recent reports. Coloby et al.13 and Stein et al.14 in retrospective analyses step-sectioned urethrocystectomy specimens of female bladder cancer patients and found respective urethral tumor involvement in 7 of 65 (10.7%) and 3 of 47 (6.4%) patients. In both studies, a strong correlation was seen with bladder cancer at the bladder neck and/or trigone, and a subtotal urethrectomy was favorably discussed. Conversely, DePaepe et al.29 in a similar study found carcinoma in situ or overt papillary tumor in the urethra in 36 % (8 of 22) of their patients and concluded that the urethra should be removed in all women undergoing radical cystectomy for TCC of the bladder. Unfortunately in this study, which included only 22 cases treated over a period of 15 years, no details regarding localization of either primary tumors in the bladder or secondary tumors in the urethra are provided. Stenzl et al.15 and Ashworth,30 looking at the incidence of secondary urethral tumors of all patients treated conservatively for bladder cancer at a single institution and followed for up to 33 years reported respective urethral tumors in 2% of 356 and 1.4% of 293 female patients. In the former study, the only consistent risk factor for secondary urethral tumors was simultaneous primary tumor involvement of the bladder neck.
From these studies, it can be concluded that the incidence of urethral cancer in female patients with bladder cancer is lower than in male patients. Orthotopic reconstruction of the lower urinary tract in the male is almost a standard operation now with a high acceptance rate both by patients and doctors and an acceptably low local recurrence rate.1, 31, 32 Superficial tumor involvement of the prostatic urethra is no longer seen as a contraindication for orthotopic urinary diversion in male patients by some authors provided intraoperative frozen section of the urethral margin shows no tumor.33 A strong argument can, therefore, be made that in female patients without bladder neck involvement of the primary bladder tumor and negative intraoperative frozen section of the urethral margin, a large portion of the urethra can be spared for an orthotopic neobladder. We do not, however, argue that the urethra can be left in all cases of cystectomy, i.e., also in patients who receive abdominal or rectosigmoidal diversion. A blind-ending urethra may have an increased incidence of tumor recurrence that can be missed.
A prospective pathologic analysis of female cystectomy specimens has shown that more than 60% of patients with bladder neck tumors had a normal tumor-free proximal urethra both on intraoperative frozen section and final permanent histology.34 The presence of tumor at the bladder neck as an exclusion criterion, therefore, may have to be revised in the presence of negative frozen-section at the urethral margin. A recent case report,35 however, demonstrated that infiltrating tumors near the outlet may bear a larger risk of local recurrence when performing urethra-sparing surgery.
The numbers regarding lymph node involvement are too small in this series to draw any conclusions. To date, based on data from the literature for male patients,36, 37 any microscopic and/or macroscopically resectable lymph node involvement is not a contraindication for performing orthotopic bladder substitutes in women for the participants of this study. None of the patients in this series had urethral recurrence. Furthermore, none of the 81 patients with TCC had pelvic recurrence. We believe that good results can be obtained by leaving a safety margin in removing the bladder neck, a urethral segment, and resecting the anterior vaginal wall. Despite the benefits of nerve preservation, however, one should refrain from that in cases where the primary tumor is thought to be close. The pelvic tumor recurrences observed in three patients were not related to sparing either the urethra or the autonomic nerves.
To our knowledge, the current series is the largest one reporting the outcome of urethra-sparing cystectomy in female patients to date. In contrast to previous studies,7, 9 cystectomy included the bladder neck and a small portion of the adjacent urethra in all patients. From an oncologic standpoint, this increases the safety margin of the primary bladder cancer. But this also may explain the better results regarding spontaneous voiding. A low-pressure ileal reservoir, which is the most common form of orthotopic urinary diversion today, may not overcome the resistance of the entire urethra and the bladder neck. Furthermore, the neobladder cannot assist urethral shortening during voiding, which is achieved with the normal bladder by dragging the vesical neck segment into the bladder.
The authors believe that to retain the function of the remaining urethral smooth musculature, the autonomic nerves to the urethra should be preserved.5, 38, 39 It has been demonstrated in animal studies that bilateral autonomic denervation of the isolated female urethra results in degeneration and defunctionalization of urethral smooth muscle cells and a decrease in urethral intraluminal pressure.40 The correlation of absent, uni- or bilateral autonomic nerve preservation with unobstructed, obstructed, and impaired voiding in this series (Table 2) demonstrated that 72% of the women where no nerve preservation was done needed to catheterize themselves compared with 9% who had bilateral nerve preservation. Sparing the autonomic nerves to the urethra, therefore, must be considered one of the reasons for the better results regarding spontaneous voiding compared with previous reports.7 We also think that autonomic nerve preservation enables removal of the bladder neck and a small portion of the urethra without negative long-term consequences for the continence rate but with better results for spontaneous micturition.
Preservation of the autonomic nerves may not always be possible on both sides because of the extension of the tumor or previous surgery. In 29% of the patients in this series, a unilateral nerve preservation was attempted for these reasons. There was no increase in urinary retention or residual urine requiring catheterization in this group. Furthermore, no case of pelvic tumor recurrence has been observed in the region of the preserved autonomic nerves.
A diurnal continence rate of 82% and a nocturnal continence rate of 72% in patients followed for 6 months or longer are comparable to reports in the literature for both male and female patients.1, 2, 6, 7, 11 These data may become better in the future because continence rates improve up to 2 years postoperatively. They also may become better because of the surgeons learning curve. In this study, it was not just the learning curve of one surgeon but that of surgeons in three different institutions.
The fact that these patients have been treated in different centers accounts for some differences in the surgical techniques creating the urinary reservoirs, mainly in configuring the ileal pouch. However, from an oncologic standpoint, there was no difference in the selection criteria, extent of surgical ablation, or handling of the remnant urethra.
- 20Continent reservoir to the urethra in the female [videotape]. J Urol 1994; 151: 204A, V-19., , .
- 22StatXact, edition 4.01. Cambridge, MA: Cytel Corporation, 1999.
- 23The ileal neobladder. J Urol 1991; 139: 39–42., , , .
- 27Life after cystectomy and orthotopic neobladder versus ileum conduit urinary diversion. Semin Urol Oncol 2000; 19(1): 18–23., , , , , , et al.