The histology of the tumors in the current series was predominantly urothelial carcinoma (80%), as shown in . One patient underwent radical cystectomy for primary cervical squamous cell carcinoma. The pathologic T-stage of the cohort is also shown in . Twenty-seven percent of the women had pathologic extravesical extension of tumor. Of the 12 women with pathologic T0, Ta, and T1 disease, 6 had clinical T2 disease at the time of transurethral resection, 1 had clinical Tis disease, and 5 had recurrent T1 tumors refractory to intravesical agents.
The mean number of lymph nodes removed was 19 (range 3-45), and 11 (20%) subjects were found to have lymph node–positive disease. In those women with positive lymph nodes, an average of 20% of the nodes contained cancer. The proportion of subjects with positive lymph nodes varied with pathologic tumor stage (). Despite negative intraoperative frozen section margins, the final pathologic urethral margin was positive for cancer in four (7%) cases, and at least one of the ureteral margins was positive in four (7%) cases.
In total, 18 (32%) patients developed disease recurrence—5 (9%) local, 15 (27%) distant, and 2 (4%) with both local and distant recurrence. In the 2 women who developed both local and distant recurrence, 1 patient recurred in the vagina and liver, and the other patient recurred in the pelvic sidewall and lung. Mean time to recurrence was 28 months (range 3-76). Of the 4 patients with positive urethral surgical margins, there were no local recurrences and 1 distant recurrence (spine). Of the 4 patients with positive ureteral margins, 2 patients recurred—both in the renal pelvis (at 24 and 58 months).
Survival was estimated by Kaplan-Meier methods, as seen in . Recurrence-free, cancer-specific, and overall 5-year survival was 54%, 57%, and 47%, respectively. For patients with lymph node–negative disease at cystectomy, cancer-specific and overall 5-year survival was 75% and 60%, respectively. Univariate regression analysis () found lymph node status as the only clinical indicator significantly correlated with cancer-specific survival (HR 14.2, 95% CI 4.6–43.8, P <.001). After adjusting for lymph node status, age at diagnosis was a significant predictor of bladder cancer–specific survival (HR 1.7, 95% CI 1.0-2.7, P = .04). This suggests that lymph node status is a confounder of the association between age and bladder cancer–specific mortality.
Kaplan-Meier estimated recurrence-free and overall survival are shown in (A) and (B), respectively. Estimated cancer-specific survival stratified by lymph node status is shown in (C).
Analysis of predictors of bladder cancer–specific mortality
At last follow-up, 32 (57%) women remained completely continent. Of the 24 subjects who continued to use pads, 18 provided information about the number of pads used (), and most women (66%) required only 1 or 2 per day (). Two (3%) women developed vaginal fistulae as the etiology for their incontinence, and three (5%) women who were initially incontinent ultimately gained complete continence (at 3, 48, and 72 months). Hypercontinence occurred in 13 (24%) women. Two initially voided but developed the need for catheterization after 6 months, and only one who had initial hypercontinence improved to normal voiding. No clinical or pathologic factors were significantly associated with continence or hypercontinence.
Functional outcomes in women undergoing cystectomy and orthotopic neobladder construction
Separate analyses of oncological and functional outcomes were completed using only patients with greater than 9 months follow-up, and results were not significantly different from those presented here before (data not shown).
ONB after radical cystectomy has been established as an accepted form of urinary diversion for men. In women, ONB has become increasingly common, with the accumulation of evidence showing its oncological safety and functional efficacy.6-8
Here, we report on a cohort of 56 women who underwent radical cystectomy with ONB for malignant disease.
Oncological characteristics and outcomes in this study are similar to the limited number of similar series previously reported. Eighty percent of subjects had urothelial carcinoma, which is the most common histology in the United States.6,8
A study from Egypt had primarily squamous cell carcinoma, likely reflecting the prevalence of Bilharzial disease in that region.7
In our series, 27% of women had pathologic stage ≥T3 disease and 20% had lymph node–positive disease. The recurrence rate of 32% is consistent with other studies, which range from 28-36%.6-8
Lymph node status at cystectomy and age at diagnosis were the only clinical or pathologic variables that predicted recurrence or survival. In this series, 5-year estimated recurrence-free, cancer-specific, and overall survival of 54%, 57%, and 47% are similar to that reported by others.6-8,13
There has been significant debate about the risk of urethral recurrence after ONB in women, given the much shorter urethra compared with men. Early investigations reported urethral involvement of bladder cancer in women to be as high as 36% and recommended urethrectomy with cystectomy as the standard of care.14
More recently, retrospective studies reported lower rates of urethral involvement and that precystectomy evidence of tumor at the bladder neck could distinguish patients at risk of urethral recurrence.15-17
In a recent review of the literature, Stein et al found that approximately 50% of women with bladder neck tumors had urethral disease, making this a risk factor but not an absolute contraindication for ONB.18
More importantly, intraoperative frozen section has been shown to provide an accurate assessment of extension into the urethra.19
At UCSF, women with bladder neck tumors are not excluded from ONB if intraoperative frozen section of the urethral margin is negative. In this series, 7% of patients had a positive permanent urethral surgical margin, despite uniformly negative frozen sections. However, none of these patients had local recurrence. In fact, no patient had urethral recurrence. These findings suggest that women with negative intraoperative urethral frozen sections can safely undergo ONB after radical cystectomy without undue risk of urethral recurrence, including those with bladder neck tumors.
The primary advantage of ONB after radical cystectomy, compared with ileal conduit or other continent cutaneous diversions, is the potential for near-normal voiding function and retained body image by voiding from the native urethra.20
However, functional outcomes after ONB in women are not as well-defined as in men. We found a similar rate of incontinence (43%) compared with previous reports, and among our patients with urinary leakage, 66% used only 1 or 2 pads per day. Our rate of incontinence was meant to be inclusive, using the intentionally broad definition of the use of any pads daily. Granberg et al, using the same definition, reported day- and nighttime incontinence rates of 10% and 43%, respectively.6
Stein et al reported questionnaire-driven rates of 23% and 34% for “frequent leakage or no urinary control whatsoever” during daytime and nighttime, respectively.8
Our hypercontinence rate of 24% is lower than reported figures ranging from 35-61%,6,8
but we continue to counsel patients preoperatively of the potential need for self catheterization. In general, approximately 35 cm of ileum (Studer and Zingg12
) and 60 cm (Hautmann et al10
) neobladders were harvested in this series, which are both smaller lengths of bowel than originally described. We believe that a smaller capacity reservoir may be responsible, in part, for this lower retention rate. However, this may come with an initially higher risk of incontinence in the first few months after surgery. We also did not observe an increase in hypercontinence with time, as has been reported previously in a study done primarily in men.21
Women are counseled both preoperatively and postoperatively about techniques of voiding and are monitored closely with assessments of postvoid residual urine volume.
The etiology of incontinence or hypercontinence after ONB is unclear but is likely multifactorial and related to patients’ baseline urinary function, surgical technique during urethral reconstruction, and postoperative changes to pelvic musculature. Multiple mechanisms for hypercontinence have been proposed, including downward migration of the neobladder-urethra angle.22,23
To address this, technical modifications, such as anterior fixation or posterior elevation of the neobladder, have been proposed, with varying success.22,24,25
We did not identify any clinical or pathologic variables significantly related to post-ONB urinary function. It seems clear, however, that rates of incontinence after ONB in women are higher than daytime (2-8%) and nighttime (20-24%) rates in men.3-5
Further study, as well as modification and refinement of surgical technique, in women undergoing ONB are needed to achieve better results.
Data regarding health-related quality of life after cystectomy and ONB were not available in our cohort. Although Gilbert et al published general results of the Bladder Cancer Index in 2007,26
the actual questionnaire was not made available until just recently.27
Using The Bladder Cancer Index, its creators reported superior urinary function
in patients undergoing ileal conduit diversion compared with ONB, but urinary bother
was equivalent in both groups.28
However, their cohort consisted primarily of men (81%), and the survey awaits further external validation.
There are several limitations to this study, including those inherent to the retrospective design. In addition, women were treated over a lengthy period by several surgeons, and therefore they represent a heterogeneous cohort. We did not analyze whether a nerve-sparing operation was performed because its impact on continence is controversial. There are also shortfalls to obtaining functional outcomes by physician interview.29
Finally, assessment of functional outcomes is partly limited by varying length of follow-up. No incontinent women developed continence with prolonged follow-up. However, it is possible that the risk of hypercontinence was underestimated in women with short follow-up.