Oncology Outcomes After Urothelial Recurrence in Bladder Cancer Patients Undergoing Radical Cystectomy Anirban P. Mitra, Mehrdad Alemozaffar, Brianna N. Harris, Anne K. Schuckman, Eila C. Skinner, and Siamak Daneshmand OBJECTIVE METHODS
RESULTS
CONCLUSION
To identify factors prognostic for survival after urothelial recurrence after radical cystectomy for bladder cancer. Of the 2029 patients with bladder cancer who underwent radical cystectomy at our institution, 80 (3.9%) patients experienced recurrence in the urothelium (upper urinary tract or urethra) and had sufficient follow-up for further analysis. Clinicopathologic characteristics were analyzed by univariate and multivariable analyses to identify factors prognostic for postrecurrence diseasespecific (PRDSS) and overall (PROS) survival. At median follow-up of 12 years, 25 (31.3%) and 55 (68.7%) patients experienced recurrence in the upper tract and urethra, respectively. Median time to recurrence, PRDSS, and PROS were 25.9, 58.4, and 48.7 months, respectively. Older age (P ¼ .018), patients with tumors that were upstaged at cystectomy compared with their clinical stage (P ¼ .049), and positive surgical margins (P ¼ .022) were associated with a lower PROS. The presence of symptoms at follow-up was associated with a poor PRDSS (P ¼ .028), which was confirmed by multivariable analysis. Patients experiencing urothelial recurrence within 2 years of cystectomy had a lower PRDSS (P ¼ .002) and PROS (P ¼ .003), which was confirmed by multivariable analysis. Site of urothelial recurrence did not influence time to recurrence (P ¼ .87), PRDSS (P ¼ .72), or PROS (P ¼ .57). Urothelial cancer relapse in the upper urinary tract or urethra has a comparable clinical course, and may be cured with extirpative surgery, with median PROS of 48.7 months after recurrence. Patients experiencing early urothelial recurrence face worse prognosis and should be considered candidates for adjuvant therapy. UROLOGY 84: 1420e1426, 2014. 2014 Elsevier Inc.
R
adical cystectomy is the gold standard treatment for patients with muscle-invasive urothelial carcinoma of the bladder (UCB), and for noninvasive disease that does not respond to intravesical therapy.1,2 Such tumors are generally aggressive in nature and lethal if the disease recurs. Although the 5-year postcystectomy recurrence-free survival probability for patients with organconfined node-negative UCB is approximately 85%, the 3year survival rate after soft tissue recurrence is only 12%.3,4 Disease recurrence in the urothelium (upper urinary tract or urethra), however, represents a unique situation that comprises the majority of late oncologic occurrences after Financial Disclosure: The authors declare that they have no relevant financial interests. From the Department of Pathology and Center for Personalized Medicine, University of Southern California and Norris Comprehensive Cancer Center, Los Angeles, CA; the Institute of Urology, University of Southern California and Norris Comprehensive Cancer Center, Los Angeles, CA; and the Department of Urology, Stanford University, Stanford, CA Address correspondence to: Siamak Daneshmand, M.D., Institute of Urology, University of Southern California and Norris Comprehensive Cancer Center, 1441 Eastlake Avenue, Suite 7416, Los Angeles, CA 90033. E-mail: [email protected] Submitted: April 8, 2014, accepted (with revisions): May 3, 2014
1420
ª 2014 Elsevier Inc. All Rights Reserved
cystectomy.5 These account for 2%-7% of all postcystectomy recurrences, and proper surveillance and prompt intervention can enhance survival in this population.6-12 Previous studies have identified prognostic covariates after soft tissue recurrence.4,13,14 However, although it is experiential knowledge that UCB recurrences in the upper tract and urethra have relatively less aggressive course than those in local and/or distant soft tissue sites, few investigations have compared features of disease relapses at these sites and prognostic factors for postrecurrence outcomes. Studies have also shown that 45%-62% of urothelial recurrences were only detected after the development of overt symptoms despite surveillance.6,9,11,12 This highlights the need to identify factors determining such recurrence to allow for early interventions, and to subsequently identify patients at risk for poor postrecurrence outcomes who may need more aggressive management. This study was therefore designed to identify the clinicopathologic features that could potentially predict time to urothelial recurrence and prognosis once a patient experienced such recurrence after cystectomy. http://dx.doi.org/10.1016/j.urology.2014.05.080 0090-4295/14
METHODS Patient Population Between 1971 and 2005, 2029 patients underwent radical cystectomy for UCB at our institution. Patient characteristics were entered into a prospectively maintained institutional review boardeapproved database. The patient database was consistently curated, maintained, and updated by a dedicated manager. Patients were proactively and meticulously followed up at specified intervals by phone calls and/or mailed correspondence to document disease course. Study inclusion criteria were (1) primary UCB treated with open radical cystectomy with the intent to cure, (2) clinical recurrence of urothelial carcinoma in the upper urinary tract or urethra after cystectomy, and (3) minimum 2-year postrecurrence follow-up if the patient was alive. Exclusion criteria were (1) presence of urethral or upper tract primaries or distant metastasis at diagnosis, (2) patients who underwent urethrectomy concurrently with cystectomy, (3) perioperative complications leading to death, and (4) UCB recurrence at any other local and/or distant soft tissue sites during the postcystectomy course. A total of 80 (3.9%) patients met the previously described criteria and were included for further analysis.
Patient-related Variables All surgical specimens were examined under a standardized pathologic protocol. Tumor staging and grading were standardized to the American Joint Committee on Cancer and World Health Organization recommendations, respectively.15,16 Pathologic stages were defined as organ confined, nonemuscle invasive (pT1N0M0); organ confined, muscle invasive (pT2N0M0); extravesical (pT3/4N0M0); and node positive (pTanyN1-3M0). Initial precystectomy clinical and final postcystectomy pathologic stages were compared to determine tumor upstaging. Urethral margin status was considered positive if tumor cells were observed in the apical urethral margin section on final pathology.17 Lymphovascular invasion and p53 status of the primary tumor were determined in 45 (56.3%) and 40 (50%) patients, respectively.18,19 Tumors that showed nuclear immunoreactivity in 10% cells were considered p53 altered.
Patient Management, Clinical Outcomes, and Analysis All patients underwent radical cystectomy, extended pelvic lymphadenectomy, and urinary diversion.3 Routine postoperative follow-up was generally at 4-month intervals in year 1, 6-month intervals in year 2, and annually thereafter. Radiographic evaluations of the urinary diversion, upper urinary tract with excretory urography, computed tomography scan or ultrasonography, and chest radiography were scheduled 4 months postoperatively and annually thereafter unless otherwise clinically indicated. The primary outcome of interest was postrecurrence diseasespecific survival (PRDSS), which was calculated from the time of urothelial recurrence to death due to disease; patients were censored at last follow-up if alive or dead due to other cause. Postrecurrence overall survival (PROS) was a secondary outcome measure, calculated from the time of urothelial recurrence to death; surviving patients were censored at last followup. Time to recurrence (TTR) was calculated from the date of cystectomy to the first documented urothelial recurrence. The Pearson chi-square test was used to examine the associations between categorical variables. The log-rank test was used to compare univariate differences in outcomes. The multivariable Cox proportional hazards modeling was used to estimate UROLOGY 84 (6), 2014
independent relationships between univariably prognostic covariates. All P values are 2 sided; P .050 was considered statistically significant and .050< P .10 was considered as a trend toward significance.
RESULTS Patient Characteristics and Disease Course Overall patient characteristics are listed in Table 1. The cohort comprised 74 (92.5%) men and 76 (95%) Caucasians; median (interquartile range [IQR]) age was 67.5 (61.8-74.3) years. A median (IQR) of 30 (19-44) lymph nodes were resected during surgery. Among the 6 (7.5%) patients with node positive disease, a median (IQR) of 2 (1-9) lymph nodes were identified with metastatic deposits; median lymph node density was 11.1%. Surgical margins were positive in 6 (7.5%) patients on final pathology, all of which corresponded to positive apical urethral margins; 5 (83.3%) of these patients recurred in the urethra. However, 136 (6.7%) patients among the principal population of 2029 patients had moderate-tosevere atypical (n ¼ 69) or positive (n ¼ 67) urethral margins; therefore, urethral recurrence rate among all patients in the population with moderate-to-severe atypical or positive urethral margins was 3.7% (5 of 136 patients). Median TTR was 25.9 months. In this cohort, older patients had shorter median TTR (P ¼ .020). Postcystectomy urothelial recurrences were documented in the upper urinary tract and urethra in 25 (31.3%) and 55 (68.7%) patients, respectively. TTR did not differ based on the site of urothelial recurrence (P ¼ .87). Thirty-two (40%) urothelial recurrences occurred within the first 2 years after cystectomy, whereas 21 (26.3%) patients recurred 4 years after surgery (Fig. 1A). Median overall follow-up was 12 years during which 55 (68.7%) patients died; 42 (76.4%) deaths were directly attributed to UCB. Median postrecurrence follow-up was 37.7 months (range, 2.7 months-32.1 years). Of 47 evaluable cases, 22 (46.8%) patients presented symptomatically at follow-up with urethral discharge, meatal itching, or gross hematuria; site of recurrence was not associated with symptomatic presentation (chi-square P ¼ .22). Postrecurrence salvage chemotherapy was administered to 19 (23.8%) patients. Twenty-four (96%) patients with upper tract recurrence subsequently underwent nephroureterectomy, of which 11 (45.8%) died of UCB at last follow-up (Fig. 1B). Forty-nine (89.1%) patients with urethral recurrence underwent urethrectomy, of which 28 (57.1%) ultimately succumbed to UCB. Three patients with urethral recurrence were managed conservatively; 2 patients were treated with intraurethral 5-fluorouracil and 1 was observed. Intervention for recurrence was initiated within 30 days in 38 (47.5%) patients. Median PRDSS and PROS were 58.4 and 48.7 months, respectively. To assess any potential variability in the clinicopathologic characteristics and outcome metrics across the 35-year cohort, patients were categorized into 1 of 2 eras: cystectomy performed on or before December 1990 1421
Table 1. Univariate associations of patient demographic, clinicopathologic, and management characteristics with time to bladder cancer recurrence Time to Recurrence Characteristic Study cohort Age, y
RESULTS
CONCLUSION
To identify factors prognostic for survival after urothelial recurrence after radical cystectomy for bladder cancer. Of the 2029 patients with bladder cancer who underwent radical cystectomy at our institution, 80 (3.9%) patients experienced recurrence in the urothelium (upper urinary tract or urethra) and had sufficient follow-up for further analysis. Clinicopathologic characteristics were analyzed by univariate and multivariable analyses to identify factors prognostic for postrecurrence diseasespecific (PRDSS) and overall (PROS) survival. At median follow-up of 12 years, 25 (31.3%) and 55 (68.7%) patients experienced recurrence in the upper tract and urethra, respectively. Median time to recurrence, PRDSS, and PROS were 25.9, 58.4, and 48.7 months, respectively. Older age (P ¼ .018), patients with tumors that were upstaged at cystectomy compared with their clinical stage (P ¼ .049), and positive surgical margins (P ¼ .022) were associated with a lower PROS. The presence of symptoms at follow-up was associated with a poor PRDSS (P ¼ .028), which was confirmed by multivariable analysis. Patients experiencing urothelial recurrence within 2 years of cystectomy had a lower PRDSS (P ¼ .002) and PROS (P ¼ .003), which was confirmed by multivariable analysis. Site of urothelial recurrence did not influence time to recurrence (P ¼ .87), PRDSS (P ¼ .72), or PROS (P ¼ .57). Urothelial cancer relapse in the upper urinary tract or urethra has a comparable clinical course, and may be cured with extirpative surgery, with median PROS of 48.7 months after recurrence. Patients experiencing early urothelial recurrence face worse prognosis and should be considered candidates for adjuvant therapy. UROLOGY 84: 1420e1426, 2014. 2014 Elsevier Inc.
R
adical cystectomy is the gold standard treatment for patients with muscle-invasive urothelial carcinoma of the bladder (UCB), and for noninvasive disease that does not respond to intravesical therapy.1,2 Such tumors are generally aggressive in nature and lethal if the disease recurs. Although the 5-year postcystectomy recurrence-free survival probability for patients with organconfined node-negative UCB is approximately 85%, the 3year survival rate after soft tissue recurrence is only 12%.3,4 Disease recurrence in the urothelium (upper urinary tract or urethra), however, represents a unique situation that comprises the majority of late oncologic occurrences after Financial Disclosure: The authors declare that they have no relevant financial interests. From the Department of Pathology and Center for Personalized Medicine, University of Southern California and Norris Comprehensive Cancer Center, Los Angeles, CA; the Institute of Urology, University of Southern California and Norris Comprehensive Cancer Center, Los Angeles, CA; and the Department of Urology, Stanford University, Stanford, CA Address correspondence to: Siamak Daneshmand, M.D., Institute of Urology, University of Southern California and Norris Comprehensive Cancer Center, 1441 Eastlake Avenue, Suite 7416, Los Angeles, CA 90033. E-mail: [email protected] Submitted: April 8, 2014, accepted (with revisions): May 3, 2014
1420
ª 2014 Elsevier Inc. All Rights Reserved
cystectomy.5 These account for 2%-7% of all postcystectomy recurrences, and proper surveillance and prompt intervention can enhance survival in this population.6-12 Previous studies have identified prognostic covariates after soft tissue recurrence.4,13,14 However, although it is experiential knowledge that UCB recurrences in the upper tract and urethra have relatively less aggressive course than those in local and/or distant soft tissue sites, few investigations have compared features of disease relapses at these sites and prognostic factors for postrecurrence outcomes. Studies have also shown that 45%-62% of urothelial recurrences were only detected after the development of overt symptoms despite surveillance.6,9,11,12 This highlights the need to identify factors determining such recurrence to allow for early interventions, and to subsequently identify patients at risk for poor postrecurrence outcomes who may need more aggressive management. This study was therefore designed to identify the clinicopathologic features that could potentially predict time to urothelial recurrence and prognosis once a patient experienced such recurrence after cystectomy. http://dx.doi.org/10.1016/j.urology.2014.05.080 0090-4295/14
METHODS Patient Population Between 1971 and 2005, 2029 patients underwent radical cystectomy for UCB at our institution. Patient characteristics were entered into a prospectively maintained institutional review boardeapproved database. The patient database was consistently curated, maintained, and updated by a dedicated manager. Patients were proactively and meticulously followed up at specified intervals by phone calls and/or mailed correspondence to document disease course. Study inclusion criteria were (1) primary UCB treated with open radical cystectomy with the intent to cure, (2) clinical recurrence of urothelial carcinoma in the upper urinary tract or urethra after cystectomy, and (3) minimum 2-year postrecurrence follow-up if the patient was alive. Exclusion criteria were (1) presence of urethral or upper tract primaries or distant metastasis at diagnosis, (2) patients who underwent urethrectomy concurrently with cystectomy, (3) perioperative complications leading to death, and (4) UCB recurrence at any other local and/or distant soft tissue sites during the postcystectomy course. A total of 80 (3.9%) patients met the previously described criteria and were included for further analysis.
Patient-related Variables All surgical specimens were examined under a standardized pathologic protocol. Tumor staging and grading were standardized to the American Joint Committee on Cancer and World Health Organization recommendations, respectively.15,16 Pathologic stages were defined as organ confined, nonemuscle invasive (pT1N0M0); organ confined, muscle invasive (pT2N0M0); extravesical (pT3/4N0M0); and node positive (pTanyN1-3M0). Initial precystectomy clinical and final postcystectomy pathologic stages were compared to determine tumor upstaging. Urethral margin status was considered positive if tumor cells were observed in the apical urethral margin section on final pathology.17 Lymphovascular invasion and p53 status of the primary tumor were determined in 45 (56.3%) and 40 (50%) patients, respectively.18,19 Tumors that showed nuclear immunoreactivity in 10% cells were considered p53 altered.
Patient Management, Clinical Outcomes, and Analysis All patients underwent radical cystectomy, extended pelvic lymphadenectomy, and urinary diversion.3 Routine postoperative follow-up was generally at 4-month intervals in year 1, 6-month intervals in year 2, and annually thereafter. Radiographic evaluations of the urinary diversion, upper urinary tract with excretory urography, computed tomography scan or ultrasonography, and chest radiography were scheduled 4 months postoperatively and annually thereafter unless otherwise clinically indicated. The primary outcome of interest was postrecurrence diseasespecific survival (PRDSS), which was calculated from the time of urothelial recurrence to death due to disease; patients were censored at last follow-up if alive or dead due to other cause. Postrecurrence overall survival (PROS) was a secondary outcome measure, calculated from the time of urothelial recurrence to death; surviving patients were censored at last followup. Time to recurrence (TTR) was calculated from the date of cystectomy to the first documented urothelial recurrence. The Pearson chi-square test was used to examine the associations between categorical variables. The log-rank test was used to compare univariate differences in outcomes. The multivariable Cox proportional hazards modeling was used to estimate UROLOGY 84 (6), 2014
independent relationships between univariably prognostic covariates. All P values are 2 sided; P .050 was considered statistically significant and .050< P .10 was considered as a trend toward significance.
RESULTS Patient Characteristics and Disease Course Overall patient characteristics are listed in Table 1. The cohort comprised 74 (92.5%) men and 76 (95%) Caucasians; median (interquartile range [IQR]) age was 67.5 (61.8-74.3) years. A median (IQR) of 30 (19-44) lymph nodes were resected during surgery. Among the 6 (7.5%) patients with node positive disease, a median (IQR) of 2 (1-9) lymph nodes were identified with metastatic deposits; median lymph node density was 11.1%. Surgical margins were positive in 6 (7.5%) patients on final pathology, all of which corresponded to positive apical urethral margins; 5 (83.3%) of these patients recurred in the urethra. However, 136 (6.7%) patients among the principal population of 2029 patients had moderate-tosevere atypical (n ¼ 69) or positive (n ¼ 67) urethral margins; therefore, urethral recurrence rate among all patients in the population with moderate-to-severe atypical or positive urethral margins was 3.7% (5 of 136 patients). Median TTR was 25.9 months. In this cohort, older patients had shorter median TTR (P ¼ .020). Postcystectomy urothelial recurrences were documented in the upper urinary tract and urethra in 25 (31.3%) and 55 (68.7%) patients, respectively. TTR did not differ based on the site of urothelial recurrence (P ¼ .87). Thirty-two (40%) urothelial recurrences occurred within the first 2 years after cystectomy, whereas 21 (26.3%) patients recurred 4 years after surgery (Fig. 1A). Median overall follow-up was 12 years during which 55 (68.7%) patients died; 42 (76.4%) deaths were directly attributed to UCB. Median postrecurrence follow-up was 37.7 months (range, 2.7 months-32.1 years). Of 47 evaluable cases, 22 (46.8%) patients presented symptomatically at follow-up with urethral discharge, meatal itching, or gross hematuria; site of recurrence was not associated with symptomatic presentation (chi-square P ¼ .22). Postrecurrence salvage chemotherapy was administered to 19 (23.8%) patients. Twenty-four (96%) patients with upper tract recurrence subsequently underwent nephroureterectomy, of which 11 (45.8%) died of UCB at last follow-up (Fig. 1B). Forty-nine (89.1%) patients with urethral recurrence underwent urethrectomy, of which 28 (57.1%) ultimately succumbed to UCB. Three patients with urethral recurrence were managed conservatively; 2 patients were treated with intraurethral 5-fluorouracil and 1 was observed. Intervention for recurrence was initiated within 30 days in 38 (47.5%) patients. Median PRDSS and PROS were 58.4 and 48.7 months, respectively. To assess any potential variability in the clinicopathologic characteristics and outcome metrics across the 35-year cohort, patients were categorized into 1 of 2 eras: cystectomy performed on or before December 1990 1421
Table 1. Univariate associations of patient demographic, clinicopathologic, and management characteristics with time to bladder cancer recurrence Time to Recurrence Characteristic Study cohort Age, y
