- Email: [email protected]
Recurrence-Free Survival After Radical Cystectomy of Patients Downstaged by Transurethral Resection
Adult Urology Recurrence-Free Survival After Radical Cystectomy of Patients Downstaged by Transurethral Resection Matthew E. Nielsen, Patrick J. Bastian, Ganesh S. Palapattu, Bruce J. Trock, Mark P. Schoenberg, Theresa Chan, and Craig G. Rogers OBJECTIVES
METHODS
RESULTS
CONCLUSIONS
The finding of bladder cancer invading the detrusor muscle on transurethral resection (TUR) is one of the clearest indications for radical cystectomy. To the extent that detrusor invasion is, in practical effect, a binary variable, the variety of outcomes after radical cystectomy in these patients belies the simplicity of this approach. In this context, we assessed bladder cancer recurrence-free survival among patients noted to have muscle-invasive urothelial carcinoma (transitional cell cancer [TCC]) on staging TUR subsequently found to have non-muscleinvasive TCC at radical cystectomy (downstaged). The records of 248 consecutive patients who underwent radical cystectomy for TCC at a single academic institution from 1994 to 2002 were retrospectively reviewed. Of these patients, 112 (45%) had documented muscle-invasive disease by TUR and were clear of gross residual tumor on cystoscopy before radical cystectomy. Of the 112 patients, 25 (22.3%) were downstaged to non-muscle-invasive disease (Stage pT1 or less) at cystectomy and 87 (77.7%) had persistent muscle-invasive disease (Stage pT2 or greater) at cystectomy. Recurrence occurred in 4 downstaged patients (16.0%) compared with 29 patients (33.3%) who were not downstaged (P ⫽ 0.094). Kaplan-Meier analysis demonstrated a statistically significant improvement in recurrence-free survival with downstaging (log-rank P ⫽ 0.008). Multivariate analysis demonstrated a threefold reduction in recurrence risk with tumor downstaging (hazard ratio 0.33, 95% confidence interval 0.10 to 1.12) that approached statistical significance (P ⫽ 0.075). Nodal status was the strongest predictor of RFS. Downstaging from muscle-invasive TCC on TUR to non-muscle-invasive TCC at radical cystectomy can be associated with a reduced risk of recurrence even after adjusting for lymph node status and adjuvant chemotherapy. UROLOGY 70: 1091–1095, 2007. © 2007 Elsevier Inc.
F
rom the initial staging system proposed by Jewett and Strong,1 it has been axiomatic that outcomes in bladder cancer (BCA) correlate with the depth of invasion of the primary tumor.1– 4 In clinical practice, invasion into the muscularis propria is, for all intents and purposes, treated as a dichotomous variable. Detrusor muscle invasion in a transurethral resection (TUR) specimen provides a clear rationale to proceed to radical primary therapy, predominantly in the form of radical cystectomy. Even if we accept that muscle invasion on clinical staging represents a practical threshold for the selection
From the James Buchanan Brady Urological Institute, Division of Epidemiology, James Buchanan Brady Urological Institute, and Department of Pathology, Johns Hopkins Hospital, Baltimore, Maryland Reprint requests: Matthew E. Nielsen, M.D., James Buchanan Brady Urological Institute, Johns Hopkins School of Medicine, 600 North Wolfe Street, Baltimore, MD 21287-2101. E-mail: [email protected] Submitted: May 16, 2006; accepted (with revisions): August 16, 2007
© 2007 Elsevier Inc. All Rights Reserved
of certain treatment strategies, it stands to reason that the extent of invasiveness behaves as a continuous variable. A distinct heterogeneity is present in the spectrum of clinical behavior for patients with tumors infiltrating beyond the boundary of the detrusor muscle. Several groups have presented different organ-sparing approaches with either radical TUR alone5,6 or TUR in combination with chemotherapy and radiotherapy7,8 for highly select subgroups of patients with clinically muscle-invasive disease. This variability is further complicated by differences in the extent of TUR. In this background, the specific question of the prognostic significance of downstaging from muscle-invasive cancer at presentation to non-muscle-invasive BCA at radical cystectomy is not well-established in published reports. We examined recurrence-free survival in patients with a clinical diagnosis of organconfined muscle-invasive BCA downstaged to non-muscle-invasive disease on final pathologic examination of the cystectomy specimen. 0090-4295/07/$32.00 doi:10.1016/j.urology.2007.08.044
1091
MATERIAL AND METHODS Patient Population From August 1994 to April 2002, a total of 248 consecutive patients (mean age 66 years, range 40.1 to 88) underwent radical cystectomy and bilateral pelvic lymphadenectomy as definitive treatment for muscle-invasive or recurrent high-grade superficial bladder transitional cell cancer (TCC) by a single surgeon at a tertiary referral center. Of these 248 patients, 121 (49%) had documented muscle-invasive disease on TUR before cystectomy without clinical evidence of extravesical disease (by bimanual examination or imaging). Patients with evidence of gross residual tumor on preoperative cystoscopy (n ⫽ 9) were excluded. This left 112 patients with muscle-invasive, highgrade TCC rendered grossly free of disease after staging TUR for analysis. Clinical and pathologic information was retrospectively obtained from individual patient charts, medical records, and approved cancer registries. No patient had known distant metastatic disease at cystectomy. The institutional review board approved the study before it began.
Pathologic Examination All radical cystectomy specimens were examined by a dedicated genitourinary pathologist. The 1997 American Joint Committee on Cancer/TNM classification system was used for pathologic staging and the World Health Organization classification was used for pathologic grading. Patients who underwent surgery before 1997 had their pathologic stage updated to reflect the 1997 TNM staging system. The TUR specimens were submitted in their entirety. Multiple sections from the tumor, bladder wall, and mucosa adjacent to, and distant from, the tumor, along with the ureters and regional lymph nodes, were evaluated. In cystectomy specimens in which no residual or minimal tumor was identified, the remaining bladder was extensively sampled.
Follow-up Patients were scheduled to undergo postoperative physical examination and routine serum chemistry studies approximately every 3 months for the first year, semiannually from the second through the fifth year, and annually thereafter. Radiographic evaluation, including intravenous pyelography, ultrasonography, and chest radiography was obtained annually unless clinically indicated. Additional studies, including bone scanning and computed tomography, were performed when clinically indicated. The median duration of follow-up was 41 months (range 1 to 120). For the purposes of this analysis, bladder cancer recurrence was defined as the development of local and/or distant recurrence. Whenever possible, recurrent lesions were biopsied to confirm the presence of bladder cancer.
Statistical Analysis The Fisher exact test and chi-square test were used to evaluate the association between categorical variables. The differences in variables with a continuous distribution between dichotomous categories were assessed using the Mann-Whitney U test. The differences in variables with a continuous distribution across ranked categories were assessed using the Kruskall-Wallis nonparametric analysis of variance. The Kaplan-Meier method was used to estimate the survival distribution, and the differences were assessed with the log-rank statistic. Multivariate survival analysis was performed with the Cox proportional 1092
Table 1. Clinical characteristics of patients, stratified by correlation of clinical to pathologic stage (pathologic stage equal to or greater than clinical stage vs. pathologic downstaging)
Variable Mean age at cystectomy* (yr) Sex Male Female Ethnicity White African American Asian Lymph node status at cystectomy Node negative Node positive Preoperative intravesical chemotherapy No chemotherapy Chemotherapy Postoperative chemotherapy No chemotherapy Chemotherapy Recurrence No Yes
Not downstaged (n ⫽ 87) 66.0 0.429 66 (75.9) 21 (24.1) 0.109 79 (90.8) 5 (5.7) 2 (2.3) 0.063
Downstaged (n ⫽ 25) 66.2 17 (68.0) 8 (32.0) 16 (64.0) 3 (12.0) 2 (8.0)
65 (74.7) 22 (25.3) 0.768
23 (92.0) 2 (8.0)
54 (62.1) 20 (23.0) 0.126
16 (64.0) 5 (20.0)
63 (72.4) 23 (26.4) 0.094 58 (66.7) 29 (33.3)
22 (88.0) 3 (12.0) 21 (84.0) 4 (16.0)
* P ⫽ 0.928.
hazards regression model. The model fit was evaluated by the likelihood ratio test. Statistical significance in this study was set as P ⱕ0.05. All reported P values are two-sided. All analyses were performed using Statistical Analysis Systems software (SAS Institute, Cary NC).
RESULTS The patient demographics are presented in Table 1, and the clinical characteristics of the cohort, stratified by downstaging, are presented in Table 2. Of the 112 patients with muscle-invasive disease on TUR, 25 (22.3%) were downstaged to non-muscle-invasive disease (pT1 or less) at cystectomy. Four patients were downstaged to Stage pT0. Recurrence occurred in 4 (16.0%) of the downstaged patients compared with 29 (33.3%) of the patients who were not downstaged (P ⫽ 0.094). The sites of recurrence included the lung, pelvis, retroperitoneum, and urethra. No patient developed subsequent upper tract TCC. The patients who were downstaged from muscle-invasive to non-muscle-invasive disease had significantly improved recurrence-free survival, as demonstrated by Kaplan-Meier analysis (Fig. 1). Univariate proportional hazards regression analysis demonstrated a statistically significant 75% decrease in the disease recurrence rate associated with tumor downstaging (hazard ratio [HR] 0.25, 95% confidence interval [CI] 0.08 to 0.84, P ⫽ 0.015). Also, a significant difference was found in recurUROLOGY 70 (6), 2007
Table 2. Multivariate Cox proportional hazards analysis recurrence in patients with muscle-invasive bladder cancer documented by TUR who underwent radical cystectomy from 1994 to 2002 Variable Model 1 (n ⫽ 112) Downstage (pT1 or less vs. pT2) Lymph nodes (positive vs. negative) Model 2 (n ⫽ 111)‡ Downstage (pT1 or less vs. pT2) Adjuvant chemotherapy (yes vs. no)
HR
95% CI
P Value*
0.33 2.68
0.10–1.12 1.31–5.47
0.075 0.007
0.29 2.69
0.09–0.97 1.33–5.42
0.044 0.007
†
TUR ⫽ transurethral resection; HR ⫽ hazard ratio; CI ⫽ confidence interval. * Wald test. † Likelihood ratio test of goodness of fit of model, chi-square (2 degrees of freedom) ⫽ 14.10, P ⫽ 0.0009. ‡ Likelihood ratio test of goodness of fit of model, chi-square (2 degrees of freedom) ⫽ 14.14, P ⫽ 0.0008.
0.29, 95% CI 0.09 to 0.97, P ⫽ 0.044; Table 2). However, owing to the high degree of collinearity between lymph node status and adjuvant chemotherapy, both variables could not be included simultaneously in a multivariate model.9 Both models gave similarly good fits to the data, indicated by the near identical values for the likelihood ratio test (Table 2). Other variables, including age, sex, carcinoma in situ at TUR or cystectomy, grade, or intravesical chemotherapy, did not significantly alter these associations.
COMMENT Figure 1. Influence of tumor downstaging after TUR on recurrence-free survival in 112 patients with muscle-invasive TCC who underwent radical cystectomy.
rence-free survival among patients downstaged to Stage pT1 versus less than Stage pT1 (pT0, pTa, pTis). Of 14 patients, 1 (7.1%) was downstaged to less than Stage pT1 recurred compared with 3 (27.3%) of 11 patients downstaged to Stage pT1 (log-rank chi-square ⫽ 4.00, P ⫽ 0.045). The effect of tumor downstaging on BCA recurrence can be confounded by lymph node status. Of the downstaged patients, 8% had positive lymph nodes compared with 25% of patients who were not downstaged (P ⫽ 0.063). Multivariate Cox proportional hazards analysis suggested a reduction in BCA recurrence risk with tumor downstaging after adjusting for lymph node status and chemotherapy (Table 2). A threefold reduction in recurrence risk with tumor downstaging that approached statistical significance was seen when lymph node status was included in the multivariate model (HR 0.33, 95% CI 0.10 to 1.12, P ⫽ 0.075). A trend was found toward adjuvant chemotherapy being preferentially administered to patients without downstaging versus downstaged patients (27% versus 12%, respectively, P ⫽ 0.126). Adjuvant chemotherapy administration correlated highly with positive lymph nodes (P ⬍0.0001). Even when postoperative chemotherapy was included in a multivariate model with downstaging, the magnitude of the risk reduction effect of downstaging was similar to the previous model, in this case achieving statistical significance (HR UROLOGY 70 (6), 2007
Discrepant clinical and pathologic staging has been described in the bladder cancer studies; however, few studies have specifically addressed the clinical outcomes of patients downstaged from muscle-invasive disease. Our results suggest that patients who are downstaged from muscle-invasive disease on TUR to non-muscle-invasive disease in the radical cystectomy specimen have improved recurrence-free survival compared with patients with persistent muscle invasion at cystectomy. Despite resecting all gross cancer by TUR, a high percentage of patients still have pathologic T2 or T3 disease. Even after adjusting for lymph node status or adjuvant chemotherapy, downstaging reduces the risk of recurrence by 67% to 71% relative to patients with persistent muscle-invasive disease in the cystectomy specimen. Although, as expected, the independent prognostic significance of downstaging is attenuated after adjustment for lymph node status, the magnitude of the effect is similar to that in the model adjusted for adjuvant chemotherapy, which did attain conventional significance levels. Furthermore, the HRs for the univariate (HR 0.25) and multivariate (HR ⫽ 0.29)-adjusted effects of downstaging are similar, suggesting that only minor confounding occurred by either lymph node status or chemotherapy. Waehre et al.10 reported a multi-institutional experience of 227 patients with muscle-invasive bladder cancer (Stage cT2-T4a) who underwent radical cystectomy after neoadjuvant radiotherapy. Downstaging to a pathologic T stage that was less than that observed at TUR occurred in 97 patients (42.7%). The 5-year cancer-specific survival rate of those patients who were not downstaged was 46% compared with 73% for patients downstaged to pT0 1093
disease (P ⬍0.0001). Thus, the investigators stated that patients with the stage reduced to pT1 or less had significantly improved survival relative to patients with evidence of muscle-invasive disease on pathologic analysis of the cystectomy specimen. An important caveat to this comparison is that the survival advantage of downstaging in this study was substantially intertwined with the effects of neoadjuvant treatment, because all patients underwent preoperative radiotherapy. It is conceivable that this could account, at least in part, for the somewhat high incidence of Stage pT0 disease. Furthermore, stage reduction was more common in the 30% of patients who received neoadjuvant chemotherapy (53% versus 40%). Other reports have also suggested a survival advantage of downstaging to pT0 by TUR.11–14 Conversely, Thrasher et al.15 found no survival advantage for patients with bladder cancer who were downstaged to Stage pT0 at radical cystectomy.15 In their population of patients undergoing cystectomy from 1969 to 1990, survival was determined by the presenting clinical stage. The 5-year cancer-specific survival of patients with clinical Stage T2 disease on TUR, who were downstaged to pT0 at radical cystectomy, was equivalent to that of the patients who had Stage pT2 on the final pathologic analysis (approximately 68%). A small number of patients received neoadjuvant radiotherapy, but this did not significantly alter their stated conclusions. These data are somewhat historical, and other potential confounders, such as lymph node status or adjuvant therapy, were not adjusted for in their analysis. Nevertheless, their data would seem to support the concept that survival after radical cystectomy is a function of occult metastatic disease already present at surgery. A similar correlation between prognosis and clinical stage was demonstrated in a recent study by Hassan et al.,16 who investigated 50 patients with pathologic stage carcinoma in situ-only on the final cystectomy specimen. The factor with the greatest influence on prognosis was clinical staging before cystectomy. The 22 patients with a TUR diagnosis of muscle invasion had a significantly greater rate of metastatic disease compared with patients without evidence of muscle invasion on TUR (22.7% versus 3.6%), despite complete resection of invasive disease. The data from Herr5,12 on the outcomes of patients with muscle-invasive bladder cancer (Stage cT2) treated only by TUR present a unique window into the natural history of bladder cancer relevant to the present discussion. In this series, the 10-year disease-specific survival rate after TUR only was 76%, similar to the 10-year disease-specific survival rate of 71% for patients who had undergone radical cystectomy as primary treatment (P ⫽ 0.3). Importantly, the disease-specific survival rate at 10 years for patients with Stage cT0 disease on restaging TUR was 82% compared with 57% for patients with clinical Stage T1 disease on restaging TUR (P ⫽ 0.003). Solsona et al.6 presented a similar series of 133 patients 1094
treated by radical TUR alone, with a 5 and 10-year cause-specific survival rate of 80.5% and 74.5%, respectively. These data suggest that select patients might benefit from TUR of their primary tumor and that downstaging by TUR can offer a survival benefit. The present study had the general limitations inherent to any retrospective series. One might logically submit that smaller bladder tumors are easier to resect completely by TUR and are, therefore, more likely to be downstaged. Large bladder tumors have been shown to be associated with worse clinical outcomes, independent of stage. Our study might have selected for patients with a smaller initial tumor burden. An additional limitation of the present study was that, inherent to the tertiary referral nature of our series, the index tumors were not routinely evaluated and resected by the operating surgeon. It is possible that unmeasured variations in the factors, including multifocality, morphology (sessile versus papillary), and primary tumor size, could have affected the composition of the study population. The relative uncertainty of this circumstance was mitigated in that each patient underwent a staging endoscopic evaluation by the operating surgeon before cystectomy, at which time no grossly evident residual tumor was noted. Patients with endoscopic or radiographic evidence of gross residual tumor were excluded from this analysis, because it was thought that they would not have had the opportunity to be downstaged. Additionally, it is possible that variations in the degree or extent of muscle invasion not captured in our data could have further refined the stratification of patients with clinical Stage T2 bladder cancer. Ultimately, a routine protocol of restaging TUR, as has been described elsewhere,5 might have more thoroughly stratified the risk groups among our patient cohort. Our results need to be validated by larger studies with longer follow-up, in which the relationship of tumor downstaging to disease recurrence, as well as overall and cancer-specific survival, can be further elucidated.
CONCLUSIONS Patients who are downstaged from muscle-invasive TCC on TUR to non-muscle-invasive TCC on radical cystectomy might have a reduced risk of disease recurrence compared with patients with persistent muscle-invasive disease in the radical cystectomy specimen. Longer follow-up and validation from larger series are needed to better characterize this issue and ascertain whether analogous relationships with improved postcystectomy overall or cancer-specific survival exist for patients with muscle-invasive tumors amenable to stage reduction by TUR. An enhanced understanding of this aspect of the natural history of bladder cancer will better inform our interpretation of the results from the variety of bladder-sparing protocols in evolution at some centers. UROLOGY 70 (6), 2007
References 1. Jewett HJ, and Strong GH: Infiltrating carcinoma of the bladder: relation of depth of penetration of the bladder wall to incidence of local extension and metastases. J Urol 55: 366 –370, 1946. 2. Herr HW: Uncertainty, stage and outcome of invasive bladder cancer. J Urol 152: 401– 402, 1994. 3. Prout GR Jr: Bladder carcinoma and a TNM system of classification. J Urol 117: 583–590, 1977. 4. Jewett HJ, and Strong GH: Infiltrating carcinoma of the bladder: relation of depth of penetration of the bladder wall to incidence of local extension and metastases. J Urol 55: 366 –370, 1946. 5. Herr HW: Transurethral resection of muscle-invasive bladder cancer: 10-year outcome. J Clin Oncol 19: 89 –93, 2001. 6. Solsona E, Iborra I, Ricos JV, et al: Feasibility of transurethral resection for muscle infiltrating carcinoma of the bladder: longterm followup of a prospective study. J Urol 159: 95–99, 1998. 7. Kaufman DS, Shipley WU, Griffin PP, et al: Selective bladder preservation by combination treatment of invasive bladder cancer. N Engl J Med 329: 1377–1382, 1993. 8. Shipley WU, Kaufman DS, Tester WJ, et al: Overview of bladder cancer trials in the Radiation Therapy Oncology Group. Cancer 97: 2115–2119, 2003. 9. McGee D, Reed D, and Yano K: The results of logistic analyses when the variables are highly correlated: an empirical example
UROLOGY 70 (6), 2007
10.
11.
12. 13.
14.
15.
16.
using diet and CHD incidence. J Chronic Dis 37: 713–719, 1984. Waehre H, Ous S, Klevmark B, et al: A bladder cancer multiinstitutional experience with total cystectomy for muscle-invasive bladder cancer. Cancer 72: 3044 –3051, 1993. Shipley WU, Kaufman DS, and Heney NM: Can chemo-radiotherapy plus transurethral tumor resection make cystectomy unnecessary for invasive bladder cancer? Oncology (Williston Park) 4: 25–34, 39, 1990. Herr HW: Staging invasive bladder tumors. J Surg Oncol 51: 217–220, 1992. Lee SE, Jeong IG, Ku JH, et al: Impact of transurethral resection of bladder tumor: analysis of cystectomy specimens to evaluate for residual tumor. Urology 63: 873– 877, 2004. Shipley WU, Kaufman DS, and Heney NM: Can chemo-radiotherapy plus transurethral tumor resection make cystectomy unnecessary for invasive bladder cancer? Oncology (Huntingt) 4: 25–34, 39, 1990. Thrasher JB, Frazier HA, Robertson JE, et al: Does of stage pT0 cystectomy specimen confer a survival advantage in patients with minimally invasive bladder cancer? J Urol 152: 393–396, 1994. Hassan JM, Cookson MS, Smith JA Jr, et al: Outcomes in patients with pathological carcinoma in situ only disease at radical cystectomy. J Urol 172: 882– 884, 2004.
1095
METHODS
RESULTS
CONCLUSIONS
The finding of bladder cancer invading the detrusor muscle on transurethral resection (TUR) is one of the clearest indications for radical cystectomy. To the extent that detrusor invasion is, in practical effect, a binary variable, the variety of outcomes after radical cystectomy in these patients belies the simplicity of this approach. In this context, we assessed bladder cancer recurrence-free survival among patients noted to have muscle-invasive urothelial carcinoma (transitional cell cancer [TCC]) on staging TUR subsequently found to have non-muscleinvasive TCC at radical cystectomy (downstaged). The records of 248 consecutive patients who underwent radical cystectomy for TCC at a single academic institution from 1994 to 2002 were retrospectively reviewed. Of these patients, 112 (45%) had documented muscle-invasive disease by TUR and were clear of gross residual tumor on cystoscopy before radical cystectomy. Of the 112 patients, 25 (22.3%) were downstaged to non-muscle-invasive disease (Stage pT1 or less) at cystectomy and 87 (77.7%) had persistent muscle-invasive disease (Stage pT2 or greater) at cystectomy. Recurrence occurred in 4 downstaged patients (16.0%) compared with 29 patients (33.3%) who were not downstaged (P ⫽ 0.094). Kaplan-Meier analysis demonstrated a statistically significant improvement in recurrence-free survival with downstaging (log-rank P ⫽ 0.008). Multivariate analysis demonstrated a threefold reduction in recurrence risk with tumor downstaging (hazard ratio 0.33, 95% confidence interval 0.10 to 1.12) that approached statistical significance (P ⫽ 0.075). Nodal status was the strongest predictor of RFS. Downstaging from muscle-invasive TCC on TUR to non-muscle-invasive TCC at radical cystectomy can be associated with a reduced risk of recurrence even after adjusting for lymph node status and adjuvant chemotherapy. UROLOGY 70: 1091–1095, 2007. © 2007 Elsevier Inc.
F
rom the initial staging system proposed by Jewett and Strong,1 it has been axiomatic that outcomes in bladder cancer (BCA) correlate with the depth of invasion of the primary tumor.1– 4 In clinical practice, invasion into the muscularis propria is, for all intents and purposes, treated as a dichotomous variable. Detrusor muscle invasion in a transurethral resection (TUR) specimen provides a clear rationale to proceed to radical primary therapy, predominantly in the form of radical cystectomy. Even if we accept that muscle invasion on clinical staging represents a practical threshold for the selection
From the James Buchanan Brady Urological Institute, Division of Epidemiology, James Buchanan Brady Urological Institute, and Department of Pathology, Johns Hopkins Hospital, Baltimore, Maryland Reprint requests: Matthew E. Nielsen, M.D., James Buchanan Brady Urological Institute, Johns Hopkins School of Medicine, 600 North Wolfe Street, Baltimore, MD 21287-2101. E-mail: [email protected] Submitted: May 16, 2006; accepted (with revisions): August 16, 2007
© 2007 Elsevier Inc. All Rights Reserved
of certain treatment strategies, it stands to reason that the extent of invasiveness behaves as a continuous variable. A distinct heterogeneity is present in the spectrum of clinical behavior for patients with tumors infiltrating beyond the boundary of the detrusor muscle. Several groups have presented different organ-sparing approaches with either radical TUR alone5,6 or TUR in combination with chemotherapy and radiotherapy7,8 for highly select subgroups of patients with clinically muscle-invasive disease. This variability is further complicated by differences in the extent of TUR. In this background, the specific question of the prognostic significance of downstaging from muscle-invasive cancer at presentation to non-muscle-invasive BCA at radical cystectomy is not well-established in published reports. We examined recurrence-free survival in patients with a clinical diagnosis of organconfined muscle-invasive BCA downstaged to non-muscle-invasive disease on final pathologic examination of the cystectomy specimen. 0090-4295/07/$32.00 doi:10.1016/j.urology.2007.08.044
1091
MATERIAL AND METHODS Patient Population From August 1994 to April 2002, a total of 248 consecutive patients (mean age 66 years, range 40.1 to 88) underwent radical cystectomy and bilateral pelvic lymphadenectomy as definitive treatment for muscle-invasive or recurrent high-grade superficial bladder transitional cell cancer (TCC) by a single surgeon at a tertiary referral center. Of these 248 patients, 121 (49%) had documented muscle-invasive disease on TUR before cystectomy without clinical evidence of extravesical disease (by bimanual examination or imaging). Patients with evidence of gross residual tumor on preoperative cystoscopy (n ⫽ 9) were excluded. This left 112 patients with muscle-invasive, highgrade TCC rendered grossly free of disease after staging TUR for analysis. Clinical and pathologic information was retrospectively obtained from individual patient charts, medical records, and approved cancer registries. No patient had known distant metastatic disease at cystectomy. The institutional review board approved the study before it began.
Pathologic Examination All radical cystectomy specimens were examined by a dedicated genitourinary pathologist. The 1997 American Joint Committee on Cancer/TNM classification system was used for pathologic staging and the World Health Organization classification was used for pathologic grading. Patients who underwent surgery before 1997 had their pathologic stage updated to reflect the 1997 TNM staging system. The TUR specimens were submitted in their entirety. Multiple sections from the tumor, bladder wall, and mucosa adjacent to, and distant from, the tumor, along with the ureters and regional lymph nodes, were evaluated. In cystectomy specimens in which no residual or minimal tumor was identified, the remaining bladder was extensively sampled.
Follow-up Patients were scheduled to undergo postoperative physical examination and routine serum chemistry studies approximately every 3 months for the first year, semiannually from the second through the fifth year, and annually thereafter. Radiographic evaluation, including intravenous pyelography, ultrasonography, and chest radiography was obtained annually unless clinically indicated. Additional studies, including bone scanning and computed tomography, were performed when clinically indicated. The median duration of follow-up was 41 months (range 1 to 120). For the purposes of this analysis, bladder cancer recurrence was defined as the development of local and/or distant recurrence. Whenever possible, recurrent lesions were biopsied to confirm the presence of bladder cancer.
Statistical Analysis The Fisher exact test and chi-square test were used to evaluate the association between categorical variables. The differences in variables with a continuous distribution between dichotomous categories were assessed using the Mann-Whitney U test. The differences in variables with a continuous distribution across ranked categories were assessed using the Kruskall-Wallis nonparametric analysis of variance. The Kaplan-Meier method was used to estimate the survival distribution, and the differences were assessed with the log-rank statistic. Multivariate survival analysis was performed with the Cox proportional 1092
Table 1. Clinical characteristics of patients, stratified by correlation of clinical to pathologic stage (pathologic stage equal to or greater than clinical stage vs. pathologic downstaging)
Variable Mean age at cystectomy* (yr) Sex Male Female Ethnicity White African American Asian Lymph node status at cystectomy Node negative Node positive Preoperative intravesical chemotherapy No chemotherapy Chemotherapy Postoperative chemotherapy No chemotherapy Chemotherapy Recurrence No Yes
Not downstaged (n ⫽ 87) 66.0 0.429 66 (75.9) 21 (24.1) 0.109 79 (90.8) 5 (5.7) 2 (2.3) 0.063
Downstaged (n ⫽ 25) 66.2 17 (68.0) 8 (32.0) 16 (64.0) 3 (12.0) 2 (8.0)
65 (74.7) 22 (25.3) 0.768
23 (92.0) 2 (8.0)
54 (62.1) 20 (23.0) 0.126
16 (64.0) 5 (20.0)
63 (72.4) 23 (26.4) 0.094 58 (66.7) 29 (33.3)
22 (88.0) 3 (12.0) 21 (84.0) 4 (16.0)
* P ⫽ 0.928.
hazards regression model. The model fit was evaluated by the likelihood ratio test. Statistical significance in this study was set as P ⱕ0.05. All reported P values are two-sided. All analyses were performed using Statistical Analysis Systems software (SAS Institute, Cary NC).
RESULTS The patient demographics are presented in Table 1, and the clinical characteristics of the cohort, stratified by downstaging, are presented in Table 2. Of the 112 patients with muscle-invasive disease on TUR, 25 (22.3%) were downstaged to non-muscle-invasive disease (pT1 or less) at cystectomy. Four patients were downstaged to Stage pT0. Recurrence occurred in 4 (16.0%) of the downstaged patients compared with 29 (33.3%) of the patients who were not downstaged (P ⫽ 0.094). The sites of recurrence included the lung, pelvis, retroperitoneum, and urethra. No patient developed subsequent upper tract TCC. The patients who were downstaged from muscle-invasive to non-muscle-invasive disease had significantly improved recurrence-free survival, as demonstrated by Kaplan-Meier analysis (Fig. 1). Univariate proportional hazards regression analysis demonstrated a statistically significant 75% decrease in the disease recurrence rate associated with tumor downstaging (hazard ratio [HR] 0.25, 95% confidence interval [CI] 0.08 to 0.84, P ⫽ 0.015). Also, a significant difference was found in recurUROLOGY 70 (6), 2007
Table 2. Multivariate Cox proportional hazards analysis recurrence in patients with muscle-invasive bladder cancer documented by TUR who underwent radical cystectomy from 1994 to 2002 Variable Model 1 (n ⫽ 112) Downstage (pT1 or less vs. pT2) Lymph nodes (positive vs. negative) Model 2 (n ⫽ 111)‡ Downstage (pT1 or less vs. pT2) Adjuvant chemotherapy (yes vs. no)
HR
95% CI
P Value*
0.33 2.68
0.10–1.12 1.31–5.47
0.075 0.007
0.29 2.69
0.09–0.97 1.33–5.42
0.044 0.007
†
TUR ⫽ transurethral resection; HR ⫽ hazard ratio; CI ⫽ confidence interval. * Wald test. † Likelihood ratio test of goodness of fit of model, chi-square (2 degrees of freedom) ⫽ 14.10, P ⫽ 0.0009. ‡ Likelihood ratio test of goodness of fit of model, chi-square (2 degrees of freedom) ⫽ 14.14, P ⫽ 0.0008.
0.29, 95% CI 0.09 to 0.97, P ⫽ 0.044; Table 2). However, owing to the high degree of collinearity between lymph node status and adjuvant chemotherapy, both variables could not be included simultaneously in a multivariate model.9 Both models gave similarly good fits to the data, indicated by the near identical values for the likelihood ratio test (Table 2). Other variables, including age, sex, carcinoma in situ at TUR or cystectomy, grade, or intravesical chemotherapy, did not significantly alter these associations.
COMMENT Figure 1. Influence of tumor downstaging after TUR on recurrence-free survival in 112 patients with muscle-invasive TCC who underwent radical cystectomy.
rence-free survival among patients downstaged to Stage pT1 versus less than Stage pT1 (pT0, pTa, pTis). Of 14 patients, 1 (7.1%) was downstaged to less than Stage pT1 recurred compared with 3 (27.3%) of 11 patients downstaged to Stage pT1 (log-rank chi-square ⫽ 4.00, P ⫽ 0.045). The effect of tumor downstaging on BCA recurrence can be confounded by lymph node status. Of the downstaged patients, 8% had positive lymph nodes compared with 25% of patients who were not downstaged (P ⫽ 0.063). Multivariate Cox proportional hazards analysis suggested a reduction in BCA recurrence risk with tumor downstaging after adjusting for lymph node status and chemotherapy (Table 2). A threefold reduction in recurrence risk with tumor downstaging that approached statistical significance was seen when lymph node status was included in the multivariate model (HR 0.33, 95% CI 0.10 to 1.12, P ⫽ 0.075). A trend was found toward adjuvant chemotherapy being preferentially administered to patients without downstaging versus downstaged patients (27% versus 12%, respectively, P ⫽ 0.126). Adjuvant chemotherapy administration correlated highly with positive lymph nodes (P ⬍0.0001). Even when postoperative chemotherapy was included in a multivariate model with downstaging, the magnitude of the risk reduction effect of downstaging was similar to the previous model, in this case achieving statistical significance (HR UROLOGY 70 (6), 2007
Discrepant clinical and pathologic staging has been described in the bladder cancer studies; however, few studies have specifically addressed the clinical outcomes of patients downstaged from muscle-invasive disease. Our results suggest that patients who are downstaged from muscle-invasive disease on TUR to non-muscle-invasive disease in the radical cystectomy specimen have improved recurrence-free survival compared with patients with persistent muscle invasion at cystectomy. Despite resecting all gross cancer by TUR, a high percentage of patients still have pathologic T2 or T3 disease. Even after adjusting for lymph node status or adjuvant chemotherapy, downstaging reduces the risk of recurrence by 67% to 71% relative to patients with persistent muscle-invasive disease in the cystectomy specimen. Although, as expected, the independent prognostic significance of downstaging is attenuated after adjustment for lymph node status, the magnitude of the effect is similar to that in the model adjusted for adjuvant chemotherapy, which did attain conventional significance levels. Furthermore, the HRs for the univariate (HR 0.25) and multivariate (HR ⫽ 0.29)-adjusted effects of downstaging are similar, suggesting that only minor confounding occurred by either lymph node status or chemotherapy. Waehre et al.10 reported a multi-institutional experience of 227 patients with muscle-invasive bladder cancer (Stage cT2-T4a) who underwent radical cystectomy after neoadjuvant radiotherapy. Downstaging to a pathologic T stage that was less than that observed at TUR occurred in 97 patients (42.7%). The 5-year cancer-specific survival rate of those patients who were not downstaged was 46% compared with 73% for patients downstaged to pT0 1093
disease (P ⬍0.0001). Thus, the investigators stated that patients with the stage reduced to pT1 or less had significantly improved survival relative to patients with evidence of muscle-invasive disease on pathologic analysis of the cystectomy specimen. An important caveat to this comparison is that the survival advantage of downstaging in this study was substantially intertwined with the effects of neoadjuvant treatment, because all patients underwent preoperative radiotherapy. It is conceivable that this could account, at least in part, for the somewhat high incidence of Stage pT0 disease. Furthermore, stage reduction was more common in the 30% of patients who received neoadjuvant chemotherapy (53% versus 40%). Other reports have also suggested a survival advantage of downstaging to pT0 by TUR.11–14 Conversely, Thrasher et al.15 found no survival advantage for patients with bladder cancer who were downstaged to Stage pT0 at radical cystectomy.15 In their population of patients undergoing cystectomy from 1969 to 1990, survival was determined by the presenting clinical stage. The 5-year cancer-specific survival of patients with clinical Stage T2 disease on TUR, who were downstaged to pT0 at radical cystectomy, was equivalent to that of the patients who had Stage pT2 on the final pathologic analysis (approximately 68%). A small number of patients received neoadjuvant radiotherapy, but this did not significantly alter their stated conclusions. These data are somewhat historical, and other potential confounders, such as lymph node status or adjuvant therapy, were not adjusted for in their analysis. Nevertheless, their data would seem to support the concept that survival after radical cystectomy is a function of occult metastatic disease already present at surgery. A similar correlation between prognosis and clinical stage was demonstrated in a recent study by Hassan et al.,16 who investigated 50 patients with pathologic stage carcinoma in situ-only on the final cystectomy specimen. The factor with the greatest influence on prognosis was clinical staging before cystectomy. The 22 patients with a TUR diagnosis of muscle invasion had a significantly greater rate of metastatic disease compared with patients without evidence of muscle invasion on TUR (22.7% versus 3.6%), despite complete resection of invasive disease. The data from Herr5,12 on the outcomes of patients with muscle-invasive bladder cancer (Stage cT2) treated only by TUR present a unique window into the natural history of bladder cancer relevant to the present discussion. In this series, the 10-year disease-specific survival rate after TUR only was 76%, similar to the 10-year disease-specific survival rate of 71% for patients who had undergone radical cystectomy as primary treatment (P ⫽ 0.3). Importantly, the disease-specific survival rate at 10 years for patients with Stage cT0 disease on restaging TUR was 82% compared with 57% for patients with clinical Stage T1 disease on restaging TUR (P ⫽ 0.003). Solsona et al.6 presented a similar series of 133 patients 1094
treated by radical TUR alone, with a 5 and 10-year cause-specific survival rate of 80.5% and 74.5%, respectively. These data suggest that select patients might benefit from TUR of their primary tumor and that downstaging by TUR can offer a survival benefit. The present study had the general limitations inherent to any retrospective series. One might logically submit that smaller bladder tumors are easier to resect completely by TUR and are, therefore, more likely to be downstaged. Large bladder tumors have been shown to be associated with worse clinical outcomes, independent of stage. Our study might have selected for patients with a smaller initial tumor burden. An additional limitation of the present study was that, inherent to the tertiary referral nature of our series, the index tumors were not routinely evaluated and resected by the operating surgeon. It is possible that unmeasured variations in the factors, including multifocality, morphology (sessile versus papillary), and primary tumor size, could have affected the composition of the study population. The relative uncertainty of this circumstance was mitigated in that each patient underwent a staging endoscopic evaluation by the operating surgeon before cystectomy, at which time no grossly evident residual tumor was noted. Patients with endoscopic or radiographic evidence of gross residual tumor were excluded from this analysis, because it was thought that they would not have had the opportunity to be downstaged. Additionally, it is possible that variations in the degree or extent of muscle invasion not captured in our data could have further refined the stratification of patients with clinical Stage T2 bladder cancer. Ultimately, a routine protocol of restaging TUR, as has been described elsewhere,5 might have more thoroughly stratified the risk groups among our patient cohort. Our results need to be validated by larger studies with longer follow-up, in which the relationship of tumor downstaging to disease recurrence, as well as overall and cancer-specific survival, can be further elucidated.
CONCLUSIONS Patients who are downstaged from muscle-invasive TCC on TUR to non-muscle-invasive TCC on radical cystectomy might have a reduced risk of disease recurrence compared with patients with persistent muscle-invasive disease in the radical cystectomy specimen. Longer follow-up and validation from larger series are needed to better characterize this issue and ascertain whether analogous relationships with improved postcystectomy overall or cancer-specific survival exist for patients with muscle-invasive tumors amenable to stage reduction by TUR. An enhanced understanding of this aspect of the natural history of bladder cancer will better inform our interpretation of the results from the variety of bladder-sparing protocols in evolution at some centers. UROLOGY 70 (6), 2007
References 1. Jewett HJ, and Strong GH: Infiltrating carcinoma of the bladder: relation of depth of penetration of the bladder wall to incidence of local extension and metastases. J Urol 55: 366 –370, 1946. 2. Herr HW: Uncertainty, stage and outcome of invasive bladder cancer. J Urol 152: 401– 402, 1994. 3. Prout GR Jr: Bladder carcinoma and a TNM system of classification. J Urol 117: 583–590, 1977. 4. Jewett HJ, and Strong GH: Infiltrating carcinoma of the bladder: relation of depth of penetration of the bladder wall to incidence of local extension and metastases. J Urol 55: 366 –370, 1946. 5. Herr HW: Transurethral resection of muscle-invasive bladder cancer: 10-year outcome. J Clin Oncol 19: 89 –93, 2001. 6. Solsona E, Iborra I, Ricos JV, et al: Feasibility of transurethral resection for muscle infiltrating carcinoma of the bladder: longterm followup of a prospective study. J Urol 159: 95–99, 1998. 7. Kaufman DS, Shipley WU, Griffin PP, et al: Selective bladder preservation by combination treatment of invasive bladder cancer. N Engl J Med 329: 1377–1382, 1993. 8. Shipley WU, Kaufman DS, Tester WJ, et al: Overview of bladder cancer trials in the Radiation Therapy Oncology Group. Cancer 97: 2115–2119, 2003. 9. McGee D, Reed D, and Yano K: The results of logistic analyses when the variables are highly correlated: an empirical example
UROLOGY 70 (6), 2007
10.
11.
12. 13.
14.
15.
16.
using diet and CHD incidence. J Chronic Dis 37: 713–719, 1984. Waehre H, Ous S, Klevmark B, et al: A bladder cancer multiinstitutional experience with total cystectomy for muscle-invasive bladder cancer. Cancer 72: 3044 –3051, 1993. Shipley WU, Kaufman DS, and Heney NM: Can chemo-radiotherapy plus transurethral tumor resection make cystectomy unnecessary for invasive bladder cancer? Oncology (Williston Park) 4: 25–34, 39, 1990. Herr HW: Staging invasive bladder tumors. J Surg Oncol 51: 217–220, 1992. Lee SE, Jeong IG, Ku JH, et al: Impact of transurethral resection of bladder tumor: analysis of cystectomy specimens to evaluate for residual tumor. Urology 63: 873– 877, 2004. Shipley WU, Kaufman DS, and Heney NM: Can chemo-radiotherapy plus transurethral tumor resection make cystectomy unnecessary for invasive bladder cancer? Oncology (Huntingt) 4: 25–34, 39, 1990. Thrasher JB, Frazier HA, Robertson JE, et al: Does of stage pT0 cystectomy specimen confer a survival advantage in patients with minimally invasive bladder cancer? J Urol 152: 393–396, 1994. Hassan JM, Cookson MS, Smith JA Jr, et al: Outcomes in patients with pathological carcinoma in situ only disease at radical cystectomy. J Urol 172: 882– 884, 2004.
1095