Guideline for the Management of Nonmuscle Invasive Bladder Cancer (Stages Ta, T1, and Tis): 2007 Update

You have accessJournal of UrologyAdult urology1 Dec 2007Guideline for the Management of Nonmuscle Invasive Bladder Cancer (Stages Ta, T1, and Tis): 2007 Update M. Craig Hall, Sam S. Chang, Guido Dalbagni, Raj Som Pruthi, John Derek Seigne, Eila Curlee Skinner, J. Stuart Wolf, and Paul F. Schellhammer M. Craig HallM. Craig Hall , Sam S. ChangSam S. Chang , Guido DalbagniGuido Dalbagni , Raj Som PruthiRaj Som Pruthi , John Derek SeigneJohn Derek Seigne , Eila Curlee SkinnerEila Curlee Skinner , J. Stuart WolfJ. Stuart Wolf , and Paul F. SchellhammerPaul F. Schellhammer View All Author Informationhttps://doi.org/10.1016/j.juro.2007.09.003AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookTwitterLinked InEmail Introduction More than 60,000 new cases of bladder cancer are diagnosed each year in the US accounting for approximately 13,000 deaths annually.1 When initially diagnosed, most bladder cancers are nonmuscle invasive (also referred to as "superficial") – i.e., either noninvasive and confined to the mucosa or invading the lamina propria but not yet invading the detrusor muscle. In 1999, the AUA published a report on the Management of Non-Muscle-Invasive Bladder Cancer (Stages Ta, T1, and Tis) (AUA Guideline) produced by the AUA's Bladder Cancer Clinical Guideline Panel.2 The report provided an evidence-based guideline for the patient with nonmuscle invasive bladder cancer and included management standards, guidelines, and options based on the strength of evidence and expected amount of variation in patient preferences. Since 1999 the field of nonmuscle invasive bladder cancer has changed substantially with regard to the understanding of the molecular biology and clinical behavior of this heterogeneous disease. In addition, the growing body and quality of clinical research methodologies have improved during this period. The more recent publication of randomized controlled trials, the gold standard of treatment evaluation, has allowed the evaluation and comparison of various treatment modalities and has further verified the heterogeneous nature of this disease process. For these reasons, the AUA's PGC elected to update the initial report by appointing a panel to develop a new guideline for the management of nonmuscle invasive bladder cancer founded on evidence-based outcomes in the literature as well as expert opinion. Only topics with sufficient evidence on which to base conclusions have been addressed in this guideline. Index patients defined by the Panel will assist in determining appropriate therapy for different risk categories and are described later in this paper. The European Association of Urology also has supported recent efforts in the development of risk stratification schemes and has included them in Oosterlinck et al's Guidelines for TaT1 (Non-muscle Invasive) Bladder Cancer3 which are similar but not identical to the updated 2007 AUA Guideline. Whereas the complete AUA Guideline for the Management of Nonmuscle Invasive Bladder Cancer: (Stages Ta, T1, and Tis): 2007 Update which includes current information on epidemiology, risk stratification, tumor markers, modes of therapy, and bladder cancer characteristics is available online (http://www.auanet.org) and may answer many questions, this manuscript will focus on the actual treatment algorithm for the clinician and patient based on currently available modalities (table 1). Table 1. Current treatment alternatives Treatment Indication(s) TURBT • Any suspected urothelial carcinoma; can be sole treatment but only in nonmuscle invasive urothelial carcinoma4 Laser ablation therapy • Treatment of select lower- and upper-tract cancer5,6 • Treatment of low-grade papillary tumors • Not appropriate for new lesions prior to tumor staging/grading Conservative management (office fulguration or cystoscopic surveillance) • Low risk and recurrent nonmuscle invasive papillary bladder tumors • Well-documented history of low-grade Ta tumors7–9 Intravesical chemotherapy and immunotherapy (details provided below) • Perioperatively or postoperatively in an adjuvant fashion • To prevent recurrence following TURBT10–18 • Adjunctive therapy in Tis where diffuse tumor prevents complete resection19–22 Intravesical therapies Mechanism of action Immunomodulatory agents BCG • Inflammatory host response; release of cytokines • May be combined with interferons23–27 Interferons • Lymphocyte activation; cytokine release; phagocyte stimulation • Antiproliferative actions • Antiangiogenic19,23 Chemotherapeutic agents Thiotepa • Alkylating agent; cross-links nucleic acids28 Mitomycin C • Antibiotic; inhibits DNA synthesis10–12 Doxorubicin, epirubicin, valrubicin • Intercalating agents; inhibits DNA synthesis29–31 Gemcitabine • Deoxycytidine analog; inhibits DNA synthesis32–36 Methodology The methodology of this guideline update was similar to that used in the previous guideline. The intention was to determine the impact of the various available treatments on the outcomes of importance to patients. The efficacy outcomes examined were recurrence of bladder tumors and progression in stage or to cystectomy. The Panel also attempted to estimate the occurrence of side effects and complications of treatments and focused on post-TURBT treatments. It was assumed that all patients had TURBT eradication of all visible tumors. The Panel examined the efficacy of alternative follow-up treatments including repeat TURBT, intravesical immuno- and chemotherapies and photodynamic therapy. The impact of tumor stage, grade, multiplicity, and recurrence status on outcomes was also considered. Excluded were treatments that were not generally available in the US and were not expected to be approved for general use by the time the guideline was published. The Panel also decided not to update outcomes for thiotepa and doxorubicin, treatments included in the previous guideline but deemed less effective than other agents by the previous panel, and as a result were not included in their analysis. Literature Search and Data Extraction The review of the evidence began with a literature search and data extraction. Articles included were identified on four MEDLINE searches beginning in October 2004 and concluding in February 2006, and supplemented with existing meta-analyses. Articles published between January 1, 1998 and December 31, 2005 were included in the analysis. The searches were limited to human subjects, English language, publication date from 1998 (the cutoff from the previous guideline) and contained the MeSH heading "bladder neoplasms." A total of 5,020 citations and abstracts were reviewed for relevance by the Panel chair and vice chair. In total, data from 322 articles were extracted by the Panel members. Inconsistencies in data recording were reconciled, extraction errors were corrected, and some articles were excluded by the Panel. A total of 158 articles were accepted for data analysis. Evidence Combination The analytic goals were expanded from the previous guideline. In addition to meta-analyzing the randomized controlled trials to determine if there were significant differences among the treatments, the Panel also decided to develop outcomes tables which actually provided estimates of outcomes for the different treatment modalities. For this guideline, the Panel elected to use the confidence-profile method,37,38 which provides methods for analyzing data from studies that are not randomized controlled trials. Three different meta-analyses of the efficacy data were performed. 1 Meta-analysis of the comparable randomized controlled trials to determine the differences between pairs of available treatments. This analysis provides estimates of the absolute differences. 2 Meta-analysis of the individual arms of the randomized controlled trials to combine all the data from such trials for each treatment. This "single-arm" analysis provides an estimate of the actual rate of occurrence of each outcome. 3 Meta-analysis of the individual arms from all studies regardless of study design. For complications and side effects, only this method was used. Hierarchical meta-analysis was used throughout due to the lack of homogeneity among the studies. The outcomes analyzed for efficacy included recurrence and progression. For recurrence, only probability of recurrence (percentage of patients with recurrence) provided sufficient data for analysis. Time between recurrences, frequency of recurrences, and number of individual tumor recurrences could not be meta-analyzed. Similarly, the Panel decided that only probability of overall progression could be analyzed. Overall progression included progression in stage or to cystectomy. Grade progression could not be analyzed. The way in which complications were grouped varied from study to study. Different names were also used for similar complications. The Panel attempted to group complications by including similar complications in a single group. Only studies that specifically reported data concerning occurrences of complications were included in the analysis. The Panel attempted to evaluate outcomes based on a variety of patient characteristics including stage, grade, tumor multiplicity, and recurrence. However, in most cases, the outcomes data were not fully or consistently stratified by these conditions. Ultimately, the Panel elected to include all studies and analyze the data based on high and low risk as well as an analysis including all studies. Low risk was defined as Grade 1 or low grade. High risk included groups that had no Grade 1 (low grade) patients or were entirely Tis and/or T1. Although the Panel originally considered a wide variety of treatments, limited data were available for many of those of interest. The Panel also considered maintenance therapy versus induction only. A wide variety of induction and maintenance schedules have been used and reported in the literature. Because the issues concerning the comparison of BCG with mitomycin C and of maintenance with induction were so important, the Panel elected to incorporate all randomized controlled trials of these agents in the analyses including those from the original guideline. After the evidence was combined and outcome tables were produced, the Panel met to review the results and identify anomalies. From the evidence in the outcome tables and expert opinion, the Panel drafted the treatment guideline. As in the previous guideline, the guideline statements were graded with respect to the degree of flexibility in their application. Although the terminology has changed slightly, the current three levels are essentially the same as in the previous guideline. A "standard" has the least flexibility as a treatment policy; a "recommendation" has significantly more flexibility; and an "option" is even more flexible. These three levels of flexibility are defined as follows: 1 Standard: A guideline statement is a standard if: (1) the health outcomes of the alternative interventions are sufficiently well known to permit meaningful decisions and (2) there is virtual unanimity about which intervention is preferred. 2 Recommendation: A guideline statement is a recommendation if: (1) the health outcomes of the alternative intervention are sufficiently well known to permit meaningful decisions, and (2) an appreciable but not unanimous majority agrees on which intervention is preferred. 3 Option: A guideline statement is an option if: (1) the health outcomes of the interventions are not sufficiently well known to permit meaningful decisions, or (2) preferences are unknown or equivocal. Options can exist because of insufficient evidence or because patient preferences are divided and may/should influence choices made. The draft was sent to 88 peer reviewers; the Panel revised the document based on the comments received from 38. The guideline was submitted for approval to the PGC of the AUA and then to the Board of Directors for final approval. Results of the Outcomes Analyses Efficacy Outcomes: Recurrence and Progression The primary treatment for nonmuscle invasive transitional cell cancer is TURBT. Although TURBT is an essential diagnostic tool and an effective therapy, 45% (CI: 37–54) of patients will have tumor recurrence within 12 months of TURBT alone. Tumor recurrence is hypothesized to be due to a combination of missed tumors, an incomplete initial resection, implantation of tumor cells shed at the time of the resection, and/or a de novo tumor occurrence from 'at-risk' urothelium. A second, less frequent, but more consequential outcome is the 3% to 15% risk of tumor progression to muscle invasive and/or metastatic bladder cancer. Progression can also include the chance of cancer-related death. In this analysis, we have examined both the early and late risk of recurrence as well as the risk of progression following intravesical chemo- or immunotherapy. Every attempt was made to report outcomes stratified by risk of recurrence based on known clinicopathological parameters (e.g., grade, stage, size of tumor, and multiplicity). Assessing the effect of intravesical therapy on overall and bladder cancer-specific survival was hindered and limited by the small number of events reported in the literature. Every attempt was made to distinguish outcomes between different types of nonmuscle invasive cancer (e.g., Ta versus T1, or versus Tis). Despite these efforts, our ability to do so was limited because many of the studies analyzed did not provide outcomes data stratified by various risk factors such as stage and grade. Furthermore, we were limited in assessing progression and survival because of a paucity of data for these endpoints, which was most likely a consequence of the comparatively short length of follow-up provided in most randomized controlled trials compared to the long, natural history of nonmuscle invasive bladder cancer. Efficacy Outcomes: Results of the Analysis For the purpose of the Guideline, the Panel chose to focus the analysis on a series of clinically relevant questions concerning specific primary therapies and maintenance therapies. Summary data relevant to these questions are presented in tables 2 to 3c and are discussed in the text that follows. The efficacy outcomes were analyzed in three ways: meta-analysis of comparable randomized controlled trials (table 2), analysis of comparable arms from randomized controlled trials (table 3), and analysis of comparable arms from all studies. Table 2. Pairwise meta-analysis of all randomized controlled trials comparing different treatments—recurrence, progression, disease-free survival and overall survival in all risk levels⁎ Recurrence Difference Progression Difference Disease Specific Survival Difference Overall Survival Difference Treatment 1 Treatment 2 S/P Est Rate % CI (2.5, 97.5)% S/P Est Rate % CI (2.5, 97.5)% S/P Est Rate % CI (2.5, 97.5)% S/P Est Rate % CI (2.5, 97.5)% TURBT + MMC Single Dose TURBT Alone 2/427 −17 (−28, −8) 1/306 3 (−4, 10) TURBT + MMC Induction TURBT Alone 2/321 −3 (−16, 10) 1/43 8 (−12, 28) TURBT + MMC + Maintenance TURBT Alone 6/559 −18 (−30, −6) 2/126 4 (−26, 32) TURBT + BCG Induction TURBT Alone 1/47 −24 (−47, −3) TURBT + BCG + Maintenance TURBT Alone 5/629 −31 (−42, −18) 3/1,680 8 (0, 15) TURBT + BCG + Maintenance TURBT + BCG Induction 4/645 −14 (−26, −1) 2/510 −14 (−37, 10) 1/126 −1 (−9,6) 2/510 1 (−7,8) TURBT + MMC + Maintenance TURBT + BCG Induction 3/1,066 −7 (−15, −0) 1/387 0 (−5, 5) TURBT + BCG + Maintenance TURBT + MMC + Maintenance 2/594 −17 (−26, −7) 1/380 −5 (−11, 1) 2/494 3 (−3,8) 1/244 −3 (−11,7) Cross study comparison analysis combined with randomized controlled trials Difference Treatment 1 Treatment 2 S/P Est Rate % CI (2.5, 97.5)% TURBT + BCG Induction TURBT + MMC Induction 3/368 −21 (−48,9) TURBT + BCG + Maintenance TURBT + MMC + Maintenance 9/1,554 −4 (−31,24) S/P, number of studies/number of pts. ⁎ Shading indicates a statistically significant difference between treatments. Negative value indicates benefit favoring treatment in column 1 (treatment 1). Blank cells indicate absence of data. Table 3a. Recurrence by treatment—randomized controlled trials: all risk groups⁎ 1-Yr 2-Yr 3-Yr 4-Yr 5-Yr Treatment G/P Est Rate % CI (2.5, 97.5)% G/P Est Rate % CI (2.5, 97.5)% G/P Est Rate % CI (2.5, 97.5)% G/P Est Rate % CI (2.5, 97.5)% G/P Est Rate % CI (2.5, 97.5)% TURBT Alone 18/1,057 45 (37,54) 21/802 56 (50,63) 12/586 56 (47,65) 4/299 44 (30,59) 5/413 41 (29,55) TURBT + MMC Single Dose 2/206 13 (1,40) 2/206 28 (11,50) 2/206 31 (14,53) 2/206 40 (28,54) 2/206 46 (37,55) TURBT + MMC Induction 3/310 40 (30,52) 3/310 53 (41,63) 3/310 59 (47,70) 3/310 66 (56,74) TURBT + MMC + Maintenance 8/890 32 (23,41) 8/890 42 (35,51) 7/853 44 (33,56) 6/728 43 (31,56) 4/574 49 (39,59) TURBT + BCG Induction 9/787 26 (17,36) 9/787 38 (30,47) 6/627 37 (28,46) 5/520 39 (29,51) 6/581 38 (27,50) TURBT + BCG + Maintenance 11/1,426 25 (17,36) 11/1,426 33 (25,41) 10/1,398 37 (30,45) 7/694 39 (31,47) 7/1,169 42 (33,51) Overall/Unspecified G/P Est Rate % CI (2.5, 97.5)% TURBT Alone 19/1,019 55 (49, 61) TURBT + MMC Single Dose TURBT + MMC Induction 1/92 46 (36, 56) TURBT + MMC + Maintenance 11/1,086 32 (25, 39) TURBT + BCG Induction 10/816 36 (27, 45) TURBT + BCG + Maintenance 19/1,427 29 (23, 36) ⁎ Please see text for data qualification. Table 3b. Recurrence by treatment—randomized controlled trials: high-risk groups⁎ 1-Yr 2-Yr 3-Yr 4-Yr 5-Yr Treatment G/P Est Rate % CI (2.5, 97.5)% G/P Est Rate % CI (2.5, 97.5)% G/P Est Rate % CI (2.5, 97.5)% G/P Est Rate % CI (2.5, 97.5)% G/P Est Rate % CI (2.5, 97.5)% TURBT Alone 2/132 60 (50,70) 2/132 76 (66,84) 2/124 77 (56,91) TURBT + MMC Induction TURBT + MMC + Maintenance 2/79 26 (16,39) 2/79 54 (29,78) 2/79 64 (40,83) 2/79 70 (50,85) 1/16 62 (38,83) TURBT + BCG Induction 4/197 15 (8,23) 4/197 25 (18,34) 4/197 29 (21,37) 3/136 33 (24,43) 3/170 32 (21,44) TURBT + BCG + Maintenance 4/322 16 (6,32) 4/322 24 (15,35) 3/294 28 (15,43) 2/229 40 (29,52) 3/294 34 (18,54) Overall/Unspecified G/P Est Rate % CI (2.5, 97.5)% TURBT Alone 11/441 60 (51,68) TURBT + MMC Induction TURBT + MMC + Maintenance 3/91 44 (32,57) TURBT + BCG Induction 4/186 34 (24,45) TURBT + BCG + Maintenance 5/573 27 (16,40) ⁎ High risk included groups that had no Grade 1 pts or were entirely carcinoma in situ and/or T1. Please see text for data qualification. Table 3c. Progression, disease specific survival, overall survival by treatment: randomized controlled trials⁎ Progression Overall Progression High Risk† Disease Specific Survival Overall Survival Overall/Unspecified Overall/Unspecified Overall/Unspecified Overall/Unspecified Treatment G/P Est Rate % CI (2.5, 97.5)% G/P Est Rate % CI (2.5, 97.5)% G/P Est Rate % CI (2.5, 97.5)% G/P Est Rate % CI (2.5, 97.5)% TURBT Alone 17/917 12 (9, 17) 2/48 17 (3,46) 4/383 94 (89,97) 5/505 84 (73,92) TURBT + MMC Single Dose 1/57 2 (0, 8) TURBT + MMC Induction 3/343 6 (2, 12) 1/92 91 (84,96) TURBT + MMC + Maintenance 9/928 11 (8, 16) 1/63 10 (4,19) 7/740 93 (91,95) 7/914 81 (71,89) TURBT + BCG Induction 8/546 10 (7, 13) 4/260 14 (9,19) 3/325 89 (79,96) 3/335 73 (56,87) TURBT + BCG + Maintenance 17/1,701 9 (7, 12) 5/341 14 (8,22) 10/1,442 95 (92,97) 13/1,557 84 (78,89) ⁎ Please see text for data qualification. † Subgroup of first column. High risk included groups that had no Grade 1 pts or were entirely carcinoma in situ and/or T1. The result of the meta-analysis of comparable studies is the central result (best estimate) of the absolute difference between the two treatments. This number is the median of the posterior distribution. Negative indicate that the first treatment has occurrences of the outcome compared to the treatment in the For the first of 2 that the absolute difference in the rate of recurrence with mitomycin C is less than with TURBT alone. the recurrence rate with TURBT was the rate with mitomycin C is expected to be The CI the for that the CI not include the estimate is statistically significantly different from at From is that both mitomycin C and BCG therapy in combination with TURBT the probability of disease recurrence in the time of the studies. treatment to progression or BCG with maintenance therapy combined with TURBT, a statistically significant in overall progression in our studies comparing an induction series of mitomycin C with an induction series of BCG were on review of the literature. In addition to the randomized controlled the Panel tables the actual probability of recurrence, progression, and survival for each of the treatments. The tables that include the results of the analysis of all randomized controlled arms are in tables to 3c. Since the estimates presented in 3 are based on a number of studies and a variety of studies than the estimates in the results are not The results in 2 are the most meaningful between treatments. The results in 3 are most randomized controlled trials are not or to provide estimates by time Clinical The Panel examined clinical questions to provide for clinical The following are 1 is the effect on recurrence of TURBT combined with a intravesical of mitomycin C versus TURBT In the initial randomized studies of thiotepa as an adjuvant the in tumor recurrence was in the studies a early This that a treatment was a alternative to randomized studies comparing different agents with TURBT the efficacy of a The Panel focused their analysis on a single of mitomycin C as is the most used in the trials were identified as for and included a combined patients. In the by et of the with TURBT had recurrences at compared to 45% of those a of 40 of mitomycin in 40 of et a statistically significant in early recurrence of to two in patients received the single of mitomycin However, a significant difference in recurrence was not at the final follow-up of When these trials were combined for meta-analysis by the a statistically significant in median recurrence rate was of the trials of a of mitomycin C our meta-analysis an effect on progression. This was due to the of with a small number of progression In the literature and the current meta-analysis the use of a single of a (e.g., mitomycin in the to the risk of recurrence in patients following an TURBT of a nonmuscle invasive bladder 2 is the effect on recurrence and progression of TURBT with an induction of intravesical chemotherapy versus TURBT combined with an induction of intravesical BCG intravesical immunotherapy has been compared to different agents in of recurrence The results have been for the include varied patient treatment different types and of tumors and/or actual efficacy differences between the treatments or individual studies have a efficacy of with maintenance therapy in patients with could not this to reported results from single trials, groups have attempted to the data with the use of meta-analyses. even the meta-analyses did not a In and reported that BCG not a recurrence intravesical However, in a of randomized trials comparing BCG maintenance therapy in trials and induction only in three to either mitomycin epirubicin, or mitomycin C and an induction maintenance in trials and induction only in et were to that the intravesical BCG induction maintenance significantly treatment in patients with The benefit of BCG was in comparison with mitomycin C and to mitomycin C only in the trials where maintenance BCG was In our meta-analysis we are to answer the intravesical chemotherapy is the most effective to recurrence and progression of bladder The meta-analysis by the 1999 AUA Bladder Guideline Panel the of the immunotherapy compared to thiotepa and However, there are randomized trials that different intravesical there are trials that mitomycin C to other intravesical chemotherapy current meta-analysis that intravesical chemotherapy agents all have in bladder cancer recurrence. were to identify trials that that these agents were to In we identified three trials with a total of patients (table combined for that mitomycin C with maintenance was to BCG induction by that mitomycin C may be the most of currently available intravesical However, this of mitomycin C to be by clinical trials. this no single chemotherapy can be considered can an induction of BCG maintenance be considered to an induction of mitomycin C Similarly, the results are not with regard to cancer progression. et could not a statistically significant for BCG versus mitomycin C for progression in Tis difference favoring a including et did a statistically significant in risk of progression in trials comparing BCG immunotherapy versus mitomycin C , two that may have the results BCG to be in the analysis included studies BCG not a