Reported use of intravesical therapy for non-muscle-invasive bladder cancer (NMIBC): results from the Bladder Cancer Advocacy Network (BCAN) survey
Matthew E. Nielsen, MD, UNC Urologic Oncology, Linebeger Comprehensive Cancer Center, 2017 Physicians Office Building, Campus Box 7235, Chapel Hill, NC, USA, 27599-7235. e-mail: email@example.com
Study Type – Therapy (patterns of practice)
Level of Evidence 2b
What's known on the subject? and What does the study add?
Claims-based analyses suggest unexplained and potentially problematic variation in treatment intensity adherence to guidelines-recommended care in NMIBC. Previous physician surveys prior to the contemporary Clinical Practice Guidelines (CPGs) reported associations between variation in NMIBC care and practice type, years in practice, and other physician-related factors.
In the largest physician survey addressing the management of NMIBC, and the first to examine these questions after the promulgation of contemporary CPGs, US urologists report grade-specific utilization consistent with CPG recommendations, at rates higher than suggested by recent claims-based analyses. As with prior studies, practice type and years in practice were significantly associated with variation in practices. Further research is needed to reconcile these findings with administrative claims data.
- • To determine self-reported practices of use of intravesical chemo- and immunotherapy for patients with non-muscle-invasive bladder cancer (NMIBC)
- • To evaluate the extent to which respondent characteristics were associated with any observed variation. Guidelines recommend intravesical therapy (IVT) in the management of NMIBC, but recent claims-based analyses suggest exceedingly low rates of use of some of these therapies.
MATERIALS AND METHODS
- • An electronic survey was developed by members of the Bladder Cancer Advocacy Network (BCAN) to elicit self-reported use of management strategies for NMIBC.
- • The survey was circulated to urologists via the American Urological Association, Society for Urologic Oncology and Large Urology Group Practice Association distribution lists.
- • In all, 512 respondents completed the survey.
- • In all, 63% reported routine perioperative mitomycin-c (MMC) after transurethral resection of bladder tumour (80% academic vs 54% private practice, P < 0.001).
- • Whereas 5% of respondents reported routine induction therapy with all new low-grade (LG) diagnoses, 99% reported routinely doing so in new high-grade (HG) cases; most commonly with single-agent bacille Calmette-Guérin (BCG) (94% vs 9% BCG/interferon and 5% MMC).
- • Reported induction therapy was higher in the setting of high-volume (77%) or frequently recurrent (44%) LG disease.
- • In all, 89% reported routinely using maintenance therapy for HG vs 29% for LG disease.
- • Routine biopsy after BCG, even with normal cystoscopy, was endorsed by 28% (39% academic vs 22% private practice, P < 0.001).
- • Urologists report grade-specific use of IVT for NMIBC, at rates higher than suggested in some claims-based analyses.
- • Further study is needed to corroborate these self-reported patterns of care with lower rates of use suggested by claims-based analyses.
non-muscle-invasive bladder cancer
Clinical Practice Guideline
Bladder Cancer Advocacy Network
Society for Urologic Oncology
Large Urology Group Practice Association
transurethral resection of bladder tumour
Intravesical therapy (IVT) represents a cornerstone of organ-sparing management of non-muscle-invasive bladder cancer (NMIBC), with the benefit of improved recurrence-free survival reported in multiple randomised clinical trials [1–3]. These trials have resulted in the inclusion of IVT in guidelines for management of certain patients with NMIBC both in the USA and Europe [4,5]. The AUA published its first Clinical Practice Guideline (CPG) for NMIBC in 1999 . The evidence base supporting best practices developed over the ensuing years, and a second Panel updated this report in 2007 . The updated CPG included more specific recommendations for scenarios in which IVT should be considered, on the basis of grade, stage, and recurrence history.
Despite guideline recommendations, several administrative claims-based studies have identified unexplained and potentially problematic variation in treatment intensity and adherence to guidelines-recommended care [7–9]. Previous physician surveys have reported associations between variation in NMIBC practices and number of years in practice, practice type and other physician-related factors, but these data are from the era before the dissemination of the modern CPG [10,11]. In this context, we conducted a survey to elicit contemporary data on practices related to the management of NMIBC to describe patterns of care and further illuminate specific areas where consensus is lacking. The principal objectives of the present study were to evaluate the extent to which self-reported practices related to IVT use varied in different NMIBC clinical scenarios, as well as the extent to which physician characteristics (academic vs private practice, number of years in practice, geographic region) were associated with any variation observed.
MATERIALS AND METHODS
An electronic multiple-choice survey was designed to elicit self-reported practices for the management of NMIBC (Appendix). The survey was pilot tested amongst a focus group of urologist members of the Bladder Cancer Think Tank and revised for clarity. Respondents did not have an opportunity to enter free text responses. The final version of the survey was sent via electronic distribution with Survey Monkey, an internet survey program, in three consecutive mailings to practicing urologists on the electronic mailing lists of the Society for Urologic Oncology (SUO), AUA and Large Urology Group Practice Association (LUGPA) via the administrative staff of each organisation. Reminder e-mails were sent to non-respondents 1 week after each initial mailing, and survey responses were collected in an electronic database. Voluntary response to the questionnaire was considered implied consent. The survey was exempt from Institutional Review Board review as all results were anonymous. We used the chi-squared test to compare proportions, and P < 0.05 was considered to indicate statistical significance.
The initial distribution to the SUO mailing list (n= 590) yielded 103 responses (17.5% response rate). The AUA mailing list (n= 5700) yielded 383 responses (6.7%) and the final distribution to the LUGPA mailing list (n= 1100) yielded 26 responses (2.4%), for a total of 512 responses (overall response rate 6.9%). Demographic characteristics of the respondents, overall and broken down by mailing, are summarised in Table 1.
Table 1. Characteristics of survey respondents
| N (%)||512 (100)||103 (20)||383 (75)||26 (5)|
|%:|| || || || |
| Practice setting:|| || || || |
| Private practice||65.||23.7||75.3||91.3|
| Non-university academic||6.0||7.2||5.5||8.7|
| University-based academic||28.2||69.1||19.2||0|
| Size of group of urologists:|| || || || |
| Years in practice:|| || || || |
| Geographic region:|| || || || |
| North Central||14.3||14.7||14.0||17.4|
| South Central||13.3||13.7||12.1||30.4|
| West Coast||17.6||11.6||19.8||8.7|
| Size of community:|| || || || |
| Big Metro, >1 million||34.6||55.7||27.1||65.2|
| Metro, 250 000–1 million||29.0||32.0||29.2||13.0|
| Urban, 50 000–250 000||22.8||11.3||26.8||8.7|
| Less Urban, 25 000–50 000||8.4||1.0||10.1||13.0|
| Rural, <25 000||5.2||0||6.8||0|
In all, 316 (62.9%) respondents reported routinely using a single-dose perioperative instillation of mitomycin-C (MMC) after initial transurethral resection of bladder tumour (TURBT). The rates of reported use of this strategy significantly varied by practice type (80.1% academic vs 53.9% private practice, P < 0.001) and years in practice (76.1% <10 years vs 65.8% 10–20 years vs 51.6% >20 years, P < 0.001). There was slight variation in this practice across the five geographic regions (58.9–66.7%), although this did not achieve statistical significance (P= 0.839). The vast majority of respondents reported administering MMC in this setting immediately after TURBT (89.9%), with 5.6% selecting ‘within 24 h’ and 4.5% selecting ‘other’.
Adjuvant induction IVT practices for patients with low-grade (LG) NMIBC are given in Table 2. As compared with new cases of high-grade (HG) NMIBC, where 99.2% of respondents reported routinely using induction IVT, only 4.7% routinely use induction IVT in new cases of LG disease (P < 0.001). However, 74.8% of respondents reported routine use in new LG cases with numerous or large tumours. The number of respondents reporting routine IVT use in recurrent LG cases was significantly higher than those reporting routine use in new LG cases (42.3% vs 4.7%, respectively, P < 0.001). There was significant variation among these practices by practice type and years in practice, with private practice urologists and those in practice >10 years endorsing relatively higher rates of routine induction IVT in recurrent LG disease (P < 0.05, Table 2) but no significant variation by geographic region (data not shown).
Table 2. Reported use of IVT in low-grade NMIBC
| N (%):||490 (96)||319 (65.8)||166 (34.2)|| ||134 (27.8)||155 (32.1)||193 (40.0)|| |
| Induction IVT at new diagnosis:|| || || || || || || || |
| Never||86 (17.3)||46 (14.4)||36 (21.8)||0.057||34 (25.4)||26 (16.8)||21 (10.9)||0.01|
| Routine||24 (4.8)||16 (5.0)||6 (3.6)||7 (5.2)||3 (1.9)||12 (6.3)|
| Only in high volume||371 (74.7)||250 (78.4)||115 (69.7)||89 (66.4)||123 (79.4)||151 (78.7)|
| Other||16 (3.2)||7 (2.2)||8 (4.9)||4 (3.0)||3 (1.9)||8 (4.2)|
| Induction IVT at recurrence:|| || || || || || || || |
| Never||16 (3.2)||7 (2.2)||9 (5.5)||0.005||10 (7.5)||3 (2.0)||3 (1.6)||<0.001|
| Routine||210 (42.3)||149 (47.0)||52 (31.5)||48 (35.8)||53 (34.4)||100 (52.4)|
| Only in high volume||247 (49.8)||149 (47.0)||95 (57.6)||67 (50.0)||90 (58.4)||84 (44.0)|
| Other||23 (4.6)||12 (3.8)||9 (5.5)||9 (6.7)||8 (5.2)||4 (2.1)|
| Routine maintenance IVT:|| || || || || || || || |
| Yes||345 (70.4)||218 (69.2)||119 (73.0)||0.388||99 (75.6)||117 (76.0)||121 (63.7)||0.017|
| No||145 (29.6)||97 (30.8)||44 (27.0)||32 (24.4)||37 (24.0)||69 (36.3)|
| IVT agent of choice:|| || || || || || || || |
| MMC||89 (18.4)||36 (11.4)||50 (32.1)||<0.001||35 (27.8)||24 (15.7)||27 (14.3)||0.005|
| BCG||365 (75.3)||263 (83.5)||96 (61.5)||83 (65.9)||119 (77.8)||153 (81.0)|
| BCG or MMC||18 (3.7)||11 (3.5)||4 (2.6)||7 (5.6)||6 (3.9)||2 (1.1)|
| Other||13 (2.7)||5 (1.6)||6 (3.9)||1 (0.8)||4 (2.6)||7 (3.7)|
Practices related to IVT for patients with high-grade HG NMIBC are given in Table 3. There was near-consensus (99.2%) for routine induction IVT after a new diagnosis of HG NMIBC, and an overwhelming majority (89.7%) endorsing routine maintenance IVT in this setting, with slightly higher rates of the latter reported by urologists in private practice (92.6% vs 85.2%, P= 0.011, Table 3). The reported usage of maintenance IVT in HG disease was significantly greater than that reported in the setting of LG disease (89.7% vs 29.6%, P < 0.001). Whereas 18.4% of overall respondents endorsed MMC as their induction IVT agent of choice in the setting of LG disease (Table 2), BCG alone or combined with other agents was endorsed in the context of HG disease by >90% of respondents, with only 2.7% endorsing MMC in this setting (Table 3). Reduced-strength BCG was rarely endorsed for routine use (3.1% overall), but more typically in the setting of poor patient tolerance of full strength instillations (80.9% overall). Nearly one-third (28%) of respondents endorsed routine biopsy after BCG, even in the absence of endoscopic or cytological abnormality, with higher rates of this practice reported among academic practitioners (38.8% vs 21.5% private practice, P < 0.001).
Table 3. Reported use of IVT in high-grade NMIBC
| N (%):|| ||315 (65.6)||165 (34.4)|| ||133 (27.9)||152 (31.9)||192 (40.2)|| |
| Induction IVT at new diagnosis|| || || || || || || || |
| Routine||490 (99.2)||312 (99.0)||165 (100)||0.209||133 (100)||152 (100)||188 (97.9)||0.050|
| Not routine||4 (0.8)||3 (1.0)||0||0||0||4 (2.1)|
| Routine maintenance IVT:|| || || || || || || || |
| Yes||433 (89.7)||286 (92.6)||138 (85.2)||0.011||115 (89.2)||138 (92.0)||167 (88.4)||0.529|
| No||50 (10.4)||23 (7.4)||24 (14.8)||14 (10.9)||12 (8.0)||22 (11.6)|
| Routine post-BCG biopsy:|| || || || || || || || |
| Yes||137 (28.0)||67 (21.5)||64 (38.8)||<0.001||43 (32.3)||30 (19.5)||57 (30.7)||0.024|
| No||352 (72.0)||244 (78.5)||101 (61.2)||90 (67.7)||124 (80.5)||129 (69.3)|
| IVT agent of choice:|| || || || || || || || |
| MMC||13 (2.7)||7 (2.2)||4 (2.5)||0.955||3 (2.3)||0||9 (4.8)||0.002|
| BCG||427 (87.3)||276 (88.5)||146 (89.6)||119 (89.5)||141 (92.2)||160 (85.6)|
| BCG + interferon||18 (3.7)||12 (3.9)||6 (3.7)||1 (0.8)||4 (2.6)||13 (7.0)|
| Multiple agents||31 (6.3)||17 (5.5)||7 (4.3)||10 (7.5)||8 (5.2)||5 (2.7)|
| Use of reduced strength BCG:|| || || || || || || || |
| Never/other||41 (8.0)||21 (6.6)||6 (3.6)||0.320||5 (3.7)||7 (4.5)||15 (7.8)||0.606|
| Routine||16 (3.1)||8 (2.5)||8 (4.8)||6 (4.5)||3 (1.9)||7 (3.6)|
| For maintenance||41 (8.0)||26 (8.2)||14 (8.4)||11 (8.2)||13 (8.4)||15 (7.8)|
| Only for poor tolerance||414 (80.9)||264 (82.8)||138 (83.1)||112 (83.6)||132 (85.2)||156 (80.8)|
In a large national sample of practicing American urologists, representing a breadth of experience in practice, geographic regions, practice size and setting, we present self-reported patterns of care related to IVT for patients with NMIBC. The patterns of care in this domain of NMIBC are relevant not only from the clinical perspective, but also in health economic terms, as an increased intensity of IVT over time was observed as an important driver of differential treatment intensity and increased average per capita expenditures in a recent study of Medicare patients with NMIBC .
Responses to the survey suggest grade-specific patterns of IVT use, with strong consensus for routine induction and maintenance therapy, predominantly with BCG, in the setting of HG disease (Table 3) and greater variation in the agent of choice and contexts associated with reported use for LG disease (Table 2). Respondents were generally less likely to use induction IVT in LG disease, except in recurrent cases or cases with numerous and/or large volume tumours, complying with the recommendations of the CPG . Consistent with previous physician surveys of NMIBC practices, there were significant associations between practice setting and number of years in practice and self-reported patterns of care (Tables 2,3) [10,11]. Although there was a modest geographic variation for a number of survey items, this variation did not reach statistical significance. This is, to our knowledge, the largest physician survey addressing the management of NMIBC, and the first to examine these questions after the publication of the contemporary AUA CPG for the management of NMIBC . This is relevant to the extent that several topics were the subject of more specific recommendations in the updated CPG as compared with the original iteration from 1999.
A discussion of the utility of single-dose perioperative chemotherapy at the time of TURBT was one important update in the 2007 AUA CPG. The 2007 CPG designated the administration of an initial single dose of immediate postoperative intravesical chemotherapy as an ‘option and not a standard by the Panel because of potential cost issues, uncertainty of pathology, side-effects and patient preference. In addition, the use of immediate intravesical chemotherapy would not be beneficial for bladder tumors that are most likely muscle invasive’. A recent claims-based analysis of patients treated from 1997 to 2004 suggested strikingly low rates of use of this evidence-based practice, with only 0.33% of patients identified as eligible having received this treatment . In the same study, the authors used modelling techniques to estimate that broader use of this treatment could result in a yearly national savings of $19.8–24.8 million/year . While these findings remain to be validated by other investigators or examined in years subsequent to the publication of the updated AUA guidelines, they raise the question of whether the use of single-dose perioperative chemotherapy may be a target for formal quality improvement efforts in NMIBC. The results of the present survey suggest a much higher rate of routine use of single-dose perioperative chemotherapy than was reflected in the claims-based study, leading to several questions, including whether: (a) the actual rates of use in contemporary practice have increased in response to the updated practice guidelines, (b) respondents to this survey are representative of current practitioners on the whole, and (c) the extent to which self-report reliably reflects actual practice. Regardless, important future directions include replication of the prior work in independent (and potentially more contemporary) administrative claims data sets, and, if low rates of use appear persistent and valid, consideration of potential interventions to increase real-world use of this important piece of the NMIBC armamentarium.
The updated CPG also more specifically addressed the use of maintenance IVT after induction. The results of the present survey suggest higher rates of use of this approach in HG vs LG disease, and relatively higher rates of use among providers in private practice (Tables 2,3). The dissemination of this particular strategy has yet to be evaluated in administrative claims, and so remains another potential topic for future investigators. Although not specifically addressed in the CPG, we also inquired about the use of routine bladder biopsy, even in the presence of normal cystoscopy and cytology, after a course of BCG for HG disease. This practice was endorsed by nearly one-third of respondents (28%), with higher rates reported by physicians in academic practices (Table 3), despite published studies calling into question the utility of this approach [13–15].
Although the present results provide novel descriptive data to expand our understanding of contemporary patterns of care in NMIBC, the limitations of this study warrant acknowledgement. The present results reflect physician responses to hypothetical scenarios. As with any survey, these results have limitations in the extent to which they reflect actual practice patterns. One specific limitation is the multiple-choice format of the survey; respondents did not have the opportunity to enter free text or alternative responses to the vignettes. It is possible that a degree of discrepancy between these results and claims data may be case-by-case judgments to modify the ultimate decision executed (reflected in claims-based analyses), despite general attitudes and intentions (reflected in self-reported practice in a multiple choice survey). The gap between intention and execution may represent the important space within which to target quality improvement efforts in this setting, as some of this gap may represent clinically sound, patient-centred decision making, whereas some may represent potentially actionable opportunities for intervention to facilitate desirable behaviour change. Future work should include prospective data collection in actual clinical practice, detailing the extent to which clinical contraindications (i.e. concern for perforation of the bladder, significant LUTS, previous intolerance of therapy) influence the actual usage rates of different guidelines-recommended interventions in practice, as these data were not captured in the present study and are not available in administrative claims datasets. These data could be informative to practicing clinicians and policymakers, building on other successful examples of surgeon-led efforts to reduce variability in processes and outcomes of care . Also, we recognise that many providers have encountered shortages of MMC in recent years; the present survey did not capture specific information on the extent to which this and/or other logistical issues interfere with providers' ability to deliver the guidelines-recommended care and may represent another source of discordance between self-reported intentions and actual care delivered.
While we had >500 respondents, larger than previous studies addressing these questions, the overall response rate was relatively low. We did not have the ability to offer financial compensation for the time devoted to the study; future studies may benefit from such incentives. There is probably substantial overlap between the organisations' electronic mailing lists, so the true denominator of unique individuals solicited for the survey is probably smaller than the summation of each mailing list, which we chose to report as the most conservative estimate. Nevertheless, to address concerns about the representativeness of the sample and the external validity of these data, a comparison of the respondents' demographic profile to data obtained from the AUA Member Profile  is presented in Table 4. As seen in the Table, although the proportion of academic urologists was slightly larger in the present survey sample, the sample was otherwise reasonably representative of the overall population of practicing urologists.
Table 4. Characteristics of survey respondents vs AUA member profile data
|Practice type:|| || || |
| Private practice||65.8||Other than full-time academic*||82.7|
| Academic practice||34.2||Full-time academic||17.3|
|Size of group:|| || || |
| ≥2 Urologists||84.6||Other||83.5|
|Years in practice:|| ||Age of urologist, years|| |
| ≤10||27.9|| Young Urologists*||21.3|
| || || ≤45||24|
| 11–20||32.2|| 46–54||14|
| 21–30||26.4|| 55–64||14|
| >30||13.7|| ≥65||19|
|Geographic region:|| || || |
| North Central||14.3||North Central||16.5|
| South Central||13.3||South Central||14.7|
| West Coast||17.6||Western||17.6|
|Size of community:|| || || |
| Big Metro, >1 million||34.6||Urban||57.2|
| Metro, 250 000–1 million||29.0||Suburbs||20.8|
| Urban, 50 000–250 000||22.8||Small Community||16.8|
| Less Urban, 25 000–50 000||8.4|| || |
| Rural, <25 000||5.2||Rural||5.1|
In conclusion, self-reported practices of the use of IVT in NMIBC are generally aligned with the 2007 AUA Guidelines for the Management of Non-muscle-invasive Bladder Cancer. Respondents reported much higher usage of single-dose perioperative chemotherapy than has been reported in claims-based studies from earlier eras, and grade-specific patterns of use for induction and maintenance IVT. As in previous physician surveys inquiring about patterns of care in this area, practice type (academic vs private practice) and number of years in practice were associated with significant variation in reported practice patterns for NMIBC.
Future work is needed to corroborate these self-reported practices and actual clinical practice, and ultimately, the association of adherence to these practices with clinical outcomes in real-world patient populations. These data reinforce the need for prospective registry studies to fill remaining gaps in the evidence and provide a foundation for quality improvement projects in this area.
The authors wish to acknowledge the Bladder Cancer Advocacy Network, the Bladder Cancer Think Tank, The Society for Urologic Oncology, the American Urological Association and the Large Urology Group Practice Association for their assistance in disseminating the electronic survey, and the survey respondents for their time.
CONFLICT OF INTEREST
Neal Shore is an investigator for Amgen, Astellas, Centocor, Dendreon, Ferring, Medivation, Nymox and Sanofi-Aventis. Yair Lotan is an investigator for Alere, Abbott and Danone.