Managing non-muscle-invasive bladder cancer

National Collaborating Centre for Cancer (UK)

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Cancer (UK). Bladder Cancer: Diagnosis and Management. London: National Institute for Health and Care Excellence (NICE); 2015 Feb. (NICE Guideline, No. 2.)

Bladder Cancer: Diagnosis and Management.

Show details

Contents

< Prev Next >

4Managing non-muscle-invasive bladder cancer

Most people with bladder cancer do not have cancer in the muscle wall of the bladder (muscle invasive bladder cancer), but have cancer that involves the surface lining of the bladder (urothelium), or the connective tissue layer (lamina propria) that connects the surface lining to the main muscle coat (non-muscle invasive bladder cancer). These cancers are designated stages pTa and pT1 respectively, and they are also classified according to whether they are regarded as not aggressive, moderately aggressive, or aggressive, grades 1, 2 and 3 respectively.

The majority of people with bladder cancer will have pTa cancers of either grade 1 or 2. These cancers may return on the lining of the bladder (recurrence), or worsen, meaning return and extend to involve the main muscle coat of the bladder or beyond (progression).

Recurrence of non-muscle invasive bladder cancer is generally not life-threatening. However people with NMIBC will need cystoscopy under anaesthesia to remove the recurrence, with the time in hospital and recovery time, and the possibility of additional treatment and follow up. Recurrence is important to the NHS because of the costs and capacity needed to treat it.

Progression, in contrast, means that the risk to life has risen and that further investigations and more invasive treatment options will be considered. If progression of the cancer to involvement of the muscle wall of the bladder occurs, 20 - 25 out of 100 such people will also have spread into their lymph glands, and their chance of cure falls sharply.

People with non-muscle invasive bladder cancer may have different experiences following their inital transurethral resection. The information people receive about what was seen and done at the operation may vary in quality, quanitity and how it is communicated and this will impact on the patients understanding of their condition and ability to make informed decisions.

There may be some form of imaging, and there will be further cystoscopy follow-up, which may be infrequent for many people. For some people there will be repeat resection and discussion of treatment options that include intravesical therapy (chemotherapy or BCG) and radical cystectomy. The subsequent pathways for people with non muscle invasive bladder cancer may therefore be very different.

The impact of this on the people involved will differ, and their concerns may be very different, but include such questions as:

Is this cancer life-threatening ?
Will I lose my bladder ?
For how long will I need to be treated?
Is recurrence a sign that the cancer has spread ?
For how long will I need to be followed-up and will my appointments be forgotten?
Will I become incontinent ?
Will my sexual function be lost ?

Some of the important issues in non-muscle invasive bladder cancer are, therefore:

Prognostic factors
Staging, including transurethral surgery and imaging
The risk of recurrence and progression, and its classification
Adjuvant treatment, including intravesical therapy and radical cystectomy
Follow-up

There is uncertainty and variation in practice in all of these areas at present.

4.1. Risk Stratification

4.1.1. Prognostic markers in non-muscle-invasive bladder cancer

Assessment of the risk of recurrence and progression is critical to choosing the optimal package of care. Prognostic markers include clinical factors such as history of recurrence and pathological characteristics including:

stage
grade
cancer size
the presence of carcinoma in situ
number of cancers
variant pathology
lymphovascular invasion.

There is no widely agreed and implemented method of assessing risk of recurrence and progression using prognostic markers.

Clinical question: Which factors determine risk of relapse and progression in newly diagnosed non-muscle invasive bladder cancer (e.g. histological grading of bladder cancer)? In addition to the factors specified in the EORTC risk tables, do urothelial cancer variants, differentiation of urothelial cancer and lymphovascular invasion predict recurrence and progression after treatment?

Clinical evidence (see also full evidence review)

Study quality and results

The NICE prognostic studies methodological checklist was used to assess the quality of the prognostic studies. All studies were assessed as being of high quality as they included the population of interest, measured the outcome adequately, and used appropriate statistical analysis. However, validation studies of the EORTC risk tables were limited by heterogeneous patient populations and treatments received and by low numbers of progression events. Studies exploring the prognostic factors of lymphovascular invasion, urothelial cancer variants and urothelial cancer differentiation were limited by small sample sizes and few patients with the factor under investigation.

Evidence statements

The EORTC risk tables (Sylvester et al., 2006) have been validated in several studies, which report that the tables successfully stratify patients into risk groups for recurrence and progression, but generally overestimate the risk of recurrence in all risk groups and the risk of progression in high risk groups (Fernandez-Gomez et al., 2011; Seo et al., 2010; Altieri et al., 2012; Hernandez et al., 2011; van Rhijn et al., 2010; Xu et al., 2013; Lammers et al., 2014).

There is some low quality evidence to suggest that the presence of lymphovascular invasion increases the risk of recurrence, progression and disease-specific survival. However, this is based on low numbers of patients with evidence of lymphovascular invasion.

One study (Brimo et al., 2013) of 86 patients reported that adverse histological variants were significantly associated with progression and recurrence on univariate analysis but were insignificant on multivariate analysis. Only four tumours were not ‘usual’ TCC. Three had features of micropapillary TCC and one had features of sarcomatoid TCC.

One study (Scosyrev et al., 2009) reported that squamous cell histologic features were associated with overall mortality and disease-specific mortality compared to TCC in patients who did not undergo cystectomy, but was not associated with increased mortality in those who were treated with cystectomy.

One study (Alkibay et al., 2009), reported that progression rates increased in patients with NMIBC and micropapillary pattern (MPP) compared with MPP-negative patients but this difference was not statistically significant (p=0.064). This analysis was based on only six patients with T1 bladder cancer and MPP, and 125 TaT1 MPP-negative patients.

Cost-effectiveness evidence

A literature review of published cost-effectiveness analyses did not identify any relevant papers for this topic. Whilst there were potential cost implications of making recommendations in this area, other questions in the guideline were agreed as higher priorities for economic evaluation. Consequently no further economic modelling was undertaken for this question.

Recommendations	Record the size and number of tumours during TURBT. Ensure that for people with non-muscle-invasive bladder cancer all of the following are recorded and used to guide discussions, both within multidisciplinary team meetings and with the person, about prognosis and treatment options: recurrence history size and number of cancers histological type, grade, stage and presence (or absence) of flat urothelium, detrusor muscle (muscularis propria), and carcinoma in situ the risk category of the person's cancer (see section 4.1.2) predicted risk of recurrence and progression, estimated using a risk prediction tool.
Relative value placed on the outcomes considered	The GDG considered the cancer-related outcomes of recurrence, disease progression, disease-specific survival and overall survival as important outcomes. Recurrence and progression lead to further treatment and potentially a worse prognosis. Survival is important for patients. Overall survival was specified as an outcome in the PICO but was not reported in the evidence.
Quality of the evidence	The quality of the evidence was assessed as high using the NICE methodology checklist for prognostic studies. However, the reviewer highlighted some issues with the evidence. Most notably, the EORTC risk calculator is limited in that it overestimates recurrence in patients treated with BCG. Validation studies of the EORTC risk tables were limited by heterogeneous patient populations and treatments received and by low numbers of progression events. Studies exploring the prognostic factors of lymphovascular invasion, urothelial cancer variants and urothelial cancer differentiation were limited by small sample sizes, with few patients with the factor under investigation. However the GDG considered that the high quality evidence of the EORTC risk factors and validation studies strengthened the case for making the recommendations.
Trade-off between clinical benefits and harms	The GDG considered that the main benefits of the recommendations are better informed decision making by the person and the implementation of existing guidelines and improvements in the quality of data collected to guide future clinical management of non-muscle invasive bladder cancer. The GDG identified no potential harms from the recommendations made.
Trade-off between net health benefits and resource use	No health economic evidence was identified and no economic model was developed for this topic. The GDG considered the potential costs and savings associated with the recommendation made. The potential costs include more staff time at MDT meetings to consider the clinical and histological prognostic factors listed in the recommendation. This was balanced against the potential savings accrued from better treatment and less need for subsequent salvage treatment. Savings will also be made from the avoidance of unnecessary treatment and follow-up.
Other considerations	No equalities issues were identified. The GDG considered that a moderate change in practice may be required to implement the recommendations. EORTC and RCPath datasets should be used routinely in local and specialist bladder MDTs, so the recommendations reflect best practice.

4.1.2. Definitions of risk

There is no widely accepted classification of risk in non-muscle invasive bladder cancer. In order to make clear recommendations for management, the GDG developed a consensus classification based on evidence reviewed and clinical opinion. For the purposes of this guideline the following definitions apply:

For this purpose, we refer only to non-muscle invasive urothelial cancer, not muscle invasive cancer or non-urothelial cancers.

Low risk NMIBC

Any of these:

Solitary pTaG1 <3cm
Solitary pTaG2 (low grade) <3cm
Any PUNLMP (papillary urothelial neoplasm of low malignant potential)

Intermediate risk NMIBC

Any tumour that is not low risk or high risk including the following:

Solitary pTaG1 ≥3cm
Multifocal pTaG1
Solitary pTaG2 (low grade) ≥3cm
Multifocal pTaG2 (low grade)
pTaG2 (high grade)
Any pTaG2 (grade not further specified)
Any low risk recurring within 12 months from last tumour occurrence

High risk NMIBC

Any of these:

pTaG3
pT1G2
pT1G3
pTis (Cis)
aggressive variants of urothelial carcinoma, for example micropapillary or nested variants

4.2. Managing non-muscle invasive bladder cancer

4.2.1. Intravesical therapy

Intravesical therapy involves the instillation into the bladder of either a chemotherapy drug (in the NHS this is typically Mytomycin C) or BCG. Intravesical chemotherapy is most often given as a single dose directly following transurethral resection of a cancer to try to prevent recurrence of non-muscle invasive bladder cancer. It can also be used on an outpatient basis as a course of treatment to try to reduce recurrence in people who have had a significant rate of recurrence.

Intravesical BCG is an immunotherapy used to treat intermediate and high-risk non-muscle invasive bladder cancer. Each treatment includes the instillation of live BCG bacteria, of which various strains are known to exist, into the bladder. Intravesical BCG is given on an outpatient basis as a course of treatment, to try to prevent recurrence and also progression in people judged to have a significant risk of these problems. In the most commonly used regimen it is given as a course of six instillations (induction BCG) followed by sets of 3 instillations over a period of up to 3 years (maintenance BCG).

Some people relapse after BCG, their management is discussed in section 4.3.2. The management of BCG-related toxicity is discussed in section 4.4.

There is wide variation in practice regarding the use of intravesical chemotherapy and intravesical BCG in the NHS at present.

Clinical question: What are the most effective adjuvant intravesical therapy (chemotherapy or immunotherapy) regimens for low-, intermediate- and high-risk non-muscle invasive bladder cancer?

Clinical evidence (see also full evidence review)

Systematic reviews and randomised trial evidence was appraised for this review. The evidence is summarised in tables 28 to 54.

Table 28

GRADE evidence profile: What are the most effective adjuvant intravesical therapy (chemotherapy or immunotherapy) regimens for low-risk/intermediate and high-risk non-muscle-invasive bladder cancer? Comparison: TUR + BCG versus TUR alone

Table 29

Table 30

Table 31

Table 32

Table 33

Table 34

Table 35

Table 36

Table 37

Table 38

Table 39

Table 40

Table 41

Table 42

Table 43

Table 44

Table 45

Table 46

Table 47

Table 48

Table 49

Table 50

Table 51

Table 52

Table 53

Table 54

Evidence statements

TUR + BCG versus TUR alone

Moderate quality evidence from a meta-analysis (Shelley et al., 2000) of 585 medium to high risk patients from six randomised trials produced an overall hazard ratio (HR) for recurrence-free survival of 0.44 (95% CI 0.34 to 0.56), indicating a 56% reduction in the risk of tumour recurrence for TUR+BCG compared to TUR alone. The main toxicities associated with BCG are urinary frequency (71%), cystitis (67%), haematuria (23%), and fever (25%). No BCG sepsis or deaths are reported.

TUR + BCG versus TUR + other treatment (chemotherapy or immunotherapy) or TUR alone

Moderate quality evidence from a meta-analysis (Pan et al., 2014) of 48 RCTs and observational cohort studies (9,482 patients) reported a pooled random effects OR for recurrence of 0.59 (95% CI 0.49 to 0.71) for TUR + BCG compared to those treated with resection alone or TUR plus another treatment other than BCG, with significant heterogeneity across studies (p<0.01). Evidence from an earlier meta-analysis (Han & Pan, 2006) suggested that the effect of BCG is less conclusive when induction BCG only is given compared to control groups (RR 0.99, 95% CI 0.77 to 1.28). In the maintenance BCG subgroup the combined random effect RR is 0.65 (95% CI 0.48 to 0.88), suggesting that maintenance BCG reduces the risk of recurrence by 35%. Moderate quality evidence from a meta-analysis of 13 trials or controlled studies comparing maintenance BCG versus no maintenance BCG for T1G3 bladder cancer, reports that overall 41% of the maintenance BCG group recurred compared to 45% in the control group (RR 0.73, 95% CI 0.61, 0.88) (Pan et al., 2008).

High quality evidence from one meta-analysis of 24 randomised trials with 4863 patients, suggests that the risk of progression was 27% lower for patients treated with BCG compared to those treated with either resection alone or TUR plus another treatment other than BCG (HR 0.73, 95% CI 0.60 to 0.88) (Sylvester et al., 2002). No reduction in the risk of progression was seen in the four trials where maintenance BCG was not used (HR 1.28, 95% CI 0.82 to 1.98). There is uncertainty of any difference for overall survival (HR 0.89, 95% CI 0.75 to 1.06) and disease-specific survival (HR 0.81, 95% CI 0.57 to 1.13) between those treated with BCG and those in the control groups. Moderate quality evidence from the two meta-analyses by Han & Pan (2006) and Pan et al. (2008) both report that drug-related and systemic toxicities are significantly more frequent in the BCG groups than chemotherapy or immunotherapy groups.

TUR + chemotherapy versus TUR alone

One systematic review and meta-analysis of 11 studies and 3,703 patients with primary bladder cancer provides a Peto Odds Ratio (pOR) of 0.56 (95% CI 0.48 to 0.65) for one-year recurrence in favour of adjuvant intravesical chemotherapy compared to TUR alone (Huncharek et al., 2000). However, significant statistical heterogeneity is reported and sensitivity analyses were conducted. The data were stratified by duration of treatment, which indicates that short-term therapy (≤2 months duration) reduces recurrence at one-year (pOR 0.70, 95% CI 0.55 to 0.90) and two-years (pOR 0.68, 95% CI 0.54 to 0.85) by approximately 30%, as compared to TUR alone (moderate quality evidence). The pOR for five trials where patients received two years of chemotherapy is 0.27 (95% CI 0.19 to 0.39), indicating a 73% reduction in the risk of recurrence at two-years for those treated with chemotherapy.

Moderate quality evidence from one meta-analysis of eight studies and 1,609 patients with recurrent bladder cancer provides a pooled OR for one-year recurrence of 0.62 (95% CI 0.51 to 0.76), in favour of chemotherapy over TUR alone, with no evidence of statistical heterogeneity (Huncharek et al., 2001). For the two- and three-year recurrence rates, significant statistical heterogeneity was reported, which was not accounted for by treatment duration. Therefore, moderate quality evidence is provided from the data when stratified into drug type (adriamycin versus other drugs). The OR for two-year recurrence of studies using adriamycin is 0.57 (95% CI 0.43 to 0.75), with no significant heterogeneity, indicating that drug type was a major contributor to outcome heterogeneity. Drugs other than adriamycin showed a reduction in two-year recurrence of 73% (versus 43% for adriamycin) with an OR of 0.27 (95% CI 0.19 to 0.37).

Another systematic review and meta-analysis provides moderate quality evidence from six randomised trials, which suggests there is uncertainty about the effect of intravesical chemotherapy on progression (HR 1.19, 95% CI 0.97 to 1.47), overall survival (HR 1.1, 95% CI 0.95 to 1.27), and disease-specific survival (HR 1.1, 95% CI not reported but effect size was non-significant), compared to TUR alone (Pawinski et al., 1996).

TUR + one post-operative instillation of chemotherapy versus TUR alone

Low to moderate quality evidence is reported from a systematic review and meta-analysis of 18 trials comparing one post-operative dose of chemotherapy with TUR alone (Abern et al., 2013). 36.6% (577/1576) of those in the TUR + chemotherapy group experienced a recurrence compared with 50.4% (769/1527) of those treated with TUR alone (RR 0.67, 95% CI 0.56 to 0.79), with significant statistical heterogeneity. This corresponds to a number needed to treat of 7 patients to avoid one recurrence. Gemcitabine and interferon α-2b does not show a benefit on recurrence, whereas the other chemotherapy agents do. The pooled RR for mitomycin C and epirubicin is 0.71 (95% CI 0.64 to 0.78), in favour of chemotherapy, with no clear dose-response relationship. Funnel plots suggest publication bias with small trials contributing disproportionately to the protective effect of chemotherapy. Progression and survival are not reported. A meta-analysis (Sylvester et al., 2004) of seven trials (1476 patients) reports mild, transient, irritative bladder symptoms including dysuria, frequency and macroscopic haematuria, in approximately 10% of patients treated with one single post-operative dose of intravesical chemotherapy.

TUR+ single dose epirubicin versus TUR + double dose Epirubicin

Low quality evidence from one randomised trial of 143 patients without CIS suggests no difference in recurrence or progression between patients treated with a single dose of 100mg epirubicin within six hours of TUR and those given a second dose of 100mg epirubicin 12-18 hours after TUR (Turkeri et al., 2010).

Moderate quality evidence from one trial of 270 patients without CIS reports that two instillations of 50mg epirubicin within 24 hours of TUR is associated with longer recurrence-free survival than TUR alone (38 months versus 13 months, p=0.004). Recurrence-free survival with two instillations of lower dose epirubicin (20mg/40ml) is not significantly longer than TUR alone (24 months versus 13 months, p=0.163). There are no significant differences between 2×50mg and 2×20mg epirubicin (p=0.146). Local grade one toxicity was reported in 22.9% of the low dose epirubicin group and 35.6% of high dose epirubicin group (RR 0.63, 95% CI 0.39 to 1.02).

Intravesical Adriamycin versus Epirubicin

Moderate quality evidence is provided by two randomised trials comparing one year treatment with adriamycin with the same schedule of epirubicin (Eto et al., 1994; Shuin et al., 1994). There were no differences in recurrence rate (RR 1.31, 95% CI 0.72 to 2.4) or local toxicities (RR 0.73, 95% CI 0.46 to 1.15) between the two treatment arms.

Adjuvant intravesical BCG versus adjuvant intravesical chemotherapy

One systematic review of nine trials and 2,261 patients (Huncharek & Kupelnick 2003) reports low quality evidence of an overall OR for one-year recurrence of 0.89 (95% CI 0.74 to 1.07), with significant heterogeneity. Heterogeneity persisted despite stratification by chemotherapy drug type. A sensitivity analysis was therefore performed stratifying by previous intravesical chemotherapy. Pooling all studies that enrolled patients with prior chemotherapy (1480 patients) provides moderate quality evidence, with an OR of 0.54 (95% CI 0.43 to 0.69) in favour of BCG. This reflects a 46% reduction in tumour recurrence at one-year among patients treated with BCG versus chemotherapy, and a lack of statistical heterogeneity. Pooling data from two studies which excluded patients previously treated with chemotherapy gives an OR of 1.82 (95% CI 1.37 to 2.41), in favour of chemotherapy. This suggests that amongst patients not previously treated, intravesical chemotherapy (MMC) reduces tumour recurrence by 82% versus BCG. Similar results were found for two-year and three-year recurrence when stratified by previous therapy.

One systematic review of eight randomised trials and 2,427 patients (Huncharek & Kupelnick 2004) randomised to either adjuvant intravesical BCG or chemotherapy provides moderate quality evidence of an OR for progression of 1.24 (95% CI 0.95 to 1.61), in favour of BCG. The confidence intervals include the value of no effect which reflects uncertainty about a difference in progression between the two treatments. The total number of events in each arm is not reported. The pooled OR of the two trials (781 patients) which excluded patients who had previously been treated with intravesical chemotherapy is 0.75 (0.45 to 1.25) in favour of MMC. In trials which included patients previously treated with chemotherapy the OR is 1.49 (1.09 to 2.03) in favour of BCG.

One meta-analysis (Sylvester et al., 2005) of nine randomised trials and 700 patients with CIS provides moderate quality evidence that 34% of complete responders treated with BCG and 50% of complete responders treated with chemotherapy recurred during follow-up (HR 0.47, 95% CI 0.31 to 0.73, in favour of BCG). 47% of patients treated with BCG and 26% treated with chemotherapy had no evidence of disease during follow-up, relating to an absolute difference of 20% and a relative reduction of 59% in the odds of treatment failure on BCG (HR 0.41, 95% CI 0.30 to 0.56). BCG is only superior to MMC in the trials where maintenance BCG was given. Data on progression were less conclusive with a HR of 0.74 (95% CI 0.45 to 1.22). Overall survival is reported in three studies (407 patients). 35.9% of patients treated with chemotherapy and 34.2% treated with BCG therapy died from any cause. Two trials reported disease-specific survival. 13.3% of patients treated with chemotherapy and 10.5% of patients treated with BCG died due to bladder cancer.

BCG versus Mitomycin C (MMC)

Moderate quality evidence is reported from one meta-analysis (Bohle et al., 2003) of 2,749 patients from nine prospective trials and two observational studies. A further trial of 92 patients was indentified and added to the pooled analysis for recurrence (Mangiarotti et al., 2008). The overall RR for recurrence is 0.77 (95% CI 0.63 to 0.95) in favour of BCG over MMC. High quality evidence from a meta-analysis of individual patient data (Malmstrom et al., 2009) including nine trials (2,820 patients) reported that in trials with BCG maintenance, there is a 32% reduction in the risk of recurrence with BCG compared to MMC (HR 0.68, 95% CI 0.58 to 8), whilst there is a 28% risk increase for BCG trials without maintenance (HR 1.28, 95% CI 1.07 to 1.52). Maintenance BCG is more effective than MMC in both patients previously treated and those not previously treated with intravesical chemotherapy.

Moderate quality evidence from one meta-analysis including 1,277 patients (Bohle & Bock 2004) reports no difference between BCG and MMC in terms of disease progression (RR 0.79, 95% CI 0.61 to 1.03). However, BCG does show superiority over MMC in the subgroup of BCG maintenance trials (RR 0.70, 95% CI 0.52 to 0.94). Moderate quality evidence from seven trials (1,880 patients) in the IPD meta-analyses reports that after a median follow-up of 4.8 years, 12% of patients progressed and 24% died (of those 30% died from bladder cancer). There are no significant differences between MMC and BCG for these end-points, even when stratified by BCG maintenance and patient risk groups.

Cystitis was more frequent in the BCG group compared to the MMC group (53.8% vs. 39.2%, p<0.001). Local and systemic toxicities were more frequent in the BCG group, except for allergy and skin reactions which were more common in MMC group. The risk of cystitis was no different between maintenance BCG and no maintenance BCG. No deaths from sepsis were reported in either arm (Bohle et al., 2003).

BCG versus Epirubicin (EPI)

Moderate quality evidence from one meta-analysis of five randomised trials (1,111 patients) (Shang et al., 2011), reports that the risk of recurrence was reduced in patients treated with BCG (35.9%) compared to EPI (51.4%) with a RR of 0.69 (95% CI 0.60 to 0.79), in favour of BCG. Low quality evidence from a subgroup analysis demonstrates no significant difference in recurrence between BCG and EPI in two trials using Pasteur strain BCG (RR 0.78, 95% CI 0.56 to 1.10). Low quality evidence for disease progression demonstrated that there are no significant differences between BCG and EPI (RR 0.78, 95% CI 0.54 to 1.13). No differences are reported for overall mortality (two studies, 769 patients) or disease-specific mortality (two studies, 769 patients). However, overall mortality is less frequent in the TICE BCG group compared to the EPI group in the study by Sylvester et al. (2010) (RR 0.79, 95% CI 0.62 to 0.99). Drug-induced cystitis (54% versus 32%), haematuria (31% versus 16%), and systemic side-effects (35% versus 1%) are significantly more frequent with BCG than EPI. However, there is significant heterogeneity between trials for systemic side-effects due to the frequency of BCG administration across studies. Moderate quality evidence from four randomised trials suggests there are no significant differences for delayed or terminated treatment due to adverse events between BCG and EPI (9% versus 7%) (RR 0.91, 95% CI 0.41 to 2.04).

BCG versus Gemcitabine

One systematic review by Jones et al. (2012) includes three studies comparing Gemcitabine with BCG. Heterogeneity between trials prevented pooling of data. One trial of 80 patients at intermediate risk of recurrence (primary Ta-T1, no CIS) provides low quality evidence that BCG (no maintenance) and Gemcitabine showed similar rates of recurrence (25% vs. 30%) and progression, with significantly more adverse effects with BCG. Moderate quality evidence is provided by one trial of 64 high risk patients, which reports that recurrence rate is higher for Gemcitabine than BCG (53% vs. 28%) and time to recurrence is shorter with Gemcitabine (25.6 months vs. 39.4 months). No patients in either group had disease progression at a mean follow-up of 44 months. Local and systemic toxicity are similar between groups. In this trial, maintenance therapy for non-recurring patients in each group was up to 36 months duration. No evidence about survival is reported.

Maintenance BCG versus induction BCG

Six trials of maintenance versus induction BCG were indentified which vary in the population included and the schedule and duration of maintenance therapy. High quality evidence from five of these trials with 686 patients, reports that 53.9% of patients in the BCG induction arm had a recurrence, compared to 37.6% in the maintenance BCG arm (RR 0.70, 95% CI 0.60 to 0.81). Moderate quality evidence from five trials (737 patients) suggests similar rates of progression (27.6% versus 31.8% for maintenance and induction BCG respectively). However, these data should be interpreted with caution due to the variation in BCG maintenance schedules and the duration of follow-up across studies. There are no differences between groups in terms of overall survival and disease-specific survival. Moderate quality evidence from two trials (126 patients) suggests that dysuria is more frequent in the maintenance arm (88.9% versus 68.3%). Rates of fever/chills are not different between groups (RR 1.47, 95% CI 0.88 to 2.44).

One trial of 53 patients reported moderate quality evidence of no significant changes in quality of life scores (EORTC-QLQ) in either group from induction treatment to 14 months after randomisation (Koga et al., 2010). Very low quality evidence from one observational study of 85 patients reports that overall quality of life was moderate, and more patients rated it as good during maintenance than during induction therapy (Mack et al., 1996).

Dose of BCG

Low dose versus standard dose BCG

Two trials provide moderate quality evidence of no difference in recurrence, progression, overall survival and disease-specific survival between one-third (27mg) dose and full dose (81mg) BCG. One trial (Martinez-Pineiro et al., 2002) included 500 patients (Ta/T1/CIS, G1-G3) and the other trial (Martinez-Pineiro et al., 2005) included 155 patients with T1G3 disease or CIS. Martinez-Pineiro et al. (2002) reports that, in patients with multifocal disease, standard dose BCG is more effective against recurrences and progression than reduced dose BCG. Local toxicity is significantly reduced in the low dose BCG arm (53% versus 67%), and fewer patients have delayed instillations or withdraw from treatment. There are no differences between groups for severe systemic toxicities (3.8% versus 2.7%).

One trial of 80 patients provides low quality evidence of no difference in recurrence, progression or cystitis between patients receiving 81mg BCG versus those receiving 54mg BCG (Yalcinkaya et al., 1998). One trial of 128 patients randomised into three arms, provides low quality evidence of no difference in recurrence rates between 120mg BCG, 80mg BCG and 40mg BCG. No patients had disease progression. Both local toxicity and systemic toxicity were reduced with lower dose of BCG (Agrawal et al., 2007).

Low dose versus very low dose BCG

Moderate quality evidence from one trial of 281 patients (Ojea et al., 2007) suggests that there are no differences in recurrence-free survival between low dose BCG (27mg) and very-low dose BCG (13.5mg) in intermediate risk patients. There are no differences in time to progression and cancer-specific survival between the two BCG treatment groups. Rates of local (65.5% vs. 64.1%) and systemic (11.3% vs. 10.8%) adverse events are also similar between the two groups.

Low dose and standard dose with 1 year or 3 year maintenance

Moderate quality evidence is provided by one trial of 1,355 patients randomised into four trial arms (Oddens et al., 2013). With a median follow-up of 7.1 years, no differences are reported for recurrence, progression, overall survival and toxicity between one-third (27mg) dose and full dose (81mg) BCG. When results are stratified by maintenance and dose, one-third dose BCG with one-year maintenance is suboptimal compared to full-dose BCG with three-year maintenance (HR for disease-free interval 0.75, 95% CI 0.59 to 0.94). In intermediate-risk patients, three years of maintenance is more effective than one year in patients receiving one-third dose (HR 1.35, 95% CI 1.03 to 1.79) but not in patients receiving full-dose (HR 0.88, 95% CI 0.64 to 1.21). In high-risk patients, three years of maintenance is more effective than one year in patients receiving full dose (HR 1.61, 95% CI 1.13 to 2.30) but not in patients receiving one-third dose BCG (HR 1.01, 95% CI 0.69 to 1.47). No significant differences are reported between treatment groups for the time to progression or overall survival.

The schedule and duration of intravesical chemotherapy

One systematic review of 23 randomised trials (Sylvester et al., 2008) which compared intravesical instillations with respect to their number, frequency, timing, duration, dose, or dose intensity concludes that the optimal schedule and duration of intravesical chemotherapy after an immediate instillation remains unknown. In low-risk patients, one immediate instillation of epirubicin may not be less effective than a delayed course of multiple instillations (3 trials, 879 patients). In patients with multiple tumours, one immediate instillation is insufficient treatment. Additional instillations may further reduce the recurrence rate; however, there is no conclusive evidence regarding their optimal duration (3 trials, 598 patients). A short intensive schedule of instillations within the first 3-4 months after an immediate instillation may be as effective as longer-term treatment schedules. Instillations during ≥1 year in intermediate-risk patients seem effective only when an immediate instillation has not been given. Higher drug concentrations and optimization of the drug's concentration in the bladder may provide better results (5 trials, 774 patients).

Chemotherapy + maintenance BCG versus maintenance BCG alone

Low quality evidence is provided by a systematic review of four randomised trials (801 patients) comparing sequential chemotherapy added to maintenance BCG with maintenance BCG alone (Houghton et al., 2012). A further study of 96 patients with CIS which compared MMC and BCG with BCG alone was also identified and added to the meta-analysis (Oosterlinck et al., 2011). The dose and duration of intravesical therapies used and the average length of follow-up varies across trials. Meta-analysis of five trials provides low quality evidence of uncertainty of a difference in recurrence between the combination arms (42.6%) and the BCG-alone arms (46.7%) (RR 0.92, 95% CI 0.79 to 1.08), but significant heterogeneity (p=0.03). Sub-group analyses provides moderate quality evidence that adding chemotherapy to maintenance BCG was associated with lower recurrence than BCG alone for Ta or T1 disease (RR 0.75, 95% CI 0.61 to 0.92), but not for CIS (RR 1.13, 95% CI 0.93 to 1.37).

Meta-analysis of five trials (897 patients) provides low quality evidence of no significant difference in progression between the combination arms (11.1%) and the BCG-alone arms (13%) (RR 0.84, 95% CI 0.59 to 1.20), but significant heterogeneity (p=0.03). Sub-group analyses provide moderate quality evidence that adding chemotherapy to maintenance BCG is associated with lower progression than BCG alone for Ta or T1 disease (RR 0.45, 95% CI 0.25 to 0.81), but not for CIS (RR 1.33, 95% CI 0.83 to 2.13). Three studies report drug-related toxicity, with no differences in cystitis, haematuria or fever between groups. The numbers of adverse events in each arm is not reported.

Cost-effectiveness evidence (see also Appendix A)

Background

Non-muscle invasive bladder cancer (NMIBC) tumours can be surgically removed using transurethral resection of bladder tumour (TURBT). However, these tumours are likely to return on the urothelium. This high risk of recurrence is a problem for patients because it raises the concern that the cancer will progress and so the patient will need to undergo further treatment (either another TURBT or diathermy).

The risk of recurrence can be reduced by the administration of chemotherapy medication into the bladder (intravesical chemotherapy), which can be done immediately, or shortly after TURBT. However, there are disadvantages to using intravesical chemotherapy as it is associated with some side effects and comes at an additional cost.

Aim of analysis

To estimate the cost-effectiveness of a single instillation of intravesical chemotherapy in addition to TURBT in comparison to TURBT alone in patients with NMIBC.

Existing Economic Evidence

A systematic literature review identified one paper related to the decision problem, a cost-utility analysis by Green et al. 2013. In the study, a decision analytic model was utilised to estimate the cost-effectiveness of fulguration compared to TURBTs with and without perioperative intravesical chemotherapy in patients with low risk NMIBC.

The authors concluded that fulguration without perioperative intravesical chemotherapy was the most cost-effective strategy for treating low-risk NMIBC. However, unusually, the authors based this conclusion upon individual cost-effectiveness calculations rather than the standard incremental calculations. When following the more standard cost-effectiveness methodology using incremental cost-effectiveness ratios (ICERs), it appears that perioperative intravesical chemotherapy plus fulguration would be the most cost-effective strategy. This strategy has an ICER of $4,169 per QALY, which is likely to fall below the cost-effectiveness threshold^a. The authors also conducted sensitivity analysis, which showed that the effectiveness of perioperative intravesical chemotherapy and the cost of TURBT were likely to be key drivers of the cost-effectiveness result.

However, Green et al. 2013 can only be deemed partially applicable to the decision problem this guideline seeks to address. The analysis considered the US healthcare system, which differs substantially from the UK system. In addition, the study only partially addressed our decision problem as it only evaluated cost-effectiveness in low risk NMIBC patients, whereas we are interested in all NMIBC risk groups.

Overall, it was considered that the current economic literature was partially useful but further analysis would be required to robustly estimate the cost-effectiveness. It should also be noted that the existing economic literature was useful for informing the development of our own economic model.

De Novo Economic Model

Since the current economic literature didn't adequately address the decision problem, a de novo economic evaluation was undertaken to assess cost-effectiveness. A Markov decision model was developed using Microsoft Excel.

The patient enters the model in a ‘disease free’ state following an initial transurethral resection of the bladder tumour (TURBT) with or without a single instillation of chemotherapy (depending upon modelled treatment arm). At each 3-monthly model cycle the patient may experience a bladder cancer recurrence. If the recurrence is detected, the patient will undergo a further TURBT (or fulguration of the tumour) and return to a disease free state. However, if the recurrence is not detected, then the patient will be at risk of progression and will have to undergo further treatment once this progression is eventually detected (cystectomy and possibly neo-adjuvant chemotherapy). The patient may also die from bladder cancer related mortality after experiencing progression and may die from other cause mortality from any health state.

Estimated total costs and quality adjusted life yefars (QALYs) are collected over the modelled 10 year time horizon for each follow-up strategy. Future costs and benefits were discounted at a rate of 3.5% per year as recommended by NICE.

The risk of recurrence and progression in patients with NMIBC was estimated using risk equations based on an analysis of 2,596 patients from seven EORTC^b trials (Sylvester et al. 2006). Patients are ‘scored’ based on a number of risk factors, such as number of tumours, tumour size, prior recurrence rate, T category, presence of CIS and grade. An individual's one year and five year risks of recurrence and progression can then be estimated based upon these scores.

For the purposes of the economic model, it was necessary to convert these five year and one year risks into 3-monthly risks. The higher risk of recurrence and progression in the first year was captured by calculating separate 3 monthly risks for the first year and subsequent years (based on the one year risk and five year EORTC risks). Furthermore, since the EORTC risk equations consider recurrence and progression independently, it was necessary to link the progression rates to the recurrence rate i.e. estimate the probability of progression given recurrence in each of the risk groups (Table 55).

Table 55

Three monthly recurrence and progression risk applied in the model.

As the modelled time horizon of 10 years exceeds the predicted risk estimates from the EORTC trials (5 years), it was also necessary to make some assumptions about the risk profile of patients in years 5-10. In the base case, it was assumed that the subsequent year rate (i.e. years 2-5) would be maintained in years 6-10 except in the case of low-risk patients in whom it was assumed that risk would be zero after 5 years (reflecting clinical practice of discharging low-risk patients from follow-up after 5 years).

The key effectiveness data utilised in the model is the reduction in recurrence risk associated with a single instillation of intravesical chemotherapy following a TURBT. According to the systematic review of the clinical evidence, the use of a single instillation of intravesical chemotherapy in addition to TURBT has a relative risk of 0.67 in comparison to TURBT alone. This treatment effect was assumed to last for two years reflecting the general consensus around its possible duration. Thereafter, the risk of recurrence was assumed to be equal to that with TURBT only. In addition, the treatment effect is not assumed to affect future recurrences if the patient has a recurrence during the two years after the single chemotherapy instillation.

Note that the single instillation of chemotherapy does not directly reduce the rates of progression. This is in line with the evidence base, which suggests that there is no treatment effect on the rates of progression. However, it should be noted that because of the model structure, a lower rate of recurrences would lead to a lower rate of progression because progression is dependent upon recurrence. Therefore, an indirect treatment effect on progression is essentially included in the model. This assumption is relaxed in a sensitivity analysis where the rates of recurrence and progression are assumed to be independent.

No comparative data on morbidity were identified in the systematic review of the clinical evidence. However a meta-analysis (Sylvester 2004) of seven trials suggested that mild irritative bladder symptoms (including dysuria, frequency and macroscopic haematuria) would occur in approximately 10% of patients treated with a single post-operative dose of intravesical chemotherapy. In addition, allergic skin reactions were reported in 1-3% of patients in two studies.

Since no data were available on morbidity in patients treated with TURBT, it was conservatively assumed that 5% would have irritative bladder symptoms and there would be no skin reactions. The treatment related morbidity rates applied in the model are shown in the table below.

The diagnostic accuracy data for flexible cystoscopy were sourced from the systematic review of the clinical evidence conducted for this guideline, with most data being sourced from a systematic review by Mowatt et al. 2010.

Bladder cancer related mortality rates were estimated using data from a systematic review by Van den Bosch et al. 2011. Using the data in the study, separate three mortality rates were estimated for patients that progressed to muscle invasive disease and those that remained non-muscle invasive following a cystectomy (3.6% and 0.5%, respectively). The lower rate in NMIBC patients reflects an assumption that patients would have to first progress to MIBC before dying of bladder cancer.

Death from other causes was captured using 2009-2011 life tables for England and Wales from the office of national statistics (ONS). These life tables give an estimate of the annual probability of death given a person's age and gender with the model assuming that 50% of patients were female and that the average age was 60 years old. These annual probabilities were converted to three-monthly probabilities for use in the model.

Costs and utilities

Modelled patients accrue costs associated with any treatment, monitoring or management strategy that they are undergoing. The costs considered in the model reflect the perspective of the analysis, thus only costs that are relevant to the UK NHS & PSS were included. These costs include drug costs, treatment costs and any other resource use that may be required (e.g. GP visit). Where possible, all costs were estimated in 2012-13 prices.

The majority of costs were sourced from NHS reference costs 2012/13 by applying tariffs associated with the appropriate HRG code. Drug costs were calculated using dosages from the British National Formulary (BNF), and unit cost information from the electronic market information tool (eMit). Where unit costs for drugs were not available from eMit, prices from the BNF were used. Resource use and cost information were obtained from the Personal Social Services Research Unit (PSSRU) and the advice of the GDG.

The model estimates effectiveness in terms of quality adjusted life years (QALYs). QALYs were estimated by combining the life year estimates with utility values (or QOL weights) associated with being in a particular health state. These utility values were identified through a search of the available literature.

Base case results

The base case results of the analysis are presented in table 56 for patients in each risk category. It can be seen that, in every risk category, a strategy of TURBT plus a single instillation of chemotherapy is more effective than a strategy of TURBT alone.

Table 56

Base case results of the model.

In the case of low and intermediate risk patients, it can also be seen that the addition of a single instillation of chemotherapy is cost saving over the modelled time horizon. This shows that the initial additional costs associated with the single chemotherapy instillation are outweighed by the cost savings associated with a reduction in recurrences (recurrence reductions of 17% and 10% were estimated over the modelled time horizon in the low and intermediate risk groups, respectively). Therefore in low and intermediate risk patients, a single instillation of chemotherapy can be considered dominant i.e. more effective and cost saving.

However, in the case of high risk patients, it can be seen that this is not the case. In high risk patients, the single instillation of chemotherapy is more costly than TURBT alone, suggesting that the potential cost savings are not as large in this group. However, it can also be seen that the addition of a single chemotherapy instillation provides an additional QALY at a cost of £6,432 and thus would be considered cost-effective using the NICE threshold (i.e. <£20,000 per QALY).

Sensitivity analysis

A series of one-way sensitivity analyses were conducted, whereby the value of an input parameter is changed and its effect on the overall outcome is recorded and assessed.

The analyses showed that the conclusion of the model is insensitive to changes in the input parameters over plausible ranges i.e. TURBT plus a single instillation of chemotherapy remains cost-effective in the all the analyses across all the risk groups.

The variations in the treatment effect duration are perhaps particularly notable as this is one of the uncertainties around the effectiveness of the single instillation of chemotherapy. The analysis shows, unsurprisingly, that the intervention is less cost-effective when the treatment effect duration is decreased. However, crucially, the single instillation of chemotherapy remains cost-effective in all analyses, even when making very pessimistic assumptions about the likely treatment effect duration (i.e. even when assuming that the chemotherapy instillation only reduces recurrences in the first 3 months after administration).

In addition to the core cost-utility analysis, the GDG were also interested in a cost analysis comparing the cost of delivering the single instillation of chemotherapy on the ward against the cost of delivering it in theatre. It was found that delivering the single instillation of chemotherapy in theatre was the cheaper of the two approaches (delivery by nurse estimated to cost an additional £23.83). This was primarily a result of the longer amount of time taken to deliver the instillation in the ward setting compared to in theatre.

A probabilistic sensitivity analysis was also conducted to assess the combined parameter uncertainty in the model. In this analysis, the mean values that were utilised in the base case were replaced with values drawn from distributions around the mean values. It was found that, at a threshold of £20,000 per QALY, TURBT plus a single instillation of chemotherapy has a very high probability of being cost-effective in the low and intermediate risk groups (100%). However, the probability is substantially lower in high risk patients at 66%, although still very much in favour of TURBT plus a single instillation of chemotherapy.

Conclusion

The results of the analysis suggest that the use of a single instillation of chemotherapy after a TURBT, in comparison to a TURBT alone, was found to be strongly cost-effective in all risk groups. It was found to be particularly cost-effective in low and intermediate risk groups, in which the strategy was cost saving as well as more effective (dominant). Furthermore, this result was found to be robust in alternative scenario analyses, one-way and probabilistic sensitivity analysis.

Recommendations	Offer people with suspected bladder cancer a single dose of intravesical mitomycin C given at the same time as the first TURBT. Offer people with newly diagnosed intermediate-risk non-muscle-invasive bladder cancer a course of at least 6 doses of intravesical mitomycin C. If intermediate-risk non-muscle-invasive bladder cancer recurs after a course of intravesical mitomycin C, refer the person's care to a specialist urology multidisciplinary team. Offer induction and maintenance intravesical BCG to people having treatment with intravesical BCG.
Relative value placed on the outcomes considered	The GDG considered progression to be an important outcome because it is associated with life-threatening complications and the need for more intensive treatment. Recurrence was also considered to be an important outcome because it leads to further treatment and patients noted the avoidance of recurrence as important. Treatment-related morbidity was considered important because intravesical therapy is associated with some side-effects. All of the outcomes from the PICO were reported by the evidence. No additional outcomes that were not specified in the PICO were used to make recommendations. The GDG considered that overall survival and disease-specific survival were not useful outcomes because there were no proven survival differences between treatments. Treatment-related mortality was not considered important because it is not applicable to this patient group as intravesical therapy is not potentially lethal. Health-related quality of life outcomes were also not considered to be useful because very little evidence was identified and it was considered to be of poor quality.
Quality of the evidence	The quality of the evidence ranged from very low to high as assessed with GRADE. Some limitations with the evidence were highlighted. For example, there were issues with applicability to current UK practice because older therapy regimens were used in some studies (some data back to the 1970s) and some study populations were not applicable to the UK. The participants risk level was not clear in some of the included studies. Also, statistical heterogeneity was present in some of the published meta-analyses that were presented. The limitations with the evidence made it difficult for the GDG to make recommendations on specific subgroups. Because some of the chemotherapy regimens in the evidence are not used in current clinical practice the GDG chose to make recommendations based on current practice. The recommendation that the immediate chemotherapy instillation should be given at the time of TURBT (in theatre) was based, partly, on the GDG's experience. The GDG considered instillation at the time of TURBT to be more convenient for clinicians and patients. It also ensures that patients receive the full benefit of this time-dependent treatment. The patient representatives on the GDG were also strongly in favour of this recommendation. The referral to SMDT in patients with recurrent intermediate disease was also based on the GDG's experience. They felt that this was important to ensure a full range of treatment options are considered. Low quality economic evidence was identified. The economist highlighted that the study was only partially applicable to the decision problem as it considered a healthcare system other than the UK (US study). Also potentially serious limitations were identified with the study with uncertainty over some of the input parameters that were used in the model. In addition, the study interpreted the economic results using an atypical approach, leading to potential misleading conclusions i.e. different conclusions might be drawn when a more conventional approach is used. The analysis was also considered to be superseded by the de novo analysis conducted by the economist, which was directly applicable and followed the methodology advised by NICE. Therefore, the published analysis was not given much consideration by the GDG when drafting the recommendations as the de novo economic analysis conducted by the economist was considered to be more appropriate.
Trade-off between clinical benefits and harms	The GDG considered the main benefit of giving a single instillation of MMC to be a reduced risk of recurrence. Giving MMC in theatre should improve access to the treatment and be more convenient for patients. The benefits of giving MMC and BCG were thought to be a reduced risk of recurrence. For BCG there was also a reduced risk of progression. The GDG compared the effectivenss of BCG and MMC and recognised that there maty be some benefit of BCG in reducing recurrence rates. However the extent of this benefit was unclear. In intermediate risk patients the GDG considered that the risk of progression were relatively low and so the reduced toxicity profile of mitomycin C was preferred. The GDG felt that referral to SMDT in patients with recurrent intermediate disease was important to ensure a full range of treatment options are considered. The GDG considered the potential harms of the recommendations made were the side-effects of intravesical treatment, particularly those associated with maintenance BCG. The GDG reached a consensus decision that many patients would rather endure the side-effects of treatment than have a cancer recurrence and receive surgical treatment.
Trade-off between net health benefits and resource use	A health economic model was developed for this topic. The results of the economic analysis were used to inform the recommendations made on the use of a single instillation of chemotherapy after an initial TURBT. The results showed that the addition of a single instillation of chemotherapy was cost-effective in all modelled risk groups. It was found to be particularly cost-effective in low and intermediate risk patients where TURBT + single chemotherapy instillation was found to be cheaper and more effective than TURBT alone (i.e. dominant)). In high risk patients, TURBT + single chemotherapy instillation was found to be more effective than TURBT alone but also more costly. However, it was shown to provide one additional QALY at a cost of £6,432, which is well below NICE's threshold of £20,000 per QALY and so it can therefore be considered cost-effective. While one-way sensitivity analysis demonstrated variation in the ICER values a single instillation of chemotherapy remained cost-effective in all modelled analyses. Furthermore, probabilistic sensitivity analysis showed that at, a threshold of £20,000 per QALY, a single instillation of chemotherapy has a very high probability of being cost-effective in the low and intermediate risk groups (100%). However, the probability is substantially lower in high risk patients at 66%, although still very much in favour of a single instillation of chemotherapy. The cost of delivering a single instillation in theatre was compared against the cost of later delivery by a nurse on the ward. Delivering it in theatre was found to be the cheaper of the two options (£23.83 cheaper). This was primarily a result of the shorter time taken by the urologist to deliver the drug in theatre. In the other areas of the topic not covered by the economic model, the GDG made the following considerations. Mitomycin C course The use of a course of mitomycin C was thought to be associated with increased costs because of the mitomycin C drug costs and the cost of treating side effects. However, there may also be potential cost savings from reduced recurrences and progression (and the further treatments that they entail). Maintenance BCG The use of maintenance BCG was thought to be associated with increased costs because of the BCG drug costs and the cost of treating side effects. However, there may also be potential cost savings from reduced recurrences and progression (and the further treatments that they entail).
Other considerations	The GDG identified no equalities issues for this topic. The GDG considered that the recommendations reflect what is currently considered best practice but there is concern that this is not universally followed. The GDG noted that there may be some additional training required to perform the procedures recommended. The GDG expect to see an increased use of single instillation MMC, MMC course and BCG maintenance. There may also be an increase in referral to SMDT following MMC failure. The GDG anticipate that there will be a greater acceptance of the need to give intravesical MMC in theatre. Regarding the recommendation of referral to SMDT following MMC failure, the GDG discussed the possibility of recommending BCG for these patients. The GDG were mindful of existing NICE guidance (Improving Outcomes in Urological Cancers) and were mindful to ensure that best practice will be universally adopted.

4.2.2. The role of biopsy in people with recurrent non-muscle invasive bladder cancer

People with non-muscle invasive bladder cancer generally have regular cystoscopic follow up to identify recurrent cancer. The likely nature of any recurrence will depend on the nature of the previous cancer.

Treatment of low risk bladder cancer recurrences is generally by transurethral resection to remove the cancer or fulguration by either electrocautery or laser energy to destroy the cancer (with or without biopsy). The former allows pathological evaluation of the cancer and may be necessary to remove tissue from large cancers, but requires regional or general anaesthesia and a rigid cystoscopy and bladder resection. Consequently, the risks of intervention are higher than for fulguration (which may be performed under local anaesthesia). However, fulguration without biopsy does not obtain tissue for analysis and could miss the minority of cases in which the cancer is becoming more aggressive. This approach is less effective at removing the cancer and so could lead to higher recurrence (or residual cancers) rates and more post-treatment symptoms.

It is likely that there is significant variation in the use of risk classification in people with non-muscle invasive bladder cancer. There is also variation in whether or not biopsy is done for apparently low risk disease. Whilst it should be standard practice to biopsy any recurrence in people with intermediate or high risk non-muscle invasive bladder cancer, the variation in the use of risk classification means that this may not always occur.

Clinical question: In patients with recurrent bladder cancer and previous low risk bladder cancer does treatment without histological sampling affect outcome?

Clinical evidence (see also full evidence review)

Evidence was provided by seven observational studies, only one of which was a comparative study. The evidence is summarised in table 57.

Table 57

GRADE evidence profile: In patients with recurrent bladder cancer and previous low risk bladder cancer does treatment with histological sampling versus treatment without histological sampling (e.g cystodiathermy).

Evidence statements

Very low quality evidence from one retrospective observational study reported on 42 patients who underwent fulguration for recurrent Ta bladder cancer and 42 matched patients who underwent TURBT. 12 patients in the fulguration group and 11 patients in the TURBT group had a recurrence during follow-up (RR 0.92, 95% CI 0.46 to 1.84) (Park et al., 2013).

Very low quality evidence from one prospective cohort study of outpatient laser ablation (OLA) in an elderly population (n=54) reported that the procedure was well tolerated with pain scores of 0-2 out of 10. The 3-month recurrence rate was 10.6% with white light OLA and 4.3% with PDD OLA (Wong et al., 2013).

One study of electromotive drug administration (EMDA) of local anaesthetic (LA) for outpatient flexible cystoscopy biopsy and cystodiathermy of recurrent low grade pTaG1-2 (Biers & Mostafid 2009) reported that there were no recurrences at the site of cystodiathermy and there were no progression events. 19% (3/16) of those with benign pathology at biopsy had a recurrence after a mean follow-up of 16.4 months. 9% (1/11) of those with TCC pathology at biopsy had a recurrence, with a time to recurrence of 15 months. Mean pain score was one, on a scale of one (no pain) to 10 (worst pain). There were no intraoperative complications (Very low quality evidence).

One study of 48 patients who were suitable for cystodiathermy under LA reported a local recurrence rate of 6% (n=3) and 15 recurrences (31%) at a different site after a median of 15 weeks follow-up (80% subsequently treated with LA cystodiathermy and 20% referred for GA cystodiathermy). No progressions were reported (Davenport et al., 2010) (Very low quality evidence).

Two studies of 192 patients (515 tumours) undergoing treatment for NMIBC recurrences with Ho:YAG laser ablation under LA with a flexible cystoscope reported a local recurrence rate of 12% (37/304) and an off-site recurrence rate of 50% (Syed et al. 2001; 2013). One study (Syed et al., 2013) reported complication rates of dysuria (4.2%), frequency (1.5%), haematuria (1.9%) and no UTIs. Mean visual pain score was one, on a scale of 0 (no pain) to 10 (worst pain) (Very low quality evidence).

In one study of 267 patients, 103 had small, low grade papillary recurrence and negative cytology and underwent cystodiathermy at least once during the study period (Donat et al., 2004). No significant differences were seen in progression of disease for patients undergoing cystodiathermy (n=103) compared to those never fulgurated in the office (n=164) (p=0.86) (Very low quality evidence).

Cost-effectiveness evidence

The primary results of the analyses by Green et al. 2013 and Wong et al. 2013 are summarised in table 58.

Table 58

Modified GRADE table showing the included evidence (Green et al. 2013 and Wong et al. 2013) for the treatment of recurrent bladder cancer and previous low risk bladder cancer with and without histological sampling.

Green et al. 2013 concluded that fulguration without perioperative intravesical chemotherapy was the most cost-effective strategy for treating low-risk NMIBC. However, unusually, the authors based this conclusion upon individual cost-effectiveness calculations rather than the standard incremental calculations. When following the more standard cost-effectiveness methodology using incremental cost-effectiveness ratios (ICERs), the strategy of perioperative intravesical chemotherapy plus fulguration would most likely be considered the most cost-effective strategy with an ICER of $4,169 per QALY.

Of particular relevance to the topic at hand, was the finding that fulguration was more cost-effective than TURBT when both were used alone or when both were used in combination with intravesical chemotherapy. In both instances fulguration was found to be more effective and cheaper than TURBT alone i.e. dominant. However, as the study is US based, these results may lack applicability to the UK healthcare system.

Wong et al. 2013 found that outpatient laser ablation was cost-effective in comparison to inpatient cystodiathermy for the treatment of NMIBC, especially in elderly patients. In the base case, outpatient laser ablation was found to be cheaper (cost reduction of $2,526) and more effective (0.12 QALYs) than inpatient cystodiathermy and is thus dominant. A further analysis showed that using PDD in addition to outpatient laser ablation was also cost-effective and indeed dominant in comparison to inpatient cystodiathermy.

Probabilistic sensitivity analysis showed that, at a threshold of £30,000 per QALY, outpatient laser ablation had approximately an 80%^c probability of being cost-effective in comparison to intravesical chemotherapy. With the addition of PDD to OLA, the strategy was more cost-effective than IC in 79.2% of simulations.

However, while the study is of some interest, it does not directly address the decision problem at hand because TURBT is not used as a comparator. The study instead compares two alternatives to TURBT and thus the key aspect of our decision problem remains unanswered by this study.

While both of these studies are somewhat useful, their lack of direct applicability to the decision problem under consideration makes it difficult to draw firm conclusions. As such, the cost-effectiveness of perioperative intravesical chemotherapy remains, to a large extent, uncertain.

Recommendations	Consider fulguration without biopsy for people with recurrent non-muscle-invasive bladder cancer if they have all of the following: no previous bladder cancer that was intermediate- or high-risk a disease-free interval of at least 6 months solitary papillary recurrence a tumour diameter of 3 mm or less.
Relative value placed on the outcomes considered	The GDG considered recurrence, progression and treatment-related morbidity to be the most important outcomes because they reflect the benefits and harms to patients of the possible change in NHS practice. Residual tumour rate and health-related quality of life were specified as outcomes in the PICO but were not reported in the evidence. No additional outcomes to those specified in the PICO were used to make recommendations.
Quality of the evidence	The quality of the evidence was very low as assessed with GRADE. The evidence was limited because there was a lack of high quality comparative studies. The included studies had a short duration of follow-up and small sample sizes. The GDG were not confident in the patient-reported pain and treatment-related morbidity data. From clinical experience the GDG considered that the pain associated with fulguration would be greater than reported in the evidence. These issues with the evidence meant that the GDG were cautious about weighing up the benefits and harms of fulguration/biopsy. The GDG used clinical experience to make a conservative recommendation about the criteria for fulguration without biopsy. The criteria are more conservative than those reported in the evidence because the GDG could not be confident in the low quality evidence presented. These recommendations were also supported by the patient/carer representatives. The GDG could not be confident in making a recommendation regarding local anaesthetic fulguration. The evidence presented did not sufficiently answer the review question so a research recommendation was made. The GDG considered that there is variation in the current practice of fulguration, so the recommendation will promote safe patient care and reduce variation in practice until there is a stronger evidence base. The research recommendation will provide an answer to review question. Very low quality health economic evidence was presented. The evidence was not directly applicable to the UK healthcare setting. Some omissions in the report make it difficult to fully appraise quality of evidence (e.g. cost inputs that were used were not fully reported). One study interpreted economic results using an atypical approach, leading to potential misleading conclusions i.e. different conclusions might be drawn when a more conventional approach is used. The GDG acknowledged the available evidence but it did not drive the decision making due to the above issues with the identified studies. The GDG set economic data as an outcome in the research recommendation due to the poor quality of the existing economic evidence.
Trade-off between clinical benefits and harms	The GDG considered that the recommendation will potentially prevent inappropriate fulgurations without biopsy which may lead to disease progressions being detected earlier. The recommendation may also lead to the avoidance of morbidity from biopsies (such as bladder perforation) and the inconvenience of biopsies in low-risk patients. The GDG also acknowledged a possible increase in the number of biopsies and its associated risks due to the conservative criteria for fulguration without biopsies. This may also lead to an increase in patient anxiety whilst waiting for biopsy results. The GDG considered avoiding under treatment from not performing a biopsy as a priority and acknowledged the extent and variation of current practice. Ensuring consistent best practice was considered to outweigh the relatively small harms to the patient. The GDG made a conservative recommendation regarding the criteria for fulguration without biopsy which is thought to outweigh the harms of a possible increase in biopsies. This was also supported by the patient/carer representatives.
Trade-off between net health benefits and resource use	The GDG acknowledged the available health economic evidence but it did not drive their decision making due to the limitations with the evidence discussed above. No health economic model or cost analysis was developed. However, the GDG considered the potential costs and savings of the recommendation made. The costs include potentially more biopsies, although the GDG noted that the extent of increase was unknown. There may also be increased costs from more patients having general anaesthetic. The savings include fewer complications from inappropriate biopsies. There may also be savings by potentially identifying progression early and the associated reduction of further treatment.
Other considerations	No equalities issues were identified. The GDG were unsure of the extent to change in practice that implementation of the recommendation would require.

Research recommendation

In people with exclusively low risk bladder cancer who experience recurrence does the addition of biopsy to fulguration or laser treatment improve progression, recurrence, morbidity and quality of life?

Why is this important

Low risk bladder cancer implies cancer at low risk of recurring within the bladder and of progressing either to more aggressive cancer or to invasive cancer. The management of recurrence of this sort of cancer has been by telescopic destruction (fulguration, resection or laser) of the recurrence, usually but not always with biopsy so that the nature of the recurrence can be confirmed and progression excluded. Biopsy generally requires cystoscopy under general or regional anaesthetic, whereas small recurrent cancers can be cleared by fulguration or laser under local anaesthesia. This may have advantages (eg, avoiding admission, reduced cost) but it risks missing progression by grade or stage.
This research could provide safety evidence for the wider use of avoidance of biopsy in recurrence of previously low risk bladder cancer, resulting in savings and reduced morbidity. It would reduce variation, has no adverse equality impact and the research is achievable.

4.2.3. Re-resection in high risk non-muscle invasive bladder cancer

People with high risk non-muscle invasive bladder cancer may have residual cancer following transurethral resection and they may actually have muscle invasive bladder cancer that was not identified at the first operation. Early repeat resection (re-resection) is used to try to ensure complete cancer clearance and improve staging. It is argued that a high quality initial resection should be sufficient and that a second procedure prolongs the pathway unnecessarily.

There is variation in practice regarding the need for re-resection and the degree of urgency with which this should be performed.

Clinical question: Does re-resection in high risk NMIBC influence outcomes?

Clinical evidence (see also full evidence review)

The evidence is summarised in table 59.

Table 59

GRADE evidence profile: Does re-resection versus no re-resection in people with high risk non-muscle invasive bladder cancer influence outcomes?

Evidence Statements

Low quality evidence (Divrik et al., 2010; Kim et al., 2012) suggests a benefit for repeat transurethral resection in patients with high risk non muscle invasive bladder cancer in terms of bladder cancer recurrence, disease progression and bladder cancer specific mortality.

Using event free survival rates from the no re-resection group in Divrik et al. (2010) trial combined with the hazard ratios reported in table 59 we could expect five year recurrence free survival rates of 63% following re-resection versus 33% without no re-resection. Estimated five-year progression-free survival would be 92% following re-resection group versus 76% without re-resection.

Low quality evidence (Divrik et al., 2010) suggests re-resection is associated with minor complications in approximately 9% of cases, including prolonged bleeding, epididymitis and transient urinary retention. Such complications could be avoided in patients who do not undergo re-resection

A systematic review of observational studies (Vianello et al., 2011) provided low quality evidence of upstaging and tumour persistence rates at re-resection. For patients with stage T1 tumours at initial TURB, approximately 32% were found to have persistent tumour of the same or lower stage at repeated TURB. Around 9% of patients with T1 tumours at initial TURB were upstaged at repeat TURB.

No evidence was found about the impact of re-resection on health related quality of life in this population.

Cost-effectiveness evidence

Recommendations	If the first TURBT shows high-risk non-muscle-invasive bladder cancer, offer another TURBT as soon as possible and no later than 6 weeks after the first resection.
Relative value placed on the outcomes considered	Upstaging and progression were considered by the GDG to be important outcomes. Both affect outcomes for patients and may change treatment decisions, for example radical treatment might be considered in cases of upstaging. Quality of life and patient-reported outcomes were specified as outcomes in the PICO but were not reported in the evidence. Recurrence and residual tumour rate were not considered useful given the evidence on upstaging.
Quality of the evidence	The evidence was assessed as being of low quality using GRADE. There were limitations in the study by Kim et al. (2012) because immediate further resection under pathology guidance was performed rather than subsequent resection, so its relevance to the review question is limited. However, the study provides some further evidence about the importance of obtaining detrusor muscle in the biopsy specimen and the outcomes from performing a further resection. Further limitations of the evidence include a lack of intention-to-treat analysis and a low number of events in the two randomised trials. The GDG considered that the lack of intention to treat analysis was unlikely to have confounded the outcome and despite the low number of events the results were still statistically significant.
Trade-off between clinical benefits and harms	The potential benefit of the recommendation made is the more effective identification of muscle invasive disease. Performing re-resection within 6 weeks could improve outcomes for patients with high-risk non-muscle invasive bladder cancer. The potential harms arise from the psychosocial and clinical morbidity associated with the delay of definitive treatment, and the risk associated with a second resection including general anaesthetic and operative risks. The GDG considered that morbidity from resection is low and the importance of accurate staging was prioritised in the decision making.
Trade-off between net health benefits and resource use	No health economic evidence was identified and no economic model was developed for this topic. The GDG were unsure of costs or savings as there is uncertainty as to what extent the recommendation varies from current practice across the UK. The GDG identified that there may be costs from increased numbers of resections and potential subsequent radical treatment for patients who are upstaged. There may be savings from reduced cost of assessing and treating patients with progressive or recurrent disease, some of which could be incurable, and from less cystoscopy follow-up in patients undergoing cystectomy.
Other considerations	No equalities issues were identified. The GDG were uncertain to what extent the recommendation varies from current practice across the UK. The GDG discussed the feasibility of performing re-resection in under 6 weeks. The studies presented in the evidence review typically reported a timeframe of 2-6 weeks, but there was no evidence comparing delay in re-resection. Therefore the GDG agreed to recommend the 6 week timeframe from the studies, and considered this to be feasible in current practice,

4.2.4. BCG or primary cystectomy in high risk non-muscle invasive bladder cancer

High risk non-muscle invasive bladder cancer has a high risk of progression to muscle invasive cancer and spread beyond the bladder. In order to reduce this risk, active treatments such as intravesical BCG or radical cystectomy are usually considered.

Intravesical BCG reduces the risk of cancer progression, and for people treated successfully with intravesical BCG, major surgery is avoided. However, recurrence and progression are common after intravesical BCG and often result in radical cystectomy. Intravesical BCG can delay the identification of worsening cancers and has a significant side effect profile.

Primary cystectomy is advocated as a more effective cancer treatment than intravesical BCG but if primary cystectomy is used routinely, patients who would have been cured by intravesical BCG alone will have had considerable over treatment with the consequent life changing effects and considerable risks associated with radical cystectomy.

There is wide variation in the use of both of these treatments, and whether a choice between them is offered.

Clinical question: For which patients with non-muscle invasive bladder cancer would primary cystectomy produce better outcomes than BCG?

Clinical evidence (see also full evidence review)

The clinical evidence is summarised in tables 60 to 62

Table 60

GRADE evidence profile: For which patients with non-muscle-invasive bladder cancer would primary cystectomy produce better outcomes than BCG? Comparison: Radiotherapy versus control (observation or intravesical therapy) for T1G3 bladder cancer

Table 61

GRADE evidence profile: For which patients with non-muscle-invasive bladder cancer would primary cystectomy produce better outcomes than BCG? Comparison: Primary cystectomy versus conservative treatment (surveillance or intravesical therapy) for high-risk (more...)

Table 62

GRADE evidence profile: For which patients with non-muscle-invasive bladder cancer would primary cystectomy produce better outcomes than BCG? Comparison: Early cystectomy versus deferred cystectomy for high-risk non-muscle invasive bladder cancer

Evidence statements

Radiotherapy versus observation or BCG therapy

Moderate quality evidence from one randomised trial of 204 patients (Harland et al., 2007) suggests uncertainty over whether radiotherapy is more or less effective than observation or BCG therapy in terms of recurrence-free survival, progression-free survival and overall survival. 5/102 (5%) of patients in the radiotherapy arm experienced long-term toxicity. 18% of the radiotherapy arm and 13% of the control arm underwent cystectomy due to recurrence or progression.

Primary cystectomy versus primary conservative treatment

Very low quality evidence from two retrospective studies (336 patients) suggests uncertainty over whether primary cystectomy is more or less effective than primary conservative treatment (observation or intravesical therapy) in terms of progression or overall survival. Conservative treatment was associated with better five-year disease-specific survival than primary cystectomy in three studies of 664 patients (Badalato et al., 2012; Park et al., 2009; Patard et al., 2001). However, in one study (Park et al., 2009) patients undergoing cystectomy were older, more likely to have proper muscle absent in the TUR specimen and included a higher proportion of gross non-papillary tumours, all of which were associated with reduced disease-specific survival. Three studies reported disease-specific mortality rates in 337 patients. There were no differences in disease-specific mortality in two studies. Low quality evidence from six studies (914 patients) reported a subsequent cystectomy rate of 26% in patients initially treated by conservative therapy.

Early cystectomy versus deferred cystectomy

Very low quality evidence from one study of 77 patients suggests uncertainty about the difference in five-year overall survival between patients treated with early cystectomy compared with patients undergoing deferred cystectomy after BCG failure (72.2% versus 73.2% five-year survival, p=0.75) (Wong et al., 2009). Three studies (583 patients) suggest reduced disease-specific survival in patients undergoing deferred cystectomy, with five-year disease-specific survival ranging from 78% to 84% across studies for early cystectomy and from 67% to 75% across studies for deferred cystectomy (Hautmann et al., 2009; Denzinger et al., 2008; Ali-el-Dein et al., 2011). Ten-year disease-specific survival ranged from 69% to 79% across studies for early cystectomy and from 51% to 64% for deferred cystectomy. Denzinger et al. (2008) reported that concomitant CIS was related to a decrease in disease-specific survival in the deferred cystectomy group only. One systematic review including 3088 patients, reported that disease-specific survival after progression from high-risk NMIBC in initially conservatively treated patients was 35% after a median follow-up of 48-123 months (van den Bosch & Alfred Witjes 2011). The disease-specific mortality rate in 1136 clinical T1G3 patients who underwent radical cystectomy was 29.8% at five years (Fritsche et al., 2010). 50% of this cohort were upstaged to pT2 or higher at cystectomy.

One study of 105 patients reported that 7% of patients had major surgical complications which were distributed equally between early and deferred cystectomy groups, including two fatal pulmonary emboli and one fatal cardiac ischaemia.

One study (Kamat et al., 2006) provides very low quality evidence from 30 patients with micro-papillary bladder cancer. 12 patients undergoing cystectomy as initial therapy had ten-year disease-specific survival of 72%, whilst in 18 patients who underwent cystectomy after progression the median disease-specific survival was 61.7 months with no patient surviving ten years. Very low quality evidence from one study of 138 patients (Cheng et al., 1999) of patients with primary CIS suggests uncertainty about a difference in 15-year progression-free survival and disease-specific survival between those treated with immediate cystectomy and those that were not (some deferred cystectomy, some intravesical therapy). Radical cystectomy performed within three months after the initial diagnosis was associated with improved overall survival, but this was not significant after controlling for age.

Cost-effectiveness evidence

The primary results of the analysis by Kulkarni et al. 2009 are summarised in table 63.

Table 63

Modified GRADE table showing the included evidence for treatments for high risk non-muscle invasive bladder cancer.

The base case results of the cost-effectiveness analysis showed that immediate cystectomy was cheaper and more effective than conservative therapy (BCG with possible delayed cystectomy) i.e. immediate cystectomy was found to be the dominant strategy.

Scenario analyses, in which age and co-morbid status were varied, showed that the optimal strategy is likely to be different for different patient subgroups. The analysis showed that immediate cystectomy was dominant in patients aged ≤55 years old regardless of co-morbid status. For patients ≥70 years old, conservative therapy was either dominant or had an ICER that was likely to be considered cost-effective (≤$32,700 per QALY). For patients between ages 60 and 70 years old, the optimal choice was dependent upon co-morbidities, with increased co-morbid burden making conservative therapy more cost-effective.

The probabilistic sensitivity analyses (PSA) showed that immediate cystectomy was found to be cost-effective in 70% and 67% of simulations at thresholds of $20,000 and $50,000 per QALY, respectively.

The results suggest that, compared with a conservative strategy using BCG, immediate radical cystectomy yielded better outcomes and lower costs for the average patient. Furthermore, the results suggest that tailoring therapy based on patient age and co-morbidity may increase survival and yield significant costs savings for the health care system.

However, there are reservations about the applicability of the analysis because it considered the Canadian health care system which may not reflect the UK setting. There were also concerns about the quality of life data that were used as they were not all patient reported and were often not drawn from patients with bladder cancer (data from prostate, lung and breast cancer were used). Potentially serious limitations were also identified as, although a systematic literature review was conducted, some of the evidence informing the model was not considered to be of high quality. Furthermore, costs were not always sourced from patients with bladder cancer, such as chemotherapy costs, which were based on patients with non-small cell lung cancer.

Recommendations	Offer the choice of intravesical BCG (Bacille Calmette-Guérin) or radical cystectomy to people with high-risk non-muscle-invasive bladder cancer, and base the choice on a full discussion with the person, the clinical nurse specialist and a urologist who performs both intravesical BCG and radical cystectomy. Include in your discussion: the type, stage and grade of the cancer, the presence of carcinoma in situ, the presence of variant pathology, prostatic urethral or bladder neck status and the number of tumours risk of progression to muscle invasion, metastases and death risk of understaging benefits of both treatments, including survival rates and the likelihood of further treatment risks of both treatments factors that affect outcomes (for example, comorbidities and life expectancy) impact on quality of life, body image, and sexual and urinary function.
Relative value placed on the outcomes considered	The outcomes of progression, survival, recurrence, cystectomy rate, and health-related quality of life were considered to be the most important. These are the main disease-specific outcomes for high risk non-muscle invasive bladder cancer. The evidence review suggested that these informed the natural history of the disease following treatment by BCG, radical radiotherapy and cystectomy. No evidence was identified for health-related quality of life.
Quality of the evidence	The quality of the evidence was assessed with GRADE as being very low to moderate. The best evidence available was a randomised trial comparing radical radiotherapy with BCG. Limitations of the evidence were that most studies were retrospective and therefore had a risk of selection bias. There were inconsistencies in the terminology used for delayed and deferred cystectomy. These issues made the evidence unreliable regarding the decision on which patients should receive BCG or cystectomy and the GDG considered this when reaching consensus. The recommendation for discussion of treatment options with health care professionals was based on clinical consensus. There was no evidence on this issue but it was considered critical to enable the patient to take an informed decision if they chose to. This is considered to be consistent with best practice. The GDG made a recommendation because patients with non-muscle invasive bladder cancer need to be treated. However, it was unclear from the available evidence which is the most effective primary treatment option. It was therefore agreed that further research into this area is needed. A research recommendation was made because of the lack of good quality evidence about which intervention is more clinically effective and cost effective. The research recommendation will also provide much needed evidence about the specific impact of these treatments on quality of life outcomes. Low quality economic evidence was identified from one Canadian study. This evidence was limited because the economic analysis was performed using Canadian costs, which may not be directly comparable with UK costs. Also, quality of life was estimated by clinicians rather than patients and the clinical effectiveness data informing the model differed from the clinical evidence review. The study also reported poorly defined clinical utility measures without reference to the information sources. These limitations meant that the GDG were unable to rely on the model to inform the recommendations. The GDG reached consensus assuming equipoise of treatments.
Trade-off between clinical benefits and harms	The GDG considered that a potential benefit of the recommendation is that patients with high risk non-muscle invasive bladder cancer should have a better informed and balanced discussion regarding their treatment. This should improve their understanding of the disease and should improve clinical outcomes. The GDG considered that there is a potential for an increase in cystectomies with the possible risk of over-treatment for some patients. Also, the discussion about treatment options could result in an overload of information for some patients, especially those who would prefer to delegate decision making. The GDG balanced the benefits against the harms by considering that patients must be given the opportunity to access full information about their prognosis and the potential benefits and risks of treatment, including the impact on quality of life. The GDG considered that giving this opportunity to all patients was of greater benefit than of giving too much information to some patients. Information and support in decision making is important for patients to make an informed decision regarding treatment, taking into account their preferences as well as prognostic information.
Trade-off between net health benefits and resource use	Low quality health economic evidence was identified from one Canadian study. However, the GDG were unable to rely on the evidence to inform the recommendations because of the limitations discussed above. The GDG reached consensus assuming equipoise of treatments. No health economic model was developed for this topic. The GDG considered that there are potential changes for working within clinical networks and some more review of patients necessary by specialist teams, which could incur extra costs to the NHS. The GDG agreed that there could be savings from reduced treatment of advanced disease due to an improved cure rate.
Other considerations	No equalities issues were identified. The GDG considered that the recommendations may alter practice in the areas served by some former cancer networks, with an increase in referral of patients to central services. The GDG considered that the evidence for BCG from section 4.2.1 would be relevant to this area. The recommendation for patients to be reviewed by a specialist performing BCG/Cystectomy was made with knowledge of current thinking of best practice. Involvement of the CNS is consistent with the NICE Urological Cancers Improving Outcomes Guidance.

Research recommendation

Is primary radical cystectomy more effective than primary intravesical BCG in high-risk non-muscle-invasive bladder cancer, in terms of quality of life and cancer-specific outcomes?

Why is this important

Options for people with high-risk non-muscle-invasive bladder cancer include cystoscopy surveillance, BCG immunotherapy or radical surgery. To date, these have not been directly compared across the same population to understand their relative benefits.
Bladder-sparing approaches avoid major surgery, but have a greater risk of cancer progression. The potential advantage of bladder-sparing approaches compared with cystectomy in maintaining quality of life may be offset by continuing concern about cancer progression and morbidity from treatment. Primary cystectomy may improve survival; however, it has high short term risks and life changing consequences. It will be overtreatment for those people whose cancer would not have progressed.

4.2.5. Treatment following failure of BCG

Failure to respond to intravesical BCG includes cancer still present after induction BCG or recurrent cancer during or after maintenance BCG treatment. Residual or recurrent cancer may be non muscle invasive or muscle invasive. Intravesical BCG failure can also include patients who did not complete their treatment due to intravesical BCG related side effects (called BCG intolerant), and therefore they may or may not be clear of cancer.

This section focuses on people with residual or recurrent non-muscle invasive bladder cancer following intravesical BCG and people who have not tolerated intravesical BCG.

The treatment options for these patients include radical cystectomy or some form of bladder sparing treatment. Radical cystectomy has the highest cure rate but may be over treatment and has life changing effects and considerable risks. The bladder sparing treatments include further intravesical BCG, intravesical chemotherapy or radical radiotherapy. These approaches avoid removal of the bladder, but carry the risk that the tumour may not respond and will progress to invasion or spread beyond the bladder. They also have side effects.

There is currently considerable variation in the management of people with non-muscle invasive bladder cancer who have failed intravesical BCG therapy.

Clinical question: What is the optimum treatment for patients with non-muscle invasive bladder cancer who have failed BCG?

Clinical evidence (see also full evidence review)

The evidence is summarised in tables 64 to 67

Table 64

GRADE evidence profile: What is the optimum treatment for patients with non-muscle-invasive bladder cancer who have failed BCG? Comparison: mitomycin C compared to gemcitabine

Table 65

GRADE evidence profile: What is the optimum treatment for patients with non-muscle-invasive bladder cancer who have failed BCG? Comparison: gemcitabine compared to BCG

Table 66

GRADE evidence profile: What is the optimum treatment for patients with non-muscle-invasive bladder cancer who have failed BCG? Comparison: BCG compared to chemotherapy

Table 67

GRADE evidence profile: What is the optimum treatment for patients with non-muscle-invasive bladder cancer who have failed BCG? Comparison: BCG alone compared to BCG plus interferon α2B

Evidence statements

Gemcitabine versus Mitomycin C

Moderate quality evidence from one randomised trial (Addeo et al., 2009) of 109 patients suggests uncertainty over the incidence of tumour recurrence in gemcitabine- versus mitomycin C-treated patients. Although incidence of tumour recurrence was lower in gemcitabine treated patients after 36 months of follow up, the 95% confidence interval around the estimated effect included both no effect and considerable benefit for gemcitabine.

Moderate quality evidence from one randomised trial of 109 patients (Addeo et al., 2010) suggests uncertainty over the incidence of tumour progression in gemcitabine- versus mitomycin C-treated patients. Incidence of tumour progression was lower in gemcitabine treated patients after 36 months of follow up, but the 95% confidence interval around the estimated effect was wide and included considerable harm, no effect and considerable benefit for gemcitabine.

Moderate quality evidence from one randomised trial of 109 patients (Addeo et al., 2010) suggested that gemcitabine treatment was associated with fewer adverse events than mitomycin C.

Gemcitabine versus intravesical BCG

Two studies (Di Lorenzo et al., 2010; Gacci et al., 2006) compared the effectiveness of gemcitabine to BCG. Meta-analysis of the results was not possible due to differences in study design and outcome definitions.

Moderate quality evidence from one randomised trial of 80 patients (Di Lorenzo et al., 2010) suggests that the incidence of tumour recurrence after 12 months is lower in patients treated with gemcitabine compared to treatment with BCG. In patients experiencing recurrence (n=56), there was no significant difference between treatment groups in the incidence of cystectomy due to disease progression. The incidence of grade two and grade three adverse events was similar for both treatments.

Very low quality evidence from one observational trial of 19 patients (Gacci et al., 2006) found no significant difference in tumour recurrence, overall survival, bladder preservation rates or adverse events between gemcitabine and BCG treatment.

BCG versus chemotherapy (MMC or epirubicin)

Very low quality evidence from one observational trial of 183 patients (Matsumoto et al., 2012) suggests that rates of recurrence-free survival (after five years of follow up) are greater in patients treated with BCG than in patients treated with chemotherapy (MMC or epirubicin).

BCG versus BCG plus interferon α2B

Very low quality evidence from one observational trial of 139 patients (Prasad et al., 2009) suggests that the incidence of disease recurrence is lower in patients treated with BCG alone compared with BCG in combination with interferon α2B.

Cost-effectiveness evidence

Recommendations	If induction BCG fails (because it is not tolerated, or bladder cancer persists or recurs after treatment with BCG), refer the person's care to a specialist urology multidisciplinary team. For people in whom induction BCG has failed, the specialist urology multidisciplinary team should assess the suitability of radical cystectomy, or further intravesical therapy if radical cystectomy is unsuitable or declined by the person, or if the bladder cancer that recurs is intermediate- or low-risk.
Relative value placed on the outcomes considered	The GDG prioritised the cancer-related outcomes of recurrence, progression, survival and treatment-related morbidity, as these are of the greatest importance to patients. Progression in particular leads to further treatment and is associated with worse prognosis. The GDG considered the outcome of bladder preservation rate to not be useful once the evidence had been appraised because evidence was only available for one possible comparison of treatments. This evidence was either of very low quality or reported only for a small number of patients. Quality of life was specified as an outcome in the PICO but was not reported in the evidence. No additional outcomes to those specified in the PICO were used to make recommendations.
Quality of the evidence	The quality of the evidence was very low to moderate as assessed with GRADE. The GDG considered potential issues with the evidence presented. Most notably, the lack of any systematic reviews and the unsuitability of any existing randomised trial evidence for meta-analysis. These issues meant that the GDG discussed the evidence in light of clinical experience and comments from patient representatives. The GDG considered that no specific intravesical therapies could be recommended due to the low quality and general lack of evidence. The GDG made the recommendation to refer patients to a SMDT for consideration of treatment options based on their clinical experience because there was no strong evidence in this area. The recommendation to consider cystectomy was prioritised based on clinical judgement and evidence of its effectiveness as a primary therapy in patients with high-risk NMIBC (presented in section 4.3.1). The GDG also made a research recommendation because of the uncertainty about which treatment is best for patients who fail BCG and who are also unsuitable for cystectomy. This research recommendation will help reduce the uncertainty about the effectiveness of radiotherapy and other novel intravesical therapies for these patients.
Trade-off between clinical benefits and harms	The GDG considered the potential benefits of the recommendation. Referral to a SMDT will ensure specialist consideration of patients with high-risk NMIBC who fail BCG treatment. This includes the consideration of appropriate treatment options, such as cystectomy or further intravesical therapy. This may also prevent under-treatment of patients in this group. The GDG considered that the recommendations will enhance patient choice and informed decision-making. The GDG noted that a possible harm of the recommendation is that potentially more patients will undergo surgery, which has associated risks and morbidity. The GDG considered that the increased probability of survival and more informed decision-making for patients would outweigh the potential increase in morbidity from surgery.
Trade-off between net health benefits and resource use	No health economic evidence was identified and no economic model was developed for this topic. The GDG considered that the potential costs of the recommendations made include increased workload for SMDTs and that potentially a greater number of patients will undergo surgery. The GDG considered that if potentially more patients undergo surgery as a result of the recommendations, then savings will be made from less intravesical therapy being administered and reduced cystoscopic follow-up. There will also be savings from reduced need for treatment of disease progression.
Other considerations	The GDG considered that cystectomy may not be an option for patients with poor manual dexterity, visual impairment or diminished mental capacity. However, the recommendations main aim is to promote equal access for all patients to specialist care. The GDG considered that the recommendations reflect best current UK practice, but acknowledged that there may be variability in adherence to this at present. The GDG therefore considered that moderate changes in practice may be required. The GDG discussed the option of radiotherapy as a treatment for this patient group. There was insufficient evidence to make a recommendation, but the GDG considered that radiotherapy could be an appropriate treatment option in a very small number of patients. The recommendation does not preclude the use of radiotherapy and a relevant research recommendation has been made. The GDG took account of the existing NICE IPG covering device-assisted Mitomycin C and a relevant recently completed but currently unpublished trial, the results of which are awaited.

Research recommendation

In people who cannot tolerate BCG or with persistent or recurrent disease after BCG, or who are not suitable for radical cystectomy is novel intravesical therapy or radiotherapy more effective than the current standard of care (for example intravesical mitomycin-C) in terms of recurrence, progression, survival and quality of life?

Why is this important

People with high risk non-muscle invasive bladder cancer are usually offered either instillation of BCG vaccine into their bladder or surgery to remove their bladder (cystectomy), because of the high risks that the cancer may worsen and spread into the muscle wall of the bladder. If BCG cannot be tolerated due to side effects, or if it fails to clear the cancer, people who are not fit enough for cystectomy, or who decline it, are at very high risk of progression of their cancer, and death. At present, further BCG or instillation of Mitomycin C are the other main treatment options.
This research would establish the efficacy and risks of radiotherapy and of novel intravesical therapy in this group who have no effective standard treatment option at present.
There would be no equality consequence, and the logistics of the research would be deliverable.

4.3. Managing side effects of treatment for non-muscle-invasive bladder cancer

Radical radiotherapy and intravesical BCG (BCG vaccine inserted into the bladder), treatments used for high risk bladder cancer that is confined to the bladder can result in patients being cured of their cancer and with their bladder preserved but with significant side effects which can result in patients having a poor quality of life.

Most people treated with intravesical BCG experience urinary frequency and urgency, visible haematuria and some pain when passing urine for 7- 10 days after each treatment. People treated with radical radiotherapy often experience similar symptoms but of lesser degree and shorter duration. However for some people these side effects continue long term.

People who experience these symptoms are usually offered simple conservative treatments, typically medication, and this is often helpful. However, as with all medication patients may experience side effects. No specific treatment has been developed for the symptoms in relation either to intravesical BCG treatment or to radical radiotherapy.

These side effects can be of a persistence and severity that interventions such as urinary catheters or occasionally even radical cystectomy may be considered. Most haematuria following intravesical BCG or radical radiotherapy will stop without any need for treatment. Treatment for persistant bleeding includes cystoscopy and diathermy, instillation of formalin or alum into the bladder. Whilst these treatments may reduce or resolve bleeding, formalin and alum can both have severe side effects. Severe bleeding can also be treated by embolisation, but this is not widely available.

Medication has been given to try to prevent or alleviate side effects in people being treated with intravesical BCG but these are not widely used. Some people are unable to complete the scheduled maintenance course of intravesical BCG because of bladder side effects and intravesical BCG schedules have been changed to improve compliance. Intravesical BCG dosage has been reduced and interval between treatments has been extended.

There is variation in the treatments that are currently offered to people who may experience or who have side effects following intravesical BCG and radical radiotherapy. Side effects are managed by a variety of different healthcare professionals in a variety of different settings.

Clinical question: What is the most effective intervention for bladder toxicity following radiotherapy or BCG therapy for bladder cancer?

Clinical evidence (see also full evidence review)

The evidence is summarised in tables 68 to 74. No evidence was identified for health-related quality of life across any of the interventions. No evidence was identified for the following interventions specified in the PICO: cystectomy, botox, alum, embolisation, catheterisation, increased time between treatments of BCG, elmiron.

Table 68

GRADE evidence profile: The effectiveness of Ofloxacin for the prevention of BCG-induced toxicity in superficial bladder cancer.

Table 69

GRADE evidence profile: The effectiveness of Isoniazid for the prevention of BCG-induced bladder toxicity in superficial bladder cancer.

Table 70

GRADE evidence profile: The effectiveness of Oxybutynin for the prevention of BCG-induced toxicity in superficial bladder cancer.

Table 71

GRADE evidence profile: The effectiveness of reduced BCG dose for BCG-induced toxicity in superficial bladder cancer: 1/3 dose versus standard dose.

Table 72

GRADE evidence profile: The effectiveness of formalin for the treatment of bladder haemorrhage secondary to radiation-induced cystitis.

Table 73

GRADE evidence profile: The effectiveness of hyperbaric oxygen therapy (HBOT) for the treatment of radiation-induced hemorrhagic cystitis.

Table 74

GRADE evidence profile: The effectiveness of sodium hyaluronate for the treatment of chemical-induced cystitis.

Evidence Statements

Ofloxacin

One randomised trial (115 participants) of moderate quality was identified comparing BCG therapy plus ofloxacin with BCG therapy plus placebo in patients with superficial bladder cancer. Treatment with 2 × 200mg ofloxacin with each BCG instillation resulted in a lower rate of mild to moderate adverse events compared to placebo between instillations four and six, and a lower rate of severe adverse events between instillations one and nine. However, the proportion of participants specifically with bladder toxicity was not reported, as the outcome of adverse events included both local and systemic symptoms.

Isoniazid

Two randomised trials (997 participants) provided moderate quality evidence on the efficacy of isoniazid for the prevention of BCG-induced bladder toxicity. In both studies the 95% confidence intervals of the effect sizes (risk ratios) included the null value, so there is no strong evidence that isoniazid has an effect on the rate of chemical cystitis, frequency or haematuria (van der Meijden et al., 2001) or bladder toxicity (including haematuria, dysuria, and frequency) (Al Khalifa et al., 2000). When toxicity was sub-grouped by severity, participants receiving isoniazid were more likely to experience mild toxicity and less likely to experience severe toxicity than the placebo group. However, it should be noted that these data were from a low number of participants.

Oxybutynin

One randomised trial (Johnson et al., 2013) of 50 participants provided low quality evidence of an increase in urinary symptoms (frequency and burning) and systemic symptoms (fever, dry mouth) in those treated with oxybutynin alongside BCG treatment compared to those in the placebo group.

Reduced BCG dose

High quality evidence from one trial (663 patients) of reduced dose BCG reported by Brausi et al. (2014) stated that there were no differences between rates of local and systemic BCG side effects between the 1/3 dose BCG group and the full-dose BCG group (RR 0.95, 95% CI 0.86 to 1.06). Reducing the dose of BCG did not decrease the percentage of patients who discontinued treatment due to side effects.

Formalin

Two case series studies (12 participants) reported the effects of intravesical formalin for treating bladder haemorrhage secondary to radiation-induced cystitis. Both studies reported that all patients had a good response to treatment with cessation of bleeding observed for three to five months (very low quality evidence).

Hyperbaric oxygen therapy (HBOT)

Seven case series studies (153 participants) provided very low quality evidence on the efficacy of HBOT for treating radiation-induced cystitis. Overall 94/153 (61%) participants showed a complete resolution of haematuria, with effectiveness ranging from 27% to 100% across studies. In most studies patients had received previous treatment for cystitis, such as alum or formalin, without success.

Sodium hyaluronate

One case series (54 patients) provided very low quality evidence on the efficacy of intravesical sodium hyaluronate for the treatment of chemical-induced cystitis in bladder cancer patients treated with Mitomycin C or BCG therapy. It is not stated whether Cystistat was the treatment used. Bladder capacity increased in all patients after treatment (mean difference 226.1 ml, 95% CI 207.1 to 245 ml). Patient-reported pain as measured by the Visual Analogue Scale (VAS) decreased in all patients (mean difference -7.7, 95% CI -8.12 to -7.31). VAS scores range from 1 to 10, with 10 indicating maximum pain tolerated.

Cost-effectiveness evidence

Recommendations	Do not offer primary prophylaxis to prevent BCG-related bladder toxicity except as part of a clinical trial. Seek advice from a specialist urology multidisciplinary team if symptoms of bladder toxicity after BCG cannot be controlled with antispasmodics or non-opiate analgesia and other causes have been excluded by cystoscopy. Seek advice from a specialist urology multidisciplinary team if symptoms of bladder toxicity after radiotherapy cannot be controlled with antispasmodics or non-opiate analgesia and other causes have been excluded by cystoscopy.
Relative value placed on the outcomes considered	Treatment-related toxicity and quality of life were both considered important outcomes despite the lack of evidence on quality of life. The GDG considered that quality of life would be improved by a reduction in bladder toxicity. Quality of life was not reported in the evidence.
Quality of the evidence	The quality of the evidence was assessed by GRADE as being of very low to high quality. Most of the publications were small case series studies, which are inadequate to assess this clinical scenario. The randomised trials that were identified were limited by a small sample size and low number of events. Many of the studies also included patients without a bladder cancer diagnosis which limits the relevance to the review question. The only outcome that was assessed as being of high quality reported no difference between reduced dose and normal dose BCG treatment. In the absence of high quality evidence about toxicity, the GDG were concerned about the detrimental effects of the interventions reported (e.g. ofloxacin and isoniazid) on the efficacy of BCG therapy. Due to this lack of evidence, the GDG based their recommendations on their clinical experience and consensus, and recommended that prophylaxis for BCG toxicity should not be offered outside of a clinical trial. The GDG agreed that there was insufficient evidence on which to base a recommendation about prophylaxis prior to radiotherapy. A recommendation for discussion with a specialist urology multidisciplinary team was made because the GDG could not make evidence-based recommendations for a specific treatment. There was no strong evidence to support a recommendation of prophylactic Ofloxacin or Isoniazid to prevent bladder toxicity, nor to reduce the dose or frequency of intravesical BCG. The GDG made a research recommendation because there is limited data that prophylactic treatment reduces BCG toxicity and there is also uncertainty about whether there could be a detrimental effect on the efficacy of the primary treatment (BCG therapy or radiotherapy). The GDG felt it would be worth exploring this with further research but noted that future studies would need to have sufficient power in order to exclude non-inferiority. The GDG made both a research recommendation and the recommendation not to offer prophylactic treatment outside the context of a clinical trial. The GDG agreed that this recommendation was made to avoid the possibility that the primary treatment (BCG or radiotherapy) may be rendered less effective by prophylactic interventions.
Trade-off between clinical benefits and harms	The GDG considered that the potential benefits of their recommendations were avoiding unknown detrimental effects of prophylactic treatments and optimising management of patients in an evidence-poor area. The GDG considered that a potential clinical benefit from the recommendation is that the skills to treat patients with bladder toxicity will be centralised in specialised teams. The GDG considered that the lack of clear advice on what to do to prevent or treat radiation toxicity is a potential harm resulting from the recommendations. However, the GDG considered that it was best not to advise the use of unproven treatments that might worsen cancer outcomes.
Trade-off between net health benefits and resource use	There was no health economic evidence and an economic model was not developed for this topic. The GDG considered that less use of unproven preventative treatments would result in lower cost. The GDG considered that there would be an additional cost associated with seeking advice from specialist teams. However, earlier specialist team involvement may reduce extended local hospital stays, community care costs and the use of ineffective treatments.
Other considerations	Implementing the recommendations is unlikely to involve any equality issues. A potential change in clinical practice was identified by the GDG because the recommendations may result in increased involvement of specialist teams for uncommon but clinically difficult problems. The GDG also considered that the involvement of specialist teams may improve expertise within clinical practice.

Research recommendation

Which interventions are effective in preventing or treating symptoms of bladder toxicity in people having BCG or radiation? A randomised trial should measure toxicity, quality of life, bladder cancer recurrence and progression.

Why is this important

Radiotherapy and intravesical BCG can be effective in controlling or curing bladder cancer. Side effects, such as urinary frequency, urgency, bladder pain or bleeding can significantly worsen quality of life. The standard maintenance course of BCG is often not completed because of these side effects. The degree of the side effects following either treatment is occasionally so profound that cystectomy may be considered to alleviate them.
There is no significant evidence that strategies commonly used to reduce side effects, such as reducing the dose, number of treatments, oral anticholinergics or prophylactic Ofloxacin, are effective. produce improved outcomes for toxicity and quality of life, without detriment to bladder cancer recurrence and progression

4.4. Follow-up after treatment for non-muscle-invasive bladder cancer

As discussed in section 4.1.2, non-muscle invasive bladder cancer can be divided into low, intermediate and high risk groups based on the risk of recurrence and progression.

Follow up of people with non-muscle invasive bladder cancer is done largely with periodic cystoscopy and the frequency of this is often adjusted according to the perceived degree of risk of the cancer. The scheduling of cystoscopy may be erratic due to lack of adherence to follow up protocols and waiting times. This adds extra stress to patients in addition to their anxiety about whether recurrence will be found.

Long term cystoscopic surveillance is expensive. The appropriate duration and frequency of cystoscopic follow up is unclear and in particular how it varies according to risk. Most follow up cystoscopies are likely to be done in people with low risk disease. Concern has been expressed whether current regimens are clinically and cost effective.

Urine cytology is also widely used in follow up of people with non-muscle invasive bladder cancer. Its sensitivity and specificity varies between risk groups but this is probably not taken into account in routine practice. In some hospitals urinary biomarkers are also used as well as or instead of urine cytology but this is not common practice.

Clinical question: What are the optimal follow-up protocols for low/intermediate risk and high-risk non-muscle invasive bladder cancer?

Clinical evidence (see also full evidence review)

The clinical evidence is summarised in tables 75 to 77.

Table 75

GRADE evidence profile: What are the optimal follow-up protocols for low/intermediate and high-risk non-muscle-invasive bladder cancer? Comparison: Frequent versus less frequent follow-up for TaG1-2 bladder cancer

Table 76

GRADE evidence profile: That are the optimal follow-up protocols for low/intermediate and high-risk non-muscle-invasive bladder cancer?

Table 77

Patient experience and preference for follow-up of NMIBC.

Evidence statements

Moderate quality evidence from one randomised trial of 97 patients (Olsen & Genster, 1995) suggests uncertainty over whether cystoscopic follow-up frequency of three months is more or less effective than follow-up with a frequency of six months in terms of recurrence, progression or overall survival.

Low quality evidence from five observational studies of patients with low-grade superficial bladder cancer report recurrence rates over long-term follow-up. Two studies including 470 patients suggest that tumour detection at the first follow-up cystoscopy is associated with a greater risk of recurrence during subsequent follow-up compared to those who are tumour-free at the first cystoscopy (Holmang & Johansson 2002; Mariappan & Smith, 2005). All studies report a reduction in the risk of recurrence over time. Some studies suggest the risk of recurrences is greatly reduced after a tumour-free period of five years or more (Mariappan & Smith, 2005; Zieger et al., 2000). In Mariappan & Smith (2005) only one (0.9%) patient had a first recurrence after being tumour-free for five years, whereas LeBlanc et al. (1999) reports recurrence rates of approximately 30% in patients after remaining tumour-free for two to ten years. Another study reports that of 20 primary Ta-T1 patients who were tumour-free for five years, seven (35%) had muscle-invasive disease (Thompson et al., 1993).

One retrospective observational study of 542 intermediate-high risk patients who had received BCG treatment reports that 338/542 (62%) patients were not tumour-free for five years or more. 22/204 (10.8%) patients had a recurrence after being tumour-free for five years or more (Holmang & Strock 2012). During the first five-years after BCG, 57 patients (10.5%) died from bladder cancer and between years six and 25, 32 patients (5.9%) died from bladder cancer.

Five observational studies report rates of upper urinary tract (UUT) recurrence ranging between 2.6% and 5.5%. Median times to UUT recurrence vary from 22 to 33 months in three studies (Miyake et al., 2006; Canales et al., 2006; Holmang et al., 1998) and one study (Hession et al., 1999) reports a mean time to recurrence of 78 months. In one study, two out of 18 UUT cancers were diagnosed by routine intravenous urography, and the other 18 presented with symptoms suggesting UUT recurrence before IVU (Miyake et al., 2006). Holmang et al. (1998) reported that IVU performed 0 to ten months before the UUT cancer was diagnosed failed to raise suspicion of a tumour in eight out of 16 patients (including three patients with initial muscle-invasive bladder cancer).

Two studies provide low quality evidence of the accuracy of ultrasound compared with cystoscopy for the detection of recurrent tumours in patients with superficial bladder cancer. In one study, three tumours detected by cystoscopy were missed by ultrasound (Stamatiou et al., 2011, and in the second study 15 patients with recurrence were not detected by ultrasound (Vallencien et al., 1986).

Low quality evidence for health-related quality of life is provided by three studies (503 patients) which report that most patients experience minimal pain (Yossepowitch et al., 2007) from undergoing cystoscopic follow-up, although the introduction of the cystoscope is rated as the most painful part of the procedure (van der Aa et al., 2008). Waiting for test results is rated as the most distressing part of follow-up by urine testing (van der Aa et al., 2008).

Cost-effectiveness evidence (see also Appendix B)

Background

There is general agreement that patients with non-muscle invasive bladder cancer (NMIBC) require regular cystoscopic surveillance of their bladder to check for recurrence. However, there is no agreement upon the optimal frequency and length of cystoscopic follow-up and, as such, there is significant variation in clinical practice.

Tailoring follow-up strategies based on risk could allow for follow-up to be safely reduced in the lower risk groups whilst ensuring that the higher risk patients are still monitored closely. In addition, the use of alternative tests to cystoscopy, such as urinary biomarkers and cytology, could have a useful role in reducing the burden of cystocopies. However, the effectiveness and cost-effectiveness of such approaches has never been reliably demonstrated.

Aims

To estimate the cost-effectiveness of reduced follow-up and/or follow-up using newer tests and techniques in comparison to the test and protocols used in current practice in NMIBC patients.

Existing Economic Evidence

A systematic literature review did not identify any cost-utility analyses that sufficiently addressed the current decision problem. However, three papers were identified that utilised modelling techniques to compare follow-up strategies; De Bekker Grob et al. 2009, Van Kessel et al. 2013 and Zhang et al. 2013.

De Bekker Grob et al. 2009 constructed a semi-Markov model to investigate two strategies; a conventional strategy consisting of cystoscopy every 3 months and a test arm consisting of microsatellite analysis of voided urine samples every 3 months with a control cystoscopy at 3, 12 and 24 months. The authors found that the probability of being without recurrence after 2 years was similar in the two groups but the total costs were higher in the test arm. Further analysis suggested that the test arm would be as effective and cost the same as the conventional arm if the sensitivity increased to ≥61%, the specificity was set to 73% and the costs were decreased from €158 to <€70. The authors concluded that cystoscopy could be partly replaced if the microsatellite analysis urine test had a higher sensitivity and its costs were reduced.

A similar analysis was conducted by Van Kessel et al. 2013, in which three surveillance strategies were compared using a Markov model; standard surveillance defined as cystoscopy every three months, minimal surveillance defined as cystoscopy at 3, 12 and 24 months and modified surveillance consisting of FGFR3 mutation analysis of voided urine samples every 3 months and cystoscopy at 3, 12 and 24 months. The authors found that the probability of no recurrence after two years of surveillance was higher for the modified surveillance than the standard or minimal surveillance arms. The total cost of surveillance was found to be lower for minimal and modified surveillance than for standard surveillance. The authors concluded that surveillance in which cystoscopy is partly replaced by FGFR3 mutation analysis of urine seems a safe, effective and cost-effective surveillance strategy.

The analysis conducted by Zhang et al. 2013 compared surveillance strategies for low risk NMIBC patients. The study was not a cost-effectiveness analysis and indeed did not even consider costs but it did estimate QALYs for each strategy. The authors developed a Markov model to compare surveillance strategies recommended in international guidelines and additional proposed strategies. The authors found that age and co-morbidities significantly affect the optimal surveillance strategy. The results suggested that younger patients should be screened more intensively than older patients and patients with co-morbidities should be screened less intensively.

De Novo Economic Model

Since the current economic literature didn't adequately address the decision problem^d, a de novo economic evaluation was undertaken to assess cost-effectiveness. A Markov decision model was developed using Microsoft Excel.

Patients were assumed to enter the model in a ‘disease free’ state following an initial transurethral resection of the bladder tumour (TURBT). At each 3-monthly model cycle the patient may experience a bladder cancer recurrence. If the recurrence is detected, the patient will undergo a further TURBT (or fulguration of the tumour) and return to a disease free state. However, if the recurrence is not detected, then the patient will be at risk of progression and will have to undergo further treatment once this progression is eventually detected (cystectomy and possibly neo-adjuvant chemotherapy). The patient may also die from bladder cancer related mortality after experiencing progression and may die from other cause mortality from any health state.

Estimated total costs and quality adjusted life years (QALYs) were collected over the modelled 10 year time horizon for each follow-up strategy. Future costs and benefits were discounted at a rate of 3.5% per year as recommended by NICE.

The risk of recurrence and progression in patients with NMIBC was estimated using risk equations based on an analysis of 2,596 patients from seven EORTC^e trials (Sylvester et al. 2006). Patients are ‘scored’ based on a number of risk factors, such as number of tumours, tumour size, prior recurrence rate, T category, presence of CIS and grade. An individual's one year and five year risks of recurrence and progression can then be estimated based upon these scores.

Table 78 shows the three monthly risks of recurrence, progression and progression given recurrence applied for each of the risk groups in the base case analysis.

Table 78

Three monthly recurrence and progression risk applied in the model.

Death from other causes was captured using 2009-2011 life tables for England and Wales from the Office of National Statistics (ONS). These life tables give an estimate of the annual probability of death given a person's age and gender with the model assuming that 50% of patients were female and that the average age was 60 years old. These annual probabilities were converted to three-monthly probabilities for use in the model.

Follow-up strategies

The variations in the frequency of follow-up that were considered in the model are summarised in table 79.

Table 79

Follow up strategies.

In addition to these variations, the use of a urinary biomarker (FISH) or cytology as a safety net to detect recurrences at the time points that would normally be checked under current practice was also considered. The diagnostic accuracy of these tests as well as cystoscopy were estimated using data from the systematic review of the clinical evidence conducted for this guideline, with most data being sourced from a systematic review by Mowatt et al. 2010.

Costs and utilities

The majority of costs were sourced from NHS reference costs 2012/13 by applying tariffs associated with the appropriate HRG code. Drug costs were calculated using dosages from the British National Formulary (BNF) and unit cost information from the electronic market information tool (eMit). Where unit costs for drugs were not available from eMit, prices from the BNF were used. Resource use and cost information were obtained from the Personal Social Services Research Unit (PSSRU) and the advice of the GDG.

Base Case Results

The base case results of the analysis for are presented in table 80 for patients in each risk category. The results are shown in the ‘dominance rank’ format as it allows for the best overall strategy to be evaluated.

Table 80

Base case cost-effectiveness result using dominance rank.

It can be seen that the optimal strategy in low and intermediate risk patients is the reduced frequency strategy. This strategy is the least effective of all the strategies but the difference is marginal and because it is substantially cheaper than the other strategies it was found to be cost-effective overall.

In the case of high risk patients, it can be seen that the reduced frequency strategy is again the cheapest strategy but it is no longer the preferred strategy in cost-effectiveness terms.

Strategies of reduced frequency with a safety net using FISH or cytology were found to be more cost-effective than this strategy with the reduced frequency follow-up strategy with FISH found to be the most cost-effective (more cost-effective than cytology because of the superior sensitivity of FISH in the base case).

Sensitivity analysis

A series of one-way sensitivity analyses were conducted, whereby an input parameter is changed, the model is re-run and the new cost-effectiveness result is recorded. This analysis is a useful way of estimating uncertainty and determining the key drivers of the model result.

The analyses showed that, in low and intermediate risk patients, reduced frequency follow-up was the most cost-effective strategy in all modelled scenarios. In the case of high risk patients, the optimal strategy remains the same as in the base case (i.e. reduced frequency with FISH) in the vast majority of the analyses. However, there are two exceptions where the reduced frequency follow-up becomes the most cost-effective strategy; one where the modelled time horizon is reduced to five years and another where the bladder cancer specific mortality rates are equivalent for NMIBC and MIBC patients.

The GDG were also interested in an analysis where only variations in follow-up frequency were considered (i.e. variations in diagnostic tests were excluded from the analysis). As in the full analysis, it was found that the optimal strategy in low and intermediate risk patients was the reduced frequency strategy. However, in the case of high risk patients, the cystoscopy frequency used in current practice was found to be the most cost-effective strategy with a cost per QALY of £9,487 in comparison to the next based strategy (Slightly reduced follow-up).

A probabilistic sensitivity analysis was also conducted to assess the combined parameter uncertainty in the model. In this analysis, the mean values that were utilised in the base case were replaced with values drawn from distributions around the mean values. It was found that, at a threshold of £20,000 per QALY, the reduced frequency follow-up strategy had a 98% and 91% probability of being cost-effective in the low and intermediate risk group, respectively. In high risk patients it was found that, at a threshold of £20,000 per QALY, the reduced follow-up strategy in combination with FISH had a 79% probability of being cost-effective.

Conclusion

The results of the analysis suggest that reducing the frequency of cystoscopic follow-up in low and intermediate risk patients is cost-effective. Furthermore, the results show that the addition of cytology or FISH as a safety net was not cost-effective in these risk groups. In high risk patients, the results of the analysis suggest that reducing cystoscopic follow-up alone is not cost-effective in comparison to current practice. However, the addition of cytology or FISH as a safety net was found to be cost-effective with a reduced frequency follow-up strategy with FISH found to be the most cost-effective strategy.

However, there are concerns about the lack of comparative data that investigates variations in follow-up and further research is required to fully assess the safety, effectiveness and cost-effectiveness of the proposed follow-up strategies.

Recommendations	Offer people with low-risk non-muscle-invasive bladder cancer cystoscopic follow-up 3 months and 12 months after diagnosis. Discharge to primary care people who have had low-risk non-muscle-invasive bladder cancer and who have no recurrence of the bladder cancer within 12 months. Do not offer routine urinary cytology or prolonged cystoscopic follow-up after 12 months for people with low-risk non-muscle-invasive bladder cancer. Offer people with intermediate-risk non-muscle-invasive bladder cancer cystoscopic follow-up at 3, 9 and 18 months, and once a year thereafter. Consider discharging people who have had intermediate-risk non-muscle-invasive bladder cancer to primary care after 5 years of disease-free follow-up. Offer people with high-risk non-muscle-invasive bladder cancer cystoscopic follow up: every 3 months for the first 2 years then every 6 months for the next 2 years then once a year thereafter. Refer people urgently to urological services if they have haematuria or other urinary symptoms and a history of non-muscle-invasive bladder cancer.
Relative value placed on the outcomes considered	The GDG considered the following outcomes to be important: Progression is associated with morbidity, mortality and cost and is readily captured; Disease specific survival and overall survival are important outcomes because it is important not to have avoidable death; Quality of life is important because it captures the patient experience of both the intervention and the disease. Patient preference, treatment-related complications, and health-related quality of life were specified as outcomes in the PICO but were not reported in the evidence. No further outcomes were used to make recommendations.
Quality of the evidence	The evidence was assessed as being of low to moderate quality using GRADE The evidence was limited by a general lack of high quality evidence. Many of the included studies were old studies and had small sample sizes, low number of events and different patient populations. The GDG considered that there was insufficient evidence to be able to support recommendations for radical changes to follow-up for patients with high-risk bladder cancer. For low and intermediate risk groups, the clinical experience of the group and the limited evidence available were felt to be sufficient to make recommendations for a change in practice. A research recommendation was made although there was a suggestion in the cost-effectiveness model that changes in follow-up in patients with high-risk disease could be safe and cost-effective. However, as there was no robust evidence in clinical practice the GDG did not feel that it could be introduced as a new standard of care and so felt that a research recommendation was appropriate. The recommendation in patients with high-risk disease results from the group's consensus estimation of conservative current practice supported by the economic model. The research recommendation sought to assess new models of follow-up.
Trade-off between clinical benefits and harms	The potential benefits of the recommendation for patients with low risk disease result from the reduced burden of cystoscopic follow-up. The GDG balanced this against the potential for harm resulting from a possible small increase in the late detection of disease recurrence and that patients may experience anxiety after discharge from follow-up. The GDG considered that reducing the burden of follow-up strongly outweighs the possible increase in late detection of recurrence. For patients with intermediate and high-risk disease, benefits may result from the wider implementation of standard practice (reduction in variation in practice), more effective identification of progression, and decreased patient anxiety from more frequent follow-up. The GDG balanced this against the possible increase in morbidity associated with cystoscopies and an increase in patient anxiety from an increased number of cystoscopies. The GDG prioritised reduction in variation in practice. The GDG also considered that minimising progression is a priority in these groups due to the adverse impact of progressive disease on patient health.
Trade-off between net health benefits and resource use	A health economic model was developed for this topic. The results of the economic analysis showed that the optimal follow-up strategy varied in each risk group: Low and intermediate risk Reduced frequency follow-up was shown to be the most cost-effective strategy in low and intermediate risk patients. It was less effective in QALY terms than the other strategies but substantially cheaper and so overall the strategy was found to be cost-effective (i.e. all other strategies have ICER > £20,000 per QALY in comparison to reduced frequency follow-up). High risk FISH with reduced frequency follow-up was shown to be the most cost-effective strategy in high risk patients. It was found to be one of the cheapest strategies and the most effective in QALY terms. In the dominance rank, it was shown to have an ICER of £5,095 per QALY in comparison to the next best strategy (cytology with reduced frequency). Owing to practical issues regarding the regular use of urinary biomarkers and cytology, the GDG were also interested in a sensitivity analysis where FISH and cytology were excluded (i.e. variations in frequency only). The results showed the current practice schedule to be the most cost-effective. It was found to be more expensive than reduced frequency schedules but was cost-effective with an ICER < £20,000 per QALY. The results of the economic model enabled the GDG to reduce the frequency and duration of cystoscopy in low and intermediate risk and informed the research recommendation in high risk patients. Overall, the GDG anticipated that the recommendations could have the following impact on costs: Low and intermediate risk Potential for increased costs associated with treating otherwise avoidable disease. Also, likely to be increased costs associated with follow-up by GPs. There will be substantial savings from reduced cystoscopic follow-up in low and intermediate risk patients High risk Potential for higher costs in some instances as the ‘current practice’ schedule may be more intensive than that used by some centres. The earlier detection of bladder cancer may lead to potential for savings through reduced treatment of advanced bladder cancer. Further savings could be made by substituting urinary tests for cystoscopy.
Other considerations	No equalities issues were identified. The GDG considered the potential change in practice resulting from these recommendations includes a substantial reduction in cystoscopic follow-up in low risk disease, an increased role in follow-up for GPs, and some reduction in cystoscopic follow-up for patients with intermediate risk disease. The GDG considered it difficult to assess the extent to change in practice required to implement the recommendation for patients with high-risk disease because of uncertainty over current practice. However, implementing the recommendations will require a risk assessment, which will be a change compared to current practice. The GDG were uncertain about what follow-up regimens are currently in place across the NHS. Strategies involving FISH were attractive in cost-effectiveness terms but there was uncertainty about their effectiveness as a substitute for cystoscopy and there was a concern about a lack of availability of the test within the NHS. After much debate, the GDG decided it was best to consider using urinary tests in a research setting rather than recommend immediate implementation. The GDG discussed how these recommendations could be audited and monitored, particularly in low risk patients.

Research recommendation

In people with high-risk non-muscle-invasive bladder cancer, are these follow-up regimens equally effective in terms of identification of progression, cost effectiveness and health-related quality of life?

Cystoscopic follow-up at 3, 6, 12, 18, 24, 36 and 48 months, and then annually, interspersed with non-invasive urinary tests.
Cystoscopic follow-up at 3, 6, 9, 12, 15, 18, 21, 24, 30, 36, 42 and 48 months and then annually thereafter.

Why is this important

Cystoscopy is currently the standard of care for follow-up of people with high-risk non-muscle-invasive bladder cancer. Regular cystoscopy may be associated with anxiety, procedural discomfort to the person and significant costs to the NHS.

Urine tests based on a variety of technologies (including cytology, fluorescence in-situ hybridization [FISH] and proteomic platforms) can detect high-grade recurrence, raising the possibility that 1 or more of these tests could be used to reduce the frequency of cystoscopy. This could improve acceptability to patients and reduce costs to the NHS without increasing the risk of disease progression.

There is a lack of evidence on the optimal frequency of follow-up and whether the frequency of cystoscopy follow-up can safely be reduced by substitution of urinary tests.

4.5. References

Abern MR, et al. Perioperative intravesical chemotherapy in non-muscle-invasive bladder cancer: a systematic review and meta-analysis. Journal of the National Comprehensive Cancer Network. 2013;11(4):477–84. [PubMed: 23584348]
Addeo R, et al. Randomized phase III trial on gemcitabine versus mytomicin in recurrent superficial bladder cancer: evaluation of efficacy and tolerance. Journal of clinical oncology: official journal of the American Society of Clinical Oncology. 2010;28(4):543–548. [PubMed: 19841330]
Agrawal MS, et al. The safety and efficacy of different doses of bacillus Calmette Guérin in superficial bladder transitional cell carcinoma. Urology. 2007;70:1075–1078. [PubMed: 18158020]
Al Khalifa M, et al. The effect of isoniazid on BCG-induced toxicity in patients with superficial bladder cancer. European Urology. 2000;37(Suppl 1):26–30. [PubMed: 10575269]
Ali-El-Dein B, et al. Survival after primary and deferred cystectomy for stage T1 transitional cell carcinoma of the bladder. Urology annals. 2011;3(3):127–132. [PMC free article: PMC3183703] [PubMed: 21976924]
Alkibay T, et al. Micropapillary pattern in urothelial carcinoma: a clinicopathological analysis. Urologia Internationalis. 2009;83(3):300–305. [PubMed: 19829030]
Altieri VM, et al. Recurrence and progression in non-muscle-invasive bladder cancer using EORTC risk tables. Urologia Internationalis. 2012;89(1):61–66. [PubMed: 22722366]
Badalament RA, et al. A prospective randomized trial of maintenance versus nonmaintenance intravesical bacillus Calmette-Guerin therapy of superficial bladder cancer. Journal of Clinical Oncology. 1987;5(3):441–449. [PubMed: 3546618]
Badalato GM, et al. Immediate radical cystectomy vs conservative management for high grade cT1 bladder cancer: is there a survival difference? BJU International. 2012;110(10):1471–1477. [PubMed: 22487512]
Biers SM, Mostafid AH. Electromotive drug administration of local anesthesia for biopsy and cystodiathermy of recurrent low grade bladder tumors. Current Urology. 2009;3(1):15–18.
Bohle A, Bock PR. Intravesical bacille Calmette-Guerin versus mitomycin C in superficial bladder cancer: formal meta-analysis of comparative studies on tumor progression. Urology. 2004;63(4):682–686. [PubMed: 15072879]
Bohle A, et al. Intravesical Bacillus Calmette-Guerin versus mitomycin C for superficial bladder cancer: a formal meta-analysis of comparative studies on recurrence and toxicity. Journal of Urology. 2003;169(1):90–95. [PubMed: 12478111]
Brausi M, et al. Side Effects of Bacillus Calmette-Guerin (BCG) in the Treatment of Intermediate- and High-risk Ta, T1 Papillary Carcinoma of the Bladder: Results of the EORTC Genito-Urinary Cancers Group Randomised Phase 3 Study Comparing One-third Dose with Full Dose and 1 Year with 3 Years of Maintenance BCG. European Urology. 2014;65(1):69–76. [PubMed: 23910233]
Brimo F, et al. Prognostic factors in T1 bladder urothelial carcinoma: the value of recording millimetric depth of invasion, diameter of invasive carcinoma, and muscularis mucosa invasion. Human Pathology. 2013;44(1):95–102. [PubMed: 22939956]
Canales BK, et al. Risk factors for upper tract recurrence in patients undergoing long-term surveillance for stage ta bladder cancer. Journal of Urology. 2006;175(1):74–77. [PubMed: 16406874]
Cheng L, et al. Survival of patients with carcinoma in situ of the urinary bladder. Cancer. 1999;85:2469–2474. [PubMed: 10357420]
Colombel M, et al. The effect of ofloxacin on bacillus calmette-guerin induced toxicity in patients with superficial bladder cancer: results of a randomized, prospective, double-blind, placebo controlled, multicenter study. Journal of Urology. 2006;176(3):935–939. [PubMed: 16890660]
Corman JM, et al. Treatment of radiation induced hemorrhagic cystitis with hyperbaric oxygen. Journal of Urology. 2003;169(6):2200–2202. [PubMed: 12771749]
Dalbagni G, et al. Clinical outcome in a contemporary series of restaged patients with clinical T1 bladder cancer. European Urology. 2009;56(6):903–910. [PMC free article: PMC8177014] [PubMed: 19632765]
Davenport K, et al. Audit of safety, efficacy, and cost-effectiveness of local anaesthetic cystodiathermy. Annals of the Royal College of Surgeons of England. 2010;92(8):706–709. [PMC free article: PMC3229385] [PubMed: 20615299]
De Berardinis, et al. T1G3 high-risk NMIBC (non-muscle invasive bladder cancer): conservative treatment versus immediate cystectomy. International Urology & Nephrology. 2011;43(4):1047–1057. [PubMed: 21442469]
De Bekker-Grob EW, et al. “Non-muscle-invasive bladder cancer surveillance for which cystoscopy is partly replaced by microsatellite analysis of urine: a cost-effective alternative? (Provisional abstract)” BJU International. 2009;104(1):41–47. [PubMed: 19500328]
Del Pizzo JJ, et al. Treatment of radiation induced hemorrhagic cystitis with hyperbaric oxygen: Long-term followup. Journal of Urology. 1998;160(3):731–733. [PubMed: 9720533]
Denzinger S, et al. Early versus deferred cystectomy for initial high-risk pT1G3 urothelial carcinoma of the bladder: do risk factors define feasibility of bladder-sparing approach? European Urology. 2008;53(1):146–152. [PubMed: 17624657]
Di Lorenzo G, et al. Gemcitabine versus bacille Calmette-Guerin after initial bacille Calmette-Guerin failure in non-muscle-invasive bladder cancer: a multicenter prospective randomized trial. Cancer. 2010;116(8):1893–1900. [PubMed: 20162706]
Divrik RT, et al. Impact of routine second transurethral resection on the long-term outcome of patients with newly diagnosed pT1 urothelial carcinoma with respect to recurrence, progression rate, and disease-specific survival: a prospective randomised clinical trial. European Urology. 2010;58(2):185–190. [PubMed: 20303646]
Donat SM, et al. Efficacy of office fulguration for recurrent low grade papillary bladder tumors less than 0.5 cm. Journal of Urology. 2004;171(2:Pt 1):636–639. [PubMed: 14713776]
Eto H, et al. Comparison of the prophylactic usefulness of epirubicin and doxorubicin in the treatment of superficial bladder cancer by intravesical instillation: a multicenter randomized trial. Cancer Chemotherapy & Pharmacology. 1994 35 Suppl:S46–S51. [PubMed: 7994786]
Fernandez-Gomez J, et al. The EORTC tables overestimate the risk of recurrence and progression in patients with non-muscle-invasive bladder cancer treated with bacillus Calmette-Guerin: external validation of the EORTC risk tables. European Urology. 2011;60(3):423–430. [PubMed: 21621906]
Fritsche HM, et al. Characteristics and outcomes of patients with clinical T1 grade 3 urothelial carcinoma treated with radical cystectomy: results from an international cohort. European Urology. 2010;57(2):300–309. [PubMed: 19766384]
Gacci M, et al. Intravesical gemcitabine in BCG-refractory T1G3 transitional cell carcinoma of the bladder: a pilot study. Urologia Internationalis. 2006;76(2):106–111. [PubMed: 16493208]
Green DA, et al. Cost-effective treatment of low-risk carcinoma not invading bladder muscle. BJU Int. 2013;111(3B):E78–E83. [PubMed: 22958598]
Han RF, Pan JG. Can intravesical bacillus Calmette-Guerin reduce recurrence in patients with superficial bladder cancer: a meta-analysis of randomized trials. Urology. 2006;67:1216–1223. [PubMed: 16765182]
Harland SJ, et al. A randomized trial of radical radiotherapy for the management of pT1G3 NXM0 transitional cell carcinoma of the bladder. The Journal of urology. 2007;178:807–813. [PubMed: 17631326]
Hautmann RE, et al. Quantification of the survival benefit of early versus deferred cystectomy in high-risk non-muscle invasive bladder cancer (T1 G3). World Journal of Urology. 2009;27(3):347–351. [PubMed: 19319539]
Hernandez V, et al. External validation and applicability of the EORTC risk tables for non-muscle-invasive bladder cancer. World Journal of Urology. 2011;29(4):409–414. [PubMed: 21190023]
Hession P, et al. Intravenous urography in urinary tract surveillance in carcinoma of the bladder. Clinical Radiology. 1999;54(7):465–467. [PubMed: 10437700]
Hinotsu S, et al. Maintenance therapy with bacillus Calmette-Guerin Connaught strain clearly prolongs recurrence-free survival following transurethral resection of bladder tumour for non-muscle-invasive bladder cancer. BJU International. 2011;108(2):187–195. [PubMed: 21176079]
Holmang S, Johansson SL. Stage Ta-T1 bladder cancer: the relationship between findings at first followup cystoscopy and subsequent recurrence and progression. Journal of Urology. 2002;167(4):1634–1637. [PubMed: 11912378]
Holmang S, Strock V. Should follow-up cystoscopy in bacillus Calmette-Guerin-treated patients continue after five tumour-free years? European Urology. 2012;61(3):503–507. [PubMed: 22119022]
Holmang S, et al. Long-term followup of a bladder carcinoma cohort: routine followup urography is not necessary. Journal of Urology. 1998;160(1):45–48. [PubMed: 9628602]
Houghton B, et al. Intravesical chemotherapy plus BCG in non-muscle invasive bladder cancer. A systematic review with meta-analysis. BJU International. 2012;111:977–983. [PubMed: 23253618]
Hudson MA, et al. Single course versus maintenance bacillus Calmette-Guerin therapy for superficial bladder tumors: a prospective, randomized trial. Journal of Urology. 1987;138(2):295–298. [PubMed: 3298694]
Huncharek M, Kupelnick B. Impact of intravesical chemotherapy versus BCG immunotherapy on recurrence of superficial transitional cell carcinoma of the bladder: metaanalytic reevaluation. American Journal of Clinical Oncology. 2003;26(4):402–407. [PubMed: 12902895]
Huncharek M, Kupelnick B. The influence of intravesical therapy on progression of superficial transitional cell carcinoma of the bladder: a metaanalytic comparison of chemotherapy versus bacilli Calmette-Guerin immunotherapy. American Journal of Clinical Oncology. 2004;27:522–528. [PubMed: 15596924]
Huncharek M, et al. Intravesical chemotherapy prophylaxis in primary superficial bladder cancer: a meta-analysis of 3703 patients from 11 randomized trials. Journal of Clinical Epidemiology. 2000;53:676–680. [PubMed: 10941943]
Huncharek M, et al. Impact of intravesical chemotherapy on recurrence rate of recurrent superficial transitional cell carcinoma of the bladder: results of a meta-analysis. Anticancer Research. 2001;21:765–769. [PubMed: 11299841]
Johnson MH, et al. Randomized controlled trial of oxybutynin extended release versus placebo for urinary symptoms during intravesical Bacillus Calmette-Guerin treatment. Journal of Urology. 2013;189(4):1268–1274. [PMC free article: PMC4150693] [PubMed: 23123375]
Jones G, et al. Intravesical gemcitabine for non-muscle invasive bladder cancer. Cochrane Database of Systematic Reviews. 2012;(Issue 1) Art. No.: CD009294. [PubMed: 22259002]
Kamat AM, et al. The case for early cystectomy in the treatment of nonmuscle invasive micropapillary bladder carcinoma. Journal of Urology. 2006;175(3):881–885. [PubMed: 16469571]
Kim W, et al. Value of immediate second resection of the tumor bed to improve the effectiveness of transurethral resection of bladder tumor. Journal of Endourology. 2012;26:1059–1064. [PubMed: 22390720]
Koga H, et al. Maintenance intravesical bacillus Calmette-Guérin instillation for Ta, T1 cancer and carcinoma in situ of the bladder: randomized controlled trial by the BCG Tokyo Strain Study Group. International Journal of Urology. 2010;17(9):759–766. [PubMed: 20604814]
Kulkarni GS, et al. Cost-effectiveness analysis of immediate radical cystectomy versus intravesical Bacillus Calmette-Guerin therapy for high-risk, high-grade (T1G3) bladder cancer. Cancer. 2009;115(23):5450–59. (Structured abstract) [PubMed: 19685529]
Kumar S, et al. Intravesical Formalin for Control of Intractable Bladder Hemorrhage Secondary to Cystitis Or Cancer. Journal of Urology. 1975;114(4):540–543. [PubMed: 1235376]
Lamm DL, et al. Maintenance bacillus Calmette-Guerin immunotherapy for recurrent TA, T1 and carcinoma in situ transitional cell carcinoma of the bladder: a randomized Southwest Oncology Group Study. Journal of Urology. 2000;163(4):1124–1129. [PubMed: 10737480]
Lammers RJ, et al. Comparison of expected treatment outcome provided by risk models and international guidelines with observed treatment outcome in a cohort of Dutch non-muscle-invasive bladder cancer patients treated with intravesical chemotherapy. BJU Int. 2014. in press. [PubMed: 24304638]
LeBlanc B, et al. Long-term followup of initial Ta grade 1 transitional cell carcinoma of the bladder. Journal of Urology. 1999;162(6):1946–1950. [PubMed: 10569544]
Lee HC, et al. Hyperbaric-Oxygen Therapy in Hemorrhagic Radiation Cystitis - A Report of 20 Cases. Undersea & Hyperbaric Medicine. 1994;21(3):321–327. [PubMed: 7950806]
Lida S, et al. Clinical outcome of high-grade non-muscle-invasive bladder cancer: a long-term single center experience. International Journal of Urology. 2009;16(3):287–292. [PubMed: 19207115]
Likourinas M, et al. Intravesical Formalin for the Control of Intractable Bladder Hemorrhage Secondary to Radiation Cystitis Or Bladder-Cancer. Urological Research. 1979;7(2):125–126. [PubMed: 573008]
Mack D. Quality of life in patients undergoing bacille Calmette-Guerin therapy for superficial bladder cancer. British Journal of Urology. 1996;78(3):369–371. [PubMed: 8881944]
Malmstrom PU, et al. An individual patient data meta-analysis of the long-term outcome of randomised studies comparing intravesical mitomycin C versus Bacillus Calmette-Guerin for non-muscle-invasive bladder cancer. Europea Urology. 2009;56:247–256. [PubMed: 19409692]
Mangiarotti B, et al. A randomized prospective study of intravesical prophylaxis in non-musle invasive bladder cancer at intermediate risk of recurrence: mitomycin chemotherapy vs BCG immunotherapy. Archivio Italiano di Urologia, Andrologia. 2008;80(4):167–171. [PubMed: 19235434]
Mariappan P, Smith G. A surveillance schedule for G1Ta bladder cancer allowing efficient use of check cystoscopy and safe discharge at 5 years based on a 25-year prospective database. Journal of Urology. 2005;173(4):1108–1111. [PubMed: 15758711]
Martínez-Piñeiro JA, et al. Long-term follow-up of a randomized prospective trial comparing a standard 81 mg dose of intravesical bacille Calmette-Guérin with a reduced dose of 27 mg in superficial bladder cancer. BJU International. 2002;89:671–680. [PubMed: 11966623]
Martínez-Piñeiro JA, et al. Has a 3-fold decreased dose of bacillus Calmette-Guerin the same efficacy against recurrences and progression of T1G3 and Tis bladder tumors than the standard dose? Results of a prospective randomized trial. The Journal of Urology. 2005;174:1242–1247. [PubMed: 16145378]
Mathews R, et al. Hyperbaric oxygen therapy for radiation induced hemorrhagic cystitis. Journal of Urology. 1999;161(2):435–437. [PubMed: 9915420]
Matsumoto K, et al. Risk of subsequent tumour recurrence and stage progression in bacille Calmette-Guerin relapsing non-muscle-invasive bladder cancer. BJU International. 2012;110(11 Pt B):E508–E513. [PubMed: 22574662]
Miyake H, et al. Limited significance of routine excretory urography in the follow-up of patients with superficial bladder cancer after transurethral resection. BJU International. 2006;97(4):720–723. [PubMed: 16536761]
Miyake M, et al. Clinical significance of subepithelial growth patterns in non-muscle invasive bladder cancer. BMC Urology. 2011;11(1):17. [PMC free article: PMC3167754] [PubMed: 21816111]
Oddens J, et al. Final Results of an EORTC-GU Cancers Group Randomized Study of Maintenance Bacillus Calmette-Guerin in Intermediate- and High-risk Ta, T1 Papillary Carcinoma of the Urinary Bladder: One-third Dose Versus Full Dose and 1 Year Versus 3 Years of Maintenance. European Urology. 2013;63(3):462–472. [PubMed: 23141049]
Oge O, et al. Proposal for changes in cystoscopic follow-up of patients with low-grade pTa bladder tumor. European Urology. 2000;37(3):271–274. [PubMed: 10720851]
Ojea A, et al. A multicentre, randomised prospective trial comparing three intravesical adjuvant therapies for intermediate-risk superficial bladder cancer: low-dose bacillus Calmette-Guerin (27 mg) versus very low-dose bacillus Calmette-Guerin (13.5 mg) versus mitomycin C. European Urology. 2007;52:1398–1406. [PubMed: 17485161]
Olsen LH, Genster HG. Prolonging follow-up intervals for non-invasive bladder tumors: a randomized controlled trial. Scandinavian Journal of Urology and Nephrology. 1995;172:33–36. [PubMed: 8578253]
Oosterlinck W, et al. Sequential intravesical chemoimmunotherapy with mitomycin C and bacillus Calmette-Guérin and with bacillus Calmette-Guérin alone in patients with carcinoma in situ of the urinary bladder: results of an EORTC genito-urinary group randomized phase 2 trial (30993). European Urology. 2011;59:438–446. [PubMed: 21156335]
Palou J, et al. Control group and maintenance treatment with bacillus Calmette-Guerin for carcinoma in situ and/or high grade bladder tumors. Journal of Urology. 2001;165(5):1488–1491. [PubMed: 11342902]
Pan J, et al. A meta-analysis of randomized trials of maintenance bacillus Calmette-Guerin instillation efficacy against recurrence of T1G3 bladder tumour. Frontiers of Medicine China. 2008;2(3):259–263.
Pan J, et al. Can intravesical bacillus Calmette-Guerin reduce recurrence in patients with non-muscle invasive bladder cancer? An update and cumulative meta-analysis. Frontiers of Medicine. 2014;8(2):241–249. [PubMed: 24810644]
Park DS, et al. An Analysis of the Efficacy, Safety, and Cost-Effectiveness of Fulguration Under Local Anesthesia for Small-Sized Recurrent Masses: A Comparative Analysis to Transurethral Resection of Bladder Tumors in a Matched Cohort. Journal of Endourology. 2013;27(10):1240–1244. [PubMed: 23964922]
Park J, et al. Prognostic significance of non-papillary tumor morphology as a predictor of cancer progression and survival in patients with primary T1G3 bladder cancer. World Journal of Urology. 2009;27(2):277–283. [PubMed: 19020879]
Parra C, et al. Management of Hemorrhagic Radiation Cystitis with Hyperbaric Oxygen Therapy. Actas Urologicas Espanolas. 2011;35(3):175–179. [PubMed: 21334100]
Patard J, et al. Tumor progression and survival in patients with T1G3 bladder tumors: multicentric retrospective study comparing 94 patients treated during 17 years. Urology. 2001;58(4):551–556. [PubMed: 11597537]
Pawinski A, et al. A combined analysis of European Organization for Research and Treatment of Cancer, and Medical Research Council randomized clinical trials for the prophylactic treatment of stage TaT1 bladder cancer. Journal of Urology. 1996;156(6):1934–1940. [PubMed: 8911360]
Prasad SM, et al. Durability of response: The achilles heel of salvage combination immunotherapy with intravesical bacillus calmette-guerin and interferon-alpha 2B in bladder cancer. Journal of Urology. 2009;181(4):72–72.
Rijkmans BG, et al. Successful treatment of radiation cystitis with hyperbaric oxygen. European Urology. 1989;16(5):354–356. [PubMed: 2776806]
Saika T, et al. Two instillations of epirubicin as prophylaxis for recurrence after transurethral resection of Ta and T1 transitional cell bladder cancer: a prospective, randomized controlled study. World Journal of Urology. 2010;28:413–418. [PubMed: 20054553]
Scosyrev E, et al. Urothelial carcinoma versus squamous cell carcinoma of bladder: is survival different with stage adjustment? Urology. 2009;73(4):822–827. [PubMed: 19193403]
Seo KW, et al. The efficacy of the EORTC scoring system and risk tables for the prediction of recurrence and progression of non-muscle-invasive bladder cancer after intravesical bacillus calmette-guerin instillation. Korean Journal of Urology. 2010;51(3):165–170. [PMC free article: PMC2855454] [PubMed: 20414391]
Serretta V, et al. A 1-year maintenance after early adjuvant intravesical chemotherapy has a limited efficacy in preventing recurrence of intermediate risk non-muscle-invasive bladder cancer. BJU International. 2010;106(2):212–217. [PubMed: 20070299]
Shang PF, et al. Intravesical Bacillus Calmette-Guérin versus epirubicin for Ta and T1 bladder cancer. Cochrane Database of Systematic Reviews. 2011;(5) Art. No.: CD006885. [PubMed: 21563157]
Shelley M, et al. Intravesical Bacillus Calmette-Guérin in Ta and T1 bladder cancer. Cochrane Database of Systematic Reviews. 2000;(4) Art. No.: CD001986. [PMC free article: PMC7017976] [PubMed: 11034738]
Shuin T, et al. A phase II study of prophylactic intravesical chemotherapy with 4'-epirubicin in recurrent superficial bladder cancer: comparison of 4'-epirubicin and adriamycin. Cancer Chemotherapy & Pharmacology. 1994 35 Suppl:S52–S56. [PubMed: 7994787]
Sommariva ML, et al. Efficacy of sodium hyaluronate in the management of chemical and radiation cystitis. Minerva Urologica e Nefrologica. 2010;62(2):145–150. [PubMed: 20562794]
Stamatiou K, et al. Accuracy of modern ultrasonographic techniques in the follow up of patients with superficial bladder carcinoma. Medical Ultrasonography. 2011;13(2):114–119. [PubMed: 21655537]
Sternberg IA, et al. Upper tract imaging surveillance is not effective in diagnosing upper tract recurrence in patients followed for nonmuscle invasive bladder cancer. Journal of Urology. 2013;190(4):1187–1191. [PubMed: 23680310]
Syed HA, et al. Holmium:YAG laser treatment of recurrent superficial bladder carcinoma: initial clinical experience. Journal of Endourology. 2001;15(6):625–627. [PubMed: 11552789]
Syed HA, et al. Flexible cystoscopy and Holmium:Yttrium aluminum garnet laser ablation for recurrent nonmuscle invasive bladder carcinoma under local anesthesia. Journal of Endourology. 2013;27(7):886–891. [PubMed: 23537221]
Sylvester RJ, et al. Intravesical bacillus Calmette-Guerin reduces the risk of progression in patients with superficial bladder cancer: a meta-analysis of the published results of randomized clinical trials. Journal of Urology. 2002;168(5):1964–1970. [PubMed: 12394686]
Sylvester RJ, et al. A single immediate postoperative instillation of chemotherapy decreases the risk of recurrence in patients with stage Ta T1 bladder cancer: a meta-analysis of published results of randomized clinical trials. Journal of Urology. 2004;171(6 Pt 1):2186–2190. [PubMed: 15126782]
Sylvester RJ, et al. Bacillus calmette-guerin versus chemotherapy for the intravesical treatment of patients with carcinoma in situ of the bladder: a meta-analysis of the published results of randomized clinical trials. Journal of Urology. 2005;174(1):86–91. [PubMed: 15947584]
Sylvester RJ, et al. Predicting recurrence and progression in individual patients with stage Ta T1 bladder cancer using EORTC risk tables: a combined analysis of 2596 patients from seven EORTC trials. European Urology. 2006;49(3):466–475. [PubMed: 16442208]
Sylvester RJ, et al. The schedule and duration of intravesical chemotherapy in patients with non-muscle-invasive bladder cancer: a systematic review of the published results of randomized clinical trials. European Urology. 2008;53(4):709–719. [PMC free article: PMC2587437] [PubMed: 18207317]
Sylvester RJ, et al. Long-term efficacy results of EORTC genito-urinary group randomized phase 3 study 30911 comparing intravesical instillations of epirubicin, bacillus Calmette-Guérin, and bacillus Calmette-Guérin plus isoniazid in patients with intermediate- and high-risk stage Ta T1 urothelial carcinoma of the bladder. Eur Urol. 2010;57(5):766–73. [PMC free article: PMC2889174] [PubMed: 20034729]
Thalmann GN, et al. Primary T1G3 bladder cancer: organ preserving approach or immediate cystectomy? Journal of Urology. 2004;172(1):70–75. [PubMed: 15201740]
Thompson RA Jr, et al. Late invasive recurrence despite long-term surveillance for superficial bladder cancer. Journal of Urology. 1993;149(5):1010–1011. [PubMed: 8483201]
Tilki D, et al. Lymphovascular invasion is independently associated with bladder cancer recurrence and survival in patients with final stage T1 disease and negative lymph nodes after radical cystectomy. BJU International. 2013;111(8):1215–1221. [PubMed: 23181623]
Türkeri L, et al. Comparison of the efficacy of single or double intravesical epirubicin instillation in the early postoperative period to prevent recurrences in non-muscle-invasive urothelial carcinoma of the bladder: prospective, randomized multicenter study. Urologia Internationalis. 2010;85:261–265. [PubMed: 20332605]
Vallancien G, et al. Can transabdominal ultrasonography of the bladder replace cystoscopy in the followup of superficial bladder tumors? Journal of Urology. 1986;136(1):32–34. [PubMed: 3520026]
van den Bosch S, Alfred Witjes J. Long-term cancer-specific survival in patients with high-risk, non-muscle-invasive bladder cancer and tumour progression: a systematic review. European Urology. 2011;60(3):493–500. [PubMed: 21664041]
van der Aa MN, et al. Patients' perceived burden of cystoscopic and urinary surveillance of bladder cancer: a randomized comparison. BJU International. 2008;101:1106–1110. [PubMed: 17888042]
van der Meijden AP, et al. Intravesical instillation of epirubicin, bacillus Calmette-Guerin and bacillus Calmette-Guerin plus isoniazid for intermediate and high risk Ta, T1 papillary carcinoma of the bladder: a European Organization for Research and Treatment of Cancer genito-urinary group randomized phase III trial. Journal of Urology. 2001;166(2):476–481. [PubMed: 11458050]
van Kessel KEM. “FGFR3 mutation analysis in voided urine samples to decrease cystoscopies and cost in nonmuscle invasive bladder cancer surveillance: A comparison of 3 strategies” Journal of Urology. 2013;189(5):1676–81. [PubMed: 23142690]
van Rhijn BW, et al. Molecular grade (FGFR3/MIB-1) and EORTC risk scores are predictive in primary non-muscle-invasive bladder cancer. European Urology. 2010;58(3):433–441. [PubMed: 20646825]
Vianello A, et al. Repeated white light transurethral resection of the bladder in nonmuscle-invasive urothelial bladder cancers: systematic review and meta-analysis. Journal of Endourology. 2011;25:1703–1712. [PubMed: 21936670]
Vriesema JL, et al. Patient opinion of urinary tests versus flexible urethrocystoscopy in follow-up examination for superficial bladder cancer: a utility analysis. Urology. 2000;56(5):793–797. [PubMed: 11068304]
Weiss JP, et al. Primary-Treatment of Radiation-Induced Hemorrhagic Cystitis with Hyperbaric-Oxygen - 10-Year Experience. Journal of Urology. 1994;151(6):1514–1517. [PubMed: 8189559]
Wong KA, et al. Outpatient laser ablation of non-muscle-invasive bladder cancer: is it safe, tolerable and cost-effective? BJU International. 2013;112(5):561–567. [PubMed: 23819486]
Wong SW, et al. Immediate versus delayed cystectomy for high-grade PT1 Transitional Cell Carcinoma of the bladder. BJU International Conference. 2009;(103 Suppl 4):2–9.
Xu T, et al. Predicting recurrence and progression in Chinese patients with nonmuscle-invasive bladder cancer using EORTC and CUETO scoring models. Urology. 2013;82(2):387–393. [PubMed: 23759377]
Yalcinkaya F, et al. Prospective randomized comparison of intravesical BCG therapy with standard dose versus low doses in superficial bladder cancer. International Urology & Nephrology. 1998;30(1):41–44. [PubMed: 9569110]
Yossepowitch O, et al. Use of urinary biomarkers for bladder cancer surveillance: patient perspectives. Journal of Urology. 2007;177(4):1277–1282. [PubMed: 17382711]
Zhang Y, et al. Comparison of surveillance strategies for low-risk bladder cancer patients. Medical Decision Making. 2013;33(2):198–214. [PubMed: 23178638]
Zieger K, et al. Long-term follow-up of noninvasive bladder tumours (stage Ta): recurrence and progression. BJU International. 2000;85(7):824–828. [PubMed: 10792160]

Footnotes

a: However, it should be noted that there is no official cost-effectiveness threshold used in the evaluation of treatments in the US health care system.
b: European Organisation for Research and Treatment of Cancer
c: Note that an approximate figure is used as two figures are presented for cost-effectiveness probability in the study (81.49% and 84.1%).
d: It should be noted that, while none of the above studies met the requirements for inclusion in the systematic review, they were nonetheless informative in helping to develop our own de novo economic model.
e: European Organisation for Research and Treatment of Cancer

Bookshelf ID: NBK356304

Contents

< Prev Next >

PubReader
Print View
Cite this Page
National Collaborating Centre for Cancer (UK). Bladder Cancer: Diagnosis and Management. London: National Institute for Health and Care Excellence (NICE); 2015 Feb. (NICE Guideline, No. 2.) 4, Managing non-muscle-invasive bladder cancer.
PDF version of this title (7.1M)

Related information

PMC
PubMed Central citations
PubMed
Links to PubMed

Recent Activity

Clear Turn Off Turn On

Managing non-muscle-invasive bladder cancer - Bladder Cancer
Managing non-muscle-invasive bladder cancer - Bladder Cancer
UNVERIFIED: Phortica variegata even-skipped-like gene, partial sequence
UNVERIFIED: Phortica variegata even-skipped-like gene, partial sequence
gi|2086095281|gb|MW425273.1|
Nucleotide
UNVERIFIED: Drosophila paraguayensis even-skipped-like gene, partial sequence
UNVERIFIED: Drosophila paraguayensis even-skipped-like gene, partial sequence
gi|2086095273|gb|MW425265.1|
Nucleotide
hypothetical protein, partial [Drosophila-associated salivary gland hypertrophy ...
hypothetical protein, partial [Drosophila-associated salivary gland hypertrophy virus]
gi|2060245492|gb|QXE96920.1|
Protein
Marmota flaviventris
Marmota flaviventris
Marmota flaviventris RefSeq Genome sequencing and assembly
BioProject

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on