BCG for Non Muscle Invasive Bladder Cancer
Clinical Question
In patients with non-muscle invasive bladder cancer does treatment with the Connaught strain of BCG compared with the TICE straight effect recurrence free survival?
Bottom Line
This study provides some evidence, albeit through a post-hoc analysis, that the Connaught and TICE strains of BCG provide a similar level of recurrence free survival in NMIBC patients
Major Points
BCG remains a main-stay of non-muscle invasive bladder cancer. The treatment is widely used around the world, but has been hampered by recent global shortages. A number of different strains of BCG exist, and this paper sets out to compare the efficacy of two of these strains, using a post-hoc analysis from a large phase II RCT. The strain of BCG used throughout the phase II study was determined by clinician preference and was given in combination with interferon. The study also considers outcomes in patients who are either BCG naive, or have previously failed BCG treatment.
Guidelines
As of January 2020, European Association of Urology (EAU) guidelines cite this paper as one piece of evidence that the strain of BCG used does not affect outcome. A further reference is made to a more extensive meta-analysis where multiple strains are compared.
Design
- Post-hoc analysis of phase II BCG/IFN-a2b trial
- N=901
- Connaught (n=292)
- TICE (n=609)
- Setting: 125 sites across the United States of America
- Enrollment: May 1999-May 2000
- Median follow-up: 24 months
- Analysis: post-hoc analysis of strain variants
- Primary outcome:Recurrence free survival
Population
Inclusion Criteria
- all forms of superficial transitional cell bladder cancer,
- cancer diagnosis could be primary or recurrent,
- cancer could be with or without carcinoma in situ (CIS),
- patients could have previous chemotherapy
- patients with previous BCG failures were eligible
- treatment had to commence 3-8 of confirmatory cystoscopy, biopsy or positive
cytology
- patients had to have had previous treatment by transurethral bladder tumor resection
Exclusion Criteria
Nil listed
Baseline Characteristics
For patients who were BCG naive (details of Connaught group provided, where p>0.05)
- Median age: 69 years
- sex: 22.6% female
- White: 96.4%
- TURBTs: <3 75.1%
- Duration: 0-2 years 84.6%
- Stage (Any CIS): 25% vs 17.8%, p=0.02
- Grade (low): 15.4%
- Tumor focality (pure CIS): 14.5% vs 7.6%, p=0.04
- Tumour size (<0.5cm): 9.2%
For patients who had previsouly had BCG (details of Connaught group provided, where p>0.05)
- Median age: 72 years
- Sex (female): 33.3 %
- White: 94.5%
- TURBTs: <3 20.9%vs 32.3%
- Duration: 0-2 years 64.2%
- Prior BCG courses (<=1): 64.6%
- Prior maintenance (No): 77.2% vs 57.4%
- Stage (Any CIS): 29%
- Grade (low): 23.2%
- Tumor focality (pure CIS): 18.7%
- Tumour size (<0.5cm): 19.4%
- BCG failure interval (<6 months): 65.7%
Interventions
This post-hoc analysis looked at the strain of BCG administered. The original trial setting is a comparison of BCG combined with interferon in the context of being BCG naïve or having a history of BCG failure.
Outcomes
Comparisons are intensive therapy vs. standard therapy.
Primary Outcomes
In multivariable analysis in patients who are BCG naïve: HR 1.21 95% CI 0.86-1.72, p=0.28
In multivariable analysis in patients who have a history of BCG failure: HR 1.11 95% CI 0.80-1.56, p=0.53
Secondary Outcomes
No secondary outcomes were reported
Subgroup Analysis
No subgroup analyses were reported
Adverse Events
Adverse events were not reported.
Criticisms
- This is a post-hoc analysis in an under powered cohort
- The number of patients included in the analysis does not match that in the original study
- Adverse events are not reported
- A head to head RCT would be better, but the paper cites a number of examples where attempts at this design have failed
Funding
Unrestricted educational grant from Schering-Plough