Intravescial BCG treatment is indicated in patients with intermediate and high risk NMIBC but not in patients with low risk NMIBC (single low grade Ta tumor <3 cm). Intermediate risk disease is defined as low grade tumor that is either greater than 3 cm, or multifocal/recurrent, while high risk tumor is defined as any T1 or high grade tumor, including carcinoma in situ.

Although the true optimal regimen for BCG remains unknown, the original 6 week induction continues to be used fairly universally [4]. Most of the controversy regarding the therapeutic regimen relates to maintenance therapy. Despite anecdotal claims to the contrary [9], the evidence is clear that maintenance therapy reduces the risk of recurrence and progression [10]. The best data supports the use of the full Southwest Oncology Group (SWOG) 8507 regimen, with 3 weekly doses of BCG at 3, 6, 12, 18, 24, 30 and 36 months after induction [5, 11]. The more recent European Organization of Research and Treatment of Cancer (EORTC) 30962 trial comparing 1 versus 3 years of maintenance and 1/3 dose versus full dose BCG, suggested that full dose BCG over 3 years is better than reduced dose or shortened maintenance in high risk NMIBC, but 1 year of full dose BCG is likely adequate for intermediate risk NMIBC [6].

More than ten different BCG strains are used around the world. These can be categorized into early strains (Russian, Moreau, Tokyo, Sweden and Birkhaug, isolated before 1930) and late strains (Danish, Glaxo, Tice, Connaught, Phipps, Frappier, Prague, RIVM and Pasteur, isolated after 1930). With the present molecular biology techniques these strains were sequenced completely and differences in the genetic variability have been identified. In addition, variations in lipid and glycolipid content of the mycobacterial cell wall have been described. Some of these differences may or may not be related to BCG viability, efficacy for inducing tumor growth inhibition and triggering cytokine production [12]. The Connaught and Tice strains are the most widely administered in North America and Europe. Although they are considered to be biosimilar, they may in fact have variable efficacy, as demonstrated in multiple in vitro and in vivo animal models, as well as in clinical trials. In a recent prospective randomized phase III trial Rentsch et al. demonstrated a stronger immune response and a better recurrence-free survival (RFS) with BCG Connaught compared to BCG Tice (5-year RFS: 74 % vs. 48 %) [13]. The efficacy of BCG Tice was also inferior to BCG RIVM in an older study [14] (5-year RFS: 36 % vs. 54 %), even though this study was underpowered to detect statistically significant outcomes. The field still lacks convincing prospective validation studies to permit a final conclusion regarding the differential efficacy of available BCG strains.

Another critical unmet need in this context is a biomarker to guide patient selection for intravesical BCG and maintenance therapy. Various biomarkers (e.g. gene signatures, methylation status, interleukin-2 gene expression, urinary interleukin-8 and 18) have been investigated in order to predict efficacy of BCG treatment [15–18]. One with potential clinical utility is multi-color fluorescent in situ hybridization (FISH) [19], although none of these markers have been adopted in routine clinical practice.

Several randomized controlled trials have confirmed the superiority of BCG over intravesical chemotherapy in the prevention of tumor recurrence. Two meta-analyses comparing BCG to intravesical chemotherapy showed that BCG more effectively reduced the risk of recurrence, and only BCG with maintenance therapy was able to decrease the risk of progression [11, 20]. Compared to TURB alone, recurrences are decreased by 30 % with the addition of BCG maintenance and by 20 % with the addition of mitomycin maintenance. Although the efficacy of BCG seems to be superior to intravesical chemotherapy, it is associated with a higher rate of side effects. For this reason, and because the risk of progression is relatively low in intermediate disease intravesical chemotherapy remains a therapeutic option in patients with intermediate risk NMIBC.

Side effects can be classified as local or systemic (Table 9.1). Serious side effects are encountered in <5 % of patients and can be treated effectively in almost all cases [21]. In the original SWOG 8507 trial of maintenance therapy, only 16 % of patients completed 36 months of therapy due to either disease recurrence/progression or side effects. More recent trials have reported higher compliance rates, with approximately one quarter of patients delaying treatment and 20 % discontinuing treatment altogether due to side effects. As a result, several secondary measures have been investigated to reduce toxicity. To prevent adverse effects, BCG should not be administered during the first 2 weeks after TURB, in patients with gross hematuria, after traumatic catheterization and with symptomatic urinary tract infection. However, the presence of leukocyturia, microscopic hematuria or asymptomatic bacteriuria is not a contraindication of BCG administration. The recent EORTC trial demonstrated that dose reduction to 1/3 dose BCG had no benefit on toxicity, so that this measure should be avoided [6]. The concomitant use of antibiotics can reduce toxicity [24], but longer follow-up and larger patient cohorts would be necessary to demonstrate that there is no detrimental effect on disease outcome [25].

Several meta-analysis have confirmed that BCG after TURB is superior to TURB alone. Regardless of patient risk category, recurrences were reduced by a quarter with the combination of TURB and BCG induction when compared to TURB alone. This rate can be increased to a third with maintenance therapy. The absolute risk reduction in progression in the Sylvester meta-analysis of BCG therapy was 4 % (13.8 % without BCG versus 9.8 % with BCG) [20].

Several common definitions of BCG failure are defined in Table 9.2 [26, 27]. The management of patients who have failed BCG therapy is complex. The risk of pogression increases with each subsequent course of intravesical therapy, and no therapies have more than an approximate 20 % response rate at 1 year. Therefore, radical cystectomy is generally recommended for high risk disease that has failed BCG. Patients relapsing more than 12 months after prior BCG therapy can be re-challenged with intravesical BCG with reasonable results [28].

If bladder preservation is sought in BCG resistant and refractory disease, several strategies are now available [29]. BCG can be combined with interferon-alpha, although there is no evidence that this is any more efficacious than repeat BCG along. Small phase II trials suggest that intravesical gemcitabine and docetaxol are well-tolerated alternatives for BCG failure with measurable but modest activity. Electromotive delivery of mitomycin alternating with BCG over 12 months proved superior to a similar, non-standard course of BCG alone in a prospective randomized clinical trial involving patients with high-risk NMIBC but not prior intravesical BCG. This method is pending evaluation in a U.S. clinical trial prior to being approved by the FDA. Due to the low response rates and the poor durability of those responses, radical cystectomy remains the standard of care for patients failing intravesical BCG therapy.