Biomedicine and Chemical Sciences

About three quarters of bladder cancers are non-muscle-invasive at diagnosis, and recurrence after transurethral resection (TURBT) is common. Intravesical BCG with maintenance is the recommended adjuvant treatment for high-risk disease and an option for intermediate-risk disease. An individual-patient-data meta-analysis of nine trials found that BCG reduced recurrence by 32% compared with mitomycin C (MMC) only when maintenance was given [1], and Cochrane reviews reached similar, low-certainty conclusions for MMC and epirubicin [2,3].

Two problems limit BCG in practice. Local and systemic toxicity leads many patients to stop treatment, which has prompted work on dose reduction and prophylaxis [4,5], and recurring global shortages mean that some patients cannot receive a full course. Recent reviews of first-line options in BCG-naïve disease have concentrated on adding immune checkpoint inhibitors or other agents to BCG [6,7]; these combinations improve recurrence-related outcomes at the cost of more severe adverse events, and they still depend on BCG supply.

Chemotherapy that replaces BCG is a different question. Earlier syntheses covered MMC, anthracyclines and single-agent gemcitabine [8,9,10], but newer randomised trials, including the first randomised comparison of sequential gemcitabine–docetaxel with BCG in a BCG-naïve population [11], have not been considered together with the older evidence. We therefore pooled RCTs comparing intravesical chemotherapy with BCG in BCG-naïve NMIBC, with recurrence as the primary outcome. The review had two objectives: first, to compare intravesical chemotherapy (single-agent or combination) with BCG; and second, to determine whether chemotherapy can be used to meet the demand created by BCG shortage. Regimens that combine chemotherapy with BCG were identified but not pooled with the chemotherapy-alone comparison (see Methods).

METHODS

The review follows the PRISMA 2020 statement [12].

2.1 Eligibility criteria

Trials in which the experimental arm also contained BCG (sequential or alternating chemo-immunotherapy) or interferon, and trials in BCG-failure populations, were excluded. These chemotherapy-plus-BCG trials were not pooled with the chemotherapy-alone comparison; they are considered in the Discussion as a possible BCG-sparing approach. Because older reports rarely describe prior BCG exposure, BCG-naïve status was classed as explicit (stated in eligibility criteria) or presumed (first-line adjuvant design with no mention of prior BCG); trials restricted to patients with prior BCG were excluded. Trials in which BCG-naïve status was not reported, or in which patients were not strictly BCG-naïve, were excluded from the primary analysis.

2.2 Search and selection

PubMed and Scopus was searched on 4 October 2026 without language limits, combining terms for BCG, bladder cancer, intravesical administration, the individual agents (mitomycin, epirubicin, doxorubicin, thiotepa, pirarubicin, gemcitabine, docetaxel, valrubicin), device-assisted delivery and randomised design. Reference lists of earlier systematic reviews were checked. Four structured queries and three author-targeted queries returned 364 records before de-duplication. Titles and abstracts were screened and potentially eligible reports were assessed against the criteria above.

2.3 Data extraction and effect measures

For each trial we recorded design, risk group, regimens, BCG schedule, follow-up, numbers randomised and outcome data. The effect measure for recurrence was the HR of chemotherapy versus BCG. When the HR and its confidence interval were not reported, ln(HR) and its standard error were derived by the methods of Parmar and Tierney [13,14], in this order of preference: reported HR with standard error from total events; log-rank or Cox p-value with total events; Kaplan–Meier survival at a fixed time with its confidence interval or p-value; and crude events per arm using the complementary log-log transformation. The method used for every trial is listed in Table 1 and in the data file. For multi-arm trials, BCG arms of different strains or schedules were combined; where two chemotherapy arms shared one BCG arm they were merged for the overall estimate and shown separately in agent-specific analyses.

2.4 Risk of bias

Risk of bias was judged with the RoB 2 tool [15] at the level of the recurrence outcome. Judgements are provisional because several trials were assessed from abstracts only.

2.5 Statistical analysis

Estimates were pooled on the log scale with the DerSimonian–Laird random-effects model [16]. Heterogeneity was described with I² and τ² [17], and Hartung–Knapp confidence intervals and 95% prediction intervals were calculated as a check. Pre-specified subgroup analyses covered chemotherapy agent, BCG maintenance in the comparator arm, explicit versus presumed BCG-naïve status and risk group. Sensitivity analyses excluded estimates that rested on assumptions or interim data (arm sizes assumed equal in two trials and a 12-month interim analysis in one), restricted the analysis to estimates derived from time-to-event statistics, substituted the 3-year MMC arm of one three-arm trial, and omitted each trial in turn. Dichotomous secondary outcomes were pooled as risk ratios (RR). Small-study effects were examined with a funnel plot and Egger's test [18]. All trials compared chemotherapy with BCG and none compared agents with each other, so the network is star-shaped; a network meta-analysis would add nothing to the agent-specific pairwise estimates and was not done. Analyses were run in Python 3 (NumPy, SciPy); forest plots are presented in Review Manager 5 layout, and an R script (meta package) that reproduces the primary analysis is provided with the data file.

RESULTS

3.1 Included trials

Thirty-four randomised trials met the design criteria. Thirteen were excluded because BCG-naïve status was not reported or not strict (eight trials of conventional chemotherapy, two of device-assisted MMC and three in CIS), leaving twenty-one eligible trials. Fourteen trials of conventional chemotherapy (3,227 participants) gave an extractable recurrence estimate: six of MMC, four of epirubicin, two of doxorubicin, one of thiotepa (the same trial as one of the doxorubicin trials, which had both doxorubicin and thiotepa arms), one of gemcitabine and one of gemcitabine–docetaxel (Table 1). Seven eligible trials could not be pooled because the accessible report gave no event counts or variance (Table 2). Trials combining chemotherapy with BCG [19,20,21] or with interferon [22,23] and a trial in BCG failure [24] were excluded.

BCG-naïve status was explicit in four trials and presumed in ten. The comparator was BCG with maintenance in eight trials and induction only in three; the schedule was mixed or unclear in three. Median follow-up ranged from 12 months to about 20 years. In the ten trials with presumed status, BCG-naïve status was inferred from a first-line adjuvant design.

Table 1. Trials contributing a recurrence estimate. HR is chemotherapy versus BCG (above 1 favours BCG).

Table 2. Eligible trials without an extractable effect size.

3.2 Risk of bias

No trial was blinded, and allocation concealment was rarely described. On provisional assessment, all trials were at high risk of bias or raised some concerns, and none was at low risk. The main reasons for a high-risk judgement were single-centre conduct with unclear randomisation, short follow-up, interim data, and effect sizes that had to be derived with assumptions.

3.3 Recurrence

Across all conventional regimens, chemotherapy was associated with a higher hazard of recurrence than BCG (HR 1.38 (95% CI 1.05–1.82), p = 0.02; Figure 2). Heterogeneity was substantial (I² = 81%, τ² = 0.20), the Hartung–Knapp interval was 0.98–1.96 and included 1, and the 95% prediction interval ran from 0.49 to 3.87, so the pooled figure is imprecise and should not be read as applying to every agent.

Results differed by agent (test for subgroup differences p = 0.002). BCG was clearly better than epirubicin (HR 2.07 (95% CI 1.36–3.13); I² = 75%, τ² = 0.13; four trials) and doxorubicin (HR 2.49 (95% CI 1.16–5.38)). For MMC the pooled estimate did not differ significantly from BCG (HR 1.05 (95% CI 0.76–1.45); I² = 73%, τ² = 0.12; six trials). The thiotepa arm of the trial that also tested doxorubicin favoured BCG (HR 3.06 (95% CI 1.39–6.73)) [40]. Single-agent gemcitabine was represented by one intermediate-risk trial, which gave a hazard similar to BCG with a wide interval (HR 0.81 (95% CI 0.35–1.87)) [41]. In the only randomised trial of gemcitabine–docetaxel, 12-month recurrence-free survival was 90.7% with gemcitabine–docetaxel and 77.1% with BCG [11], giving a derived HR of 0.38 (95% CI 0.12–1.18). The two gemcitabine-containing trials together gave HR 0.61 (95% CI 0.30–1.26).

Figure 2. Recurrence with conventional intravesical chemotherapy versus BCG, by agent. Random-effects model. The overall estimate uses the combined chemotherapy arms of Martínez-Piñeiro 1990.

3.4 Subgroup and sensitivity analyses

The BCG schedule explained much of the heterogeneity (p for subgroup difference < 0.001; Figure 3). Where BCG was given with maintenance, chemotherapy was inferior (HR 1.53 (95% CI 1.20–1.95); I² = 54%, τ² = 0.06; eight trials). Where BCG was given as a 6-week induction only, the estimate favoured chemotherapy; all three trials tested MMC (HR 0.77 (95% CI 0.62–0.96); I² = 0%, τ² = 0.00). Three trials with mixed or unclear schedules were not included in this comparison. The estimate was HR 1.23 (95% CI 0.77–1.97) in the four explicitly BCG-naïve trials and HR 1.47 (95% CI 1.03–2.11) in the ten trials with presumed BCG-naïve status (p for difference = 0.55). The single trial confined to high-risk disease gave HR 1.40 (95% CI 1.02–1.93) and the two trials confined to intermediate-risk disease HR 0.93 (95% CI 0.58–1.50).

Excluding the three estimates that rested on assumptions or interim data (arm sizes assumed equal in two trials, a 12-month interim analysis in one) gave HR 1.59 (95% CI 1.20–2.11). Restricting to the six estimates derived from hazard ratios, log-rank statistics or Kaplan–Meier data gave HR 1.58 (95% CI 1.26–1.98) with less heterogeneity (I² = 54%, τ² = 0.04). Using the 3-year MMC arm of Friedrich 2007, which had far fewer recurrences than 6-week BCG (HR 0.35, 95% CI 0.19–0.65) [32], gave an overall HR of 1.30 (95% CI 0.95–1.77). Omitting one trial at a time (Figure 6) moved the pooled HR between 1.28 and 1.48, and the lower confidence limit fell to 0.99 when EORTC 30911, Cheng 2005, Hinotsu 2011 or Martínez-Piñeiro 1990 was omitted. The funnel plot was roughly symmetrical (Egger p = 0.92; Figure 5), although with only fourteen trials the test has little power.

Figure 3. Recurrence by BCG schedule in the comparator arm. Cheng 2005, Hinotsu 2011 and Bendary 2011 are not shown because their BCG schedule was mixed or unclear.

3.5 Secondary outcomes

Progression was reported with counts in three trials [11,26,49] and did not differ significantly, although it tended to be more frequent with chemotherapy (RR 1.49 (95% CI 0.98–2.26); I² = 0%, τ² = 0.00; Figure 4a). In the CIS stratum of SWOG 8216 [39], failure to achieve a complete response was more likely with doxorubicin than with BCG (RR 2.21 (95% CI 1.46–3.35)), in line with an earlier meta-analysis [50].

Adverse events, as defined in each trial, tended to be less frequent with chemotherapy (RR 0.73 (95% CI 0.52–1.05); I² = 53%, τ² = 0.05; three trials [11,33,35]; Figure 4b). Discontinuation because of adverse events was reported in only two trials, which pointed in opposite directions: it was less frequent with epirubicin in EORTC 30911 (RR 0.32 (95% CI 0.19–0.52)) [35], but one MMC trial reported more withdrawals for chemical cystitis and hypersensitivity (RR 5.50 (95% CI 1.29–23.45)) [33]. Grade 3 or higher events were too sparsely reported to pool; in the gemcitabine–docetaxel trial there were none with chemotherapy and three with BCG [11]. Survival was rarely reported. EORTC 30911 found longer overall and disease-specific survival with BCG than with epirubicin [36], whereas the long-term MMC trials found no survival difference [26,49].

Figure 4a. Progression.

Figure 4b. Adverse events (any or local, as defined in each trial).

Figure 5. Funnel plot for the primary analysis.

Figure 6. Leave-one-out sensitivity analysis for the primary outcome. The dashed line is the overall pooled HR.

3.6 Certainty of evidence

Table 3. Summary of findings.

DISCUSSION

Taken as a class, intravesical chemotherapy was less effective than BCG at preventing recurrence in patients who had not previously received BCG, but the class effect hides clear differences between agents. The class estimate was nevertheless borderline: the Hartung–Knapp interval included 1, and the lower confidence limit fell to 0.99 when some single trials were omitted. BCG was consistently better than epirubicin and doxorubicin. For MMC the comparison depended on the BCG schedule: maintenance BCG was superior, whereas induction-only BCG was less effective than MMC, which is the pattern described in the individual-patient-data meta-analysis [1] and in earlier work on progression [51]. The two gemcitabine-based trials gave estimates compatible with BCG, with a tendency to fewer adverse events.

The seven eligible trials that could not be pooled matter for the MMC result. SWOG 8795 [42] and CUETO 95011 [44] both favoured BCG, and their absence probably biases our MMC estimate towards the null. Our MMC estimate should therefore be read alongside the Cochrane review, which had access to these data and reported an HR of 0.88 (0.71–1.09) for BCG versus MMC [2]. The interim data used for one large trial [27,28] and the equal-arm assumption for another [29] add further uncertainty; removing assumption-based estimates did not change the direction of the overall result.

The newer evidence is the main addition here. One small trial of single-agent gemcitabine in intermediate-risk disease found a recurrence rate similar to BCG [41]. A trial of gemcitabine against one-third-dose BCG also reported no difference at one year but gave no counts [47]. The randomised trial of sequential gemcitabine–docetaxel reported fewer recurrences and fewer adverse events than BCG at 12 months [11]; with 15 recurrences in 91 patients and no description of the randomisation procedure in the accessible record, the confidence interval includes both a large benefit and no difference. These findings agree with a recent pooled analysis of gemcitabine that mixed randomised and observational studies [10], and they support the rationale for the ongoing phase 3 BRIDGE trial (ECOG-ACRIN EA8212) comparing gemcitabine–docetaxel with BCG [52], whose results were still awaited when the most recent review of first-line regimens was completed [7], but they do not establish equivalence.

For practice, the results support current guidance: maintenance BCG remains the reference for high-risk disease, anthracyclines are not an adequate substitute, and MMC with maintenance is a reasonable alternative for intermediate-risk disease. When BCG is unavailable or not tolerated, gemcitabine-based chemotherapy is a possible alternative in intermediate-risk patients, but the randomised evidence is limited to one small trial of gemcitabine and one of gemcitabine–docetaxel, follow-up is short and progression data are few.

Meeting the BCG shortage. Applied to the two objectives of this review, the pooled comparison shows that chemotherapy as a class tended to be less effective than BCG for preventing recurrence (HR 1.38, 95% CI 1.05–1.82), so it cannot replace BCG across the board, and that estimate was borderline. It can, however, reduce the number of patients who need BCG. The single trial confined to high-risk disease gave HR 1.40 (95% CI 1.02–1.93), whereas the two trials confined to intermediate-risk disease gave HR 0.93 (95% CI 0.58–1.50); although this comparison is underpowered (p for difference = 0.17), it fits the view that scarce BCG should be reserved for high-risk patients and that chemotherapy (MMC with maintenance or a gemcitabine-based regimen) can be offered to intermediate-risk patients. Chemotherapy also tended to cause fewer adverse events (RR 0.73, 95% CI 0.52–1.05) and may be easier to complete in patients who tolerate BCG poorly. Regimens that combine chemotherapy with BCG [19,20,21] were excluded from the pooled analysis. In CUETO 93009, sequential mitomycin C plus BCG was more effective but more toxic than BCG alone [20]. Whether such combinations reduce the BCG needed per patient, or only add toxicity, was outside the scope of this review and needs a separate analysis. BCG supply can also be stretched rather than replaced: in a real-world cohort of 563 patients treated with adequate BCG, one-third-dose BCG was not associated with worse time to recurrence (p = 0.449) or progression (p = 0.716) than full dose, with higher-risk patients prioritised for full dose [53]. That study was retrospective, and the one-third-dose group had more low-grade, solitary tumours without carcinoma in situ.

4.1 Strengths and limitations

The review brings the 2023–2026 trials together with the older literature on a single hazard-ratio scale and separates the comparison by agent and by BCG schedule. Its limitations are substantial. Only PubMed was searched and screening was not done in duplicate. Most effect sizes were derived from summary statistics in abstracts, and three relied on assumptions or interim data; full-text verification may change individual estimates. Seven eligible trials could not be included quantitatively. The pooled evidence for gemcitabine and for gemcitabine–docetaxel rests on one trial each, and the secondary outcomes (progression and adverse events) on three trials each. Thirteen trials were excluded because BCG-naïve status was not reported or not strict, including three single-agent gemcitabine trials, so the pooled results may not reflect all of the available randomised evidence. BCG-naïve status was explicit in only four of the fourteen pooled trials, so the population in older trials is BCG-naïve by inference (the estimate in the explicit trials, HR 1.23 (95% CI 0.77–1.97), did not differ significantly from that in the presumed trials, HR 1.47 (95% CI 1.03–2.11); p = 0.55, but this comparison has low power); some included patients who had previously received intravesical chemotherapy. BCG strain, dose and schedule, chemotherapy dose and dwell time, the use of re-resection and the definition of recurrence all varied, and adverse-event definitions were not uniform. Risk-of-bias judgements are provisional.

CONCLUSIONS

In BCG-naïve NMIBC, BCG with maintenance gives better protection against recurrence than epirubicin or doxorubicin and, less consistently, than mitomycin C. Gemcitabine and gemcitabine–docetaxel (one trial each) produced recurrence estimates compatible with BCG and tended to cause fewer adverse events, but these results come from small, short trials at high risk of bias. Adequately powered trials with at least two years of follow-up, standard BCG maintenance in the control arm and reporting of progression are needed before chemotherapy can be recommended in place of BCG for high-risk disease.

Summary of meta-analysis results. Across 14 RCTs (3,227 participants) in which BCG-naïve status was explicit or presumed, intravesical chemotherapy carried a higher hazard of recurrence than BCG (HR 1.38, 95% CI 1.05–1.82; I² = 81%), although the Hartung–Knapp interval (0.98–1.96) included 1. The effect depended on the agent: epirubicin HR 2.07 (95% CI 1.36–3.13), doxorubicin HR 2.49 (95% CI 1.16–5.38), mitomycin C HR 1.05 (95% CI 0.76–1.45), gemcitabine HR 0.81 (95% CI 0.35–1.87) and gemcitabine–docetaxel HR 0.38 (95% CI 0.12–1.18), the last two from one trial each. BCG with maintenance was superior (HR 1.53, 95% CI 1.20–1.95), whereas induction-only BCG was less effective than MMC (HR 0.77, 95% CI 0.62–0.96). Progression did not differ significantly (RR 1.49, 95% CI 0.98–2.26; three trials), and adverse events tended to be less frequent with chemotherapy (RR 0.73, 95% CI 0.52–1.05; three trials). The certainty of the evidence was low to very low.

Meeting the BCG shortage with chemotherapy. (1) Comparison: chemotherapy tended to be less effective than BCG overall, clearly so for epirubicin and doxorubicin, whereas mitomycin C, gemcitabine and gemcitabine–docetaxel gave estimates that were compatible with BCG. (2) Meeting the demand: when BCG is scarce, the available evidence is consistent with reserving it for high-risk disease and offering intermediate-risk patients mitomycin C with maintenance; gemcitabine and gemcitabine–docetaxel are promising but each rests on one small trial and needs phase 3 confirmation. Chemotherapy combined with BCG was not pooled here and needs its own analysis. Stretching the BCG supply with one-third-dose BCG is a complementary strategy supported by real-world evidence [53]. Because BCG-naïve status was explicit in only four of the fourteen pooled trials, these conclusions apply to BCG-naïve disease by inference in most of the evidence.

Declarations

Funding: Nil

Conflicts of interest: Nil

Author contribution: Literature review, manuscript writing, publishing the article.

Ethics: not required for a review of published data. Data availability: the extraction file and analysis code accompany this manuscript.

Use of AI: Nil

REFERENCES

1. Malmström PU, Sylvester RJ, Crawford DE, Friedrich M, Krege S, Rintala E, et al. An individual patient data meta-analysis of the long-term outcome of randomised studies comparing intravesical mitomycin C versus bacillus Calmette-Guérin for non-muscle-invasive bladder cancer. Eur Urol. 2009;56(2):247-56. doi:10.1016/j.eururo.2009.04.038
2. Schmidt S, Kunath F, Coles B, Draeger DL, Krabbe LM, Dersch R, et al. Intravesical Bacillus Calmette-Guérin versus mitomycin C for Ta and T1 bladder cancer. Cochrane Database Syst Rev. 2020;1(1):CD011935. doi:10.1002/14651858.CD011935.pub2
3. Shang PF, Kwong J, Wang ZP, Tian J, Jiang L, Yang K, et al. Intravesical Bacillus Calmette-Guérin versus epirubicin for Ta and T1 bladder cancer. Cochrane Database Syst Rev. 2011;2011(5):CD006885. doi:10.1002/14651858.CD006885.pub2
4. Azuri W, Jaunarena JH, Camean JJ, Chemi J, Villaronga A, Daneshmand S, et al. The impact of dose reduction of bacillus Calmette-Guerin on oncological outcomes and toxicity in non-muscle invasive bladder cancer: a systematic review and meta-analysis. Bladder Cancer. 2023;9(3):227-36. doi:10.3233/BLC-230044
5. Wong CH, Ko IC, Leung DK, Yuen SK, Siu B, Yuan Y, et al. A Systematic Review and Meta-analysis of the Clinical Impact of Prophylactic Quinolones with Adjuvant Bacillus Calmette-Guérin Instillation for Non-muscle-invasive Bladder Cancer. Eur Urol Oncol. 2025;8(2):571-580. doi:10.1016/j.euo.2024.12.013
6. Roessler N, Maiorano BA, Miszczyk M, Miyajima K, Inoue S, von Deimling M, et al. First-line treatments for BCG-naïve non-muscle invasive bladder cancer: a systematic review and meta-analysis. World J Urol. 2026;44(1):76. doi:10.1007/s00345-025-06180-5
7. Pattou M, Rouprêt M, Kamat A, Boorjian SA, Gontero P, Shariat SF, et al. Novel BCG combination or modified BCG regimens in BCG-naïve high-risk non–muscle-invasive bladder cancer: a systematic review and meta-analysis. Eur Urol Focus. 2026 (in press). doi:10.1016/j.euf.2026.06.019
8. Chou R, Selph S, Buckley DI, Fu R, Griffin JC, Grusing S, et al. Intravesical Therapy for the Treatment of Nonmuscle Invasive Bladder Cancer: A Systematic Review and Meta-Analysis. J Urol. 2016;197(5):1189-1199. doi:10.1016/j.juro.2016.12.090
9. Han MA, Maisch P, Jung JH, Hwang JE, Narayan V, Cleves A, et al. Intravesical gemcitabine for non-muscle invasive bladder cancer. Cochrane Database Syst Rev. 2021;6(6):CD009294. doi:10.1002/14651858.CD009294.pub3
10. Kim DK, Jeon J, Kwon JK, Bang S, Lee KS, Koo KC, et al. Efficacy and Toxicity in Scheduled Intravesical Gemcitabine Versus Bacillus Calmette-Guérin for Bladder Cancer: A Systematic Review and Meta-Analysis. Cancers (Basel). 2026;18(6). doi:10.3390/cancers18060990
11. Srivastava A, Trivedi S, Singh Y, Kumar U, Kumar L, Data S, et al. Prospective comparative evaluation of clinical efficacy and safety of Intravesical gemcitabine-docetaxel versus BCG in adjuvant management of BCG-naive non-muscle invasive urothelial bladder cancer in Indian population. Int Urol Nephrol. 2026;58(9):3491-3502. doi:10.1007/s11255-026-05067-7
12. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71.
13. Parmar MK, Torri V, Stewart L. Extracting summary statistics to perform meta-analyses of the published literature for survival endpoints. Stat Med. 1998;17(24):2815-34.
14. Tierney JF, Stewart LA, Ghersi D, Burdett S, Sydes MR. Practical methods for incorporating summary time-to-event data into meta-analysis. Trials. 2007;8:16.
15. Sterne JAC, Savović J, Page MJ, Elbers RG, Blencowe NS, Boutron I, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. 2019;366:l4898.
16. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7(3):177-88.
17. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327(7414):557-60.
18. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629-34.
19. Hayne D, Zhang AY, Thomas H, Krieger L, Ischia J, Anderson P, et al. Bacillus Calmette-Guérin Plus Mitomycin Versus Bacillus Calmette-Guérin Alone for Bacillus Calmette-Guérin-naïve Non-muscle-invasive Bladder Cancer: A Randomised Phase 3 Trial (ANZUP 1301). Eur Urol. 2026;89(5):446-453. doi:10.1016/j.eururo.2026.01.009
20. Solsona E, Madero R, Chantada V, Fernandez JM, Zabala JA, Portillo JA, et al. Sequential combination of mitomycin C plus bacillus Calmette-Guérin (BCG) is more effective but more toxic than BCG alone in patients with non-muscle-invasive bladder cancer in intermediate- and high-risk patients: final outcome of CUETO 93009, a randomized prospective trial. Eur Urol. 2014;67(3):508-16. doi:10.1016/j.eururo.2014.09.026
21. Di Stasi SM, Giannantoni A, Giurioli A, Valenti M, Zampa G, Storti L, et al. Sequential BCG and electromotive mitomycin versus BCG alone for high-risk superficial bladder cancer: a randomised controlled trial. Lancet Oncol. 2006;7(1):43-51. doi:10.1016/S1470-2045(05)70472-1
22. Marttila T, Järvinen R, Liukkonen T, Rintala E, Boström P, Seppänen M, et al. Intravesical Bacillus Calmette-Guérin Versus Combination of Epirubicin and Interferon-α2a in Reducing Recurrence of Non-Muscle-invasive Bladder Carcinoma: FinnBladder-6 Study. Eur Urol. 2016;70(2):341-7. doi:10.1016/j.eururo.2016.03.034
23. Duchek M, Johansson R, Jahnson S, Mestad O, Hellström P, Hellsten S, et al. Bacillus Calmette-Guérin is superior to a combination of epirubicin and interferon-alpha2b in the intravesical treatment of patients with stage T1 urinary bladder cancer. A prospective, randomized, Nordic study. Eur Urol. 2009;57(1):25-31. doi:10.1016/j.eururo.2009.09.038
24. Di Lorenzo G, Perdonà S, Damiano R, Faiella A, Cantiello F, Pignata S, et al. Gemcitabine versus bacille Calmette-Guérin after initial bacille Calmette-Guérin failure in non-muscle-invasive bladder cancer: a multicenter prospective randomized trial. Cancer. 2010;116(8):1893-900. doi:10.1002/cncr.24914
25. Rintala E, Jauhiainen K, Alfthan O, Hansson E, Juusela H, Kanerva K, et al. Intravesical chemotherapy (mitomycin C) versus immunotherapy (bacillus Calmette-Guérin) in superficial bladder cancer. Eur Urol. 1991;20(1):19-25. doi:10.1159/000471653
26. Järvinen R, Kaasinen E, Sankila A, Rintala E. Long-term efficacy of maintenance bacillus Calmette-Guérin versus maintenance mitomycin C instillation therapy in frequently recurrent TaT1 tumours without carcinoma in situ: a subgroup analysis of the prospective, randomised FinnBladder I study with a 20-year follow-up. Eur Urol. 2009;56(2):260-5. doi:10.1016/j.eururo.2009.04.009
27. DeBruyne FM, van der Meijden AP, Geboers AD, Franssen MP, van Leeuwen MJ, Steerenberg PA, et al. BCG (RIVM) versus mitomycin intravesical therapy in superficial bladder cancer. First results of randomized prospective trial. Urology. 1988;31(3 Suppl):20-5.
28. Witjes JA, v d Meijden AP, Collette L, Sylvester R, Debruyne FM, van Aubel A, et al. Long-term follow-up of an EORTC randomized prospective trial comparing intravesical bacille Calmette-Guérin-RIVM and mitomycin C in superficial bladder cancer. EORTC GU Group and the Dutch South East Cooperative Urological Group. European Organisation for Research and Treatment of Cancer Genito-Urinary Tract Cancer Collaborative Group. Urology. 1998;52(3):403-10. doi:10.1016/s0090-4295(98)00212-x
29. Vegt PD, Witjes JA, Witjes WP, Doesburg WH, Debruyne FM, van der Meijden AP. A randomized study of intravesical mitomycin C, bacillus Calmette-Guerin Tice and bacillus Calmette-Guerin RIVM treatment in pTa-pT1 papillary carcinoma and carcinoma in situ of the bladder. J Urol. 1995;153(3 Pt 2):929-33.
30. Lundholm C, Norlén BJ, Ekman P, Jahnson S, Lagerkvist M, Lindeborg T, et al. A randomized prospective study comparing long-term intravesical instillations of mitomycin C and bacillus Calmette-Guerin in patients with superficial bladder carcinoma. J Urol. 1996;156(2 Pt 1):372-6. doi:10.1016/s0022-5347(01)65853-1
31. Malmström PU, Wijkström H, Lundholm C, Wester K, Busch C, Norlén BJ. 5-year followup of a randomized prospective study comparing mitomycin C and bacillus Calmette-Guerin in patients with superficial bladder carcinoma. Swedish-Norwegian Bladder Cancer Study Group. J Urol. 1999;161(4):1124-7.
32. Friedrich MG, Pichlmeier U, Schwaibold H, Conrad S, Huland H. Long-term intravesical adjuvant chemotherapy further reduces recurrence rate compared with short-term intravesical chemotherapy and short-term therapy with Bacillus Calmette-Guérin (BCG) in patients with non-muscle-invasive bladder carcinoma. Eur Urol. 2007;52(4):1123-29. doi:10.1016/j.eururo.2007.02.063
33. Mangiarotti B, Trinchieri A, Del Nero A, Montanari E. A randomized prospective study of intravesical prophylaxis in non-musle invasive bladder cancer at intermediate risk of recurrence: mitomycin chemotherapy vs BCG immunotherapy. Arch Ital Urol Androl. 2008;80(4):167-71.
34. Melekos MD, Zarakovitis IE, Fokaefs ED, Dandinis K, Chionis H, Bouropoulos C, et al. Intravesical bacillus Calmette-Guérin versus epirubicin in the prophylaxis of recurrent and/or multiple superficial bladder tumours. Oncology. 1996;53(4):281-8. doi:10.1159/000227574
35. van der Meijden AP, Brausi M, Zambon V, Kirkels W, de Balincourt C, Sylvester R. 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. J Urol. 2001;166(2):476-81. doi:10.1097/00005392-200108000-00016
36. Sylvester RJ, Brausi MA, Kirkels WJ, Hoeltl W, Calais Da Silva F, Powell PH, 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. 2009;57(5):766-73. doi:10.1016/j.eururo.2009.12.024
37. Cheng CW, Chan SF, Chan LW, Chan CK, Ng CF, Cheung HY, et al. Twelve-year follow up of a randomized prospective trial comparing bacillus Calmette-Guerin and epirubicin as adjuvant therapy in superficial bladder cancer. Int J Urol. 2005;12(5):449-55. doi:10.1111/j.1442-2042.2005.01064.x
38. Hinotsu S, Akaza H, Naito S, Ozono S, Sumiyoshi Y, Noguchi S, et al. Maintenance therapy with bacillus Calmette-Guérin Connaught strain clearly prolongs recurrence-free survival following transurethral resection of bladder tumour for non-muscle-invasive bladder cancer. BJU Int. 2010;108(2):187-95. doi:10.1111/j.1464-410X.2010.09891.x
39. Lamm DL, Blumenstein BA, Crawford ED, Montie JE, Scardino P, Grossman HB, et al. A randomized trial of intravesical doxorubicin and immunotherapy with bacille Calmette-Guérin for transitional-cell carcinoma of the bladder. N Engl J Med. 1991;325(17):1205-9. doi:10.1056/NEJM199110243251703
40. Martínez-Piñeiro JA, Jiménez León J, Martínez-Piñeiro L, Fiter L, Mosteiro JA, Navarro J, et al. Bacillus Calmette-Guerin versus doxorubicin versus thiotepa: a randomized prospective study in 202 patients with superficial bladder cancer. J Urol. 1990;143(3):502-6. doi:10.1016/s0022-5347(17)40002-4
41. Bendary L, Khalil S, Shahin A, Nawar N. Intravesical gemcitabine versus bacillus Calmette-Guérin (BCG) in treatment of non-muscle invasive bladder cancer: short term comparative study [abstract]. J Urol. 2011;185(4 Suppl):e664-5. [conference abstract; citation details to be verified by the authors]
42. Lamm DL, Blumenstein BA, Crawford ED, Crissman JD, Lowe BA, Smith JA, et al. Randomized intergroup comparison of bacillus Calmette-Guerin immunotherapy and mitomycin C chemotherapy prophylaxis in superficial transitional cell carcinoma of the bladder: a Southwest Oncology Group study. Urol Oncol. 1995;1(3):119-26. doi:10.1016/1078-1439(95)00041-f
43. Krege S, Giani G, Meyer R, Otto T, Rübben H. A randomized multicenter trial of adjuvant therapy in superficial bladder cancer: transurethral resection only versus transurethral resection plus mitomycin C versus transurethral resection plus bacillus Calmette-Guerin. Participating Clinics. J Urol. 1996;156(3):962-6. doi:10.1016/s0022-5347(01)65673-8
44. Ojea A, Nogueira JL, Solsona E, Flores N, Gómez JM, Molina JR, 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. Eur Urol. 2007;52(5):1398-406. doi:10.1016/j.eururo.2007.04.062
45. Melekos MD, Chionis HS, Paranychianakis GS, Dauaher HH. Intravesical 4'-epi-doxorubicin (epirubicin) versus bacillus Calmette-Guérin. A controlled prospective study on the prophylaxis of superficial bladder cancer. Cancer. 1993;72(5):1749-55. doi:10.1002/1097-0142(19930901)72:5<1749::aid-cncr2820720539>3.0.co;2-8
46. Hinotsu S, Akaza H, Isaka S, Kanetake H, Kubota Y, Kuroda M, et al. Sustained prophylactic effect of intravesical bacille Calmette-Guérin for superficial bladder cancer: a smoothed hazard analysis in a randomized prospective study. Urology. 2006;67(3):545-9. doi:10.1016/j.urology.2005.09.045
47. Gontero P, Oderda M, Mehnert A, Gurioli A, Marson F, Lucca I, et al. The impact of intravesical gemcitabine and 1/3 dose Bacillus Calmette-Guérin instillation therapy on the quality of life in patients with nonmuscle invasive bladder cancer: results of a prospective, randomized, phase II trial. J Urol. 2013;190(3):857-62. doi:10.1016/j.juro.2013.03.097
48. Sachan A, Nayyar R, Pethe S, Singh P, Seth A. A 3-arm randomized control trial to compare the efficacy of re-circulant hyperthermic intravesical chemotherapy versus conventional intravesical mitomycin C and BCG therapy for intermediate-risk non-muscle invasive bladder cancer. World J Urol. 2024;42(1):609. doi:10.1007/s00345-024-05338-x
49. Gårdmark T, Jahnson S, Wahlquist R, Wijkström H, Malmström PU. Analysis of progression and survival after 10 years of a randomized prospective study comparing mitomycin-C and bacillus Calmette-Guérin in patients with high-risk bladder cancer. BJU Int. 2007;99(4):817-20. doi:10.1111/j.1464-410X.2006.06706.x
50. Sylvester RJ, van der Meijden AP, Witjes JA, Kurth K. 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. J Urol. 2005;174(1):86-91; discussion 91-2. doi:10.1097/01.ju.0000162059.64886.1c
51. Böhle A, Bock PR. Intravesical bacille Calmette-Guérin versus mitomycin C in superficial bladder cancer: formal meta-analysis of comparative studies on tumor progression. Urology. 2004;63(4):682-6; discussion 686-7. doi:10.1016/j.urology.2003.11.049
52. Kates M, Chu X, Hahn N, Pietzak E, Smith A, Shevrin DH, et al. Background and Update for ECOG-ACRIN EA8212: A Randomized Phase 3 Trial of Intravesical Bacillus Calmette-Guérin (BCG) Versus Intravesical Docetaxel and Gemcitabine Treatment in BCG-naïve High-grade Non-muscle-invasive Bladder Cancer (BRIDGE). Eur Urol Focus. 2023;9(4):561-563. doi:10.1016/j.euf.2023.06.006
53. Lobo N, Bree KK, Hensley PJ, Nogueras-Gonzalez GM, Abraham P, Navai N, et al. Reduced-dose bacillus Calmette-Guérin (BCG) in an era of BCG shortage: real-world experience from a tertiary cancer centre. BJU Int. 2022;130(3):323-30. doi:10.1111/bju.15661