Bladder cancer is a common malignancy that ranks as the 9th most frequently diagnosed malignancy worldwide. This type of cancer is also ranked 13th in terms of mortality rates, with developing countries having higher mortality rates than developed countries (1). Non-muscle invasive bladder cancer (NMIBC) is a subset of bladder cancer that comprises tumors of stage Ta, T1 and carcinoma in situ (CIS). It is estimated that NMIBC accounts for 70-75% of all diagnosed cases of bladder cancer (2,3).

NMIBC is known to have a high recurrence and progression rate that depends on the tumor risk profile with the chance of recurrence at 1 year ranging between 15-61% and 31-78% at 5 years. Progression rates also vary significantly, with the 1-year progression rate into muscle invasive metastatic bladder cancer (MIBC) ranging from 0.2-17%, and increasing to 0.8-45% at 5 years (2). Mortality rates associated with NMIBC are relatively lower than those associated with MIBC, and increases with higher-risk tumor features. In patients with low-grade Ta tumors, the 15-year progression-free survival is 95% with no cancer-specific mortality. This decreases to 61% in Ta high-grade Ta tumors, with a disease-specific mortality of 26%. Patients with T1 tumors have a progression-free survival rate of 44%, with a disease-specific mortality reaching 38% (4). When progression to MIBC occurs, patients are expected to have a poor prognosis, with a 5-year mortality rate of 50-70% even following radical cystectomy (5).

Bladder cancer is among the types of cancer affecting the elderly with the highest treatment costs, with an estimated economic burden of approximately US $4 billion per year. This high cost is attributed to the need for lifelong cystoscopic surveillance and multiple treatment modalities (6). Transurethral resection of bladder tumor (TURBT) is the initial treatment of choice, where the tumor is resected using a resectoscope inserted via the urethra. In this procedure, it is recommended that all visible tumors be resected along with the underlying detrusor muscle (7). It is known that the risk of upstaging NMIBC (≥T2) increases up to 49% in the case that the detrusor muscle is not obtained during resection (8). TURBT is also associated with a high recurrence rate, with a 5-year recurrence reaching 42% for T1 tumors (9). To mitigate this risk, clinicians usually recommend a second TURBT and adjuvant therapy. The instillation of intravesical chemotherapy is recommended by the European Association of Urologist (EAU) to reduce the recurrence rate (10). A previous meta-analysis demonstrated that immediate postoperative intravesical chemotherapy reduced recurrence to 37% compared to 48% with TURBT alone (11). Several agents, such as mitomycin C, epirubicin, doxorubicin and gemcitabine are used for this purpose. Mitomycin C is the most commonly studied drug; however, it is currently not available in Indonesia (12).

Immunotherapy using bacillus Calmette-Guérin (BCG) is a widely-recognized first line agent for managing high-risk NMIBC, according to the European Organization for Research and Treatment of Cancer (EORTC) risk calculator (13,14). Currently, The EAU Guidelines suggest that BCG following TURBT is more effective in preventing recurrence than TURBT alone or TURBT with intravesical chemotherapy (10). Despite its efficacy, BCG is associated with a greater number of side-effects than chemotherapy (15).

Given the high economic burden associated with NMIBC, the careful and evidence-based selection of treatment modalities is essential for optimizing outcomes and efficiency. There is a need to compare intravesical chemotherapy agents other than mitomycin C with first-line BCG immunotherapy in the Indonesian setting. The present study aimed to investigate the efficacy and safety of gemcitabine-based intravesical chemotherapy vs. BCG immunotherapy in order to provide a clearer perspective on which treatment provides a better prognostic value for patients with NMIBC.

The present systematic review and meta-analysis included four RCTs comprising 344 patients and one retrospective cohort study with 103 patients. The earliest RCT was conducted by Porena et al (17) in 2010 among high-risk superficial bladder cancer patients based on the EAU risk classification system. BCG was administered as a 6-weekly instillation of 5x10⁸ CFU of Tice strain BCG retained intravesically for 2 h (17). Gemcitabine was administered as a 6-weekly instillation of 2,000 mg, also retained for 2 h. In that study, BCG was associated with a lower recurrence rate (28.1 vs. 53.1%) and a longer recurrence-free survival than gemcitabine (17). These findings are in contrast to those of the other three RCTs included in the present study.

Di Lorenzo et al (18) compared these treatments in patients with high-risk NMIBC (per EORTC scoring system) who had previously failed BCG therapy. In their study, BCG (Connaught strain, 81 mg/50 ml) was administered 4-6 weeks after the first treatment failure, over a 6-week induction followed by maintenance at 3, 6 and 12 months. The gemcitabine group received 2,000 mg/50 ml over a period of 6 weeks and similar maintenance at 2, 6 and 12 months (18). In contrast to the study by Porena et al (17), the study by Di Lorenzo et al (18) found lower recurrence rates and a higher 2-year recurrence-free survival in the gemcitabine group, with no significant differences in time-to-recurrence or progression. This was the only study to include BCG-refractory patients, which may explain the divergent findings from that of Porena et al (17).

In their study, Bendary et al (19) included patients with primary Ta-T1 tumors without CIS and reported no significant differences in recurrence or progression rates between BCG and gemcitabine. In that study, BCG was administered at 6x10⁸ CFU (strain unspecified) and gemcitabine at 2,000 mg in 50 ml saline, both over 6 weekly 2-h instillations (19). Similarly, Gontero et al (16) found no significant difference in recurrence, recurrence-free survival, or progression in intermediate-risk patients with NMIBC. Their BCG regimen used one-third dose (27 mg) of Connaught strain over a 6-week induction period. Gemcitabine was administered identically across studies (2,000 mg/50 ml, 6-week induction) (16).

Prasanna et al (20) conducted a retrospective cohort study on patients with CIS, pTa and pT1 tumors, stratified into EAU-defined risk groups. BCG (Oncotice strain, 5x10⁸ CFU) and gemcitabine (2,000 mg) were administered as 6 weekly instillations. Despite baseline differences in risk-group distribution, adjusted multivariate analysis revealed no difference in disease-free survival between the groups (20). All studies agreed that gemcitabine had a more favorable safety profile, with fewer local and systemic adverse events, leading to better tolerability and quality of life.

The variability in the findings across these five studies is largely attributable to clinical and methodological heterogeneity. Porena et al (17) included only high-risk patients per EAU criteria, which included patients with NMIBC with CIS, all T1 tumors without CIS, Ta LG/G2 without CIS with three risk factors, Ta HG/G3 without CIS and with two risk factors, and T1 G2 without CIS and with one risk factor (10). Conversely, Gontero et al (16) focused on intermediate-risk NMIBC, and accordingly reported no difference in outcomes between gemcitabine and BCG. Di Lorenzo et al (18) used the EORTC scoring system to define high risk (recurrence score >9 and/or progression score >13). Although EORTC risk classification may slightly outperform EAU in predicting recurrence (c-index 0.64 vs. 0.62; P=0.035), the key distinguishing feature in the study by Di Lorenzo et al (18) was the inclusion of BCG-refractory patients, not the risk model used.

Bendary et al (19) and Prasanna et al (20) included broader and less well-defined populations. In the study by Bendary et al (19), the inclusion of all Ta-T1 tumors without CIS spanned a wide risk spectrum but lacked proper risk stratification. By contrast, Prasanna et al (20) used similarly broad criteria but stratified patients into EAU risk groups, strengthening the interpretation of findings. BCG regimens also varied. Although strain and dose differences existed, a previous meta-analysis of 65 trials involving >12,000 patients concluded that no strain demonstrated clear superiority (21). Gontero et al (16) used a one-third dose of the Connaught strain in intermediate-risk patients to reduce toxicity without sacrificing efficacy. Gemcitabine regimens were consistent across studies. Only Di Lorenzo et al (18) included patients with prior BCG failure, and only in this context did gemcitabine demonstrate superior efficacy.

The findings presented herein align with those of prior reviews. Shelley et al (22) reported that gemcitabine was comparable to BCG in intermediate-risk patients, inferior in high-risk, and superior in BCG-refractory NMIBC. However, the present systematic review included more representative populations from Bendary et al (19) and Prasanna et al (20), both demonstrating no efficacy difference across wider risk groups. Similarly, the Cochrane review by Jones et al (23) found gemcitabine less effective in high-risk NMIBC, comparable in intermediate-risk, and more effective in BCG-refractory patients. A separate meta-analysis also revealed no significant difference in recurrence (HR, 1.04; 95% CI, 0.38-2.89) or progression (HR, 0.01; 95% CI, -0.75 to 0.77) between the two agents (24).

Overall, the findings of the present systematic review and meta-analysis suggest that intravesical gemcitabine was non-inferior to BCG immunotherapy in terms of efficacy, while it resulted in lower rates of dysuria, fever, urinary frequency and itching. However, intravesical gemcitabine appears particularly useful for patients with BCG-refractory NMIBC. In this population, the EAU recommends radical cystectomy (11), although it remains an invasive procedure with significant morbidity and mortality, and entails major lifestyle changes (25). Intravesical gemcitabine may therefore serve as a non-invasive alternative for NMIBC patients unresponsive to BCG immunotherapy.

The superiority of gemcitabine over BCG in BCG-refractory NMIBC may be explained by several factors. BCG, as with other immunotherapies, relies on an intact immune system to activate innate immunity and initiate BCG-specific and tumor-specific T-cell responses. Although the exact mechanisms remain unclear, BCG failure is considered to result from complex interactions between host immunity and tumor microenvironment, leading to immune dysregulation and inadequate immune activation. Specifically, non-responders have been shown to be associated with increased levels of CD25⁺ regulatory T-cells and tumor-associated macrophage, enrichment of exhausted CD8⁺PDL-1(+) T-cells, and reduced levels of Th2-predominant CD4⁺ T-cells within the tumor microenvironment (26). By contrast, gemcitabine inhibits tumor DNA synthesis, leading to apoptosis, and is therefore less dependent on host immune function (23). Its lower risk of adverse events may be attributed to its immune-selective cytotoxicity and high plasma clearance, allowing any drug that enters systemic circulation to be rapidly eliminated (23). Furthermore, BCG is a live-attenuated strain of Mycobacterium bovis that activates a strong immune response and induces inflammation following urothelial invasion (27), which contributes to a higher incidence of local and systemic side-effects (27,28).

In conclusion, the present systematic review and meta-analysis demonstrated that intravesical gemcitabine was non-inferior to BCG immunotherapy in the treatment of patients with NMIBC and may serve as a viable option for those unresponsive to BCG. Although the overall risk of AEs was comparable, gemcitabine was associated with lower rates of dysuria, fever, urinary frequency and itching. Further high-quality RCTs with larger sample sizes, standardized dosing, treatment intervals, and follow-up durations are warranted to confirm these findings.