1. Introduction

Bladder cancer represents a substantial global health burden, ranking as the 10th most prevalent malignancy worldwide, with more than 500,000 new cases diagnosed annually¹. Urothelial carcinoma (UC) is the predominant histologic type, and at initial diagnosis, approximately 75–80% of patients present with non–muscle-invasive bladder cancer (NMIBC)². This category encompasses a heterogeneous spectrum of disease, ranging from low-grade Ta tumors to high-grade T1 lesions and carcinoma in situ (CIS), each carrying distinct risks for recurrence and progression. For decades, the standard of care for high-risk NMIBC has remained transurethral resection of the bladder tumor, followed by adjuvant intravesical immunotherapy with Bacillus Calmette–Guérin (BCG)^2-4. When successful, this regimen significantly reduces recurrence rates and delays progression to muscle-invasive disease^5-8.

However, despite adequate BCG therapy, treatment failure remains a critical challenge in urologic oncology. Approximately 30–40% of patients will experience disease recurrence, and up to 15% will progress to muscle-invasive bladder cancer (MIBC)⁹. Historically, the management of these patients has been complicated by inconsistent definitions of treatment failure. To address this, regulatory bodies such as the United States Food and Drug Administration (FDA) and the European Association of Urology have established strict criteria for “BCG-unresponsive” disease^2,10. This subset includes patients with high-grade T1 recurrence within six months of adequate BCG therapy, persistent or recurrent CIS within 12 months, or the development of high-grade tumors during maintenance therapy².

The distinction between “BCG-unresponsive” and “BCG-relapsing” is clinically pivotal. While relapsing patients may still benefit from BCG re-challenge, patients with BCG-unresponsive disease derive minimal benefit from further BCG instillations and face a heightened risk of progression. For this high-risk cohort, the gold standard treatment is radical cystectomy (RC) with urinary diversion. Yet, RC is associated with significant perioperative morbidity and profound alterations in quality of life, leading many patients to refuse surgery or be deemed medically ineligible due to age and comorbidities^11,12.

Consequently, there is an urgent need for effective bladder-sparing therapies. This need is further exacerbated by recurring global shortages of BCG, which have forced clinicians to ration doses and seek alternative first-line and salvage strategies^13,14. Fortunately, the therapeutic landscape is rapidly evolving. In recent years, multiple bladder-sparing approaches have been investigated, including cytotoxic intravesical treatments, device-assisted drug instillation techniques, intravesical immunotherapeutic modalities, systemic immunotherapy, and gene-based therapies^15-18. In addition, various combination strategies, such as integrated chemotherapy with either intravesical or systemic immunotherapy, as well as innovative intravesical delivery platforms for cytotoxic agents, are being explored^19,20.

This narrative review aims to provide a comprehensive update on the management of BCG-unresponsive NMIBC. In this review, we explore the mechanisms of BCG resistance, clarify the definitions guiding clinical decision-making, and critically evaluate the efficacy and safety of current and emerging bladder-preserving therapies that are reshaping contemporary standards of care.

This is a narrative, non-systematic review based on a targeted search of PubMed/MEDLINE, ClinicalTrials.gov, and major guideline sources (last accessed: December 2025). Search terms included “BCG-unresponsive,” “non–muscle-invasive bladder cancer,” “pembrolizumab,” “nadofaragene,” “CG0070/cretostimogene,” “N-803,” and “TAR-200.” Studies were selected through expert appraisal, prioritizing late-phase trials, regulatory approvals, and clinically relevant evidence; no formal systematic screening was performed. Where available, late-phase and practice-defining evidence (regulatory approvals/phase III) was emphasized; early-phase signals were summarized briefly as hypothesis-generating.

2. Mechanisms of action and resistance to Bacillus Calmette–Guérin

To understand why therapy fails, it is essential to first appreciate the complex mechanism of action of BCG. BCG is not merely a local irritant but a potent immunotherapy that relies on a multi-step interaction between the bacteria, urothelial cells, and the host immune system. Upon instillation, BCG attaches to the bladder wall via fibronectin, a crucial step for the subsequent internalization of bacteria into urothelial tumor cells²¹. Once internalized, BCG releases mycobacterial antigens, such as Ag85 and MPT64, which are presented on the cell surface²². This triggers a robust innate immune response characterized by the activation of Toll-like receptors (TLRs; e.g., TLR2, TLR4, and TLR9) and the release of pro-inflammatory cytokines, including interleukin (IL)-2, IL-12, tumor necrosis factor-α, and interferon (IFN)-γ²³. This “cytokine storm” recruits neutrophils, macrophages, and natural killer (NK) cells to the tumor microenvironment²⁴. Crucially, BCG induces the polarization of macrophages toward the antitumor M1 phenotype and stimulates a Th1-mediated adaptive immune response, leading to the direct lysis of tumor cells by cytotoxic cluster of differentiation (CD)8⁺ T lymphocytes.

Resistance to BCG is multifactorial, involving both tumor-intrinsic factors and mechanisms of adaptive immune evasion. Internalization failure is a key resistance mechanism; since internalization is required for antigen presentation, loss of fibronectin expression or defects in endocytosis can render tumor cells “invisible” to the immune system. A primary mechanism of adaptive resistance is the upregulation of immune checkpoints. Exposure to BCG can induce the expression of programmed death-ligand 1 (PD-L1) on tumor cells via the mitogen-activated protein kinase pathway. This interaction with programmed cell death protein-1 (PD-1) on T cells inhibits the cytotoxic response, allowing the tumor to escape immune surveillance. Additionally, chronic stimulation may lead to T-cell exhaustion, characterized by the expression of inhibitory receptors such as T-cell immunoreceptor with Ig and ITIM domains and lymphocyte-activation gene 3, which further dampen the antitumor response^25,26.

3. Defining the clinical challenge: Bacillus Calmette–Guérin-unresponsive disease

Historically, the management of BCG failure was hindered by vague terminology, with “BCG failure” often serving as a catch-all term that complicated clinical decision-making. To address this challenge, regulatory bodies such as the FDA and the European Association of Urology have established rigorous definitions for “BCG-unresponsive” disease to standardize patient care and trial design¹⁰. A foundational criterion for this classification is the receipt of “adequate BCG,” defined as the completion of at least five of six doses of an initial induction course, plus either at least two of three doses of maintenance therapy or at least two of six doses of a second induction course²⁷. Patients meeting these exposure criteria are stratified based on the timing and nature of recurrence: “BCG-unresponsive” disease represents the highest-risk category and includes persistent or recurrent CIS within 12 months of adequate therapy, recurrent high-grade Ta/T1 tumors within six months, or persistent high-grade T1 disease at the first evaluation (three months) after induction. This is distinct from “BCG-relapsing” disease, where high-grade recurrence occurs after these specific windows (e.g., >6 months for papillary tumors or >12 months for CIS). Distinguishing these categories is clinically pivotal because, unlike relapsing patients who may still respond to a BCG re-challenge, those with unresponsive disease derive minimal benefit from further BCG and are prioritized for RC or novel bladder-sparing therapies (Figure 1) ²⁸.

4. Systemic immunotherapy: The checkpoint inhibitor era

The elucidation of immune escape mechanisms in bladder cancer, particularly the upregulation of the PD-1/PD-L1 pathway following BCG exposure, has established a biological rationale for the use of immune checkpoint inhibitors (ICIs) in the BCG-unresponsive setting. While BCG relies on a localized immune response, systemic immunotherapy aims to overcome adaptive resistance by restoring the cytotoxic activity of T-cells that have been exhausted or inhibited by the tumor microenvironment^29,30. This approach represents a significant departure from traditional intravesical therapies, offering a systemic solution to a local problem and reshaping the treatment algorithm for patients ineligible for or refusing RC.

Pembrolizumab, a humanized monoclonal antibody targeting the PD-1 pathway, was the first systemic agent to receive FDA approval for the treatment of BCG-unresponsive high-risk NMIBC (Table 1). PD-1 is an inhibitory immune checkpoint receptor expressed on activated T lymphocytes and other immune effector cells; its interaction with the ligands PD-L1 and PD-L2—which are frequently upregulated in the tumor microenvironment—leads to attenuation of T-cell–mediated antitumor activity and facilitates immune evasion. By blocking PD-1 and preventing ligand binding, pembrolizumab restores antitumor immune surveillance and promotes prolonged T-cell activation. The clinical efficacy of pembrolizumab in this setting was established in the pivotal KEYNOTE-057 trial—a single-arm, phase II study enrolling patients with BCG-unresponsive CIS, with or without concomitant papillary disease, who were either ineligible for or declined RC³¹.

In the pivotal KEYNOTE-057 trial, patients received pembrolizumab 200 mg intravenously every three weeks for up to 24 months or until disease recurrence. It is worth noting that while the trial utilized this three-weekly schedule, regulatory approvals now also permit a dosing alternative of 400 mg every six weeks to reduce the frequency of hospital visits. While the FDA approval is specific to this CIS-containing population (Cohort A), the trial also included a separate cohort (Cohort B) for patients with papillary-only high-risk disease, though results in this subgroup have not yet led to a regulatory indication. The results from Cohort A demonstrated a complete response (CR) rate of 41% at three months. Although recurrence is common in this high-risk population, the response durability was notable, with a median duration of response reaching 16.2 months³¹. Furthermore, 46% of responders maintained their CR for at least 12 months, and importantly, progression to muscle-invasive disease was uncommon during the study period. However, the systemic nature of the treatment introduces a distinct toxicity profile compared to intravesical therapies. Approximately 15% of patients experienced Grade 3 or higher immune-related adverse events, necessitating careful patient selection and monitoring. Despite these risks, pembrolizumab has become a benchmark for bladder-sparing strategies, particularly for patients with CIS who have exhausted intravesical options.

Beyond its established role in BCG-unresponsive disease, pembrolizumab is currently being investigated in combination with intravesical BCG as a potential strategy in earlier stages of high-risk NMIBC. Given that BCG and pembrolizumab exert antitumor effects through complementary immunologic mechanisms, their combined use may offer synergistic enhancement of both innate and adaptive immune responses. A non-randomized phase II study (NCT03504163) is evaluating the combination of pembrolizumab and BCG as a first-line therapy in patients with high-grade, BCG-naïve NMIBC, with the primary endpoint being the proportion of patients achieving disease-free status; results from this trial are currently awaited³². In parallel, the randomized Phase III KEYNOTE-676 trial (NCT03711032) is comparing BCG plus pembrolizumab versus BCG alone in patients with BCG-naïve as well as BCG-recurrent or persistent high-risk NMIBC, aiming to define whether combination therapy can improve oncologic outcomes over standard intravesical treatment³³. Although the efficacy and safety of this combinatorial approach remain to be established, these studies reflect a broader shift toward immunotherapy-based combination strategies in bladder cancer.

In this evolving landscape, other ICIs, including atezolizumab, nivolumab, and the investigational agent sasanlimab, are also under active evaluation as monotherapies or in combination with BCG, underscoring the growing interest in checkpoint inhibition and biomarker-driven strategies, particularly PD-L1 expression, as potential determinants of treatment response.

Atezolizumab, a monoclonal antibody that inhibits the PD-L1/PD-1 pathway and blocks CD80 receptors, was evaluated in the SWOG S1605 phase II trial for patients with BCG-unresponsive high-risk NMIBC¹⁹. The study reported a CR rate of 27% at six months and 26% at 18 months in patients with CIS. Although these results demonstrated clinical activity, response rates were somewhat lower than those observed with pembrolizumab, and the trial reported a comparable rate of serious immune-related adverse events, occurring in approximately 16% of patients¹⁹. Consequently, while atezolizumab demonstrates activity in this disease state, it has not yet replaced pembrolizumab as the standard of care in this setting.

4.1 Shifting paradigms: Early integration in Bacillus Calmette–Guérin-naïve disease

However, the investigation has expanded into the BCG-naïve landscape to determine if upfront combination therapy offers superior control. The Phase Ib/II BladderGATE trial (NCT04134000) assessed atezolizumab combined with induction and maintenance BCG in patients who were either treatment-naïve or had not received BCG for at least two years and yielded encouraging early results³⁴. Interim safety data from 34 patients indicated that the regimen was generally well-tolerated, with no dose-limiting toxicities; however, immune-related adverse events led to discontinuation in seven participants³⁵. Preliminary efficacy findings were promising, demonstrating a two-year disease-free survival (DFS) of 72.8% and a low progression rate to muscle-invasive disease of 8%³⁶. However, the definitive Phase III ALBAN trial (NCT03799835), which randomized over 500 BCG-naïve patients to receive BCG with or without intravenous atezolizumab, failed to confirm these benefits³⁷. In stark contrast to the earlier phase findings, the trial did not meet its primary endpoint, showing no significant difference in event-free survival (EFS) between the combination and BCG alone (hazard ratio [HR] 0.98; p = 0.9106)³⁷. Furthermore, the addition of atezolizumab increased the rate of grade ≥3 treatment-related adverse events without providing oncological benefit, suggesting that the indiscriminate addition of systemic PD-L1 inhibition to standard BCG is ineffective in an unselected BCG-naïve population.

Following the precedent set by pembrolizumab, other checkpoint inhibitors are being rigorously evaluated to expand the therapeutic armamentarium for BCG-unresponsive and persistent disease. While current approvals focus on the unresponsive setting, significant effort is now directed toward improving outcomes in BCG-naïve patients to prevent the development of resistance. Early Phase I data have shown promising activity for the combination of durvalumab and BCG, prompting the initiation of two major Phase III trials^38-40. The POTOMAC study (NCT03528694) is a large-scale, randomized, open-label trial enrolling 1,018 BCG-naïve high-risk patients, assigned to one of three arms: (i) durvalumab + BCG induction followed by maintenance with both agents; (ii) durvalumab + BCG induction followed by BCG-only maintenance; or (iii) standard BCG induction and maintenance alone^{39, 41}. At a median follow-up of approximately 61 months, durvalumab combined with BCG induction and maintenance significantly improved DFS compared with BCG alone (HR = 0.68, 95% confidence interval = 0.50–0.93, p = 0.015), corresponding to a 32% reduction in the risk of recurrence or death. The estimated 24-month DFS rate was 86.5% in the combination arm versus 81.6% with BCG alone. In contrast, durvalumab combined with BCG induction did not significantly improve DFS, underscoring the importance of BCG maintenance. Grade ≥3 treatment-related adverse events occurred more frequently with durvalumab (21%) than with BCG alone (4%), although no treatment-related deaths were observed. These findings support PD-L1 inhibition as an effective strategy in the BCG-naïve high-risk setting, while highlighting the balance between incremental efficacy and increased immune-related toxicity⁴¹.

In parallel, the open-label, single-arm PATAPSCO trial is evaluating the safety and tolerability of adding monthly durvalumab to BCG induction (with or without maintenance) over a 12-month period, with adverse events serving as the primary outcome measure⁴⁰. While the ALBAN trial dampened enthusiasm for PD-L1 combinations and POTOMAC results are still maturing, the subcutaneous PD-1 inhibitor sasanlimab has recently provided the first definitive Phase III evidence of benefit in this specific population. The Phase III CREST trial (NCT04165317) randomized over 1,000 patients with BCG-naïve high-risk NMIBC to receive standard BCG induction and maintenance either alone or in combination with subcutaneous sasanlimab⁴². The first published results revealed that the combination significantly prolonged EFS compared to BCG alone (HR = 0.68; p = 0.0095), with a 36-month EFS rate of 82.1% versus 74.8% in the control arm⁴³. The benefit was consistent across key subgroups, including patients with T1 disease and those with CIS. Notably, an exploratory analysis presented at EMUC25 highlighted patients with very high-risk disease—defined as those with multiple recurrences, large tumors, or concurrent T1 and CIS with additional risk factors⁴⁴. In this notoriously difficult-to-treat population, the combination achieved a remarkable separation in outcomes, with a 36-month EFS of 84.3% versus 61.5% for BCG alone (HR = 0.34), suggesting that systemic priming may be most critical for patients with the highest burden of disease. Although the addition of sasanlimab increased the rate of grade ≥3 treatment-related adverse events (29.1% vs. 6.3%), the safety profile was considered manageable. These findings position sasanlimab as a potential practice-changing addition to the standard of care for BCG-naïve patients, particularly those with very high-risk features who may require systemic immune priming to prevent early failure.

The divergence between the positive results of the POTOMAC/CREST trials and the negative results of the ALBAN trial raises critical mechanistic questions. It is hypothesized that the specific timing of checkpoint inhibition relative to BCG instillation, or differences in the immune microenvironment of the enrolled populations, may dictate response. The failure of atezolizumab in the ALBAN trial suggests that simply adding systemic PD-L1 inhibition to BCG is not a universal solution, and that patient selection—potentially excluding “immune-cold” tumors that lack sufficient T-cell infiltration—may be required.

5. Novel intravesical therapies: Gene therapy and interleukin-15 superagonists

While systemic checkpoint inhibitors address the need for new mechanisms of action, they introduce systemic toxicity profiles that can be challenging for the elderly, comorbid population often affected by bladder cancer. Consequently, the development of novel intravesical agents that confine potent biological activity to the bladder has been a major focus of recent translational research. This effort has culminated in the approval of innovative gene therapies and cytokine superagonists that leverage distinct molecular pathways to overcome BCG resistance.

Gene therapy has finally moved from theoretical promise to clinical reality with the advent of nadofaragene firadenovec (Adstiladrin). This non-replicating adenoviral vector is engineered to deliver the human IFN-α2b gene directly into the urothelium^45-47. Once instilled, the viral vector transduces the bladder wall cells, effectively turning them into local “factories” that produce high concentrations of IFN-α2b protein, which then exerts direct cytotoxic and anti-angiogenic effects on the tumor.

In a pivotal Phase III trial, patients were stratified into two efficacy cohorts: those with CIS (with or without concomitant high-grade Ta or T1 disease) and those with high-grade papillary disease only. In the primary CIS efficacy population, nadofaragene firadenovec achieved a CR rate of 51% at three months, with a median duration of response of 9.7 months²⁰. Notably, 46% of responders maintained their disease-free status for at least 12 months. Nadofaragene firadenovec was well tolerated, with no dose-limiting toxicity or clinically significant treatment-related side effects. Most adverse events were transient and of mild-to-moderate severity (grade 1–2), and Grade 3–4 adverse events were reported in 29 patients; only six events were considered treatment-related. Based on these results, the FDA approved nadofaragene firadenovec for the treatment of BCG-unresponsive NMIBC with CIS (Table 1).

Additionally, secondary data indicate that combining post-treatment titers with the baseline fold-change of serum anti-human adenovirus type-5 antibodies may serve as a predictive marker for treatment efficacy⁴⁸. Beyond its efficacy, the agent offers a significant logistical advantage; it is administered on a less frequent dosing schedule of once every three months, reducing the burden of frequent catheterizations associated with standard induction and maintenance regimens. Building on the success of the pivotal Phase III trial, the evaluation of nadofaragene firadenovec has expanded beyond regulatory approval to assess its utility in broader clinical contexts. In 2023, a real-world study was conducted to track treatment patterns, clinician experiences, and oncologic outcomes among high-risk, BCG-unresponsive patients receiving therapy in routine practice⁴⁹. To further define the agent’s therapeutic scope, Ferring Pharmaceuticals subsequently unveiled the comprehensive “ABLE” clinical program in 2024. This initiative includes the ABLE-22 trial, which investigates potential synergies by comparing nadofaragene monotherapy against combinations with either intravesical gemcitabine/docetaxel or systemic pembrolizumab in patients with BCG-unresponsive carcinoma in situ⁵⁰. Simultaneously, the ABLE-32 study targets the intermediate-risk population, comparing the gene therapy against active surveillance, with a primary endpoint of 24-month recurrence-free survival (RFS)⁵¹. Although a third study regarding retreatment strategies for initial non-responders (ABLE-42) was originally proposed, it was subsequently withdrawn to prioritize resources for the active cohorts⁵².

Another breakthrough in localized therapy utilizes oncolytic viruses to selectively target tumor vulnerabilities. Cretostimogene grenadenorepvec (CG0070) is an oncolytic immunotherapy with a dual mechanism of action. It is a replication-competent oncolytic adenovirus designed to selectively replicate in and destroy cells with defective retinoblastoma (Rb) pathways, a common defect in bladder cancer⁵³. While specific RB1 gene mutations are found in a subset of patients, functional inactivation of the broader Rb pathway is ubiquitous in high-grade UC, occurring in more than 80–90% of cases. This widespread defect allows the virus to target the vast majority of high-grade tumors, thereby highlighting the need for pre-treatment biomarker selection in routine clinical practice.

The virus was also engineered to express granulocyte–macrophage colony-stimulating factor, which further stimulates the immune system following tumor cell lysis^54,55. CG0070 has been evaluated in a Phase II, open-label, single-arm, multicenter trial (BOND-002; NCT02365818) involving 45 patients with high-risk, BCG-unresponsive NMIBC who declined RC⁵⁶. Following a six-week induction course of weekly intravesical instillations, with subsequent monthly maintenance, the study reported an overall CR rate of 47% at six months. Outcomes appeared more favorable in patients with CIS, with approximately half achieving a CR, suggesting greater activity in this subgroup⁵⁶. In contrast, responses were limited among patients with papillary-only disease, particularly those with T1 tumors, in whom no CRs were observed at six months. Disease progression was infrequent among responders. Response assessment was based on cystoscopic evaluation, urine cytology, and biopsy, including mandatory random bladder biopsies at six months. Treatment-related adverse events were predominantly low grade (1–2), most commonly bladder spasm, hematuria, dysuria, and urinary urgency⁵⁶. These findings led to the Phase III trial BOND-003⁵⁷.

The administration schedule in BOND-003 consisted of weekly intravesical induction for six weeks, followed by a maintenance phase of three weekly instillations administered every three months for the first year, and then every six months for the second year. This maintenance schedule aims to sustain local immune activation while balancing patient convenience. Updated results from the Phase III BOND-003 trial, specifically involving patients with BCG-unresponsive CIS (±Ta/T1), have been notable, demonstrating a 75% CR rate at any time point in patients with BCG-unresponsive high-risk NMIBC.⁵⁸

A separate cohort is evaluating efficacy in patients with papillary-only disease, a group often distinct in biological response. The therapy has shown a favorable safety profile with no Grade ≥3 adverse events related to treatment, and approximately 46% of patients maintained a CR at the 12-month mark, validating the oncolytic approach as a potent bladder-sparing option⁵⁸. These data supported the FDA’s decision to assign Fast Track designation to CG0070 monotherapy in BCG-unresponsive CIS, including cases with concurrent Ta/T1 disease. Given the anticipation that cretostimogene grenadenorepvec (CG0070) could establish itself as a backbone treatment for NMIBC, research efforts have expanded to explore its synergistic potential in combinatorial regimens. Specifically, the CORE-001 trial, a Phase II open-label study, was designed to evaluate the efficacy of combining intravesical CG0070 with systemic pembrolizumab in patients with BCG-unresponsive CIS. The final results from this cohort demonstrated a robust CR rate of 57% at the 12-month landmark, meeting the primary endpoint⁵⁹. Notably, the overall CR rate at any time point was 83%, with highly durable responses observed; nearly all patients who were disease-free at one year maintained their status at the 24-month assessment⁵⁹. Furthermore, the combination exhibited a favorable safety profile with no evidence of synergistic toxicity, suggesting that adding checkpoint inhibition does not compound the local adverse events associated with oncolytic viral therapy.

Beyond these advanced-stage viral platforms, the pipeline includes investigational non-viral gene therapy candidates currently in early-phase development. EG-70 is a novel non-viral nanoparticle plasmid. Unlike viral vectors, which can induce neutralizing antibodies, EG-70 encodes IL-12 and activators of the retinoic acid-inducible gene-1 receptor⁶⁰. This dual-mechanism approach stimulates an innate immune response via retinoic acid-inducible gene-1 and an adaptive response through IL-12 secretion⁶⁰. The clinical utility of this agent is currently under investigation in the multicenter Phase I/II LEGEND trial (NCT04752722). In the completed Phase I dose-escalation segment involving patients with BCG-unresponsive CIS, EG-70 demonstrated a favorable safety profile and achieved a promising overall CR rate of 73% in this small Phase I cohort⁶¹. Consequently, the study has advanced to its registration-enabling Phase II stage, utilizing a recommended dose of 800 µg administered via intravesical instillation on weeks 1, 2, 5, and 6 of 12-week cycles. This expansion aims to validate the agent’s efficacy, specifically evaluating the CR rate at 48 weeks, while confirming its ability to elicit local immune activation with minimal systemic toxicity.

Complementing these gene-based approaches is the relatively recent FDA approval of nogapendekin alfa-inbakicept (Anktiva or N-803) in April 2024, marking the arrival of IL-15 superagonists in the urologic armamentarium⁶². This first-in-class fusion protein mimics the physiological trans-presentation of IL-15, selectively activating NK cells and CD8⁺ memory T-cells without stimulating regulatory T-cells, which can dampen the immune response^62,63. In the Phase II/III QUILT-3.032 trial, the combination of intravesical Anktiva and BCG achieved an approximately 71% CR rate in Cohort A, which enrolled patients with BCG-unresponsive CIS (with or without papillary tumors)⁶³—the FDA approval is currently limited to this population, distinct from Cohort B, which evaluated patients with papillary-only disease. The most significant finding is the durability of this response; the median duration of response reached 26.6 months, with a 24-month progression-free survival rate of over 84%. The ability of this combination to help over 90% of responders avoid RC underscores the potential of cytokine superagonists to synergize with BCG and restore immune sensitivity in previously resistant tumors. Crucially, this oncologic control does not appear to come at the expense of the patient’s well-being. A secondary analysis of the QUILT-3.032 trial focused on patient-reported outcomes and found that global health and physical function scores remained stable relative to baseline throughout the treatment period⁶⁴. This preservation of quality of life, taken together with the robust efficacy data, reinforces a favorable risk-benefit profile for the N-803 + BCG regimen.

6. Device-assisted therapies and novel drug delivery systems

Beyond novel pharmacological agents, recent advancements have focused on enhancing the delivery and efficacy of established chemotherapeutics through mechanical or physical means. Standard intravesical chemotherapy often fails due to limited drug penetration into the urothelium and rapid washout during voiding. To address these pharmacokinetic limitations, device-assisted therapies such as hyperthermic intravesical chemotherapy and sustained-release implantable devices have emerged as valuable tools in the BCG-unresponsive landscape.

Hyperthermia has been utilized to synergize with chemotherapy by increasing cell membrane permeability and impairing DNA repair mechanisms in tumor cells⁶⁵. Systems such as the COMBAT BRS recirculate heated chemotherapy, typically mitomycin C (MMC), maintaining bladder wall temperatures between 41°C and 43°C. Recent data indicate that this approach is effective—a multicenter study of hyperthermic intravesical chemotherapy with MMC reported a 24-month RFS rate of 57.4% even in the difficult-to-treat BCG-unresponsive population⁶⁶. Similarly, research into microwave-induced hyperthermia highlights the importance of patient selection. In a Phase III randomized controlled trial involving high-risk NMIBC patients who failed BCG induction, the combination of MMC and microwave hyperthermia yielded a two-year DFS of 35%¹⁷. Crucially, a pre-planned sub-analysis revealed that efficacy is highly dependent on tumor morphology. While the treatment showed lower response rates in CIS, it demonstrated a significantly improved DFS in non-CIS papillary tumors compared with controls (53% vs. 24%). Distinct from thermal approaches, Electromotive Drug Administration has shown promise in similar settings. A prospective Phase II study of Electromotive Drug Administration-MMC in BCG-refractory patients reported durable responses that also varied by tumor type, with three-year disease-free rates reaching 75% for Ta Grade 3 tumors and 71.4% for T1 Grade 3, but dropping to 50% for CIS and 25% for mixed disease¹⁶.

Parallel to hyperthermia, novel intravesical drug delivery systems are revolutionizing how chemotherapy is administered. The TAR-200 system is an indwelling, pretzel-shaped device inserted into the bladder as an office-based procedure that provides the continuous, sustained release of gemcitabine over a 21-day period. This method bypasses the “dwell-time” limitations of standard instillations. In the Phase IIb SunRISe-1 trial, TAR-200 monotherapy demonstrated a CR rate of 82.4% in patients with BCG-unresponsive CIS (with or without papillary disease), with a median duration of response reaching 25.8 months⁶⁷. Furthermore, in patients with papillary-only disease, the device maintained a 12-month DFS rate of 70.2%, confirming its utility across different high-risk phenotypes. The device was well-tolerated and addresses the issue of treatment adherence and drug exposure duration. Following the positive results from the SunRISe-1 trial, the gemcitabine intravesical system (TAR-200) was approved by the FDA in September 2025 for the treatment of patients with BCG-unresponsive NMIBC with CIS.

Building on these promising Phase II data, two pivotal Phase III trials are currently underway to validate this technology in broader populations. The SunRISe-3 study targets patients with BCG-naïve high-risk disease, randomizing them into three distinct arms to compare TAR-200 monotherapy and TAR-200 combined with systemic cetrelimab against the standard of care, BCG induction, and maintenance⁶⁸. Concurrently, the SunRISe-5 trial specifically addresses BCG-unresponsive patients presenting with papillary-only tumors, aiming to recruit approximately 250 participants to evaluate TAR-200 against investigator-selected intravesical chemotherapy. Together, these large-scale investigations seek to establish sustained drug delivery as a foundational strategy for managing both early and refractory high-risk NMIBC.

Similarly, the TAR-210 system is being developed to deliver erdafitinib for patients with specific FGFR alterations, representing the convergence of precision medicine and advanced device engineering. Early clinical data for this device have shown preliminary activity, demonstrating CR rates of 82% in BCG-refractory patients and 87% in those with intermediate-risk recurrences^69,70. It is important to note that these high response rates are based on preliminary early-phase data; definitive efficacy and long-term survival benefits must be confirmed. To validate these findings, the Phase III MoonRISe-1 trial (NCT06319820) is now underway to compare TAR-210 directly with standard intravesical chemotherapy in patients with intermediate-risk NMIBC harboring these specific genetic mutations.

7. Targeted therapies and antibody-drug conjugates

The molecular characterization of UC has unveiled specific mutations and surface markers that can be exploited for targeted therapy. Approximately 65–75% of NMIBC cases harbor mutations in the FGFR3 gene, making it an attractive therapeutic target⁷¹. Erdafitinib, a pan-FGFR tyrosine kinase inhibitor already approved for metastatic disease, is now showing efficacy in the non–muscle-invasive setting⁷². The Phase II, open-label, randomized THOR-2 trial was designed to rigorously evaluate oral erdafitinib against investigator-selected intravesical chemotherapy in patients with high-risk, BCG-unresponsive NMIBC harboring specific FGFR alterations⁷³. While the standard starting dose for metastatic UC is 8 mg daily with uptitration to 9 mg, the THOR-2 trial (Cohort 1) utilized a modified oral dose of 6 mg once daily to improve tolerability in this specific non–muscle-invasive population while maintaining efficacy. Although enrollment in Cohort 1 was curtailed prior to reaching the full target, the data from the 73 participants demonstrated compelling efficacy. While the median RFS was not yet reached in the erdafitinib arm, the survival curves showed early and sustained separation: the 6- and 12-month RFS rates were 96% and 77% for patients treated with erdafitinib, respectively, compared to 73% and 41% for those receiving standard chemotherapy⁷³. This benefit appeared consistent across various subgroups, regardless of prior BCG exposure or tumor stage. However, this systemic efficacy comes with a trade-off, as serious adverse events were more frequent in the erdafitinib group (22%) than in the chemotherapy arm (13%), underscoring the need to carefully balance oncologic control with toxicity in future investigations. Specifically, the systemic side effects of oral FGFR inhibitors, such as hyperphosphatemia and ocular toxicity, have spurred interest in intravesical formulations—such as the aforementioned TAR-210—to maximize local efficacy while minimizing systemic exposure.

Efforts to exploit the FGFR pathway have also extended to rogaratinib, a broad-spectrum inhibitor targeting FGFR1–4^74,75. While this agent demonstrated modest activity in broader UC cohorts, achieving an objective response rate of 21% and survival outcomes comparable to chemotherapy, its translation to the non–muscle-invasive setting proved challenging. A dedicated Phase II trial (BLASST‐3) specifically designed for BCG-refractory, high-risk patients with FGFR overexpression was initiated but ultimately withdrawn due to an inability to complete enrollment. This outcome underscores the logistical challenges inherent in conducting biomarker-driven trials within this specific patient population.

Parallel to these kinase inhibitors, APL-1202 (Nitroxoline) is emerging as a potential oral alternative, addressing the quality-of-life burden associated with frequent catheterizations. APL-1202 is a reversible inhibitor of methionine aminopeptidase 2 (MetAP2), an enzyme critical for tumor angiogenesis and cell growth⁷⁶. In a Phase II trial involving BCG-unresponsive patients, this oral agent demonstrated a one-year recurrence-free rate of 54.3%, suggesting that methionine aminopeptidase 2 may be a potential therapeutic target. By offering a systemic, non-invasive route of administration, it represents a significant departure from standard local therapies⁷⁷. However, current evidence is limited to Phase II findings, and larger randomized studies are necessary to confirm these RFS rates and establish the drug’s comparative efficacy against standard intravesical regimens.

Antibody-drug conjugates are also migrating from the metastatic to the NMIBC setting. These agents consist of a monoclonal antibody linked to a potent cytotoxic payload, allowing for the precise delivery of chemotherapy to cells expressing specific antigens. Enfortumab vedotin, which targets Nectin-4, and disitamab vedotin, which targets human epidermal growth factor receptor 2, are currently under investigation. For instance, disitamab vedotin is being evaluated in combination with checkpoint inhibitors (TRUCE-04 trial, NCT05495724), based on the hypothesis that the release of cytotoxic agents within the tumor microenvironment will further stimulate the immune response⁷⁸.

Furthermore, targeted toxins such as oportuzumab monatox (Vicinium) have been rigorously explored for their ability to deliver cytotoxic payloads directly to epithelial cell adhesion molecule-expressing urothelial cells. This recombinant fusion protein utilizes a Pseudomonas exotoxin payload that, upon internalization via endocytosis, irreversibly arrests protein synthesis to induce cell death⁷⁹. In the pivotal Phase III VISTA trial (NCT02449239), which treated 134 BCG-unresponsive patients using an intensive 12-week induction schedule, the agent achieved a CR rate of 40% at three months⁸⁰. The durability was notable among responders, with 52% maintaining their response at one year and a median duration of response of 9.4 months. Outcomes were also evaluated in a distinct papillary-only cohort, in which 50% of patients remained recurrence-free at the 12-month mark.

Despite these signals of efficacy, the FDA ultimately declined approval for use in BCG-unresponsive NMIBC, leading to a pause in its monotherapy development. The regulatory decision reflected concerns that, while the drug showed clinical activity, the CR rates and durability did not meet the rigorous thresholds required to demonstrate a meaningful clinical benefit over existing therapies in this refractory setting. Nevertheless, investigation continues into combinatorial strategies, most notably pairing Vicinium with the systemic PD-L1 inhibitor durvalumab in BCG-unresponsive NMIBC. The final results from a Phase I trial (NCT03258593) demonstrated that this combination achieved a DFS rate of 47% at the 12-week primary evaluation, with 40% of patients maintaining their response at six months⁸¹. The median RFS was reported at 13.2 months, and biomarker analysis indicated a corresponding decrease in urinary epithelial cell adhesion molecule levels alongside an upregulation of PD-L1, suggesting active biological engagement⁸¹. While 27% of patients experienced Grade 3 or higher adverse events, the regimen was generally deemed safe and tolerable for this high-risk population.

To assist in navigating this complex therapeutic landscape, Table 1 provides a comprehensive overview of the key agents discussed, summarizing their mechanisms of action, pivotal trial outcomes, and current regulatory status. In parallel, Figure 1 presents a simplified, phenotype-based clinical decision framework to support therapy selection in high-risk and post-BCG NMIBC.

Figure 1. Phenotype-driven clinical decision framework guiding therapeutic selection in patients with high-risk and post-BCG NMIBC. Abbreviations: BCG: Bacillus Calmette–Guérin; CIS: Carcinoma in situ; FDA: Food and Drug Administration; FGFR: Fibroblast growth factor receptor; NMIBC: Non–muscle-invasive bladder cancer.

8. Therapeutic efficacy and safety and real-world implementation challenges

A critical consideration in the modern post-BCG therapeutic landscape is the balance between therapeutic efficacy and safety and real-world feasibility. Systemic immunotherapies, while clinically effective, introduce a distinct toxicity profile characterized by immune-related adverse events. As reported in trials such as KEYNOTE-057 and POTOMAC, approximately 15–20% of patients receiving ICIs experience Grade ≥3 adverse events, including colitis, pneumonitis, and endocrinopathies^31,41. Given the advanced age and comorbidity burden typical of the NMIBC population, these systemic risks require careful patient selection, robust monitoring, and multidisciplinary management.

In contrast, emerging intravesical therapies such as nadofaragene firadenovec, cretostimogene grenadenorepvec, and TAR-200 largely confine toxicity to the genitourinary tract (e.g., dysuria, urgency, hematuria), thereby minimizing systemic morbidity. However, these bladder-directed approaches introduce distinct logistical and resource challenges. While nadofaragene offers a comparatively convenient once-every-three-months dosing schedule, other platforms—such as cretostimogene and N-803—require intensive induction and maintenance regimens that demand high patient adherence and clinic infrastructure. Device-assisted systems such as TAR-200 further require procedural expertise and staff training. Consequently, therapeutic decision-making in NMIBC is increasingly shaped not only by efficacy but also by patient frailty, tolerance for repeated catheterization, and the operational capacity of the treating center.

9. Future directions: Biomarkers and precision medicine

As the therapeutic landscape expands, the “one-size-fits-all” approach is becoming obsolete. Future progress relies on integrating biomarkers to guide treatment selection. PD-L1 expression, while extensively studied, remains a complex marker in NMIBC. Currently, it serves more as a prognostic indicator of high-risk disease rather than a definitive binary predictor of checkpoint inhibitor response, highlighting the need for standardized assays. In contrast, precision oncology has been validated through FGFR3 genomic profiling, which is now required for patient selection for erdafitinib and crucial for triaging patients to trials of TAR-210. Furthermore, the rise of intravesical gene therapies has underscored the importance of host immune factors; specifically, high baseline anti-adenovirus antibody titers may limit the efficacy of viral vectors such as nadofaragene firadenovec, suggesting that pre-treatment serological testing could identify patients better suited for alternative delivery platforms.

10. Conclusion

The management of BCG-unresponsive NMIBC is undergoing a historic transformation. For decades, the treatment algorithm was binary: BCG or RC. Today, the landscape has expanded into a diverse mosaic of options that allow for personalized, bladder-sparing strategies. We have moved from a “one-size-fits-all” immunotherapy approach to a nuanced era involving systemic checkpoint inhibition, precision gene therapy, cytokine superagonists, and advanced device-assisted delivery systems.

The FDA approvals of pembrolizumab, nadofaragene firadenovec, and, most recently, Anktiva, provide clinicians with validated tools to delay or prevent cystectomy in high-risk patients. However, this abundance of options introduces a new challenge: sequencing and selection. Future research must prioritize the identification of predictive biomarkers, such as PD-L1 status, FGFR alterations, or urinary cytokine profiles, to determine which patients will benefit from which therapy. Furthermore, as shortages of BCG continue to plague the urologic community, the definition of “standard of care” may shift, with novel agents potentially moving earlier in the treatment paradigm to the BCG-naïve setting.

Looking ahead, the potential efficacy of treatments is expected to improve through combinatorial approaches that synergize local cytotoxicity with systemic immune modulation, though toxicity trade-offs must be carefully monitored. To accelerate these developments, adaptive trial designs are essential to reduce the cost and duration of drug development. Furthermore, the ongoing global BCG shortage compels a re-evaluation of the treatment paradigm, potentially driving the early integration of these novel agents into the BCG-naïve setting to stabilize the supply chain.

Finally, as these sophisticated therapies enter routine practice, economic and access considerations must be addressed to ensure that logistical burdens and costs do not widen the disparity between academic centers and community practices. Ultimately, the goal remains clear: to preserve the bladder and quality of life without compromising oncological survival, a goal that is now closer to reality than ever before.