NMIBC is often first treated with Bacillus Calmette-Guerin (BCG), a biotic therapy originally used as a killed tuberculosis vaccine strain, to train the immune system and evasion of this therapy may result in tumor progression ^[1][21]. A first-line adjuvant therapy, BCG treatment is effective in about 50% of patients, with the rest either failing to respond, relapsing, or having adverse events ^[2][22]. Originally thought to be wholly immune-centered around CD8⁺ T cells, evidence exists that also demonstrates some direct tumoricidal activity of the bacteria (putatively via oxidative stress or necrotic pathway activation) but increases in PD-L1 expression on tumor cell surfaces might explain BCG evasion ^[3][23]. However, for non-responders, valrubicin or pembrolizumab (anti-PD-L1) are the only FDA/European Medicines Agency-approved therapies for NMIBC if BCG is ineffective ^[4][24].

Although radical cystectomy with pelvic lymph node dissection (RC/PLND) is the current gold standard for MIBC, can result in an R0 (complete cure) condition, and has additional benefits in preventing metastasis, it has side effects that can be devastating for quality of life (incontinence, impotence, neurologic damage) and some patients are unsuitable for such surgery. It is the main line of treatment for progressive or de novo MIBC; however, RC alone is also not a definitive treatment, as 5-year survival rates of 40–60% and recurrence in as little as 12 months have been reported ^[5][25]. For this reason, RC is usually combined with neo-adjuvant chemotherapy (NAC) to maximize tumor control before and after surgery. Neo-adjuvant cisplatin with RC is the currently recommended multimodal treatment standard for MIBC ^[6][26].

While the use of chemotherapy with RC seems obvious, NAC has been reported as underutilized, with less than 20% of patients in the US between 2004–2014 having received it ^[7][27]. Meanwhile, a recent meta-study of 35,738 patients in 13 reports found that only 17.2% of patients underwent NAC regimens ^[8][28]. Since complete, partial, and downstaged response rates were 16.6%, 14.6%, and 45.0% in that study, NAC may be an important weapon against UC progression ^[8][28]. Another meta-study of 8 reports found NAC + RC was superior to RC alone with regard to overall survival (OS; HR 0.79; 95% CI: 0.68–0.92, p = 0.002), bolstering the utility of NAC to provide improved MIBC prognoses ^[9][29].

Trimodal therapy (TMT), consisting of combined chemo- and radiotherapies plus TURBT, has been adopted as an alternative to RC/PLND in selected patients ^[10][30]. Such bladder-sparing improves quality of life but requires optimal patient selection with regard to co-morbidities and tumor status (ideally cT2 with no carcinoma in situ) ^[11][31]. However, an insufficient number of clinical trials prevents the full clarification of survival and quality-of-life improvements that may be possible with TMT. A recent report simulated 500,000 patients with a 2-D Markov model and, in comparisons between RC and TMT, found that TMT had slightly higher quality of life in elderly patients while RC had better overall survival and life quality in younger patients ^[10][30]. Similarly, a US study of 2306 military veterans with MIBC found that TMT was associated with comparable survival to RC with neoadjuvant chemotherapy but only in patients older than 65 years of age ^[12][32]. Thus, the utility of TMT may be comparable at best in some patients but worse in older patients. Additionally, a 2018 meta-study of 57 total studies and 30,293 patients found that, while TMT mean 10-year OS was insignificantly lower than RC (30.9% TMT vs. 35.1% RC, p = 0.32), it was chemotherapy response that determined the best survival results with TMT ^[13][33]. The concept of RC for long-term survival superiority was also questioned by a 2020 metasudy by Ding et al that found superior OS results for RC after 10 years but only when Charlson comorbidity scores were 0 ^[14][34]. Furthermore, within that 10-year timeframe, TMT was comparable, indicating that TMT is a valid therapeutic option for patients who are unsuited for RC or who do not wish to undergo radical resection ^[14][34].

These data indicate that, while R0 resection is considered curative, long-term survival also depends on a complete response to chemotherapy and selection of chemoagents by using predictive models for response is therefore a critical component of improving TMT performance. Several proposed and current trials that may finely tune response rate predictive models are currently exploring combinations of immune checkpoint inhibitors and radiation for cisplatin and RC-ineligible patients ^[15][16][35,36]. As such, maintaining a maximum level of tumor control with TMT requires accurate and precise biomarkers to select the proper chemotherapy agents as well as to monitor progress after treatment. The current requirement of frequent CT/MRI or cystoscopy to monitor progress is not clinically feasible and cannot accurately predict resistance to therapy.

Since atezolizumab was first approved by FDA for metastatic UC in 2016, immune checkpoint inhibitors (ICIs) to PD-1, PD-L1, or CTLA-4, which target angiogenesis and sensitize the immune response to limit both tumor growth and metastasis, are frequently employed as an adjuvant therapy combined with chemotherapy. Anti-PD-L1 therapy on cancer cells prevents binding of PD-1 on T cells to increase apoptosis while CTLA-4 blocks CD80 and CD86 activity, downregulating Treg-mediated immunoregulation and increasing CD8⁺ cytotoxic activity ^[17][44]. In this fashion, antitumor T cell activity is maximized. Both types of drugs are often combined (ICI-ICI) to starve the tumor and increase immune effectiveness since a metastudy of 2 RCTs with 1518 total patients found ICI treatment alone (atezolizumab or nivolumab) was not significantly useful in high-risk muscle-invasive UC ^[18][45]. Conversely, a study (CheckMate 032) of combined PD-1 and CTLA-4 inhibitors (nivolumab 1 mg/kg plus ipilimumab 3 mg/kg) found response rate of 38.0% in 92 patients receiving both drugs ^[19][46]. Adverse events must also be considered, as a metastudy of 21 reports with 11,454 patients total who received nivolumab, pembrolizumab, atezolizumab, or ipilimumab found a higher incidence of non-fatal adverse events ^[20][47]. With the promising results of adjuvant immunotherapy, clinical trials of neoadjunctive immunotherapy have been intensively carried out.

Much effort has been made to look for reliable biomarkers for predicting response to ICIs, such as PD-L1 or CD8 expression by immunohistochemistry ^{[21][22][23][24]}[48,49,50,51] and/or tumor mutation burden (TMB) ^[25][52]. Unfortunately, these biomarkers have not yet to be verified in large clinical trials ^[24][26][51,53].

As mentioned in Section 2, FGFR3 is a common mutation found in patients with MIBC ^[27][56]. Erdafitinib, a pan-FGFR tyrosine kinase inhibitor, is the first FDA-approved targeted therapy for mUC with susceptible FGFR2/3 alterations following platinum-containing chemotherapy. A phase II trial of 99 enrolled patients with local advanced and unresectable/metastatic UC with an FGFR3 mutation or FGFR2/3 fusion observed disease progression in all patients following chemotherapy ^[28][57]. The confirmed response rate was 40% while an additional 39% of patients were stabilized. For 22 patients with previous immunotherapy, the erdafitinib response rate was 59%. At 24 months median follow-up, the median overall survival was 11.3 months. Adverse events of >grade 3 related to treatment occurred in 46% of patients while 13% had to discontinue erdafitinib due to adverse events ^[28][57]. Based on those results, several other FGFR inhibitors are being evaluated, including infigratinib, which has demonstrated promising activity ^[29][58].

Even with an absence of viable biomarkers for prediction and treatment management, development of new modalities for bladder cancer continue to increase specificity by combining drugs with antibody conjugates (antibody-conjugated drugs; ADC). Unlike ICIs, these drugs are manufactured to deliver cytotoxic molecules via antibodies (usually IgG) engineered for strong interaction with specific tumor antigens and very little cross-reactivity ^[30][59]. Connecting links between the antibody and payload are usually constructed of disulfide- or protease-dependent bonds in order to prevent premature release of the payload and to exploit the acidic pH and ROS-intensive microenvironments of tumors that are likely to sever the linker ^[30][59]. Then, cytotoxic molecules such as auristatins (microtubule destabilizers), maytansinoids, DNA alkylators, or proteotoxins (protein synthesis inhibitors) are connected to the scaffold as effector payloads to slow tumor cell growth and prevent replication. Other antibody-based fusion techniques concurrently being developed to stimulate immunogenic responses (e.g., ALT-803) and deliver current ICIs with higher specificity (e.g., ATOR-1015) have been extensively reviewed by Bogen et al. ^[31][60].

With the discovery of fucolsylated glycans as potential biomarkers for pancreatic, MIBC, and other cancers, the potential of lectin-targeted payload delivery to MIBC has become feasible ^[32][33][61,62]. Lectin specificity to these glycans is possible with recombinant technology and future sequencing studies on diverse bladder cancer specimens could allow for a glycan-lectin binding library to be constructed for targeting both non-invasive and invasive bladder cancers ^[34][63]. Additionally, these lectins are amenable to conjugation with nontoxic, photoreactive dyes that respond to near-infrared (NIR) light by conformational changes that induce necrotic death in cells ^[35][64]. A recent paper by Kuroda et al. has demonstrated the feasibility of this system for pancreatic cancer in a murine model ^[36][65]. Since NIR is harmless to human tissue and cystoscopes already have illumination capability, addition of NIR light and lectin-conjugated dye systems to TURBT may be a possibility that removes residual disease and promotes complete response.