Neoadjuvant treatment patterns and biomarker selection in muscle-invasive bladder cancer

Abstract

As the feasibility of risk-adaptive bladder-sparing treatment is increasingly validated, the prospects for neoadjuvant therapy in muscle-invasive bladder cancer (MIBC) are rapidly evolving. For patients seeking effective and tolerable treatment options, platinum-based chemotherapy, particularly dose-dense MVAC (ddMVAC), remains the preferred standard. However, the emergence of novel interventions such as immune checkpoint inhibitors (ICIs), FGFR inhibitors, and antibody–drug conjugates (ADCs) offers promising alternatives, especially for those ineligible for cisplatin-based regimens. Ongoing clinical trials, including KEYNOTE-B15, RC48-C017, and NIAGARA, are actively investigating the efficacy of combining these agents with existing neoadjuvant therapies, aiming to establish new first-line treatment options. Although predictive models based on histological features, DNA damage repair (DDR) genes, molecular subtyping, liquid biopsies, and in vitro organoids have demonstrated potential in guiding treatment selection, the clinical translation process remains slow. There is a pressing need to accelerate the exploration of genetic heterogeneity in MIBC and to validate the clinical utility of emerging biomarkers to optimize patient selection for neoadjuvant therapy. This review will comprehensively examine the evolution of neoadjuvant treatment paradigms, focusing on high-quality evidence from evidence-based medicine and translational clinical research, with the aim of enhancing and updating readers' knowledge of neoadjuvant therapy for MIBC and providing insights for future practice and research directions.

Highlights

• Cisplatin-based neoadjuvant chemotherapy remains the standard treatment for patients with localized MIBC recommended by major guidelines, with a 5-year overall survival benefit of 5-8% and the opportunity for sequential bladder preservation. However, concerns about its toxicity, drug resistance, and delayed timing of surgery have led to a change in the treatment landscape.

• The commonly utilized neoadjuvant chemotherapy regimens are dose-dense MVAC (ddMVAC) and gemcitabine-cisplatin (GC). For patients who can tolerate higher toxicity, ddMVAC may provide better long-term benefits, whereas GC is somewhat more favorable in terms of cost control. There is currently no consensus on the optimal duration of treatment.

• ICI, targeted drugs, ADC, etc. have been approved for the treatment of mUC one after another. Most of these emerging MIBC neoadjuvant alternative drugs have shown promising prospects in phase 2 clinical trials, supporting the neoadjuvant combination therapy around ICI to become a new first-line option in the future.

• With the deepening of understanding of multi-omics phenotypic characteristics of patients, neoadjuvant therapy needs to move towards the era of "precision" to bring the best benefits to patients in different strata.

• Well-designed clinical trials are urgently needed to evaluate and validate potential biomarkers and accelerate the integration of reliable and practical efficacy stratification tools into standard patient care.

Introduction

Bladder cancer (BCa) is a malignant tumor of the urinary system that predominantly affects men, with an age-standardized incidence rate of 20 per 100,000 men [1, 2]. Recent data reveal that BCa is now the ninth most common cancer worldwide, with over 610,000 new cases and an estimated 220,000 deaths annually [3]. Histologically, BCa comprises approximately 75% pure urothelial carcinoma (UC) and 25% other “variant” types (including squamous differentiation, glandular differentiation, micropapillary, plasmacytoid, nested, sarcomatoid, or neuroendocrine variants), highlighting the complexity of BCa diagnosis and treatment [4]. Among patients, 70–75% present with non-muscle-invasive bladder cancer (NMIBC), which generally has a favorable prognosis with a 5-year survival rate of over 90%. However, 50–70% of these patients experience recurrence after transurethral resection of bladder tumor (TURBT), and 21% of high-grade T1 cases progress to muscle-invasive bladder cancer (MIBC) within 5 years [5–7]. Compared to NMIBC, MIBC is more prone to lymphatic and distant organ metastasis, with up to 50% of cases developing distant metastases after radical cystectomy (RC), resulting in a contemporary 5-year survival rate of less than 60% [8, 9].

Neoadjuvant therapy for MIBC began in the 1980s, when the MVAC regimen (methotrexate, vinblastine, doxorubicin, and cisplatin) elicited a response in 63% of 41 patients [10]. Subsequent phase III trials such as Nordic Trial I, Nordic Trial II, SWOG-8710 and BA06 30894 followed up the potential survival benefits of several different cisplatin-based neoadjuvant chemotherapy (NAC) regimens in the long term [11–13]. Over the past decade, cisplatin-based NAC has become the standard therapy for resectable T2-4a cN0 M0 patients, with commonly used regimens including three to six cycles of dose-dense MVAC (ddMVAC), or up to four cycles of GC (gemcitabine and cisplatin) [14] (Fig. 1). For patients suitable for cisplatin, NAC provides an absolute 5–8% overall survival (OS) benefit at 5 years [15]. However, due to inherent frailty, comorbidities, and renal impairment in the BCa patient population, up to 50% of patients are deemed ineligible for NAC, leading to a paradox where less than 25% of patients receive cisplatin-based NAC in clinical practice, despite level I evidence supporting its use [16].

Fig. 1.

Timeline of the development of neoadjuvant therapy for MIBC. In the 1980s, MVAC became the first chemotherapy regimen used for neoadjuvant therapy in MIBC. Over the following two decades, the results of several large Phase III randomized trials established NAC as the standard treatment for localized MIBC. Recently, ICI, targeted therapies, and ADCs have been approved for mUC treatment, leading to updates in the neoadjuvant treatment landscape for MIBC. BCa Bladder Cancer; MVAC Methotrexate, Vinblastine, Doxorubicin, and Cisplatin; mUC Metastatic Urothelial Carcinoma; CA Cisplatin and Doxorubicin; RT Radiotherapy; GC Gemcitabine and Cisplatin; ddMVAC Dose-Dense MVAC; SCUC Small Cell Urothelial Carcinoma; CMV Cisplatin, Methotrexate, Vinblastine; ddGC Dose-Dense GC; ICI Immune Checkpoint Inhibitor; ADC Antibody–Drug Conjugate; NAC Neoadjuvant Chemotherapy; RCT Randomized Controlled Trial; MIBC Muscle-Invasive Bladder Cancer (Created with BioRender.com)

Immune checkpoint inhibitors (ICIs) as an alternative or adjunct to NAC are increasingly being tested in numerous trials. Pembrolizumab and atezolizumab have both been approved by the FDA as first-line alternatives for cisplatin-ineligible patients whose tumors express PD-L1 [17–19]. The latest ABACUS trial results report a 2-year disease-free survival (DFS) and OS of 68% and 77%, respectively, in cisplatin-ineligible patients treated with single-agent atezolizumab as neoadjuvant therapy [20]. Antibody–drug conjugates (ADCs) have recently shown significant efficacy in improving outcomes for patients with locally advanced or metastatic UC, with FDA approval [8, 21]. Enfortumab vedotin (EV) combined with pembrolizumab significantly extended OS in patients with unresectable advanced UC, with a median OS of 31.5 months compared to 16.1 months with chemotherapy [22]. The continuous updates in clinical evidence underscore the ongoing optimization of treatment strategies and the complexity of selecting the best neoadjuvant therapy for different patients. Additionally, risk-adaptive neoadjuvant therapy offers hope for sequential bladder-sparing approaches, which could improve patients' quality of life. However, there is still a lack of consensus regarding the combinations of neoadjuvant regimens, their specific details, and the selection of optimal benefit populations. This review will comprehensively explore the evolution of neoadjuvant treatment landscapes, with a particular focus on the latest high-quality evidence from evidence-based medicine and translational clinical research. In addition to meta-analyses of efficacy and safety, we will also discuss the potential benefits of neoadjuvant therapy for different patient groups, aiming to provide readers with a quick understanding of neoadjuvant therapy for MIBC and insights for future practice and research directions.

Conventional neoadjuvant first-line therapy

The 2024 guidelines from the European Association of Urology (EAU), American Urological Association (AUA), and National Comprehensive Cancer Network (NCCN) uniformly advocate for cisplatin-based NAC for patients with non-metastatic MIBC who are suitable candidates for cisplatin-based treatment [2, 23–25]. The AUA’s endorsement of NAC as the standard therapy for MIBC is primarily supported by two substantial phase III randomized controlled trials (RCT), SWOG 8710 (MVAC) and BA06 30894 (CMV, cisplatin, methotrexate, and vinblastine), both of which demonstrated that cisplatin-based combination chemotherapy significantly improved pathological complete response (pCR) rates and 5-year OS compared to RC alone [13, 26]. A meta-analysis involving 15 RCTs, encompassing 3,285 patients, further corroborates the significant survival benefit of NAC, reporting a hazard ratio (HR) of 0.87 with a 95% confidence interval (CI) of 0.79 to 0.96 [15]. The contribution of NAC to prognosis may also be closely related to its expected advantages in early eradication of micrometastatic disease, facilitating early evaluation of chemotherapy efficacy, and improving patient compliance with surgery.

Additionally, compared to patients who did not receive NAC, the administration of NAC does not appear to increase the risk of short-term postoperative complications. After adjusting for baseline clinical characteristics, NAC remained insignificantly associated with short-term complications (OR 1.06, 95% CI 0.82–1.39) and mortality (OR 0.75, 95% CI 0.36–1.55) [27]. A recent intriguing study suggests that the cumulative 90-day complication rate following novel immunotherapy-based neoadjuvant treatment appears higher than previously reported for robot-assisted radical cystectomy (RARC) alone or for chemotherapy-based neoadjuvant regimens, particularly with a notably high rate of infectious complications (59.6%) [28]. This indicates that the use of neoadjuvant immunotherapy warrants careful consideration. It is important to highlight that the recommendation for neoadjuvant immunotherapy as a first-line treatment remains confined to the context of clinical trials, with a strong recommendation strength from the EAU [2].

On the other hand, concerns regarding the toxicity and drug resistance of NAC persist. A significant number of patients may be excluded due to their intolerance to cisplatin, while others, even after undergoing NAC, might face delayed curative surgery due to their insensitivity to chemotherapy [29]. Regarding the general critical time frame for performing RC following NAC, a recent meta-analysis has indicated significant heterogeneity in terms of survival benefits, although a 3-month threshold is generally considered acceptable. While there is currently no prospective data suggesting that surgical delays caused by NAC adversely affect survival, retrospective analyses suggest that such delays should not exceed 8 to 12 weeks in order to mitigate the risks of potential disease progression or reduced OS [30–34].

Combination regimen and clinical evidence of NAC

MVAC

MVAC was the earliest regimen applied in the neoadjuvant treatment of MIBC. In 1980s, the oncological outcomes of 30 patients with MIBC who received preoperative MVAC chemotherapy followed by RC were first reported, showing a 33% pCR rate and 17% downstaging to < ypT2N0 [10]. Phase III RCT SWOG-8710 observed that three cycles of neoadjuvant MVAC significantly increased the number of patients with no residual disease compared to the RC-alone group (38% vs. 15%, p < 0.001) and was associated with improved prognosis [26]. Another randomized phase III study, JCOG0209, although not confirming a difference in OS between two cycles of neoadjuvant MVAC and RC alone, demonstrated a significant increase in the proportion of ypT0 patients with MVAC [35]. However, the toxicity associated with MVAC, including bone marrow suppression and neutropenia, raises significant concerns, with up to 82% of patients exposed to MVAC experiencing grade 3/4 toxic reactions [36, 37]. Due to these safety and tolerability issues, the role of MVAC as the standard of care has been largely supplanted by GC and ddMVAC, which offer better safety profiles and tolerability.

GC

GC has shown favorable efficacy and tolerability in patients with metastatic disease and is currently the most commonly used standard regimen in BCa NAC [38]. A meta-analysis that included 754 patients with MIBC treated with the GC regimen between 2007 and 2015 reported a 30.0% pCR rate and a 48.67% rate of pathological downstaging to < ypT2 [39]. Compared to the traditional MVAC regimen, the GC regimen demonstrated equivalent survival benefits with lower toxicity in clinical practice [36, 40, 41]. Modified regimens, such as dose-dense GC (ddGC), have been investigated in prospective phase II trials. A study reported a pT0 rate of 32% after 3 cycles of ddGC, which is comparable to standard GC, with a 27% incidence of grade 3 treatment-related adverse events (TRAEs) [42]. However, the trial was prematurely terminated due to a higher-than-expected incidence of vascular events (23%) following chemotherapy [42]. Another multicenter study evaluated the efficacy of a 6-cycle ddGC regimen in patients with MIBC, with the primary endpoint being the pathologic downstaging rate (pDS) [43]. Among 46 evaluable patients, 57% achieved < ypT2N0, while 37% experienced grade ≥ 3 TRAEs, with anemia (12%) being the most common toxicity [43]. For the rare high-risk upper tract urothelial carcinoma (UTUC), NAC regimens are generally extrapolated from MIBC-related studies. Unlike the 30–40% pCR rate observed in BCa following NAC, the pCR rate for neoadjuvant UTUC may be less than 10% [44]. Recent prospective data have shed light on the efficacy of 4 cycles of split-dose GC in high-grade UTUC. Among the 57 evaluable patients, 36 (63%) demonstrated a pathological response (95% CI, 49%-76%), and 11 patients (19%) achieved a pCR [45].

ddMVAC

In order to overcome the toxicity associated with standard doses of MVAC, the ddMVAC regimen of high-dose-intensity MVAC combined with granulocyte colony-stimulating factor (G-CSF) began to appear in 2001 [46, 47]. A study in 2014 reported the responsiveness and toxicity of ddMVAC in neoadjuvant treatment of MIBC, and the results showed that 15 out of 40 patients achieved pCR, with the majority (82%) experiencing only grade 1 to 2 treatment-related toxicities, and no grade 3 or 4 nephrotoxicity or treatment-related deaths were reported [48]. Another study reported in the same year included 39 MIBC patients with a median follow-up of 2 years [49]. 49% of patients achieved ≤ ypT1N0M0, 10% experienced ≥ grade 3 TRAEs, and the 1-year DFS rate was 89%, demonstrating comparable efficacy and improved tolerability in contrast to standard-dose MVAC [49]. In a prospective phase II trial involving 30 patients with high-grade UTUC, four cycles of ddMVAC achieved a pCR rate of 13.8%, with more than 60% of patients achieving a final pathological stage ≤ ypT1 and a grade 3–4 toxicity rate of 23% [50].

CMV

The 1999 BA06 30894 randomized controlled trial involving 976 MIBC patients assessed the impact of neoadjuvant treatment with CMV [51]. The study demonstrated that four cycles of neoadjuvant CMV provided an absolute 3-year OS benefit of 5.5% [51]. A long-term analysis of the BA06 30894 trial indicated that neoadjuvant CMV reduced the risk of death in MIBC patients by 16%, with a 10-year OS rate of 36%, a 10-year metastasis-free survival (MFS) rate of 33%, and a 10-year DFS rate of 27% [13]. However, due to the lack of sufficient survival benefit and more data support, this regimen is no longer routinely used.

Alternative regimens for cisplatin intolerance

The current prioritization of NAC regimens is primarily determined by cisplatin eligibility. Cisplatin, a heavy metal compound targeting DNA, is taken up by cells through mechanisms such as passive diffusion and copper-specific transport proteins (CTR) [52, 53]. Once inside the cell, the structure of cisplatin (cis-[Pt(NH3)2Cl2]) loses two chloride ions and, through electrostatic attraction, binds to nuclear DNA, forming Pt-DNA adducts that obstruct DNA transcription and replication, ultimately leading to cytotoxic events [54, 55]. Since its FDA approval in 1978, cisplatin has become widely used in the treatment of various malignancies [56–58]. Cisplatin is cleared through glomerular filtration and tubular secretion and induces nephrotoxicity in a dose-dependent manner through mechanisms involving DNA damage, oxidative stress, and apoptosis [59, 60]. Galsky et al. [61] first defined cisplatin eligibility for metastatic UC based on criteria such as a WHO or ECOG performance status of less than 2, a Karnofsky Performance Status (KPS) greater than 60–70%, a creatinine clearance of ≥ 60 mL/min, grade < 2 peripheral neuropathy (CTCAE v5.0), grade < 2 hearing loss (CTCAE v5.0), and grade < 3 heart failure (NYHA). These criteria have become the standard thresholds in clinical trials over the past decade [61–63].

However, due to concerns that some patients may be inappropriately excluded and thus undertreated, research is ongoing to explore updated criteria for cisplatin eligibility, particularly for patients with borderline renal function (creatinine clearance of 40–60 mL/min). This includes the evaluation of eligibility through estimated glomerular filtration rate (eGFR) equations (preferably CKD-EPI) and the application of appropriate mitigation strategies (such as diuresis, hydration, and dose fractionation) to expand cisplatin indications [16, 62, 64, 65]. Cisplatin eligibility criteria for metastatic UC are often extrapolated to MIBC, yet the specific criteria for cisplatin eligibility in MIBC have not been formally agreed upon in clinical guidelines [2, 24, 25, 66].

Patients who are ineligible for cisplatin but meet the criteria for platinum-based chemotherapy (ECOG PS > 3; CrCl < 30 mL/min) as first defined by Gupta et al. [67, 68] in 2019, can receive carboplatin-based chemotherapy. Based on the results from the EORTC 30986 study, carboplatin plus gemcitabine (GCa) is considered the standard alternative for patients ineligible for cisplatin in the treatment of metastatic or locally advanced unresectable UC [69, 70]. This regimen has even become the most commonly used first-line treatment in real-world settings. Some retrospective evaluations have explored the potential value of GCa in the neoadjuvant setting. One study found that patients receiving neoadjuvant GCa had a significantly better 5-year OS compared to those who only underwent RC (79.5% vs. 53.8%, p < 0.001) [71]. Another study indicated that there was no significant difference in complete clinical response (cCR) between neoadjuvant GC and GCa regimens (38.7% vs. 36.2%) [72]. A large multicenter retrospective analysis of 747 patients showed that cisplatin-ineligible individuals receiving at least 3 cycles of neoadjuvant/induction GCa achieved survival benefits comparable to those from neoadjuvant GC [73]. Despite these findings, the application of GCa in the preoperative setting has not been recommended by clinical guidelines due to concerns over carboplatin-related toxicity and a lack of high-quality evidence.

Uncertainties remain in NAC therapy

GC vs. ddMVAC

Currently, in clinical practice, ddMVAC has largely replaced the traditional MVAC regimen and has become one of the most commonly used treatments for MIBC [74]. Recent studies on NAC have focused on comparing the efficacy of ddMVAC with GC (Table 1). ddMVAC has shown higher response rates in metastatic disease, suggesting it may also be more effective in the perioperative setting. A meta-analysis of retrospective studies indicated that ddMVAC significantly outperforms GC in terms of pCR and OS [75].

Table 1.

Pivotal clinical trial of cisplatin-based neoadjuvant chemotherapy for MIBC

Trial Name	Study Design	Treatment Arms	Patient Population	pCR rate (%)	OS rate(%)	≥ 3 TRAEs (%)	Updated Year
SWOG 8710	Phase III randomized	3 cycles of MVAC* vs No NAC	Cisplatin-eligible, cT2–T4aN0, n = 317	38* vs 15	57* vs 43 (5-year)	72*	2003
BA06 30894	Phase III randomized	3 cycles of CMV* vs No NAC	Cisplatin-eligible, cT2G3–T4aN0–Nx, n = 976	NA	36* vs 30 (10-year)	NA	2011
JCOG0209	Phase III randomized	2 cycles of MVAC* vs No NAC	Cisplatin-eligible, cT2–T4aN0, n = 130	34* vs 9	72* vs 62 (5-year)	NA	2014
NCT01031420	Phase II, single-arm	3 cycles of ddMVAC	Cisplatin-eligible, cT2–T4aN0, n = 44	38	83 (1.8-year)	12	2014
NCT01589094	Phase II, single-arm	6 cycles of ddGC	Cisplatin-eligible, cT2–T4aN0, n = 49	15	> 80 (2-year)	37	2018
SWOG S1314	Phase II, randomized	4 cycles of ddMVAC* vs 4 cycles of GC	Cisplatin-eligible, cT2–T4aN0, n = 227	28* vs 30	70* vs 63 (5-year)	38* vs 44	2023
VESPER	Phase III, Randomized	6 cycles of ddMVAC* vs 4 cycles of GC	Cisplatin-eligible, cT2–T4aN0, n = 493	42* vs 36	64* vs 56 (5-year)	52* vs 55	2024

MIBC muscle-invasive bladder cancer; pCR pathological complete response; OS overall survival; ≥ 3 TRAEs the incidence of Grade 3 or higher treatment-related adverse events; MVAC methotrexate, vinblastine, doxorubicin, and cisplatin; CMV cisplatin, methotrexate, vinblastine; NAC neoadjuvant chemotherapy; NA not applicable; ddMAVC dose-dense MVAC; GC gemcitabine and cisplatin; ddGC dose-dense GC; *, the linked data

The Phase III VESPER trial, which included 500 MIBC patients from February 2013 to March 2018, compared 6 cycles of ddMVAC with 4 cycles of GC [76–78]. Secondary endpoint analysis reported in 2021 found that the ddMVAC group had higher rates of weakness and gastrointestinal toxicity, as well as a higher local control rate (pCR, tumor shrinkage, or organ confinement) (p = 0.021) [77]. The primary endpoint analysis released in 2022 showed a significantly higher 3-year PFS for the neoadjuvant ddMVAC group (66% vs. 56%, HR 0.70 [95% CI 0.51; 0.96], p = 0.025) [78]. The latest follow-up data observed that the 5-year OS in the ddMVAC group (HR 0.79 [95% CI 0.52; 0.97]) and cancer-specific survival (CSS) (HR 0.55 [95% CI 0.39; 0.78]) were improved compared to the GC group, supporting the use of 6 cycles of ddMVAC over 4 cycles of GC in the neoadjuvant setting [76].

The Phase II SWOG S1314 trial showed that in intention-to-treat analysis (n = 227), the ypT0 rates for 4 cycles of ddMVAC and 4 cycles of GC were 28% and 30% (p = 0.75), respectively; downstaging rates were 47% and 40% (p = 0.27) [79]. Long-term results revealed no significant differences in 5-year OS (HR 0.87, 95% CI 0.54–1.40; p = 0.57) or event-free survival (EFS) (HR 0.86, 95% CI 0.59–1.26; p = 0.45) between ddMVAC and GC [80].

A 2023 study systematically reviewed and conducted a network meta-analysis (NMA) of Phase III RCTs comparing different NAC regimens, including GC, CMV, and ddMVAC [81]. Bayesian analysis and SUCRA treatment ranking suggested that ddMVAC is highly likely to be the optimal NAC regimen based on mortality rates [81]. Recent meta-analyses have shown that neoadjuvant ddMVAC improves OS (HR 0.71 [95% CI 0.56; 0.90]), PFS (HR 0.76 [95% CI 0.60; 0.97]), and partial pathological response rate (pPR) (OR 1.34 [95% CI 1.01; 1.78]) compared to GC, although the difference in pCR rates was not statistically significant [82]. Furthermore, ddMVAC was associated with a higher incidence of ≥ Grade 3 TRAEs [82].

The advantage of ddMVAC over GC in OS remains debated. For MIBC patients who can tolerate higher toxicity, ddMVAC is likely to be more effective than GC. However, from a healthcare resource utilization and cost perspective, GC has been shown to reduce per-patient costs by approximately 41% compared to ddMVAC when adjusted for ECOG performance status and age (95% CI 28%–52%; p < 0.001) [83]. Thus, the choice of regimen should consider both cost and benefit factors.

NAC treatment cycle

Currently, there is no established consensus on the optimal number of NAC cycles for MIBC. Based on clinical experience, the standard NAC regimen typically consists of 3 to 4 cycles. A retrospective study involving 693 patients with cT2-4N0M0 urothelial carcinoma, who received NAC followed by radical cystectomy, reported ypT0 and ypT1 rates of 26.8% and 54.8% for 4 cycles, and 33.3% and 62.3% for 3 cycles, respectively, with no statistically significant difference between the two durations [84]. OS also showed no difference between 4 and 3 cycles (HR 1.00 [95% CI 0.57 to 1.74], p [0.99]) [84]. A meta-analysis similarly suggested that 3 and 4 cycles of NAC yield comparable pathological responses and prognoses, though 4 cycles had a better cancer-specific survival rate (HR = 1.31) [85]. Another meta-analysis comparing 3 and 4 cycles of neoadjuvant GC included 1091 patients and found that 4 cycles led to higher pCR (OR = 0.66) and pDS (OR = 0.63) compared to 3 cycles [86]. Tumor recurrence rates, OS, and CSS were not significantly different, and the incidence of NAC-related adverse effects was similar between the two regimens [86]. Existing evidence does not clearly indicate the superiority of 3 versus 4 cycles of NAC. Therefore, it is strongly recommended to make a prudent decision on the appropriate NAC regimen in clinical practice, considering patient preferences, physical condition, compliance, and economic burden after thorough communication with the patient.

Histological response characteristics

Histological differentiation or variant forms of BCa have been associated with a more aggressive clinical course in several studies, correlating with an increased risk of disease recurrence, cancer-specific mortality, and overall mortality [87–91]. Due to their rarity, patients with non-pure UC histology are often excluded from clinical trials, resulting in less updated treatment knowledge. Recent studies have begun to address this gap. A meta-analysis from 2024 reported that NAC for variant histology BCa patients might achieve better survival outcomes and higher pDS rates compared to pure UC, although their analysis included various histological variants, introducing inherent biases [92]. In fact, the benefits of different histological variants in NAC require more detailed discussion [4]. A retrospective study classified patients into three histological groups: pure UC (PUC), UC with divergent differentiation (UDD, including squamous or glandular differentiation), and variant histology (VH, including micropapillary, plasmacytoid, nested, sarcomatoid, or neuroendocrine variants) [93]. Patients received either pure RC or NAC + RC. The results indicated that NAC provided overall survival benefits for the PUC group (HR 0.71, 95% CI 0.51–0.98, p = 0.0013), cancer-specific survival benefits for the VH group (HR 0.55, 95% CI 0.30–0.99, p = 0.0495), whereas the UDD group had poorer survival outcomes regardless of NAC treatment [93]. Another meta-analysis found that NAC improved OS and CSS only in the sarcomatoid and neuroendocrine subgroups [90]. Post hoc analysis of the prospective VESPER trial assessed pathological responses and PFS in VH/DD patients and found no evidence of benefit associated with NAC [94]. However, UC with ≥ 50% squamous differentiation (HR adjusted: 2.11, 95% CI 1.01–4.38) and micropapillary subtype (HR adjusted: 2.03, 95% CI 0.98–4.22) were associated with worse PFS [94]. The current confusion regarding NAC benefits for non-pure UC may be attributed to insufficient consideration of histological heterogeneity. Combining limited subgroup analyses with previously established evidence suggests that pure neuroendocrine carcinoma responds better to NAC, while pure squamous carcinoma shows a poorer response. Updated results from PURE-01 evaluating 3 cycles of neoadjuvant pembrolizumab for VH patients showed that 6 of 7 squamous cell carcinoma cases (86%) were downstaged to ypT ≤ 1, with 1 case achieving ypT0, offering potential hope for variant histology types that may otherwise respond poorly to NAC [95].

Lymphovascular invasion (LVI) is characterized by tumor cells invading lymphatic vessels and capillaries, detected in 34.6% of RC specimens, indicating micro-metastasis and leading to lymph node metastasis and poorer prognosis in MIBC patients [96]. Earlier studies indicated that LVI was associated with non-responsiveness to NAC and worse OS following NAC [97, 98]. However, a recent retrospective analysis of Swedish multicenter data found that patients with LVI-positive had a 26.2% cancer-specific survival benefit from NAC, whereas no such benefit was observed in the LVI-negative cohort [99]. These results suggest that subgroup-specific NAC benefits may be confounded by baseline differences within the subgroups. It is essential to design well-considered prospective studies to further validate the value of LVI and its alternative biomarkers.

Emerging options for neoadjuvant therapy

Alternative treatment options such as ICIs, targeted therapies, or ADCs should be considered to overcome the limitations of chemotherapy (Table 2).

Table 2.

Pivotal clinical trial of novel neoadjuvant therapy for MIBC

Trial Name	Study Design	Treatment Arms	Patient Population	pCR rate (%)	OS rate (%)	≥ 3 TRAEs (%)	Updated year
LCCC1520	Phase II, single-arm	GC + pembrolizumab	Cisplatin-eligible, cT2–T4aN0, n = 39	36	91 (1-year)	74	2021
NCT02989584	Phase II, single-arm	GC + atezolizumab	Cisplatin-eligible, cT2–T4aN0, n = 44	41	NA	59	2022
ABACUS	Phase II, single-arm	2 cycles of atezolizumab	Cisplatin-ineligible, cT2–T4aN0, n = 95	31	77 (2-year)	NA	2022
PURE-01	Phase II, single-arm	3 cycles of pembrolizumab	RC-eligible, cT2–T4N0, n = 155	42	84 (3-year)	6	2022
NEOBLADE	Phase II, randomized	GC + nintedanib * vs GC + placebo	Cisplatin-eligible, cT2–T4N0, n = 120	37* vs 32	60* vs 49 (5-year)	93* vs 79	2022
KEYNOTE-869	Phase Ib/II, single-arm	3 cycles of EV	Cisplatin-ineligible, cT2–T4aN0, n = 22	36.4	NA	18.2	2023
RC48-C017	Phase II, single-arm	DV + toripalimab	cT2–T4aN0, HER2+, n = 44	62.1	NA	15.9	2024
NCT05723991	Phase II, single-arm	DV + gemcitabine	Cisplatin-ineligible, cT2–T4aN0, HER2+, n = 17	46.7	NA	0	2024
SunRISe-4	Phase II, randomized	TAR-200 + CET* vs CET alone	Cisplatin-ineligible, cT2–T4aN0, n = 120	42* vs 23	NA	NA	2024

MIBC muscle-invasive bladder cancer; pCR pathological complete response; OS overall survival; ≥ 3 TRAEs the incidence of Grade 3 or higher treatment-related adverse events; GC gemcitabine and cisplatin; NA not applicable; EV Enfortumab vedotin; DV disitamab vedotin; CET cetrelimab; *, the linked data

Immune checkpoint inhibitors

The use of ICIs has transformed the treatment paradigm for metastatic UC. Pembrolizumab, nivolumab, atezolizumab, durvalumab, and avelumab have received FDA approval for the treatment of metastatic UC. Based on the results of two large Phase III randomized controlled trials—Checkmate-274 and IMvigor010—ICIs are becoming an important component in the perioperative treatment of MIBC [73, 100, 101]. ICIs, as monoclonal antibodies, can block checkpoint molecules such as cytotoxic T lymphocyte-associated protein 4 (CTLA-4) and programmed cell death receptor 1 (PD-1)/programmed cell death ligand 1 (PD-L1) from binding to their respective receptors, thereby enhancing T cell activation and preventing immune evasion by tumor cells [102–104] (Fig. 2).

Fig. 2.

The mechanisms of neoadjuvant therapy drugs for mIBC and potential combination strategies. The cisplatin-based neoadjuvant chemotherapy (NAC) impedes DNA transcription and replication by targeting DNA to form Pt–DNA cross-links, leading to cytotoxic events. Radiotherapy (RT) employs high-energy ionizing radiation, which damages DNA directly or generates charged species within cells, inducing impairment and apoptosis in cancer cells. Furthermore, RT is thought to possess immunomodulatory effects and exhibits a synergistic relationship with immune checkpoint inhibitors (ICI). ICIs activate or restore T-cell antitumor activity by blocking the interaction of checkpoint molecules such as cytotoxic T lymphocyte-associated protein 4 (CTLA-4) or programmed cell death receptor 1 (PD-1)/programmed cell death ligand 1 (PD-L1) with their respective receptors. Bladder cancer (BCa) cells are known to highly express surface proteins such as Nectin-4, HER2, and Trop-2. Antibody–drug conjugates (ADC) combine specificity and potency by chemically linking monoclonal antibodies targeting these markers with cytotoxic drugs, enabling selective delivery of therapeutically effective payloads to tumor sites. Target agents (TA) selectively inhibit pathway alterations associated with specific gene mutations in BCa cells, effectively blocking tumor progression. The aforementioned combinations of neoadjuvant therapeutic options are currently being evaluated in clinical trials. MVAC Methotrexate, Vinblastine, Doxorubicin, and Cisplatin; GC Gemcitabine and Cisplatin; ddMVAC Dose-Dense MVAC; CMV Cisplatin, Methotrexate, Vinblastine (Created with BioRender.com)

The Phase II, single-arm PURE-01 trial investigated the activity of 3 cycles of pembrolizumab as neoadjuvant immunotherapy in 50 patients with cT ≤ 3bN0 disease, showing a pCR rate of 42% (95% CI 28.2% to 56.8%) [105]. Notably, patients with high PD-L1 expression (combined positive score (CPS) ≥ 10) had a pCR rate of up to 54.3% [105]. The incidence of Grade 3 or higher TRAEs was 6%, indicating manageable tolerability [105]. The most recent update from the PURE-01 trial, with a median follow-up of 3 years, reported 36-month EFS and OS rates of 74.4% (95% CI 67.8%–81.7%) and 83.8% (95% CI 77.8%–90.2%), respectively [106, 107]. EFS was significantly stratified by PD-L1 tertiles [106, 107].

The single-arm, Phase II ABACUS trial evaluated 2 cycles of atezolizumab in 95 MIBC patients, finding a pCR rate of 31% (27/88). The 2-year DFS and OS rates were 68% and 77%, respectively, with a 2-year DFS of 85% for those who achieved pCR [20, 108]. In the Phase I RJBLC-I2N003 trial, 20 MIBC patients received neoadjuvant toripalimab, resulting in a pCR rate of 40% (8/20), demonstrating comparable pathologic response to standard neoadjuvant chemotherapy [109]. A recent study retrospectively compared neoadjuvant pembrolizumab with standard neoadjuvant chemotherapy, finding that patients receiving pembrolizumab monotherapy had superior response rates and survival outcomes, regardless of cisplatin eligibility [110].

CTLA-4 inhibitors are often combined with PD-1/PD-L1 inhibitors in dual ICI therapy [111]. The single-arm, Phase I NABUCCO trial reported the feasibility of 2 cycles of ipilimumab combined with nivolumab in 24 patients with Stage III UC [111–113]. Although 41% of patients experienced Grade 3–4 immune-related adverse events (irAEs) and 46% achieved complete pathologic response, the results demonstrated dose dependency and lacked mature long-term follow-up [111–113].

Another study systematically reviewed the efficacy and safety of neoadjuvant ICIs, showing an overall pCR rate of 36%. pCR rates were 27% for ICI monotherapy, 41% for ICI dual therapy, and 43% for ICI plus neoadjuvant chemotherapy [114]. The overall rate of Grade 3 or higher TRAEs was 15%, with rates of 7%, 31%, and 17% for ICI monotherapy, ICI dual therapy, and ICI plus neoadjuvant chemotherapy, respectively [114]. Further results from randomized controlled trials are needed to support updates to the standards for neoadjuvant immunotherapy.

Targeted drugs

Fibroblast growth factor receptors (FGFRs) have recently emerged as novel therapeutic targets in cancer treatment due to their oncogenic potential, with approximately 15% of MIBC exhibiting molecular alterations in the FGF axis [115, 116]. Erdafitinib, a tyrosine kinase inhibitor targeting FGFR1-4, has been approved by the FDA for the treatment of advanced UC with susceptible FGFR2 or FGFR3 genetic alterations, particularly in patients whose disease has progressed during or after platinum-based therapy [117, 118]. Given erdafitinib’s demonstrated antitumor activity and manageable safety profile in locally advanced or metastatic UC, two Phase II trials, NERA (NCT05564416) and SOGUG-NEOWIN (NCT06511648), are currently evaluating the pCR of neoadjuvant erdafitinib, either as monotherapy or in combination therapy, in cisplatin-ineligible MIBC patients with susceptible FGFR3/2 genetic alterations [119]. Nintedanib is a small molecule inhibitor targeting the tyrosine kinases PDGFR, FGFR-1 and VEGFR-2, the double-blind, randomized phase II NEOBLADE trial investigated the impact of adding nintedanib or placebo to neoadjuvant GC on response and prognosis in MIBC patients [120]. The pCR rate was 37% in the nintedanib group and 32% in the placebo group. In terms of safety, the incidence of grade ≥ 3 TRAEs reached 93% in the nintedanib group, with thromboembolic events being the most common serious adverse event [120]. The study results did not support an improvement in the primary endpoint (pCR rate) with nintedanib [120]. However, post-hoc analysis suggested that the addition of nintedanib to NAC might improve survival benefits, warranting further investigation in future studies [120]. Between 55 and 75% of MIBC specimens exhibit overexpression of epidermal growth factor receptor (EGFR) [121]. However, downstream pathways of EGFR appear to be bypassed by other receptor tyrosine kinases (RTKs), making EGFR-targeted therapies seem promising but ultimately yielding limited benefit for MIBC patients [122]. A recently presented abstract at the ASCO conference highlighted the significant potential of nimotuzumab combined with concurrent chemoradiotherapy (CRT) as a bladder-sparing or neoadjuvant treatment in 21 MIBC patients [123]. The reported 3-year OS and DFS rates were 100% and 90.4%, respectively, with only 2 RC performed and a 90% organ preservation rate [123]. Notably, no Grade 3 or 4 TRAEs were reported [123].

Antibody–drug conjugates

ADCs combine specificity with potency by chemically linking a monoclonal antibody to a cytotoxic drug, enabling the selective delivery of the therapeutic payload to tumor sites while sparing normal cells, thereby reducing overall toxicity [124, 125]. Currently, ADCs developed for UC target several antigens, including Nectin-4, human epidermal growth factor receptor 2 (HER2), trophoblast cell-surface antigen 2 (Trop-2), SLIT and NTRK-like family member 6 (SLITRK6), tissue factor (TF), and epithelial cell adhesion molecule (EpCAM).

Nectin-4

Nectin-4 is a cell surface protein overexpressed in various malignancies, with 87% of NMIBC and 58% of MIBC patients testing positive for Nectin-4 [126]. In 2019, the FDA granted accelerated approval to EV for the treatment of patients with locally advanced or metastatic UC [127]. This ADC is composed of a monoclonal antibody targeting Nectin-4 linked to the microtubule inhibitor MMAE, positioning it as a leading drug in ADC development [128, 129]. Results from the global, open-label, randomized, multicenter phase III EV-301 trial (NCT03474107) demonstrated a significantly longer OS in the EV group compared to the chemotherapy group (median OS of 12.88 months vs. 8.97 months) [130]. Additionally, PFS was markedly better in the EV group (median PFS of 5.55 months vs. 3.71 months). The incidence of grade 3 or higher TRAEs was comparable between the two groups, at 51.4% and 49.8%, respectively [130]. At the 2023 ASCO Annual Meeting, updated results from the EV-103 trial were presented, in which 22 cisplatin-ineligible MIBC patients received three cycles of EV as neoadjuvant therapy [131]. The primary endpoint was the pCR rate, with key secondary endpoints including pDS, EFS based on investigator assessment, and safety [131]. Preliminary results showed a pCR rate of 36.4%, a pDS rate of 50%, and 18.2% of patients experienced grade 3 or higher TRAEs [131].

HER2

HER2 is a growth-promoting tyrosine kinase receptor involved in cell proliferation and tumorigenesis [132]. A large-scale study involving 37,992 patients indicated that the incidence of HER2 overexpression in BCa (12.4%) is even higher than in breast cancer (10.5%) [133]. Disitamab vedotin (DV, RC48), a novel HER2-targeted ADC, has shown promising efficacy and good tolerability in HER2-positive advanced BCa [134]. Recent reports from a combined analysis of two multicenter, phase II, single-arm studies conducted in China, RC48-C005 and RC48-C009, involved 107 patients with HER2-positive (immunohistochemistry 3+ or 2+) locally advanced or metastatic UC who had failed at least one systemic chemotherapy [135, 136]. The overall objective response rate (ORR) was 50.5%, with median PFS and OS of 5.9 months (95% CI 4.3 to 7.2) and 14.2 months (95% CI 9.7 to 18.8), respectively [135, 136]. The most common TRAEs included peripheral sensory neuropathy, leukopenia, elevated AST, and neutropenia, with a ≥ grade 3 TRAE incidence of 54.2% [135, 136]. The abstract released at ESMO 2024 reported a multicenter phase II trial that enrolled 17 HER2-positive, platinum-intolerant MIBC patients, who underwent 3 cycles of neoadjuvant treatment with DV combined with gemcitabine [137]. The pCR rate reached 46.7%. The most common TRAEs of all grades were rash (11.76%) and fever (17.65%), with no grade ≥ 3 TRAEs observed [137].

Trop-2

Trop-2 is a transmembrane glycoprotein highly expressed in urothelial carcinoma, playing a crucial role in cell proliferation, apoptosis, cell adhesion, epithelial-mesenchymal transition, and both tumorigenesis and progression. It has emerged as a novel and promising therapeutic target [138–140]. Sacituzumab govitecan (SG) is a novel ADC that targets Trop-2 and is conjugated with the topoisomerase I inhibitor SN-38 [141, 142]. The phase II open-label TROPHY-U-01 study investigated the activity and safety of SG in patients with metastatic UC who had progressed following platinum-based chemotherapy and checkpoint inhibitors [143, 144]. The 2024 updated safety and efficacy results revealed an ORR of 28% with a median PFS of 5.4 months at a median follow-up of 10.5 months. The median OS was 10.9 months. Key ≥ grade 3 TRAEs included neutropenia, leukopenia, anemia, diarrhea, and febrile neutropenia, with 6% of patients discontinuing treatment due to TRAEs [144]. The phase II SURE-01 trial aims to evaluate the efficacy of neoadjuvant SG monotherapy in patients with MIBC [145]. The interim results, presented at the 2024 ASCO meeting, included 21 patients with MIBC who were either ineligible for or refused NAC, with 33.3% (7/21) being cT3-4N0. Among the cohort, 47.6% exhibited mixed histology, with a cCR rate of 47.6%, defined by negative MRI, cystoscopy, and circulating tumor DNA (ctDNA) testing. A ypT0N0-x response was achieved in 37.5% (6/16) of cases, indicating promising activity of neoadjuvant SG in MIBC patients who did not meet standard treatment criteria [146].

Neoadjuvant combination therapy

NAC + ICI

The therapeutic efficacy and safety of combining ICIs with NAC have been evaluated in several Phase II trials. The Phase II LCCC1520 trial (NCT02690558) investigated the addition of pembrolizumab to gemcitabine and split-dose cisplatin for MIBC neoadjuvant treatment [147]. This study reported a pathological downstaging rate of 56% (22/39), surpassing the 45%–50% observed with gemcitabine-cisplatin alone in historical controls [147]. The pCR rate was 36%, with manageable toxicity [147]. The Phase II BLASST-1 trial reported that 41 MIBC patients treated with nivolumab plus GC had a pCR rate of 49%, and 66% had pathological downstaging to ≤ ypT1N0 [148]. Grade 3 or higher TRAEs occurred in 20% of patients, primarily neutropenia, thrombocytopenia, and renal dysfunction [148]. In another Phase II trial, 76 MIBC patients received nivolumab plus GC, with 33 achieving cCR [149]. Of these, 32 chose to forego immediate cystectomy [149]. Grade 3 or higher TRAEs were observed in 75% of patients, mostly anemia, neutropenia, and urinary tract infections [149]. A Phase II trial involving 44 cisplatin-eligible MIBC patients assessed neoadjuvant treatment with atezolizumab plus GC [150]. This trial observed a pCR rate of 41% (16/39) and 69% (27/39) with downstaging to < ypT2N0 [150]. Twenty-six patients (59%) experienced Grade 3 or higher TRAEs [150]. In a multicenter, single-arm Phase II study, MIBC patients received neoadjuvant treatment with camrelizumab plus GC [151]. Among 30 evaluable patients, 13 (43.3%) achieved pCR, and 16 (53.3%) had pathological downstaging [151]. Grade 3 or higher TRAEs were reported in 34.9% of patients, with no events leading to death [151].

A multicenter real-world retrospective study recruited 253 patients undergoing neoadjuvant therapy, with 98 receiving tislelizumab-GC, 107 receiving GC, and 48 receiving tislelizumab monotherapy [152]. The pCR rate was significantly higher in the tislelizumab-GC cohort (31.6%) compared to the GC (11.2%) and tislelizumab (14.6%) cohorts (p = 0.001) [152]. Response rates were also higher in the tislelizumab-GC group (60.2%) compared to the GC (43.0%) and tislelizumab (35.4%) groups [152]. However, the combined treatment group experienced a higher frequency of Grade 3 or higher adverse events, including anemia, neutropenia, thrombocytopenia, nausea, and hypothyroidism [152]. A meta-analysis of 22 studies involving 843 patients found an overall pCR rate of 42.6% for NAC-ICI, higher than for neoadjuvant ICI monotherapy (24%) and neoadjuvant dual ICI therapy (32.1%) [153]. Subgroup analysis showed Grade 3 or higher irAEs rates of 7.4% (95% CI 4.3%–10.5%) for ICI monotherapy, 30.3% (95% CI 15.3%–45.3%) for dual ICI therapy, and 14.5% (95% CI 3.5%–25.4%) for NAC-ICI [153].

Due to limitations in follow-up duration and cohort size, long-term survival benefits of NAC-ICI remain unclear. The available data, though limited in quality, suggest that compared to chemotherapy alone, NAC-ICI may increase pathological downstaging, but it has not shown significant improvement in the intermediate endpoint of pCR (30–44%). Furthermore, the pathological response advantage of NAC-ICI is not notable when compared to trials of neoadjuvant ICI monotherapy, and the associated increase in toxicity requires further attention and management. Unless higher-level evidence can demonstrate comparable or superior pCR and survival benefits, or the identification of more suitable patient populations, this treatment combination may face challenges in maintaining its role in the neoadjuvant treatment landscape for MIBC.

Promisingly, the interim results of the randomized phase III NIAGARA trial, recently presented at ESMO 2024, aimed to evaluate the efficacy of perioperative durvalumab combined with NAC compared to NAC alone[154]. The formal analysis of pCR showed no statistical significance; however, the durvalumab treatment group demonstrated significantly longer EFS and OS, with a safety profile comparable to that of monotherapy, showing clinical significance worthy of further exploration [154]. Another randomized phase II trial, SunRISe-4, revealed its interim analysis results at ESMO 2024 [155]. As of May 31, 2024, the study had enrolled 120 cisplatin-eligible MIBC patients, who were randomly assigned in a 5:3 ratio to receive either the TAR-200 intravesical drug delivery system (designed for localized sustained release of gemcitabine) combined with cetrelimab (C1 group) or cetrelimab monotherapy (C2 group) [155]. Among evaluable patients, the pCR rate was significantly higher in the C1 group compared to the C2 group (42% vs. 23%) [155]. In terms of safety, the incidence of TRAEs was 72% in the C1 group and 44% in the C2 group [155].

ICI + ADC

The exploration of ADCs in combination with ICIs represents the latest advancements in neoadjuvant therapy. Ongoing clinical trials, such as KEYNOTE-905/EV-303, KEYNOTE-B15/EV-304, and VOLGA (NCT04960709), are evaluating the clinical efficacy of combining EV with ICIs for MIBC in the neoadjuvant setting [156–158]. Combining DV with ICIs has also shown promising results. A real-world study reviewed data on neoadjuvant treatment of MIBC patients with either DV monotherapy or in combination with ICIs. With a median follow-up of 19.0 months, the pathological objective response rate among 11 patients was 81.8%, and a bladder preservation rate of 72.7% was maintained at 16 months [159]. Adverse events were manageable [159].

At the 2024 ASCO meeting, a retrospective single-arm study reported that among 47 MIBC patients treated with DV plus toripalimab, 29 achieved cCR (61.7%), and 37 achieved pDS (78.7%) [160]. 25% of patients experienced Grade 3 or higher TRAEs, primarily anemia, ALT elevation, and fatigue [160]. Bladder-preserving surgery was chosen by 30 patients, with a 1-year bladder intact disease-free survival (BI-DFS) rate of 100% [160]. The Phase Ib/II Hope-03 trial enrolled 51 HER2-positive MIBC patients treated with DV plus tislelizumab. Preliminary results from 18 patients showed a 68.75% cCR rate (T0/Ta/Tis) and a 90.91% disease control rate (DCR), with no Grade 3 or higher TRAEs reported, indicating encouraging efficacy and safety [161]. The Phase II RC48-C017 trial (NCT05297552) reported preliminary results for DV plus toripalimab in HER2-expressing MIBC [162]. The pCR rate was 62.1% (18/29), and the pathological response rate was 75.9% (22/29) [162]. 15.9% of patients reported Grade 3–4 TRAEs, with no Grade 3–5 DV-related peripheral neuropathy [162]. Survival data are still immature [162]. Another single-arm study explored DV combined with cadonilimab (AK104, a PD-1/CTLA-4 bispecific antibody) in HER2-positive MIBC perioperative therapy, achieving a 100% DCR among 5 patients who completed neoadjuvant treatment, with good tolerability [163]. The Phase II SURE-02 trial aims to investigate the efficacy of SG combined with pembrolizumab in neoadjuvant therapy for MIBC, with primary endpoints including pCR, and secondary endpoints including EFS, cCR, and OS [146]. Mid-term results have yet to be published [146].

Initial data highlight the excellent response rates and tolerability of ADC-ICI combinations in neoadjuvant therapy for specific MIBC populations. However, most studies have yet to release mature results. Further detailed pathological assessments, long-term survival data, and additional meta-analyses are anticipated to support the combination’s potential as a novel first-line neoadjuvant regimen in the near future.

Neoadjuvant therapy response prediction biomarkers

The FDA has approved the use of PD-L1 expression on tumor or immune cells, high microsatellite instability (MSI-H), and tumor mutational burden (TMB) to predict responses to ICIs across various cancers [164–166]. Additionally, studies evaluating the expression of commonly targeted therapeutic markers (e.g., FGFR, HER2) as predictors of treatment response are not discussed here.

DNA damage repair genes

Cisplatin exerts its therapeutic effects by inducing DNA damage in tumors. Therefore, alterations in DNA damage repair (DDR) genes, such as ATM, RB1, FANCC, and ERCC1/2, may increase tumor vulnerability to cisplatin, enhancing its sensitivity [167–169]. A small prospective study indicated that tissue BRCA1, ERCC1, and CTR gene mRNA expression are independent predictors of NAC response [170]. Recent retrospective analyses have confirmed that bladder cancer patients with ERCC2 gene mutations generally respond to chemotherapy [171]. In treatment-naïve samples from two independent clinical cohorts receiving cisplatin-based NAC, alterations in one or more DDR genes (ATM, RB1, FANCC) predicted pathological response to the ddMVAC regimen (p < 0.001; sensitivity 87%, specificity 100%) and improved OS (p = 0.007) [172]. However, in a validation cohort using the ddGC regimen, changes in these DDR genes did not show a significant trend for improvement in PFS and OS [172]. Similarly, the same research team demonstrated in a larger validation set, SWOG S1314, that mutations in ATM, RB1, FANCC, or ERCC2 prior to NAC predicted pCR (ypT0) at surgery, but there was no evidence of interaction between treatment regimens (ddMVAC vs. GC) and gene mutations [173]. Additionally, alterations in any DDR gene were independently associated with higher response rates to PD-1/PD-L1 blockade (67.9% vs 18.8%; P < 0.001) [174]. Current evidence suggests that DDR-related neoadjuvant efficacy markers show promising potential.

Liquid biopsy markers

Liquid biopsy facilitates rapid, non-invasive, and repeatable acquisition of genetic profile characteristics. In an exploratory analysis of the SWOG S1314 trial, researchers profiled circulating free DNA (cfDNA) methylation to generate biomarkers associated with NAC response, successfully predicting 79% of patients’ pathological outcomes [175]. Similarly, a Danish study reported the potential of combining blood and urine DNA analysis to predict NAC response [176]. By collecting body fluids before, during, and after treatment—with a median follow-up of 41.3 months post-RC—it was observed that tumor DNA kinetics in these fluids correlated with response rates, allowing for patient stratification by outcomes (p = 0.003) [176]. Additionally, the multicenter CirGuidance study used the CELLSEARCH system to assess the presence of circulating tumor cells (CTCs) from blood. CTC-positive patients (≥ 1 detectable CTC) were recommended for NAC followed by RC, whereas CTC-negative patients underwent RC without NAC [177]. The primary endpoint was 2-year OS in the CTC-negative group, and an unapproved exploratory analysis suggested that CTC-positive patients receiving NAC had a 22.8% improvement in 2-year OS compared to those not receiving NAC [177].

The phase I trial RJBLC-I2N003 evaluated the utility of urine tumor DNA (utDNA) based minimal residual disease (MRD) detection in neoadjuvant ICIs [109]. Results indicated that longitudinal utDNA analysis outperformed radiological assessment and traditional biomarkers in predicting pathological results of ICIs [109]. The final results of ABACUS showed that continuous ctDNA levels had high prognostic predictive value for neoadjuvant atezolizumab treatment [20]. Phase 1 NABUCCO data supported that ctDNA depletion in plasma could predict PFS with dual neoadjuvant ICIs [112]. Exploratory analyses from the Phase III VOLGA trial indicated that for cisplatin-ineligible patients with MIBC who received three cycles of neoadjuvant durvalumab + tremelimumab + EV, plasma ctDNA clearance during treatment was associated with improved clinical outcomes [178]. The key advantage of liquid biopsy lies in its ability to dynamically monitor the response to neoadjuvant therapy, enabling more personalized care.

Molecular subtypes

Molecular subtypes reflect different stages of urothelial differentiation based on characteristic gene expression [179, 180]. However, the diversity of subtype sets hinders their clinical application. Recent consensus systems have identified six molecular categories for MIBC: luminal papillary (24%), luminal nonspecified (8%), luminal unstable (15%), stroma-rich (15%), basal/squamous (35%), and neuroendocrine-like (3%), paving the way for identifying specific subtypes enriched for chemotherapy or immune response [181]. A retrospective cohort from Sweden confirmed that different molecular subtypes receive varying clinical benefits from NAC [182]. Under consensus classification, high expression of SPP1 in the Ba/Sq subtype showed lower pCR rates and poorer survival outcomes, whereas the stroma-rich subtype exhibited improved CSS [182]. A comprehensive multi-omics analysis of 300 MIBC patients receiving chemotherapy (neoadjuvant or first-line) revealed that high genomic instability (e.g., allele imbalance, large insertions/deletions, extensive SBS5 mutations, BRCA2 somatic mutations) and non-Ba/Sq subtypes are most likely to respond to chemotherapy [168]. Another multicohort study of 601 MIBC patients using a commercial molecular subtype classifier assessed the impact of subtype differences on NAC benefits, showing that non-luminal subtypes benefit most from NAC, with a 10% increase in 3-year OS (71% vs. 61%), while luminal subtypes showed minimal benefit (63% vs. 65%) [183]. An analysis of expression data from the PURE-01 cohort evaluated the relationship between immune features and molecular subtypes (cancer genome atlas, consensus model, and gene expression subtype classifier [GSC]) with CR and PFS [184]. Results showed the neuroendocrine-like subtype had the worst 2-year OS (33%), whereas the GSC claudin-low subtype had the best 2-year OS [184]. Recent study presented at ESMO 2024 analyzed TURBT samples from MIBC patients through proteomics, identifying three molecular subtypes with distinct features. Among these, the basal-like subgroup, characterized by high expression of markers KRT5, KRT6A, KRT14, and CD44, demonstrated increased DNA replication and cell cycle activity, exhibiting a higher pCR rate after NAC (56%) compared to the other two groups [185]. In contrast, the LiL subgroup, which exhibited higher immune and cytoskeletal activity, had a pCR rate of 36%, suggesting that increased activity of the cytoskeleton, extracellular matrix, and collagen proteins may be associated with NAC resistance [185].

Other biomarkers

Combining predictive biomarkers with recognized markers may yield complementary performance. Cyclin-dependent kinase 6 (CDK6) is an important regulator of the cell cycle [186, 187]. A retrospective analysis of 933 MIBC patients showed that high CDK6 expression led to poor prognosis, good response to platinum-based chemotherapy, but poor response to ICIs [188]. Ultimately, a response scoring model based on CDK6, PD-L1, and TMB effectively predicted responses to ICIs and chemotherapy [188].

As sequencing resolution improves, the degree of infiltration of cellular subpopulations has been found to indicate efficacy. A study using single-nucleus RNA sequencing (snSeq) on 25 newly diagnosed MIBC patients reported that CDH12-enriched tumor subpopulations were resistant to NAC but exhibited superior responses to ICIs [189]. CDH12 enrichment outperformed BCa subtype prediction in patient stratification [189]. Longitudinal analysis from the NCT02108652 cohort showed that the expansion of new antigen-reactive CD8+ T cells (NART) between pre-treatment and 3 weeks post-treatment could distinguish between disease control and progression, with peripheral NART from disease control patients exhibiting a PD1+Ki67+ effector phenotype and elevated CD39 levels[190].

Adding ICIs to NAC to enhance patient responsiveness has shown promise. Understanding tumor microenvironment (TME) characteristics may aid in predicting ICI-NAC response. A 2024 study analyzed protein expression in 18 responders and 18 non-responders from two trials adding ICIs to neoadjuvant GC [LCCC1520 (pembrolizumab) and BLASST-1 (nivolumab)] using digital spatial proteomics (DSP) [191]. Using spatially anchored TME features, the study achieved good predictive performance for NAC-ICI response (AUC = 0.827) [191]. Another study performed a post hoc immune feature analysis of the LCCC1520 and ABACUS trials [192]. Results indicated that plasma IL-9 might be a potential predictive biomarker for ICI-NAC response, while tumor IL-8 characteristics and the stroma-rich subtype were potential predictive biomarkers for benefits from neoadjuvant ICIs [192].

These studies support that developing new efficacy prediction biomarkers based on gene expression profiles, molecular subtypes, cellular subpopulation infiltration, and spatial TME characteristics may offer additional guidance for optimizing neoadjuvant treatment strategies for heterogeneous patient populations. Well-designed clinical trials are needed to evaluate these emerging biomarkers to validate their utility.

Organoids

Commonly used 2D cell culture methods overlook the interactions between tumors and their TME or matrix components, leading to unreliable drug testing results. In contrast, in vitro organoids, which are directly derived from patient pathological tissues and replicate the characteristics of the parent tumor, are emerging as a powerful tool for studying tumor heterogeneity and guiding treatment decisions. Comparative studies of primary cell lines, patient-derived xenografts (PDX), and patient-derived organoid (PDO) systems have revealed that both PDX and PDO models exhibit a high degree of concordance with actual patient treatment responses in predicting drug sensitivity [193]. A study proposed a machine learning framework capable of identifying robust drug biomarkers using pharmacogenomic data derived from organoid models, which accurately predicted the drug responses of 77 BCa patients treated with cisplatin [194]. Another study developed a BCa PDO-microfluidic chip drug sensitivity prediction system, facilitating high-throughput prediction of individualized patient treatment responses [195]. PDOs faithfully recapitulate the histopathological and clinical characteristics of the parent tumors and demonstrate advantages over PDX models in terms of cost, efficiency, and ethical considerations. The integration of organoid platforms with multi-omics feature analysis, artificial intelligence algorithms, and microfluidic chip technology is expected to advance personalized neoadjuvant therapy, enhancing the accuracy and efficiency of drug screening and efficacy prediction. However, it is important to highlight that PDOs still face challenges related to low establishment, passage, and maintenance success rates. There is an urgent need to develop standardized culture conditions and create large-scale, renewable PDO biobanks to overcome these limitations and accelerate their clinical translation in predicting neoadjuvant treatment efficacy.

Emerging clinical trials and what remains unknown

The rapid development of neoadjuvant therapies for MIBC presents significant advancements, yet several unresolved issues remain. The increasing diversity of treatment choices, while promising, necessitates the integration of current evidence and a reassessment of treatment sequencing. This includes updating optimal first-line combinations to maximize patient survival benefits (Fig. 3).

Fig. 3.

Evolving paradigms of neoadjuvant therapy for MIBC. In current practice, platinum-based neoadjuvant chemotherapy (NAC) is the standard treatment for patients with muscle-invasive bladder cancer (MIBC). Neoadjuvant immunotherapy remains limited to the clinical trial setting. Patients who receive neoadjuvant treatment and achieve clinical/pathological remission have a better prognosis and a higher chance of bladder preservation. The criteria for neoadjuvant treatment eligibility are being optimized, and with the advent of the era of immunotherapy, combination regimens involving immune checkpoint inhibitors (ICI) are expected to replace single-agent neoadjuvant chemotherapy as the new standard of care. Precision medicine will further guide individualized clinical decision-making, utilizing information from multi-omics phenotypes and organoid drug sensitivity predictions to help patients achieve the greatest benefit. ADC, antibody–drug conjugates; RC, radical cystectomy (Created with BioRender.com)

Currently, although traditional first-line treatments are based on cisplatin-containing chemotherapy regimens, high-quality data on new treatment options in clinical applications is still emerging (Table 3). The phase III, randomized, double-blind study KEYNOTE-866 (NCT03924856) is evaluating the efficacy of perioperative Pembrolizumab (MK-3475) combined with NAC versus perioperative placebo combined with NAC in cisplatin-eligible MIBC patients. The primary endpoint is the 60-month EFS, with completion anticipated by 2025. Additionally, the phase III, randomized, open-label, multicenter, global study NIAGARA (NCT03732677) aims to determine the efficacy and safety of durvalumab combined with GC-based neoadjuvant therapy in MIBC, with primary endpoints including the 6-month pCR rate and 48-month EFS, expected to conclude by late 2026. The phase I trial NCT04610671 is assessing adverse events during the neoadjuvant therapy of intravesical GC combined with intravenous nivolumab for cisplatin-ineligible MIBC patients.

Table 3.

Ongoing clinical trials investigating neoadjuvant therapies for MIBC

Trial name (NCT)		Study design		Treatment arms		Patient population		Primary endpoint	Status		Estimated completion date		Remarks
NCT05723991		Phase IV, single-arm		GC + DV		Cisplatin-eligible, cT2–T4aN0, HER2+ n = 36		pCR	Recruiting		2025		NAC-ADC
NCT05822934		Phase III, randomized		GCa vs GC		Cisplatin-eligible, cT2–T4aN0-N3, n = 20		Clinical response rate	Active, not recruiting		2024		Carboplatin
KEYNOTE-866 (NCT03924856)		Phase III, randomized		GC + pembrolizumab vs GC + placebo		Cisplatin-eligible, cT2–T4aN0, n = 907		EFS	Active, not recruiting		2025		NAC-ICI
KEYNOTE-B15 (NCT04700124)		Phase III, Randomized		EV + pembrolizumab vs GC		Cisplatin-eligible, cT2–T4aN0, n = 784		EFS	Active, not recruiting		2026		ICI-ADC
NIAGARA (NCT03732677)		Phase III, Randomized		GC + durvalumab vs Durvalumab		Cisplatin-eligible, cT2–T4aN0-N1, n = 1063		pCR, EFS	Active, not recruiting		2026		NAC-ICI
NCT05776758		Phase III, Randomized		NAC + RC vs RC alone		Cisplatin-eligible, cT2–T4N0, n = 236		OS	Recruiting		2027		NAC survival benefit
NEO-BLAST (NCT06537154)		Phase II/III, Randomized		Active surveillance vs definitive bladder treatment		cT2–T4aN0, Received NAC, n = 688		Proportion of patients found to have a cCR and randomized on the number of patients enrolled, MFS	Not yet recruiting		2028		Active surveillance: utDNA, ctDNA, MRI and TURBT
NeoSTOP-IT (NCT06571708)		Phase II, Randomized		GC + cemiplimab + fianlimab vs GC + cemiplimab		Cisplatin-eligible, cT2–T3N0, n = 36		cCR	Not yet recruiting		2028		NAC-ICI-ICI
NCT06059547		Phase II, cohort study		Nivolumab + visugromab vs Nivolumab + placebo		Cisplatin-ineligible, cT2–T4aN0, n = 30		pCR	Recruiting		2026		ICI-GDF-15 monoclonal antibody
SOGUG-NEOWIN (NCT06511648)		Phase II, cohort study		Erdafitinib vs Erdafitinib + cetrelimab		Cisplatin-ineligible, cT2–T4aN0-N1, FGFR alteration, n = 90		pCR, pDS	Recruiting		2029		ICI-FGFR inhibitor
NCT05016973		Phase II, single-arm		DV + triplizumab		cT2–T3bN0, HER2+, n = 25		PaR	Not yet recruiting		2024		ICI-ADC
NCT05328336		Phase II, Single-arm		Nab-paclitaxel + tislelizumab		cT2–T4Nx, n = 74		CR	Recruiting		2025		NAC-ICI
RAD-VACCINE (NCT05241340)		Phase II, single-arm		Sasanlimab + stereotactic body radiation therapy		Cisplatin-ineligible, cT2–T4aN0, n = 33		Feasibility and safety, pCR	Recruiting		2025		ICI-RT
RJBLC-NEBC001 (NCT06091124)		Phase II, single-arm		Adebrelimab + etoposide + cisplatin		Cisplatin-eligible, Neuroendocrine carcinoma n = 26		pCR, safety and tolerability	Recruiting		2026		Neoadjuvant therapy for VH
NCT06364956		Phase Ib/II, single-arm		Tislelizumab + palbociclib		Cisplatin-ineligible, cT2–T4aN0, n = 36		The safety dose of the combination, pCR	Recruiting		2026		ICI-CDK4/6 inhibitor
NCT06341400		Phase Ib/II, Single-arm		DV + toripalimab		Cisplatin-ineligible, cT2–T4aN0, HER2+, n = 55		pCR	Recruiting		2027		ICI-ADC
STAR-EV (NCT06394570)		Phase I/II, single-arm		EV + stereotactic body radiation therapy		Cisplatin-ineligible, cT2–T4aN0, n = 19		MTD, pCR	Not yet recruiting		2027		ADC-RT
TASUC-Neo (NCT05839119)		Phase I, single-arm		GC + degarelix		Cisplatin-eligible, cT2–T4N0, AR+, n = 32		pCR	Recruiting		2030		NAC-GnRH receptor antagonist
NCT06325423		Observational		NAC		cT2–T4N0-N1, Received NAC, n = 50		Response to NAC in relation to: NETs expression, CD8 expression, NET/CD8 ratio and baseline clinicopathological features	Not yet recruiting		2027		Predicting response to NAC
NCT06373055	Observational		NAC		cT2–T4aN0, Received NAC, n = 250		Identification of responders to NAC by spatial transcriptome profiling			Recruiting		2033	Identification of biomarkers predicting treatment response

DV disitamab vedotin; GC gemcitabine and cisplatin; pCR pathological complete response; NAC neoadjuvant chemotherapy; ADC antibody–drug conjugate; GCa carboplatin plus gemcitabine; EFS event-free survival; ICI immune checkpoint inhibitor; EV Enfortumab vedotin; RC radical cystectomy; OS overall survival; cCR complete clinical response; utDNA urine tumor DNA; ctDNA circulating tumor DNA; MRI magnetic resonance imaging; TURBT transurethral resection of bladder tumor; pDS pathological downstaging; PaR pathological response rate per central pathology review; CR clinical response; RT radiation therapy; VH variant histology; MTD maximum tolerated dose; NETs neutrophil extracellular traps

The phase III, randomized, open-label study EV-304 (NCT04700124), which is comparing the efficacy of EV combined with pembrolizumab versus GC alone as neoadjuvant therapy for cisplatin-eligible MIBC patients, with completion expected by late 2026. The single-arm, prospective, exploratory study NCT06341400, which is investigating the pCR of RC48 combined with toripalimab for cisplatin-ineligible MIBC patients.

Other trials include the phase II, open-label, multicenter, multinational interventional trial SOGUG-NEOWIN (NCT06511648) is assessing the efficacy and safety of Erdafitinib alone or in combination with cetrelimab for FGFR-mutant MIBC patients who are either ineligible for or have refused cisplatin-based NAC, with expected completion in 2029. APL-1202, a selective human methionine aminopeptidase II (MetAP2) inhibitor, has shown anti-tumor angiogenesis and anti-tumor activity in preclinical studies. The open-label, multicenter study EXPECTATE (NCT04813107) aims to compare the efficacy of oral APL-1202 combined with tislelizumab versus tislelizumab alone as neoadjuvant therapy for MIBC. The RADIANT trial (NCT04543110) is evaluating the efficacy and safety of sequential radiotherapy combined with durvalumab as neoadjuvant therapy for MIBC patients.

Additionally, the prospective observational study NCT05767528 aims to evaluate the accuracy of 3D in vitro PDO drug sensitivity testing in predicting the efficacy of neoadjuvant therapy for MIBC. NCT06325423 is investigating the correlation between neutrophil extracellular traps (NETs) expression, CD8 expression, and baseline pathological features with NAC response to identify patient populations that may benefit.

Last but not least, while specific histological subtypes exhibit lower NAC response rates or poorer prognosis, the treatment needs of these patients should not be overlooked. Currently, there is limited understanding of the specific benefits of neoadjuvant therapies for different histological subtypes. Furthermore, while numerous biomarkers have shown promising predictive performance, their clinical translation remains challenging. The cost-effectiveness of these emerging biomarkers must be considered, with the goal of integrating reliable and practical patient stratification tools into standard care as soon as possible.

Conclusion

The neoadjuvant treatment landscape for MIBC is undergoing rapid transformation, signaling a promising future. While cisplatin-based neoadjuvant chemotherapy, particularly ddMVAC, remains the standard for patients who can tolerate its toxicity, emerging therapies such as ICIs, FGFR inhibitors, and ADCs are redefining treatment paradigms. These novel interventions offer viable alternatives for cisplatin-ineligible patients and hold potential as new first-line options through combination therapies, as ongoing clinical trials continue to explore their efficacy.

A thorough evaluation of safety profiles and toxicity management remains critical. Advances in understanding the histological and genetic heterogeneity of MIBC will propel the shift toward personalized medicine. The development of innovative liquid biopsy techniques, including cfDNA and CTCs, is anticipated to refine the prediction and monitoring of treatment responses. Looking ahead, the integration of these tailored approaches promises to enhance treatment precision, optimizing survival outcomes for each patient based on their unique characteristics.

Box 1 The role of neoadjuvant therapy in bladder preservation:

• Bladder preservation therapies aim to achieve long-term survival while maintaining optimal quality of life, serving as an alternative or complement to RC.

• Common bladder preservation strategies include:

  • Radical dose radiotherapy,
  • Concurrent chemoradiotherapy,
  • Trimodal therapy (TMT) combining maximal TURBT with chemoradiotherapy,
  • Partial cystectomy (PC).

• Multicenter propensity score matching and weighted analyses have shown no statistically significant difference between TMT and RC in 5-year cancer-specific survival and metastasis-free survival, supporting their comparable long-term efficacy [196].

• In clinical practice, bladder preservation requires close multidisciplinary collaboration (MDT) to identify optimal candidates and tailor individualized treatment plans.

• Patients with localized MIBC (cT2-4aN0M0) who are eligible for neoadjuvant therapy are potential candidates for bladder preservation. However, not all patients are suited for immediate TMT. Even with TMT, about 20% may require salvage RC due to treatment failure or recurrence [197].

• Neoadjuvant therapy-driven bladder preservation offers a more dynamic clinical decision-making approach, allowing sequential treatments based on the patient’s response and preferences, thus improving bladder retention outcomes.

• The role of neoadjuvant therapy-driven bladder preservation is gaining prominence. Retrospective cohort data presented at ESMO 2024 demonstrated that bladder-sparing treatments following neoadjuvant immunotherapy can achieve 2-year DFS comparable to TMT. However, stronger evidence is needed to confirm its long-term efficacy [198].

Author contributions

S.Z. and C.X. conceived and designed the study. W.H., J.X., and Z.W. consulted and summarized the literature. W.H. and J.X. drafted and revised the manuscript. Figures were prepared by W.H., M.W., Q.C., C.Z., Y.W. (Yue Wang), X.Y. (Xufeng Yu), Y.W. (Yi Wang), H.Z., X.Y. (Xuncheng Yang),and Z.Z. All authors reviewed the manuscript.

Funding

This research was financed by grants from National Natural Science Foundation of China (Nos. 81802515, 81801854, 82172871, 81972391, 82272950), Discipline Development Plan of Changhai Hospital (No. 2019YXK041), Science and Technology Commission of Shanghai Municipality (Nos. 20Y11904800, 22140903700), Shanghai Municipal Health Commission (No. 2022YQ010).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This project is not applicable to our research.

Consent for publication

This project is not applicable to our research.

Competing interests

The authors declare no competing interests.