Perioperative systemic therapy in muscle invasive bladder cancer: Current standard method, biomarkers and emerging strategies

Abstract

Bladder cancer ranks as the 10th most common cancer type globally, and muscle-invasive disease accounts for approximately 25% of newly diagnosed bladder cancers. Despite definitive treatment, 50% of patients with muscle-invasive bladder cancer (MIBC) develop metastasis within 2 years, leading to death. Perioperative systemic therapy is generally recommended to control local relapse or distant metastasis after surgical resection for patients with MIBC. Cisplatin-based neoadjuvant chemotherapy followed by radical cystectomy is the current standard treatment to improve oncologic control and survival outcomes. Adjuvant chemotherapy is recommended for patients with pathological T3-4 or positive lymph nodes after radical cystectomy if no neoadjuvant chemotherapy was given. Nonetheless, perioperative systemic therapy is not applied widely because of its toxicity, and less than 25% of patients receive cisplatin-based neoadjuvant chemotherapy. Therefore, the development of predictive biomarkers for neoadjuvant chemotherapy efficacy and alternative effective regimens for cisplatin-ineligible patients are important. Furthermore, recently, novel anticancer agents such as immune checkpoint inhibitors and antibody-drug conjugates have proven survival benefits in the metastatic setting, thereby expanding their therapeutic applications to the perioperative setting for non-metastatic MIBC. Herein, we discuss the current status and future perspectives of perioperative systemic strategies for MIBC.

INTRODUCTION

Bladder cancer (BC) is the tenth most commonly diagnosed malignancy worldwide, with approximately 573,000 new cases and 213,000 cancer-related deaths reported in 2020 globally [1]. BC can be divided into non-muscle-invasive BC (NMIBC) and muscle-invasive BC (MIBC). At initial diagnosis, approximately 70%–75% and 25%–30% of cases are diagnosed as NMIBC and MIBC, respectively [2,3]. If left untreated, only 15% of MIBC patients would survive beyond 2 years [4]. A recent report has indicated even worse results, with median overall survival (OS) of 9 months and 5-year OS rate of 5% for untreated MIBC [5].

Patients with MIBC are managed using a multimodal approach consisting of neoadjuvant chemotherapy (NAC), followed by radical cystectomy (RC) with pelvic lymph node dissection (PLND) [6]. Post-RC recurrence is a major cause of mortality in MIBC patients, and most recurrences occur during the first 2–3 years post-RC [7]. The distant metastasis rate after RC is 22%–38% [8,9], and accompanying micrometastasis contributes to disease progression after definitive therapy [10]. As approximately 10%–35% of clinically node-negative patients have micrometastasis, and RC with PLND alone provide insufficient control [11,12], effective micrometastatic disease eradication with perioperative systemic therapy is required [13].

Currently, the standard treatment for patients with non-metastatic MIBC eligible for cisplatin treatment is cisplatin-based NAC followed by RC [14,15]. The combination of cisplatin-based NAC and RC has shown potential in increasing the pathological complete response (pCR) or downstaging to NMIBC, recurrence-free survival (RFS), cancer-specific survival (CSS) and OS in MIBC patients [16,17,18]. Despite improvements in pCR and survival rates with NAC, recurrence occurs within 2–3 years after RC leading to death of patients [19]. Additionally, approximately 50% of patients undergoing RC are cisplatin-ineligible, and NAC with cisplatin is administered in <25% of patients in actual clinical practice [20].

Immune checkpoint inhibitors (ICIs) including anti-programmed cell death-1 (PD-1)/programmed cell death-ligand 1 (PD-L1) antibodies [21,22,23], fibroblast growth factor receptor 3 inhibitors, and antibody-drug conjugates (ADCs) [24] have recently been introduced into therapeutic landscape of locally advanced or metastatic BC, and their promising clinical benefits against platinum-resistant BC have been validated through multiple large randomized clinical trials (RCTs). Several studies researching the perioperative roles of ICIs and ADCs are ongoing [24,25,26]. Herein, we discuss the current status and future perspectives of perioperative systemic strategies for MIBC.

EVIDENCE OF CISPLATIN-BASED NAC FOR MIBC

European Association of Urology guidelines strongly recommend cisplatin-based combination NAC for cisplatin-eligible patients with non-metastatic MIBC (cT2-4aN0M0) [14]. Furthermore, National Comprehensive Cancer Network guidelines recommend cisplatin-based combination NAC as standard treatment (category 1) for cT2–4a or cN1 MIBC [15]. Well-designed RCTs have shown the survival benefit of cisplatin-based NAC [17,27]. Grossman et al. [17] investigated the role of 3 cycles of methotrexate, vinblastine, doxorubicin, and cisplatin (MVAC) regimen in a phase 3 RCT. A total of 317 MIBC patients (cT2–4aN0M0) were enrolled, with 153 and 154 patients assigned to MVAC and control groups, respectively. Median survival was prolonged in the NAC group, as compared with the control group (77 months vs. 46 months, p=0.06). Moreover, the NAC group had higher pCR rates than did the control group after RC (38% vs. 15%, p<0.001). Similarly, an international, multicenter RCT (BA06 30894) reported long-term results regarding the efficacy of 3 cycles of cisplatin, methotrexate, and vinblastine [27]. A total of 976 patients were randomly allocated to NAC (n=491) or control arms (n=485). Patients in the NAC arm received 3 cycles of cisplatin, methotrexate, and vinblastine every 3 weeks prior to local treatment; RC, radiotherapy (RTx), or RTx followed by RC. This NAC regimen improved OS (hazard ratio [HR], 0.84; p=0.037), metastasis-free survival (HR, 0.77; p=0.001) and disease-free survival (DFS) (HR, 0.82; p=0.008) over local treatment alone (Table 1) [17,27,28,29,30,31].

Table 1. Evidence on neoadjuvant chemotherapy in muscle-invasive bladder cancer.

Trial	Phase	Population	Sample size	Study design	Regimen of NAC		pCR rate (%)
Grossman et al. (2003) (SWOG 8710 trial) [17]	3	T2-T4a, N0, M0	153 in NAC group 154 in control group	NAC+Cx Cx	Methotrexate 30 mg/m2 Vinblastine 3 mg/m2 Doxorubicin 30 mg/m2 Cisplatin 70 mg/m2	3 cycles q4w	38
International Collaboration (2011) (BA06 30894) [27]	3	T2-T4a, N0-Nx, M0	491 in NAC group 485 in control group	NAC+Cx and/or RTx Cx and/or RTx	Methotrexate 30 mg/m2 Vinblastine 4 mg/m2 Cisplatin 100 mg/m2	3 cycles q3w	33
Choueiri et al. (2014) (NCT00808639) [28]	2	T2-T4a, N0-N1, M0	39	NAC+Cx	Methotrexate 30 mg/m2 Vinblastine 3 mg/m2 Doxorubicin 30 mg/m2 Cisplatin 70 mg/m2	4 cycles q2w	26
Malmstrom et al. (1996) (Nordic I) [29]	3	T1-T4a, Nx, M0	124 in NAC group 122 in control group	NAC+RTx+Cx RTx+Cx	Cisplatin 70 mg/m2 Doxorubicin 30 mg/m2	2 cycles q3w	-
Sherif et al. (2002) (Nordic II) [30]	3	T2-T4a, Nx, M0	155 in NAC group 154 in control group	NAC+Cx Cx	Cisplatin 100 mg/m2 Methotrexate 250 mg/m2	3 cycles q3w	-
Plimack et al. (2014) (NCT01031420) [31]	2	T2-T4a, N0-N1, M0	44	NAC+Cx	Methotrexate 30 mg/m2 Vinblastine 3 mg/m2 Doxorubicin 30 mg/m2 Cisplatin 70 mg/m2	3 cycles q2w	38

NAC, neoadjuvant chemotherapy; pCR, pathological complete response; RTx, radiotherapy; Cx, cystectomy; q3w, every 3 weeks; q4w, every 4 weeks; q2w, every 2 weeks.

Several meta-analyses have also identified a NAC role in MIBC treatment [7,18]. The Advanced Bladder Cancer Meta-analysis Collaboration analyzed the data of 3,005 patients with MIBC from 11 RCTs [7]. An OS benefit was observed in patients treated with platinum-based combination NAC (HR, 0.86; 95% confidence interval [CI], 0.77–0.95), representing a 14% reduction in mortality risk, as compared with that in controls. This was equivalent to a 5% absolute 5-year OS benefit. Another meta-analysis, which involved 3,285 patients from 15 RCTs, confirmed OS benefit of cisplatin-based NAC in MIBC patients (HR, 0.87; 95% CI, 0.79–0.96) [18]. The survival benefit was more prominent with a cisplatin-based combination NAC than with a single-agent NAC (HR, 0.84 vs. 1.10).

OPTIMAL NAC REGIMEN: GEMCITABINE PLUS CISPLATIN VS. DOSE-DENSE MVAC

Since the 1980s, conventional MVAC regimen with a 4-week schedule has been used as an active treatment for metastatic BC [28]. Despite the established survival benefit of MVAC regimens, concerns regarding MVAC toxicity have limited its clinical use and contributed to the development of more effective and tolerable chemotherapeutic regimens. In a phase 3 RCT comparing gemcitabine plus cisplatin (GC) with MVAC regimen for metastatic BC, GC showed comparable efficacy but less toxicity when compared with MVAC [32,33]. GC regimen with a 3-week schedule has been integrated into neoadjuvant MIBC treatment, without prospective RCT data. Dash et al. [34] retrospectively evaluated neoadjuvant GC and reported that 4 cycles of neoadjuvant GC achieved a similar pCR rate compared with neoadjuvant MVAC (26% and 28% for GC and MVAC, respectively). Multiple single-arm prospective studies also reported that GC followed by RC was effective and well-tolerated, as compared with neoadjuvant MVAC [35,36,37].

As neoadjuvant therapy, dose-dense or accelerated regimens have been developed to reduce toxicity and improve benefits for BC patients. Plimack et al. [31] investigated the efficacy and safety of accelerated neoadjuvant MVAC. A total of 44 MIBC patients (cT2–4aN0–1M0) received accelerated MVAC (3 cycles of 30 mg/m² of methotrexate, 3 mg/m² of vinblastine, 30 mg/m² of doxorubicin, and 70 mg/m² of cisplatin every 2 weeks) with pegfilgrastim support before RC. They reported a pCR rate of 38% and <pT2 downstaging rate of 53% with a tolerable toxicity profile, which were comparable to those of conventional 12-week MVAC and GC regimens. In a phase 2 RCT (SWOG S1314 trial), Flaig et al. [38] compared the efficacy of neoadjuvant GC and dose-dense MVAC (ddMVAC) and reported that pCR rate (30% vs. 28%, p=0.75) and pathological downstaging to ≤T1 rate (40% vs. 47%, p=0.27) were similar between neoadjuvant GC and ddMVAC regimens, with the rates of adverse events (AEs) ≥grade 3 of 44% (50/114) and 38% (41/109) in GC and ddMVAC arms, respectively.

In 2022, the results of GETUG-AFU V05 VESPER trial, a multicenter phase 3 RCT comparing GC and ddMVAC perioperative regimens, were presented [39]. Out of 493 MIBC patients with cT2–4aN0M0 disease, 437 received 4 cycles of GC (n=219) or 6 cycles of ddMVAC (n=218) in a neoadjuvant setting. The primary and secondary endpoints were 3-year progression-free survival (PFS) and pathological response rate, respectively. The ddMVAC group exhibited greater improvement in 3-year PFS than did the GC group (66% vs. 56%, HR=0.70, p=0.025), although OS data were immature. Similar pCR rates were observed in the ddMVAC (42%) and GC (36%) groups (p=0.2). With respect to safety, grade ≥3 toxicities were more frequently observed in the ddMVAC than in the GC arm, including severe anemia (p<0.001), nausea/vomiting (p=0.003) and asthenia (p<0.001). More deaths occurred in the ddMVAC (n=3) than in the GC (n=1) arm (Table 2) [34,35,36,37,38,39]. To summarize, GC and ddMVAC are the most actively used neoadjuvant cytotoxic chemotherapeutic regimens for MIBC. Long-term OS data is needed and safety profiles should be considered in choosing a NAC regimen between GC and ddMVAC.

Table 2. Optimal regimen for neoadjuvant chemotherapy in muscle-invasive bladder cancer.

Trial	Phase	Population	Sample size	Study design	Regimen		pCR rate (%)
Dash et al. (2008) [34]	Retrospective	T2-T4a, N0, M0	42 in GC group	GC+Cx	Gemcitabine 1,000 mg/m2 Cisplatin 70 mg/m2 (single dose) or Cisplatin 35 mg/m2 (split dose)	4 cycles q3w	26
			54 in MVAC group	MVAC+Cx	Methotrexate 30 mg/m2 Vinblastine 3 mg/m2 Doxorubicin 30 mg/m2 Cisplatin 70 mg/m2	4 cycles q4w	28
Iyer et al. (2018) (NCT01589094) [35]	2	T2-T4a, N0, M0	46 in ddGC group	ddGC+Cx	Gemcitabine 2,500 mg/m2 Cisplatin 35 mg/m2	6 cycles q2w	15
Niedersüss-Beke et al. (2017) (NCT01801644) [36]	-	T2-T4a, N0-N1, M0	87 in GC group	GC+Cx	Gemcitabine 1,000 mg/m2 Cisplatin 70 mg/m2	3 cycles q3w or q4w	22.5
Herchenhorn et al. (2007) [37]	2	T2-T4a, N0-N1, M0	21 in GC group	GC+Cx	Gemcitabine 1,200 mg/m2 Cisplatin 75 mg/m2	3 cycles q3w	20
Flaig et al. (2021) (SWOG S1314; NCT02177695) [38]	2	T2-T4a, N0, M0	112 in ddMVAC group	ddMVAC+Cx	Methotrexate 30 mg/m2 Vinblastine 3 mg/m2 Doxorubicin 30 mg/m2 Cisplatin 70 mg/m2	4 cycles q2w	28
			115 in GC group	GC+Cx	Gemcitabine 1,000 mg/m2 Cisplatin 70 mg/m2	4 cycles q3w	30
Pfister et al. (2022) (GETUG-AFU V05 VESPER; NCT01812369) [39]	3	T2-T4a, N0, M0	218 in ddMVAC group	ddMVAC+Cx	Methotrexate 30 mg/m2 Vinblastine 3 mg/m2 Doxorubicin 30 mg/m2 Cisplatin 70 mg/m2	6 cycles q2w	42
			219 in GC group	GC+Cx	Gemcitabine 1,250 mg/m2 Cisplatin 70 mg/m2	4 cycles q3w	36

GC, gemcitabine-cisplatin; pCR, pathological complete response; Cx, cystectomy; MVAC, methotrexate-vinblastine-doxorubicin-cisplatin; q2w, every 2 weeks; q3w, every 3 weeks; q4w, every 4 weeks; ddGC, dose-dense GC; ddMVAC, dose-dense MVAC.

NAC FOR PATIENTS WITH IMPAIRED RENAL FUNCTION

Although the clinical benefit of cisplatin-based combination NAC for MIBC has long been established, less than 25% of patients are treated with cisplatin-based NAC [20,40]. Renal dysfunction is among the most common causes for cisplatin ineligibility, as cisplatin, a well-known nephrotoxic drug, results in renal dysfunction in up to 40% of cisplatin-based chemotherapy recipients [41].

To reduce the nephrotoxic effects of cisplatin in patients with renal dysfunction, alternative regimens using carboplatin have been evaluated. In a phase 2 trial, the efficacy of a neoadjuvant paclitaxel, carboplatin, and gemcitabine (PCG) regimen was evaluated in patients with T2–4a urothelial cancer (UC) without distant metastasis [42]. The PCG regimen was proven ineffective with the 2-year OS of 59% and a high rate of remnant disease after RC [42]. A retrospective study by Peyton et al. [43] also demonstrated that pCR and 2-year OS for carboplatin- and cisplatin-based NAC regimens were 9% vs. 24% and 35% vs. 62%–73%, respectively. Neoadjuvant carboplatin for MIBC was proven suboptimal.

GC regimen with split dose cisplatin may be an alternative option for patients with diminished renal function. In a phase 2 study that included 23 MIBC patients (cT2-4N0M0) with glomerular filtration rate (GFR) ≥40 mL/min, Hussain et al. [44]. reported that a NAC regimen with fractionated cisplatin dosing (1,000 mg/m² of gemcitabine plus 35 mg/m² of cisplatin on days 1 and 8 every 3 weeks) followed by RC or RT, was well-tolerated, with pCR rate of 48%. AEs ≥grade 3 occurred in 4 cases, including neutropenic sepsis, nausea, vomiting and diarrhea. No significant deterioration in renal function was observed even in patients with impaired renal function (GFR of 40–60 mL/min). In a retrospective study, Osterman et al. [45] compared the efficacy of cisplatin-based NAC regimens in MIBC (cT2-4) patients with GFR ≥40 mL/min. Forty patients received 3–4 cycles of split-schedule GC (1,000 mg/m² of gemcitabine plus 35 mg/m² of cisplatin on days 1 and 8, every 3 weeks) or ddMVAC (35 mg/m² of cisplatin plus 30 mg/m² of methotrexate on day 1, 35 mg/m² of cisplatin plus 30 mg/m² of doxorubicin and 3 mg/m² of vinblastine on day 2, every 2 weeks), and another 40 patients received conventional cisplatin-based GC or ddMVAC. pCR (pT0) and partial response (<pT2) rates were lower in the split than conventional schedule (17.5% vs. 32.5%, p=0.21 for pCR; 27.5% vs. 45%, p=0.21 for partial response), but the differences were non-significant. The AE rates of any grade, including nephrotoxicity (increased serum creatinine level >1.5 times baseline), were not significantly different between the two regimens. To summarize, NAC with split-dose cisplatin may be an alternative option in selected cases with borderline renal dysfunction (GFR of 40–60 mL/min) with comparable efficacy and safety.

BIOMARKERS FOR PREDICTING RESPONSE TO NAC

Despite the survival benefit of cisplatin-based NAC in MIBC, the absolute DFS benefit is modest at 7% and potential treatment delay should be considered [20,46]. A percentage of patients fail to respond to NAC and may experience delayed RC, while incurring treatment toxicity [47,48]. The prolonged time to RC (>10 weeks) after NAC adversely affected OS and CSS [47]. Therefore, more accurate stratification of patients undergoing NAC is necessary to identify patients more likely to respond. The development of clinically reliable predictors of response to NAC is an unmet need. By the same logic, it could also facilitate selection of patients who could safely avoid the attendant morbidity and mortality of RC without compromising oncologic outcomes.

1. Variant histology and UC with divergent differentiation

Most specimens reveal pure UC in up to 33% RC specimens [49] and pure UC is generally associated with greater pCR rates at RC than are mixed tumors [50]. UC with divergent differentiation is commonly associated with high grade and advanced disease [49], and patients with UC with divergent differentiation experienced significantly worse survival outcomes regardless of NAC use pre-RC [51]. In addition, around 25% of MIBC cases have associated variant histology (VH) [52]. Although the unique nature of VH BC suggests a need to identify treatment plans tailored to specific VHs [53], there are no definitive data to guide therapy for those with major or pure VH components [52]. Several studies have evaluated NAC in this setting with varying results for each VH to provide further direction on its appropriate use. For example, NAC was associated with pathological downstaging for all MIBC with VH, with improved OS for patients with UC, sarcomatoid variant UC and neuroendocrine carcinoma [54]. Similarly, based on a retrospective review, patients with micropapillary, sarcomatoid, plasmacytoid, mixed and neuroendocrine variant histologies may benefit from NAC; although those with squamous and adenocarcinoma did not [55]. In contrast, in a subgroup analysis from the RCT SWOG 8710, patients with mixed histology (squamous or glandular) had a survival benefit with NAC (HR=0.46) compared with primary RC [56]. These results are consistent with the findings of Kaimakliotis et al. [57], who reported similar oncological outcomes of NAC for patients with pure UC and VH, with the latter being mostly represented by squamous or glandular differentiation.

Current evidence does not indicate a clinically significant survival benefit in using NAC for sarcomatoid variants [52,53,58]. For example, sarcomatoid variants are associated with adverse pathological features, 5-year CSS rates as low as 28.4%, and a poor response to NAC [58]. Therefore, immediate RC has been proposed as first-line therapy for sarcomatoid variants. Data on the benefit of NAC for micropapillary histology are mixed [53,59]. A series from the Memorial Sloan Kettering Cancer Center found that 45% of patients with any micropapillary disease component were downstaged to pT0 after NAC with GC, although there was no significant OS benefit [59]. Another group reported worse survival for patients receiving NAC compared to immediate RC (32% vs. 71% of 5-year OS) [60]. Results from various studies are summarized in a recent meta-analysis that investigated the effects of NAC in the presence of the micropapillary variant [61]. Authors reported pCR ranging from 11% to 55% without significantly affecting RFS, CSS, or OS [61]. Finally, in plasmacytoid UC, the role of NAC is unclear due to small retrospective studies with different chemotherapy regimens [52].

2. Molecular subtype

Recent studies suggest that molecular subtype classification holds promise for predicting NAC response and/or survival benefit in this setting. Basal/squamous (Ba/Sq) or luminal non-specified tumors may benefit from NAC despite the aggressiveness; luminal tumors are less aggressive but more NAC-resistant; while some subtypes with characteristics of p53-like or claudin-low subtypes have the worst prognosis despite NAC administration [13,48,62,63,64,65]. For example, in patients with MIBC, a commercially available molecular subtyping assay revealed non-luminal tumors received the greatest benefit from NAC, with 10% greater OS at 3 years (71% vs 61%), while patients with luminal tumors experienced minimal survival benefit (63% vs 65%) for NAC vs. non-NAC [64]. Especially, the Ba/Sq subtype, identified in the TCGA, UNC, and MD Anderson classifications, is associated with more aggressive tumors, worse prognosis, and poor response to NAC, which is inconsistent with previous reports [62,63,65]. For example, a recent study investigated how molecular subtypes impact pathological response and survival in patients receiving preoperative cisplatin-based NAC [65]. Patients with the Ba/Sq subtype had lower pCR proportions (21%), vs. 52% and 31% with the genomically unstable (GU) and urothelial-like (Uro) tumors, respectively, following NAC and RC [65]. Molecular subtype was independently associated with improved survival for patients with GU tumors (HR 0.29) and Uro (subtype C) tumors (HR 0.37) compared with Ba/Sq tumors, adjusting for clinical stage [65], suggesting that UC of the Ba/Sq subtype consistently displayed both less frequent pCR and worse survival outcomes following NAC [65].

3. Genomic alterations, gene expression signatures and cell free DNA

Cisplatin enters the cell through several membrane transporters, becomes activated, and causes DNA damage either as single-strand breaks (SSBs), double-strand breaks (DSBs), or interstrand-crosslinks, which prevents cell replication and induces cell death [48,62]. There are four described mechanisms of cisplatin resistance: (1) pretarget resistance, defined as reduced intracellular drug accumulation (i.e., CTR1) or sequestration; (2) on-target resistance, defined as the ability to repair or tolerate DNA damage, mediated mainly by nucleotide excision repair (NER) proteins (i.e., ERCC1), mismatch repair proteins and homologous recombination (HR) systems (i.e., BRCA); (3) post-target resistance involves signaling pathway alterations leading to apoptosis (i.e., p53, BLC-2); and (4) off-target resistance occurs when signaling pathways not directly influenced by cisplatin are involved (i.e., ERBB2/EGFR) [48].

The most interesting results come from studies evaluating DNA damage repair (DDR) genes [48]. SSBs can be repaired by base excision, mismatch, or NER, while DSBs can be repaired by non-homologous end joining or HR [62]. DDR gene alterations were associated with improved chemosensitivity and survival in cisplatin-based NAC [13]. Therefore, the presence of mutations in DDR pathways can identify patients likely to respond to cisplatin-based NAC [62]. For example, the NER pathway, which repairs SSBs, includes the ERCC excision repair 1 (ERCC1) and ERCC excision repair 2 (ERCC2) proteins [62]. Especially, ERCC2 mutations are found in 12% of BCs and have been associated with a loss of NER pathway function in preclinical studies [66]. ERCC2 mutations are the most validated single gene alterations in predicting response to cisplatin-based NAC [66,67,68]. Other studies have also correlated mutations in the ATM serine/threonine kinase (ATM), RB transcriptional corepressor 1 (RB1) and FA complementation group C (FANCC) repair genes with efficacy (pathological response, PFS and OS) of cisplatin-based NAC in MIBC [69]. Similarly, a recent phase II trial in 49 patients treated with neoadjuvant GC analyzed a panel of 29 DDR genes and found that the presence of deleterious mutations, including ERCC2 mutations, was associated with response to chemotherapy, with a positive predictive value of 89% and a two-year RFS of 100% [35].

Furthermore, several novel gene signatures have been developed as promising biomarkers to predict chemoresponsiveness of MIBC patients. In the prospective SWOG S1314 study, Co-expression Exploration (COXEN) program was undertaken to primarily evaluate COXEN, a gene expression prediction model, as a predictive biomarker in MIBC patients treated with neoadjuvant ddMVAC and GC [70]. The COXEN score was significantly associated with downstaging in combined ddMVAC and GC arms (p=0.02) [70]. The ranked guided iterative feature elimination machine learning algorithm, previously demonstrated to have the ability to select biomarkers with high predictive power, also identified a 9-gene signature (CNGB1, GGH, HIST1H4F, IDO1, KIF5A, MRPL4, NCDN, PRRT3, and SLC35B3) able to select responders from non-responders to NAC with 100% predictive accuracy [71].

The role of cell free DNA (cfDNA) methylation for predicting the response to NAC was also investigated in 73 patients who received NAC in SWOG S1314 trial [72]. Lu et al. [72] reported the predictive accuracy for pathologic response was 79% using methylation-based response score and circulating bladder DNA fraction, which were calculated based on cfDNA methylation data.

ROLE OF NEOADJUVANT IMMUNOTHERAPY

Several immunotherapeutic agents have been approved as first- and second-line treatments for cisplatin-ineligible or refractory locally advanced or metastatic MIBC patients [14,15]. The first ICI applied in the neoadjuvant setting for MIBC was ipilimumab, an anti-cytotoxic T-lymphocyte associated protein 4 (CTLA-4) antibody [73]. Neoadjuvant ipilimumab therapy lowered disease stage after RC in localized BC (cT1-2N0M0) patients. Recently, there is increasing interest in novel ICI use as neoadjuvant therapy for MIBC.

1. ICI therapy in the neoadjuvant setting

PURE-01 trial is a single-arm phase 2 study evaluating the efficacy of neoadjuvant pembrolizumab in 50 MIBC patients (cT≤3bN0-1) [74]. Three cycles of pembrolizumab (200 mg) were administered every 3 weeks, followed by RC. The results indicated that 42% of pCR and 54% of downstaging to <pT2 were achieved, meeting study endpoints. In biomarker analyses, increased levels of PD-L1 expression (combined positive score ≥10%) and tumor mutational burden (TMB; >15 mutations/Mb) were positively associated with pCR. Thyroid dysfunction (18%) was the most frequent AE.

In ABACUS trial, 2 cycles of atezolizumab (1,200 mg) with 3-week intervals were administered to 95 cisplatin-ineligible MIBC patients with cT2–4N0M0 tumors [75]. Eighty-seven patients underwent cystectomy. The pCR and 1-year RFS rates were 31% and 79%, respectively, thus achieving the primary and secondary endpoints. The pCR rate was higher in PD-L1-positive (≥5% of immune cell) patients, reaching 37%. Preexisting activated T-cells were associated with better oncological outcomes; higher pCR and 1-year RFS rates in patients with high presence of CD8 positive intraepithelial cells. Grade 3 or 4 AEs occurred in 10 (11%) patients.

NABUCCO trial, examined the neoadjuvant efficacy of ipilimumab plus nivolumab combination regimen [76]. Twenty-four patients with stage 3 UC (cT3-4aN0M0 or cT2-4aN1-3M0), who were cisplatin-ineligible or refused cisplatin-based chemotherapy, were enrolled and received 3 mg/kg of ipilimumab on day 1, 3 mg/kg of ipilimumab plus 1 mg/kg of nivolumab on day 22, and 3 mg/kg of nivolumab on day 43 followed by RC. Primary endpoint was the feasibility of surgical resection within 12 weeks from neoadjuvant treatment initiation. Twenty-three (96%) patients underwent RC within 12 weeks, and 11 (46%) patients had a pCR. In 55% of patients, grade 3 or 4 immune-related AEs were identified including increased lipase/aminotransferase levels, fatigue and diarrhea (Table 3) [74,75,76].

Table 3. Clinical trials on neoadjuvant immune checkpoint inhibitor treatment in muscle-invasive bladder cancer.

Trial	Phase	Population	Sample size	Regimen	pCR rate (%)
Necchi et al. (PURE-01; NCT02736266) [74]	2	T≤3bN0	50	3 cycles of pembrolizumab (200 mg IV q3w)	42
Powles et al. (ABACUS; NCT02662309) [75]	2	T2-T4, N0, M0	95	2 cycles of atezolizumab (1,200 mg IV q3w)	31
Van Dijk et al. (NABUCCO; NCT03387761) [76]	1b	T3-4a, N0, M0 or T2-4a, N1–3, M0	24	Ipilimumab 3 mg/kg (day 1) Ipilimumab 3 mg/kg plus nivolumab 1 mg/kg (day 22) Nivolumab 3 mg/kg (day 43)	46

pCR, pathological complete response; q3w, every 3 weeks.

2. ICI plus chemotherapy combination in the neoadjuvant setting

Cytotoxic chemotherapy has been shown to modulate immune response via several mechanisms; stimulation of T-cell activation and dendritic cell maturation, immunogenic cell death induction, and reduction in the immunosuppressive function of immune cells [77]. Combining a PD1/PD-L1 antagonist with cytotoxic agents may provide complementary benefits in increasing effective antitumor immunity [77,78]. Based on this theoretical background, the potential preoperative role of combination therapy with ICI and conventional cytotoxic chemotherapy has been actively investigated (Table 4) [79,80,81,82,83].

Table 4. Clinical trials on neoadjuvant immune checkpoint inhibitor treatment plus chemotherapy in muscle-invasive bladder cancer.

Trial	Phase	Population	Sample size	Regimen		pCR rates (%)	Downstaging rate (<pT2N0M0) (%)
Gupta et al. (2020) (BLASST-1; NCT03294304) [79]	2	T2-4a, N0-1, M0	41	Nivolumab 360 mg (day 8) Cisplatin 70 mg/m2 (day 1) Gemcitabine 1,000 mg/m2 (day 1, 8)	4 cycles q3w	34	65.8
Hoimes et al. (2020) (HCRN GU14-188; NCT02365766) [80]	1b/2	T2-4a, N0, M0	43 (cohort 1)	Nivolumab 360 mg (day 8) Cisplatin 70 mg/m2 (day 1) Gemcitabine 1,000 mg/m2 (day 1, 8)	5 cycles q3w (starting at day 8) 4 cycles q3w	44.4	61.1
Funt et al. (2022) (NCT02989584) [81]	2	T2-4a, N0, M0	39	Atezolizumab 1,200 mg (day 8) Cisplatin 70 mg/m2 (day 1) Gemcitabine 1,000 mg/m2 (day 1, 8) or Split-dose cisplatin 35 mg/m2 (day 1, 8) Gemcitabine 1,000 mg/m2 (day 1, 8)	6 cycles q3w 4 cycles q3w	41	69
Cathomas et al. (2020) (SAKK 06/17; NCT03406650) [82]	2	T2-4a, N0-1, M0	34	Durvalumab 1,500 mg	4 cycles q3w (starting at cycle 2) (*10 cycles of durvalumab continue after surgery)	30	50
				Cisplatin 70 mg/m2 (day 1) Gemcitabine 1,000 mg/m2 (day 1, 8)	4 cycles q3w
Rose et al. (2021) (NCT02690558) [83]	2	T2-4a, NX-0, M0	39	Pembrolizumab 200 mg (day 1) Split-dose cisplatin 35 mg/m2 (day 1, 8) Gemcitabine 1,000 mg/m2 (day 1, 8)	4 cycles q3w	36	56

pCR, pathological complete response; q3w, every 3 weeks.

In BLASST-1 trial, the efficacy and safety of neoadjuvant nivolumab plus GC were investigated [79]. Forty-one MIBC patients (cT2–cT4a, N0–N1, M0) received 4 cycles of nivolumab plus GC every 3 weeks. RC was performed within 8 weeks of neoadjuvant therapy. A pathological response ≤pT1N0, primary endpoint, was observed in 27 (65.8%) patients. The rate of grade 3–4 AEs, including thrombocytopenia, neutropenia and renal dysfunction, was 20% mainly caused by GC chemotherapy.

In GU14-188 trial, including cisplatin-eligible [80] and cisplatin-ineligible [84] non-metastatic MIBC patients (cT2–T4aN0M0), the tolerability and efficacy of chemo-immunotherapy were investigated. Forty-three cisplatin-eligible patients received 5 doses of pembrolizumab plus 4 cycles of GC every 3 weeks prior to RC [80]. In 36 patients who underwent RC, the pathological complete (ypT0N0) and partial (≤ypT1N0) response rates were 44% and 61%, respectively. In the cisplatin-ineligible arm, five doses of pembrolizumab in combination with three cycles of gemcitabine were administered to 37 patients [84]. The pathological response rate for ypT0N0 and ≤ypT1N0 were 45% and 52%, respectively. In both arms, the feasibility and efficacy of pembrolizumab and GC or gemcitabine chemotherapy were identified, with manageable treatment-related toxicity.

In another single-arm, phase 2 trial, neoadjuvant GC plus atezolizumab (anti-PD-L1) was investigated [81]. Forty-four cisplatin-eligible MIBC patients (cT2-4aN0M0) received a lead-in dose of atezolizumab, followed 2 weeks later by four cycles of GC with 3-week intervals plus atezolizumab and one additional atezolizumab dose 3 weeks after GC. RC is recommended within 4–8 weeks of the last dose of GC. The rates of downstaging to <pT2N0 and pCR were 69% (27/39) and 41% (16/39), respectively, achieving endpoints. Grade 3 AEs were observed in 11% of patients.

SAKK 06/17 trial investigated the feasibility of perioperative durvalumab in combination with GC [82]. Primary and secondary endpoints were 2-year event-free survival and safety/pathological response rates, respectively. Four cycles each of durvalumab and GC were administered as neoadjuvant treatment for 34 muscle-invasive urothelial carcinoma patients (cT2-4aN0–N1M0; 33 BC and one upper tract cancer). Ten cycles of durvalumab maintenance continued postoperatively. In the interim analysis, four cycles of combination neoadjuvant treatment were completed in all patients, with 30% of pCR rate and tolerable AEs (Table 4) [79,80,81,82,83].

ADJUVANT THERAPY FOR MIBC

1. Platinum-based adjuvant chemotherapy

Adjuvant chemotherapy has some benefits, including accurate staging to reduce overtreatment and preventing RC delay. Nonetheless, evidence supporting the oncological benefits of adjuvant chemotherapy is less clear in MIBC. Some large retrospective studies have reported prolonged survival owing to adjuvant chemotherapy [85,86,87], despite the lack of well-designed RCTs. In the first prospective trial (n=91) testing adjuvant chemotherapy efficacy, Skinner et al. [88] compared 3-year PFS rates between adjuvant chemotherapy and RC-only groups in MIBC patients (pT3-4 or pN+). Four cycles of a cisplatin-doxorubicin-cyclophosphamide regimen were administered as adjuvant chemotherapy. They reported better 3-year PFS in the adjuvant chemotherapy arm (70%) than in the cystectomy-only arm (46%; p=0.001). In another phase 3 trial, the Spanish Oncology Genitourinary Group reported that four cycles of a paclitaxel-gemcitabine-cisplatin regimen significantly prolonged OS in the adjuvant therapy arm compared to the observation arm (5-year OS: 60% vs. 31%, respectively) [89]. In a recent meta-analysis of 945 patients from nine RCTs, the authors reported a survival benefit in OS (HR=0.77, p=0.049) and DFS (HR=0.66, p=0.010) [90]. However, the level of evidence appears low due to methodological flaws.

2. Adjuvant immunotherapy

While evidence for routine conventional chemotherapy use in adjuvant settings is lacking, adjuvant immunotherapy, especially agents that block the PD-1/PD-L1 axis, is emerging as a novel perioperative treatment for MIBC. Recently, several phase 3 RCTs have tested adjuvant ICIs in MIBC.

In CheckMate 274 trial, a phase 3 multicenter RCT included pT3-4a, or pN+ patients who were not eligible for, or declined cisplatin-based NAC, and ypT2-4a or ypN+ patients who underwent cisplatin-based NAC. The results showed significantly improved DFS for 353 patients receiving adjuvant nivolumab (240 mg every 2 weeks for up to 1 year or until disease progression), compared to 356 patients who received a placebo (median DFS 20.8 vs. 10.8 months, p<0.001) [91]. Survival benefit of adjuvant nivolumab was also valid in the population with PD-L1 expression level ≥1% meeting the primary endpoint, with 6-month DFS rates of 74.5% vs. 55.7% (p<0.001). AE ≥grade 3 were found in 18% and 7% of the nivolumab and placebo groups, respectively.

However, in another phase 3 RCT, IMvigor010 trial, which tested adjuvant atezolizumab efficacy (1,200 mg every 3 weeks for 16 cycles) over observation post-RC, no definite benefit in DFS was found in adjuvant atezolizumab group over placebo group (median DFS 19.4 vs. 16.6 months, p=0.24) in the intention to treat population (Table 5) [91,92]. There is an ongoing phase 3 trial (AMBASSADOR trial) evaluating adjuvant pembrolizumab, whose results are as yet unreported.

Table 5. Clinical trials on adjuvant immune checkpoint inhibitor treatment in muscle-invasive bladder cancer.

Variable	Bajorin et al. (2021) (CheckMate 274; NCT02632409) [91]	Bellmunt et al. (2021) (IMvigor010; NCT02450331) [92]
Phase	3	3
Population	Urothelial carcinoma (% of UTUC=21.0%) pT3-4a or pN+(without NAC) or ypT2-4a or ypN+(with NAC)	Urothelial carcinoma (% of UTUC=6.7%) pT3-4a or pN+(without NAC) or ypT2-4a or ypN+(with NAC)
Sample Size	Total 709 353 in nivolumab group 356 in placebo group	Total 809 406 in atezolizumab group 403 in observation group
Regimen of adjuvant treatment	Nivolumab group: Nivolumab 240 mg IV every 2 wk for up to 1 y Placebo group: Placebo every 2 wk for up to 1 y	Atezolizumab group: Atezolizumab 1,200 mg IV every 3 wk for up to 1 y Observation group: observation
Primary endpoint	DFS in ITT population Nivolumab group: 20.8 (95% CI, 16.5–27.6) mon Placebo group: 10.8 (95% CI, 8.3–13.9) mon HR=0.70 (p<0.001)	DFS in ITT population Atezolizumab group: 19.4 (95% CI, 15.9–24.8) mon Observation group: 16.6 (95% CI, 11.2–24.8) mon HR=0.89 (p=0.24)
Rate of TRAEs≥grade 3	Nivolumab group: 17.9% Placebo group: 7.2%	Atezolizumab group: 16% Observation group: -

UTUC, upper tract urothelial carcinoma; NAC, neoadjuvant chemotherapy; IV, intravenous; DFS, disease-free survival; ITT, intention to treat; CI, confidence interval; HR, hazard ratio; TRAE, treatment-related adverse event.

EVOLVING PARADIGM OF PERIOPERATIVE TREATMENT FOR MIBC

1. Immunotherapy in the neoadjuvant and adjuvant setting

Several studies have investigated neoadjuvant chemoimmunotherapy in combination with adjuvant immunotherapy in resectable MIBC. ENERGIZE trial is a phase 3 study evaluating the neoadjuvant nivolumab efficacy with or without linrodostat, an irreversible indoleamine 2, 3-dioxygenase 1 (IDO1) inhibitor, combined with GC in the patients with cT2-4aN0M0 MIBC [26]. IDO1 converts tryptophan to kynurenine, which upregulates regulatory T-cells and inhibits immune cell infiltration. After RC, adjuvant nivolumab with or without linrodostat was scheduled. Primary endpoints are pCR and EFS. NIAGARA trial is another phase 3 study that tested the efficacy and safety of neoadjuvant durvalumab plus GC, followed by adjuvant durvalumab treatment in the patients with cT2-4aN0-1M0 MIBC [25]. Primary endpoints are pCR and EFS in the population without renal dysfunction. In PIVOT IO 009 trial, a phase 3 RCT, the efficacy of neoadjuvant and adjuvant nivolumab in combination with bempegaldesleukin, a CD 122 preferential interleukin-2 agonist, in cisplatin-ineligible MIBC patients (cT1-4aN1M0) is under investigation [93]. Primary endpoints are pCR and EFS. These clinical trials are examples of numerous ongoing studies evaluating novel perioperative treatments for MIBC. Recently attempts to find biomarkers for immunotherapy have been made. PD-L1 level, TMB, immune cell infiltration and DDR gene alteration are the candidates of biomarkers for immunotherapy [94]. However, none of these has shown consistent results to be included in routine clinical practice.

2. Novel therapeutic agent: ADC

ADCs are monoclonal antibodies covalently linked to cytotoxic payloads [95]. ADCs selectively target specific antigens on the cancer surface and deliver cytotoxic drugs to tumor cells. The commonly used target antigens include nectin-4, human epidermal growth factor receptor 2 and trophoblast cell surface antigen 2 [96]. The ability to selectively deliver optimal cytotoxic drugs to different cancer cell types is a prominent advantage, leading to increased efficacy and reduced toxicity. The efficacy of ADCs as novel anticancer agents for diverse solid tumors, including metastatic BC, has been actively investigated [96].

Several ADCs have been tested in diverse clinical settings in the MIBC field. Enfortumab vedotin (EV) is the first ADC to show promising results. EV targets nectin-4, a heavily expressed cell adhesion molecule in BC; nectin-4 positive in 87% of NMIBC and 58% of MIBC [97]. The payload, monomethyl auristatin E, exhibits antimitotic activity by blocking tubulin polymerization [96]. In locally advanced or metastatic UC, EV treatment demonstrated efficacy, with a confirmed objective response rate of 44%, including 12% CR [98]. Recently, the OS benefit of EV treatment over chemotherapy for locally advanced or metastatic UC has been identified (median OS 12.88 vs. 8.97 months, p<0.001) [99]. A trial investigating the role of perioperative EV is currently in progress. Cohort L of EV-103 trial currently enrolls cisplatin-ineligible MIBC patients (cT2–T4aN0M0 or cT1-T4aN1M0). The primary endpoint is pCR rate [100].

Several phase 3 trials testing the EV-plus-ICI regimen efficacy are ongoing. VOLGA trial is investigating perioperative combination therapy with EV and durvalumab with or without tremelimumab (anti-CTLA-4 antibody) for cisplatin-ineligible MIBC patients. The perioperative role of EV combined with pembrolizumab was investigated in phase 3 trials, particularly KEYNOTE-095/EV-303 trial in cisplatin-ineligible patients and KEYNOTE-B15/EV-304 trial in cisplatin-eligible MIBC patients [24].

CONCLUSIONS

The recurrence rate of MIBC is substantial despite radical surgical treatment, eventually leading to metastatic disease and death. Perioperative systemic therapy, particularly cisplatin-based NAC, has shown survival benefits, becoming the current standard of care. Nonetheless, widespread perioperative chemotherapy use has been limited by chemotherapeutic agent toxicity. The recent ICI emergence has changed the MIBC treatment landscape in palliative and perioperative settings. ICIs alone or combined with conventional chemotherapy have demonstrated promising results with manageable toxicity. Furthermore, novel therapeutic agents such as ADCs (e.g., EV) are being tested by several trials in the perioperative setting based on the success of palliative treatment.