Emerging immunotherapies for bladder cancer

Abstract

Purpose of review

Inhibition of immune escape mechanisms, such as the programed death-ligand 1 pathway, has demonstrated rapid, durable responses in multiple tumor types, including advanced urothelial carcinoma. This review discusses emerging immunotherapies for urothelial carcinoma in various stages of clinical development.

Recent findings

Urothelial carcinoma has a high mutational burden, which may increase the number of tumor antigens and potentially enhance the ability of the immune system to recognize tumor cells as foreign. However, urothelial carcinoma can evade the immune system by downregulating tumor–antigen presentation, upregulating various immune checkpoints, and inactivating cytotoxic T cells. Immunotherapies for urothelial carcinoma target each of these steps to restore immune-mediated cytotoxicity. Many of these agents are in clinical trials for urothelial carcinoma.

Summary

Immunotherapies are active in urothelial carcinoma, but only in a fraction of patients, implying the presence of persistent immune escape. Identifying the mechanisms of immune escape and developing rational combinatorial regimens may make the benefit of immunotherapy accessible to a broader population.

INTRODUCTION

Bladder cancer is the most common malignancy of the urinary tract. Approximately 380 000 new cases are diagnosed each year worldwide, resulting in about 150 000 deaths [1]. It is the fourth most common cancer among men in developed countries. Bladder tumors can originate in the urinary bladder (90%), renal pelvis (8%), or ureter and urethra (2%). Epidemiological studies have identified a range of environmental risk factors. Smoking and chemical carcinogens such as aromatic amines, polycyclic hydrocarbons, and arsenic are associated with the development of urothelial carcinoma, and chronic infection with Schistosoma hematobium has been associated with squamous cell carcinoma of the bladder [1]. Genome-wide association studies have also identified several germline single-nucleotide polymorphisms that contribute to bladder cancer risk [2].

Clinically, bladder cancer is classified as non-muscle-invasive (NMIBC), muscle-invasive (MIBC), and metastatic bladder cancer. The majority (~70%) of patients present with NMIBC, which is usually managed with transurethral resection of the bladder tumor (TURBT) and may be followed by a course of intravesical Bacillus Calmette-Guerin (BCG) to decrease the risk of recurrence and progression in high-risk cases. Approximately 15–60% of bladder tumors recur at 1 year and 30–80% recur at 5 years [3]. BCG-refractory NMIBC is often managed with repeat TURBT and intravesical instillation with chemotherapy agents such as mitomycin, thiotepa, valrubicin, doxorubicin, and gemcitabine. In selected cases, radical cystectomy may be warranted. MIBC, a potentially lethal phenotype, occurs in about 25–45% of patients [4]. Approximately 20–25% of bladder cancer patients present with de-novo MIBC stage T2 or above [5].

Radical cystectomy is the mainstay of treatment for MIBC. Cisplatin-based combination chemotherapy administered before radical cystectomy or definitive local therapy improved survival in multiple randomized controlled trials [6,7] and meta-analyses [8,9]. To date, no adjuvant therapy after radical cystectomy has demonstrated improvement in overall survival (OS). The quality of data available on adjuvant therapy is inferior to level-1 evidence supporting neoadjuvant cisplatin-based chemotherapy [5]. However, a large proportion of patients with MIBC are ineligible for cisplatin-based chemotherapy because of multiple comorbidities, including hearing loss, cardiac disease, poor performance status, and renal insufficiency.

Bladder-preservation approaches provide a reasonable alternative for patients who cannot tolerate the risks associated with surgery, elderly patients, or those with significant comorbidities. A recent study indicated that patients receiving chemotherapy and radiotherapy after TURBT have a 5-year OS rate of 56% and a 5-year bladder-intact survival rate of 42% [10]. Approximately 50% of MIBC patients will eventually progress to metastatic disease despite aggressive multimodal therapy.

Multiple randomized controlled trials have established cisplatin-based combination chemotherapy as the standard, first-line treatment for metastatic bladder cancer [11-16], with a median OS of 12–15 months. Carboplatin is a common substitute for patients who cannot tolerate cisplatin [17]. An EORTC phase-III trial comparing gemcitabine/carboplatin with methotrexate/carboplatin/vinblastine in cisplatin-ineligible patients demonstrated a median survival of 8–9 months for both regimens [18]. Currently, there is no standard second-line therapy for patients who progress following platinum-based therapy.

EMERGING IMMUNOTHERAPIES FOR BLADDER CANCER

In recent years, immunotherapy has gained traction as a strategy for cancer treatment. (See Table 1 for emerging immunotherapeutic targets and ongoing clinical trials in bladder cancer.) Sipuleucel-T, a therapeutic vaccine, was approved by the US Food and Drug Administration (FDA) in 2010 for use in asymptomatic or minimally symptomatic metastatic castration-resistant prostate cancer [19]. Ipilimumab, a fully human monoclonal antibody that targets cytotoxic T-lymphocyte antigen-4 (CTLA-4), was approved for unresectable or metastatic melanoma in 2011 [20]. In 2014, the FDA approved pembrolizumab and nivolumab for advanced melanoma and blinatumomab for acute lymphocytic leukemia, and granted breakthrough therapy designation for the immune checkpoint inhibitors MPDL3280A, an antiprogramed cell death protein ligand 1 (anti-PD-L1) antibody for metastatic urothelial bladder cancer and nivolumab, an anti-PD-1 antibody, for Hodgkin’s lymphoma.

Table 1.

Selected emerging immunotherapies for bladder cancer

Class	Agent(s)	Target	Phase	Setting	Trial
Immune checkpoint inhibitor	MPDL3280A	PD-L1	II	Second-line advanced metastatic	NCT02302807
	Pembrolizumab	PD-1	III	Second-line advanced metastatic
	Pembrolizumab/BCG	PD-1	I	NMIBC	NCT02324582
	Nivolumab ± ipilimumab	PD1 CTLA4	I	Second-line advanced metastatic	NCT01928394
	Nivolumab + cabozantinib ± ipilimumab	PD-L1 cMET/VEGFR2 CTLA4	I	Second-line advanced metastatic	NCT02308943
	MEDI4736	PD-L1	I	Second-line advanced metastatic	NCT01693562
	AMP-514	PD-L1	I	Second-line advanced metastatic	NCT02013804
	MSB0010718C	PD-L1	I	Second-line advanced metastatic	NCT01943461
	MGA271	B7-H3	I	Second-line advanced metastatic	NCT01391143
Immune agonist	MEDI6383	OX40	I	Second-line advanced metastatic	NCT02221960
	MOXR0916	OX40	I	Second-line advanced metastatic	NCT02219724
Vaccine	DN24-02	HER2/neu	II	Adjuvant	NCT01353222
	AdHER2 DC	HER2/neu	I	Adjuvant and second-line advanced metastatic	NCT01730118
	HS410	Heat shock protein gp96	I	NMIBC	NCT02010203
	PANVAC	CEA and MUC-1	I	NMIBC	NCT02015104
	MAGE-A3	MAGE-A3	II	Adjuvant	NCT01435356
	DEC-205-NY-ESO-1 fusion protein vaccine	NY-ESO-1	I	Second-line advanced metastatic	NCT01522820
Adoptive T-cell transfer	Chimeric antigen receptor	NY-ESO-1	II	Second-line advanced metastatic (NY-ESO-positive)	NCT01967823
	Anti-MAGE-A3-DP4 TCR	MAGE-A3-DP4	I/II	Second-line advanced metastatic (MAGE-A3+)	NCT02111850

BCG, Bacillus Calmette-Guerin; PD-1, programed death 1; PD-L1, programed death-ligand 1.

Bladder cancer has long been known to be immune-responsive [21]. Intravesical instillation of BCG induces infiltration of cytotoxic T lymphocytes (CTLs) and mediates cell-mediated cytotoxicity against bladder tumors in patients with NMIBC. The emergence of BCG-refractory disease in a subset of patients suggests that BCG resistance may be mediated by a complex mechanism of immune escape [22^■]. An effective antitumor immune response involves a series of events: first, dying cancer cells release cancer antigens; second, dendritic cells (DCs) and antigen-presenting cells (APCs) present these antigens; third, APCs and T cells are primed and activated; fourth, CTLs traffic to and infiltrate tumors; and fifth, CTLs recognize and kill cancer cells. This process provides a framework for understanding the mechanisms of response and resistance to cancer therapy (Fig. 1). Divergence from any of these steps facilitates immune escape, whereas optimizing each of these steps provides new therapeutic opportunities. For instance, although intravesical chemotherapies work by direct cytotoxic effects and release of cancer-cell antigens, intravesical BCG works by causing T cells to infiltrate tumors. Although therapeutic cancer vaccines and anti-CTLA-4 antibodies work by priming, activating, and expanding T cells, immune checkpoint inhibitors such as anti-PD-1/PD-L1 antibodies restore effector T-cell function against cancer cells at the tumor site [23,24].

FIGURE 1.

Urothelial carcinoma-specific antitumor T-cell immunity (modified from Chen and Mellman [22^■]). 1. Released bladder tumor antigens. 2. Antigen processing and presentation. 3. Tumor antigen-specific T-cell priming, activation, and expansion. 4. Cytotoxic T-cell trafficking and infiltration to the tumor microenvironment. 5. T-cell recognition and immune-mediated tumor-cell killing.

Recent clinical trials have shown durable antitumor responses following blockade of the PD-1/PD-L1 pathway, particularly in patients with PD-L1-expressing tumors or immune subsets [25,26,27^■■-29^■■]. Ongoing clinical trials are investigating immune checkpoint blockade in multiple cancers, including bladder cancer. Here, we summarize results from recent clinical trials of immune therapies in bladder cancer and provide an overview of current challenges and future directions in immunotherapy for bladder cancer.

IMMUNE CHECKPOINT INHIBITORS

Tumors use immune checkpoint pathways, which regulate T-cell activation, to escape antitumor immunity. Many immune checkpoint molecules are involved in this mechanism, including CTLA-4, PD-1 and its ligands PD-L1 and PD-L2, T-cell immunoglobulin mucin-3, and lymphocyte activation gene-3 [24].

Cancer cells often produce immunosuppressive cytokines such as TGF-β, IL-10, indoleamine 2,3-dioxygenase, and vascular endothelial growth factor (VEGF), and recruit or expand immunosuppressive cells such as regulatory T cells (Tregs), tumor-associated macrophages, and myeloid-derived suppressor cells (MDSCs) [30]. Chemotherapies such as cyclophosphamide can deplete these immunosuppressive cells, whereas novel immunomodulators such as anti-PD-1 or anti-CTLA-4 antibodies may counteract immunosuppressive elements in the tumor microenvironment.

Inhibition of the programed death 1/programed death-ligand 1 pathway

PD-1 is an immune inhibitory receptor expressed on several immune-cell subsets, particularly cytotoxic T cells [31]. PD-1 interacts with two ligands, PD-L1 (B7-H1, CD274) and PD-L2 (B7-DC). PD-L1 is expressed on tumor and immune cells, including monocytes, macrophages, DCs, and T cells, whereas PD-L2 is expressed primarily on macrophages and DCs [32,33]. When these ligands bind to PD-1 receptors on T cells, a negative signal turns off T-cell activation and cytokine production. Although ligation of PD-1 with PD-L1 and PD-L2 during infection or inflammation in normal tissue is critically important in maintaining immune homeostasis, this interaction in the tumor microenvironment allows tumor cells to escape T-cell-mediated cytotoxicity.

Recent studies have demonstrated that upregulation of PD-L1 is a key mechanism of immune escape in early-stage bladder cancer [34-37]. Overexpression of PD-L1 in urothelial carcinomas has been shown to correlate with high-grade tumors and worse clinical outcome. Inman et al. [35] detected abundant expression of PD-L1 in the BCG-induced bladder granulomata of patients with BCG-refractory bladder cancer, as well as an association between PD-L1 expression and increased tumor-infiltrating mononuclear cells (odds ratio = 5.5, P = 0.004). Their data support the hypothesis that PD-L1 promotes disease progression of BCG-refractory bladder cancer by inactivating cytotoxic T cells.

PD-L1 expression on DCs and macrophages in the tumor microenvironment, in addition to expression of PD-L1 on tumor cells, may play an important role in immune escape. Mullane et al. [38] analyzed 160 bladder tumor samples and found PD-L1 expression on mononuclear cells (37%) and tumor cells (20%). Interestingly, PD-L1 expression on infiltrating mononuclear cells was significantly associated with longer OS in patients who developed metastatic disease (P = 0.04), but PD-L1 on tumor cells was not. IFN-γ, which is produced by activated T cells, can lead to upregulation of PD-L1 on both immune and tumor cells [33,39], suggesting that PD-L1 expression on immune cells may be upregulated by T-cell activation to maintain immune homeostasis. These findings suggest that T-cell antitumor immunity in bladder cancer may be limited by increased expression of PD-L1 on mononuclear cells.

A phase-I multicohort study (PCD4989G) evaluates MPDL3280A, a high-affinity, humanized monoclonal IgG1 antibody against PD-L1, in advanced solid tumors. Because PD-L1 is expressed on activated T cells, MPDL3280A was engineered with a modification in the fragment crystallizable (Fc) region that eliminates antibody-dependent cellular cytotoxicity to prevent depletion of T cells expressing PD-L1 [28^■■]. PCD4989G has multiple expansion cohorts, including one for metastatic urothelial bladder cancer. Powles et al. [29^■■] reported on the safety and antitumor activity in the first 68 patients enrolled with metastatic urothelial bladder cancer, and provided the first compelling evidence that the PD-L1 pathway is a valid bladder cancer target. Heavily pretreated patients tolerated MPDL3280A therapy very well and had an objective response rate (ORR) of 26%. ORR was even more remarkable (43%) among patients with PD-L1⁺ tumors. Even among patients whose tumors were PD-L1⁻, the response rate was 11% by RECIST v1.1. Median time to first response was 42 days (range, 38–85 days). Median duration of response was not yet reached at the time of data analysis. At data cutoff, 16 of 17 responses were ongoing. On the basis of these results, the FDA granted MPDL3280A breakthrough therapy designation for accelerated clinical development. A single-arm phase-II trial of MPDL3280A in patients with metastatic urothelial bladder cancer (NCT02108652) has recently finished enrollment, with a primary endpoint of ORR.

KEYNOTE-012 is a phase-Ib multicohort study of pembrolizumab, a humanized IgG4 high-affinity anti-PD-1 antibody engineered to remove antibody-dependent cellular cytotoxicity activity, in patients with PD-L1⁺ advanced solid tumors. Plimack et al. reported on safety and antitumor activity among the first 33 patients enrolled with metastatic urothelial bladder cancer. PD-L1 expression in stroma or at least 1% of tumor cells was required for study entry. Among 33 patients enrolled, 30 had urothelial carcinoma histology and three had nonurothelial carcinoma or mixed histology. Adverse events (≥1 drug-related), most commonly fatigue, peripheral edema, and nausea, were reported in 61% of patients; four (12%) reported grade 3–4 drug-related adverse events; only rash was seen in more than one patient. Of 29 evaluable patients, the ORR was 24.1%, with 10.3% complete responses; 14% of patients had stable disease. KEYNOTE-045, a randomized phase-III study of pembrolizumab vs. a taxane or vinflunine, is ongoing with primary endpoints of OS and progression-free survival.

Other agents that target the PD-1/PD-L1 pathway are being evaluated in patients with urothelial bladder cancer, including nivolumab (anti-PD-1 antibody), and MEDI-4736 and MSB0010718C (anti-PD-L1 antibodies). Results are pending.

The impressive antitumor activities that result from PD-1/PD-L1 blockade provide compelling evidence that immune checkpoint molecules are viable therapeutic targets for bladder cancer treatment. Although tumor regression in response to anti-PD-1/PD-L1 antibodies was rapid and durable in a fraction of patients, approximately 50% of patients did not achieve a response even if their tumor tested positive for the drug’s target, implying immune escape mediated by a mechanism separate from the PD-L1 pathway.

As mentioned above, tumor cells can evade immune attack by multiple mechanisms. The immune system can suppress tumor growth, but may also promote tumor progression either by selecting for tumor cells more fit to survive or by establishing conditions within the tumor microenvironment that facilitate tumor growth [40,41]. Thus, a therapy that targets aspects of the tumor microenvironment other than the PD-1/PD-L1 pathway may have therapeutic potential.

B7-H3

B7-H3 is an attractive emerging target for immunotherapy of bladder cancer. A member of the B7 family identified by Chapoval et al. [42] B7-H3 is overexpressed in a variety of solid tumors, including urothelial, prostate, and renal cell carcinomas. First described as a costimulatory molecule, it is now known to deliver both costimulatory and coinhibitory signals [43]. B7-H3 inhibits T-cell activation and cytokine production. In a murine model, B7-H3-deficient mice developed increased T helper type 1-mediated lung inflammation and autoimmune encephalomyelitis [44], suggesting B7-H3’s role as a negative regulator of T helper type 1 responses. Boorjian et al. [45] used immunohistochemistry to evaluate expression of multiple T-cell coinhibitory molecules (B7-H3, B7-H1, and PD-1) in urothelial carcinomas. On examination of 318 urothelial bladder cancer samples obtained postradical cystectomy, 222 of 314 (70.7%) stained positive for B7-H3. Expression of B7-H3 in urothelial bladder cancer was significantly increased compared with adjacent nontumor urothelium, as a median of 70% of tumor cells expressed B7-H3 compared with 20% of cells in nontumor specimens (P < 0.001). The increase in B7-H3 expression showed a trend toward tumor-stage independence (P = 0.13). Moreover, patients with prior intravesical BCG tended to show increased expression of B7-H3 (P = 0.023). A phase-I study (NCT01391143) of MGA271, a first-in-class humanized IgG1-κ monoclonal antibody that recognizes human B7-H3, is currently ongoing in multiple tumor types, including urothelial carcinoma.

Cytotoxic T-lymphocyte antigen-4

Ipilimumab, a therapeutic cancer vaccine, targets CTLA-4, a coinhibitory molecule that regulates T-cell priming and expansion during T-cell activation. Carthon et al. [46] conducted a pilot study of ipilimumab to assess the safety and immunologic effects of CTLA-4 blockade in localized bladder cancer. CTLA-4 blockade was associated with increased frequency of CD4⁺ICOS^hi T cells and increased perivascular infiltration of activated T cells in tumor tissue. Preliminary data from this small pilot study suggested a correlation with increased OS. A multi-institutional phase-II study (NCT01524991) evaluates the activity of first-line gemcitabine, cisplatin, and ipilimumab in metastatic bladder cancer, with a primary endpoint of 1-year OS.

AGONISTIC ANTIBODY THERAPIES TARGETING COSTIMULATORY MOLECULES

OX40 is a costimulatory receptor expressed by activated CD4⁺ and CD8⁺ T cells [47]. OX40 engagement by ligands present on DCs increases the proliferation, effector function, and survival of T cells. Preclinical and clinical studies have shown that OX40 agonists increase antitumor immunity [48,49]. Two clinical trials of agonistic OX40 antibodies in patients with advanced or metastatic solid tumors, including bladder cancer, are currently ongoing (NCT02221960 and NCT02219724).

CD40, a member of the tumor necrosis factor receptor superfamily, is expressed on DCs, B cells, monocytes, and MDSCs, as well as a range of tumors. Intratumoral injection of adenoviral vectors expressing CD40 ligand has yielded positive results in experimental and clinical bladder cancer [50]. In addition, agonistic CD40 monoclonal antibodies may generate anticancer immunity by activating APCs and promoting antitumor T-cell responses [51]. An agonistic anti-CD40 antibody has been studied in preclinical and clinical models [52-54]. Several in-vivo studies have shown that systemically administered agonistic CD40 antibodies can induce robust antitumor immune responses. However, serious side-effects have been reported, including cytokine release syndrome and liver function abnormalities [52]. Recent studies have investigated the potential of delivering a nontoxic dose of agonistic anti-CD40 antibody to the bladder tumor region and draining lymph nodes. Local administration of agonistic CD40 antibody may cure localized and distant experimental bladder cancer [54-56].

THERAPEUTIC CANCER VACCINES

A therapeutic cancer vaccine acts primarily by altering presentation of tumor antigens. It usually targets host DCs for effective antigen presentation and subsequent priming of cytotoxic T cells and helper T cells. Tumor-specific T effector cells, along with other innate immune cells, work to destroy cancer cells. Sipuleucel-T is the first therapeutic cancer vaccine approved by the FDA for use in humans [19]. Data from several prospective vaccine clinical trials indicate the positive prognostic value of immune responses [57,58]. It is believed that the patients most likely to derive clinical benefits from vaccine therapy are those with slow-growing disease and low tumor burden [59,60].

Several vaccine trials have been undertaken in patients with NMIBC and low tumor burden. One approach in these patients is to test the vaccine in combination with intravesical BCG because of the potential for synergistic immune stimulation. Another interesting approach would be to combine vaccine with an immune checkpoint inhibitor to evaluate the possible negative feedback mechanism of these immune-stimulating therapies. Potential long-term side-effects would be of concern in this patient population. Another ideal setting for testing vaccine therapies is the adjuvant setting after radical cystectomy, in which no standard therapy exists and the risk of recurrence is high. Several tumor antigens associated with bladder cancer have been investigated as vaccine targets or are currently in clinical development, including MUC-1, HER2, MAGE, and NY-ESO-1, among others. However, the fact that therapeutic cancer vaccines cannot affect preexisting immunosuppressive conditions in the tumor microenvironment presents an ongoing challenge.

PANVAC

PANVAC is a recombinant poxviral vaccine engineered to express the tumor antigens CEA and MUC-1, as well as three costimulatory molecules (B7.1, ICAM-1, and LFA-3). A phase-I study of PANVAC demonstrated promising activity in advanced solid tumors that expressed the vaccine target [61]. MUC-1 is commonly expressed in urothelial bladder cancers [62]. A randomized phase-II study (NCT02015104) investigates the efficacy of intravesical BCG plus subcutaneous PANVAC vs. BCG alone in BCG-refractory, high-risk NMIBC, with the primary endpoint of disease-free survival.

HS-410

HS-410 (vesigenurtacel-l) is an allogeneic cell-based vaccine modified to express a recombinant secretory heat shock protein (gp96) fused with a fragment crystallizable domain of an immunoglobulin. A phase-I/II trial (NCT02010203) studies the safety and tolerability of HS-410 in combination with BCG in patients with NMIBC who have failed previous BCG therapy.

HER2

In reported studies, HER2 overexpression in bladder cancer ranged from 9.2 [63] to 60.3% in high-grade invasive tumors [64]. A meta-analysis reported an average HER2 expression of about 40% in bladder cancers, and an association with poor prognosis [65].

DN24-02 is an investigational, autologous cell-based immunotherapy that targets HER2 and uses the same platform as sipuleucel-T. For both, the vaccine products are manufactured from patients’ own immune cells by culturing them with the recombinant antigen BA7072, which has components of the intracellular and extracellular domains of the HER2 antigen linked to GM-CSF. A randomized, placebo-controlled phase-II study of DN24-02 (NCTO1353222) in postradical cystectomy patients whose tumor samples expressed HER2 has completed enrollment; its primary endpoint is OS. Of 61 patients screened, 49 (80%) had a HER2 expression score of at least 1+, 27 (44%) had a score of at least 2+, and three (5%) had a 3+ score in the primary tumor. Of 23 patients who had HER2 expression levels evaluated in tumor from involved lymph nodes, 20 (87%) had a HER2 expression score of at least 1+, 12 (52%) had a score of at least 2+, and three (15%) had a 3+ score in the lymph nodes [66]. Survival data are pending.

Another vaccine targeting HER2, AdHER2/neu DC, is being studied in a phase-I trial (NCT01730118). AdHER2/neu DC is an investigational autologous DC-based vaccine transfected with an adenoviral vector expressing HER2 (AdHER2). Like DN24-02, AdHER2/neu DC uses ex-vivo maturation to produce the vaccine from the patient’s own immune cells. However, AdHER2/neu DC requires only a single leukapheresis to make all the required vaccine products, which are administered intradermally. The trial is designed for two cohorts: first, postcystectomy in the adjuvant setting in patients with MIBC, and second, metastatic bladder cancer in patients who progress after at least first-line systemic chemotherapy.

Melanoma antigen family A3

Melanoma antigen family A3 (MAGE-A3) is a cancer/testis antigen expressed in a variety of malignancies, including bladder cancer. A recent study showed that cancer stem cells express MAGE-A3 antigen, which may be an immunotherapeutic target in bladder cancer [67]. MAGNOLIA (NCT01435356) is a phase-II, randomized, placebo-controlled trial evaluating the safety and efficacy of adjuvant MAGE-A3 plus AS-15 antigen-specific cancer vaccine vs. placebo in patients with MAGE-A3⁺ MIBC after cystectomy.

NY-ESO-1

NY-ESO-1 is a cancer/testis antigen expressed in 30–45% of bladder cancers [68-70]. Sharma et al. [71] tested a recombinant NY-ESO-1 protein vaccine administered with GM-CSF and BCG as immunologic adjuvants in urothelial carcinoma patients. Six patients with NY-ESO-1-expressing MIBC were vaccinated after radical cystectomy. NY-ESO-1-specific antibody, CD8⁺ T-cell, and CD4⁺ T-cell responses were reported in five of six, one of six, and six of six patients, respectively.

DEC-205-NY-ESO-1 is a fusion protein vaccine composed of full-length NY-ESO-1 fused to the C terminus of the two human monoclonal antibodies against the human mannose receptor and DEC-205, a specific DC marker implicated in antigen cross-presentation [72]. This vaccine platform induces both CD8⁺ and CD4⁺ T-cell responses with broad antigen specificity. A phase-I trial (NCT01522820) of DEC-205-NY-ESO-1 with or without sirolimus is enrolling patients with NY-ESO-1⁺ tumors, including bladder cancer.

ADOPTIVE T-CELL TRANSFER THERAPY

Cytotoxic T cells are the primary agents mediating antitumor activity. Adoptive T-cell transfer therapy (ACT) induces tumor regression in multiple tumor types [73,74,75^■]. One promising strategy for expanding the range of ACT is to administer T cells that have been genetically engineered to express tumor-specific antigen receptors. These may be chimeric antigen receptors composed of an extracellular domain derived from a tumor-specific antibody linked with an intracellular domain of T-cell receptors and T-cell costimulatory receptors, which recognize cell-surface antigens in a nonmajor histocompatibility complex-restricted manner. They may also be traditional α/β T-cell receptors, which recognize epitopes of intracellular antigens presented by nonmajor histocompatibility complex molecules [75^■].

ACT for bladder cancer is in the early stages of development. A phase-II study (NCT01967823) of T-cell receptor immunotherapy targeting NY-ESO-1 in patients with NY-ESO-1⁺ cancer (including bladder cancer) is currently accruing. Although this approach has shown remarkable activity in other NY-ESO-1⁺ tumors such as synovial sarcoma and melanoma [76], its activity in bladder cancer has not been reported. Another ACT phase-I/II trial (NCT02111850) targeting MAGE-A3 in patients with metastatic cancer (including bladder cancer) who are HLA-DP0401⁺ is also enrolling patients.

COMBINATION STUDIES

Because tumors use a complex series of mechanisms to evade the host immune system, altering one pathway may not be enough to prevent immune escape. A preclinical study showed that dual immune checkpoint blockade of CTLA-4 and PD-1 resulted in more pronounced antitumor activity than either blockade alone [77]. On the basis of these results, Wolchok et al. [27^■■] evaluated concurrent and sequential administration of ipilimumab and nivolumab in patients with melanoma and found that concurrent administration resulted in an ORR of 40% (95% confidence interval: 27–55) by modified WHO criteria. These responses were ‘rapid and deep’. A survival analysis [78] showed 1-year and 2-year OS rates of 85 and 79% (95% confidence interval: not reached). In patients with metastatic renal cell carcinoma, the combination led to an ORR of 43-48% at two dosing schedules [79]. In both settings, responses were observed regardless of PD-L1 expression, which may suggest that adding ipilimumab may overcome negative PD-L1 expression status. An ongoing phase-I/II study of nivolumab with or without ipilimumab (NCT01928394) evaluates the safety and activity of this approach in several advanced solid tumors, including metastatic urothelial carcinoma.

Preliminary results of a phase-II study of cabozantinib, which targets multiple tyrosine kinases (primarily MET and VEGFR2), in heavily pretreated metastatic urothelial carcinoma patients showed single-agent activity in patients with soft tissue and bone metastasis [80]. We reported that cabozantinib decreases Tregs and CTLA-4 on Tregs in vitro and in patients with urothelial carcinoma [81]. In this study, increased Treg CTLA-4 expression, which plays a key role in Treg immunosuppressive function, after cabozantinib significantly correlated with worse progression-free survival and OS. A preclinical study indicated that cabozantinib not only altered the phenotype of MC38-CEA murine tumor cells, rendering them more sensitive to immune-mediated killing, but also altered the frequency of immune subpopulations in the periphery and the tumor microenvironment, generating a more permissive immune environment [82]. An ongoing phase-I study combining cabozantinib and nivolumab with or without ipilimumab (NCT02308943) includes comprehensive immune profiling using peripheral blood and tumor tissues. A recent study reported that VEGF-A enhances expression of immune checkpoint receptors (including PD-1) on CD8⁺ T cells in murine tumors [83], suggesting that combining antiangiogenic agents targeting VEGF-A/VEGFR and immune checkpoint blockade may have a synergistic antitumor effect. Several studies have evaluated the activity and safety of this combination. A study of nivolumab, combined with either sunitinib or pazapanib, showed a promising ORR of 52 and 45%, respectively, in patients with metastatic renal cell carcinoma [84].

Emerging evidence demonstrates that conventional therapies increase antitumor immunity [85,86]. For example, gemcitabine has been shown to increase expression of human leukocyte antigen (HLA) class I on cancer cells, enhance cross-presentation of tumor antigens to CD8⁺ T cells, and selectively kill MDSCs, indicating gemcitabine’s potential to facilitate T-cell-dependent anticancer immunity. A recent study suggested that cisplatin stimulates HLA-I expression and inhibits signal transducer and activator of transcription 6-regulated expression of PD-L2 on APCs and cancer cells [87]. Cisplatin has also been shown to increase the permeability of tumor cells to granzyme B, which suggests that cancer cells may be susceptible to CTL-mediated lysis even if they do not express HLA-I or tumor antigen recognized by CTLs [88].

Tregs play a pivotal role in maintaining self-tolerance and inhibiting antitumor immunity. We showed that TRC105, a human/murine chimeric IgG1-κ monoclonal antibody to CD105 (endoglin), significantly decreased Tregs on day 15 after TRC105 administration in metastatic castration-resistant prostate cancer (NCT01090765) [89]. This trend was also observed in a phase-II clinical trial of TRC105 (NCT01328574) in patients with advanced or metastatic urothelial carcinoma. Thus, the combination of immunotherapy and anticancer drugs that have immunomodulatory properties may potentially enhance antitumor activity.

CONCLUSION

Patients with BCG-resistant or BCG-refractory bladder cancer are at significant risk for morbidity and mortality. This implies that immune escape plays a major role in therapy resistance and disease progression. Improved understanding of the mechanisms of immune escape has led to an array of emerging therapeutic strategies for bladder cancer. The clinical activity of PD-1/PD-L1 pathway blockade provides compelling evidence that immune escape mechanisms are viable targets for bladder cancer treatment. Several other immune checkpoint inhibitors are in preclinical and clinical development. The combination of immunotherapy and conventional therapy that has an immunomodulatory function is a promising strategy for maximizing clinical activity. Careful selection of agents and rational clinical trial design are essential to the development of agents with a high potential for clinical efficacy.

KEY POINTS.

• Urothelial carcinoma is immunotherapy-responsive.

• Immunotherapy’s activity in only a fraction of patients, as well as the development of immunotherapy resistance, implies immune escape.

• The immune checkpoint molecule programmed death-ligand 1 appears to be a key pathway of immune escape and a valid therapeutic target for urothelial carcinoma.

• Clinical trials are investigating other immune checkpoint molecules, such as B7-H3, and other immune escape mechanisms.

• Given the complexity of immune escape, rational combinatorial regimens appear highly promising.