Purpose:

The CANON [CAVATAK in NON–muscle-invasive bladder cancer (NMIBC)] study evaluated a novel ICAM-1–targeted immunotherapeutic-coxsackievirus A21 as a novel oncolytic agent against bladder cancer.

Patients and Methods:

Fifteen patients enrolled in this “window of opportunity” phase I study, exposing primary bladder cancers to CAVATAK prior to surgery. The first 9 patients received intravesical administration of monotherapy CAVATAK; in the second stage, 6 patients received CAVATAK with a subtherapeutic dose of mitomycin C, known to enhance expression of ICAM-1 on bladder cancer cells. The primary endpoint was to determine patient safety and maximum tolerated dose (MTD). Secondary endpoints were evidence of viral replication, induction of inflammatory cytokines, antitumor activity, and viral-induced changes in resected tissue.

Results:

Clinical activity of CAVATAK was demonstrated by induction of tumor inflammation and hemorrhage following either single or multiple administrations of CAVATAK in multiple patients, and a complete resolution of tumor in 1 patient. Whether used alone or in combination with mitomycin C, CAVATAK caused marked inflammatory changes within NMIBC tissue biopsies by upregulating IFN-inducible genes, including both immune checkpoint inhibitory genes (PD-L1 and LAG3) and Th1-associated chemokines, as well as the induction of the innate activator RIG-I, compared with bladder cancer tissue from untreated patients. No significant toxicities were reported in any patient, from either virus or combination therapy.

Conclusions:

The acceptable safety profile of CAVATAK, proof of viral targeting, replication, and tumor cell death together with the virus-mediated increases in “immunological heat” within the tumor microenvironment all indicate that CAVATAK may be potentially considered as a novel therapeutic for NMIBC.

Translational Relevance

This report describes the first “window-of-opportunity” study of the tumor microenvironment of non–muscle-invasive bladder cancer (NMIBC) following exposure to a novel oncolytic virus, CVA21. Posttreatment tissue was directly visualized and available for study after resection. No clear differences were seen in tumors receiving CVA21 versus prior low-dose mitomycin-C, an agent used to increase expression of the CVA21 entry molecule, ICAM-1. Intravesical CVA21 therapy was extremely well-tolerated, was highly tumor-selective, and replicated in NMIBC causing inflammatory changes (immunological “heat”) in tumor but not normal bladder. Upregulation of IFN response genes and immune checkpoint genes (PD-L1 and LAG3) was evident, supporting combination studies with immune checkpoint inhibitor antibodies. NMIBC is regarded as a malignancy that responds well to immunotherapy. Oncolytic viral therapy, alone or in combination with immune checkpoint blockade, may offer an alternative to the 40-year-old standard-of-care, BCG therapy, but without its limiting toxicities.

Non–muscle-invasive bladder cancer (NMIBC) is a highly prevalent cancer with lifelong risk of occurrence. Transurethral resection (TUR) of all visible lesions is a standard treatment for NMIBC (1), but is accompanied with a high tumor recurrence rate ranging from 50% to 70% as well as a high tumor progression rate between 10% and 20% over a period of 2 to 5 years (2, 3). Thus, guidelines recommend intravesical chemotherapy and immunotherapy in the management of NMIBC to reduce these risks of recurrence and progression (4). Immunotherapy with Bacille Calmette–Guerin (BCG) decreases frequency and delays time to cancer recurrence and progression in patients with NMIBC (1, 5). Unfortunately, one third of NMIBC patients experience serious side effects of local and systemic BCG infection, and one third do not respond (6, 7). Treatment options for BCG-refractory patients are limited, and patients often undergo cystectomy. Combined with a potential worldwide shortage of BCG (8), there is an urgent need to develop novel therapies for this disease.

The success of BCG treatment for NMIBC stems, uniquely, from the BCG-induced immune response. Antigen-presenting cells in the urothelium can phagocytize BCG, followed by the presentation of antigen to BCG-specific CD4+ T cells. Proinflammatory cytokines are released, resulting in a predominant Th1-cell–induced immunity with an enhanced recognition of cancer cells through activated macrophages, CD8+ T cells, natural killer cells, and other effector cells (9, 10). Oncolytic viruses are emerging immunotherapeutic agents for a broad range of malignancies. As well as their ability to preferentially replicate in and lyse cancer cells directly, it is their induction of host immunity which is increasingly recognized to be the major component of their antitumor efficacy (11). Coxsackievirus A21 (CVA21), a naturally occurring common cold-producing enterovirus, is one such effective oncolytic agent against a range of solid tumors (12–15). CVA21 specifically targets and lytically infects susceptible cells expressing the CVA21 cellular receptors, intercellular adhesion molecule-1 (ICAM-1), and decay-accelerating factor (DAF; ref. 16). Our group recently demonstrated the susceptibility of bladder cancer cell lines to CVA21, the ability to enhance oncolysis by modulating expression of the viral receptor ICAM-1 by low doses of mitomycin C treatment, and the induction of immunogenic cell death in CVA21-treated cell lines capable of generating long-lasting protective antitumor immunity in the bladder mucosa (17). These results provided the rationale for a phase I/II clinical trial (CANON) to investigate the therapeutic potential of CVA21 as a new immunotherapy approach for the treatment of NMIBC.

This trial determined safety, feasibility, and biological effects of escalating intravesical doses of a novel bioselected formulation of CVA21 (CAVATAK) administered alone or in combination with mitomycin C in 15 first-line NMIBC patients prior to TURBT surgery.

Participants

Eligible patients were ≥18 years of age, with a clinical diagnosis of NMIBC based on cystoscopic appearance, and which was suitable for TUR. Eastern Cooperative Oncology Group performance status ≤ 2, ANC > 1,500/mm³; Hb >9.0 g/dL; Platelet >100,000/mm³, and serum creatinine ≤ 1.5 mg/dL were required. Patients with prior local or systemic treatments for NMIBC were excluded.

The study was conducted in accordance with UK-recognized ethical guidelines and received approval from National Research Ethics Committee (Ref 14/LO/1804). The patients provided written-informed consent for treatment and analysis of their biological samples. The trial is registered on ClinicalTrials.gov (identifier: NCT02316171). Supplementary Table S1 outlines the patients and treatment characteristics.

Study design and outcome measures

CANON was a phase I, two-part, open-label, dose-escalation study designed to evaluate the safety and clinical activity of intravesical CAVATAK (Coxsackievirus A21, CVA21) alone and in sequential combination with low-dose mitomycin C in 16 first-line patients with NMIBC who were candidates for and were planning to undergo TUR for treatment and staging of their disease (see Supplementary Materials and Methods for inclusion and exclusion criteria). This gave a relatively homogeneous study population and facilitated collection of resected tumor tissue for histologic, pharmacodynamics, and pharmacokinetic analyses.

The study consisted of two sequential parts. Part 1 (VLA012A) was a study of the safety and tolerability of CVA21 administered via intravesical instillation as a single agent in subjects with NMIBC scheduled to undergo TURBT for treatment and staging of their disease. Three cohorts consisting of 3 subjects each received CVA21 on day 1 in Cohorts A1 (1 × 108 TCID50) and A2 (3 × 108 TCID50) with subject in Cohorts A3 receiving to doses of CVA21 (3 × 108 TCID50) on days 1 and 2. As these doses were well tolerated, Part 2 (VLA012B) commenced. Part 2 (VLA012B) evaluated the safety and tolerability of CVA21 administered in sequential combination with low-dose mitomycin C in the same subject population. Two dose levels and two schedules of administration of CVA21 were evaluated with a fixed dose of mitomycin C (10 mg) in two cohorts of subjects. Mitomycin C was administered by intravesical instillation on day 1, and intravesical instillation of CVA21 (3 × 108 TCID50) occurred 4 hours after instillation of mitomycin C. In subjects scheduled to receive two instillations of CVA21 (3 × 108 TCID50), the instillations were at approximately the same time each day. Both parts of the study were open-label, with ascending doses and increased frequency of CVA21 dosing in a standard 3+3 design. After the MTD of the combination was established in VLA012B, up to 10 additional subjects were to be enrolled at the MTD to further explore the safety and PD of the combination. This part of the study was not executed due to the objectives of the study being met and the sponsor deciding to terminate the study. The MTD was defined in the protocol as the highest intravesical CVA21 dose at which no more than one subject in each 3-subject cohort experienced a dose-limiting toxicity (DLT). The period of observation for DLTs was from the time of first administration to 7 days later.

The highest dose of intravesical CVA21 used in this study (3 × 108 TCID50) on days 1 and 2 was well tolerated and may be used in future studies. The safety and tolerability of the treatment of CVA21 with and without low-dose mitomycin C were assessed through daily physical examinations, vital signs, and review of hematology, serum chemistry, and urinalysis data. Treatment-emergent adverse events were also reviewed (Supplementary Materials and Methods and Supplementary Table S2). The mitomycin C dose (Supplementary Materials and Methods) was based on in vitro studies where ICAM-1 expression was upregulated without cytopathic effects (CPE) on tumor cells. Cystoscopy photography was performed before and after treatment. Intravesical treatment was followed by TURBT surgery after 8 to 11 days which allowed tissues to be analyzed for virus replication, apoptosis, evidence of viral-induced changes in immune cell infiltrates, and immune checkpoint molecules. Serum and urine were collected on day 1 (before virus instillation), 3, 5, and 8 after virus treatment.

RNA extraction and viral RNA detection

Aliquots of urine were clarified by centrifugation (1 minute, 10,000 RPM) and viral RNA extracted from 140 μL of supernatant using Qiagen QIAmp Viral RNA Mini Kit (Qiagen). Control samples spiked with 1 × 103 and 1 × 106 copies were also extracted as internal controls. For detection of viral RNA, 5 μL extracted RNA was tested in triplicate with 20 μL of Qiagen's Quantifast pathogen RT-PCR + IC kit. Primer and Probe sequences are as follows: Forward primer (KKVP3fwd): 5′-GAGCTAAACCACCAACCAATCG-‘3’; Reverse primer (KKVP3rev): 5′-CGGTGCAACCATGGAACAA-‘3’; Probe (KKVP3): 6FAM-CACACACATCATCTGGGA-MGB. Samples that tested positive were further tested by TCID50 assay to confirm replication-competent virus.

TCID50 assay

Aliquots of urine samples were thawed, vortexed, and centrifuged (1 minute, 10,000 RPM) and supernatant serially diluted (DMEM, 2% FBS) and incubated on 70% to 90% confluent monolayers of SKMEL-28 cells, prepared in 96-well plates, 24 hours previously. Plates were incubated for 5 days before scoring wells positive/negative for CPE and calculating the viral titer using the Karber method.

Neutralizing anti-coxsackievirus antibody assay

SK-MEL-28 cells were plated 24 hours prior to the assay. Test serum was serially diluted in the range 1:4 to 1:32768, and 100 μL of each dilution was mixed with 100 μL of media containing 100 TCID50 CVA21 and incubated at 37°C for 1 hour. Following incubation, the sera were added to SK-MEL-28 cells and incubated for 72 hours with cytopathic effect (CPE) scored visually at the endpoint and 50% neutralizing titer calculated. Positive and negative control serum was also tested, and viral titer was confirmed by TCID50 with each assay.

Immunohistochemistry

Note that 4 μm sections were deparaffinized before antigen retrieval, and endogenous peroxidase blocking was performed. Slides were incubated with mouse anti-enterovirus antibody (DakoCytomation), rabbit anti-cleaved Caspase 3 (Cell Signaling Technology), rabbit anti-HMGB1 antibody (Abcam), and mouse anti–ICAM-1 antibody (Santa Cruz Biotechnology). Subsequently, slides were incubated with a horseradish peroxidase–labeled secondary antibody (DakoCytomation), followed by detection with diaminobenzidine solution and counterstained with hematoxylin (VectorLabs).

Total RNA extraction from formalin-fixed paraffin-embedded tumor tissues

Total RNA was isolated from paraffin-embedded tumor tissues using a Norgen FFPE RNA Purification kit (Norgen Biotek) as per the manufacturer's instructions. RNA concentration and purity were measured using an Agilent 2100 BioAnalyzer.

Human Cell Death PathwayFinder RT² Profiler PCR Array

cDNAs were reverse transcribed from RNA extracted from CAVATAK-treated bladder tumor tissues (see above) using the RT² First Strand Kit (SABiosciences). Comparison of the relative expression of 12 cell death–related genes between the samples was made with PCR Array: the cDNAs from those samples were characterized using a custom-designed Cell Death PathwayFinder RT² Profiler PCR Array (SABiosciences) and the RT2 SYBR Green/Rox PCR master mix (SABiosciences) on a Stratagene Mx3000P Thermal Cycler. ATP5B and GAPDH housekeeping genes were used for normalization (18), and data were analyzed with the ΔΔCt method.

Multispectral IHC analysis

Tissue sections were cut at 4 μm from formalin-fixed paraffin-embedded blocks. All sections were deparaffinized and subjected to heat-induced epitope retrieval in citrate buffer (pH 6.0; Biogenex). Note that 6-plex panel IHC was performed for each tissue slide using the following antibodies: anti-FoxP3 (clone 236A/E7, Abcam), anti–PD-L1 (clone E1L3N, Cell Signaling Technology), anti-CD8 (clone SP16, Spring Bioscience), anti-CD3 (clone SP7, Spring Bioscience), anti-CD163 (clone MRQ26, Ventana), and anti-Cytokeratin (clone AE1/AE3, DAKO). Antigen–antibody binding was visualized with TSA-Cy5, TSA-Cy3, TSA-FITC, TSA-Cy5.5, TSA-Coumarin (PerkinElmer), and TSA-Alexa594 (Life Technologies). Microwave treatment in citrate buffer (pH 6.0) was performed between antibody detection to prevent cross-reactivity. Tissue slides were counterstained with DAPI and coverslipped with VectaShield mounting media (Vector Labs). Control tissue samples were stained for each different marker. Hematoxylin and eosin staining was performed for each sample and reviewed by a pathologist to ensure the representativity of the tissue sample. For a detailed IHC protocol for multispectral analysis, see study by Feng and colleagues (19).

Microscopy and image analysis of multiplexed IHC: phenotype cell quantification in high-resolution images

Digital images were captured with PerkinElmer Vectra 2.0 platform following hot spot lymphocyte assessment: Tumor areas with the highest immune cell (CD3+CD8+) infiltrates were scanned at 20X and selected for analysis. Three images of 0.36 mm2 each per tissue sample were analyzed with InForm Software (PerkinElmer). The total number of cells were enumerated for the following phenotypes: PD-L1+ tumor cells, PD-L1+ other cells, CD3+PD-L1+, CD3+PD-L1- FoxP3+, CD3+CD8+PD-L1+, CD3+CD8+PD-L1-, CD163+PD-L1+, CD163+PD-L1- in the stroma and tumor compartment.

Detection of urinary HMGB1

Urine samples of patients and controls were tested using a commercially available HMGB1 ELISA kit (ST51011; IBL International GmbH). The assay was conducted according to the manufacturer's instructions. Urinary HMGB1 was expressed as HMGB1/Cr ratio (mg/μmolCr) to correct for differences in dilution.

Detection of urinary cytokine levels

Quansys Biosciences was contracted to test the urinary cytokine levels in 12 CANON patients, 15 untreated NMIBC patients, and 13 healthy controls using their Q-Plex Human Cytokine multiplexed ELISA array. Thawed samples were diluted with the appropriate Quansys sample dilution buffer at a ratio of (sample:buffer) 1:2 (50%). Polypropylene low-binding 96-well plates were used to prepare the samples and standards prior to loading the Q-Plex plate. Each dilution was measured in triplicate, a total of 3 wells per sample. The intensity of chemiluminescence from each array was measured using the Q-View chemiluminescent imager (Quansys Biosciences) and quantified using the Q-View software (Quansys Biosciences).

Nanostring

All RNA samples included in the study passed quality control requirements (as assessed by the RNA integrity number) of the platform. Digital multiplexed NanoString nCounter analysis system (NanoString Technologies)-based gene expression profiling was performed on 100 ng total RNA from each sample as input material according to the manufacturer's instructions. Nanostring RNA analysis of 700 immune-related genes was performed using the nCounter GX Human PanCancer Immune profiling Kit (XT) on the nCounter Analysis System. Analysis and normalization of the raw Nanostring data were performed using nSolver Analysis Software v1.1 (Nanostring Technologies).

Statistical analysis

Correlations were evaluated by the Pearson tests. All P values were calculated using two-tailed test. P values < 0.05 were considered statistically significant. Analyses were performed using GraphPad Prism.

Patients, study design, and toxicity of the trial

Fifteen patients were recruited into this study. All patients presented with cystoscopically visible bladder tumors positive for urine cytology and were scheduled to undergo a TUR of their bladder tumor (TURBT) as part of their standard clinical care. The patients' clinical characteristics are shown in Supplementary Table S1, and the design of the clinical trial (CANON) is illustrated in Fig. 1A and in Supplementary Materials and Methods. Intravesical administration of CAVATAK either as a single agent or in combination with mitomycin C was generally well tolerated with no grade 2 or higher product-related adverse events observed (Supplementary Table S2). Six patients developed urinary tract infections which were attributed to displacement of bacteria from the urethra into the bladder during catheterization. This was prevented in further patients by a 5-day course of oral amoxicillin commencing the day of first catheter insertion. The primary endpoint of this trial was to determine patient safety and MTD. Secondary objectives were assessment of viral replication, antitumor activity, and viral-induced changes in immune cell infiltrates.

Figure 1.

A and B, Tumor response: pre- and posttreatment cystoscopy. Cystoscopic photography was performed in all subjects to record the appearance of their tumor before and after intravesical CAVATAK treatment. Representative images pre- and post-CAVATAK treatment are shown for 2 patients (B001 and B008).

Figure 1.

A and B, Tumor response: pre- and posttreatment cystoscopy. Cystoscopic photography was performed in all subjects to record the appearance of their tumor before and after intravesical CAVATAK treatment. Representative images pre- and post-CAVATAK treatment are shown for 2 patients (B001 and B008).

Close modal

Tumor response: pre- and posttreatment cystoscopy

The clinical trial design of CANON (Fig. 1A), administering CAVATAK a week before the patients underwent a TURBT from which tissue was then available to study, gave a unique window of opportunity to study the effect of this oncolytic virus on bladder cancer. Response was defined as complete (disappearance of tumor, i.e., pathologic complete response) or macroscopic increase of hemorrhage compared with pretreatment. Due to the heterogeneous size, shape, and preexisting hemorrhage/necrosis, only observational comparison was possible. Cystoscopic photography was performed in all subjects to record the appearance of their tumor before and after intravesical CAVATAK treatment. As shown in Fig. 1B, 8 to 11 days following intravesical CAVATAK treatment surface hemorrhage and inflammation of tumors was observed in a number of the patients, and one complete response was confirmed by histology. Supplementary Fig. S1 shows the urine cytology slide from this complete responder (patient B008) confirming he did have bladder cancer prior to the intravesical CAVATAK treatment.

Viral-induced antitumor activity in TUR tissue

All patients underwent surgery 8 to 11 days after treatment with CAVATAK monotherapy or in combination with mitomycin C, which allowed us to evaluate specific tumor targeting of the virus in the resected tumors. IHC staining for enterovirus protein indicated viral infection in 12 of 14 tissues available for study with varying amounts of positive tumor staining observed (Fig. 2A). No viral staining was detected in surrounding stromal areas or in areas of tumor tissues displaying normal glandular change within urothelium (Fig. 2B) consistent with selective CAVATAK targeting to, and/or replication in, malignant cells. Viral protein positivity in the tumor appeared to be correlated with those tumors that displayed at least localized areas of ICAM-1 positivity, in keeping with our previous work showing the necessity of ICAM-1 for CVA21 infection of bladder cancer cells (17). One patient (B004) displayed extensive homogeneous ICAM-1 expression by the tumor cells, n = 5 displayed < 30%, n = 4 < 10% ICAM-1 tumor positivity, and n = 4 no ICAM-1 tumor positivity was detected. IHC analysis of the other CVA21 cellular receptor, DAF, showed strong homogeneous DAF expression by the tumor cells in all patient cases (Supplementary Fig. S2). Hematoxylin and eosin–stained sections of the viral-positive tissues showed areas of nonviable tumor and apoptotic bodies which was further confirmed by staining for cleaved caspase 3. Tissues that demonstrated minimal to no viral infection also displayed little to no cleaved caspase 3 positivity and were confirmed as viable tumor from the hematoxylin and eosin stains. The detection of significant areas of nonviable tumor based on hematoxylin and eosin staining as well as cleaved caspase 3 was not observed in untreated historic bladder cancer control tissues (data not shown) providing further evidence that the increased tumor cell apoptosis observed in the viral-positive tumors was due to CAVATAK-induced cell death.

Figure 2.

Intravesical CAVATAK selectively targets tumor cells within the bladder. A, IHC images showing representative cases displaying ICAM-1 positivity, enterovirus protein (VP-1), and cleaved caspase 3 (positive staining shown in brown). The hematoxylin and eosin images of the same cases show the extent of tumor necrosis/apoptotic bodies outlined by the broken black line. Magnification, x20. B, Left (Patient B006): Anti-enterovirus positivity by tumor cells compared with absence of anti-enterovirus staining on normal urothelium (indicated by broken line). Right (Patient B004): Anti-enterovirus staining (brown) only present on tumor cells and not on stromal cells and lymphocytic infiltrate.

Figure 2.

Intravesical CAVATAK selectively targets tumor cells within the bladder. A, IHC images showing representative cases displaying ICAM-1 positivity, enterovirus protein (VP-1), and cleaved caspase 3 (positive staining shown in brown). The hematoxylin and eosin images of the same cases show the extent of tumor necrosis/apoptotic bodies outlined by the broken black line. Magnification, x20. B, Left (Patient B006): Anti-enterovirus positivity by tumor cells compared with absence of anti-enterovirus staining on normal urothelium (indicated by broken line). Right (Patient B004): Anti-enterovirus staining (brown) only present on tumor cells and not on stromal cells and lymphocytic infiltrate.

Close modal

Previous preclinical work from our group showed that the mode of CAVATAK-induced cell death in bladder cancer cell lines was predominantly by apoptosis (17). To ascertain the cell death route taken by bladder cancer cells in response to CAVATAK infection in this human clinical setting, we analyzed the total RNA extracted from the CAVATAK-treated tumors using a custom Human Cell Death PathwayFinder RT2 Profiler PCR Array. Using this array, the expression of key genes important for the central mechanisms of cellular death, apoptosis, autophagy, and necrosis, was profiled, and their expression levels were compared with archival untreated bladder cancers. As shown in Table 1, genes encoding the intrinsic apoptotic cell death pathway (BCL2L1, BAK1, and caspase-9) were predominantly upregulated compared with untreated bladder cancer in the majority of patient tumors studied. Three of the tumors also displayed upregulation of one of the main players in programmed necrotic cell death, the serine/threonine kinase receptor-interacting protein 3 (RIPK3). In addition, seven of the tumors displayed a significant increase in the immunogenic cell death marker, Calreticulin.

Table 1.

CAVATAK induces the intrinsic apoptotic cell death pathway in bladder tumors

CAVATAK induces the intrinsic apoptotic cell death pathway in bladder tumors
CAVATAK induces the intrinsic apoptotic cell death pathway in bladder tumors

Increases in infectious virus in patient urine following intravesical CVA21 administration

To measure viral replication within the patients' tumors, level of viral shedding into the urine was assessed by analysis of viral copy number by RT-PCR, and retrieval of replication-competent virus was determined by adding urine to CVA21-sensitive SKMEL-28 cells in a tissue culture infectious dose (TCID50) assay. Between days 2 and 5 following the initial instillation of CAVATAK, all patients from both the monotherapy and combination cohorts showed an increase in viral levels with several patients (4, 5, 6, 7, 8, 11, and 15), demonstrating a second peak at later time points suggesting more than one cycle of viral replication within the bladder of these patients (Fig. 3).

Figure 3.

Increased levels of CAVATAK in patient urine indicate secondary viral replication within the bladder. Viral shedding was assessed in the urines of patients before and between days 2 and 5 after initial intravesical treatment with CAVATAK. The blue line represents the analysis of viral copy number by RT-PCR, and the red line depicts the levels of replication-competent virus as determined by a TCID50 assay. Indications of secondary viral replication were evident in patients 4, 5, 6, 7, 8, 11, and 15.

Figure 3.

Increased levels of CAVATAK in patient urine indicate secondary viral replication within the bladder. Viral shedding was assessed in the urines of patients before and between days 2 and 5 after initial intravesical treatment with CAVATAK. The blue line represents the analysis of viral copy number by RT-PCR, and the red line depicts the levels of replication-competent virus as determined by a TCID50 assay. Indications of secondary viral replication were evident in patients 4, 5, 6, 7, 8, 11, and 15.

Close modal

Urinary cytokine profile in CAVATAK-treated patients

We investigated the ability of CAVATAK to modulate the immune response in the NMIBC microenvironment as reflected by shed urinary cytokines in 12 CANON patients, and compared with 15 untreated NMIBC patients and 13 age-matched healthy controls. A 17-plex quantitative ELISA-based chemiluminescent assay was used to screen the urines of CANON trial patients before treatment at day 1 and then after treatment on days 3, 5, and 8 (with the exception of patient B005 where no d8 urine sample was obtained). Figure 4A shows the kinetics of the CAVATAK-induced cytokine response to 8 of the 17 cytokines studied in 6 representative patients. Although the cytokine levels did not vary significantly over the treatment period in the majority of patients, patient B004 was the exception displaying a peak in the level of virally induced cytokines (IL6, IL1a, IL1b, and TNFα) at day 3 after infection which then declined with a second increase from day 5. Notably, 4 of the virus-treated patients (B001, B005, B008, and B010), including the patient (B008) that demonstrated a complete response to CAVATAK, displayed increased levels of the proinflammatory cytokine IL23 following treatment which was consistently undetectable in the untreated NMIBC patients and healthy controls (data not shown).

Figure 4.

A, Urinary cytokine levels in patients following CAVATAK treatment. The Q-Plex Human Cytokine multiplexed ELISA array (Quansys Biosciences) was used to assess the level and kinetics of the CAVATAK-induced cytokine response in the urine of CAVATAK-treated patients on days 1, 3, 5, and 8 after viral treatment. Representative results are shown for 6 patients. B and C, CAVATAK-specific induction of HMGB1. B, Urinary HMGB1 levels were measured in the trial patients on day 1 (before CAVATAK treatment) and then on days 3, 5, and 8/9 after viral treatment using an HMGB1 ELISA kit (ST51011; IBL International GmbH). Urinary HMGB1 was expressed as HMGB1/Cr ratio (mg/μmolCr) to correct for differences in dilution. Six of the 11 patients tested showed a marked increase in HMGB1 levels on day 3, 48 hours after viral treatment. C, Screening of HMGB1 expression (brown staining) in untreated non–muscle-invasive bladder cancer (3 representative cases shown: control 1–3) and CAVATAK-treated bladder cancer (3 representative cases shown: B004, B003, and B007) by IHC. The untreated bladder cancer cases showed nuclear localization of HMGB1 consistent with that seen in the control tissue, normal human kidney. In contrast, the CAVATAK-treated bladder cancers expressed HMGB1 within the cytoplasm (magnification, x20; inset images, x40).

Figure 4.

A, Urinary cytokine levels in patients following CAVATAK treatment. The Q-Plex Human Cytokine multiplexed ELISA array (Quansys Biosciences) was used to assess the level and kinetics of the CAVATAK-induced cytokine response in the urine of CAVATAK-treated patients on days 1, 3, 5, and 8 after viral treatment. Representative results are shown for 6 patients. B and C, CAVATAK-specific induction of HMGB1. B, Urinary HMGB1 levels were measured in the trial patients on day 1 (before CAVATAK treatment) and then on days 3, 5, and 8/9 after viral treatment using an HMGB1 ELISA kit (ST51011; IBL International GmbH). Urinary HMGB1 was expressed as HMGB1/Cr ratio (mg/μmolCr) to correct for differences in dilution. Six of the 11 patients tested showed a marked increase in HMGB1 levels on day 3, 48 hours after viral treatment. C, Screening of HMGB1 expression (brown staining) in untreated non–muscle-invasive bladder cancer (3 representative cases shown: control 1–3) and CAVATAK-treated bladder cancer (3 representative cases shown: B004, B003, and B007) by IHC. The untreated bladder cancer cases showed nuclear localization of HMGB1 consistent with that seen in the control tissue, normal human kidney. In contrast, the CAVATAK-treated bladder cancers expressed HMGB1 within the cytoplasm (magnification, x20; inset images, x40).

Close modal

Increases in urinary HMGB1 levels in CAVATAK-treated patients

Having previously shown the release of HMGB1 by CVA21-treated bladder cancer cells in vitro and the importance of this potent cytokine for directing an immune response (17, 20), HMGB1 levels were assessed in the urine of the virus-treated patients. Figure 4B shows the increases in urinary levels of HMGB1 in selected patients (6 of 11 assessed) after CAVATAK treatment using creatinine-normalized values of HMGB1 (pg HMGB1/mg creatinine).

Interestingly, IHC analysis of this cytokine showed increased expression of cytoplasmic HMGB1 in CAVATAK-treated tumor tissues when compared with untreated bladder cancer controls (Fig. 4C). The untreated bladder cancer control tissues displayed a predominantly nuclear localization of HMGB1 consistent with that seen in the positive control tissue, normal kidney. This result suggests that CAVATAK infection may trigger the translocation of HMGB1 to the cell cytoplasm whence, upon cell death, it can be passively released to the extracellular environment, in keeping with the increased levels of urinary HMGB1 in virus-treated patients.

CAVATAK induces upregulation of PD-L1 in the NMIBC microenvironment

A multispectral IHC method which allowed the simultaneous detection of 7 markers was employed to investigate changes in the immune microenvironment of CAVATAK-treated NMIBCs compared with control-untreated archival cases of NMIBC and normal bladder tissue. Quantitation of the CD8+ T-cell infiltration revealed no significant differences between the virus-treated tumors and untreated bladder cancer or normal bladder tissue in either the stromal regions or intraepithelial regions (Supplementary Fig. S3). In addition, despite having shown in a previous preclinical study an enhanced uptake of the virus in mitomycin C–treated bladder cancer cell lines, this treatment did not result in an increased CD8+ T-cell infiltration in the patient tumors compared with control tumor tissues. However, notably, the patient (B008) who experienced a complete response to CAVATAK treatment showed a considerable immune infiltrate within the biopsy taken from the bladder area where the tumor had previously been identified (Supplementary Fig. S4). The 7-marker multispectral analysis also revealed an increase in PD-L1 within the stromal areas of the CAVATAK-treated tumors compared with the control bladder tissues, although this did not reach statistical significance (Supplementary Fig. S3).

CAVATAK induces upregulation of immune response genes

To further explore the immune response to CAVATAK, Nanostring Pan-Cancer Immune profiling was performed on RNA derived from 12 CAVATAK-treated and 7 untreated NMIBCs. CAVATAK led to elevated expression of the IFN-inducible genes IFIT1, IFIH1, OAS3, and MX1, compared with levels seen in untreated NMIBC control tissues. Furthermore, CAVATAK led to higher expression of the IFNγ-induced chemokine genes encoding CXCL9, CXCL10, and CXCL11 (Fig. 5A) associated with a Th1-mediated immune response.

Figure 5.

CAVATAK treatment upregulates IFN-induced genes and immune checkpoint molecules within the microenvironment of NMIBC tissue. Nanostring Pan-Cancer Immune profiling was performed on RNA derived from 12 CAVATAK-treated and 7 untreated NMIBCs. A, CAVATAK treatment induced elevated expression of the viral RNA response genes IFIT1, IFIH1, OAS3, and MX1, and IFNγ-induced chemokine genes encoding CXCL9, CXCL10, and CXCL11 compared with levels seen in untreated NMIBC control tissues. Of the immune checkpoint inhibitory and immunosuppressive genes, PD-L1, PD-L2, and LAG3 were upregulated in the CAVATAK-treated tumors compared with untreated NMIBC controls. B, CAVATAK treatment induced elevated expression of the RIG-I–like receptor dsRNA helicase enzyme that is encoded by the DDX58 gene. The levels of significance as determined by an unpaired t test which refer to the average expression of upregulated genes in the CAVATAK-treated patients compared with the average expression of untreated NMIBC control tissues are indicated: *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 5.

CAVATAK treatment upregulates IFN-induced genes and immune checkpoint molecules within the microenvironment of NMIBC tissue. Nanostring Pan-Cancer Immune profiling was performed on RNA derived from 12 CAVATAK-treated and 7 untreated NMIBCs. A, CAVATAK treatment induced elevated expression of the viral RNA response genes IFIT1, IFIH1, OAS3, and MX1, and IFNγ-induced chemokine genes encoding CXCL9, CXCL10, and CXCL11 compared with levels seen in untreated NMIBC control tissues. Of the immune checkpoint inhibitory and immunosuppressive genes, PD-L1, PD-L2, and LAG3 were upregulated in the CAVATAK-treated tumors compared with untreated NMIBC controls. B, CAVATAK treatment induced elevated expression of the RIG-I–like receptor dsRNA helicase enzyme that is encoded by the DDX58 gene. The levels of significance as determined by an unpaired t test which refer to the average expression of upregulated genes in the CAVATAK-treated patients compared with the average expression of untreated NMIBC control tissues are indicated: *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Close modal

Due to the known barriers that hamper oncolytic virus-induced antitumor immune responses, we were interested to explore the expression levels of immune checkpoint molecules and immunosuppressive enzymes within the CAVATAK-treated tumors compared with untreated archival NMIBCs. As shown by the histograms in Fig. 5A, in CAVATAK-treated tumors, there was an increase in the expression level of the immune checkpoint molecules PD-L1 and LAG3 as well as the amino acid–depleting enzyme, indoleamine 2,3-dioxygenase (IDO). These immunosuppressive mechanisms could be dampening T-cell responses and thus limiting the efficacy of CAVATAK, thus providing a rationale for future combination therapies using immune checkpoint inhibition or IDO-1 inhibition.

Another notable result from the Nanostring immune profiling was the significant upregulation in CAVATAK-treated tumors of the gene DDX58 which encodes the RIG-I–like receptor dsRNA helicase enzyme (Fig. 5B). This specific upregulation of DDX58 further confirms the infection by and sensing of CAVATAK within these tumors and is in keeping with the induction of a type I IFN response.

ICAM-1 expression is essential for productive CAVATAK infection

Previous work from our group has clearly shown the importance of ICAM-1 expression levels on bladder cancer cell lines in order to obtain a sufficient level of CAVATAK infection and subsequent oncolytic effect (17). This finding is further exemplified in this clinical study by one particular patient (B004) who, despite receiving the lowest dose of virus (single CAVATAK dose of 1 × 108), demonstrated significant viral infection and viral-induced changes in their bladder tumor. Although all the patients' tumors expressed DAF, many of the patients' tumors only showed focal regions of ICAM-1 tumor positivity. Notably, patient B004's resected CAVATAK-treated tumor displayed a high level of ICAM-1 expression throughout the tumor, enabling a high level of virus infection and apoptotic tumor cell death (Supplementary Fig. S5B). Furthermore, the virus-treated tumor displayed cytoplasmic HMGB1 expression, a high level of urinary cytokines (Fig. 4A), immune infiltration, increased levels of stromal PD-L1, and a high level of perforin expression indicating a clear immune response to this productive infection (Supplementary Fig. S5B). Perforin-positive cells were also detected in four other CANON patients (B006, B007, B008, and B010), but not at the same high prevalence (Supplementary Fig. S5A). Such perforin-positive cells were not detected in untreated bladder cancer (n = 10 tissues studied) and normal bladder tissues (n = 10 tissues studied).

NMIBC remains a highly prevalent and significant health problem worldwide, requiring intrusive surveillance and high health economic costs. It is the prototypical example of successful immunotherapy with high primary response rates to BCG likely involving a combination of innate and adaptive immune responses initially to BCG, followed by tumor-specific T-cell responses (9). The unrestricted BCG-induced vigorous inflammatory response in the entire bladder mucosa is highly problematic and often results in incomplete treatment programs.

Oncolytic viruses are emerging as highly potent cancer immunotherapeutics, which in many ways have similar properties to BCG (21). They target cancers through direct cytotoxicity and also induce cancer-specific adaptive immunity. Their ability to convert “cold” tumors into “hot” tumors through inflammation makes them potential partners for combination with immune checkpoint inhibitors (22), and there are currently a large number of clinical trials ongoing with this combination. This includes delivery of CVA21 into metastatic melanoma deposits alone (CALM, ClinicalTrials.gov Identifier: NCT01227551) or combined with anti–CTLA-4 antibody (MITCI ClinicalTrials.gov Identifier: NCT02307149) and anti–PD-1 (CAPRA, ClinicalTrials.gov Identifier: NCT02565992). These studies confirmed that CVA21 exposure causes a marked immune inflammatory and very durable local and abscopal antitumor response. NMIBC is an ideal model for oncolytic immunotherapy with easy and direct access to bladder cancer, their visualization, delivery of high doses of virus, no risk of neutralizing antibody limitation, complete control of length of exposure through catheterization, and sequencing of repeat doses and/or combination treatments based on existing schedules used for BCG therapy.

Our preclinical study had indicated ICAM-1–targeted cytotoxicity after CVA21 exposure and the induction of immunogenic cell death in bladder cancer cell lines, enhanced by exposure to low, noncytotoxic doses of mitomycin-C through induction of ICAM-1 expression (17). DAF expression was not dysregulated across numerous cell lines and was not predictive of CAVATAK infection. In the current neoadjuvant clinical study, 15 patients received intravesical CVA21 ± low-dose mitomycin C prior to tumor resection. The purpose of the study was to determine safety and tolerability for this approach, but also biological endpoints as a “window of opportunity” approach including tumor response, changes in the immune environment, and evidence of virus replication. Importantly, the study was not attempting to compare efficacy or immunogenicity with BCG, but aimed to provide a clear picture of immunomodulatory effects of CAVATAK in treatment-naïve tissue. All patients had detailed prior imaging to exclude locally advanced and metastatic disease. All patients had positive urine cytology, but evidence of muscle invasion would have only been possible after TURBT: this was not evident in any patient post-TURBT, and despite the cohort all being urine cytology positive, not all were high-grade tumors. The treatment was extremely well-tolerated with minor toxicities reported. Objectively, clinical photography indicated virus-induced hemorrhagic and inflammatory changes in 3 patients, but notably there was no evidence of a general inflammation of the normal bladder, in contrast to what would be expected with BCG therapy. The patient with complete response showed clear evidence of acute inflammation within the preexisting area of tumor within the bladder providing further evidence for the role of the immune response in the therapeutic outcome of this viral treatment.

We found evidence of secondary viral replication in most patients in urine samples between 2 and 5 days after CVA21 treatment by both viral copy number by RT-PCR and TCID50, representing replication-competent virus production. IHC staining for enteroviral protein was shown to be specifically associated with cancer cells and not normal urothelium in all but 2 patients (in whom the tissue was nonevaluable due to damage at resection).

After intravesical delivery of CVA21, no virus was detected in serum in any patient, nor was there evidence of induction of a systemic neutralizing antibody response in either cohort (Supplementary Table S3). After systemic delivery of CVA21 in an ongoing study (STORM, https://clinicaltrials.gov/ct2/show/NCT02043665), approximately 80% of patients demonstrated a neutralizing antibody response at day 8 after treatment (personal communication with Viralytics Ltd.).

We previously demonstrated the importance of ICAM-1 expression for CVA21 uptake in bladder cancer cell lines (17). In this study, ICAM-1 expression was mostly focal, in 8 of the 14 patient tumors studied. This proportion was lower than expected and may be explained by the paucity and quality of tissue after TURBT, and heat artifacts to adjacent tissue. Indeed, the importance of ICAM-1 expression for a productive CVA21 infection was clearly demonstrated by one particular patient (B004) whose tumor displayed high, widespread ICAM-1 expression which was associated with high levels of virus protein expression, apoptosis, HMGB-1 in urine, urinary cytokines, and immune infiltration with PD-L1 induction.

Virus infection was associated with apoptosis, determined by cleaved caspase 3 expression in patient tumor tissues. In preclinical studies, we found the mode of cell death in bladder cancer lines was through the intrinsic apoptosis pathway; in patient tissue, we found this was also the case in the majority of cases with additional evidence of necroptosis in 3 patients. Using bladder cancer cell lines, we had previously determined that CVA21 induces a type of cancer cell death that is immunogenic (17, 23). In this clinical study, 6 of 11 patients, despite wide intra-and interpatient variations in urine pH, specific gravity, and metabolite content, showed increased levels of one of the hallmarks of immunogenic cell death, high mobility group box 1 (HMGB1) in the urine post-CVA21 treatment. Furthermore, IHC analysis of HMGB1 expression in the tumor tissues of the study patients revealed a greater degree of cytoplasmic expression. Unfortunately, we were unable to correlate the cytoplasmic HMGB1 expression revealed in tissue with peaks in urinary HMGB1 as we excluded any patient urine samples that displayed hematuria to prevent false-positive results. HMGB-1 is one of the most abundantly secreted DAMPs following oncolytic virus infection (24–28) and is thought to aid the therapeutic efficacy of oncolytic viral therapy by acting as a potent immunostimulatory molecule and chemoattractant for monocytic cellular infiltration during virus infection (29).

In addition to levels of HMGB1 in the urine, the levels of inflammatory cytokines were also evaluated to see whether they reflected local immunomodulation caused by CVA21 infection. Such an analysis of urinary cytokine profiles has previously been used to determine responders and nonresponders to BCG therapy (30). Looking at the kinetics of the cytokine responses in individual patients, there were increases in classical virally induced cytokines (31) on day 3 after infection, often with a second peak at day 8. Interestingly, one cytokine that did differ between bladder cancer controls and CVA21-treated patients was the levels of IL23 which was consistently undetectable in the untreated NMIBC patients, but was elevated in four of the virus-treated patients, including the patient that demonstrated a complete response to CAVATAK. IL23 has been shown to have significant antitumor effects in various models of cancer and is associated with the promotion of cell-mediated immune responses and activation of cytotoxic T lymphocytes (CTL) or natural killer (NK) cells (32–35).

One of the key aims of the study was to induce immunological “heat” into the NMIBC tumor microenvironment we have previously observed as relatively “cold.” Multispectral IHC analysis did not show significant differences in the quantitation of CD8+ T cells in the stromal or intraepithelial regions between the CAVATAK-treated and untreated bladder tumors. Although CAVATAK treatment did not appear to result in increased TILs, further analysis revealed that the CD8+ TILs that were present in the virus-treated tumors displayed a more activated phenotype based on perforin expression, and several cases revealed higher PD-L1 expression in the stroma. Interestingly, although NK cell activation has been shown to be a prominent feature of the immune response to BCG (36, 37), CD56 IHC staining of these virus-treated tumors did not show any evidence for significant NK cell involvement in the immune response to CAVATAK in the bladder tumor microenvironment.

An even more comprehensive analysis of the expression of immune response genes in patient tissue compared with tissue from untreated, surgically resected NMIBC of similar tumor stage and grade was performed using Nanostring analysis. Untreated NMIBC had significant expression of genes associated with local immunosuppression such as IDO and ARG-2. Data have already indicated that such IDO gene expression is a feature of more aggressive NMIBC, emphasizing the potential immunosuppressive role of IDO (36). As expected, as a result of virus infection/replication, the induction of IFN-inducible genes (IFIT1, IFIH1, MX1) and IFNg-induced chemokine genes was observed. In keeping with CVA21-associated induction of IFN, we also found marked increases in expression of immune checkpoint molecules PD-L1 and LAG-3 as well as the amino acid–depleting enzyme, IDO presumably triggered to dampen down the immune response to CVA21. The pattern of response is very similar to changes observed after intratumoral injection of metastatic malignant melanoma in the studies mentioned earlier (CALM and CAPRA. We demonstrated activation of innate immunity by CVA21-infected tumors by the induction of RIG-I gene (retinoic acid–inducible gene I) expression. RIG- is a RIG-I–like receptor dsRNA helicase enzyme that is encoded by the DDX58 gene. RIG-I is part of the RIG-I–like receptor family, which also includes MDA5 and LGP2, and functions as a pattern recognition receptor which is a sensor for a number of viruses (38).

Although this study has obvious limitations in terms of size, some heterogeneity of patient tumor size, and distribution, it does provide valuable insights into the potential efficacy and utility of oncolytic immunotherapy for this condition. The study ended at the point of TURBT resection so although no safety concerns arose from any patient at any time after the study, formal follow-up tumor evaluations were not possible within the scope of the study. The scheduling of mitomycin-C/CAVATAK was based on in vitro demonstration of ICAM-1 upregulation following mitomycin C (within 24 hours), but may have been suboptimal in the trial context. The circa 11 days was judged by the ethics committee as the limit of an acceptable temporary delay of the TURBT procedure. The translational aspects were limited by tissue quality following surgery, and measuring urinary markers such as cytokines and HMBG-1 is challenging. We were unable to take distant random bladder biopsies for comparison at surgery due to the risk of tumor seeding. Although Nanostring allowed evaluation of all of the available tumors, multispectral analysis is limited by its selectivity and focus on very small areas within the whole tumor sample. The role and value of mitomycin C for ICAM-1 modulation were still unclear.

Previously, both replicating and nonreplicating viruses have been evaluated in preclinical models of bladder cancer (39–43). In human trials, intravesical vaccinia (Dryvax) resulted in immune infiltration of both malignant and normal tissue (44), and an oncolytic adenovirus CG0070 expressing granulocyte-macrophage colony-stimulating factor is being evaluated in a randomized study phase II/III study after demonstrating objectives responses of 48% to 63% in different dosing regimens (45, and www.clinicaltrials.gov, NCT01438112). These studies all indicate the approach is feasible but have focused mainly on recurrent disease after BCG failure, where the natural history of the cancer has been altered. None of the studies has provided detailed mechanistic evaluation of virus protein expression, viral replication kinetics, and the effect on the immune microenvironment.

The lack of significant toxicity in any patient gives us great scope to design more prolonged dosing schedules for future studies. The study suggested no useful role for low-dose mitomycin C in the context of potential ICAM-1 upregulation. In keeping with evolving data from clinical trials of CVA21 in malignant melanoma, the virus-induced immune checkpoint molecules, presumably as a result of IFN induction, would suggest obvious combination therapy with a checkpoint inhibitor (46). These agents are already under evaluation, as single treatment, in patients with NMIBC failing BCG therapy who are at high risk of relapse. Therefore, the next stage of clinical evaluation of CVA21 in NMIBC would logically move to combination therapy, sequencing the checkpoint inhibitor after CVA21 therapy to provide a potentially alternative effective treatment for this disease to BCG.

K.J. Harrington reports receiving commercial research grants from AstraZeneca, Boehringer-Ingelheim, MSD, and Replimune; reports receiving speakers bureau honoraria from AstraZeneca, Bristol-Myers Squibb, Boehringer-Ingelheim, Merck-Serono, MSD, Pfizer, and Replimune; and is a consultant/advisory board member for AstraZeneca, Bristol-Myers Squibb, Boehringer-Ingelheim, Merck-Serono, MSD, Pfizer, and Replimune. B. Davies holds ownership interest (including patents) in Viralytics. M. Grose is an employee of Viralytics. B. Fox is an employee of UbiVac, Argos, Bayer, Bristol-Myers Squibb, CellDex Therapeutics, Definiens, Janssen/Johnson & Johnson, Macrogenics, AstraZeneca, Akoya/PerkinElmer, and PrimeVax; reports receiving commercial research grants from Viralytics/Merck, Bristol-Myers Squibb, Bayer, Definiens, Janssen/Johnson & Johnson, Macrogenics, NanString, OncoSec, Akoya/PerkinElmer, and Shimadzu; and holds ownership interest (including patents) in UbiVac and PrimeVax. D. Shafren reports receiving commercial research grants from, holds ownership interest (including patents) in, and is a consultant/advisory board member for Viralytics. No potential conflicts of interest were disclosed by the other authors.

Conception and design: G. Au, B. Fox, R. Vile, H.S. Pandha

Development of methodology: N.E. Annels, B. Davies, B. Fox, R. Vile

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): D. Mansfield, M. Arif, S.S. Sandhu, K.J. Harrington, B. Davies, M. Grose, B. Fox, R. Vile, H. Mostafid, H.S. Pandha

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): N.E. Annels, D. Mansfield, C. Ballesteros-Merino, G.R. Simpson, M. Denyer, A.A. Melcher, K.J. Harrington, I. Bagwan, B. Fox, R. Vile, H.S. Pandha

Writing, review, and/or revision of the manuscript: N.E. Annels, D. Mansfield, C. Ballesteros-Merino, G.R. Simpson, A.A. Melcher, K.J. Harrington, G. Au, M. Grose, I. Bagwan, B. Fox, R. Vile, H. Mostafid, D. Shafren, H.S. Pandha

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): N.E. Annels, M. Denyer, G. Au, M. Grose

Study supervision: N.E. Annels, M. Grose, H.S. Pandha

Others (perform, analysis, and report the IHC part): C. Ballesteros-Merino

Viralytics Ltd. funded this study.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Babjuk
M
,
Burger
M
,
Zigeuner
R
,
Shariat
SF
,
van Rhijn
BW
,
Comperat
E
, et al
EAU guidelines on non-muscle-invasive urothelial carcinoma of the bladder: update 2013
.
Eur Assoc U Eur Urol
2013
;
64
:
639
53
.
2.
Chen
SY
,
Du
LD
,
Zhang
YH
. 
Pilot study of intravesical instillation of two new generation anthracycline antibiotics in prevention of superficial bladder cancer recurrence
.
Chin Med J (Engl)
2010
;
123
:
3422
6
.
3.
Addeo
R
,
Caraglia
M
,
Bellini
S
,
Abbruzzese
A
,
Vincenzi
B
,
Montella
L
, et al
Randomized phase III trial on gemcitabine versus mytomicin in recurrent superficial bladder cancer: evaluation of efficacy and tolerance
.
J Clin Oncol
2010
;
28
:
543
8
.
4.
Babjuk
M
,
Oosterlinck
W
,
Sylvester
R
,
Kaasinen
E
,
Böhle
A
,
Palou-Redorta
J
, et al
European Association of Urology (EAU). EAU guidelines on non-muscle-invasive urothelial carcinoma of the bladder, the 2011 update
.
Eur Urol
2011
;
59
:
997
1008
.
5.
Braasch
MR
,
Bohle
A
,
O'Donnell
MA
. 
Risk-adapted use of intravesical immunotherapy
.
BJU Int
2008
;
102
:
1254
64
.
6.
Bassi
P
. 
BCG (Bacillus of Calmette Guerin) therapy of high-risk superficial bladder cancer
.
Surg Oncol
2002
;
11
:
77
83
.
7.
Fuge
O
,
Vasdev
N
,
Allchorne
P
,
Green
JSA
. 
Immunotherapy for bladder cancer
.
Res Rep Urol
2015
;
7
:
65
79
.
8.
Mostafid
AH
,
Palou Redorta
J
,
Sylvester
R
,
Witjes
JA
. 
Therapeutic options in high-risk non-muscle-invasive bladder cancer during the current worldwide shortage of bacille Calmette-Guérin
.
Eur Urol
2015
;
67
:
359
60
.
9.
Kitamura
H
,
Tsukamoto
T
. 
Immunotherapy for urothelial carcinoma. Current status and perspectives
.
Cancers
2011
;
3
:
3055
71
.
10.
Abebe
F
. 
Is interferon-gamma the right marker for Bacille Calmette-Guérin-induced immune protection? The missing link in our understanding of tuberculosis immunology
.
Clin Exp Immunol
2012
;
169
:
213
9
.
11.
Melcher
A
,
Parato
K
,
Rooney
CM
,
Bell
JC
. 
Thunder and lightning: immunotherapy and oncolytic viruses collide
.
Mol Ther
2011
;
19
:
1008
16
.
12.
Shafren
DR
,
Au
GG
,
Nguyen
T
,
Newcombe
NG
,
Haley
ES
,
Beagley
L
, et al
Systemic therapy of malignant human melanoma tumors by a common cold-producing enterovirus, Coxsackievirus A21
.
Clin Cancer Res
2004
;
10
:
53
60
.
13.
Au
GG
,
Lindberg
AM
,
Barry
RD
,
Shafren
DR
. 
Oncolysis of vascular malignant human melanoma tumors by Coxsackievirus A21
.
Int J Oncol
2005
;
26
:
1471
6
.
14.
Berry
LJ
,
Au
GG
,
Barry
RD
. 
Potent oncolytic activity of human enteroviruses against human prostate cancer
.
Prostate
2008
;
68
:
577
87
.
15.
Skelding
KA
,
Barry
RD
,
Shafren
DR
. 
Systemic targeting of metastatic human breast tumor xenografts by Coxsackievirus A21
.
Breast Cancer Res Treat
2009
;
113
:
21
30
.
16.
Shafren
DR
,
Dorahy
DJ
,
Ingham
RA
,
Burns
GF
,
Barry
RD
. 
Coxsackievirus A21 binds to decay-accelerating factor but requires intercellular adhesion molecule 1 for cell entry
.
J Virol
1997
;
71
:
4736
43
.
17.
Annels
NE
,
Arif
M
,
Simpson
GR
,
Denyer
M
,
Moller-Levet
C
,
Mansfield
D
, et al
Oncolytic immunotherapy for bladder cancer using Coxsackie A21 virus
.
Mol Ther Oncolytics
2018
;
9
:
1
12
.
18.
Andersen
CL
,
Jensen
JL
,
Ørntoft
TF
. 
Normalization of real-time quantitative reverse transcription-PCR data: a model-based variance estimation approach to identify genes suited for normalization, applied to bladder and colon cancer data sets
.
Cancer Res
2004
;
64
:
5245
50
.
19.
Feng
Z
,
Puri
S
,
Moudgil
T
,
Wood
W
,
Hoyt
CC
,
Wang
C
, et al
Multispectral imaging of formalin-fixed tissue predicts ability to generate tumor-infiltrating lymphocytes from melanoma
.
J ImmunoTher Cancer
2015
;
3
:
47
.
20.
Bianchi
ME
,
Crippa
MP
,
Manfredi
AA
,
Mezzapelle
R
,
Rovere Querini
P
,
Venereau
E
. 
High mobility group box 1 protein orchestrates responses to tissue damage via inflammation, innate and adaptive immunity, and tissue repair
.
Immunol Rev
2017
;
280
:
74
82
.
21.
Grigg
C
,
Blake
Z
,
Gartrell
R
,
Sacher
A
,
Taback
B
,
Saenger
Y
. 
Talimogene laherparepvec (T-Vec) for the treatment of melanoma and other cancers
.
Semin Oncol
2016
;
43
:
638
46
.
22.
Meyers
DE
,
Wang
AA
,
Thirukkumaran
CM
,
Morris
DG
. 
Current immunotherapeutic strategies to enhance oncolytic virotherapy
.
Front Oncol
2017
;
7
:
114
.
23.
Kroemer
G
,
Galluzzi
L
,
Kepp
O
,
Zitvogel
L
. 
Immunogenic cell death in cancer therapy
.
Annu Rev Immunol
2013
;
31
:
51
72
.
24.
Wang
H
,
Ward
MF
,
Fan
XG
,
Sama
AE
,
Li
W
. 
Potential role of high mobility group box 1 in viral infectious diseases
.
Viral Immunol
2006
;
19
:
3
9
.
25.
Donnelly
OG
,
Errington-Mais
F
,
Steele
L
,
Hadac
E
,
Jennings
V
,
Scott
K
, et al
Measles virus causes immunogenic cell death in human melanoma
.
Gene Ther
2013
;
20
:
71
5
.
26.
Borde
C
,
Barnay-Verdier
S
,
Gaillard
C
,
Hocini
H
,
Marechal
V
,
Gozlan
J
. 
Stepwise release of biologically active HMGB1 during Hsv-2 infection
.
PLoS ONE
2011
;
6
:
e16145
.
27.
Huang
B
,
Sikorski
R
,
Kirn
DH
,
Thorne
SH
. 
Synergistic anti-tumor effects between oncolytic vaccinia virus and paclitaxel are mediated by the IFN response and HMGB1
.
Gene Ther
2011
;
18
:
164
72
.
28.
Guo
ZS
,
Naik
A
,
O'Malley
ME
,
Popovic
P
,
Demarco
R
,
Hu
Y
, et al
The enhanced tumor selectivity of an oncolytic vaccinia lacking the host range and antiapoptosis genes SPI-1 and SPI-2
.
Cancer Res
2005
;
65
:
9991
8
.
29.
Hosakote
YM
,
Brasier
AR
,
Casola
A
,
Garofalo
RP
,
Kurosky
A
. 
Respiratory syncytial virus infection triggers epithelial HMGB1 release as a damage-associated molecular pattern promoting a monocytic inflammatory response
.
J Virol
2016
;
90
:
9618
31
.
30.
Saint
F
,
Patard
JJ
,
Maille
P
,
Soyeux
P
,
Hoznek
A
,
Salomon
L
, et al
T helper 1/2 lymphocyte urinary cytokine profiles in responding and nonresponding patients after 1 and 2 courses of Bacillus Calmette-Guerin for superficial bladder cancer
.
J Urol
2001
;
166
:
2142
7
.
31.
Morgensen
TH
,
Paludan
SR
. 
Molecular pathways in virus-induced cytokine production
.
Microbiol Mol Biol Rev
2001
;
65
:
131
50
.
32.
Kaiga
T
,
Sato
M
,
Kaneda
H
,
Iwakura
Y
,
Takayama
T
,
Tahara
H
. 
Systemic administration of IL-23 induces potent antitumor immunity primarily mediated through Th1-type response in association with the endogenously expressed IL-12
.
J Immunol
2007
;
178
:
7571
80
.
33.
Reay
J
,
Kim
SH
,
Lockhart
E
,
Kolls
J
,
Robbins
PD
. 
Adenoviral-mediated, intratumor gene transfer of interleukin 23 induces a therapeutic antitumor response
.
Cancer Gene Ther
2009
;
16
:
776
85
.
34.
Kuramoto
T
,
Fujii
R
,
Nagai
H
,
Belladonna
ML
,
Yoshimoto
T
,
Kohjimoto
Y
, et al
IL-23 gene therapy for mouse bladder tumour cell lines
.
BJU Int
2011
;
108
:
914
21
.
35.
Choi
IK
,
Li
Y
,
Oh
E
,
Kim
J
,
Yun
CO
. 
Oncolytic adenovirus expressing IL-23 and p35 elicits IFN-γ- and TNF-α-co-producing T cell-mediated antitumor immunity
.
Plos One
2013
;
8
:
e67512
.
36.
García-Cuesta
EM
,
Esteso
G
,
Ashiru
O
,
López-Cobo
S
,
Álvarez-Maestro
M
,
Linares
A
, et al
Characterization of a human anti-tumoral NK cell population expanded after BCG treatment of leukocytes
.
Oncoimmunology
2017
;
6
:
e1293212
.
37.
Brandau
S
,
Riemensberger
J
,
Jacobsen
M
,
Kemp
D
,
Zhao
W
,
Zhao
X
, et al
NK cells are essential for effective BCG immunotherapy
.
Int J Cancer
2001
;
92
:
697
702
.
38.
Patel
JR
,
García-Sastre
A
. 
Activation and regulation of pathogen sensor RIG-I
.
Cytokine Growth Factor Rev
2014
;
25
:
513
23
.
39.
Bochner
BH
. 
Gene therapy in bladder cancer
.
Curr Opin Urol
2008
;
18
:
519
23
.
40.
Ramesh
N
,
Ge
Y
,
Ennist
DL
,
Zhu
M
,
Mina
M
,
Ganesh
S
, et al
CG0070, a conditionally replicating granulocyte macrophage colony-stimulating factor—armed oncolytic adenovirus for the treatment of bladder cancer
.
Clin Cancer Res
2006
;
12
:
305
13
.
41.
Cozzi
PJ
,
Malhotra
S
,
McAuliffe
P
,
Kooby
DA
,
Federoff
HJ
,
Huryk
B
, et al
Intravesical oncolytic viral therapy using attenuated, replication-competent herpes simplex viruses G207 and Nv1020 is effective in the treatment of bladder cancer in an orthotopic syngeneic model
.
FASEB J
2001
;
15
:
1306
8
.
42.
Simpson
GR
,
Horvath
A
,
Annels
NE
,
Pencavel
T
,
Metcalf
S
,
Seth
R
, et al
Combination of a fusogenic glycoprotein, pro-drug activation and oncolytic HSV as an intravesical therapy for superficial bladder cancer
.
Br J Cancer
2012
;
106
:
496
507
.
43.
Hanel
EG
,
Xiao
Z
,
Wong
KK
,
Lee
PWK
,
Britten
RA
,
Moore
RB
. 
A novel intravesical therapy for superficial bladder cancer in an orthotopic model: oncolytic reovirus therapy
.
J Urol
2004
;
172
:
2018
22
.
44.
Gomella
LG
,
Mastrangelo
MJ
,
McCue
PA
,
Maguire
HC
,
Mulholland
SG
,
Lattime
EC
. 
Phase 1 study of intravesical vaccinia virus as a vector for gene therapy of bladder cancer
.
J Urol
2001
;
166
:
1291
5
.
45.
Burke
JM
,
Lamm
DL
,
Meng
MV
,
Nemunaitis
JJ
,
Stephenson
JJ
,
Arseneau
JC
, et al
A first in human phase 1 study of CG0070, a GM-CSF expressing oncolytic adenovirus, for the treatment of nonmuscle invasive bladder cancer
.
J Urol
2012
;
188
:
2391
7
.
46.
Ribas
A
,
Dummer
R
,
Puzanov
I
,
VanderWalde
A
,
Andtbacka
RHI
,
Michielin
O
, et al
Oncolytic virotherapy promotes intratumoral T cell infiltration and improves anti-PD-1 immunotherapy
.
Cell
2017
;
170
:
1109
19
.

Supplementary data