Reduction of Muscle-Invasive Tumors by Photodynamic Therapy with Tetrahydroporphyrin-Tetratosylat in an Orthotopic Rat Bladder Cancer Model

This article is featured in Highlights of This Issue, p. 731

Bladder cancer is the ninth most common cancer worldwide (1), with a mortality rate of 2.3 (GLOBOCAN 2012; http://www.globocan.iarc.fr/). Approximately 70% of cases are initially diagnosed as non–muscle-invasive urothelial carcinomas (stage Ta, T1, CIS) and 30% as muscle-invasive urothelial carcinomas. Non–muscle-invasive bladder cancer is usually treated by transurethral resection of the tumor followed by instillation of Bacillus Calmette–Guérin or chemotherapy, respectively. In high-grade T1 tumors or after failure of first-line therapy, radical cystectomy should prevent further progression (2). For muscle-invasive urothelial cancer (≥ pT2), radical cystectomy including lymph node dissection with or without neoadjuvant chemotherapy is recommended (3). To date, there are no options of organ preserving therapies in advanced bladder cancer.

Photodynamic therapy (PDT) is a promising option for minimal-invasive bladder preserving treatment of urothelial cancer. The principle of PDT is based on photoactivation of a nontoxic light reactive agent, so-called photosensitizer (PS), by a specific wavelength (4). In the presence of tissue oxygen, photoactivation leads to intracellular oxidative stress by means of reactive oxygen species and singlet oxygen (¹O₂), which can destroy tumors by three different effects: (i) tumor cell death (apoptosis/necrosis), (ii) coagulation of microvessels resulting in vascular shutdown (nutritional cutoff), and (iii) stimulation of the host immune system (5, 6). An ideal PS for tumor therapy has no dark toxicity and selectively accumulates in tumor cells (7). The photodynamic principle is implemented in diagnostics and therapies of several tumor entities. Several systemically applied PSs (e.g., photofrin, haematoporphyrin derivates) have been used for PDT of superficial urothelial cancer and carcinoma in situ (CIS) (7–9). Systemic administration of these drugs resulted in undesirable side effects such as severe skin photosensitization. Studies with locally applied 5-aminolevulinic acid (ALA) had complete response rates up to 52%–60% in CIS patients without skin photosensitization, but caused bladder wall fibrosis due to high irradiation intensities used (10, 11). To date, no standard PDT for superficial urothelial carcinomas could be developed, and PDT of muscle-invasive bladder cancer is hampered by the insufficient tissue penetration due to short excitation wavelength of the commonly used PSs (12).

Tetrahydroporphyrin-tetratosylat (THPTS) is a novel water-soluble and positively charged PS. THPTS has an absorption maximum at 760.5 nm, which lies within the so-called phototherapeutic window, thus allowing up to 15-mm tissue penetration (13, 14). THPTS strongly accumulates in tumor cells with a tumor/normal tissue ratio from 1.4 to 25 (14). Pharmacokinetic examinations showed rapid clearance from healthy tissue (half-life 30 hours), which avoid generalized skin photosensitivity. Reduced energy required for photoactivation minimizes tissue fibrosis following heat damage (14). Cytotoxic effects of THPTS-PDT have been described in different tumors (15–17), including human urothelial carcinoma cells in vitro. THPTS-PDT on human bladder carcinoma cells showed significant tumor cell toxicity while sparing normal human detrusor myocytes. (18). To confirm our recent results of THPTS-PDT efficacy in human urothelial cancer cells, we here evaluate the effects of THPTS-PDT in an orthotopic rat bladder cancer model (19). We applied THPTS to rat urothelial cancer cells (AY-27) in vitro to analyze pharmacokinetics, subcellular distribution, and cell death mechanisms; and we used a Fischer F344 rat model to evaluate the effects of THPTS-PDT on tumor growth in vivo.

The near-infrared (760 nm) PS THPTS (C₇₂H₇₀N₈O₁₂S₄; molecular weight: 1,367.66 Da; purity ≥ 95%) was obtained from TetraPDT GmbH. THPTS is a cationic bacteriochlorin derivate (Fig. 1), which shows no change in absorption spectra within 1 month when kept in aqueous solutions in the dark at 4°C (14). The physicochemical properties of THPTS have been extensively described in previous studies (14, 20).

Rat AY-27 urothelial carcinoma cells are not commercially available. Cells were kindly provided by Prof. F. Guillemin and Dr. M.A. D'Hallewin (Centre de Lutte Contre le Cancer de Lorraine, Nancy, France). The cell line was initially derived from an N-(4-[5-nitro-2-furyl]-2-thiazolyl) formamide (FANFT, 0.2%)-induced carcinoma of the urinary bladder in male F344 Fischer rats (Charles River; ref. 21). Cells were cultured in RPMI-1640 w/o phenol red supplemented with 20% fetal calf serum (FCS), 5% penicillin/streptomycin, and 200 mmol/L glutamine (Biochrom). AY-27 were characterized as poorly differentiated cells of urothelial origin by cytokeratin staining (cytokeratin-pan, cytokeratin-7: positive; cytokeratin-13, cytokeratin-20: negative; Supplementary Fig. S1). Mycoplasma testing was performed monthly using PCR Mycoplasma Test Kit (AppliChem GmbH).

Uptake analyses were performed on a BD LSR II flow cytometer (Becton-Dickinson). Prior to measurement AY-27 cells were incubated with THPTS (200 μmol/L, 2 hours) at 37°C or 4°C. THPTS fluorescence was monitored during 2 hours using an octagon laser (excitation: 488 nm–20 mW) and a PE-Cy7-A filter set (detection: BP750/60 nm). Dead cells were excluded from analysis by propidium iodide staining. Very small cells (probably representing cell detritus) and duplets were excluded in a two-step gating strategy using FSC-H/FSC-A scatters.

Cells were cultured on collagen-A coated 35-cm dishes with glass bottom (ibidi). For subcellular staining, AY-27 cells were treated with CellLight reagents BacMam 2.0 (Early/Late Endosomes-GFP), LysoTracker-Green or MitoTracker-Green (Thermo Fisher Scientific). After THPTS incubation (200 μmol/L, 37°C, 2 hours), cells were washed and covered with calcium-free Krebs-Ringer solution and analyzed by confocal laser scanning microscopy (LSM 5 Pascal, Carl Zeiss).

Prior to THPTS-PDT, cells were grown in 96-well plates (Sarstedt). At 70% to 80% confluency, samples were treated 2 hours with THPTS (0, 12.5, 25, 50, 100, and 200 μmol/L) and irradiated at 760 nm using a 760 ± 5 nm diode laser (LAMI-Helios; TetraPDT GmbH). Gentle light dose of 10 J/cm² was chosen based on its effective and dose-dependent cytotoxicity in initial irradiation experiments using various light doses (10, 30, and 60 J/cm²; Supplementary Fig. S2). Laser intensity was measured and adjusted using a laser power meter (Laserstar). After irradiation, cells were incubated for indicated time periods and assayed for cytotoxic effects.

Cells were grown in 24-well plates (Greiner Bio-One). After THPTS-PDT (12.5 μmol/L), 24-hour time-lapse recordings of AY-27 cells were analyzed for membrane blebbing and nuclear condensation using an EVOS XL digital inverted microscope (AMG).

Cells were seeded into 96-well-plates. After irradiation, plates were incubated for 2 and 24 hours. Cell survival was analyzed with the CellTiter-Blue Cell Viability Assay (ex/em: 560/590). The apoptotic activity was measured with the Apo-ONE Homogeneous Caspase-3/7 Assay (ex/em: 485/527) and the loss of membrane integrity was determined via measurement of lactate dehydrogenase (LDH) release using the CytoTox-ONE Homogeneous Membrane Integrity Assay (ex/em: 560/590). Assays were obtained from Promega and measured on a microplate reader (SpectraMax M5, Molecular Devices).

All animal experiments were approved by the Landesdirektion Sachsen (TVV 47/11) and complied with requirements of the ARRIVE guidelines. Pilot study (n = 11) was performed for establishing the bladder cancer model prior to treatment. Animals were sacrificed after indicated time periods and bladders were excised (Table 1).

We followed the method of Xiao and colleagues (19, 22). Female F344 Fischer rats (150–180 g; Charles River Germany) were anesthetized i.p. with a mixture of 90 mg/kg Ketamine (Ketamin 10%, WDT) and 10 mg/kg Xylazin (Rompun 2%, Bayer Health Care). Transurethral catheterization was done with an 18G catheter (BD InsyteAutogard18GA). Bladder conditioning was done by washing with 0.5 mL of 0.1 N HCl for 15 seconds and 0.5 mL of 0.1 N NaOH for 15 seconds. Tumor inoculation was done by instillation of 2 × 10⁶ AY-27 cells (in RPMI-1640, w/o phenol red, 20% FCS, 5% penicillin/streptomycin, 200 mmol/L glutamine) for 60 minutes. The animals were kept for 10 days under standard conditions. Tramal (40 mg/kg; in drinking water; Grünenthal) was administered for pain management. Animals temperature and body weight were checked daily. Ten days after instillation, animals were sacrificed for pilot study or used for PDT treatment.

Animals were anesthetized and catheterized as described above. Bladders were instilled with THPTS (100 μmol/L in PBS) for 2 hours. Controls were instilled with PBS. Following rinsing, bladders were filled with 0.5 mL PBS and irradiated transurethrally via a glass fiber at 760 nm ± 5 nm with light dose of 10 J/cm² (LAMI-Helios, TetraPDT GmbH). After treatment, animals were kept for 14 days as described above. When applicable, tumor growth and treatment procedure was monitored by sonography (Venue 40, GE Healthcare).

The bladders were excised and opened ventrally by a median incision. Bladders were fixed (buffered formaldehyde 4% solution, pH 7.2; VWR) for 24 hours. After macroscopy, paraffin-embedded bladders were serially sectioned (7 μm; Supplementary Fig. S3). Every 20th section was stained by hematoxylin/eosin (H&E) and analyzed microscopically for tumor staging. Therefore, current TNM classification of malignant tumors (TNM; ref. 23) was slightly modified to our needs (Supplementary Table S1). H&E-stained slides were scanned (EPSON Perfection V750 Pro, Epson) to determine tumor size. Tumor volume was calculated from tumor area of manually selected regions on the histologic sections using ImageJ and the spacing of the serial sections.

Complete data analysis was performed using Prism 6.0 (GraphPad Software Inc.) statistical software. In vitro results present data from at least three independent experiments. Differences were analyzed by χ² test, Wilcoxon signed-rank test and Mann–Whitney test. P ≤ 0.05 was considered statistically significant.

Cellular uptake of THPTS into rat AY-27 urothelial cancer cells was analyzed by flow cytometry after incubation with 200 μmol/L THPTS at 37°C. THPTS uptake was highly temperature dependent. At 37°C uptake was rapid, started within 1 minute and increased linearly during 2 hours of incubation. THPTS uptake was significantly decreased to almost nonexistent at 4°C (Fig. 2A and B). Localization of THPTS in endosomes and lysosomes, but not mitochondria, was observed using confocal laser microscopy, revealing colocalization of labels after treatment with cell compartment stains (Fig. 2C–K).

Cytotoxicity and cell death mechanisms were analyzed after THPTS treatment and irradiation with a dose of 10 J/cm². Time-lapse analyses (Fig. 3) demonstrated early onset of apoptosis by induction of membrane blebbing within 5 minutes after THPTS-PDT of AY-27 cells. A maximum of blebbing was seen after 10 minutes (Fig. 3B and C). Cell detachment was observed immediately after PDT (Fig. 3D and E). Significant nuclear condensation was seen as early as 4 hours after irradiation and increased steadily during 24 hours of observation (Fig. 3G and H). Cytotoxic effects were additionally determined by cell-based assays 2 and 24 hours after PDT using various THPTS concentrations. THPTS-PDT led to a significant dose-dependent decrease of AY-27 cell survival, resulting in 82% loss of viability 2 hours after PDT (25 μmol/L THPTS) and 97% after 24 hours (25 μmol/L THPTS). Caspase-3/7 activity was significantly increased after 2 hours with a maximum at 12.5 μmol/L. Release of LDH indicating loss of membrane integrity significantly increased 24 hours after PDT with a maximum at 25 μmol/L THPTS. At that point, apoptotic caspase-3/7 activity was not significantly changed (Supplementary Fig. S4).

Pilot study (n = 11) was performed for establishing the orthotopic bladder cancer model in Fischer rats prior to treatment. Animals were sacrificed after indicated time periods and bladders were excised (Table 1).

Bladder tumor induction was done according to Xiao and colleagues (19). In the pilot study, all 11 rats developed an urothelial cell carcinoma (Fig. 4). Intravesical global AY-27 administration resulted in solid tumors 10 days after inoculation (Fig. 4A and C). H&E staining revealed pT1 tumors invading the subepithelial connective tissue (Fig. 4B and D) with total tumor volume ranging from 0.02 to 0.48 mm³ (mean 0.10 mm³). Relative tumor volume normalized to bladder wall volume ranged from 0.40% to 6.51% (mean, 1.77%; Fig. 4F). Minimal variations of animal body weight complied with legal requirements for establishment of an animal model (Fig. 4E). Animal numbers of treatment studies were determined by power analysis (G*-Power V.3.1; University of Düsseldorf, Germany). Because adequate group size of 10 rats was confirmed (power = 0.970), animal numbers were minimized in respect to the 3R principle (24). Treatment was done 10 days after tumor induction by global THPTS administration (PBS in controls) and transurethral light application. Bladders were extracted 14 days after treatment and evaluated by macroscopy and histology. There were no irradiation-related deaths in the groups. On dissection, bladders were covered by intraperitoneal fat and no perforations were observed. In addition, there was no macroscopic evidence of abscess in the kidneys to suggest serious local infections. Histology of serial sections allowed quantification of bladder tumor volumes (Supplementary Table S2). Treatment with THPTS-PDT led to significantly reduced numbers of ≥ pT2 tumors (P = 0.024; χ² test). In the treatment group, 2 of 10 animals had muscle-invasive (≥ pT2a) tumors and 8 of 10 animals exhibited non–muscle-invasive bladder cancer (pT1). In contrast, in the control group, 7 of 10 animals showed ≥ pT2a tumors and 3 of 10 non–muscle-invasive tumors. Quantification revealed significant reduction of tumor volume in the THPTS-PDT treatment group (Fig. 5A). Relative tumor volume was reduced from 17.42% ± 4.43% to 6.99% ± 4.07% (mean ± SD; P = 0.023; Mann–Whitney test). Relative tumor volume of ≥ pT2 classified tumors was reduced from 13.79% ± 4.68% to 4.51% ± 4.13% (mean ± SD; P = 0.034; Mann–Whitney test). Interestingly, pT1 tumors were not significantly reduced in volume.

Photodynamic effects on tissue architecture are shown in Fig. 5B and C. There was fragmentary damage of the urothelium in the tumor area with regions of complete urothelial destruction and superficial ulceration. Tumor-surrounding areas showed acute inflammation by infiltration with mononucleated round cells. Extensive full-thickness damage of the bladder wall was not seen. There were no signs indicating phototoxic side effects (e.g., fibrosis, inflammation) in the flanking normal urothelium, which consisted of 3 to 6 tightly opposed cell layers. The lamina propria contained loose connective tissue with isolated cellular components and a few vascular structures extending to the underlying muscle layers (Fig. 5C).

PDT could be a promising alternative in bladder preserving treatment of bladder cancer. Due to selective tumoricidal phototoxicity with minor side effects in clinical experience, research concentrated on the prodrug ALA/hexaminolevulinate (HAL) as possible photosensitizers (25, 26). The major disadvantage of these agents is low therapeutic invasion depth. The present study aimed to evaluate the near-infrared excitable THPTS for treatment of invasive bladder carcinoma in an orthotopic rat model.

Investigations in cultured rat AY-27 cells gave insights into cellular mechanisms of THPTS-PDT. Cellular THPTS accumulation is comparable to results on human TCC in our previous study (18). As in human urothelial carcinoma cells, THPTS pharmacokinetics showed strong dependency on temperature with linear uptake at 37°C and almost no uptake at 4°C, speaking in favor of an active uptake mechanism (Fig. 2A and B). Also, THPTS colocalized to endosome and lysosome trackers, but did not show mitochondrial accumulation (Fig. 2C–K). This indicates cell type–specific subcellular transport mechanisms, because THPTS preferentially accumulated in mitochondria in retinoblastoma cells (16). Cytotoxic effects of PDT could be due to direct release of lysosomal enzymes after illumination (27). THPTS-PDT led to early onset of apoptosis as demonstrated in time lapse (Fig. 3) and cytotoxicity assays (Supplementary Fig. S4), which is in line with findings in human urothelial cancer cells (18). Therefore, AY-27 cells are suitable to investigate THPTS-PDT efficacy in bladder cancer.

We established the orthotopic rat model initially described in 1999 by Xiao and colleagues (19). Various trial setups resulted in 80% to 100% tumor induction (19, 22, 28). We could show tumor induction in 11 of 11 rats (all pT1) 10 days after instillation of 2 × 10⁶ AY-27 cells (Fig. 4). To overcome tumor volume variations observed in pilot study, we developed a strict method for studies in the treatment groups. Accurate evaluation with serial sections (7 μm) and examination of every 20th section was not achieved in any former studies (e.g., serial sections with only two stains; ref. 29).

Animal studies confirmed the in vitro results and indicate tumoricidal phototoxicity and high tumor selectivity of THPTS in vivo. Laser light exposure after intravesical THPTS application did not cause skin photosensitization, irradiation-related deaths, or extensive full-thickness bladder damage. Other second-generation PSs (e.g., ALA/HAL) showed comparable tumor selectivity and low bladder wall phototoxicity in the F344 bladder cancer model. For instance, there were no deaths or histopathologic changes in the bladder muscular layer after intravesical ALA-PDT (30). In addition, selective urothelial cancer necrosis could be achieved with ALA/HAL-PDT and Hypericin-PDT without major effects to the underlying musculature in F344 (31, 32). However, these PSs have a therapeutic absorption maximum at shorter wavelengths, a less favorable characteristic for deep tumoricidal action. Therefore, research concentrated on PDT of superficial urothelial cancer. Due to its absorption maximum at 760.5 nm, THPTS-PDT could overcome these limitations of low tissue penetration. In the present study, we evaluated THPTS-PDT in vivo. Our findings are in accordance with previous in vivo results of effective THPTS-PDT on heterotopic cholangiocarcinoma in SCID mice (13). Additionally, selective necrosis with an invasion depth of 13 mm could be induced by THPTS-PDT on heterotopic C26 colon carcinoma in BALB/c mice (14). Here, we revealed significant reduction of total tumor volume and ≥ pT2 tumor number after singular intravesical THPTS-PDT on AY-27–induced bladder tumors in an immunocompetent animal host without side effects (e.g., skin photosensitization, bladder shrinkage; Fig. 5). Reasons for lacking complete tumor remission could be numerous. Future research should focus on optimizing drug and light dose and should explore repeated treatment regimens resembling current clinical standards to improve treatment effects. Nevertheless, this is the first time PDT was successfully evaluated for treatment of muscle-invasive bladder cancer. Effective reduction to non–muscle-invasive bladder cancer was shown, which might offer bladder-preserving treatment option by transurethral resection.

Our results support the idea of THPTS-PDT as a promising alternative in antitumor therapies. We confirmed cell culture results in human urothelial carcinoma cell lines for rat cells. For the first time, its possible therapeutic use was shown in an immunocompetent animal model. THPTS-PDT may overcome major problems of currently used photosensitizers and might offer a new treatment strategy in advanced bladder cancer therapy.

No potential conflicts of interest were disclosed.

Conception and design: L.-C. Horn, J. Neuhaus

Development of methodology: M. Berndt-Paetz, P. Schulze, A. Weimann, R. Hermann, J. Neuhaus

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M. Berndt-Paetz, P. Schulze, P.C. Stenglein, A. Weimann, J. Neuhaus

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M. Berndt-Paetz, P. Schulze, P.C. Stenglein, Q. Wang, L.-C. Horn, J. Griebel, J. Neuhaus

Writing, review, and/or revision of the manuscript: M. Berndt-Paetz, P. Schulze, L.-C. Horn, Y.M. Riyad, J. Griebel, A. Glasow, J.-U. Stolzenburg, J. Neuhaus

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J. Neuhaus

Study supervision: J. Neuhaus

We thank Prof. F. Guillemin and Dr. M.A. D'Hallewin (Centre de Lutte le Cancer de Lorraine, Nancy, France) for providing the rat urothelial cancer cell line AY-27. The study was supported by the Deutsche Forschungsgemeinschaft (grant: NE425/6-1 to J. Neuhaus).

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.