The concept of foam as a drug carrier for intraperitoneal chemotherapy, feasibility, cytotoxicity and characteristics

For decades, intraperitoneal chemotherapy (IPC) was delivered into the abdominal cavity as a liquid solution. This preliminary study aims to evaluate foam as a potential new drug carrier for IPC delivery. Foam-based intraperitoneal chemotherapy (FBIC) was produced with taurolidine, hydrogen peroxide, human serum, potassium iodide and doxorubicin/ oxaliplatin for both ex vivo and in vitro experiments. Analysis of FBIC efficacy included evaluation of cytotoxicity, tissue penetration, foam stability, temperature changes and total foam volume per time evaluation. FBIC showed penetration rates of about 275 ± 87 µm and higher cytotoxicity compared to controls and to conventional liquid IPC (p < 0.005). The volume of the generated foam was approximately 50-times higher than the initial liquid solution and temporarily stable. Foam core temperature was measured and increased to 47 °C after 9 min. Foam ingredients (total protein content) were evenly distributed within different locations. Our preliminary results are quite encouraging and indicate that FBIC is a feasible approach. However, in order to discuss a possible superior effect over conventional liquid or aerosolized chemo applications, further studies are required to investigate pharmacologic, pharmacodynamic and physical properties of FBIC.

www.nature.com/scientificreports www.nature.com/scientificreports/ exhibiting similar cytotoxic levels as oxaliplatin 12 . Hydrogen peroxide has also demonstrated a wide range of specific antitumoral activity. Both endogenously produced and exogenously added hydrogen peroxide display an antitumor effect [13][14][15] . Unfortunately, the delivery of hydrogen peroxide into solid tumors is much more challenging than its common use as a surface applicant in the treatment of skin cancer 16 . Nevertheless, hydrogen peroxide has recently been tested in the treatment of solid tumors, among other substances 17,18 . Although its various antitumoral effects are increasingly understood and appreciated 19 , the implementation of a potentially widely applicable hydrogen peroxide solution for oncologic purposes remains challenging.

Methods
foam-based intraperitoneal chemotherapy (fBic). The ratio of foam ingredients were experimentally determined. To create the FBIC solution of taurolidine (Taurolin ® Ringer 0.5%, Berlin-Chemie AG, Berlin, Germany), hydrogen peroxide (30% hydrogen peroxide solution, Chempur, Piekary Śląskie, Poland), human serum (from human male AB plasma, Sigma-Aldrich; Merck KgaA, Darmstadt, Germany) and potassium iodide (Sigma-Aldrich; Merck KgaA) was used. The initial liquid solution consisted of 0.045% taurolidine, 22.8% hydrogen peroxide, 12.5% human serum and 12 mM potassium iodide. Additionally, doxorubicin (doxorubicin hydrochloride purchased from PFS ® , 2 mg/ml, Pfizer, Sandwich, United Kingdom) and oxaliplatin (Medoxa, Medac GmbH, Wedel, Germany) were added in both ex vivo and in vitro models, respectively. The applied dosage of doxorubicin was chosen based on dosages used in PIPAC, e.g. 3 mg of doxorubicin was used to create foam covering a 4-liter cavity 11,20,21 . Oxaliplatin was applied at a total concentration of 26.8 µg/ml. Ex vivo model. The experiments were performed in a standard ex vivo model on commercially available tissue samples, therefore no approval of the Institutional Review Board and no consent of the Local Board on Animal Care was required. The ex vivo model has been well established and previously described in many studies 22,23 . A commercially available hermetic plastic box with a total volume of 4 liters was used, mimicking the abdominal cavity. The plastic box was closed during each individual procedure. On the cover of the plastic box, a 5 mm trocar (Kii ® Balloon Blunt Tip System, Applied Medical, Rancho Santa Margarita, CA, USA) was placed. Using one trocar, a medical intravenous catheter was introduced. The catheter was used to apply the foam by injecting a previously mixed starting solution (containing taurolidine, hydrogen peroxide, human serum and doxorubicin) and potassium iodide to induce foam formation. Three fresh tissue specimen of peritoneum (German landrace pigs), each measuring 3.0 ×3.0 ×0.5 cm, were placed at the side wall (Fig. 1). Drug-exposure time to FBIC was 30 min.
Evaluation of foam stability, measurement of temperature and gravitational effects. 30 ml of the initial FBIC solution was introduced in a measuring jug to produce 4 liters of foam. FBIC volume was measured after initiating foam creation by addition of potassium iodide, which served as a catalyst. Central and peripheral temperatures were measured via a temperature probe at different time points of the foam formulation process (Fig. 1). Time points for temperature measurements were equivalent to 0, 25, 50, 75 and 100% of the estimated maximal total volume of the foam produced. When the foam reached its maximum expansion, volume samples (0.25 ml) were taken from the bottom, middle and top of the foam. To evaluate possible gravitational effects in the foam composition, the protein fraction of the human serum was used as an indicator of possible concentration imbalances. Using the Bradford method (Sigma-Aldrich), the total content of human serum protein was determined to detect possible concentration differences between different areas of the foam. www.nature.com/scientificreports www.nature.com/scientificreports/ Microscopic analysis. After treatments, all tissue samples were rinsed with sterile NaCl 0.9% solution to eliminate superficial cytostatic agents and immediately frozen in liquid nitrogen. Cryosections (7 µm) were prepared from different areas of the specimen. Sections were mounted with VectaShield containing 1.5 µg/ml 4' ,6-diamidino-2-phenylindole (ProLong ® Gold Antifade Reagent with DAPI, Thermo Fisher Scientific) to stain nuclei. Penetration depth of doxorubicin was measured using a Nikon Eclipse 80i fluorescence microscope (Nikon Instruments Europe B.V. Amsterdam, Netherlands). The distance between the peritoneal surface and the innermost positive staining for doxorubicin accumulation was reported in micrometers.
Next, to investigate the effect of FBIC treatment on tumor cell cytotoxicity, the medium was aspirated from each well. The attached cells were treated with four different solutions: The exposure time was 1 hour at 36°C with 5% CO 2 . After this period, the content of the wells was aspirated, and 0.5 ml of fresh medium was added. Cells were incubated for 48 hours under the same conditions and an MTS proliferation assay was performed.

Foam stability, measurement of temperature and gravitational effects.
The formation of a temporarily stable foam was possible. The foam reached its maximal volume after 12 minutes, after which volume size constantly decreased (Fig. 2). The volume of the foam produced was approximately 50 times greater than the initial liquid solution. The foam was stable for 150 minutes. The foam volume decreased by 50% after 74 minutes and 75% after 113 minutes. The temperature of central and peripheral parts of the formulated foam was measured at time points equivalent to 0, 25, 50, 75 and 100% of the estimated maximum total volume of the foam. Core (temperature probe 2) and peripheral temperatures (temperature probes 1 and 3) of the foam rose up to 47 and 46 °C, respectively, after 9 min (Fig. 3A). Foam total protein content was evenly distributed on the bottom, in the middle and at the top of the box. No significant differences in the concentration of the protein could be detected (p > 0.05, Fig. 3B).

Microscopic analysis of doxorubicin depth penetration in the ex vivo model. All samples had
contact with the FBIC in the established ex vivo model. A minimal amount of foam remained constantly on the sample surface at all times. A partial discoloration of the superficial peritoneal layer could be observed after removing the foam layer. The mean depth of doxorubicin penetration detected via fluorescence microscopy was found to be 275 ± 87 µm (Fig. 4). Measured penetration depths are higher at peripheral locations in FBIC than in tissue samples treated with liquids and pressurized aerosol 8,24 . The volume of the foam produced decreased with time, as the foam structure destabilized. www.nature.com/scientificreports www.nature.com/scientificreports/

Cytotoxicity of FBIC on colon cancer cells (HT-29) in vitro.
FBIC showed significant cytotoxicity compared to untreated controls in an in vitro model (p < 0.05, Fig. 5). Interestingly, FBIC seems to have even higher cytotoxicity than medium with oxaliplatin (p < 0.005). This effect remains unchanged even if oxaliplatin is added to FBIC. No additional cytotoxicity is observed in that case.   www.nature.com/scientificreports www.nature.com/scientificreports/

Discussion
Intraperitoneal administration of anticancer drug solutions is an established method of treating PM, as highly concentrated drug particles are put into contact with tumor nodules in the peritoneal cavity. However, limitations such as inhomogeneous drug distribution and limited penetration into the peritoneal tissue have been described with both liquid as well as aerosol-based IPC 6,8 , even when applied with new, improved techniques 20,24,25 . Additionally, novel substances delivered intraperitoneally have displayed limitations in drug tissue penetration 26 and modified technical applications have been suggested to improve IPC 27 .
Foam displays some unique characteristics which, to the best of our knowledge, have not been tested as a drug carrier for intraperitoneal chemo applications. Foam-based IPC could be a technically feasible option for treatment of PM, using doxorubicin and oxaliplatin as in this study or other chemotherapeutic agents. Penetration depths of chemotherapy are higher compared to liquid or aerosol-based IPC, especially when compared to penetration levels at more peripheral locations. This study also demonstrates that foam volume decreases with time. Slow degradation allows for extended drug contact time with the peritoneal tissue and a higher drug diffusion gradient (Fig. 6). This is an interesting feature, since length of contact of a chemotherapeutic drug with the peritoneal tissue enhances drug availability and efficiency.
Foam has many advantages over both aerosol and liquid applications. It expands differently than liquids and gas and displays a higher drug-carrying capacity than gas. Thus, even a low total drug dosage can create high  www.nature.com/scientificreports www.nature.com/scientificreports/ drug concentrations as more than 95% of the actual volume is air. Aerosol chemotherapy has been shown to achieve much higher drug concentrations than regular liquid solutions. However, this increase in drug concentration is not without consequence, as aerosol chemotherapy displays increased inhomogeneity compared to liquid applications.
The unique characteristics of foam might significantly improve the response of PM to IPC. In our study, foam containing hydrogen peroxide and taurolidine has demonstrated cytotoxic properties which may be sufficient for the treatment of PM without the need for additional chemotherapeutic agents. However, further studies are needed to evaluate the clinical applications of taurolidine and hydrogen peroxide foam in the treatment of PM.
Our data indicate that foam might be a possible carrier for IPC and could offer increased drug penetration and more homogenous drug distribution than conventional liquids and pressurized aerosol. However, further research is required to assess its potential in IPC application. To the best of our knowledge, no clinical experience for foam-based applications in IPC has been previously collected or published in peer-reviewed literature. While this study presents preliminary data, it gives important insight into the potential of FBIC to improve PM treatment and encourages further studies to evaluate FBIC's full efficacy and biodistribution.

Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.