Biomimetic device and foreign body reaction cooperate for efficient tumour cell capture in murine advanced ovarian cancer

ABSTRACT Metastasis is facilitated by the formation of pre-metastatic niches through the remodelling of the extracellular matrix (ECM) promoted by haematopoietic and stromal cells. The impact of these primed sites is pronounced for intraperitoneal metastases, where the cavity-exposed ECM supports the attachment of the disseminating tumour cells. Likewise, implantation of biomaterial scaffolds influences metastatic progression systemically through a foreign body reaction (FBR). In this study, we integrated the concept of creating an artificial niche to capture tumour cells actively disseminating in the peritoneal cavity with a therapeutic strategy modulating the interactions of metastatic cells with the ECM. The aim was to transform a disseminated disease into a focal disease. For this, we designed and developed a ‘biomimetic’ ECM composed of a nonresorbable three-dimensional scaffold with collagen coating and characterized the FBR to the implanted biomaterial. We also analysed the safety of the implanted devices and their ability to capture tumour cells in different murine preclinical models of advanced ovarian cancer. Implantation of the biomimetic devices resulted in an initial inflammatory reaction that transformed progressively into a fibrous connective tissue response. The adhesive capabilities of the scaffold were improved with the ancillary effect of the FBR and showed clinical utility in terms of the efficacy of capture of tumour cells, disease focalization and survival benefit. These results demonstrated the performance and safety of this ‘biomimetic’ ECM in preclinical models of advanced ovarian cancer. Translated into the clinical setting, this new therapeutic strategy represents the possibility for control of peritoneal carcinomatosis upon primary ovarian debulking surgery and to expand the percentage of patients who are candidates for second rescue surgeries at the time of relapse.


Figure S2
In vitro biomimetic tumor cell capture efficacy was assessed in an orbital adhesion assay that mimics peritoneal dissemination in ovarian cancer (de la Fuente et al., J Natl Cancer Inst. 2015;107(9)). Biomimetic devices, including the bare scaffold without collagen coating and devices with increasing (low and high) concentrations of collagen (25 and 250 mg per device, respectively), were immobilized in pre-treated p100 cell culture plates (one device per plate) and one million of SKOV3 cells, labelled with the fluorescent marker calcein (4 µM/l calcein acetoxymethyl ester (Invitrogen, Paisley, UK), according to manufacturer´s protocol), were added to each plate before placed on an orbital shaker at 90rpm and incubated for 24, 48 and 72 hours at 37ºC in 5% CO2. After incubation, biomimetic devices were recovered from the plates and the amount of cells captured by the bare scaffold and by scaffolds coated with increasing amounts of collagen was evaluated. Quantification of captured fluorescent-labelled SKOV3 cells was performed using a fluorimeter. Fluorescence intensity of calcein is recorded at 485nm wavelength and expressed in arbitrary units. The amount of tumor cells, expressed as mean average ± standard deviation, captured by Biomimetic bare scaffold (Empty Group), and bare scaffold coated with collagen at low (Biomimetic Low Collagen Concentration Group) and high (Biomimetic High Collagen Concentration Group) collagen concentrations at 24 (black bars), 48 (dark grey bars) and 72 (light grey bars) hours of incubation in the adhesion orbital assay, confirmed the principal adhesive abilities of the scaffold and the ancillary effect of the collagen coating (p<0.0001), in a collagen concentration and time-dependent manner (p<0.0001).  The efficacy of the biomimetic devices to capture different clinically relevant ovarian cancer cell types was evaluated in the murine model of ovarian cancer peritoneal dissemination at one week post-implantation, complementing the results observed with the SKOV3 adenocarcinoma cell line. Two different cell lines were evaluated: TOV112 (serous origin) and OV90 (endometroid origin), as representative cell lines of the most common histology subtypes of ovarian carcinomas diagnosed in the clinical setting. Additionally, primary cancer cells isolated from ascitic fluid of ovarian cancer patients were also studied, representative of patient derived tumors. In this study, one million cells were injected intraperitoneally. One week later, mice were sacrificed and the pattern of tumor cell dissemination was evaluated by bioluminescence for luciferase-expressing

Figure S8
Tumor burden at time of sacrifice of mice included in the survival preclinical study that simulates its intended clinical use evaluated as global bioluminescence signal from tumor cells in the peritoneal metastasis and in the biomimetic devices for the different groups included in the study.
Normalized photons indicate that no differences were found between the control, the biomimetic post-injection and the biomimetic groups, further supporting the conclusion that tumor capture efficacy by the biomimetic technology promotes a remodelling of the pattern of metastasis and the focalization of the disease. Finally, bioluminescence signal from mice included in the biomimetic removal group represents residual disease after removal of the devices.  represents the concentration of a drug at which 50% of its maximal effect was observed (Carboplatin TEVA (10mg/ml; lot number: 13E28KB); and Paclitaxel TEVA (6mg/ml; lot number 3620112). EC50 dose for Carboplatin was 10µM, and for Paclitaxel was 4nM; EC50/2 represents half of the EC50 drug concentration. One week after biomimetic device implantation, one million luciferase-expressing SKOV3 ovarian cancer cells were injected intraperitoneally, and 24 hours later chemotherapy was administrated intraperitoneally. The pattern of peritoneal metastasis was

Figure S10
Tumor cell capture efficacy of the biomimetic device in immune-competent mice. To evaluate the impact of the immune system on the tumor cell capture efficacy of the biomimetic technology, immune-competent mice (C57BL/6) were implanted with the biomimetic devices as described in M&M, and one week later injected with 3x10 6 DiD-fluorescently labelled ID8 ovarian cancer cells.
This cell line is originated from mice tumors so the use of immune competent preclinical models is favoured. One week after tumor cell injection, mice were sacrificed and the pattern of peritoneal dissemination evaluated as described in M&M, in the presence (biomimetic group; n=2) and absence (control group; n=2) of the biomimetic device. As shown by these preliminary results, the Adequate studies with humanized preclinical models are ongoing to appropriately characterise the impact of the immune components on the efficacy of the biomimetic technology.