Feasibility and guidelines for the use of an injectable fiducial marker (BioXmark^®) to improve target delineation in preclinical radiotherapy studies using mouse models.

BioXmark^®

BioXmark^® is a liquid FM produced by Nanovi A/S (Kongens Lyngby, Denmark). It is a sterile, ready-to-inject liquid composed of biodegradable sucrose acetate isobutyrate (SAIB), iodinated SAIB and ethanol. This formulation ensures that when BioXmark^® is injected into soft tissue the ethanol partly diffuses out of the marker, increasing its viscosity, resulting in the formation of a semi-solid gel.³⁶ BioXmark^® is visible on magnetic resonance imaging (MRI) and ultrasonography due to the SAIB component and visible on X-ray imaging modalities due to the electron-dense iodinated SAIB component.³² Evidence of the visibility of BioXmark^® on multiple imaging modalities in clinical, preclinical and phantom studies is presented in Table 1.

Like other FMs, it is recommended clinically to avoid the injection of BioXmark^® into necrotic tissue, highly vascularised tumour tissue or air-filled cavities.³⁶ Preclinical injection of BioXmark^® should also avoid these tissues and always be through the least invasive procedure possible and actively avoid the need for surgical implantation. Volumes over 50 μl are not recommended for use in mouse models due to negative imaging artefacts on preclinical CBCT scans. An optimum marker volume is suggested between 10-20 μl, this will allow visualisation on CBCT scans without hindering visualisation of small anatomical structures.³⁵ Larger volumes may be more applicable for clinical use or larger animal species.

Animal models

All experimental procedures were carried out in accordance with the Home Office Guidance on the Operation of the Animals (Scientific Procedures) Act 1986, and approved by the Queen’s University Belfast Animal Welfare and Ethical Review Body (PPL2813). Animals were euthanized by Schedule 1 procedures. Animal studies are reported in compliance with the ARRIVE guidelines.³⁷

All mice were obtained from Charles River Laboratories (Oxford, UK). A mix of mice ages and genders were used due to animal availability from other ongoing studies. Female 12–15-week-old C57BL/6J mice were used for subcutaneous (n=3) and intra-peritoneal injections (n=2), male 14-18-week-old C57BL/6J mice were used for intra-tumoral study (n=21 BioXmark^® injected tumours, n=21 control tumours) and male 10-15-week-old C57BL/6J mice were used for prostate injections (n=2). These numbers were used to assess the feasibility of injection types. No criteria were set to exclude animals from these studies. Mice were randomized to receive either a subcutaneous, or intra-peritoneal injections using an online random sequence generator. Blinding was not possible for intra-tumoral and prostate injections. Blinding was also not possible for image analysis as the injected FM, BioXmark^®, was clearly visualized on CBCT scans.

All mice were housed under controlled conditions (12-hour light–dark cycle, 21°C) in standard caging and received a standard laboratory diet and water ad libitum. To improve the welfare of mice environmental enrichment tools were placed in all cages including cardboard tubes for exploration, softwood blocks to encourage gnawing to prevent teeth overgrowth, nesting material for comfort and mouse swings for added cage complexity and exercise. Mice were also handled gently using refined cupping methods and frequently from a young age to reduce stress.

Mice were anaesthetised with injectable ketamine and xylazine (100 mg/kg and 10 mg/kg) prior to BioXmark^® injection and CBCT imaging. All mice were placed in a heat-box (37°C) for recovery and monitored closely after injections until conscious and returned to normal behaviour.

BioXmark^® uptake

A 1 ml micro-dose syringe (Vlow Medical, Netherlands) was used for the uptake and injection of BioXmark^®. Micro-dose syringes are recommended as they ensure a controlled injection volume. A loading 18-gauge needle was used to fill the syringe slowly from the glass ampoule. BioXmark^® is a viscous liquid so thinner needles may require a longer time to fill the syringe. Air bubbles were checked for through visual inspection prior to injection and were removed by gently tapping the side of the syringe and any excess or leakage BioXmark^® was cleaned using ethanol wipes. The loading needle was removed after uptake into the syringe and replaced with a smaller needle (25-gauge) for each injection. Needles were replaced prior to each injection.

Subcutaneous injection of BioXmark^®

BioXmark^® was loaded into a micro-dose syringe as detailed above. For this study, we used a micro-dose syringe with options for the injection of 10, 20 and 40 μl. Mice were anaesthetised with injectable ketamine and xylazine (100 mg/kg and 10 mg/kg) before injection. A small area of fur on each flank was shaved and cleaned with an ethanol wipe for injection. 25-gauge needles were used for subcutaneous injections. The skin of the mouse was tented using the thumb and finger and a preselected volume of BioXmark^® (10 μl (n=1), 20 μl (n=1) or 40 μl (n=1)) was injected under consistent pressure. Each mouse received two injections of BioXmark^®, one on each flank, e.g. 10 μl on the left and 10 μl on the right. As BioXmark^® is very viscous the needle was held for 30 seconds after the volume was injected and then slowly removed. All mice were placed in a heat-box for recovery (37°C) and monitored closely until fully recovered from the anaesthetic and then returned to their original cage.

Intraperitoneal injection of BioXmark^®

For this study, we used a micro-dose syringe (20 and 40 μl) and 25-gauge needles for injections (20 μl, n=1, 40 μl, n=1). Mice were anaesthetised with injectable ketamine and xylazine (100 mg/kg and 10 mg/kg) before the study, and the injection point sterilized with ethanol before intraperitoneal injection. As BioXmark^® is very viscous, the needle was held for 30 seconds following injection and then slowly removed. No mice presented with bleeding or leakage of BioXmark^®, if there was leakage of BioXmark^® this could be removed with an ethanol wipe. Mice were placed in a heat-box for recovery (37°C) and monitored closely until fully recovered from the anaesthetic before returning to their original cage.

Intra-tumoral injection of BioXmark^®

A previous study reported that it was not possible to mix BioXmark^® with tumour cells and a solid tumour must be established prior to injection of BioXmark^®.³⁴ In this study, we trialled an intra-tumoral injection of BioXmark^® as outlined in Brown et al.³⁵

Tumour xenograft studies were performed using MC38 colon cancer cells (originate from James W. Hodge) cultured in DMEM media supplemented with 10% foetal bovine serum (FBS) and 1% penicillin/streptomycin. Cells were maintained at 37°C in a humidified atmosphere of 5% CO₂ and subcultured every 3–4 days to maintain exponential growth.

MC38 cells were cultured in vitro and prepared in PBS (1 × 10⁵ cells per 100 μl). Subsequently, 100 μl was injected subcutaneously into the flank of each C57BL/6J mouse (n=42). Mice were anesthetized using inhalant isoflurane (0.5 – 2%) for implant and placed in a heat box for recovery. Mice were then returned to conventional housing and closely monitored. Tumour volume was determined three times a week using calliper measurements in three orthogonal dimensions. Once tumours grew to a volume of 100 mm³ mice were randomised using an online random sequence generator into control (n=21) and BioXmark^® (n=21) cohorts. This injection method was used to assess the radiobiological effect of BioXmark^® with tumours treated with single (16 Gy) or fractionated (2×8 Gy or 3×4 Gy) doses of RT. These mice were treated with RT as outlined in Brown et al.³⁵

For the intra-tumoral injection of BioXmark^®, the FM was loaded into a micro-syringe as detailed above and 25-gauge needles were used for injection. Mice were anaesthetised with injectable ketamine and xylazine (100 mg/kg and 10 mg/kg) and the injection point sterilized with ethanol before injection. The needle was carefully placed into the middle of the tumour (estimated through needle insertion) and 20 μl of BioXmark^® injected under consistent pressure. To prevent leakage or bleeding the needle was held for 30 seconds before removal. No mice presented with bleeding or leakage of BioXmark^®. All mice were placed in a heat-box for recovery (37°C) and monitored closely until fully recovered from the anaesthetic.

Orthotopic prostate injection of BioXmark^®

Intra-prostate injections of BioXmark^® were performed under aseptic conditions (n=2). Mice were anaesthetised using inhalant isoflurane (0.5–2%) throughout the procedure and administered analgesia (buprenorphine 0.015 mg/ml (0.05 mg/kg dose)) via IP injection before the surgical procedure and 6 hours post. Protocol adapted from Pavese et al.³⁸ A small portion skin was shaved and the prostate exposed through a low midline abdominal incision of 5 mm (roughly 1 cm above external genitals). The anterior lobe was identified, and a 26-gauge needle was used to inject 10 μl of BioXmark^® under consistent pressure and the needle held in place for 30 seconds to prevent leakage. Due to the delicate procedure, we trialled a finer gauge needle and found it feasible to inject BioXmark^® effectively with a 26-gauge needle. Successful injection was confirmed by the formation of a small rounding or mound shape. The injection area was checked for bleeding or leakage of BioXmark^®, none was found to leak, before carefully placing the prostate back into the abdominal cavity and the small wound was sealed with tissue glue and stitches. Mice were placed in a heat box for recovery (37°C) and healing, body weight and behaviour were closely monitored after the surgical procedure and then returned to original caging.

Injection protocols

BioXmark^® uptake

Injection of BioXmark^®

Intra-tumoral injection of BioXmark^®

A previous study trialled mixing BioXmark^® with tumour cells prior to implant for targeting, this was not feasible with the tumour needing to be establish prior to BioXmark^® injection.³⁴ In this study we trialled an intra-tumoral injection of BioXmark^® as outlined in Brown et al 2020.³⁵ Mice were randomised into control of BioXmark^® injected cohorts once sucutaneous tumours reached a suitable size for treatment (100 mm³). This protocol was used to determine the radiobiological effect of BioXmark^® on tumour response, cytotoxicity and dose perturbation.³⁵ For tumour targeting multiple subcutaneous injections of BioXmark^® around the tumour would be more suitable (Figure 1).

Orthotopic prostate injection of BioXmark^®

Mice were anaesthetised using inhalant isoflurane throughout the procedure and administered analgesia (buprenorphine 0.015 mg/ml (0.05 mg/kg dose)) prior to surgical procedure and 6 hours post. Protocol adapted from Pavese et al 2013.³⁹

Preclinical imaging

Imaging was completed using the Small Animal Radiation Research Platform (SARRP) (Xstrahl Life Sciences, Camberley UK) onboard CBCT imaging. An imaging energy of 60 kV was used with 0.5 mm Al filtration for all in vivo models. The acquired CBCT scans were transformed into material properties by defining five discrete windows for these materials; air, lung, fat, tissue and bone.³⁹ Imaging was completed immediately after the injection of BioXmark^®, while the mice were still anaesthetised, for subcutaneous, IP and intra-tumoral injections. Mice injected with BioXmark^® into the prostate were allowed to recover overnight in the home cage and imaged the following day. These mice were anaesthetised with injectable ketamine and xylazine (100 mg/kg and 10 mg/kg) for imaging.

CERR analysis

Computational Environment for Radiological Research (CERR) software (https://cerr.github.io/CERR/) was used to complete additional image analysis. Image analysis could not be blinded as BioXmark^® was visible on the CBCT scans. For contouring of BioXmark^®, all relevant DICOM images and structure sets were exported into CERR software within MATLAB (Version 2019b).⁴⁰ Structures of interest were contoured manually using the CERR in-house contouring interface. Contours were created slice-by-slice in the coronal plane with corrections made using the sagittal and axial planes. To reduce variations in contours all structures were contoured by the same user.

Statistical analysis

Statistical differences were calculated using unpaired two-tailed student t-tests, or one-way ANOVA tests where appropriate, with a significance threshold of p < 0.05 using Prism GraphPad Prism 7 (Version 7.01, GraphPad Software, Inc.). Data is presented either as the average for the entire experimental arm ± standard error (SEM).

In vivo injection techniques

This study has provided methods for the application of BioXmark^® in small animal models. Mice of both male and female sexes were used and no behavioural or toxic side-effects due to BioXmark^® were observed.

Figure 1A shows CBCT scans from mice injected subcutaneously with BioXmark^® at two separate points on the flank (10, 20 or 40 μl). BioXmark^® is easily visualised in bright white on all CBCT scans and could be easily differentiated from anatomical structures (Figure 1A). Capsules of BioXmark^® formed in the cavity under the skin with all volumes having a similar 3-D shape (Figure 1B & Supplementary Figure 1). Further analysis of the long-term (5-month) stability of BioXmark^® injected subcutaneously is detailed in Brown et al.³⁵

Figure 1. In vivo CBCT imaging analysis of the different volumes of BioXmark^® following subcutaneous injection.

Panel A: CBCT scans of 10 (i), 20 (ii) and 40 µl (iii) of BioXmark^® (left to right) injected subcutaneously at two points on the flank of mice (n =1). Panel B: Volumes of BioXmark^® on CBCT scans were contoured using CERR software and 3-D shape of each marker visualized.

Due to differences in marker shape between subcutaneous (Figure 1A) and intraperitoneal injections (Figure 2A) we assume that BioXmark^® moulds around anatomical structures before solidifying. There is greater diversity in the shape after injection into the intraperitoneal cavity with the most evidenced by the larger volume of 40 μl (Figure 2A & B), indicating that the intraperitoneal injection is less stable than other methods with the marker breaking up and potentially solidifying in a different area than injected.

Figure 2. In vivo CBCT imaging analysis of the intraperitoneal injection of different volumes of BioXmark^®.

Panel A: CBCT scans of 20 µl (i) and 40 µl (ii) of BioXmark^® injected via intraperitoneal injection (n=1 per dose). Panel B: 3-D visualisation of 20 µl (i) and 40 µl (ii) of BioXmark^® after intraperitoneal injection.

Tumour volumes have different shapes and sizes and are heterogeneous causing differences in diffusion of a substance, such as drugs or contrast agents, and potentially BioXmark^®. Figure 3A shows the injection of 20 μl of BioXmark^® into a subcutaneous MC38 colon carcinoma tumour. This was completed to trial the feasibility of injecting BioXmark^® into dense tissue. Intra-tumoral injections were visualised as a hyperdense structure in the centre of the tumour easily differentiated from the tumour tissue (Figure 3A). 3-D reconstructions of injected BioXmark^® which underwent fractionated RT (3×4 Gy) are shown in Figure 3B. Minimal changes were observed in the shape of BioXmark^®. No radiobiological effects of BioXmark^® were determined, further detailed in Brown et al.³⁵

Figure 3. In vivo CBCT imaging analysis of the intra-tumoral injection of 20 µl of BioXmark^®.

20 µl was injected into MC38 subcutaneous tumours once they reached a volume of 100 mm³. Panel A: Mice were imaged on CBCT at 60 kV prior to receiving radiotherapy (n=21). Panel B: Volumes of BioXmark^® on CBCT scans were contoured using CERR software and the 3-D shape of each marker was visualised throughout the dose schedule.

The injection of 10 μl of BioXmark^® into an orthotopic prostate model is shown in Figure 4A. This method is the most technical and clinically accurate for placement of a FM for RT treatment targeting. The orientation of the marker can be monitored through 3-D analysis providing additional information on the location of the prostate. BioXmark^® can be clearly visualised from CBCT scans but also differentiated from soft tissue through CBCT values for tissue (p=0.0093), air (p=0.0072) and bone (p=0.0073) (Figure 4C). Injected BioXmark^® can help to identify the prostate to improve defining a target volume for beam delivery. This model could be used for normal tissue targeting of the prostate or for targeting orthotopic prostate tumours.

Figure 4. In vivo CBCT imaging analysis of the prostate orthotopic injection of 10 µl of BioXmark^®.

Panel A: Mice were imaged on CBCT at 60 kV after injection of BioXmark^® (n=2). Panel B: 3-D shape of injected BioXmark^®. Panel C: Viability of the marker was quantified through CBCT numbers for BioXmark^®, tissue, air and bone. Data presented are mean values ± SEM; statistical significance is reported as *< 0.05, ** < 0.01.