Animal Model
Experiments were performed using male New Zealand white rabbits obtained from Charles River Institution (Wilmington, Massachusetts), weighing 2.5-3.0 kg in accordance with the University Health Network/University of Toronto guidelines for the humane use of animal care. This study was reviewed and approved by the ethics committee under the Animal Resource Centre, University Health Network, University of Toronto under Animal Utilization Protocol #2931.19. A total of 7 rabbits were involved in this study. Rabbits were naïve to previous treatment / experiments and were assessed by the veterinarians prior to the study to ensure fitness. No inclusion/exclusion criteria were used for selection of these rabbits. Rabbits were individually caged in climate-controlled rooms with a 12-hour light cycle and provided with food and water ad libitum and allowed to acclimatize to facilities for at least 7 days prior to experimental start date. The care and maintenance of these animals were performed in a humane manner with strict compliance to the animal care experimental protocol approved by the institutional animal care and use committee of the University Health Network, University of Toronto.
To induce VX2 buccal tumors, 1 cm3 of VX2 tumor was passed through a 100 µm filter to create a tumor cells suspension using PBS. Seven rabbits were injected with the cell suspension of VX2 squamous cell carcinoma into the right buccal mucosa. Tumor progression was evaluated by clinical examination twice a week and imaging performed weekly. Rabbits were clearly marked via marker pen along the ear and separate dates scheduled for monitoring, imaging and treatment to minimize confounders. At endpoint, rabbits were euthanized following the approved protocol, consisting of anesthesia using isofluorane followed by overdose with potassium chloride. Outcomes for this study involved the quantification of ctDNA amd CTC from serial serological evaluation following tumor induction and radiation treatment. No animals were used as a control group, as this study was a performed to demonstrate proof of concept. All data attained from the animals was included in the final analysis.
CT Imaging and Image analysis
CT imaging was performed at day 5 post VX2 cell injection to ascertain the presence of a viable tumor. Thereafter repeat CT was repeated on the first, third and fifth day of radiation of each cycle. Following completion of the planned course of radiation, imaging was acquired weekly thereafter. CT imaging (Locus Ultra, GE Healthcare, Milwaukee, Wisconsin, USA) was performed weekly (80 kVp, 50 mA). CT image analysis was performed using Microview (GE Healthcare, Milwaukee, Wisconsin, USA) and custom in-house program written using MATLAB (MathWorks®, Natick, Massachusetts, USA). The tumor volumes were contoured using a semi-automated threshold-based method, the mean and standard deviation of the voxel signal distribution within each VOI were calculated.
Radiotherapy Intervention
Radiation for ctDNA
Four rabbits underwent primary radiation when tumor was approximately 1cm in size. This usually occurred 10 days after inoculation of tumor. They each received two cycles of 4 Gy for 5 days with a treatment break of 2 days between the first and second cycle (Figure 1). Treatment breaks over the weekend was designed in compatibility to standard radiation treatment in most centers. The decision for early radiation at day 10 was made to mimic early disease that can be treated with unimodality radiation treatment. Furthermore, it has been noted that the presence of ctDNA fluctuates with tumor necrosis, which for the purpose of this study obscures the effects of radiation. Blood was taken for analysis on days of radiation and selected time points after radiation. On days where radiation treatment was planned, all blood samples were acquired prior to the radiation procedure. In summary, blood taking was performed at days 10-15, 17-21, 25, 27, 31 after tumor injection. The rabbits were sacrificed on day 31 and histological sections obtained.
Radiation for CTC
Three rabbits underwent primary radiation approximately 17 days after inoculation of tumor. All rabbits received a total of 20Gy (4Gy per fraction, one fraction a day for 5 days); no treatment breaks were allocated. These rabbits were monitored by imaging in the form of CT scans and blood sampling via the ear veins for CTCs before radiation and every other day during and following radiation. These rabbits underwent radiation at a later date compared to the rabbits from the ctDNA cohort, in view of the greater tumor volume necessary to detect CTCs. Rabbits were sacrificed 3 days after completion of radiation treatment, imaging was acquired prior to sacrifice.
Quantification of ctDNA
The VX2 tumor model is associated with the cottontail rabbit papillomavirus (CRPV). This is similar to the human papilloma virus (HPV) which has been established as a key oncovirus for cancers of the oropharynx. Thus, this model closely resembles the management of a patient with HPV derived oropharynx cancer. The target for ctDNA detection utilizes primer probes for oncogene E6 or E7. For the purpose of this study, plasma ctDNA was detected by means of a previously validated qPCR assay based on a 58-bp E6 assay derived from the E6 open reading frame (ORF).(14) From prior studies, the 58-bp E6 assay has shown to be sensitive enough to achieve robust detection of CRPV DNA sequences form as little as 3.7 pg of VX2 genomic DNA. Specificities for this test also remain high showing no evidence in samples of rabbits that were not inoculated with VX2. In addition to ctDNA, the collection of cell-free DNA (cfDNA) and total DNA in the serum was collected for further analysis.
Quantification of CTC
CTC quantification was performed using immunomagnetic nanoparticles within a microfluidic device that has been previously described.(13) In summary, blood taken from the rabbit was incubated with MACS anti-CTC nanoparticles for 10 minutes. Thereafter, this blood was then introduced into a fabricated microchip and PBS-EDTA used to rinse out red blood cells before fixation with 100% formaldehyde. Following this, immunostaining was performed via CTC-specific antibodies. This method of CTC measurement, which will be referred to as TxViva, was preferred due to the ability to distinguish the differential expression of these epithelial markers (Figure 2). In addition, it also allows for capture of multiple surface markers, such as Epithelial Growth Factor Receptor (EGFR), Human Epidermal Growth Factor 2 (HER-2), Prostate Specific Membrane Antigen (PSMA) and Epithelial Cell Adhesion Molecule (Epcam) (Fig 1). Comparisons of CTC detection was performed between EpCAM alone and TxViva from a 2ml blood sample.
Statistical analysis
Statistical analysis was performed using GraphPad Prism 7 (GraphPad Software, La Jolla, CA, USA). A P-value of < 0.05 was considered significant.