Abstract Book

S1288

ESTRO 37

Material and Methods The squamous cell carcinoma (SCC) cell line FaDu and the immortalized untransformed RPE cells were used. Cells were seeded in agarose pre-coated multi-well plates to generate tumour spheroids. Prior to irradiation (4h), medium containing 10µM pimonidazole and 10µM BrdU was added to the culture and immediately exchanged back to normal medium after irradiation to assess the microenvironmental parameters at the time of irradiation. Irradiated and mock-treated spheroids were fixed in 4% formaldehyde and embedded into paraffin. Spheroid cross-sections with 4 µm distance between them were used for analysis. Microenvironmental parameters were assessed in BrdU and pimonidazole stained section while radiation-induced damage was visualized in the consecutive section stained for γH2AX and cell nuclei were counterstained with dapi. We have developed a macro-programming of the publically available Fiji (imageJ) software that allows automated selection of every nuclei of the spheroid cross-section based on dapi signal segmentation. Quantification of single foci within each nucleus taking into account parameters of both foci size and signal intensity is subsequently performed. The exact location of each segmented nucleus in relation to the outer rim of the spheroid cross-section (based on image co-ordinates) is recorded and used to generate a distance heat-map. We aim to merge the distance map with proliferation and hypoxia-mask allowing documentation of proliferative and hypoxic status of each cell both by means of BrdU/pimonidazole staining and live cell hypoxia reporter (ongoing). Results Cross-section of RPE spheroids could allow for segmentation of all viable cells (mean cell number analyzed per section 1500 cells). We observed a statistically significant inverse-linear correlation between the number of γH2AX foci per nucleus and the distance from the outer spheroid rim 1h post-4 Gy irradiation (slope: -0.02, r 2 : 0.10, p<0.0001), while no correlation was existing in mock-irradiated controls (p=0.29). Conclusion Herein, we present an automated method that allows quantification of DNA DSBs in every single cell in spheroid cross-sections which can be coupled with assessment of the proliferation or hypoxic status of each single cell. EP-2335 3D tumour spheroids as an alternative to clonogenic assays for predicting radiation response in vivo S.C. Brueningk 1 , M. Costa 1 , C. Box 1 , I. Rivens 1 , S. Nill 1 , U. Oelfke 1 , G. Ter Haar 1 1 The Institute of Cancer Research and The Royal Marsden NHS Foundation Trust, Joint Department of Physics at the Institute of Cancer Research and the Royal Marsden NHS Foundation Trust, Sutton, United Kingdom Purpose or Objective Whilst clonogenic assays have traditionally been used to estimate radiation response in vitro , there are now more physiological models and assays available that have not been fully explored for prediction of in vivo response. Tumour spheroids are 3D cancer cell aggregates cultured in vitro . In terms of cell-cell contact, and nutrient and oxygen gradients they have a more physiological cellular microenvironment than 2D cultures. This study investigates whether tumour spheroids provide a more predictive model for evaluation of radiation response in vivo than 2D cultures.

Material and Methods Three human cancer cell lines, (LICR-LON-HN5 (HN5), CAL R , HCT116) were tested for their radiation response to single radiation doses up to 20Gy in 2D and 3D cultures in vitro , and at 20Gy in subcutaneous tumours in vivo . Conventional 2D clonogenic assays were performed and growth curves acquired (resazurin assay). For 3D tumour spheroids, volume growth was assessed using a cytometer, and cell viability was measured with a chemiluminescence assay. Subcutaneous tumours in female FOXnu n1 mice were irradiated at 200mm 3 in a small animal radiation research platform (SARRP). Tumour volumes were monitored for up to 60 days post treatment. Histological sections of both tumours and spheroids were stained at time of treatment with either haematoxylin & eosin (to visualise tissue morphology), or pimonidazole (to assess the spatial distribution of hypoxia). Results In 2D assays HN5 and CAL R responded similarly to radiation whereas HCT116 cells were more radiosensitive (SF 2HN5 =0.60, SF 2CALR =0.52, SF 2HCT116 =0.38). In 3D spheroids, CAL R and HCT116 radiation responses were comparable, with notable volume decreases in irradiated samples. HN5 tumour spheroids did not display significant volume changes over time. Differences in growth delay were best observed for spheroids irradiated with 5Gy where time to regrowth was shortest for HCT116 spheroids but comparable for the other two cell lines (figure 1). These results were reflected well in the in vivo 20Gy radiation responses, with relatively constant volumes for irradiated HN5 tumours, but tumour regression in both irradiated CAL R and HCT116 models (figure 2). A higher proportion of HCT116 tumours regrew which may be correlated with the fast growth seen in 3D spheroids. Histological sample analysis allowed comparison of the cellular microenvironment within spheroid and tumour sections.

Conclusion 3D Spheroid assays capture a combination of the growth and radiation response behaviours of the cell lines used. They may hence be more predictive of radiation response in vivo than traditional 2D assays and could be used to design preclinical in vivo studies of combination treatments of other treatment modalities with radiotherapy.