Abstract
BACKGROUND: The chick chorioallantoic membrane (CAM) tumor model has garnered significant interest due to its robust applications in research. However, there is a lack of understanding regarding the characteristics of nodules at different developmental stages and their corresponding imaging features. This study aimed to establish correlations between nodule attributes and magnetic resonance imaging (MRI) features via an ex ovo pancreatic cancer CAM model. Furthermore, this study explored the potential of functional MRI techniques in quantitatively discriminating pathological characteristics. METHODS: A total of 50 chicken embryos were cultured in ovo through embryonic day (ED) 3, then transferred to a customized chamber for ex ovo culture and subsequent experiments. On ED7, cancer cell mixtures [pancreatic cancer cells mixed with Matrigel and Dulbecco's modified Eagle medium (DMEM)] were implanted on the CAM to establish the cancer model. At each observation time point (ED10, ED12, and ED14), eight models were randomly selected for multi-parametric MRI (mpMRI) via a 3.0 Tesla (T) magnetic resonance (MR) scanner and pathological examination. The remaining models were observed on ED16. Quantitative assessments including tumor nodule volume, apparent diffusion coefficient (ADC), and T1 values were further correlated with histopathological findings. RESULTS: Among 50 embryos, 90% (45/50) underwent tumor implantation, with 68% (34/50) models completing mpMRI scanning and pathological examination (n=8 at each of the first three time points; n=10 at the last time point). MRI could clearly observe differences in tumor signals, except for the T1 mapping sequence. Nodule volumes were 60.40±12.18 mm(3) on ED12, 87.44±17.45 mm(3) on ED14, and 108.13±5.45 mm(3) on ED16, showing significant progression (P<0.001 for all intergroup comparations). ADC values were (1.47±0.48)×10⁻(3) mm(2)/s on ED12, (1.32±0.13)×10⁻(3) mm(2)/s on ED14 (P=0.29 vs. ED12), and (0.96±0.07)×10⁻(3) mm(2)/s on ED16 (P<0.05 vs. both earlier time points). T1 relaxation times were measured (1.64±0.31)×10(3) ms on ED12, (1.73±0.16)×10(3) ms on ED14, and (1.74±0.29)×10(3) ms on ED16, with no significant variations (P=0.69 for all intergroup comparisons). Pathological examination confirmed viable tumor cells in all specimens, demonstrating ring-like tumor tissue by ED12 that progressively replaced the Matrigel over subsequent time points. CONCLUSIONS: The ex ovo CAM serves as a reliable and reproducible experimental platform. Conventional 3.0 T MRI provides high-resolution morphological assessment of tumor dynamics, whereas functional imaging techniques offer promising capabilities for quantifying inter-nodular heterogeneity.