Investigation of the chicken chorioallantoic membrane (CAM) xenograft model in theranostics development

Abstract

Background: During drug screening, there is a critical bridge between in vitro and pre-clinical in vivo testing. While laboratory mice are routinely used for the evaluation of radiopharmaceutical compound, their use is constrained by high costs, ethical considerations, and limited throughput, highlighting the demand for high-throughput methodologies to enhance preclinical screening efficiency. Aim: This study investigates the potential of chicken chorioallantois membrane (CAM) as a cost- and time-efficient intermediate for drug evaluation between in vitro and preclinical testing. More specifically, the study explores the CAM model in the context of testing a radioligand designed to diagnose the childhood cancer neuroblastoma. Results: The final timeline of the CAM model development included a windowing procedure at embryonic day 3 (E3), a cell inoculation procedure at E9 and the final experimental day at E14. The tumour xenografts were found to be influenced by the timing of cell inoculation and the proliferative rate of the cells lines. Cell inoculation with lens tissue and cells in 20uL Matrigel at E9 was the most successful and resulted in a vascularised tumour xenograft on the CAM. The self-pulled glass needle injection system enabled the injection of radioligand in the CAM vessel at E14. Three different radioligands were injected in the CAM model: two neuroblastoma targets VHH nanobodies (NCAM1) and antibodies (Dinutuximab), and for “proof-of-concept” FC-VHH nanobodies with its isotype control. Each compound exhibited different biodistribution pattern based on its size with the smaller compound (VHH, 15kDa) being found to have faster blood clearance than the larger compounds (FC-VHH 60kDa and antibody 150kDa). Uptake was found to be highest for all radioligands specifically in the excretory and metabolic organ (i.e. liver, spleen and kidney). The VHH nanobody and FC-VHH also showed to have high binding affinity to the xenografted tumour expressing the target of interest based on their blood-tumour and muscle-tumour uptake ratios. Conclusion: The CAM xenograft model offers a promising, cost-effective, high-throughput, and rapid alternative to mouse models for evaluating new radioligands. The biodistribution patterns observed in the CAM model closely ressembled those in mice and were correlated with the size of the injected radioligand. Radioligand tumour uptake was effectively measurable in the CAM model, and neuroblastoma xenografts further validated in vitro NCAM1 binding studies, highlighting the model’s potential for targeting neuroblastoma cells. The primary limitations of the CAM model are related to the developmental stage of the chickem embryo and species-specific differences.