Analysis of drug sensitivity of human high-grade osteosarcoma in a chick chorioallantoic membrane (CAM) model: a proof of principle study

Analysis of drug sensitivity of human high-grade osteosarcoma in a chick chorioallantoic membrane (CAM) model: a proof of principle study:





Abstract

Objective

Multi-agent chemotherapy is an important pillar in treatment of high-grade osteosarcoma. In an effort to improve patient survival, it is imperative to determine the effectiveness of new substances. The objective of this study was to investigate whether the chick chorioallantoic membrane (CAM) model can be used to analyze drug sensitivity in high-grade osteosarcoma.



Results

Spare biopsy tissue from five patients diagnosed with high-grade osteosarcoma was transferred into non-immortalized primary cell culture. After a pre-incubation period of 10 days, fertilized chick eggs were inoculated with primary tumor cells suspended in extracellular matrix gel. On day 16, treatment with 20 µmol/l doxorubicin (n = 4) or 25 µl of culture medium (n = 6) was performed for 24 h. CAM membranes were documented macroscopically, harvested and examined histologically. Transfer of biopsy specimens into primary cell culture was successful in all cases. 50% (n = 10) of eggs died after inoculation with tumor cells and before application of doxorubicin. No deaths occurred after application of doxorubicin. Histological examination found a response to doxorubicin in all four specimens. Based upon these results, the CAM model represents a promising preclinical alternative to animal experiments to determine drug sensitivity of osteosarcoma cells. Further research with regard to other substances and dosages appear justified.



Background

High-grade osteosarcoma is a rare disease, mostly affecting children, adolescents and young adults. A combination of chemotherapy and wide tumor resection are the main pillars for successful treatment of these patients. The significance of chemotherapy becomes clear when considering how overall survival rates improved from 20% to more than 60% with the introduction of multi-agent chemotherapy [1]. Standard chemotherapeutic agents include doxorubicin, cisplatin, ifosfamide and methotrexate [2]. Patients are usually enrolled in multimodal clinical trials in an effort to improve treatment outcomes by comparing new approaches to a standard regimen. Current protocols intend patients to undergo neoadjuvant chemotherapy to evaluate response to chemotherapy as the percentage of tumor regression found in the resection specimen of the primary tumor. As a result, good and poor responders are assigned to different adjuvant chemotherapy groups, which vary with regard to duration and composition of remaining treatments [2]. However, evidence regarding the efficacy of this risk-stratified treatment approach is still lacking [3, 4]. Therefore, in case of poor response to chemotherapy, an antedated operation or modulation of administered chemotherapy might be desirable. However, no significant developments have been made with regard to new drugs in osteosarcoma protocols over the last few decades which may also be due to the lack of suitable preclinical individualized models having the power to reliably predict response [1, 3, 4].



Preclinical research currently relies on in vitro and mouse models for drug screens to identify promising substances that might be translated into a clinical trial [5]. The latter in vivo xenograft models tend to be lengthy and unable to yield short-term information that might be able to influence individual treatment approaches. Beyond that, there has been a paradigm shift in our scientific society: ethical recommendations challenge scientists to develop models that do not rely on animal experiments whenever possible. In this context, the chorioallantoic membrane (CAM) model is particularly interesting as the chick embryo is not considered to be a living animal until day 17 of development in most countries and therefore does not fall under animal experiment [6, 7]. Though successful engraftment of musculoskeletal tumors including human osteosarcoma cell lines has been reported and though there is some preliminary evidence that it may serve as a preclinical screening assay predictive for disease outcome, it is not in common use yet [8, 9].



Therefore, it is the objective of this study to analyze whether the CAM model may serve as a reliable preclinical drug-screening assay and if so, may even be predictive of response of the tumor to chemotherapy seen in the resected specimen of the primary tumor in a follow-up study.