A 3D Tumor‐Mimicking In Vitro Drug Release Model of Locoregional Chemoembolization Using Deep Learning‐Based Quantitative Analyses

• Published in: Advanced science 2023-04, Vol.10 (11), p.e2206195-n/a
• Main Authors: Liu, Xiaoya, Wang, Xueying, Luo, Yucheng, Wang, Meijuan, Chen, Zijian, Han, Xiaoyu, Zhou, Sijia, Wang, Jiahao, Kong, Jian, Yu, Hanry, Wang, Xiaobo, Tang, Xiaoying, Guo, Qiongyu
• Format: Article
• Language: English
• Subjects: 3D drug release model; Aqueous solutions; Cancer therapies; Carcinoma, Hepatocellular - therapy; Chemoembolization, Therapeutic; Chemotherapy; decellularized organ; Deep Learning; Drug Liberation; Embolization; Extracellular matrix; hepatocellular carcinoma; Humans; Liver cancer; Liver Neoplasms - therapy; Pharmaceuticals; Quantitative analysis; transarterial chemoembolization; Veins & arteries
• ISSN: 2198-3844; 2198-3844
• DOI: 10.1002/advs.202206195

Primary liver cancer, with the predominant form as hepatocellular carcinoma (HCC), remains a worldwide health problem due to its aggressive and lethal nature. Transarterial chemoembolization, the first‐line treatment option of unresectable HCC that employs drug‐loaded embolic agents to occlude tumor‐feeding arteries and concomitantly delivers chemotherapeutic drugs into the tumor, is still under fierce debate in terms of the treatment parameters. The models that can produce in‐depth knowledge of the overall intratumoral drug release behavior are lacking. This study engineers a 3D tumor‐mimicking drug release model, which successfully overcomes the substantial limitations of conventional in vitro models through utilizing decellularized liver organ as a drug‐testing platform that uniquely incorporates three key features, i.e., complex vasculature systems, drug‐diffusible electronegative extracellular matrix, and controlled drug depletion. This drug release model combining with deep learning‐based computational analyses for the first time permits quantitative evaluation of all important parameters associated with locoregional drug release, including endovascular embolization distribution, intravascular drug retention, and extravascular drug diffusion, and establishes long‐term in vitro–in vivo correlations with in‐human results up to 80 d. This model offers a versatile platform incorporating both tumor‐specific drug diffusion and elimination settings for quantitative evaluation of spatiotemporal drug release kinetics within solid tumors. A 3D drug release model combined with deep learning‐based computational analyses enables quantitative evaluation of spatiotemporal drug release behaviors within tumor tissues, showing the promise for use as a powerful research tool to facilitate the development and optimization of translational drug compositions while leveraging comprehensive information of the whole drug release process from minimum organ samples for various locoregional treatments.