AUTHOR=Gao Yang , Zhou Rong , Huang Jun-Feng , Hu Bo , Cheng Jian-Wen , Huang Xiao-Wu , Wang Peng-Xiang , Peng Hai-Xiang , Guo Wei , Zhou Jian , Fan Jia , Yang Xin-Rong 

TITLE=Patient-Derived Xenograft Models for Intrahepatic Cholangiocarcinoma and Their Application in Guiding Personalized Medicine

JOURNAL=Frontiers in Oncology

VOLUME=Volume 11 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2021.704042

DOI=10.3389/fonc.2021.704042

ISSN=2234-943X

ABSTRACT=Background: Intrahepatic cholangiocarcinoma (ICC) remains one of the most intractable malignancies. The development of effective drug treatments for ICC is seriously hampered by the lack of reliable tumor models. At present, patient derived xenograft (PDX) models prove to accurately recapitulate the genetic and biological diversity required to decipher tumor biology and therapeutic vulnerabilities. This study was designed to investigate the establishment and potential application of PDX models for guiding personalized medicine and identifying potential biomarker for lenvatinib resistance.
Methods: We generated PDX models from 89 patients with ICC and compared the morphological and molecular similarities of parental tumors and passaged PDXs. The clinicopathologic features affecting PDX engraftment and the prognostic significance of PDX engraftment were analyzed. Drug treatment responses were analyzed in PDX models and patients was analyzed. Finally, lenvatinib treatment response was examined in PDX models and potential drug resistance was revealed. 
Results: Forty-nine PDX models were established (take rate: 55.1%). Successful PDX engraftment was associated with negative HbsAg (P = 0.031), presence of mVI (P = 0.001), poorer tumor differentiation (P = 0.023), multiple tumor number (P = 0.003) and later TNM stage (P = 0.039). Moreover, patients whose tumor engrafted had significantly shorter time to recurrence (TTR) (P < 0.001) and worse overall survival (OS) (P < 0.001). Multivariate analysis indicated that PDX engraftment was an independent risk factor for shortened TTR (HR = 1.84; 95% CI, 1.05–3.23; P = 0.034) and OS (HR = 2.13; 95% CI, 1.11–4.11; P = 0.024). PDXs were histologically and genetically similar to their parental tumors. We also applied PDX drug testing results to guide clinical treatment for patients with ICC and found similar treatment responses. PDX models also facilitated personalized medicine for patients with ICC based on drug screening results using whole exome sequencing data. Additionally, PDX models reflected the heterogeneous sensitivity to lenvatinib treatment and CDH1 might be vital to lenvatinib-resistance. 
Conclusion: PDX models provide a powerful platform for preclinical drug discovery, and potentially facilitate the implementation of personalized medicine and improvement of cancer patient survival.