389PThe establishment of patient-derived organoid models and drug response of resectable non-small cell lung cancer

Abstract Background Patient-derived cancer organoid (PDOs) models have proven with powerful research value and significant clinical application prospects. However, we still know little about organoid models of non-small cell lung cancer (NSCLC). This study aims to characterize the consistency of genomic variations between the primary tumours and PDOs, and to explore utility of PDOs as preclinical models to predict the treatment response for the precision medicine. Methods Tumour samples were collected for organoid culture. Primary tumour and PDO samples were analysed by whole exon sequencing (WES). C-MET overexpression of tissue was test by immunohistochemistry. Antineoplastic drugs were tested by the PDOs. Cell viability was measured by Cell Titer Glo assay 7-10 days after drug treatment. Heatmap of log IC50 values were calculated from drug response analyses of PDOs by applying nonlinear regression (curve fit). Results A total of 7 patients (pts) (I-III stage) were enrolled. 7 paired surgical tumour and PDOs were analysed, respectively. Comparison of gene mutations of top 20 ranked genes related with lung cancer, the concordance were over 80% between tumour and PDOs in 5 pts. The concordances of the other 2 pts were less than 50%. Both tissues and PDOs harbored driver mutations in 4 pts (2 EGFR L858R, 1 EGFR EX20 ins and 1 KRAS G12C ). Drug screen was carried out by using 26 antineoplastic drugs in the 7 PDOs in vitro. Of the 2 PDOs with EGFR L858R, one displayed the most significant response to Gefitinib, the other showed resistance to Gefitinib but significant response to Osimertinib, whose matched tissue showed c-MET overexpression indicating a mechanism of resistant to Gefitinib. The PDO with EGFR EX20 ins also indicated resistance to Gefitinib but significant response to Osimertinib in accordance with public articles. Conclusions Patient-derived lung cancer organoids could provide us a practical model system for studying NSCLC and predict treatment response for personal precision medicine. Legal entity responsible for the study The authors. Funding Has not received any funding. Disclosure All authors have declared no conflicts of interest.