In vitro drug sensitivity (IDS) of patient-derived primary osteosarcoma cells as an early predictor of the clinical outcomes of osteosarcoma patients.

Abstract

Early prediction of clinical response to conventional chemotherapy is a significant factor in determining an overall treatment strategy for osteosarcoma. Cells were extracted from treatment-naïve biopsies from 16 osteosarcoma patients who received a doxorubicin and cisplatin-based neoadjuvant chemotherapy regimen and their sensitivities to doxorubicin and cisplatin were measured as IC50 values. Associations of in vitro drug sensitivity (IDS) levels and clinical outcomes were examined. Primary osteosarcoma cells responded to doxorubicin and cisplatin with IC50 values of 0.088 ± 0.032µM and 16.7 ± 8.5µM, respectively. The patients with a non-metastatic phenotype and surviving patients showed significantly lower IC50 values for both drugs. ROC analysis defined the optimal IC50 cut-off values for doxorubicin (IDSdox) and cisplatin (IDScpt) as 0.05µM (AUC 0.82) and 14µM (AUC 0.87), respectively. Survival analysis found significantly longer disease-free survival (DFS, n = 14) and overall survival (OS, n = 16) times in the patients with low IDSdox (p = 0.0064 for DFS and p = 0.0102 for OS) and low IDScpt (p = 0.0204 for DFS and p = 0.0021 for OS). Interestingly, when the patients with low IDScpt and those with low IDSdox were combined (Group 1), significant associations with prolonged DFS (p = 0.0042, C-statistic 0.78) and OS (p = 0.0010, C-statistic 0.79) were found. In this cohort, histological response to neoadjuvant chemotherapy could predict only OS. This study indicates that IDS analysis could potentially be a practical, rapid, and reliable technique for predicting clinical outcomes. It could also be used to identify patients for whom conventional chemotherapy is most appropriate and, in the future, help advance personalized therapy.